Browse the corpus

therwise young and healthy German Landswine undergoing a similar surgical instrumentation. German Landswine (open box-and-whisker plots), FBM swine with (dark gray box-and-whisker plots), and without hyperchloesteremic diet (light gray box-and-whisker plots). All data are median (quartiles, range; n = 5 in each group). Discussion This study was to test the hypothesis whether the PPAR-β/δ agonist GW0742 would attenuate kidney injury during long-term, resuscitated, polymicrobial porcine septic shock. Since GW0742 had been effective in animals with obesity and diabetes, we studied swine with hyperlipidemia and ubiquitous atherosclerosis [15]. The major findings were that (1) GW0742 failed to attenuate sepsis-induced organ dysfunction and histological damage and (2) did not affect the parameters of inflammation and oxidative and nitrosative stress.

Editorial Critical care is a rapidly growing specialty. As a mark of its increasing maturity and health, its practitioners are increasingly challenging long-held dogma and revising practice guidelines. Critical care certainly punches well beyond its weight, accounting for a disproportionately large number of clinical trials reported in the highest impact factor journals. An impressive level of altruistic collaboration has often driven such studies, even on shoestring budgets. Outcomes have improved, in large part due to improvements in general care, early recognition of deterioration, and avoidance of iatrogenic harm. Notwithstanding these achievements, few novel drugs or other interventions that have impacted positively on patient outcomes have come to the fore over the last 20 to 30 years. This disconnect highlights the crucial disparity between the clinical and biological phenotypes of critical illness. Treatment strategies based on clinical syndromic presentations, e.g. for sepsis, acute respiratory distress syndrome, and acute kidney injury, have repeatedly fallen short of their initial promise. The heterogeneity of the population with respect to age, gender, underlying comorbidity, trajectory and severity of illness, time to presentation, infecting organism, and the type and number of affected organs adds further complexity to trialing within an already challenging patient cohort. Some interventions, albeit based on a seemingly sound rationale, have even been associated with patient harm.

bidity, trajectory and severity of illness, time to presentation, infecting organism, and the type and number of affected organs adds further complexity to trialing within an already challenging patient cohort. Some interventions, albeit based on a seemingly sound rationale, have even been associated with patient harm. It is starkly obvious that improving our still limited grasp of the underlying pathophysiology is key to making major advances in patient management. Yet, basic/translational research has been traditionally relegated to a minor role and is generally undervalued by the clinical community. This is reflected by a relatively low profile at (inter)national congresses, fewer funding opportunities, and comparatively few basic scientists specifically focusing on critical care conditions and pathologies. Even the specialist journals generally eschew basic/translational papers, particularly those involving animal or laboratory models. However, in their defence, the quality of such work is somewhat variable, thus creating a perfect Catch 22 situation of struggling to attract the funding and quality researchers needed to improve the overall calibre of the research output.

/translational papers, particularly those involving animal or laboratory models. However, in their defence, the quality of such work is somewhat variable, thus creating a perfect Catch 22 situation of struggling to attract the funding and quality researchers needed to improve the overall calibre of the research output. With this in mind, Intensive Care Medicine, the official journal of the European Society of Intensive Care Medicine (ESICM), has taken the bold step of splitting into two separate but closely linked sister journals. Intensive Care Medicine will continue as a solely clinical journal, reporting clinical trials, process of care, ethical issues, health economics, and so forth. Intensive Care Medicine Experimental (ICMx) will focus on experimental research, stretching from cell and in silico models, through in vivo and ex vivo laboratory studies, to human volunteer and patient studies, where the emphasis is on biology rather than on clinical outcomes. The risk is that clinicians will choose to ignore this stand-alone offering. Yet, we hope that offering a dedicated platform for such research will promote its importance, elevate standards, improve transferability of findings from bench to bedside (and back), and stimulate clinicians, basic scientists and, particularly, the next generations of researchers to actively participate. In conjunction, other initiatives will be taken by ESICM to elevate basic and translational research and encourage engagement.

improve transferability of findings from bench to bedside (and back), and stimulate clinicians, basic scientists and, particularly, the next generations of researchers to actively participate. In conjunction, other initiatives will be taken by ESICM to elevate basic and translational research and encourage engagement. Intensive Care Medicine Experimental (icm-experimental.com) will be an online-only electronic journal that, by being open access, will guarantee rapid and widespread dissemination of data, but with the assurance of decent, high-quality peer review [1]. Being part of the SpringerOpen stable, it provides quality processing of manuscripts and numerous portals to publicize the papers (including social media). This will be complemented by the ESICM website that will promote the journal on its front page. A strong endorsement has come from the ready acceptance by a cadre of 40 top-flight researchers to take on active roles as Senior Editors and Editors, with commitments to lend their significant expertise to the peer review process, to submit their research, and to offer enthusiastic promotion in support of this fledgling journal. These researchers are a healthy mix of basic scientists and clinician scientists whose research activities have led them to embrace experimental research in distinct areas of critical care (inflammation, immunity, respiratory, metabolic, neurological, endocrine, trauma, and so forth). ICMx will thus, uniquely, be driven by the critical care community with a sole focus on basic to translational research.

research activities have led them to embrace experimental research in distinct areas of critical care (inflammation, immunity, respiratory, metabolic, neurological, endocrine, trauma, and so forth). ICMx will thus, uniquely, be driven by the critical care community with a sole focus on basic to translational research. As with any new offering, the first few years will be challenging. We are, however, quietly confident that it will succeed and sincerely hope the scientific community will join in this exciting journey.

Introduction The prevalence of acute kidney injury (AKI) in septic patients is high, approximately 20% to 50%, and the AKI-related mortality rate is 75% in septic shock compared to 45% without sepsis [1, 2]. The pathogenesis and pathophysiology of sepsis-induced AKI is highly complex, and current treatment strategies are mainly supportive rather than curative [3]. There is a growing body of evidence that microcirculatory dysfunction accompanied by tissue dysoxia might play a key role in the development of septic organ damage [4, 5]. Excessive release of pro-inflammatory mediators and disturbances in the coagulation system are believed to be involved in the pathogenesis of sepsis; both leading to microcirculatory dysfunction and consequent organ failure [6, 7]. Activated protein C (APC) is an important endogenous protein that modulates coagulation and inflammation by promoting fibrinolysis and inhibiting thrombosis and inflammation [8–10]. Different experimental and clinical studies could demonstrate that the administration of APC improved outcome of severe sepsis [11–15]. With respect to the protective effects of APC on the kidney during sepsis, especially the group of Gupta et al. has identified specific mechanisms of action of APC in rat models of experimentally induced septic AKI [16–19]. Another group, furthermore, showed that APC reduced ischemia/reperfusion (I/R)-induced renal injury in rats [20].

he protective effects of APC on the kidney during sepsis, especially the group of Gupta et al. has identified specific mechanisms of action of APC in rat models of experimentally induced septic AKI [16–19]. Another group, furthermore, showed that APC reduced ischemia/reperfusion (I/R)-induced renal injury in rats [20]. Besides its potentially beneficial effects, there are serious concerns regarding the safety and efficacy of APC treatment in critically ill septic patients [21]. These concerns led to discontinuation of all ongoing clinical trials using APC for treatment of severe sepsis. APC is no longer suggested for the treatment of severe sepsis or septic shock. Despite these developments, investigating the effects of APC in experimental studies may provide more insights into the development of septic AKI. The above mentioned studies have provided key insights into the beneficial effects of APC in sepsis; however, none of these studies have investigated the effects of APC on renal oxygenation and function in terms of tubular sodium reabsorption. In the present study, we tested whether continuous recombinant human APC administration would be able to protect renal oxygenation and function during the acute phase of endotoxemia and fluid resuscitation.

s have investigated the effects of APC on renal oxygenation and function in terms of tubular sodium reabsorption. In the present study, we tested whether continuous recombinant human APC administration would be able to protect renal oxygenation and function during the acute phase of endotoxemia and fluid resuscitation. Materials and methods All experiments in this study were approved by the institutional Animal Experimentation Committee of the Academic Medical Center of the University of Amsterdam (institutional protocol number: DFL 100404). Care and handling of the animals were in accordance with the guidelines for Institutional and Animal Care and Use Committees. The experiments were performed on 32 Wistar male rats (Harlan Laboratories, Inc., Boxmeer, The Netherlands) with mean ± SD body weight of 318 ± 15 g. Surgical preparation The rats were anesthetized with an intraperitoneal injection of a mixture of 100 mg/kg ketamine (Nimatek®, Eurovet, Bladel, The Netherlands), 0.5 mg/kg medetomidine (Domitor, Pfizer Inc., New York, NY, USA), and 0.05 mg/kg atropine sulfate (Centrafarm Pharmaceuticals B.V., Etten-Leur, The Netherlands). After tracheotomy, the animals were mechanically ventilated with a FiO2 of 0.4. Body temperature was maintained at 37°C ± 0.5°C during the entire experiment by external warming. The ventilator settings were adjusted to maintain end-tidal PCO2 between 30 and 35 mmHg, and arterial PCO2 between 35 and 40 mmHg.

therlands). After tracheotomy, the animals were mechanically ventilated with a FiO2 of 0.4. Body temperature was maintained at 37°C ± 0.5°C during the entire experiment by external warming. The ventilator settings were adjusted to maintain end-tidal PCO2 between 30 and 35 mmHg, and arterial PCO2 between 35 and 40 mmHg. The vessels were cannulated with polyethylene catheters (outer diameter = 0.9 mm; B. Braun Melsungen AG, Melsungen, Germany) for drug and fluid administration and hemodynamic monitoring. A catheter in the right carotid artery was connected to a pressure transducer to monitor the mean arterial blood pressure (MAP) and heart rate. The right femoral artery was cannulated for blood sampling. The right femoral vein was cannulated for continuous infusion of Ringer’s lactate (15 mL/kg/h; Baxter B.V., Utrecht, The Netherlands) and ketamine (50 mg/kg/h; Nimatek®; Eurovet). The left kidney was exposed, decapsulated, and immobilized in a Lucite kidney cup (K. Effenberger, Pfaffingen, Germany) via a 4-cm incision in the left flank. The renal vessels were carefully separated under preservation of nerves and adrenal gland. A perivascular ultrasonic transient time flow probe was placed around the left renal artery (type 0.7 RB; Transonic Systems Inc., Ithaca, NY, USA) and connected to a flow meter (T206; Transonic Systems Inc.) to continuously measure the renal blood flow (RBF). An estimation of the renal vascular resistance (RVR) was made as RVR (dynes s−1 cm−5) = (MAP / RBF). The left ureter was isolated, ligated, and cannulated with a polyethylene catheter for urine collection.

) and connected to a flow meter (T206; Transonic Systems Inc.) to continuously measure the renal blood flow (RBF). An estimation of the renal vascular resistance (RVR) was made as RVR (dynes s−1 cm−5) = (MAP / RBF). The left ureter was isolated, ligated, and cannulated with a polyethylene catheter for urine collection. After the surgical protocol (approximately 60 min), one optical fiber was placed 1 mm above the decapsulated kidney and another optical fiber 1 mm above the renal vein to measure oxygenation in the renal microvasculature and renal vein, respectively, using phosphorimetry [22–24]. A small piece of aluminum foil was placed on the dorsal site of the renal vein to prevent contribution of underlying tissue to the phosphorescence signal in the venous oxygenation measurement. Oxyphor G2 (a two-layer glutamate dendrimer of tetra-(4-carboxy-phenyl) benzoporphyrin; Oxygen Enterprises Ltd., Philadelphia, PA, USA) was subsequently infused (6 mg/kg IV over 5 min) followed by a 30-min stabilization period. A short description of phosphorimetry is given below, and a more detailed description of the technology has been provided elsewhere [22, 23].

tra-(4-carboxy-phenyl) benzoporphyrin; Oxygen Enterprises Ltd., Philadelphia, PA, USA) was subsequently infused (6 mg/kg IV over 5 min) followed by a 30-min stabilization period. A short description of phosphorimetry is given below, and a more detailed description of the technology has been provided elsewhere [22, 23]. Experimental protocol After baseline measurements were performed 30 min after Oxyphor G2 infusion, endotoxemic shock was induced in three groups of rats (n = 8/group) by a bolus of lipopolysaccharide (LPS, 10 mg/kg, serotype 0127:B8, Sigma-Aldrich, Zwijndrecht, The Netherlands). One hour after the LPS bolus, fluid resuscitation (5 mL/kg followed by 5 mL/kg/h; Voluven®, 6% HES 130/0.4; Fresenius Kabi, Schelle, Belgium) was started and continued for 2 h. In addition to the fluid resuscitation, one group received 10 μg/kg/h APC and one group received 100 μg/kg/h APC (recombinant human activated protein C; Drotrecogin Alpha, Xigris®, Eli Lilly and Company, Indianapolis, IN, USA). A fourth group of rats (n = 8) did not receive LPS or fluid resuscitation and served as a sham-operated time control group. All the rats, except for those in the time control group, received the same fluid volume. The experiments were terminated by infusion of 1 mL of 3 M potassium chloride (KCl), after which the kidneys were removed and weighted.

) did not receive LPS or fluid resuscitation and served as a sham-operated time control group. All the rats, except for those in the time control group, received the same fluid volume. The experiments were terminated by infusion of 1 mL of 3 M potassium chloride (KCl), after which the kidneys were removed and weighted. Blood variables Arterial blood samples (0.5 mL) were taken from the femoral artery at the following time points: (1) baseline; (2) 1 h after the LPS bolus, before the start of fluid resuscitation; and (3) after 2 h of fluid resuscitation. The blood samples were replaced by the same volume of Voluven®. The samples were analyzed for blood gas values (ABL505 blood gas analyzer; Radiometer Medical ApS, Copenhagen, Denmark), hemoglobin concentration, and hemoglobin oxygen saturation (OSM3; Radiometer Medical ApS). Additionally, plasma creatinine concentrations were determined in all the samples.

same volume of Voluven®. The samples were analyzed for blood gas values (ABL505 blood gas analyzer; Radiometer Medical ApS, Copenhagen, Denmark), hemoglobin concentration, and hemoglobin oxygen saturation (OSM3; Radiometer Medical ApS). Additionally, plasma creatinine concentrations were determined in all the samples. Renal microvascular and venous oxygenation Microvascular oxygen tension in the renal cortex (CμPO2), outer medulla (MμPO2), and renal venous oxygen tension (PrvO2) were measured by oxygen-dependent quenching of phosphorescence lifetimes of the systemically infused albumin-targeted (and therefore circulation-confined) phosphorescent dye Oxyphor G2 [24]. Oxyphor G2 (a two-layer glutamate dendrimer of tetra-(4-carboxy-phenyl) benzoporphyrin) has two excitation peaks (λexcitation1 = 440 nm, λexcitation2 = 632 nm) and one emission peak (λemission = 800 nm). These optical properties allow (near) simultaneous lifetime measurements in microcirculation of the kidney cortex and the outer medulla due to different optical penetration depths of the excitation light [25]. For the measurement of renal venous PO2 (PrvO2), a mono-wavelength phosphorimeter was used [26]. Oxygen measurements based on phosphorescence lifetime techniques rely on the principle that phosphorescence can be quenched by energy transfer to oxygen, resulting in shortening of the phosphorescence lifetime. A linear relationship between reciprocal phosphorescence lifetime and oxygen tension (given by the Stern-Volmer relation) allows quantitative measurement of PO2 [27].

hniques rely on the principle that phosphorescence can be quenched by energy transfer to oxygen, resulting in shortening of the phosphorescence lifetime. A linear relationship between reciprocal phosphorescence lifetime and oxygen tension (given by the Stern-Volmer relation) allows quantitative measurement of PO2 [27]. Renal oxygen delivery and consumption Arterial oxygen content (AOC) was calculated by (1.31 × hemoglobin × SaO2) + (0.003 × PaO2), where SaO2 is the arterial oxygen saturation, and PaO2 is the arterial partial pressure of oxygen. Renal venous oxygen content (RVOC) was calculated as (1.31 × hemoglobin × SrvO2) + (0.003 × PrvO2), where SrvO2 is the venous oxygen saturation, and PrvO2 is the renal vein partial pressure of oxygen (measured using phosphorimetry). Renal oxygen delivery per gram of renal tissue was calculated as DO2 (mL/min/g) = RBF × AOC. Renal oxygen consumption per gram of renal tissue was calculated as VO2 (mL/min/g) = RBF × (AOC − RVOC). The renal oxygen extraction ratio was calculated as O2ER (%) = VO2/DO2 × 100.

e of oxygen (measured using phosphorimetry). Renal oxygen delivery per gram of renal tissue was calculated as DO2 (mL/min/g) = RBF × AOC. Renal oxygen consumption per gram of renal tissue was calculated as VO2 (mL/min/g) = RBF × (AOC − RVOC). The renal oxygen extraction ratio was calculated as O2ER (%) = VO2/DO2 × 100. Renal function For the analysis of urine volume, creatinine concentration, and sodium (Na+) concentration at the end of the protocol, urine samples from the left ureter were collected for 10 min. Creatinine clearance rate (CLcrea) per gram of renal tissue was calculated with standard formula: CLcrea (mL/min/g) = (U × V) / P, where U is the urine creatinine concentration, V is the urine volume per unit time, and P is the plasma creatinine concentration. Renal sodium reabsorption (TNa+, (mmol/min)) was calculated as TNa+ = (PNa+ × CCR) − (UNa+ × V), where UNa+ is the urine sodium concentration, and PNa+ is the plasma sodium concentration. The renal oxygen consumption efficiency for sodium transport (VO2/TNa+) was assessed as the ratio of the renal VO2 over the total amount of sodium reabsorbed (TNa+, (mmol/min)). Data analysis Statistical analysis was performed using GraphPad Prism version 5.0 for Windows (GraphPad Software Inc., San Diego, CA, USA). Data are presented as median (25% to 75% percentiles). The statistical significance of differences between groups was tested using two-way ANOVA with Bonferroni post hoc tests. P values < 0.05 were considered significant.

rformed using GraphPad Prism version 5.0 for Windows (GraphPad Software Inc., San Diego, CA, USA). Data are presented as median (25% to 75% percentiles). The statistical significance of differences between groups was tested using two-way ANOVA with Bonferroni post hoc tests. P values < 0.05 were considered significant. Results Table 1 shows the systemic and renal hemodynamic variables: mean arterial pressure (MAP), renal blood flow (RBF), renal vascular resistance (RVR), arterial hemoglobin level (Hb), renal oxygen delivery (DO2), renal oxygen consumption (VO2), and microvascular oxygen tensions in the renal cortex (CμPO2) and medulla (MμPO2) at baseline (BL); 1 h after the LPS bolus, before the start of fluid resuscitation (LPS); and after 2 h of fluid resuscitation (FR). Figure 1 shows the renal DO2 and VO2 and TNa+, renal oxygen handling efficacy (VO2/TNa+), and creatinine clearance rate at the end of the protocol. No differences in any of these variables between groups were present at baseline.Figure 1 DO 2 , VO 2 , T Na+ , VO 2 / T Na+ , and creatinine clearance rate at the end of protocol. Data are presented as Whisker boxes and range. T p < 0.05 vs time control; F p < 0.05 vs LPS + FR. (A) Renal oxygen delivery and consumption, (B) renal sodium reabsorption, (C) renal oxygen handling, and (D) renal creatinine clearance rate. Table 1 Systemic and renal hemodynamic variables in the renal cortex and medulla at three sampling points

Results Table 1 shows the systemic and renal hemodynamic variables: mean arterial pressure (MAP), renal blood flow (RBF), renal vascular resistance (RVR), arterial hemoglobin level (Hb), renal oxygen delivery (DO2), renal oxygen consumption (VO2), and microvascular oxygen tensions in the renal cortex (CμPO2) and medulla (MμPO2) at baseline (BL); 1 h after the LPS bolus, before the start of fluid resuscitation (LPS); and after 2 h of fluid resuscitation (FR). Figure 1 shows the renal DO2 and VO2 and TNa+, renal oxygen handling efficacy (VO2/TNa+), and creatinine clearance rate at the end of the protocol. No differences in any of these variables between groups were present at baseline.Figure 1 DO 2 , VO 2 , T Na+ , VO 2 / T Na+ , and creatinine clearance rate at the end of protocol. Data are presented as Whisker boxes and range. T p < 0.05 vs time control; F p < 0.05 vs LPS + FR. (A) Renal oxygen delivery and consumption, (B) renal sodium reabsorption, (C) renal oxygen handling, and (D) renal creatinine clearance rate. Table 1 Systemic and renal hemodynamic variables in the renal cortex and medulla at three sampling points BL (t= 0 min) LPS (t= 60 min) FR (t= 180 min) MAP (mmHg) Time control 99 (97–103) 99 (94–99) 90 (85–94) LPS + FR 102 (96–109) 77 (73–93) T 68 (44–79) T APC10 100 (99–102) 76 (69–84) T 72 (59–78) T APC100 102 (102–103) 78 (76–92) T 81 (70–89) T RBF (mL/min) Time control 6.2 (5.8-6.3) 5.9 (4.7-6.1) 5.1 (4.7-5.8) LPS + FR 6.8 (6.0-6.8) 3.1 (3.1-3.3) T 4.6 (2.2-5.1) T APC10 6.6 (5.0-7.1) 3.2 (1.7-4.5) T 4.1 (2.8-5.1) T APC100 6.0 (5.9-6.8) 3.2 (2.3-4.2) T 4.2 (2.8-4.7) T

