Browse the corpus
Walk the evidence base by book and chapter — the raw source passages that ground Ask, Differential, and the rest.
500 passages (showing first 500)
Editor’s key points Positive end-expiratory pressure is widely used in mechanically ventilated patients with the acute respiratory distress syndrome (ARDS), but the effect of PEEP on tissue oxygen delivery is not known. The authors investigated the effects of PEEP on tissue oxygen delivery in ARDS. Increasing PEEP increased arterial oxygen tension but decreased tissue oxygen delivery. The incidence of acute respiratory distress syndrome (ARDS) has been estimated at ∼70 per 100 000 patients yr−1,1,2 with a lethal outcome in 55% of patients.3 The global burden of the disease was estimated as 5.5 million patients yr−1 requiring intensive care unit admission and mechanical pulmonary ventilation.4
ichotomous variable of ‘sex’ is reported, as the participants were not asked to self-report their gender. Reporting follows the Strengthening the Reporting of Observational Studies in Epidemiology guidelines.64Fig 1 Flow chart of participant numbers and assessments. CPM, conditioned pain modulation; F, female; M, male. Fig 1 Conditioned pain modulation protocol CPM was assessed using pressure pain threshold (PPT) on the knee as a variable test stimulus and cold water immersion of the contralateral hand as a conditioning stimulus. PPT has good-to-excellent reliability as a variable test stimulus,1, 5, 25, 26 and use on the contralateral lower limb ensures engagement of ascending–descending long tract activity and not just segmental spinal inhibitory effects.1 PPT testing with cold conditioning is reproducible, sensitive to change,1 and has a good test–retest reliability with a smaller sample size than alternative test (electrical and cuff PPT) or conditioning (cuff algometry) stimuli.27 As generalised sensitivity may be either increased or decreased after preterm birth depending on participant age, and the type and intensity of test stimulus,12, 28 PPT was used to provide a reliable linear measure of increases or decreases in baseline threshold. Repeat measures of PPT were performed both in parallel with (15 s), and after cessation of, the conditioning stimulus (50 and 90 s after initial immersion) to assess the duration of effect and minimize the likelihood that the effects are caused by distraction during hand immersion. A cold (5°C) conditioning stimulus with hand immersion for 30 s was chosen, as this evoked inhibitory CPM in healthy 12–17 yr olds,7 but is shorter than many protocols, as preterm young adults may have reduced cold-pressor tolerance (5 of 31 withdrew hand before 30 s).29
Increasing PEEP increased arterial oxygen tension but decreased tissue oxygen delivery. The incidence of acute respiratory distress syndrome (ARDS) has been estimated at ∼70 per 100 000 patients yr−1,1,2 with a lethal outcome in 55% of patients.3 The global burden of the disease was estimated as 5.5 million patients yr−1 requiring intensive care unit admission and mechanical pulmonary ventilation.4 The rationale for the application of PEEP during mechanical ventilation of the lungs of patients with ARDS is to prevent alveolar collapse, reducing injurious alveolar shear stresses and improving ventilation–perfusion matching, and thus, arterial oxygenation.5,6 Studies investigating the effect of PEEP have consistently shown an improvement in oxygenation and pulmonary compliance.5,7–10 Survival benefit was seen in patients when PEEP was assigned based on oxygen requirements in combination with low vs traditional tidal volume ventilation,11 and some risk reduction was shown in a pooled subgroup analysis when patients were stratified based on ARDS severity.12 Despite this, the results of four large studies examining high-PEEP strategies in ARDS have demonstrated an equivocal effect on mortality,13–16 confirmed after meta-analysis.12 Consequently, it is unclear who might benefit from the application of PEEP; clinical investigation has, to date, failed to provide a conclusive answer.
this, the results of four large studies examining high-PEEP strategies in ARDS have demonstrated an equivocal effect on mortality,13–16 confirmed after meta-analysis.12 Consequently, it is unclear who might benefit from the application of PEEP; clinical investigation has, to date, failed to provide a conclusive answer. High levels of positive airway pressure throughout the respiratory cycle have the potential to impair cardiac performance, manifested as a reduced cardiac output.17–20 This is a result of increased right ventricular afterload, reduced left ventricular preload, and reduced biventricular compliance.21 It is credible that PEEP-induced reduction in cardiac output may outweigh the benefit of improved arterial oxygenation, resulting in reduced organ oxygen delivery (DO2), and this was suggested by the early work by Suter and colleagues22 examining the effect of PEEP on lung compliance. Given the difficulty of accurately measuring oxygen delivery in vivo, arterial oxygenation is the usual, clinical target for ventilatory optimization; thus, we may remain unaware of the quantitative effect that PEEP might have with respect to oxygen delivery. Acute respiratory distress syndrome is a heterogeneous disease process, with widely varying cause and progression. This makes it an ideal candidate for high-fidelity modelling studies that can investigate the benefits and hazards of PEEP across individuals in the safe, cost-effective, reproducible in silico environment.
Acute respiratory distress syndrome is a heterogeneous disease process, with widely varying cause and progression. This makes it an ideal candidate for high-fidelity modelling studies that can investigate the benefits and hazards of PEEP across individuals in the safe, cost-effective, reproducible in silico environment. Methods Our study uses a highly integrated computer simulation model of the pulmonary and cardiovascular systems that has recently been developed by our group.23,24 The model includes 100 independently configured alveolar compartments and 19 cardiac compartments. Aspects of the model related to pulmonary pathophysiology have been validated previously,23,25–28 including ARDS.29,30 This model was integrated with a multicompartmental, contractile cardiovascular model with pulsatile blood flow and ventilation-affected, transalveolar blood flow. The cardiac section of the model consists of two contractile ventricles, with atria modelled as non-contractile, low-resistance, high-compliance compartments. The mathematical principles underpinning the model are detailed in the Supplementary appendix.
ile blood flow and ventilation-affected, transalveolar blood flow. The cardiac section of the model consists of two contractile ventricles, with atria modelled as non-contractile, low-resistance, high-compliance compartments. The mathematical principles underpinning the model are detailed in the Supplementary appendix. Cardiopulmonary interactions are modelled in a number of ways. Ventricular contractility is modelled as a truncated sine-wave varying ventricular elastance over time. Intrapulmonary pressure is transmitted variably across ventricular walls such that lung inflation reduces ventricular compliance; in this study, 85% of intrapulmonary pressure was transmitted across the ventricles. Transalveolar blood flow is governed by pulmonary arterial pressure and by transalveolar vasoresistance; this resistance is affected dynamically in each alveolar compartment by alveolar volume (causing longitudinal stretch) and alveolar pressure (causing compression).
ary pressure was transmitted across the ventricles. Transalveolar blood flow is governed by pulmonary arterial pressure and by transalveolar vasoresistance; this resistance is affected dynamically in each alveolar compartment by alveolar volume (causing longitudinal stretch) and alveolar pressure (causing compression). Published, clinical data were used to validate the responses of the integrated model against those of individual ARDS patients. Global optimization algorithms were used to configure the model parameters (e.g. microbronchial resistances, transalveolar vasoresistances, and alveolar compliance) against published clinical trial data on tidal volume, respiratory rate, haemoglobin concentration, metabolic rate, and pulmonary shunt fraction, in order to replicate arterial blood gas values. Once these static configurations were determined, cardiovascular settings in the model (e.g. ventricular contractility, compartmental blood volumes, arterial resistances, and ventricular splinting) were manually tuned to match reported dynamic changes in cardiovascular performance during PEEP variation. Where there were deficiencies in the published data sets, historically appropriate patient characteristics and clinical data were used to create a plausible estimate of the missing data values. For example, if haemoglobin concentration was not reported, it was estimated to be 100 g litre−1 (this being a common clinical target before the publication of the TRICC study in 1999);31 where ventilation mode was not specified, it was assumed to be volume controlled with a constant inspiratory flow and inspiratory-to-expiratory ratio of 1:2 (this being reported in many clinical trials investigating ventilation strategies for ARDS).11
linical target before the publication of the TRICC study in 1999);31 where ventilation mode was not specified, it was assumed to be volume controlled with a constant inspiratory flow and inspiratory-to-expiratory ratio of 1:2 (this being reported in many clinical trials investigating ventilation strategies for ARDS).11 Initial matching was against data derived from a single ARDS patient reported by Dantzker and colleagues.17 This patient had severe ARDS, with arterial oxygen partial pressure (PaO2)-to-fractional inspired O2 ratio of 87 mm Hg (11.6 kPa), and underwent a four-stage incremental PEEP trial. Thereafter, the same process was carried out using the clinical data sets of Jardin and colleagues18 and Pinsky and colleagues;20 for each of these clinical studies, the model was matched to the average cardiopulmonary state of each study population and subjected to the published incremental PEEP trials. Model outputs were compared against the data collected in the clinical studies. Model simulation and comparison of results with the historical data were carried out by two independent investigators in different universities, allowing corroborated evaluation of the simulations. Full details of the methodology used in validating the model against clinical data are provided in the Supplementary appendix. When accurate reproduction of results was achieved, we were able to proceed to prospective testing of the effects of an incremental PEEP trial on DO2 in a variety of matched in silico subjects with various ARDS disease configurations.
Model simulation and comparison of results with the historical data were carried out by two independent investigators in different universities, allowing corroborated evaluation of the simulations. Full details of the methodology used in validating the model against clinical data are provided in the Supplementary appendix. When accurate reproduction of results was achieved, we were able to proceed to prospective testing of the effects of an incremental PEEP trial on DO2 in a variety of matched in silico subjects with various ARDS disease configurations. PEEP trial simulation protocol An in silico bank of 12 ARDS subjects was created by configuring parameters in the model to match published pathophysiological data from several publications.17,18,20,32,33 An additional ‘healthy’ subject was configured to provide a baseline response (Table S3 in the Supplementary appendix). The cardiovascular model parameters used in the validation process were used to initialize the cardiovascular system model for prospective testing, and these parameters are reported in the Supplementary appendix (Table S6).
PEEP trial simulation protocol An in silico bank of 12 ARDS subjects was created by configuring parameters in the model to match published pathophysiological data from several publications.17,18,20,32,33 An additional ‘healthy’ subject was configured to provide a baseline response (Table S3 in the Supplementary appendix). The cardiovascular model parameters used in the validation process were used to initialize the cardiovascular system model for prospective testing, and these parameters are reported in the Supplementary appendix (Table S6). Modelled patients received pulmonary ventilation using settings in line with recommendations from the ARDSnet study;11 using square-wave pressure-controlled ventilation, with a constant inspiratory-to-expiratory ratio of 1:2 and ventilatory rate of 10 bpm. Inspiratory pressure was adjusted to maintain a tidal volume between 450 and 600 ml (6–8 ml kg−1) to maintain arterial carbon dioxide partial pressure (PaCO2) between 4 and 10 kPa based on the findings of early tidal volume and permissive hypercapnia studies.34–36 The initial fractional inspired O2 from the matching process was fixed and kept constant throughout; this ensured that any observed increase in oxygenation would be as a result of alveolar recruitment. The PEEP started at 0 cm H2O and increased by 5 cm H2O every 10 min to a maximum of 20 cm H2O, before finally returning to 0 cm H2O (see Table 1). Table 1 Summary of findings and trial protocol. Data presented are the mean average (sd) values at each PEEP value for all 12 simulated patients. CO, cardiac output; DO2, oxygen delivery; MAP, mean aortic arterial pressure; PaCO2, arterial partial pressure of carbon dioxide; PaO2, arterial partial pressure of oxygen; SaO2, arterial oxygen saturation; SvO2, mixed venous oxygen saturation; VT, tidal volume
) values at each PEEP value for all 12 simulated patients. CO, cardiac output; DO2, oxygen delivery; MAP, mean aortic arterial pressure; PaCO2, arterial partial pressure of carbon dioxide; PaO2, arterial partial pressure of oxygen; SaO2, arterial oxygen saturation; SvO2, mixed venous oxygen saturation; VT, tidal volume Parameter PEEP 0 cm H2O PEEP 5 cm H2O PEEP 10 cm H2O PEEP 15 cm H2O PEEP 20 cm H2O PEEP 0 cm H2O VT (ml) 488 (68.6) 503 (63.3) 507 (78.9) 520 (74.4) 545 (72.6) 488 (68) MAP (mm Hg) 89.4 (2.12) 85.4 (1.70) 82.2 (1.22) 79.4 (1.05) 76.9 (0.99) 89.3 (2.11) CO (ml min−1) 6523 (197) 6008 (153) 5552 (95.8) 5106 (77.4) 4669 (74.3) 6520 (199) SaO2 (%) 89.2 (7.26) 90.2 (6.26) 89.5 (7.17) 93.1 (7.52) 93.1 (8.47) 89.1 (7.33) PaO2 (kPa) 8.90 (2.88) 9.54 (2.99) 9.86 (3.79) 11.25 (3.86) 15.8 (8.78) 9.36 (2.77) PaCO2 (kPa) 5.63 (0.99) 6.26 (1.54) 6.61 (1.83) 6.59 (1.89) 6.37 (1.96) 6.80 (1.92) SvO2 (%) 56.0 (1.54) 54.3 (1.50) 50.6 (1.66) 50.0 (1.75) 48.3 (2.00) 56.0 (1.54) DO2 (ml min−1) 792 (271) 736 (242) 676 (223) 636 (200) 591 (185) 791 (272) Recruitment (%) 66.2 (12.8) 70.3 (12.3) 73.0 (12.4) 78.1 (11.2) 83.7 (11.4) 66.3 (12.8) Dynamic strain (ratio) 0.74 (0.28) 0.60 (0.18) 0.56 (0.17) 0.49 (0.16) 0.43 (0.18) 0.77 (0.27)
4) 54.3 (1.50) 50.6 (1.66) 50.0 (1.75) 48.3 (2.00) 56.0 (1.54) DO2 (ml min−1) 792 (271) 736 (242) 676 (223) 636 (200) 591 (185) 791 (272) Recruitment (%) 66.2 (12.8) 70.3 (12.3) 73.0 (12.4) 78.1 (11.2) 83.7 (11.4) 66.3 (12.8) Dynamic strain (ratio) 0.74 (0.28) 0.60 (0.18) 0.56 (0.17) 0.49 (0.16) 0.43 (0.18) 0.77 (0.27) The following outputs were recorded every 10 ms: arterial haemoglobin oxygen saturation (SaO2), mixed venous haemoglobin oxygen saturation (SvO2), PaO2, PaCO2, arterial pH, alveolar recruitment (the fraction of alveoli receiving non-zero ventilation), cardiac output, aortic blood pressure, arterial oxygen delivery (DO2), and dynamic alveolar strain (as a surrogate for the risk of alveolar injury) as suggested by Protti and colleagues.37 Data are presented as averages throughout the ninth minute after each change in PEEP (i.e. during the minute preceding the next PEEP value). Model simulations were run on a 64-bit Intel Core i7 3.7 GHz Windows 7 personal computer, running Matlab version R2014a (8.3.0.532; MathWorks Inc., Natick, MA, USA). Research ethics committee approval was not sought, because the data used for model development and validation were sourced from studies that had already received ethical approval. Simulation protocols were performed purely in silico without participation from human subjects.
8.3.0.532; MathWorks Inc., Natick, MA, USA). Research ethics committee approval was not sought, because the data used for model development and validation were sourced from studies that had already received ethical approval. Simulation protocols were performed purely in silico without participation from human subjects. Results Results of the initial calibration are shown in the Supplementary appendix; Fig. S4 shows the close fit of the model to data from patient-8 from Dantzker and colleagues17 against cardiac output, ventricular stroke volume, pulmonary vascular resistance, and PaO2. Figures S5 and S6 in the Supplementary appendix show simulation performance against the clinical results from Jardin and colleagues18 and Pinsky and colleagues,20 respectively. Modelled results were consistently very close to those of clinical studies, indicating acceptable validity of the models in reproducing dynamic, in vivo, multi-organ behaviour. Table 1 shows the average value of each measured parameter in all 12 in silico subjects at each PEEP setting during the implemented PEEP trial simulation protocol. The results presented show the mean and sd for all 12 in silico subjects at each PEEP setting for each reported parameter. Figures displaying the behaviour of each subject in the group are provided in the Supplementary appendix.
silico subjects at each PEEP setting during the implemented PEEP trial simulation protocol. The results presented show the mean and sd for all 12 in silico subjects at each PEEP setting for each reported parameter. Figures displaying the behaviour of each subject in the group are provided in the Supplementary appendix. In all but the healthy subject, PEEP increased arterial oxygenation. At 20 cm H2O PEEP, in comparison with the value at 0 cm H2O PEEP, the following changes were observed; each is expressed as mean (sd, range): PaO2 increased by 6.9 kPa (8.77, −0.47 to 27.0 kPa; Fig. 1); SaO2 increased by 3.9% (6.36, −4.33 to 19.4 kPa); recruited alveolar compartments increased by 18% (10.3, 0–42.8%; Fig. 2); mean arterial pressure reduced by 22 mm Hg (1.79, 9.20–15.5 mm Hg); cardiac output reduced by 1.85 litres min−1 (0.17, 1.46–2.11 litres min−1); and oxygen delivery reduced by 0.20 litres min−1 (0.20, 0.07–0.46 litres min−1; Fig. 3). Fig 1 Arterial partial pressure of oxygen (PaO2) over time during incremental PEEP trial. Filled blocks with errors bars represent group mean and sd for all 12 in silico patients (left axis). Dashed lines show the average ventilation metrics during the incremental PEEP trial (PaWP, peak airway pressure; MaWP, mean airway pressure; right axis). Fig 2 Percentage of recruited alveolar compartments over time during incremental PEEP trial. Filled blocks with errors bars represent group mean and sd for all 12 in silico patients.
In all but the healthy subject, PEEP increased arterial oxygenation. At 20 cm H2O PEEP, in comparison with the value at 0 cm H2O PEEP, the following changes were observed; each is expressed as mean (sd, range): PaO2 increased by 6.9 kPa (8.77, −0.47 to 27.0 kPa; Fig. 1); SaO2 increased by 3.9% (6.36, −4.33 to 19.4 kPa); recruited alveolar compartments increased by 18% (10.3, 0–42.8%; Fig. 2); mean arterial pressure reduced by 22 mm Hg (1.79, 9.20–15.5 mm Hg); cardiac output reduced by 1.85 litres min−1 (0.17, 1.46–2.11 litres min−1); and oxygen delivery reduced by 0.20 litres min−1 (0.20, 0.07–0.46 litres min−1; Fig. 3). Fig 1 Arterial partial pressure of oxygen (PaO2) over time during incremental PEEP trial. Filled blocks with errors bars represent group mean and sd for all 12 in silico patients (left axis). Dashed lines show the average ventilation metrics during the incremental PEEP trial (PaWP, peak airway pressure; MaWP, mean airway pressure; right axis). Fig 2 Percentage of recruited alveolar compartments over time during incremental PEEP trial. Filled blocks with errors bars represent group mean and sd for all 12 in silico patients. Fig 3 Oxygen delivery over time during incremental PEEP trial. Filled blocks with errors bars represent group mean and sd for all 12 in silico patients.
Fig 2 Percentage of recruited alveolar compartments over time during incremental PEEP trial. Filled blocks with errors bars represent group mean and sd for all 12 in silico patients. Fig 3 Oxygen delivery over time during incremental PEEP trial. Filled blocks with errors bars represent group mean and sd for all 12 in silico patients. Average dynamic alveolar strain (i.e. alveolar tidal volume/end-expiratory volume) decreased with the incremental addition of PEEP, with an absolute average reduction in strain ratio of 0.314, representing a relative average reduction in strain of 43% across the group; however, the exact relationship between PEEP and strain varied within the group (Fig. 4). The relationship between alveolar unit recruitment and dynamic alveolar strain ratio for all PEEP settings in all 12 in silico subjects is shown in Fig. S7 in the Supplementary appendix. Fig 4 Dynamic lung strain over time during incremental PEEP trial. Filled blocks with errors bars represent group mean and sd for all 12 in silico patients. Discussion The use of a highly integrated pulmonary and cardiovascular computer simulation allowed systematic investigation of the effects of incremental PEEP on the cardiopulmonary systems in an in silico population of ARDS patients.
Average dynamic alveolar strain (i.e. alveolar tidal volume/end-expiratory volume) decreased with the incremental addition of PEEP, with an absolute average reduction in strain ratio of 0.314, representing a relative average reduction in strain of 43% across the group; however, the exact relationship between PEEP and strain varied within the group (Fig. 4). The relationship between alveolar unit recruitment and dynamic alveolar strain ratio for all PEEP settings in all 12 in silico subjects is shown in Fig. S7 in the Supplementary appendix. Fig 4 Dynamic lung strain over time during incremental PEEP trial. Filled blocks with errors bars represent group mean and sd for all 12 in silico patients. Discussion The use of a highly integrated pulmonary and cardiovascular computer simulation allowed systematic investigation of the effects of incremental PEEP on the cardiopulmonary systems in an in silico population of ARDS patients. In silico simulation allows repeated, systematic investigation, without confounding by unquantified (‘silent’) interpatient variability or variability over time. The use of a population of modelled subjects allows greater confidence in extrapolating our findings to patients. The standardization allowed by an in silico population reduced ‘data noise’, and we anticipate that pragmatic and ethical issues would prevent a similar investigation in vivo.
lity or variability over time. The use of a population of modelled subjects allows greater confidence in extrapolating our findings to patients. The standardization allowed by an in silico population reduced ‘data noise’, and we anticipate that pragmatic and ethical issues would prevent a similar investigation in vivo. From previous work,30 it was expected that improved oxygenation would arise from alveolar recruitment. There is currently no standardized duration for a clinical trial of incremental PEEP; however, there are recommendations for the duration of recruitment manoeuvres lasting from a few seconds38 to several minutes.5,6 It was observed that after ventilation changes, the model required a period of equilibration, during which alveolar compartment recruitment occurred, and cardiac performance and gas exchange stabilized. On this basis, the duration of each level of PEEP was set at 10 min, approximately twice the time required for 98% equilibration.24 Increasing PEEP markedly increased PaO2; however, the accompanying increase in haemoglobin saturation was smaller, reflecting the finite oxygen-binding capacity of haemoglobin. The greatest increase in PaO2 was observed in patients with the worst hypoxaemia at 0 cm H2O PEEP, in whom the supplemental airway pressure caused the greatest recruitment of collapsed alveoli (Fig. 2); this was supported by the demonstration that PEEP-induced improvement in oxygenation reversed when PEEP returned to 0 cm H2O.
test increase in PaO2 was observed in patients with the worst hypoxaemia at 0 cm H2O PEEP, in whom the supplemental airway pressure caused the greatest recruitment of collapsed alveoli (Fig. 2); this was supported by the demonstration that PEEP-induced improvement in oxygenation reversed when PEEP returned to 0 cm H2O. Cardiac output decreased relatively consistently in all subjects in response to incremental PEEP. This was attributable to a combination of ventricular splinting by the distended lung, reducing right ventricular preload, and increased pulmonary vascular resistance, increasing right ventricular afterload; both serving to reduce right ventricular ejection and left ventricular filling.24 In all patients, the PEEP-induced reduction in cardiac output outweighed the induced improvement in oxygenation, such that increasing PEEP decreased DO2 in every modelled circumstance. There was a decrease in dynamic strain ratios during the PEEP trial, with return to baseline levels on removal of PEEP. The degree of dynamic strain reduction appeared to be related to the presence of collapsed but recruitable alveolar units, and this is best illustrated graphically, where strain is plotted against recruitment (Fig. S7 in the Supplementary appendix). Examination of the individual patient data shows that an increase in PEEP caused strain to increase, until the threshold opening pressure was achieved for the collapsed lung units; when new alveolar units opened, the total distending force was distributed amongst more alveoli, thereby reducing the average strain for the whole lung.
f the individual patient data shows that an increase in PEEP caused strain to increase, until the threshold opening pressure was achieved for the collapsed lung units; when new alveolar units opened, the total distending force was distributed amongst more alveoli, thereby reducing the average strain for the whole lung. Death in patients with ARDS rarely appears to be attributable to respiratory failure per se; rather, the majority of deaths are attributable to the underlying ARDS trigger or disease progression to systemic inflammatory response syndrome or multiple organ failure.39 This is caused in part by biological lung trauma (‘biotrauma’) caused by alveolar epithelial damage, resulting in the release of pulmonary cytokines,39 but may also be compounded by systemic release of inflammatory cytokines secondary to inadequate organ perfusion. Findings from this investigation offer an explanation for the apparent lack of mortality benefit in studies examining high-PEEP strategies in patients with ARDS,13–16 and for the fact that subgroup analyses indicate that those with the severest disease may benefit most from PEEP.12 Careful examination of each in silico patient demonstrates that those with the worst starting hypoxaemia and greatest number of recruitable alveolar compartments exhibited the greatest improvement in oxygenation compared with reduction in cardiac output. Likewise, those with the largest number of recruitable compartments also exhibited the greatest reduction in the average dynamic lung strain.
he worst starting hypoxaemia and greatest number of recruitable alveolar compartments exhibited the greatest improvement in oxygenation compared with reduction in cardiac output. Likewise, those with the largest number of recruitable compartments also exhibited the greatest reduction in the average dynamic lung strain. The recent mediation analysis of large randomized control studies by Amato and colleagues40 has suggested that high-PEEP strategies may not always be protective, and high plateau pressures may not necessarily add mortality risk in ARDS. The protocol for our modelling study was based on stepwise escalation of PEEP with fixed driving pressure (using a pressure-controlled mode of ventilation), and we were therefore unable to investigate variation in our recordings based on fixed PEEP vs matching plateau pressure. However, the results do demonstrate different proportions of alveolar compartment recruitment and dynamic lung strain on an individual patient basis for an almost uniform set of driving pressures across the group. This is most evident in Figs S25, S26, and S27 in the Supplementary appendix. Modelling may provide the ideal opportunity to examine further the variation in driving pressure during ventilation in ARDS and its relationship to oxygenation, ventilation, oxygen delivery, and lung mechanics.
ing pressures across the group. This is most evident in Figs S25, S26, and S27 in the Supplementary appendix. Modelling may provide the ideal opportunity to examine further the variation in driving pressure during ventilation in ARDS and its relationship to oxygenation, ventilation, oxygen delivery, and lung mechanics. Our demonstration of a consistent decrease in DO2 in response to PEEP in an in silico population is novel. The influence of PEEP on cardiovascular performance outweighed, in every patient, the improvement in oxygenation in an experimental population with deliberately varied ventilation–perfusion mismatching and alveolar collapse. The notion that PEEP can be categorized in terms of improvement in oxygenation, reduction of shunt fraction, and impact on the cardiovascular system is not in itself new.41 There seems to be agreement that PEEP should be an addition to ventilation strategy in ARDS,42–44 although there is still much debate concerning the determination of the correct level of PEEP for use in patients with ARDS.45 Titration of PEEP to cardiovascular performance requires estimation of the cardiac output; however, the use of the ‘gold-standard’ pulmonary artery catheter has decreased dramatically over the last decade,46,47 and despite surveys of practice indicating the facility to use flow monitoring in critical care, only a small proportion of critical care units seem to do so on a routine basis.48
the cardiac output; however, the use of the ‘gold-standard’ pulmonary artery catheter has decreased dramatically over the last decade,46,47 and despite surveys of practice indicating the facility to use flow monitoring in critical care, only a small proportion of critical care units seem to do so on a routine basis.48 Our study had several limitations that should be noted. The model was calibrated and validated against data from historical studies. It is plausible that the cardiovascular side-effects of historical drugs and dosage required to produce levels of sedation deep enough to allow the traditional high tidal volume ventilation, with or without neuromuscular block, may completely obtund normal cardiovascular system baroreceptor reflexes. Indeed, the aforementioned historical studies consistently report that heart rate did not change significantly throughout the duration of their interventions;17–20 it was therefore deemed appropriate to fix the heart rate at 100 beats min−1 for this investigation. The importance of atrial contraction for ventricular filling is poorly understood. The majority of experimental data relate to lower mammalian studies, human studies with small numbers of volunteers, and stable patients, with possible confounding factors from study design and timing of follow-up.49 Consequently, in this study we used a lumped model of atrial and ventricular contraction; in the context of a fixed heart rate and the assumption of sinus rhythm and normal heart valves, we feel that this has minimal impact on the applicability of our results.
unding factors from study design and timing of follow-up.49 Consequently, in this study we used a lumped model of atrial and ventricular contraction; in the context of a fixed heart rate and the assumption of sinus rhythm and normal heart valves, we feel that this has minimal impact on the applicability of our results. Conclusion The highly integrated cardiopulmonary model used in this study was able to match accurately the cardiorespiratory interactions of individual patients with ARDS receiving mechanical ventilation, allowing in-depth and controlled investigation of key outcome parameters using data that may not be routinely scrutinized in daily clinical practice. Our results show that changing the ventilation strategy to improve commonly monitored oxygenation indices and increase alveolar protection by preserving the open lung may, counterintuitively, be at the hidden expense of reducing global oxygen delivery, particularly in patients with less severe ARDS. In clinical practice, PEEP-response trials should ideally include measurement and titration to DO2 in order to allow personalized application of optimal PEEP to maximize alveolar protection while minimizing the reduction in global oxygen delivery. Such a personalized approach might yield substantial improvements in the effectiveness of our existing therapeutic strategies. Authors’ contributions Design of pulmonary model: J.G.H. Pulmonary model development and modelling of disease: A.D., W.W., D.G.B., J.G.H. Development and integration of cardiovascular model: M.H., J.G.H., A.D., M.C. Patient matching: A.D., W.W., D.G.B.
Our results show that changing the ventilation strategy to improve commonly monitored oxygenation indices and increase alveolar protection by preserving the open lung may, counterintuitively, be at the hidden expense of reducing global oxygen delivery, particularly in patients with less severe ARDS. In clinical practice, PEEP-response trials should ideally include measurement and titration to DO2 in order to allow personalized application of optimal PEEP to maximize alveolar protection while minimizing the reduction in global oxygen delivery. Such a personalized approach might yield substantial improvements in the effectiveness of our existing therapeutic strategies. Authors’ contributions Design of pulmonary model: J.G.H. Pulmonary model development and modelling of disease: A.D., W.W., D.G.B., J.G.H. Development and integration of cardiovascular model: M.H., J.G.H., A.D., M.C. Patient matching: A.D., W.W., D.G.B. Writing and final approval of manuscript: M.C., A.D., M.H., W.W., D.G.B., J.G.H. Declaration of interest None declared. Funding UK Medical Research Council (grant number MR/K019783/1). Supplementary material Supplementary material is available at British Journal of Anaesthesia online.
Editor’s key points Experimental evidence supports a benefit to red blood cell (RBC) washing to reduce inflammatory factors before transfusion. In a randomized trial of washed and standard unwashed RBCs in high-risk cardiac surgery patients, the experimental benefit was not replicated. Owing to the limited power of this trial, larger studies are necessary to test the hypothesis that RBC washing is beneficial.
Experimental evidence supports a benefit to red blood cell (RBC) washing to reduce inflammatory factors before transfusion. In a randomized trial of washed and standard unwashed RBCs in high-risk cardiac surgery patients, the experimental benefit was not replicated. Owing to the limited power of this trial, larger studies are necessary to test the hypothesis that RBC washing is beneficial. Organ injury associated with red blood cell (RBC) transfusion has been attributed to a ‘storage lesion’; a progressive disruption of erythrocyte homeostasis associated with depletion of high-energy phosphates during storage that results in accumulation of microparticles and other inflammatory substances in the supernatant of RBCs.1 Experimental studies implicate platelet and monocyte activation by RBC microparticles, and endothelial dysfunction as a consequence of altered haem metabolism, in transfusion-associated organ injury,2–5 and suggest that removal of the storage supernatant by cell washing attenuates inflammatory responses and organ dysfunction.26 In support of these findings, washing of allogeneic RBCs has been shown to attenuate inflammation in children undergoing cardiac surgery.7 We tested the hypothesis that allogeneic RBC washing attenuates inflammation and organ failure in adult cardiac surgery patients receiving large-volume transfusions. In a prespecified substudy, we tested the hypothesis that RBC washing attenuates platelet and leucocyte activation by removing inflammatory RBC microparticles. We also assessed whether cell-free haemoglobin (Hgb) release by RBCs after washing results in endothelial activation.
ients receiving large-volume transfusions. In a prespecified substudy, we tested the hypothesis that RBC washing attenuates platelet and leucocyte activation by removing inflammatory RBC microparticles. We also assessed whether cell-free haemoglobin (Hgb) release by RBCs after washing results in endothelial activation. Methods The Red Cell Washing for the Attenuation of Organ Injury Following Cardiac Surgery (REDWASH) trial was a multicentre, single-blinded, parallel-group, randomized controlled trial of washing of allogeneic RBCs before transfusion vs standard care (no washing). The trial had ethical approval (REC Reference 12/EM/0475) and was registered (ISRCTN 27076315). The trial protocol has been published;8 changes to the study design after trial commencement are listed in the online Supplementary material. The main trial was terminated by the funder in March 2015 because of slow recruitment. This report includes the results of a prespecified mechanistic substudy planned for the first 60 patients recruited.8 A detailed description of the study methods is available as online Supplementary material. Patients Adults (≥16 yr of age) undergoing cardiac surgery with blood cardioplegia identified as representing a high-risk group for large-volume blood transfusion (LVBT) using a modified risk score9 (score ≥25) were eligible for inclusion. Exclusion criteria are listed in Supplementary Table S1.
Methods The Red Cell Washing for the Attenuation of Organ Injury Following Cardiac Surgery (REDWASH) trial was a multicentre, single-blinded, parallel-group, randomized controlled trial of washing of allogeneic RBCs before transfusion vs standard care (no washing). The trial had ethical approval (REC Reference 12/EM/0475) and was registered (ISRCTN 27076315). The trial protocol has been published;8 changes to the study design after trial commencement are listed in the online Supplementary material. The main trial was terminated by the funder in March 2015 because of slow recruitment. This report includes the results of a prespecified mechanistic substudy planned for the first 60 patients recruited.8 A detailed description of the study methods is available as online Supplementary material. Patients Adults (≥16 yr of age) undergoing cardiac surgery with blood cardioplegia identified as representing a high-risk group for large-volume blood transfusion (LVBT) using a modified risk score9 (score ≥25) were eligible for inclusion. Exclusion criteria are listed in Supplementary Table S1. Randomization and blinding Subjects were randomly assigned with concealed allocation using an Internet-based randomization system (Sealed Envelope Ltd, Medicines Healthcare Regulatory Authority (MHRA) recognized facility). Randomization was stratified by study site and type of procedure. Outcome assessors were blinded to allocation.
and blinding Subjects were randomly assigned with concealed allocation using an Internet-based randomization system (Sealed Envelope Ltd, Medicines Healthcare Regulatory Authority (MHRA) recognized facility). Randomization was stratified by study site and type of procedure. Outcome assessors were blinded to allocation. Intervention Eligible subjects who consented to participate were randomly allocated, in a 1:1 ratio, to receive either standard care (unwashed prestorage leucodepleted allogenic red blood cells) or washed red blood cells, between the commencement of surgery and 48 h after surgery. Allogeneic saline–adenine–glucose–mannitol (SAGM) stored RBC units used in the trial were issued by National Health Service Blood and Transplant (NHSBT) as per standard care. The Continuous AutoTransfusion System (CATS™; Fresenius AG, Bad Homburg, Germany) was used on the basis that low g force centrifugation with this device minimizes RBC trauma.1011 For the intervention, RBCs were washed in theatre or at the patient’s bedside with saline using the Quality Mode and immediately administered to the patient. Washed units were not stored for future use. The haematocrit threshold for transfusion was 23. A major protocol violation was defined as receipt of only unwashed blood for subjects randomized to receive washed RBCs, or the receipt of only washed blood in patients randomized to receive standard care.
red to the patient. Washed units were not stored for future use. The haematocrit threshold for transfusion was 23. A major protocol violation was defined as receipt of only unwashed blood for subjects randomized to receive washed RBCs, or the receipt of only washed blood in patients randomized to receive standard care. Outcomes The primary outcome was severity of the systemic inflammatory response as indicated by serum interleukin (IL)-8 measured at baseline and at four postsurgery time points. We have previously shown that IL-8 is increased in transfused patients.812 Secondary outcomes are described in Supplementary Table S2. For the mechanism study, serum, platelet-poor plasma, and urine samples were collected at baseline and a serial time points. Microparticles in storage supernatant and platelet-poor plasma were characterized using flow cytometry (Cyan ADP; Beckman Coulter, Brea, CA USA). Cell-free Hgb and plasma total and non-transferrin-bound iron were measured in the supernatant of RBC units and in plasma as described.13–15 Platelet activation [platelet activating complex (PAC)-1 (BD Biosciences, Abingdon, Oxford, UK), P selectin/CD62P (Abcam, Cambridge, UK), and PE-coupled CD41 (Affymetrix, Santa Clara, CA, USA)] and leucocyte activation markers (CD64, CD163; Affymetrix) were measured using flow cytometry (Cyan ADP; Beckman Coulter) in whole blood. Platelet activation was also measured indirectly in whole blood using Multiplate aggregometry (Roche Diagnostics International Ltd, Rotkreuz, Switzerland). Serum bilirubin was measured on the Siemens Advia 2400 Chemistry System (Siemens, Frimley, UK). Hepcidin was measured using an enzyme-linked immunosorbent assay (Abbexa, Cambridge, UK). Serum intercellular adhesion molecule (ICAM)-1 was measured using multiplex assays on the MAGPIX (Luminex Corporation, Austin, TX, USA). Reactive oxygen species concentrations, protein carbonyl content, and thiobutiric acid reactive substances (TBARS) were measured with the following commercially available kits: OxiSelect (Cell Biolabs, Inc., San Diego, CA, USA), Parameter TBARS assay (R&D Systems, Abingdon, Oxford, UK), and carbonyl content assay kit (Abcam).
e oxygen species concentrations, protein carbonyl content, and thiobutiric acid reactive substances (TBARS) were measured with the following commercially available kits: OxiSelect (Cell Biolabs, Inc., San Diego, CA, USA), Parameter TBARS assay (R&D Systems, Abingdon, Oxford, UK), and carbonyl content assay kit (Abcam). Statistical analyses As trial recruitment was terminated early, the comparison of the primary end point (IL-8) is reported here. Secondary outcomes are reported in the online Supplementary material. The mechanism substudy was exploratory; therefore, no sample size calculation was performed. The analysis was performed on an intention-to-treat basis on all randomized subjects who entered the trial (underwent surgery) and had the primary outcome measured at one time point or more (including baseline). Data are presented as the mean (sd) for normally distributed data, median (interquartile range, IQR) for non-normally distributed data, or n (%). Means for continuous outcomes for clinical and experimental data (transformed logarithmically if required) were compared using mixed effects models, adjusting for baseline values where available. Treatment estimates were reported as effect sizes with 95% confidence intervals (CIs). A statistical analysis plan was written before database lock (see online Supplementary material). The analysis was performed with SAS version 9.4 (SAS Institute Inc., Cary, NC, USA).
ls, adjusting for baseline values where available. Treatment estimates were reported as effect sizes with 95% confidence intervals (CIs). A statistical analysis plan was written before database lock (see online Supplementary material). The analysis was performed with SAS version 9.4 (SAS Institute Inc., Cary, NC, USA). Results Trial cohort Sixty subjects were recruited and randomized at three UK centres between May 2013 and March 2015. After withdrawals (Fig. 1A), the analysis population comprised 29 allocated to the standard care arm and 27 to the washing arm. There were four severe protocol deviations, two in the standard care group and two in the washing group. Fifty-four subjects were eligible for follow-up at 3 months. Overall, the mean age of participants was 74 yr [Range 51-90 years], and 28 (50.0%) were female (Supplementary Table S3). Generally, there was good balance between groups; however, by chance there was a lower baseline estimated glomerular filtration rate (eGFR ml .min-1. 1.73m-2) in the washing group [68 (24) vs 90 (30)] in the standard care arm. The types of procedures and the bypass and cross-clamp times were well matched between groups (Supplementary Table S4). Fig 1 (A) CONSORT diagram of REDWASH trial. (B) Haemoglobin concentrations [mean (sd)] in samples collected from REWASH participants. (C) Number of red blood cell (RBC) units transfused during the trial, during cardiac surgery, and after surgery. (D) Histogram of the transfused RBC unit age distribution.
ary Table S4). Fig 1 (A) CONSORT diagram of REDWASH trial. (B) Haemoglobin concentrations [mean (sd)] in samples collected from REWASH participants. (C) Number of red blood cell (RBC) units transfused during the trial, during cardiac surgery, and after surgery. (D) Histogram of the transfused RBC unit age distribution. Transfusion and blood loss The median number of units transfused was 3 (2–5) in the standard care arm compared with 4 (2–6) in the washing arm (P=0.04; Supplementary Table S5). In the washing arm, one subject did not receive any red cells and two received only unwashed RBCs. The distribution of timings (intra- vs postoperative) of RBC transfusions was similar between groups (Fig. 1B). The mean storage age of the RBCs was 20 (4.1) and 22 (5.6) days in the standard care and washing arms, respectively (P=0.078; Fig. 1C and Supplementary Table S5). There was no difference between groups with respect to blood loss (4 and 12 h; Supplementary Table S5), serial haemoglobin concentrations (Fig. 1D), or exposure to non-RBC blood components (Supplementary Table S5).
) days in the standard care and washing arms, respectively (P=0.078; Fig. 1C and Supplementary Table S5). There was no difference between groups with respect to blood loss (4 and 12 h; Supplementary Table S5), serial haemoglobin concentrations (Fig. 1D), or exposure to non-RBC blood components (Supplementary Table S5). Primary and secondary outcomes There was no difference between groups for the primary outcome, serum IL-8 [adjusted mean difference, standard vs washing, 0.239 (−0.231, 0.709), P=0.318; Fig. 2A]. There was no difference between groups with respect to alternative biomarkers of leucocyte and endothelial activation (Supplementary Table S6). There was no difference between groups with respect to MOD scores, worst MOD scores, serial arterial partial pressure of oxygen/inspired oxygen fraction (PaO2/FIO2) ratios, or serial troponin measurements (Fig. 2B–D). Serial creatinine values and urinary neutrophil gelatinase-associated lipocalcin values [logarithmically adjusted mean difference 0.435 (95% CI 0.022, 0.849), P=0.039; Fig. 2E] were significantly higher in the washing group, and calculated creatinine clearance was lower (Fig. 2F). However, differences in kidney function and injury were not statistically significant when adjusted for the difference in baseline eGFR (Supplementary Table S7). Clinical outcomes and resource use in the analysis population and adverse events in the safety population are reported for completeness16 in Supplementary Tables S8 and S9.. Fig 2 Inflammation and organ injury. (A) Serum interleukin-8 (IL-8). (B) Multiple organ dysfunction scores (MODS). (C) Arterial partial pressure of oxygen/inspired oxygen fraction (PaO2/FIO2) ratio. (D) Serum troponin. (E) Urine neutrophil gelatinase-associated lipocalcin (NGAL). (F) Serum creatinine clearance. Values represent the mean (sd). Findings were not different in prespecified per protocol and safety analyses.
ores (MODS). (C) Arterial partial pressure of oxygen/inspired oxygen fraction (PaO2/FIO2) ratio. (D) Serum troponin. (E) Urine neutrophil gelatinase-associated lipocalcin (NGAL). (F) Serum creatinine clearance. Values represent the mean (sd). Findings were not different in prespecified per protocol and safety analyses. Fig 3 Effects of washing on the red blood cell (RBC) storage lesion in vitro. (A) Fractional changes in CD235a/annexin V (AV)-positive microparticles during storage and after washing; eight RBC units were analysed. (B) Fractional changes in RBC-derived microparticles in 21-day-stored RBC bags in response to washing (6 units analysed before and 3 units after washing). (C) Cell-free haemoglobin (Hgb) in washed RBC units. (D) Haemoglobin concentrations after washing at day 21. (E) Adenosine triphosphate (ATP) concentrations in 21-day-old RBC bags before and after washing (6 units analysed before and 3 units after washing). (F) Osmotic fragility of 21-day-old RBCs (6 units analysed). Values represent means (sd). Fig 4 Effects of red blood cell (RBC) washing on microparticle concentrations and platelet and leucocyte activation. (A) Microparticle concentration in collected plasma samples. (B and C) Fractional content of annexin V- (AV) and CD235a-positive microparticles in plasma samples. (D) Activation of monocytes measured by expression of CD64 on monocytes (CD163). (E) Expression of integrin αIIb/β3 (PAC-1) on activated platelets. (F) Expression of CD62P on platelets (CD41). Values represent means (sd).
and C) Fractional content of annexin V- (AV) and CD235a-positive microparticles in plasma samples. (D) Activation of monocytes measured by expression of CD64 on monocytes (CD163). (E) Expression of integrin αIIb/β3 (PAC-1) on activated platelets. (F) Expression of CD62P on platelets (CD41). Values represent means (sd). Mechanism analyses Effect of washing on stored RBCs Microparticles positive for CD235a (glycophorin A, an RBC antigen) and annexin V (AV), which labels phosphatidylserine and oxidized lipids, accumulated in the supernatant of RBC units throughout storage (Fig. 3A). Mechanical washing of RBCs significantly reduced concentrations of CD235a/AV microparticle and total microparticle concentrations in units stored for 41 (Fig. 3A) and 21 days (Fig. 3B). Cell-free Hgb concentrations also increased progressively in the supernatant of RBCs during storage (Fig. 3C). Mechanical washing did not reduce concentrations of free Hgb in RBCs stored for 35 days (Fig. 3C) and doubled free Hgb concentrations in RBCs stored for 21 days (Fig. 3D). Washing did not significantly alter RBC adenosine triphosphate concentrations or osmotic fragility (Fig. 3D and E).
f RBCs during storage (Fig. 3C). Mechanical washing did not reduce concentrations of free Hgb in RBCs stored for 35 days (Fig. 3C) and doubled free Hgb concentrations in RBCs stored for 21 days (Fig. 3D). Washing did not significantly alter RBC adenosine triphosphate concentrations or osmotic fragility (Fig. 3D and E). Effects of RBC washing on microparticles and platelet and leucocyte activation There was no difference between groups with respect to serial measures of total, AV-positive, or RBC-derived (CD235a positive) microparticles in plasma up to 48 h postsurgery (Fig. 4A–C). There was no difference between groups with respect to concentrations of CD64/CD163-positive (activated) leucocytes, PAC-1, or CD40/CD62P-positive (activated) platelets, as determined by flow cytometry (Fig. 4D–F), or platelet activation as assessed by Multiplate aggregometry (data not shown).
postsurgery (Fig. 4A–C). There was no difference between groups with respect to concentrations of CD64/CD163-positive (activated) leucocytes, PAC-1, or CD40/CD62P-positive (activated) platelets, as determined by flow cytometry (Fig. 4D–F), or platelet activation as assessed by Multiplate aggregometry (data not shown). Effects of RBC washing on cell-free haemoglobin, oxidative stress, and endothelial activation Plasma cell-free Hgb concentrations were increased significantly in both washing and standard care groups immediately postsurgery (Fig. 5A). However, there was no difference between groups in plasma free Hgb, iron concentrations, serum haematocrit, serum total bilirubin, non-transferrin-bound iron, hepcidin concentrations, nitric oxide bioavailability, or indicators of endothelial activation [serum ICAM-1 and CD144 (vascular endothelial cadherin)-positive microparticles; Fig. 5B–H]. Specifically, endothelium-derived microparticles did not coexpress integrin β1 (VLA5; Fig. 5I), a putative marker of haem-mediated endothelial activation. Fig 5 Effects of washing on oxidative stress markers in REDWASH subjects. (A) Plasma cell-free haemoglobin. (B) Plasma total iron concentrations. (C) Serial haematocrit. (D) Total bilirubin. (E) Non-transferrin-bound iron. (F) Serum hepcidin. (G) Nitric oxide bioavailability. (H) Serum intercellular adhesion molecule (ICAM)-1. (I) Endothelial (CD144) microparticle concentrations in plasma. (J) Oxidative potential of cell-free haemoglobin (Hgb) derived from microparticle-free storage supernatant from stored RBCs (HS Sup) or derived from osmotically lysed stored RBCs or lysed fresh RBCs. ROS, reactive oxygen species. (K) Serum protein carbonylation. (L) Serum thiobarbituric acid reactive substances (TBARS, lipid peroxidation). Values represent means (sd).
Hgb) derived from microparticle-free storage supernatant from stored RBCs (HS Sup) or derived from osmotically lysed stored RBCs or lysed fresh RBCs. ROS, reactive oxygen species. (K) Serum protein carbonylation. (L) Serum thiobarbituric acid reactive substances (TBARS, lipid peroxidation). Values represent means (sd). The greater part of plasma free Hgb immediately postsurgery was probably not altered by RBC washing but was most probably attributable to lysis of recipient RBCs by the cardiopulmonary bypass (CPB) circuit. However, free Hgb released after washing of stored RBCs is more reactive than that released by lysed fresh RBCs, as generated by the CPB circuit (Fig. 5J), implying that similar plasma free Hgb concentrations in groups might bely qualitative differences in biological activity. However, we did not detect differences in measures of oxidative stress in serum, a putative mechanism by which free Hgb damages cells, as measured by protein carbonylation or lipid peroxidation (Fig. 5K and L).
t similar plasma free Hgb concentrations in groups might bely qualitative differences in biological activity. However, we did not detect differences in measures of oxidative stress in serum, a putative mechanism by which free Hgb damages cells, as measured by protein carbonylation or lipid peroxidation (Fig. 5K and L). Discussion Red blood cell-derived (CD235a-positive) microparticles accumulate progressively in RBC units from the start of storage and are effectively removed by mechanical washing. In the REDWASH trial, however, transfusion of washed RBCs did not reduce plasma concentrations of CD235a-positve microparticles or levels of platelet or leucocyte activation in recipients compared with standard RBCs. Cell-free Hgb also increases in the supernatant of RBC units during storage and is increased by washing. However, in the REDWASH trial there was no difference in plasma free Hgb concentrations or markers of haem/iron metabolism, endothelial activation, or oxidative stress between groups.
ts compared with standard RBCs. Cell-free Hgb also increases in the supernatant of RBC units during storage and is increased by washing. However, in the REDWASH trial there was no difference in plasma free Hgb concentrations or markers of haem/iron metabolism, endothelial activation, or oxidative stress between groups. This is the first randomized trial to evaluate the risks and benefits of RBC washing in adults. We recruited a high-risk population of cardiac surgery patients that should in principle have benefited from safer RBC transfusion. Patients in both groups received significant volumes of RBCs, 1000 ml, within 48 h of surgery and experienced significant morbidity; 86% of patients experienced a composite clinical outcome of death, sepsis, or organ failure. Subjects in the washing group received more RBC units despite similar perioperative levels of blood loss and serial serum haemoglobin concentrations. This is consistent with previous clinical and experimental studies that report loss of 10–25% of red cells during mechanical washing.610 The chief limitations of the study are the limited power of our observations and the early termination of the main trial. For these reasons, we can neither accept nor refute our primary hypotheses. Furthermore, by chance, there was a difference in baseline eGFR between the groups. These limitations notwithstanding, the randomization generated two groups that received large volumes of washed or unwashed RBCs and, after adjustment for baseline eGFR, demonstrated no difference for any measured biomarker of inflammation or organ injury. On balance, our findings argue against important clinical benefits for RBC washing. The REDWASH trial did not evaluate immunomodulation, and therefore, we are unable to comment on the effects of washing on these processes. The results of a similar trial (NCT02094118) evaluating the clinical efficacy of RBC washing will provide greater insights into these results.
ant clinical benefits for RBC washing. The REDWASH trial did not evaluate immunomodulation, and therefore, we are unable to comment on the effects of washing on these processes. The results of a similar trial (NCT02094118) evaluating the clinical efficacy of RBC washing will provide greater insights into these results. Experimental work has documented both potential risks and benefits of RBC washing. We have shown that RBC-derived microparticles activate platelets and leucocytes in vitro. Furthermore, we have shown that mechanical RBC washing in a swine model effectively removes microparticles and attenuates platelet and leucocyte sequestration in lungs and the clinical and histological correlates of transfusion-associated lung injury ( MJ Wozniak, GJ Murphy, unpublished observations). Likewise, we have shown that cell-free Hgb activates endothelial cells by a pathway previously described only in human aortic endothelial cells in the presence of minimally modified low-density lipoprotein.17 The activation is most probably mediated by lipids oxidized by increased concentrations of free Hgb18 that occurs either in the blood bags or after transfusion.19 Here, an alternative pathway results in VLA5 (integrin-α5/β1) expression, retention of alternatively spliced CS-1 fibronectin on the surface of endothelial cells, and endothelial cell–monocyte interaction.20 In swine, as in humans, RBC washing results in accelerated release of cell-free Hgb, and this results in acute endothelial injury, oxidative stress, and acute kidney injury in experimental subjects (unpublished observations). Studies in human cells have also highlighted potential sublethal damage to RBCs during mechanical washing that could potentially offset any benefits derived from reductions in RBC-derived microparticles.21
endothelial injury, oxidative stress, and acute kidney injury in experimental subjects (unpublished observations). Studies in human cells have also highlighted potential sublethal damage to RBCs during mechanical washing that could potentially offset any benefits derived from reductions in RBC-derived microparticles.21 Our experimental results were not replicated in the REDWASH trial. This could be attributable to differences between species, differences in the time period over which RBCs were transfused, or both, in porcine experiments and trial subjects. Typically, 4 units were transfused during 2 h in swine, whereas a similar volume, on average, was transfused during 2 days in the clinical trial. The porcine experiments were also conducted without CPB. These factors might have diminished the clinical benefits of washing; neither microparticle depletion nor the release of free Hgb observed after washing was reflected by differences in plasma microparticle concentrations or cell-free Hgb in subjects or by differences in inflammation and organ injury. Moreover, the peak mean IL-8 concentration in our porcine experiments after transfusion with older stored RBCs was 10 pg ml−1. In the standard care patients in the REDWASH trial, the peak IL-8 concentration was 220 pg ml−1. We speculate that the effects of washing were superseded by greater platelet and leucocyte activation, free Hgb release, and endothelial injury attributable to surgery and CPB.22 Alternatively, the mixed risks and benefits from RBC washing might have resulted in no overall difference between trial groups. Manual washing techniques that minimize RBC trauma are unlikely to address this limitation because these are time consuming. Indeed, the delay attendant on bedside washing contributed to protocol non-compliance in the present study. An alternative is RBC rejuvenation.23 This maximizes the benefits of washing by preventing production of microparticles after washing, minimizes risks by preventing RBC damage and free Hgb release, and attenuate transfusion-mediated organ injury in experimental studies (unpublished observations).
in the present study. An alternative is RBC rejuvenation.23 This maximizes the benefits of washing by preventing production of microparticles after washing, minimizes risks by preventing RBC damage and free Hgb release, and attenuate transfusion-mediated organ injury in experimental studies (unpublished observations). In conclusion, the results of the mechanism substudy of the REDWASH trial did not support our hypothesis that mechanical washing of transfused RBCs would attenuate platelet and leucocyte activation and organ injury in cardiac surgery patients. Authors’ contributions Conceived the trial and wrote the application for funding (with others): G.J.M. Designed the trial: G.J.M., M.J.W., A.H.G. Managed the conduct of the trial: W.D., S.K.B., N.B., T.K. Managed the data during the trial: W.D., T.K., T.M. Laboratory analyses: M.J.W., R.C., M.W., N.S. Statistical analyses: T.M., M.N. Drafted the report: M.J.W., G.J.M., A.H.G. All authors reviewed the report for important intellectual content and approved the final version. All authors, external and internal, had full access to all of the data (including statistical reports and tables) and can take responsibility for the integrity of the data and the accuracy of the data analysis. Supplementary material Supplementary material is available at British Journal of Anaesthesia online. Supplementary Material Supplementary Data Click here for additional data file.
All authors, external and internal, had full access to all of the data (including statistical reports and tables) and can take responsibility for the integrity of the data and the accuracy of the data analysis. Supplementary material Supplementary material is available at British Journal of Anaesthesia online. Supplementary Material Supplementary Data Click here for additional data file. Acknowledgements The authors would like to thank research teams at the recruiting centres, including Mr Mark Hickey, Mr Leon Hadjinikolaou, Dr Viktor Zlocha, and Dr Jacek Szostek, all at the Glenfield Hospital in Leicester, and members of the Cardiac Surgery Programme Steering Committee who assisted in the governance of the trial; Professor Alison Goodall, Professor Nigel Brunskill, Dr Karl Herbert, Mr Sunil Bhudia, Mr Alan Philipps, and Mr Anthony Locke. Declaration of interest None declared. The views expressed are those of the authors and do not necessarily reflect those of the British Heart Foundation, the National Institute for Health Research, or the Department of Health. Funding National Institute for Health Research Programme Grant for Applied Research (RP-PG-0407-10384 for The REDWASH trial); British Heart Foundation (RG/13/6/29947, CH/12/1/29419, and PG/11/95/29173 to G.J.M., M.W., W.D., T.K. and N.S.); Leicester Cardiovascular Biomedical Research Unit (G.J.M., M.W., W.D., T.K. and N.S.).
Editor's key points • Early-life experience, including preterm birth, may have long-term effects on pain processing. • Young adults from a longitudinal cohort study of preterm infants were matched to healthy controls. • Conditioned pain modulation was used to assess descending modulatory effects on pain processing. • Extremely preterm birth and female sex both affected baseline pain sensitivity and descending modulatory effects. • The impact of early-life experience and sex on chronic pain vulnerability needs further study. Conditioned pain modulation (CPM) assesses the ability of a noxious ‘conditioning stimulus’ to alter the sensitivity to a ‘test stimulus’ at a distant body site. Reduced sensitivity or inhibition is the most common response, but a continuum from inhibition to facilitation is possible. Differences in the directionality or degree of CPM have been suggested as a biomarker to predict persistent pain after surgery, risk of chronic pain, or individual differences in treatment response.1, 2, 3, 4 The reliability of CPM is influenced by study methodology.1 Modulation is quantified by either an alteration in pain threshold (for a variable mechanical, thermal, or electrical test stimulus) or change in pain intensity (to a fixed test stimulus).1, 2 Cold is a reliable conditioning stimulus,1, 5 but a range of temperatures (1–13°C) and immersion durations (20–180 s) have been utilized.5, 6, 7, 8 The degree of modulation can also be influenced by age,7 gender,9, 10 psychological factors,2 differences in baseline sensitivity,1 and intercurrent chronic pain.3, 11
1, 2 Cold is a reliable conditioning stimulus,1, 5 but a range of temperatures (1–13°C) and immersion durations (20–180 s) have been utilized.5, 6, 7, 8 The degree of modulation can also be influenced by age,7 gender,9, 10 psychological factors,2 differences in baseline sensitivity,1 and intercurrent chronic pain.3, 11 The evaluation of CPM after preterm birth has specific relevance. Early-life pain and tissue injury have been associated with long-term changes in somatosensory processing that can differ with the degree of prematurity, duration of hospitalisation, and pain exposure, and are also influenced by the type and intensity of the subsequent test stimulus.12 Anxiety and pain catastrophising scores are higher in preterm than age-matched term-born young adults,13, 14 and these psychological factors may also influence CPM. Reported associations between preterm birth and chronic pain in later life vary,12, 15 but CPM may improve the identification of groups with increased risk. One study reported lack of inhibitory modulation in a group of preterm children with higher neonatal pain exposure,16 but no studies have assessed CPM in a large cohort of young adults born preterm.
rm birth and chronic pain in later life vary,12, 15 but CPM may improve the identification of groups with increased risk. One study reported lack of inhibitory modulation in a group of preterm children with higher neonatal pain exposure,16 but no studies have assessed CPM in a large cohort of young adults born preterm. In an observational cohort study, we compared CPM in extremely preterm (EP) and term-born control (TC) young adults. The primary outcome was identification of descending modulatory effects (inhibition, facilitation, and no change) in EP and TC participants. Changes in test stimulus threshold over time identified the directionality and duration of CPM. As the outcome after EP birth is worse in males17 and females may have increased chronic pain prevalence and sensitivity to experimental pain stimuli,18 sex-dependent differences were assessed. Secondarily, we calculated the percentage change from baseline to quantify the CPM effect. In line with recent recommendations,12 correlations between the degree of CPM effect and current pain experience, medication use, and anxiety and pain catastrophising scores were explored.
sex-dependent differences were assessed. Secondarily, we calculated the percentage change from baseline to quantify the CPM effect. In line with recent recommendations,12 correlations between the degree of CPM effect and current pain experience, medication use, and anxiety and pain catastrophising scores were explored. Methods Study population The EPICure cohort study recruited all infants born EP before 26 weeks gestational age (GA) across 276 maternity units in the UK and Ireland from March through December 1995. Of 1185 live births, 811 reached the neonatal intensive care unit (NICU), 497 died in the hospital, 314 were discharged home, and 9 have subsequently died.19, 20 The neurodevelopmental and health outcomes have been longitudinally assessed in EP participants, with recruitment at 30 months,19 6 yr,21 11 yr,22 and now at 19 yr20 as described previously. Whilst there has been loss to follow-up because of loss of contact details or participant preference, retention over 19–20 yr has been relatively high (92% at 30 months, 68% at 6 yr, 71% at 11 yr, and 42% at 19 yr).20 Age-matched TCs were recruited at 6 and 11 yr, and have also provided longitudinal data.21 The current study at 19 yr (EPICure@19) was approved by the National Research Ethics Committee Hampshire ‘A’ (Reference: 13/SC/0514). After a written consent, the participants completed general health and cognitive questionnaires, plus respiratory, cardiovascular, and neuroimaging assessments at the University College London Hospitals Clinical Research Facility between February 2014 and October 2015. Current data for demographic variables, cognitive measures, general health, and psychological questionnaires were extracted from the main EPICure database, along with neonatal data for EP participants [weight and GA at birth, clinical risk index for babies (CRIB) score on admission to NICU, and duration of hospital stay]. CPM was performed in conjunction with quantitative sensory testing (QST) on the hand and chest wall, which, along with additional data from the pain history, is reported separately.23 Pain and sensory thresholds at 11 yr of age were previously evaluated by the same investigator (S.M.W.) in a subset of the cohort living within 2 h travel of London, but the evaluation at this younger age did not include CPM.24
est wall, which, along with additional data from the pain history, is reported separately.23 Pain and sensory thresholds at 11 yr of age were previously evaluated by the same investigator (S.M.W.) in a subset of the cohort living within 2 h travel of London, but the evaluation at this younger age did not include CPM.24 CPM was assessed in 98 EP born and 48 TC participants (Fig. 1) in a dedicated sensory testing facility at UCL Great Ormond Street Institute of Child Health, London. As the assessment included a standardised questionnaire regarding previous and current pain experiences and evaluation of sensory dysfunction related to neonatal scars, the investigator was not blinded to the group. The same standardised verbal instructions were used for testing in the same sequence by a single investigator (S.M.W.).23 The participants self-reported VAS measures on linear scales, had control of all response functions (i.e. pressing button for pressure threshold or removing hand from conditioning), and were informed that they could decline or withdraw from testing at any time. All testing was performed at the same time of the day and in the same temperature-controlled room. The participants were offered cool water, but not caffeinated drinks in the 90 min before CPM testing. Throughout this paper, the dichotomous variable of ‘sex’ is reported, as the participants were not asked to self-report their gender. Reporting follows the Strengthening the Reporting of Observational Studies in Epidemiology guidelines.64Fig 1 Flow chart of participant numbers and assessments. CPM, conditioned pain modulation; F, female; M, male.
ichotomous variable of ‘sex’ is reported, as the participants were not asked to self-report their gender. Reporting follows the Strengthening the Reporting of Observational Studies in Epidemiology guidelines.64Fig 1 Flow chart of participant numbers and assessments. CPM, conditioned pain modulation; F, female; M, male. Fig 1
are caused by distraction during hand immersion. A cold (5°C) conditioning stimulus with hand immersion for 30 s was chosen, as this evoked inhibitory CPM in healthy 12–17 yr olds,7 but is shorter than many protocols, as preterm young adults may have reduced cold-pressor tolerance (5 of 31 withdrew hand before 30 s).29 Baseline PPT was the mean of three repetitions of ascending stimuli applied over the head of the right fibula. A computer-controlled handheld 1 cm2 algometer (Somedic SENSEBox®, Sosdala, Sweden) incorporating an optical feedback system ensured a standardised increase in pressure (ramp of 40 kPa s−1 to a maximum 1000 kPa). The participants pressed a response button when pain/discomfort was perceived. For values at 15 (during conditioning), and 50 and 90 s (after cessation of cold immersion), a single ascending stimulus was applied, with the participants asked to press the button when pain/discomfort was experienced at the same level as baseline. The left hand was immersed up to the wrist with the palm down and fingers spread into a 5°C circulating water bath (Techne TE-10D Thermoregulator B-8 Bath and RU-200 Dip Cooler; Techne, Burlington, NJ, USA). The participants were instructed to leave their hand in the water for 30 s, or until the stimulus became too uncomfortable/painful.7 The duration of immersion was recorded and subjects rated the intensity of hand discomfort (0–10 verbal rating scale) on removal. Survival curves for cold-pressor tolerance are reported,23 but, here, data were split based on durations of more, or less than, 20 s.
or until the stimulus became too uncomfortable/painful.7 The duration of immersion was recorded and subjects rated the intensity of hand discomfort (0–10 verbal rating scale) on removal. Survival curves for cold-pressor tolerance are reported,23 but, here, data were split based on durations of more, or less than, 20 s. Questionnaire-based outcomes The participants marked VASs (100 mm line) to score average pain in the last week and pretest anxiety (following description of CPM protocol).7 Analgesic-use data were extracted from the pain history (S.M.W.),23 and additional data, including medications, were extracted from general health (J.B.), and cognitive and psychology questionnaires (H.O.) collected at the University College London Hospital. The Health Utilities Index Mark 3 (HUI-3)30 includes self-reported pain ranked as 1=free of pain/discomfort, 2=mild/moderate pain that prevents no activities, 3=moderate pain that prevents a few activities, 4=moderate/severe pain that prevents some activities, and 5=severe pain that prevents most activities. The Diagnostic and Statistical Manual of Mental Disorders anxiety t-score (range: 50–100; score ≥70 clinically significant) was obtained from the Achenbach Adult Self-Report.31 The pain catastrophising scale (PCS)32 rates rumination, magnification, and helplessness (total score: 0–52). Wechsler Abbreviated Scale of Intelligence Full Scale IQ (FSIQ) scores were obtained for all participants.33
range: 50–100; score ≥70 clinically significant) was obtained from the Achenbach Adult Self-Report.31 The pain catastrophising scale (PCS)32 rates rumination, magnification, and helplessness (total score: 0–52). Wechsler Abbreviated Scale of Intelligence Full Scale IQ (FSIQ) scores were obtained for all participants.33 Statistical analysis As the EPICure study aimed to recruit the maximum available subjects from this longitudinal cohort and multiple health outcomes were being assessed, no a priori power analysis was performed for individual evaluations, such as CPM. In previous CPM studies in healthy young adults, a 5.3% change in PPT with the same algometer was deemed a meaningful CPM effect,6 and the current methodology (variable pressure test stimulus and cold conditioning stimulus) had high reliability and the lowest sample size (n=17) for detecting significant CPM effects during conditioning (90% power; α=0.5).27 The 95% confidence intervals (CIs) for observed changes in PPT are reported in the paper.
,6 and the current methodology (variable pressure test stimulus and cold conditioning stimulus) had high reliability and the lowest sample size (n=17) for detecting significant CPM effects during conditioning (90% power; α=0.5).27 The 95% confidence intervals (CIs) for observed changes in PPT are reported in the paper. Data were analysed using SPSS® version 23 (IBM, Portsmouth, UK), and plotted in Prism version 7 (GraphPad, San Diego, CA, USA); P<0.05 was considered statistically significant. Categorical data were compared with Pearson's χ2 test or Fisher's exact for smaller samples. After the assessment of normality (D'Agostino and Pearson test), group comparisons were analysed with Student's unpaired two-tailed t-test or Mann–Whitney U-test. Consistent with previous reports in adolescents and adults,34, 35 raw PPT data (kPa) were log-transformed. The resultant normally distributed data were analysed for main effects with three-way mixed-design analysis of variance (anova) with two between-subject factors (EP status and sex), and a repeated measures factor of time; degrees of freedom were corrected with Huynh–Feldt estimates of sphericity, and P values with Bonferroni adjustment for multiple comparisons.6 As sex influences the neurodevelopmental outcome after preterm birth (increased mortality and morbidity in EP males contribute to higher number of females in the sample)17 and pain response,18 we also evaluated sex differences. To display the time course of modulation and sex-dependent group differences, changes in ln PPT are graphed separately in males and females to identify the change from baseline PPT, and analysed by two-way repeated measures anova with group and time as variables, and multiplicity-adjusted P values and Bonferroni post hoc comparisons are reported. The percentage change from baseline PPT (kPa) was calculated [(PPTx seconds – PPTbaseline/PPTbaseline) × 100] as previously reported,6 and provided a normally distributed measure of the degree and direction of change (i.e. ‘CPM effect’). For CPM effect (% change in PPT at 15 s), stepwise linear regression models included candidate variables from Spearman's correlations and prior literature.8, 36
PPTbaseline/PPTbaseline) × 100] as previously reported,6 and provided a normally distributed measure of the degree and direction of change (i.e. ‘CPM effect’). For CPM effect (% change in PPT at 15 s), stepwise linear regression models included candidate variables from Spearman's correlations and prior literature.8, 36 Results CPM was assessed in 98 EP [born between 22.1 and 25.9 (24.9, 0.8; mean, SD) weeks GA, at weight 732,128 g] and 48 TC participants (Table 1). From those attending for somatosensory evaluation, three male EP participants with variable baseline responses (two reported difficulty with numerical scales and prior neonatal surgery, and one was tired and reported difficulty concentrating) had sensory data excluded, and one EP female with Raynaud's symptoms declined cold water immersion.Table 1 Participant characteristics and outcomes based on preterm status and conditioning tolerance. Data are presented as: mean (standard deviation), median [inter-quartile range], or (%). ADHD, attention deficit hyperactivity disorder medication (methylphenidate); antidepr., antidepressant medications (citalopram, fluoxetine, and mirtazapine); Anxiety (Ach), anxiety total score Achenbach Youth Self-Report scale; EP, extremely preterm; F, female; FSIQ, full-scale intelligence quotient Wechsler Abbreviated Scale of Intelligence; Internalising, subscale score Achenbach Youth Self-Report scale; M, male; PCS, pain catastrophising scale total score; PPT, pressure pain threshold; TC, term-born control; VRS, verbal rating scale. ∗Group demographic data (EP vs TC) are presented for all EP (n=102) and TC (n=48) in a separate paper23 with a subgroup analysis based on sex-dependent differences rather than conditioning stimulus tolerance. †Surgery included: patent ductus arteriosus ligation (8F; 4M), inguinal hernia repairs (1F; 7M), laparotomy (3F; 2M), and others (1F; 2M). ‡Co-codamol, paracetamol, and codeine. ¶Others: migraine prophylaxis (immersion >20 s); azathioprine for Crohn's disease and associated abdominal pain (immersion <20 s). §P-values from Student's two-tailed unpaired t-test. ||P-values from two-sided Fisher's exact test. #P-values by Mann–Whitney U-test. ∗∗For measures with missing data, available numbers are listed
: migraine prophylaxis (immersion >20 s); azathioprine for Crohn's disease and associated abdominal pain (immersion <20 s). §P-values from Student's two-tailed unpaired t-test. ||P-values from two-sided Fisher's exact test. #P-values by Mann–Whitney U-test. ∗∗For measures with missing data, available numbers are listed Table 1 Extremely preterm Term-born control P Extremely preterm P Term-born control P Conditioning ≥20 s Conditioning <20 s Conditioning ≥20 s Conditioning <20 s n=98 n=48 n=67 n=31 n=41 n=7 Characteristics∗ Age (yr) [range] 19.3 (0.6) [18.4–20.5] 19.2 (0.5) [18.1–20.1] 0.3§ 19.2 (0.6) [18.4–19.3] 19.3 (0.8) [18.5–20.5] 0.5§ 19.1 (0.5) [18.3–20.1] 19.2 (0.6) [18.1–19.8] 0.6§ Height (cm) 163.6 (9.2) 167.3 (8.9) 0.02§ 165.2 (9.1) 160.1 (8.5) 0.01§ 167.3 (9.3) 167.3 (7.1) 0.9§ Weight (kg) 63.0 (13.9) 67.8 (15.6) 0.06§ 64.1 (13.2) 60.7 (15.2) 0.3§ 67.3 (14.5) 70.9 (22.3) 0.6§ BMI (kg m−2) 23.5 (4.5) 24.1 (4.7) 0.4§ 23.5 (4.7) 23.5 (4.3) 0.9§ 23.9 (4.1) 25.3 (7.5) 0.5§ Gender; F:M (%F) 61:40 (60%) 29:19 (60%) 0.9|| 35:32 (52%) 25:6 (81%) <0.01|| 25:16 (61%) 4:3 (57%) 0.9|| Neonatal surgery 28 (13F; 15M) 0 <0.01|| 17 (5F; 12M) 11 (8F; 3M) 0.3|| 0 0 Sensory data Baseline PPT (kPa) 241 [153–376] 213 [160–315] 0.6# 336 [290–382] 154 [107–235] <0.01# 206 [158–310] 299 [181–424] 0.3# Baseline PPT, ln 5.5 (0.6) 5.4 (0.5) 0.6§ 5.6 (0.6) 5.1 (0.5) <0.01§ 5.4 (0.5) 5.6 (0.5) 0.3§ Immersion time (s) 30 [14–30] 30 [28–30] 0.02# 30 [30–30] range=20–30 12 [8–14] <0.01# 30 [30–30] range=21–30 12 [11–13] <0.01# Immersion pain, VRS 0–10 8 [7.3–10] 8 [7–9] 0.06# 8 [7–10] 10 [8–10] <0.01# 8 [7–9] 8 [7–10] 0.7# Questionnaires Average pain last week, VAS 0–100 mm 16 (23) 14 (18) 0.5§ 14 (20) 21 (28) 0.3§ 14 (19) 16 (13) 0.1§ Pretest anxiety, VAS 0–100 mm 7 (16) 1 (3) 0.014§ 6 (16) 7 (16) 0.8§ 0.6 (2) 3 (6) 0.6§ FSIQ score 88 (14) 104 (10) <0.01§ 88 (14) 90 (14) 0.5§ 105 (10) 98 (7) 0.06§ Pain catastrophising (PCS) 5 [0–14.5] (n=89)∗∗ 5 [0–14] (n=45)∗∗ 0.5# 8.8 [5.8–11.8] (n=61)∗∗ 6.5 [0–15.5] (n=28)∗∗ 0.6# 5 [0–12] (n=39)∗∗ 4.5 [0–28] (n=6)∗∗ 0.8# DSM Anxiety (Ach) 56 [50–58] (n=93)∗∗ 50 [50–54] (n=45)∗∗ 0.011# 52 [50–58] (n=65)∗∗ 55 [50–61] (n=28)∗∗ 0.6# 50 [50–54] (n=39)∗∗ 50 [50–57] (n=6)∗∗ 0.8# Medication Analgesia use None, 70 (71%); occasional, 20 (20%); regular, 6 (6%) (n=97)∗∗ None, 37 (77%); occasional, 10 (21%); regular, 1 (2%) 0.4|| None, 50 (75%); occasional, 11 (15%); regular, 5 (8%) None, 19 (%); occasional, 9 (%); regular, 2 (%) (n=30) 0.3|| None, 30 (73%); occasional, 10
∗ 50 [50–57] (n=6)∗∗ 0.8# Medication Analgesia use None, 70 (71%); occasional, 20 (20%); regular, 6 (6%) (n=97)∗∗ None, 37 (77%); occasional, 10 (21%); regular, 1 (2%) 0.4|| None, 50 (75%); occasional, 11 (15%); regular, 5 (8%) None, 19 (%); occasional, 9 (%); regular, 2 (%) (n=30) 0.3|| None, 30 (73%); occasional, 10 (22%); regular, 1 (2%) None, 7; occasional, 0 (%); regular, 0 (%) 0.3|| Analgesia type Paracetamol, 19; NSAID, 5; gabapentin and co-codamol,‡ 1; others,¶ 2 Paracetamol, 6; NSAID, 4; NSAID and co-codamol, 1 Paracetamol, 12; NSAID, 3; gabapentin and co-codamol, 1; others, 1 Paracetamol, 7; NSAID, 2; gabapentin and co-codamol, 1; others, 1 Paracetamol, 6; NSAID, 4; NSAID + co-codamol, 1 None Psychotropic medication 9 (10%) (7F, 2M) (n=92)∗∗ 1 (2%) (1M) (n=44)∗∗ 0.17|| Antidepr., 5 (4F; 1M); ADHD, 1F Antidepr., 3 (2F; 1M) Antidepr., 1 (M) None
AID, 3; gabapentin and co-codamol, 1; others, 1 Paracetamol, 7; NSAID, 2; gabapentin and co-codamol, 1; others, 1 Paracetamol, 6; NSAID, 4; NSAID + co-codamol, 1 None Psychotropic medication 9 (10%) (7F, 2M) (n=92)∗∗ 1 (2%) (1M) (n=44)∗∗ 0.17|| Antidepr., 5 (4F; 1M); ADHD, 1F Antidepr., 3 (2F; 1M) Antidepr., 1 (M) None Baseline sensitivity Baseline PPT over the fibula head did not differ significantly between the EP and TC groups (Table 1). However, within the EP group, baseline PPT was lower in females than males [5.2 ln kPa (95% CI: 5.1–5.4) vs 5.8 (95% CI: 5.7–6.0)], with the greatest difference in EP males with prior neonatal surgery [6.0 ln kPa (95% CI: 5.7–6.2); n=15], as also reported for PPT on the digit in these participants.23 Tolerance of cold conditioning was significantly shorter in EP participants, and a higher proportion of EP [EP vs TC: 31/98 (32%) vs 7/48 (85%)] participants withdrew the hand before 20 s (P=0.03) (Table 1). Shorter duration of immersion correlated with a higher pain score on hand removal and lower baseline PPT (Table 2), predominantly in EP participants (Supplementary Table S2).Table 2 Correlations between sensory variables, current pain, psychological variables, and medication use (all participants; n=146). CPM, conditioned pain modulation; DSM-Anxiety, anxiety t-score Achenbach Youth Self-Report scale; HUI-3, Health Utilities Index Mark 3; PCS, pain catastrophising scale total score; PPT, pressure pain threshold; VRS, verbal rating scale. Data = two-tailed Spearman's rho bivariate correlation coefficient *P<0.05 **P<0.01
CPM, conditioned pain modulation; DSM-Anxiety, anxiety t-score Achenbach Youth Self-Report scale; HUI-3, Health Utilities Index Mark 3; PCS, pain catastrophising scale total score; PPT, pressure pain threshold; VRS, verbal rating scale. Data = two-tailed Spearman's rho bivariate correlation coefficient *P<0.05 **P<0.01 Table 2 Baseline PPT Immersion time Immersion pain CPM % Pain ranking Regular analgesia PCS Anxiety Regular psychotropics Baseline PPT (ln kPa) 1.0 Immersion time (s) 0.30∗∗ 1.0 Immersion pain (VRS) –0.27∗∗ –0.36∗∗ 1.0 Conditioned pain modulation % (15 s) –0.30∗∗ –0.20∗∗ –0.06 1.0 Pain ranking (HUI-3) [n=139] –0.12 –0.08 0.05 –0.01 1.0 Regular analgesia –0.11 –0.09 0.07 0.03 0.31∗∗ 1.0 Pain catastrophising [n=134] –0.15 –0.04 0.12 –0.16 0.23∗∗ 0.26∗∗ 1.0 DSM-Anxiety [n=138] –0.08 –0.05 0.11 –0.02 0.27∗∗ 0.20∗ 0.40∗∗ 1.0 Regular psychotropics [n=135] –0.06 –0.01 0.16 0.10 0.19∗ 0.20∗ 0.14 0.33∗∗ 1.0
n ranking (HUI-3) [n=139] –0.12 –0.08 0.05 –0.01 1.0 Regular analgesia –0.11 –0.09 0.07 0.03 0.31∗∗ 1.0 Pain catastrophising [n=134] –0.15 –0.04 0.12 –0.16 0.23∗∗ 0.26∗∗ 1.0 DSM-Anxiety [n=138] –0.08 –0.05 0.11 –0.02 0.27∗∗ 0.20∗ 0.40∗∗ 1.0 Regular psychotropics [n=135] –0.06 –0.01 0.16 0.10 0.19∗ 0.20∗ 0.14 0.33∗∗ 1.0 Detection of modulatory effect Twenty seconds was chosen as a cut-off for adequate conditioning. This is the minimum duration previously reported in CPM protocols, and ensured hand removal occurred after, rather than during or before, the parallel 15 s PPT measurement. The duration of immersion between 20 and 30 s reliably evoked modulation (at least 10% change in PPT) in all TC participants (38, inhibition; 3, facilitation) and in 64 EP participants (64, inhibition; 3, no change). After 20 s of conditioning, group data confirm significant increases at 15 s in both TC [5.4 (95% CI: 5.2–5.5) to 5.8 (95% CI: 5.6–6.0) ln kPa] and EP [5.6 (95% CI: 5.5–5.8) to 6.0 (5.9–6.2)] groups, which were maintained beyond cessation of the stimulus (Fig. 2a). There was a significant main effect of EP status (F1,104=4.8; P=0.03), time (F2.8,265=76.7; P<0.001), and sex (F1,104=17.7; P<0.001) on PPT ln kPa (Fig. 2b and c).Fig 2 Effect of preterm birth, sex, and time on pressure pain threshold during conditioned pain modulation (CPM). (a) Change in mechanical pressure pain threshold over the right fibula head (PPT, raw data, kPa) in term control (TC) and extremely preterm (EP)-born young adults during (15 s) and after (50 and 90 s) a conditioning stimulus (0–30 s). Females unable to tolerate immersion until the parallel PPT measurement (<15 s) had no significant change in PPT with time. Additional groups with <15 s immersion were too small for analysis (five EP males, four TC females, and two TC males). Data points=mean [95% confidence interval (CI)]. (b) and (c) Change in log-normalised PPT (ln kPa) with time in (b) males and (c) females. For EP and TC participants tolerating at least 20 s conditioning, PPT is significantly increased above baseline at 15 and 50 s. In EP females with <15 s immersion, a minor increase in threshold is seen only at 15 s. Data points=mean (95% CI). ###P<0.001, #P<0.05, ***P<0.001, **P<0.01, *P<0.05, §P<0.05; two-way repeated measures analysis of variance with Bonferroni post hoc comparisons of within-group change compared with baseline.