BL (t= 0 min) LPS (t= 60 min) FR (t= 180 min) MAP (mmHg) Time control 99 (97–103) 99 (94–99) 90 (85–94) LPS + FR 102 (96–109) 77 (73–93) T 68 (44–79) T APC10 100 (99–102) 76 (69–84) T 72 (59–78) T APC100 102 (102–103) 78 (76–92) T 81 (70–89) T RBF (mL/min) Time control 6.2 (5.8-6.3) 5.9 (4.7-6.1) 5.1 (4.7-5.8) LPS + FR 6.8 (6.0-6.8) 3.1 (3.1-3.3) T 4.6 (2.2-5.1) T APC10 6.6 (5.0-7.1) 3.2 (1.7-4.5) T 4.1 (2.8-5.1) T APC100 6.0 (5.9-6.8) 3.2 (2.3-4.2) T 4.2 (2.8-4.7) T RVR (dyn s−1 cm−5) Time control 1,319 (1,251-1,373) 1,347 (1,301-1,579) 1,332 (1,280-1,558) LPS + FR 1,276 (1,183-1,407) 1,901 (1,620-2,673) T 1,311 (828–2,314) APC10 1,267 (1,124-1,635) 2,086 (1,353-3,222) T 1,371 (938–2,038) APC100 1,331 (1,203-1,407) 2,024 (1,761-2,743) T 1,636 (1,383-1,864) Hb (g/dL) Time control 0.19 (0.18-0.20) 0.18 (0.17-0.20) 0.18 (0.15-0.19) LPS + FR 0.19 (0.17-0.20) 0.17 (0.17-0.18) 0.13 (0.12-0.14) T APC10 0.19 (0.18-0.20) 0.19 (0.17-0.20) 0.13 (0.12-0.14) T APC100 0.19 (0.18-0.19) 0.18 (0.17-0.19) 0.13 (0.11-0.14) T DO2 (mL O2/min/g) Time control 0.90 (0.82-0.95) 0.80 (0.60-0.93) 0.69 (0.56-0.78) LPS + FR 0.87 (0.77-0.96) 0.39 (0.37-0.43) T 0.46 (0.22-0.48) T APC10 0.89 (0.81-0.99) 0.41 (0.26-0.63) T 0.40 (0.25-0.52) T APC100 0.85 (0.80-0.97) 0.45 (0.28-0.61) T 0.37 (0.29-0.49) T

APC100 0.19 (0.18-0.19) 0.18 (0.17-0.19) 0.13 (0.11-0.14) T DO2 (mL O2/min/g) Time control 0.90 (0.82-0.95) 0.80 (0.60-0.93) 0.69 (0.56-0.78) LPS + FR 0.87 (0.77-0.96) 0.39 (0.37-0.43) T 0.46 (0.22-0.48) T APC10 0.89 (0.81-0.99) 0.41 (0.26-0.63) T 0.40 (0.25-0.52) T APC100 0.85 (0.80-0.97) 0.45 (0.28-0.61) T 0.37 (0.29-0.49) T VO2 (mL O2/min/g) Time control 0.16 (0.12-0.20) 0.18 (0.08-0.26) 0.18 (0.16-0.19) LPS + FR 0.13 (0.12-0.17) 0.11 (0.08-0.15) 0.17 (0.11-0.18) APC10 0.18 (0.12-0.19) 0.13 (0.07-0.16) 0.14 (0.11-0.20) APC100 0.17 (0.13-0.24) 0.12 (0.10-0.17) 0.16 (0.14-0.21) CμPO2 (mmHg) Time control 86 (82–87) 85 (78–92) 71 (66–79) LPS + FR 83 (77–87) 69 (65–77) T 49 (43–51) T APC10 82 (81–87) 75 (65–83) T 56 (49–59) T APC100 80 (75–88) 67 (64–75) T 68 (54–71) MμPO2 (mmHg) Time control 65 (59–66) 62 (57–67) 58 (53–60) LPS + FR 55 (51–66) 55 (50–57) T 40 (35–44) T APC10 62 (55–64) 53 (47–60) T 45 (39–49) T APC100 59 (55–69) 55 (49–59) T 55 (49–58) Data are presented as median (25% to 75% percentiles). MAP, mean arterial pressure; RBF, renal blood flow; RVR, renal vascular resistance; Hb, arterial hemoglobin level; DO2, renal oxygen delivery; VO2, renal oxygen consumption; CμPO2, microvascular oxygen tension in the renal cortex; microvascular oxygen tension in the medulla (MμPO2); BL, baseline (BL); LPS, lipopolysaccharide; FR, fluid resuscitation. T p < 0.05 vs time control.

al vascular resistance; Hb, arterial hemoglobin level; DO2, renal oxygen delivery; VO2, renal oxygen consumption; CμPO2, microvascular oxygen tension in the renal cortex; microvascular oxygen tension in the medulla (MμPO2); BL, baseline (BL); LPS, lipopolysaccharide; FR, fluid resuscitation. T p < 0.05 vs time control. Systemic and renal hemodynamic variables The bolus of LPS (10 mg/kg) induced a significant drop in MAP and RBF, and a rise in RVR. One hour after LPS administration, all the rats received the same amount of Voluven® during the resuscitation protocol, i.e., a bolus of 5 mL/kg followed by 5 mL/kg/h for 2 h. Fluid resuscitation could not improve MAP, RBF, and RVR back to baseline level. No additional effects of APC supplementation (10 or 100 μg/kg/h) on systemic and renal hemodynamic variables were observed.

e same amount of Voluven® during the resuscitation protocol, i.e., a bolus of 5 mL/kg followed by 5 mL/kg/h for 2 h. Fluid resuscitation could not improve MAP, RBF, and RVR back to baseline level. No additional effects of APC supplementation (10 or 100 μg/kg/h) on systemic and renal hemodynamic variables were observed. Renal oxygenation variables In line with RBF, renal DO2 decreased after the LPS bolus in all groups. Fluid resuscitation led to a reduction in the arterial hemoglobin concentration, and therefore, in contrast to RBF, renal DO2 did not improve due to the hemodilution. At the end of the protocol, despite differences in MAP and RBF between the groups, there were no significant differences in DO2. The reduced renal DO2 was also reflected by the reduced microvascular oxygenation in the renal cortex and medulla in all groups. In the APC100 group, the cortical and medullar PO2 was slightly higher than those in other experimental groups; however, this difference was not statistically significant. Even though renal DO2 was decreased after LPS and fluid administration, renal VO2 was maintained throughout the entire protocol. No additional effects of APC supplementation on renal DO2 and VO2 were observed.

ly higher than those in other experimental groups; however, this difference was not statistically significant. Even though renal DO2 was decreased after LPS and fluid administration, renal VO2 was maintained throughout the entire protocol. No additional effects of APC supplementation on renal DO2 and VO2 were observed. Renal function parameters At the end of the protocol, creatinine clearance rate and sodium reabsorption had decreased in all endotoxemic groups. Supplementation with APC improved both parameters, albeit not to baseline level. Renal oxygen handling efficacy, as expressed as the amount of oxygen consumed by the kidney per sodium reabsorbed (VO2/TNa+), was increased fourfold in the endotoxemic group receiving only fluids without APC and less than two-old in the groups receiving APC. Figure 2 shows that renal sodium reabsorption was closely correlated to renal microvascular oxygenation during endotoxemia and resuscitation, and APC supplementation partially protects both renal parameters.Figure 2 Renal sodium reabsorption (T Na+ ) versus the microvascular oxygenation. In the renal cortex (A) and medulla (B) at the end of the protocol in all groups.

closely correlated to renal microvascular oxygenation during endotoxemia and resuscitation, and APC supplementation partially protects both renal parameters.Figure 2 Renal sodium reabsorption (T Na+ ) versus the microvascular oxygenation. In the renal cortex (A) and medulla (B) at the end of the protocol in all groups. Discussion It has been shown in rat models of experimentally induced sepsis that APC has protective effects on the kidney [16–19]. These studies have provided important insights into both the pathophysiology of septic AKI and the potential role of APC in its prevention and/or treatment. In the present study, we aimed to test whether continuous recombinant human APC administration would be able to protect renal oxygenation and function during the acute phase of lipopolysaccharide-induced endotoxemia and fluid resuscitation. Although APC has been withdrawn from the market, investigating its effects in studies like this is still very relevant as it might provide more insight into the development of septic AKI.

to protect renal oxygenation and function during the acute phase of lipopolysaccharide-induced endotoxemia and fluid resuscitation. Although APC has been withdrawn from the market, investigating its effects in studies like this is still very relevant as it might provide more insight into the development of septic AKI. In our model, endotoxemia and fluid resuscitation led to progressive renal vasoconstriction (increased RVR and decreased RBF) and a decrease in renal DO2 and microvascular oxygenation, a fall in glomerular filtration rate (decreased creatinine clearance), and a fourfold rise in the amount of oxygen consumed by the kidney per sodium reabsorbed (VO2/TNa+). Our main findings regarding the effects of APC were that APC did not have significant effects on the systemic and renal hemodynamic and oxygenation variables or creatinine clearance. In the fluid resuscitation group, MAP was slightly lower than those in the other experimental groups, however not significant. Despite lower MAP values, RBF was higher than those in the APC10 and APC100 groups. There were no statistical differences in DO2 between groups at the end of the protocol. Particularly, in the APC100 group, both renal cortical and medullary microvascular oxygenation were better preserved than in fluid resuscitation alone. However, this difference was not statistically significant. In addition, sodium reabsorption and oxygen consumption per sodium reabsorbed (VO2/TNa+) were preserved in the APC10 and APC100 groups as compared to fluid resuscitation alone. Renal sodium reabsorption was closely correlated to renal microvascular oxygenation during endotoxemia and resuscitation, suggesting a better renal oxygen handling and less renal damage. Nevertheless, there was no assessment of cellular hypoxia or damage to confirm the suggestion above.

ed to fluid resuscitation alone. Renal sodium reabsorption was closely correlated to renal microvascular oxygenation during endotoxemia and resuscitation, suggesting a better renal oxygen handling and less renal damage. Nevertheless, there was no assessment of cellular hypoxia or damage to confirm the suggestion above. Although the mechanisms underlying acute renal failure are not completely defined, in the early stage of sepsis, impairment of the renal microcirculation is believed to be a key complication potentially leading to renal failure [25, 28–30]. In addition to an imbalance between physiological vasoactive compounds, it has been suggested that hypoxic microvascular areas might arise in the renal cortex in untreated endotoxemia [5]. These hypoxic areas are considered to reflect shunting of weak microcirculatory units [31, 32]. In the present study, 100 μg/kg APC-treated rats had less reduction of microvascular oxygenation than that observed in the other groups. As mentioned above, this was not statistically significant due to small-sized groups. Also, MAP was slightly higher in APC-treated rats, compared to that with fluid resuscitation alone. It might be suggested that the changes in the renal microvascular oxygenation are related to those in systemic blood pressure. However, RBF and DO2 were not influenced by the mild changes in systemic blood pressure.

ups. Also, MAP was slightly higher in APC-treated rats, compared to that with fluid resuscitation alone. It might be suggested that the changes in the renal microvascular oxygenation are related to those in systemic blood pressure. However, RBF and DO2 were not influenced by the mild changes in systemic blood pressure. Furthermore, APC treatment did not affect renal DO2 or VO2. This could also be explained by reduced microvascular shunting in the APC100 group: when blood is shunted from the microcirculation, the remaining microcirculatory blood would deoxygenate more rapidly, while venous oxygenation would be maintained as this is mixed microcirculatory blood and shunted blood from the arterial side. It has been shown that serine protease protein C plays an important role in controlling thrombosis and inflammation and that it exhibits cytoprotective properties [26]. There is evidence that in septic patients, a reduced plasma level of protein C is prognostic for clinical outcome [33]. With respect to the protective effects of APC on the kidney during sepsis, especially, the group of Gupta et al. has identified specific mechanisms of action of APC in rat models of experimentally induced septic AKI [16, 19]. In these studies, rats simultaneously received LPS and 10, 30, or 100 μg/kg APC. In the present study, the rats first received an LPS bolus, and fluid resuscitation was started 1 h later with 0, 10, or 100 μg/kg/h APC. We did not investigate the direct acting mechanisms of APC in our present study but merely focused on the acute effects of APC-supplemented fluid resuscitation in the kidneys of endotoxemic rats. Mimicking the clinical use of APC, at a rate of 24 μg/kg/h for 96 h was neither practical nor was our aim in this study.

PC. We did not investigate the direct acting mechanisms of APC in our present study but merely focused on the acute effects of APC-supplemented fluid resuscitation in the kidneys of endotoxemic rats. Mimicking the clinical use of APC, at a rate of 24 μg/kg/h for 96 h was neither practical nor was our aim in this study. The increase in VO2/TNa+ following LPS and fluid resuscitation without APC could either indicate less efficient oxygen use for ATP production for Na+ reabsorption or that oxygen is used for other purposes than ATP production such as ROS generation. In a recently published review [34], we described that ischemia/reperfusion injury, also arises during hypotensive (septic) shock and resuscitation, is associated with intrarenal microcirculatory dysfunction caused by an imbalance between vasoconstrictors and vasodilators, endothelial damage and endothelium-leukocyte interactions [35–37], oxidative stress [38, 39], and oxygen handling [40–42]. Alterations in oxygen transport pathways can result in cellular hypoxia and/or dysoxia [33]. This condition is associated with mitochondrial failure and/or activation of alternative pathways for oxygen consumption [34, 43]. This could explain the observed rise in VO2/TNa+ here.

ss [38, 39], and oxygen handling [40–42]. Alterations in oxygen transport pathways can result in cellular hypoxia and/or dysoxia [33]. This condition is associated with mitochondrial failure and/or activation of alternative pathways for oxygen consumption [34, 43]. This could explain the observed rise in VO2/TNa+ here. Another explanation for the observed increase of VO2/TNa+ is the back leak phenomenon. Renal I/R injury is shown to cause derangements of the tubule cell cytoskeleton, altered integrity of tight junctions between cells and loss of epithelial polarity, ultimately providing a pathway for back leak of filtrate. These impairments are suggested to cause uncoupling of renal sodium transport and oxygen consumption, leading to inefficient sodium reabsorption. In this view, the above described results of this study might also be explained by prevention of I/R-induced renal cell injury in APC-treated rats and protective effects of APC on cellular integrity and tight junction structure [44, 45].

dium transport and oxygen consumption, leading to inefficient sodium reabsorption. In this view, the above described results of this study might also be explained by prevention of I/R-induced renal cell injury in APC-treated rats and protective effects of APC on cellular integrity and tight junction structure [44, 45]. There is some evidence that traditional HES solutions can impair renal function and should be used with caution in patients with renal insufficiency [46]. In contrast, the latest generation of HES with low molecular weight and low degree of substitution (such as HES 130/0.4) is suggested to have minimal influence on renal function and coagulation. In 2006, Johannes et al., using phosphorescence lifetime technique, studied the influence of fluid resuscitation and fluid of choice on renal microvascular oxygenation in a similar model of endotoxemia [47]. HES 130/0.4 had least influence on renal VO2 and restored renal function. In this view, we have decided to choose HES 130/0.4 for fluid resuscitation during our study.

e technique, studied the influence of fluid resuscitation and fluid of choice on renal microvascular oxygenation in a similar model of endotoxemia [47]. HES 130/0.4 had least influence on renal VO2 and restored renal function. In this view, we have decided to choose HES 130/0.4 for fluid resuscitation during our study. We are aware that our study suffers from some limitations. First, no markers of systemic or renal inflammation or coagulation disorders or oxidative stress were measured. Instead, we merely focused on the acute physiological effects of APC-supplemented fluid resuscitation on renal oxygenation, oxygen handling, and function. Second, endotoxemic models may not reflect all the situations encountered in human sepsis and may lack relevance in gram-positive sepsis. However, it is a reproducible model of acute inflammation that involves similar pathways and thus allows us to study the pathophysiology and potential treatment of endotoxemia-induced AKI. Extrapolation of this model to clinical scenarios in terms of treatment strategies should be made with utmost caution. Instead, our study should be regarded as adding to our understanding of the factors contributing to renal microcirculatory failure and potential treatment strategies. Third, the present study only allows assessment of the acute effects of LPS, fluid resuscitation, and APC supplementation in this short-term rat model. APC was infused for 2 h, and the experiments only lasted 3 h in total which may be insufficient to see any hemodynamic effect. Fourth, the only drug specifically approved for sepsis (recombinant human APC) has been withdrawn from the market. However, investigating its effects in experimental studies is still very relevant as it potentially provides new insights into the development and treatment of septic AKI.

nt to see any hemodynamic effect. Fourth, the only drug specifically approved for sepsis (recombinant human APC) has been withdrawn from the market. However, investigating its effects in experimental studies is still very relevant as it potentially provides new insights into the development and treatment of septic AKI. Conclusions In conclusion, our data suggest that renal sodium reabsorption is closely correlated to renal microvascular oxygenation during experimentally induced endotoxemia. APC supplementation to standard resuscitation protocol partially protected both renal parameters. The specific mechanisms responsible for these protective effects of APC warrant further study. Authors’ original submitted files for images Below are the links to the authors’ original submitted files for images.Authors’ original file for figure 1 Authors’ original file for figure 2 Emre Almac, Tanja Johannes contributed equally to this work. Competing interests All authors declare that they have no competing interests. Authors’ contributions EA, TJ, EGM, and CI participated in the research design. EA conducted the experiments. EA, TJ, and RB performed data analysis. EA, TJ, RB, ABJG, and CI contributed to the writing of the manuscript. RB, KEU, and CI provided supervision. All authors read and approved the final manuscript. Acknowledgements This study was partially supported by a grant from Eli Lilly and Co. as part of an investigator-initiated investigation. Eli Lilly and Co. was not involved in the design, analysis, or interpretation of the study or in the decision to publish the data.

Background Ample evidence is available that the activation of the peroxisome proliferator-activated receptors (PPAR), ligand-activated transcription factors of the nuclear hormone receptor family, presenting as PPAR-α, PPAR-γ, and PPAR-β/δ, has beneficial effects in various shock models. The highly selective synthetic PPAR-β/δ agonist GW0742 [1] blunted shock-induced organ injury as a result of attenuated inflammation and oxidative and nitrosative stress and decreased activation of the nuclear transcription factor κB (NF-κB) [2–8]. These organ-protective properties were also present in animals with obesity [9] and diabetes [10], most likely as a result of enhanced insulin sensitivity and, consequently, improved glucose utilization [11], as well as attenuated endothelial dysfunction [12]. However, all these data originate from short-term, un-resuscitated rodent models characterized by hypotension and low cardiac output. Therefore, we tested the hypothesis whether GW0742 may attenuate kidney dysfunction during long-term, resuscitated, porcine fecal peritonitis-induced septic shock [13, 14]. Given the beneficial effects of GW0742 on glucose homoeostasis [11] and vascular function [12], we investigated swine with hyperlipidemia and ubiquitous atherosclerosis [15].

pothesis whether GW0742 may attenuate kidney dysfunction during long-term, resuscitated, porcine fecal peritonitis-induced septic shock [13, 14]. Given the beneficial effects of GW0742 on glucose homoeostasis [11] and vascular function [12], we investigated swine with hyperlipidemia and ubiquitous atherosclerosis [15]. Methods The University of Ulm Animal Care Committee and the Federal authorities for animal research had approved the experiments, which were performed in adherence to National Institutes of Health Guidelines on the Use of Laboratory Animals. Twenty-two adult, castrated, male pigs (age 15 to 30 months, median (interquartile range) body weight of 72 (65 to 81) kg) were used. The pig strain is a cross-breed of Rapacz farm pigs homozygous for the R84C low-density lipoprotein (LDL) receptor mutation with smaller strains ('FBM’), with hypercholesteremia due to an atherogenic diet [15].

ted, male pigs (age 15 to 30 months, median (interquartile range) body weight of 72 (65 to 81) kg) were used. The pig strain is a cross-breed of Rapacz farm pigs homozygous for the R84C low-density lipoprotein (LDL) receptor mutation with smaller strains ('FBM’), with hypercholesteremia due to an atherogenic diet [15]. Animal preparation Anesthesia and surgical instrumentation have been described in detail previously [13–15]. Briefly, anesthesia was induced with atropine, propofol, and ketamine to allow endotracheal intubation and was maintained thereafter with pentobarbitone, buprenorphine, and pancuronium. Ventilator settings were fraction of inspired O2 (FiO2) 0.35, positive end-expiratory pressure (PEEP) 10 cmH2O, tidal volume 8 mL·kg-1, respiratory rate 10 to 12 breaths·min-1 adjusted to maintain arterial PCO2 = 35 to 40 mmHg, inspiratory (I)/expiratory (E) ratio 1:1.5, peak airway pressure <40 cmH2O, and modified to I/E ratio 1:1 and PEEP 12 or 15 cmH2O, respectively, if the ratio of arterial O2 partial pressure (PaO2)/FiO2 is <300 or <200 mmHg [13–15]. The right jugular vein and carotid artery were exposed for the insertion of a central venous catheter sheath and the placement of a balloon-tipped pulmonary artery catheter to measure central venous (CVP), pulmonary arterial (MPAP), and pulmonary artery occlusion pressures (PAOP), and a thermistor-tipped arterial catheter for blood pressure (MAP) recording and transpulmonary single indicator thermodilution-cardiac output (CO) measurement. The right kidney and a femoral vein were surgically exposed, and a catheter was advanced into the inferior vena cava and manually guided into a right renal vein under visual control [14, 15]. A catheter in the urinary bladder allowed urine collection. Two tubes were placed through the abdominal wall for peritonitis induction. Ringer's solution was continuously infused as maintenance fluid (10 mL·kg-1·h-1). As needed, animals received hydroxyethyl starch to maintain cardiac filling pressures during surgery.

A catheter in the urinary bladder allowed urine collection. Two tubes were placed through the abdominal wall for peritonitis induction. Ringer's solution was continuously infused as maintenance fluid (10 mL·kg-1·h-1). As needed, animals received hydroxyethyl starch to maintain cardiac filling pressures during surgery. Experimental protocol After instrumentation and an 8-hour recovery, baseline data were collected. Thereafter, the supernatant (3 mL·kg-1) of 1.0 g·kg-1 autologous feces incubated in 500 mL 0.9% saline for 12 h at 38°C was injected into the abdominal cavity via the abdominal drainage tubes. Hydroxyethyl starch (10 mL∙kg-1∙h-1, 5 mL∙kg-1∙h-1 if CVP or PAOP is >18 mmHg) allowed maintaining hyperdynamic hemodynamics. If necessary, norepinephrine was infused and titrated to maintain MAP at baseline values (no further increase if the heart rate is ≥160 min-1 to avoid tachycardia-induced myocardial ischemia) [13–16]. Animals randomly received i.v. GW0742 (0.03 mg∙kg-1, n = 10, body weight 71 (66 to 80) kg) or vehicle (DMSO; n = 12, body weight 75 (66 to 81) kg) at 6, 12, and 18 h after the induction of peritonitis. The same GW0742 dose attenuated renal dysfunction in murine endotoxic shock [2] and organ injury after kidney ischemia/reperfusion injury in diabetic rats [10]. The timing of the GW0742 administration was chosen, because10-day survival was doubled in mice injected with this dose at 6.5 and 12.5 h after cecal ligation and puncture-induced sepsis [2]. After additional data collection at 12 and 24 h after the induction of peritonitis, animals were sacrificed under deep anesthesia.

he timing of the GW0742 administration was chosen, because10-day survival was doubled in mice injected with this dose at 6.5 and 12.5 h after cecal ligation and puncture-induced sepsis [2]. After additional data collection at 12 and 24 h after the induction of peritonitis, animals were sacrificed under deep anesthesia. Measurements and calculations Hemodynamics, gas exchange (calorimetric O2 uptake and CO2 production, arterial and mixed venous blood gases), glucose, lactate, creatinine, renal venous nitrite + nitrate (NO2- + NO3-), tumor necrosis factor-α (TNFα), and interleukin-6 (IL-6) concentrations were determined as described previously [13–16]. Endogenous glucose production and glucose oxidation were derived from plasma 1,2,3,4,5,6-13C6-glucose and the mixed expiratory 13CO2 isotope enrichment, respectively, during continuous glucose isotope infusion [13, 14]. Urinary and blood creatinine and Na+ levels were analyzed to calculate creatinine clearance and fractional Na+-excretion [9] together with blood neutrophil gelatinase-associated lipocalin (NGAL) [15]. At the end of the experiment, immediate post-mortem kidney tissue samples were analyzed for the expression of the inducible nitric oxide synthase (iNOS), heme oxygenase-1 (HO-1), and cleaved caspase-3 as well as for the activation of the nuclear transcription factor κB (NF-κB) as described in detail previously [15, 16]. Pyramid-shaped kidney specimens showing kidney cortex, medulla, renal papilla, and the corresponding renal calyx were dissected for histopathological examination, performed by an experienced pathologist (A.S.) blinded for the sample grouping [15]. Histopathological alterations were analyzed for the degree of 'glomerular tubularization’, dilatation of Bowman's space, and swelling of Bowman's capsule, cellular edema of the proximal tubule, distal tubular dilatation and elongation, tubular protein cylinders, and tubular necrosis as described in detail previously [15].

topathological alterations were analyzed for the degree of 'glomerular tubularization’, dilatation of Bowman's space, and swelling of Bowman's capsule, cellular edema of the proximal tubule, distal tubular dilatation and elongation, tubular protein cylinders, and tubular necrosis as described in detail previously [15]. Immunohistochemistry allowed quantifying the formation of nitrotyrosine (rabbit anti-Nitrotyrosine, Millipore, Schwalbach, Germany) [16] and the expression of the PPAR-β/δ (rabbit anti-PPAR delta antibody #ab23673, Abcam plc, Cambridge, UK), the method being described in detail in the supplement. The latter was determined on formalin-fixed, paraffin-embedded kidney biopsies, which had been taken during surgical instrumentation both in FBM and young and healthy German Landswine in a previous study [15], as well as age-matched FBM pigs that had not been fed with the atherogenic diet (n = 5 in each group). Results are presented as mean densitometric sum red. Statistical analysis Data are presented as median (quartiles). After the exclusion of normal distribution using the Kolmogorov-Smirnoff test, differences within groups were analyzed by a Friedmann analysis of variance on ranks and a subsequent Dunn's test with Bonferroni correction. Inter-group differences were tested using a Mann–Whitney rank sum test.

nted as median (quartiles). After the exclusion of normal distribution using the Kolmogorov-Smirnoff test, differences within groups were analyzed by a Friedmann analysis of variance on ranks and a subsequent Dunn's test with Bonferroni correction. Inter-group differences were tested using a Mann–Whitney rank sum test. Results One animal in the control group died 15 h after the induction of peritonitis; therefore, data at the end of the experiment originate from 11 vehicle-treated animals only. Colloid and norepinephrine requirements were comparable in the two groups (Additional file 1: Table A). MAP progressively decreased and CO increased despite aggressive circulatory support, ultimately resulting in impaired pulmonary gas exchange and lactic acidosis, however, without inter-group difference (Table 1). Sepsis caused a progressive deterioration of renal function, coinciding with increased renal venous concentrations of pro-inflammatory cytokines, NO metabolites, and isoprostanes, again without any inter-group difference (Table 2). Western blotting confirmed the findings on blood biomarkers, while tissue expression of HO-1, iNOS, and activated caspase 3 were comparable to those from animals that had undergone surgical instrumentation only, sepsis nearly tripled NF-κB activation (Additional file 1: Table B). There was, however, no treatment effect.Table 1 Systemic hemodynamics, gas exchange, and metabolism Before peritonitis 12-h peritonitis 24-h peritonitis Body temperature (°C) DMSO 36.5 (35.7, 37.4) 37.6 (37.0, 38.4)a 38.4 (37.5, 38.7)a

Results One animal in the control group died 15 h after the induction of peritonitis; therefore, data at the end of the experiment originate from 11 vehicle-treated animals only. Colloid and norepinephrine requirements were comparable in the two groups (Additional file 1: Table A). MAP progressively decreased and CO increased despite aggressive circulatory support, ultimately resulting in impaired pulmonary gas exchange and lactic acidosis, however, without inter-group difference (Table 1). Sepsis caused a progressive deterioration of renal function, coinciding with increased renal venous concentrations of pro-inflammatory cytokines, NO metabolites, and isoprostanes, again without any inter-group difference (Table 2). Western blotting confirmed the findings on blood biomarkers, while tissue expression of HO-1, iNOS, and activated caspase 3 were comparable to those from animals that had undergone surgical instrumentation only, sepsis nearly tripled NF-κB activation (Additional file 1: Table B). There was, however, no treatment effect.Table 1 Systemic hemodynamics, gas exchange, and metabolism Before peritonitis 12-h peritonitis 24-h peritonitis Body temperature (°C) DMSO 36.5 (35.7, 37.4) 37.6 (37.0, 38.4)a 38.4 (37.5, 38.7)a GW0742 36.9 (36.5, 37.3) 38.1 (37.7, 38.2)a 38.0 (37.7, 38.3)a Heart rate (min-1) DMSO 83 (70, 99) 144 (121, 149)a 149 (142, 163)a GW0742 69 (63, 92) 134 (126, 146)a 158 (150, 162)a Mean arterial pressure (mmHg) DMSO 103 (92, 108) 90 (87, 101)a 75 (65, 98)a GW0742 110 (106, 114) 97 (93, 100)a 83 (75, 94)a

Before peritonitis 12-h peritonitis 24-h peritonitis Body temperature (°C) DMSO 36.5 (35.7, 37.4) 37.6 (37.0, 38.4)a 38.4 (37.5, 38.7)a GW0742 36.9 (36.5, 37.3) 38.1 (37.7, 38.2)a 38.0 (37.7, 38.3)a Heart rate (min-1) DMSO 83 (70, 99) 144 (121, 149)a 149 (142, 163)a GW0742 69 (63, 92) 134 (126, 146)a 158 (150, 162)a Mean arterial pressure (mmHg) DMSO 103 (92, 108) 90 (87, 101)a 75 (65, 98)a GW0742 110 (106, 114) 97 (93, 100)a 83 (75, 94)a Mean pulmonary artery pressure (mmHg) DMSO 25 (22, 26) 38 (28, 42)a 39 (35, 40)a GW0742 27 (24, 27) 35 (34, 38)a 42 (38, 47)a Central venous pressure (mmHg) DMSO 13 (9, 14) 15 (12, 17)a 18 (17, 18)a GW0742 13 (11, 15) 15 (13, 17)a 19 (17, 22)a Pulmonary artery occluded pressure (mmHg) DMSO 13 (10, 14) 15 (14, 17)a 17 (17, 19)a GW0742 15 (11, 17) 16 (14, 18)a 17 (16, 20)a Cardiac output (L·min-1) DMSO 4.4 (3.8, 5.2) 6.4 (5.7, 7.9)a 8.0 (5.7, 9.3)a GW0742 4.4 (3.7, 5.2) 7.1 (6.1, 8.5)a 7.6 (7.1, 7.8)a Hemoglobin (g·L-1) DMSO 94 (80, 100) 111 (99, 123)a 115 (110, 121)a GW0742 91 (80, 102) 104 (107, 119)a 110 (107, 119)a Arterial PO2 (mmHg) DMSO 176 (161, 185) 151 (121, 161) 118 (91, 144)a GW0742 180 (171, 181) 163 (125, 172) 118 (110, 139)a PaO2/FiO2 ratio (mmHg) DMSO 550 (502, 583) 459 (362, 503)a 244 (109, 398)a GW0742 536 (487, 611) 470 (392, 574)a 236 (205, 386)a Arterial PCO2 (mmHg) DMSO 36 (34, 38) 38 (37, 41) 37 (35, 40) GW0742 35 (33, 37) 38 (38, 39) 36 (34, 38) Arterial pH DMSO 7.45 (7.44, 7.46) 7.45 (7.39, 7.46) 7.35 (7.22, 7.41)a GW0742 7.45 (7.43, 7.46) 7.45 (7.40, 7.47) 7.36 (7.29, 7.41)a Arterial base excess (mmol·L-1) DMSO 1.2 (-0.3, 1.7) 1.7 (-1.1, 4.3) -4.3 (-11.7, -0.8)a

Arterial PCO2 (mmHg) DMSO 36 (34, 38) 38 (37, 41) 37 (35, 40) GW0742 35 (33, 37) 38 (38, 39) 36 (34, 38) Arterial pH DMSO 7.45 (7.44, 7.46) 7.45 (7.39, 7.46) 7.35 (7.22, 7.41)a GW0742 7.45 (7.43, 7.46) 7.45 (7.40, 7.47) 7.36 (7.29, 7.41)a Arterial base excess (mmol·L-1) DMSO 1.2 (-0.3, 1.7) 1.7 (-1.1, 4.3) -4.3 (-11.7, -0.8)a GW0742 0.4 (-0.5, 0.6) 2.5 (0.5, 4.0) -5.3 (-7.7, -1.6)a Arterial lactate (mmol·L-1) DMSO 0.8 (0.7, 1.4) 1.0 (0.8, 2.0) 4.1 (1.7, 8.3)a GW0742 1.1 (0.9, 1.3) 1.0 (.7, 1.4) 2.6 (2.4, 4.5)a O2 uptake (mL·kg-1·min-1) DMSO 2.5 (2.2, 2.9) 3.0 (2.7, 3.1) 4.1 (3.5, 4.3)a GW0742 2.6 (2.5, 2.9) 2.7 (2.3, 3.5) 3.8 (3.4, 4.6)a CO2 production (mL·kg-1·min-1) DMSO 2.4 (2.2, 2.7) 2.8 (2.5, 3.0) 3.2 (3.1, 3.6)a GW0742 2.3 (2.0, 2.6) 2.6 (2.4, 3.0) 3.2 (2.8, 3.6)a Blood glucose (mg·dL) DMSO 118 (109, 127) 76 (64, 88)a 94 (73, 120) GW0742 117 (111, 122) 68 (58, 73)a 75 (69, 83)a Glucose production (mg·kg-1·min-1] DMSO 1.3 (1.3, 1.5) 2.1 (2.0, 2.5) 3.0 (2.4, 3.3) GW0742 1.3 (1.2, 1.8) 2.3 (2.1, 2.8) 3.0 (2.8, 3.5) Glucose oxidation (mg·kg-1·min-1) DMSO 0.4 (0.3, 0.4) 1.3 (1.2, 1.4) 1.7 (1.6, 2.1) GW0742 0.3 (0.3, 0.4) 1.4 (1.3, 1.6) 1.9 (1.6, 2.3) All data are presented as median (quartiles); vehicle (DMSO): n = 12 (n = 11 during 12- to 24-h peritonitis), GW0742: n = 10; a p < 0.05 vs. before peritonitis. Table 2 Parameters of renal O 2 exchange, metabolism, function (NGAL - neutrophil gelatinase-associated lipocalin), as well as renal venous biomarkers of inflammation (NO 2 - + NO 3 - - nitrite plus nitrate) and oxidative stress (n.d. - not determined)

Glucose production (mg·kg-1·min-1] DMSO 1.3 (1.3, 1.5) 2.1 (2.0, 2.5) 3.0 (2.4, 3.3) GW0742 1.3 (1.2, 1.8) 2.3 (2.1, 2.8) 3.0 (2.8, 3.5) Glucose oxidation (mg·kg-1·min-1) DMSO 0.4 (0.3, 0.4) 1.3 (1.2, 1.4) 1.7 (1.6, 2.1) GW0742 0.3 (0.3, 0.4) 1.4 (1.3, 1.6) 1.9 (1.6, 2.3) All data are presented as median (quartiles); vehicle (DMSO): n = 12 (n = 11 during 12- to 24-h peritonitis), GW0742: n = 10; a p < 0.05 vs. before peritonitis. Table 2 Parameters of renal O 2 exchange, metabolism, function (NGAL - neutrophil gelatinase-associated lipocalin), as well as renal venous biomarkers of inflammation (NO 2 - + NO 3 - - nitrite plus nitrate) and oxidative stress (n.d. - not determined) Before peritonitis 12-h peritonitis 24-h peritonitis Renal venous PO2 (mmHg) DMSO 54 (45, 61) 60 (50, 61) 56 (47, 61) GW0742 53 (47, 55) 58 (52, 63) 56 (48, 60) Renal venous PCO2 (mmHg) DMSO 40 (38, 41) 44 (41, 45) 45 (41, 47) GW0742 38 (36, 40) 44 (42, 45) 42 (37, 46) Renal venous pH DMSO 7.42 (7.41, 7.44) 7.42 (7.36, 7.44) 7.30 (7.12, 7.39)a GW0742 7.43 (7.40, 7.45) 7.42 (7.38, 7.44) 7.31 (7.23, 7.40)a Renal venous base excess (mmol1·L-1) DMSO 1.4 (0.0, 2.5) 2.4 (-0.2, 4.7) -4.4 (-12.3, 1.5)a GW0742 1.0 (0.1, 1.7) 3.3 (1.3, 4.4) -4.7 (-8.2, -2.1)a Renal venous lactate (mmol1·L-1) DMSO 0.9 (0.7, 1.4) 1.0 (0.9, 1.9) 4.1 (1.3, 9.2)a GW0742 1.0 (0.7, 1.0) 1.3 (1.2, 1.5) 3.2 (2.1, 4.3)a Arterial creatinine (μmol1·L-1) DMSO 90 (84, 96) 81 (75, 96) 126 (98, 141)a GW0742 93 (87, 95) 81 (80, 88) 124 (94, 137)a Blood NGAL (ng·L-1) DMSO 65 (55, 72) n.d. 364 (270, 400)a GW0742 56 (53, 61) n.d. 365 (276, 400)a

Renal venous lactate (mmol1·L-1) DMSO 0.9 (0.7, 1.4) 1.0 (0.9, 1.9) 4.1 (1.3, 9.2)a GW0742 1.0 (0.7, 1.0) 1.3 (1.2, 1.5) 3.2 (2.1, 4.3)a Arterial creatinine (μmol1·L-1) DMSO 90 (84, 96) 81 (75, 96) 126 (98, 141)a GW0742 93 (87, 95) 81 (80, 88) 124 (94, 137)a Blood NGAL (ng·L-1) DMSO 65 (55, 72) n.d. 364 (270, 400)a GW0742 56 (53, 61) n.d. 365 (276, 400)a Urine output (mL·kg-1·h-1) DMSO 8.8 (7.7, 10.4) 3.2 (1.8, 5.1)a GW0742 10.0 (8.2, 10.7) 3.6 (2.4, 6.7)a Creatinine clearance (mL1·min-1) DMSO 120 (91, 134) 56 (39, 96)a GW0742 129 (114, 140) 78 (66, 81)a Fractional Na+ excretion (%) DMSO 10 (8, 11) 5 (4, 8)a GW0742 8 (7, 9) 4 (3, 4)a Renal venous interleukin-6 (ng·L-1) DMSO 96 (77, 139) 1,710 (1,242, 4,807)a 8,405 (2,565, 23,153)a GW0742 97 (77, 107) 2,656 (1,186, 2,910)a 3,457 (3,348, 3,489)a Renal venous tumor necrosis factor-α (ng·L-1) DMSO 34 (28, 52) 70 (55, 105)a 125 (91, 172)a GW0742 49 (42, 126) 63 (50, 84)a 183 (131, 239)a Renal venous NO2 - + NO3 - (μmol·L-1) DMSO 7 (5, 14) 10 (8, 15) 14 (12, 21)a GW0742 9 (6, 10) 9 (8, 14) 17 (10, 21)a Renal venous 8-isoprostane (ng·L-1) DMSO 85 (80, 131) 96 (78, 132) 123 (90, 166)a GW0742 83 (72, 109) 91 (70, 113) 123 (87, 264)a Data on urine output, creatinine clearance, and fractional Na+ excretion refer to 0 to 12 and 12 to 24 h of peritonitis, respectively. All data are presented as median (quartiles); vehicle (DMSO): n = 12 (n = 11 at 24-h peritonitis), GW0742: n = 10; a p < 0.05 vs. before peritonitis.