15 and 50 s. In EP females with <15 s immersion, a minor increase in threshold is seen only at 15 s. Data points=mean (95% CI). ###P<0.001, #P<0.05, ***P<0.001, **P<0.01, *P<0.05, §P<0.05; two-way repeated measures analysis of variance with Bonferroni post hoc comparisons of within-group change compared with baseline. Fig 2 Increased sensitivity in a subgroup of extremely premature females To identify the factors associated with reduced conditioning tolerance, we compared measures in participants who did, or did not, tolerate 20 s cold immersion (Table 1). EP young adults with reduced conditioning tolerance (<20 s) also had higher pain scores on hand removal and lower baseline PPT, and a higher proportion were females (Table 1). Twenty-eight EP (and no TC) participants had required neonatal surgery and were distributed across both immersion durations (see Table 1 for sex distribution and type of surgery).
ioning tolerance (<20 s) also had higher pain scores on hand removal and lower baseline PPT, and a higher proportion were females (Table 1). Twenty-eight EP (and no TC) participants had required neonatal surgery and were distributed across both immersion durations (see Table 1 for sex distribution and type of surgery). As PPT may not be accurate if the participants are removing and reporting VAS in one hand, whilst they are pressing a PPT response button with the contralateral hand, data related to immersion times of 15–19 s (five EP females, one EP male, and one TC male) were not further analysed. The remaining 31 participants (20 EP females, 5 EP males, 4 TC females, and 2 TC males) tolerated less than 15 s immersion and removed the hand before the first PPT. Only EP females comprised a sufficient sample for further analysis. Brief immersion failed to produce modulatory effects in this group, as there was no change in raw PPT with time (Fig. 2a), and only a minor increase in normalised PPT (P<0.05) at 15 s that was not maintained at 50 s (Fig. 2c).
and before the first PPT. Only EP females comprised a sufficient sample for further analysis. Brief immersion failed to produce modulatory effects in this group, as there was no change in raw PPT with time (Fig. 2a), and only a minor increase in normalised PPT (P<0.05) at 15 s that was not maintained at 50 s (Fig. 2c). Degree of conditioned pain modulation To compare the degree of CPM effect, the percentage change from baseline was calculated for individual participants tolerating at least 20 s immersion. The degree of CPM effect varied with time, with a similar maximal change at 15 s in both EP and TC groups [53% (95% CI: 41–65) vs 57% (95% CI: 42–71)] (Fig. 3). Whilst PPT was higher in EP males than EP females at all time points and the absolute change during conditioning was slightly greater [133 kPa (inter-quartile range: 94–225) vs 89 (51–196); P=0.048], once expressed as percentage change from the higher baseline, there were no sex differences in CPM effect. Baseline PPT (ln kPa) was negatively correlated with CPM effect (% change at 15 s) (Table 2), and when separated by group, in EP [r=–0.45 (95% CI: –0.6 to –0.18; P<0.01] (Supplementary Table S1B), but not TC (r=–0.18; P=0.3) participants. In regression analysis with participants tolerating 20–30 s immersion (n=108), the duration of conditioning did not influence CPM effect (therefore, these durations were grouped together), but PPT had a significant effect (Table 3). Reduced cold tolerance is most marked in EP females, particularly those with prior neonatal surgery,23 and shorter immersion time correlates with lower PPT in EP, but not TC participants (Supplementary Table S2). Including all participants in the regression model (n=148) highlighted the impact of shorter immersion time, as this variable now had a significant impact on calculated CPM effect, but there were no marked changes related to other variables (Supplementary Table S3).Fig 3 Degree of conditioned-pain-modulation effect after at least 20 s conditioning stimulus in extremely preterm (EP) and term-born control (TC) males and females. (a) The percentage change in pressure pain threshold (PPT) during the conditioning stimulus is not significantly different across groups based on EP status or sex. Individual data points, bars=mean [95% confidence interval (CI)]. (b) Raised PPT during the conditioning stimulus (15 s) is maintained at 50 and 90 s.
ales. (a) The percentage change in pressure pain threshold (PPT) during the conditioning stimulus is not significantly different across groups based on EP status or sex. Individual data points, bars=mean [95% confidence interval (CI)]. (b) Raised PPT during the conditioning stimulus (15 s) is maintained at 50 and 90 s. Data points=mean (95% CI); ***P<0.001; **P<0.01 all groups increase vs baseline; #P<0.05 TC males, EP females, and EP males vs baseline; two-way repeated measures analysis of variance with Bonferroni post hoc comparisons. Fig 3Table 3 Linear model of conditioned-pain-modulation effect (% change in PPT at 15 s) for participants with conditioning tolerance ≥20 s (n=108). DSM-Anxiety, anxiety t-score Achenbach Youth Self-Report scale; HUI-3, Health Utilities Index Mark 3; PCS, pain catastrophising scale; PPT, pressure pain threshold; SE, standard error Table 3Variables Step 1 (n=108) Step 2 (n=108) Step 3 (n=96) B SE B β P B SE B β P B SE B β P Baseline PPT –27 7.1 –0.35 <0.001 –30 7.9 –0.39 <0.001 –35 8.6 –0.43 <0.001 Immersion time –1.9 1.5 –0.12 0.19 –1.7 1.5 –0.10 0.27 –0.16 1.6 –0.01 0.92 Extremely premature status –1.8 9.0 –0.02 0.84 2.6 9.7 0.02 0.79 Sex –8.9 9.3 –0.09 0.34 –14 10 –0.14 0.19 Pain (HUI-3 ranking) –5.4 7.5 0.08 0.47 Regular analgesics 21 24 0.09 0.38 Catastrophising (PCS) –0.58 0.56 –0.13 0.30 DSM-Anxiety (Ach) –0.57 0.59 –0.12 0.33 Regular psychotropics 65 23 0.29 0.01 R2 0.15 0.16 0.29 F for R2 F2,105=9.3; P=0.001 F4,103=4.8; P=0.001 F9,87=3.9; P=0.001
–8.9 9.3 –0.09 0.34 –14 10 –0.14 0.19 Pain (HUI-3 ranking) –5.4 7.5 0.08 0.47 Regular analgesics 21 24 0.09 0.38 Catastrophising (PCS) –0.58 0.56 –0.13 0.30 DSM-Anxiety (Ach) –0.57 0.59 –0.12 0.33 Regular psychotropics 65 23 0.29 0.01 R2 0.15 0.16 0.29 F for R2 F2,105=9.3; P=0.001 F4,103=4.8; P=0.001 F9,87=3.9; P=0.001 Current pain, medication, and psychological outcomes are related, but do not influence CPM effect Higher self-reported pain (average VAS in last week or HUI-3 ranking), regular analgesia use, higher anxiety and catastrophising scores, and regular psychotropic medications were inter-related, but did not correlate with CPM variables (baseline PPT, immersion duration, immersion pain, or CPM effect) (Table 2). Medication use is listed in Table 1. No participants had taken analgesia on the day of testing, and 27/97 EP and 11/48 TC reported use of occasional or regular analgesia, most often paracetamol or an NSAID (Table 1). Pain-related conditions had required specialist management in four EP females (steroid injection for knee pain, two pain clinic reviews and gabapentin for persistent post-surgical pain or fibromyalgia, and rheumatology review and physiotherapy for back pain), two EP males (neurologist and migraine prophylaxis, and gastroenterologist and azathioprine for Crohn's disease and associated abdominal pain), and one TC female (rheumatologist and physiotherapy for hypermobility).
persistent post-surgical pain or fibromyalgia, and rheumatology review and physiotherapy for back pain), two EP males (neurologist and migraine prophylaxis, and gastroenterologist and azathioprine for Crohn's disease and associated abdominal pain), and one TC female (rheumatologist and physiotherapy for hypermobility). Self-reported medication use (general health questionnaire) was available for 136 participants. Ten participants (nine EP and one TC) reported medications with psychoactive properties (antidepressants or medications used for attention deficit hyperactivity disorder; Table 1). Participants taking antidepressant medications had higher anxiety scores (P<0.01) (Supplementary Table S1). Seven of these participants tolerated at least 20 s immersion and tended to have a higher CPM effect, but the variability is wide (one, no change; six, inhibitory response; 99±106% at 15 s). Inclusion of psychotropic medications had a significant effect in the regression model (Table 3; Supplementary Table S3), but numbers are small, and significance was lost when an outlier with a very low baseline PPT (48 kPa, EP female) and high percentage change during conditioning (PPT 200 kPa; 332% increase; see Fig. 3a) was excluded.
otropic medications had a significant effect in the regression model (Table 3; Supplementary Table S3), but numbers are small, and significance was lost when an outlier with a very low baseline PPT (48 kPa, EP female) and high percentage change during conditioning (PPT 200 kPa; 332% increase; see Fig. 3a) was excluded. Although FSIQ was lower in EP participants (Table 1), FSIQ did not correlate with conditioning tolerance, baseline PPT, or CPM effect (Supplementary Table S2). Height and weight were lower in EP than TC participants, but BMI did not differ (Table 1) and did not influence CPM parameters. Prior neonatal surgery influenced baseline PPT in EP males and conditioning tolerance in females,23 but not CPM effect. Neonatal variables (birth weight, CRIB score on admission to intensive care, and duration of hospital stay) also did not correlate with CPM parameters (Supplementary Table S2A).
luence CPM parameters. Prior neonatal surgery influenced baseline PPT in EP males and conditioning tolerance in females,23 but not CPM effect. Neonatal variables (birth weight, CRIB score on admission to intensive care, and duration of hospital stay) also did not correlate with CPM parameters (Supplementary Table S2A). Discussion Inhibitory CPM was demonstrated in the majority of young adults, but differences in conditioning stimulus tolerance influenced the ability to quantify CPM. In participants tolerating conditioning, there were significant main effects of EP status, sex, and time on PPT during and after hand immersion, with inhibitory modulation evoked in 64/98 EP (3, no change) and 38/48 TC (3, facilitation). Identification and quantification of CPM in EP, but not TC, participants were influenced by sex-dependent differences in sensitivity to both the test (reduced sensitivity in EP males) and conditioning stimulus (increased sensitivity in EP females). One-third of EP females had low baseline PPT and reduced cold-pressor tolerance, and the brief conditioning did not alter the subsequent PPT. Current pain, anxiety, and pain catastrophising scores did not correlate with CPM magnitude.
itivity in EP males) and conditioning stimulus (increased sensitivity in EP females). One-third of EP females had low baseline PPT and reduced cold-pressor tolerance, and the brief conditioning did not alter the subsequent PPT. Current pain, anxiety, and pain catastrophising scores did not correlate with CPM magnitude. Baseline sensitivity and conditioning tolerance were significantly influenced by EP status and sex. Previous comparison to TCs reported lower PPT in very preterm (VP; mean: 31 weeks GA) adolescents, predominantly as a result of increased sensitivity in females and minimal difference in males.28 Here, PPT on the head of the fibula was lower in EP females than EP males, and correlated with sex-dependent differences measured on the middle digit of the hand in this cohort.23 Reduced cold-pressor tolerance has also been reported in young adults born EP (mean: 26.8 weeks GA) with 5 of 31 withdrawing the hand before 30 s, and females were more sensitive.29 In VP (mean: 31 weeks GA) young adults, female sex and neonatal necrotising enterocolitis reduced the likelihood of tolerating cold immersion at 19 yr,37 but details of surgery for this or other conditions are not reported. We similarly found reduced sensitivity in EP females, particularly those that had undergone neonatal surgery,23 but the proportion tolerating less than 20 s immersion was larger than anticipated, and brief immersion hampered our ability to reliably quantify CPM effect.
ery for this or other conditions are not reported. We similarly found reduced sensitivity in EP females, particularly those that had undergone neonatal surgery,23 but the proportion tolerating less than 20 s immersion was larger than anticipated, and brief immersion hampered our ability to reliably quantify CPM effect. CPM can clearly identify differences in the proportion of participants with inhibition, no change, or facilitation.38, 39 For participants tolerating immersion, significant inhibitory CPM was identified in the majority of both EP and TC participants, and the smaller numbers with no change or facilitation did not differ between groups. Whilst these descriptive differences are generalisable across studies, different methods and time points have been used to compare the magnitude of CPM across groups.4, 16, 36, 40, 41 Despite clear group increases in PPT after conditioning, there was a wide within-group variability in calculated percent CPM, as also seen in some previous evaluations in healthy adolescents36 and adults.42 This limited our ability to identify group differences in CPM magnitude. Whilst there are no generally accepted ‘normative’ data for the magnitude of CPM effect, our data (mean increase in PPT of 50–65% at 15 s, and 31–36% at 50 s in TC) are consistent with studies using similar methodology in adults9, 43 and adolescents,7, 36 and persistence beyond the conditioning stimulus suggests CPM is not solely related to non-specific distraction.3
’ data for the magnitude of CPM effect, our data (mean increase in PPT of 50–65% at 15 s, and 31–36% at 50 s in TC) are consistent with studies using similar methodology in adults9, 43 and adolescents,7, 36 and persistence beyond the conditioning stimulus suggests CPM is not solely related to non-specific distraction.3 Preterm birth has been associated with persistent alterations in somatosensory function and pain response,12, 23 but CPM has only previously been assessed in a small group of VP children. CPM efficacy varies with age,1 and weak inhibitory effects in childhood become more robust throughout adolescence.7 Mechanisms underlying this delayed maturation of descending inhibition have been identified in rodents,12 and this normal developmental trajectory can be altered by neonatal tissue injury (hind-paw carrageenan inflammation44 or incision45). At 7–11 yr, inhibitory CPM was enhanced (greater decrease in heat pain intensity after cold conditioning) in a ‘low-pain’ group of six children born VP (28–32 weeks gestation), but absent in a ‘high-pain’ group of seven with longer NICU admission and increased procedural pain exposure.16 The EPICure cohort participants were born at an earlier GA, had longer hospital stay, and 28% required neonatal surgery, suggesting they would constitute a high-pain group.16 However, in these EP young adults, failure to tolerate the conditioning stimulus and female sex, rather than neonatal variables per se, were the predominant factors associated with ‘absence’ of CPM.
er GA, had longer hospital stay, and 28% required neonatal surgery, suggesting they would constitute a high-pain group.16 However, in these EP young adults, failure to tolerate the conditioning stimulus and female sex, rather than neonatal variables per se, were the predominant factors associated with ‘absence’ of CPM. Increased inhibitory CPM has been reported in males.9, 18, 42 Whilst the relative change in PPT was higher in EP males, once expressed as percentage change from the higher PPT, this difference was lost. Females tend to have lower PPT6, 46 and reduced cold-pressor tolerance,47, 48 and CPM identified a large subset of EP females with increased sensitivity to both pressure and cold immersion. This ‘lack’ of CPM response is likely to reflect failure of brief immersion to engage descending modulatory effects.49 There is currently no standardised reporting for ‘non-responders’ or subjects with no change in CPM,1, 50 and data from subjects with reduced conditioning tolerance are often excluded.6, 36 As this included a large number and proportion of our EP females, data from this group are presented separately. The degree to which the duration and intensity of the conditioning stimulus influence the CPM effect is debated.42 A CPM paradigm that alters the conditioning intensity based on individual sensitivity40 would have advantages for groups, such as EP young adults, who have marked variability in conditioning tolerance.
ly. The degree to which the duration and intensity of the conditioning stimulus influence the CPM effect is debated.42 A CPM paradigm that alters the conditioning intensity based on individual sensitivity40 would have advantages for groups, such as EP young adults, who have marked variability in conditioning tolerance. Psychological factors influence descending modulatory pathways and interact with similar neurotransmitter systems as CPM.2, 39 A meta-analysis found no overall correlation between CPM and psychological variables in healthy or pain populations, but a secondary analysis showed modality-specific correlations between increased CPM effect and higher anxiety using pressure-based testing and higher pain catastrophising using an electrical test stimulus.2 Children born VP had higher pain catastrophising scores,14 and altered patterns of functional MRI activation by a prolonged thermal heat stimulus at 11–16 yr included differences in brainstem modulatory regions.51 Consistent with existing literature,13, 52 our EP young adults self-reported more internalising and anxiety, and pain catastrophising was higher in females, but whilst these measures were associated with increased self-reported pain, they did not correlate with cold tolerance or CPM effect. A large Norwegian population-based registry found adults born VP or EP were more likely to be taking psychotropic medications (antidepressants, anxiolytics, and hypnotics), with overall greater use by females, and EP males more likely to be taking medication for attention deficit hyperactivity disorder.53 Here, the EP participants taking psychotropic medications had higher self-reported anxiety and pain, and inclusion of this variable influenced the regression model, but the small numbers and wide variability limited the reliability of relationships with the CPM effect.
on for attention deficit hyperactivity disorder.53 Here, the EP participants taking psychotropic medications had higher self-reported anxiety and pain, and inclusion of this variable influenced the regression model, but the small numbers and wide variability limited the reliability of relationships with the CPM effect. Reduced inhibitory CPM3, 11 or a shift to facilitation has been reported in adults38, 40 and youth41 with chronic pain. Here, there were no clear associations between current pain experience and the degree or directionality of CPM. Whilst restoration of inhibition after treatment suggests that impaired CPM is a reversible effect of chronic pain,1 reduced inhibitory CPM in adults54 and facilitation rather than inhibition in adolescents8 predicted persistent musculoskeletal pain. Similarly, impaired preoperative CPM has predicted persistent post-surgical pain after different types of surgery.4, 55, 56, 57 EP females with increased sensitivity to pressure and noxious cold, in whom robust inhibition could not be quantified, may be at increased risk of persistent pain after surgery or injury in the future. An ongoing assessment in large cohorts is required to further quantify the risk and evaluate the potential preventive interventions.
h increased sensitivity to pressure and noxious cold, in whom robust inhibition could not be quantified, may be at increased risk of persistent pain after surgery or injury in the future. An ongoing assessment in large cohorts is required to further quantify the risk and evaluate the potential preventive interventions. The limitations of this study include potential selection bias, as not all EPICure cohort participants were assessed at 19–20 yr. EPICure@19 participants did not differ in birth weight, GA, or sex from those lost to follow-up, and those attending for QST did not differ from the remaining participants undergoing other assessments at 19 yr.23 Long-term follow-up tends to favour NICU survivors with a relatively favourable outcome,58 and EPICure@19 participants had higher socio-economic status and higher mean IQ scores at earlier assessments than non-participants,20 suggesting that the effects may be underestimated. Ethnicity was not assessed, as the majority of subjects were Caucasian, and fewer EP males were tested, but with a matched proportion of controls. In females, the results were not stratified by menstrual phase or use of oral/implanted contraceptive hormones, although some reports suggest these factors have minimal effect on CPM magnitude,9 pressure threshold, or cold-pressor sensitivity.59 The analysis of CPM effect tends to focus, as here, on bulbospinal control of descending inhibition/facilitation; however, additional mechanisms that may also modulate pain response and be influenced by preterm birth include alterations in hypothalamic–pituitary–adrenal axis function60 and autonomic nervous system activation.38, 61
of CPM effect tends to focus, as here, on bulbospinal control of descending inhibition/facilitation; however, additional mechanisms that may also modulate pain response and be influenced by preterm birth include alterations in hypothalamic–pituitary–adrenal axis function60 and autonomic nervous system activation.38, 61 In summary, descending inhibitory modulation was identified in the majority of participants, with increases in PPT during conditioning maintained beyond the stimulus. For those tolerating cold immersion, the degree and directionality of CPM did not differ between EP and term-born young adults. However, the ability to quantify and compare the CPM effect was influenced by sensitivity to the test and the conditioning stimulus. Improvements in neonatal intensive care are now resulting in increased numbers of EP born children reaching adulthood, and identifying risk factors for future illness is a major focus of longitudinal outcome studies.62, 63 The CPM protocol identified a high proportion of EP females with a persistent increased sensitivity to pressure and noxious cold that may influence future pain experience or risk of persistent pain.23 As sex/gender and preterm birth influence conditioning and test stimulus sensitivity, these factors should be considered when choosing the methodology and analysis of CPM to predict or assess persistent post-surgical pain or chronic pain. Authors' contributions Study design/planning: S.M.W., H.O., J.B., N.M. Data analysis: S.M.W., H.O., J.B. Drafting and writing paper: S.M.W.
In summary, descending inhibitory modulation was identified in the majority of participants, with increases in PPT during conditioning maintained beyond the stimulus. For those tolerating cold immersion, the degree and directionality of CPM did not differ between EP and term-born young adults. However, the ability to quantify and compare the CPM effect was influenced by sensitivity to the test and the conditioning stimulus. Improvements in neonatal intensive care are now resulting in increased numbers of EP born children reaching adulthood, and identifying risk factors for future illness is a major focus of longitudinal outcome studies.62, 63 The CPM protocol identified a high proportion of EP females with a persistent increased sensitivity to pressure and noxious cold that may influence future pain experience or risk of persistent pain.23 As sex/gender and preterm birth influence conditioning and test stimulus sensitivity, these factors should be considered when choosing the methodology and analysis of CPM to predict or assess persistent post-surgical pain or chronic pain. Authors' contributions Study design/planning: S.M.W., H.O., J.B., N.M. Data analysis: S.M.W., H.O., J.B. Drafting and writing paper: S.M.W. Revision and approval of the final manuscript: S.M.W., H.O., J.B., N.M. Overall planning and conduct of evaluations at 19: EPICure@19 study group. Declaration of interest The authors declare that they have no conflicts of interest.
Authors' contributions Study design/planning: S.M.W., H.O., J.B., N.M. Data analysis: S.M.W., H.O., J.B. Drafting and writing paper: S.M.W. Revision and approval of the final manuscript: S.M.W., H.O., J.B., N.M. Overall planning and conduct of evaluations at 19: EPICure@19 study group. Declaration of interest The authors declare that they have no conflicts of interest. Funding Medical Research Council, UK (G0401525) to N.M.; Department of Health, National Institute for Health Research, Biomedical Research Centre funding scheme at the University College London Hospitals/UCL to N.M.; Great Ormond Street Hospital Children's Charity (Projects V2818 and W1071H) to S.M.W. Appendix A Supplementary data The following are the supplementary data related to this article:Multimedia component 1 Multimedia component 1 Multimedia component 2 Multimedia component 2
Funding Medical Research Council, UK (G0401525) to N.M.; Department of Health, National Institute for Health Research, Biomedical Research Centre funding scheme at the University College London Hospitals/UCL to N.M.; Great Ormond Street Hospital Children's Charity (Projects V2818 and W1071H) to S.M.W. Appendix A Supplementary data The following are the supplementary data related to this article:Multimedia component 1 Multimedia component 1 Multimedia component 2 Multimedia component 2 Acknowledgements The authors gratefully acknowledge the contribution of the participants and their families to the current and previous follow-up studies, and all current and past members of the EPICure research team. This study is presented on behalf of the EPICure@19 Study Group Investigators that include Neil Marlow, EGA UCL Institute for Women's Health (principal investigator); John Cockcroft, Cardiff University; Xavier Golay, UCL Institute of Neurology; John Hurst, University College London; Samantha Johnson, University of Leicester; Sebastien Ourselin, University College London; Suellen Walker, UCL Great Ormond Street Institute of Child Health; Dieter Wolke, University of Warwick. The study was supported by the National Institute for Health Research University College London Hospitals Clinical Research Facility. This article is accompanied by an editorial: Back To The Future: lifelong changes in pain processing in ‘ageing of prematurity’ by McCarthy & Colvin, Br J Anesth 2018:121:529–531, doi: 10.1016/j.bja.2018.06.017
Acknowledgements The authors gratefully acknowledge the contribution of the participants and their families to the current and previous follow-up studies, and all current and past members of the EPICure research team. This study is presented on behalf of the EPICure@19 Study Group Investigators that include Neil Marlow, EGA UCL Institute for Women's Health (principal investigator); John Cockcroft, Cardiff University; Xavier Golay, UCL Institute of Neurology; John Hurst, University College London; Samantha Johnson, University of Leicester; Sebastien Ourselin, University College London; Suellen Walker, UCL Great Ormond Street Institute of Child Health; Dieter Wolke, University of Warwick. The study was supported by the National Institute for Health Research University College London Hospitals Clinical Research Facility. This article is accompanied by an editorial: Back To The Future: lifelong changes in pain processing in ‘ageing of prematurity’ by McCarthy & Colvin, Br J Anesth 2018:121:529–531, doi: 10.1016/j.bja.2018.06.017 Appendix A Supplementary data related to this article can be found at https://doi.org/10.1016/j.bja.2018.05.066.
Editor's key points • Long-term impact of early life experience on pain responses is poorly understood. • Participants (who had been born preterm) were recruited from an established database, with term controls. Using somatosensory testing, brain imaging, and validated questionnaires, pain and associated factors were comprehensively assessed. • Preterm participants showed persistent changes in somatosensory processing and brain structure, with sex differences. Preterm birth is an acknowledged health care priority because of its increasing prevalence,1 acute morbidity, and persistent impact on multiple health outcomes.2 Exposure to repeated painful procedures and surgical interventions during neonatal intensive care, particularly after extreme preterm birth (<28 weeks gestation), is occurring at a time when the developing nervous system is vulnerable to altered levels of activity.3 Improved recognition of pain is a research priority for neonates born preterm4 to guide management and minimise acute distress, but the longer-term impact of increased procedural pain exposure and neonatal surgery on brain structure and connectivity5, 6, 7 and adverse neurodevelopmental outcome8, 9 is increasingly recognised. However, the degree to which biological effects associated with preterm birth persist into adulthood or are modulated by subsequent experience and psychosocial factors can vary.8
posure and neonatal surgery on brain structure and connectivity5, 6, 7 and adverse neurodevelopmental outcome8, 9 is increasingly recognised. However, the degree to which biological effects associated with preterm birth persist into adulthood or are modulated by subsequent experience and psychosocial factors can vary.8 Understanding effects of preterm birth and neonatal surgery on both somatosensory and affective components of pain response is necessary to identify factors that influence current pain experience, influence future risk, or both.3, 10 Persistent alterations in somatosensory function have been demonstrated in preterm-born children,11, 12, 13 but may be influenced by the subsequent age- and sex-dependent changes in sensory thresholds throughout adolescence.14, 15 Psychological factors that influence pain experience, such as increased anxiety persist into early adulthood after extreme preterm birth,2, 16 and higher pain catastrophising was noted in preterm children.12 Reported associations between preterm birth and chronic pain prevalence vary, but the different methodologies and populations in epidemiological and cohort studies, and limited details about the type, severity, and impact of pain, hamper comparison across studies.17, 18, 19, 20
ophising was noted in preterm children.12 Reported associations between preterm birth and chronic pain prevalence vary, but the different methodologies and populations in epidemiological and cohort studies, and limited details about the type, severity, and impact of pain, hamper comparison across studies.17, 18, 19, 20 This observational cohort study compared somatosensory function and pain experience in extremely preterm-born (EP; <26 weeks gestation) and healthy term-born young adults. We hypothesised that group differences in thermal sensitivity and the added impact of neonatal surgery previously identified at 11 yr in this cohort13 would persist at 19 yr. In addition, we explored associations with neuroanatomical factors, current pain experience, cognitive function, anxiety, and pain catastrophising. As male sex is an independent risk factor for adverse neurodevelopmental outcome after preterm birth,21, 22 and sex/gender influences experimental pain sensitivity and chronic pain prevalence in adulthood,23, 24 outcomes were compared in males and females.
erience, cognitive function, anxiety, and pain catastrophising. As male sex is an independent risk factor for adverse neurodevelopmental outcome after preterm birth,21, 22 and sex/gender influences experimental pain sensitivity and chronic pain prevalence in adulthood,23, 24 outcomes were compared in males and females. Methods Participants Participants were recruited from the UK EPICure population-based cohort of infants born extremely preterm in the UK and Ireland from March to December 1995. Although extreme preterm birth is defined as <28 weeks gestation, the EPICure cohort restricted recruitment to earlier high-risk births at <26 weeks gestation. Of 811 infants of the correct gestational age admitted to neonatal intensive care, 497 died in hospital and 314 were discharged home.25 Participation in longitudinal evaluation at 30 months,25 6 yr,26 11 yr,27 and at 19 yr has been previously described.22 The current study was approved by the National Research Ethics Committee Hampshire ‘A’ (Reference: 13/SC/0514), described on the cohort website (EPICure@19; www.epicure.ac.uk), and potential participants received written information. Non-participants had previously asked not to be contacted, declined participation, or were uncontactable. EP participants in EPICure@19 did not differ in birth weight, gestational age, or sex from those lost to follow-up, but had higher mean full-scale intelligence quotient (FSIQ) scores at earlier assessments and higher socio-economic backgrounds than non-participants.22 After giving written consent, participants underwent a 2 day evaluation at the University College London Hospital, Clinical Research Facility (London, UK) between February 2014 and October 2015. Pain and somatosensory function were evaluated in 102 EP and 48 term-born control (TC) young adults (Fig. 1) in a dedicated sensory testing facility at University College London Great Ormond Street Institute of Child Health (London, UK). Additional data related to neonatal variables, participant characteristics, and questionnaires at 18–20 yr were extracted from the main EPICure database. Data related to conditioned pain modulation are reported in the companion manuscript (Walker and colleagues,28 Br J Anaesth in press). Reporting is in accordance with the STROBE (Strengthening the Reporting of Observational studies in Epidemiology) Checklist for cohort studies.Fig 1 EPICure recruitment and assessment flowchart.
ted to conditioned pain modulation are reported in the companion manuscript (Walker and colleagues,28 Br J Anaesth in press). Reporting is in accordance with the STROBE (Strengthening the Reporting of Observational studies in Epidemiology) Checklist for cohort studies.Fig 1 EPICure recruitment and assessment flowchart. E@19, EPICure at 19 yr study; F, female; M, male; QST, quantitative sensory testing. Fig 1 Assessments A standardised clinical pain history included: site, intensity (0–10 verbal rating scale, VRS), frequency, and duration of recurrent pain; impact on function and activity; interference with usual activity due to recurrent pain (0–10 VRS); and analgesic use. Overall pain report was graded by a pain clinician (S.M.W.; 0=no regular pain, 1=infrequent pain, does not limit activities, 2=more frequent pain with some impact on function, 3=more severe pain that limits activity). Participants used visual analogue scales (0–10 cm) to report current pain intensity (right now; average in the past week; worst pain in the past week), interference with usual activities because of pain, and anticipatory anxiety before testing.29
n with some impact on function, 3=more severe pain that limits activity). Participants used visual analogue scales (0–10 cm) to report current pain intensity (right now; average in the past week; worst pain in the past week), interference with usual activities because of pain, and anticipatory anxiety before testing.29 Quantitative sensory testing Somatosensory function was assessed with a standardised protocol30, 31 adapted to match previous preterm-born cohort studies.11, 13 Evaluation was performed by a single investigator (S.M.W.) in the same temperature-controlled room with standardised verbal instructions. Before data acquisition, tests were demonstrated and participants advised they could decline or cease testing at any point. Testing was performed on the thenar eminence of the self-reported non-dominant hand to evaluate generalised thresholds and then on the chest wall. Localised testing adjacent to neonatal scars was restricted to thoracic dermatomes (high proportion of EP but no TC participants had chest scars when previously evaluated13). Participants without scars had testing on the lateral chest wall within the second to sixth thoracic dermatomes. Thermal thresholds were not obtained in two of 38 EP females because of equipment malfunction. The need to ask about prior surgery, and the site and nature of neonatal scars, precluded the investigator being blinded to group.
ut scars had testing on the lateral chest wall within the second to sixth thoracic dermatomes. Thermal thresholds were not obtained in two of 38 EP females because of equipment malfunction. The need to ask about prior surgery, and the site and nature of neonatal scars, precluded the investigator being blinded to group. Modalities included: i) cool (CDT) and warm detection (WDT), cold (CPT) and heat (HPT) pain thresholds using a handheld 18×18 mm contact thermode (baseline 32°C, 1°C/s, limits 10°C and 50°C; Senselab MSA Thermal Stimulator; Somedic, Sosdala, Sweden) to match testing at 11 yr;13 ii) mechanical detection threshold (MDT) with von Frey hairs (geometric mean of 10 appearance and disappearance thresholds); iii) mechanical pricking pain threshold (MPT) with ascending PinPrick Stimulators (8–512 mN) until discomfort/pain rated 0–10 (VRS1) then after 1 s−1 train of 10 repeated stimuli (VRS10) to calculate wind-up ratio (WUR=VRS10–VRS1);11 and iv) pressure pain threshold (PPT) mean of three values on middle phalanx of middle finger with hand-held 1 cm2 algometer and optical feed-back (ramp 40 kPa s−1, maximum 1000 kPa; SENSEBox; Somedic, Sosdala, Sweden). As static thermal thresholds demonstrated reduced sensitivity in children after preterm birth, but a prolonged thermal stimulus unmasked increased sensitivity,11 cold pressor testing was also evaluated (see also conditioned pain modulation protocol; Walker et al.28 Br J Anaesth, in press). The hand was immersed to the wrist with the fingers spread into a 5°C circulating water bath (TE-10D Thermoregulator, B-8 Bath, RU-200 Dip Cooler; Techne, Burlington, VT, USA) and immersion duration (maximum 30 s) recorded.
aluated (see also conditioned pain modulation protocol; Walker et al.28 Br J Anaesth, in press). The hand was immersed to the wrist with the fingers spread into a 5°C circulating water bath (TE-10D Thermoregulator, B-8 Bath, RU-200 Dip Cooler; Techne, Burlington, VT, USA) and immersion duration (maximum 30 s) recorded. Questionnaires Self-report questionnaires (investigators H.O. and J.B.) included: i) Pain Catastrophizing Scale (PCS; total 0–52, subscales rumination, magnification, helplessness)32; ii) Diagnostic and Statistical Manual (DSM) anxiety t-score (range 50–100; ≥70 clinically significant) and internalising problems t-score (range 50–100; ≥64 clinically significant) extracted from Achenbach Adult Self-Report Questionnaire33; and iii) FSIQ using the Wechsler Abbreviated Scale of Intelligence Second Edition (WASI-II; mean: 100, sd: 15).34 MRI We acquired 3D T1-weighted MPRAGE (TR/TE=6.93/3.14 ms) volumes at 1 mm isotropic resolution on a Philips 3T Achieva (Philips, Amsterdam, Netherlands) MRI scanner and carried out a multi-class tissue segmentation of the white matter volume using combined multi-atlas and Gaussian mixture model segmentation routines.35 This method produces a state-of-the-art segmentation and region labelling by voxel-wise voting between several propagated atlases guided by the local image similarity. This algorithm automatically estimates thalamus and amygdala volumes. See Supplementary material for pathway specific tissue properties (fractional anisotropy and average intra-axonal volume fractions).
on and region labelling by voxel-wise voting between several propagated atlases guided by the local image similarity. This algorithm automatically estimates thalamus and amygdala volumes. See Supplementary material for pathway specific tissue properties (fractional anisotropy and average intra-axonal volume fractions). Statistical analysis As this descriptive cohort study aimed to recruit the maximum available subjects, no a priori power calculation was performed. Statistically significant group differences in thermal thresholds were found when 43 EP and 44 TC participants from the current cohort were tested at age 11 yr.13
on and region labelling by voxel-wise voting between several propagated atlases guided by the local image similarity. This algorithm automatically estimates thalamus and amygdala volumes. See Supplementary material for pathway specific tissue properties (fractional anisotropy and average intra-axonal volume fractions). Statistical analysis As this descriptive cohort study aimed to recruit the maximum available subjects, no a priori power calculation was performed. Statistically significant group differences in thermal thresholds were found when 43 EP and 44 TC participants from the current cohort were tested at age 11 yr.13 Statistical analyses included: group-wise comparisons with Mann–Whitney U-test or two-tailed Student's t-test; two-way ANOVA with group (TC, EP, EP+surgery) and sex as variables for normally-distributed or log-transformed mechanical data36; two-sided χ2 test for categorical data; two-tailed Spearman's rho (rs) for bivariate correlations; and log rank Mantel–Cox for survival curves. Truncated regression models evaluated generalised thermal sensitivity [GTS: time to HPT, 32–50°C at 1°C s−1)+(time to CPT, 32 to 10°C)+(cold pressor duration)] with higher values reflecting increased thermal tolerance (i.e. decreased sensitivity; maximum=18+22+30=70). For quantitative sensory testing (QST) profiles, sex-matched Z-transformed scores were calculated z=[(XEP participant–Meancontrols)/sdcontrols] and adjusted so >0 indicates increased sensitivity and <0 decreased sensitivity.30 Analyses was performed with SPSS Version 23 (IBM, Portsmouth, UK) and Prism Version 7 (GraphPad, San Diego, CA, USA). P values are reported with Bonferroni adjustment for multiple comparisons.
=[(XEP participant–Meancontrols)/sdcontrols] and adjusted so >0 indicates increased sensitivity and <0 decreased sensitivity.30 Analyses was performed with SPSS Version 23 (IBM, Portsmouth, UK) and Prism Version 7 (GraphPad, San Diego, CA, USA). P values are reported with Bonferroni adjustment for multiple comparisons. Results Participant characteristics One hundred and two EP and 48 age- and sex-matched TC participants underwent pain and somatosensory assessment (Fig. 1). EP participants had lower height and weight, but the same BMI as TC (Table 1). FSIQ scores were lower in the EP group, but did not differ between QST and remaining EPICure@19 participants.22 Thirty EP participants had required neonatal surgery (12 closure patent ductus arteriosus, seven laparotomy, 10 inguinal hernia repairs, and one ventricular drain). The surgery subgroup had longer initial hospitalisation, but did not differ in birth weight, gestational age or risk index score on neonatal ICU (NICU) admission (Supplementary Table 1). QST results were excluded because of variability in three EP males (two had difficulty with numerical scales; one reported tiredness and difficulty concentrating). Chest wall testing was declined in three EP subjects (time; scar allodynia; tired), and one EP female with Raynaud's symptoms declined cold evaluation (Fig. 1). No participant reported distress during testing.Table 1 Demographic data: group and sex differences. *Sample size for full group; for outcomes where data was not available for all participants, the number of participants (n=) is included below the result. †Obtained using Mann–Whitney U-test; ‡P values by two-sided χ2; §Female neonatal surgery: closure patent ductus arteriosus (PDA) n=8; laparotomy n=4; inguinal hernia repair, IH, n=1; ¶Male neonatal surgery: IH n=9, laparotomy n=3; PDA n=2, PDA+IH, n=2; CSF drain n=1. mod/sev, moderate or severe; IQR, inter-quartile range; MSK, musculoskeletal pain; occas., occasional; PCS, Pain Catastrophizing Scale; Ach, Achenbach Scale; VAS, visual analogue scale 0–10 cm; VRS, verbal rating scale (0=no pain; 10 = worst pain can imagine)
IH n=9, laparotomy n=3; PDA n=2, PDA+IH, n=2; CSF drain n=1. mod/sev, moderate or severe; IQR, inter-quartile range; MSK, musculoskeletal pain; occas., occasional; PCS, Pain Catastrophizing Scale; Ach, Achenbach Scale; VAS, visual analogue scale 0–10 cm; VRS, verbal rating scale (0=no pain; 10 = worst pain can imagine) Table 1Characteristic EPICure cohort Female Male Extremely preterm (n=102)* Term control (n=48)* P-value Extremely preterm (n=61)* Term control (n=29)* P Extremely preterm (n=41)* Term control (n=19)* P-value Participant characteristics Age (yr), mean (range) 19.3 (18.4–20.5) 19.2 (18.1–20.1) 0.29† 19.3 (18.4–20.3) 19.1 (18.1–20.1) 0.52† 19.3 (18.3–20.5) 19.2 (18.2–20.1) 0.45† Height (cm), mean (sd) 163 (9.5) 167 (8.9) 0.02† 158 (6.5) 162 (5.8) 0.004† 172 (6.7) 175 (6.4) 0.052† Weight (kg), mean (sd) 62.7 (13.9) 67.8 (15.6) 0.048† 57.8 (11.5) 63.7 (15.1) 0.06† 69.8 (14.2) 74.1 (14.5) 0.31† Body mass index (kg m−2), mean (sd) 23.4 (4.5) 24.1 (4.7) 0.35† 23.2 (4.2) 24.1 (4.8) 0.36† 23.7 (5.0) 24.1 (4.7) 0.73† Male sex, n (%) 40 (60) 19 (60) 0.94‡ Prior surgery Neonatal/initial admission, n (%) 30 (29) 0 (0) <0.01‡ 13 (21)§ 0 (0) <0.01‡ 17 (41)¶ 0 (0) <0.01‡ Subsequent surgery, n/N (%) 41/93 (44) 15/46 (33) 0.21‡ 26/55 (47) 9/27 (33) 0.9‡ 15/38 (40) 6/19 (32) 0.77‡ Pain history Intensity worst pain in past week VAS, median (IQR) 2.7 (0–5) n=97 1.4 (0–4.5) n=48 0.66† 2.1 (0–6) n=59 1.3 (0–5) n=29 0.80† 0.8 (0–2.8) n=38 1.5 (0–3.5) n=19 0.25† Incidence recurrent pain, %, (n/N) 54 (55/101) 58 (28/48) 0.31‡ 56 (34/60) 62 (18/29) 0.32‡ 51 (21/41) 53 (10/19) 0.31‡ Primary pain site, % MSK 31 headache 22 other 1 MSK 37 headache 19 other 2 MSK 28 headache 28 MSK 31 headache 31 MSK 34 headache 13 other 2 MSK 47 other 5 Pain ranking, % no/mild 78 ≥ mod/sev 22 no/mild 92 ≥ mod/sev 8 0.04‡ no/mild 73 ≥ mod/sev 27 no/mild 86 ≥ mod/sev 14 0.17‡ no/mild 85 ≥ mod/sev 15 no/mild 100 0.08‡ Recurrent pain intensity VRS, mean (sd) 6.2 (2.6) n=55 5.7 (2.5) n=28 0.65† 6.3 (2.8) n=34 5.8 (2.3) n=18 0.88† 6.1 (2.2) n=21 5.5 (2.9) n=10 0.57† Interference because of pain VRS, mean (sd) 3.3 (3.8) 1.4 (2.6) 0.02† 3.3 (4.1) 1.1 (2.2) 0.03† 3.2 (3.4) 1.8 (3.4) 0.30† Analgesia use, % none 74 occas. 20 regular 7 none 77 occas. 19 regular 2 0.47‡ none 64 occas. 26 regular 8 none 72 occas. 24 regular 3 none 73 occas. 10 regular 5 none 89 occas. 11 Questionnaires PCS total score, median (IQR), n.
S, mean (sd) 3.3 (3.8) 1.4 (2.6) 0.02† 3.3 (4.1) 1.1 (2.2) 0.03† 3.2 (3.4) 1.8 (3.4) 0.30† Analgesia use, % none 74 occas. 20 regular 7 none 77 occas. 19 regular 2 0.47‡ none 64 occas. 26 regular 8 none 72 occas. 24 regular 3 none 73 occas. 10 regular 5 none 89 occas. 11 Questionnaires PCS total score, median (IQR), n. 5 (5–14) n=91 5 (0–14) n=45 0.53† 7 (1–16) n=56 6.5 (0–19) n=28 0.93† 2 (0–14) n=35 0 (0–7) n=17 0.23† DSM anxiety T score Ach, median (IQR), n 52 (50–58) n=95 50 (50–54) n=45 0.01† 52 (50–60) n=57 50 (50–54.2) n=28 0.10† 52 (50–58) n=38 50 (50–54) n=17 0.06† Full-scale intelligence quotient score, mean (sd) 87.2 (14.9) 103.8 (10.1) <0.01† 89.2 (14.5) 102 (8.1) <0.001† 84.2 (15.0) 106 (12.5) <0.01†
35 0 (0–7) n=17 0.23† DSM anxiety T score Ach, median (IQR), n 52 (50–58) n=95 50 (50–54) n=45 0.01† 52 (50–60) n=57 50 (50–54.2) n=28 0.10† 52 (50–58) n=38 50 (50–54) n=17 0.06† Full-scale intelligence quotient score, mean (sd) 87.2 (14.9) 103.8 (10.1) <0.01† 89.2 (14.5) 102 (8.1) <0.001† 84.2 (15.0) 106 (12.5) <0.01† Thermal thresholds and cold tolerance Thenar eminence sensitivity for all thermal modalities (CDT, WDT, CPT, HPT) was reduced in the EP vs TC group (Fig. 2a; Supplementary Table 2). Consistent with previous group differences at 11 yr,13 median CPT was lower (–3.8°C, 95%CI –5 to –0.6, P=0.01) and HPT was higher (2.6°C 95%CI 0.2–3.6, P=0.03) in EP vs TC participants. This was on a background of age-related increase in threshold in both TC (median HPT at 19 vs 11 yr +4.4°C, 95%CI 2.6–5.9) and EP participants (+4.3°C, 95%CI 2.5–5.8; Supplementary Table 3). Within-subject sensitivity to heat and cold was inversely correlated in both TC (HPT and CPT: rs=–0.80, 95%CI –0.91 to –0.63, P<0.01) and EP participants (rs=–0.82, 95%CI –0.87 to –0.73, P<0.01).Fig 2 Thermal sensitivity is influenced by EP status, sex, and stimulus intensity. (a) Raw thermal threshold data in extreme preterm (EP) and term control (TC) participants show group differences in cold detection threshold (CDT), warm detection threshold (WDT), cold pain threshold (CPT), and heat pain threshold (HPT). Scatter plot and median (inter-quartile range); *P<0.01, †P<0.05. (b) Generalised thermal sensitivity (composite of time to HPT, time to CPT, and duration of cold pressor immersion; maximum score 70) demonstrates increased tolerance of thermal stimuli (decreased sensitivity) in EP males with prior neonatal surgery (EP+surg) but increased sensitivity in females. Data points=mean [95%CI] two-way ANOVA *P<0.01 male vs female EP+surgery. (c–h) Thermal survival curves for HPT in females (c) and males (d) and CPT in females (e) and males (f) demonstrate decreased sensitivity in males, particularly after neonatal surgery (EP+surg). Cold pressor tolerance was significantly reduced in EP females (g), particularly after neonatal surgery, but did not differ in males (h). Log rank Mantel–Cox comparison: ‡P<0.05, TC vs EP; ¶P<0.05, TC vs EP; §P<0.01 TC vs EP + surg.
onstrate decreased sensitivity in males, particularly after neonatal surgery (EP+surg). Cold pressor tolerance was significantly reduced in EP females (g), particularly after neonatal surgery, but did not differ in males (h). Log rank Mantel–Cox comparison: ‡P<0.05, TC vs EP; ¶P<0.05, TC vs EP; §P<0.01 TC vs EP + surg. Fig 2 When evaluating static thermal thresholds, more EP participants reached thermal test limits without experiencing discomfort/pain. Twenty-six (27%) EP and 2 (4%) TC had HPT >49°C, and 26 (27%) EP and 5 (10%) TC had CPT <11°C. Survival curves evaluated subgroup effects at the limits of testing (Fig. 2c–f), with failure to reach HPT or CPT most common in EP males with neonatal surgery. Raw data analyses also identify sex-dependent differences related to EP status and neonatal surgery (Supplementary Table 4).
) EP and 5 (10%) TC had CPT <11°C. Survival curves evaluated subgroup effects at the limits of testing (Fig. 2c–f), with failure to reach HPT or CPT most common in EP males with neonatal surgery. Raw data analyses also identify sex-dependent differences related to EP status and neonatal surgery (Supplementary Table 4). In response to a more prolonged noxious cold stimulus, EP participants were more likely to withdraw the hand before 30 s of cold pressor testing (OR=2.2, 95%CI 1.1–4.4), particularly EP surgery females (Fig. 2g). In EP males, cold pressor tolerance did not differ from TC, and there was a relative left-shift compared with threshold survival curves. GTS provided a summary measure incorporating time to HPT and CPT and duration of cold pressor tolerance, with higher scores (range 0–70 s) representing reduced sensitivity. Truncated regression modelling identified significant interactions between EP surgery and sex (Supplementary Table 5), with decreased sensitivity in EP surgery males (69 s, 95%CI 53–85) but increased sensitivity in EP surgery females (39 s 95%CI 30–48; Fig. 2b).
r scores (range 0–70 s) representing reduced sensitivity. Truncated regression modelling identified significant interactions between EP surgery and sex (Supplementary Table 5), with decreased sensitivity in EP surgery males (69 s, 95%CI 53–85) but increased sensitivity in EP surgery females (39 s 95%CI 30–48; Fig. 2b). Thermal sensitivity and amygdala volume Imaging data were available for 39 TC and 72 EP QST participants, including 16 of 30 EP neonatal surgery participants. The volume of pain-relevant brain regions was influenced by preterm status, sex, or both (Fig. 3a and b; Supplementary Fig. 1), with significant correlations with thermal sensitivity for the thalamus and amygdala (Supplementary Table 6). Amygdala volume was lower in EP than TC participants, with a significant main effect of EP status (F1,111=50, P<0.01) and sex (F1,111=23, P<0.01). Amygdalothalamic tract fractional anisotropy differed between TC females and EP females, but there were no differences in axonal volume across groups (Supplementary Fig. 2) and no difference in tissue composition using T2 relaxometry has been reported in this cohort.37 Lower amygdala volume sex-dependently correlated with reduced thermal sensitivity (HPT, CPT, and cold pressor tolerance) in males, but increased sensitivity in females (Supplementary Table 6). In EP participants, amygdala volume was negatively correlated with HPT in males (rs=–0.43, P=0.03) but positively in females (rs=0.44, P<0.01; Fig. 3e). Adjusting for amygdala volume increased effect sizes in the GTS model. FSIQ was not a significant predictor and therefore excluded (Supplementary Table 5).Fig 3 Amygdala volume and thermal sensitivity. (a,b) Volume of the amygdala (a) and thalamus (b) is influenced by EP status and sex. Scatter plot and mean (95%CI) §P<0.01 EP < TC; ¶P<0.01 female < male. (c) In EP participants, higher heat pain threshold is negatively correlated with amygdala volume in males (R2=0.17, P=0.038) and positively correlated in EP females (R2=0.14, P=0.009). (f) There is no significant relationship or sex difference in TC. Scatter plot, regression line (95%CI; See Supplementary Fig. 2 for amgdalothalamic tract fractional anisotropy and average intra-axonal volume fraction).
gdala volume in males (R2=0.17, P=0.038) and positively correlated in EP females (R2=0.14, P=0.009). (f) There is no significant relationship or sex difference in TC. Scatter plot, regression line (95%CI; See Supplementary Fig. 2 for amgdalothalamic tract fractional anisotropy and average intra-axonal volume fraction). Fig 3
gdala volume in males (R2=0.17, P=0.038) and positively correlated in EP females (R2=0.14, P=0.009). (f) There is no significant relationship or sex difference in TC. Scatter plot, regression line (95%CI; See Supplementary Fig. 2 for amgdalothalamic tract fractional anisotropy and average intra-axonal volume fraction). Fig 3 Thenar and chest wall sensory profiles Differences from TC data are expressed as z-scores to illustrate sensory profiles across thermal and mechanical modalities (Fig. 4). Decreases in thermal mechanical detection (MDT) and pressure pain sensitivity (PPT) in EP males were statistically significant in the neonatal surgery subgroup (Fig. 4b; Supplementary Table 2). Sensory thresholds on the unscarred chest wall are consistent with thenar values (i.e. no difference in females, reduced sensitivity in males; Fig. 4c and d; Supplementary Table 7).Fig 4 Somatosensory profiles on hand and chest. (a,b) Thenar sensory profiles in extremely preterm (EP) females (a) show minor differences in z-score (normalised to term controls, TC). (b) In EP males with neonatal surgery (EP+surg), differences from TC extend across thermal and mechanical modalities. (c,d) Adjacent to neonatal thoracic scars (EP+scar), minor differences in warm and mechanical detection are seen in females (c) but in EP males there are generalised reductions in threshold sensitivity on the chest wall that are more marked in the EP+scar group (d). Scar-related perceptual sensitisation (positive wind-up ratio) and dynamic mechanical allodynia (DMA; numerical rating scale, NRS 0–10) to brush is observed in females (neonatal scars on chest wall or other body sites) and males. Data = z-score mean (95%CI) with increased sensitivity represented as positive and decreased sensitivity as negative values. EP vs TC: ¶P<0.05 ||P<0.01; EP+surgery or EP+scar vs TC: †P<0.05 *P<0.01.
cal rating scale, NRS 0–10) to brush is observed in females (neonatal scars on chest wall or other body sites) and males. Data = z-score mean (95%CI) with increased sensitivity represented as positive and decreased sensitivity as negative values. EP vs TC: ¶P<0.05 ||P<0.01; EP+surgery or EP+scar vs TC: †P<0.05 *P<0.01. Fig 4
cal rating scale, NRS 0–10) to brush is observed in females (neonatal scars on chest wall or other body sites) and males. Data = z-score mean (95%CI) with increased sensitivity represented as positive and decreased sensitivity as negative values. EP vs TC: ¶P<0.05 ||P<0.01; EP+surgery or EP+scar vs TC: †P<0.05 *P<0.01. Fig 4 Localised sensory change adjacent to neonatal scars Testing on the unscarred lateral chest wall was performed in all TC and 63 EP participants. Thirty-three EP participants (22 female, 11 male) had clearly visible thoracic dermatome scars related to open surgery (n=16) or surgical vascular access and chest drain insertion (n=11). Localised decreases in static thermal and mechanical detection thresholds adjacent to neonatal thoracic scars were apparent in EP females (Fig. 4c) but were more marked and on a background of generalised differences in EP males (Fig. 4d). Mechanical detection threshold (MDT) was higher on the chest than the hand (Supplementary Tables 4 and 7), with good correlation between the sites (rs=0.67 for TC; rs=0.68 for EP). Normalised data show a main effect of group (TC vs EP vs EP+scar; F2,135=13, P<0.01), but not sex (F1,135=0.5, P=0.5), with thresholds adjacent to scars higher than TC in both females and males (Supplementary Table 7). This is consistent with the scar-related localised decrease in static mechanical and thermal sensitivity in this cohort at 11 yr.13 A small number of participants in all groups reported either rapid change in perceived thermal intensity (TC vs EP vs EP+scar: 6/48 vs 13/61 vs 10/33) or paradoxical hot/cold sensations (TC vs EP vs EP+scar: 4/48 vs 10/61 vs 4/33).
calised decrease in static mechanical and thermal sensitivity in this cohort at 11 yr.13 A small number of participants in all groups reported either rapid change in perceived thermal intensity (TC vs EP vs EP+scar: 6/48 vs 13/61 vs 10/33) or paradoxical hot/cold sensations (TC vs EP vs EP+scar: 4/48 vs 10/61 vs 4/33). Mechanical perceptual sensitisation (positive wind-up ratio) was more common adjacent to scars [23/31, 75% vs unscarred EP (31/63, 49%) or TC (19/48, 40%); χ2 P<0.01]. Allodynia to brush (DMA rated as VRS 2–10/10) was reported over thoracic (8/31 EP) and other neonatal scars (additional four EP participants VRS 2–6/10; Fig. 4c and d). Within the surgery subgroup, higher scar-related brush allodynia correlated with a lower GTS score (i.e. increased sensitivity; rs=–0.49, P<0.05). Three EP participants declined testing adjacent to scars because of persistent sensitivity. No participants reported brush allodynia on the unscarred chest wall or thenar eminence. Cognitive function and sensory thresholds There was a significant effect of group on FSIQ score (TC, EP, EP+surgery; F2,144=32; P<0.01), but no main effect of sex (F1,144=0.09; P=0.81). Neonatal surgery had a similar added effect in both males (EP vs EP+surgery, 87.4, 13.6 vs 79.6, 16.1; mean, sd) and females (EP vs EP+surgery, 91.3, 14.5 vs 81.1, 12.3). Lower FSIQ correlated with lower brain region volumes in both males and females, but not with sensory thresholds (Supplementary Table 6).
9; P=0.81). Neonatal surgery had a similar added effect in both males (EP vs EP+surgery, 87.4, 13.6 vs 79.6, 16.1; mean, sd) and females (EP vs EP+surgery, 91.3, 14.5 vs 81.1, 12.3). Lower FSIQ correlated with lower brain region volumes in both males and females, but not with sensory thresholds (Supplementary Table 6). Current pain, pain catastrophising, and anxiety Regular pain was common, particularly mild musculoskeletal pain related to work or sporting activity. Moderate-severe pain requiring analgesia or impairing function was more common in EP (22/101; 22%) than TC (4/48; 8%) participants (χ2 P=0.04). For those with regular pain, self-reported interference with activity because of pain was higher in EP participants (Table 1). Higher anxiety and pain catastrophising scores correlated weakly with thermal pain thresholds and more strongly with increased pain severity in EP participants (Supplementary Table 8). No participants had taken analgesia on the test day. More females than males reported headache (26/89; 29% vs 6/60; 10%) and use of analgesia (32% vs 13%), but these outcomes were not influenced by EP status. Prevalence data exclude menstruation pain as many did not spontaneously report this or were taking hormone treatment for symptom management or contraception. In those specifically asked, the mean intensity of period pain was 7.1, 2.3 (VRS 0–10; mean, sd) with 12/30 EP and 5/18 TC females reporting problematic pain that reduced activity.
clude menstruation pain as many did not spontaneously report this or were taking hormone treatment for symptom management or contraception. In those specifically asked, the mean intensity of period pain was 7.1, 2.3 (VRS 0–10; mean, sd) with 12/30 EP and 5/18 TC females reporting problematic pain that reduced activity. After demonstration of sensory tests, pretest anxiety was low and did not correlate with thermal thresholds (Supplementary Table 8). DSM anxiety scores were higher in EP participants (Table 1) with clinically significant scores ≥70 in one of 38 EP males, five of 57 EP females, and two of 28 TC females. All pain catastrophising subscales had high internal consistency (Cronbach's α>0.8) in TC (0.91; subscales 0.81–0.92) and EP (0.91; subscales 0.82–0.91) participants. Overall, pain catastrophising scores were influenced by female sex (P=0.028), and current pain experience (HUI-3 pain score; P=0.032), but not EP status or FSIQ.
catastrophising subscales had high internal consistency (Cronbach's α>0.8) in TC (0.91; subscales 0.81–0.92) and EP (0.91; subscales 0.82–0.91) participants. Overall, pain catastrophising scores were influenced by female sex (P=0.028), and current pain experience (HUI-3 pain score; P=0.032), but not EP status or FSIQ. Discussion This is the first comprehensive evaluation of sex- and modality-dependent somatosensory function in young adults who had been born extremely preterm. Sensitivity to static thermal thresholds was reduced in EP males, but prolonged noxious cold unmasked increased sensitivity in EP females, with the greatest difference in neonatal surgery subgroups. The degree and sex-dependent directionality of altered thermal sensitivity in EP participants correlated with reduced amygdala volume but not with current cognitive function, suggesting the amygdala plays a sex-dependent role in central modulation of experimental pain stimuli. In contrast to these generalised changes, a mixed pattern of sensory loss and sensory gain was localised to neonatal scars in both males and females. EP participants were more likely to report current pain of at least moderate severity, with increased pain intensity also associated with higher anxiety and pain catastrophising scores.
these generalised changes, a mixed pattern of sensory loss and sensory gain was localised to neonatal scars in both males and females. EP participants were more likely to report current pain of at least moderate severity, with increased pain intensity also associated with higher anxiety and pain catastrophising scores. Extremely preterm babies undergo repeated procedural interventions as part of intensive care management and up to a third require surgery to manage complications or congenital anomalies.8, 38 Cumulative pain exposure is difficult to quantify and is confounded by comorbidity. Duration of mechanical ventilation or NICU stay have been used as proxy measures of pain exposure39, 40 and higher numbers of tissue breaking procedures correlate with worse outcome.9 We used neonatal surgery as an indicator of increased tissue injury, although this may also be confounded by disease severity or perioperative instability,41 and specific effects of analgesia or anaesthesia42 cannot be determined from the available data. As also seen here, surgery during initial hospitalisation has a persistent impact on cognitive outcome.8 However, FSIQ scores did not differ between our male and female EP surgical participants, and do not account for differences in the degree or directionality of altered thermal sensitivity in males and females.
data. As also seen here, surgery during initial hospitalisation has a persistent impact on cognitive outcome.8 However, FSIQ scores did not differ between our male and female EP surgical participants, and do not account for differences in the degree or directionality of altered thermal sensitivity in males and females. Temperature detection is mediated by multiple thermosensitive channels responsive to both stimulus intensity and duration.43 In children born very preterm (VP, <32 weeks gestation) thermal threshold sensitivity was no different39, 44 or decreased.11 Our EP participants were born at an earlier gestational age (24.9, 0.8 weeks; mean, sd) and required longer hospital admission (134, 63 days), and the reduced thermal threshold sensitivity and added impact of neonatal surgery noted at 11 yr13 had persisted. This was on a background of expected age-related increase in threshold,31 but clear sex-dependent differences had now emerged. The interindividual variability in thermal pain thresholds is consistent with previous reports,24 but within-subject consistencies included: discrimination of stimulus intensity (heat at higher temperature than warm, cold lower temperature than cool); reduced sensitivity to both hot and cold; and correlations across different body sites. In contrast to these measures of static thermal thresholds, more prolonged and noxious thermal stimuli activate descending modulatory pathways that can shift the balance between inhibition or facilitation of spinal inputs and influence perceived pain intensity.45 Therefore, in addition to measures of static thermal threshold, we also performed cold pressor testing to assess sensitivity to a more prolonged and intense thermal stimulus. Previously, VP children were shown to have reduced threshold sensitivity, but prolonged heat unmasked increased perceptual sensitisation11 and increased activation in pain-relevant brain regions, including primary somatosensory cortex, thalamus, and basal ganglia.46 Reduced cold pressor tolerance has also been previously reported in EP young adults.40 Routine QST profiles do not include prolonged thermal stimuli, but a composite measure including time to thermal thresholds and cold tolerance (GTS) highlighted decreased sensitivity in EP males, increased sensitivity in EP females, and the added impact of neonatal surgery in both.
iously reported in EP young adults.40 Routine QST profiles do not include prolonged thermal stimuli, but a composite measure including time to thermal thresholds and cold tolerance (GTS) highlighted decreased sensitivity in EP males, increased sensitivity in EP females, and the added impact of neonatal surgery in both. We postulate that increased tissue injury and pain in early life contributes to activity-dependent alterations in thermal nociceptive signalling, that are also influenced by sex-dependent differences in central modulation.
iously reported in EP young adults.40 Routine QST profiles do not include prolonged thermal stimuli, but a composite measure including time to thermal thresholds and cold tolerance (GTS) highlighted decreased sensitivity in EP males, increased sensitivity in EP females, and the added impact of neonatal surgery in both. We postulate that increased tissue injury and pain in early life contributes to activity-dependent alterations in thermal nociceptive signalling, that are also influenced by sex-dependent differences in central modulation. Experimental pain sensitivity has been correlated with altered structure and connectivity in central sensory-discriminative (e.g. thermal sensitivity and somatosensory cortical thickness47) and emotional/affective pathways (e.g. visceral sensitivity and thalamus and amygdala volume48), with sex differences in fMRI response predominantly in regions encoding affective pain response.49 In EP participants, thermal sensitivity correlated with amygdala volume. The amygdala attaches emotional significance to sensory information relayed from the thalamus, and altered amygdala connectivity has been associated with pain-related fear in adolescents50 and pain catastrophising in adults.51 Importantly for evaluation of future risk, alterations in amygdala volume and connectivity also predicted the transition from acute to chronic back pain in adults.52 After preterm birth, alterations in brain structure and connectivity persist beyond adolescence,2, 37 and functional correlates include reduced cognitive ability53 and poorer psychosocial functioning.54 More specifically, differences in amygdala volume and connectivity influenced fear processing and emotion recognition after preterm birth.55, 56, 57, 58 Here, amygdala volume correlated with both the degree and directionality of altered thermal sensitivity (i.e. decreased in males, increased in females). As sex-dependent differences in amygdala activation also emerge during adolescence,59, 60 divergence in thermal sensitivity between males and females may be clearer in early adulthood than at younger ages. Alterations in socio-emotional circuits, which are influenced by biological vulnerability, early life adversity, and parenting, have been proposed as a link between preterm birth and subsequent psychosocial and emotional outcomes,56 and we suggest extending this model to include effects on experimental pain sensitivity in EP young adults. These exploratory associations require further evaluation in functional imaging studies.
ty, and parenting, have been proposed as a link between preterm birth and subsequent psychosocial and emotional outcomes,56 and we suggest extending this model to include effects on experimental pain sensitivity in EP young adults. These exploratory associations require further evaluation in functional imaging studies. Neonatal scars were associated with decreased static thresholds but increased dynamic mechanical sensitivity in both males and females, suggesting a different localised effect related to peripheral tissue injury. Comparison across multiple modalities is facilitated by conversion to z-scores, and differences from large reference control datasets identify specific sensory profiles in adults with peripheral neuropathic pain.30, 61 Here, we restricted comparison to contemporaneous age- and sex-matched controls and used a protocol that facilitated comparison with previous preterm cohorts. Despite the relatively small subgroups and limited effect size for some modalities, the sensory profiles illustrate sex-dependent effects, the added impact of neonatal surgery, and a different pattern of generalised and localised sensory change adjacent to neonatal scars. Similar mixed patterns of sensory gain, loss, or both have been reported after inguinal or thoracic surgery in children62, 63 and adults.64, 65 While scar-related sensory changes do not always correlate with reported pain,66, 67 several EP participants had marked brush allodynia or declined testing because of scar-related sensitivity, which may predispose to increased pain after re-injury.68 Repeat surgery in the same dermatome as prior neonatal surgery increased pain scores and analgesic requirements in infants.69 Our laboratory studies in rodents identified long-term alterations after neonatal hindpaw incision that include enhanced re-incision hyperalgesia in adulthood.70, 71 Importantly, prevention by peri-incision local anaesthetic suggests activity-dependent mechanisms that can be modulated by clinically-relevant analgesic interventions.3, 72 Although UK paediatric anaesthetists in 1995 reported regular use of opioids and local anaesthetic techniques for neonates requiring surgery,73 specific data for preterm neonates and this cohort are not available. Additional clinical studies are required to compare the ability of different systemic or regional analgesic techniques to modulate the long-term impact of neonatal surgery.
of opioids and local anaesthetic techniques for neonates requiring surgery,73 specific data for preterm neonates and this cohort are not available. Additional clinical studies are required to compare the ability of different systemic or regional analgesic techniques to modulate the long-term impact of neonatal surgery. Pain is a complex sensory and emotional experience, requiring a biopsychosocial approach to evaluation and management.74 Psychological comorbidities are common and are effective targets for intervention in adolescents and adults with chronic pain.75, 76 While some psychosocial factors can increase resilience or be protective (e.g. social support, active coping), others (e.g. fear of pain, anxiety, catastrophising) increase vulnerability,77, 78 and contribute to sex differences in experimental pain sensitivity.79 After preterm birth, children reported higher pain catastrophising,12 and increased anxiety persists into early adulthood.16 Here, higher anxiety and catastrophising scores in EP young adults correlated with both increased thermal sensitivity and more intense current pain. Detailed pain phenotyping, which incorporates history, QST, anxiety, and pain catastrophising has been suggested for clinical trials,80 and along with neuroimaging,52, 81 may enhance prediction of persistent pain risk and improve personalised pain management.
increased thermal sensitivity and more intense current pain. Detailed pain phenotyping, which incorporates history, QST, anxiety, and pain catastrophising has been suggested for clinical trials,80 and along with neuroimaging,52, 81 may enhance prediction of persistent pain risk and improve personalised pain management. Epidemiological studies associate early life adversity and childhood somatic symptoms with increased risk of chronic pain in adulthood.82 While preterm birth (<37 weeks gestation) in 1958 had a minor impact on prevalence of widespread pain at 45 yr,83 EP survivors now reaching adulthood had more invasive NICU management at much earlier gestational ages. Longitudinal evaluations in extreme preterm cohorts have identified persistent effects on cognitive, mental health and system-specific health outcomes,16, 84 but pain experience is not consistently reported. Based on quality of life or general health care questionnaires, current pain prevalence in VP or EP young adults has been reported as no different,17, 19, 85 decreased,86 or increased.87 Here, we found no difference in overall prevalence, as mild pain was common and the study was not adequately powered for this outcome. However, an increased proportion of EP participants reported moderate–severe recurrent pain that required analgesia and influenced activity. In VP and very low birth weight cohorts, self-reported pain increased throughout the third decade18, 20, 88 when chronic pain generally becomes more prevalent, particularly in women.23 Psychological interventions that encourage adaptive coping and improve self-management of pain have been suggested for preterm-born adults,18 and may be particularly advantageous if high-risk subgroups can be identified, such as females with both altered pain coping style and enhanced sensitivity to noxious stimuli. Standardised use of outcomes that incorporate type of pain, impact on function, and use of health resources by males and females would facilitate comparison across cohorts and more clearly delineate the impact of differing neonatal exposures and preterm birth on subsequent pain experience.
nsitivity to noxious stimuli. Standardised use of outcomes that incorporate type of pain, impact on function, and use of health resources by males and females would facilitate comparison across cohorts and more clearly delineate the impact of differing neonatal exposures and preterm birth on subsequent pain experience. Study limitations include potential selection bias as not all eligible EPICure subjects attended. As long-term follow-up tends to recruit NICU survivors with a relatively favourable outcome89 and EPICure@19 participants had higher mean FSIQ and socioeconomic status than non-participants,22 results may under-estimate overall effects. Some participants did not complete all tests, either because of participant preference, time or test availability, but sample sizes for analyses based on available data are noted. Only half of the neonatal surgery group underwent MRI, which limited the ability to analyse subgroup effects for this outcome. Fewer EP males were tested but with a matched proportion of controls. The vast majority of subjects were Caucasian and differences related to ethnicity were not assessed. As subjects were not asked to self-report gender, dichotomous sex-differences are reported for males and females.
subgroup effects for this outcome. Fewer EP males were tested but with a matched proportion of controls. The vast majority of subjects were Caucasian and differences related to ethnicity were not assessed. As subjects were not asked to self-report gender, dichotomous sex-differences are reported for males and females. Extreme preterm birth affects 0.5–1% of the population1 and in the postsurfactant era more survivors are now reaching adulthood. For this vulnerable group, even modest increases in risk for future illness may represent significant healthcare burdens.84, 90 Understanding persistent biological changes in nociceptive pathways and the psychosocial factors that modulate the risk and impact of persistent pain in later life will enhance awareness and recognition of targets for intervention84, 90 to improve outcome throughout the lifespan. Early life experience and sex should be considered during clinical evaluations of somatosensory function or chronic pain, and when evaluating risk factors for persistent pain. Authors' contributions Study design/planning: S.M.W., S.O., N.M. Study conduct and data acquisition: S.M.W., A.M., H.O’R., J.B., Z.E.-R. Data analysis: S.M.W., H.O’R., A.M. Writing paper: S.M.W. with review: N.M. Review and approval of final manuscript: all authors. Overall planning and conduct of evaluations: EPICure@19 Study Group. Declaration of interest The authors declare that they have no conflicts of interest.
Study conduct and data acquisition: S.M.W., A.M., H.O’R., J.B., Z.E.-R. Data analysis: S.M.W., H.O’R., A.M. Writing paper: S.M.W. with review: N.M. Review and approval of final manuscript: all authors. Overall planning and conduct of evaluations: EPICure@19 Study Group. Declaration of interest The authors declare that they have no conflicts of interest. Funding Medical Research Council, UK (G0401525 to N.M., EPICure@19 Study Group). Department of Health, National Institute for Health Research Biomedical Research Centre funding scheme at University College London Hospital/University College London (part-funding to N.M.). Great Ormond Street Hospital Children's Charity (Projects V2818 and W1071H to S.W.). Supplementary data The following is the supplementary data related to this article:Multimedia component 1 Multimedia component 1 Acknowledgements The authors and EPICure Study Group gratefully acknowledge the contribution of all participants and their families to the current and previous evaluations in this cohort. We also acknowledge the important contributions of all researchers and administrative staff involved in the EPICure@19 study and, in particular, assistance with statistical analysis by Kate Bennett. The EPICure@19 Study Group Investigators include: Neil Marlow, EGA UCL Institute for Women's Health (Principal Investigator); John Cockcroft, Cardiff University; Xavier Golay, UCL Institute of Neurology; John Hurst, UCL; Samantha Johnson, University of Leicester; Sebastien Ourselin, UCL; Suellen Walker, UCL GOS Institute of Child Health; Dieter Wolke, University of Warwick.
arlow, EGA UCL Institute for Women's Health (Principal Investigator); John Cockcroft, Cardiff University; Xavier Golay, UCL Institute of Neurology; John Hurst, UCL; Samantha Johnson, University of Leicester; Sebastien Ourselin, UCL; Suellen Walker, UCL GOS Institute of Child Health; Dieter Wolke, University of Warwick. This article is accompanied by an editorial: Back To The Future: lifelong changes in pain processing in ‘ageing of prematurity‘ by McCarthy & Colvin, Br J Anesth 2018:121:529–531, doi: 10.1016/j.bja.2018.06.017 Supplementary data related to this article can be found at https://doi.org/10.1016/j.bja.2018.03.035.
Editor's key points • Infusing large volumes of crystalloid solutions to replace fluid can be harmful. • The composition of the ‘ideal’ fluid remains elusive. • Solutions containing acetate, lactate, or chloride were given to rats subjected to removal of blood volume. • Some differences in acid-base status were seen but outcomes were similar in all groups. • Composition of resuscitation fluid may be less important than volume. While many crystalloid solutions are commercially available, none have the tonicity and electrolyte content that exactly matches that found in normal plasma. Two litres of 0.9% saline (n-saline) may produce significant hyperchloraemic metabolic acidosis.1, 2, 3 This may be associated with pathogenic effects such as a decrease in glomerular filtration,4, 5 disruption of coagulation,6, 7, 8 and impaired immune function.9, 10 Conversely, both Hartmann's and Ringer's lactate solutions are hypotonic, with a sodium content well below that of plasma, an osmolarity of 276 mOsm litre−1 and osmolality of 257 mOsm litre−1.11 This may be unsuitable in patients with intracranial pathology because of an increased risk of cerebral oedema12 and can also contribute to hyponatraemia.13
Ringer's lactate solutions are hypotonic, with a sodium content well below that of plasma, an osmolarity of 276 mOsm litre−1 and osmolality of 257 mOsm litre−1.11 This may be unsuitable in patients with intracranial pathology because of an increased risk of cerebral oedema12 and can also contribute to hyponatraemia.13 Various anion combinations—including lactate, acetate, bicarbonate, gluconate, and malate—have been developed in an attempt to produce satisfactory balanced salt solutions. Infusion of an appropriate solution should cause minimal adverse physiological impact, particularly in the setting of severe haemorrhage and resuscitation. The optimal choice of resuscitation fluid is still a matter of debate. The American College of Surgeons recommends the use of normal saline or Ringer's lactate for initial management of shock,14 whereas acetate administration is gaining in popularity, particularly in Europe.15 Though some animal models support the use of balanced solutions over saline16, 17 in the treatment of severe haemorrhagic shock, others have been equivocal.18 A common problem in the design of these studies is that the resuscitation solution volumes were fixed (×3 bled volume). Determination of intravascular volume in the context of haemodynamic instability can be difficult. Under-resuscitation results in hypoperfusion, whereas fluid overloading increases complications. Thus, fluid responsiveness should ideally be based on dynamic indices measured before and after a fluid bolus.19
ed volume). Determination of intravascular volume in the context of haemodynamic instability can be difficult. Under-resuscitation results in hypoperfusion, whereas fluid overloading increases complications. Thus, fluid responsiveness should ideally be based on dynamic indices measured before and after a fluid bolus.19 We investigated the effects of three crystalloid solutions on the balance of plasma electrolytes and acid-base status during isovolaemic haemodilution, with resuscitation volumes based on real-time monitoring of ventricular filling. We were unable to find any prior study that specifically addresses this question using a matched osmolar, chloride-rich solution with concurrent measures of circulatory haemodynamics and regional oxygenation. We hypothesised that solutions containing metabolisable anions (acetate and lactate) would result in less change in plasma bicarbonate (our primary endpoint) and better preserved cardiac function and a longer time to death (secondary endpoints). Materials and methods Methods and results are reported according to relevant ARRIVE guidelines20 and compliant with EU directive 2010/63/EU.21 All experiments were performed according to local University College London ethics committee approval and Home Office (United Kingdom) guidelines under the 1986 Scientific Procedures Act. Male Wistar rats (∼300 g body weight) were used in all experiments. Before instrumentation, animals were housed in cages of six on a 12 h:12 h light-dark cycle at 20–25 °C. Access to food and water was unrestricted.
committee approval and Home Office (United Kingdom) guidelines under the 1986 Scientific Procedures Act. Male Wistar rats (∼300 g body weight) were used in all experiments. Before instrumentation, animals were housed in cages of six on a 12 h:12 h light-dark cycle at 20–25 °C. Access to food and water was unrestricted. Animals were anaesthetised with isoflurane (5% in room air) but remained spontaneously breathing throughout. Adequate depth of anaesthesia was ensured by assessing the stability of arterial pressure and heart rate, and lack of pedal withdrawal response to a nociceptive stimulus. Rectal temperature was maintained at 37°C by placing the animals on a heated mat. Cannulation of the left common carotid artery was performed for blood pressure monitoring, blood sampling, and removal, and of the right internal jugular vein to enable fluid administration for haemodilution. A tracheostomy was sited and connected to a T-piece. The bladder was cannulated for drainage of urine. An oxygen sensor (Oxylite™, Oxford Optronix, Didcot, UK), pre-calibrated by the manufacturer, was inserted into the left vastus lateralis muscle for continuous monitoring of tissue PO2 (tPO2), as previously described.22
tracheostomy was sited and connected to a T-piece. The bladder was cannulated for drainage of urine. An oxygen sensor (Oxylite™, Oxford Optronix, Didcot, UK), pre-calibrated by the manufacturer, was inserted into the left vastus lateralis muscle for continuous monitoring of tissue PO2 (tPO2), as previously described.22 After instrumentation, animals remained anaesthetised with isoflurane throughout the experiment with adequacy of anaesthesia regularly monitored by lack of flexor response to paw pinching. Euvolaemia was achieved by administering 4 ml kg−1 n-saline (0.9% sodium chloride; Baxter Healthcare, Thetford, UK) over 10 s followed by a continuous infusion of 15 ml kg−1 h−1. This regimen had been previously determined from our previous studies where mean arterial blood pressure was not altered by more than 10% and ensured adequate filling at baseline.23 The study plan is shown in Fig 1. After a minimum of 30 min stabilisation, the n-saline infusion was terminated, and baseline haemodynamics and tissue oxygenation were recorded (t = 0). Transthoracic echocardiography was performed using a 14 MHz probe scanning at 0–2 cm depth (Vivid 7 Dimension, GE Healthcare, Bedford, UK) by an experienced operator (N.J.E.), in particular, taking note of the left ventricular internal diameter in diastole (LVIDD). Aortic blood flow velocities were determined in the aortic arch using pulsed-wave Doppler. Stroke volume was determined as the product of velocity–time integral and vessel cross-sectional area. Heart rate was determined by measuring the time between cardiac cycles. Ten percent of estimated blood volume was then removed (based on a total of 70 ml kg−1) from the arterial line and replaced by twice the volume of a (blinded) test solution. This was repeated at 15-min cycles. The composition of the three test solutions is given in Table 1. Removed blood was sampled at 15-min intervals for plasma ions, acid-base status, haemoglobin and glucose, and derivation of bicarbonate (ABL800FLEX, Radiometer, Copenhagen, Denmark). The animals were randomised to one of three study groups (n = 15/group) using an Excel-generated (Microsoft Corp. Redmond, WA, USA) table of random numbers.
ampled at 15-min intervals for plasma ions, acid-base status, haemoglobin and glucose, and derivation of bicarbonate (ABL800FLEX, Radiometer, Copenhagen, Denmark). The animals were randomised to one of three study groups (n = 15/group) using an Excel-generated (Microsoft Corp. Redmond, WA, USA) table of random numbers. Sample size was calculated assuming an HCO3− difference of −10 mmol litre−1 (saline), −7 mmol litre−1 (lactate), and −5 mmol litre−1 (acetate) with standard deviation (sd) 2.5 mmol litre−1; 10 rats per group would be sufficient to demonstrate significant differences in bicarbonate at an alpha = 0.05 and beta = 0.1 (90% power), and 15 per group would be sufficient to show a 20 (10)% difference in fluid volume, and 20 (5)% difference in time to death. A pilot study of five rats suggested that this model was robust and should answer the relevant question.Fig 1 Study plan. Bv, 10% blood volume; LVIDD, left ventricular internal diameter in diastole. Table 1 Fluid constituents. Ions reported as mEq litre−1, osmolarity as mOsm litre−1 Fluid Na+ K+ Cl- Mg2+ Lactate Acetate Calculated osmolarity Chloride solution 140 0 140 1 0 0 282 Lactate solution 140 4 112 1 34 0 291 Acetate solution 137 4 110 1.5 0 34 286.5
Sample size was calculated assuming an HCO3− difference of −10 mmol litre−1 (saline), −7 mmol litre−1 (lactate), and −5 mmol litre−1 (acetate) with standard deviation (sd) 2.5 mmol litre−1; 10 rats per group would be sufficient to demonstrate significant differences in bicarbonate at an alpha = 0.05 and beta = 0.1 (90% power), and 15 per group would be sufficient to show a 20 (10)% difference in fluid volume, and 20 (5)% difference in time to death. A pilot study of five rats suggested that this model was robust and should answer the relevant question.Fig 1 Study plan. Bv, 10% blood volume; LVIDD, left ventricular internal diameter in diastole. Table 1 Fluid constituents. Ions reported as mEq litre−1, osmolarity as mOsm litre−1 Fluid Na+ K+ Cl- Mg2+ Lactate Acetate Calculated osmolarity Chloride solution 140 0 140 1 0 0 282 Lactate solution 140 4 112 1 34 0 291 Acetate solution 137 4 110 1.5 0 34 286.5 Constituents of each study solution (mmol litre−1) are shown in Table 1. The chloride (CS) and lactate (LS) solutions were prepared by the Department of Chemistry at UCL and the acetate solution (AS) was supplied by Fresenius−Kabi (Bad Homburg, Germany). The osmolality of the solutions was measured using freezing point depression. Strong ion difference (SID) was calculated as SID = (Na++K+)−(Cl−+lactate).