GW0742 83 (72, 109) 91 (70, 113) 123 (87, 264)a Data on urine output, creatinine clearance, and fractional Na+ excretion refer to 0 to 12 and 12 to 24 h of peritonitis, respectively. All data are presented as median (quartiles); vehicle (DMSO): n = 12 (n = 11 at 24-h peritonitis), GW0742: n = 10; a p < 0.05 vs. before peritonitis. Kidney histopathology showed only mild to moderate glomerular and tubular damage without inter-group difference (Additional file 1: Table C). Immunohistochemistry showed marked nitrotyrosine formation, again without inter-group difference (Additional file 1: Table B). Immunohistochemistry of the renal PPAR-β/δ showed that expression was nearly ten times lower than both in young and healthy German Landswine and age-matched FBM pigs that had not been fed with the atherogenic diet (Figure 1).Figure 1 Examples and results of the quantitative image analysis of the PPAR-β/δ immunohistochemistry in biopsies. Examples (A) (magnification ×20; left panel: German Landswine, right panel: FBM swine) and results of the quantitative image analysis (B) of the PPAR-β/δ immunohistochemistry in biopsies taken during surgical instrumentation in comparison to biopsies taken in otherwise young and healthy German Landswine undergoing a similar surgical instrumentation. German Landswine (open box-and-whisker plots), FBM swine with (dark gray box-and-whisker plots), and without hyperchloesteremic diet (light gray box-and-whisker plots). All data are median (quartiles, range; n = 5 in each group).

in animals with obesity and diabetes, we studied swine with hyperlipidemia and ubiquitous atherosclerosis [15]. The major findings were that (1) GW0742 failed to attenuate sepsis-induced organ dysfunction and histological damage and (2) did not affect the parameters of inflammation and oxidative and nitrosative stress. The lacking efficacy of GW0742 is in contrast to previous studies in polymicrobial sepsis [2, 7], which, however, report data from young and otherwise healthy rodents. We studied septic shock in FBM swine, a strain presenting with ubiquitous hypercholesteremia-induced atherosclerosis [15]. In addition, these animals showed a several-fold reduction of the kidney PPAR-β/δ expression as compared to young and healthy German Landswine. In obese, insulin-resistant mice, weight loss increased PPAR-γ expression [17], and both starvation and endurance training activated PPAR-δ in healthy animals [18, 19]. PPAR-δ knockout mice are glucose-intolerant [11], and PPAR-β/δ activation normalized the diabetes-related endothelial dysfunction [12]. Our FBM swine presented with both hypercholesteremia as well as impaired glucose homoeostasis; at baseline, direct, aerobic glucose oxidation was significantly lower (<30% vs. 60% to 75% of glucose production) when compared to young (age 3 to 4 months) and healthy German Landswine undergoing the same protocol of resuscitated fecal peritonitis [13, 14]. Hence, the pre-existing atherosclerosis together with the 'metabolic syndrome’ may have caused GW0742 inefficacy due to PPAR-β/δ down-regulation. It could be argued in this context that GW0742 did attenuate ischemia/reperfusion injury in obese and diabetic rats [9, 10]. These data, however, originate from un-resuscitated, short-term models in rats, whereas we studied long-term, fully resuscitated porcine septic shock.

inefficacy due to PPAR-β/δ down-regulation. It could be argued in this context that GW0742 did attenuate ischemia/reperfusion injury in obese and diabetic rats [9, 10]. These data, however, originate from un-resuscitated, short-term models in rats, whereas we studied long-term, fully resuscitated porcine septic shock. Our findings of reduced PPAR-β/δ expression were obtained from biopsies taken during surgical instrumentation, i.e., prior to any inflammatory challenge [15]. Since the abdominal cavity had been closed again after the surgical instrumentation, we could not obtain kidney biopsies after induction of peritonitis. Therefore, we cannot exclude that sepsis further aggravated PPAR-β/δ down-regulation and thus contributed to the lacking efficacy of GW0742. In rodents, any shock-related effect on PPAR expression seems to be organ-specific; hemorrhage, injection of endotoxin, or cecal ligation and puncture significantly reduced pulmonary [20], myocardial [21], and hepatic [22] PPAR-α, -γ, or -δ protein expression and/or mRNA. However, endotoxemia only decreased renal PPAR-α mRNA, whereas PPAR-γ and PPAR-δ mRNA remained unchanged [23].

o be organ-specific; hemorrhage, injection of endotoxin, or cecal ligation and puncture significantly reduced pulmonary [20], myocardial [21], and hepatic [22] PPAR-α, -γ, or -δ protein expression and/or mRNA. However, endotoxemia only decreased renal PPAR-α mRNA, whereas PPAR-γ and PPAR-δ mRNA remained unchanged [23]. Limitations of the study Since we did not study the effect of GW0742 in young and healthy animals, we do not have direct evidence that the PPAR-β/δ down-regulation associated with the underlying hypercholesteremia and atherosclerosis rather than other potentially confounding factors contributed to our findings, e.g., the use of adult swine per se, the integration of standard intensive care procedures into the experimental design, the timing and dosing of the treatment, and/or the duration of the study. To our knowledge, GW0742 has only been studied in rodents. Nevertheless, GW501516, another highly specific PPAR-β/δ agonist with a chemical structure very close to that of GW0742 [1], attenuated endotoxin-induced NF-κB activation, and cytokine release not only in rodent [24–26] but also in human tissues [26, 27]. So far, a modulation of the PPAR-β/δ expression related to nutritional and/or metabolic interventions has only been studied in rodents as well [18, 19]. However, in swine, a similar diet-induced hypercholesteremia as in our experiments [28] was associated with a markedly reduced PPAR-α and PPAR-γ expression [29], coinciding with impaired glucose tolerance [28, 29], arterial hypertension [28], increased transaminase activities [29], and aggravated oxidative stress [28]. Treatment with resveratrol restored PPAR-α and PPAR-γ expression [29], going along with improved left heart perfusion and function during chronic myocardial ischemia [30, 31]. It could be argued that all rodent data on GW0742 were obtained in young animals, while we studied adult swine. In fact, ischemia/reperfusion injury and hemorrhagic shock induced PPAR-γ expression only in young but not in adult or old rodents [32, 33], which in turn impaired adaptive autophagy [33]. There are no data available on PPAR-β/δ expression under these conditions, but the above-mentioned studies on hypercholesteremia-induced PPAR-α and PPAR-γ down-regulation also originate from adult swine. Moreover, tissue PPAR-β/δ expression in kidneys from adult FBM swine that had not been fed with cholesterol-enriched diet was similar to that in young and healthy German Landswine (Figure 1B).

ut the above-mentioned studies on hypercholesteremia-induced PPAR-α and PPAR-γ down-regulation also originate from adult swine. Moreover, tissue PPAR-β/δ expression in kidneys from adult FBM swine that had not been fed with cholesterol-enriched diet was similar to that in young and healthy German Landswine (Figure 1B). We only studied the effect of 0.03 mg∙kg-1 of GW0742. This dose was chosen because it attenuated kidney dysfunction in murine endotoxin- and cecal ligation and puncture-induced septic shock at plasma concentrations selectively activating the PPAR-β/δ without any cross-reactivity on the other PPAR isoforms [2]. Moreover, in rats, this dose reduced organ injury after kidney ischemia/reperfusion injury [10], and a ten times higher dose did not further influence myocardial ischemia/reperfusion injury [3]. It should be noted in this context that equipotent drug doses are usually much lower on a per kilogram basis in large species (e.g., swine, dogs) than in rodents [34]. FBM swine present with reduced creatinine clearance and moderate histological damage of the kidney already under baseline conditions [15]. Nevertheless, any pre-existing organ dysfunction most likely did not influence our results; due to fluid resuscitation and catecholamine infusion, creatinine clearance was comparable to that in young and healthy German Landswine undergoing the same protocol of resuscitated fecal peritonitis [13, 14].

e conditions [15]. Nevertheless, any pre-existing organ dysfunction most likely did not influence our results; due to fluid resuscitation and catecholamine infusion, creatinine clearance was comparable to that in young and healthy German Landswine undergoing the same protocol of resuscitated fecal peritonitis [13, 14]. Conclusions In swine with pre-existing atherosclerosis, the PPAR-β/δ agonist GW0742 failed to attenuate septic shock-induced kidney dysfunction and organ damage, most likely due to reduced receptor expression associated with cardiovascular and metabolic co-morbidity. Electronic supplementary material Additional file 1: Supplementary information. Table A: Norepinephrine infusion rate, hydroxyethyl starch infusion rate and urine output. Table B: Activation of NF-κB, protein expression of the iNOS, HO-1, cleaved caspase-3, and nitrotyrosine in post mortem kidney specimen. Table C: Kidney histopathology score. Figure A: Representative examples of the histopathological items analyzed. (DOCX 1 MB) Martin Wepler, Sebastian Hafner contributed equally to this work. Competing interests The authors declare that they have no competing interests. Authors’ contributions

Electronic supplementary material Additional file 1: Supplementary information. Table A: Norepinephrine infusion rate, hydroxyethyl starch infusion rate and urine output. Table B: Activation of NF-κB, protein expression of the iNOS, HO-1, cleaved caspase-3, and nitrotyrosine in post mortem kidney specimen. Table C: Kidney histopathology score. Figure A: Representative examples of the histopathological items analyzed. (DOCX 1 MB) Martin Wepler, Sebastian Hafner contributed equally to this work. Competing interests The authors declare that they have no competing interests. Authors’ contributions PR, EC, CT, AK, and MM conceived the study and designed the experiment. MW, SH, FS, and JM were responsible for anesthesia and surgery and, together with MR, MGr, FG, and AS, for data collection and statistical analysis. FW and BS carried out the immune biology measurements. OM, ASch, and PM were responsible for histology and immunohistochemistry. UW and JV performed the isotope measurements. PR, MGe, CT, and PM drafted the manuscript. All authors read and approved the final manuscript. Acknowledgements This study was supported by the Else-Kröner-Fresenius-Stiftung (AZ 2011_A18). Dr. M. Matejovic was supported by the Charles University Research Fund (project number P36). Very special thanks are dedicated to Rosemarie Mayer, Rosa Maria Engelhardt, Tanja Schulz, and Ingrid Eble for their skillful technical assistance.

Review Introduction Severe sepsis is characterized by a dysregulated systemic inflammatory response to infection resulting in acute multiple organ dysfunction (MOF) and a high mortality rate. The pathophysiology of sepsis-induced MOF remains incompletely understood but a growing body of evidence supports impairment of cellular oxygen utilization as a key mechanism [1–4]. Considerable enthusiasm has recently surrounded the potential beneficial effect of fibrates, thiazolidinediones (TZD) and, particularly, statins as adjunctive therapies for sepsis [5–9]. Some experimental studies also suggest a role for resveratrol [10–14]. However, most of these positive outcomes have been generated in laboratory models of sepsis such as caecal ligature and puncture or injection of endotoxin. The discrepancies observed to date between human and experimental studies may relate to the difficulty in reproducing the complexity of human sepsis and/or the use of doses far in excess of those currently used in clinical practice. Prospective clinical trial data are insufficient, particularly for fibrates and TZDs [15–17], and non-existent for resveratrol.

en human and experimental studies may relate to the difficulty in reproducing the complexity of human sepsis and/or the use of doses far in excess of those currently used in clinical practice. Prospective clinical trial data are insufficient, particularly for fibrates and TZDs [15–17], and non-existent for resveratrol. Statins, fibrates and TZDs modulate lipid and glucose metabolism. Resveratrol, a phenol constituent of red wine, is not available as a stand-alone medication but has been reported to slow down carcinogenesis, cardiovascular disease and ischaemic injury [18]. All four classes exert pleiotropic effects through mechanisms that remain incompletely understood [6, 8, 18, 19]. Immune-inflammatory modulation is a common property; however, several authors have also underlined effects on mitochondrial function [14, 20–23]. This may be highly pertinent to critical illness as bioenergetic dysfunction is implicated in the pathophysiology of sepsis-induced multi-organ failure [4]. In view of this burgeoning interest, it is timely to summarize the main recognized mechanisms of these agents, including their actions on mitochondria, and to offer a critical review of currently available experimental and clinical data.

Statins, fibrates and TZDs modulate lipid and glucose metabolism. Resveratrol, a phenol constituent of red wine, is not available as a stand-alone medication but has been reported to slow down carcinogenesis, cardiovascular disease and ischaemic injury [18]. All four classes exert pleiotropic effects through mechanisms that remain incompletely understood [6, 8, 18, 19]. Immune-inflammatory modulation is a common property; however, several authors have also underlined effects on mitochondrial function [14, 20–23]. This may be highly pertinent to critical illness as bioenergetic dysfunction is implicated in the pathophysiology of sepsis-induced multi-organ failure [4]. In view of this burgeoning interest, it is timely to summarize the main recognized mechanisms of these agents, including their actions on mitochondria, and to offer a critical review of currently available experimental and clinical data. Modes of action Statins In addition to lowering low-density lipoprotein (LDL) cholesterol, statins exhibit a wide range of other biological effects. Statins inhibit 3-hydroxy-3-methyl-glutaryl-coenzymeA (HMGCoA) reductase, a key enzyme in the mevalonate pathway. Mevalonate is a precursor for cholesterol, ubiquinone and isoprenoids (Figure 1) [24]. Thus, statins can decrease all three end products but, as a consequence of enzyme affinity, mainly reduce cholesterol production.Figure 1 Schema showing mechanisms involved in mitochondrial dysfunction. During sepsis and potential points of modulation by statins, PPAR agonists and resveratrol. ROS, reactive oxygen species; NO, nitric oxide; PPAR, peroxisome proliferator-activated receptor; CoQ10, coenzyme Q10, ubiquinone.

cholesterol production.Figure 1 Schema showing mechanisms involved in mitochondrial dysfunction. During sepsis and potential points of modulation by statins, PPAR agonists and resveratrol. ROS, reactive oxygen species; NO, nitric oxide; PPAR, peroxisome proliferator-activated receptor; CoQ10, coenzyme Q10, ubiquinone. Ubiquinone is both an electron carrier within the mitochondrial electron transport chain that generates ATP and a powerful anti-oxidant [21]. While several clinical and experimental studies have reported that statins decreased the ratio of plasma ubiquinone to LDL (its natural carrier), its effects on tissue ubiquinone levels are more controversial. It may be an important mechanism underlying statin-induced myopathy [25].

TP and a powerful anti-oxidant [21]. While several clinical and experimental studies have reported that statins decreased the ratio of plasma ubiquinone to LDL (its natural carrier), its effects on tissue ubiquinone levels are more controversial. It may be an important mechanism underlying statin-induced myopathy [25]. Isoprenoids (farnesyl pyrophosphate and geranyl pyrophosphate) serve as lipid attachments and activators for various signalling molecules such as G-protein and GTP-binding protein (Ras and Ras-like protein) [24], which have been associated with reactive oxygen species (ROS) production and activation of pro-inflammatory pathways (reviewed by Blanco-Colio et al. [26]). The pleiotropic effects of statins have been associated with decreased levels of these small proteins [27]. Both in vivo and in vitro studies show that statins can induce cellular accumulation of endothelial nitric oxide synthase, inhibit expression of adhesion molecules and chemokines that recruit inflammatory cells, inhibit expression of pro-coagulant factors, induce production of anti-coagulant substances, increase apoptosis, decrease oxidative stress, and modulate the adaptive immune system (reviewed by Terblanche et al. [8]). In a volunteer study, pre-treatment with simvastatin prior to lipopolysaccharide (LPS) attenuated the upregulation of Toll-like receptor 2 and 4 surface expression on circulating monocytes [28]. How many of these effects are related to lowering LDL cholesterol remains uncertain. Of note, squalestatin, a selective inhibitor of the synthesis of sterol derivatives, has no anti-inflammatory effect compared to statins [24].

upregulation of Toll-like receptor 2 and 4 surface expression on circulating monocytes [28]. How many of these effects are related to lowering LDL cholesterol remains uncertain. Of note, squalestatin, a selective inhibitor of the synthesis of sterol derivatives, has no anti-inflammatory effect compared to statins [24]. Statins can affect skeletal muscle mitochondria in vitro by inhibiting respiratory chain complexes and oxidative capacity [29, 30], decreasing mitochondrial membrane potential [30], uncoupling oxidative phosphorylation [30], inducing mitochondrial swelling and apoptosis [30] and decreasing mitochondrial density [31] (Figure 1). However, no clear relationship has been documented between a decrease in ubiquinone and alterations in mitochondrial function. Hydrophilic statins (e.g. pravastatin) are much less ‘mitotoxic’ compared to lipophilic statins such as cerivastatin, fluvastatin, atorvastatin and simvastatin [30]. It is noteworthy that the toxic effects of atorvastatin are mostly observed with doses that are much higher than those prescribed to patients. The delayed metabolism of statins seen in critical illness may result in very high plasma levels so the risk of toxicity would potentially increase; however, this would be difficult to distinguish in an unstable patient with concurrent multi-organ failure [32]. As discussed later, these ‘toxic’ effects may, paradoxically, offer some protective effects during sepsis. Recently, Bouitbir et al. reported that statin-treated patients who underwent cardiac surgery had decreased oxidative stress, enhanced oxidative capacity, and a marked augmentation of mRNA expression of the peroxisome proliferator-activated receptor (PPAR) gamma co-activators 1α (PGC-1α) and 1β (PGC-1β). PGC-1α is the main regulator of mitochondrial biogenesis, i.e. new protein turnover [20]. This study raises new insights regarding the action of statins, but the clinical impact remains to be explored.

A expression of the peroxisome proliferator-activated receptor (PPAR) gamma co-activators 1α (PGC-1α) and 1β (PGC-1β). PGC-1α is the main regulator of mitochondrial biogenesis, i.e. new protein turnover [20]. This study raises new insights regarding the action of statins, but the clinical impact remains to be explored. Fibrates and thiazolidinediones PPARs are ligand-activated transcription factors that belong to the nuclear receptor superfamily. Once activated by ligands, PPARs form a heterodimer with the retinoic X receptor (RXR) that allows recruitment of a set of co-activators or co-repressors [6, 33]. This heterodimer binds to the PPAR response element within the promoter region of its target genes, provoking either expression or repression. PPAR inhibits expression of pro-inflammatory cytokines through direct or indirect actions on pro-inflammatory transcription factors (NF-κB, STAT, AP-1) [6, 34]. Fibrates are synthetic ligands of peroxisome proliferator-activated receptor-α (PPAR-α). Fibrates are used for treating dyslipidaemias, lowering both triglyceride and LDL cholesterol levels. They also ameliorate insulin resistance and glucose intolerance [5, 35]. The PPAR-α receptor is expressed mainly in brown fat and liver but has been found in many other cells [5, 36].

tor-activated receptor-α (PPAR-α). Fibrates are used for treating dyslipidaemias, lowering both triglyceride and LDL cholesterol levels. They also ameliorate insulin resistance and glucose intolerance [5, 35]. The PPAR-α receptor is expressed mainly in brown fat and liver but has been found in many other cells [5, 36]. TZDs are synthetic ligands of peroxisome proliferator-activated receptor-γ (PPAR-γ). While most of their effects appear dependent upon PPAR activation, TZDs could also exert anti-inflammatory effects in macrophages via a PPAR-independent pathway [37]. Thiazolidinediones are drugs used for managing type II diabetes mellitus and the metabolic syndrome. Rosiglitazone, pioglitazone, troglitazone, rivoglitazone and ciglitazone are members of this therapeutic class. However, most have been withdrawn from the market due to adverse effects on heart. The PPAR-γ receptor is highly expressed in adipose tissue and, to a lesser extent, in intestine, immune and stem cells [6]. Activation of these receptors decreases insulin resistance and modifies lipid storage. A time-dependent downregulation of PPAR-γ expression has been reported in experimental sepsis that is partially restored by TZDs [34].

ghly expressed in adipose tissue and, to a lesser extent, in intestine, immune and stem cells [6]. Activation of these receptors decreases insulin resistance and modifies lipid storage. A time-dependent downregulation of PPAR-γ expression has been reported in experimental sepsis that is partially restored by TZDs [34]. After various inflammatory insults, in vivo and in vitro studies have shown that both fibrates and TZDs improved endothelial dysfunction [38, 39], inhibited expression of adhesion molecules and inflammatory cytokines [40, 41] and decreased oxidative stress and nitric oxide production [39, 42]. Fibrates can inhibit coagulation [38, 43] and may improve haemorheologic parameters [44]. TZDs increase plasma adiponectin levels [45] and may initiate macrophage apoptosis via caspase-3 activation [46]. Both drug groups also impair mitochondrial function, at least in vitro, via direct inhibition of mitochondrial respiration (mainly complex I) [47], by membrane depolarization [48] and through increases in uncoupled respiration [49, 50] (Figure 1). At lower doses, TZDs enhanced mitochondrial potential and promoted lymphocyte survival [51]. Several studies report that PPAR-γ agonists induce mitochondrial biogenesis by mechanisms that are not fully understood but could be mediated, at least in part, via activation of PGC-1α [52, 53]. None of these studies have been carried out in sepsis, where early activation of mitochondrial biogenesis has been associated with survival [54].

t that PPAR-γ agonists induce mitochondrial biogenesis by mechanisms that are not fully understood but could be mediated, at least in part, via activation of PGC-1α [52, 53]. None of these studies have been carried out in sepsis, where early activation of mitochondrial biogenesis has been associated with survival [54]. Statins, fibrates and TZDs can act synergistically to affect some of the pathways previously described [55]. For example, patients with cardiovascular disease showed additive anti-inflammatory effect when statins were given in combination with PPAR-γ agonists [56]. Resveratrol Resveratrol is a natural phenol present in many plants but especially abundant in red grapes, peanuts and mulberries [18]. It exerts beneficial effects in experimental sepsis when administered either before or shortly after the septic insult. The mechanisms involved are not yet clearly defined. Recent research has demonstrated that resveratrol activates the silent mating type information regulator 2 homolog 1 (SIRT1), which is a key regulator of cellular defenses and cell survival in response to stress [57]. Of specific interest in sepsis is the interaction between SIRT1 and mitochondrial biogenesis [23, 57]. Resveratrol has also been shown to downregulate the pro-inflammatory response [14, 58] and to have anti-oxidant properties [11, 57].

s a key regulator of cellular defenses and cell survival in response to stress [57]. Of specific interest in sepsis is the interaction between SIRT1 and mitochondrial biogenesis [23, 57]. Resveratrol has also been shown to downregulate the pro-inflammatory response [14, 58] and to have anti-oxidant properties [11, 57]. Experimental and clinical evidence of statins, fibrates and thiazolidinedione effects in sepsis Statins In a murine model of sepsis, simvastatin pre-treatment markedly improved survival times (median 128 h versus 28 h, p < 0.005) [59]. Even treatment commencing after the onset of sepsis improved survival times, though less impressively (median 37 h versus 23 h for placebo, p < 0.05) [60]. In another study, 3 days of simvastatin pre-treatment improved survival and reduced sepsis-induced acute kidney injury by direct effects on the renal microvasculature, reversal of tubular hypoxia and a decrease in systemic inflammation [61]. In a rat model of peritonitis, 30 days' pre-treatment with high doses of simvastatin or atorvastatin prevented hepatic mitochondrial enzyme dysfunction; however, no improvement was seen in liver dysfunction while mortality differences were not reported [62].

tubular hypoxia and a decrease in systemic inflammation [61]. In a rat model of peritonitis, 30 days' pre-treatment with high doses of simvastatin or atorvastatin prevented hepatic mitochondrial enzyme dysfunction; however, no improvement was seen in liver dysfunction while mortality differences were not reported [62]. Several observational studies have reported significant survival improvement in large cohorts of patients on statin therapy who suffer from bacterial [63–66] or viral infections [67]. However, some authors argue this simply represents a healthy user effect [68, 69]. An association with harm was even reported in a study of infections post-stroke [70]. Four double-blind, placebo-controlled, randomized clinical studies have been performed to examine the impact of either introducing [71–73] or continuing [74] statin therapy in patients with sepsis. Statin treatment was not associated with a significant decrease in pro-inflammatory markers compared to placebo [71, 73, 74]. In a different context, healthy volunteers randomized to statin pre-treatment and subsequently challenged with inhaled LPS manifested less pulmonary and systemic inflammation [75]. While one study showed a reduction in the progression to severe sepsis in patients taking statins, albeit with a similar rate of intensive care admissions between the statin and control groups [72], this was not confirmed by other studies [71, 73, 74]. Finally, stopping statin therapy in septic patients was not associated with worse outcomes [74]. As mentioned previously, statins should be used with caution in critically ill patients due to their unpredictable pharmacokinetics and the amplifying effect related to co-administration of drugs with cytochrome P450 inhibitory effects [32]. Opinions on the utility of statins in sepsis thus remain mixed.

outcomes [74]. As mentioned previously, statins should be used with caution in critically ill patients due to their unpredictable pharmacokinetics and the amplifying effect related to co-administration of drugs with cytochrome P450 inhibitory effects [32]. Opinions on the utility of statins in sepsis thus remain mixed. Fibrates A few studies have reported improved outcomes following fibrate therapy in experimental sepsis [38, 76]. For example, influenza-infected mice pre-treated with gemfibrozil had a 54% reduction in mortality [76]. In patients with chronic hepatitis C and hyperlipidaemia, bezafibrate given as adjunctive therapy decreased plasma virus titres and improved liver dysfunction [16]. Fibrates also prevented muscular atrophy [76], a major problem occurring with sepsis. Protective mechanisms need to be elucidated but a decrease in atrogen and myostatin expression has been shown with fibrates therapy in a rat model of chronic inflammation [77]. Children with severe sepsis had decreased leukocyte PPAR-α expression, and this was related to disease severity [78]. In a randomized controlled trial in children with severe burn injury, in which condition mitochondrial dysfunction has been previously observed [79], mitochondrial biogenesis and oxidative phosphorylation were improved with fenofibrates therapy commenced within a week post-burn [35].

s related to disease severity [78]. In a randomized controlled trial in children with severe burn injury, in which condition mitochondrial dysfunction has been previously observed [79], mitochondrial biogenesis and oxidative phosphorylation were improved with fenofibrates therapy commenced within a week post-burn [35]. Thiazolidinediones In different models of sepsis, pre- or post-treatment with TZD improved outcomes, blunted pro-inflammatory cytokine production and reduced organ injury [34, 42, 45, 80, 81]. A recent study in endotoxic mice revealed protection from the PPAR-γ agonist, rosiglitazone with less reduction in mitochondrial content, improved cardiac dysfunction and better survival rates [82]. On the other hand, the PPAR-α agonist WY-14643 offered no protection. Pioglitazone given to healthy volunteers after endotoxin challenge did not affect plasma levels of TNFα, IL-6 levels or the adhesion molecule VCAM ([83] abstract). An improvement in blood pressure was reported in one study investigating ciglitazone administration following caecal ligation and puncture in rats, but there was no effect on survival [34]. No observational clinical trials are available with TZDs; however, users appear to be at higher risk of pneumonia or lower respiratory tract infection [84]. The risk of heart attack described with this drug has discouraged its widespread use and thus the possibility of conducting large-scale observational studies [85]. Several studies reported a significant reduction in weight loss with TZDs following different types of infectious challenge [42, 86, 87]. This observation has not been specifically studied but TZDs block activation of NF-κB [34], a main activator of muscle wasting [88], in addition to their effects on stimulating PGC-1α and mitochondrial biogenesis [89].

nt reduction in weight loss with TZDs following different types of infectious challenge [42, 86, 87]. This observation has not been specifically studied but TZDs block activation of NF-κB [34], a main activator of muscle wasting [88], in addition to their effects on stimulating PGC-1α and mitochondrial biogenesis [89]. The role of TDZs is not limited to bacterial infection. In a randomized placebo-controlled trial of 140 patients suffering from non-severe malaria, Boggild et al. reported in faster blood parasite clearance and a more rapid decrease in inflammatory biomarker levels (IL-6 and monocyte chemoattractant protein-1) in those given rosiglitazone as adjunctive therapy [15]. After a lethal dose of C. Albicans, mice post-treated with pioglitazone had better outcomes and less renal dysfunction [87]. Rosiglitazone also dramatically improved survival in a murine influenza model [86, 90] and reduced HIV-1 replication in lymphocytes and brain macrophages in an experimental model of HIV-1 encephalitis [91]. Resveratrol Resveratrol improved sepsis-induced acute organ injury [10, 12, 13], but the effect on mortality was uncertain [12, 14]. In different cell or tissue types, resveratrol upregulated PGC-1α and decreased mitochondrial ROS production with a consequent increase in mitochondrial size, DNA content and mitochondrial respiratory enzymatic activity [14, 23].

duced acute organ injury [10, 12, 13], but the effect on mortality was uncertain [12, 14]. In different cell or tissue types, resveratrol upregulated PGC-1α and decreased mitochondrial ROS production with a consequent increase in mitochondrial size, DNA content and mitochondrial respiratory enzymatic activity [14, 23]. Conclusions Through their pleiotropic actions, statins, fibrates, thiazolidinediones and resveratrol can target multiple mechanisms involved in sepsis. These are summarized in the schema shown in Figure 1. Their actions on mitochondrial function and, particularly mitochondrial biogenesis, are of interest in a pathological state where mitochondrial dysfunction may play a key role in the development of organ dysfunction. We can speculate that inhibitory effects of these agents, in particular statins, on mitochondrial function in otherwise healthy patients may potentially offer some benefit in the septic state. By decreasing mitochondrial activity and membrane potential, production of mitochondrial reactive oxygen species would decrease and this may result in a greater degree of cell and mitochondrial protection. Further clinical and experimental studies are warranted to reveal whether benefit can be shown in septic patient populations or in those at high risk of developing sepsis. Abbreviations HMGCoA3-hydroxy-3-methyl-glutaryl-coenzymeA LDLlow-density lipoprotein LPSlipopolysaccharide PGC-1βperoxisome proliferator activated receptor γ coactivator 1 β PGC-1αperoxisome proliferator activated receptor γ coactivator 1 α PPARperoxisome proliferator-activated receptor

Conclusions Through their pleiotropic actions, statins, fibrates, thiazolidinediones and resveratrol can target multiple mechanisms involved in sepsis. These are summarized in the schema shown in Figure 1. Their actions on mitochondrial function and, particularly mitochondrial biogenesis, are of interest in a pathological state where mitochondrial dysfunction may play a key role in the development of organ dysfunction. We can speculate that inhibitory effects of these agents, in particular statins, on mitochondrial function in otherwise healthy patients may potentially offer some benefit in the septic state. By decreasing mitochondrial activity and membrane potential, production of mitochondrial reactive oxygen species would decrease and this may result in a greater degree of cell and mitochondrial protection. Further clinical and experimental studies are warranted to reveal whether benefit can be shown in septic patient populations or in those at high risk of developing sepsis. Abbreviations HMGCoA3-hydroxy-3-methyl-glutaryl-coenzymeA LDLlow-density lipoprotein LPSlipopolysaccharide PGC-1βperoxisome proliferator activated receptor γ coactivator 1 β PGC-1αperoxisome proliferator activated receptor γ coactivator 1 α PPARperoxisome proliferator-activated receptor TZDthiazolidinediones ROSreactive oxygen species SIRT1silent mating type information regulator 2 homolog-1. Competing interests The authors declare that they have no competing interests. Authors’ contributions JM and MS wrote the article. Both authors read and approved the final manuscript.

Background Acute respiratory distress syndrome (ARDS) is characterized by a major decrease in lung aerated volume. End-expiratory lung volume measurement by the nitrogen washin-washout technique (EELVWI-WO) [1] is available at the bedside from an ICU ventilator, and may help titrating PEEP during mechanical ventilation of ARDS patients. Validation of this technique has been previously performed in mechanically ventilated patients using computed tomography (CT) as gold standard [2], but at relatively low PEEP levels (5 cm H2O), low fraction of inspired oxygen (FiO2) and respiratory rate (RR), and with tidal volume (VT) 8 ± 1 mL · kg-1 in the upper range of current experts’ recommendations for ARDS management [3]. Such ventilatory settings may have insufficiently challenged the validity of this technique, which requires a constant inhomogeneity in alveolar gas throughout the measurement, and may be less precise at FiO2 greater than 0.7 [1]. Furthermore, the WI-WO technique is particularly suitable for repeated EELV assessment, and hence to identify EELV trends, but has never been formally validated for this purpose. The aims of this study were to evaluate (1) the reliability of EELVWI-WO measurement at variable PEEP and VT, at high RR and FiO2 during experimental ARDS, using CT as a reference and (2) the trending ability of WI-WO technique to detect change in EELV associated with PEEP and VT variations.

Furthermore, the WI-WO technique is particularly suitable for repeated EELV assessment, and hence to identify EELV trends, but has never been formally validated for this purpose. The aims of this study were to evaluate (1) the reliability of EELVWI-WO measurement at variable PEEP and VT, at high RR and FiO2 during experimental ARDS, using CT as a reference and (2) the trending ability of WI-WO technique to detect change in EELV associated with PEEP and VT variations. Methods This study was approved by our Institutional Review Board for the care of animal subjects (Comité d’experimentation animale de l′université Lyon I), and carried out in 14 pigs (28 ± 2 kg). Animal preparation Pigs were anesthetized with propofol and fentanyl, tracheotomized and mechanically ventilated in volume-controlled mode, with constant inspiratory flow, VT 10 mL · kg-1, inspired fraction of oxygen (FiO2) 0.21, zero end-expiratory pressure, and RR adjusted to achieve normocapnia using Engström Carestation® ventilator (General Electric Healthcare, Madison, WI, USA). Muscle relaxation was obtained with pancuronium bromide. Right jugular vein was cannulated with a 3-lumen 8.5-Fr catheter for drug administration. Carotid artery was cannulated with an 8.5 Fr catheter. FiO2 was increased to 1 at the end of animal preparation.

ventilator (General Electric Healthcare, Madison, WI, USA). Muscle relaxation was obtained with pancuronium bromide. Right jugular vein was cannulated with a 3-lumen 8.5-Fr catheter for drug administration. Carotid artery was cannulated with an 8.5 Fr catheter. FiO2 was increased to 1 at the end of animal preparation. Measurements Air flow was measured using a small volume pneumotachograph (PN 281637, Hamilton medical AG, Bonaduz, Switzerland). Pressure at the airway opening was measured using a connecting tube with lateral aperture connected between the endotracheal tube and the pneumotachograph. Signals of arterial blood pressure, pressure at the airway opening, and air flow were read by transducers (Becton Dickinson, Sandy, UT, USA), connected to an A/D card (MP 100; Biopac Systems, Santa Barbara, CA, USA), acquired at 200 Hz and analyzed with Acknowledge® software (Biopac Systems, Santa Barbara, CA, USA). Tracheal pressure was measured through an air filled catheter introduced down the endotracheal tube, positioned 2 cm distal to the tube tip, and connected to the ventilator, to obtain alveolar pressure [4]. EELVWI-WO was assessed by the ventilator, by using the nitrogen washout/washin technique [1], from continuous measurement of end-tidal O2 and CO2 during a 0.1 change of FiO2 using pediatric sensors (Pedi-lite+, Dahtex-Ohmeda Inc, Madison, WI, USA). The average value of the washout and washin measurements during 1 to 0.9 and 0.9 to 1 FiO2 changes was given by the ventilator.

itrogen washout/washin technique [1], from continuous measurement of end-tidal O2 and CO2 during a 0.1 change of FiO2 using pediatric sensors (Pedi-lite+, Dahtex-Ohmeda Inc, Madison, WI, USA). The average value of the washout and washin measurements during 1 to 0.9 and 0.9 to 1 FiO2 changes was given by the ventilator. EELVCT was calculated using lung CT, as previously described [2]. CT calibration using the manufacturer phantom was performed before each CT study. The CT scanner (Biograph mCT/S, Siemens, Munich, Germany) was set as follows: interval 1 mm, voltage 120 kV, pitch 1.2 mm, and field of view 300 mm. Whole lung CT images were taken during 15 s end-expiratory. CT raw data were reconstructed as 1-mm-thick contiguous slices using a medium smooth filter (B31f). Image segmentation was manually performed over the whole lung using Turtleseg® software [5, 6] (http://www.turtleseg.org). Gas volume in each lung voxel was computed from the CT number using the following formulas [2]: Gas volume = 0 for lung voxels with CT number > 0. EELVCT was computed as the sum of gas volume in all the voxels defined by lung segmentation. Expected EELV on zero end-expiratory pressure was deemed as 33 mL · kg-1 body weight as previously published in normal anesthetized pigs [7].

EELVCT was calculated using lung CT, as previously described [2]. CT calibration using the manufacturer phantom was performed before each CT study. The CT scanner (Biograph mCT/S, Siemens, Munich, Germany) was set as follows: interval 1 mm, voltage 120 kV, pitch 1.2 mm, and field of view 300 mm. Whole lung CT images were taken during 15 s end-expiratory. CT raw data were reconstructed as 1-mm-thick contiguous slices using a medium smooth filter (B31f). Image segmentation was manually performed over the whole lung using Turtleseg® software [5, 6] (http://www.turtleseg.org). Gas volume in each lung voxel was computed from the CT number using the following formulas [2]: Gas volume = 0 for lung voxels with CT number > 0. EELVCT was computed as the sum of gas volume in all the voxels defined by lung segmentation. Expected EELV on zero end-expiratory pressure was deemed as 33 mL · kg-1 body weight as previously published in normal anesthetized pigs [7]. Protocol ARDS was performed by saline lavage at ventilatory settings mentioned above. Intra-tracheal instillations of 1,000 mL aliquots of 0.9% sodium chloride warmed at 37°C were repeated until PaO2/FiO2 ratio was <100 mmHg. RR may be increased up to 35 breaths per min to maintain pH above 7.20, then kept constant except at the end of experiment, where at the highest VT, it was decreased to maintain peak airway pressure below 100 cm H2O.

0 mL aliquots of 0.9% sodium chloride warmed at 37°C were repeated until PaO2/FiO2 ratio was <100 mmHg. RR may be increased up to 35 breaths per min to maintain pH above 7.20, then kept constant except at the end of experiment, where at the highest VT, it was decreased to maintain peak airway pressure below 100 cm H2O. Then, PEEP was set to 20 cm H2O, VT to 6 mL · kg-1, and a recruitment maneuver was performed by applying a continuous airway pressure of 40 cm of H2O over 40 s. A decremental PEEP trial was then performed from 20 to 2 cm H2O by 2 cm H2O steps of 10 min each. At the end of the decremental PEEP trial, animals were randomized into three PEEP groups, for which PEEP level was set according to either highest compliance (n = 4), or highest EELVWI-WO (n = 5), or PEEP-FiO2 table (n = 4) [8]. This randomization was used to deliver a wide PEEP range during the final part of the study, in order to obtain multiple combinations of PEEP and VT, so as to perform a multivariate analysis adjusted for PEEP and VT. One pig died just after the PEEP trial before randomization, and was kept in the final analysis. The selected PEEP was applied for 1 h, and VT was adjusted to maintain plateau pressure of the respiratory system ≤30 cm of H2O.

ombinations of PEEP and VT, so as to perform a multivariate analysis adjusted for PEEP and VT. One pig died just after the PEEP trial before randomization, and was kept in the final analysis. The selected PEEP was applied for 1 h, and VT was adjusted to maintain plateau pressure of the respiratory system ≤30 cm of H2O. After 1 h of applied selected PEEP, eight levels of VT (4, 5, 6, 7, 8, 10, 15, 20 mL · kg-1) ranging from 100 to 625 mL, were applied for 2 min leaving PEEP level unchanged. EELVWI-WO and EELVCT were measured immediately after ARDS onset, at the end of each PEEP step during the PEEP trial, and at the end of each VT change. A 15-s end-inspiratory pause was performed to check the absence of air leak in each experimental condition. Some experimental conditions were not available since pneumothorax occurred in several pigs at high VT or since EELVWI-WO values were lacking for technical reasons, ending up in 218 data points in final analysis (see Additional file 1: Table S1 for description of lacking data points). Statistical analysis Statistical analyses were performed using R software [9], with packages nlme [10], MethComp [11], pROC [12], OptimalCutpoints [13], and multcomp [14]. Values were expressed as mean ± standard deviation (SD). The level of statistical significance was set below 0.05. EELVWI-WO and EELVCT were compared using a linear mixed-effects model, and Bland and Altman representation [15]. Limits of agreement were computed using alternating regression [16] since bias was non-constant and the experimental design involved repeated measurements.

Statistical analysis Statistical analyses were performed using R software [9], with packages nlme [10], MethComp [11], pROC [12], OptimalCutpoints [13], and multcomp [14]. Values were expressed as mean ± standard deviation (SD). The level of statistical significance was set below 0.05. EELVWI-WO and EELVCT were compared using a linear mixed-effects model, and Bland and Altman representation [15]. Limits of agreement were computed using alternating regression [16] since bias was non-constant and the experimental design involved repeated measurements. To control for an effect of confounding variables on the bias, a linear mixed-effects model was built using PEEP, VT, EELVCT at ARDS onset and their interactions as factors with fixed effect, pigs as factor with random effect [17], and bias as dependent variable. Model simplification was performed using a backward stepwise algorithm. Percentage error was computed as × SDBias/meanEELV[18]. As percentage error was not reported in the two previously published studies that compared EELVWI-WO and EELVCT[2, 19], Cartesian data of these studies were reanalyzed, being uplifted using a scientific program allowing extraction of individual data points from a digitalized graph (DataThief III®[20]) as follows: a digital copy of each regression plot was analyzed with DataThief from the portable document format file of the journal articles, and the extracted data were exported as two columns of X-Y coordinates, with each row representing an extracted data point, allowing computation of percentage error of each study.

follows: a digital copy of each regression plot was analyzed with DataThief from the portable document format file of the journal articles, and the extracted data were exported as two columns of X-Y coordinates, with each row representing an extracted data point, allowing computation of percentage error of each study. Changes in EELV between consecutive measurements were computed for EELVWI-WO (ΔEELVWI-WO) and EELVCT (ΔEELVCT). Ability of the WI-WO technique to track changes in EELV was assessed using four-quadrant and polar plots. The four-quadrant plot relates ΔEELVWI-WO and ΔEELVCT, with upper right and lower left quadrants being quadrants of agreement (in which both EELVWI-WO and EELVCT have the same directional changes) and lower right and upper left quadrants being quadrants of disagreement (in which EELVWI-WO and EELVCT have opposite directional changes). Concordance rate was defined as the percentage of data points falling into one of the two quadrants of agreement, expressed as a percentage of the total number of data points [21]. The main drawback of the four-quadrant plot is the lack of quantification of the distance between each data point and the line of identity, leading to the development of polar plot analysis [21]. Polar plot is obtained by a 45° clockwise rotation of the four-quadrant plot, changing the dimensions of the radius to mean ΔEELV [22], lining up the line of identity along the horizontal axis. Data points with positive and negative directional changes are located on the right and the left side of the polar plot, respectively, and the polar angle represents the angle of each data point with line of identity. A 0° polar angle depicts a perfect agreement between ΔEELVWI-WO and ΔEELVCT, while polar angles in the range 45° to 135° and 225° to 315° depict disagreement between directional changes of EELVWI-WO and EELVCT. The following variables are computed from polar plots: (1) angular bias as the mean angle between all data points and polar axis [21], reflects the difference in calibration between the reference and test methods; (2) radial limits of agreement as the radial sector containing 95% of the data points, after conversion of negative deflections to positive ones, is a polar version of the 95% confidence limits and is similar to the limits of agreement in Bland and Altman analysis [21].

libration between the reference and test methods; (2) radial limits of agreement as the radial sector containing 95% of the data points, after conversion of negative deflections to positive ones, is a polar version of the 95% confidence limits and is similar to the limits of agreement in Bland and Altman analysis [21]. Bias and angular bias were compared to zero using Mann-Whitney U test. Multiple comparisons were performed with Dunnett’s test using PEEP 0 as a reference. The ability of the WI-WO technique to detect a change in EELV greater than 100, 150, 200, 250, and 300 mL was tested by computations of area under receiver operating characteristic (AUC) curve. The optimal cut-off points were computed using the Youden J statistic.

Bias and angular bias were compared to zero using Mann-Whitney U test. Multiple comparisons were performed with Dunnett’s test using PEEP 0 as a reference. The ability of the WI-WO technique to detect a change in EELV greater than 100, 150, 200, 250, and 300 mL was tested by computations of area under receiver operating characteristic (AUC) curve. The optimal cut-off points were computed using the Youden J statistic. Results Ventilatory settings and arterial blood gases during the whole experiment are reported in Additional file 2: Table S2. Figure 1 depicts the evolution of EELVWI-WO and EELVCT over time. Immediately after ARDS onset, EELVCT was very low (236 ± 143 mL (25% ± 15% of its theoretical value, range 104 to 668 mL)), and increased to 1,206 ± 185 mL (range 957 to 1,528 mL) at PEEP 20.Figure 1 EELV WI-WO and EELV CT at each experimental condition (upper panel), with corresponding RR, PEEP, and V T (lower panel). Values are mean ± standard deviation. ARDS, experimental acute respiratory distress syndrome; EELVWI-WO, end-expiratory lung volume assessed with the nitrogen washin-washout technique; EELVCT, end-expiratory lung volume assessed by computed tomography; optimal PEEP, optimal PEEP level according to one of the three methods (see text for details); PEEP, positive end-expiratory pressure; RR, respiratory rate; V T, tidal volume. EELCCT and EELVWI-WO values were very close at PEEP 0, but their difference progressively increased with the PEEP level.