S) and lactate (LS) solutions were prepared by the Department of Chemistry at UCL and the acetate solution (AS) was supplied by Fresenius−Kabi (Bad Homburg, Germany). The osmolality of the solutions was measured using freezing point depression. Strong ion difference (SID) was calculated as SID = (Na++K+)−(Cl−+lactate). To ensure comparable ventricular filling with the different fluids, echocardiography was repeated 3 min after each fluid bolus and filling volume estimated by the LVIDD. If LVIDD was below baseline at 5 min post-fluid bolus, half of the initial fluid dose was given again. This was repeated at 3-min intervals until either LVIDD≥baseline or a maximum of 5× volume of the blood withdrawn was given. This cycle was repeated at 15-min intervals until death. Death was defined as irreversible loss of cardiac contractility on echocardiography. Animals remained anaesthetised throughout.
ven again. This was repeated at 3-min intervals until either LVIDD≥baseline or a maximum of 5× volume of the blood withdrawn was given. This cycle was repeated at 15-min intervals until death. Death was defined as irreversible loss of cardiac contractility on echocardiography. Animals remained anaesthetised throughout. Statistics All haemodynamic and blood gas data are presented as mean and sd unless otherwise stated. Statistical analysis of parametric data was performed using repeated measures two-way analysis of variance with Bonferroni post hoc correction. Data for total volume fluid administered are shown as median, inter-quartile range, and range, and were analysed using Kruskal–Wallis testing followed by Dunn's test for post hoc comparisons. Survival curves were generated according to the Kaplan–Meier method and were compared with the log-rank test. All statistical analyses were performed using Prism 5.0 software (GraphPad Software, San Diego, CA, USA). Probability values <0.05 were considered significantly different.
Dunn's test for post hoc comparisons. Survival curves were generated according to the Kaplan–Meier method and were compared with the log-rank test. All statistical analyses were performed using Prism 5.0 software (GraphPad Software, San Diego, CA, USA). Probability values <0.05 were considered significantly different. Results All groups had similar physiological values after volume optimisation and 30 min stabilisation. Mean survival times between the groups were similar (Fig 2a). No significant difference was seen between the three groups in the total volume of fluid given per animal [CS: 56 (3) ml; LS: 62 (3) ml; AS 65 (3) ml; Fig 2b].Fig 2 Impact of chloride-, acetate-, and lactate-based fluids on (a) Kaplan–Meier plot of survival time with progressive haemodilution, (b) volume required to maintain ventricular filling (median, inter-quartile and full range), (c) haemoglobin (mean and sd), and (d) serum glucose (mean and sd). AS, acetate-based solution; CS, chloride-based solution; LS, lactate-based solution. n = 15 per group. As the first animal died before Cycle 7, the beginning of Cycle 6 was defined as the last possible point for comparison of the haematological, biochemical, cardiovascular, and tissue oxygen parameters between the three groups. For clarity, all subsequent figures are shown up until this time point. There was a stepwise decrease in haemoglobin (Fig 2c) and an increase in glucose (likely stress-related) (Fig 2d) that was consistent between groups.
aematological, biochemical, cardiovascular, and tissue oxygen parameters between the three groups. For clarity, all subsequent figures are shown up until this time point. There was a stepwise decrease in haemoglobin (Fig 2c) and an increase in glucose (likely stress-related) (Fig 2d) that was consistent between groups. Plasma chloride rose progressively in the CS group but remained unchanged in the LS and AS groups (Fig 3a). There was a corresponding decrease in plasma bicarbonate in the CS group, whereas bicarbonate fell later in the lactate group but was unchanged at the end of Cycle 6 in the acetate group (Fig 3b). Arterial lactate concentrations increased with progressive haemodilution. While the increase was greater in the LS group, this did not differ significantly from the other two groups.Fig 3 Impact of chloride-, acetate-, and lactate-based fluids on serum levels of (a) chloride, (b) bicarbonate, (c) lactate, (d) arterial pH, (e) strong ion difference (SID), and (f) anion gap. All data are mean and sd. *Significant differences between groups, P < 0.05. AS, acetate-based solution; CS, chloride-based solution; LS, lactate-based solution. n = 15 per group.
d fluids on serum levels of (a) chloride, (b) bicarbonate, (c) lactate, (d) arterial pH, (e) strong ion difference (SID), and (f) anion gap. All data are mean and sd. *Significant differences between groups, P < 0.05. AS, acetate-based solution; CS, chloride-based solution; LS, lactate-based solution. n = 15 per group. Despite significant decreases in bicarbonate and SID with chloride (Fig 3d and e), pH remained constant because of compensatory hyperventilation and hypocapnia (Fig 4a). Sodium and potassium concentrations stayed normal throughout (Fig 4b and c). The arterial pH, however, rose markedly with acetate, despite a decrease in SID and no change in anion gap (Fig 3f), but remained unchanged with lactate despite a decrease in SID. Of note, PaCO2 remained similar between groups while the calculated anion gap only showed a late increase with lactate that was significantly different with respect to acetate but not bicarbonate.Fig 4 Impact of chloride-, acetate-, and lactate-based fluids on (a) arterial carbon dioxide tension, (b) serum potassium, and (c) serum sodium. All data are mean and sd. AS, acetate-based solution; CS, chloride-based solution; LS, lactate-based solution. n = 15 per group.
nt with respect to acetate but not bicarbonate.Fig 4 Impact of chloride-, acetate-, and lactate-based fluids on (a) arterial carbon dioxide tension, (b) serum potassium, and (c) serum sodium. All data are mean and sd. AS, acetate-based solution; CS, chloride-based solution; LS, lactate-based solution. n = 15 per group. With increasing haemodilution, there was a stepwise decrease in systemic blood pressure (Fig 5c), however, stroke volume and heart rate remained constant because of the fluid replacement (Fig 5a and b). Skeletal muscle tissue PO2 gradually decreased (Fig 5e), consistent with the progressive decrease in oxygen supply to the muscle with the haemodilution. In terms of haemodynamics, there was no between-group difference in any variable measured.Fig 5 Impact of chloride-, acetate-, and lactate-based fluids on (a) stroke volume, (b) heart rate, (c) blood pressure, (d) arterial oxyhaemoglobin saturation, and (e) skeletal muscle tissue oxygen tension. All data are mean and sd. AS, acetate-based solution; LS, lactate-based solution; S, chloride-based solution; tPO2, muscle tissue oxygen tension. n = 15 per group.
ased fluids on (a) stroke volume, (b) heart rate, (c) blood pressure, (d) arterial oxyhaemoglobin saturation, and (e) skeletal muscle tissue oxygen tension. All data are mean and sd. AS, acetate-based solution; LS, lactate-based solution; S, chloride-based solution; tPO2, muscle tissue oxygen tension. n = 15 per group. Discussion A balanced crystalloid solution should ideally contain constituent electrolytes that impose no physiological stress or disturbance to normal physiology. The challenge in developing a more appropriately balanced crystalloid solution revolves around difficulty in providing a suitable anion to balance the cation content, as a significant amount of the plasma anionic buffer is normally provided by albumin. Using an animal model of progressive haemodilution, we performed a comprehensive assessment of the physiological and biochemical effects of resuscitation with fluids containing the most widely used anions: chloride, lactate, and acetate combined with similar concentrations of sodium. Despite distinct biochemical effects produced by the three separate crystalloids, there were no differences in haemodynamics, tissue oxygenation, total volume of fluid required, or time-to-death.
luids containing the most widely used anions: chloride, lactate, and acetate combined with similar concentrations of sodium. Despite distinct biochemical effects produced by the three separate crystalloids, there were no differences in haemodynamics, tissue oxygenation, total volume of fluid required, or time-to-death. Significant separation of plasma chloride was observed early in the chloride group, as expected. Although this did not cause a decrease in arterial pH, a significant difference in SID was seen. Acetate maintained bicarbonate concentrations albeit with a significant metabolic alkalosis compared with the other two groups. Acetate is metabolised by several organs, but mainly by muscle where it is converted to the central intermediary metabolite, acetyl coenzyme A. In contrast, lactate is mainly converted to either glucose or bicarbonate in the liver; in conditions of disrupted hepatic perfusion, its metabolism may be disturbed. This may account for both the greater decrease in bicarbonate seen with lactate compared with acetate, and the lack of a pH alkalisation effect seen with infusion of lactate. The reduced bicarbonate concentration explains the raised anion gap seen in the lactate group, while the anion gap is maintained in the chloride group because of the induced hyperchloraemia. There was no inter-group difference in serum glucose, thus the effect of organic anions (acetate and lactate) on glucose metabolism described elsewhere24 was not seen.
explains the raised anion gap seen in the lactate group, while the anion gap is maintained in the chloride group because of the induced hyperchloraemia. There was no inter-group difference in serum glucose, thus the effect of organic anions (acetate and lactate) on glucose metabolism described elsewhere24 was not seen. The differences in acid-base status did not impact on either survival or the total volume of resuscitation fluid infused. In a rodent model of severe haemorrhagic shock, acetated Ringer's solution (27 mM acetate) improved metabolic acidosis and prolonged median survival compared with lactated Ringer's or n-saline.18 Tissue injury, assessed by plasma enzyme activities, was most pronounced with lactated Ringer's, medium with n-saline, and least with acetated Ringer's solution. In contrast, in a swine model of severe haemorrhage, resuscitation with Ringer's lactate solution provided the best survival rate followed by n-saline and then Plasmalyte (pH 7.4, 27 mM acetate), yet pH was highest with the AS.25 On the other hand, a comparison of Ringer's lactate and n-saline in another pig haemorrhage model showed no difference in survival but only a possible vasodilatatory effects of n-saline.26 A recent study compared Ringer's lactate, Ringer's acetate, PlasmaLyte, and n-saline in a rat model of haemorrhagic shock in the presence or absence of a 70% partial liver resection.27 Arterial bicarbonate content and pH were higher in animals receiving Ringer's acetate and PlasmaLyte, although haemodynamics (including renal blood flow and blood pressure) and renal oxygenation were similar across all four groups.
del of haemorrhagic shock in the presence or absence of a 70% partial liver resection.27 Arterial bicarbonate content and pH were higher in animals receiving Ringer's acetate and PlasmaLyte, although haemodynamics (including renal blood flow and blood pressure) and renal oxygenation were similar across all four groups. Observational data in patients suggested that restricted use of chloride-rich fluids was associated with a decreased risk of renal complications.28 However, a recent prospective randomised multi-centre trial in intensive care unit patients failed to show any effect on renal injury or other outcomes comparing a balanced crystalloid solution vs n-saline, although the volumes of crystalloid administered were rather small.29 Prophylactic addition of sodium bicarbonate during cardiac surgery to alkalinise the urine failed to prevent acute kidney injury, and was associated with an increase in mortality.30 Our results support a recent Cochrane analysis of randomised trials comparing balanced with unbalanced fluid solutions for use in surgery, which concluded that saline-based fluids were equivalent to the use of balanced solutions in terms of clinical outcome.31
injury, and was associated with an increase in mortality.30 Our results support a recent Cochrane analysis of randomised trials comparing balanced with unbalanced fluid solutions for use in surgery, which concluded that saline-based fluids were equivalent to the use of balanced solutions in terms of clinical outcome.31 The recognition of a metabolic acidosis after administration of chloride-rich solutions often induces clinical concern, with reflex administration of further fluid in the mistaken interpretation that the acidosis signifies ongoing hypovolaemia rather than from iatrogenic hyperchloraemic aetiology. Whether one is a proponent of the classical Henderson–Hasselbalch equation or the Stewart hypothesis, the use of acid-base balance as a specific surrogate of hypovolaemia is flawed and may lead to adverse effects. All crystalloids lack oxygen-carrying capacity,32 and may have potential pro-inflammatory and oxidative effects.33 Ideally, as used in this study, minimally-invasive, real-time dynamic haemodynamic monitoring (e.g. using change in stroke volume or end-diastolic ventricular diameter) should assess the impact of manoeuvres that alter venous return. This permits accurate titration of fluid that should minimise problems related to fluid under- or overload.
his study, minimally-invasive, real-time dynamic haemodynamic monitoring (e.g. using change in stroke volume or end-diastolic ventricular diameter) should assess the impact of manoeuvres that alter venous return. This permits accurate titration of fluid that should minimise problems related to fluid under- or overload. Limitations We used a short-term model of progressive haemodilution to assess the impact of the different crystalloid fluids, with no attempt to return shed blood. In clinical practice, ongoing severe haemorrhage would prompt transfusion of blood. Whether this would result in any difference remains to be seen. However, our objective was to determine the acid-base effects of the three crystalloids in the absence of any confounding variables induced by blood transfusion. The animals were anaesthetised with isoflurane, which carries potential vasodilatatory and negative inotropic consequences that may impact upon tissue perfusion. However, experience over many years with this and similar models has revealed isoflurane to offer the greatest degree of cardiorespiratory stability in comparison with i.v. agents. Urinalysis was not performed as the animals became oliguric and then anuric early in the haemodilution protocol. In view of the short-term nature of the study, urea or creatinine was not measured. While differences likely exist between rat and man in terms of physiology, blood buffering capacity, and blood composition, comparison data in the stress environment are sparse. Similarly, differences in handling of the different anions also needs to be elucidated. Of note, acid-base homeostasis is remarkably similar across all non-hibernating mammalian species. Rats adequately reflect the human response to exercise,34 and we consider the response to haemorrhage would likely be similar. Finally, it would be of interest to repeat the study with the animals under heavy sedation with/without paralysis and controlled mechanical ventilation to maintain fixed PaCO2 values, acknowledging the impact of positive airways pressures and decreased sympathetic activation on the stress response.
d likely be similar. Finally, it would be of interest to repeat the study with the animals under heavy sedation with/without paralysis and controlled mechanical ventilation to maintain fixed PaCO2 values, acknowledging the impact of positive airways pressures and decreased sympathetic activation on the stress response. Conclusions We compared resuscitative effects of chloride-, acetate- and lactate-containing crystalloid solutions in a rodent model of sequential haemodilution. The CS was associated with hyperchloraemia and the largest decreases in bicarbonate and SID, while the AS maintained serum bicarbonate but induced a significant metabolic alkalosis. However, all solutions had equivalent effects on haemodynamics and tissue oxygenation, volumes required to maintain diastolic filling at baseline levels, and time-to-death. Our results from the current study suggest that the metabolisable anions, acetate, and lactate, did not have an impact on either survival or total volume administered when compared with chloride. Regardless of composition, the main challenge in fluid therapy is the identification of the optimal volume that needs to be infused to correct tissue hypoperfusion adequately, but avoid overload. This should be the main goal when considering safe administration of resuscitation fluid.35 Authors' contributions Designed the study: A.D., M.F.M.J., M.M., M.S. Performed the experiments: N.J.E., P.H., A.D. Interpreted the data: N.J.E., M.F.M.J., M.S. Wrote the initial manuscript: N.J.E., M.S. Comment and contribution to the initial manuscript: A.D., P.H., M.F.M.J., M.M.
Conclusions We compared resuscitative effects of chloride-, acetate- and lactate-containing crystalloid solutions in a rodent model of sequential haemodilution. The CS was associated with hyperchloraemia and the largest decreases in bicarbonate and SID, while the AS maintained serum bicarbonate but induced a significant metabolic alkalosis. However, all solutions had equivalent effects on haemodynamics and tissue oxygenation, volumes required to maintain diastolic filling at baseline levels, and time-to-death. Our results from the current study suggest that the metabolisable anions, acetate, and lactate, did not have an impact on either survival or total volume administered when compared with chloride. Regardless of composition, the main challenge in fluid therapy is the identification of the optimal volume that needs to be infused to correct tissue hypoperfusion adequately, but avoid overload. This should be the main goal when considering safe administration of resuscitation fluid.35 Authors' contributions Designed the study: A.D., M.F.M.J., M.M., M.S. Performed the experiments: N.J.E., P.H., A.D. Interpreted the data: N.J.E., M.F.M.J., M.S. Wrote the initial manuscript: N.J.E., M.S. Comment and contribution to the initial manuscript: A.D., P.H., M.F.M.J., M.M. Declaration of interest M.M. is a member of the Editorial Board of the British Journal of Anaesthesia and a consultant for Edwards Lifesciences and Deltex. M.M. has organised educational meetings that have received financial support from Fresenius–Kabi (www.ebpom.org). M.S. is developing a bladder tissue oxygen sensor with Oxford Optronix (clinical study funded by Department of Health and Wellcome Trust) and is a consultant for Deltex Medical and New B Innovation. M.F.M.J., M.M., and M.S. have received speaker fees from Fresenius.
inancial support from Fresenius–Kabi (www.ebpom.org). M.S. is developing a bladder tissue oxygen sensor with Oxford Optronix (clinical study funded by Department of Health and Wellcome Trust) and is a consultant for Deltex Medical and New B Innovation. M.F.M.J., M.M., and M.S. have received speaker fees from Fresenius. Funding UK Medical Research Council Studentship Award (to N.J.E.). Fresenius–Kabi provided an unrestricted grant to support the study. Oxford Optronix provided the tissue PO2 probes. M.S. is a National Institute of Health Research (NIHR) Senior Investigator. The work was performed at University College London Hospitals/University College London (UCLH/UCL) which receives support from the NIHR Biomedical Research Centre funding scheme.
Editor's key points • Tidal recruitment/derecruitment is considered as a mechanism of ventilator-induced lung injury in acute respiratory distress syndrome. • Respiratory oscillations in PaO2 have been hypothesised to be indicative of tidal recruitment/derecruitment. • In a porcine model for lung injury, the authors investigated whether an association between tidal recruitment/decruitment and PaO2 oscillations existed. • There was no change in amplitude of PaO2 oscillations during the respiratory cycle, and an association with tidal recruitment/decruitment was therefore not demonstrated. • These findings challenge the hypothesis that respiratory oscillations in PaO2 are indicative of the lung collapse observed in lung injury. Respiratory oscillations in the partial pressure of arterial oxygen (PaO2) have been hypothesised to indicate tidal recruitment/derecruitment (R/D) in acute respiratory distress syndrome (ARDS).1, 2 Notably, R/D is one of the proposed mechanisms of ventilator-induced lung injury in this condition.3, 4, 5 In addition, PaO2 oscillations per se could potentially cause or augment organ damage by exposing organs to cyclically varying O2 levels.6, 7, 8, 9 Despite the leading hypothesis that PaO2 oscillations are caused by tidal R/D, no study to date has demonstrated their relationship by simultaneous dynamic measurements of both PaO2 and R/D. Moreover, oscillations in PaO2 have been found in mechanically ventilated uninjured lungs with no tendency to collapse.10
7, 8, 9 Despite the leading hypothesis that PaO2 oscillations are caused by tidal R/D, no study to date has demonstrated their relationship by simultaneous dynamic measurements of both PaO2 and R/D. Moreover, oscillations in PaO2 have been found in mechanically ventilated uninjured lungs with no tendency to collapse.10 Therefore, the aims of the present study were to explore whether R/D is the underlying reason for PaO2 oscillations by simultaneously measuring PaO2 and R/D dynamically in a porcine, collapse-prone ARDS model, and to examine whether larger increases in lung collapse during breath-hold manoeuvres are associated with a larger reduction in PaO2. For this purpose, we used newly developed fluorescence-quenching fibreoptic PaO2 probes with a response time of less than 100 ms that allow measurement of PaO2 oscillations in real time in vitro and in vivo11, 12, 13, 14 together with a single-slice dynamic CT (dCT) with a sampling interval of 250 ms affording measurements of R/D during mechanical ventilation.15 Methods Ethical approval This study of five domestic pigs (three males and two females; mean weight [standard deviation (sd)]=29.6 (1.7) kg) at the Hedenstierna laboratory, Uppsala University, Sweden was approved by the regional animal welfare ethics committee (Ref: C98/16) and adhered to Animal Research: Reporting of In Vivo Experiments guidelines.16 Measurements undertaken on the uninjured lungs of animals reported in this study have been published elsewhere.10
Hedenstierna laboratory, Uppsala University, Sweden was approved by the regional animal welfare ethics committee (Ref: C98/16) and adhered to Animal Research: Reporting of In Vivo Experiments guidelines.16 Measurements undertaken on the uninjured lungs of animals reported in this study have been published elsewhere.10 Animal preparation Table 1 shows the baseline characteristics of each animal. The animals were premedicated with i.m. xylazine 2 mg kg−1, ketamine 20 mg kg−1, and midazolam 0.5 mg kg−1, and underwent induction of anaesthesia with i.v. propofol titrated to effect (1–3 mg kg−1). The trachea was intubated and mechanical ventilation subsequently commenced. During preparation and before commencement of the study protocol, the animals were ventilated with volume-controlled ventilation (VCV) at 20–25 breaths per minute (bpm) [to maintain end-tidal CO2 (EtCO2) 4.5–6 kPa], with a tidal volume (VT) of 10 ml kg−1, positive end-expiratory pressure (PEEP) of 5 cm H2O, and an inspiratory:expiratory ratio (I:E) of 1:2. The ventilator tubing and tracheal tube were checked for leaks by analysis of the spirometry data. Anaesthesia was maintained with continuous i.v. ketamine 32 mg kg−1 h−1, fentanyl 4 μg kg−1 h−1, and midazolam 0.16 mg kg−1 h−1. General anaesthesia was confirmed by absence of spontaneous movements and by absence of reaction to painful stimulation between the front hooves. After confirmation of general anaesthesia, muscle relaxation was achieved with an initial bolus of rocuronium 0.2 mg kg−1 followed by 0.1 mg kg−1 boluses when spontaneous ventilatory efforts were detected from the airway gas and pressure traces. The adequacy of anaesthesia was determined during the periods of muscle relaxation by the absence of cardiovascular signs of sympathetic stimulation (increases in heart rate or arterial BP). Maintenance fluids were administered i.v. in the form of isotonic electrolyte solution (Ringerfundin; B. Braun Melsungen AG, Melsungen, Germany) at a rate of 10 ml kg−1 h−1 during the instrumentation phase and 7 ml kg−1 h−1 for the rest of the protocol. Once anaesthetised, bilateral surgical dissections of the neck were performed. The exposed right internal jugular vein was cannulated with a pulmonary artery catheter used for continuous pulmonary artery pressure monitoring, intermittent thermodilution cardiac output monitoring, and core temperature monitoring.
otocol. Once anaesthetised, bilateral surgical dissections of the neck were performed. The exposed right internal jugular vein was cannulated with a pulmonary artery catheter used for continuous pulmonary artery pressure monitoring, intermittent thermodilution cardiac output monitoring, and core temperature monitoring. The right and left internal carotid arteries were cannulated with 20G Leadercath Arterial cannulae (Vygon, Swindon, UK) for the introduction of fibreoptic PaO2 probes.Table 1 Baseline characteristics and post-injury blood gas data for each animal. Pre-lavage blood gas values were within normal limits. Blood gas data presented were measured post-saline lavage. ♂, male; ♀, female. CO, cardiac output; FiO2, fraction of inspired O2; Hb, haemoglobin; PFR, PaO2:FiO2 ratio; sd, standard deviation Table 1Variable Animal number Mean (sd) 1 2 3 4 5 Sex ♂ ♀ ♂ ♀ ♂ — Weight (kg) 31.1 29.0 29.8 31.2 26.7 29.6 (1.7) FiO2 0.5 0.8 0.7 0.7 0.9 0.7 (0.1) PFR 288 285 232 105 276 237 (77) pH 7.29 7.35 7.32 7.23 7.37 7.31 (0.05) PaO2 (kPa) 19.2 38.1 21.7 9.8 33.0 24.4 (11.2) PaCO2 (kPa) 8.2 7.3 7.8 8.9 6.8 7.8 (0.8) Hb (g L−1) 91 80 81 86 76 82 (5) CO (L min−1) 3.7 4.8 3.5 4.2 3.2 3.9 (0.6)
Weight (kg) 31.1 29.0 29.8 31.2 26.7 29.6 (1.7) FiO2 0.5 0.8 0.7 0.7 0.9 0.7 (0.1) PFR 288 285 232 105 276 237 (77) pH 7.29 7.35 7.32 7.23 7.37 7.31 (0.05) PaO2 (kPa) 19.2 38.1 21.7 9.8 33.0 24.4 (11.2) PaCO2 (kPa) 8.2 7.3 7.8 8.9 6.8 7.8 (0.8) Hb (g L−1) 91 80 81 86 76 82 (5) CO (L min−1) 3.7 4.8 3.5 4.2 3.2 3.9 (0.6) Lung injury A collapse-prone lung injury was induced with a technique modified from Lachmann and colleagues.17 Preoxygenation with a fraction of inspired O2 (FiO2) of 1.0 preceded the ventilator disconnection and lavage of the lungs by instillation of 0.9% saline solution (at 37°C) via the tracheal tube. After 30 s, the saline was drained out of the lungs and ventilation recommenced. This process was repeated until a PaO2:FiO2 ratio (PFR) of <300 mm Hg (40 kPa) was achieved. Data collection and processing Cardiorespiratory variables, including peripheral O2 saturations (SpO2), ECG, invasive arterial BP (AS/3 Multi-Parameter Patient Monitor; Datex-Ohmeda, Madison, WI, USA), airway gas composition, flow, and pressure (Capnomac Ultima; Datex-Ohmeda), were continuously monitored and recorded as analogue signals throughout the protocol. PaO2 signals from the fibreoptic probes were continuously collected with OxyLitePro monitors (Oxford Optronix, Abingdon, UK), converted to digital form using PowerLab (ADInstruments, Dunedin, New Zealand) and displayed/recorded with LabChart version 8.1.5 (ADInstruments) with a sampling rate of 10 Hz. Physiological data were processed using R version 3.4.1 (R Core Team, Vienna, Austria).18
OxyLitePro monitors (Oxford Optronix, Abingdon, UK), converted to digital form using PowerLab (ADInstruments, Dunedin, New Zealand) and displayed/recorded with LabChart version 8.1.5 (ADInstruments) with a sampling rate of 10 Hz. Physiological data were processed using R version 3.4.1 (R Core Team, Vienna, Austria).18 Study protocol The animals were positioned in dorsal recumbency on the CT scanner table. FiO2 [mean (sd)=0.7 (0.1)] was set depending on the lung injury achieved with saline lavages (see Table 1), and PaO2 was recorded continuously. A first set of measurements considered tidal ventilation, when animals were ventilated in both pressure-controlled ventilation (PCV) and VCV modes at I:E ratios of 1:2, 2:1, 1:4, and 4:1 to explore the effects of different ventilatory modes on PaO2 and its dynamic changes. Upon each change of ventilator setting, the PaO2 trace was monitored until its mean value was stable, and then recorded for 120 s. The dCT images were recorded in the last 30 s of this period. A second set of experiments considered breath-hold manoeuvres, when whole-lung CT scans were recorded during the first and last 5 s of an imposed 30 s breath hold at:(i) End expiration: the expiratory and inspiratory valves closed at an initial airway pressure of 5 cm H2O airway pressure (Ve) (ii) End inspiration: the valves closed after inspiration of VT 10 ml kg−1 (VT10) (iii) End large inspiration: the valves closed after inspiration of VT 20 ml kg−1 (VT20).
A second set of experiments considered breath-hold manoeuvres, when whole-lung CT scans were recorded during the first and last 5 s of an imposed 30 s breath hold at:(i) End expiration: the expiratory and inspiratory valves closed at an initial airway pressure of 5 cm H2O airway pressure (Ve) (ii) End inspiration: the valves closed after inspiration of VT 10 ml kg−1 (VT10) (iii) End large inspiration: the valves closed after inspiration of VT 20 ml kg−1 (VT20). Breath-hold manoeuvres were repeated multiple times in each animal in sequences designed to ensure all permutations of manoeuvre-order were achieved. The anaesthetised animals were euthanised with a bolus dose of potassium chloride (1–2 mmol kg−1) upon completion of the study protocol. CT image acquisition A SOMATOM Definition Flash or SOMATOM Definition Edge (Siemens, Munich, Germany) were used to acquire all images as series of transverse sections with a reconstituted voxel size of 0.5 × 0.5 × 5 mm. Scans of a single juxta-diaphragmatic thoracic slice were acquired at 50 ms intervals with a 70 kV tube voltage, 246 mA current, and collimation of 64 × 0.6 mm in order to analyse dynamic changes during ventilation. A whole-lung scan was conducted at the start and immediately before the end of each breath-hold manoeuvre using a tube voltage of 80 kV, 364 mA current, and 64 × 60 mm collimation.
intervals with a 70 kV tube voltage, 246 mA current, and collimation of 64 × 0.6 mm in order to analyse dynamic changes during ventilation. A whole-lung scan was conducted at the start and immediately before the end of each breath-hold manoeuvre using a tube voltage of 80 kV, 364 mA current, and 64 × 60 mm collimation. CT image analysis CT images were segmented using 3D Slicer version 4.6.219 (http://www.slicer.org) with exclusion of the mediastinum, diaphragm, inferior vena cava, and hilar vessels. Exclusion of intrapulmonary vessels within regions of increased voxel density was not possible, and these, along with the conducting airways up to the level of the clavicles, were included in the analysis. Every fifth image was analysed producing a final temporal resolution of 250 ms. All images were then sub-segmented according to voxel density:20, 21, 22(i) Collapse: –100 to +100 Hounsfield units (HU) (ii) Poorly aerated lung: –500 to –101 HU (iii) Normally aerated lung: –900 to –501 HU (iv) Overdistended lung: –1000 to –901 HU. The mass of each lung fraction (e.g. collapsed) was then calculated using the mean density and volume of each fraction assuming the lung is composed solely of air and water.20 The fractional mass of each region was then calculated as: (1) (mass of fraction)*100%/(total mass of all fractions). Tidal R/D was defined as the difference between the maximum and minimum measured mass of the collapsed lung during the course of a single breath.
The mass of each lung fraction (e.g. collapsed) was then calculated using the mean density and volume of each fraction assuming the lung is composed solely of air and water.20 The fractional mass of each region was then calculated as: (1) (mass of fraction)*100%/(total mass of all fractions). Tidal R/D was defined as the difference between the maximum and minimum measured mass of the collapsed lung during the course of a single breath. Statistical analysis Statistical analyses were performed in GraphPad Prism (version 7.00 for Windows, GraphPad Software; La Jolla, CA, USA; https://www.graphpad.com). Before analysis, all data were tested for normality and homogeneity of variance. Parametric data were compared with paired, two-tailed, Student's t-test, and non-parametric with Wilcoxon matched-pairs signed-rank test. The level of significance was set at P<0.05 for all tests. Tidal ventilation CT measurements from all animals were compared using a one-way analysis of variance (anova) with multiple comparisons and Greenhouse–Geisser correction (parametric), or using a Wilcoxon matched-pairs signed-rank test (non-parametric). A two-way anova with Šidák correction for multiple comparisons was used for the analysis of CT measurements of inter- and intra-animal variability during tidal ventilation under different conditions.
ons and Greenhouse–Geisser correction (parametric), or using a Wilcoxon matched-pairs signed-rank test (non-parametric). A two-way anova with Šidák correction for multiple comparisons was used for the analysis of CT measurements of inter- and intra-animal variability during tidal ventilation under different conditions. Breath-hold manoeuvres The effect of type of breath-hold manoeuvre on the change in lung collapse was compared using a Kruskal–Wallis test with Dunn's correction for multiple comparisons. A two-way anova was used to examine the effects of individual animal and type of breath-hold manoeuvre on the change in PaO2. Spearman correlations were used to assess the relationship between change in collapse and change in PaO2.
as compared using a Kruskal–Wallis test with Dunn's correction for multiple comparisons. A two-way anova was used to examine the effects of individual animal and type of breath-hold manoeuvre on the change in PaO2. Spearman correlations were used to assess the relationship between change in collapse and change in PaO2. Results Tidal R/D measured by dCT was detected when the expiratory time exceeded the inspiratory time during tidal ventilation Figure 1 shows the changes in compartmental mass over the course of a single breath. The maximum fraction of collapse was significantly larger than the minimum fraction when I:E was below 1 [PCV 1:2 (11.6–19.5%), PCV 1:4 (11.5–19.9%), VCV 1:2 (12.6–20.4%), VCV 1:4 (12.9–20.6%); P<0.05], as quantified by single-slice dCT at a temporal resolution of 250 ms. This effect was consistent between PCV and VCV. There was no difference between the maximum and minimum fractions of collapse when I:E was higher than 1.Fig 1 Changes in compartmental mass over the course of a single breath. Red, atelectasis; yellow, poorly aerated; and green, normally aerated. I:E, inspiratory:expiratory ratio; PCV, pressure-controlled ventilation; VCV, volume-controlled ventilation. Error bars represent standard deviation. Only in conditions where the expiratory time exceeded inspiratory time there was a significant difference between the mean maximum and minimum fractions of collapse, PCV 1:2 (11.6–19.5%), PCV 1:4 (11.5–19.9%), VCV 1:2 (12.6–20.4%), VCV 1:4 (12.9–20.6%). Overdistended mass represented <2% of total mass and remained unchanged throughout the breath in all conditions (not shown).
iratory time there was a significant difference between the mean maximum and minimum fractions of collapse, PCV 1:2 (11.6–19.5%), PCV 1:4 (11.5–19.9%), VCV 1:2 (12.6–20.4%), VCV 1:4 (12.9–20.6%). Overdistended mass represented <2% of total mass and remained unchanged throughout the breath in all conditions (not shown). Fig 1
iratory time there was a significant difference between the mean maximum and minimum fractions of collapse, PCV 1:2 (11.6–19.5%), PCV 1:4 (11.5–19.9%), VCV 1:2 (12.6–20.4%), VCV 1:4 (12.9–20.6%). Overdistended mass represented <2% of total mass and remained unchanged throughout the breath in all conditions (not shown). Fig 1 Respiratory PaO2 oscillation amplitude was not clearly related to R/D A total of n=148 different 30 s sections of respiratory PaO2 oscillation data were analysed, all of which were with VT=10 ml kg−1 and ventilatory frequency=12 bpm. Figure 2a illustrates the relationship between the change in lung collapse and the respiratory PaO2 oscillation amplitude. The mean PaO2 and respiratory PaO2 oscillation amplitude for each ventilator condition are shown in Table 2 and Figure 2b. There was not a strong correlation between mean airway pressure and mean PaO2 during tidal ventilation, with a 0.55 kPa reduction in PaO2 for 1 cm H2O increase in airway pressure (r=−0.46). A one-way repeated measures anova with multiple comparisons supported a significant effect of ventilatory mode on mean PaO2 during tidal ventilation (F(3,57)=7.6; P=0.0002).Fig 2 Mean respiratory PaO2 oscillation amplitude during tidal ventilation under different ventilatory conditions. (a) Correlation between the mean respiratory PaO2 oscillation amplitude (kPa) recorded during CT scanning and the relevant associated CT-measured change in fractional collapse during that ventilatory condition. The linear regression analysis results gave: Pig 1: r2=0.44, gradient=2.33; Pig 2: r2=0.31, gradient=0.62; Pig 3: r2=0.00, gradient=0.00; Pig 4: r2=0.23, gradient=3.06; Pig 5: r2=0.15, gradient=0.96. (b) Mean amplitude (kPa) with error bars representing standard deviation (black dots and lines). Amplitudes are calculated from tidal ventilation both before and during CT for each ventilator condition. Each animal is represented by a different coloured symbol. I:E, inspiratory:expiratory ratio; PCV, pressure-controlled ventilation; VCV, volume-controlled ventilation; x-axis ratios in (b) represent different I:E ratios.
udes are calculated from tidal ventilation both before and during CT for each ventilator condition. Each animal is represented by a different coloured symbol. I:E, inspiratory:expiratory ratio; PCV, pressure-controlled ventilation; VCV, volume-controlled ventilation; x-axis ratios in (b) represent different I:E ratios. Fig 2Table 2 Mean PaO2, amplitude of respiratory PaO2 oscillations, and mean airway pressure during different ventilatory conditions. Values shown are mean (standard deviation) Table 2I:E ratio Pressure-controlled ventilation Volume-controlled ventilation Mean PaO2 (kPa) Oscillation amplitude (kPa) Mean airway pressure (cm H2O) Mean PaO2 (kPa) Oscillation amplitude (kPa) Mean airway pressure (cm H2O) 1:2 25.2 (4.7) 2.6 (0.8) 11 (8) 26.4 (4.6) 2.1 (0.6) 8 (5) 2:1 19.7 (6.6) 1.6 (0.5) 16 (8) 25.9 (5.5) 2.1 (0.7) 11 (6) 1:4 27.1 (2.5) 2.8 (1.0) 8 (7) 26.3 (3.6) 2.1 (0.5) 7 (5) 4:1 22.7 (3.9) 2.0 (0.5) 18 (6) 26.4 (3.2) 2.3 (0.9) 13 (5) Respiratory PaO2 oscillation amplitude was significantly lower during PCV 2:1 when compared to PCV 1:2, PCV 1:4, VCV 1:2 and significantly higher during PCV 1:4 compared to PCV 4:1, PCV 2:1 and VCV 1:4.
1:4 27.1 (2.5) 2.8 (1.0) 8 (7) 26.3 (3.6) 2.1 (0.5) 7 (5) 4:1 22.7 (3.9) 2.0 (0.5) 18 (6) 26.4 (3.2) 2.3 (0.9) 13 (5) Respiratory PaO2 oscillation amplitude was significantly lower during PCV 2:1 when compared to PCV 1:2, PCV 1:4, VCV 1:2 and significantly higher during PCV 1:4 compared to PCV 4:1, PCV 2:1 and VCV 1:4. Whole-lung collapse did not decrease during a 30 s breath-hold manoeuvre with large tidal volume The analysis of n=74 breath-hold manoeuvres undertaken with simultaneous CT showed a significant increase in the fraction of collapse from the start to the end of an imposed breath-hold manoeuvre, as shown in Table 3. The mean (sd) airway pressure decreased concurrently during breath-hold manoeuvres at end expiration by 4 (2) cm H2O, VT10 by 11 (6) cm H2O, and VT20 by 11 (6) cm H2O. The fractional increase in collapse during the end-expiratory breath-hold manoeuvres was significantly larger compared with the other two conditions (14% vs 0.9–2.1%; P<0.0001). There was no difference between the change in collapse with VT10 and VT20 end-inspiratory breath-hold manoeuvres. The change in PaO2 could not be predicted from the change in collapse with simple linear regression (r2=0.23, 0.15, 0.14 for Ve, VT10, and VT20 breath-hold manoeuvres, respectively).Table 3 Mean fractional mass of collapsed lung measured by CT at the start and end of breath-hold manoeuvres. Ve, end expiratory; VT10, end-inspiratory (10 ml kg−1); VT20, end-inspiratory (20 ml kg−1). ∗Ve: t(24)=12; P<0.0001. †VT10: t(24)=3.2; P<0.005. ‡VT20: 95% confidence interval: 0.6–1.5%
manoeuvres, respectively).Table 3 Mean fractional mass of collapsed lung measured by CT at the start and end of breath-hold manoeuvres. Ve, end expiratory; VT10, end-inspiratory (10 ml kg−1); VT20, end-inspiratory (20 ml kg−1). ∗Ve: t(24)=12; P<0.0001. †VT10: t(24)=3.2; P<0.005. ‡VT20: 95% confidence interval: 0.6–1.5% Table 3Animal number Fractional collapse (%) Ve VT10 VT20 Start End Start End Start End 1 18.4 (2.1) 23.8 (2.0) 16.8 (1.4) 17.5 (2.1) 15.3 (1.7) 16.0 (1.3) 2 37.1 (2.9) 55.3 (3.6) 22.5 (7.7) 28.3 (9.4) 13.0 (2.3) 15.8 (1.8) 3 25.4 (1.1) 37.8 (1.4) 19.9 (0.7) 22.5 (0.5) 16.1 (0.6) 17.0 (1.4) 4 19.8 (1.2) 36.8 (3.0) 8.7 (1.6) 10.2 (2.5) 7.0 (0.6) 9.6 (3.2) 5 32.8 (0.9) 53.1 (0.7) 25.6 (3.4) 27.1 (0.7) 19.1 (1.1) 20.0 (1.4) Combined 26.0 (7.1) 40.1 (12.0) 18.6 (6.5) 20.7 (7.1) 14.8 (4.1) 16.1 (3.7) Difference 14.1∗ (6.1) 2.1† (3.0) 0.9 (0.6–1.5)‡
4 (1.1) 37.8 (1.4) 19.9 (0.7) 22.5 (0.5) 16.1 (0.6) 17.0 (1.4) 4 19.8 (1.2) 36.8 (3.0) 8.7 (1.6) 10.2 (2.5) 7.0 (0.6) 9.6 (3.2) 5 32.8 (0.9) 53.1 (0.7) 25.6 (3.4) 27.1 (0.7) 19.1 (1.1) 20.0 (1.4) Combined 26.0 (7.1) 40.1 (12.0) 18.6 (6.5) 20.7 (7.1) 14.8 (4.1) 16.1 (3.7) Difference 14.1∗ (6.1) 2.1† (3.0) 0.9 (0.6–1.5)‡ Significant variation in the PaO2 change for different breath-hold manoeuvres within and between individual animals Figure 3 shows the mean continuous PaO2 recordings from the animals (n=5) during each breath-hold manoeuvre studied before and during CT scanning (n=146). For each individual manoeuvre, the change in PaO2 was calculated from the start of the breath-hold manoeuvre (measured from the start of the imposed airway pressure change) to the subsequent nadir in the trace. The mean (sd) change in PaO2 was –16.5 (6.3) kPa during end-expiratory breath-hold manoeuvres, and –10.5 (5.0) kPa and –4.8 (3.4) kPa for VT10 and VT20 end-inspiratory breath-hold manoeuvres, respectively. A repeated measures anova with Greenhouse–Geisser correction determined that these PaO2 changes were significantly different between each condition (F(2,38)=110; P<0.0001). Supplementary Table S1 shows post hoc multiple comparisons, and Supplementary Figure S1 shows all recorded traces for each animal and breath-hold manoeuvre. These demonstrate high variability between each animal and manoeuvre in the majority of cases.Fig 3 PaO2 and airway pressure traces during breath-hold manoeuvres. The left column shows end-expiratory breath-hold manoeuvres (Ve), the middle column 10 ml kg−1 end-inspiratory breath-hold manoeuvres (VT10), and the right column 20 ml kg−1 end-inspiratory breath-hold manoeuvres (VT20). Red represents PaO2 and blue represents mean airway pressure. Solid lines represent mean of n=5 animal manoeuvres associated with CT imaging. The shaded area represents standard deviation. PaO2 traces have been corrected for the effect of O2 uptake (V˙O2) over time.
l kg−1 end-inspiratory breath-hold manoeuvres (VT20). Red represents PaO2 and blue represents mean airway pressure. Solid lines represent mean of n=5 animal manoeuvres associated with CT imaging. The shaded area represents standard deviation. PaO2 traces have been corrected for the effect of O2 uptake (V˙O2) over time. Fig 3 Discussion This study investigated the relationship between dynamic changes in PaO2 and collapse in anaesthetised, mechanically ventilated pigs with saline-lavage lung injury. We demonstrated dynamic respiratory PaO2 oscillations during tidal ventilation, which increased when CT markers of tidal R/D increased. However, the magnitude of this association was not as large as expected. We found a significant tidal R/D only when I:E <1. Additionally, our study showed an increase in lung collapse even during a 30 s large inspiratory breath-hold manoeuvre and that the increase in collapse was associated with a significant reduction in PaO2 during these manoeuvres, including when the effect of continuous O2 uptake (V˙O2) was considered.
only when I:E <1. Additionally, our study showed an increase in lung collapse even during a 30 s large inspiratory breath-hold manoeuvre and that the increase in collapse was associated with a significant reduction in PaO2 during these manoeuvres, including when the effect of continuous O2 uptake (V˙O2) was considered. The analysis of the respiratory PaO2 oscillation amplitude showed that there were differences in the amplitudes for some conditions; however, these did not match the conditions, in which R/D was detected by dCT. The minimum and maximum amounts of mean fractional collapse (mean (SD)) were 14 (7) % and 17 (9) % measured in PCV 2:1 and PCV 1:4 respectively. Given the small differences and large variability in these values, it is likely that, whilst technically measurable, they do not represent a meaningful physiological or clinical difference. These results suggest a lack of a strong association between R/D and respiratory PaO2 oscillations. This finding does not support the hypothesised strong causal relationship between them,1, 2, 23, 24 and suggests the presence of other contributing determinants of variable shunt fraction within each breath. This proposition is supported by results from studies in the uninjured porcine lung, where the presence of respiratory PaO2 oscillations was demonstrated in the absence of R/D.10 The measured mass of collapse [mean (SD) = 16 (9)%] was lower than that reported in other studies using dCT25, 26, 27 in lung injury, although some of these studies considered different HU ranges. The measured collapse in our study, however, was 88 (59)% higher than that measured in studies examining a similar protocol in the uninjured lung using the same species and model,10 and consistent with findings from intra-vital microscopy.28 This result suggests that other determinants of respiratory PaO2 oscillations should be considered.
in our study, however, was 88 (59)% higher than that measured in studies examining a similar protocol in the uninjured lung using the same species and model,10 and consistent with findings from intra-vital microscopy.28 This result suggests that other determinants of respiratory PaO2 oscillations should be considered. Whilst R/D is likely to cause an increase in the amplitude of respiratory PaO2 oscillations, in the context of a complex physiological system, the effect may be obscured by other competing variables, such as the redistribution of pulmonary blood flow to regions with different ventilation:perfusion ratios.29 This hypothesis is supported by the finding in our study that mean PaO2 did not increase with an increase in mean airway pressure, as reported previously.30, 31 Additionally, CT-measured voxel density used as a surrogate marker of collapse assumes that the higherdensity voxels are true collapse and not another high radiodense material, such as fluid (alveolar flooding) or blood, and so may overstate the degree of true collapse.
an airway pressure, as reported previously.30, 31 Additionally, CT-measured voxel density used as a surrogate marker of collapse assumes that the higherdensity voxels are true collapse and not another high radiodense material, such as fluid (alveolar flooding) or blood, and so may overstate the degree of true collapse. The observed amplitude of oscillations was smaller than had been demonstrated elsewhere.1, 2 This may be partially explained by the smaller VT (10 vs ∼30 ml kg−1) and lower peak PAW (<35 vs >45 cm H2O) during tidal ventilation in our study. Whilst the VT is larger than the recommended human clinical VT,32 it is smaller than those used in studies showing respiratory PaO2 oscillations with amplitudes of >13 kPa, and equivalent to VT measured in spontaneously ventilating pigs.33 In addition, peak pressure was <30 cm H2O throughout the experiments, in contrast to previous studies where peak pressure exceeded 45 cm H2O.1
,32 it is smaller than those used in studies showing respiratory PaO2 oscillations with amplitudes of >13 kPa, and equivalent to VT measured in spontaneously ventilating pigs.33 In addition, peak pressure was <30 cm H2O throughout the experiments, in contrast to previous studies where peak pressure exceeded 45 cm H2O.1 The amount of lung collapse, unexpectedly, did not decrease during imposed inspiratory (VT=10 and 20 ml kg−1) breath-hold manoeuvres. In the context of evidence demonstrating that the majority of recruitment occurs within the first few seconds of application of inflation pressure,15, 33, 34 this result may represent CT measurement of ‘starting’ collapse being taken at a time point already on the plateau of the recruitment curve. In fact, the starting collapse mass recorded in Ve was 31% greater than that recorded in VT20 (26.0–14.8%). However, it is important to recognise that the airway pressure is not maintained during a prolonged pause, as both the expiratory and inspiratory ventilator valves close at the start of manoeuvre, and the pressure in the lungs decreases as a result of the ‘pendelluft’ phenomena35 and oxygen consumption (V˙O2), although we attempted to correct for the effect of V˙O2 in our analysis by subtracting the calculated V˙O2 at each 100 ms time point. The reduction in airway pressure will increase both the amount of poorly aerated regions and lung collapse, which in turn will reduce PaO2 by increasing shunt and V/Q mismatch. Indeed, we found a decrease in airway pressure during all breath-hold manoeuvres.
by subtracting the calculated V˙O2 at each 100 ms time point. The reduction in airway pressure will increase both the amount of poorly aerated regions and lung collapse, which in turn will reduce PaO2 by increasing shunt and V/Q mismatch. Indeed, we found a decrease in airway pressure during all breath-hold manoeuvres. The main limitations of our study are that the porcine model does not comprise all the features observed in human ARDS, and that the PFRs attained were consistent with only mild to moderate lung injury. However, the lavage model is very prone to collapse and is easily recruitable. Thus, this model of lung injury would exaggerate the R/D phenomena and possible R/D dependent PaO2 oscillations. In conclusion, to the best of our knowledge, this is the first study to measure contemporaneously dynamic R/D and PaO2 in a collapse-prone ARDS model. We found a very limited association between R/D and respiratory PaO2 oscillations, certainly much smaller than expected from the published literature. These results challenge the accepted hypothesis that R/D is the main determinant of respiratory PaO2 oscillations in ARDS, where reduction of PaO2 oscillation amplitude is mostly expected from reduction of R/D. Our study warrants further investigation into the dynamic, often overlooked role of pulmonary perfusion within the complex context of pulmonary responses to mechanical ventilation.10, 36 Authors' contributions Study design: FF, GH, CH, AF. Study conduct: FF, NB. Data analysis: DC. Data interpretation: DC, FF, JC, AF, AL, GH. Writing of paper: DC, FF. Critical revision: all authors.
In conclusion, to the best of our knowledge, this is the first study to measure contemporaneously dynamic R/D and PaO2 in a collapse-prone ARDS model. We found a very limited association between R/D and respiratory PaO2 oscillations, certainly much smaller than expected from the published literature. These results challenge the accepted hypothesis that R/D is the main determinant of respiratory PaO2 oscillations in ARDS, where reduction of PaO2 oscillation amplitude is mostly expected from reduction of R/D. Our study warrants further investigation into the dynamic, often overlooked role of pulmonary perfusion within the complex context of pulmonary responses to mechanical ventilation.10, 36 Authors' contributions Study design: FF, GH, CH, AF. Study conduct: FF, NB. Data analysis: DC. Data interpretation: DC, FF, JC, AF, AL, GH. Writing of paper: DC, FF. Critical revision: all authors. Financial support: FF, AF, CH, AL. Declaration of interest The authors declare that they have no conflicts of interest. Funding Wellcome Trust Translation Award (HMRXGK00) to AF and CH; Swedish Heart and Lung Foundation (20170531); Swedish Research Council (K2015-99X-2273101-4) to AL; Oxford University Medical Research Fund (MRF/LSV2014/2091) to FF; Whitaker International Fellow Grant to NB. Appendix A Supplementary data The following are the Supplementary data to this article:Multimedia component 1 Multimedia component 1 Multimedia component 2 Multimedia component 2
Funding Wellcome Trust Translation Award (HMRXGK00) to AF and CH; Swedish Heart and Lung Foundation (20170531); Swedish Research Council (K2015-99X-2273101-4) to AL; Oxford University Medical Research Fund (MRF/LSV2014/2091) to FF; Whitaker International Fellow Grant to NB. Appendix A Supplementary data The following are the Supplementary data to this article:Multimedia component 1 Multimedia component 1 Multimedia component 2 Multimedia component 2 Acknowledgements The authors are grateful to H. McPeak, G. Fioroni, A. Roneus, K. Ahlgren, M. Swälas, M. Andersson, and M. Segelsjö for technical support; M. C. Tran for invaluable assistance with data analysis; and L. Camporota for helpful discussions; additionally, OxSTaR, St Peter's College, and the Nuffield Division of Anaesthetics office for their ongoing support of DC. Appendix A Supplementary data to this article can be found online at https://doi.org/10.1016/j.bja.2018.09.011.
Fatty-acid amide hydrolase (FAAH) is the major catabolic enzyme for a range of bioactive lipids called fatty-acid amides (FAAs).1, 2 These FAAs include N-acyl ethanolamines, such as anandamide (AEA), that act as endogenous ligands for cannabinoid receptors (i.e. endocannabinoids). Other substrates of FAAH include palmitoylethanolamide (PEA), oleoylethanolamine (OEA), and N-acyl-taurines. 2-Arachidonoylglycerol (2-AG) is another related endocannabinoid and FAA, but is metabolised mostly by monoacylglycerol lipase (MAGL). AEA has roles in nociception, fear-extinction memory, anxiety, and depression.3, 4 FAAH knockout mice have elevated brain concentrations of AEA, display an analgesic phenotype in response to acute thermal stimuli, and show reduced pain in formalin and carrageenan inflammatory models.5, 6 FAAH is therefore an attractive drug target for treating pain, anxiety, and depression, although recent clinical trials with FAAH inhibitors were unsuccessful.7, 8
ions of AEA, display an analgesic phenotype in response to acute thermal stimuli, and show reduced pain in formalin and carrageenan inflammatory models.5, 6 FAAH is therefore an attractive drug target for treating pain, anxiety, and depression, although recent clinical trials with FAAH inhibitors were unsuccessful.7, 8 The human FAAH gene contains a commonly carried hypomorphic single-nucleotide polymorphism (SNP) (C385A; rs324420; C allele frequency 74%, A 26%) that significantly reduces the activity of the FAAH enzyme.9 Genetic association studies have investigated the link between this and other FAAH SNPs and pain sensitivity.10, 11, 12 Notably, homozygous carriers of the hypomorphic SNP (A allele) in a cohort of women undergoing breast cancer surgery were less sensitive to cold pain and had a reduced need for postoperative analgesia.10 Furthermore, a mouse knock-in model of the human SNP showed that both the mouse and human SNP carriers display enhanced fear-extinction learning and decreased anxiety-linked behaviours.13 Here, we describe a pain-insensitive patient with a non-anxious disposition presenting with a novel genetic disorder associated with loss of function of FAAH.
-in model of the human SNP showed that both the mouse and human SNP carriers display enhanced fear-extinction learning and decreased anxiety-linked behaviours.13 Here, we describe a pain-insensitive patient with a non-anxious disposition presenting with a novel genetic disorder associated with loss of function of FAAH. Case report A 66-yr-old Caucasian female presented to Raigmore Hospital in Inverness, Scotland for orthopaedic surgery, specifically a trapeziectomy with ligament reconstruction and tendon interposition and extensor pollicis longus realignment after a diagnosis of bilateral pantrapezial osteoarthritis. There was significant deformity and deterioration in the use of the right thumb, which was reported as painless before operation. The pre-assessment note classed her as ASA physical status 1, but highlighted that she had a history of vomiting after intake of morphine.
diagnosis of bilateral pantrapezial osteoarthritis. There was significant deformity and deterioration in the use of the right thumb, which was reported as painless before operation. The pre-assessment note classed her as ASA physical status 1, but highlighted that she had a history of vomiting after intake of morphine. For the surgery, she received general anaesthesia with an ultrasound-guided axillary nerve block. She received fentanyl 50 μg i.v., propofol 200 mg i.v., ondansetron 4 mg i.v. intraoperatively, and levobupivacaine 0.25% (20 ml) for the axillary nerve block. After operation, her pain intensity score was 0/10 until the next day when she was discharged home. The only postoperative analgesic she received in hospital was paracetamol 1 g i.v. in the PACU on the day of her surgery. She also received cyclizine 50 mg i.v. twice. Extraordinarily, she required no postoperative analgesics other than paracetamol for this known painful surgery (trapeziectomy), even after the axillary nerve block had worn off. She showed no pain from pinching or from peripheral i.v. cannula manipulation, which led to further investigations.
clizine 50 mg i.v. twice. Extraordinarily, she required no postoperative analgesics other than paracetamol for this known painful surgery (trapeziectomy), even after the axillary nerve block had worn off. She showed no pain from pinching or from peripheral i.v. cannula manipulation, which led to further investigations. The patient had been diagnosed with osteoarthritis of the hip, which she reported as painless, which was not consistent with the severe degree of joint degeneration. At 65 yr of age, she had undergone a hip replacement and was administered only paracetamol 2 g orally on Postoperative days 1 and 2, reporting that she was encouraged to take the paracetamol, but that she did not ask for any analgesics. She was also administered a single dose of morphine sulphate 10 mg orally on the first postoperative evening that caused severe nausea and vomiting for 2 days. After operation, her pain intensity scores were 0/10 throughout except for one score of 1/10 on the first postoperative evening. Her past surgical history was notable for multiple varicose vein and dental procedures for which she has never required analgesia. She also reported a long history of painless injuries (e.g. suturing of a laceration and left wrist fracture) for which she did not use analgesics. She reported numerous burns and cuts without pain (Supplementary Fig. S1), often smelling her burning flesh before noticing any injury, and that these wounds healed quickly with little or no residual scar. She reported eating Scotch bonnet chilli peppers without any discomfort, but a short-lasting ‘pleasant glow’ in her mouth. She described sweating normally in warm conditions.
tary Fig. S1), often smelling her burning flesh before noticing any injury, and that these wounds healed quickly with little or no residual scar. She reported eating Scotch bonnet chilli peppers without any discomfort, but a short-lasting ‘pleasant glow’ in her mouth. She described sweating normally in warm conditions. The patient lives with her husband, and has a daughter and a son from her previous marriage. Her family history is unremarkable for neuropathy or painful conditions. Her mother and daughter appear to perceive pain normally. Her father (now deceased) had little requirement for pain killers. Her son also reports of having some degree of pain insensitivity, but not to the same extent as her. She does not take any medication at present, and is fit and active with no medical conditions apart from arthritis (Supplementary Fig. S2). She is talkative and happy with an optimistic outlook. On the Generalized Anxiety Disorder-7 anxiety questionnaire taken at age 70, she scored 0/21, classified as mild (the lowest category).14 Likewise, on the Patient Health Questionnaire-9 for depression, she scored 0/29, classified as mild.15 She reported long-standing memory lapses (e.g. frequently forgetting words mid-sentence and placement of keys). She also reported never panicking, not even in dangerous or fearful situations, such as in a recent road traffic accident.
Patient Health Questionnaire-9 for depression, she scored 0/29, classified as mild.15 She reported long-standing memory lapses (e.g. frequently forgetting words mid-sentence and placement of keys). She also reported never panicking, not even in dangerous or fearful situations, such as in a recent road traffic accident. After the painless trapeziectomy surgery and a history of ‘painless operations’, she was referred to and further investigated by pain genetics teams from University College London and the University of Oxford at age 67 yr. Ethical approval was granted from both institutions, and written consent taken from the patient, her two children, and mother. On clinical examination, she had multiple scars around the arms and on the back of her hands. Quantitative sensory testing (Supplementary Fig. S3) demonstrated hyposensitivity to noxious heat both in the hands and feet (see Supplementary data for further clinical details).
rom the patient, her two children, and mother. On clinical examination, she had multiple scars around the arms and on the back of her hands. Quantitative sensory testing (Supplementary Fig. S3) demonstrated hyposensitivity to noxious heat both in the hands and feet (see Supplementary data for further clinical details). Genetic tests identify a microdeletion downstream of FAAH Genomic DNA was isolated from the patient, her two children, and her mother for exome sequencing. After filtering of variants, four candidate mutations in the patient and her son were identified, but none were considered likely to be causal for the phenotype (see Supplementary data). We broadened our genetic analyses and searched for cytogenetic copy number changes across the genome using the CytoScan™ HD Array (Thermo Fisher Scientific, UK). This identified an ∼8 kb heterozygous microdeletion on Chromosome 1 that began ∼4.7 kb downstream from the 3′ end of FAAH (Fig 1a; Supplementary Fig. S4). Polymerase chain reaction and sequencing analyses confirmed that the patient co-inherited the microdeletion and FAAH hypomorphic SNP allele (rs324420) (Fig 1b). Her unaffected mother and daughter did not carry the microdeletion, but her son, who also has some pain-sensitivity deficits, was heterozygous for the microdeletion (Supplementary Fig. S5), but did not carry the hypomorphic SNP allele. One Colombian male (HG01353) (pain phenotype unknown) out of 5008 alleles screened in the 1000 Genomes Project also carries a similar-sized microdeletion (esv3585936 in Supplementary Table S1), but is homozygous wild type for FAAH SNP rs324420.Fig 1 (a) Genomic map showing FAAH, FAAH-OUT, and microdeletion. Human chromosome 1 showing the gene footprints of FAAH and FAAH-OUT. Exons are denoted by blue boxes and the direction of transcription shown by arrows. FAAH single-nucleotide polymorphism (SNP) rs324420 maps to Exon 3 (indicated by an asterisk). The 8131 bp microdeletion detected in the patient is shown flanked by Alu repeated sequences (green boxes), which likely predispose the genomic region to rearrangements. The promoter region and Exons 1 and 2 of FAAH-OUT map to the deleted sequence. (b) Genotype summary. The proband carries both the FAAH-OUT microdeletion and the hypomorphic FAAH SNP, and displayed a full hypoalgesic phenotype. Her son carries the FAAH-OUT microdeletion and had a partial hypoalgesic phenotype. Neither the unaffected mother nor daughter carries the microdeletion.
deleted sequence. (b) Genotype summary. The proband carries both the FAAH-OUT microdeletion and the hypomorphic FAAH SNP, and displayed a full hypoalgesic phenotype. Her son carries the FAAH-OUT microdeletion and had a partial hypoalgesic phenotype. Neither the unaffected mother nor daughter carries the microdeletion. (c–f) Circulating anandamide (AEA), palmitoylethanolamide (PEA), oleoylethanolamine (OEA), and 2-arachidonoylglycerol (2-AG) concentrations. Concentrations of AEA, PEA, OEA, and 2-AG were measured by mass spectrometry from blood samples obtained from the patient and four unrelated normal controls. AEA, PEA, and OEA are substrates for FAAH; 2-AG is not. Controls A and B are homozygous wild type for the hypomorphic SNP; Controls C and D are heterozygous carriers. Average values for the controls were AEA (1.2 pmol ml−1), PEA (43.4 pmol ml−1), OEA (5.1 pmol ml−1), and 2-AG (42.2 pmol ml−1), which is consistent with previous data using a similar measurement protocol.16 Average values for the patient (two measurements) were AEA (2.0 pmol ml−1), PEA (113.1 pmol ml−1), OEA (17.3 pmol ml−1), and 2-AG (45 pmol ml−1). Figure 1
(c–f) Circulating anandamide (AEA), palmitoylethanolamide (PEA), oleoylethanolamine (OEA), and 2-arachidonoylglycerol (2-AG) concentrations. Concentrations of AEA, PEA, OEA, and 2-AG were measured by mass spectrometry from blood samples obtained from the patient and four unrelated normal controls. AEA, PEA, and OEA are substrates for FAAH; 2-AG is not. Controls A and B are homozygous wild type for the hypomorphic SNP; Controls C and D are heterozygous carriers. Average values for the controls were AEA (1.2 pmol ml−1), PEA (43.4 pmol ml−1), OEA (5.1 pmol ml−1), and 2-AG (42.2 pmol ml−1), which is consistent with previous data using a similar measurement protocol.16 Average values for the patient (two measurements) were AEA (2.0 pmol ml−1), PEA (113.1 pmol ml−1), OEA (17.3 pmol ml−1), and 2-AG (45 pmol ml−1). Figure 1 Given the extraordinary phenotype in the patient and the vicinity of the microdeletion to FAAH, we investigated how the microdeletion could be pathogenic. Molecular cloning experiments (see Supplementary data) identified novel 5′exons of an expressed FAAH pseudogene, herein called FAAH-OUT (2.845 kb cDNA; KU950306), that mapped to within the microdeletion (Fig 1a). Tissue expression analyses showed FAAH-OUT to be expressed in a wide range of tissues, including fetal and adult brain, and in dorsal root ganglia (DRG; Supplementary Fig. S6). FAAH-OUT likely encodes a long non-coding RNA (Supplementary Fig. S7). We considered that the microdeletion may negate the normal function of FAAH through a reduction in neural expression of FAAH-OUT or through loss of a critical genomic regulatory element for FAAH, and hence, obtained blood samples to measure FAAH-regulated lipids.
y encodes a long non-coding RNA (Supplementary Fig. S7). We considered that the microdeletion may negate the normal function of FAAH through a reduction in neural expression of FAAH-OUT or through loss of a critical genomic regulatory element for FAAH, and hence, obtained blood samples to measure FAAH-regulated lipids. Elevated FAA concentrations in blood To determine the effects of carrying both the microdeletion in FAAH-OUT and the hypomorphic FAAH SNP, we measured the circulating FAA concentrations from blood samples from the patient and four controls, two of which were heterozygous carriers of the SNP. Circulating concentrations of AEA were increased by 70% and of OEA and PEA approximately tripled compared with controls (Fig. 1c–e). The concentrations of 2-AG, another endocannabinoid that is mainly degraded by MAGL and not FAAH, were largely unaltered (Fig 1f). These results are consistent with FAAH having a significant loss of function in the patient.
re increased by 70% and of OEA and PEA approximately tripled compared with controls (Fig. 1c–e). The concentrations of 2-AG, another endocannabinoid that is mainly degraded by MAGL and not FAAH, were largely unaltered (Fig 1f). These results are consistent with FAAH having a significant loss of function in the patient. Discussion The endocannabinoid system is an important physiological system that performs a wide array of homeostatic functions and is important for pain perception.17 FAAH is a critical enzyme for the breakdown of a range of bioactive lipids (including the endocannabinoid AEA and related FAAs and N-acyl-taurines) with diverse physiological roles. Mouse modelling of FAAH loss of function mutations and pharmacological inhibition studies have shown a range of phenotypes, including hypoalgesia, accelerated skin wound healing, enhanced fear-extinction memory, reduced anxiety, and short-term memory deficits.6, 13, 18, 19, 20, 21 Furthermore, human hypomorphic FAAH SNPs are associated with a reduced need for postoperative analgesia, increased postoperative nausea and vomiting induced by opioids, and decreased anxiety-linked behaviours.10, 13, 16, 22, 23, 24
extinction memory, reduced anxiety, and short-term memory deficits.6, 13, 18, 19, 20, 21 Furthermore, human hypomorphic FAAH SNPs are associated with a reduced need for postoperative analgesia, increased postoperative nausea and vomiting induced by opioids, and decreased anxiety-linked behaviours.10, 13, 16, 22, 23, 24 Here, we report a new human genetic disorder in a patient with hypoalgesia, altered fear and memory symptoms, and a non-anxious disposition. This disorder is attributable to co-inheritance of a microdeletion in a novel pseudogene and a known FAAH hypomorphic SNP. The microdeletion is flanked by repeated sequences that likely predispose the region to genomic rearrangements, as seen in other genomic disorders.25 Consequently, there are likely to be additional similar individuals in the general population. The likelihood that this disorder has been under-reported is highlighted by the fact that the patient was diagnosed at age 66 yr despite a recurrent history of painless injuries. Lipid profiling in peripheral blood showed significant increases in AEA, OEA, and PEA, which could be further exaggerated in the brain and DRG. Further work is needed to understand which FAA is the major contributor to the painless phenotype.
was diagnosed at age 66 yr despite a recurrent history of painless injuries. Lipid profiling in peripheral blood showed significant increases in AEA, OEA, and PEA, which could be further exaggerated in the brain and DRG. Further work is needed to understand which FAA is the major contributor to the painless phenotype. The microdeletion removes the promoter and first two exons of FAAH-OUT, but how this disrupts the function of FAAH is still to be elucidated. A hypothesis is that the FAAH-OUT transcript normally functions as a decoy for microRNAs as a result of the high sequence homology, and protects FAAH mRNA from degradation (Supplementary Fig. S7).26 Alternatively, FAAH-OUT may have an epigenetic role in regulating FAAH transcription, or the deletion removes a critical transcriptional regulatory element.25, 27 Future work will help us to understand whether targeting FAAH-OUT by viral shRNA or gene editing techniques is an effective analgesic/anxiolytic drug development strategy. This patient provides new insights into the role of the endocannabinoid system in analgesia and more specifically on the FAAH genomic locus, and highlights the importance of the adjacent, previously uncharacterised FAAH-OUT gene to pain sensation. Given the previous failure of FAAH-inhibitor analgesic drug trials, this report has significance, as it provides a new route to developing FAAH-related analgesia through targeting of FAAH-OUT. Authors' contributions Clinical work: HH, JDR, DLHB, DS. Molecular genetics: AMH, ALO, MCL, SL, SJG, JJC. Exome sequencing data analyses: JTB.
This patient provides new insights into the role of the endocannabinoid system in analgesia and more specifically on the FAAH genomic locus, and highlights the importance of the adjacent, previously uncharacterised FAAH-OUT gene to pain sensation. Given the previous failure of FAAH-inhibitor analgesic drug trials, this report has significance, as it provides a new route to developing FAAH-related analgesia through targeting of FAAH-OUT. Authors' contributions Clinical work: HH, JDR, DLHB, DS. Molecular genetics: AMH, ALO, MCL, SL, SJG, JJC. Exome sequencing data analyses: JTB. Bioinformatics: AMH, ALO, JTB, MCL, JJC. Blood preparation and anandamide analyses: MNH, MvD, MM. Research design: DS, JJC. Wrote the manuscript with help from all authors: JJC. Approved the final manuscript: all authors. Supplementary material Supplementary material is available at British Journal of Anaesthesia online. Declaration of interest The authors declare that they have no conflicts of interest.
Blood preparation and anandamide analyses: MNH, MvD, MM. Research design: DS, JJC. Wrote the manuscript with help from all authors: JJC. Approved the final manuscript: all authors. Supplementary material Supplementary material is available at British Journal of Anaesthesia online. Declaration of interest The authors declare that they have no conflicts of interest. Funding Medical Research Council (Career Development Award G1100340 to JJC); Wellcome Trust (200183/Z/15/Z to JJC, 095698Z/11/Z and 202747/Z/16/Z to DLHB); Alzheimer's Society (research fellowship to JTB), University of Cambridge Academic Foundation Programme (to MCL); Molecular Nociception Group (to MCL); National Institutes of Health (Bethesda, MD, USA) Ruth L. Kirschstein Institutional National Research Service Award (to MCL); Wellcome Trust funded London Pain Consortium (to JDR); Colciencias through a Francisco Jose de Caldas Scholarship (LASPAU, Harvard University) (to JDR); Canadian Institutes of Health Research (CIHR; to MNH); CIHR (postdoctoral funding to MM).
) Ruth L. Kirschstein Institutional National Research Service Award (to MCL); Wellcome Trust funded London Pain Consortium (to JDR); Colciencias through a Francisco Jose de Caldas Scholarship (LASPAU, Harvard University) (to JDR); Canadian Institutes of Health Research (CIHR; to MNH); CIHR (postdoctoral funding to MM). Appendix A Supplementary data The following are the Supplementary data to this article:Multimedia component 1. Multimedia component 1. Multimedia component 2. Multimedia component 2. Multimedia component 3. Multimedia component 3. Multimedia component 4. Multimedia component 4. Multimedia component 5. Multimedia component 5. Multimedia component 6. Multimedia component 6. Multimedia component 7. Multimedia component 7. Multimedia component 8. Multimedia component 8. Multimedia component 9. Multimedia component 9. Multimedia component 10. Multimedia component 10. Multimedia component 11. Multimedia component 11.
t 5. Multimedia component 5. Multimedia component 6. Multimedia component 6. Multimedia component 7. Multimedia component 7. Multimedia component 8. Multimedia component 8. Multimedia component 9. Multimedia component 9. Multimedia component 10. Multimedia component 10. Multimedia component 11. Multimedia component 11. Acknowledgements The authors would like to thank the family for participation in this study, and the volunteers for donating blood samples for analyses. The authors thank John N. Wood for helpful discussions and advice throughout this project. The authors also thank Patrick Fox, Hamish Hay, and Louise Reid (Raigmore Hospital, Inverness, UK); Iain Jones (Southern General Hospital, Glasgow, Scotland); and Judith Singleton (Leith Walk Surgery, Edinburgh, UK) for help with blood sampling, and Sylvie Rose (formerly Addenbrooke's Hospital, Cambridge, UK) for help and advice regarding the cytogenetics analyses. The authors thank the Southern Alberta Mass Spectrometry Centre, located in and supported by the Cumming School of Medicine, University of Calgary, for their services in targeted liquid chromatography–tandem mass spectrometry. Appendix A Supplementary data to this article can be found online at https://doi.org/10.1016/j.bja.2019.02.019.
Editor's key points • The inspired sinewave method can be used to determine cardiac output noninvasively in mechanically ventilated subjects. • This proof-of-concept study aimed to determine the level of agreement and trending ability of this technique with thermodilution-based cardiac output in a pig model. • Under control conditions, there was a good level of agreement but with wide limits of agreements between both methods, with marginal–good trending ability. • The trending ability was reduced by lung lavage-induced acute lung injury. • Further studies should reveal whether inspired sinewave technology could be developed as an accurate cardiac output monitoring technique. Cardiac output monitoring can be used to manage surgical patients at high risk of haemorrhage or haemodynamic instability, and is an essential part of goal-directed therapy.1, 2, 3 Bolus thermodilution remains the current ‘gold standard’ for measuring cardiac output, but the required pulmonary artery catheterisation is highly invasive and has the potential to lead to life-threatening complications.4 Indeed, it is still uncertain whether its application in critically ill patients improves clinical outcomes.5, 6, 7 As such, the use of pulmonary artery catheterisation has declined in favour of techniques that are less invasive, such as transpulmonary thermodilution and those utilising transoesophageal Doppler and pulse contour analysis.8
ncertain whether its application in critically ill patients improves clinical outcomes.5, 6, 7 As such, the use of pulmonary artery catheterisation has declined in favour of techniques that are less invasive, such as transpulmonary thermodilution and those utilising transoesophageal Doppler and pulse contour analysis.8 The ideal cardiac output monitor should be noninvasive and automated, and generate continual, accurate, and precise measurements. Measuring cardiac output via a respiratory based method such as the inspired sinewave technique (IST) has the potential to meet these criteria. The theoretical and experimental basis for the IST originates from the work of Zwart and colleagues9 and Hahn and colleagues,10, 11 whereby the concentration of a tracer gas (e.g. nitrous oxide [N2O]) is sinusoidally oscillated in inspired air. A mathematical model of the lung processes the resultant amplitude/phase of the expired sinewave signal and recovers values for cardiopulmonary variables such alveolar volume and pulmonary blood flow (Q˙IST).12, 13, 14 This proof-of-concept study aimed at validating the Q˙IST technique in anaesthetised, mechanically ventilated pigs during pharmacologically induced haemodynamic changes and a saline lavage model of acute lung injury. We particularly examined the trending ability and agreement of Q˙IST with invasive cardiac output measurements based on thermodilution.
d at validating the Q˙IST technique in anaesthetised, mechanically ventilated pigs during pharmacologically induced haemodynamic changes and a saline lavage model of acute lung injury. We particularly examined the trending ability and agreement of Q˙IST with invasive cardiac output measurements based on thermodilution. Methods Inspired sinewave technique Details of the IST and inspired sinewave device (ISD) have been detailed elsewhere,15 but in brief: at the start of each inhalation a small volume of N2O is injected into inspired gas by a mass flow controller (Alicat Scientific, Inc., Tucson, AZ, USA). Airflow is measured using an ultrasonic flowmeter (VenThor 22/2A; Thor Laboratories, Budapest, Hungary), and N2O concentration using a rapid mainstream infrared sensor (Square One Technology, Boulder, CO, USA). The volume injected is proportional to inspiratory flow, and over time (i.e. successive breaths) the inspired N2O concentration oscillates sinusoidally around a fixed mean (4%), with a predetermined amplitude (3%) and period (60 s). Without venous recirculation of the N2O sinewave, which is assumed negligible at short periods (i.e. <3 min),16 the end-tidal N2O concentration also oscillate in a sinewave pattern (Fig. 1). A single-compartment tidal ventilation model of the lung uses this resulting amplitude and phase of the expired sinewave to estimate cardiopulmonary variables such as pulmonary blood flow (Q˙IST). The mathematical principles of the IST and the hardware layout of the ISD are provided in Supplementary Fig. S1.Fig 1 A typical example of recorded data, with inspired N2O concentration oscillating sinusoidally around a fixed mean (4%), with a predetermined amplitude (3%) and period (180 s). Green line is the expired N2O concentration; the blue and red crosses are the N2O concentrations in inspired gas and end-tidal gas, respectively; the blue and red lines are the inspired and expired N2O sinewaves, respectively (reproduced from Bruce and colleagues15).
h a predetermined amplitude (3%) and period (180 s). Green line is the expired N2O concentration; the blue and red crosses are the N2O concentrations in inspired gas and end-tidal gas, respectively; the blue and red lines are the inspired and expired N2O sinewaves, respectively (reproduced from Bruce and colleagues15). Fig 1 Animal preparation Measurements were made in eight female large white pigs with a weight of 37–53 kg (mean=43.6 kg). All procedures in the study were approved by a local ethics committee and the UK Home Office, London, UK (PPL: 30/2960) and performed at the school of Veterinary Sciences, University of Bristol, Bristol, UK. At the end of each study, the animals were killed under anaesthesia with an overdose of pentobarbital (∼100 mg kg−1). Relevant sections of the Animal Research: Reporting of In Vivo Experiments (ARRIVE) guidelines were followed.