Results Ventilatory settings and arterial blood gases during the whole experiment are reported in Additional file 2: Table S2. Figure 1 depicts the evolution of EELVWI-WO and EELVCT over time. Immediately after ARDS onset, EELVCT was very low (236 ± 143 mL (25% ± 15% of its theoretical value, range 104 to 668 mL)), and increased to 1,206 ± 185 mL (range 957 to 1,528 mL) at PEEP 20.Figure 1 EELV WI-WO and EELV CT at each experimental condition (upper panel), with corresponding RR, PEEP, and V T (lower panel). Values are mean ± standard deviation. ARDS, experimental acute respiratory distress syndrome; EELVWI-WO, end-expiratory lung volume assessed with the nitrogen washin-washout technique; EELVCT, end-expiratory lung volume assessed by computed tomography; optimal PEEP, optimal PEEP level according to one of the three methods (see text for details); PEEP, positive end-expiratory pressure; RR, respiratory rate; V T, tidal volume. EELCCT and EELVWI-WO values were very close at PEEP 0, but their difference progressively increased with the PEEP level. Comparison of EELVWI-WO and EELVCT EELVWI-WO and EELVCT were significantly correlated (R2 = 0.63, p < 0.001). The regression equation between EELVWI-WO and EELVCT, had an intercept of 96 mL (p < 0.001) and a slope of 0.58 mL-1 (p < 0.001; Figure 2). Bland and Altman representation exhibited a non-constant bias, decreasing toward more negative values as mean EELV increased (Figure 3). The difference between EELVWI-WO and EELVCT was related to their mean value by the following equation:Figure 2 Relationship between EELV WI-WO and EELV CT . Each symbol represents a concomitant measurement of end-expiratory lung volume assessed with either nitrogen washin-washout technique (EELVWI-WO) or computed tomography (EELVCT). Solid line is the regression line. Dashed line is the line of identity.

owing equation:Figure 2 Relationship between EELV WI-WO and EELV CT . Each symbol represents a concomitant measurement of end-expiratory lung volume assessed with either nitrogen washin-washout technique (EELVWI-WO) or computed tomography (EELVCT). Solid line is the regression line. Dashed line is the line of identity. Figure 3 Bias and limits of agreement between EELV CT and EELV WI-WO , using Bland and Altman representation. Each symbol represents a concomitant measurement of EELVWI-WO and EELVCT. Horizontal continuous line and horizontal broken lines are the mean bias and 95% prediction interval limits of the bias between EELVWI-WO and EELVCT, respectively. EELVWI-WO, end-expiratory lung volume assessed with the nitrogen washin-washout technique; EELVCT, end-expiratory lung volume assessed by computed tomography; 95% p.i., 95% prediction interval of the bias between EELVWI-WO and EELVCT. 1

Figure 3 Bias and limits of agreement between EELV CT and EELV WI-WO , using Bland and Altman representation. Each symbol represents a concomitant measurement of EELVWI-WO and EELVCT. Horizontal continuous line and horizontal broken lines are the mean bias and 95% prediction interval limits of the bias between EELVWI-WO and EELVCT, respectively. EELVWI-WO, end-expiratory lung volume assessed with the nitrogen washin-washout technique; EELVCT, end-expiratory lung volume assessed by computed tomography; 95% p.i., 95% prediction interval of the bias between EELVWI-WO and EELVCT. 1 Limits of agreement of the bias were ±398 mL (Figure 3), and percentage error was computed to 57%. A significant interaction between PEEP, VT, EELV baseline value on the bias between methods was identified (Table 1) and reported (Figure 4). The bias between methods was strongly influenced by PEEP level, increasing at higher PEEP regardless the VT level. The bias further increased when high PEEP was combined to low VT, when EELV at baseline was low.The bias at PEEP 0 amounted to -54 ± 101 mL, was not significantly different from 0, and was compared to bias values at higher PEEP. As shown in Figure 5, the bias at PEEP 10 and higher did significantly differ from PEEP 0, while non-significant differences were found for lower PEEP.Table 1 Statistical modeling of the bias between EELV WI-WO and EELV CT as a function of confounding variables

nt from 0, and was compared to bias values at higher PEEP. As shown in Figure 5, the bias at PEEP 10 and higher did significantly differ from PEEP 0, while non-significant differences were found for lower PEEP.Table 1 Statistical modeling of the bias between EELV WI-WO and EELV CT as a function of confounding variables AIC Statistical significance Model 1: No explanatory variable 2992 Model 2: Adjusting for V T 2994 V T: p = 0.55 Model 3: Adjusting for PEEP 2868 PEEP: p < 0.0001 Model 4: Adjusting for EELVBase 2980 EELVBase: p < 0.0001 Model 5: Final model adjusting for V T, PEEP, EELVBase and their three-way interaction 2821 three-way interaction: p < 0.0001 V T × PEEP interaction: p < 0.0001 V T × EELVBase: p < 0.001 PEEP × EELVBase: p < 0.0001 V T: p < 0.001 PEEP: p < 0.0001 EELVBase: p < 0.05 AIC, Akaike information criterion; EELV, end-expiratory lung volume; EELV Base, EELV baseline value (ARDS onset, PEEP 0 cm H2O); PEEP, positive end-expiratory pressure; V T, tidal volume.

AIC Statistical significance Model 1: No explanatory variable 2992 Model 2: Adjusting for V T 2994 V T: p = 0.55 Model 3: Adjusting for PEEP 2868 PEEP: p < 0.0001 Model 4: Adjusting for EELVBase 2980 EELVBase: p < 0.0001 Model 5: Final model adjusting for V T, PEEP, EELVBase and their three-way interaction 2821 three-way interaction: p < 0.0001 V T × PEEP interaction: p < 0.0001 V T × EELVBase: p < 0.001 PEEP × EELVBase: p < 0.0001 V T: p < 0.001 PEEP: p < 0.0001 EELVBase: p < 0.05 AIC, Akaike information criterion; EELV, end-expiratory lung volume; EELV Base, EELV baseline value (ARDS onset, PEEP 0 cm H2O); PEEP, positive end-expiratory pressure; V T, tidal volume. Figure 4 Interaction plot between V T , PEEP, and EELV Base . V T, PEEP, and EELVBase were classified as high or low, as a function of their relationship with their median value. The following cut-off values were identified: V T 170 mL; PEEP 10 cm H2O; EELVBase = 157 mL. Bars are mean values, and error bars, standard deviations. Bias, mean difference between EELVWI-WO and EELVCT in each subgroup; EELVBase, end-expiratory lung volume at baseline (ARDS onset, PEEP 0); EELVWI-WO, end-expiratory lung volume assessed with the nitrogen washin-washout technique; EELVCT, end-expiratory lung volume assessed by computed tomography; PEEP, positive end-expiratory pressure; V T, tidal volume. Figure 5 Mean difference in bias between each PEEP level from 2 to 20 and PEEP 0 cm H 2 O. Closed circles are mean differences and bars are 95% confidence intervals.

Figure 4 Interaction plot between V T , PEEP, and EELV Base . V T, PEEP, and EELVBase were classified as high or low, as a function of their relationship with their median value. The following cut-off values were identified: V T 170 mL; PEEP 10 cm H2O; EELVBase = 157 mL. Bars are mean values, and error bars, standard deviations. Bias, mean difference between EELVWI-WO and EELVCT in each subgroup; EELVBase, end-expiratory lung volume at baseline (ARDS onset, PEEP 0); EELVWI-WO, end-expiratory lung volume assessed with the nitrogen washin-washout technique; EELVCT, end-expiratory lung volume assessed by computed tomography; PEEP, positive end-expiratory pressure; V T, tidal volume. Figure 5 Mean difference in bias between each PEEP level from 2 to 20 and PEEP 0 cm H 2 O. Closed circles are mean differences and bars are 95% confidence intervals. Assessment of trending ability of the WI-WO technique ΔEELVWI-WO values adequately tracked ΔEELVCT changes against time (Figure 6). Concordance rate over all measurements amounted to 79%, and slightly increased after exclusion of small changes in EELV which do not reflect trending ability (Table 2).Figure 6 Four quadrants plot relating ΔEELV WI-WO with ΔEELV CT between consecutive measurements. Continuous black lines are quadrant limits. Dotted line is the regression line. Dashed line is the line of identity. Each data point is the change in end-expiratory lung volume (EELV) between consecutive measurements assessed with the nitrogen washin-washout technique (ΔEELVWI-WO) or computed tomography (ΔEELVCT).

nts. Continuous black lines are quadrant limits. Dotted line is the regression line. Dashed line is the line of identity. Each data point is the change in end-expiratory lung volume (EELV) between consecutive measurements assessed with the nitrogen washin-washout technique (ΔEELVWI-WO) or computed tomography (ΔEELVCT). Table 2 Concordance rate, angular bias, and radial limits of agreement in different data subsets No exclusion zone Exclusion threshold 100 mL Exclusion threshold 150 mL Exclusion threshold 200 mL Exclusion threshold 300 mL Concordance rate 79% 82% 82% 86% 86% Angular bias ± SD (°) -4 ± 37 3 ± 25 6 ± 25 1 ± 26 -1 ± 25 Radial limits of agreement (°) ±78 ±51 ±48 ±48 ±50 Exclusion threshold refers to exclusion of data points with change in end-expiratory lung volume between consecutive time points below or equal to 100, 150, 200, and 300 mL, respectively. SD, standard deviation.

ngular bias ± SD (°) -4 ± 37 3 ± 25 6 ± 25 1 ± 26 -1 ± 25 Radial limits of agreement (°) ±78 ±51 ±48 ±48 ±50 Exclusion threshold refers to exclusion of data points with change in end-expiratory lung volume between consecutive time points below or equal to 100, 150, 200, and 300 mL, respectively. SD, standard deviation. Results of the polar plot analysis are reported in Figure 7, and in Table 2. The angular bias amounted to -4° ± 37° for all measurements. After exclusion of EELV changes ≤100 mL, the angular bias amounted to 3° ± 25°, and was not statistically different from 0. Radial limits of agreement were wide when all measurements were taken into account (±78°), but where narrowed to ±51° after exclusion of EELV changes ≤100 mL. Increasing the exclusion threshold of EELV changes up to 300 mL did not improve the radial limits of agreement (Table 2). Diagnostic performance of EELVWI-WO to detect absolute EELV changes greater than 100, 150, 200, 250, and 300 mL is presented in Table 3. Diagnosis accuracy was fair for detection of absolute EELV changes above 200 mL (AUC 0.79 (CI 95% 0.70 to 0.89)), and good for detection of absolute EELV changes above 300 mL (AUC 0.89 (CI 95% 0.83 to 0.95)).Figure 7 Polar plots assessing trending ability of EELV WI-WO to track changes in EELV. Panel A refers to the whole set of measurements, and panel B is restricted to data related to changes in EELV greater than 100 mL since a small change in EELV does not reflect trending ability but mainly random error measurement. The radial axis joining 0 to 180° is a 45° clockwise rotation of the line of identity in the four-quadrant plots, and represents agreement. The better the agreement between ΔEELV measurements, the closer data pairs will lie along the horizontal radial axis. The distance from the center of each plot represents the mean change in EELV between methods (mean ΔEELV) at each consecutive time point. Data points located between 315° and 45° refer to time points in which both EELVCT and EELVWI-WO increased (upper right quadrant of the four quadrant plot), while data points located between 135° and 225° refer to consecutive time points in which both EELVCT and EELVWI-WO decreased (lower left quadrant of the four quadrant plot). Data points located between 45 and 135° or 225 and 315° correspond to disagreement in the directional change of EELV between the washin-washout technique and computed tomography.

135° and 225° refer to consecutive time points in which both EELVCT and EELVWI-WO decreased (lower left quadrant of the four quadrant plot). Data points located between 45 and 135° or 225 and 315° correspond to disagreement in the directional change of EELV between the washin-washout technique and computed tomography. Continuous line represents the angular bias, while dashed lines represent radial limits of agreement. EELVWI-WO, end-expiratory lung volume assessed with the nitrogen washin-washout technique; EELVCT, end-expiratory lung volume assessed by computed tomography; ΔEELV, change in EELV between consecutive measurements. Table 3 Diagnostic performance of EELV WI-WO to detect variations in ELLV CT at different thresholds ΔEELVCT threshold for AUC computation (mL) AUC (CI 95%) Optimal ΔEELVWI-WO cut-off (mL) Se Sp PPV NPV PLR NLR Youden index 100 0.58 (0.50 to 0.66) 42 0.84 0.33 0.57 0.66 1.25 0.48 0.17 150 0.73 (0.64 to 0.81) 166 0.65 0.75 0.46 0.87 2.61 0.47 0.40 200 0.79 (0.70 to 0.89) 166 0.80 0.75 0.42 0.94 3.26 0.27 0.55 250 0.87 (0.79 to 0.94) 169 0.90 0.77 0.41 0.97 3.89 0.13 0.67 300 0.89 (0.83 to 0.95) 169 0.93 0.77 0.41 0.98 4.05 0.09 0.70

index 100 0.58 (0.50 to 0.66) 42 0.84 0.33 0.57 0.66 1.25 0.48 0.17 150 0.73 (0.64 to 0.81) 166 0.65 0.75 0.46 0.87 2.61 0.47 0.40 200 0.79 (0.70 to 0.89) 166 0.80 0.75 0.42 0.94 3.26 0.27 0.55 250 0.87 (0.79 to 0.94) 169 0.90 0.77 0.41 0.97 3.89 0.13 0.67 300 0.89 (0.83 to 0.95) 169 0.93 0.77 0.41 0.98 4.05 0.09 0.70 AUC, area under receiver operating characteristic curve; CI 95%, 95% confidence interval; ΔEELV CT, absolute change in end-expiratory lung volume assessed by computed tomography between consecutive time points; ΔEELV WI-WO , absolute change in end-expiratory lung volume assessed with the nitrogen washin-washout technique between consecutive time points; NLR, negative likehood ratio; NPV, negative predictive value; PLR, positive likehood ratio; PPV, positive predictive value; Se, sensitivity, Sp, specificity. Discussion The main findings of the present study are that, in experimental ARDS, (1) EELVWI-WO underestimates EELVCT, and this underestimation increases linearly as EELV increases; (2) this underestimation is dependent on ventilatory settings (mainly PEEP); (3) the precision of this technique is poor with a percentage error as high as 57%; (4) this technique is however reliable to detect an EELV change greater than 200 mL.

tes EELVCT, and this underestimation increases linearly as EELV increases; (2) this underestimation is dependent on ventilatory settings (mainly PEEP); (3) the precision of this technique is poor with a percentage error as high as 57%; (4) this technique is however reliable to detect an EELV change greater than 200 mL. A formal comparison between EELVWI-WO EELVCT has already been performed in a pig model of pleural effusion [19] and in mechanically ventilated patients [2]. While Chiumello et al. found a slightly positive constant bias between EELVWI-WO and EELVCT[2], and Graf et al. a slightly negative constant bias, a non-constant linear bias was found in the present study. Beside differences in species, experimental protocol or mechanism of lung injury, the likely explanation of this discrepancy is related to the higher PEEP level applied in the present study (up to 20 cm H2O), while PEEP was set to 5 cm H2O in the human study [2] and ≤10 cm H2O in the animal study [19]. The dependence of bias on PEEP level was further emphasized in our study by multivariate analysis (Table 1, Figure 4), more specifically for PEEP levels greater than 10 cm H2O. Finally, a detrimental effect of PEEP levels greater than 16 cm H2O on EELVWI-WO accuracy was also recently pointed out by Dellamonica et al. [23]. Nevertheless, in spite of a non-constant bias, the true EELV value may still be assessed by the WI-WO technique using Equation 1, as pointed out by Bland and Altman [24].

. Finally, a detrimental effect of PEEP levels greater than 16 cm H2O on EELVWI-WO accuracy was also recently pointed out by Dellamonica et al. [23]. Nevertheless, in spite of a non-constant bias, the true EELV value may still be assessed by the WI-WO technique using Equation 1, as pointed out by Bland and Altman [24]. The detrimental effect of high PEEP levels on the bias may have multiple explanations. First, the WI-WO technique measures volume of lung regions that may be reached by nitrogen (and hence are ventilated), while CT measures aeration of both ventilated and non-ventilated regions. One could then hypothesize that regional ventilation (and hence nitrogen during EELV measurement) at high PEEP may be preferentially directed toward non-overinflated regions and that overinflated regions may not be detected in the EELV measurement. However, this explanation is not supported by our data, since the amount of overinflated area was very low even at the highest PEEP level in our study (1.1 ± 0.89 mL). Nevertheless, the observed difference in EELV between methods at high PEEP may be related to the fact that the WI-WO method measures a functional EELV, while CT measures anatomical EELV. Occurrence of leaks at high PEEP is another hypothesis to explain the non-constant bias, but this explanation is unlikely since plateau pressure was maintained during the prolonged end-inspiratory pauses performed for CT acquisition. Finally, there may be insufficient nitrogen mixing within the aerated lung, during the time allocated for measurement by the ventilator, since animals were ventilated at high RR and low VT, which may have increased the time to reach equilibrium during the WI-WO measurement. An extended time between experimental stages and/or a lengthened WI-WO period may have narrowed the bias between methods.

during the time allocated for measurement by the ventilator, since animals were ventilated at high RR and low VT, which may have increased the time to reach equilibrium during the WI-WO measurement. An extended time between experimental stages and/or a lengthened WI-WO period may have narrowed the bias between methods. Regarding precision, comparison between aforementioned studies, using limits of agreement is hindered by heterogeneity of EELV values across studies. Assessment of percentage error may overcome this problem, but was unavailable in the two previously published studies [2, 19]. Using Data Thief 3.0 to extract Cartesian data from these studies, we have computed from the author’s published figures, the percentage error which amounted to 28% in the human study and 46% in the pig study, versus 57% in our study. The relatively lower precision computed from our data may be a consequence of ventilatory settings, that may have particularly challenged the validity of the WI-WO technique. Indeed, FiO2 greater than 0.7 precludes the computation of the respiratory quotient (RQ) required for EELV measurements, and a default RQ of 0.85 is assumed by the Engström Carestation® ventilator.

a may be a consequence of ventilatory settings, that may have particularly challenged the validity of the WI-WO technique. Indeed, FiO2 greater than 0.7 precludes the computation of the respiratory quotient (RQ) required for EELV measurements, and a default RQ of 0.85 is assumed by the Engström Carestation® ventilator. However, using a metabolically active lung model, Olegard et al. have nicely demonstrated that errors in RQ computation have a negligible effect on the precision of EELV measurements [1], suggesting that the high FiO2 used in our study may have only marginally influenced our results. Regarding trending, the lack of significant angular bias suggests that calibration of EELVWI-WO is in agreement with the reference method. However, the relatively wide radial limits of agreement suggest that external factors may account for the variability of the relationship between ΔEELVWI-WO and ΔEELVCT. We could speculate that this phenomenon is mainly related to the effect of PEEP on the bias between EELVWI-WO and ΔEELVCT.

with the reference method. However, the relatively wide radial limits of agreement suggest that external factors may account for the variability of the relationship between ΔEELVWI-WO and ΔEELVCT. We could speculate that this phenomenon is mainly related to the effect of PEEP on the bias between EELVWI-WO and ΔEELVCT. Our study has several strengths. Since multiple combinations of PEEP and VT were evaluated, a systematic analysis of the effect of ventilator parameters on the reliability of the technique could be performed using multivariate analysis, and was able to identify the PEEP level as an independent risk factor for measurement error. Furthermore, this is the first study having assessed trending ability of the WI-WO technique, and provided with cut-off values above which EELV changes may be considered as meaningful. Finally, the present study, while performed with particularly challenging ventilatory settings (high FiO2, high PEEP, high RR, and low VT), demonstrates that the validity of the WI-WO technique may be extended to the sickest ARDS patients.

ed with cut-off values above which EELV changes may be considered as meaningful. Finally, the present study, while performed with particularly challenging ventilatory settings (high FiO2, high PEEP, high RR, and low VT), demonstrates that the validity of the WI-WO technique may be extended to the sickest ARDS patients. Our study has nevertheless some methodological issues that must be addressed. First, it was conducted using pediatric sensors for EELVWI-WO measurements, which may be less accurate for the highest VT. Nevertheless, a subset analysis after exclusion of VT > 300 mL led to similar results (Additional file 3: Figure S1). Furthermore, the bias was not increased at high VT, as compared to low VT, as shown by our interaction plot (Figure 4). Another limitation, in the perspective of extrapolation of these results to ARDS patients, is related to the relatively low EELV achieved in the pigs of this study in some experimental conditions. However, 62% of the EELVCT measurements were greater than the first EELV quartile observed at low PEEP in a recent study on 30 ARDS patients [23], suggesting that most of the measurements performed in our study are in the range of clinically plausible values for EELV in ARDS patients. Another potential limitation is that the ARDS model used in the present study is particularly recruitable with PEEP. However, 50% of ARDS patients are considered recruiters by PEEP [23, 25], and early ARDS share similar features as saline lavage regarding response to PEEP. The 2-min interval between VT changes and EELV measurements may have been too short for CO2 equilibration and achievement of both progressive recruitment and blood flow redistribution. However, a subset analysis limited to data acquired during the PEEP trial (with 10 min between measurements) led to similar results (see Additional file 4: Figure S2).