960) and performed at the school of Veterinary Sciences, University of Bristol, Bristol, UK. At the end of each study, the animals were killed under anaesthesia with an overdose of pentobarbital (∼100 mg kg−1). Relevant sections of the Animal Research: Reporting of In Vivo Experiments (ARRIVE) guidelines were followed. All procedures have been detailed elsewhere,17 but in brief: the pigs lay in dorsal recumbency and received i.v. anaesthesia using ketamine, fentanyl, and midazolam. The trachea was intubated, and ventilation was provided in volume-controlled mode using an Engstrӧm Carestation Ventilator (GE Healthcare, Madison, WI, USA), which delivered a tidal volume of 8 ml kg−1, with positive end-expiratory pressure set at 5 cm H2O. Ventilatory frequency was adjusted to maintain PaCO2 between 4.5 and 8 kPa, and it then remained fixed throughout the protocol. Adequacy of general anaesthesia was established with the absence of movement/reflexes the absence of a reaction to painful stimulation between the front hooves, and haemodynamic monitoring. Pancuronium was then administered for muscle relaxation, and Hartmann's solution (10 ml kg−1 h−1) was used for fluid maintenance. Additional boluses of the muscle relaxant were administered when spontaneous ventilatory efforts were detected from the airway gas and pressure traces. The adequacy of anaesthesia was determined during the periods of muscle relaxation by the absence of cardiovascular signs of sympathetic stimulation (increases in heart rate or arterial BP). Anaesthesia was maintained with continuous i.v. ketamine 32 mg kg-1 hr-1, fentanyl 4 mg kg-1 hr-1, and midazolam 0.16 mg kg-1 hr-1.
e adequacy of anaesthesia was determined during the periods of muscle relaxation by the absence of cardiovascular signs of sympathetic stimulation (increases in heart rate or arterial BP). Anaesthesia was maintained with continuous i.v. ketamine 32 mg kg-1 hr-1, fentanyl 4 mg kg-1 hr-1, and midazolam 0.16 mg kg-1 hr-1. A pulmonary artery catheter (PAC; Edward Lifesciences, Irvine, CA, USA) was inserted into the internal jugular vein and advanced into the pulmonary artery. Assessment of pulmonary artery pressure (PAP) and wedge waveforms confirmed the correct position of the PAC. Arterial pressure (via a peripheral artery), PAP, ECG, %SO2, and end-tidal CO2 were continuously recorded, whereas Q˙T was measured intermittently (see below). Variables were monitored with standard patient monitors (Datex Ohmeda Capnomac Ultima; multi-parameter patient monitor, Datex AS3, Madison, WI, USA) and recorded on a computer via PowerLab and LabChart (AD Instruments, Dunedin, New Zealand). Experimental protocol After a stabilisation period of approximately 20 min, baseline measurements were recorded for 10 min, where Q˙T measurements did not fluctuate by more than 15%. Cardiac output was then pharmacologically manipulated via infusions of esmolol (300–400 μg kg−1 min−1), reducing Q˙T by at least 1 L min−1, followed by infusions of dobutamine (15–20 μg kg−1 min−1) increasing Q˙T by at least 3 L min−1 above baseline measurements.
ere Q˙T measurements did not fluctuate by more than 15%. Cardiac output was then pharmacologically manipulated via infusions of esmolol (300–400 μg kg−1 min−1), reducing Q˙T by at least 1 L min−1, followed by infusions of dobutamine (15–20 μg kg−1 min−1) increasing Q˙T by at least 3 L min−1 above baseline measurements. Measurements of Q˙IST are fully automated and continued throughout the entire protocol. Q˙IST is calculated every 30 s as a rolling average of the previous 3 min. Q˙T was calculated using the Datex cardiac output module, and was obtained as the average of three 10 ml injections of normal saline through the PAC. Q˙T measurements were obtained every 3–5 min, and only three measurements ≤10% apart were used to form the average. As each Q˙IST measurement is a rolling average, linear regression analysis between paired Q˙T and Q˙IST measurements was used to determine any time delay, as described elsewhere.18 In these experiments, the time delay of Q˙IST measurements was shown to be around 60 s. After correction, paired cardiac output measurements were used for the subsequent analysis. After these measurements, repeated warm saline lavages were performed (30 ml kg−1 0.9% NaCl at approximately 38°C) until a PaO2/FIO2 ratio of 200–300 mm Hg was achieved, as determined from arterial blood gas analysis. The above experimental protocol was then repeated.
Measurements of Q˙IST are fully automated and continued throughout the entire protocol. Q˙IST is calculated every 30 s as a rolling average of the previous 3 min. Q˙T was calculated using the Datex cardiac output module, and was obtained as the average of three 10 ml injections of normal saline through the PAC. Q˙T measurements were obtained every 3–5 min, and only three measurements ≤10% apart were used to form the average. As each Q˙IST measurement is a rolling average, linear regression analysis between paired Q˙T and Q˙IST measurements was used to determine any time delay, as described elsewhere.18 In these experiments, the time delay of Q˙IST measurements was shown to be around 60 s. After correction, paired cardiac output measurements were used for the subsequent analysis. After these measurements, repeated warm saline lavages were performed (30 ml kg−1 0.9% NaCl at approximately 38°C) until a PaO2/FIO2 ratio of 200–300 mm Hg was achieved, as determined from arterial blood gas analysis. The above experimental protocol was then repeated. Statistical analysis The experiments were designed in accordance with Home Office principles of RRR, essentially minimising animal use whilst preserving data fidelity. Sample sizes were estimated using Mead’s resource equation, as is typical for laboratory animal studies. We chose the sample size to give a value of E (the error degree of freedom) between 10 and 20. Since we used the same animals for measurements before and after lung injury (paired), and a quasi-factorial ‘treatment’ design (i.e. changing cardiac output from resting to higher to lower), a sample size of 8 animals yields an E value of 20. Results are displayed as means 95% confidence interval unless otherwise stated. The agreement of paired absolute Q˙T and Q˙IST values were assessed via calculating their interchangeability rate.19 Values were considered interchangeable if the difference was less than a maximal acceptable value: Q˙T+Q˙IST2xRQ˙T2+RQ˙T2, where RQT is the repeatability coefficient for Q˙T measurements. The agreement between absolute Q˙T and Q˙IST, and between the changes in Q˙T and Q˙IST from baseline (ΔQ˙T and ΔQ˙IST, respectively) were examined via Bland–Altman analysis, and adjusted for multiple measurements within subjects.20 The trending ability of Q˙IST vs Q˙T was examined by calculating the Pearson correlation coefficient (r) for paired ΔQ˙ from the previous paired measurements and assessing the concordance (% of paired ΔQ˙ with same directional change) using a four-quadrant plot. A concordance >92% is considered a good trending ability.21, 22 In addition to assessing the agreement of the direction of ΔQ˙ with a four-quadrant plot, the agreement of the magnitude of ΔQ˙ has been examined with a half-circle polar plot—both calculated with an exclusion zone of 15% mean Q˙.18, 21, 22 With polar plots, the radial length of each data point represents the mean of paired ΔQ˙T and ΔQ˙IST, and the polar angle (°) signifies the agreement of the magnitude of ΔQ˙ between the two methods.
rk has shown that IST measures of cardiopulmonary function are both accurate and reproducible in young healthy volunteers,15 but further work is necessary to validate the technique in human patients/participants by comparing the trending ability of Q˙IST with an appropriate reference method during haemodynamic changes. There are several limitations with the statistical approaches used in the current study. The exclusion zone in the four-quadrant plot and polar analysis aims to remove small ΔQ˙ which may be attributed to noise, but its definition (e.g. an absolute value or % change) is rather arbitrary. Furthermore, the polar analysis exclusion zone may inadvertently remove the most discordant measurements, that is those that have similar absolute values but in opposite directions.18 The concordance rate calculation in the four-quadrant plot does not account for the range of variation as, for example, a plot will be still be considered concordant if one method records ΔQ˙ of 100% and the other only 1%.18 The TIM23 may redress some of the limitations; however, as four-quadrant plots and polar analysis are the most commonly used approaches, and we want to compare the performance of IST with other techniques, they form the basis of our data analysis.
e of ΔQ˙ has been examined with a half-circle polar plot—both calculated with an exclusion zone of 15% mean Q˙.18, 21, 22 With polar plots, the radial length of each data point represents the mean of paired ΔQ˙T and ΔQ˙IST, and the polar angle (°) signifies the agreement of the magnitude of ΔQ˙ between the two methods. A mean polar angle (angular bias) of approximately ±5° with a concordance of >95% (% data points within ±30°) (radial limits of agreement) is considered to have a good trending ability, and a concordance of 90–95% is considered marginal.21 Trending ability was also assessed via the trend interchangeability method (TIM).23 First, changes in Q˙T are determined as ‘interpretable’ if no overlap exists between the confidence intervals (Q˙T±Q˙T×repeatabilitycoefficient) of consecutive measurements. Each interpretable ΔQ˙T is then either determined as interchangeable with ΔQ˙IST, non-interchangeable, or situated within an uncertain interchangeable zone.23 Changes are regarded interchangeable if the second pair of measurements (Q˙T and Q˙IST) are found within a predicted precision interval (i.e. within confidence intervals of interchangeable changes), calculated via a predicted line of identity from the first pair of measurements (Q˙T1 and Q˙IST1) and the repeatability coefficient of Q˙T. These confidence intervals can be defined as follows: X=Y(1+RQT)+(1+RQT)(Q˙T1−Q˙IST1)andX=Y(1-RQT)+(1+RQT)(Q˙T1−Q˙IST1). ΔQ˙ is considered uncertain if its interval of precision intersects one of the confidence intervals, but the value itself does not lie within them.
d Q˙IST1) and the repeatability coefficient of Q˙T. These confidence intervals can be defined as follows: X=Y(1+RQT)+(1+RQT)(Q˙T1−Q˙IST1)andX=Y(1-RQT)+(1+RQT)(Q˙T1−Q˙IST1). ΔQ˙ is considered uncertain if its interval of precision intersects one of the confidence intervals, but the value itself does not lie within them. Differences between baseline variables were assessed via two-tailed unpaired Student's t-tests, with statistical significance set at P<0.05. All analyses were completed using standard statistical software (Sigma plot version 13; Systat Software, Erkrath, Germany).
d Q˙IST1) and the repeatability coefficient of Q˙T. These confidence intervals can be defined as follows: X=Y(1+RQT)+(1+RQT)(Q˙T1−Q˙IST1)andX=Y(1-RQT)+(1+RQT)(Q˙T1−Q˙IST1). ΔQ˙ is considered uncertain if its interval of precision intersects one of the confidence intervals, but the value itself does not lie within them. Differences between baseline variables were assessed via two-tailed unpaired Student's t-tests, with statistical significance set at P<0.05. All analyses were completed using standard statistical software (Sigma plot version 13; Systat Software, Erkrath, Germany). Results Uninjured lung Baseline haemodynamic and blood gas measurements for each animal are shown in Table 1. All animals survived the protocol, which resulted in marked haemodynamic changes (Fig. 2). A total of 248 paired Q˙IST and Q˙T measurements were recorded and the ΔQ˙ values from baseline are presented in Figure 2a–h. Linear regression analysis produced the regression equation ΔQ˙IST=0.68×ΔQ˙T−0.45, with r=0.88. Bland–Altman analysis of absolute Q˙T and Q˙IST values (Fig. 3a) shows a mean bias of 0.79 L min−1 (0.51, 1.07) with limits of agreement (LOA) ranging from 5.5 (4.6, 6.4) to –3.9 L min−1 (–2.8, –4.7). Despite this wide LOA between data points from all animals, the LOA within each animal is substantially narrower: the LOA for each animal ranged between 0.31 and 0.8 L min−1 (mean=0.61 L min−1). This is reflected in the Bland–Altman analysis of ΔQ˙IST vs ΔQ˙T (Fig. 3b), where data were normalised to the individual mean baseline value, showing a mean bias of –0.52 (–0.41, –0.63) with LOA ranging from –2.4 (–2, –2.6) to 1.3 (0.9, 2.6), and linear regression analysis reveals an equation of ΔQ˙IST−ΔQ˙T=0.27×meanΔQ˙−0.56, r=0.48. The interchangeability rate was calculated as 41%. As only poor interchangeability was observed, estimating the range of interchangeable measurements was inappropriate.Table 1 Baseline and respiratory and haemodynamic variables from each animal before repeated saline lavages. DBP, diastolic blood pressure; FIO2, fraction of inspired O2; Hb, haemoglobin; HR, heart rate; IST, inspired sinewave technique; Q˙T, cardiac output from PAC thermodilution; Q˙IST, cardiac output from IST; PAC, pulmonary artery catheter; PaCO2, arterial CO2 partial pressure; PaO2, arterial O2 partial pressure; PAP, mean pulmonary artery pressure; Paw Peak, peak airway pressure; PFR, PaO2/FiO2 ratio; SaO2, arterial oxygen saturation; SBP, systolic blood pressure
put from PAC thermodilution; Q˙IST, cardiac output from IST; PAC, pulmonary artery catheter; PaCO2, arterial CO2 partial pressure; PaO2, arterial O2 partial pressure; PAP, mean pulmonary artery pressure; Paw Peak, peak airway pressure; PFR, PaO2/FiO2 ratio; SaO2, arterial oxygen saturation; SBP, systolic blood pressure Table 1Uninjured Parameter Animal number Mean 1 2 3 4 5 6 7 8 Weight (kg) 46.5 45 39 53 37 37 47 44.5 43.6 HR (beats min−1) 75 70 100 87 60 87 70 102 81 SBP (mm Hg) 113 101 105 80 130 95 105 95 103 DBP (mm Hg) 87 62 71 60 90 85 75 65 74 Q˙T (L min−1) 2.9 6.2 5 4.4 3.9 3.2 6.7 3.7 4.5 Q˙IST (L min−1) 3.9 6 4.1 4.9 8.1 5.6 7.2 9.2 6.1* PAP (mm Hg) 20 16 18 15 17 18 16 21 18 Haemoglobin (g dl−1) 7.8 8.2 8.9 8.3 10.2 8.3 8.8 9.6 8.8 FIO2 0.6 0.28 0.4 0.4 0.35 0.4 0.4 0.4 0.4 SaO2 (%) 99 99 100 99 95 97 100 97 98 pH 7.37 7.48 7.49 7.38 7.31 7.33 7.33 7.24 7.36 PaO2 (mm Hg) 304 115 163 160 186 196 164 175 182 PaCO2 (mm Hg) 59 45 44 46 60 56 58 60 53.5 PFR 507 411 408 400 531 490 410 438 449 Paw peak (cm H20) 15 14 16 18 16 16 15 13 15 * significant difference from Q˙T (P<0.05) Fig 2 (a–h) ΔQ˙ from baseline in all eight animals throughout the protocol. Dark blue diamonds are Q˙T measurements, and light blue dots and lines are Q˙IST measurements. Fig 2Fig 3 Bland–Altman analysis of the agreement between paired measurements of (a) Q˙ISTvsQ˙T and (b) ΔQ˙ISTvs ΔQ˙T from baseline. Colours represent each animal (1–8). Solid line, mean difference (bias); dotted lines, limits of agreement (LOA). Fig 3
It is possible that a potential relationship between chronotropic incompetence and myocardial injury may have been confounded by intraoperative hypotension. However, when we repeated the primary analysis adding intraoperative vasopressor use (a surrogate marker of hypotension) as a co-variate, the results were similar. Conclusions Chronotropic incompetence was associated with both impaired cardiopulmonary function and elevated NT pro-BNP (indicating subclinical heart failure). However, in contrast to parasympathetic measures, chronotropic incompetence was not linked to myocardial injury. These data suggest that a mechanistic role for sympathetic dysregulation in myocardial injury is unlikely, and adds further support to the hypothesis that cardiac vagal dysfunction is the predominant autonomic influence in determining myocardial injury and perioperative outcome.5, 8, 31, 32, 41, 42 Authors' contributions Hypothesis conception: TEFA, GLA Analysis plan design: TEFA, RMP, BHC, DW, GLA Data analysis: TEFA, GLA Writing paper: TEFA, GLA with input from RMP Revising paper: all authors
Fig 2 (a–h) ΔQ˙ from baseline in all eight animals throughout the protocol. Dark blue diamonds are Q˙T measurements, and light blue dots and lines are Q˙IST measurements. Fig 2Fig 3 Bland–Altman analysis of the agreement between paired measurements of (a) Q˙ISTvsQ˙T and (b) ΔQ˙ISTvs ΔQ˙T from baseline. Colours represent each animal (1–8). Solid line, mean difference (bias); dotted lines, limits of agreement (LOA). Fig 3 Trending ability was assessed by calculating the ΔQ˙ from the previous paired measurement values throughout the protocol, with ΔQ˙ <15% mean Q˙ excluded. Linear regression shows a high correlation between ΔQ˙IST and ΔQ˙T of 0.84 (P<0.01). The concordance was 92.5% as assessed using four-quadrant analysis (Fig. 4), where data points located in either quadrant of agreement were considered concordant. Linear regression analysis of the four-quadrant plot shows an equation of ΔQ˙IST=0.65×ΔQ˙T+0.01. Half-circle polar plot analysis (Fig. 5) revealed an angular bias of 5.98° (–24.4°, 36.3°), and a concordance of 92.3% (% of data points within ±30° limits). Half-circle polar plot analysis of the %ΔQ˙ (Supplementary Fig. S3) revealed an angular bias of 4.75° (26.8, 36.2) and a concordance of 90.5%. Trending ability assessed via the TIM demonstrated the following: 65.1% of paired measurements were uninterpretable, 13.2% interchangeable, 9.7% uncertain, and 12% non-interchangeable.Fig 4 Four-quadrant plot analysis of ΔQ˙ISTvs ΔQ˙T throughout the protocol, with an exclusion zone of 15% mean Q˙ (0.67 L min−1). Linear regression analysis reveals an equation of ΔQ˙IST=0.65×ΔQ˙T+0.01, where r=0.84. Data points located in either quadrant of agreement were considered concordant (92.5%).
interchangeable.Fig 4 Four-quadrant plot analysis of ΔQ˙ISTvs ΔQ˙T throughout the protocol, with an exclusion zone of 15% mean Q˙ (0.67 L min−1). Linear regression analysis reveals an equation of ΔQ˙IST=0.65×ΔQ˙T+0.01, where r=0.84. Data points located in either quadrant of agreement were considered concordant (92.5%). Fig 4Fig 5 Half-circle polar-plot analysis of Q˙ISTvsQ˙T throughout the protocol, with an exclusion zone of 15% mean cardiac output (0.67 L min−1). Mean angular bias=5.98° (radial limits, ±30.2°). Data points located within ±30° limits were considered concordant (92.3%). Data points distributed near the polar axis (0°) indicate good trending. Fig 5
Fig 4Fig 5 Half-circle polar-plot analysis of Q˙ISTvsQ˙T throughout the protocol, with an exclusion zone of 15% mean cardiac output (0.67 L min−1). Mean angular bias=5.98° (radial limits, ±30.2°). Data points located within ±30° limits were considered concordant (92.3%). Data points distributed near the polar axis (0°) indicate good trending. Fig 5 Injured lung model Repeated saline lavages resulted in significant reductions in the mean baseline PaO2/FIO2 ratio (250.4 mm Hg [201.4, 299.4] vs 449 mm Hg [413, 455]; P<0.05). Trending ability was examined using the same approaches discussed above (Supplementary Figs S4 and S5). Four-quadrant analysis revealed a concordance of 89.4%, and polar plot analysis showed a mean angular bias of 21.8° (–4.2°, 47.6°) and a concordance of 69.7%. There was a reduced slope in the linear regression equation of the four-quadrant plot (ΔQ˙IST=0.3×ΔQ˙T+0.1), highlighting that impairments in trending ability after lung injury was the result of underestimations in ΔQ˙IST relative to ΔQ˙T. This impaired trending ability was correlated with shunt fraction (Supplementary Fig. S6; r=0.79, P<0.05). The ΔQ˙ from baseline in the injured animals are presented in Supplementary Figure S7a–h. Trending ability assessed via TIM demonstrated: 53.7% of paired measurements were uninterpretable, 7.1% interchangeable, 7.6% uncertain, and 31.5% non-interchangeable.
d with shunt fraction (Supplementary Fig. S6; r=0.79, P<0.05). The ΔQ˙ from baseline in the injured animals are presented in Supplementary Figure S7a–h. Trending ability assessed via TIM demonstrated: 53.7% of paired measurements were uninterpretable, 7.1% interchangeable, 7.6% uncertain, and 31.5% non-interchangeable. Discussion This proof-of-concept study investigated a novel, noninvasive, and fully automated method to continuously monitor changes in cardiac output using the IST. We showed that a prototype device that utilised the IST could be integrated in a breathing circuit of mechanical ventilator, suggesting that it is feasible to use within the perioperative setting. We further demonstrated that a cardiac output based on the IST demonstrated a ‘marginal–good’ trending ability, based upon four-quadrant plot and polar analysis, compared with the pulmonary artery catheter thermodilution reference method.21
ilator, suggesting that it is feasible to use within the perioperative setting. We further demonstrated that a cardiac output based on the IST demonstrated a ‘marginal–good’ trending ability, based upon four-quadrant plot and polar analysis, compared with the pulmonary artery catheter thermodilution reference method.21 Thermodilution is widely considered the gold standard technique for cardiac output monitoring. However, because of its highly invasive nature and the associated complications, its clinical use is becoming less frequent.6, 8 In contrast, the use of less invasive techniques such as those utilising pulse contour analysis or Doppler ultrasound have increased, yet they remain dependent on the skill of the user and their application still requires a somewhat invasive procedure such as an arterial line8 (although arterial vascular access is often routinely required in critically ill patients). Respiratory techniques, such as the IST, have the potential to provide an entirely noninvasive method to monitor cardiac output. The IST also has other qualities of an idealised cardiac output monitor,24 such as its ability to deliver automated and continual measurements that are independent from the skill of the operator. In addition the IST, like other respiratory-based methods,25 can simultaneously estimate other aspects of cardiopulmonary function such as effective lung volume,15 a parameter which may provide additional clinically important information for mechanically ventilated patients.
independent from the skill of the operator. In addition the IST, like other respiratory-based methods,25 can simultaneously estimate other aspects of cardiopulmonary function such as effective lung volume,15 a parameter which may provide additional clinically important information for mechanically ventilated patients. The trending ability of the IST compares well with other noninvasive or minimally invasive methods of monitoring cardiac output. Examining concordance using a four-quadrant plot against thermodilution has been commonly used. The concordance of >92% in the current investigation is commonly regarded as a ‘good’ or clinically acceptable trending ability,21, 22 and this is similar to that observed in human participants with transoesophageal Doppler26, 27 and certain pulse contour techniques,28, 29 and is superior to non-calibrated pulse contour.30, 31 Furthermore, analysis via TIM revealed a similar percentage of paired interchangeable measurements in comparison with pulse contour techniques such LiDCO and PiCCO.23
human participants with transoesophageal Doppler26, 27 and certain pulse contour techniques,28, 29 and is superior to non-calibrated pulse contour.30, 31 Furthermore, analysis via TIM revealed a similar percentage of paired interchangeable measurements in comparison with pulse contour techniques such LiDCO and PiCCO.23 The findings are also similar to other recently developed respiratory-based approaches utilising the Fick principle,32, 33 the trending ability of which have also been examined in pigs. In the current investigation polar plot analysis revealed a mean angular bias that was comparable with other respiratory-based methods,32, 33 and a concordance that was similar to other respiratory-based techniques and considered a ‘marginal’ trending ability.21 Altogether, IST seems to have a marginal–good trending ability relative to PAC thermodilution, and is comparable with other respiratory-based or minimally invasive techniques.
ry-based methods,32, 33 and a concordance that was similar to other respiratory-based techniques and considered a ‘marginal’ trending ability.21 Altogether, IST seems to have a marginal–good trending ability relative to PAC thermodilution, and is comparable with other respiratory-based or minimally invasive techniques. Despite the encouraging trending ability, the absolute agreement between Q˙IST and Q˙T was weaker. This can be easily observed from the examination of the baseline variables where offset errors have occurred, and from the poor interchangeability rate. Bland–Altman analysis of absolute values revealed an acceptable bias. A wide LOA could make it impossible to achieve good trending ability between ΔQ˙IST and ΔQ˙T, but a closer inspection reveals that the majority of bias variability occurs between animals, not within each animal. Indeed, the LOA for each individual animal was low, and so normalisation of values to each animal results in a lower bias and narrower LOA and a good trending ability, which clearly demonstrates the potential of the technique. From a practical perspective, when monitoring patients in many clinical settings absolute values only offer limited information, whereas the temporal variation in values (e.g. the % change in cardiac output) from what is ‘normal’ for each individual patient provides greater clinical insight.34 Indeed, polar plot analysis reveals an even smaller angular bias when ΔQ˙IST and ΔQ˙T are presented as percentage changes.
only offer limited information, whereas the temporal variation in values (e.g. the % change in cardiac output) from what is ‘normal’ for each individual patient provides greater clinical insight.34 Indeed, polar plot analysis reveals an even smaller angular bias when ΔQ˙IST and ΔQ˙T are presented as percentage changes. Bland–Altman analysis of ΔQ˙ from baseline also shows the existence of greater degrees of Q˙IST underestimation with higher mean ΔQ˙. Indeed, it is likely that the LOA at these higher ΔQ˙ is underestimated. This scaling error could be a consequence of the recirculation of the N2O sinewave emerging at higher cardiac outputs and so conflicts with assumptions made by the lung model. Clearly, further work and adjustments to the model are needed to improve the trending (and absolute agreement) of Q˙IST vs Q˙T and, as a first step, simulations using single and multi-compartmental tidal lung models are being developed to investigate these questions.
o conflicts with assumptions made by the lung model. Clearly, further work and adjustments to the model are needed to improve the trending (and absolute agreement) of Q˙IST vs Q˙T and, as a first step, simulations using single and multi-compartmental tidal lung models are being developed to investigate these questions. Data from the IST measurements are averaged over a 3 min rolling window and updated to the monitor every 30 s. The time of 3 min was chosen because it captures three full sinewave periods and gives robust solutions, although a shorter period could be used depending on the degree of stability required. Theoretically this approach has two potential clinical consequences. The first is that it will dampen the dynamic component of the cardiac output time series, and the second is that it results in a lag between real-time cardiac output changes and their registration. Given that even under the most challenging clinical circumstances (other than sudden cardiac arrest), cardiac output changes over a time course of minutes and the ultra-dynamic component of the time series is small; so the 3 min moving average should have so significant impact. In contrast, the ‘lag’ effect will clearly result in a delay in the registration of real-time CO values. In the data presented here, this was measured as being approximately 60 s (cf. contemporaneously measured Q˙T). This property is common to many other similar monitors where averaging is used (Lidco, Deltex) and needs to be taken into consideration.
l clearly result in a delay in the registration of real-time CO values. In the data presented here, this was measured as being approximately 60 s (cf. contemporaneously measured Q˙T). This property is common to many other similar monitors where averaging is used (Lidco, Deltex) and needs to be taken into consideration. The current experiment collected data from mechanically ventilated animals and so may not necessarily apply to human participants or those spontaneously breathing. Previous work has shown that IST measures of cardiopulmonary function are both accurate and reproducible in young healthy volunteers,15 but further work is necessary to validate the technique in human patients/participants by comparing the trending ability of Q˙IST with an appropriate reference method during haemodynamic changes.
ed concordant if one method records ΔQ˙ of 100% and the other only 1%.18 The TIM23 may redress some of the limitations; however, as four-quadrant plots and polar analysis are the most commonly used approaches, and we want to compare the performance of IST with other techniques, they form the basis of our data analysis. As with other respiratory-based approaches, the presence of lung disease or injury will likely impair aspects of Q˙IST performance. The saline lavage model of acute lung injury in the current investigation reduced the trending ability of Q˙IST and the concordance from a four-quadrant plot became slightly poorer. The reduced slope clearly revealed that impairments in trending ability were the result of underestimations in ΔQ˙IST relative to ΔQ˙T. Any cardiac output that bypasses ventilated regions of the lung cannot be quantified by approaches relying on gas exchange measurements, and we have shown that the impaired trending of ΔQ˙IST was significantly correlated with shunt fraction. Further refinement of the lung model is required to address this issue; nevertheless, the estimation of ‘effective’ cardiac output—that is, blood flow that participates in pulmonary gas exchange—by respiratory-based techniques might also provide important clinical information in the critical care setting. For example, as increases in Q˙IST should improve oxygenation and CO2 elimination independently of any change in cardiac output, Q˙IST might provide important additional information in the optimisation of ventilator settings. Authors' contributions Conception of the protocol: AF.
As with other respiratory-based approaches, the presence of lung disease or injury will likely impair aspects of Q˙IST performance. The saline lavage model of acute lung injury in the current investigation reduced the trending ability of Q˙IST and the concordance from a four-quadrant plot became slightly poorer. The reduced slope clearly revealed that impairments in trending ability were the result of underestimations in ΔQ˙IST relative to ΔQ˙T. Any cardiac output that bypasses ventilated regions of the lung cannot be quantified by approaches relying on gas exchange measurements, and we have shown that the impaired trending of ΔQ˙IST was significantly correlated with shunt fraction. Further refinement of the lung model is required to address this issue; nevertheless, the estimation of ‘effective’ cardiac output—that is, blood flow that participates in pulmonary gas exchange—by respiratory-based techniques might also provide important clinical information in the critical care setting. For example, as increases in Q˙IST should improve oxygenation and CO2 elimination independently of any change in cardiac output, Q˙IST might provide important additional information in the optimisation of ventilator settings. Authors' contributions Conception of the protocol: AF. Design of the protocol: RB, PP, AF. Acquisition of data: RB, DC, AM, FF, PP, AF. Analysis and interpretation of the data: all authors. Development of the new technique: RB, MT, PP. Writing paper: RB. Contributed to the production of the manuscript: DC, AM, MT, FF, PP, AF.
As with other respiratory-based approaches, the presence of lung disease or injury will likely impair aspects of Q˙IST performance. The saline lavage model of acute lung injury in the current investigation reduced the trending ability of Q˙IST and the concordance from a four-quadrant plot became slightly poorer. The reduced slope clearly revealed that impairments in trending ability were the result of underestimations in ΔQ˙IST relative to ΔQ˙T. Any cardiac output that bypasses ventilated regions of the lung cannot be quantified by approaches relying on gas exchange measurements, and we have shown that the impaired trending of ΔQ˙IST was significantly correlated with shunt fraction. Further refinement of the lung model is required to address this issue; nevertheless, the estimation of ‘effective’ cardiac output—that is, blood flow that participates in pulmonary gas exchange—by respiratory-based techniques might also provide important clinical information in the critical care setting. For example, as increases in Q˙IST should improve oxygenation and CO2 elimination independently of any change in cardiac output, Q˙IST might provide important additional information in the optimisation of ventilator settings. Authors' contributions Conception of the protocol: AF. Design of the protocol: RB, PP, AF. Acquisition of data: RB, DC, AM, FF, PP, AF. Analysis and interpretation of the data: all authors. Development of the new technique: RB, MT, PP. Writing paper: RB. Contributed to the production of the manuscript: DC, AM, MT, FF, PP, AF. Declaration of interests ADF and PP have a patent pending which is relevant to this work (Method and apparatus for measurement of cardiopulmonary function; European patent no. EP 3122249 A1, US patent no. US20170100043 A1).
Development of the new technique: RB, MT, PP. Writing paper: RB. Contributed to the production of the manuscript: DC, AM, MT, FF, PP, AF. Declaration of interests ADF and PP have a patent pending which is relevant to this work (Method and apparatus for measurement of cardiopulmonary function; European patent no. EP 3122249 A1, US patent no. US20170100043 A1). Funding The UK Department of Health, National Institute of Health Research (II-LA-0214-20005) grant funded the development of the sensing unit used in the technology. A Wellcome Trust (HMRXGK00) grant funded the experimental costs. Appendix A Supplementary data The following is the Supplementary data to this article:Multimedia component 1 Multimedia component 1 Acknowledgements The authors are grateful for the important work from veterinary anaesthetists Christian Dancker and William McFadzean, and for the technical assistance from Hanne McPeak and Kristina Britchford. Appendix A Supplementary data to this article can be found online at https://doi.org/10.1016/j.bja.2019.02.025.
Editor's key points • Xenon has neuroprotective effects, but its impact on long-term outcomes after traumatic brain injury has not been defined. • In a mouse model of traumatic brain injury, short-term post-injury treatment with xenon reduced secondary injury and improved long-term neurological recovery and survival. • This translational study suggests that xenon should be evaluated in human brain injury patients. Traumatic brain injury (TBI) is a complex and heterogeneous disorder representing a significant global healthcare burden.1 In Europe, the annual incidence of TBI has been estimated at 326 per 100 000 population.2 Those patients who survive TBI frequently suffer from debilitating conditions that limit their ability to work or re-integrate to society, and have an increased risk of death. A recent long-term observational study on TBI patients aged 15–54 yr calculated a 7-fold increase in risk of death up to 13 yr after hospital admission.3, 4 In Europe, ∼7.7 million people are living with TBI-related disability.5 It is becoming clear that even a single TBI early in life is a risk factor for developing cognitive dysfunction, Alzheimer's disease (AD), and other neurodegenerative conditions later in life.6
isk of death up to 13 yr after hospital admission.3, 4 In Europe, ∼7.7 million people are living with TBI-related disability.5 It is becoming clear that even a single TBI early in life is a risk factor for developing cognitive dysfunction, Alzheimer's disease (AD), and other neurodegenerative conditions later in life.6 Current clinical treatment for TBI patients is largely supportive, centred on non-specific endpoints such as management of tissue oxygenation, cerebral perfusion pressure and intracranial pressure.7, 8 At present, there are no clinically proven specific neuroprotective drugs for TBI.9, 10, 11 There is a need for neuroprotective treatments that minimise brain damage after TBI and improve long-term outcomes and survival.
uch as management of tissue oxygenation, cerebral perfusion pressure and intracranial pressure.7, 8 At present, there are no clinically proven specific neuroprotective drugs for TBI.9, 10, 11 There is a need for neuroprotective treatments that minimise brain damage after TBI and improve long-term outcomes and survival. Xenon is an anaesthetic noble gas that is neuroprotective in models of ischaemic brain injury. Xenon has recently been shown to be effective in reducing white matter damage in humans after ischaemic brain injury resulting from out-of-hospital cardiac arrest.12 Xenon is protective in in vitro models of brain trauma,13, 14, 15 but much less is known about its effects in brain trauma in vivo. We previously showed in an animal model that administration of xenon after TBI improves short-term (up to 4 days) and long-term outcomes (28 days) in mice.16 The current study is a longer-term follow-up that evaluates the effect of xenon administered to young adult male mice shortly after experimental brain trauma, investigating cognitive function, brain histology, and survival up to 20 months after TBI, which is approaching the end of a healthy animal's life span. We used the rodent controlled cortical impact (CCI) brain trauma model to test the following hypotheses: (1) xenon treatment after TBI improves very long-term cognitive and histological outcomes (at 20 months); and (2) xenon treatment after TBI improves survival at (i) 12 months and (ii) the end of the study (20 months).
used the rodent controlled cortical impact (CCI) brain trauma model to test the following hypotheses: (1) xenon treatment after TBI improves very long-term cognitive and histological outcomes (at 20 months); and (2) xenon treatment after TBI improves survival at (i) 12 months and (ii) the end of the study (20 months). Methods All experiments were approved by the Animal Ethics Committee of the Landestuntersuchungsamt Rheinland-Pfalz (protocol number: G12-1-010). We designed our study to comply with ARRIVE (Animal Research: Reporting of In Vivo Experiments) guidelines.17 Adult male C57BL/6N mice (n=77), aged 2.5 months, mean weight (standard error of the mean [sem]) 23.9 (0.1) g at the time of surgery or perfusion for naïve animals, were obtained from Charles River Laboratory (Sulzfeld, Germany). Animals were housed in filter-top cages in a pathogen-free facility in a 12:12 light–dark cycle (7 am–7 pm light) at 22°C with ad libitum access to food and water. Animals were closely monitored in the postoperative period for at least 4 h, and then early the following day. Long-term survival animals were checked at least once daily throughout the study.
ges in a pathogen-free facility in a 12:12 light–dark cycle (7 am–7 pm light) at 22°C with ad libitum access to food and water. Animals were closely monitored in the postoperative period for at least 4 h, and then early the following day. Long-term survival animals were checked at least once daily throughout the study. Experimental groups, randomisation, and blinding Animals were randomly assigned to CCI primary injury (no treatment) or CCI followed by 75% Xe:25% O2 or CCI followed by 75% N2:25% O2 (control gas) or sham surgery followed by 75% N2:25% O2 (control gas) groups. The experimenter performing the CCI surgery was blinded to treatment. A separate experimenter, blinded to groups and treatment, performed behavioural tests. For cohort 1 (n=22), there were nine animals in each 24 h group (TBI control; TBI xenon) and four animals in the primary injury group. Animals were allowed to survive for 15 min (primary injury group) or 24 h. In cohort 2 (n=50), long-term (20 month) survival experiments had 20 animals each in the TBI xenon and TBI control groups, and 10 animals in the sham-surgery group. For cohort 3 (n=5), a naïve group of five animals aged 2.5 months was included for comparison of hemisphere volume and cerebral white matter with the long-term cohort.
(n=50), long-term (20 month) survival experiments had 20 animals each in the TBI xenon and TBI control groups, and 10 animals in the sham-surgery group. For cohort 3 (n=5), a naïve group of five animals aged 2.5 months was included for comparison of hemisphere volume and cerebral white matter with the long-term cohort. Traumatic brain injury Animals were anaesthetised with 3.5 vol% sevoflurane in an air/oxygen mixture (40% O2:60% N2) supplied via a facemask in spontaneously breathing animals with buprenorphine analgesia (0.1 mg kg−1). Core body temperature was monitored and maintained at 37°C for the duration of the surgery by means of a rectal probe and feedback-controlled heating pad (Hugo Sachs, March-Hugstetten, Germany). Traumatic injury was performed using the CCI model, as described previously.16 Animals were fixed in a stereotactic frame (Kopf Instruments, Tujunga, CA, USA) and a 4×5 mm craniotomy window was created using a saline-cooled, high-speed drill along the coronal and lambdoid sutures and laterally as close as possible to the temporalis muscle insertion. The bone flap was lifted exposing the dura above the right parietal cortex, between the sagittal, lambdoid, and coronal sutures. The tip of a custom-built CCI device (L. Kopacz, Mainz, Germany) was positioned above the intact dura in the centre of the craniotomy window (1 mm from sagittal suture and 1 mm from lambdoid suture). The angle of the impactor, typically 25° from the sagittal plane, was adjusted such that the tip was perpendicular to the dural surface. The impactor tip was flat, with a diameter of 3 mm, impact velocity of 8 m s−1, impact duration of 150 ms, and penetration depth of 1.0 mm. Our CCI impact parameters and the functional and histological outcomes are similar to those classified as a moderate–severe injury.18 After CCI surgery, the craniotomy was closed with the bone flap, sealed with tissue glue (Histoacryl; Braun-Melsungen, Melsungen, Germany), and the skin sutured. Sham-surgery animals underwent identical anaesthesia, temperature control, placement in stereotaxic frame, surgical skin incision to reveal the surface of the skull, which was drilled superficially, but no craniotomy was performed. The duration of the sham surgery and anaesthesia was identical to that of the CCI animals. In order to avoid any confounding effects from the anaesthesia and analgesia we were careful to ensure that the sham group received exactly the same drugs.
skull, which was drilled superficially, but no craniotomy was performed. The duration of the sham surgery and anaesthesia was identical to that of the CCI animals. In order to avoid any confounding effects from the anaesthesia and analgesia we were careful to ensure that the sham group received exactly the same drugs. The choice of anaesthetic and analgesic drugs in animal TBI studies may also have an impact on how secondary injury develops. For our study we chose to combine sevoflurane and buprenorphine. Sevoflurane has been shown to have minimal effects on secondary injury development in TBI models.19 Buprenorphine is a highly effective and safe analgesic in rodents.20, 21 Despite some controversy surrounding opioid effects on TBI outcomes in rodents,22, 23 evidence is limited. Interestingly, buprenorphine has been shown to have no impact on injury development in rodent models of brain ischaemia.24, 25 Overall, the benefits of appropriate intraoperative and postoperative analgesia are well described, and mandatory for ethical and animal welfare. Animals were returned to their individual home cages in a heated incubator (33°C, 35% humidity; IC8000; Draeger, Lübeck, Germany) and allowed to recover for 15 min before treatment, breathing room air.
propriate intraoperative and postoperative analgesia are well described, and mandatory for ethical and animal welfare. Animals were returned to their individual home cages in a heated incubator (33°C, 35% humidity; IC8000; Draeger, Lübeck, Germany) and allowed to recover for 15 min before treatment, breathing room air. Xenon or control gas administration Gas treatments were administered to spontaneously breathing animals in a series of custom-made xenon exposure chambers linked in a closed circuit, for a total duration of 3 h, starting 15 min after CCI injury. Gas concentrations inside the circuit were monitored continuously via a xenon meter (model 439 EX; Nyquist Ltd, Congleton, UK) and an oxygen meter (Oxydig; Draeger). Carbon dioxide was removed from the system using soda lime pellets. Additional volumes of gases were added as necessary to maintain concentrations of 21–25% for oxygen and 70–75% for xenon throughout the 3 h administration period. Gases were circulated at 700 ml min−1 by a small animal ventilator (Inspira ASV; Harvard Apparatus, Holliston, MA, USA). Xenon (BOC HiQ 74.96% xenon: 25.04% oxygen) was obtained from BOC Ltd (Guildford, UK). The exposure chambers were housed inside a heated incubator. After the 3 h treatment with xenon or control gas, animals were returned to the home cage where they breathed room air.
Inspira ASV; Harvard Apparatus, Holliston, MA, USA). Xenon (BOC HiQ 74.96% xenon: 25.04% oxygen) was obtained from BOC Ltd (Guildford, UK). The exposure chambers were housed inside a heated incubator. After the 3 h treatment with xenon or control gas, animals were returned to the home cage where they breathed room air. Quantification of functional outcome Neuroscore Functional outcome before and after CCI injury was determined using a 15-point neurological outcome score evaluating locomotor ability, vestibulomotor function, and general behaviour.16, 26 The neuroscore was performed in real time before CCI surgery and repeated 24 h after injury by an experimenter blinded to the surgical and treatment groups. Because the sham-surgery animals had the same skin incision and sutures as the CCI animals, they were not visually distinguishable.
n, and general behaviour.16, 26 The neuroscore was performed in real time before CCI surgery and repeated 24 h after injury by an experimenter blinded to the surgical and treatment groups. Because the sham-surgery animals had the same skin incision and sutures as the CCI animals, they were not visually distinguishable. Contextual fear conditioning Two weeks and 20 months after the injury, cognitive function was assessed in the same cohort of animals by a contextual fear conditioning test implemented using a multi-conditioning system (TSE Systems GmbH, Bad Homburg, Germany). The test measures hippocampus-dependent contextual place memory. For the 2 week and 20 month tests, entirely different contexts (chambers) were used in order to test memory of novel environments. These tests were carried out by another blinded observer. For context conditioning training, an animal was placed inside a conditioning context Plexiglas chamber (36×20×20 cm) with a removable shock grid made of stainless steel rods (4 mm in diameter, spaced 6 mm apart). After 2 min, animals received a first electrical shock (0.4 mA, 2 s) and 15 s later a second shock with the same characteristics. Animal behaviour was recorded by a video camera to monitor freezing behaviour, defined as the lack of movement (excluding respiratory movements). The analysis of the freezing behaviour was performed using computerised video software (EthovisionXT software; Noldus Information Technology, Wageningen, The Netherlands). The contextual memory test was performed 24 h after context conditioning training. Mice were monitored for freezing for 2 min in the same context chamber that was used for training. The cumulative duration (s) of freezing behaviour during the 2 min of testing was used.
ormation Technology, Wageningen, The Netherlands). The contextual memory test was performed 24 h after context conditioning training. Mice were monitored for freezing for 2 min in the same context chamber that was used for training. The cumulative duration (s) of freezing behaviour during the 2 min of testing was used. Survival analysis The long-term survival cohort was kept for 20 months after CCI surgery to assess survival. Animals were monitored on a daily basis. Deaths that occurred up to the end of the observation protocol were spontaneous, and no animals had to be euthanised.
ormation Technology, Wageningen, The Netherlands). The contextual memory test was performed 24 h after context conditioning training. Mice were monitored for freezing for 2 min in the same context chamber that was used for training. The cumulative duration (s) of freezing behaviour during the 2 min of testing was used. Survival analysis The long-term survival cohort was kept for 20 months after CCI surgery to assess survival. Animals were monitored on a daily basis. Deaths that occurred up to the end of the observation protocol were spontaneous, and no animals had to be euthanised. Histological processing At 15 min and 24 h animals were anaesthetised with sevoflurane and euthanised by cervical dislocation. Brains were carefully removed, frozen on powdered dry ice, and stored at –80°C. At 20 months the surviving cohort of animals were terminally anaesthetised with pentobarbital and transcardially perfused with 20 ml of cold phosphate-buffered saline (PBS; Biochrom GmbH, Berlin, Germany) followed by 30 ml of cold 4% paraformaldehyde (DAC paraformaldehyde; Merck KGaA, Darmstadt, Germany). Brains were carefully removed from the skull and post-fixed in 4% paraformaldehyde (in PBS) overnight at 4°C, then transferred to 30% sucrose in PBS until the brains sank, before being frozen as described above. Naïve animals were terminally anaesthetised and perfused in the same manner. Frozen brains were embedded in Optimal Cutting Temperature mounting medium (Cell Path Ltd, Newton, Powys, UK) and cut in the coronal plane with a cryostat tissue slicer (Microm HM 560 [15 min, 24 h] or CryoStar NX70 [20 month, naïve]; Thermo Scientific, Walldorf, Germany). To quantify lesion volume, for each brain a total of 16–18 sections (10 μm thick) spanning the entire lesion were collected on Superfrost® Plus microscope slides (Thermo Fisher Scientific, Braunschweig, Germany) every 500 μm, starting at Bregma +3.14 mm. For immunohistochemistry on the long-term brains, 20 μm-thick slices were also obtained at a brain point 4.5 mm from the anterior pole of each brain (because of variable atrophy in the aged brains, these slices ranged from Bregma –1.25 and –2.15 mm when mapped to the Allen mouse brain atlas).27, 28
at Bregma +3.14 mm. For immunohistochemistry on the long-term brains, 20 μm-thick slices were also obtained at a brain point 4.5 mm from the anterior pole of each brain (because of variable atrophy in the aged brains, these slices ranged from Bregma –1.25 and –2.15 mm when mapped to the Allen mouse brain atlas).27, 28 Quantification of contusion volume Slices (10 μm) were stained with cresyl violet (Merck Millipore, Darmstadt, Germany) as described.16 Slices were imaged with a digital camera (Scopetek DCM510; Scopetek Opto-Eletric Co. [Hangzhou, China] or Zeiss Axiocam 105 colour; Zeiss [Jena, Germany]) attached to a stereomicroscope (Wild model M8, Heerbrugg, Switzerland or Zeiss Stemi 305, Zeiss). The contusion was evident from a clear difference in the intensity of the cresyl staining. The area of the contusion was measured using image analysis software (Scopephoto 3.1, Scopetek Opto-Eletric Co. or Zen, Zeiss) by an investigator blinded to the experimental groups. Contusion volume was calculated by multiplying contusion areas, A, by the distance between brain sections, d (500 μm), according to the following formula: d2∗(A1+An)+d*(A2+A3+…+An−1) Secondary injury volume at 24 h was calculated by subtracting the mean primary injury contusion volume at 15 min from the total contusion volume measured at 24 h.
Quantification of contusion volume Slices (10 μm) were stained with cresyl violet (Merck Millipore, Darmstadt, Germany) as described.16 Slices were imaged with a digital camera (Scopetek DCM510; Scopetek Opto-Eletric Co. [Hangzhou, China] or Zeiss Axiocam 105 colour; Zeiss [Jena, Germany]) attached to a stereomicroscope (Wild model M8, Heerbrugg, Switzerland or Zeiss Stemi 305, Zeiss). The contusion was evident from a clear difference in the intensity of the cresyl staining. The area of the contusion was measured using image analysis software (Scopephoto 3.1, Scopetek Opto-Eletric Co. or Zen, Zeiss) by an investigator blinded to the experimental groups. Contusion volume was calculated by multiplying contusion areas, A, by the distance between brain sections, d (500 μm), according to the following formula: d2∗(A1+An)+d*(A2+A3+…+An−1) Secondary injury volume at 24 h was calculated by subtracting the mean primary injury contusion volume at 15 min from the total contusion volume measured at 24 h. Quantification of area of myelinated fibres in the corpus callosum Slices (10 μm) were stained overnight with 1% w/v Luxol Fast Blue in 95% isopropanol:5% acetic acid (10%) at 60°C in a humidified water bath. Slices were washed briefly in 95% isopropanol followed by distilled water. Slices were differentiated in 0.5% w/v lithium carbonate for 30 s, followed by 70% isopropanol for 30 s followed by distilled water, then counterstained with cresyl violet. Slices (Bregma –1.94 mm) were imaged with a digital camera (Zeiss Axiocam 105 colour; Zeiss) attached to a stereomicroscope (Zeiss Stemi 305, Zeiss). The area of the contralateral corpus callosum was measured by a blinded investigator from the midline to a point directly above the distal end of the dentate gyrus (DG) granule cell layer29 using the area measurement function in ImageJ (FIJI).30
05 colour; Zeiss) attached to a stereomicroscope (Zeiss Stemi 305, Zeiss). The area of the contralateral corpus callosum was measured by a blinded investigator from the midline to a point directly above the distal end of the dentate gyrus (DG) granule cell layer29 using the area measurement function in ImageJ (FIJI).30 Immunofluorescence staining and analysis Twenty micrometre-thick slices from the perfused brains of the surviving cohort of animals were used for immunofluorescence staining for glial fibrillary associated protein (GFAP, reactive astrocytes), ionised calcium-binding adapter molecule 1 (Iba1, microglia), neuronal nuclei (NeuN, neurones), and 4′,6-diamidino-2-phenylindole (DAPI) (nuclei). GFAP: slices were washed in PBS for 3 min and blocked for 1 h with 5% normal goat serum, 0.5% bovine serum albumin (diluted in PBS–1% Tween). Sections were incubated overnight at 4°C with rat anti-mouse GFAP (1:500 in blocking solution, 13-0300 Invitrogen/Life Technologies, Carlsbad, CA, USA). The following day, sections were washed with PBS–1% Tween and incubated for 2 h at room temperature with Alexa Fluor® 568 goat anti-rat (A-11077 Invitrogen/Life Technologies). Sections were washed in PBS–1% Tween, incubated with DAPI (1:10000 in PBS–1% Tween) for 5 min and mounted using Shandon Immunomount® (Thermofisher, Schwerte, Germany). Iba1 and NeuN: after antigen retrieval in 0.015 M citric acid, pH 6.0, at 75°C for 20 min, slices were allowed to cool, and washed in PBS for 1 min then blocked for 1 h with 5% normal goat serum, 1% bovine serum albumin (diluted in PBS–0.1% TritonX). Sections were incubated overnight at 4°C with rabbit anti-Iba1 (1:500 in blocking solution, 019-19741; Wako, Neuss, Germany) and mouse anti-NeuN (1:500 in blocking solution, MAB377 Millipore, Darmstadt, Germany). The following day, sections were washed with PBS and incubated for 2 h at room temperature with Alexa Fluor® 568 goat anti-rabbit (1:500, A11011 Invitrogen/Life Technologies) and Alexa Fluor® 488 goat anti-mouse (1:500, A11001 Invitrogen/Life Technologies). Sections were washed in PBS–1% Tween, incubated with DAPI (1:10000 in PBS–0.1% TritonX) for 5 min, then washed in PBS and mounted using Shandon Immunomount® (Thermofisher). The stained sections were imaged using an Axio Observer Z1 widefield microscope (Carl Zeiss AG, Oberkochen, Germany) with a 20× (GFAP) or 10× (NeuN; Iba1) objective (20× Zeiss Plan-Apochromat NA 0.80, WD 0.55 mm; 10× Zeiss EC Plan-Neofluar NA 0.30, WD 5.20 mm).
PBS and mounted using Shandon Immunomount® (Thermofisher). The stained sections were imaged using an Axio Observer Z1 widefield microscope (Carl Zeiss AG, Oberkochen, Germany) with a 20× (GFAP) or 10× (NeuN; Iba1) objective (20× Zeiss Plan-Apochromat NA 0.80, WD 0.55 mm; 10× Zeiss EC Plan-Neofluar NA 0.30, WD 5.20 mm). Whole sections were imaged using the multi-photon acquisition function of Zeiss Zen software (excitation and emission wavelengths: DAPI 335–383, 420–470 nm; Iba1 and GFAP 538–562, 570–640 nm; NeuN 450–490, 500–550 nm). Images were analysed by blinded investigators using ImageJ (FIJI). The GFAP positive area was quantified as percentage area (%-area) in the corpus callosum and in defined regions of interests (ROIs) (Fig. 1a) in the hypothalamus, retrosplenial cortex, amygdala, and hippocampal CA1, CA2, CA3, and DG regions, after rolling ball background subtraction and thresholding (equally for all images) to best match the raw image. For quantification of GFAP positive scar bordering the lesion cavity, images were thresholded (equally for all images) to best match the raw image and the area of interest delineated with a polygon selection in the retrosplenial and the somatosensory cortex. The GFAP-positive total stained area (μm2) and the length (μm) of the ROI were used to calculate an average scar thickness (μm) at the lesion cavity using ImageJ (FIJI). GFAP is upregulated when astrocytes become reactive, and is a reliable and sensitive marker of most reactive astrocytes.31 Because we were interested in quantifying reactive astrogliosis we chose to measure the percentage area of GFAP positive immunoreactivity, rather than count the absolute number of astrocytes. The number of Iba1 positive cells was manually counted in the same defined ROIs as above (Fig. 1a). NeuN positive cells were manually counted in the same ROIs as above for hypothalamus, retrosplenial cortex, and amygdala. For quantification of NeuN positive cells in the hippocampus, we used rectangular ROIs, parallel to and centred on the pyramidal cell layers—CA1 300×50 μm, CA2 200×50 μm, CA3 150×300 μm, and DG 2× 300×80 μm boxes—one in each arm. In all of the long-term histology, TBI control and TBI xenon groups were compared with the sham group that was the same age and had been treated identically to the TBI groups but without impact, in order to ensure that any effects are independent of age-related changes in particular cell types or immunoreactivity.Fig.
In all of the long-term histology, TBI control and TBI xenon groups were compared with the sham group that was the same age and had been treated identically to the TBI groups but without impact, in order to ensure that any effects are independent of age-related changes in particular cell types or immunoreactivity.Fig. 1 (a) A typical GFAP stained brain slice at Bregma –1.75 mm from a TBI control animal 20 months after injury. GFAP positive area was measured in the retrosplenial cortex (R), hypothalamus (Hy), amygdala (Am) and the hippocampus (CA1, CA2, CA3, and DG regions). The lesion on the right side of the cortex is clearly visible on the right of the image. For the retrosplenial cortex the dimensions of the region of interest (ROI) was an ellipse of 240 μm (width) × 605 μm (height); for the hypothalamus the ROI was a circle of diameter 600 μm; for the amygdala the ROI was a circle of diameter 750 μm; the ROIs for the hippocampus were drawn manually for each slice (using the Allen mouse brain atlas28 as a reference). Scale bar, 1000 μm. (b) Acute phase injury development. The controlled cortical impact model produces a primary lesion, clearly seen 15 min after trauma (hatched blue bar), with injury developing significantly 24 h after impact (control gas treatment); (n=4, TBI 15 min; n=9, TBI 24 h). (c) Xenon (red bar) significantly reduces secondary injury volume at 24 h compared with control gas (blue bar). Secondary injury is calculated by subtracting primary injury (15 min) from total contusion volume at 24 h (n=9, TBI control; n=9, TBI xenon). (d) Xenon improves short-term neurological outcome 24 h after injury (n=9, TBI control; n=9, TBI xenon). Control gas-treated animals (blue solid bars) received 75% nitrogen:25% oxygen. Xenon-treated animals (red bars) received 75% xenon:25% oxygen. Treatment was started 15 min after the impact and was administered for 3 h. Bars represent mean values and error bars are standard errors. *P<0.05; **P<0.01; ***P<0.001, compared with TBI control at 24 h. GFAP, glial fibrillary associated protein; DG, dentate gyrus; TBI, traumatic brain injury.
ed 75% xenon:25% oxygen. Treatment was started 15 min after the impact and was administered for 3 h. Bars represent mean values and error bars are standard errors. *P<0.05; **P<0.01; ***P<0.001, compared with TBI control at 24 h. GFAP, glial fibrillary associated protein; DG, dentate gyrus; TBI, traumatic brain injury. Fig. 1
ed 75% xenon:25% oxygen. Treatment was started 15 min after the impact and was administered for 3 h. Bars represent mean values and error bars are standard errors. *P<0.05; **P<0.01; ***P<0.001, compared with TBI control at 24 h. GFAP, glial fibrillary associated protein; DG, dentate gyrus; TBI, traumatic brain injury. Fig. 1 Statistics Data were assessed for normality using the Shapiro–Wilk test. We assessed significance of differences in contusion volume, secondary injury volume, and neurological outcome score between TBI control and TBI xenon groups using a two-tailed Student's t-test. For the fear conditioning test and hemisphere volume data, we compared sham, TBI control, and TBI xenon groups using two-way analysis of variance (anova) or one-way anova, respectively, with Bonferroni's post hoc test. Data on corpus callosum area, GFAP positive area, and Iba1 in the corpus callosum were analysed using one-way anova with Bonferroni's post hoc test. In order to test the hypothesis that xenon treatment reduces astrocytic scarring, the GFAP positive scar data were analysed using a one-tailed Student's t-test. Normality tests, anova, and Student's t-tests were implemented using Graphpad Prism Version 7 software (GraphPad Software Inc., La Jolla, CA, USA). Some of the NeuN, Iba1, and GFAP positive distributions in the ROIs in the hypothalamus, retrosplenial cortex, amygdala, and hippocampus were significantly different from a normal distribution, and could not be transformed into a normal distribution. Therefore, the long-term histology of these areas in the TBI control, TBI xenon and sham groups were compared using a Kruskal–Wallis (KW) test with Benjamini–Yekutieli correction for multiple comparisons implemented using the statistical program Stata (Version 15; StataCorp, College Station, TX, USA). As the null statistics for the KW test are known not to follow a χ2 distribution for small numbers especially in the region of the 0.95 and 0.99 quantile, results from the KW test were compared with the exact results for a KW test using a program written in Mathematica (Mathematica 11.3.0.0; Wolfram Research Inc., Champaign, IL, USA).32 For long-term survival data, cumulative event curves were constructed for the survival of animals with the use of the Kaplan–Meier method using GraphPad Prism.
t were compared with the exact results for a KW test using a program written in Mathematica (Mathematica 11.3.0.0; Wolfram Research Inc., Champaign, IL, USA).32 For long-term survival data, cumulative event curves were constructed for the survival of animals with the use of the Kaplan–Meier method using GraphPad Prism. A one-tailed Gehhan–Breslow–Wilcoxon test, with the Bonferroni–Holm correction, was used to assess differences between survival curves for TBI control and TBI xenon groups at 12 and 20 months after injury. The hazard ratio was calculated using the method of Mantel–Haenszel. A P-value of 0.05 or less was taken to indicate a significant difference. Values are quoted as mean (sem) for normally distributed data or median (inter-quartile range) if data are not normally distributed. Sample sizes (n) are indicated in the figure legends.
d ratio was calculated using the method of Mantel–Haenszel. A P-value of 0.05 or less was taken to indicate a significant difference. Values are quoted as mean (sem) for normally distributed data or median (inter-quartile range) if data are not normally distributed. Sample sizes (n) are indicated in the figure legends. Results Xenon treatment reduces secondary injury and improves outcome acutely after TBI The injury parameters we used create a primary injury such that secondary injury develops significantly (P<0.001), increasing by 98% between 15 min and 24 h after trauma (Fig. 1b) in the TBI control group. Xenon treatment, starting 15 min after injury for 3 h, resulted in a significant reduction in secondary injury by 38% (5%) (P<0.05) at 24 h after injury (Fig. 1c). The reduction in secondary injury volume at 24 h translated into significant improvement in neurological outcome, with a 67% (11%) reduction (P<0.01) in neurological impairment score (Fig. 1d). Neurological outcome score was zero in all groups before CCI or sham surgery, and in the sham-surgery group 24 h later (data not shown).
ion in secondary injury volume at 24 h translated into significant improvement in neurological outcome, with a 67% (11%) reduction (P<0.01) in neurological impairment score (Fig. 1d). Neurological outcome score was zero in all groups before CCI or sham surgery, and in the sham-surgery group 24 h later (data not shown). TBI in young adult mice results in late-onset cognitive impairment that is prevented by xenon treatment We investigated development of the cognitive phenotype at different time points after TBI in the same experimental cohort. We therefore investigated cognitive performance at 2 weeks and 20 months after injury, using a different contextual fear conditioning paradigm at each time point to investigate associative memory. At 2 weeks after injury, duration of immobilisation or ‘freezing’ in the pre-trial before fear conditioning was low for all groups (∼6 s), and there was no significant difference between groups (Fig. 2a). At 24 h after the fear conditioning in context trial, all groups exhibited increased ‘freezing’ behaviour (∼12 s) compared with pre-trial (P<0.001), indicating that they had remembered the context. However, at 2 weeks after injury there was no difference in the duration of freezing between groups (Fig. 2a), indicating all groups recalled the context to the same extent. The lack of difference in freezing time between sham and either of the TBI groups indicates that there was no early memory impairment. At the 20 month time point freezing time in the pre-trial was also low (<6 s) for all groups with no significant differences between groups. However, at the 20 month time point, in the context trial there was a significant reduction (P<0.05) in freezing duration in the TBI control group [29 (4) s] compared with the uninjured sham group [47 (3) s], indicating a late-onset memory impairment in the TBI control group (Fig. 2b). Interestingly, at the 20 month time point the xenon-treated TBI group showed no memory impairment as indicated by the fact that freezing duration [41 (5) s] was not significantly different to that of the sham group (Fig. 2b). This suggests that early xenon treatment after TBI prevents late-onset hippocampus-dependent memory impairment.Fig. 2 Xenon treatment shortly after TBI prevents the development of late-onset TBI-related memory deficits in the same cohort.
n [41 (5) s] was not significantly different to that of the sham group (Fig. 2b). This suggests that early xenon treatment after TBI prevents late-onset hippocampus-dependent memory impairment.Fig. 2 Xenon treatment shortly after TBI prevents the development of late-onset TBI-related memory deficits in the same cohort. (a) At 2 weeks after injury, in a contextual fear conditioning test, no differences between groups in freezing behaviour were observed in the context-trial period indicating there were no memory impairments in either TBI group. (b) At 20 months after injury there was a significant memory impairment in the untreated TBI control group compared with the sham group. Xenon treatment prevented the onset of this impairment. No differences between groups were observed in the pre-trial period. During contextual trial, there was a significant reduction in freezing time in the TBI control group, indicating a memory deficit, compared with the uninjured sham group. Freezing time in TBI xenon group was not significantly different to the sham group. TBI control (blue bars) and sham operated animals (white bars) received 75% nitrogen:25% oxygen; TBI xenon animals (red bars) received 75% xenon:25% oxygen. Treatment was started 15 min after the impact or sham procedure and was administered for 3 h. Bars represent mean values and error bars are standard errors (2 weeks: n=10, sham; n=20, TBI control; n=20, TBI xenon; 20 months: n=7, sham; n=9, TBI control; n=13, TBI xenon). *P<0.05 compared with sham. TBI, traumatic brain injury. Fig. 2
(a) At 2 weeks after injury, in a contextual fear conditioning test, no differences between groups in freezing behaviour were observed in the context-trial period indicating there were no memory impairments in either TBI group. (b) At 20 months after injury there was a significant memory impairment in the untreated TBI control group compared with the sham group. Xenon treatment prevented the onset of this impairment. No differences between groups were observed in the pre-trial period. During contextual trial, there was a significant reduction in freezing time in the TBI control group, indicating a memory deficit, compared with the uninjured sham group. Freezing time in TBI xenon group was not significantly different to the sham group. TBI control (blue bars) and sham operated animals (white bars) received 75% nitrogen:25% oxygen; TBI xenon animals (red bars) received 75% xenon:25% oxygen. Treatment was started 15 min after the impact or sham procedure and was administered for 3 h. Bars represent mean values and error bars are standard errors (2 weeks: n=10, sham; n=20, TBI control; n=20, TBI xenon; 20 months: n=7, sham; n=9, TBI control; n=13, TBI xenon). *P<0.05 compared with sham. TBI, traumatic brain injury. Fig. 2 Brain volume and lesion size 20 months after injury Compared with young adult naïve brains the brain hemisphere volumes of all groups were significantly (P<0.0001) smaller (Supplementary Fig. S1). The left and right hemisphere volumes in the sham group were respectively 28.1% (0.5%) and 30.5% (0.2%) smaller compared with the young naïve group. This is most likely caused by brain atrophy as a result of ageing, but we cannot exclude the possibility that anaesthesia and sham surgery are contributory factors. In the TBI control and TBI xenon groups, the contralateral (left) hemispheres were 30.3% (0.4%) and 27.8% (0.2%) smaller than the young naïve group. Traumatic lesions were present at 20 months in both TBI groups, but there was no difference in volume between the TBI control group [17.7 (1.9) mm3, n=8] and the xenon-treated TBI group [17.6 (1.6) mm3, n=11].
s, the contralateral (left) hemispheres were 30.3% (0.4%) and 27.8% (0.2%) smaller than the young naïve group. Traumatic lesions were present at 20 months in both TBI groups, but there was no difference in volume between the TBI control group [17.7 (1.9) mm3, n=8] and the xenon-treated TBI group [17.6 (1.6) mm3, n=11]. Xenon treatment reduces long-term white matter degeneration after TBI In order to determine the effect of injury and xenon treatment on neuronal white matter 20 months after injury, we measured the area of myelinated fibre tracts stained with Luxol Fast Blue in the contralateral corpus callosum (Fig. 3a and b). The area of the corpus callosum in the sham group at 20 months was 0.23 (0.02) mm2, which was not significantly different to that of the naïve group, aged 2.5 months, 0.20 (0.04) mm2 (data not shown), indicating that in the absence of trauma there was no age-related neurodegeneration. However, in the TBI control group we observed a significant (P<0.001) reduction of 67% (13%) in white matter compared with the uninjured sham group. In contrast, the area of the contralateral corpus callosum myelinated fibre tracts in the TBI xenon group was significantly (P<0.05) larger, by 100% (14%), than the TBI control group and was not significantly different to the uninjured sham group. The reduction in white matter in the TBI control group was accompanied by a significant (P<0.05) increase in GFAP positive astrocytes in the corpus callosum compared with the sham group, whereas in the xenon-treated TBI group GFAP positive staining was not significantly different to the sham group (Fig. 3c). There was a significant (P<0.05) increase in Iba1 positive microglia in the corpus callosum in both the TBI control and TBI xenon groups compared with the sham group (Fig. 3d).Fig. 3 Xenon treatment after TBI reduces long-term white matter degeneration in the contralateral corpus callosum 20 months after injury. (a) Typical Luxol Fast Blue stained slices showing myelinated fibres (dark blue) in the contralateral corpus callosum. Scale bar, 200 μm. r, retrosplenial cortex; c, corpus callosum myelinated fibres; so, stratum oriens (CA1); p, pyramidal layer (CA1); sr, stratum radiatum (CA1). (b) There was a significant reduction in the area of myelinated fibres in contralateral corpus callosum in the TBI control group compared with the sham group.
Scale bar, 200 μm. r, retrosplenial cortex; c, corpus callosum myelinated fibres; so, stratum oriens (CA1); p, pyramidal layer (CA1); sr, stratum radiatum (CA1). (b) There was a significant reduction in the area of myelinated fibres in contralateral corpus callosum in the TBI control group compared with the sham group. Xenon treatment after TBI reduced the white matter loss in corpus callosum (corpus callosum area: n=7, sham; n=8, TBI control; n=11, TBI xenon). (c) There was a significant increase in GFAP positive staining in the contralateral corpus callosum in the TBI control group compared with the sham group, consistent with astrogliosis. The TBI xenon group was not significantly different to the sham group (GFAP area: n=7, sham; n=7, TBI control; n=11, TBI xenon). (d) The number of Iba1 positive cells was significantly increased in both TBI control and TBI xenon groups compared with the sham group (Iba1: n=6, sham; n=7, TBI control; n=10, TBI xenon). TBI control animals (blue bars) and sham animals (white bars) received 75% nitrogen:25% oxygen. TBI xenon animals (red bars) received 75% xenon:25% oxygen. Treatment was started 15 min after the impact or sham procedure and was administered for 3 h. The bars indicate mean values and the error bars are sem. *P<0.05, ***P<0.001 compared with sham; #P<0.05 compared with TBI control. GFAP, glial fibrillary associated protein; TBI, traumatic brain injury; sem, standard error of the mean. Fig. 3
Xenon treatment after TBI reduced the white matter loss in corpus callosum (corpus callosum area: n=7, sham; n=8, TBI control; n=11, TBI xenon). (c) There was a significant increase in GFAP positive staining in the contralateral corpus callosum in the TBI control group compared with the sham group, consistent with astrogliosis. The TBI xenon group was not significantly different to the sham group (GFAP area: n=7, sham; n=7, TBI control; n=11, TBI xenon). (d) The number of Iba1 positive cells was significantly increased in both TBI control and TBI xenon groups compared with the sham group (Iba1: n=6, sham; n=7, TBI control; n=10, TBI xenon). TBI control animals (blue bars) and sham animals (white bars) received 75% nitrogen:25% oxygen. TBI xenon animals (red bars) received 75% xenon:25% oxygen. Treatment was started 15 min after the impact or sham procedure and was administered for 3 h. The bars indicate mean values and the error bars are sem. *P<0.05, ***P<0.001 compared with sham; #P<0.05 compared with TBI control. GFAP, glial fibrillary associated protein; TBI, traumatic brain injury; sem, standard error of the mean. Fig. 3 Xenon treatment reduces chronic perilesional astrocytic scarring after TBI We measured the thickness of the astrocytic scar bordering the lesion cavity in the retrosplenial cortex and somatosensory cortex (Fig. 4a and b) at 20 months after injury. The thickness of the astrocytic scar in the xenon-treated TBI group was significantly reduced compared with the TBI control group, by 45% (8%) (P<0.01) and 39% (8%) (P<0.05) in the retrosplenial cortex and somatosensory cortex, respectively (Fig. 4c and d).Fig. 4 Xenon-treatment reduces astrocytic scarring 20 months after injury. Typical images of GFAP positive staining bordering lesion cavity in (a) retrosplenial cortex and (b) somatosensory cortex. Scale bars are 100 μm. The average thickness of the astrocytic scar bordering the lesion cavity was reduced in the TBI xenon group compared with the TBI control group in the (c) retrosplenial cortex and (d) somatosensory cortex. Control gas-treated animals (blue bars) received 75% nitrogen:25% oxygen. Xenon-treated animals (red bars) received 75% xenon:25% oxygen. Treatment was started 15 min after the impact and was administered for 3 h. The bars indicate mean values and the error bars are sem (n=7, TBI control; n=11, TBI xenon). *P<0.05; **P<0.01 compared with TBI control. GFAP, glial fibrillary associated protein; TBI, traumatic brain injury; sem, standard error of the mean.
Treatment was started 15 min after the impact and was administered for 3 h. The bars indicate mean values and the error bars are sem (n=7, TBI control; n=11, TBI xenon). *P<0.05; **P<0.01 compared with TBI control. GFAP, glial fibrillary associated protein; TBI, traumatic brain injury; sem, standard error of the mean. Fig. 4
Treatment was started 15 min after the impact and was administered for 3 h. The bars indicate mean values and the error bars are sem (n=7, TBI control; n=11, TBI xenon). *P<0.05; **P<0.01 compared with TBI control. GFAP, glial fibrillary associated protein; TBI, traumatic brain injury; sem, standard error of the mean. Fig. 4 Xenon reduces the chronic increase in GFAP expression induced by TBI The GFAP stained area at 20 months after TBI was measured in four specific brain regions involved in contextual fear conditioning: the hypothalamus, amygdala, hippocampus and retrosplenial cortex.33, 34, 35, 36 We found significantly increased GFAP stained area in both ipsilateral and contralateral hypothalamus and in the right retrosplenial cortex in the TBI control group compared with the sham group (P<0.05 and P<0.01 for the different regions, respectively). In the TBI xenon group, GFAP positive area was significantly reduced (P<0.05) in the hypothalamus compared with the TBI control, with no significant difference between sham and TBI xenon groups (Fig. 5a and d). In the right retrosplenial cortex GFAP positive area was reduced in the xenon-treated TBI group compared with the TBI control group (Fig. 5d). We did not find significant differences between any of the groups in the other brain regions examined, amygdala or hippocampus (Fig. 5d; Supplementary Fig. S2a). In the sham group we observed differential expression of GFAP in different brain areas, with greatest expression in hippocampal CA1 and least in the retrosplenial cortex. Because we compared the same brain regions of the injured groups and sham group, variability between different brain regions did not impact the results.Fig. 5 Xenon reduces neuroinflammation and neuronal loss 20 months after injury. Typical examples of immunostaining images for sham, TBI control and TBI xenon groups for (a) astrocytes (GFAP positive) in the hypothalamus, (b) microglia (Iba1 positive) in the amygdala, and (c) neurones (NeuN positive) in the CA1 region of the hippocampus. Scale bars in (a) and (c): 25 μm; scale Bar in (b): 30 μm.
y. Typical examples of immunostaining images for sham, TBI control and TBI xenon groups for (a) astrocytes (GFAP positive) in the hypothalamus, (b) microglia (Iba1 positive) in the amygdala, and (c) neurones (NeuN positive) in the CA1 region of the hippocampus. Scale bars in (a) and (c): 25 μm; scale Bar in (b): 30 μm. Quantification of: (d) GFAP positive area (n=7, sham; n=7, TBI control; n=11, TBI xenon, except for CA1: n=6, sham; CA1: n=6, TBI control); (e) number of Iba1 positive cells (n=6, sham; n=8, TBI control; n=11, TBI xenon, for all regions except the following hippocampal regions: CA1: n=5, sham; n=7, TBI control; n=10, TBI xenon; CA2: n=7, TBI control; DG: n=6, sham, n=8, TBI control, n=11, TBI xenon); and (f) number of NeuN positive cells in sham, TBI control and TBI xenon animals in (i) left and right hypothalamus, (ii) left and right amygdala, (iii) left hippocampus CA1 and DG, and (iv) left and right retrosplenial cortex (n=6, sham; n=8, TBI control; n=11, TBI xenon for all regions except the following: right retrosplenial cortex n=10, TBI xenon; right amygdala n=10, TBI xenon; CA1: n=4, sham, CA1: n=7, TBI control, CA1: n=10, TBI xenon, DG: n=5, sham, DG: n=6, TBI control, DG: n=9, TBI xenon). TBI control animals (blue boxes) and sham animals (white boxes) received 75% nitrogen:25% oxygen. TBI xenon animals (red boxes) received 75% xenon:25% oxygen. Treatment was started 15 min after the impact or sham procedure and was administered for 3 h. The boxes show median and inter-quartile range, the whiskers indicate the data range. *P<0.05, **P<0.01 compared with sham #P<0.05 compared with TBI control. GFAP, glial fibrillary associated protein; Iba1, ionised calcium-binding adapter molecule 1 ; NeuN, neuronal nuclei; TBI, traumatic brain injury; DG, dentate gyrus.
e boxes show median and inter-quartile range, the whiskers indicate the data range. *P<0.05, **P<0.01 compared with sham #P<0.05 compared with TBI control. GFAP, glial fibrillary associated protein; Iba1, ionised calcium-binding adapter molecule 1 ; NeuN, neuronal nuclei; TBI, traumatic brain injury; DG, dentate gyrus. Fig. 5 Xenon reduces microglial proliferation in the ipsilateral amygdala after TBI We measured the number of Iba1-positive microglia in the hypothalamus, amygdala, hippocampus, and retrosplenial cortex at 20 months after injury (Fig. 5b and e). We observed a significant (P<0.05) increase in the number of Iba-positive microglial cells in the TBI control group in the right amygdala, whereas the TBI xenon group was not different to the sham group in this region (Fig. 5e). No other significant differences from sham were observed (Fig. 5e, Supplementary Fig. S2b). Xenon reduces neuronal loss in the contralateral CA1 and DG after TBI We measured the number of NeuN-positive neurones in the hypothalamus, amygdala, hippocampus, and retrosplenial cortex (Fig. 5c and f). In the CA1 and DG of the hippocampus there was a significant (P<0.05) decrease in the number of neurones in the TBI control group compared with the sham group, whereas the TBI xenon group was not different to the sham group (Fig. 5f). In the hypothalamus, retrosplenial cortex, amygdala (Fig. 5f), and CA2 and CA3 regions of the hippocampus (Supplementary Fig. 2c), there was no significant difference in the number of neurones in the TBI groups or the sham group.
the sham group, whereas the TBI xenon group was not different to the sham group (Fig. 5f). In the hypothalamus, retrosplenial cortex, amygdala (Fig. 5f), and CA2 and CA3 regions of the hippocampus (Supplementary Fig. 2c), there was no significant difference in the number of neurones in the TBI groups or the sham group. Effect of xenon treatment on survival after TBI Our hypotheses were that xenon treatment improves survival after TBI at 12 and 20 months after injury. All of the deaths that occurred up until the end of the observation period of 20 months were spontaneous. We analysed the data using Kaplan–Meier survival curves (Fig. 6). At both the 6 and 12 month time points, there was 20% mortality in the TBI control group whereas in the sham and TBI xenon groups mortality was 0%. The survival curve for the TBI xenon group was significantly different (P<0.05) to that of the TBI control group at 12 months. Animals in the TBI control group were more likely to die than the TBI xenon group in the first 12 months with a hazard ratio of 8.38 (95% confidence interval [CI], 1.13–61.95). At the end of the observation period of 20 months, the mortality in the TBI control group was 60%, whereas in the sham group it was 30% and in the TBI xenon group it was 45%. At the end of the observation period (20 months), the survival curves for the TBI control and TBI xenon groups were not significantly different (P=0.09); and the hazard ratio was 1.68 (95% CI, 0.70–4.00). The data as a whole show that xenon treatment significantly improves survival in the first 12 months after TBI.Fig. 6 Kaplan–Meier survival curves up to 20 months after injury. The untreated TBI control animals started to die earlier, with 20% mortality at 12 months, whereas in the TBI xenon group and sham groups there was 100% survival up to 12 months. At 12 months after injury the survival curve for the TBI xenon group was significantly different (P<0.05) to the TBI control group. At the end of the observation period mortality was 60% in TBI control, 30% in sham and 45% in TBI xenon groups. At 20 months the survival curve for the xenon-treated group was not significantly different to the TBI control group (P=0.09). TBI control (blue line) and sham operated (black line) animals received 75% nitrogen:25% oxygen; TBI xenon (red line) received 75% xenon:25% oxygen. Treatment was started 15 min after the impact or sham procedure and was administered for 3 h.
ated group was not significantly different to the TBI control group (P=0.09). TBI control (blue line) and sham operated (black line) animals received 75% nitrogen:25% oxygen; TBI xenon (red line) received 75% xenon:25% oxygen. Treatment was started 15 min after the impact or sham procedure and was administered for 3 h. Beginning of observation period: n=10, sham; n=20, TBI control; n=20, TBI xenon; End of observation period: n=7, sham; n=8, TBI control; n=11, TBI xenon. TBI, traumatic brain injury. Fig. 6 Discussion TBI is recognised as a dynamic condition with long-lasting sequelae that evolve in the weeks, months, and years after injury.1 Many of the long-term cognitive and motor impairments experienced by TBI patients result from the developing secondary injury arising after the primary insult. Survivors of TBI have a significantly increased risk of death,4 and it is known that even a single TBI early in life is a risk factor for developing AD, chronic traumatic encephalopathy (CTE), and other cognitive impairments later in life.6 The mechanistic link between TBI, AD, and CTE is not fully understood, but chronic neuroinflammation and neurodegeneration are thought to play key roles.6 Our aim was to investigate the effects of xenon treatment on secondary injury in both the short- and very long-term on behavioural outcomes longitudinally in an ageing cohort, and on survival, after a single TBI.
s not fully understood, but chronic neuroinflammation and neurodegeneration are thought to play key roles.6 Our aim was to investigate the effects of xenon treatment on secondary injury in both the short- and very long-term on behavioural outcomes longitudinally in an ageing cohort, and on survival, after a single TBI. Experimental paradigm The rodent CCI model is a widely used, highly reproducible, pre-clinical model of blunt TBI.37 With a few notable exceptions,29, 38, 39, 40 the majority of studies using this model have investigated outcomes in the time frame of 1–28 days. Our current study is one of the longest to evaluate a neuroprotective treatment in an animal model of TBI as the study duration was very close to the end of a healthy animal's life span. The CCI injury parameters we used result in a clear early secondary injury, evidenced by the lesion volume approximately doubling in the first 24 h. This early secondary injury is accompanied by a moderate neurological impairment at this time point. Xenon treatment after TBI reduces acute secondary injury and improves short-term neurological impairment Our results show that xenon is effective in significantly reducing the developing secondary injury in the first 24 h. We also found a significant improvement in acute neurological outcome at 24 h in the xenon-treated group. These results are consistent with a previous study where we showed that xenon reduced total contusion volume, but we did not examine secondary injury separately.16
Conclusions Chronotropic incompetence was associated with both impaired cardiopulmonary function and elevated NT pro-BNP (indicating subclinical heart failure). However, in contrast to parasympathetic measures, chronotropic incompetence was not linked to myocardial injury. These data suggest that a mechanistic role for sympathetic dysregulation in myocardial injury is unlikely, and adds further support to the hypothesis that cardiac vagal dysfunction is the predominant autonomic influence in determining myocardial injury and perioperative outcome.5, 8, 31, 32, 41, 42 Authors' contributions Hypothesis conception: TEFA, GLA Analysis plan design: TEFA, RMP, BHC, DW, GLA Data analysis: TEFA, GLA Writing paper: TEFA, GLA with input from RMP Revising paper: all authors Declaration of interest The Measurement of Exercise Tolerance before Surgery study funding sources had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; and preparation or approval of the article. RMP holds research grants, and has given lectures and performed consultancy work for Nestlé Health Sciences, B. Braun, Medtronic, GlaxoSmithKline, and Edwards Lifesciences, and is a member of the associate editorial board of the British Journal of Anaesthesia. GLA is a member of the editorial advisory board for Intensive Care Medicine Experimental, is an Editor for the British Journal of Anaesthesia, and has undertaken consultancy work for GlaxoSmithKline. TEFA is a committee member of the Perioperative Exercise Testing and Training Society. MPWG is Vice President of CPX International, serves on the medical advisory board for Sphere Medical Ltd, and is joint editor-in-chief of Perioperative Medicine. There are no other relationships or activities that could appear to have influenced the submitted work.
the developing secondary injury in the first 24 h. We also found a significant improvement in acute neurological outcome at 24 h in the xenon-treated group. These results are consistent with a previous study where we showed that xenon reduced total contusion volume, but we did not examine secondary injury separately.16 Xenon treatment prevents late-onset memory deficits after TBI A single TBI early in life increases the risk of cognitive decline in later life.6, 41 We examined a cognitive function, hippocampus-dependent memory, in the same cohort, at 2 weeks and 20 months after traumatic injury. At the 2 week time point, no memory impairment was evident in either the TBI control or TBI xenon groups. In contrast, at 20 months in the TBI control group there was a significant memory impairment compared with the uninjured sham group. Remarkably, this late-onset cognitive impairment was absent in the TBI xenon group. The contextual fear conditioning paradigm we used is robust and one of the most comprehensively characterised rodent models of learning and memory. This test allowed us to use novel visuospatial contexts for the two time points and is not confounded by motor deficits. Although memory is a cognitive function, it is only one component of global cognition that can be assessed in patients. Nevertheless, our findings are of particular clinical relevance because of the association of TBI with increased risk of AD and other neurodegenerative conditions in which delayed onset memory impairment is observed.6, 42 Our results suggest that xenon treatment may prevent development of these persistent cognitive deficits in TBI patients.
ings are of particular clinical relevance because of the association of TBI with increased risk of AD and other neurodegenerative conditions in which delayed onset memory impairment is observed.6, 42 Our results suggest that xenon treatment may prevent development of these persistent cognitive deficits in TBI patients. Histopathological correlates of late-onset cognitive impairment are reduced by xenon Long-term loss of cerebral white matter after TBI is associated with cognitive impairment in patients. In the TBI control group we observed chronic neurodegeneration of white matter in the contralateral corpus callosum that was associated with reactive astrogliosis and an increase in microglia, similar to findings reported by Pischiutta et al29 and Loane et al39 12 months after CCI in mice. Remarkably xenon treatment reduced loss of white matter and chronic astrogliosis in the corpus callosum. Interestingly, microglia were significantly increased in the corpus callosum of both TBI control and TBI xenon groups, suggesting involvement of toxic A1-type reactive astrocytes43 induced by activated microglia in the chronic neurodegeneration. In addition to reducing reactive astrogliosis in the contralateral corpus callosum, xenon treatment resulted in a significant reduction in the thickness of astrocytic scarring bordering the lesion cavity in the retrosplenial and somatosensory cortex. We suggest that the reduced scarring limits neurodegeneration and promotes axonal repair in the perilesional tissue.
n the contralateral corpus callosum, xenon treatment resulted in a significant reduction in the thickness of astrocytic scarring bordering the lesion cavity in the retrosplenial and somatosensory cortex. We suggest that the reduced scarring limits neurodegeneration and promotes axonal repair in the perilesional tissue. The neuronal circuit involved in contextual fear conditioning involves a number of brain regions, including the hippocampus, amygdala, retrosplenial cortex, and hypothalamus.33, 34, 35, 36 We found neuronal loss in the CA1 and DG regions of the contralateral hippocampus in the TBI control group 20 months after TBI, consistent with the results of another study at 12 months after injury.39 Both of these regions are involved in episodic memory, and this neuronal loss in the TBI control group correlates with the late-onset memory impairment. Remarkably, xenon treatment reduced this long-term neurodegeneration, with the number of neurones in the CA1 region and the DG regions in the xenon-treated TBI group not different to the sham group, correlating with xenon preventing late-onset impairment of hippocampus-dependant memory. We found increased astrocyte reactivity in the hypothalamus and retrosplenial cortex of TBI control animals compared with the sham group. This is consistent with similar findings reported in the thalamus and cortex of mice 1 yr after CCI injury.29 Interestingly, in xenon-treated animals, the GFAP positive area was decreased in the retrosplenial cortex and was not different from the sham group in the hypothalamus. Hypothalamic dysfunction has been found in other rodent models of TBI,44, 45 including evidence of persistent hypothalamic astrocytosis 2 months after CCI.46 However, we did not observe significant changes in astrocyte reactivity in the amygdala or hippocampus. We observed increases in number of microglia in the right amygdala in the TBI control group but not in the xenon-treated group. The amygdala is involved in memory recall involving anxiety and fear responses, and this observation may play a role in xenon preventing long-term cognitive impairment. Our findings suggest that the learning and memory deficits developing late after the injury result from a combination of discrete injury, involving key brain areas such as CA1 and DG regions of the hippocampus, the hypothalamus, retrosplenial cortex, and amygdala, and reduction of connectivity resulting from loss of white matter.
s suggest that the learning and memory deficits developing late after the injury result from a combination of discrete injury, involving key brain areas such as CA1 and DG regions of the hippocampus, the hypothalamus, retrosplenial cortex, and amygdala, and reduction of connectivity resulting from loss of white matter. Overall, xenon treatment preserved neuronal white matter, and reduced neuronal loss in the hippocampal CA1 and DG regions, and this may explain how xenon treatment prevents the late-onset cognitive impairment. Xenon reduced chronic GFAP immunoreactivity, a well-known marker of astrocyte activation and neuroinflammation31 in the hypothalamus and retrosplenial cortex. Given the recent evidence of the role astrocytes play in regulating neuronal activity, it is worth noting that even in the absence of cytotoxic effects, changes in astrocyte number and reactivity may result in neuronal dysfunction. Hypothalamic dysfunction associated with TBI is not yet fully understood; however, this is known to contribute to cognitive deficits in TBI patients,47, 48 and disruption of sleep and circadian rhythm are observed in many TBI patients.
e Perioperative Exercise Testing and Training Society. MPWG is Vice President of CPX International, serves on the medical advisory board for Sphere Medical Ltd, and is joint editor-in-chief of Perioperative Medicine. There are no other relationships or activities that could appear to have influenced the submitted work. Funding Medical Research Council and British Journal of Anaesthesia clinical research training fellowship (MR/M017974/1 to TEFA); UK National Institute for Health Research Professorship (to RMP); British Journal of Anaesthesia / Royal College of Anaesthetists basic science career development award, British Oxygen Company research chair grant in anaesthesia from the Royal College of Anaesthetists, and British Heart Foundation Programme Grant (RG/14/4/30736 to GLA); merit awards from the Department of Anesthesia at the University of Toronto (to BHC and DW); New Investigator Award from the Canadian Institutes of Health Research (to DW) ; --> NIHR Southampton Biomedical Research Centre (to MPWG) Appendix A Supplementary data The following is the Supplementary data to this article:Multimedia component 1 Multimedia component 1 Appendix A Supplementary data to this article can be found online at https://doi.org/10.1016/j.bja.2019.03.022.
ects, changes in astrocyte number and reactivity may result in neuronal dysfunction. Hypothalamic dysfunction associated with TBI is not yet fully understood; however, this is known to contribute to cognitive deficits in TBI patients,47, 48 and disruption of sleep and circadian rhythm are observed in many TBI patients. Xenon treatment improves survival after injury TBI is associated with an increased risk of death in those who survive the initial injury. A longitudinal study following a large cohort of patients for up to 13 yr after TBI found that by the end of the study mortality was 40%.3, 4 When compared with a control group, this represents an increase in mortality of 2.8-fold overall (all ages) and a 7-fold increase for those sustaining a TBI at 15–54 yr.34 We aimed to determine whether xenon treatment would increase survival in a cohort of mice that had sustained a single TBI early in life. Our cohort was aged 2.5 months (young adult) when they were subjected to CCI, and our objective was to study them until old age. Given the reported life span of male C57BL/6 mice of 23–30 months,49, 50 we studied them for 20 months after injury (age 23 months), and determined whether there was an effect of treatment at 12 months and at the end of the study. We observed a greater number of deaths in the TBI control group compared with the TBI xenon group or sham group at all time points. Early deaths (6 months after injury) were observed in the TBI control group but not in any other group, which we attribute to an increased risk of death resulting from the TBI. Xenon treatment improved survival significantly in the first 12 months after injury. Interestingly, no animals in the TBI xenon group or the sham group died in the first 12 months. At 12 months the hazard ratio of the TBI control group compared with the TBI xenon group was 8.34, similar to the 7-fold increased risk of death in TBI patients suffering a TBI early in life. At 20 months there were more survivors in the xenon-treated group, with a hazard ratio of 1.68 (P=0.09). The lack of a significant difference between TBI control and TBI xenon groups at the end of the study likely reflects that this time point is very close to the normal expected life span. Nevertheless, these results show that xenon improves early survival after TBI, and further preclinical studies to investigate this are merited.
f a significant difference between TBI control and TBI xenon groups at the end of the study likely reflects that this time point is very close to the normal expected life span. Nevertheless, these results show that xenon improves early survival after TBI, and further preclinical studies to investigate this are merited. Clinical relevance Our aim was to evaluate xenon's potential as a treatment for TBI in clinically relevant settings, with xenon treatment 15 min after the trauma, based on a scenario where a TBI patient might receive medical attention at the scene of injury within 15–20 min. Although there may be limited circumstances when xenon could be administered within this time frame, it is possible that xenon could be given by first responders. The most likely treatment scenario would be on arrival at hospital, and xenon is effective even when treatment start time is delayed up to 3 h after injury.16 The 3 h duration of xenon treatment used is relatively short, and it is plausible that extending treatment time might result in greater neuroprotection. The mechanisms by which xenon reduces secondary injury may be pleotropic, but preventing or reducing glutamate excitotoxicty51, 52, 53, 54, 55, 56 is likely to play an important role. Glutamate levels in TBI patients peak 24–72 h after injury.57 It is possible that xenon treatment given during the entire period that glutamate is elevated would be more efficacious. Current treatment for TBI patients is largely supportive, with no clinically proven treatments specifically targeting neuronal loss and neurodegeneration.11 Xenon is approved for clinical use as a general anaesthetic and has recently undergone clinical trials as a neuroprotectant for ischaemic brain injury in neonates and adult cardiac arrest patients.12, 58, 59 We show here that xenon treatment after TBI in mice reduces chronic neurodegeneration of cerebral white matter, reduces loss of hippocampal neurones, prevents late-onset TBI-related cognitive impairment, and improves long-term survival. These long-term improvements in clinically relevant outcomes and survival provide further support to the idea that xenon could be used as an early neuroprotective treatment in TBI patients.
tter, reduces loss of hippocampal neurones, prevents late-onset TBI-related cognitive impairment, and improves long-term survival. These long-term improvements in clinically relevant outcomes and survival provide further support to the idea that xenon could be used as an early neuroprotective treatment in TBI patients. Authors' contributions Study design/planning: RCP, KR, SCT, RD. Study conduct: RCP, TH, KR, RD. Data analysis: all authors. Drafting of the paper: RCP, SCT, RD. Revision of the paper: all authors. Declaration of interest NPF is a named inventor on patents covering the use of xenon as a neuroprotectant and a shareholder in Neuroprotexeon Ltd., a spin-out company that seeks to commercialise the use of xenon as a neuroprotectant. The other authors declare that they have no conflicts of interest.
Revision of the paper: all authors. Declaration of interest NPF is a named inventor on patents covering the use of xenon as a neuroprotectant and a shareholder in Neuroprotexeon Ltd., a spin-out company that seeks to commercialise the use of xenon as a neuroprotectant. The other authors declare that they have no conflicts of interest. Funding The Medical Research Council, London, UK (MR/N0277361/1; MC_PC_13064); European Society for Anaesthesiology, Brussels, Belgium; The Gas Safety Trust, London, UK; Royal British Legion Centre for Blast Injury Studies, Imperial College London, UK; Royal Centre for Defence Medicine, Birmingham, UK; The Association of Anaesthetists of Great Britain and Ireland, London, UK; National Institute of Academic Anaesthesia, London, UK. We acknowledge the financial support of the Royal British Legion. Rita Campos-Pires was the recipient of a doctoral training award from the Fundação para a Ciência e a Tecnologia, Lisbon, Portugal (SFRH/BD/78886/2011), and a scholarship from the Deutscher Akademischer Austauschdienst (DAAD), German Academic Exchange Service, Bonn, Germany (A/13/72143). Tobias Hirnet is funded by the German Federal Ministry of Education and Research, Berlin, Germany (grant BMBF 01EO1003). Flavia Valeo is the recipient of an Imperial College London President's PhD scholarship, London, UK.
Akademischer Austauschdienst (DAAD), German Academic Exchange Service, Bonn, Germany (A/13/72143). Tobias Hirnet is funded by the German Federal Ministry of Education and Research, Berlin, Germany (grant BMBF 01EO1003). Flavia Valeo is the recipient of an Imperial College London President's PhD scholarship, London, UK. Appendix A Supplementary data The following are the supplementary data related to this article:figs1 Healthy brain hemisphere volumes in naïve animals aged 2.5 months (grey bars) and in sham (white bars), TBI control (blue bars) and TBI xenon groups (red bars) aged 22.5 months. TBI xenon animals received 75% xenon:25% oxygen. Sham and TBI control animals received 75% nitrogen:25% oxygen. Treatment was started 15 minutes after the impact or sham procedure and was administered for 3 hours. Bars represent mean values and error bars are standard errors (n=5, naïve; n=7, sham; n=8, TBI control; n=11, TBI xenon). ∗∗∗∗P<0.0001.
Editor's key points • There is considerable interest in electroencephalographic signatures for identifying the cortical effects of various anaesthetics. • The EEG parameters correlating with pharmacokinetic–pharmacodynamic modelling of the hypnotic actions of ketamine were studied in human volunteers. • Changes in slow wave activity after a hypnotic dose of ketamine were well fitted by a standard sigmoid model • Onset, but not offset, of slow wave activity was consistently associated with loss of behavioural responsiveness.
% xenon:25% oxygen. Sham and TBI control animals received 75% nitrogen:25% oxygen. Treatment was started 15 minutes after the impact or sham procedure and was administered for 3 hours. Bars represent mean values and error bars are standard errors (n=5, naïve; n=7, sham; n=8, TBI control; n=11, TBI xenon). ∗∗∗∗P<0.0001. figs1figs2 Quantification of reactive astrocytes (GFAP positive area) (A), number of microglia (Iba1 positive) (B) and number of neurons (NeuN positive), (C) in the hippocampal CA2 and CA3 areas, 20 months after injury. TBI control animals (blue bars) and sham animals (white bars) received 75% nitrogen:25% oxygen TBI xenon animals (red bars) received 75% xenon:25% oxygen. Treatment was started 15 minutes after the impact or sham procedure and was administered for 3 hours. The boxes show median and interquartile range, the whiskers indicate the data range (GFAP: n=7, sham; n=7, TBI control; n=11, TBI xenon; except CA2: n=10, TBI xenon. Iba1: n=6, sham; n=8, TBI control; n=11, TBI xenon, for all regions except CA2: n=7, TBI control; NeuN: n=6, sham; n=8, TBI control; n=11, TBI xenon, for all regions except CA2: n=5, sham; CA2: n=6, TBI control; CA2: n=10, TBI xenon). figs2
figs1figs2 Quantification of reactive astrocytes (GFAP positive area) (A), number of microglia (Iba1 positive) (B) and number of neurons (NeuN positive), (C) in the hippocampal CA2 and CA3 areas, 20 months after injury. TBI control animals (blue bars) and sham animals (white bars) received 75% nitrogen:25% oxygen TBI xenon animals (red bars) received 75% xenon:25% oxygen. Treatment was started 15 minutes after the impact or sham procedure and was administered for 3 hours. The boxes show median and interquartile range, the whiskers indicate the data range (GFAP: n=7, sham; n=7, TBI control; n=11, TBI xenon; except CA2: n=10, TBI xenon. Iba1: n=6, sham; n=8, TBI control; n=11, TBI xenon, for all regions except CA2: n=7, TBI control; NeuN: n=6, sham; n=8, TBI control; n=11, TBI xenon, for all regions except CA2: n=5, sham; CA2: n=6, TBI control; CA2: n=10, TBI xenon). figs2 Acknowledgements We thank Frida Kornes, laboratory manager, and Christina Gölz, Department of Anaesthesiology, Medical Centre of Johannes Gutenberg University, Mainz, Germany, for advice and technical support, Cristine Ostwald, Department of Neurosurgery, Medical Centre of Johannes Gutenberg University, Mainz, Germany, for assistance with Luxol Fast Blue staining, Michael Schaefer, Department of Anaesthesiology, Medical Centre of Johannes Gutenberg University, Mainz, Germany, for advice on immunostaining, Rushdie Abuhamdah, Department of Life Sciences, Imperial College London, UK, for assistance with data blinding, and Stephen Rothery, National Heart and Lung Institute, Imperial College London, UK, for advice on widefield microscopy and image analysis.
Gutenberg University, Mainz, Germany, for advice on immunostaining, Rushdie Abuhamdah, Department of Life Sciences, Imperial College London, UK, for assistance with data blinding, and Stephen Rothery, National Heart and Lung Institute, Imperial College London, UK, for advice on widefield microscopy and image analysis. This article is accompanied by an editorial: Xenon for traumatic brain injury: a noble step forward and a wet blanket by Terrando & Warner, Br J Anaesth 2019:123:9–11, doi: 10.1016/j.bja.2019.04.004 Appendix A Supplementary data related to this article can be found at https://doi.org/10.1016/j.bja.2019.02.032.
Editor's key points • Many moderate- to high-risk surgical patients have cardiac autonomic dysfunction, which can manifest as either parasympathetic or sympathetic dysfunction, or as a combined disorder. • The inability to increase HR sufficiently quickly during exercise (chronotropic incompetence) reflects sympathetic dysfunction, and failure to decrease HR sufficiently quickly after termination of exercise (impaired HR recovery) points to parasympathetic dysfunction. • In this multicentre cohort study, including 1325 moderate- to high-risk noncardiac surgery patients, all of whom underwent cardiopulmonary exercise testing, approximately 30% had preoperative chronotropic incompetence and 40% had impaired HR recovery. • Chronotropic incompetence (sympathetic dysfunction) was related to preoperative biomarker indicators of heart failure, but, when adjusting for known confounders, was not significantly associated with 3-day postoperative myocardial injury or 1-yr mortality. • In contrast, in a multivariable analysis, preoperative impaired HR recovery (parasympathetic dysfunction), in isolation or in combination with chronotropic incompetence, was significantly associated with 3-day postoperative myocardial injury.
• Chronotropic incompetence (sympathetic dysfunction) was related to preoperative biomarker indicators of heart failure, but, when adjusting for known confounders, was not significantly associated with 3-day postoperative myocardial injury or 1-yr mortality. • In contrast, in a multivariable analysis, preoperative impaired HR recovery (parasympathetic dysfunction), in isolation or in combination with chronotropic incompetence, was significantly associated with 3-day postoperative myocardial injury. Around 30% of patients undergoing noncardiac surgery sustain asymptomatic myocardial injury, which is strongly associated with mortality during hospital admission.1, 2, 3 Myocardial injury is more likely to occur in patients with preoperative cardiac vagal (parasympathetic) dysfunction, identified by impaired HR recovery (i.e. decrease) after exercise.4 Cardiac vagal autonomic impairment is a common feature in deconditioned surgical patients,5, 6, 7 in whom preoperative cardiopulmonary exercise testing (CPET) also reveals physiological features of cardiac failure, including lower peak oxygen consumption and higher resting HR.8
recovery (i.e. decrease) after exercise.4 Cardiac vagal autonomic impairment is a common feature in deconditioned surgical patients,5, 6, 7 in whom preoperative cardiopulmonary exercise testing (CPET) also reveals physiological features of cardiac failure, including lower peak oxygen consumption and higher resting HR.8 Whilst cardiac vagal activity reduces HR after exercise, maximal aerobic exercise is facilitated by increases in HR, principally driven by the sympathetic nervous system.9, 10, 11, 12, 13, 14 The impaired ability to increase HR,4 which is required for increased activity or demand, is broadly defined as chronotropic incompetence.12 In cardiac failure, high circulating levels of catecholamines result in decreased β-adrenoceptor density and desensitisation, which limit β-agonist-mediated contractility.15 Consequently, chronotropic incompetence is a robust predictor of mortality in patients with overt, clinically diagnosed cardiac failure.12 Here, we hypothesised that chronotropic incompetence, identified during preoperative CPET, was associated with myocardial injury within 3 days after noncardiac surgery. In order to assess whether sympathetic or parasympathetic dysfunction was more strongly associated with myocardial injury, we also evaluated the relationships between impaired (i) HR increase with exercise (i.e. chronotropic incompetence), and (ii) HR recovery from exercise and myocardial injury.
er noncardiac surgery. In order to assess whether sympathetic or parasympathetic dysfunction was more strongly associated with myocardial injury, we also evaluated the relationships between impaired (i) HR increase with exercise (i.e. chronotropic incompetence), and (ii) HR recovery from exercise and myocardial injury. Methods Study design and setting This was a predefined secondary analysis of the Measurement of Exercise Tolerance before Surgery (METS) study, an international, prospective, observational cohort study of preoperative assessment before noncardiac surgery at 25 hospitals in the UK, Canada, New Zealand, and Australia. The study protocol and the main study results were published previously.2, 16 Research ethics committees reviewed the study, and it was conducted in accordance with the principles of the Declaration of Helsinki and the Research Governance Framework.
surgery at 25 hospitals in the UK, Canada, New Zealand, and Australia. The study protocol and the main study results were published previously.2, 16 Research ethics committees reviewed the study, and it was conducted in accordance with the principles of the Declaration of Helsinki and the Research Governance Framework. Participants Participants were aged 40 yr or older, undergoing elective noncardiac surgery under general anaesthesia or regional anaesthesia with a planned overnight stay in a hospital, and with at least one of the following perioperative risk factors: intermediate- or high-risk surgery, coronary artery disease, heart failure, cerebrovascular disease, diabetes mellitus, preoperative renal insufficiency, peripheral arterial disease, hypertension, a history of tobacco smoking within the previous year, or be aged 70 yr or older. The exclusion criteria were planned procedure using only endovascular technique, use of CPET for risk stratification as part of routine care, insufficient time for CPET before surgery, presence of an implantable cardioverter defibrillator, known or suspected pregnancy, previous enrolment in the study, severe hypertension (systolic pressure >180 mm Hg or diastolic pressure >100 mm Hg), active cardiac conditions, or other contraindications precluding CPET.16, 17 The participants gave written informed consent to take part before surgery.
rter defibrillator, known or suspected pregnancy, previous enrolment in the study, severe hypertension (systolic pressure >180 mm Hg or diastolic pressure >100 mm Hg), active cardiac conditions, or other contraindications precluding CPET.16, 17 The participants gave written informed consent to take part before surgery. Study conduct and data collection Researchers collected data directly from participants and their medical record. A detailed and standardised data set was collected before surgery, during the hospital stay, and after surgery. One year after surgery, the participants were contacted by telephone and underwent a short interview. Each participant underwent CPET and had blood sampled for N-terminal pro-B-type natriuretic peptide (NT pro-BNP) before surgery, and routine blood sampling for cardiac troponin on the 1st, 2nd, and 3rd days after surgery.
year after surgery, the participants were contacted by telephone and underwent a short interview. Each participant underwent CPET and had blood sampled for N-terminal pro-B-type natriuretic peptide (NT pro-BNP) before surgery, and routine blood sampling for cardiac troponin on the 1st, 2nd, and 3rd days after surgery. Cardiopulmonary exercise testing The participants underwent preoperative symptom-limited CPET using a standardised incremental ramp protocol with electromagnetically braked cycle ergometers.18 The test protocol consisted of spirometry in the seated position, followed by 3 min of rest sitting on the ergometer, followed by 3 min of unloaded pedalling, followed by pedalling with progressively increasing workload. Once the participants reached their peak performance, the exercise test was stopped, the workload reduced to 20 W, and the participants continued to pedal for 5 min in order to cool down. The participants were encouraged to pedal at a steady rate of 60 rev min−1. Work rates increased by 10 W min−1 in untrained participants, and by 20–30 W min−1 in trained participants or those undertaking regular physical activity according to a specific algorithm. Cardiopulmonary function was monitored continuously via electrocardiogram; pulse oximetry; and breath-by-breath measurement of minute ventilation, carbon dioxide production, and oxygen consumption. Non-invasive blood pressure was monitored every 3 min. The investigators at each site interpreted each CPET and collected a standardised data set. Peak oxygen consumption was calculated as the mean oxygen consumption during the final 20 s of incremental exercise.19 The anaerobic threshold was identified using the modified V-slope method, followed by the ventilatory equivalent and excess carbon dioxide methods.20 Clinicians at each site were blinded to the results of CPET, except where there was a safety concern according to predefined criteria.16
l 20 s of incremental exercise.19 The anaerobic threshold was identified using the modified V-slope method, followed by the ventilatory equivalent and excess carbon dioxide methods.20 Clinicians at each site were blinded to the results of CPET, except where there was a safety concern according to predefined criteria.16 Exposures The exposure of interest was chronotropic incompetence, defined as chronotropic index (CI) <0.6 using the method described by Dobre and colleagues.21 This threshold is associated with mortality in patients with severe heart failure.21 CI is the ratio of measured increase in HR during exercise to the age-predicted maximal increase in HR.12 HR was measured at rest and at peak oxygen consumption during CPET to give the measured increase in HR. The most widely accepted method for calculating age-predicted maximal HR is 220–age, as described by Astrand.22, 24 The CI for the main analysis was calculated using the formula: (1) CI=(peak HR–resting HR)/(age-predicted maximal HR–resting HR)
at peak oxygen consumption during CPET to give the measured increase in HR. The most widely accepted method for calculating age-predicted maximal HR is 220–age, as described by Astrand.22, 24 The CI for the main analysis was calculated using the formula: (1) CI=(peak HR–resting HR)/(age-predicted maximal HR–resting HR) However, as various population-dependent thresholds have been derived,23 it has been suggested to use a CI equation generated in a population most closely matching the population of interest. The equation suggested by Tanaka and colleagues23 is recommended for apparently healthy persons, whilst other equations are recommended for those with known or suspected cardiovascular disease. In this study, we primarily used the Astrand24 method, and supplemented this with two post hoc sensitivity analyses: firstly, the calculated CI using the Tanaka and colleagues23 method, and second, using the CI (Astrand24) as a continuous variable.
e recommended for those with known or suspected cardiovascular disease. In this study, we primarily used the Astrand24 method, and supplemented this with two post hoc sensitivity analyses: firstly, the calculated CI using the Tanaka and colleagues23 method, and second, using the CI (Astrand24) as a continuous variable. Primary outcome The primary outcome measure was myocardial injury, defined as blood troponin T or I concentration greater than the limit of the reference range (99th centile) for each assay, within 72 h after surgery. Troponin assays differed between participating hospitals, and are listed in Supplementary Table S1. The secondary outcome was all-cause mortality at 1 yr after surgery. Predefined explanatory variables were preoperative NT pro-BNP >300 pg ml−1, a threshold used to predict postoperative cardiovascular events in surgical patients25 and heart failure in community cohorts,26 and preoperative CPET-derived markers of subclinical heart failure (ventilatory efficiency slope [minute ventilation/carbon dioxide production] ≥34, peak oxygen consumption ≤14 ml kg−1 min−1, and HR recovery ≤6 beats min−1 decrease at 1 min after the end of exercise).8
tients25 and heart failure in community cohorts,26 and preoperative CPET-derived markers of subclinical heart failure (ventilatory efficiency slope [minute ventilation/carbon dioxide production] ≥34, peak oxygen consumption ≤14 ml kg−1 min−1, and HR recovery ≤6 beats min−1 decrease at 1 min after the end of exercise).8 Statistical analysis We used STATA version 14 (StataCorp LP, College Station, TX, USA) to analyse the data. We excluded the small number of participants without a record of the exposure or outcome. We ranked the sample by CI and dichotomised it according to a threshold of <0.6, to define groups with and without chronotropic incompetence. We presented the baseline characteristics for the whole cohort and stratified by chronotropic incompetence. Firstly, we used univariable logistic regression analysis to test for the association between chronotropic incompetence and myocardial injury. Secondly, we constructed multivariable logistic regression models, adjusted for co-variates that are known to be associated with perioperative myocardial injury and routinely used for preoperative risk assessment: age >70 yr, male sex, preoperative renal insufficiency, peripheral vascular disease, existing diagnosis of heart failure, coronary artery disease, hypertension, diabetes mellitus, chronic obstructive pulmonary disease, cerebrovascular disease, high-risk surgery, and pre-existing atrial fibrillation.27, 28, 29, 30, 31, 32 We used backwards stepwise selection to identify variables for inclusion in the final model, with a Type 1 error threshold of <0.1. Missing data were handled by list-wise deletion. The results of logistic regression analyses were presented as odds ratios (ORs) with 95% confidence intervals. Normally distributed data were expressed as mean (standard deviation), and non-normally distributed data were expressed as median (inter-quartile range). Binary data were expressed as percentages. The threshold for statistical significance was P<0.05.
s were presented as odds ratios (ORs) with 95% confidence intervals. Normally distributed data were expressed as mean (standard deviation), and non-normally distributed data were expressed as median (inter-quartile range). Binary data were expressed as percentages. The threshold for statistical significance was P<0.05. Secondary analyses We repeated the primary analysis using mortality within 1 yr after surgery, a binary categorical variable, as the outcome measure. We previously described a relationship between preoperative resting HR and subclinical heart failure.8 To explore whether chronotropic incompetence is associated with a phenotype of heart failure in this cohort, we repeated the primary analysis using the following outcome measures, which are biomarkers known to be predictive of poor clinical outcome in overt heart failure: NT pro-BNP >300 pmol L−1, peak oxygen consumption ≤14 ml kg−1 min−1, ventilatory efficiency slope (minute ventilation/carbon dioxide production) at the anaerobic threshold ≥34, and HR recovery ≤6 beats min−1 decrease.8 We have demonstrated previously that parasympathetic autonomic dysfunction is associated with myocardial injury.4 In order to draw direct comparisons between sympathetic and parasympathetic dysfunction, we examined the prevalence of physiological markers of impaired sympathetic and parasympathetic functions using chronotropic incompetence and HR recovery, respectively. We used a widely accepted definition of parasympathetic dysfunction, HR recovery <12 beats min−1 decrease during the first minute after the end of exercise.9
mined the prevalence of physiological markers of impaired sympathetic and parasympathetic functions using chronotropic incompetence and HR recovery, respectively. We used a widely accepted definition of parasympathetic dysfunction, HR recovery <12 beats min−1 decrease during the first minute after the end of exercise.9 Sensitivity analyses Resting HR and the HR response to exercise can be influenced by medications, such as beta blockers and rate-limiting calcium channel antagonists, which may influence the results of our analysis. We addressed this in three ways. Firstly, we repeated the primary analysis, including beta blockers and diltiazem/verapamil as co-variates in the multivariable model. Secondly, we repeated the primary analysis, excluding patients receiving beta blockers and diltiazem/verapamil. Thirdly, we examined whether the use of beta blockade/calcium channel blocker altered the participants' ability to exceed a respiratory exchange ratio (RER) >1.05, as RER <1.05 indicates sub-maximal effort, or that the test was terminated prematurely.12
xcluding patients receiving beta blockers and diltiazem/verapamil. Thirdly, we examined whether the use of beta blockade/calcium channel blocker altered the participants' ability to exceed a respiratory exchange ratio (RER) >1.05, as RER <1.05 indicates sub-maximal effort, or that the test was terminated prematurely.12 Our main analysis used the Astrand24 method. We also performed a post hoc sensitivity analysis, which repeated the primary analysis using age-predicted maximal HR calculated by the Tanaka and colleagues23 method, as we suspect that a substantial number of surgical patients have subclinical cardiac failure.8 We primarily defined chronotropic incompetence as CI <0.6, as described in studies of patients with heart failure. However, studies in other populations have defined chronotropic incompetence as CI <0.8.33 Therefore, we repeated the primary analysis using CI <0.8 as the exposure, and examined the CI as a continuous variable. Sample size estimation As this was a planned secondary analysis of prospectively collected data, the sample size was determined based on the comparisons being made in the principal analysis, which has been published previously.2 We estimated that chronotropic incompetence may be present in up to ∼30% participants. Overall, 12.6% of participants in METS sustained perioperative myocardial injury. If participants with chronotropic incompetence had a higher incidence of ∼16%, at least 1305 participants' data would be required to detect a clinically significant difference (α=0.05; 1-β=80%).
may be present in up to ∼30% participants. Overall, 12.6% of participants in METS sustained perioperative myocardial injury. If participants with chronotropic incompetence had a higher incidence of ∼16%, at least 1305 participants' data would be required to detect a clinically significant difference (α=0.05; 1-β=80%). Results Patients (1741) were recruited into the METS study between March 1, 2013 and March 25, 2016. After predefined exclusions of patients, we analysed data obtained from 1325 participants (Fig. 1). CI <0.6 was present in 396/1325 (29.9%) study participants, of whom 816/1325 (61.7%) were male (Table 1).Fig 1 Patient flow diagram showing the number of cases included in the analysis. CPET, cardiopulmonary exercise testing. Fig 1Table 1 Baseline patient characteristics. Descriptive data stratified by preoperative chronotropic incompetence (defined as chronotropic index [CI] <0.6). Data are presented as n (%) or mean (standard deviations; sd). Continuous data are reported to one decimal place, and categorical data are rounded to the nearest whole number.
seline patient characteristics. Descriptive data stratified by preoperative chronotropic incompetence (defined as chronotropic index [CI] <0.6). Data are presented as n (%) or mean (standard deviations; sd). Continuous data are reported to one decimal place, and categorical data are rounded to the nearest whole number. Table 1 Whole cohort CI <0.6 CI ≥0.6 Number of cases, n 1324 396 928 Mean age (sd) 64.2 (10.4) 64.8 (10.5) 64.0 (10.3) Age ≥70 yr (%) 447 (33.8) 149 (37.6) 298 (32.1) Male sex (%) 817 (61.7) 236 (59.6) 581 (62.6) Co-morbid disorder (%) Atrial fibrillation 50 (3.8) 23 (5.8) 27 (2.9) Diabetes mellitus 243 (18.4) 90 (22.7) 153 (16.5) Hypertension 725 (54.8) 238 (60.1) 487 (52.5) Diagnosis of congestive cardiac failure 16 (1.2) 10 (2.5) 6 (0.7) Coronary artery disease 153 (11.6) 72 (18.2) 81 (8.7) Peripheral vascular disease 39 (3.0) 18 (4.6) 21 (2.3) Previous stroke or transient ischaemic attack 52 (3.9) 25 (6.3) 27 (2.9) Chronic obstructive pulmonary disease 163 (12.3) 74 (18.7) 89 (9.6) Preoperative estimated glomerular filtration rate <60 ml min−1 (1.73 m2)−1 108 (8.2) 45 (11.4) 63 (6.8) Surgical procedure type (%) Vascular 23 (1.7) 12 (3.0) 11 (1.2) Intraperitoneal or retroperitoneal 29 (2.2) 5 (1.3) 24 (2.6) Urological or gynaecological 437 (33.0) 131 (33.1) 306 (33.0) Intra-thoracic 306 (23.3) 107 (27.0) 199 (21.4) Orthopaedic 398 (30.1) 106 (26.8) 292 (31.5) Head and neck 87 (6.6) 23 (5.8) 64 (6.9) Other 39 (3.0) 11 (2.8) 28 (3.0) High-risk surgery (%) 756 (57.1) 221 (55.8) 535 (57.7) ASA physical status (%) 1 99 (7.5) 24 (6.1) 75 (8.1) 2 772 (58.4) 207 (52.3) 565 (61.0) 3 433 (32.8) 159 (40.2) 274 (29.6) 4 18 (1.4) 6 (1.5) 12 (1.3) Preoperative medication (%) Beta blockers 213 (16.1) 137 (34.6) 76 (8.2) Diltiazem or verapamil 25 (1.9) 11 (2.8) 14 (1.5) Haemodynamic variables Resting heart rate (beats min−1) 77 (14.3) 75 (15.2) 78 (3.7) Resting systolic blood pressure (mm Hg) 129 (18.1) 127 (9.0) 130 (17.6) Resting pulse pressure (mm Hg) 51 (16.5) 51 (17.9) 52 (15.8)
ion (%) Beta blockers 213 (16.1) 137 (34.6) 76 (8.2) Diltiazem or verapamil 25 (1.9) 11 (2.8) 14 (1.5) Haemodynamic variables Resting heart rate (beats min−1) 77 (14.3) 75 (15.2) 78 (3.7) Resting systolic blood pressure (mm Hg) 129 (18.1) 127 (9.0) 130 (17.6) Resting pulse pressure (mm Hg) 51 (16.5) 51 (17.9) 52 (15.8) Markers of severe cardiac failure Chronotropic index <0.6 was associated with elevated preoperative NT pro-BNP >300 pg ml−1 (OR: 1.57 [1.11–2.23]; P<0.001), adjusted for potentially confounding factors. CI <0.6 was also associated with three independent measures of moderate-to-severe heart failure (Table 3). CI <0.6 was more commonly found in patients with ventilatory efficiency slope (minute ventilation/carbon dioxide production) ≥34 (OR: 1.40 [1.09–1.81]; P=0.009), peak oxygen consumption ≤14 ml kg−1 min−1 (OR: 7.57 [5.50–10.43]; P<0.001), and HR recovery ≤6 beats min−1 decrease during the first minute after the end of exercise (OR: 2.63 [1.97–3.52]; P<0.001).Table 2 Chronotropic incompetence and 1 yr mortality. The independent variable was chronotropic incompetence (defined as chronotropic index [CI] <0.6). The dependent variable was mortality within the 1 yr follow-up period. Results of two separate analyses are presented: firstly, univariable (unadjusted) logistic regression analysis, and secondly, multivariable logistic regression adjusting for three variables found to be associated with the dependent variable. The following variables were excluded from the final multivariable model: diabetes mellitus, peripheral vascular disease, atrial fibrillation, high-risk surgery, previous stroke or transient ischaemic attack, clinical diagnosis of heart failure, and preoperative renal insufficiency. Results are presented as odds ratios with 95% confidence intervals and associated P-values.
ltivariable model: diabetes mellitus, peripheral vascular disease, atrial fibrillation, high-risk surgery, previous stroke or transient ischaemic attack, clinical diagnosis of heart failure, and preoperative renal insufficiency. Results are presented as odds ratios with 95% confidence intervals and associated P-values. Table 2Co-variates Mortality Odds ratio P-value Univariable analysis Chronotropic incompetence 2.26 (1.13–4.51) 0.02 Multivariable analysis Male sex 2.22 (0.98–5.03) 0.06 History of stroke or transient ischaemic attack 3.00 (0.99–9.03) 0.05 History of chronic obstructive pulmonary disease 2.61 (1.17–5.86) 0.02 CI <0.6 1.98 (0.97–4.02) 0.06 Table 3 Chronotropic incompetence and markers of heart failure. The independent variable was chronotropic incompetence (defined as chronotropic index [CI] <0.6). The dependent variables were N-terminal pro-B-type natriuretic peptide (NT pro-BNP) >300 pg ml−1, ventilatory equivalent for carbon dioxide (VE/VCO2) at the anaerobic threshold ≥34, peak oxygen consumption (VO2) ≤14 ml kg−1 min−1, and HR recovery (HRR) ≤6 beats min−1 decrease within the first minute after the end of exercise. Results of univariable (unadjusted) and multivariable (adjusted) logistic regression analyses are presented as odds ratios with 95% confidence intervals and associated P-values. Variables were selected for inclusion in the multivariable model using stepwise selection.
within the first minute after the end of exercise. Results of univariable (unadjusted) and multivariable (adjusted) logistic regression analyses are presented as odds ratios with 95% confidence intervals and associated P-values. Variables were selected for inclusion in the multivariable model using stepwise selection. Table 3Co-variates NT pro-BNP >300 pmol L−1 VE/VCO2 ≥34 VO2 peak ≤14 ml kg−1 min−1 HRR ≤6 beats min−1 Odds ratio P-value Odds ratio P-value Odds ratio P-value Odds ratio P-value Univariable analysis CI <0.6 2.11 (1.56–2.86) <0.001 1.57 (1.23–2.00) <0.001 6.44 (4.82–8.59) <0.001 2.81 (2.11–3.74) <0.001 Multivariable analysis Age ≥70 yr 2.82 (2.01–3.95) <0.001 2.58 (2.02–3.29) <0.001 1.33 (0.95–1.84) 0.093 1.53 (1.13–2.05) 0.005 Male sex — — 0.55 (0.44–0.71) <0.001 0.17 (0.12–0.23) <0.001 0.63 (0.47–0.84) 0.002 History of atrial fibrillation 11.43 (5.71–22.88) <0.001 — — — — — — History of heart failure 7.42 (2.01–27.40) 0.003 — — — — — — History of coronary artery disease 2.56 (1.67–3.93) <0.001 — — — — — — History of peripheral vascular disease — — — — 2.62 (1.20–5.73) 0.015 — — History of hypertension 1.46 (1.02–2.10) 0.039 — — — — — — History of stroke or transient ischaemic attack — — — — — — — — History of chronic obstructive pulmonary disease — — 1.36 (0.96–1.93) 0.082 — — — — History of diabetes mellitus — — 1.31 (0.97–1.77) 0.078 — — 1.46 (1.02–2.07) 0.037 Preoperative estimated glomerular filtration rate <60 ml min−1 (1.73 m2)−1 3.68 (2.29–5.91) <0.001 1.67 (1.10–2.53) 0.016 1.90 (1.14–3.15) 0.013 1.67 (1.05–2.66) 0.029 High-risk surgery — — — — 1.44 (1.05–1.98) 0.025 — — CI <0.6 1.57 (1.11–2.23) 0.011 1.40 (1.09–1.81) 0.009 7.57 (5.50–10.43) <0.001 2.63 (1.97–3.52) <0.001
operative estimated glomerular filtration rate <60 ml min−1 (1.73 m2)−1 3.68 (2.29–5.91) <0.001 1.67 (1.10–2.53) 0.016 1.90 (1.14–3.15) 0.013 1.67 (1.05–2.66) 0.029 High-risk surgery — — — — 1.44 (1.05–1.98) 0.025 — — CI <0.6 1.57 (1.11–2.23) 0.011 1.40 (1.09–1.81) 0.009 7.57 (5.50–10.43) <0.001 2.63 (1.97–3.52) <0.001 Primary outcome: myocardial injury Within 3 days after surgery, 162/1325 (12.2%) patients sustained myocardial injury, which occurred in 50/396 (12.6%) patients with CI <0.6 and 112/928 (12.1%) patients with CI ≥0.6. There was no difference in the odds of myocardial injury amongst patients with CI <0.6 compared with those with CI >0.6 (unadjusted OR: 1.05 [0.74–1.50]; P=0.78). In the multivariable analysis, CI <0.6 was not associated with myocardial injury (P>0.60). Secondary outcome: sympathetic vs parasympathetic measures and myocardial injury We examined the prevalence of physiological markers of impaired sympathetic and parasympathetic functions using chronotropic incompetence and HR recovery, respectively. We found that 169 (12.8%) had low CI alone, 294 (22.2%) had HR recovery <12 beats min−1 alone, and 227 (17.2%) had both HR recovery <12 beats min−1 and CI <0.6. When we repeated the primary analysis using HR recovery <12 beats min−1 and CI <0.6 as the exposures, we found that only HR recovery <12 beats min−1 was associated with myocardial injury (Supplementary Table S1).
%) had HR recovery <12 beats min−1 alone, and 227 (17.2%) had both HR recovery <12 beats min−1 and CI <0.6. When we repeated the primary analysis using HR recovery <12 beats min−1 and CI <0.6 as the exposures, we found that only HR recovery <12 beats min−1 was associated with myocardial injury (Supplementary Table S1). Secondary outcome: postoperative mortality Within 1 yr of surgery, 33/1325 (2.5%) patients had died. With univariable analysis, postoperative mortality was more frequent amongst patients with CI <0.6 (16/396 [4.0%]) compared with patients without CI <0.6 (17/928 [1.8%]; unadjusted OR: 2.26 [1.13–4.51]; P=0.02. However, on multivariable analysis, CI <0.6 and mortality were not significantly associated (OR: 1.98 [0.97–4.02]; P=0.06; Table 2 and Fig. 2).Fig 2 Kaplan–Meier survival plot for chronotropic incompetence (chronotropic index <0.6) vs no chronotropic incompetence (chronotropic index ≥0.6). Fig 2
Secondary outcome: postoperative mortality Within 1 yr of surgery, 33/1325 (2.5%) patients had died. With univariable analysis, postoperative mortality was more frequent amongst patients with CI <0.6 (16/396 [4.0%]) compared with patients without CI <0.6 (17/928 [1.8%]; unadjusted OR: 2.26 [1.13–4.51]; P=0.02. However, on multivariable analysis, CI <0.6 and mortality were not significantly associated (OR: 1.98 [0.97–4.02]; P=0.06; Table 2 and Fig. 2).Fig 2 Kaplan–Meier survival plot for chronotropic incompetence (chronotropic index <0.6) vs no chronotropic incompetence (chronotropic index ≥0.6). Fig 2 Sensitivity analyses When we repeated the primary and secondary analyses, including the preoperative use of beta blockers, diltiazem, or verapamil as co-variates, the results were very similar (Supplementary Table S2). Similar proportions of patients receiving these drugs achieved RER (carbon dioxide produced/oxygen consumed) >1.05. When we repeated the primary analysis excluding patients receiving beta blockers, diltiazem, or verapamil, CI <0.6 was not associated with myocardial injury (OR: 7.20 [0.60–87.02]; P=0.12) in the univariable analysis. The multivariable model did not converge because of collinearity between variables. We could not complete the regression analysis for mortality because an insufficient number of patients died. When we repeated the primary analysis using age-predicted maximum HR calculated using the method described by Tanaka and colleagues,23 CI <0.6 was not associated with myocardial injury (OR: 1.10 [0.78–1.53]; P=0.59) or mortality (OR: 1.50 [0.75–2.99]; P=0.25) in the univariable analysis. In multivariable analysis, CI <0.6 was removed from the stepwise models at the P>0.56 level for both outcomes. When we repeated the analysis using CI <0.8 as the exposure, CI <0.8 was not associated with myocardial injury (OR: 0.87 [0.62–1.21]; P=0.41) or mortality (OR: 1.15 [0.56–2.36]; P=0.70) in the univariable analysis. In the multivariable analysis, CI <0.8 was removed from the stepwise model at the P>0.19 level (myocardial injury) and the P>0.97 level (mortality). When we repeated the analysis using CI as a continuous variable, chronotropic incompetence was not significantly associated with myocardial injury (OR: 1.20 [0.68–2.09]; P=0.53) or mortality (OR: 0.78 [0.38–1.62]; P=0.51).
from the stepwise model at the P>0.19 level (myocardial injury) and the P>0.97 level (mortality). When we repeated the analysis using CI as a continuous variable, chronotropic incompetence was not significantly associated with myocardial injury (OR: 1.20 [0.68–2.09]; P=0.53) or mortality (OR: 0.78 [0.38–1.62]; P=0.51). Discussion The principal finding of this analysis was that preoperative chronotropic incompetence—an impaired ability to increase HR in response to exercise—was not associated with myocardial injury within 3 days after surgery. Impaired HR recovery (indicative of parasympathetic dysfunction) after exercise, rather than impaired HR increase during exercise (indicative of sympathetic dysfunction), was associated with postoperative myocardial injury. We also confirmed the deconditioned phenotype of subclinical cardiac failure in preoperative patients, as chronotropic incompetence was associated with elevated preoperative NT pro-BNP, a preoperative risk factor for postoperative cardiovascular morbidity and a biomarker for heart failure in the general population. Moreover, we found a strong association between chronotropic incompetence and CPET-derived markers for heart failure.8, 34 These data confirm our previous findings in a large prospective cohort that almost one-third of patients undergoing noncardiac surgery exhibit a phenotype of subclinical cardiac failure that is frequently accompanied by significant autonomic impairment.8, 31
incompetence and CPET-derived markers for heart failure.8, 34 These data confirm our previous findings in a large prospective cohort that almost one-third of patients undergoing noncardiac surgery exhibit a phenotype of subclinical cardiac failure that is frequently accompanied by significant autonomic impairment.8, 31 We defined chronotropic incompetence using an established threshold of CI, which is prognostically associated with increased mortality in longitudinal cohorts of patients with heart failure.21 Our results do not support a link between beta-adrenoceptor dysfunction, as identified using chronotropic incompetence, and myocardial injury. The inability to increase HR in patients with chronotropic incompetence suggests that a direct link between HR and supply-demand mismatch is unlikely to underpin myocardial injury. However, it is plausible that chronotropic incompetence could promote myocardial injury through indirect links. The failure to increase cardiac output under certain perioperative circumstances, which require HR elevation, may be linked to myocardial injury, as suggested by the Perioperative Ischemic Evaluation (POISE) trial of perioperative metoprolol.35 Similarly, failure to meet metabolic demands during surgery may drive organ injury, which in turn could increase the risk of myocardial injury. The precise mechanism leading to a decrease in β1-adrenergic receptor expression and desensitisation in cardiac failure is unclear, but may involve oxidative stress36 driven by chronic systemic inflammation.37 Our finding that chronotropic incompetence is associated with reduced survival after noncardiac surgery is consistent with similar observations in patients with heart failure,12, 21, 22, 38, 39, 40 supporting the hypothesis that there is a cohort of surgical patients with severe, yet subclinical, heart failure.8
ion.37 Our finding that chronotropic incompetence is associated with reduced survival after noncardiac surgery is consistent with similar observations in patients with heart failure,12, 21, 22, 38, 39, 40 supporting the hypothesis that there is a cohort of surgical patients with severe, yet subclinical, heart failure.8 A notable strength of our study is that the results have high external validity attributable to the prospective, international, multicentre nature of the study cohort, which makes our findings readily generalisable to the majority of intermediate- and high-risk surgical patients. The primary outcome, myocardial injury, is an objective, biomarker defined endpoint and not subject to observer bias. Clinicians at each participating hospital were blinded to the results of the preoperative CPET. Therefore, the measurement of chronotropic incompetence did not influence perioperative care. Our analysis also has several limitations. As with any observational study, it is possible that our results may be influenced by unmeasured confounding. The primary outcome was myocardial injury, and the sample size for the study, which was based on cardiovascular outcomes, was appropriate for this outcome. However, the study was not powered to detect differences in mortality, and therefore, we advise that inferences regarding mortality should be with caution.
ounding. The primary outcome was myocardial injury, and the sample size for the study, which was based on cardiovascular outcomes, was appropriate for this outcome. However, the study was not powered to detect differences in mortality, and therefore, we advise that inferences regarding mortality should be with caution. It is possible that our results could have been influenced by the definition of chronotropic incompetence. We defined this as CI <0.6, which is an indicator of poor prognosis in patients with heart failure.8 However, some studies in other populations have used a different threshold of CI (<0.8).33 When we repeated the analysis using CI <0.8, the results were similar. CI was calculated as the proportion of age-predicted maximum HR reached during preoperative exercise. As with any pragmatic study of exercise, there is an underlying assumption that the HR recorded at peak exertion is an accurate measure of maximal HR. Because of the clinical nature of the study, we were unable to confirm this with repeated measurements, so there is a possibility that some measurements of maximum HR might not represent true maximal values. However, more than 80% of the cohort achieved an end-exercise RER (carbon dioxide produced/oxygen consumed) of >1.05, which is generally accepted to represent peak effort.12
o confirm this with repeated measurements, so there is a possibility that some measurements of maximum HR might not represent true maximal values. However, more than 80% of the cohort achieved an end-exercise RER (carbon dioxide produced/oxygen consumed) of >1.05, which is generally accepted to represent peak effort.12 There are several methods for calculating age-predicted maximum HR, which could potentially influence the results. We chose the method described by Astrand,24 which is the most widely accepted, as the primary method. However, we recalculated age-predicted maximum HR using the Tanaka and colleagues23 method, and the results were similar.23 When we repeated the analysis using CI as a continuous variable, we did not identify a relationship with myocardial injury. However, this method assumes a linear relationship between CI and the risk myocardial injury, which may not be true.
um HR using the Tanaka and colleagues23 method, and the results were similar.23 When we repeated the analysis using CI as a continuous variable, we did not identify a relationship with myocardial injury. However, this method assumes a linear relationship between CI and the risk myocardial injury, which may not be true. Resting HR or change in HR may be influenced by rate-limiting medications. When we repeated the analysis after removing the 224 patients receiving beta blockers or rate-limiting calcium channel antagonists, our results were similar. We also repeated the multivariable analysis including treatment with beta blockers or rate-limiting calcium channel anatagonists as separate terms in the model, and our results were similar. CPET were conducted and interpreted by investigators at 24 participating hospitals, so there is a potential for observer bias and measurement error between centres. However, this was mitigated through the prospective use of a standardised CPET protocol and case report form.16 It is possible that a potential relationship between chronotropic incompetence and myocardial injury may have been confounded by intraoperative hypotension. However, when we repeated the primary analysis adding intraoperative vasopressor use (a surrogate marker of hypotension) as a co-variate, the results were similar.
• The EEG parameters correlating with pharmacokinetic–pharmacodynamic modelling of the hypnotic actions of ketamine were studied in human volunteers. • Changes in slow wave activity after a hypnotic dose of ketamine were well fitted by a standard sigmoid model • Onset, but not offset, of slow wave activity was consistently associated with loss of behavioural responsiveness. The electroencephalographic (EEG) effects of ketamine have been studied for more than 50 yr.1 In an early paper, Schwartz and colleagues1 reported that hypnotic doses of ketamine produced strong increases in high frequency (beta–gamma, 20–45 Hz) and theta (4–8 Hz) waves, punctuated by episodic slow (delta, 0.25–1.5 Hz) waves (see Fig 1, Fig 2 in their paper, and subsequent work by Schüttler and colleagues2). Using multi-spectral techniques on EEG obtained from frontal electrodes, Akeju and co-workers3 have recently rediscovered, and more elegantly quantified, the ketamine-induced EEG pattern, in which bursts of high frequency activity alternate with slow waves. They termed this a ‘gamma burst’ pattern. However, they had only looked at frontal electrodes, and had also given most of the patients adjunctive midazolam and fentanyl. They suggested further work with high-density EEG collection needed to be done. Using such a system, a previous paper by Vlisides and colleagues4 compared sub-hypnotic infusions of ketamine with hypnotic doses, and found strikingly increased EEG theta power (with regional phase locking), decreased alpha power, and loss of anterior-to-posterior alpha connectivity. They also noted an increase in delta power.Fig 1 Changes in spectral power in response to intravenous injection of 1.5 mg kg−1 ketamine (administered at the white/grey vertical lines) for one individual. Times of loss of behavioural response (LOBR; black rectangle) and recovery of behavioural response (ROBR; blue line). We use a rectangle for LOBR because the subject could have lost responsiveness at any time in the 30 s window between the verbal commands from the audio loop. (a) Spectrogram of the EEG (yellow is high power, red medium power, and black low power). The green line shows the time course of the calculated effect-site concentration of ketamine (μg ml−1×10). (b) Time course of the slow wave activity (SWA) and theta power, and (c, d) examples of the hysteresis loops of effect-site concentrations vs SWA for two different t1/2ke0 values, where t1/2ke0 is the half-time of equilibration between blood and effect-site.
alculated effect-site concentration of ketamine (μg ml−1×10). (b) Time course of the slow wave activity (SWA) and theta power, and (c, d) examples of the hysteresis loops of effect-site concentrations vs SWA for two different t1/2ke0 values, where t1/2ke0 is the half-time of equilibration between blood and effect-site. LOBR and ROBR could have occurred over a 30 s period, hence the black and blue lines. Data are from channel 52, which corresponds to P3 in the 10–20 system. Fig 1Fig 2 Time course of slow wave activity, theta, and beta–gamma power after induction (time=0 s) for channel 46 (T3) for all subjects. Individual trajectories are in grey, and the thick green line is the median at each time point. The black dots are the point of detection of loss of behavioural response to command, and the blue dots are the point of recovery of behavioural response to command. (For graphical clarity, dots have been used—the actual points of change in behavioural response could have occurred up to 30 s before the dots.) Fig 2
Fig 1Fig 2 Time course of slow wave activity, theta, and beta–gamma power after induction (time=0 s) for channel 46 (T3) for all subjects. Individual trajectories are in grey, and the thick green line is the median at each time point. The black dots are the point of detection of loss of behavioural response to command, and the blue dots are the point of recovery of behavioural response to command. (For graphical clarity, dots have been used—the actual points of change in behavioural response could have occurred up to 30 s before the dots.) Fig 2 Slow wave power is associated with loss of perception in subjects given gamma aminobutyric acid (GABA)-ergic drugs such as propofol or volatile anaesthetics.5, 6 Ketamine has different molecular targets; but if the EEG slow waves produced by ketamine are associated with loss of responsiveness, it suggests that slow waves might be causally mediating change in consciousness. A few pharmacokinetic–pharmacodynamic (PKPD) studies of ketamine have been reported, mainly looking at the perceived analgesic effect of ketamine7 or clinical sedation8 as the primary outcome. It is unclear which EEG parameter is optimal for the hypnotic PKPD modelling of ketamine, or which EEG parameter is most closely linked to loss of behavioural responsiveness (LOBR). The median frequency of the EEG has been related to serum ketamine concentrations using PKPD models.2 This is a composite measure, reflecting the balance between the divergent ketamine effects on the EEG, namely ketamine-induced increased slow wave power and ketamine-induced increase in higher frequency power. We performed a formal PKPD analysis of the high density EEG and ketamine data exploring dose-related changes in theta, beta–gamma, and slow wave activity (SWA). We hypothesised that the changes in SWA with a hypnotic dose of ketamine could be fitted by a standard sigmoid PKPD model, and their onset could be consistently associated with LOBR.
KPD analysis of the high density EEG and ketamine data exploring dose-related changes in theta, beta–gamma, and slow wave activity (SWA). We hypothesised that the changes in SWA with a hypnotic dose of ketamine could be fitted by a standard sigmoid PKPD model, and their onset could be consistently associated with LOBR. Methods Data collection The details of the data collection have been published.4, 9 In brief, after written informed consent and approval by the University of Michigan Medical School Institutional Review Board, Ann Arbor, MI, USA (HUM00061087), 15 volunteers (7 male/8 female, American Society of Anesthesiologists [ASA] physical status 1, 20–40 yr of age, BMI <30 kg m−2) were given a sub-anaesthetic i.v. infusion of ketamine 0.5 mg kg−1 administered over 40 min, followed by a 30 min pause for rest and psychometric testing.9 Then a hypnotic induction i.v. bolus dose of ketamine 1.5 mg kg−1 was administered. The times of loss and recovery of behavioural responsiveness (LOBR/ROBR) were estimated by an audioloop command to squeeze the right or left hand every 30 s. The EEG was obtained using 128-channel system (HydroCel nets, Net Amps 400 amplifiers, and Net Station 4.5 software; Electrical Geodesics, Inc., Eugene, OR, USA) and digitised at 500 Hz using a vertex reference.
ponsiveness (LOBR/ROBR) were estimated by an audioloop command to squeeze the right or left hand every 30 s. The EEG was obtained using 128-channel system (HydroCel nets, Net Amps 400 amplifiers, and Net Station 4.5 software; Electrical Geodesics, Inc., Eugene, OR, USA) and digitised at 500 Hz using a vertex reference. Signal processing and analysis Basic spectral and connectivity patterns have been published for the first 10 participants in this study across both sedative and anaesthetic periods.4 Here, we report a separate PKPD analysis on the second (bolus) part of the study, where all participants (n=15) lost responsiveness. This occurred 30 min after cessation of the low-dose sub-hypnotic infusion. All processing was done using the Chronux (http://chronux.org/) and EEGLAB10 toolboxes, and purpose-written Matlab (MathWorks, Natick, MA, USA) scripts. The output of the EEG system produces a virtual DC signal. Channels were therefore re-referenced to an average reference, down-sampled to 125 Hz, and bandpass-filtered (0.1–50 Hz) using a fifth-order Butterworth filter and the ‘filtfilt.m’ phase preserving filter function. The spectral power (in dB) was calculated using the Chronux ‘mtspecgramc.m’ function on a moving 4 s segment of data (3 s overlap), time–bandwidth product of two and three tapers. SWA was calculated as mean power from 0.25 to 1.5 Hz, and this was smoothed using a median filter. Maximum theta (4–8 Hz) and beta–gamma (20–45 Hz) power were similarly calculated.
d using the Chronux ‘mtspecgramc.m’ function on a moving 4 s segment of data (3 s overlap), time–bandwidth product of two and three tapers. SWA was calculated as mean power from 0.25 to 1.5 Hz, and this was smoothed using a median filter. Maximum theta (4–8 Hz) and beta–gamma (20–45 Hz) power were similarly calculated. Pharmacokinetic-pharmacodynamic modelling Effect-site concentrations of ketamine were estimated using parameters derived from published work.11 Because the previous low-dose infusion will have slightly loaded the peripheral compartment, the time course of concentrations were calculated on the basis of whole ketamine administration (i.e. both the sub-hypnotic infusion and hypnotic bolus). The concentration of ketamine at the start of the hypnotic bolus was at a level that has no hypnotic effects (39 ng ml−1). To see if the results were robust to the choice of pharmacokinetic model, they were also analysed using the model of Clements and Nimmo.12 The results were not significantly different (P=0.47). In a separate unpublished pilot study, blood samples for plasma ketamine were obtained at the end of the slow infusion in seven similar subjects. The mean ketamine concentration was 183 ng ml−1, which was within the 95% confidence limits of the mean calculated using our PKPD model (124–195 ng ml−1). This confirmed that our PKPD model was well calibrated to the population mean. However, as in all PKPD modelling, individual variability in plasma concentrations was large. In the pilot study, the actual ketamine concentrations ranged from 62 to 440 ng ml−1.
f the mean calculated using our PKPD model (124–195 ng ml−1). This confirmed that our PKPD model was well calibrated to the population mean. However, as in all PKPD modelling, individual variability in plasma concentrations was large. In the pilot study, the actual ketamine concentrations ranged from 62 to 440 ng ml−1. To see if different regions of the brain showed different sensitivity to ketamine, we compared PKPD models of SWA from five 10–20 system channels (Fz, F3, T3, P3, Pz) chosen to represent medial, lateral, frontal, and parietal cortices. We also compared the PKPD modelling of theta and beta–gamma band powers with that of the SWA. For the modelling, we used 480 s of EEG data from the time of ketamine injection to fit the PKPD model. This time frame was chosen to concentrate on the period of unresponsiveness and minimise the various movement artifacts present around the time of ROBR. For the same reason, we set the baseline SWA, theta, and beta–gamma as the minimum for the period between ketamine injection and LOBR.
injection to fit the PKPD model. This time frame was chosen to concentrate on the period of unresponsiveness and minimise the various movement artifacts present around the time of ROBR. For the same reason, we set the baseline SWA, theta, and beta–gamma as the minimum for the period between ketamine injection and LOBR. Modelling was done in three stages. Firstly, for each subject, channel, and frequency band, we ran an individual PKPD model (‘nlinfit.m’) using 50 different ke0 values equidistantly spaced on a logarithmic scale. This resulted in values for the half-time for equilibration between blood and effect-site (t1/2ke0) ranging from 8 to 139 s. Thus, we obtained the ke0 value for each subject and channel that gave the best model fit to the SWA, theta, and beta–gamma power as measured by coefficient of variation (R2). This effectively identified the ke0 required to collapse the hysteresis loop for each frequency band of interest (see Fig. 1c and d for SWA).
s. Thus, we obtained the ke0 value for each subject and channel that gave the best model fit to the SWA, theta, and beta–gamma power as measured by coefficient of variation (R2). This effectively identified the ke0 required to collapse the hysteresis loop for each frequency band of interest (see Fig. 1c and d for SWA). The drug effect on EEG power was fitted for each frequency band using a standard sigmoid function: (1) EEGpower=β1+β21+exp(−CeKetamine−β3β4), where β1 determines the baseline SWA, β2 the plateau SWA, β3 and β4 control the slope and position of the sigmoid, and CeKetamine is the effect-site concentration of ketamine. The time courses of both the raw data and the fitted model were then overlaid and checked to confirm that the model produced a single peaked function that plausibly tracked the raw data and ketamine concentrations before entry into the mixed-effects model. This was because of convergence problems in the mixed-effects model if a significant proportion of the fits followed unphysiological trajectories. The second stage of analysis was to estimate the sigmoid parameters using a non-linear mixed-effects regression model (‘nlmefitsa.m’) that used a stochastic expectation maximisation algorithm.13 Population parameters describing drug concentration vs EEG power (i.e. SWA/theta/beta–gamma) were included as fixed effects. Between-subject variation around the population means was included as a random effect. We also used a constant error model.
efitsa.m’) that used a stochastic expectation maximisation algorithm.13 Population parameters describing drug concentration vs EEG power (i.e. SWA/theta/beta–gamma) were included as fixed effects. Between-subject variation around the population means was included as a random effect. We also used a constant error model. Finally, we statistically compared parameters and goodness-of-fit for different channels, subjects, and frequency bands using analysis of variance (anova). Slow wave morphology The alternating delta–gamma burst pattern reported by Akeju and colleagues3 suggested an episodic phenomenon that might not necessarily be captured well using these frequency domain methods. We therefore also examined the morphology of individual slow waves in the time domain. To minimise filter-induced distortion of the waveform, we removed the baseline drift by subtracting a 4 s median filtered waveform. We then identified the time, width, and amplitude of any slow waves that had an amplitude of >4 inter-quartile ranges (IQRs) from baseline, a duration of >0.04 s and were not within 1.6 s of the previous wave.
duced distortion of the waveform, we removed the baseline drift by subtracting a 4 s median filtered waveform. We then identified the time, width, and amplitude of any slow waves that had an amplitude of >4 inter-quartile ranges (IQRs) from baseline, a duration of >0.04 s and were not within 1.6 s of the previous wave. Statistical analysis For normally distributed data (probability distribution tested using the Kolmogorov–Smirnoff test), we report observational data as mean (standard deviation, sd). To examine the influence of subject, channel, and frequency band on the model parameters and R2 values, we used single anova, with subject as the between-subject group, and channels and frequency bands (modelled as within-subject fixed effects). We used the Bonferroni test for post hoc group comparisons. Otherwise, we report median (IQR) and use the Wilcoxon rank sum test (or sign rank test for paired data) for comparisons. Results In response to a hypnotic dose of ketamine, the EEG patterns for all subjects were similar to those described.1, 2, 3 A typical spectrogram is shown in Figure 1a, and demonstrates increased broadband gamma and beta power, and the appearance of narrowband theta oscillations after ketamine injection (white vertical line). A strong period of SWA can also be seen from around the point of loss of behavioural response (black rectangle) to about 400 s (Fig. 1b).
is shown in Figure 1a, and demonstrates increased broadband gamma and beta power, and the appearance of narrowband theta oscillations after ketamine injection (white vertical line). A strong period of SWA can also be seen from around the point of loss of behavioural response (black rectangle) to about 400 s (Fig. 1b). The hysteresis curves for SWA for two example ke0s (Fig. 1c and d) show that the collapse of the curve occurs with a shorter t1/2ke0 of 26 s. Overall, the best fit was achieved with mean t1/2ke0 for SWA of 23 (4) s (range 13–39 s). The t1/2ke0 for beta–gamma power was variable (mean 98 [72] s, range 9–138 s), and there were 33 out of 75 channels for which no good fit could be obtained. Subject (P<0.001), frequency band (P<0.001), and channel (P=0.002) all significantly influenced the t1/2ke0. In the post hoc analysis, t1/2ke0 for theta power was significantly longer (47 [22] s, range 13–138 s) than those of SWA and beta–gamma, which reflects the slower onset and offset of theta power. As regards channel effects, the only significant effect was that the t1/2ke0 for T3 was shorter than Fz and P3. Relationship of responsiveness to observed power changes As suggested by the significant difference in ke0 values, the time course of increases in SWA and theta power were also different. The median SWA and beta–gamma increases had a fast onset and rapid decay within about 5 min, whereas the increase in theta power was more prolonged. Five subjects also showed a clear bimodal pattern for theta power (Fig. 2).
ference in ke0 values, the time course of increases in SWA and theta power were also different. The median SWA and beta–gamma increases had a fast onset and rapid decay within about 5 min, whereas the increase in theta power was more prolonged. Five subjects also showed a clear bimodal pattern for theta power (Fig. 2). Subjects lost responsiveness at 79 (20) s, and regained responsiveness after a wide variation in time (682 [212] s). For the Pz electrode, subjects experienced LOBR on average 33 (32) s before the maximum value of SWA was reached. However, four of them had LOBR around the same time as their maximum SWA, although it is possible that they might have experienced LOBR up to 30 s before, because of the 30 s interval for questioning. The maximum in theta power was significantly later (95 [98] s, P=0.04). Maximum beta–gamma power tended to occur close to LOBR (22 [58] s) but was variable with six subjects having maximum beta–gamma up to 46 s before LOBR.
hey might have experienced LOBR up to 30 s before, because of the 30 s interval for questioning. The maximum in theta power was significantly later (95 [98] s, P=0.04). Maximum beta–gamma power tended to occur close to LOBR (22 [58] s) but was variable with six subjects having maximum beta–gamma up to 46 s before LOBR. As shown in Table 1, subjects lost responsiveness at higher estimated effect-site concentrations of ketamine than at ROBR (1.64 [0.17] μg.ml−1 vs 1.06 [0.21] μg.ml−1, P<0.0001 paired t-test). ROBR occurred a long time after the EEG maxima (569 [215] s after the SWA peak, 508 [195] s after the theta peak, and 580 [233] s after the beta–gamma peak). LOBR was associated with increased SWA (13.8 [7.2] dB) than at ROBR (9.3 [2.5] dB, P=0.04 paired t-test).Table 1 Power (dB) at various time points for the three frequency bands, and concomitant calculated ketamine effect-site concentrations (μg ml−1). The power at return of behavioural response (ROBR) is significantly less than at loss of behavioural response (LOBR) for all frequency bands (P=0.025 SWA, P<0.001 theta, P=0.0017 beta–gamma). Power at LOBR and ROBR is significantly less than the maximum power for all wave bands (P<0.001). Data for T3 channel are shown. Data shown as mean (sd). Table 1 Gamma Theta Slow wave activity CeKetamine LOBR –3.09 (2.58) 9.25 (3.52) 12.55 (6.33) 1.64 (0.17) Maximum power 0.95 (0.99) 25.7 (45.9) 18.1 (0.56) – ROBR –6.56 (2.77) 4.94 (3.79) 7.96 (3.07) 1.06 (0.21)
As shown in Table 1, subjects lost responsiveness at higher estimated effect-site concentrations of ketamine than at ROBR (1.64 [0.17] μg.ml−1 vs 1.06 [0.21] μg.ml−1, P<0.0001 paired t-test). ROBR occurred a long time after the EEG maxima (569 [215] s after the SWA peak, 508 [195] s after the theta peak, and 580 [233] s after the beta–gamma peak). LOBR was associated with increased SWA (13.8 [7.2] dB) than at ROBR (9.3 [2.5] dB, P=0.04 paired t-test).Table 1 Power (dB) at various time points for the three frequency bands, and concomitant calculated ketamine effect-site concentrations (μg ml−1). The power at return of behavioural response (ROBR) is significantly less than at loss of behavioural response (LOBR) for all frequency bands (P=0.025 SWA, P<0.001 theta, P=0.0017 beta–gamma). Power at LOBR and ROBR is significantly less than the maximum power for all wave bands (P<0.001). Data for T3 channel are shown. Data shown as mean (sd). Table 1 Gamma Theta Slow wave activity CeKetamine LOBR –3.09 (2.58) 9.25 (3.52) 12.55 (6.33) 1.64 (0.17) Maximum power 0.95 (0.99) 25.7 (45.9) 18.1 (0.56) – ROBR –6.56 (2.77) 4.94 (3.79) 7.96 (3.07) 1.06 (0.21) PKPD modelling of slow wave activity response There was inter-individual variation in intensity of response and, on the basis of visual inspection, we removed one channel out of 75 from the SWA model, four from the theta model, and 33 from the beta–gamma model. The fixed-effect model parameters and goodness-of-fit (R2) for each frequency band and channel are shown in Table 2. The individual time course of the raw and modelled data are shown in Supplementary information. Overall, the model fitted well and comparably for SWA and theta frequencies. The goodness-of-fit was slightly better for the theta frequency band model (R2, 86 vs 80 vs 79), but more channels had been withdrawn before the theta model fit than the SWA fit. When the beta–gamma time course was reliable, the model could be fitted well, but about half the beta–gamma records had to be withdrawn from the analysis because they were too noisy to model. Unsurprisingly, the parameters differed significantly between different frequency bands. The effects of channel were only significant for β4 (i.e. position of sigmoid), which was significantly larger in Pz, vs Fz and T3 on post hoc analysis.Table 2 Estimated parameters for the pharmacokinetic–pharmacodynamic (PKPD) slow wave activity model from five different channels (Fz, Pz, F3, T3, P3). Values represent mean (standard error of the mean, sem) across individuals. Sigmoid parameters: β1, baseline activity; β2, plateau; β3 and β4 control the slope and position of the sigmoid. R2 describes the goodness of fit.
harmacodynamic (PKPD) slow wave activity model from five different channels (Fz, Pz, F3, T3, P3). Values represent mean (standard error of the mean, sem) across individuals. Sigmoid parameters: β1, baseline activity; β2, plateau; β3 and β4 control the slope and position of the sigmoid. R2 describes the goodness of fit. Table 2Slow wave activity Parameter Fz (n=14) Pz (n=15) F3 (n=15) T3 (n=15) P3 (n=15) β1 6.99 (0.88) 4.93 (0.81) 2.86 (0.53) 4.46 (0.51) 6.13 (0.46) β2 25.40 (4.87) 34.06 (4.78) 20.63 (3.90) 28.83 (3.57) 35.48 (5.54) β3 1.72 (0.04) 1.78 (0.04) 1.71 (0.02) 1.79 (0.05) 1.81 (0.05) β4 0.05 (0.01) 0.08 (0.02) 0.08 (0.02) 0.14 (0.03) 0.15 (0.04) R2 0.79 (0.04) 0.86 (0.02) 0.79 (0.04) 0.77 (0.04) 0.79 (0.02) Theta Parameter Fz (n=15) Pz (n=14) F3 (n=15) T3 (n=13) P3 (n=14) β1 3.68 (0.92) 2.36 (0.83) 1.85 (0.62) 1.14 (1.64) 1.24 (1.50) β2 11.86 (2.56) 14.37 (3.02) 11.14 (2.87) 15.64 (3.70) 18.02 (3.61) β3 1.25 (0.12) 1.39 (0.07) 1.55 (0.07) 1.34 (0.16) 1.27 (0.14) β4 0.15 (0.06) 0.22 (0.07) 0.16 (0.03) 0.28 (0.07) 0.53 (0.12) R2 0.86 (0.03) 0.86 (0.04) 0.87 (0.02) 0.84 (0.02) 0.84 (0.04) Beta–gamma Parameter Fz (n=14) Pz (n=5) F3 (n=5) T3 (n=10) P3 (n=8) β1 –7.52 (0.77) –4.66 (0.63) –3.60 (1.80) –6.52 (0.83) –7.49 (0.61) β2 11.47 (2.71) 11.78 (1.63) 16.83 (11.70) 8.45 (1.67) 13.8 (4.38) β3 1.60 (0.16) 1.87 (0.27) 1.37 (0.08) 1.72 (0.06) 1.69 (0.20) β4 0.21 (0.05) 0.30 (0.09) 0.04 (0.02) 0.16 (0.07) 0.23 (0.07) R2 0.76 (0.03) 0.67 (0.12) 0.85 (0.04) 0.81 (0.04) 0.74 (0.09)
–7.52 (0.77) –4.66 (0.63) –3.60 (1.80) –6.52 (0.83) –7.49 (0.61) β2 11.47 (2.71) 11.78 (1.63) 16.83 (11.70) 8.45 (1.67) 13.8 (4.38) β3 1.60 (0.16) 1.87 (0.27) 1.37 (0.08) 1.72 (0.06) 1.69 (0.20) β4 0.21 (0.05) 0.30 (0.09) 0.04 (0.02) 0.16 (0.07) 0.23 (0.07) R2 0.76 (0.03) 0.67 (0.12) 0.85 (0.04) 0.81 (0.04) 0.74 (0.09) It was rare for the SWA to achieve a plateau at the 1.5 mg kg−1 dose of ketamine, so the estimation of the top part of the sigmoid (i.e. β2 parameter) was least accurate (Fig. 1d). The modelled concentration–response curves, and examples of the best, worst, and median fits are shown in Figure 3. The curves tend to be very steep—saturating over the 1.7–2 μg ml−1 concentration range, and reach very high values of SWA for some subjects.Fig 3 Pharmacokinetic–pharmacodynamic modelling for slow wave activity in the Pz channel. Modelled concentration–effect curves for each subject (a), and examples of the best (b), worst (c), and median (d) modelled slow wave activity (SWA) time course vs real SWA time course. Fig 3
It was rare for the SWA to achieve a plateau at the 1.5 mg kg−1 dose of ketamine, so the estimation of the top part of the sigmoid (i.e. β2 parameter) was least accurate (Fig. 1d). The modelled concentration–response curves, and examples of the best, worst, and median fits are shown in Figure 3. The curves tend to be very steep—saturating over the 1.7–2 μg ml−1 concentration range, and reach very high values of SWA for some subjects.Fig 3 Pharmacokinetic–pharmacodynamic modelling for slow wave activity in the Pz channel. Modelled concentration–effect curves for each subject (a), and examples of the best (b), worst (c), and median (d) modelled slow wave activity (SWA) time course vs real SWA time course. Fig 3 Slow wave activity morphology Although our study was not designed as a formal comparison of different SWA morphologies, the slow waves induced by ketamine (Figs. 1d 4) look quite different in shape and size to those typically described during propofol or sevoflurane anaesthesia, or in natural sleep, which tend to be continuous in nature, and smaller in amplitude compared with the episodic ketamine-induced waves described below. Some subjects showed only small, short-lived increases in slow wave power. For those participants who showed an obvious strong response across electrode channels (n=9), the slow waves were characterised by between 10 and 45 intense stereotypical hyperpolarisations or depolarisations (absolute amplitude deviation from baseline 140 [58] μV) lasting about 0.3–1 s (median 0.26 s, IQR 0.29 s) and occurring every 3–10 s (median 5.9 s, IQR 6.8 s). These slow-wave episodes occur as interruptions on an underlying ketamine-induced theta–gamma EEG pattern. They start around 40 s after the ketamine bolus and resolve around 300–400 s. Two subjects showed depolarisations that mirrored the hyperpolarisation pattern. These were considered to be a manifestation of a phase reversal/reference contamination phenomenon.14 We can see the wave starting in the medial prefrontal region and rapidly enlarging to cover most of the front of the cortex before resolving (Fig. 4b and c). These slow waves are almost always maximal in the midline frontal–prefrontal region (82%). The mirror image (red) 100 μV positive wave seen posteriorly is probably largely an artifact of the average reference montage.Fig 4 Example of the morphology of ketamine-induced slow waves. (a) Time course of the waves (channel Fz) in relation to ketamine injection (grey vertical line), LOBR (black rectangle) and ROBR (blue line). (b) Close-up view of a single slow wave showing the surrounding theta–gamma oscillations. (c) At the six time points, shown by the vertical red lines, a topological map of the spatial distribution of the instantaneous EEG amplitude is shown.
e injection (grey vertical line), LOBR (black rectangle) and ROBR (blue line). (b) Close-up view of a single slow wave showing the surrounding theta–gamma oscillations. (c) At the six time points, shown by the vertical red lines, a topological map of the spatial distribution of the instantaneous EEG amplitude is shown. They were chosen as: (1) pre-wave, (2) wave-initiation, (3) maximum-wave, (4) resolving wave, (5) end-wave, and (6) post-wave time points. Colour bar is in units of μV. LOBR, loss of behavioural responsiveness; ROBR, recovery of behavioural responsiveness. Fig 4 Discussion Our results showed that a hypnotic dose of ketamine causes atypical large episodic slow waves that occur around or after LOBR, when the effect-site ketamine concentration was more than ∼1.5 μg ml−1. The waves are predominantly medial frontal, and usually consist of large, almost synchronous, hyperpolarisations. These waves were first described by Schwartz and colleagues1 in 1974 (see Fig 1, Fig 2 in their paper), but have been relatively ignored since then because they are somewhat obscured by traditional spectral analysis methods that tend to emphasise the obvious ketamine-induced increases in theta and beta–gamma power. The cortical hyperpolarisation pattern seen with these waves has been confirmed using intracellular recordings in cats given ketamine and an alpha-2 adrenergic agonist.15
mewhat obscured by traditional spectral analysis methods that tend to emphasise the obvious ketamine-induced increases in theta and beta–gamma power. The cortical hyperpolarisation pattern seen with these waves has been confirmed using intracellular recordings in cats given ketamine and an alpha-2 adrenergic agonist.15 In previous experiments with propofol,6 LOBR typically occurred as SWA increased. Mechanistically it is plausible that slow waves cause disruptions in normal cortical function that would interrupt perception. The medial–frontal origin of most of the slow waves is worthy of note. There is some evidence that the medial frontal cortex plays an important role mediating arousal from propofol and volatile agent anaesthesia.16, 17 Maximal SWA may cause unconsciousness via loss of anterior–posterior functional connectivity, or on the basis of functional MRI evidence of thalamocortical isolation and consequent loss of perception and self-awareness.6 It is possible that ketamine reduces consciousness via a similar network-level mechanism. However, the SWA pattern disappeared long before ROBR. This suggests that SWA is capable of disrupting consciousness, but, conversely, the absence of SWA is not sufficient for the return of wakeful connected consciousness. It is likely that the period between loss of the SWA and ROBR was marked by a return of some sort of disconnected consciousness, as manifest by dreaming or hallucinations that are common with ketamine. The effect-site concentration required for LOBR is similar to that found in previous experiments. Schüttler and co-workers2 gave five subjects a larger dose of ketamine (250 mg administered as a rapid infusion over a few minutes) to achieve maximal EEG slowing. They achieved peak serum concentrations of ketamine ∼4 μg ml−1, and using the median EEG frequency found similar response curves (IC50, 2.0 [0.5] μg ml−1) for racemic ketamine. Idvall and colleagues18 found similar concentrations, and also hysteresis, whereby ROBR occurred at lower ketamine concentrations and smaller SWA than for LOBR.
k serum concentrations of ketamine ∼4 μg ml−1, and using the median EEG frequency found similar response curves (IC50, 2.0 [0.5] μg ml−1) for racemic ketamine. Idvall and colleagues18 found similar concentrations, and also hysteresis, whereby ROBR occurred at lower ketamine concentrations and smaller SWA than for LOBR. Flores and co-workers19 modelled the time course of high frequency (∼140 Hz) electrocorticogram oscillations in rats after i.p. ketamine. These oscillations showed an initial increase in power that preceded loss of righting reflex, but they did not report on slow waves. Subsequently, ketamine caused a secondary peak in high frequency oscillations that was maximum around the point of regaining righting reflex and outlasted it by about 45 min. They successfully modelled this phenomenon as the mutually antagonistic effects of excitatory N-methyl-d-aspartate (NMDA) receptor blocking activity and a second inhibitory (non-NMDA receptor mediated) activity of ketamine. Ketamine has a number of molecular targets, but its actions on hyperpolarisation-activated, cyclic nucleotide-gated subtype 1 (HCN1) currents has been linked to its hypnotic and analgesic effects.20 There is increasing evidence that the increase in high frequency power caused by NMDA antagonists is not associated with loss of behavioural response, but is related to psychotomimetic behaviour.21, 22 Insofar as scalp-recorded gamma activity in humans may be correlated with the very high frequency electrocorticogram oscillations in rats, our results were broadly in agreement with these patterns; and that gamma power should not be used to model drug effects related to LOBR. In our 15 subjects, theta power had an inconsistent relationship with behavioural responsiveness, but probably should be included as part of any subsequent large-scale study of PKPD modelling of ketamine's effects on responsiveness.
atterns; and that gamma power should not be used to model drug effects related to LOBR. In our 15 subjects, theta power had an inconsistent relationship with behavioural responsiveness, but probably should be included as part of any subsequent large-scale study of PKPD modelling of ketamine's effects on responsiveness. The generators of theta waves seen with ketamine are not well understood, and it is not clear why the time course is slower in onset and longer in duration than for SWA and beta–gamma power. We speculate that theta involves slower acting molecular targets and mechanisms downstream from fast ion channels. Increased theta resonance may represent network disequilibrium between anterior and posterior regions with impaired information flow.4 This speculative interpretation requires targeted, follow-up investigation (e.g. anatomical source analysis) to advance understanding of such theta oscillations. There are divergent views on the time course of the onset of the actions of ketamine.23 Clinically the time course has been thought comparable with the LOBR seen with propofol and thiopentone.24 A detailed study in sheep that compared EEG changes with a mass balance measure of actual brain drug uptake suggested a long t1/2ke0 of around 120 s, which is similar to that of propofol.25 However, most other studies suggest a more rapid uptake of ketamine to the effect-site. Using the clinical endpoint of LOBR in children, a mean (range) t1/2ke0 of 11 (7–20) s was reported.8 Our EEG-based estimate agrees with a short t1/2ke0 similar to that of methohexitone.26
s, which is similar to that of propofol.25 However, most other studies suggest a more rapid uptake of ketamine to the effect-site. Using the clinical endpoint of LOBR in children, a mean (range) t1/2ke0 of 11 (7–20) s was reported.8 Our EEG-based estimate agrees with a short t1/2ke0 similar to that of methohexitone.26 A limitation to this study is that the ketamine plasma levels were not measured directly; accuracy of the model could only be indirectly inferred from dosing and EEG responses. Similarly, it is unclear whether the very steep dose–response curves are a manifestation of a true pharmacodynamic threshold phenomenon (such as a phase-change transition to unconsciousness), or whether they are an exaggerated pharmacokinetic drug diffusion effect caused by the bolus dose. Another limitation is the fact that we had to remove some channels in order to achieve physiologically reasonable model fitting, especially for the beta–gamma analysis. Although not ideal, this does accurately reflect the real-world problems of separating muscle artifact from EEG in a significant proportion of ketamine patients.
mitation is the fact that we had to remove some channels in order to achieve physiologically reasonable model fitting, especially for the beta–gamma analysis. Although not ideal, this does accurately reflect the real-world problems of separating muscle artifact from EEG in a significant proportion of ketamine patients. Conclusions We found that ketamine induced slow wave activity in the electroencephalogram at brain concentrations above ∼1.5 μg ml−1, and this was associated with loss of behavioural responsiveness. However, loss of SWA did not correspond to recovery of behavioural responses. As measured by SWA, the time for ketamine diffusion into the brain effect-site (23 s) is much faster than that reported for propofol. The slow waves seen with ketamine are quite different in morphology to those seen with propofol and sevoflurane, and are predominantly medio-frontal in distribution, probably reflecting hyperpolarisations in the medial default mode network (anterior cingulate cortex). Authors' contributions Study design: all authors. Analysis structure: all authors. Interpretation: all authors. Writing of the manuscript: all authors. Approved of the final version of the manuscript: all authors. Matlab programming: JS, RP. Declaration of interest CEW is named on a patent for the use of slow wave activity as a measure of depth of anaesthesia.