anges and EELV measurements may have been too short for CO2 equilibration and achievement of both progressive recruitment and blood flow redistribution. However, a subset analysis limited to data acquired during the PEEP trial (with 10 min between measurements) led to similar results (see Additional file 4: Figure S2). The present study may have important clinical implication. Indeed, as shown in Table 3, a change in EELVWI-WO greater than 166 mL would give an important clue to the clinician that the true EELV has changed by more than 200 mL. Furthermore, despite the non-constant bias of the EELV measurement by the WI-WO technique, the true EELV value may still be assessed using Equation 1, provided that the absolute EELV value is relevant for the clinician [24]. Conclusion The reliability of the WI-WO technique is critically dependent on ventilatory settings, but sufficient to accurately detect EELV change over time greater than 200 mL. Electronic supplementary material Additional file 1: Table S1: Reasons for lack of data as a function of each of the 3 experimental stages. Values are number of lacking data/total number of data (%). EELVWI-WO, end-expiratory lung volume assessed with the nitrogen washout-washin technique; EELVCT, end-expiratory lung volume assessed by computed tomography; V T, tidal volume. (DOCX 15 KB)

for lack of data as a function of each of the 3 experimental stages. Values are number of lacking data/total number of data (%). EELVWI-WO, end-expiratory lung volume assessed with the nitrogen washout-washin technique; EELVCT, end-expiratory lung volume assessed by computed tomography; V T, tidal volume. (DOCX 15 KB) Additional file 2: Table S2: Ventilatory settings and arterial blood gases in each experimental condition. Values are number of mean ± standard deviation (range). ALI, acute lung injury onset; PEEP, positive end-expiratory pressure; RR, respiratory rate; V T, tidal volume. (DOCX 17 KB) Additional file 3: Figure S1: Bias and limits of agreement between EELVCT and EELVWI-WO, using Bland and Altman representation in a subset of the data (exclusion of V T > 300 mL). Each symbol represents a concomitant measurement of EELVWI-WO and EELVCT. Horizontal continuous line and horizontal broken lines are the mean bias and 95% prediction interval limits of the bias between EELVWI-WO and EELVCT, respectively. EELVWI-WO, end-expiratory lung volume assessed with the nitrogen washout-washin technique; EELVCT, end-expiratory lung volume assessed by computed tomography; 95% p.i., 95% prediction interval of the bias between EELVWI-WO and EELVCT; V T, tidal volume. (DOCX 278 KB)

the bias between EELVWI-WO and EELVCT, respectively. EELVWI-WO, end-expiratory lung volume assessed with the nitrogen washout-washin technique; EELVCT, end-expiratory lung volume assessed by computed tomography; 95% p.i., 95% prediction interval of the bias between EELVWI-WO and EELVCT; V T, tidal volume. (DOCX 278 KB) Additional file 4: Figure S2: Bias and limits of agreement between EELVCT and EELVWI-WO, using Bland and Altman representation in a subset of the data, acquired during the PEEP trial (with 10 min between measurements). Each symbol represents a concomitant measurement of EELVWI-WO and EELVCT. Horizontal continuous line and horizontal broken lines are the mean bias and 95% prediction interval limits of the bias between EELVWI-WO and EELVCT, respectively. EELVWI-WO, end-expiratory lung volume assessed with the nitrogen washin-washout technique; EELVCT, end-expiratory lung volume assessed by computed tomography; 95% p.i., 95% prediction interval of the bias between EELVWI-WO and EELVCT; V T, tidal volume. (DOCX 137 KB) Abbreviations ARDSacute respiratory distress syndrome AUCarea under curve ΔEELVWI-WOchange in end-expiratory lung volume between consecutive measurements by the nitrogen washin-washout technique ΔEELVCTchange in end-expiratory lung volume between consecutive measurements by computed tomography EELVend-expiratory lung volume EELVWI-WOend-expiratory lung volume measurement by the nitrogen washin-washout technique EELVCTend-expiratory lung volume measurement by computed tomography FiO2fraction of inspired oxygen PEEPpositive end-expiratory pressure RRrespiratory rate

ΔEELVCTchange in end-expiratory lung volume between consecutive measurements by computed tomography EELVend-expiratory lung volume EELVWI-WOend-expiratory lung volume measurement by the nitrogen washin-washout technique EELVCTend-expiratory lung volume measurement by computed tomography FiO2fraction of inspired oxygen PEEPpositive end-expiratory pressure RRrespiratory rate SDstandard deviation VTtidal volume WI-WOnitrogen washin-washout Competing interests JCR, CP, AMP, JST, MO, BN, MHH, and FL have no competing interests. CG received a grant of 9600 € by GE Healthcare as study funding. CG has no other ties with this company, and the manuscript was submitted without review by this company. Authors’ contributions JCR and CG have made substantial contributions to the conception, design, acquisition of data, and analysis and interpretation of data; AMP, JST, MO, BN, and MHH to the analysis and interpretation of data; MO to the conception, design, and analysis and interpretation of data; CP to the conception, acquisition of data, and analysis and interpretation of data; and FL to the acquisition of data. JCR has drafted the submitted article. CP, AMP, JST, MO, BN, MHH, FL, and CG have revised the draft critically for important intellectual content. All authors have provided the final approval of the version to be published; and agreed to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.

draft critically for important intellectual content. All authors have provided the final approval of the version to be published; and agreed to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. Role of the sponsors The study was founded by General Electric who provided with the ventilator and disposables, technical assistance, and funded the study (9,600 €). The investigators had no ties with this company. The study was also funded by grant no. C11S01 from the French-Colombian program ECOS-Nord. These sponsors had no role in the design, collection, analysis, and interpretation of data; in the writing of the manuscript; and in the decision to submit the manuscript for publication. AMP was supported by a Colombian doctoral grant from Colciencias and by a Rhône-Alpes Region grant CMIRA.

Background Despite improvements in initial resuscitation and supportive care, septic shock remains the major cause of mortality in intensive care units (ICU) worldwide [1]. The failure of 15 years of clinical trials assessing adjunctive therapies together with recent research advances in pathophysiology [2,3] suggest the need to better define the disease and stratify patients allowing for appropriate treatments and targeted therapy. Success in septic shock treatment will rely on a more mediator-directed therapy and thus will require specific and sensitive monitoring tools. Many studies have aimed at dissecting mechanisms implicated in the disease, and it is now well accepted that injuries in ICU patients lead not only to the initiation of an uncontrolled and exacerbated pro-inflammatory response but also to a deep activation of anti-inflammatory processes [4]. However, the exact chronology of this process remains elusive [5]. It is well accepted that both systemic inflammatory response syndrome (SIRS) and compensatory anti-inflammatory response syndrome (CARS) occur [6,7]. They may be responsible for multiple-organ dysfunction syndrome (MODS) and cell reprograming/immunosuppression [8,9]. From a clinical perspective, several studies have highlighted that the first hours are critical for the response to injury and consequently for the patient’s care and outcome [10–12]. This enforces the importance of characterizing patients early in the process of the disease.

cell reprograming/immunosuppression [8,9]. From a clinical perspective, several studies have highlighted that the first hours are critical for the response to injury and consequently for the patient’s care and outcome [10–12]. This enforces the importance of characterizing patients early in the process of the disease. Microarray-based expression profiling provides an interesting opportunity to gain a broader genome level ‘picture’ of a complex and heterogeneous clinical syndromes such as septic shock [13]. These powerful genomic approaches are recommended by opinion leaders to identify specific pathways that could be targeted by different drugs depending on patients’ subset [3]. It is exemplified by a recent genomic description in trauma patients [14], showing that mRNA genes of both immune activation and immune suppression were concomitantly expressed early after injury. As trauma and septic shock share similarities in pathophysiology and as the understanding of the global host response in sepsis would have a major impact on the successful use of targeted therapies [3], the main objective of the study was to analyze the genome-wide expression patterns of blood leukocytes sequentially during the first 48 h after septic shock. We also investigated whether patterns of gene expression within the first 48 h after septic shock were different in the two extremes of clinical severity using a stratification based on the Simplified Acute Physiology Score II (SAPSII).

patterns of blood leukocytes sequentially during the first 48 h after septic shock. We also investigated whether patterns of gene expression within the first 48 h after septic shock were different in the two extremes of clinical severity using a stratification based on the Simplified Acute Physiology Score II (SAPSII). Methods Patients Twenty-eight patients, 18 years and above, admitted to two ICUs of a university hospital for septic shock were included. Their clinical characteristics are shown in Table 1. The diagnosis of septic shock was based on the ACCP/SCCM criteria [15].Table 1 Patients’ characteristics at admission SAPSII-low n = 14 (%) SAPSII-high n = 14 (%) p value Total n = 28 (%)

Methods Patients Twenty-eight patients, 18 years and above, admitted to two ICUs of a university hospital for septic shock were included. Their clinical characteristics are shown in Table 1. The diagnosis of septic shock was based on the ACCP/SCCM criteria [15].Table 1 Patients’ characteristics at admission SAPSII-low n = 14 (%) SAPSII-high n = 14 (%) p value Total n = 28 (%) Female n (%) 4 (28.57) 5 (35.72) ns 9 (32.1) Male n (%) 10 (71.43) 9 (64.28) ns 19 (67.9) Age median (Q1-Q3) 59 (46-69) 74 (58-79) ns 62 (54-76) Non-survivor in 28 days n (%) 1 (7) 4 (28.5) ns 5 (17.9) Charlson median (Q1-Q3) 2 (0-2.8) 2.5 (1.3-3.8) ns 2 (0.75-3.25) SAPSII on admission median (Q1-Q3) 34 (29-40) 56 (49-63) <0.0001 45 (34-56) Duration length in ICU median (Q1-Q3) 10 (5-11) 11 (6-30) ns 10 (5-14) SOFA H6 11 (9-13) 10 (9-13) ns 10 (9-13) Comorbidity n (%) ns 0 9 (64.3) 7 (50) 16 (57.1) 1 4 (28.6) 5 (35.7) 9 (32.1) >2 1 (7.1) 2 (14.3) 3 (10.7) Type of admission n (%) ns Surgery 5 (35.7) 8 (57.1) 13 (46.4) Medical 9 (64.3) 6 (42.9) 15 (53.6) Type of infection n (%) ns Community acquired 6 (42.86) 9 (64.29) 15 (53.5) Hospital acquired 8 (57.14) 5 (35.71) 13 (46.4) Suspected infection n (%) Clinically documented diagnosis 1 (7) 3 (21.4) 4 (16.6) Microbiologically documented diagnosis 13 (93) 11 (79) 24 (86) Bacilli Gram (−) 10 (77) 7 (64) ns 17 (61) Cocci Gram (+) 5 (38) 7 (64) ns 12 (43) Fungi 2 (15) 0 2 (7) Cell count White blood cells (giga/L) 13.16 (7.52-16.93) 6.97 (3.23-15.68) ns 11.07 (5.9-16.3) Lymphocytes 0.73 (0.32-1.22) 0.68 (0.4-1.33) ns 0.72 (0.35-1.3) Polymorphonuclear cells 10.14 (6.6-14.31) 5.45 (2.5-12.61) ns 9.56 (4.21-13.12) Monocytes 0.57 (0.49-1.66) 0.45 (0.19-0.66) ns 0.55 (0.38-0.67) ns, not significant. Significance was designated at the p < 0.05 level of confidence. Values represent effectives and percentage indicated in parentheses or median and first and third quartile indicated in parentheses.

5.45 (2.5-12.61) ns 9.56 (4.21-13.12) Monocytes 0.57 (0.49-1.66) 0.45 (0.19-0.66) ns 0.55 (0.38-0.67) ns, not significant. Significance was designated at the p < 0.05 level of confidence. Values represent effectives and percentage indicated in parentheses or median and first and third quartile indicated in parentheses. Sepsis is defined as the combination of a SIRS and infection diagnosed macroscopically and microbiologically. SIRS is defined as a clinical situation involving at least two of the following clinical criteria: hypothermia (<36°C) or hyperthermia (>38°C), tachycardia (>90/min), tachypnea (>20 breaths/min) and/or arterial PCO2 of 32 mmHg or lower and/or mechanical ventilation, and leukocytosis (>12,000/mm3) or leukopenia (<4,000/mm3). Septic shock was defined as acute circulatory failure (systolic blood pressure <90 mmHg, mean arterial pressure <65 mmHg, or a reduction in systolic blood pressure >40 mmHg from baseline) despite adequate volume resuscitation. In order to limit potential confounding effects of other conditions affecting immunity, patients with one or more severe comorbidities (i.e., the human immunodeficiency syndrome, hematologic malignancies evolving from insulin-dependent diabetes, the dialyzed chronic renal failure, chronic liver disease stages III and more) or patients receiving immunosuppressive therapy were excluded.

affecting immunity, patients with one or more severe comorbidities (i.e., the human immunodeficiency syndrome, hematologic malignancies evolving from insulin-dependent diabetes, the dialyzed chronic renal failure, chronic liver disease stages III and more) or patients receiving immunosuppressive therapy were excluded. The onset of the septic shock was defined as the beginning of vasopressor therapy. All patients were treated similarly according to the standardized recommendations of our ICU. Severity was assessed using the SAPSII [16]. The first blood sample was collected at the onset of shock (i.e., within 30 min after the beginning of vasoactive treatment: 0, 24, and 48 h after (H0, H24, and H48)). Septic shock patients were divided into SAPSII-low and SAPSII-high groups according to the median of SAPSII score (SAPSII scores of <45 and >45, respectively). Twenty-five healthy volunteers (median age [Q1-Q3], 48 [40-52] years) with no known comorbidities were also included to provide a panel of control values for mRNA expression. Protocol was approved by Comité consultatif de Protection de Personnes (CPP) de Lyon A, and informed consent forms were signed by patients (or by a third party based on the patient’s state of consciousness).

Twenty-five healthy volunteers (median age [Q1-Q3], 48 [40-52] years) with no known comorbidities were also included to provide a panel of control values for mRNA expression. Protocol was approved by Comité consultatif de Protection de Personnes (CPP) de Lyon A, and informed consent forms were signed by patients (or by a third party based on the patient’s state of consciousness). Sample collection, processing, and microarray hybridization Peripheral blood samples were collected in PAXgene™ Blood RNA tubes (PreAnalytix, Hombrechtikon, Switzerland) in order to stabilize mRNA [17]. Total RNA was extracted according to the manufacturer’s instructions. Briefly, total RNA was isolated using the PAXgeneTM blood RNA kit (PreAnalytix). The residual genomic DNA was digested using the RNase-Free DNase set (Qiagen Valencia, CA, USA). The integrity of the total RNA was assessed using Agilent 2100 Bioanalyzer (Agilent Technologies, Santa Clara, CA, USA). The 104 microarray experiments (corresponding to the three time points of the 28 septic shock and the 25 healthy volunteers) were performed as previously described [18]. Briefly, gene expressions were generated using GeneChip® Human Genome U133 Plus 2.0 arrays (Affymetrix, Sta. Clara, CA, USA) according to manufacturer’s protocol. Affymetrix GeneChip Operating Software version 1.4 (GCOS) was used to manage GeneChip array data and to automate the control of GeneChip fluidics stations (FS450) and scanner (GeneChip® Scanner 3000). Data from this experiment have been deposited in the National Center for Biotechnology Information (NCBI) and are available in the GEO DataSets site under accession number GSE57065.

used to manage GeneChip array data and to automate the control of GeneChip fluidics stations (FS450) and scanner (GeneChip® Scanner 3000). Data from this experiment have been deposited in the National Center for Biotechnology Information (NCBI) and are available in the GEO DataSets site under accession number GSE57065. Reverse transcription and quantitative PCR Total RNA was reverse transcribed into cDNA using the SuperScript III reverse transcription-polymerase chain reaction (PCR) system (Invitrogen Life Technologies, Carlsbad, CA, USA) according to the manufacturer’s instructions. mRNA expression was quantified using quantitative real-time polymerase chain reaction. Briefly, PCR reactions were performed using a LightCycler instrument with the Fast-Start DNA Master SYBR Green I real-time PCR kit according to the manufacturer’s instructions (Roche, Basel, Switzerland). Thermocycling was performed in a final volume of 20 μL containing 3 mM of MgCl2 and 0.5 μM each of the required primers. PCR was performed with an initial denaturation step of 10 min at 95°C followed by 40 cycles of a touchdown PCR protocol (10 s at 95°C, 10 s annealing at 68°C to 58°C, and 16 s extension at 72°C). mRNA expression of genes was investigated using specific cDNA standards. The cDNA standard was prepared from purified PCR amplicons obtained for each candidate genes: TBX21 (FP: 5′-TGTGACCCAGATGATTGTGCT-3′, RP: 5′-AGCTGAGTAATCTCGGCATTC-3′), GATA3 (FP: 5′-AAGCGAAGGCTGTCTGCAGC-3′, RP: 5′-GGGTCTGTTAATATTGTGAAGC-3′), CX3CR1 (FP: 5′-AGTCTGAGCAGGACAGGGTG-3′, RP: 5′-GTCCCAAAGACCACGATGTCC-3′), HLA-DRA (FP: 5′-GCCTCTTCTCAAGCACTGGGA-3′, RP: 5′-CCACCAGACCCACAGTCAGG-3′), IRAK-3 (FP: 5′-CTCGGAATTTCTCTGCCAAG-3′, RP: 5′-GTGGGAGGATCTTCAGCAAA-3′), and for the housekeeping gene peptidylpropyl isomerase B encoding for cyclophilin B; PPIB (FP: 5′-GGAGATGGCACAGGAGGAAAGA-3′, RP: 5′-GGGAGCCGTTGGTGTCTTTG-3′). The second derivative maximum method was used with the LightCycler software to automatically determine the crossing point for individual samples. Standard curves were generated by quadruplicate cDNA standard. Relative standard curves describing the PCR efficiency of selected genes were created and used to perform efficiency-corrected quantification with the LightCycler Relative Quantification Software (Roche Molecular Biochemicals). The results were expressed as a concentration ratio between the target gene mRNA and peptidylpropyl isomerase B mRNA levels.

ing the PCR efficiency of selected genes were created and used to perform efficiency-corrected quantification with the LightCycler Relative Quantification Software (Roche Molecular Biochemicals). The results were expressed as a concentration ratio between the target gene mRNA and peptidylpropyl isomerase B mRNA levels. Data analysis Univariate analysis was performed to compare characteristics between groups using either the chi-squared or Mann-Whitney U test.

ing the PCR efficiency of selected genes were created and used to perform efficiency-corrected quantification with the LightCycler Relative Quantification Software (Roche Molecular Biochemicals). The results were expressed as a concentration ratio between the target gene mRNA and peptidylpropyl isomerase B mRNA levels. Data analysis Univariate analysis was performed to compare characteristics between groups using either the chi-squared or Mann-Whitney U test. Gene expression data were imported into Partek Genomics Suite 6.5 (Partek, St Louis, MO, USA) as .CEL files using default parameters. Transcriptomic data were normalized with gc-Robust Multi-array Average (gcRMA) algorithm. The RMA method [19] consists of three steps: background adjustment, quantile normalization [20], and probe set summarization of the log-normalized data applying a median polishing procedure. Differential expression analysis was performed using analysis of variance (ANOVA). A step-up false discovery rate (FDR) was applied to p values from the linear contrasts to determine a cutoff for significantly differentially expressed genes. Gene lists were created using cutoff of FDR <0.05 and twofold change. Hierarchical clustering was performed using the gene expression module from Partek. Euclidian distance method after normalization by shift mean columns to mean of zero and scale to standard deviation of 1 was used. Gene ontology, functional enrichment, and canonical pathways analyses were performed using Ingenuity Pathway Analysis (IPA) [12,21] (www.ingenuity.com). Fisher’s exact test was used to calculate the p value for determining the probability that each function or pathway assigned to the dataset was due to chance alone. The Human Genome U133 Plus 2.0 array was used as the reference.

analyses were performed using Ingenuity Pathway Analysis (IPA) [12,21] (www.ingenuity.com). Fisher’s exact test was used to calculate the p value for determining the probability that each function or pathway assigned to the dataset was due to chance alone. The Human Genome U133 Plus 2.0 array was used as the reference. Results Patients’ clinical characteristics The patient’s characteristics at admission are presented in Table 1. The age and sex distribution was similar to what is usually observed in septic shock patients’ cohorts, with a percentage of male patients (63.3%) higher than female and a median age of 62 years. The SAPSII and the Sequential Organ Failure Assessment (SOFA) scores were high (median [Q1-Q3]: 45 [34-56] and 11 [9–13] respectively), although the mortality rate was lower (18%) than usually described in the literature (>30%) for such severe patients [22].

ents (63.3%) higher than female and a median age of 62 years. The SAPSII and the Sequential Organ Failure Assessment (SOFA) scores were high (median [Q1-Q3]: 45 [34-56] and 11 [9–13] respectively), although the mortality rate was lower (18%) than usually described in the literature (>30%) for such severe patients [22]. Gene expression patterns in septic shock patients over time Our microarray data showed clearly that septic shock patients developed major genomic alterations during the first 48 h after the onset of shock affecting more than 71% of the human genome. As shown in Figure 1A, most of these alterations were already detectable within the first 30 min following admission since more than 60% of the human genome was already altered at H0.Figure 1 Genes differentially expressed (GDE) between septic shock patients and healthy volunteers. (A) Results summary table representing the percentage of modulated genes using a false discovery rate (FDR) adjusted probability <0.05 over the 48 h compared to the human genome (Entrez genes included in HG-U133 plus 2.0 microarray). The table also shows the percentage of significant (FDR <0.05 and FC >2) down- and upregulated genes. (B) Venn diagram representing the GDE between septic shock and healthy volunteers using a false discovery rate (FDR) adjusted probability <0.05. Values represent the number of unique Entrez genes found significantly changed in whole blood of patients at the onset of shock (H0) and 24 and 48 h after (H24 and H48).