Authors' contributions Study design: all authors. Analysis structure: all authors. Interpretation: all authors. Writing of the manuscript: all authors. Approved of the final version of the manuscript: all authors. Matlab programming: JS, RP. Declaration of interest CEW is named on a patent for the use of slow wave activity as a measure of depth of anaesthesia. Funding LUMINOUS Project and the MRC Development Pathway funding Scheme (award ref MR/R006423/1) to CEW. The LUMINOUS project: European Union's Horizon 2020 research and innovation programme H2020-FETOPEN-2014-2015- RIA under agreement No. 686764). The Wellcome Centre for Integrative Neuroimaging: core funding from Wellcome Trust (203139/Z/16/Z). U.S. National Institutes of Health (T32GM103730 and R01GM111293) and the Department of Anesthesiology, University of Michigan Medical School, Ann Arbor, MI, USA, to PV. University of Auckland, School of Medicine E G Shrimpton Fund, and the James S. McDonnell Foundation (Grant #220023046) to RP. Appendix A Supplementary data The following is the Supplementary data to this article:Multimedia component 1 Multimedia component 1 Acknowledgements The authors thank Amy McKinney for leading research coordination efforts and Bryan Kunkler for assistance with EEG data acquisition. Appendix A Supplementary data to this article can be found online at https://doi.org/10.1016/j.bja.2019.07.021.
Editor's key points • Noble gases have shown neuroprotective effects in experimental models of cerebral ischaemia. • An in vitro model of cerebral ischaemia was used to compare the neuroprotective efficacy of the full series of noble gases. • Whereas xenon and argon were similarly neuroprotective, helium, neon, and krypton were without a protective effect. • Reversal of neuroprotection by xenon, but not by argon, by elevated glycine suggests distinct protective mechanisms. • Further translational studies to evaluate these two noble gases as neuroprotectants are warranted by these findings. Neurological injuries resulting from hypoxia–ischaemia are leading causes of morbidity and mortality worldwide.1, 2, 3 Hypoxic–ischaemic brain injury has a variety of aetiologies including stroke, cardiac arrest, neonatal hypoxic–ischaemic encephalopathy (HIE), drowning and exposure to asphyxiant gases and carbon monoxide. Many who survive a hypoxic–ischaemic brain injury have persisting disability, with long-term care and rehabilitation costs.4 Treatment options are limited to thrombolytic drugs and clot removal for ischaemic stroke, and therapeutic cooling (or hypothermia) for cardiac arrest and neonatal HIE. Currently, there are no clinically proven treatments specifically targeted at preventing or limiting neuronal death resulting from ischaemia.
n costs.4 Treatment options are limited to thrombolytic drugs and clot removal for ischaemic stroke, and therapeutic cooling (or hypothermia) for cardiac arrest and neonatal HIE. Currently, there are no clinically proven treatments specifically targeted at preventing or limiting neuronal death resulting from ischaemia. There is a need to develop neuroprotective treatments for hypoxic–ischaemic brain injury. Currently there is interest in the noble gases as novel treatments for ischaemic and traumatic brain injury.5, 6, 7, 8, 9 Attention has focused on xenon, which has already undergone clinical trials for HIE10, 11, 12 and out-of-hospital cardiac arrest,13, 14 but there is also interest in the use of argon and helium, which have been evaluated in in vitro and in vivo models.15, 16, 17, 18, 19, 20 Neuroprotection by particular noble gases has been reported under different conditions.7, 21, 22, 23, 24, 25 Few studies, however, have evaluated neuroprotection by krypton or neon,26, 27 and investigation of the whole series of noble gases in hypoxic–ischaemic brain injury under the same conditions has been limited to dissociated cell cultures.27 We report the neuroprotective efficacy of helium, neon, argon, krypton, and xenon under identical conditions using organotypic hippocampal brain-slice cultures subjected to oxygen-glucose deprivation (OGD), an experimental model of cerebral ischaemia. We tested the hypothesis that the N-methyl-d-aspartate (NMDA) receptor glycine site is involved in noble gas neuroprotection against hypoxic–ischaemic brain injury in vitro.
ons using organotypic hippocampal brain-slice cultures subjected to oxygen-glucose deprivation (OGD), an experimental model of cerebral ischaemia. We tested the hypothesis that the N-methyl-d-aspartate (NMDA) receptor glycine site is involved in noble gas neuroprotection against hypoxic–ischaemic brain injury in vitro. Methods Unless otherwise stated, chemicals were obtained from Sigma-Aldrich Ltd (Gillingham, Dorset, UK). All gases were obtained from BOC Ltd (Guildford, Surrey, UK); pure noble gases were N5.0 grade (99.999%).
ons using organotypic hippocampal brain-slice cultures subjected to oxygen-glucose deprivation (OGD), an experimental model of cerebral ischaemia. We tested the hypothesis that the N-methyl-d-aspartate (NMDA) receptor glycine site is involved in noble gas neuroprotection against hypoxic–ischaemic brain injury in vitro. Methods Unless otherwise stated, chemicals were obtained from Sigma-Aldrich Ltd (Gillingham, Dorset, UK). All gases were obtained from BOC Ltd (Guildford, Surrey, UK); pure noble gases were N5.0 grade (99.999%). Hippocampal organotypic slices Experiments were performed in compliance with the Animal Welfare and Ethical Review Body of Imperial College London and the Animals (Scientific Procedures) Act of 1986. Animals (pups and their dams) were housed in individually ventilated cages in a pathogen-free facility in a 12:12 h light–dark cycle (7:00 am–7:00 pm light) at 22°C with ad libitum access to food and water. Animals were checked at least once daily. Organotypic hippocampal slice cultures were prepared as described24, 26, 28, 29 from male and female 7-day-old C57BL/6 mouse pups (Harlan Ltd, Bicester, Oxfordshire, UK). Briefly, after euthanasia, brains were removed and placed in ice-cold ‘preparation’ medium that contained Gey's balanced salt solution, 33 mM d-glucose (Fisher Scientific, Loughborough, Leicestershire, UK) and 1% antibiotic–antimycotic suspension. The hippocampi were removed, and 400 μm thick transverse slices were prepared using a McIllwain tissue chopper. Slices were transferred into ice-cold preparation medium, gently separated and then placed on tissue culture inserts (Millicell-CM; Millipore Corporation, Carrigtwohill, Co. Cork, Ireland) that were inserted into a six-well tissue culture plate. The wells contained ‘growth’ medium consisting of 50% (v:v) Minimal Essential Medium Eagle, 25% Hank's balanced salt solution, 25% inactivated horse serum, 2 mM l-glutamine, 32 mM d-glucose, and 1% antibiotic–antimycotic suspension. Slices were incubated at 37°C in a 95% air:5% CO2 humidified atmosphere. The growth medium was changed every 3 days. Experiments were carried out after 14 days in culture. Cell culture inserts containing four to seven slices were randomly assigned to sham, OGD control, or OGD noble gas treatment groups.
tic suspension. Slices were incubated at 37°C in a 95% air:5% CO2 humidified atmosphere. The growth medium was changed every 3 days. Experiments were carried out after 14 days in culture. Cell culture inserts containing four to seven slices were randomly assigned to sham, OGD control, or OGD noble gas treatment groups. Oxygen-glucose deprivation and hyperbaric gas chamber The growth medium was changed to serum-free ‘experimental’ medium consisting of 75% Minimal Essential Medium Eagle, 25% Hank's balanced salt solution, 2 mM l-glutamine, 33 mM d-glucose, 1% antibiotic–antimycotic suspension, and 4.5 μM propidium iodide (PI). One hour after transfer to experimental media, slices were imaged to assess viability before OGD. Typically, slices exhibited very little PI fluorescence, an indicator of healthy slices. A small number of slices were excluded from further analysis because they failed to meet objective viability criteria at this time point (t=0); either there were regions of dense staining, or there were more than 20 pixels at intensity levels above 80, or tissue fragments were visible, indicating compromised viability, presumably as a result of mechanical damage during slice preparation. Immediately after initial imaging, experimental medium was exchanged for ‘OGD medium’, 120 mM NaCl, 5 mM KCl, 1.25 mM NaH2PO4, 2 mM MgSO4, 2 mM CaCl2, 25 mM NaHCO3, 10 mM sucrose, 20 mM HEPES, pH 7.25 or ‘sham medium,’ which had the same composition, except that sucrose was replaced with 10 mM d-glucose. OGD medium was deoxygenated before use by bubbling for 45 min at 50 ml min−1 with 95% N2:5% CO2, in a Dreschel bottle using a fine-sintered glass bubbler and filter-sterilised using a 0.2 μm filter. Sham medium was treated in the same way except it was bubbled with 20% O2:75% N2:5% CO2. After solution exchange, culture dishes were transferred to a small chamber (Fig. 1a) that contained a high-speed fan for rapid gas mixing. The chamber was housed in an incubator at 37°C. The chamber (gas volume 0.925 L) was flushed with humidified gas (95% N2:5% CO2 or 20% O2:75% N2:5% CO2) for 5 min at 5 L min−1 ensuring better than 99.99% gas replacement. After flushing, the chamber was sealed for a set period of 30 min, constituting the duration of OGD (or sham treatment).Fig 1 (a) Diagram of the chamber used for oxygen-glucose deprivation (OGD) and gas treatment. Organotypic hippocampal slice cultures in six-well cell-culture dishes were placed in the chamber.
cement. After flushing, the chamber was sealed for a set period of 30 min, constituting the duration of OGD (or sham treatment).Fig 1 (a) Diagram of the chamber used for oxygen-glucose deprivation (OGD) and gas treatment. Organotypic hippocampal slice cultures in six-well cell-culture dishes were placed in the chamber. A small fan (shown in black) ensured mixing of the gases. (b) Schematic showing the experimental timeline. (c) Typical propidium iodide fluorescence images at of slices (i) sham, (ii) OGD, and (iii) maximal injury. Scale bars=500 μm. (d) Intensity histogram of slices from sham (black) and OGD (red) groups. (e) Quantification of injury in sham (white bar), OGD (brown bar), and maximal injury (dark red bar) slices at 24 h after injury or sham procedure. Slices were exposed to control gas (75% N2:20% O2:5% CO2) with 0.5 atm helium for 24 h after OGD or sham procedure. The intensity histograms are the mean of 10 (sham) and 25 (OGD) slices. Pixel numbers have been normalised to the median of the OGD slices. Bars are median values, error bars are the 95% confidence interval. n=191, sham; n=326, OGD, n=125 max injury. ****P<0.0001, Kruskal–Wallis test with Dunn's correction for multiple comparisons. Fig 1
A small fan (shown in black) ensured mixing of the gases. (b) Schematic showing the experimental timeline. (c) Typical propidium iodide fluorescence images at of slices (i) sham, (ii) OGD, and (iii) maximal injury. Scale bars=500 μm. (d) Intensity histogram of slices from sham (black) and OGD (red) groups. (e) Quantification of injury in sham (white bar), OGD (brown bar), and maximal injury (dark red bar) slices at 24 h after injury or sham procedure. Slices were exposed to control gas (75% N2:20% O2:5% CO2) with 0.5 atm helium for 24 h after OGD or sham procedure. The intensity histograms are the mean of 10 (sham) and 25 (OGD) slices. Pixel numbers have been normalised to the median of the OGD slices. Bars are median values, error bars are the 95% confidence interval. n=191, sham; n=326, OGD, n=125 max injury. ****P<0.0001, Kruskal–Wallis test with Dunn's correction for multiple comparisons. Fig 1 After the period of OGD, slices were removed from the chamber and medium was replaced with experimental medium (in experiments with added glycine [100 μM], this was added for the first time at this stage). Slices were returned to the chamber which was flushed with 20% O2:75% N2:5% CO2 as before and sealed. In the helium experiments at 1.0 atm the chamber was flushed with 20% O2:75% He:5% CO2. In experiments with xenon, krypton, argon, and neon, after flushing with 20% O2:75% N2:5% CO2 an additional 0.5 atm of noble gas was added after sealing the chamber, with helium used as control for the effects of pressure. Treatment with noble gases was started 10 min after OGD. For all gas mixtures (except during OGD), the partial pressures of oxygen and carbon dioxide were fixed at 0.2 and 0.05 atm, respectively. During OGD, the partial pressures were 0.95 atm nitrogen and 0.05 atm carbon dioxide. The chamber fan was left on for 5 min to achieve mixing of gases. After 24 h in the chamber, slices were imaged using a fluorescent microscope (see section ‘Quantifying cell injury’). The experimental timeline is shown in Fig. 1b.
ly. During OGD, the partial pressures were 0.95 atm nitrogen and 0.05 atm carbon dioxide. The chamber fan was left on for 5 min to achieve mixing of gases. After 24 h in the chamber, slices were imaged using a fluorescent microscope (see section ‘Quantifying cell injury’). The experimental timeline is shown in Fig. 1b. Quantifying cell injury PI only enters cells with compromised cellular membranes and becomes fluorescent after binding to nucleic acids, allowing quantification of cell injury.30, 31, 32 The PI assay does not distinguish between different cell types or grey and white matter, as would be possible with histopathology, but PI has the advantage in that real-time quantification of injury can take place in the same slices at different time points (in this case, the viability assessment at t=0 h before injury and at t=24 h after OGD or sham procedure). An epifluorescence microscope (Nikon Eclipse 80; Kingston upon Thames, Surrey, UK), with a low-power (2×) objective was used to quantify PI fluorescence. A digital video camera and software (Micropublisher 3.3 RTV camera and QCapture Pro software; Qimaging Inc, Surrey, BC, Canada) were used to capture the images. Image intensity analysis of the red channel was performed using ImageJ software,33 with the distribution of intensities plotted as a histogram over 256 intensity levels. Uninjured sham slices under control conditions, incubated in the chamber for 24 h at 37°C with 20% O2:75% N2:5% CO2, showed little PI fluorescence (Fig. 1c[i]) compared with OGD injured slices (Fig. 1c[ii]) that exhibited bright PI fluorescence. In order to determine the relative magnitude of the OGD injury we determined maximal cell death by incubating some slices in 70% ethanol overnight at 4°C (Fig. 1c[iii]). To quantify the injury we integrated the number of pixels above a threshold of 100, which provides a robust quantitative measurement of cell injury (Fig. 1d).24 Absolute pixel values were normalised to the median value of the control OGD slices (Fig. 1e).
ting some slices in 70% ethanol overnight at 4°C (Fig. 1c[iii]). To quantify the injury we integrated the number of pixels above a threshold of 100, which provides a robust quantitative measurement of cell injury (Fig. 1d).24 Absolute pixel values were normalised to the median value of the control OGD slices (Fig. 1e). Statistical analysis Data were tested for normality using the Shapiro–Wilk test and found to be non-normal. Results are shown as median values with error bars representing 95% confidence intervals. We assessed significance using the Kruskal–Wallis test with Dunn's correction for multiple comparisons. A P-value of <0.05 was taken to indicate a significant difference between groups. Statistical tests were performed using GraphPad Prism v 7.04 (GraphPad Inc., La Jolla, CA, USA). Results Oxygen-glucose deprivation results in sub-maximal injury To determine the relative intensity of our OGD injury we compared uninjured sham slices with slices subjected to OGD and slices subjected to maximal injury. Compared with uninjured shams, slices subjected to OGD exhibited a bright PI fluorescence, which was sub-maximal (Fig. 1c and d). Injury in the OGD slices was greater than sham and less than maximal injury (Fig. 1e).
we compared uninjured sham slices with slices subjected to OGD and slices subjected to maximal injury. Compared with uninjured shams, slices subjected to OGD exhibited a bright PI fluorescence, which was sub-maximal (Fig. 1c and d). Injury in the OGD slices was greater than sham and less than maximal injury (Fig. 1e). Helium has no effect on hypoxic–ischaemic injury We determined the effect of 1.0 atm helium (75% He:20% O2:5% CO2) on sham and OGD slices (Fig. 2). There was no significant difference between sham slices with or without helium. Injury was low in sham slices, with median values 4.9% and 8.3% of the median value of the control OGD in the absence and presence of helium, respectively. After OGD, injury developed significantly (p < 0.0001) at 24 h compared with shams, in both the absence and presence of helium. However, there was no significant difference between OGD slices treated with helium and control OGD slices treated with 75% N2:20% O2:5% CO2.Fig 2 Helium at atmospheric pressure has no effect after injury or sham procedure. Slices were exposed to either control gas (75% N2:20% O2:5% CO2) or helium (75% He:20% O2:5% CO2) for 24 h after OGD or sham procedure. Pixel numbers have been normalised to the median of the control OGD slices. Bars are median values, error bars are the 95% confidence interval. n=31, sham; n=35, sham helium; n=54, OGD; n=46, helium OGD. ****p<0.0001 compared with OGD, Kruskal–Wallis test with Dunn's correction for multiple comparisons. Ns, not significant; OGD, oxygen-glucose deprivation. Fig 2
Helium has no effect on hypoxic–ischaemic injury We determined the effect of 1.0 atm helium (75% He:20% O2:5% CO2) on sham and OGD slices (Fig. 2). There was no significant difference between sham slices with or without helium. Injury was low in sham slices, with median values 4.9% and 8.3% of the median value of the control OGD in the absence and presence of helium, respectively. After OGD, injury developed significantly (p < 0.0001) at 24 h compared with shams, in both the absence and presence of helium. However, there was no significant difference between OGD slices treated with helium and control OGD slices treated with 75% N2:20% O2:5% CO2.Fig 2 Helium at atmospheric pressure has no effect after injury or sham procedure. Slices were exposed to either control gas (75% N2:20% O2:5% CO2) or helium (75% He:20% O2:5% CO2) for 24 h after OGD or sham procedure. Pixel numbers have been normalised to the median of the control OGD slices. Bars are median values, error bars are the 95% confidence interval. n=31, sham; n=35, sham helium; n=54, OGD; n=46, helium OGD. ****p<0.0001 compared with OGD, Kruskal–Wallis test with Dunn's correction for multiple comparisons. Ns, not significant; OGD, oxygen-glucose deprivation. Fig 2 Xenon and argon prevent hypoxic–ischaemic injury, whereas krypton and neon have no effect As helium was without effect, we investigated the effect of 0.5 atm of the noble gases xenon, krypton, argon, and neon on OGD injury (Fig. 3). As these experiments used mild hyperbaric conditions, we used 0.5 atm helium in the control OGD to control for any effects of pressure. Sham slices exhibited very little injury (Fig. 3a[i]) compared with control OGD (Fig. 3a[ii]), whereas treatment with xenon (Fig. 3a[iii]) or argon (Fig. 3a[iv]) after OGD reduced injury. Xenon and argon were equally effective at reducing OGD injury, both reducing injury significantly (p < 0.0001) by 96% (Fig. 3b). The OGD slices treated with xenon and argon were not significantly different to each other or to the uninjured sham group (Fig. 3b). Thus both of these noble gases can prevent injury development in this in vitro model. We found that krypton and neon were without significant effect on OGD injury (Fig. 3b).Fig 3 Xenon and argon prevent after OGD injury whereas other noble gases have no protective effect. (a) Typical propidium iodide fluorescence images of slices (i) sham, (ii) OGD + 0.5 atm helium, (iii) OGD + 0.5 atm xenon, (iv) OGD + 0.5 atm argon. Scale bars are 500 μm. (b) Quantification of injury at 24 h, in sham (white bar), OGD + 0.5 atm xenon (red bar), OGD + 0.5 atm argon (blue bar), OGD + 0.5 atm neon (green bar), OGD + 0.5 atm krypton (purple bar), OGD control (brown bar). OGD control and sham slices were exposed to 0.5 atm helium. All slices were also exposed to 1.0 atm control gas (75% N2:20% O2:5% CO2) with total partial pressure 1.5 atm. Pixel numbers have been normalised to the median of the OGD + 0.5 atm helium slices. Bars are median values, error bars are the 95% confidence interval. n=191, sham; n=95, OGD + xenon; n=52, OGD + argon; n=89, OGD + neon; n=108, OGD + krypton; n=326, OGD control. ****P<0.0001 compared with OGD, Kruskal–Wallis test with Dunn's correction for multiple comparisons. Ns, not significant; OGD, oxygen-glucose deprivation.
an values, error bars are the 95% confidence interval. n=191, sham; n=95, OGD + xenon; n=52, OGD + argon; n=89, OGD + neon; n=108, OGD + krypton; n=326, OGD control. ****P<0.0001 compared with OGD, Kruskal–Wallis test with Dunn's correction for multiple comparisons. Ns, not significant; OGD, oxygen-glucose deprivation. Fig 3
an values, error bars are the 95% confidence interval. n=191, sham; n=95, OGD + xenon; n=52, OGD + argon; n=89, OGD + neon; n=108, OGD + krypton; n=326, OGD control. ****P<0.0001 compared with OGD, Kruskal–Wallis test with Dunn's correction for multiple comparisons. Ns, not significant; OGD, oxygen-glucose deprivation. Fig 3 Glycine reverses the protective effect of xenon, but not of argon, against hypoxic–ischaemic injury In order to determine the role of the NMDA receptor in the protective effect of xenon and argon, we investigated the effect of glycine on neuroprotection by these noble gases. The addition of 100 μM glycine had no significant effect on control OGD injury with helium (Fig. 4). The protective effect of argon was unaffected by addition of glycine, with a 91% reduction with glycine compared with 96% reduction without glycine. In contrast, addition of glycine completely reversed the protective effect of xenon, with a 96% reduction in injury without and a 0.4% reduction in injury with added glycine. This is consistent with xenon neuroprotection being mediated by the NMDA receptor glycine binding site. These findings indicate that xenon and argon provide neuroprotection against hypoxic–ischaemic injury by different mechanisms.Fig 4 Neuroprotection by xenon but not argon is reversed by elevated glycine. The addition of 100 μM glycine has no significant effect on the control OGD injury, but reverses the protective effect of 0.5 atm xenon. Quantification of injury at 24 h in sham (white bar), OGD + 0.5 atm xenon (red bar), OGD + 0.5 atm xenon + glycine (red hatched bar), OGD + 0.5 atm argon (blue bar), OGD + 0.5 atm argon + glycine (blue hatched bar), OGD control (brown bar), OGD control + glycine (brown hatched bar). OGD control and sham slices were exposed to 0.5 atm helium. All slices were also exposed to 1.0 atm control gas (75% N2:20% O2:5% CO2) with total partial pressure 1.5 atm. Pixel numbers have been normalised to the median of the OGD control slices. Bars are median values, error bars are the 95% confidence interval. n=191, sham; n=95, OGD + xenon; n=91, OGD + xenon + glycine; n=52, OGD + argon; n=48, OGD + argon + glycine; n=326, OGD control; n=69, OGD control + glycine. ****P<0.0001 compared with OGD, Kruskal–Wallis test with Dunn's correction for multiple comparisons. Ns, not significant; OGD, oxygen-glucose deprivation; gly, glycine.
sham; n=95, OGD + xenon; n=91, OGD + xenon + glycine; n=52, OGD + argon; n=48, OGD + argon + glycine; n=326, OGD control; n=69, OGD control + glycine. ****P<0.0001 compared with OGD, Kruskal–Wallis test with Dunn's correction for multiple comparisons. Ns, not significant; OGD, oxygen-glucose deprivation; gly, glycine. Fig 4
sham; n=95, OGD + xenon; n=91, OGD + xenon + glycine; n=52, OGD + argon; n=48, OGD + argon + glycine; n=326, OGD control; n=69, OGD control + glycine. ****P<0.0001 compared with OGD, Kruskal–Wallis test with Dunn's correction for multiple comparisons. Ns, not significant; OGD, oxygen-glucose deprivation; gly, glycine. Fig 4 Discussion Oxygen-glucose deprivation model Organotypic hippocampal slice cultures (OHSCs) were subjected to OGD, with injury quantified by PI fluorescence. This preparation retains a variety of cell types (e.g. different types of neurones and glia) with cellular organisation and synaptic connectivity similar to in vivo,34, 35 and is widely used as an intermediate between dissociated cell cultures and whole-animal models.20, 24, 26, 28, 36, 37, 38, 39, 40 We chose 30 min as the duration of OGD because we previously showed that this produced a reliable and robust injury.24 OGD results in a diffuse global injury, and the injury produced by 30 min OGD was sub-maximal. The OHSC model we used has advantages and limitations. An in vitro model allows us to control the slice environment. Organotypic brain slice cultures exposed to OGD are a widely used model of cerebral hypoxia–ischaemia,41, 42, 43 and in vitro OGD causes disruption of cellular function that is similar to hypoxia–ischaemia in vivo.44, 45, 46 We measured cell death and neuroprotection in the hippocampal slice as a whole in order to avoid subjectivity associated with precisely defining the boundaries of CA1, CA3, and dentate gyrus. In humans with ischaemic brain injury, hippocampal sub-regions exhibit differential sensitivity, with CA1 being particularly vulnerable.47 In our study, we observed qualitatively that the CA1 region appeared more sensitive to ischaemic injury, in agreement with clinical data and previous in vitro studies,24, 48 and this likely reflects the density of NMDA receptors. Nevertheless, xenon and argon appeared to reduce injury to a similar degree in different hippocampal areas as observed for other neuroprotective drugs.24, 26, 42, 43, 49
tive to ischaemic injury, in agreement with clinical data and previous in vitro studies,24, 48 and this likely reflects the density of NMDA receptors. Nevertheless, xenon and argon appeared to reduce injury to a similar degree in different hippocampal areas as observed for other neuroprotective drugs.24, 26, 42, 43, 49 Lack of effect of helium There are few studies of helium as a neuroprotectant, and these have produced contradictory results. Helium was found to be neuroprotective in a rat model of ischaemic brain injury; however, this was shown not to be a pharmacological effect but because of hypothermia resulting from breathing helium at room temperature (because of the high thermal conductivity of helium).5, 21, 50 In an in vitro OGD model in isolated cell cultures, a detrimental effect of helium was reported.27 We previously showed that mild hyperbaric helium (0.5 atm partial pressure at 1.5 atm) had no effect on OGD injury in hippocampal slices.24 In the current study, we investigated the effects of 0.75 atm helium under normobaric conditions (1 atm) and found that normobaric helium had no effect on uninjured sham slices or on OGD slices after 24 h of treatment. Our treatment is given inside a temperature-controlled incubator at 37°C; hence, we can exclude the effect of hypothermia as found in the in vivo studies. In our OHSC model at 37°C, helium is devoid of any observable effect against OGD.
um had no effect on uninjured sham slices or on OGD slices after 24 h of treatment. Our treatment is given inside a temperature-controlled incubator at 37°C; hence, we can exclude the effect of hypothermia as found in the in vivo studies. In our OHSC model at 37°C, helium is devoid of any observable effect against OGD. Xenon and argon are equally effective as neuroprotectants, whereas other noble gases are without effect In contrast to helium, both xenon and argon resulted in significant neuroprotection. Both gases prevented injury development; OGD slices treated with 0.5 atm xenon or 0.5 atm argon were not significantly different to uninjured sham slices. Interestingly, we found that xenon and argon were equally effective in the OGD model, in contrast to an in vitro model of traumatic brain injury where argon was less effective than xenon.26 The reasons why argon appears to be as effective as xenon in this ischaemic injury model are not clear, but this suggests that different secondary injury mechanisms may be involved in ischaemic and traumatic brain injury. The comparable efficacy of xenon and argon we observe in vitro does not necessarily mean that similar long-term functional improvements will be observed for these gases in vivo. The relative efficacy of xenon and argon on clinically relevant long-term functional outcomes after hypoxia–ischaemia in vivo remains to be determined. Krypton and neon were without any protective effect, consistent with findings on isolated cell cultures subjected to OGD and OHSCs that had experienced a traumatic insult.26, 27
ive efficacy of xenon and argon on clinically relevant long-term functional outcomes after hypoxia–ischaemia in vivo remains to be determined. Krypton and neon were without any protective effect, consistent with findings on isolated cell cultures subjected to OGD and OHSCs that had experienced a traumatic insult.26, 27 Neuroprotection by xenon, but not argon, involves NMDA receptor inhibition at the glycine site Xenon inhibits the NMDA receptor by competing for the binding of the co-agonist glycine, and xenon inhibition can be prevented by elevating glycine concentrations.51, 52 In the current study we found that addition of glycine had no effect on the control OGD injury. The simplest explanation for this observation is that the concentration of endogenous glycine is just below saturating on the concentration–effect curve. However, the neuroprotective effect of xenon was completely reversed by the addition of glycine, consistent with inhibition of the NMDA receptor glycine site mediating xenon's protective effect. In contrast, addition of glycine had no effect on neuroprotection by argon, indicating that argon acts via a different mechanism. The reversal of neuroprotection by xenon but not argon with added glycine is consistent with what we observed in an in vitro model of traumatic brain injury.26 Xenon acts at other targets, such as the two pore-domain potassium channel TREK-1 and the ATP-sensitive potassium (K-ATP) channel, but has no effect on N-type calcium channels53, 54, 55; however, our findings indicate that inhibition of the NMDA receptor glycine site is likely to play a major role in the neuroprotective effect of xenon.24, 26, 51, 52, 56, 57 The targets mediating the neuroprotective effect of argon are less clear; we have shown that argon does not inhibit NMDA receptors or activate TREK-1 channels.26 Nevertheless, other in vitro and in vivo studies with argon have identified activation of signalling pathways involving MEK-ERK 1/2 and PI3K/AKT, with up-regulation of heme-oxygenase-1.17, 58, 59 A recent study has also identified Nrf2 and the mammalian target of rapamycin (mTOR) signalling pathway as targets for argon.60 Although these studies clearly identify changes in these signalling pathways after argon treatment, it is not clear whether argon is acting on the upstream targets of these pathways.
7, 58, 59 A recent study has also identified Nrf2 and the mammalian target of rapamycin (mTOR) signalling pathway as targets for argon.60 Although these studies clearly identify changes in these signalling pathways after argon treatment, it is not clear whether argon is acting on the upstream targets of these pathways. Clinical relevance of our findings There is currently much interest in the clinical use of noble gases as neuroprotectants in ischaemic brain injuries. Xenon is licenced for use as a general anaesthetic and has already undergone clinical trials in neonatal HIE, ischaemic brain damage after cardiac arrest, coronary artery bypass graft surgery, and orthopaedic surgery in the elderly.10, 11, 13, 14, 61, 62, 63 The most recent study in ischaemic brain injury in cardiac arrest patients found reduced early white-matter damage in the xenon-treated group,14 and a larger multi-centre study is underway. Relative to the other inert gases, xenon is more expensive and is used in closed or semi-closed circuits to conserve gas. There has been interest in evaluating other less expensive noble gases as neuroprotectants that could be given in open circuits. Helium and oxygen mixtures have been used medically for respiratory conditions such as asthma and chronic obstructive pulmonary disease (COPD) in adults and bronchiolitis and croup in children, but systematic reviews conclude that the currently available evidence does not support its use in these conditions.64, 65, 66, 67 Our finding that helium has no effect on hypoxic–ischaemic injury in vitro is consistent with a lack of pharmacological effect of helium at normobaric pressures. The finding that helium has neuroprotective properties in animal models of hypoxia–ischaemia via the physical mechanism of inducing cooling5, 21 is an interesting observation, but more straightforward and controllable techniques for therapeutic cooling are available. Our finding that argon is neuroprotective agrees with a large body of in vitro and in vivo evidence.7, 18, 68 Interestingly, in this OGD model we found 0.5 atm argon to be as effective as 0.5 atm xenon, in contrast to our earlier work with an in vitro model of traumatic brain injury where argon was less effective than xenon.
nding that argon is neuroprotective agrees with a large body of in vitro and in vivo evidence.7, 18, 68 Interestingly, in this OGD model we found 0.5 atm argon to be as effective as 0.5 atm xenon, in contrast to our earlier work with an in vitro model of traumatic brain injury where argon was less effective than xenon. Argon has been shown not to affect cerebral circulation in humans,69 and there are proposals to evaluate argon as a neuroprotectant in patients.6 One obstacle, perhaps, to its clinical use is an unambiguous identification of its mode of action but, on the other hand, it is the most abundant of the noble gases and the cheapest to produce. On a positive note, the fact that argon and xenon do not act by the same mechanism means that combinations of these gases may have a synergistic effect, or that argon and xenon are effective treatments for different forms of ischaemic neurological injury. The observation that in a mouse model of traumatic brain injury, xenon is able to prevent development of very late-onset traumatic brain injury-related memory deficits with reduced white matter loss in the corpus callosum is of great interest,9 but it remains to be seen if this will translate into similar findings in humans. In conclusion our findings that argon and xenon are equally effective neuroprotective in hypoxic–ischaemic injury but act via different mechanisms will prompt further translational studies to evaluate these two noble gases as neuroprotectants, either singly or in combinations. Authors' contributions Study design/planning: RD, KH, ILW. Study conduct: MK, KH.
In conclusion our findings that argon and xenon are equally effective neuroprotective in hypoxic–ischaemic injury but act via different mechanisms will prompt further translational studies to evaluate these two noble gases as neuroprotectants, either singly or in combinations. Authors' contributions Study design/planning: RD, KH, ILW. Study conduct: MK, KH. Data analysis: RD, CJE, MK, KH. Drafting of paper: RD, CJE. Revision of and responsibility for paper contents: all authors. Declarations of interest NF has disclosed being a named inventor on a number of patents relating to the use of xenon as a neuroprotectant. He has a financial interest in the use of xenon as a neuroprotectant. The remaining authors have disclosed that they do not have any potential conflicts of interest. Funding St Peter's Hospital, Chertsey, UK. Medical Research Council, London, UK (MR/N027736/1; MC_PC_13064). The Gas Safety Trust, London, UK. British Journal of Anaesthesia/Royal College of Anaesthetists, London UK. Royal Centre for Defence Medicine, Birmingham, UK, Royal British Legion Centre for Blast Injury Studies, Imperial College London, UK. The European Society for Anaesthesiology. MK was recipient of a Rector's PhD Studentship from Imperial College London. KH was the recipient of a PhD studentship from the Westminster Medical School Research Trust, London, UK. We acknowledge funding from The Royal British Legion.
ury Studies, Imperial College London, UK. The European Society for Anaesthesiology. MK was recipient of a Rector's PhD Studentship from Imperial College London. KH was the recipient of a PhD studentship from the Westminster Medical School Research Trust, London, UK. We acknowledge funding from The Royal British Legion. Acknowledgements We thank Perrine Pluchon and Ala Mejaddam for assistance with preliminary experiments, Raquel Yustos, research technician, Department of Life Sciences, Imperial College London, for technical support, Neal Powell of the Department of Physics, Imperial College London, for artwork, Paul Brown, workshop manager and Steve Nelson, workshop technician, Department of Physics, Imperial College London, for building the inert gas chamber.
The use of cell salvage in obstetrics remains controversial. There is a concern of potential contamination of the salvaged blood with amniotic fluid (AF), which may cause AF embolus (anaphylactoid syndrome of pregnancy),1 and the risk of alloimmunization of the mother with any fetal red blood cells aspirated. Previous studies investigated the use of cell salvage in obstetrics and focused on the feasibility of the washing process in removing potential contaminants. After the use of a leucodepletion filter in cancer surgery, these filters were introduced into obstetric practice.1 A later study using a newer advanced model demonstrated complete removal or a further reduction in contaminants by these filters.2 Along with these results and later reports, the National Institute of Health and Clinical Excellence (NICE) published guidelines in 2005 suggesting that these filters can be considered for use in obstetrics.3 A common assumption was that two suction devices should be used, with as much AF aspirated to waste before any lost blood is collected.14 Although current evidence supports the use of cell salvage in obstetrics, a recent review suggested that more clinical data are required to add to this evidence and make the use of cell salvage more commonplace in obstetrics.4
A common assumption was that two suction devices should be used, with as much AF aspirated to waste before any lost blood is collected.14 Although current evidence supports the use of cell salvage in obstetrics, a recent review suggested that more clinical data are required to add to this evidence and make the use of cell salvage more commonplace in obstetrics.4 The aim of the study was to add results to the current clinical evidence to support the use of cell salvage. We measured AF and fetal red cell contamination, and residual heparin levels of washed salvaged maternal blood when using either one or two suction devices, filtering the washed product through a leucodepletion filter for cell salvage (RS1 VAE PALL® Medical). Methods Information to perform power analysis and produce a well-designed prospective study is lacking due to the limited data available.4 This study was set up as an exploratory study to investigate the safety and feasibility of cell salvage in obstetrics. Ethics Committee opinion was sought, but approval was not required. The sample size was dependent on the number of patients accessible over a set time period. Written informed consent was obtained from 34 patients undergoing elective lower segment Caesarean section at the Royal Cornwall Hospital over a 4 month period. Patients were alternatively assigned to one of two groups before surgery. Group 1 (n=17) involved only one suction where all AF and blood was collected into the cell salvage machine, with Group 2 (n=17) using a second suction to aspirate as much AF to waste before any lost blood was collected.
al over a 4 month period. Patients were alternatively assigned to one of two groups before surgery. Group 1 (n=17) involved only one suction where all AF and blood was collected into the cell salvage machine, with Group 2 (n=17) using a second suction to aspirate as much AF to waste before any lost blood was collected. All patients had a spinal anaesthetic apart from one who had an epidural top up; all spinals were performed using a Whitacre 25 gauge needle, with hyperbaric bupivacaine 0.5% and morphine 0.1 mg plus in some cases fentanyl 10–20 µg. In two cases, spinals had to be performed twice due to incomplete blocks and one was converted to a general anaesthetic. During the surgical procedure, any blood lost was salvaged into a Dideco Electa Autotransfusion Cell Separator (Sorin Group, Milan, Italy). Aspirated blood was mixed with heparin 30 000 IU litre−1 of saline. Suction was set at 100–150 mm Hg with a prime at 100 ml min−1 and wash volumes of 900 ml on a continuing wash at 100 ml min−1. The set was emptied at 150 ml min−1. A 55 ml Latham bowl was used in all cases. In addition, any swabs from the surgical procedure were soaked in 1 litre of saline and ‘compressed’ at the end of case and solution aspirated, maximizing red cell collection. Processed washed blood was then gravity run through a Pall RS1 VAE leucodepletion filter (Pall Medical, Portsmouth, UK) into another re-infusion bag.
dition, any swabs from the surgical procedure were soaked in 1 litre of saline and ‘compressed’ at the end of case and solution aspirated, maximizing red cell collection. Processed washed blood was then gravity run through a Pall RS1 VAE leucodepletion filter (Pall Medical, Portsmouth, UK) into another re-infusion bag. Pre-wash samples were taken from the collection reservoir, and post-wash and post-filtration samples were taken from the re-infusion packs, which were inverted gently several times to ensure even mixing before sampling. The cell salvage machine use and sampling, in all cases, was performed by the same member of staff. Pre-wash, post-wash, and post-filtration samples from both groups were tested for markers of AF contamination and heparin levels. Post-filtration samples were also tested to quantify any fetal red blood cells. Six millilitres of blood were collected into a specimen tube and processed for alpha-fetoprotein (AFP) on an E170 Modular Analytics Analyser (Roche Diagnostics, Basel, Switzerland). Four millilitres of blood were collected and centrifuged at 1600g for 5 min and processed for heparin levels on an Instrumentation Laboratory Futura Advance Coagulometer (Instrumentation Laboratory, MA, USA).
Six millilitres of blood were collected into a specimen tube and processed for alpha-fetoprotein (AFP) on an E170 Modular Analytics Analyser (Roche Diagnostics, Basel, Switzerland). Four millilitres of blood were collected and centrifuged at 1600g for 5 min and processed for heparin levels on an Instrumentation Laboratory Futura Advance Coagulometer (Instrumentation Laboratory, MA, USA). Six millilitres of blood were collected for fetal squames cell levels. After centrifugation at 1200g for 5 min, one part plasma was mixed with three parts Cytolyt Solution (Cytyc Corporation, West Sussex, UK). Vials were then centrifuged at 1200g for 5 min and supernatant discarded. Two to three drops of the cell bullet were then transferred to 20 ml PreservCyt Solution (Cytyc Corporation) and processed on a ThinPrep®2000 Processor (Cytyc Corporation) using blue-box non-gynae filters (Cytyc Corporation), generating one alcohol-fixed slide per sample. Slides were stained using the Papanicolaou technique. To avoid any potential carryover, pre-wash, post-wash, and post-filtration slides were stained separately using fresh reagents. Each slide was then examined under light microscopy for presence of fetal squames cells, with the whole slide being counted to exclude any variables, giving a result of number of cells per slide. Two millilitres of blood were collected into EDTA for a full blood count to provide haemoglobin (Hb), haematocrit (HCT), and white blood cell (WBC) levels on a Bayer Advia 120 analyser (Bayer, Newbury, UK).
Six millilitres of blood were collected for fetal squames cell levels. After centrifugation at 1200g for 5 min, one part plasma was mixed with three parts Cytolyt Solution (Cytyc Corporation, West Sussex, UK). Vials were then centrifuged at 1200g for 5 min and supernatant discarded. Two to three drops of the cell bullet were then transferred to 20 ml PreservCyt Solution (Cytyc Corporation) and processed on a ThinPrep®2000 Processor (Cytyc Corporation) using blue-box non-gynae filters (Cytyc Corporation), generating one alcohol-fixed slide per sample. Slides were stained using the Papanicolaou technique. To avoid any potential carryover, pre-wash, post-wash, and post-filtration slides were stained separately using fresh reagents. Each slide was then examined under light microscopy for presence of fetal squames cells, with the whole slide being counted to exclude any variables, giving a result of number of cells per slide. Two millilitres of blood were collected into EDTA for a full blood count to provide haemoglobin (Hb), haematocrit (HCT), and white blood cell (WBC) levels on a Bayer Advia 120 analyser (Bayer, Newbury, UK). Post-filtration samples were also tested for fetal red blood cell quantification. Five millilitres of blood were collected into EDTA and processed with a monoclonal antibody to fetal Hb (IQProducts, Groningen, The Netherlands), with a Becton Dickinson FACSCalibur (NJ, USA) flow cytometer at 540 nm. Note: it was planed to perform all 34 cases for fetal red cells by flow cytometry, but due to unforeseen circumstances, the first seven cases were analysed using the Kleihauer–Betke technique.
y to fetal Hb (IQProducts, Groningen, The Netherlands), with a Becton Dickinson FACSCalibur (NJ, USA) flow cytometer at 540 nm. Note: it was planed to perform all 34 cases for fetal red cells by flow cytometry, but due to unforeseen circumstances, the first seven cases were analysed using the Kleihauer–Betke technique. Pre- and post-wash samples and post-wash and post-filtration samples were compared. Parametric data were analysed using a paired t-test, and non-parametric data analysed using a one-sample Wilcoxon test. P<0.05 was considered statistically significant. The observed power of the study was calculated after the collection of the data. Results Mean (sd) (range) age of patients in Group 1 was 32.1 (19–46) yr, and weight 82.7 (16.2) kg. In Group 2, the age was 32.1 (24–41) yr and body weight 82.4 (12.6) kg. Table 1 gives the reason for Caesarean section along with any coexisting diseases as shown in Table 2. Table 1 Reasons for Caesarean section Reason Number Previous section 23 Twins IVF 2 Back injury/pain 2 Breech 2 Failure to progress/Category 3 1 IUGR 1 Polyartetitisnodosa 1 Raised BP/pregnancy-induced hypertension 1 Placenta praevia 1 Table 2 Coexisting diseases
Results Mean (sd) (range) age of patients in Group 1 was 32.1 (19–46) yr, and weight 82.7 (16.2) kg. In Group 2, the age was 32.1 (24–41) yr and body weight 82.4 (12.6) kg. Table 1 gives the reason for Caesarean section along with any coexisting diseases as shown in Table 2. Table 1 Reasons for Caesarean section Reason Number Previous section 23 Twins IVF 2 Back injury/pain 2 Breech 2 Failure to progress/Category 3 1 IUGR 1 Polyartetitisnodosa 1 Raised BP/pregnancy-induced hypertension 1 Placenta praevia 1 Table 2 Coexisting diseases Co-morbidity Number Smoker 5 Diabetic 2 Asthma 2 Epilepsy 1 Protein S deficiency 1 Polyartetitisnodosa 1 There were no surgical problems during the procedures. Four cases did not lose enough blood to obtain post-filtration samples (one from Group 1 and three from Group 2), with three further cases having partially filled bowls. A partial bowl was defined as a bowl that was not completely filled in the automatic mode of the cell salvage machine. These three bowls provided erroneous post-filtration results and, therefore, were excluded from analysis. The final blood product in these cases (one from Group 1 and two from Group 2) had a reduced Hb level of 7.1–10.3 g dl−1 and HCT of 21–31%, with a higher level of fetal red blood cells: 10.5% compared with 1.51%. Apart from fetal red blood cells in one case (Group 2), the values for Hb, HCT, and fetal red blood cells all were considerably above the upper 95% confidence limit for the full bowls.
) had a reduced Hb level of 7.1–10.3 g dl−1 and HCT of 21–31%, with a higher level of fetal red blood cells: 10.5% compared with 1.51%. Apart from fetal red blood cells in one case (Group 2), the values for Hb, HCT, and fetal red blood cells all were considerably above the upper 95% confidence limit for the full bowls. The collected blood volume from Group 1 [1782 (354) ml] was slightly higher than that collected in Group 2 [1497 (562) ml], but the difference was not statistically significant (P=0.14). Similarly, on average, the red blood cell volume was higher in Group 1 [168 (77) ml] than in Group 2 [135 (94) ml], but again the difference was not statistically significant (P=0.30). Table 3 shows the mean Hb, AFP, squames cells, and fetal red cell levels in the groups, along with statistical analysis. Table 3 Analysis and comparison for Group 1 and Group 2. AFP, alpha-fetoprotein. Values are mean (95% CI). *We acknowledge that the observed power results are low; this, however, is one of the largest studies to date in the obstetric setting. †Mean HCT post-filtration was 45% in Group 1 and 42% in Group 2
Table 3 shows the mean Hb, AFP, squames cells, and fetal red cell levels in the groups, along with statistical analysis. Table 3 Analysis and comparison for Group 1 and Group 2. AFP, alpha-fetoprotein. Values are mean (95% CI). *We acknowledge that the observed power results are low; this, however, is one of the largest studies to date in the obstetric setting. †Mean HCT post-filtration was 45% in Group 1 and 42% in Group 2 Group 1 Group 2 P-value Observed power* AFP pre-wash (IU ml−1) 253.72 (183.64–323.80) 381.60 (193.63–569.57) 0.19 0.26 AFP post-wash (IU ml−1) 2.71 (0.80–4.62) 3.14 (1.45–4.83) 0.72 0.06 AFP post-filtration (IU ml−1) 2.58 (0.54–4.62) 3.53 (1.29–5.76) 0.51 0.10 Hb pre-wash (g dl−1) 2.14 (1.52–2.75) 1.83 (1.17–2.50) 0.48 0.11 Hb post-wash (g dl−1) 16.95 (16.28–17.62) 16.20 (15.26–17.14) 0.18 0.27 Hb post-filtration† (g dl−1) 15.77 (15.07–16.47) 14.74 (13.70–15.78) 0.17 0.27 Squames cells pre-wash (cells per slide) 1.12 (−0.02 to 2.25) 0.94 (0.01–1.74) 0.80 0.06 Squames cells post-wash (cells per slide) 34.59 (10.89–58.28) 43.71 (14.42–73.00) 0.61 0.08 Squames cells post-filtration (cells per slide) 0.27 (−0.11 to 0.66) 0 0.18 0.26 Fetal red blood cells (%) 1.34 (0.56–2.12) 1.76 (0.95–2.57) 0.43 0.12 Dependent on the manufacturer, the reference range for AFP is up to 10–15 IU ml−1. The post-wash levels were significantly reduced (P<0.001) with no further reduction seen post-filtration (P=0.996). On average, there was a 98.7% reduction of pre-wash levels by the washing stage.
4 (0.56–2.12) 1.76 (0.95–2.57) 0.43 0.12 Dependent on the manufacturer, the reference range for AFP is up to 10–15 IU ml−1. The post-wash levels were significantly reduced (P<0.001) with no further reduction seen post-filtration (P=0.996). On average, there was a 98.7% reduction of pre-wash levels by the washing stage. Mean heparin levels pre-wash were 1.33 IU ml−1 (95% CI 0.91–1.75 IU ml−1) in Group 1 and 1.44 IU ml−1 (95% CI 0.82–2.05 IU ml−1) in Group 2. Levels were reduced by 100% in both the groups (P<0.001) by the washing stage with no heparin being detected post-wash or post-filtration in all 34 cases. Fetal squames cells were present in 12 out of 34 cases (six from each group) pre-wash, present in all post-wash samples, with a significant reduction post-filtration (P<0.001). Squames were present in two post-filtration samples, with two cells seen in each case. Both cases were from Group 1. Pre-wash samples were heavily diluted with AF and heparinized saline from surgery explaining why 22 samples were negative. The mean percentage fetal red cells post-filtration from all cases (both groups) was 1.51% (95% CI 0.98–2.05), with a range of 0.13–4.35%. The first seven cases (three from Group 1 and four from Group 2) performed by the Kleihauer technique generated a mean 1.72% (95% CI −0.65 to 4.08), whereas the remaining cases performed by flow cytometry resulted in mean 1.47% (95% CI 0.92–2.01).
(both groups) was 1.51% (95% CI 0.98–2.05), with a range of 0.13–4.35%. The first seven cases (three from Group 1 and four from Group 2) performed by the Kleihauer technique generated a mean 1.72% (95% CI −0.65 to 4.08), whereas the remaining cases performed by flow cytometry resulted in mean 1.47% (95% CI 0.92–2.01). Discussion This study has shown the efficiency of the washing stage of the cell salvage machine, when used in combination with a leucodepletion filter, in significantly reducing levels of AF contaminants. AFP was significantly reduced post-wash to levels well within the normal range for the general population, confirming the results from previous studies.156 During the study, the amount of heparin used was higher than with other surgical disciplines due to the hypercoagulable state of pregnancy. It would be expected that pre-wash samples were heavily contaminated with heparin, and this was confirmed during the study. It has been previously reported that residual levels of heparin remain in salvaged blood,4 but we have now demonstrated the complete removal of heparin by the washing process in all 34 cases. The role of fetal squames cells in AF embolus is still debated. In one study,1 fetal squames cells were still present even with the additional step of filtering the washed blood through a leucodepletion filter, with only two out of 27 post-filtration cases being negative. Advanced filters may reduce this contamination.2
The role of fetal squames cells in AF embolus is still debated. In one study,1 fetal squames cells were still present even with the additional step of filtering the washed blood through a leucodepletion filter, with only two out of 27 post-filtration cases being negative. Advanced filters may reduce this contamination.2 Of the 27 cases in our study, only two were positive post-filtration with two cells seen in each slide for each case. These were found when one suction was used. However, it is difficult to differentiate between fetal and adult squames cells. There is no evidence to show that fetal squames cells routinely enter the maternal circulation,7 and so the significance of this contamination is difficult to quantify and in fact may have no clinical significance.
en one suction was used. However, it is difficult to differentiate between fetal and adult squames cells. There is no evidence to show that fetal squames cells routinely enter the maternal circulation,7 and so the significance of this contamination is difficult to quantify and in fact may have no clinical significance. Fetal red blood cells were still present in the final product, consistent with previous studies,26 and so could be significant in cases of red cell antigen incompatibilities between the mother and fetus. Rh(D) incompatibilities, although clinically significant, are generally avoided by routine prophylactic anti-D treatment throughout the pregnancy. However, fetal hyperbilirubinaemia and anaemia in future pregnancies can occur when antibodies have been formed to other red cell antigen incompatibilities. Examples of other clinically relevant antibodies that have been implicated in haemolytic disease of the newborn include anti-K, anti-c, anti-Fy(a), and anti-Jk(a).8 Nevertheless, it must be appreciated that there is still a risk of alloimmunization of the mother either from transfusion of allogenic blood or a sensitizing event during pregnancy. Current treatment of obstetric haemorrhage is with allogenic blood transfusion; however, there are concerns with allogenic blood. To date, there have been four confirmed cases of vCJD transmission via allogeneic blood, confirming that vCJD can be transmitted through blood transfusion.9
Fetal red blood cells were still present in the final product, consistent with previous studies,26 and so could be significant in cases of red cell antigen incompatibilities between the mother and fetus. Rh(D) incompatibilities, although clinically significant, are generally avoided by routine prophylactic anti-D treatment throughout the pregnancy. However, fetal hyperbilirubinaemia and anaemia in future pregnancies can occur when antibodies have been formed to other red cell antigen incompatibilities. Examples of other clinically relevant antibodies that have been implicated in haemolytic disease of the newborn include anti-K, anti-c, anti-Fy(a), and anti-Jk(a).8 Nevertheless, it must be appreciated that there is still a risk of alloimmunization of the mother either from transfusion of allogenic blood or a sensitizing event during pregnancy. Current treatment of obstetric haemorrhage is with allogenic blood transfusion; however, there are concerns with allogenic blood. To date, there have been four confirmed cases of vCJD transmission via allogeneic blood, confirming that vCJD can be transmitted through blood transfusion.9 These potential risks of transfusion and reduction in availability of blood make it important to establish the use of cell salvage in obstetrics. It has been shown in more than 400 documented cases where cell salvaged blood has been returned to mothers with no significant adverse results.41011 Several key studies have commented on the use of cell salvage.12–14
eduction in availability of blood make it important to establish the use of cell salvage in obstetrics. It has been shown in more than 400 documented cases where cell salvaged blood has been returned to mothers with no significant adverse results.41011 Several key studies have commented on the use of cell salvage.12–14 This current study is one of the largest to date, with results that add to the growing body of evidence, showing there is little or no possibility for AF contamination to enter the re-infusion system, when used in conjunction with a leucodepletion filter. The role of the leucodepletion filter in this study has confirmed that it is required to remove AF contamination in cell salvage. WBC, platelets, and squames cells were still present post-wash but were significantly reduced by the filter. Whether the squames are of fetal or maternal origin is perhaps irrelevant as the filters have shown that they can remove both. As expected, AFP and heparin levels were not reduced by the filter, but heparin was completely removed in the washing process.
ll present post-wash but were significantly reduced by the filter. Whether the squames are of fetal or maternal origin is perhaps irrelevant as the filters have shown that they can remove both. As expected, AFP and heparin levels were not reduced by the filter, but heparin was completely removed in the washing process. Regarding cell salvage in obstetrics, these results could be used to change the guidelines and allow routine re-transfusion of salvaged blood as opposed to overriding guidance in extreme emergency, and therefore allowing cell salvage to be used for elective and emergency cases. Even though it was not our intention to re-transfuse any salvaged blood in this study, ∼500 ml was transfused to two cases. No changes in clinical state, heart rate, or clinical complications were noted, showing that in elective cases unexpected large blood loss can still occur and cell salvage can have a role to play in this situation. This initial project acted as an exploratory study, and after the collection of these data, we have now implemented a comprehensive programme of cell salvage within our Trust to all women undergoing elective Caesarean sections, and further work is ongoing. We conclude that one suction may be used in the obstetric setting, and washed filtered blood from partially filled bowls should not be re-transfused, regardless of the clinical situation. To obtain maximum washing efficiency in removing AF and fetal contaminants, only complete bowls in the automatic mode should be accepted. Funding This study was made possible by an educational grant from Pall Medical.
We conclude that one suction may be used in the obstetric setting, and washed filtered blood from partially filled bowls should not be re-transfused, regardless of the clinical situation. To obtain maximum washing efficiency in removing AF and fetal contaminants, only complete bowls in the automatic mode should be accepted. Funding This study was made possible by an educational grant from Pall Medical. Acknowledgements The authors wish to thank Cytyc Corporation for donation of the blue-box non-gynae filters for the ThinPrep® 2000 Processor and Sorin Group for the loan of the cell salvage machine.
collection: one participant was excluded from all study procedures as they returned from theatre in a critical condition; one participant was excluded as it was not possible to insert the pressure measuring catheter; and three were excluded after the pressure measuring catheter became dislodged before data collection. On return from theatre, while sedated and ventilated, a 10 F catheter was inserted into the nasopharynx via the nose. The catheter was secured in place and remained in situ overnight. Participants received all standard ICU care. The morning after surgery, once awake and extubated, participants were routinely mobilized to a chair and their pain levels assessed. Analgesia was provided where necessary, as per unit protocol, to ensure that each participant was able to breathe deeply and comfortably. For consistency, all measurements were performed by one of the researchers trained in this technique.
Patients with respiratory failure are typically treated with three main respiratory support strategies, depending on the severity of their illness. These are traditional oxygen therapy, non-invasive ventilation, and invasive mechanical ventilation.1 A new respiratory support therapy has recently been introduced into the adult arena. NHF allows the delivery of up to 60 litre min−1 of heated and humidified gas via a wide bore nasal cannula. However, the effect of delivering such high-flow rates into the nasopharynx remains unclear. In neonatal care, the delivery of relatively high flows of heated and humidified gas via a nasal cannula has gained increasing acceptance in the treatment of respiratory conditions.2–4 Studies of NHF in this patient population have demonstrated an effect comparable with nasal continuous positive airway pressure (CPAP).4–6 In adults, there have been reports that NHF may be beneficial in the treatment of obstructive sleep apnoea, attributable to a flow-related pressure effect.7 There has been a similar pressure effect reported in healthy adult volunteers8–9 where a positive relationship between flow and airway pressure has also been described. However, to date there have been no published reports evaluating pressure effects in the adult ICU population. The aim of this study was to quantify the airway pressure effect associated with NHF in an adult patient cohort.
In neonatal care, the delivery of relatively high flows of heated and humidified gas via a nasal cannula has gained increasing acceptance in the treatment of respiratory conditions.2–4 Studies of NHF in this patient population have demonstrated an effect comparable with nasal continuous positive airway pressure (CPAP).4–6 In adults, there have been reports that NHF may be beneficial in the treatment of obstructive sleep apnoea, attributable to a flow-related pressure effect.7 There has been a similar pressure effect reported in healthy adult volunteers8–9 where a positive relationship between flow and airway pressure has also been described. However, to date there have been no published reports evaluating pressure effects in the adult ICU population. The aim of this study was to quantify the airway pressure effect associated with NHF in an adult patient cohort. Methods A prospective study was conducted in a 16-bed cardiothoracic and vascular ICU to test the hypothesis that a positive airway pressure is generated. The study was approved by the regional Ethics Committee. Written informed consent was obtained from the participants before operation.
The aim of this study was to quantify the airway pressure effect associated with NHF in an adult patient cohort. Methods A prospective study was conducted in a 16-bed cardiothoracic and vascular ICU to test the hypothesis that a positive airway pressure is generated. The study was approved by the regional Ethics Committee. Written informed consent was obtained from the participants before operation. Participants who were undergoing elective cardiac surgery were eligible for inclusion in this study. They were excluded if they had a history of sinus problems, nasal trauma, or a markedly deviated septum. In total, 20 participants were recruited, of which 15 completed the study. Five participants were excluded before data collection: one participant was excluded from all study procedures as they returned from theatre in a critical condition; one participant was excluded as it was not possible to insert the pressure measuring catheter; and three were excluded after the pressure measuring catheter became dislodged before data collection.
participants were routinely mobilized to a chair and their pain levels assessed. Analgesia was provided where necessary, as per unit protocol, to ensure that each participant was able to breathe deeply and comfortably. For consistency, all measurements were performed by one of the researchers trained in this technique. Before performing measurements, correct placement of the catheter was confirmed using end-tidal carbon dioxide (CO2) monitoring. If necessary, the catheter was suctioned and manipulated until a clear respiratory trace was achieved. A visual check was also performed to locate the tip of the catheter behind the uvula. The catheter was then connected to the Honeywell precision pressure transducer (PPT-0001 DWWW2VA-B, Honeywell International Ltd, NJ, USA) using a laptop computer interface (Figs 1 and 2). The Optiflow™ system (MR880 heated humidifier, RT241 heated delivery tube, Fisher and Paykel Healthcare Ltd, Auckland, New Zealand) with an air/oxygen blender (Bird high-flow blenders, Cardinal Healthcare, IL, USA) was used for all measurements. Fig 1 Optiflow™ system set-up. a, Optiflow™ RT034 cannula; b, heater delivery tube RT241; c, MR880 heated humidifier; d, laptop interface; e, pressure transducer. Fig 2 Standard facemask set-up. g, Mask adapter (commercially unavailable).
Before performing measurements, correct placement of the catheter was confirmed using end-tidal carbon dioxide (CO2) monitoring. If necessary, the catheter was suctioned and manipulated until a clear respiratory trace was achieved. A visual check was also performed to locate the tip of the catheter behind the uvula. The catheter was then connected to the Honeywell precision pressure transducer (PPT-0001 DWWW2VA-B, Honeywell International Ltd, NJ, USA) using a laptop computer interface (Figs 1 and 2). The Optiflow™ system (MR880 heated humidifier, RT241 heated delivery tube, Fisher and Paykel Healthcare Ltd, Auckland, New Zealand) with an air/oxygen blender (Bird high-flow blenders, Cardinal Healthcare, IL, USA) was used for all measurements. Fig 1 Optiflow™ system set-up. a, Optiflow™ RT034 cannula; b, heater delivery tube RT241; c, MR880 heated humidifier; d, laptop interface; e, pressure transducer. Fig 2 Standard facemask set-up. g, Mask adapter (commercially unavailable). Once the system temperature stabilized (target; 37°C with an absolute humidity of 44 mg H2O litre−1), therapy was commenced at 35 litre min−1 using the Optiflow™ wide bore nasal cannula (Fig. 1; RT034 nasal cannula, Fisher and Paykel Healthcare Ltd). As per the manufacturer's recommendations, the investigators ensured that no more than half the internal diameter of the nares was taken up by the cannula. Participants were given ∼15 min on the system to become accustomed to the feeling of increased flow and to allow breathing patterns to settle before measurements were commenced. Measurements were then repeated with a standard facemask (Fig. 2; Medium Adult SEE-THRU® Oxygen Mask, Hudson Respiratory Care Inc., NC, USA) at 35 litre min−1 (target; 37°C with an absolute humidity of 44 mg H2O litre−1) connected by way of an adapter to the heated delivery tube. This ensured that all equipment and conditions were standardized with the exception of the interface. Oxygen concentrations were titrated to meet participants' requirements. Measurements were recorded with the participant's mouth open and mouth closed for each interface. Each recording was taken over 1 min of quiet breathing. A washout period of 5 min was allowed between each of the measurements to ensure no carry-over effect between therapies. During this time, the patency and position of the catheter was re-checked.
th the participant's mouth open and mouth closed for each interface. Each recording was taken over 1 min of quiet breathing. A washout period of 5 min was allowed between each of the measurements to ensure no carry-over effect between therapies. During this time, the patency and position of the catheter was re-checked. At the end of the procedure, the nasopharyngeal catheter was removed and the participant returned to their original oxygen delivery device. Nasopharyngeal pressure profiles were recorded for each participant over 1 min. An example of one participant's trace is shown in Figure 3. Data analysis determined the mean nasopharyngeal airway pressure. This required averaging the pressure from the peak of inspiration of the first breath to the peak of inspiration of the last breath within a 1 min recording. This allowed the entire pressure profile of each breath within that 1 min period to be included in the pressure calculation. All data analysis was performed using Microsoft® Office Excel 2003. Data are presented as mean (sd). Paired t-test was used to compare mean differences between NHF and standard facemask, and between the mouth open and mouth closed measurements for each interface. Fig 3 Breathing pressure profile of one participant over 1 min. Nasopharyngeal pressure (cm H2O) generated at 35 litre min−1 using NHF with the mouth closed (NHFMC) is shown in black. Nasopharyngeal pressure (cm H2O) generated at 35 litre min−1 using a facemask with the mouth closed (FMMC) is shown in grey.
Nasopharyngeal pressure profiles were recorded for each participant over 1 min. An example of one participant's trace is shown in Figure 3. Data analysis determined the mean nasopharyngeal airway pressure. This required averaging the pressure from the peak of inspiration of the first breath to the peak of inspiration of the last breath within a 1 min recording. This allowed the entire pressure profile of each breath within that 1 min period to be included in the pressure calculation. All data analysis was performed using Microsoft® Office Excel 2003. Data are presented as mean (sd). Paired t-test was used to compare mean differences between NHF and standard facemask, and between the mouth open and mouth closed measurements for each interface. Fig 3 Breathing pressure profile of one participant over 1 min. Nasopharyngeal pressure (cm H2O) generated at 35 litre min−1 using NHF with the mouth closed (NHFMC) is shown in black. Nasopharyngeal pressure (cm H2O) generated at 35 litre min−1 using a facemask with the mouth closed (FMMC) is shown in grey. Results Data from 15 participants were analysed. The participants were 17 males and two females of mean (range) age 63 (41–79) yr, weight 86 (67–107) kg, and height 175 (156–186) cm.
Fig 3 Breathing pressure profile of one participant over 1 min. Nasopharyngeal pressure (cm H2O) generated at 35 litre min−1 using NHF with the mouth closed (NHFMC) is shown in black. Nasopharyngeal pressure (cm H2O) generated at 35 litre min−1 using a facemask with the mouth closed (FMMC) is shown in grey. Results Data from 15 participants were analysed. The participants were 17 males and two females of mean (range) age 63 (41–79) yr, weight 86 (67–107) kg, and height 175 (156–186) cm. Significantly higher nasopharyngeal airway pressures were recorded with NHF in the mouth closed position when compared with mouth open [2.7 (1.04) vs 1.2 (0.76) cm H2O (P=0.001)] (Table 1). There was no significant difference in nasopharyngeal airway pressure generated with facemask mouth open [0.11 (0.39) cm H2O] and facemask mouth closed [0.2 (0.63) cm H2O] (P=0.5). The corresponding values of nasopharyngeal pressures with NHF, during mouth open or mouth closed, were significantly higher than those with a facemask (P=0.001). Table 1 Measurements from the 15 participants showing individual nasopharyngeal pressures (cm H2O) and mean nasopharyngeal pressures (cm H2O) with sd generated with NHF and facemask at 35 litre min−1 with the open mouth (NHFMO and FMMO) and closed mouth positions (NHFMC and FMMC)
Significantly higher nasopharyngeal airway pressures were recorded with NHF in the mouth closed position when compared with mouth open [2.7 (1.04) vs 1.2 (0.76) cm H2O (P=0.001)] (Table 1). There was no significant difference in nasopharyngeal airway pressure generated with facemask mouth open [0.11 (0.39) cm H2O] and facemask mouth closed [0.2 (0.63) cm H2O] (P=0.5). The corresponding values of nasopharyngeal pressures with NHF, during mouth open or mouth closed, were significantly higher than those with a facemask (P=0.001). Table 1 Measurements from the 15 participants showing individual nasopharyngeal pressures (cm H2O) and mean nasopharyngeal pressures (cm H2O) with sd generated with NHF and facemask at 35 litre min−1 with the open mouth (NHFMO and FMMO) and closed mouth positions (NHFMC and FMMC) Participant Mean nasopharyngeal pressure (cm H2O) NHFMC NHFMO FMMC FMMO 1 1.7 1.6 1.3 0.3 2 2.7 1.7 — 1.3 3 — — — — 4 1.5 0.9 −0.2 −0.5 5 1.6 0.2 −0.01 0.02 6 2.2 0.5 0.005 −0.01 7 3.4 2.1 0.02 0.06 8 2.7 1.6 −0.003 0.04 9 3.4 0.7 −0.09 0.02 10 2.2 0.9 0.02 −0.04 11 5.3 2.3 1.7 0.4 12 1.7 0.1 −0.5 0.02 13 2.5 2.4 0.6 −0.06 14 3.0 0.7 0.006 0.04 15 3.7 1.5 −0.1 0.03 Mean airway pressure at 35 litre min−1 2.7 1.2 0.2 0.1 sd 1.04 0.76 0.63 0.39 Discussion This study demonstrates that a significant positive airway pressure effect is delivered with NHF using the Optiflow™ system and a wide bore nasal cannula. The pressure effect was shown to be most evident with the participant's mouth closed, but was still significant with the mouth open, when compared with a facemask interface at the same gas flow rate.
nificant positive airway pressure effect is delivered with NHF using the Optiflow™ system and a wide bore nasal cannula. The pressure effect was shown to be most evident with the participant's mouth closed, but was still significant with the mouth open, when compared with a facemask interface at the same gas flow rate. The mean nasopharyngeal airway pressure rather than the end-expiratory airway pressure was analysed. This was considered to be a more accurate assessment of overall therapy effect. The terms CPAP and PEEP have been avoided as it is not clear to us how these terms relate to NHF. For the purposes of this study, a nasopharyngeal catheter was used rather than an oesophageal balloon catheter, due to the perceived patient risks and discomfort associated with the latter technique. Nasopharyngeal pressure has been taken as the most feasible surrogate measure of transpulmonary pressure. It should be noted that during analysis, the pressure profiles from the first four participants (Table 1) were noted to be dampened. Indeed, participant number 3 had no analysable pressure traces. These errors were assumed to be technique-related, but mean data have been reported where possible. To assess for difference, analysis was carried out including, and excluding, data belonging to the participants 1–4. Results showed that the difference in treatment effect remained similar whether these participants were included or not. Previous research has shown comparable results with those obtained in this study.89
For the purposes of this study, a nasopharyngeal catheter was used rather than an oesophageal balloon catheter, due to the perceived patient risks and discomfort associated with the latter technique. Nasopharyngeal pressure has been taken as the most feasible surrogate measure of transpulmonary pressure. It should be noted that during analysis, the pressure profiles from the first four participants (Table 1) were noted to be dampened. Indeed, participant number 3 had no analysable pressure traces. These errors were assumed to be technique-related, but mean data have been reported where possible. To assess for difference, analysis was carried out including, and excluding, data belonging to the participants 1–4. Results showed that the difference in treatment effect remained similar whether these participants were included or not. Previous research has shown comparable results with those obtained in this study.89 An overall effect on airway pressure was observed when individual pressure profiles recorded in this study were examined. As can be seen in Figure 3, the pressure tracing recorded with a facemask tends to rotate around zero (atmospheric pressure). With NHF, the entire pressure profile of the breath is elevated. Visual analysis of these profiles would also seem to suggest that the expiratory phase is prolonged when using NHF. This could be explained by the effect of breathing out against the incoming gas flow.
k tends to rotate around zero (atmospheric pressure). With NHF, the entire pressure profile of the breath is elevated. Visual analysis of these profiles would also seem to suggest that the expiratory phase is prolonged when using NHF. This could be explained by the effect of breathing out against the incoming gas flow. A large interpatient variability was observed in this study. When participants using NHF breathed with mouth closed, the mean nasopharyngeal pressure generated ranged from 1.54 to 5.34 cm H2O (Table 1). We hypothesize that this variability may be due to differences in leak around the outside of the nasal cannula and the wide variability in nare size among the study population. A smaller leak may create an increased resistance to expiration resulting in higher nasopharyngeal pressure. Previous studies support this theory.510 Physiological differences in airway anatomy may also explain some of the variability. It is uncertain whether the presence of a nasopharyngeal catheter to measure airway pressure as used in this and other studies89 has an effect on the pressure generated in the upper airway.
re. Previous studies support this theory.510 Physiological differences in airway anatomy may also explain some of the variability. It is uncertain whether the presence of a nasopharyngeal catheter to measure airway pressure as used in this and other studies89 has an effect on the pressure generated in the upper airway. In spite of these limitations, it is evident that a positive airway pressure is generated using NHF and this may have important clinical implications. A number of clinical benefits are associated with conventional pressure-generating devices, including improved oxygenation; improved ventilation perfusion matching; reduced airways resistance; reduced work of breathing; and the balancing of intrinsic PEEP.11 Studies to assess the extent to which NHF is associated with these clinical benefits are required. In conclusion, NHF is a new respiratory support modality into which little clinical research has been conducted. We have shown that NHF can significantly increase mean nasopharyngeal airway pressure in an adult patient population. This may be an important factor in determining the most appropriate respiratory support therapy for a particular patient. Areas for future research would include the investigation of airway pressure measurements at differing flow rates, clinical applications, and the influence of NHF on patient outcomes. Funding Research in the Cardiothoracic and Vascular Intensive Care Unit is supported in part by Fisher and Paykel Healthcare Ltd.
In conclusion, NHF is a new respiratory support modality into which little clinical research has been conducted. We have shown that NHF can significantly increase mean nasopharyngeal airway pressure in an adult patient population. This may be an important factor in determining the most appropriate respiratory support therapy for a particular patient. Areas for future research would include the investigation of airway pressure measurements at differing flow rates, clinical applications, and the influence of NHF on patient outcomes. Funding Research in the Cardiothoracic and Vascular Intensive Care Unit is supported in part by Fisher and Paykel Healthcare Ltd. Acknowledgement Special thanks to Steven Korner, Fisher and Paykel Healthcare Ltd, for assistance in refining the measurement technique.
Pentraxins are a group of acute phase proteins which are produced in response to inflammatory conditions in vivo and which play a key role in the innate immune system.12 C-reactive protein (CRP) is a classical short pentraxin routinely used in the diagnosis and monitoring of inflammation and infection. Pentraxin-3 (PTX3) is the first member of the long pentraxin family group and is expressed in several cell types after exposure to pro-inflammatory stimuli, including specific microbial constituents and cytokines.3–5 Sepsis is the systemic generalized activation of the innate immune response due to infection, involving cellular inflammatory responses triggered by intracellular oxidative stress. Clinically, causative organisms are predominantly bacterial, but the use of indwelling central venous catheters and treatment with broad-spectrum antibiotics are additional risk factors for fungal infection and Candida spp. is now the fourth-most-common organism to be isolated from the bloodstream of hospitalized patients6 with attributable mortality of candidiasis of around 38%.7
t the use of indwelling central venous catheters and treatment with broad-spectrum antibiotics are additional risk factors for fungal infection and Candida spp. is now the fourth-most-common organism to be isolated from the bloodstream of hospitalized patients6 with attributable mortality of candidiasis of around 38%.7 The promoter region of PTX3 contains binding sites for the redox-sensitive transcription factor nuclear factor κB (NFκB),89 which is involved in the regulation pathway of many inflammatory mediators important in innate immunity. Oxidative stress has consistently been demonstrated in patients with sepsis10–14 and acts as a trigger for inflammation. We hypothesized that antioxidants may affect expression of PTX3. Using bacterial cell components, cytokines, or killed Candida albicans cells to mimic different conditions of sepsis, we determined the effect of antioxidant treatment on PTX3 expression in human endothelial cells. In addition, we measured PTX3 in relation to total antioxidant capacity and lipid hydroperoxides in plasma from patients with severe sepsis.
tokines, or killed Candida albicans cells to mimic different conditions of sepsis, we determined the effect of antioxidant treatment on PTX3 expression in human endothelial cells. In addition, we measured PTX3 in relation to total antioxidant capacity and lipid hydroperoxides in plasma from patients with severe sepsis. Methods Patient study In this pilot study, 20 consecutive patients were recruited from the intensive care unit (ICU) within 24 h of fulfilling the criteria for sepsis, after local ethical approval and obtaining written informed consent from the patient or assent from a close relative. The criteria used were those recommended by the Consensus Meeting of the American Thoracic Society and the American Society of Critical Care Medicine,15 namely clinical suspicion of infection plus two of the following: tachycardia (>100 beats min−1), tachypnoea (>20 bpm or ventilated), or leucocyte count <4 or >12×109 litre−1. Patients <16 yr old, who were pregnant or lactating, HIV positive, on steroid or immunosuppression therapy, who had any form of cancer or autoimmune disease or who were taking statins, were excluded. Five patients were subsequently excluded; two were found subsequently not to have sepsis and another three were taking statins. Heparinized blood was obtained from an indwelling cannula and immediately centrifuged and the plasma stored at −80°C for PTX3 analysis. Acute physiological and chronic health evaluation (APACHE) II score was also recorded. Blood samples were also obtained with consent from 11 healthy laboratory and research staff (age range 25–50 yr) using heparinized vacutainers and samples were treated as for patients.
nd the plasma stored at −80°C for PTX3 analysis. Acute physiological and chronic health evaluation (APACHE) II score was also recorded. Blood samples were also obtained with consent from 11 healthy laboratory and research staff (age range 25–50 yr) using heparinized vacutainers and samples were treated as for patients. In vitro study All reagents were obtained from Sigma Aldrich Ltd, Poole, Dorset, UK, unless stated otherwise. The human umbilical vein endothelial cell line HUVEC-C was used (American Type Culture Collection, Manassas, VA, USA). Cells were cultured in 6-well plates as we have previously described15 in Dulbecco's Modified Eagle's Medium (Lonza Wokingham Ltd, Berkshire, UK) containing heat-inactivated fetal calf serum 10%, gentamicin 50 µg ml−1, and amphotericin B 250 µg ml−1. Yeast cells of the human pathogenic fungus C. albicans wild-type derivative strain NGY152 were used.16 Candida albicans was grown in Sabouraud broth overnight at 30°C with shaking. The overnight culture was diluted 1:100 in fresh broth, grown until spectrophotometry showed an absorbance of 0.5 at 600 nm, then harvested. The pellet was washed three times in phosphate-buffered saline (PBS) and resuspended to a final concentration of 1×108 yeast cells ml−1 in PBS. The cells were heat-killed by incubation at 56°C for 2 h.
was diluted 1:100 in fresh broth, grown until spectrophotometry showed an absorbance of 0.5 at 600 nm, then harvested. The pellet was washed three times in phosphate-buffered saline (PBS) and resuspended to a final concentration of 1×108 yeast cells ml−1 in PBS. The cells were heat-killed by incubation at 56°C for 2 h. Endothelial cells were cultured for 24 h at 37°C in the presence of lipopolysaccharide (LPS) 2 µg ml−1 plus peptidoglycan G (PepG) 20 µg ml−1, tumour necrosis factor α (TNFα, PeproTech EC Ltd, London, UK) 10 ng ml−1, interleukin-1β (IL-1β, PeproTech) 20 ng ml−1, TNFα and IL-1β combined, or killed C. albicans cells at a multiplicity of infection (MOI) of 1–10, along with the antioxidants 6-hydroxy-2,5,7,8-tetramethylchroman-2-carboxylic acid (trolox, a water-soluble vitamin E analogue, 100 mM), N-acetylcysteine (NAC, a glutathione precursor, 25 mM) or 2,3-dimethoxy-5-methyl-6-(10-hydroxydecyl)-1,4-benzoquinone (idebenone, a water-soluble co-enzyme Q10 analogue, 1 µM, a kind gift from Dr M.P. Murphy, MRC-Dunn Nutrition Unit, Cambridge, UK), or appropriate solvent controls. Culture supernatants were stored at –80°C for subsequent PTX3 measurement. In separate experiments, endothelial cells were treated with IL-1β and TNFα plus antioxidants as above and nuclear extracts were prepared for NFκB assay. All measurements on cells were corrected for viable cell number which was determined using acid phosphatase activity.17 The precision of this assay (% coefficient of variation) was 1.06%.
Endothelial cells were cultured for 24 h at 37°C in the presence of lipopolysaccharide (LPS) 2 µg ml−1 plus peptidoglycan G (PepG) 20 µg ml−1, tumour necrosis factor α (TNFα, PeproTech EC Ltd, London, UK) 10 ng ml−1, interleukin-1β (IL-1β, PeproTech) 20 ng ml−1, TNFα and IL-1β combined, or killed C. albicans cells at a multiplicity of infection (MOI) of 1–10, along with the antioxidants 6-hydroxy-2,5,7,8-tetramethylchroman-2-carboxylic acid (trolox, a water-soluble vitamin E analogue, 100 mM), N-acetylcysteine (NAC, a glutathione precursor, 25 mM) or 2,3-dimethoxy-5-methyl-6-(10-hydroxydecyl)-1,4-benzoquinone (idebenone, a water-soluble co-enzyme Q10 analogue, 1 µM, a kind gift from Dr M.P. Murphy, MRC-Dunn Nutrition Unit, Cambridge, UK), or appropriate solvent controls. Culture supernatants were stored at –80°C for subsequent PTX3 measurement. In separate experiments, endothelial cells were treated with IL-1β and TNFα plus antioxidants as above and nuclear extracts were prepared for NFκB assay. All measurements on cells were corrected for viable cell number which was determined using acid phosphatase activity.17 The precision of this assay (% coefficient of variation) was 1.06%. Pentraxin-3 and CRP measurement PTX3 expression from plasma or culture supernatants was measured by enzyme immunoassay (R&D Systems Europe Ltd, Abingdon, Oxon, UK). Briefly, 96-well plates were coated with anti-PTX3 monoclonal antibody and incubated overnight at 4°C. After incubation, plates were washed with PBS containing Tween 20 0.05% and blocked with skim milk powder 10% for 1 h at 37°C. Plates were washed again and recombinant human PTX3 as a calibration standard, or plasma or culture supernatant was added to each well. After 2 h incubation at 37°C, plates were washed and biotinylated anti-PTX3 polyclonal antibody was added for a further hour at 37°C followed by 40 min incubation with streptavidin–horseradish peroxidase (HRP) then a chromogen substrate. The reaction was stopped using sulphuric acid and quantified spectrophotometrically. The precision of this assay (% coefficient of variation) was 2.9%.
i-PTX3 polyclonal antibody was added for a further hour at 37°C followed by 40 min incubation with streptavidin–horseradish peroxidase (HRP) then a chromogen substrate. The reaction was stopped using sulphuric acid and quantified spectrophotometrically. The precision of this assay (% coefficient of variation) was 2.9%. CRP was measured colorimetrically using a Siemens ADVIA 2400 autoanalyzer (Siemens Diagnostics, Tarrytown, NY, USA). Nuclear factor κB Nuclear extraction was performed using the NucBuster kit (Novagen, Merk Chemicals, Nottingham, UK) according to the manufacturer's instructions. Briefly, cells were suspended in NucBuster extraction reagent 1 and vortexed for 15 s, incubated on ice for 5 min and vortexed again. Nuclei were sedimented by centrifugation at 13 000g at 4°C, for 5 min, the supernatant was removed and the nuclei pellet was resuspended in Nucbuster extraction reagent 2 containing protease inhibitor cocktail and dithiothreitol 1.28 mM. The samples were then vortexed, incubated on ice, and centrifuged as before. The supernatant (nuclear extract) was then used in the NFκB assay.
or 5 min, the supernatant was removed and the nuclei pellet was resuspended in Nucbuster extraction reagent 2 containing protease inhibitor cocktail and dithiothreitol 1.28 mM. The samples were then vortexed, incubated on ice, and centrifuged as before. The supernatant (nuclear extract) was then used in the NFκB assay. NFκB was measured using the NoShift transcription factor immunoassay kit (Novagen, Merk Chemicals). First, nuclear extract was incubated on ice for 30 min with binding buffer containing poly(dI-dC), salmon sperm DNA, and wild-type DNA (biotinylated, 10 pmol ml−1). After incubation, the reaction mixture was made up to 100 µl with binding buffer before being transferred to a streptavidin-coated 96-well plate and incubated at 37°C for 1 h. The plate was then washed, detection antibody added, and incubated again at 37°C for 1 h. The plate was again washed before the addition of goat-anti-mouse IgG HRP conjugate for 30 min before being washed five times. Chromogen substrate was then added to the wells and incubated at room temperature until colour developed. This reaction was stopped with the addition of hydrochloric acid, quantified spectrophotometrically, and expressed in terms of nuclear protein content, measured using the Bradford reagent. The precision of this assay (% coefficient of variation) was 1.4%.
wells and incubated at room temperature until colour developed. This reaction was stopped with the addition of hydrochloric acid, quantified spectrophotometrically, and expressed in terms of nuclear protein content, measured using the Bradford reagent. The precision of this assay (% coefficient of variation) was 1.4%. Total antioxidant capacity and lipid peroxides Total antioxidant capacity was measured using a commercially available kit (Cayman Chemical, Ann Arbor, MI, USA) based on the ability of plasma antioxidants to inhibit oxidation of 2,2′-azino-di-[3-ethylbenzothiaolinesulphonate] (ABTS) to the ABTS radical by metmyoglobin.18 The capacity of plasma to prevent oxidation of ABTS is compared with that of trolox and is quantified as molar trolox equivalents. The precision of this assay (% coefficient of variation) was 3.4%. Total lipid hydroperoxides were measured using a spectrophotometric technique.12 Lipid hydroperoxides are extracted into chloroform, which eliminates interference by hydrogen peroxide or endogenous ferric ions in the sample. The precision of this assay (% coefficient of variation) was 0.63%.
Total antioxidant capacity and lipid peroxides Total antioxidant capacity was measured using a commercially available kit (Cayman Chemical, Ann Arbor, MI, USA) based on the ability of plasma antioxidants to inhibit oxidation of 2,2′-azino-di-[3-ethylbenzothiaolinesulphonate] (ABTS) to the ABTS radical by metmyoglobin.18 The capacity of plasma to prevent oxidation of ABTS is compared with that of trolox and is quantified as molar trolox equivalents. The precision of this assay (% coefficient of variation) was 3.4%. Total lipid hydroperoxides were measured using a spectrophotometric technique.12 Lipid hydroperoxides are extracted into chloroform, which eliminates interference by hydrogen peroxide or endogenous ferric ions in the sample. The precision of this assay (% coefficient of variation) was 0.63%. Data analysis For the in vitro study, four replicate independent experiments in triplicate were performed. No assumptions were made about the distribution of data which were analysed using the Kruskal–Wallis with Mann–Whitney post hoc testing and Bonferroni's correction as appropriate. Patients' plasma PTX3 and total antioxidant capacity were compared with those of healthy subjects using the Mann–Whitney U-test. All data are presented as median (range). A P-value of <0.05 was considered to be significant.
he Kruskal–Wallis with Mann–Whitney post hoc testing and Bonferroni's correction as appropriate. Patients' plasma PTX3 and total antioxidant capacity were compared with those of healthy subjects using the Mann–Whitney U-test. All data are presented as median (range). A P-value of <0.05 was considered to be significant. Results Fifteen patients were included (six females and nine males, aged 22–85 yr) (Table 1). The median (range) APACHE II score was 21 (9–32). PTX3 levels in plasma were significantly higher in patients with sepsis than in healthy people [26 (1–202) ng ml−1 compared with 6 (1–12) ng ml−1, P=0.01, Fig. 1a]. PTX3 levels were also significantly lower in men than in women [16.7 (1–42) ng ml−1 compared with 55 (24–202) ng ml−1, P<0.003, Fig. 1a]. When PTX3 levels in patients were plotted according to APACHE II score quartile, PTX3 levels were related to APACHE score quartile (Fig. 1b). Median (range) CRP was 164 (20–370) mg ml−1 and was unrelated to APACHE II score quartile. Fig 1 (a) Plasma PTX3 concentrations in 15 patients with sepsis and 11 healthy subjects. Squares are females, circles are males. (b) Data sets compared using Mann–Whitney U-test. Plasma PTX3 concentrations according to APACHE II score quartile in 15 patients with sepsis. Table 1 Characteristics of patients with sepsis. UTI, urinary tract infection; CBD, common bile duct
Fig 1 (a) Plasma PTX3 concentrations in 15 patients with sepsis and 11 healthy subjects. Squares are females, circles are males. (b) Data sets compared using Mann–Whitney U-test. Plasma PTX3 concentrations according to APACHE II score quartile in 15 patients with sepsis. Table 1 Characteristics of patients with sepsis. UTI, urinary tract infection; CBD, common bile duct Sex M/F Age (yr) APACHE II score ICU survival Hospital survival Infection site F 75 15 N N Pancreatic abscess F 56 24 Y Y UTI F 70 32 Y N Pneumonia M 49 17 Y Y Pneumonia M 45 31 Y Y Pneumonia M 73 19 N N Pneumonia F 54 30 Y Y Pneumonia F 81 32 Y N UTI M 74 18 Y Y Sepsis CBD stent M 85 27 N N Peritonitis M 32 25 Y Y Pneumonia M 83 21 Y Y Pneumonia M 39 12 Y Y Pneumonia M 22 9 Y Y Pneumonia F 55 24 Y Y Pneumonia To assess oxidative stress, we measured total plasma antioxidant capacity and plasma lipid hydroperoxides. Antioxidant capacity was significantly lower in patients with sepsis than healthy controls [0.99 (0.1–1.7) mM compared with 2.2 (1.3–3.3) mM, P=0.01, Fig. 2]. Lipid hydroperoxides were below the limit of detection in plasma from all healthy subjects. In patients with sepsis, lipid hydroperoxide levels were 3.32 (0.3–10.6) nM. There was no relationship between PTX3 and either antioxidant capacity or lipid hydroperoxides. Total antioxidant capacity and lipid hydroperoxide were not different between males and females. Fig 2 Plasma total antioxidant capacity in 15 patients with sepsis and 11 healthy subjects. Data sets compared using Mann–Whitney U-test.
Sex M/F Age (yr) APACHE II score ICU survival Hospital survival Infection site F 75 15 N N Pancreatic abscess F 56 24 Y Y UTI F 70 32 Y N Pneumonia M 49 17 Y Y Pneumonia M 45 31 Y Y Pneumonia M 73 19 N N Pneumonia F 54 30 Y Y Pneumonia F 81 32 Y N UTI M 74 18 Y Y Sepsis CBD stent M 85 27 N N Peritonitis M 32 25 Y Y Pneumonia M 83 21 Y Y Pneumonia M 39 12 Y Y Pneumonia M 22 9 Y Y Pneumonia F 55 24 Y Y Pneumonia To assess oxidative stress, we measured total plasma antioxidant capacity and plasma lipid hydroperoxides. Antioxidant capacity was significantly lower in patients with sepsis than healthy controls [0.99 (0.1–1.7) mM compared with 2.2 (1.3–3.3) mM, P=0.01, Fig. 2]. Lipid hydroperoxides were below the limit of detection in plasma from all healthy subjects. In patients with sepsis, lipid hydroperoxide levels were 3.32 (0.3–10.6) nM. There was no relationship between PTX3 and either antioxidant capacity or lipid hydroperoxides. Total antioxidant capacity and lipid hydroperoxide were not different between males and females. Fig 2 Plasma total antioxidant capacity in 15 patients with sepsis and 11 healthy subjects. Data sets compared using Mann–Whitney U-test. We also measured PTX3 secretion from human endothelial cells in vitro. In the absence of an inflammatory stimulus, endothelial cells produced detectable basal levels of PTX3 (Fig. 3). When we exposed cells to either LPS plus PepG, TNFα, IL-1β, TNFα plus IL-1β, or C. albicans, PTX3 levels in culture medium were significantly higher than in untreated cells (Fig. 3). There was no difference in PTX3 expression in endothelial cells exposed to C. albicans between 1 and 10 MOI (data not shown). PTX3 expression was greatest in cells exposed to both TNFα and IL-1α and was ∼10-fold higher than in control cells (P<0.001, Fig. 3). When we treated cells concurrently with antioxidants, PTX3 levels were lower, independent of the cell stimulus, than without antioxidants.
albicans between 1 and 10 MOI (data not shown). PTX3 expression was greatest in cells exposed to both TNFα and IL-1α and was ∼10-fold higher than in control cells (P<0.001, Fig. 3). When we treated cells concurrently with antioxidants, PTX3 levels were lower, independent of the cell stimulus, than without antioxidants. Fig 3 The effect of different inflammatory stimuli on PTX3 expression by human endothelial cells in vitro. Note the different scales on the two graphs. Data are from four replicate experiments conducted in triplicate. LPS, lipopolysaccharide (2 µg ml−1); PepG, peptidoglycan G (20 µg ml−1); TNFα, tumour necrosis factor α (10 ng ml−1); IL-1β, interleukin-1β (20 ng ml−1); C. albicans, heat-killed Candida albicans cells (MOI=3). Nuclear expression of NFκB was higher in cells treated with TNFα and IL-1β than controls (P<0.001, Fig. 4), and was lower in cells also treated with trolox or NAC, but not idebenone, compared with those treated with TNFα plus IL-1β alone (Fig. 4). Fig 4 The effect of antioxidants on PTX3 and NFκB expression in human endothelial cells treated with TNFα and IL-1β. Data are from four replicate experiments conducted in triplicate and were analysed by the Kruskal–Wallis with Mann–Whitney U post hoc testing. *Significantly lower than cells without antioxidant (P<0.001).
Fig 4 The effect of antioxidants on PTX3 and NFκB expression in human endothelial cells treated with TNFα and IL-1β. Data are from four replicate experiments conducted in triplicate and were analysed by the Kruskal–Wallis with Mann–Whitney U post hoc testing. *Significantly lower than cells without antioxidant (P<0.001). Discussion We showed that PTX3 was up-regulated in human endothelial cells in response to bacterial cell proteins, heat-killed Candida cells, or cytokines, with the greatest response occurring with a combination of TNFα plus IL-1β. Endothelial PTX3 response to an inflammatory stimulus was significantly reduced by treatment with NAC, trolox, or idebenone, and was associated with lower nuclear NFκB expression in NAC and trolox-treated cells. Idebenone had no effect on NFκB expression. We also showed that circulating PTX3 concentrations are higher in patients with sepsis compared with healthy subjects and are higher in female patients than males. PTX3 was also related to APACHE II score quartile. In addition, total antioxidant capacity was low and lipid hydroperoxide levels were elevated in the patients with sepsis, indicating oxidative stress. However, PTX3 concentrations in patients with sepsis were not related to antioxidant capacity, despite the regulation of PTX3 expression by antioxidants in vitro.
rtile. In addition, total antioxidant capacity was low and lipid hydroperoxide levels were elevated in the patients with sepsis, indicating oxidative stress. However, PTX3 concentrations in patients with sepsis were not related to antioxidant capacity, despite the regulation of PTX3 expression by antioxidants in vitro. PTX3 is a member of the pentraxin superfamily, a family of proteins characterized by a multimeric, usually pentameric structure.12 CRP, the prototype of the pentraxin family, is an acute-phase protein produced in the liver in response to inflammatory signals such as IL-6. PTX3 is the prototype of the long pentraxin family and has some similarities with the short pentraxins but differs with respect to cellular source, and ligand recognition. PTX3 is produced in response by a variety of cell types, including macrophages and monocytes, dendritic cells, fibroblasts, epithelial, and endothelial cells, in response to inflammatory signals such as TNFα, IL-1β, microbial proteins including LPS, lipoarabinomannans, and PepG, but not IL-6.1–5
and ligand recognition. PTX3 is produced in response by a variety of cell types, including macrophages and monocytes, dendritic cells, fibroblasts, epithelial, and endothelial cells, in response to inflammatory signals such as TNFα, IL-1β, microbial proteins including LPS, lipoarabinomannans, and PepG, but not IL-6.1–5 We found that exposure of endothelial cells to LPS, PepG, TNFα, or IL-1β resulted in PTX3 up-regulation, with an additive response when cells were treated with both TNFα and IL-1β, as previously reported.3–57 We also found that a ratio of between 1 and 10 heat-killed Candida cells to one endothelial cell resulted in an up-regulation of PTX3 expression by endothelial cells. Infection of mice with intra-cerebroventricular injection of live C. albicans was previously shown to result in up-regulation of PTX3 in the brain,19 but PTX3 expression by endothelial cells in response to C. albicans has not been described. Endothelial cells phagocytose live Candida cells, leading to secretion of cytokines including TNFα and IL-1β.20–23 The up-regulation of PTX3 expression found in our study may be related to TNFα and IL-1α produced as a result of phagocytosis of heat-killed Candida cells by the endothelial cells, although it has been reported that Candida cells killed with sodium periodate, although phagocytosed, are unable to elicit this cytokine response.2021 However, in another study, heat-killed Candida cells were able to stimulate monocytes to release pro-inflammatory cytokines,24 suggesting that the means of killing the cells may be relevant.
reported that Candida cells killed with sodium periodate, although phagocytosed, are unable to elicit this cytokine response.2021 However, in another study, heat-killed Candida cells were able to stimulate monocytes to release pro-inflammatory cytokines,24 suggesting that the means of killing the cells may be relevant. PTX3 blood levels are usually <10 ng ml−1 in healthy subjects and increase rapidly during inflammation and infection, leading to investigation of the potential of PTX3 as an early biomarker for sepsis. In a heterogeneous group of critically ill patients ranging from those with no infection to those with septic shock, it was reported that PTX3 levels were higher in the more severely ill patients.25 We also found that PTX3 levels were related to severity of illness as assessed by APACHE II score. The APACHE scores were not presented in that report,25 making further comparisons difficult. PTX3 has also been found to be increased in patients with acute respiratory distress syndrome.26 In a recent study in children with meningococcal disease, high PTX3 and low CRP concentrations at admission discriminated between the presence and the absence of shock.27 PTX3 did not correlate with paediatric risk of mortality, whereas CRP correlated negatively. PTX3 blood levels increase more rapidly than CRP and only loose2526 or no27 correlations are found between circulating levels of PTX3 and CRP. PTX3 has also been shown to be up-regulated in patients with dengue virus infection.28
X3 did not correlate with paediatric risk of mortality, whereas CRP correlated negatively. PTX3 blood levels increase more rapidly than CRP and only loose2526 or no27 correlations are found between circulating levels of PTX3 and CRP. PTX3 has also been shown to be up-regulated in patients with dengue virus infection.28 Marked oxidative stress has been consistently reported in patients with sepsis8202930 and acts as a trigger for up-regulation of many inflammatory responses. The human PTX3 proximal promoter contains binding sites for several transcription factors, including activator protein-1 (AP-1), NFκB, and nuclear factor-IL-6. The NFκB proximal site is essential for induction by IL-1β and TNFα,67 and PTX3 was shown to be regulated by NFκB.31 NFκB is redox-sensitive32 and has been shown to be involved in regulation of inflammatory responses in patients with sepsis: we and others have found elevated activation of NFκB linked to increased mortality3334 and inhibitable by NAC treatment, resulting in down-regulation of inflammatory mediators.35 We therefore investigated whether antioxidants were able to down-regulate PTX3 expression by endothelial cells in vitro. We found that NAC, trolox, and idebenone treatment all resulted in lower PTX3 levels on exposure to an inflammatory stimulus. In the case of NAC and trolox, this was accompanied by lower nuclear expression of NFκB. In a study using macrophages, the antioxidant pyrrolidine dithiocarbamate inhibited LPS-induced PTX3 expression in murine macrophages via an NFκB-dependent mechanism.31 However, there are other redox-sensitive transcription factors besides NFκB, such as AP-1,36 which may also be important for regulation of PTX3. Exogenous and endogenous antioxidants have been shown to be effective in blocking activation of NFκB either directly or indirectly, including NAC and trolox.3536 Although co-enzyme Q10 can inhibit NFκB,37 we can find no reports of the effect of idebenone on transcription, and so the mechanism of the down-regulation of PTX3 expression by idebenone remains unknown.
en shown to be effective in blocking activation of NFκB either directly or indirectly, including NAC and trolox.3536 Although co-enzyme Q10 can inhibit NFκB,37 we can find no reports of the effect of idebenone on transcription, and so the mechanism of the down-regulation of PTX3 expression by idebenone remains unknown. Oxidative stress can be defined as an imbalance between the production of reactive oxygen species and their removal by protective antioxidants. We measured lipid hydroperoxide levels and total antioxidant capacity in our patients with sepsis. Lipid peroxidation is the oxidative degradation of lipids and has been reported previously to be elevated in critically ill patients.8101230 Total antioxidant capacity reflects the ability of plasma to resist oxidative damage in vitro and has been widely used to assess the antioxidant status of patients with sepsis.8929 We found that lipid hydroperoxides were increased and total antioxidant capacity was decreased in our patients. The classical view of lipid peroxidation products is that they are essentially harmful, inducing and propagating chronic inflammatory reactions. However, it has been suggested that lipid peroxide breakdown products may promote cellular defence mechanisms,3839 such as induction of signalling pathways, resulting in up-regulation of anti-inflammatory mediators, and inhibition of signalling pathways which control pro-inflammatory mediator expression. Thus, the role of lipid hydroperoxides in inflammatory responses in sepsis is unclear. We found that neither lipid hydroperoxides nor antioxidant capacity was related to PTX3 levels, despite the effect of antioxidants on PTX3 expression in vitro. Although this was a small pilot study, there was a range of PTX values and oxidative stress was apparent, and so we would have expected some evidence of a relationship. These results may reflect the complex nature of the role of redox state involved in regulation of inflammatory pathways in vivo.
3 expression in vitro. Although this was a small pilot study, there was a range of PTX values and oxidative stress was apparent, and so we would have expected some evidence of a relationship. These results may reflect the complex nature of the role of redox state involved in regulation of inflammatory pathways in vivo. Other factors also influence PTX3 levels. In a study of more than 2000 healthy Japanese subjects, plasma PTX3 levels were reported to be lower in men than in women, higher in older age groups, and inversely correlated with triglycerides.40 We also found that PTX3 levels in patients with sepsis were highest in the female patients, but the skewed age range of patients with sepsis did not allow assessment of any age. The previous study of critically ill patients25 did not report sex distribution. It remains unclear as to the importance of the contribution of these confounding factors. In conclusion, we report that PTX3 expression in endothelial cells can be up-regulated by a variety of inflammatory mediators, including bacterial cell proteins, cytokines, and C. albicans and is down-regulated by NAC and trolox through an NFκB-mediated process. PTX3 levels are increased in patients with sepsis, and related to APACHE II score. Although we found evidence of oxidative stress in our patients, the regulation of PTX3 by antioxidants in vitro was not reflected clinically in this pilot study. Further work is needed to clarify regulatory mechanisms for PTX3 and its use as a biomarker in sepsis.
n patients with sepsis, and related to APACHE II score. Although we found evidence of oxidative stress in our patients, the regulation of PTX3 by antioxidants in vitro was not reflected clinically in this pilot study. Further work is needed to clarify regulatory mechanisms for PTX3 and its use as a biomarker in sepsis. Funding A.L.H. was an Anaesthetic Research Society vacation scholar. N.R.W. and H.F.G. are funded by the Medical Research Council, the British Journal of Anaesthesia, and the Intensive Care Society. N.A.R.G. acknowledges funding from the Wellcome Trust.
EEG is the method of choice in measuring the hypnotic component of anaesthesia. During the last decade, several measures have been introduced for this purpose. The most widely adopted EEG measure of anaesthetic drug effect is Bispectral Index (BIS™ monitor, Aspect Medical Systems, Newton, MA, USA).1 BIS utilizes a sensor placed in the forehead and temple of the patient to collect EEG. Although the approach is easy to use for anaesthetists, the placement is prone to facial EMG contamination,2 which is shown to represent impending arousal or nociception,3 even during deep hypnosis with propofol.4 Therefore, a single-number index like BIS can never tell with certainty the impact of EMG on the information received from the analysed signal. To at least partly overcome the problem described above, another EEG monitoring device, the M-ENTROPY™ module (GE Healthcare, Helsinki, Finland), was introduced.5 The M-ENTROPY™ module calculates the characteristics of the upper facial biosignal with an analysis of time–frequency balanced spectral entropy, taking into account the special characteristics of suppressed EEG signal. The resulting index of EEG and EMG activity is called Entropy, which has been shown to be a valid indicator of the hypnotic effect of propofol, thiopental, isoflurane, sevoflurane, and desflurane.6
sis of time–frequency balanced spectral entropy, taking into account the special characteristics of suppressed EEG signal. The resulting index of EEG and EMG activity is called Entropy, which has been shown to be a valid indicator of the hypnotic effect of propofol, thiopental, isoflurane, sevoflurane, and desflurane.6 Entropy consists of two parameters: State Entropy (SE) and Response Entropy (RE). SE is computed over the frequency range from 0.8 to 32 Hz, which represents the EEG-dominant part of the frequency spectrum. Therefore, SE reflects the cortical state of the patient. RE is computed over the frequency range from 0.8 to 47 Hz, covering both the EEG-dominant and the EMG-dominant areas of the spectrum. Consequently, RE−SE difference serves at least partly as an indicator of upper facial EMG activation.5 The RE and SE are both calculated with 5 s intervals from the preceding 1.92 and 15 s segments of EEG, respectively, and the new values are shown on the screen and output from the monitor. The appearance of upper facial EMG indicates that the patient is responding to an external stimulus,7 which is usually nociceptive. Therefore, RE−SE difference may indicate nociception or inadequate anaesthesia, an assumption supported by one recently published article,8 but disputed by another.9
The RE and SE are both calculated with 5 s intervals from the preceding 1.92 and 15 s segments of EEG, respectively, and the new values are shown on the screen and output from the monitor. The appearance of upper facial EMG indicates that the patient is responding to an external stimulus,7 which is usually nociceptive. Therefore, RE−SE difference may indicate nociception or inadequate anaesthesia, an assumption supported by one recently published article,8 but disputed by another.9 The present study was designed to prospectively investigate RE, SE, RE−SE difference, and upper facial EMG during propofol–nitrous oxide or propofol–nitrous oxide–remifentanil anaesthesia in patients without neuromuscular blocking agent (NMBA) medication (i) before and after intubation and (ii) before and after beginning of laparoscopic surgery. The primary aim of the study was to evaluate the behaviour of RE, SE, and RE−SE difference between the groups. Secondary aims were to compare the Entropy indices between intubation and surgery, to study the impact of visually verified EMG on the indices, and to make comparisons between heart rate (HR) and RE−SE difference. As remifentanil is a very potent opioid, we hypothesized that smaller increase in RE, and therefore, smaller RE−SE difference would be seen in patients receiving remifentanil as an adjuvant of their propofol−nitrous oxide anaesthesia. We also hypothesized that EMG would react more precisely than EEG to nociceptive stimuli, and that changes of RE−SE difference would associate with alterations in HR. To our knowledge, this is the first clinical study where the appearance of EMG is verified visually both from the original biosignal and from its spectral presentation, and compared with the behaviour of the Entropy parameters.
eptive stimuli, and that changes of RE−SE difference would associate with alterations in HR. To our knowledge, this is the first clinical study where the appearance of EMG is verified visually both from the original biosignal and from its spectral presentation, and compared with the behaviour of the Entropy parameters. Methods Patients This study followed the design of a prospective clinical study. The local Ethics Committee and Finnish National Agency for Medicines approved the study protocol. All patients gave their written informed consent. Inclusion criteria were the following: patients undergoing elective laparoscopic gynaecological surgery (expected duration >30 min) between ages 18 and 60 yr, and ASA physical status I or II. Patients were excluded if they had a disease or injury affecting the central nervous system, alcohol or drug abuse, or BMI >28. All patients fasted overnight before surgery. A total of 33 patients were enrolled.
naecological surgery (expected duration >30 min) between ages 18 and 60 yr, and ASA physical status I or II. Patients were excluded if they had a disease or injury affecting the central nervous system, alcohol or drug abuse, or BMI >28. All patients fasted overnight before surgery. A total of 33 patients were enrolled. Electroencephalogram acquisition The forehead biosignal was collected with a disposable electrode strip (Entropy Sensor, GE Healthcare) for Entropy measurement. After degreasing of the forehead skin using isopropanol 70%, the strip was positioned as recommended by the manufacturer. The signal was acquired from two electrodes of the strip: one frontally in the midline, 2 cm above the eyebrows, and the other 2 cm laterally from the outer canthus of the left eye. The first electrode is on the frontal muscle, the second on the orbicularis oculi and temporal muscles. Entropy was collected with an M-ENTROPY™ module of the S/5™ Anesthesia Monitor (GE Healthcare). The sampling rate was 400 Hz. High- and low-pass filters of 0.5 and 118 Hz (−3 dB; 10 dB per decade), respectively, were applied. Power line artifact was not filtered. All the monitored values were collected on a laptop computer.
py was collected with an M-ENTROPY™ module of the S/5™ Anesthesia Monitor (GE Healthcare). The sampling rate was 400 Hz. High- and low-pass filters of 0.5 and 118 Hz (−3 dB; 10 dB per decade), respectively, were applied. Power line artifact was not filtered. All the monitored values were collected on a laptop computer. Anaesthesia and study protocol RE, SE, and vital signs were collected under two anaesthetic regimens, both of which included premedication with diazepam 10 mg p.o. An i.v. route was established for all patients and an infusion of isotonic saline was started. Intermittent non-invasive arterial pressure was recorded every 5 min. Electrocardiogram, inspired fractions and end-tidal concentrations of anaesthetic gases and CO2, and peripheral oxygen saturation were continuously monitored with the Datex-Ohmeda S/5™ Anesthesia Monitor.
d an infusion of isotonic saline was started. Intermittent non-invasive arterial pressure was recorded every 5 min. Electrocardiogram, inspired fractions and end-tidal concentrations of anaesthetic gases and CO2, and peripheral oxygen saturation were continuously monitored with the Datex-Ohmeda S/5™ Anesthesia Monitor. Entropy monitoring started before induction of anaesthesia and continued uninterrupted until the end of the study. Seventeen patients were anaesthetized with propofol 3 mg kg−1 i.v., followed by manually controlled ventilation of nitrous oxide 67% in oxygen via face mask. Endotracheal intubation was performed, as gently as possible, 150 s later. If intubation difficulties were met, mask ventilation was re-started and another intubation attempt was done 1–3 min later. If the second attempt was unsuccessful, rocuronium was given and the patient was discarded from the study. After securing the airway, controlled mechanical ventilation was started with a fresh flow of 6 litre min−1 (67% nitrous oxide in oxygen), and propofol infusion was started. The infusion rate of propofol was adjusted to keep the Entropy parameters between 40 and 60, the target being 50.
nt was discarded from the study. After securing the airway, controlled mechanical ventilation was started with a fresh flow of 6 litre min−1 (67% nitrous oxide in oxygen), and propofol infusion was started. The infusion rate of propofol was adjusted to keep the Entropy parameters between 40 and 60, the target being 50. Sixteen patients were anaesthetized the same way as described above, but target-controlled infusion (TCI, Asena™ PK, Alaris Medical Systems, Basingstoke, UK) of remifentanil was added to the anaesthetic regimen. Remifentanil was infused with the estimated effect-site concentration (Ce) of 4.0 ng ml−1 from the beginning of anaesthetic induction (the pharmacokinetic model of Minto and colleagues)10 until the end of the study. Laryngoscopy for endotracheal intubation was initiated after reaching the remifentanil Ce of 4.0 ng ml−1. After endotracheal intubation, patients were not touched or otherwise disturbed for 5 min to discover the magnitude of RE−SE difference, and the presence or absence of EMG, during anaesthesia without surgery. After 5 min, the patient was prepared for operation and permission to start laparoscopy was granted. A propofol bolus of 50 mg was given as a rescue medication if the Entropy values exceeded 60, and repeated 90 s later, if needed. The study was completed 1 min after setting the first laparoscopy trochar. Thereafter, the administration of nitrous oxide was discontinued and patient's lungs were ventilated with air/oxygen. Fentanyl (the propofol group) and rocuronium were given according to clinical needs.
ed 60, and repeated 90 s later, if needed. The study was completed 1 min after setting the first laparoscopy trochar. Thereafter, the administration of nitrous oxide was discontinued and patient's lungs were ventilated with air/oxygen. Fentanyl (the propofol group) and rocuronium were given according to clinical needs. Analyses of the biosignal RE and SE values were analysed off-line as a mean of 15 s (three consecutive readings) at the following time points: awake, 90 s after anaesthetic induction, 30 s before intubation, 30 s after intubation, after a 5 min equilibrium period, 60 s before skin incision, 30 s after skin incision, 30 s after setting of the needle of Veress, 30 s after beginning of gas insufflation, at the end of gas insufflation, and 30 s after setting of the first laparoscopy trochar. RE−SE was calculated by subtracting the SE value from the corresponding RE value.
um period, 60 s before skin incision, 30 s after skin incision, 30 s after setting of the needle of Veress, 30 s after beginning of gas insufflation, at the end of gas insufflation, and 30 s after setting of the first laparoscopy trochar. RE−SE was calculated by subtracting the SE value from the corresponding RE value. Spectrogram, that is, a presentation of the spectral content of a biosignal, was produced with Somnologica™ sleep analysis program (Medcare Flaga, Reykjavik, Iceland). Power spectra of consecutive 1.0 s samples of the signal are calculated and presented vertically along y-axis, plotted against time in x-axis. The density (i.e. ‘darkness’) of spectrogram reveals the amount of activity at respective frequency. Thus, spectrogram presents the same information as successive power spectra, but in a compressed form. Power line artifact is seen in spectrogram as a sharp activity band at 50 or 60 Hz, and ECG or EOG artifacts are typically located at rather low frequency range. Therefore, EMG is virtually the only artifact that is displayed over a wide frequency range.
nformation as successive power spectra, but in a compressed form. Power line artifact is seen in spectrogram as a sharp activity band at 50 or 60 Hz, and ECG or EOG artifacts are typically located at rather low frequency range. Therefore, EMG is virtually the only artifact that is displayed over a wide frequency range. Both the raw biosignal and a spectrogram at the time points of interest were inspected visually off-line, without knowledge of the behaviour of the Entropy indices, by an experienced clinical neurophysiologist (V.J.), to judge the presence or absence of EMG. Later on, the traces and classifications were analysed jointly by all authors, to ensure the agreement of the detections. To demonstrate the effect of EMG on the EEG spectrum, spectral analyses before and after intubation or commencement of surgery were drawn in four patients. In analyses, a window of 15 s was used. All patients were interviewed during the first postoperative day, regarding their possible anaesthesia and surgery-related memories and intubation-associated sequelae.
Both the raw biosignal and a spectrogram at the time points of interest were inspected visually off-line, without knowledge of the behaviour of the Entropy indices, by an experienced clinical neurophysiologist (V.J.), to judge the presence or absence of EMG. Later on, the traces and classifications were analysed jointly by all authors, to ensure the agreement of the detections. To demonstrate the effect of EMG on the EEG spectrum, spectral analyses before and after intubation or commencement of surgery were drawn in four patients. In analyses, a window of 15 s was used. All patients were interviewed during the first postoperative day, regarding their possible anaesthesia and surgery-related memories and intubation-associated sequelae. Statistical methods The study was designed to have a power of 80% to detect statistical significance, assuming two-sided α-level of 0.05, with surgery-associated mean RE−SE differences of 7.0 and 2.0 units (sd 4.5 units) between the propofol and the propofol−remifentanil groups, respectively. The power calculation was based on the previous unpublished Entropy data of our own. To meet the criteria of power calculation, 14 patients per group were needed. All statistical analyses were performed using SPSS for Windows software (version 16.0, SPSS, Chicago, IL, USA). One-way analysis of variance followed by t-tests was used for parametric data, and Mann–Whitney and χ2 tests were used for non-parametric comparisons, where appropriate. Pearson's correlation analysis was used to compare the change of HR with that of RE−SE difference at endotracheal intubation. P<0.05 was considered statistically significant.
ance followed by t-tests was used for parametric data, and Mann–Whitney and χ2 tests were used for non-parametric comparisons, where appropriate. Pearson's correlation analysis was used to compare the change of HR with that of RE−SE difference at endotracheal intubation. P<0.05 was considered statistically significant. Results Two enrolled patients, both receiving propofol−nitrous oxide without remifentanil, were excluded because of problems in laryngoscopy and intubation. All others were successfully intubated with one or two attempts. Therefore, 15 patients in the propofol−nitrous oxide group and 16 patients in the propofol−nitrous oxide−remifentanil group were studied. Patient characteristics did not differ between the groups. The mean ages of patients not receiving remifentanil and those with remifentanil were 39 vs 32 yr, mean weight was 62 vs 61 kg, and mean height was 166 vs 165 cm, respectively. The RE or SE values did not differ between the groups along the study. The RE−SE difference showed significant between-group difference awake and from skin incision to the end of the study. Awake, the mean of propofol−nitrous oxide group was 1.22 units higher (P<0.042). Therefore, the groups were baseline-corrected by subtracting 1.22 units from all means in the propofol−nitrous oxide group. After such correction, the only detected between-group difference was after setting the first laparoscopy trochar. The behaviour of RE, SE, baseline-corrected RE−SE difference, and HR in both groups is presented in Figure 1.
aseline-corrected by subtracting 1.22 units from all means in the propofol−nitrous oxide group. After such correction, the only detected between-group difference was after setting the first laparoscopy trochar. The behaviour of RE, SE, baseline-corrected RE−SE difference, and HR in both groups is presented in Figure 1. Fig 1 The behaviour of RE, SE, baseline-corrected RE−SE, and HR [mean (sd)] during the study in patients receiving propofol–nitrous oxide or propofol−nitrous oxide−remifentanil anaesthesia. 1, awake; 2, 90 s after anaesthetic induction; 3, 30 s before intubation; 4, 30 s after intubation; 5, after a 5 min equilibrium period; 6, 60 s before skin incision; 7, 30 s after skin incision; 8, 30 s after setting of the needle of Veress; 9, 30 s after beginning of gas insufflation; 10, at the end of gas insufflation; 11, 30 s after setting of the first laparoscopy trochar. ♣P<0.05, ♣♣P<0.01 within the propofol–nitrous oxide group, respectively. #P<0.05, ##P<0.01, and *P<0.05, **P<0.01 within the propofol–nitrous oxide–remifentanil group, and between the groups, respectively. The SE and RE values increased more commonly after intubation than after skin incision (Table 1). RE increased ≥10 units in 27 events. This increase in RE values was rapidly followed by ≥10 units increase of SE in 21/27 cases, thereby decreasing the calculated RE−SE difference.
Fig 1 The behaviour of RE, SE, baseline-corrected RE−SE, and HR [mean (sd)] during the study in patients receiving propofol–nitrous oxide or propofol−nitrous oxide−remifentanil anaesthesia. 1, awake; 2, 90 s after anaesthetic induction; 3, 30 s before intubation; 4, 30 s after intubation; 5, after a 5 min equilibrium period; 6, 60 s before skin incision; 7, 30 s after skin incision; 8, 30 s after setting of the needle of Veress; 9, 30 s after beginning of gas insufflation; 10, at the end of gas insufflation; 11, 30 s after setting of the first laparoscopy trochar. ♣P<0.05, ♣♣P<0.01 within the propofol–nitrous oxide group, respectively. #P<0.05, ##P<0.01, and *P<0.05, **P<0.01 within the propofol–nitrous oxide–remifentanil group, and between the groups, respectively. The SE and RE values increased more commonly after intubation than after skin incision (Table 1). RE increased ≥10 units in 27 events. This increase in RE values was rapidly followed by ≥10 units increase of SE in 21/27 cases, thereby decreasing the calculated RE−SE difference. Table 1 Number of patients with small (<10), moderate (10 to ≤30) or large (>30 units) change in SE or RE values at intubation or skin incision. The mean of three consecutive index values before intubation or skin incision is subtracted from that of three consecutive values 1 min after the event. *P=0.006, **P=0.001 for SE and RE change values, respectively, between intubation and skin incision (two-tailed Mann–Whitney test)
lues at intubation or skin incision. The mean of three consecutive index values before intubation or skin incision is subtracted from that of three consecutive values 1 min after the event. *P=0.006, **P=0.001 for SE and RE change values, respectively, between intubation and skin incision (two-tailed Mann–Whitney test) Magnitude of change in index values Intubation*, ** Incision Propofol Propofol–remifentanil Propofol Propofol–remifentanil SE RE SE RE SE RE SE RE <10 units 7 4 8 7 9 9 16 15 10 to ≤30 units 4 7 6 5 5 5 0 1 >30 units 4 4 2 4 1 1 0 0 Σ 15 15 16 16 15 15 16 16 When all blindly analysed episodes of interventions were pooled together, the EMG was present more frequently in recordings of patients in the propofol group (79 EMG findings in 165 episodes) than in the propofol–remifentanil group (33 findings in 176 episodes). Incidence of EMG in both groups and the association between EMG and Entropy RE−SE difference are given in Table 2. In some patients, EMG changed the power spectrum of the biosignal from 15 Hz up to ∼150 Hz, that is, to the highest detectable frequencies. The presence of EMG in the raw biosignal of an example patient, and its effect on power spectrum, Entropy values, and spectrogram, is depicted in Figure 2.
ce are given in Table 2. In some patients, EMG changed the power spectrum of the biosignal from 15 Hz up to ∼150 Hz, that is, to the highest detectable frequencies. The presence of EMG in the raw biosignal of an example patient, and its effect on power spectrum, Entropy values, and spectrogram, is depicted in Figure 2. Fig 2 (a) The appearance of EMG activity in EEG signal after laryngoscopy (arrow downward) and attempted intubation in a patient not receiving remifentanil. EEG signal remains suppressed, despite strong EMG contamination. (b) Power spectrum of the EEG signal before (black line) and after (grey line) laryngoscopy in the same situation as in (a). EMG activity changes the spectrum after laryngoscopy. The spectrum with EMG impact starts to change already below 20 Hz. Dotted vertical line represents the 32 Hz frequency, which is used in Entropy calculation to differentiate EEG activity (<32 Hz) from EMG activity (>32 Hz). In Entropy calculation, low SE value increases rapidly after appearance of EMG, because the power at <32 Hz area increases. (c) EEG spectrogram (frequency vs time) of the same patient as in (a) and (b). Fast activity disappears in the beginning of anaesthesia (0–30 s). Continuous activity below 30 Hz is merely EEG. Placement of oropharyngeal airway (OA) at 1 min 40 s elicits short EMG response, shown as a vertical bar up to 150 Hz. Laryngoscopy and attempted intubation (IA) at arrow is associated with longer EMG response. Replacement of oropharyngeal airway at 3 min 50 s, and successful intubation (Intub) at 4 min 30 s. Long-lasting EMG activity ensues.
al airway (OA) at 1 min 40 s elicits short EMG response, shown as a vertical bar up to 150 Hz. Laryngoscopy and attempted intubation (IA) at arrow is associated with longer EMG response. Replacement of oropharyngeal airway at 3 min 50 s, and successful intubation (Intub) at 4 min 30 s. Long-lasting EMG activity ensues. Table 2 The occurrence of visible EMG in EEG signal along the course of study in patients receiving propofol–nitrous oxide or propofol–remifentanil–nitrous oxide anaesthesia, and the association of EMG with RE–SE difference
al airway (OA) at 1 min 40 s elicits short EMG response, shown as a vertical bar up to 150 Hz. Laryngoscopy and attempted intubation (IA) at arrow is associated with longer EMG response. Replacement of oropharyngeal airway at 3 min 50 s, and successful intubation (Intub) at 4 min 30 s. Long-lasting EMG activity ensues. Table 2 The occurrence of visible EMG in EEG signal along the course of study in patients receiving propofol–nitrous oxide or propofol–remifentanil–nitrous oxide anaesthesia, and the association of EMG with RE–SE difference Time point EMG present/absent (n) RE–SE [mean (sd)] Propofol Propofol–remifentanil P-value (Fisher's exact test) EMG present EMG absent P-value (t-test) Awake 15/0 16/0 1.0 8.96 (1.70) Induction+90 s 5/10 4/12 0.704 2.44 (2.35) 1.72 (1.83) 0.43 Before intubation 1/14 0/16 0.484 12 (0) 1.73 (1.86) <0.001 After intubation 15/0 13/3 0.226 9.96 (7.6) 0.67 (0.33) <0.001 Steady state 4/11 0/16 0.043 8 (1.63) 2.93 (3.6) 0.001 Before skin incision 2/13 0/16 0.226 12 (1.41) 2.52 (2.73) 0.027 After skin incision 6/9 0/16 0.007 12.17 (3.49) 1.96 (2.13) <0.001 Needle of Veress 6/9 0/16 0.007 13 (6.48) 2.28 (3.01) 0.009 Start of gas insufflation 7/8 0/16 0.002 7.57 (5.16) 1.83 (2.44) 0.025 End of gas insufflation 8/7 0/16 0.001 8.38 (4.5) 1.52 (1.81) 0.003 Laparoscopy trochar 10/5 0/16 <0.001 9.1 (2.92) 1.62 (1.96) <0.001 Figure 3 presents the spectrograms from the whole study period in two patients, one in the propofol group and another in the propofol−remifentanil group. In some patients, the EMG was seen on the top of suppressed EEG signal in association with noxious stimulus (Fig. 4).
r 10/5 0/16 <0.001 9.1 (2.92) 1.62 (1.96) <0.001 Figure 3 presents the spectrograms from the whole study period in two patients, one in the propofol group and another in the propofol−remifentanil group. In some patients, the EMG was seen on the top of suppressed EEG signal in association with noxious stimulus (Fig. 4). Fig 3 EEG spectrograms presenting the whole study periods in two patients: (a) without and (b) with remifentanil infusion. Fast activity disappears in the beginning of anaesthesia. EEG activity is seen in both recordings as a continuous activity below 30 Hz. EMG contamination is seen as vertical bars up to 150 Hz. AI, attempted intubation; I, intubation; PP, patient positioning; SI, skin incision. Laryngoscopy and intubation are associated with strong EMG activity in both recordings. Thereafter, no EMG is seen in (b), that is, the patient receiving remifentanil. In (a), both noxious and non-noxious stimuli (like positioning the patient) elicit EMG activity. Horizontal bars at 50 and 150 Hz are power line (50 Hz) artifacts and its harmonic (150 Hz). Vertical bars (ImpCh) indicate automatic Entropy sensor impedance checks every 10 min. Fig 4 Two successive 40 s samples of EEG which show development of EMG after intubation (arrow upward) during EEG burst suppression. The box below is an enlarged 4 s piece of the signal, showing the beginning of EMG activity.
Fig 3 EEG spectrograms presenting the whole study periods in two patients: (a) without and (b) with remifentanil infusion. Fast activity disappears in the beginning of anaesthesia. EEG activity is seen in both recordings as a continuous activity below 30 Hz. EMG contamination is seen as vertical bars up to 150 Hz. AI, attempted intubation; I, intubation; PP, patient positioning; SI, skin incision. Laryngoscopy and intubation are associated with strong EMG activity in both recordings. Thereafter, no EMG is seen in (b), that is, the patient receiving remifentanil. In (a), both noxious and non-noxious stimuli (like positioning the patient) elicit EMG activity. Horizontal bars at 50 and 150 Hz are power line (50 Hz) artifacts and its harmonic (150 Hz). Vertical bars (ImpCh) indicate automatic Entropy sensor impedance checks every 10 min. Fig 4 Two successive 40 s samples of EEG which show development of EMG after intubation (arrow upward) during EEG burst suppression. The box below is an enlarged 4 s piece of the signal, showing the beginning of EMG activity. The mean time intervals from induction of anaesthesia to intubation were 161 and 168 s for groups without and with remifentanil, respectively (NS). The same figures for intervals from anaesthetic induction to skin incision were 29 and 32 min, respectively (NS).
Fig 4 Two successive 40 s samples of EEG which show development of EMG after intubation (arrow upward) during EEG burst suppression. The box below is an enlarged 4 s piece of the signal, showing the beginning of EMG activity. The mean time intervals from induction of anaesthesia to intubation were 161 and 168 s for groups without and with remifentanil, respectively (NS). The same figures for intervals from anaesthetic induction to skin incision were 29 and 32 min, respectively (NS). HR was similar in both groups, awake and 90 s after induction of anaesthesia. From intubation to the end of the study period, HR was higher in patients without remifentanil (Fig. 1). The correlation between the change of HR and that of RE−SE difference at endotracheal intubation was poor: r=0.179 (P=0.336). Rescue medication due to elevated SE values was needed 18 times in 12 patients (nine in the propofol group and three in the propofol–remifentanil group). None of the patients, however, had recall, sore throat, or other complaints in the postoperative interview.
HR was similar in both groups, awake and 90 s after induction of anaesthesia. From intubation to the end of the study period, HR was higher in patients without remifentanil (Fig. 1). The correlation between the change of HR and that of RE−SE difference at endotracheal intubation was poor: r=0.179 (P=0.336). Rescue medication due to elevated SE values was needed 18 times in 12 patients (nine in the propofol group and three in the propofol–remifentanil group). None of the patients, however, had recall, sore throat, or other complaints in the postoperative interview. Discussion The first finding in our study was that RE, SE, and RE−SE difference increased in both groups during laryngoscopy and endotracheal intubation, along with continuous EMG activity in the biosignal, in spite of remifentanil (4.0 ng ml−1) in one group. The second finding was that the intubation-associated increase in RE−SE difference was only transient, because the increase in RE was often rapidly followed by an increase in SE. These findings suggest that RE−SE difference cannot be used as a long-lasting, reliable indicator of nociception. Although regularly detected in the initial phase of stimulation, RE−SE difference soon disappears due to increasing SE.
transient, because the increase in RE was often rapidly followed by an increase in SE. These findings suggest that RE−SE difference cannot be used as a long-lasting, reliable indicator of nociception. Although regularly detected in the initial phase of stimulation, RE−SE difference soon disappears due to increasing SE. The small, though statistically significant, difference in RE−SE awake was most likely due to random factors,11 because no abnormal EEG was seen in visual evaluation. However, this difference in RE−SE awake was taken into account by normalizing the groups in the way described in Results. After baseline correction according to the awake values, the RE−SE difference was higher in patients not receiving remifentanil only at setting of the first trochar. The third finding was that intubation was associated with greater increases in RE, SE, and RE−SE difference than commencement of surgery. None of the indices showed increased values at skin incision. This is in line with earlier literature, ranking endotracheal intubation as a stronger nociceptive stimulus than skin incision.12 The fourth and perhaps the most significant finding was that increased index values during nociceptive stimuli were strongly associated with the presence of EMG, confirmed by visual analysis of the raw biosignal and by the analysis of the spectrogram; both performed without knowledge of anaesthesia regimen. Spectral analysis of the biosignal demonstrated an EMG-associated change in the spectrum at <20–100 Hz.
ociceptive stimuli were strongly associated with the presence of EMG, confirmed by visual analysis of the raw biosignal and by the analysis of the spectrogram; both performed without knowledge of anaesthesia regimen. Spectral analysis of the biosignal demonstrated an EMG-associated change in the spectrum at <20–100 Hz. After endotracheal intubation, no continuous EMG activity in the raw signal, or significant RE−SE difference in Entropy values, was noted in patients receiving remifentanil. EMG was more frequently present during surgical manipulation in patients without remifentanil. The reactivity of upper facial biosignal was seen in nearly all patients during intubation, in spite of lower RE and SE values than was the case during commencement of surgery. This could be explained by the high stimulus intensity of laryngoscopy and intubation1314 that elicits EMG activity also in patients receiving remifentanil. Another postulation might be the modelling of remifentanil pharmacokinetics: as the estimated effect-site remifentanil concentration is based on the cortical electrical activity,10 it is theoretically possible that the drug concentration was lower than assumed in the areas controlling EMG. The EMG recorded is likely to be both from facial- and trigeminal-innervated muscles as the electrodes are on frontal and temporal muscles. The pathways to frontal and temporal muscles are the facial and trigeminal nerves, respectively, both originating from the brain stem.15 The Ke0 for remifentanil effect at the brain stem might differ from that measured as a change in EEG. The mean duration of 30 min from anaesthetic induction to skin incision, on the other hand, was sufficient to complete the saturation of remifentanil, abolishing the reactivity of EMG. Finally, as one of the electrodes in the Entropy strip is placed on the patient's temple, it is possible that the EMG activity during laryngoscopy and intubation comes partly from the temporal muscle, instead of the frontal muscle. Further research is needed to elucidate the behaviour of EMG during intubation.
of EMG. Finally, as one of the electrodes in the Entropy strip is placed on the patient's temple, it is possible that the EMG activity during laryngoscopy and intubation comes partly from the temporal muscle, instead of the frontal muscle. Further research is needed to elucidate the behaviour of EMG during intubation. In analysis of Entropy and raw EEG signal, the presence of EMG dictated Entropy values during nociceptive events. Typically, RE values increased with short delay, as described earlier by Vakkuri and colleagues.6 In the majority of cases, the increase of RE was soon followed by an increase in SE values, decreasing the RE−SE difference. The increase in SE values was most probably due to the increase in biosignal activity below 32 Hz, as depicted in Figure 2. While all activity below 32 Hz is regarded as EEG in the analysis of Entropy,5 strong EMG impact starts to change the spectrum already at 20 Hz.16 This may explain the unexpected behaviour of SE during nociceptive stimuli, when anaesthesia regimen does not include adequate anti-nociception or effective neuromuscular block. Nociception-associated EMG activity may increase SE values, because the content of the biosignal starts to change already below 20 Hz (Fig. 2).
his may explain the unexpected behaviour of SE during nociceptive stimuli, when anaesthesia regimen does not include adequate anti-nociception or effective neuromuscular block. Nociception-associated EMG activity may increase SE values, because the content of the biosignal starts to change already below 20 Hz (Fig. 2). HR is a standard indicator of nociception. In this study, HR differed between the groups from intubation to the end of the study. Higher values were seen in patients without remifentanil. This may indicate stronger nociceptive input in these patients, although the direct effect of remifentanil on HR should also be considered.17 The poor correlation between HR and RE−SE difference was most probably due to the widespread nociception-elicited alterations in EEG, reducing the RE−SE difference, as described above. On the basis of our results, it is appropriate to conclude that Entropy RE−SE difference may increase during strong nociceptive stimuli, but this increase is often transient, due to increasing SE values. Therefore, although high SE values serve as an indicator of impending awareness, EMG contamination can associate with high SE values during otherwise sufficient hypnotic state of anaesthesia. Adequate anti-nociceptive medication, however, decreases the risk of EMG contamination, as seen in our patients receiving remifentanil.
re, although high SE values serve as an indicator of impending awareness, EMG contamination can associate with high SE values during otherwise sufficient hypnotic state of anaesthesia. Adequate anti-nociceptive medication, however, decreases the risk of EMG contamination, as seen in our patients receiving remifentanil. One might argue that increasing SE levels in our study indicate less adequate anaesthetic depth and impending arousal. From a theoretical point of view, that is well possible. However, the connection between EMG, spectral characteristics of EEG, and SE was well demonstrated in the present study. EMG is generated at brain stem level; therefore, the presence of EMG is not a direct indicator of awareness. Moreover, appearance of EMG does not necessarily alter the underlying EEG signal during anaesthesia. As an example, in Figure 4, the burst suppression EEG signal remained unaltered, although strong EMG activity contaminated the signal. Finally, none of our patients had recall despite high SE values, although the study remains underpowered to examine awareness. The biosignal was visually analysed by a single neurophysiologist with considerable experience on anaesthesia-related EEG. No artifacts that could contribute significantly to the frequency bands above 32 Hz were detected in this analysis. As shown in Figure 3, electrical noise appears as a narrow band at 50 Hz, whereas electro-oculogram and ECG artifacts are limited at low frequency areas, instead of widespread EMG activity. With the aid of spectral presentation, EMG is reliably detected in EEG.
requency bands above 32 Hz were detected in this analysis. As shown in Figure 3, electrical noise appears as a narrow band at 50 Hz, whereas electro-oculogram and ECG artifacts are limited at low frequency areas, instead of widespread EMG activity. With the aid of spectral presentation, EMG is reliably detected in EEG. The intraoperative EMG contamination of EEG is challenging for an anaesthetist, because such low-amplitude activity is not always readily seen on monitor, even though the displaying of ‘raw’ EEG signal is highly recommended. The EMG-associated increase in SE value does not directly possess the risk of unintentional awareness during anaesthesia, because high SE value usually leads to re-adjustment of hypnotics. When remaining undetected, the situation could, however, cause inappropriate and potentially harmful use of anaesthetics.
ommended. The EMG-associated increase in SE value does not directly possess the risk of unintentional awareness during anaesthesia, because high SE value usually leads to re-adjustment of hypnotics. When remaining undetected, the situation could, however, cause inappropriate and potentially harmful use of anaesthetics. The anaesthetic regimen of this study does not reflect our routine clinical practice and necessitates a comment. Our aim was to study the magnitude of RE−SE difference with or without strong anti-nociceptive medication, and without confounding impact of NMBAs. Nitrous oxide, however, was used as a baseline anti-nociceptive agent in our intubated patients under controlled ventilation. In most cases, endotracheal intubation and commencement of surgery went smoothly, whereas difficulties in intubation were faced in two patients, who were discarded from the study. Owing to the gentle manipulating techniques and used rescue medication, none of the patients reported any anaesthesia-related sequelae. With this technique, we were able to demonstrate the effects of EMG on the Entropy indices RE and, especially, SE. As neuromuscular blockers were not used, the impact of them remains to be studied later on.
nipulating techniques and used rescue medication, none of the patients reported any anaesthesia-related sequelae. With this technique, we were able to demonstrate the effects of EMG on the Entropy indices RE and, especially, SE. As neuromuscular blockers were not used, the impact of them remains to be studied later on. In conclusion, we showed that Entropy RE−SE difference cannot reliably be used as an indicator of nociception in patients anaesthetized with propofol, nitrous oxide, and remifentanil without NMBAs. EMG activity can contaminate the interpretation, especially by increasing SE values. Further research is needed to elucidate the effects of EMG in detail, and to study the effect of neuromuscular blocking agents on EMG signal.
Complex cardiac surgery is frequently accompanied by excessive perioperative bleeding because of coagulation system impairment, inadequate surgical haemostasis, or both.1 Bleeding increases the risk of re-exploration, allogeneic blood transfusion, or perioperative myocardial infarction, and consequently, associated morbidity and mortality.2 Aortic valve operation and ascending aorta replacement (AV–AA) typically involves hypothermia, prolonged cardiopulmonary bypass (CPB), and large graft anastomoses, and is associated with an increased risk of intra- and postoperative blood loss and high transfusion rates.34 Conventional haemostatic therapy consists of transfusion of allogeneic blood products that include fresh-frozen plasma (FFP), platelet concentrate, and cryoprecipitate. However, although the use of these products was developed empirically, their haemostatic efficacy has not been evaluated thoroughly in the surgical setting.56
ional haemostatic therapy consists of transfusion of allogeneic blood products that include fresh-frozen plasma (FFP), platelet concentrate, and cryoprecipitate. However, although the use of these products was developed empirically, their haemostatic efficacy has not been evaluated thoroughly in the surgical setting.56 Haemocomplettan® P (brand name in Europe)/Riastap (brand name in USA) (CSL Behring, Marburg, Germany) is a highly purified, lyophilized, virus-inactivated fibrinogen concentrate obtained from human plasma that can be rapidly reconstituted without the need for thawing and cross-matching, which are necessary for FFP and cryoprecipitate. The administration of fibrinogen concentrate was originally reserved for replacement therapy in congenital fibrinogen deficiency, and in the USA, Riastap is only approved for this indication. In the meantime, European reports on haemostatic therapy with Haemocomplettan® P in acquired perioperative deficiency of fibrinogen have been published.7–11 Acquired fibrinogen deficiency occurring during and after CPB is associated with increased bleeding after cardiac surgery.1213 However, the haemostatic efficacy of fibrinogen concentrate in correcting such deficiency in complex cardiac surgery has not been investigated to date.
cy of fibrinogen have been published.7–11 Acquired fibrinogen deficiency occurring during and after CPB is associated with increased bleeding after cardiac surgery.1213 However, the haemostatic efficacy of fibrinogen concentrate in correcting such deficiency in complex cardiac surgery has not been investigated to date. To reduce blood component transfusion in cardiac surgery, point-of-care methods such as thrombelastography/thromboelastometry have been applied in algorithms supporting bleeding management in relation to blood clotting quality.14–16 Thromboelastometry (ROTEM®; Pentapharm GmbH, Munich, Germany) assesses the viscoelasticity of whole blood. One of the ROTEM® tests, the FIBTEM test, provides prompt information on the clot strength specifically attributed to fibrin/fibrinogen using cytochalasin-D-induced inactivation of platelets in vitro.17 This test may be used to guide the administration of fibrinogen concentrate for prompt haemostatic therapy.9–11 We hypothesized that postoperative haemostasis could be improved by increasing plasma fibrinogen concentrations, since bleeding complications were observed to be lower in patients with high perioperative fibrinogen concentrations.1213 The primary aim of this pilot study was to evaluate whether FIBTEM-guided intraoperative fibrinogen repletion was able to reduce the use of allogeneic blood products and postoperative bleeding in patients undergoing AV–AA.
ions were observed to be lower in patients with high perioperative fibrinogen concentrations.1213 The primary aim of this pilot study was to evaluate whether FIBTEM-guided intraoperative fibrinogen repletion was able to reduce the use of allogeneic blood products and postoperative bleeding in patients undergoing AV–AA. Methods The study protocol was approved by the institutional review board of the Hannover Medical School, and informed written consent was obtained from patients enrolled in the prospective part of the study. The inclusion criterion was elective AV–AA throughout the study. Exclusion criteria for both the retrospective and the prospective parts of the study were: any known congenital or acquired bleeding disorders, severe liver disease or heparin-induced anticoagulation effects, despite protamine therapy, age under 18 yr, pregnancy or nursing, redo surgery, emergency operation, and positive anamnesis for intake of platelet aggregation inhibitors within 5 days of surgery.
were: any known congenital or acquired bleeding disorders, severe liver disease or heparin-induced anticoagulation effects, despite protamine therapy, age under 18 yr, pregnancy or nursing, redo surgery, emergency operation, and positive anamnesis for intake of platelet aggregation inhibitors within 5 days of surgery. Transfusion algorithm Retrospective data from all 42 patients undergoing elective AV–AA in 2006 selected according to the inclusion and exclusion criteria were obtained from medical records (Group A). Patients had been transfused without a standardized transfusion protocol or point-of-care laboratory testing, and had received on average 4 u of FFP and 2 u of platelet concentrate during and the first 24 h after the operation. On the basis of these data and on algorithms described in the literature,1415 a two-step blood products transfusion algorithm for patients undergoing AV–AA was developed (Fig. 1). Fig 1 Flow chart of transfusion algorithm. FFP, fresh-frozen plasma; PC, platelet concentrates; PLT, platelet count (×103 µl−1); u, unit. In addition, we developed a method for quantifying mediastinal bleeding after the completion of heparin neutralization and surgical haemostasis after weaning from CPB. The method was also applied after each therapy step. The surgical field was thoroughly covered with sterile, dry surgical swabs of known weight after all blood had been removed using a suction device. Surgical swabs were carefully removed after 5 min and blood loss was determined by weighing the swabs and measuring the weight increase.18
applied after each therapy step. The surgical field was thoroughly covered with sterile, dry surgical swabs of known weight after all blood had been removed using a suction device. Surgical swabs were carefully removed after 5 min and blood loss was determined by weighing the swabs and measuring the weight increase.18 On the basis of preliminary measurements, the cut-off value chosen for clinically relevant diffuse, microvascular bleeding was 60 g. If between 60 and 250 g had been absorbed (i.e. 720–3000 g of blood h−1), two-step transfusion therapy was initiated (Fig. 1). Patients with blood loss >250 g were surgically re-explored and blood loss re-evaluated in the same way. The first step of haemostatic therapy was administered based on the platelet count. If the platelet count measured at the removal of the aortic clamp was >100×103 µl−1, patients were initially transfused with 4 u of FFP; if it was ≤100×103 µl−1, therapy was initiated with a transfusion of 2 u of platelet concentrate. Each transfusion was to be completed in 15 min and followed by blood loss measurement. If blood loss was not reduced to <60 g, patients who received FFP in the first therapy step then received 2 u of platelet concentrate, and those who initially received platelet concentrate were given 4 u of FFP (Fig. 1). If diffuse, microvascular bleeding persisted, patients received a further 2 u of FFP and 1 u of platelet concentrate consecutively. After successful haemostatic therapy, defined as a 5 min blood loss <60 g in subsequent assessments, the thorax was closed.
ly received platelet concentrate were given 4 u of FFP (Fig. 1). If diffuse, microvascular bleeding persisted, patients received a further 2 u of FFP and 1 u of platelet concentrate consecutively. After successful haemostatic therapy, defined as a 5 min blood loss <60 g in subsequent assessments, the thorax was closed. Prospective treatment groups Fifteen patients undergoing AV–AA were prospectively enrolled into Groups B and C. Five consecutive patients in Group B received transfusion according to the pre-defined blood products transfusion algorithm. Patients in Group C received fibrinogen concentrate before being transfused according to the transfusion algorithm. Fibrinogen concentrate dose was determined based on maximum clot firmness (MCF) in the FIBTEM test performed at the removal of the aortic clamp. With the goal of increasing FIBTEM MCF in Group C to ∼22 mm, the following formula was established: Therefore, the dose of fibrinogen concentrate equalled (22–FIBTEM MCF)×body weight/140. According to this formula, a patient of 70 kg requires a fibrinogen concentrate dose of ∼0.5 g to elevate FIBTEM MCF by ∼1 mm. The dose was rounded to a whole number of grams; the maximum fibrinogen concentrate dose was arbitrarily set at 6 g. After fibrinogen administration, patients in Group C received transfusion according to the algorithm applied to Group B, if bleeding persisted (Fig. 1).
Therefore, the dose of fibrinogen concentrate equalled (22–FIBTEM MCF)×body weight/140. According to this formula, a patient of 70 kg requires a fibrinogen concentrate dose of ∼0.5 g to elevate FIBTEM MCF by ∼1 mm. The dose was rounded to a whole number of grams; the maximum fibrinogen concentrate dose was arbitrarily set at 6 g. After fibrinogen administration, patients in Group C received transfusion according to the algorithm applied to Group B, if bleeding persisted (Fig. 1). For both Groups B and C, the transfusion of red blood cells (RBC) was administered to maintain haematocrit values between 23% and 25% on CPB, reaching 28% after CPB when the blood from the extracorporeal circulation system was re-infused into the patient. The primary endpoint of the study was transfusion of allogeneic blood products after CPB in the 24 h postoperative period; the secondary endpoint was the 24 h postoperative blood loss. Postoperative complications were documented until the patient was discharged. Intraoperative management All patients underwent general anaesthesia induced with etomidate, fentanyl, and cisatracurium. For maintenance of anaesthesia, sevoflurane was titrated to an end-tidal concentration of 1–2% until aortic cross-clamping on CPB. For the duration of CPB, propofol was infused and additional boluses of fentanyl were given every 30 min during the operation. All patients received 500 ml of Ringer's lactate solution and 500 ml of gelatine polysuccinate (Gelafundin® 0.026, Serumwerk, Bernburg, Germany) at the start of anaesthesia.
ss-clamping on CPB. For the duration of CPB, propofol was infused and additional boluses of fentanyl were given every 30 min during the operation. All patients received 500 ml of Ringer's lactate solution and 500 ml of gelatine polysuccinate (Gelafundin® 0.026, Serumwerk, Bernburg, Germany) at the start of anaesthesia. After aortic cannulation and administration of heparin 400 IU kg−1 (Heparin-Natrium-25000-ratiopharm®, Merckle GmbH, Blaubeuren, Germany), an extracorporeal circulation system was established, and the ascending aorta was replaced with an artificial graft (Hemashield Gold™ or Platin™, Woven Double Velour Vascular Graft, Boston Scientific International SA, Boston, MA, USA). Moderate hypothermia of 32°C was used routinely in all patients. Before CPB, 1 million kallikrein-inhibiting units (KIU) of aprotinin were administered, with an additional 1 million KIU in the CPB priming solution. After the initial anticoagulation, additional doses of heparin were given to maintain activated clotting time over 480 s. The system was primed with 1000 ml of Ringer's lactate solution, 500 ml of sodium chloride, and 40 ml of 8.4% sodium bicarbonate. After aorta replacement, patients were re-warmed to a bladder temperature of 36.5°C and weaned from CPB. Heparin was neutralized with protamine sulphate (Protamin Valeant, Valeant Pharmaceuticals GmbH, Eschborn, Germany) immediately after CPB. After surgery, patients were transferred to the intensive care unit (ICU).
. After aorta replacement, patients were re-warmed to a bladder temperature of 36.5°C and weaned from CPB. Heparin was neutralized with protamine sulphate (Protamin Valeant, Valeant Pharmaceuticals GmbH, Eschborn, Germany) immediately after CPB. After surgery, patients were transferred to the intensive care unit (ICU). Haematological evaluations Blood samples were drawn serially from the radial artery catheter (20 gauge) into commercially available pre-filled collection vials (Sarstedt, Nuembrecht, Germany), which contained heparin, citrate, or ethylenediamine tetracetic acid as the anticoagulant. Blood was sampled before operation (before induction of anaesthesia), at removal of the aortic clamp, at the end of CPB, after intraoperative haemostatic therapy, and 24 h after the operation. Activated partial thromboplastin time (aPTT; Kaolin, Stago Diagnostica, Asnières, France), prothrombin time (PT; Neoplastine®, Stago Diagnostica), and fibrinogen concentration (Clauss method: optical read-out) were determined using the STA-R® Analyzer (Stago Diagnostica & Roche, Germany). Platelet count and haematocrit were measured using the Sysmex XE-2100 (Roche Diagnostics, Mannheim, Germany). Platelet counts were available within 10–15 min of testing.
stica), and fibrinogen concentration (Clauss method: optical read-out) were determined using the STA-R® Analyzer (Stago Diagnostica & Roche, Germany). Platelet count and haematocrit were measured using the Sysmex XE-2100 (Roche Diagnostics, Mannheim, Germany). Platelet counts were available within 10–15 min of testing. A four-channel ROTEM® device (Pentapharm, Munich, Germany) was used to perform thromboelastometric analyses of whole blood samples as described previously.9101617 ROTEM® analyses were performed using 300 µl of whole blood and 20 µl of 0.2 M calcium chloride together with specific activators. In the EXTEM test, the activator used was rabbit brain thromboplastin. In the FIBTEM test, cytochalasin-D was added to rabbit brain thromboplastin in order to inhibit the contribution of platelets to the formation of the fibrin clot. The following parameters were recorded for the ROTEM® tests: clotting time [CT (s); time from the start of the test until a clot firmness of 2 mm is detected] and MCF (mm).
est, cytochalasin-D was added to rabbit brain thromboplastin in order to inhibit the contribution of platelets to the formation of the fibrin clot. The following parameters were recorded for the ROTEM® tests: clotting time [CT (s); time from the start of the test until a clot firmness of 2 mm is detected] and MCF (mm). Blood count, coagulation factors, and thromboelastometry were measured at the start of the procedure, after CPB, after coagulation therapy, and 24 h after operation. Only platelet count and thromboelastometry at the time point of unclamping the aorta were relevant for the planning of the coagulation therapy, so no other measures were recorded at this stage. ROTEM® results were concealed from the attending anaesthetists, surgeons, and intensive care physicians. The measurements were performed by a member of the anaesthesiology department not involved in patient operation management.
the planning of the coagulation therapy, so no other measures were recorded at this stage. ROTEM® results were concealed from the attending anaesthetists, surgeons, and intensive care physicians. The measurements were performed by a member of the anaesthesiology department not involved in patient operation management. Statistical analysis The differences between the groups were analysed with regard to patient characteristics, intraoperative and 24 h postoperative transfusion of allogeneic blood products, and 24 h postoperative blood loss. The primary endpoint, the use of allogeneic blood products, was compared between the retrospective Group A and the prospective fibrinogen therapy Group C. The secondary endpoint, the 24 h postoperative blood loss, was compared between the same two groups. Group B was used to assess whether significant differences in transfusion parameters would be induced by standardizing the transfusion practice compared with the retrospective Group A. Because detailed coagulation analyses were not available in Group A (historical control) during CPB and at the end of CPB, data obtained from the prospective conventional therapy (Group B) were compared with those from Group C. On the basis of our previous experience with ROTEM® assays, a minimal sample size of 5 was needed to detect a 30% change in MCF values with a β-value of >0.8 and an α-value of <0.05. Data are presented as mean (sd). Continuous variables were analysed with a Mann–Whitney U-test; categorical variables were analysed using the χ2 test. A P-value of ≤0.05 was considered to be statistically significant.
e of 5 was needed to detect a 30% change in MCF values with a β-value of >0.8 and an α-value of <0.05. Data are presented as mean (sd). Continuous variables were analysed with a Mann–Whitney U-test; categorical variables were analysed using the χ2 test. A P-value of ≤0.05 was considered to be statistically significant. Results All patient groups were similar with regard to preoperative characteristics (Table 1). For the prospective groups (B and C), the main parameters guiding the initial treatment step were comparable, namely platelet count at removal of the aortic cross-clamping (mean 135×103 and 137×103 µl−1, respectively) and the 5 min blood loss assessment after weaning from CPB [mean 137 (54) and 133 (55) g, respectively]. Table 1 Characteristics in patients undergoing AV–AA. Data presented as mean (range), mean (sd) or absolute
Results All patient groups were similar with regard to preoperative characteristics (Table 1). For the prospective groups (B and C), the main parameters guiding the initial treatment step were comparable, namely platelet count at removal of the aortic cross-clamping (mean 135×103 and 137×103 µl−1, respectively) and the 5 min blood loss assessment after weaning from CPB [mean 137 (54) and 133 (55) g, respectively]. Table 1 Characteristics in patients undergoing AV–AA. Data presented as mean (range), mean (sd) or absolute Group A (retrospective) (n=42) Group B (n=5) Group C (+fibrinogen) (n=10) Age (yr) 57 (33–89) 61 (47–76) 57 (25–76) Weight (kg) 84 (14) 94 (8) 90 (20) Body mass index (kg m−2) 27 (5) 27 (2) 29 (5) Female (n) 14 0 2 Coronary heart disease (n) 6 1 1 Peripheral vascular disease (n) 3 0 2 Cerebrovascular disease (n) 2 0 0 Five patients were recruited to the prospective, blood products therapy group (B). Mean 5 min blood loss was 84 (12) g after the first therapy step. As a result of a defective FFP bag, one patient only received 3 u of FFP. One patient had blood loss below 60 g, and therefore no additional haemostatic therapy was required. Intraoperative bleeding was successfully managed after the second therapy step in all patients; the mean 5 min blood loss decreased to 49 (6) g. The retrospective and the prospective group treated with allogeneic blood products, Groups A and B, were comparable regarding bleeding and transfusion parameters, that is, these were not influenced by the introduction of the standardized transfusion algorithm. The patients in these groups were also comparable with regard to postoperative parameters in the ICU. Regarding standard laboratory data, preoperative levels of haematocrit appeared lower in Group B than in Group A, but the difference was not significant (Table 2). Other laboratory parameters, including platelet count, were comparable at all times.
ere also comparable with regard to postoperative parameters in the ICU. Regarding standard laboratory data, preoperative levels of haematocrit appeared lower in Group B than in Group A, but the difference was not significant (Table 2). Other laboratory parameters, including platelet count, were comparable at all times. Table 2 Laboratory parameters in patients undergoing AV–AA. Data presented as mean (sd). aPTT, activated partial thromboplastin time; CPB, cardiopulmonary bypass; PT, prothrombin time. *P<0.05 Group C vs Group A; †P<0.05 Group C vs Group B (P<0.05); there were no statistically significant differences between Groups A and B
e 2 Laboratory parameters in patients undergoing AV–AA. Data presented as mean (sd). aPTT, activated partial thromboplastin time; CPB, cardiopulmonary bypass; PT, prothrombin time. *P<0.05 Group C vs Group A; †P<0.05 Group C vs Group B (P<0.05); there were no statistically significant differences between Groups A and B Laboratory data Group A (n=42) Group B (n=5) Group C (+fibrinogen) (n=10) Preoperative laboratory parameters Haematocrit (%) (normal range: 41.5–50.4) 40 (5) 35 (6) 39 (5) PT (s) (normal range: 11–13.5) 15 (3) 14 (1) 14 (1) aPTT (s) (normal range: 26–35) 32 (8) 29 (2) 29 (3) Platelet count (103 µl−1) (normal range: 150–450) 202 (63) 204 (46) 196 (33) Fibrinogen (g litre−1) (normal range: 2.0–4.5) 3.4 (0.6) 3.2 (1.0) 3.3 (1.0) Removal of aortic clamp Platelet count (103 µl−1) 135 (43) 137 (38) End of CPB Haematocrit (%) 29 (2) 29 (4) PT (s) 19 (1) 20 (2) aPTT (s) 31 (3) 32 (2) Platelet count (103 µl−1) 103 (26) 104 (23) Fibrinogen (g litre−1) 2.1 (0.6) 2.2 (0.6) After coagulation therapy Haematocrit (%) 28 (3) 25 (3) 28 (4)† PT (s) 17 (1) 18 (1) 18 (1) aPTT (s) 33 (5) 32 (3) 32 (2) Platelet count (103 µl−1) 128 (40) 142 (24) 115 (31) Fibrinogen (g litre−1) 2.2 (0.4) 2.1 (0.3) 3.6 (0.6)*,† First postoperative day Haematocrit (%) 31 (3) 31 (4) 31 (5) PT (s) 17 (2) 16 (1) 16 (1) aPTT (s) 40 (16) 36 (5) 36 (7) Platelet count (103 µl−1) 135 (31) 129 (41) 132 (33) Fibrinogen (g litre−1) 4.4 (0.6) 4.3 (0.9) 4.4 (0.7) The amount of RBC concentrate used on CPB was comparable between the groups (Table 3). The 10 patients recruited to the prospective Group C received a mean dose of 5.7 (0.7) g fibrinogen concentrate. This effectively reduced the 5 min blood loss from 133 (55) g after weaning from CPB to 32 (18) g (<60 g in all patients). Therefore, according to the transfusion algorithm, no additional intraoperative administration of FFP or platelet concentrates was necessary after the end of CPB in this group. After the operation and during the first 24 h in the ICU, only two patients treated with fibrinogen concentrate required transfusion [mean 0.7 (range 0–4) u in Group C vs 8.5 (5.3) and 8.2 (2.3) u in Groups A and B, respectively] (Table 3). Group C had lower 24 h drainage than the retrospective Group A (Table 3). Intubation time and ICU stay duration were shorter and complication rates lower in Group C (Table 3).
gen concentrate required transfusion [mean 0.7 (range 0–4) u in Group C vs 8.5 (5.3) and 8.2 (2.3) u in Groups A and B, respectively] (Table 3). Group C had lower 24 h drainage than the retrospective Group A (Table 3). Intubation time and ICU stay duration were shorter and complication rates lower in Group C (Table 3). Table 3 Intra- and postoperative parameters in patients undergoing AV–AA. Data presented as mean (sd) or absolute. CPB, cardiopulmonary bypass; ICU, intensive care unit; prolonged ventilatory support, ventilatory support >40 h. *P<0.05 Group C vs Group A; †P<0.05 Group C vs Group B
gen concentrate required transfusion [mean 0.7 (range 0–4) u in Group C vs 8.5 (5.3) and 8.2 (2.3) u in Groups A and B, respectively] (Table 3). Group C had lower 24 h drainage than the retrospective Group A (Table 3). Intubation time and ICU stay duration were shorter and complication rates lower in Group C (Table 3). Table 3 Intra- and postoperative parameters in patients undergoing AV–AA. Data presented as mean (sd) or absolute. CPB, cardiopulmonary bypass; ICU, intensive care unit; prolonged ventilatory support, ventilatory support >40 h. *P<0.05 Group C vs Group A; †P<0.05 Group C vs Group B Parameters Group A (n=42) Group B (n=5) Group C (+fibrinogen) (n=10) Intraoperative Aortic clamp time (min) 72 (21) 72 (19) 68 (31) CPB time (min) 107 (25) 108 (29) 100 (40) Lowest temperature on CPB (°C) 31.3 (2.2) 31.6 (2.5) 32.8 (2.5) Red blood cells on CPB (u) 1.1 (1.8) 0.8 (1.3) 0.7 (1.1) 5 min blood loss (ml) after weaning from CPB N/A 137 (54) 133 (55) fibrinogen concentrate N/A N/A 32 (18) first therapy step N/A 84 (12) N/A second therapy step N/A 49 (6) N/A Postoperative Patients who did not receive any allogeneic blood after CPB and on first day ICU (n) 1 0 8*,† ICU time to extubation (h) 13 (12) 12 (5) 9 (5) ICU time (h) 36 (26) 31 (21) 20 (5)*,† Re-exploration for bleeding (n) 2 1 0 Postoperative atrial fibrillation (n) 6 1 1 Prolonged ventilatory support (n) 1 0 0 Major neurological events (n) 0 0 0 30 day mortality (n) 0 0 0 Postoperative hospitalization (days) 10 (3) 12 (12) 10 (2) Units transfused/volume drained after CPB and during the first 24 h in ICU Red blood cells (u) 2.4 (2.5) 2.4 (1.1) 0.5 (1.1)*,† Fresh-frozen plasma (u) 4.5 (2.1) 4.2 (1.1) 0.2 (0.6)*,† Platelet concentrate (u) 1.6 (1.7) 1.6 (0.9) 0.0 (0.0)*,† Total blood cell concentrates (u) 8.5 (5.3) 8.2 (2.3) 0.7 (1.5)*,† Drainage volume (ml) 793 (560) 716 (219) 366 (199)*,† Standard laboratory analyses showed that preoperative values were comparable across the groups (Table 1). For all groups, the coagulation parameters were similarly affected during CPB, reflected by the analysis performed upon removal of the aortic clamp (Tables 2 and 4). Analysis performed after weaning from bypass showed no significant differences between the groups. After the final therapy step, laboratory values showed a higher plasma concentration of fibrinogen in Group C than in Groups A and B. However, after 24 h, the fibrinogen plasma concentration was uniformly high in all the groups.
s performed after weaning from bypass showed no significant differences between the groups. After the final therapy step, laboratory values showed a higher plasma concentration of fibrinogen in Group C than in Groups A and B. However, after 24 h, the fibrinogen plasma concentration was uniformly high in all the groups. Table 4 ROTEM® values in patients undergoing AV–AA. Data presented as mean (sd). CPB, cardiopulmonary bypass; EXTEM®, ROTEM® test with extrinsic activation of coagulation; FibTEM®, ROTEM® test with extrinsic activation of coagulation and platelet inhibition with cytochalasin D; CT, clotting time; MCF, maximal clot firmness. *P<0.05 Group C vs Group B ROTEM® values Group B (n=5) Group C (+fibrinogen) (n=10) Preoperative EXTEM® CT (s) (normal range: 35–80) 71 (7) 69 (7) EXTEM® MCF (mm) (normal range: 53–72) 64 (5) 62 (5) FibTEM® MCF (mm) (normal range: 9–25) 15 (3) 14 (4) Removal of the aortic clamp EXTEM® CT (s) 104 (32) 132 (73) EXTEM® MCF (mm) 58 (6) 56 (6) FibTEM® MCF (mm) 12 (2) 11 (3) End of CPB EXTEM® CT (s) 87 (10) 92 (14) EXTEM® MCF (mm) 57 (4) 55 (5) FibTEM® MCF (mm) 12 (2) 11 (3) After coagulation therapy EXTEM® CT (s) 76 (4) 69 (10) FibTEM® MCF (mm) 12 (2) 20 (3)* First postoperative day EXTEM® CT (s) 7 (9) 73 (11) EXTEM® MCF (mm) 64 (4) 63 (5) FibTEM® MCF (mm) 22 (6) 21 (3) Preoperative ROTEM® data were comparable between Groups B and C (Table 4). After therapy, Groups B and C had similar EXTEM values, but Group C had higher FIBTEM values (Table 4). However, after 24 h, FIBTEM and EXTEM values were comparable and within the normal range in both groups.
3 (5) FibTEM® MCF (mm) 22 (6) 21 (3) Preoperative ROTEM® data were comparable between Groups B and C (Table 4). After therapy, Groups B and C had similar EXTEM values, but Group C had higher FIBTEM values (Table 4). However, after 24 h, FIBTEM and EXTEM values were comparable and within the normal range in both groups. Discussion In this pilot study, haemostatic therapy with fibrinogen concentrate targeting a high plasma fibrinogen level in AV–AA patients resulted in a reduction in transfusion of allogeneic blood products and drainage volume compared with a historical control group that received conventional haemostatic therapy. Before haemostatic therapy, at the end of CPB, coagulation disturbances seen in laboratory tests were comparable between the conventional therapy groups (A and B) and prospective fibrinogen therapy group (Group C). Both groups had prolonged PT, decreased platelet counts, and decreased fibrinogen levels compared with baseline. The administered fibrinogen concentrate [5.7 (0.7) g, representing ∼285 ml] restored fibrinogen plasma to baseline levels without affecting PT and platelet count. Despite the lower transfusion of RBC, FFP, and platelet concentrate in Group C, the laboratory data on haematocrit, platelet count, PT/aPTT, and fibrinogen were comparable among the three groups at 24 h after surgery.
enting ∼285 ml] restored fibrinogen plasma to baseline levels without affecting PT and platelet count. Despite the lower transfusion of RBC, FFP, and platelet concentrate in Group C, the laboratory data on haematocrit, platelet count, PT/aPTT, and fibrinogen were comparable among the three groups at 24 h after surgery. There are currently ongoing discussions in the literature concerning the critical level of plasma fibrinogen in relation to perioperative bleeding.12131920 There are experimental and clinical data describing a protective effect of high plasma fibrinogen levels. A study by Velik-Salchner and colleagues21 reported that the use of fibrinogen concentrate (250 mg kg−1) was more effective than platelet concentrate in a porcine hepatic laceration model in the presence of thrombocytopenia (platelet count <30×103 mm−3). There are ∼40 000–80 000 glycoprotein IIb/IIIa receptors on a single, activated platelet, and the number of these receptors is relatively constant after CPB.22 Thrombin generation is decreased after CPB,23 but one molecule of thrombin can cleave up to 1680 molecules of fibrinogen.24 Assuming a simple enzyme–substrate reaction (Michaelis–Menten equation) between thrombin and fibrinogen, the Michaelis constant (Km) value of fibrinogen at 2 g litre−1 (6 µM) represents half of the maximal reaction rate between thrombin and fibrinogen.25 According to the Michaelis–Menten equation, the targeted fibrinogen level of 3.6 g litre−1 (or 10.7 µM) would nearly maximize the interaction between fibrinogen and the amount of thrombin available after CPB, resulting in improved haemostasis. The threshold level of fibrinogen was also evaluated in obstetric patients who developed severe post-partum haemorrhage, in which a fibrinogen level ≤2 g litre−1 had 100% positive predictive value for bleeding and a level >4 g litre−1 had 79% negative predictive value for bleeding.19 Other clinical studies have shown that a low fibrinogen concentration better predicts increased bleeding after prolonged CPB.1213 In another clinical setting, Heindl and colleagues11 previously described the use of fibrinogen (7–8 g) in patients with major traumatic bleeding refractory to standard coagulation therapy.
Other clinical studies have shown that a low fibrinogen concentration better predicts increased bleeding after prolonged CPB.1213 In another clinical setting, Heindl and colleagues11 previously described the use of fibrinogen (7–8 g) in patients with major traumatic bleeding refractory to standard coagulation therapy. Since this first report, fibrinogen concentrate has been shown to improve haemostasis in acquired hypofibrinogenaemia associated with cardiac surgery, liver transplantation, trauma, placental abruption, dilutional coagulopathy during complex orthopaedic procedures, and in disseminated intravascular coagulation as a result of massive blood loss and transfusion.7–10 Using fibrinogen concentrate as a first-line therapy to correct postoperative bleeding and to reduce the use of FFP, platelet concentrate, or both seems to be a reasonable approach as these allogeneic blood products are associated with various adverse outcomes.2627 In addition, as the average concentrate of fibrinogen in FFP is 2.5 g litre−1, FFP cannot be used for haemostatic therapy targeting a plasma fibrinogen level higher than this.
elet concentrate, or both seems to be a reasonable approach as these allogeneic blood products are associated with various adverse outcomes.2627 In addition, as the average concentrate of fibrinogen in FFP is 2.5 g litre−1, FFP cannot be used for haemostatic therapy targeting a plasma fibrinogen level higher than this. Fibrinogen concentrate administration was guided by FIBTEM (clot strength in the presence of platelet inhibition) using the empirical target MCF of 22 mm with an arbitrary limit for the maximal dose set at 6 g fibrinogen concentrate. This strategy resulted in a mean plasma fibrinogen concentration increase (from 2.2 to 3.6 g litre−1) and an increase in mean FIBTEM MCF from 11 to 20 mm. The decision whether to administer fibrinogen concentrate had to be made within 10 min after CPB when diffuse bleeding was diagnosed. The FIBTEM test is rapid and requires no centrifugation of the sample, a time-consuming step otherwise necessary in the standard laboratory-based assessment of fibrinogen concentration. In contrast to the optical read-out of the Clauss method, which measures the time to change in turbidity caused by fibrin formation and estimates fibrinogen plasma concentration from a calibration curve, the FIBTEM provides information on the mechanical strength of the clot. For all these reasons, we considered the FIBTEM (ROTEM®) to be the optimal bedside assay to guide the dosage of fibrinogen concentrate in this setting.
used by fibrin formation and estimates fibrinogen plasma concentration from a calibration curve, the FIBTEM provides information on the mechanical strength of the clot. For all these reasons, we considered the FIBTEM (ROTEM®) to be the optimal bedside assay to guide the dosage of fibrinogen concentrate in this setting. Because fibrinogen is an acute-phase protein, its level increases gradually after surgical procedures.13 Even though Group C received a mean of 5.7 g of fibrinogen as haemostatic therapy after weaning from CPB, similar plasma fibrinogen levels in the three groups and similar ROTEM® MCF values in Groups B and C were noted on postoperative day 1 (Tables 2 and 4). This finding may be relevant to the assessment of the safety of administration of fibrinogen concentrate in this setting. In addition, no immediate neurological and cardiorespiratory complications were observed in either group (Table 3).
® MCF values in Groups B and C were noted on postoperative day 1 (Tables 2 and 4). This finding may be relevant to the assessment of the safety of administration of fibrinogen concentrate in this setting. In addition, no immediate neurological and cardiorespiratory complications were observed in either group (Table 3). Other therapeutic options for diffuse bleeding after weaning from CPB may be considered. In countries where fibrinogen concentrate is not available, cryoprecipitate may be used, as it contains a higher concentration of fibrinogen than FFP.28 Unlike the fibrinogen concentrate we used (Haemocomplettan® P/Riastap, which is pasteurized for 20 h), viral inactivation is not generally applied to cryoprecipitate. Therapy with cryoprecipitate therefore carries a risk of viral transmission equivalent to that of FFP administration.29 Recombinant activated factor VII (rFVIIa) has been increasingly considered an ‘off-label’ rescue haemostatic agent in cardiac surgery.30 Although thrombin generation is decreased after CPB,23 the balance between thrombin inhibitors such as antithrombin (also decreased after prolonged CPB)23 and thrombin could be disturbed by adding a drug that generates a ‘thrombin burst’. A review of the US Food and Drug Administration's Adverse Event Reporting System found a total of 431 adverse event reports for rFVIIa from 1999 to 2004, including 185 thromboembolic events, 90% of which related to off-label use in patients without haemophilia.31
rbed by adding a drug that generates a ‘thrombin burst’. A review of the US Food and Drug Administration's Adverse Event Reporting System found a total of 431 adverse event reports for rFVIIa from 1999 to 2004, including 185 thromboembolic events, 90% of which related to off-label use in patients without haemophilia.31 This preliminary study has limitations. First, it was not randomized or blinded and was underpowered to confirm the efficacy and safety of fibrinogen replenishment in complex cardiac surgery. The present data were obtained in a specific population (AV–AA surgery); therefore, our findings may not be appropriate for the management of every type of post-cardiac surgical bleeding diathesis. Secondly, the major endpoints in this study only included the amount of allogeneic blood product use and the 24 h postoperative blood loss. A longer follow-up of large numbers of patients would be necessary to confirm the efficacy of fibrinogen concentrate for haemostatic therapy and to assess safety parameters such as neurological, cardiorespiratory, and infectious complications.
mount of allogeneic blood product use and the 24 h postoperative blood loss. A longer follow-up of large numbers of patients would be necessary to confirm the efficacy of fibrinogen concentrate for haemostatic therapy and to assess safety parameters such as neurological, cardiorespiratory, and infectious complications. In summary, the present data indicate that fibrinogen concentrate may be effective in reducing both the use of allogeneic blood products and postoperative bleeding in aortic surgical patients. To our knowledge, this was the first time that patients with fibrinogen levels within the normal range (mean 2.2 g litre−1) were substituted with fibrinogen concentrate to achieve an upper normal range (mean 3.6 g litre−1). Compared with the allogeneic blood products, such as FFP, cryoprecipitate, and platelet concentrate, fibrinogen concentrate can be potentially time-saving by precluding the need for cross-matching, thawing, or both. The FIBTEM MCF was an appropriate parameter for dosing fibrinogen concentrate in this setting. A validation study with a prospective, randomized, placebo-controlled design is currently underway. Funding The study was supported by CSL Behring, Marburg, Germany. Acknowledgement The authors are indebted to Gerald Hochleitner (CSL Behring) for excellent technical advice.
Uncontrolled bleeding associated with major trauma and surgery is often life-threatening. Acquired coagulopathy of trauma is responsible for the majority of postoperative traumatic haemorrhagic fatalities, and the onset of acute coagulopathy is associated with increased overall mortality.1 The coagulopathy of trauma is a manifestation of the combined effects of blood loss and dilution, coagulation factor and platelet consumption, hypothermic platelet dysfunction, acidosis-induced decreases in coagulation factor activity, and fibrinolysis.2
ute coagulopathy is associated with increased overall mortality.1 The coagulopathy of trauma is a manifestation of the combined effects of blood loss and dilution, coagulation factor and platelet consumption, hypothermic platelet dysfunction, acidosis-induced decreases in coagulation factor activity, and fibrinolysis.2 A recognized complication of massive transfusion is dilutional coagulopathy, which occurs when lost blood is replaced with fluids that do not contain coagulation factors. Dilutional coagulopathy is distinct from the more recently recognized phenomenon of acute traumatic coagulopathy, which is independent of i.v. fluid administration and appears to be mediated through activation of the protein C pathway.3 The shift from whole blood transfusion in the past to the current practice of specific blood component therapy has resulted in earlier occurrence of thrombocytopenia and clotting factor deficiency in trauma patients.45 Trauma resuscitation usually starts with crystalloid or colloid solutions followed by red blood cell concentrates, with resultant coagulation factor dilution.2 A vicious cycle can ensue, in which further haemorrhage necessitates additional resuscitation fluid with resultant exacerbation of haemodilution, coagulopathy, and blood loss. A recent German analysis demonstrated a high frequency of established coagulopathy in multiple injury patients upon emergency room admission that was associated with the amount of prior i.v. fluids administered.6 Consequently, a treatment strategy of earlier coagulation factor replacement has been advocated.478
d loss. A recent German analysis demonstrated a high frequency of established coagulopathy in multiple injury patients upon emergency room admission that was associated with the amount of prior i.v. fluids administered.6 Consequently, a treatment strategy of earlier coagulation factor replacement has been advocated.478 Fresh frozen plasma (FFP) is a source of all the coagulation factors, including the labile factors, and its use has been recommended either in massive bleeding or significant bleeding complicated by coagulopathy, as indicated by >1.5-fold prolongation of prothrombin time (PT) and activated partial thromboplastin time (aPTT).9 However, FFP has certain drawbacks for rapid reversal of coagulopathy. FFP may require ABO compatibility, and is typically available only after the blood type has been established and plasma has been thawed. These preliminary procedures may take more than an hour in some facilities and may jeopardize the opportunity to prevent rapid deterioration in the clinical condition of the patient.2 Although the overall risks of FFP are low, complications can include immunological reactions such as allergy/anaphylaxis, transfusion-related acute lung injury (TRALI), and haemolysis due to anti-A or anti-B if FFP is transfused across ABO groups.1011 TRALI, an acute syndrome of dyspnoea, hypoxia, and pulmonary ‘white-out’, is a major cause of transfusion-related death.10
ns can include immunological reactions such as allergy/anaphylaxis, transfusion-related acute lung injury (TRALI), and haemolysis due to anti-A or anti-B if FFP is transfused across ABO groups.1011 TRALI, an acute syndrome of dyspnoea, hypoxia, and pulmonary ‘white-out’, is a major cause of transfusion-related death.10 Delivery of coagulation factors in concentrated form may offer advantages in overcoming the coagulopathic effects of large resuscitation fluid volumes.2 Prothrombin complex concentrate (PCC), which contains vitamin K-dependent coagulation factors, can be administered rapidly without the need for matching the blood group or thawing the product. PCC has increasingly been employed for rapid reversal of coumarin oral anticoagulant therapy.12–21 The most recent update from the British Committee for Standards in Haematology recommends that, for reversal of anticoagulation in patients with major bleeding, PCC should be administered in preference to FFP.22 The use of PCC in trauma patients has not been reported.
versal of coumarin oral anticoagulant therapy.12–21 The most recent update from the British Committee for Standards in Haematology recommends that, for reversal of anticoagulation in patients with major bleeding, PCC should be administered in preference to FFP.22 The use of PCC in trauma patients has not been reported. In coagulopathic trauma patients, a major obstacle to optimal coagulation factor replacement therapy has been the lack of suitable assays that can reflect the overall haemostatic balance. The calibrated automated thrombin generation assay (TGA) is a global coagulation test being used to monitor both hypocoagulable and hypercoagulable states. Thrombin generation, which plays a central role in activating coagulation enzymes, inhibitors, and platelets and in cleaving fibrinogen to fibrin monomers, has been proposed as an overall function test of the haemostatic–thrombotic system.23 Thrombin generation has been investigated as a means to monitor treatment with bypassing agents such as factor VIII inhibitor bypass activity (FEIBA), which is an activated PCC, and recombinant factor VIIa in haemophilia patients with inhibitors.24–27 PCC-mediated oral anticoagulant reversal in vitro has recently been evaluated using TGA.28
s been investigated as a means to monitor treatment with bypassing agents such as factor VIII inhibitor bypass activity (FEIBA), which is an activated PCC, and recombinant factor VIIa in haemophilia patients with inhibitors.24–27 PCC-mediated oral anticoagulant reversal in vitro has recently been evaluated using TGA.28 An animal model of dilutional coagulopathy and major haemorrhage in pigs has been developed.29 In that study, PCC was effective in normalizing coagulation and improving haemostasis compared with an inactive placebo; however, an active control fluid containing coagulation factors was not evaluated. The model was used in the present study to compare the effects of PCC and FFP on haemorrhage after femur or spleen injury. In addition, the impact of those agents on coagulation function in vivo, including thrombin generation, was examined. Methods Animals As described previously in detail, dilutional coagulopathy was induced in anaesthetized mildly hypothermic (36°C) pigs.29 Forty-seven castrated male pigs (large white×German noble) were obtained from a local breeding farm (Schlosser, Schwalmtal, Germany). The animals were 3–4 months old and weighed 21–32 kg. The study was performed in accordance with the German Animal Welfare law and approved by the regional government authorities. Animal housing and care were furnished in compliance with current regulations of the European Union.
farm (Schlosser, Schwalmtal, Germany). The animals were 3–4 months old and weighed 21–32 kg. The study was performed in accordance with the German Animal Welfare law and approved by the regional government authorities. Animal housing and care were furnished in compliance with current regulations of the European Union. Anaesthesia After an overnight fast with free access to water, pigs were premedicated i.m. with a mixture of 2 mg kg−1 azaperone (Stresnil®, Janssen-Cilag GmbH, Neuss, Germany), 15 mg kg−1 ketamine (Ketavet, Pharmacia & Upjohn, Erlangen, Germany), and 0.02 mg kg−1 atropine sulfate (Atropinsulfate, B. Braun, Melsungen, Germany). Anaesthesia was induced by 10 mg kg−1 thiopental sodium via an ear vein. The animals were placed on a Heyer Access ventilator after tracheal intubation. Inhaled anaesthesia was maintained by isoflurane (Forane, Abbott Laboratories Inc., Abbott Park, IL, USA) at a concentration of 1–2%. A 20-gauge catheter was placed into a femoral artery for continuous arterial blood pressure measurements, and body temperature was monitored by a rectal thermometer. Attainment and maintenance of deep anaesthesia were confirmed by an absent pedal withdrawal reflex.
nc., Abbott Park, IL, USA) at a concentration of 1–2%. A 20-gauge catheter was placed into a femoral artery for continuous arterial blood pressure measurements, and body temperature was monitored by a rectal thermometer. Attainment and maintenance of deep anaesthesia were confirmed by an absent pedal withdrawal reflex. Haemodilution Vascular access was secured through a 14-gauge catheter in the external jugular vein. Phased blood withdrawal (total 60 ml kg−1, 65–70% of calculated total blood volume), salvaged erythrocyte retransfusion (total 20 ml kg−1), and volume substitution with a total of 40 ml kg−1 hydroxyethyl starch 200/0.5 (Infukoll 6%, Schwarz Pharma AG, Mannheim, Germany) were performed over a period of 80 min after induction of anaesthesia.29 Treatment Study treatments were administered 115 min after initiation of blood withdrawal (25 min after completion of phased haemodilution). The pigs were randomly assigned to receive 15 ml kg−1 isotonic saline, 25 IU kg−1 PCC (Beriplex® P/N, CSL Behring GmbH, Marburg, Germany), or standard-dose (15 ml kg−1) or high-dose (40 ml kg−1) porcine FFP via the indwelling jugular catheter. The 40 ml kg−1 FFP dose was selected to maximize efficacy without unacceptable haemodynamic disturbances. Beriplex P/N is a biochemically well-characterized balanced PCC containing coagulation factors II (FII), VII (FVII), IX (FIX), and X (FX), as well as the anticoagulant proteins C and S.30 The administered PCC dose of 25 IU kg−1, based on FIX content, is equivalent to 0.8 ml kg−1.
unacceptable haemodynamic disturbances. Beriplex P/N is a biochemically well-characterized balanced PCC containing coagulation factors II (FII), VII (FVII), IX (FIX), and X (FX), as well as the anticoagulant proteins C and S.30 The administered PCC dose of 25 IU kg−1, based on FIX content, is equivalent to 0.8 ml kg−1. The respective mean (sd) concentrations of FII, FVII, FIX, and FX in Beriplex P/N are 31.0 (3.4), 16.2 (1.9), 28.9 (2.2), and 40.5 (3.3) IU ml−1.31 Corresponding values in porcine plasma are 0.50 (0.04), 0.48 (0.06), 2.3 (0.42), and 0.58 (0.07) U ml−1.29 Therefore, the administered 0.8 ml kg−1 PCC volume contained doses of those coagulation factors, respectively, 3.3, 1.8, 0.67, and 3.7-fold those of the 15 ml kg−1 porcine PCC volume and 1.2, 0.68, 0.25, and 1.4-fold the 40 ml kg−1 volume. Porcine rather than human FFP was selected as control fluid, because in pilot experiments unpredictable and sometimes serious transfusion reactions were encountered in pigs receiving human FFP. For preparation of porcine FFP, blood was withdrawn from the carotid artery of healthy donor pigs through custom-made connection tubing. The anticoagulated blood was centrifuged 10 min at 3000 g, and the plasma was frozen and stored at −70°C until treatment. Prior to porcine FFP infusions, haemagglutination and haemolysis of individual animal erythrocyte specimens were tested to lessen the risk of transfusion reactions.
gh custom-made connection tubing. The anticoagulated blood was centrifuged 10 min at 3000 g, and the plasma was frozen and stored at −70°C until treatment. Prior to porcine FFP infusions, haemagglutination and haemolysis of individual animal erythrocyte specimens were tested to lessen the risk of transfusion reactions. Experimental trauma At 5 min following administration of study treatments, standardized wounds were created either by drilling a 3 mm hole into the femur or making a 7 cm long and 1 cm deep incision into the spleen. Time to haemostasis and blood loss were monitored for 120 min after the experimental injuries were inflicted. Groups of 5–7 pigs had received each of the test treatments before femur injury. Group size for spleen injury was six animals, with the exception that seven animals had received saline placebo prior to spleen incision.
haemostasis and blood loss were monitored for 120 min after the experimental injuries were inflicted. Groups of 5–7 pigs had received each of the test treatments before femur injury. Group size for spleen injury was six animals, with the exception that seven animals had received saline placebo prior to spleen incision. Skin bleeding time (SBT) was determined in duplicate from a 5 mm long and 1 mm deep incision at a shaved inner site on the ear using a standard cutting device (Surgicutt®, International Technidyne Corp., Edison, NJ, USA). Shed blood was blotted with a filter paper from the edge of the wound. The time from incision to cessation of blood flow was recorded as the SBT. Determinations of time to haemostasis, blood loss, and SBT were made by observers blinded to the group assignments of the animals. While remaining under deep anaesthesia, the animals were humanely killed upon conclusion of the study experimental procedures by injecting embutramide, mebenzonium iodide, and tetracaine hydrochloride (T61®, Intervet Deutschland GmbH, Unterschleißheim, Germany).
bservers blinded to the group assignments of the animals. While remaining under deep anaesthesia, the animals were humanely killed upon conclusion of the study experimental procedures by injecting embutramide, mebenzonium iodide, and tetracaine hydrochloride (T61®, Intervet Deutschland GmbH, Unterschleißheim, Germany). Laboratory assays Blood samples were collected from the carotid artery at baseline, after the completion of phased haemodilution (80 min) and 5 min after study treatment administration (120 min). Coagulation factors were measured in coagulation factor-deficient plasma (Dade Behring, Marburg, Germany) with a Schnitger and Gross coagulometer (Heinrich Amelung GmbH, Lemgo, Germany). FII, FVII, and FX were determined by PT assay, and FIX was measured by aPTT assay. PT was determined with a Schnitger & Gross coagulometer using the Thromborel reagent (Dade Behring). TGA was performed by calibrated automated thrombinography (CAT, Thrombinoscope B.V., Maastricht, the Netherlands) in diluted plasma according to the method of Hemker and colleagues.32 The concentrations of recombinant relipidated tissue factor (Dade Behring) and phospholipids were 5 pM and 4 µM, respectively. The peak molar quantity of thrombin present in clotting plasma was calculated using the Thrombinoscope software version 3.0.0.29.
uted plasma according to the method of Hemker and colleagues.32 The concentrations of recombinant relipidated tissue factor (Dade Behring) and phospholipids were 5 pM and 4 µM, respectively. The peak molar quantity of thrombin present in clotting plasma was calculated using the Thrombinoscope software version 3.0.0.29. Statistical analysis Median differences and their 95% confidence intervals (CI) were determined by exact Hodges–Lehmann estimation. Time to haemostasis was analysed by the Kaplan–Meier product-limit method and exact logrank test. Between-group differences in the volume of blood lost were evaluated by exact Wilcoxon test. Analyses were performed using R version 2.6.2 (The R Foundation for Statistical Computing, Vienna, Austria) and StatXact 7.0 (Cytel Software Corp., Cambridge, MA, USA) statistical software. Results Coagulation factors Upon completion of phased haemodilution (80 min), median rectal temperature fell to 36.2°C with an interquartile range (IQR) of 35.9–36.9°C from 37.8°C (IQR, 37.6–38.5°C) at baseline. Haemodilution profoundly diminished circulating concentrations of FII, FVII, FIX, and FX, with median values declining more than 50% (Fig. 1). The impact of standard- or high-dose FFP administration was minor, and all four coagulation factors remained markedly lower than at baseline. In contrast, PCC fully normalized FII, FVII, and FX. While the effect was smaller, PCC did also increase FIX concentration by a median of 7.8% (CI 6.4–9.4%) from the post-haemodilution level.
standard- or high-dose FFP administration was minor, and all four coagulation factors remained markedly lower than at baseline. In contrast, PCC fully normalized FII, FVII, and FX. While the effect was smaller, PCC did also increase FIX concentration by a median of 7.8% (CI 6.4–9.4%) from the post-haemodilution level. Fig 1 Levels of (a) FII, (b) FVII, (c) FIX, and (d) FX at baseline and after haemodilution and subsequent administration of FFP or PCC. Concentration values expressed as a percentage of the baseline median. In each panel is shown the median change from baseline in animals receiving FFP and PCC and corresponding CI. Horizontal lines within boxes indicate the medians, lower and upper box boundaries the 25th and 75th percentiles, respectively, and lower and upper error bars the 10th and 90th percentiles. Individual animal data points and horizontal lines representing the medians are displayed for the 40 ml kg−1 FFP group. CI, 95% confidence interval; FFP, fresh frozen plasma; FII, factor II; FVII, factor VII; FIX, factor IX; FX, factor X; PCC, prothrombin complex concentrate. Haemodilution was also attended by a major decrease in circulating fibrinogen concentration, from a median of 3.34 to 1.82 g litre−1 (Table 1). Nonetheless, fibrinogen level equalled or exceeded 1 g litre−1 in all animals subsequent to haemodilution. After FFP or PCC administration, median fibrinogen remained near the post-haemodilution level. In no recipient of either FFP or PCC did fibrinogen level fall below 1 g litre−1.
of 3.34 to 1.82 g litre−1 (Table 1). Nonetheless, fibrinogen level equalled or exceeded 1 g litre−1 in all animals subsequent to haemodilution. After FFP or PCC administration, median fibrinogen remained near the post-haemodilution level. In no recipient of either FFP or PCC did fibrinogen level fall below 1 g litre−1. Table 1 Fibrinogen. FFP, fresh frozen plasma; IQR, interquartile range; PCC, prothrombin complex concentrate. *Fibrinogen concentration was not determined in the placebo group. Baseline fibrinogen was not evaluable for one animal in the PCC group due to technical difficulties with the plasma sample Category n Fibrinogen (g litre−1), median (IQR) Baseline 31* 3.34 (2.94–3.82) Haemodilution 32 1.82 (1.52–2.04) 15 ml kg−1 FFP 14 2.04 (2.01–2.18) 40 ml kg−1 FFP 5 2.33 (2.26–2.37) PCC 13 1.79 (1.54–1.91) Prothrombin time Haemodilution prolonged PT (Fig. 2a). Saline administration further lengthened PT, whereas FFP partially reversed the haemodilution-induced PT prolongation. Of the tested fluids, PCC was the only one to normalize PT fully. Fig 2 (a) Prothrombin time, (b) peak thrombin generation, and (c) ETP after haemodilution and subsequent administration of saline, FFP, or PCC. Graphic conventions as in Figure 1. CI, 95% confidence interval; FFP, fresh frozen plasma; ETP, endogenous thrombin potential; PCC, prothrombin complex concentrate.
Category n Fibrinogen (g litre−1), median (IQR) Baseline 31* 3.34 (2.94–3.82) Haemodilution 32 1.82 (1.52–2.04) 15 ml kg−1 FFP 14 2.04 (2.01–2.18) 40 ml kg−1 FFP 5 2.33 (2.26–2.37) PCC 13 1.79 (1.54–1.91) Prothrombin time Haemodilution prolonged PT (Fig. 2a). Saline administration further lengthened PT, whereas FFP partially reversed the haemodilution-induced PT prolongation. Of the tested fluids, PCC was the only one to normalize PT fully. Fig 2 (a) Prothrombin time, (b) peak thrombin generation, and (c) ETP after haemodilution and subsequent administration of saline, FFP, or PCC. Graphic conventions as in Figure 1. CI, 95% confidence interval; FFP, fresh frozen plasma; ETP, endogenous thrombin potential; PCC, prothrombin complex concentrate. Platelets Platelet counts were markedly reduced by haemodilution. The median value declined to 161×109 litre−1 (IQR 138–197×109 litre−1) from 444×109 litre−1 (IQR 374–506×109 litre−1) at baseline. Substantial depletion in platelets persisted after administration of all test fluids. Thrombin generation Haemodilution strongly attenuated peak thrombin generation (Fig. 2b). The haemodilution-induced decline in peak thrombin generation was not reversed by either saline or FFP (Fig. 2b). In contrast, PCC restored peak thrombin generation to a level not significantly different from that at baseline.
Platelets Platelet counts were markedly reduced by haemodilution. The median value declined to 161×109 litre−1 (IQR 138–197×109 litre−1) from 444×109 litre−1 (IQR 374–506×109 litre−1) at baseline. Substantial depletion in platelets persisted after administration of all test fluids. Thrombin generation Haemodilution strongly attenuated peak thrombin generation (Fig. 2b). The haemodilution-induced decline in peak thrombin generation was not reversed by either saline or FFP (Fig. 2b). In contrast, PCC restored peak thrombin generation to a level not significantly different from that at baseline. Despite the observed attenuation of the thrombin peak, the impact of haemodilution on endogenous thrombin potential (ETP) was negligible (Fig. 2c). This observation reflected the persistence of higher post-peak thrombin concentrations, thus offsetting the reduction in peak values. The effects of saline and FFP on ETP were both minimal. PCC, however, augmented ETP by a median of nearly 600 nM min above baseline. Haemostasis After either experimental bone or spleen trauma, PCC significantly shortened time to haemostasis compared with FFP (Fig. 3). Haemostasis was achieved by all 13 PCC-treated animals. Over the course of the 120 min observation period, bleeding failed to cease in four of 14 saline recipients (29%), five of 13 animals in the 15 ml kg−1 FFP group (38%), and three of four evaluable pigs (75%) treated with 40 ml kg−1 FFP. Compared with FFP, PCC also significantly lowered the volume of blood lost following either bone or spleen trauma (Fig. 4).
on period, bleeding failed to cease in four of 14 saline recipients (29%), five of 13 animals in the 15 ml kg−1 FFP group (38%), and three of four evaluable pigs (75%) treated with 40 ml kg−1 FFP. Compared with FFP, PCC also significantly lowered the volume of blood lost following either bone or spleen trauma (Fig. 4). Fig 3 Time to haemostasis following experimental (a) bone or (b) spleen trauma in animals treated with saline, FFP, or PCC. FFP, fresh frozen plasma; PCC, prothrombin complex concentrate. Fig 4 Individual animal blood losses following experimental (a) bone or (b) spleen trauma in groups treated with saline, FFP or PCC. Horizontal lines depict the medians. FFP, fresh frozen plasma; PCC, prothrombin complex concentrate. From a baseline median of 117 s, SBT was more than doubled by haemodilution (Table 2). Saline placebo and FFP showed little effect in shortening the prolonged SBT. However, PCC normalized SBT. Table 2 Skin bleeding time. FFP, fresh frozen plasma; IQR, interquartile range; PCC, prothrombin complex concentrate; SBT, skin bleeding time. *One animal excluded due to the development of hypotension
From a baseline median of 117 s, SBT was more than doubled by haemodilution (Table 2). Saline placebo and FFP showed little effect in shortening the prolonged SBT. However, PCC normalized SBT. Table 2 Skin bleeding time. FFP, fresh frozen plasma; IQR, interquartile range; PCC, prothrombin complex concentrate; SBT, skin bleeding time. *One animal excluded due to the development of hypotension Category n SBT (s), median (IQR) Baseline 47 117 (108–127) Haemodilution 47 305 (279–329) Placebo 14 288 (273–311) 15 ml kg−1 FFP 15 250 (219–280) 40 ml kg−1 FFP 4* 234 (177–244) PCC 13 137 (126–152) Discussion Acquired coagulopathy following trauma is an important determinant of outcome. Coagulopathic patients, including those with head injury, experience worse outcomes than patients with the same injury severity but no clotting disturbance.9 Therefore, a major clinical need exists for modalities that quickly and effectively reverse the dilutional coagulopathy occurring over the course of trauma treatment. Pharmacological treatment options consist of antifibrinolytics, fresh whole blood and plasma, FFP, cryoprecipitate, and coagulation factor concentrates.2933 This is the first study to compare PCC with FFP for correction of coagulopathy under conditions of haemodilution and haemorrhage in a porcine trauma model.
atment. Pharmacological treatment options consist of antifibrinolytics, fresh whole blood and plasma, FFP, cryoprecipitate, and coagulation factor concentrates.2933 This is the first study to compare PCC with FFP for correction of coagulopathy under conditions of haemodilution and haemorrhage in a porcine trauma model. Time to haemostasis and blood loss following either femur or spleen injury were both significantly diminished by PCC compared with FFP treatment. The ineffectiveness of FFP with respect to these primary endpoints appears to reflect persistent coagulation factor deficiency after administration of a standard FFP dose (15 ml kg−1) comparable to the typical adult dose used in clinical practice, as well as after a much higher dose (40 ml kg−1).34 In the absence of ongoing coagulation factor consumption or loss, a dose of 10–15 ml kg−1 FFP is expected to raise factor levels by 25% in humans.33 In the porcine model, smaller increases were observed, possibly due to the use of human coagulation factor-deficient plasma in the porcine coagulation factor assays. The availability of species-specific coagulation factor-deficient plasma would be needed to assess the contribution, if any, of assay method to the relatively small increases in coagulation factor levels following FFP administration. On the other hand, the normalization of FII, FVII, and FX levels by PCC in the study animals suggests that assay inaccuracy may not explain the lack of FFP effect in restoring levels of those coagulation factors. In any case, high-dose FFP proved to be no more effective than the standard dose in speeding haemostasis and restricting blood loss after bone injury.
, FVII, and FX levels by PCC in the study animals suggests that assay inaccuracy may not explain the lack of FFP effect in restoring levels of those coagulation factors. In any case, high-dose FFP proved to be no more effective than the standard dose in speeding haemostasis and restricting blood loss after bone injury. In patients, greater restoration of coagulation factor levels with FFP is hampered by the risk of fluid overload. Additionally, during preparation of FFP for clinical use, coagulation factors are diluted by ∼15% with citrate, and further losses are believed to occur during freezing and thawing.3536 Due to the high coagulation factor concentrations in PCC, FII, FVII, and FX rebounded fully to baseline levels after PCC administration in the model system. The small FIX rise produced by PCC was similar in magnitude to that by 15 ml kg−1 porcine FFP. This observation reflects the comparatively low FIX dose in the administered 0.8 ml kg−1 PCC volume. Whereas the doses of the other three coagulation factors in the administered PCC volume were 2–4-fold those in 15 ml kg−1 porcine FFP, the corresponding dose of FIX was 33% lower. In any case, PCC did produce a small but statistically significant increase in FIX which, in concert with full normalization of the other three coagulation factors, proved sufficient to enhance haemostasis.
ministered PCC volume were 2–4-fold those in 15 ml kg−1 porcine FFP, the corresponding dose of FIX was 33% lower. In any case, PCC did produce a small but statistically significant increase in FIX which, in concert with full normalization of the other three coagulation factors, proved sufficient to enhance haemostasis. Another advantage of PCC is viral safety. Most FFP preparations are not subjected to viral inactivation. Prepared from plasma screened by polymerase chain reaction, Beriplex P/N is pasteurized and nanofiltrated to eliminate viruses.3037 In clinical trials there has been no evidence of viral transmission following Beriplex P/N administration.2031 Although the coagulation factor increase was relatively small, FFP did partially reverse the haemodilution-induced PT prolongation. PCC, however, entirely normalized PT. These findings are consistent with prior data. In a prospective audit of FFP transfusion in patients with mild PT prolongation (13–17 s), halfway normalization of PT was accomplished in only 15% of cases and full correction in <1%.38 In contrast, Beriplex P/N administration effectively normalized PT in critically ill patients with moderately reduced coagulation activity.39 In a porcine model of dilutional coagulopathy similar to that in the present study, PCC plus fibrinogen normalized PT, while normal saline was ineffective.40
correction in <1%.38 In contrast, Beriplex P/N administration effectively normalized PT in critically ill patients with moderately reduced coagulation activity.39 In a porcine model of dilutional coagulopathy similar to that in the present study, PCC plus fibrinogen normalized PT, while normal saline was ineffective.40 It is generally recognized that both PT and aPTT are insensitive measures for detecting hypocoagulant conditions.36 The value of SBT in monitoring the coagulation system and directing therapeutic inventions remains to be established. In this study, the SBT results closely coincided with those for time to haemostasis and volume of blood loss after bone or spleen trauma. While platelet function is a major determinant of SBT, the coagulation cascade may also play a role. In rats, the direct thrombin inhibitor melagatran has been shown to prolong SBT, and the prolongation could be reversed by PCC.41 TGAs are being actively investigated for their utility in characterizing and monitoring both hypocoagulable and hypercoagulable states. In the present study, TGA data were consistent with the bleeding endpoint results. Thus, PCC normalized the thrombin peak, whereas FFP had negligible effect.
It is generally recognized that both PT and aPTT are insensitive measures for detecting hypocoagulant conditions.36 The value of SBT in monitoring the coagulation system and directing therapeutic inventions remains to be established. In this study, the SBT results closely coincided with those for time to haemostasis and volume of blood loss after bone or spleen trauma. While platelet function is a major determinant of SBT, the coagulation cascade may also play a role. In rats, the direct thrombin inhibitor melagatran has been shown to prolong SBT, and the prolongation could be reversed by PCC.41 TGAs are being actively investigated for their utility in characterizing and monitoring both hypocoagulable and hypercoagulable states. In the present study, TGA data were consistent with the bleeding endpoint results. Thus, PCC normalized the thrombin peak, whereas FFP had negligible effect. TGA has been used to monitor the response to FFP administration in surgery patients.36 Thrombin generation parameters and fibrinogen levels were higher in post-transfusion plasma from patients who stopped bleeding than from those with ongoing haemorrhage. An in vitro study has demonstrated the ability of PCC to restore thrombin generation in plasma from orally anticoagulated patients.42 TGA has also been used to monitor the effects of activated PCC.254344
were higher in post-transfusion plasma from patients who stopped bleeding than from those with ongoing haemorrhage. An in vitro study has demonstrated the ability of PCC to restore thrombin generation in plasma from orally anticoagulated patients.42 TGA has also been used to monitor the effects of activated PCC.254344 Both platelet count and fibrinogen concentration were substantially reduced by haemodilution and did not recover in response to PCC or FFP treatment. However, the platelet number and fibrinogen concentration were not below the thresholds to sustain the competence of the coagulation system, as demonstrated by the ability of PCC to control femur and spleen bleeding. The porcine model of dilutional coagulopathy employed in the present study has been previously established for evaluating the effectiveness of i.v. fluids in normalizing coagulation function and reducing bleeding resulting from a subsequent haemorrhagic challenge in the form of a standardized experimental femur or spleen injury.29 The model simulates clinical situations in which, after initial traumatic haemorrhagic shock and resuscitation, haemostasis has been secured, but dilutional coagulopathy needs to be corrected to prevent excessive bleeding during further surgical interventions the patient may require. A similar porcine dilutional coagulopathy model has been described, in which placebo or coagulation factors were administered prior to a standardized hepatic laceration, and subsequent blood loss and survival were assessed.40
d to prevent excessive bleeding during further surgical interventions the patient may require. A similar porcine dilutional coagulopathy model has been described, in which placebo or coagulation factors were administered prior to a standardized hepatic laceration, and subsequent blood loss and survival were assessed.40 In this porcine trauma model of dilutional coagulopathy and haemorrhage, PCC proved superior to FFP in normalizing PT, SBT, and peak thrombin generation and controlling bleeding. These findings support a potential role for PCC in coagulopathic trauma and surgery patients. In view of the unmet clinical need for more efficacious haemostatic agents in such patients, clinical studies are now justified to confirm the observed favourable effects of PCC in the present preclinical model system. Funding The authors received an unrestricted grant from CSL Behring. Funding to Pay the Open Access Charge was provided by CSL Behring. Acknowledgements Prothrombin complex concentrate for use in this study was provided by CSL Behring GmbH, Marburg, Germany, the commercial supplier of that product and sponsor of the study.
There is evidence to support the use of opioids for carefully selected patients with chronic non-malignant pain.15282963 An important and controversial issue concerning the management of chronic pain is assessment and treatment of breakthrough pain. Although there is currently no unanimous definition of breakthrough pain in chronic malignant or non-malignant pain,64 a consensus panel recommendation from 2005 suggested that breakthrough pain should be defined as ‘an abrupt, short-lived, and intense pain that “breaks through” the around-the-clock (ATC) analgesia that controls persistent pain’.6 Subtypes of breakthrough pain include incident (often predictable and precipitated by activity or movement), idiopathic or spontaneous, and end-of-dose pain.664
pain should be defined as ‘an abrupt, short-lived, and intense pain that “breaks through” the around-the-clock (ATC) analgesia that controls persistent pain’.6 Subtypes of breakthrough pain include incident (often predictable and precipitated by activity or movement), idiopathic or spontaneous, and end-of-dose pain.664 Breakthrough pain is best described in patients with malignant pain, leading to a number of adverse effects including a more severe pain syndrome.135052 However, studies have shown that breakthrough pain is also prevalent among opioid-treated patients with chronic non-malignant pain conditions and impacts negatively on their quality of life.82449596673 In a recent survey of chronic non-malignant pain, 74% of patients (most commonly, low back pain) with opioid-controlled baseline pain reported breakthrough pain, reaching maximum intensity within a median of 10 min and lasting for a median of 60 min.49 In these prevalence studies, the majority of cases of breakthrough pain were precipitated (incident pain), although end-of-dose pain was also reported.84959 It was proposed recently that peripheral, central, or both sensitization may be a common component of breakthrough pain in both malignant and non-malignant diseases.64
hese prevalence studies, the majority of cases of breakthrough pain were precipitated (incident pain), although end-of-dose pain was also reported.84959 It was proposed recently that peripheral, central, or both sensitization may be a common component of breakthrough pain in both malignant and non-malignant diseases.64 The primary treatment for breakthrough pain in malignant pain is with immediate-release (IR), short-acting opioids on a pre-emptive or as-needed basis, in addition to the usual opioid regimen.736 This strategy has been adopted for chronic non-malignant pain. However, there is little evidence that using short-acting opioids as rescue medication is an optimal long-term treatment strategy in chronic non-malignant pain.
-acting opioids on a pre-emptive or as-needed basis, in addition to the usual opioid regimen.736 This strategy has been adopted for chronic non-malignant pain. However, there is little evidence that using short-acting opioids as rescue medication is an optimal long-term treatment strategy in chronic non-malignant pain. Evidence is needed for the routine long-term use of short-acting opioids for breakthrough pain in patients with chronic non-malignant pain, particularly as there is a school of thought that exposure to short-acting opioids might increase the risk of abuse, opioid tolerance and the need for dose escalation, or inadequate use of opioids. These are important unknown factors in our understanding of opioid usage in pain relief. In addition, many short-acting opioids are ineffective for certain types of breakthrough pain, as their onset of action is outside the window of maximum pain intensity of the breakthrough pain episode.495052 Rapid onset formulations (e.g. oral transmucosal fentanyl) have been developed to address this.3651606266 We are not aware of any randomized controlled studies comparing the efficacy and tolerability of long-acting opioid treatment in chronic non-malignant pain patients who have access to IR opioids for breakthrough pain and those who do not. We have systematically reviewed the literature to approach this question. We compared the analgesic efficacy and incidence of common opioid side-effects between studies of long-acting opioids in chronic non-malignant pain that did and did not allow the use of IR opioid rescue medication using meta-regression analyses.
We have systematically reviewed the literature to approach this question. We compared the analgesic efficacy and incidence of common opioid side-effects between studies of long-acting opioids in chronic non-malignant pain that did and did not allow the use of IR opioid rescue medication using meta-regression analyses. Methods The primary objective of this review was to compare clinical studies, in patients with chronic non-malignant pain, of long-acting opioids that allowed IR opioid rescue medication (‘rescue’ studies) with those that did not (‘no rescue’ studies) to determine the impact of opioid rescue medication use on the analgesic efficacy of chronic opioid therapy among. As a secondary objective, the impact of opioid rescue medication use on the tolerability of chronic opioid therapy in terms of common opioid side-effects (nausea, constipation, and somnolence/sedation) was also investigated.
pact of opioid rescue medication use on the analgesic efficacy of chronic opioid therapy among. As a secondary objective, the impact of opioid rescue medication use on the tolerability of chronic opioid therapy in terms of common opioid side-effects (nausea, constipation, and somnolence/sedation) was also investigated. Literature search The following electronic databases were searched for articles relevant to this systematic review: MEDLINE (1950 to July 2006) and EMBASE (1974 to July 2006). In the MEDLINE search, terms for long-acting opioid analgesics [i.e. (‘analgesics-opioid’, ‘opioid’, ‘narcotic’, ‘fentanyl’, ‘morphine’, ‘hydromorphone’, ‘hydrocodone’, ‘oxycodone’, ‘oxymorphone’, ‘codeine’, ‘dihydrocodeine’, ‘pethidine’, ‘meperidine’, or ‘tramadol’) and (‘delayed-action preparations’, ‘long-acting’, ‘contin’, ‘OROS’, ‘SODAS’, ‘TIMERX’, ‘sustained-release/action’, ‘controlled-release/action’, ‘delayed-release/action’, ‘extended-release/action’, ‘slow-release/action’, ‘timed-release/action’, ‘modified-release/action’, ‘continuous-release/action’, ‘transdermal’, ‘TTS’, ‘TDS’, ‘ER’, ‘CR’, or ‘SR’) or ‘buprenorphine’ or ‘methadone’] were combined with terms for non-malignant (i.e. ‘noncancer’, ‘nonmalignant’, ‘nononcologic’, ‘nontumour’, ‘multimorbidity’, ‘low back’, ‘chronic musculoskeletal’, ‘osteogenic’, ‘phantom limb’, ‘vascular-diseases/disorders’, ‘chronic pancreatitis’, ‘coronary arteriosclerosis’, ‘coronary atherosclerosis’, ‘neuralgia-postherpetic’, ‘trigeminal-neuralgia’, ‘diabetic neuropathies’, ‘amyloid-neuropathies’, ‘brachial-plexus-neuropathies’, ‘mononeuropathies’, ‘polyneuropathies’, ‘neuropathy’, ‘neuralgia’, ‘arthritis-rheumatoid’, ‘osteoarthritis’, or ‘osteoporosis’) and pain (i.e. ‘pain’ or ‘analgesia’) or terms for specific non-malignant pain disorders (i.e. ‘low back pain’ or ‘complex regional pain syndrome’). A number of study design search terms were also included in the strategy. Equivalent search terms specific to EMBASE were used for the EMBASE search. The full MEDLINE and EMBASE search strategies are given in Supplementary material, Appendix 1.
pain disorders (i.e. ‘low back pain’ or ‘complex regional pain syndrome’). A number of study design search terms were also included in the strategy. Equivalent search terms specific to EMBASE were used for the EMBASE search. The full MEDLINE and EMBASE search strategies are given in Supplementary material, Appendix 1. All English, German, French, Spanish, or Italian language full-text research articles were eligible for inclusion, as long as they met the criteria defined below. All published, prospective, blinded, or open-label clinical trials with either a randomized or controlled design were eligible. Prospective observational studies were also accepted. Case reports, conference proceedings, or retrospective studies (surveys or audits) were excluded from this review. Eligible patients were adults aged ≥18 yr with chronic non-malignant pain (such as chronic musculoskeletal pain, vascular disorders, chronic pancreatitis, lower back pain, osteogenic pain, coronary artery disease, phantom limb pain, post-herpetic neuralgia, trigeminal neuralgia, diabetic neuropathy, or neuropathic pain). Studies that included patients with malignant pain were eligible only if less than one-third of the study population were patients with malignant pain. Studies in both opioid-naïve patients and patients already on opioids were eligible.
erpetic neuralgia, trigeminal neuralgia, diabetic neuropathy, or neuropathic pain). Studies that included patients with malignant pain were eligible only if less than one-third of the study population were patients with malignant pain. Studies in both opioid-naïve patients and patients already on opioids were eligible. Treatments included in the review were long-acting opioids or opioid formulations [such as methadone, transdermal fentanyl, or sustained-release (SR) morphine] including the weak opioids, tramadol, codeine, and slow-release dihydrocodeine. All routes of administration, other than intrathecal or intraspinal, were eligible. Since intrathecal use of opioids is an invasive method often used as a last resort in a selection of patients unresponsive to other treatments, we felt that that patient group is not comparable with those receiving oral or transdermal opioids. All studies in which one or more opioids was administered for at least 4 weeks, including those that compared one opioid with placebo, other opioids, or other active controls were eligible for inclusion. Studies had to have a flexible dosing regimen (at least during the maintenance phase); studies with specific criteria for dose escalation or dose escalation to a maximum dose were eligible. Whether an IR opioid was used as rescue medication or not had to be defined. If this information was not given explicitly in the publication, a single attempt was made to contact the authors to obtain this information before the study was excluded.
Treatments included in the review were long-acting opioids or opioid formulations [such as methadone, transdermal fentanyl, or sustained-release (SR) morphine] including the weak opioids, tramadol, codeine, and slow-release dihydrocodeine. All routes of administration, other than intrathecal or intraspinal, were eligible. Since intrathecal use of opioids is an invasive method often used as a last resort in a selection of patients unresponsive to other treatments, we felt that that patient group is not comparable with those receiving oral or transdermal opioids. All studies in which one or more opioids was administered for at least 4 weeks, including those that compared one opioid with placebo, other opioids, or other active controls were eligible for inclusion. Studies had to have a flexible dosing regimen (at least during the maintenance phase); studies with specific criteria for dose escalation or dose escalation to a maximum dose were eligible. Whether an IR opioid was used as rescue medication or not had to be defined. If this information was not given explicitly in the publication, a single attempt was made to contact the authors to obtain this information before the study was excluded. Studies reporting at least one of the following outcome measures were eligible. A measure of analgesic efficacy, such as pain relief or pain intensity measured by the Brief Pain Inventory, a visual analogue scale, or equivalent, or patient and investigator global assessment of analgesic efficacy.
Whether an IR opioid was used as rescue medication or not had to be defined. If this information was not given explicitly in the publication, a single attempt was made to contact the authors to obtain this information before the study was excluded. Studies reporting at least one of the following outcome measures were eligible. A measure of analgesic efficacy, such as pain relief or pain intensity measured by the Brief Pain Inventory, a visual analogue scale, or equivalent, or patient and investigator global assessment of analgesic efficacy. Tolerability: number or percentage of patients reporting either nausea, constipation, or somnolence/sedation (or drowsiness if somnolence/sedation not reported).
Studies reporting at least one of the following outcome measures were eligible. A measure of analgesic efficacy, such as pain relief or pain intensity measured by the Brief Pain Inventory, a visual analogue scale, or equivalent, or patient and investigator global assessment of analgesic efficacy. Tolerability: number or percentage of patients reporting either nausea, constipation, or somnolence/sedation (or drowsiness if somnolence/sedation not reported). Selection of studies Titles and abstracts for studies identified in the literature search were reviewed for inclusion according to the selection criteria described above. Those studies that clearly failed to meet the selection criteria were discarded immediately. The full-text articles were obtained for all studies that appeared to meet the criteria and for those where insufficient information was available in the abstract to determine eligibility. All full-text articles were reviewed independently by two reviewers (H.W. and either J.D. or U.R.) to ensure that all the inclusion criteria were met adequately. Any disagreements between the two reviewers were resolved by discussion and consensus without the need to consult a third party. Where it was not clear whether the use of rescue medication had been allowed during the study or not, a single e-mail (no reminders) was sent out to the corresponding author requesting this information before the article was excluded. For usual care or non-interventional trials that did not explicitly report the use of rescue medication, it was assumed that rescue medication was allowed as per usual clinical practice. Eighty-seven potential studies were identified by MEDLINE and EMBASE and the full text was reviewed; of these, 47 were found on closer inspection to meet all the study selection criteria. The majority of excluded studies failed to meet more than one of the selection criteria and were therefore not excluded on the basis of one criterion. Factors commonly leading to exclusion included: failure to confirm whether opioid rescue medication was allowed or not, study duration <4 weeks, incorrect intervention (e.g. intrathecal administration), or inflexible dosing regimen. An additional study46 that we were aware was in press at the time of the search and therefore not picked up by the search strategy was also included, as it met all the selection criteria.
was allowed or not, study duration <4 weeks, incorrect intervention (e.g. intrathecal administration), or inflexible dosing regimen. An additional study46 that we were aware was in press at the time of the search and therefore not picked up by the search strategy was also included, as it met all the selection criteria. Data extraction and outcome measures The following data were extracted from each of the included studies: study design; type of non-malignant pain or condition studied; patient numbers; intervention (type of opioid and route of administration); duration of treatment; the primary or first reported (in sufficient detail for analysis) analgesic outcome (scale, measure at baseline and endpoint, change in measure or difference vs placebo, and measures of variation as appropriate); number, per cent, or both of patients reporting nausea, constipation, or somnolence/sedation (or drowsiness, if somnolence/sedation not reported); and number or percentage of patients using rescue medication (for rescue studies only). We applied the Jadad criteria for randomized controlled trials to each study to determine the methodological quality and validity of each study.25 The criteria were applied independently by two reviewers (H.W. and N.M.). There were three disagreements between the two independent reviewers, which were resolved by a third party (A.J.). In brief, a study received one point for each ‘yes’ or zero points for each ‘no’ given in reply to the questions provided in Appendix 2 in the Supplementary material (maximum number of points is 5).
We applied the Jadad criteria for randomized controlled trials to each study to determine the methodological quality and validity of each study.25 The criteria were applied independently by two reviewers (H.W. and N.M.). There were three disagreements between the two independent reviewers, which were resolved by a third party (A.J.). In brief, a study received one point for each ‘yes’ or zero points for each ‘no’ given in reply to the questions provided in Appendix 2 in the Supplementary material (maximum number of points is 5). Statistical analysis The primary outcome was analgesic efficacy and the secondary outcomes were the percentage of patients with nausea, constipation, or somnolence. All outcome variables were analysed in a meta-regression analysis.68 Meta-analysis would normally be used to investigate the treatment effect within studies, but since there are no studies directly comparing the effect of opioid rescue medication use on long-term analgesic efficacy, we have used meta-regression to make indirect comparisons between the ‘rescue’ and the ‘no rescue’ studies. For the purposes of the analysis, certain study designs were treated in the following way. Studies comparing two long-acting opioids were treated as two separate uncontrolled studies. In two crossover studies comparing transdermal fentanyl with SR morphine, only the experimental transdermal fentanyl arm was used in the adverse event (AE) analysis (these studies were both excluded from the primary analysis).
For the purposes of the analysis, certain study designs were treated in the following way. Studies comparing two long-acting opioids were treated as two separate uncontrolled studies. In two crossover studies comparing transdermal fentanyl with SR morphine, only the experimental transdermal fentanyl arm was used in the adverse event (AE) analysis (these studies were both excluded from the primary analysis). In studies comparing long- and short-acting opioids, the long-acting opioid arm was treated as an uncontrolled study. The rationale for this is that in equivalence studies of SR opioids, short-acting opioids are often used to show efficacy of the SR formulation. In these studies, patients take the short-acting opioids regularly; therefore, you would expect similar pain control as with the SR opioid. Since placebo-controlled studies often allow continuation of non-opioid analgesics during the study period, studies comparing a long-acting opioid with a non-opioid analgesic were treated as placebo-controlled studies.
In studies comparing long- and short-acting opioids, the long-acting opioid arm was treated as an uncontrolled study. The rationale for this is that in equivalence studies of SR opioids, short-acting opioids are often used to show efficacy of the SR formulation. In these studies, patients take the short-acting opioids regularly; therefore, you would expect similar pain control as with the SR opioid. Since placebo-controlled studies often allow continuation of non-opioid analgesics during the study period, studies comparing a long-acting opioid with a non-opioid analgesic were treated as placebo-controlled studies. For each study, the primary pain outcome or the first pain outcome reported in sufficient detail was used in the meta-regression analysis. A measure of the treatment effect, which was the difference between opioid and control in controlled studies and the difference from baseline in uncontrolled studies, was calculated based on the published data. The standard error (se) of the treatment effect was also calculated, using standard formulas if the necessary information was available. The following assumptions were made for uncontrolled studies, if the necessary information for calculating the SE was not available. Mean at baseline or endpoint=median at baseline or endpoint. Inter-quartile range (IQR)=1.35×standard deviation (sd) if IQR is specified and sd is missing. If measure of variability of the change was unavailable and sd of the endpoint value was available, the sd of the change was assumed to be equal to the sd of the endpoint value.
For each study, the primary pain outcome or the first pain outcome reported in sufficient detail was used in the meta-regression analysis. A measure of the treatment effect, which was the difference between opioid and control in controlled studies and the difference from baseline in uncontrolled studies, was calculated based on the published data. The standard error (se) of the treatment effect was also calculated, using standard formulas if the necessary information was available. The following assumptions were made for uncontrolled studies, if the necessary information for calculating the SE was not available. Mean at baseline or endpoint=median at baseline or endpoint. Inter-quartile range (IQR)=1.35×standard deviation (sd) if IQR is specified and sd is missing. If measure of variability of the change was unavailable and sd of the endpoint value was available, the sd of the change was assumed to be equal to the sd of the endpoint value. The treatment effect for each study was converted where necessary onto a scale of 0–100 (higher scores equal worse pain or less pain relief). This was done by multiplying the treatment effect by 100/maximum scale score. Analgesic effect was analysed in a meta-regression analysis using Stata 9.2.68 Meta-regression was used to investigate the effect of rescue medication on treatment effect, while controlling for other potentially confounding variables, namely study design, Jadad score, and type of opioid. The regression coefficient from the analysis is the estimated average difference in analgesic efficacy (number of points on a 0–100 scale) between the ‘rescue’ and the ‘no rescue’ medication studies; negative values are in favour of rescue medication. For other variables, negative scores are in favour of the specific category in question and positive values are in favour of the reference category.
in analgesic efficacy (number of points on a 0–100 scale) between the ‘rescue’ and the ‘no rescue’ medication studies; negative values are in favour of rescue medication. For other variables, negative scores are in favour of the specific category in question and positive values are in favour of the reference category. The percentage of patients treated with long-acting opioid with nausea, constipation, or somnolence/sedation was calculated for each study and the se of the percentage was calculated as the width of the exact binomial confidence interval divided by 2×1.96 (this is more appropriate for small sample sizes than the normal approximation to the se). For studies in which the incidence of AEs was reported separately for different parts of the study, the highest incidence values were used. As for the efficacy outcome, meta-regression was used to investigate the effect of rescue medication, while controlling for other potentially confounding variables. The logit of the percentage of patients with each AE was also used as a sensitivity analysis, but the results were similar and are therefore not presented. The logit and se were calculated by fitting an intercept-only logistic regression model for each outcome. The regression coefficient from the analysis is the estimated average difference in AE incidence (% of patients) between the ‘rescue’ and the ‘no rescue’ medication studies; negative values are in favour of rescue medication. For other variables, negative scores are in favour of the specific category in question and positive values are in favour of the reference category.
timated average difference in AE incidence (% of patients) between the ‘rescue’ and the ‘no rescue’ medication studies; negative values are in favour of rescue medication. For other variables, negative scores are in favour of the specific category in question and positive values are in favour of the reference category. The entire analysis was done again, this time excluding studies with a Jadad score of 0. The analysis was also run excluding the weak opioids, namely tramadol, codeine, tilidine, and dihydrocodeine. In addition, the studies that included any patients with malignant pain were excluded and the analysis was re-run. This was done as a sensitivity analysis to ensure that there was an assessment on pure non-malignant pain. Results Study characteristics and quality Forty-eight studies met all of the study selection criteria and were included in the review; of these, 24 studies allowed the use of rescue medication (‘rescue’ studies) and 24 did not (‘no rescue’ studies). The characteristics of each included study are summarized in the Supplementary material.1–49–1214–1618–232730–3537–4854–5861656770–7275 The frequency of the different types of study design (for the purposes of analysis), control group (controlled studies only), and opioid type along with the quality of the included articles in terms of Jadad score is given for the ‘rescue’ and ‘no rescue’ studies and overall in Table 1. The maximum time of treatment was 2 yr.
quency of the different types of study design (for the purposes of analysis), control group (controlled studies only), and opioid type along with the quality of the included articles in terms of Jadad score is given for the ‘rescue’ and ‘no rescue’ studies and overall in Table 1. The maximum time of treatment was 2 yr. Table 1 Summary of study characteristics and quality. *How the study was treated for the purposes of analysis and does not necessarily reflect the original design of the study. †One study in which the control was non-steroidal anti-inflammatory drug and one study in which the control was morphine-free patients; both were considered as placebo-controlled studies for the purposes of the analysis. ‡Controlled studies only (n=13). §n=26 for the rescue studies, as two of the studies comparing two long-acting opioids were counted twice; each arm was treated as a separate uncontrolled study
ch the control was morphine-free patients; both were considered as placebo-controlled studies for the purposes of the analysis. ‡Controlled studies only (n=13). §n=26 for the rescue studies, as two of the studies comparing two long-acting opioids were counted twice; each arm was treated as a separate uncontrolled study Characteristic No rescue (n=24) Rescue (n=24) Total (n=48) Study design*, n (%) Placebo-controlled crossover 3 (12.5) 0 (0) 3 (6.3) Placebo-controlled, parallel-group† 8 (33.3) 2 (8.3) 10 (20.8) Uncontrolled 13 (54.2) 22 (91.7) 35 (72.9) Control group, n (%)‡ Active placebo 2 (18.2) 0 (0) 2 (15.4) Morphine-free 0 (0) 1 (50.0) 1 (7.7) NSAID 0 (0) 1 (50.0) 1 (7.7) Placebo 9 (81.8) 0 (0) 9 (69.2) Type of opioid, n (%)§ Morphine 4 (16.7) 4 (15.4) 8 (16.0) Oxycodone 6 (25.0) 2 (7.7) 8 (16.0) Fentanyl 3 (12.5) 13 (50.0) 16 (32.0) Tramadol 6 (25.0) 1 (3.9) 7 (14.0) Other 5 (20.8) 6 (23.1) 11 (22.0) Jadad score, n (%) 0 2 (8.3) 2 (8.3) 4 (8.3) 1 11 (45.8) 21 (87.5) 32 (66.7) 2 0 (0) 1 (4.2) 1 (2.1) 3 0 (0) 0 (0) 0 (0) 4 3 (12.5) 0 (0) 3 (6.3) 5 8 (33.3) 0 (0) 8 (16.7) Mean (sd) Jadad score 2.6 (2.0) 0.96 (0.36) 1.8 (1.7) The majority of both the ‘no rescue’ and the ‘rescue’ studies were uncontrolled. All but two of the placebo-controlled, parallel-group studies and all the crossover studies (all placebo-controlled) did not allow rescue medication use. The most common control used in the controlled studies was placebo and the most common opioids in the ‘no rescue’ group were oxycodone and tramadol and in the ‘rescue’ group, fentanyl. As may be expected, there was an imbalance in study quality between the ‘rescue’ and the ‘no rescue’ studies. Most ‘rescue’ studies were of a low quality (Jadad score of 0 or 1), whereas just under half of the ‘no rescue’ studies were of high quality (Jadad score of 4 or 5). For this reason, Jadad score was included as a term in the meta-regression analysis.
ce in study quality between the ‘rescue’ and the ‘no rescue’ studies. Most ‘rescue’ studies were of a low quality (Jadad score of 0 or 1), whereas just under half of the ‘no rescue’ studies were of high quality (Jadad score of 4 or 5). For this reason, Jadad score was included as a term in the meta-regression analysis. Of the 24 studies included that allowed short-acting opioid rescue medication, seven reported the number or percentage of patients who actually took rescue medication. In these seven studies, the percentage of patients who actually took rescue medication ranged from 11% to 100% (median 57%).1192033354245 Analgesic efficacy analysis Of the 48 studies, 40 were included in the meta-regression analysis of analgesic efficacy (Table 2). Eight studies were excluded from the analysis, as it was not possible to calculate either the effect size (n=4) or the se for the effect size (n=4) based on the published data. Table 2 Analgesic efficacy results. *Reference category is placebo-controlled. †Reference category is morphine. CI, confidence interval
Analgesic efficacy analysis Of the 48 studies, 40 were included in the meta-regression analysis of analgesic efficacy (Table 2). Eight studies were excluded from the analysis, as it was not possible to calculate either the effect size (n=4) or the se for the effect size (n=4) based on the published data. Table 2 Analgesic efficacy results. *Reference category is placebo-controlled. †Reference category is morphine. CI, confidence interval Regression coefficient 95% CI P-value Unadjusted analysis Rescue medication −6.3 −16.9, 4.2 0.24 Adjusted for study design Rescue medication −0.3 −11.6, 11.0 0.96 Uncontrolled studies* −16.8 −30.0, −3.7 0.01 Crossover studies* −6.9 −28.7, 14.9 0.54 Adjusted for study design and opioid type Rescue medication −5.2 −17.2, 6.9 0.40 Uncontrolled studies* −16.1 −29.6, −2.6 0.02 Crossover studies* −10.8 −32.7, 11.0 0.33 Oxycodone† −5.6 −22.7, 11.5 0.52 Fentanyl† −2.7 −17.7, 12.3 0.73 Tramadol† −21.4 −39.2, −3.5 0.02 Other† −3.5 −20.6, 13.7 0.69 The results of the unadjusted analysis indicated an average difference of six points on a 0–100 scale for efficacy in favour of rescue medication use, although this difference was not significant (P=0.24) (Table 2). As a sensitivity analysis, studies involving any patients with malignant pain were excluded and the analysis was re-run. Four studies were identified and excluded.19214275 After excluding studies involving any patients with malignant pain, the average difference was −4.5; 95% CI: −15.38, 6.46; P=0.424.
0.24) (Table 2). As a sensitivity analysis, studies involving any patients with malignant pain were excluded and the analysis was re-run. Four studies were identified and excluded.19214275 After excluding studies involving any patients with malignant pain, the average difference was −4.5; 95% CI: −15.38, 6.46; P=0.424. After adjusting for potentially confounding variables, namely study design and both study design and type of opioid, the difference in analgesic efficacy between the ‘rescue’ and the ‘no rescue’ studies was not significant, with regression coefficients close to 0 and 95% confidence intervals that excluded an effect of more than 18 points in each case. The movement of the regression coefficient towards 0 after adjustment for study design is to be expected due to the imbalance in study design between the ‘rescue’ and the ‘no rescue’ studies. Excluding studies that involved any patients with malignant pain and adjusting for study design did not affect the results (regression coefficient 1.2; 95% CI: −10.31, 12.82; P=0.831). Very similar results were obtained when the analysis was adjusted for Jadad score and both Jadad score and study design. This is to be expected, as Jadad score is a measure of study quality based on the study design. Excluding studies with a Jadad score of 0 did not change the outcome of the analgesic efficacy analysis. Excluding weak opioids from the analysis also did not change the findings.
After adjusting for potentially confounding variables, namely study design and both study design and type of opioid, the difference in analgesic efficacy between the ‘rescue’ and the ‘no rescue’ studies was not significant, with regression coefficients close to 0 and 95% confidence intervals that excluded an effect of more than 18 points in each case. The movement of the regression coefficient towards 0 after adjustment for study design is to be expected due to the imbalance in study design between the ‘rescue’ and the ‘no rescue’ studies. Excluding studies that involved any patients with malignant pain and adjusting for study design did not affect the results (regression coefficient 1.2; 95% CI: −10.31, 12.82; P=0.831). Very similar results were obtained when the analysis was adjusted for Jadad score and both Jadad score and study design. This is to be expected, as Jadad score is a measure of study quality based on the study design. Excluding studies with a Jadad score of 0 did not change the outcome of the analgesic efficacy analysis. Excluding weak opioids from the analysis also did not change the findings. In the analyses adjusting for study design, a significantly greater analgesic effect was seen for uncontrolled studies vs placebo-controlled, parallel-group studies, as would be expected (regression coefficient: −16.8; 95% CI: −30.0, −3.7; P=0.01). This was also true after adjusting for study design and opioid type. Interestingly, significantly better analgesic efficacy was seen in studies of tramadol compared with morphine studies (regression coefficient: −21.4; 95% CI: −39.2, −3.5; P=0.02).