enes. (B) Venn diagram representing the GDE between septic shock and healthy volunteers using a false discovery rate (FDR) adjusted probability <0.05. Values represent the number of unique Entrez genes found significantly changed in whole blood of patients at the onset of shock (H0) and 24 and 48 h after (H24 and H48). In addition, the transcriptome disruptions observed at H0 appeared to be particularly stable during the study period (Figure 1B). Indeed, over a total of 14,341 unique Entrez genes differentially expressed (FDR adjusted probability <0.05), 8,885 were in common between the three time points (i.e., H0, H24, and H48). While the mRNA expression of most of these genes was already altered at H0, their number seemed to decrease slowly up to H48, indicating that most of the phenomenon had already occurred at admission. When compared to healthy volunteers, the percentage of downregulated genes at H0 (56.1%) was higher than the percentage of upregulated ones (43.9%). The percentage of downregulated genes trended toward a decrease, while the percentage of upregulated ones increased over the first 48 h (53.7% and 47.8% of the downregulated genes at H24 and H48, respectively).

e percentage of downregulated genes at H0 (56.1%) was higher than the percentage of upregulated ones (43.9%). The percentage of downregulated genes trended toward a decrease, while the percentage of upregulated ones increased over the first 48 h (53.7% and 47.8% of the downregulated genes at H24 and H48, respectively). Biological functions and pathways involved in septic shock response The gene ontology analysis performed on 8,885 genes commonly modulated at the three time points showed an enrichment of infectious disease processes (Table 2). In addition, most of the molecular functions deregulated during the first 48 h were associated with cellular rearrangement.Table 2 Top biological functions enriched by genomic variations in septic shock patients Name p value Number of genes Percentage in tGDE Percentage in function

Biological functions and pathways involved in septic shock response The gene ontology analysis performed on 8,885 genes commonly modulated at the three time points showed an enrichment of infectious disease processes (Table 2). In addition, most of the molecular functions deregulated during the first 48 h were associated with cellular rearrangement.Table 2 Top biological functions enriched by genomic variations in septic shock patients Name p value Number of genes Percentage in tGDE Percentage in function Diseases and disorders Infectious disease 2.70E−15 to 3.60E−03 1,097 12.3 32.5 Respiratory disease 9.97E−12 to 4.57E−03 194 2.2 5.7 Organismal injury and abnormalities 2.56E−10 to 5.01E−03 316 3.6 8.6 Dermatological diseases and conditions 5.33E−10 to 3.80E−03 295 3.3 9.9 Renal and urological disease 7.73E−10 to 7.73E−10 206 2.3 8.7 Molecular and cellular functions Cellular development 3.15E−14 to 5.88E−03 933 10.5 15.6 RNA post-transcriptional modification 4.22E−14 to 8.93E−04 196 2.2 45.1 Cell death 3.53E−13 to 6.16E−03 1,880 21.2 30.7 Cellular function and maintenance 8.28E−12 to 5.94E−03 739 8.3 13.4 Cell morphology 1.16E−09 to 4.80E−03 400 4.5 8.9 Physiological system development and function Hematological system development and function 3.88E−15 to 6.17E−03 1,158 13.0 29.7 Hematopoiesis 3.88E−15 to 4.66E−03 681 7.7 38.4 Lymphoid tissue structure and development 5.06E−11 to 5.10E−03 528 5.9 35.4 Cell-mediated immune response 2.16E−10 to 1.80E−03 305 3.4 33.0 Tissue morphology 2.52E−10 to 4.84E−03 628 7.1 14.0 Results are obtained from the 8,885 common genes differentially expressed between septic shock patients and healthy volunteers over time using false discovery rate <0.05. The ‘p value’ column represents the range of significance for the specific functions included in the high level categories described. The p value is a measure of the likelihood that the association between a set of genes in the dataset and a related function is due to random association. p values <0.05 indicate a statistically significant, non-random association. The p value is calculated by the Fisher’s exact test. The ‘percentage in tGDE’ column represents the percentage of genes found in a function compared to the total number of gene differentially expressed (n = 8,885). The ‘percentage in function’ column represents the percentage of genes found in a function compared to the total number of genes described by IPA to be associated with the function.

’ column represents the percentage of genes found in a function compared to the total number of gene differentially expressed (n = 8,885). The ‘percentage in function’ column represents the percentage of genes found in a function compared to the total number of genes described by IPA to be associated with the function. Among the most increased canonical pathways enriched at H0 (Figure 2A), 6 out of 10 were related to molecules involved in the endotoxin tolerance or pathogen recognition as well as cytokines/cytokine receptors, leading to increased inflammation and innate immune response. For example, the expressions of IL1R1, IL1R2, IL1RAP, IL1RN, IL18, IL18RAP, IL4R, IL10, IFNGR1, TGFBR1, IRAK3, MAP2K6, MAPK1, MAPK14, SOCS3, S100A8, MMP9, LY96, JAK2, JAK3, and NFKBIA were significantly increased after shock.Figure 2 Top canonical pathways affected in septic shock. (A and B) The 10 top up- and downregulated canonical pathways enriched by genes differentially expressed over time. The graph shows the −log10 (p value) of the enrichment of the pathway.

MMP9, LY96, JAK2, JAK3, and NFKBIA were significantly increased after shock.Figure 2 Top canonical pathways affected in septic shock. (A and B) The 10 top up- and downregulated canonical pathways enriched by genes differentially expressed over time. The graph shows the −log10 (p value) of the enrichment of the pathway. In parallel, most of the top 10 gene families suppressed after septic shock were related to T cell signaling and antigen-mediated response (Figure 2B). For example, expressions of gene part of the T cell receptor (TCR) complex such as CD247, CD3E, CD3G, and CD3D were markedly decreased, as well as antigen-presentation genes like HLA-DMA, HLA-DOA, HLA-DRA, HLA-DQA1, CIITA, HLA-DRB1, HLA-DOB, HLA-DMB, CD74, HLA-DPB1, TAP2, and HLA-DPA1. Importantly, the transcriptional modulation observed was very consistent over the time period as illustrated by the fold change of these selected genes (Additional file 1: Table S1).

tigen-presentation genes like HLA-DMA, HLA-DOA, HLA-DRA, HLA-DQA1, CIITA, HLA-DRB1, HLA-DOB, HLA-DMB, CD74, HLA-DPB1, TAP2, and HLA-DPA1. Importantly, the transcriptional modulation observed was very consistent over the time period as illustrated by the fold change of these selected genes (Additional file 1: Table S1). Similarly, we looked at the global host response with the predicted activation or inhibition status of these enriched functions (Additional file 2: Table S2). The inflammatory response process was the most enriched category over time, including functions predicted to be inhibited as well as activated (Figure 3A). For example, 52 genes included in the function ‘activation of mononuclear leukocytes’ showed expression direction consistent with the decrease of the function, therefore suggesting an anergy of those cells. In particular, the gene whose expression increased the most was SAMSN1 (fold change 9.04) (Figure 3B). Inversely, the gene whose expression decreased the most was CD247 (i.e., CD3 zeta chain, fold change −5.19). In details, among these 52 genes included in the function activation of mononuclear leukocytes, several could be directly linked to the T cell receptor signaling pathway (CD247, CD28, CD3D, CD3G, CD4, CD8A, FYN, ITK, LCK, NFATC2, NFATC3, PAG1, PRKCQ, and VAV2) or to some activation pathway like glucocorticoid receptor signaling (BCL2, CCL5, CD163, CD247, CD3D, CD3G, IL8, NFATC2, NFATC3, and PRKAA1) or p53 signaling (BCL2, PTEN, THBS1, and TP53). Most of them were transmembrane receptors like CD molecules (CD2, CD3D, CD3G, CD4, CD8A, CD27, CD28, CD47, CD163, and CD247), MHC genes (CD74, HLA-DRA, and HLA-DRB1), or transcription factors (GATA3, NFATC2, NFATC3, STAT4, TBX21, and TP53).Figure 3 Enriched biological category analysis illustrated by inflammatory response function. (A) Inflammatory response functions (high-level functional category) found to be enriched by genes differentially expressed between septic shock patients and healthy volunteers over the 48 h. The z-score value predicts the direction of change for the function. An absolute z-score of ≥2 is considered significant. A function is predicted to be increased if the z-score is ≥2 and decreased if the z-score is less than or equal to −2. Data for the comparison of septic shock patients versus healthy volunteers are represented at the onset of shock (H0) and 24 and 48 h after (H24, H48).

absolute z-score of ≥2 is considered significant. A function is predicted to be increased if the z-score is ≥2 and decreased if the z-score is less than or equal to −2. Data for the comparison of septic shock patients versus healthy volunteers are represented at the onset of shock (H0) and 24 and 48 h after (H24, H48). (B) Highlight on the 52 genes that have expression direction consistent with decreases in activation of mononuclear leukocytes process. Inversely, the function ‘phagocytosis of cells’ was predicted to be activated during the kinetic. The 22 genes involved in this function were found with expressions consistent with its activation, either increased (e.g., PROS1, MERTK, FCGR1A, THBS1, or ITGAM), or decreased (e.g., CD47, FOXP1, or FYN). Five genes selected within these functional categories were validated by qRT-PCR. The gene expressions measured by either microarray or qRT-PCR were highly correlated (Additional file 3: Figure S1).

Inversely, the function ‘phagocytosis of cells’ was predicted to be activated during the kinetic. The 22 genes involved in this function were found with expressions consistent with its activation, either increased (e.g., PROS1, MERTK, FCGR1A, THBS1, or ITGAM), or decreased (e.g., CD47, FOXP1, or FYN). Five genes selected within these functional categories were validated by qRT-PCR. The gene expressions measured by either microarray or qRT-PCR were highly correlated (Additional file 3: Figure S1). Functional analysis of genes in severe versus less severe patients SAPSII is a reliable score for severity assessment in septic patients and is therefore frequently used to stratify patients in clinical studies [23]. In order to investigate whether the two extremes of severity after septic shock could be associated with different genomic responses and overtime evolutions, we used the median SAPSII values to stratify patients in our cohort in severe (SAPSII-high group, SAPSII >45) and less severe (SAPSII-low group, SAPSII <45) groups. The clinical characteristics for the SAPSII-high and SAPSII-low groups are presented in Table 1. A difference, although not significant, was observed in mortality rate between the two severity groups. We also observed a small difference between cell counts at admission, with decreased number of white blood cells in SAPSII-high group at admission (6.97 versus 13.16 in SAPSII-low group). This difference tended to decrease over the time period, with 13.98 and 11.47 giga/L white blood cells at H48 for SAPSII-low and SAPSII-high groups, respectively (data not shown).

l counts at admission, with decreased number of white blood cells in SAPSII-high group at admission (6.97 versus 13.16 in SAPSII-low group). This difference tended to decrease over the time period, with 13.98 and 11.47 giga/L white blood cells at H48 for SAPSII-low and SAPSII-high groups, respectively (data not shown). Interestingly, an equal number of genes was significantly modulated (FDR <0.05; fold change >2) in SAPSII-high and SAPSII-low groups compared to healthy volunteers at the onset of shock (more than 67% of genes differentially expressed were in common - Figure 4B). The highest difference in the response between SAPSII-low and SAPSII-high groups was observed at H48 (1,735 genes differentially expressed in SAPSII-high analysis compared to 1,161 in SAPSII-low group leading to only 44% of commonly regulated genes at this time point) (Figure 4). Although the two severity groups seemed to have the same genomic response at H0, the number of altered genes in SAPSII-low patients decreased over the 48 h study period, returning to control values, whereas it remained stable in the SAPSII-high group even with a trend toward increase (Figure 4C).Figure 4 GDE between SAPSII-high and SAPSII-low patients versus healthy volunteers. (A) Venn diagram representing the GDE using FDR-adjusted probability <0.05. Values represent the number of unique Entrez genes found significantly changed in whole blood of SAPSII-high and SAPSII-low groups at the onset of shock (H0) and 24 and 48 h after (H24 and H48) compared to healthy volunteers. (B) Results summary table representing the percentage of common genes in both groups over the 48 h. (C) Number of SAPSII-low and SAPSII-high significant upregulated genes (FDR <0.05, fold change >2) compared to healthy volunteers at H0, H24, and H48.

) and 24 and 48 h after (H24 and H48) compared to healthy volunteers. (B) Results summary table representing the percentage of common genes in both groups over the 48 h. (C) Number of SAPSII-low and SAPSII-high significant upregulated genes (FDR <0.05, fold change >2) compared to healthy volunteers at H0, H24, and H48. A functional analysis showed that the number of significantly enriched inhibited functions was much higher than the number of activated ones (≈50 versus 15) (Figure 5A,B). In the SAPSII-low patients group, a steady decrease in the number of statistically enriched functions was observed (both inhibited and activated), while the number of activated functions remained quite stable in the SAPSII-high patients group (from 13 at H0 to 10 at H48) and the number of inhibited functions increased over time (50 at H0 to 70 at H48).Figure 5 Enriched functions of SAPSII-low and SAPSII-high patients compared to healthy volunteers. (A and B) The number of SAPSII-high and SAPSII-low significantly decreased and increased functions (absolute z-score ≥2) compared to healthy volunteers at H0, H24, and H48. (C) Table of principal enriched functions in SAPSII-low and SAPSII-high groups compared to healthy volunteers (HV) at the onset of shock 0, 24, and 48 h after. For each category, the total number and the number of enriched functions predicted to be activated or inhibited are indicated.

y volunteers at H0, H24, and H48. (C) Table of principal enriched functions in SAPSII-low and SAPSII-high groups compared to healthy volunteers (HV) at the onset of shock 0, 24, and 48 h after. For each category, the total number and the number of enriched functions predicted to be activated or inhibited are indicated. The same functions were enriched in both groups reflecting that the overall changes (direction of the responses) in gene expression at H0 between the two severity groups was very similar (Figure 5C). For example, cellular and metabolic functions (e.g., cellular development or hematological systems) were found significantly enriched and mostly inhibited in both groups. Conversely, some differences were observed in the predicted activation state of functions. For example at H0, in the ‘cell death and viability’ category, a number of functions were predicted to be activated in SAPSII-low group whereas these were predicted to be inhibited in SAPSII-high group. Most of the activated functions were related to cell death and apoptosis, trending toward a decrease during the kinetic in SAPSII-low group. Therefore, the predicted z-score for cell death of mononuclear leukocytes decreased from 2.9 to 1 in SAPSII-low group whereas it increased from 1.7 to 2.4 in SAPSII-high group (Additional file 4: Table S3). In this function, PTEN gene, whose main function consists in the induction of apoptosis via caspase activation, was overexpressed in the SAPSII-low group at H0 (FC = 2.26 and 1.52 compared to healthy volunteers for SAPSII-low and SAPSII-high, respectively). In agreement, the expression of pro-apoptotic CASP1 and CASP4 was more increased in SAPSII-low group (FC = 2.15 and 2.07 compared to SAPSII-high FC = 1.55 and 1.14, respectively).

verexpressed in the SAPSII-low group at H0 (FC = 2.26 and 1.52 compared to healthy volunteers for SAPSII-low and SAPSII-high, respectively). In agreement, the expression of pro-apoptotic CASP1 and CASP4 was more increased in SAPSII-low group (FC = 2.15 and 2.07 compared to SAPSII-high FC = 1.55 and 1.14, respectively). The inhibiting functions exhibited by the SAPSII-high patient group were mostly related to cytotoxicity of cells, cytolysis, and viability functions. Thus, the predicted z-scores of ‘cytotoxicity of lymphocytes’ functions were −2.9, −2.5, and −3.2 at H0, H24, and H48, respectively.

verexpressed in the SAPSII-low group at H0 (FC = 2.26 and 1.52 compared to healthy volunteers for SAPSII-low and SAPSII-high, respectively). In agreement, the expression of pro-apoptotic CASP1 and CASP4 was more increased in SAPSII-low group (FC = 2.15 and 2.07 compared to SAPSII-high FC = 1.55 and 1.14, respectively). The inhibiting functions exhibited by the SAPSII-high patient group were mostly related to cytotoxicity of cells, cytolysis, and viability functions. Thus, the predicted z-scores of ‘cytotoxicity of lymphocytes’ functions were −2.9, −2.5, and −3.2 at H0, H24, and H48, respectively. Genes differentially expressed between severe and less severe patients Among the genes differentially expressed between septic shock patients and healthy volunteers (twofold difference - FDR <0.05), only 142 probe sets (corresponding to 122 unique Entrez gene) allowed to significantly distinguish SAPSII-low and SAPSII-high groups of patients (FDR <0.05). Figure 6 represents a heat map of these 142 probe sets.Figure 6 Differences in gene expression between SAPSII-low and SAPSII-high groups. (A) Heat map of 142 probe sets whose expression was (i) greater than or equal to twofold different (FDR <0.05) when comparing healthy volunteers with SAPSII-low (Low) or SAPSII-high (High) groups and (ii) also significantly different (FDR <0.05) between both groups of septic shock patients. Clustering was done using Euclidian distance method after normalization (shift probe set expression to mean of zero and scale to standard deviation of 1). (B) A cluster illustrating genes that are downmodulated in patients versus controls, with genes remaining downmodulated over time in SAPSII-high group only. (C) A cluster illustrating genes that are upmodulated in patients versus controls, with genes remaining upmodulated over time in SAPSII-high group only. (D) A cluster illustrating a downmodulation in SAPSII-low group only. (E) A cluster illustrating an upmodulation in SAPSII-low group only.

e in SAPSII-high group only. (C) A cluster illustrating genes that are upmodulated in patients versus controls, with genes remaining upmodulated over time in SAPSII-high group only. (D) A cluster illustrating a downmodulation in SAPSII-low group only. (E) A cluster illustrating an upmodulation in SAPSII-low group only. Most of the genes that were differentially expressed between the two groups were implicated in inflammatory and infection responses (Additional file 5: Table S4). The trend of the response was similar in both groups, and differences were observed only at the level and the duration of the genomic response between the two severity groups. These 142 genes were hierarchically arranged into 14 clusters (Additional file 6: Figure S2 and Figure 6). Those 14 clusters followed four distinct patterns of expressions illustrated by the 4 clusters displayed in Figure 6 B,C,D,E.

Most of the genes that were differentially expressed between the two groups were implicated in inflammatory and infection responses (Additional file 5: Table S4). The trend of the response was similar in both groups, and differences were observed only at the level and the duration of the genomic response between the two severity groups. These 142 genes were hierarchically arranged into 14 clusters (Additional file 6: Figure S2 and Figure 6). Those 14 clusters followed four distinct patterns of expressions illustrated by the 4 clusters displayed in Figure 6 B,C,D,E. For example, clusters 1 to 6 (Figure 6B) consisted in a panel of genes whose expressions were less decreased and recovered rapidly to normal value during the 48 h in SAPSII-low group but were decreased in SAPSII-high group with a delayed recovery. The genes following this trend of expression were related to leukocytes activation and immune response (BCL11B, BTN3A1, BTN3A2, BTN3A3, CX3CR1, DDX58, F2RL1, HLA-DQA1, HLA-DQB1, HLA-DPB1, IGHM, IGKC, IL7R, IRF1, LCK, LTB, LY75, MME, MMP12, PDE4B, PECAM1, PI3, PRKCH, PSMB9, RASGRP1, RORA, SFN, STAT1, and ZFP36L2) and mostly linked to cell death (APOL6, AQP3, BCL11B, CAMK1D, CX3CR1, DDX58, F2RL1, FGL2, GIMAP4, GZMA, IGHM, IGKC, IL32, IL7R, IRF1, ITGA4, LCK, LEF1, LTB, MME, MX1, PARP14, PDE4B, PECAM1, PI3, POLB, PRKCH, PTGER4, RASGRP1, SEMA4D, SGK1, STAT1, TAP2, XAF1, and ZFP36L2).

, PSMB9, RASGRP1, RORA, SFN, STAT1, and ZFP36L2) and mostly linked to cell death (APOL6, AQP3, BCL11B, CAMK1D, CX3CR1, DDX58, F2RL1, FGL2, GIMAP4, GZMA, IGHM, IGKC, IL32, IL7R, IRF1, ITGA4, LCK, LEF1, LTB, MME, MX1, PARP14, PDE4B, PECAM1, PI3, POLB, PRKCH, PTGER4, RASGRP1, SEMA4D, SGK1, STAT1, TAP2, XAF1, and ZFP36L2). Clusters 13 and 14 (Figure 6C) illustrated genes with a small increase in gene expression and a rapid recovery in SAPSII-low group of patients and a high increase over time in SAPSII-high patients. Among these, we observed genes related to infection and cell viability (ALOX12, ANKRD9, CA2, LCN2, MS4A4A, OLFM4, PF4, PPBP, RETN, and TCN1). The last two patterns of expressions are illustrated by clusters 7 to 11 and 12 (Figure 6D,E). In contrast, to the previous patterns (clusters 1 to 6 and 13 and 14), they showed an earlier and more pronounced response (over- and under-expressed) in the SAPSII-low group. Genes found in these clusters were mostly implicated in bacterial or viral infections (AGTRAP, CARD16, CASP1, CD274, CD46, CFD, CLEC2B, DDX60L, HSP90AB1, MAN1A1, PKN2, and SLPI) and cell death or viability (AGTRAP, BAG1, CASP1, CD274, CD46, CFD, HSP90AB1, HSPB1, LRRK2, PKN2, RBM3, SLAMF7, SLC6A8, SLPI, and SNCA).

roup. Genes found in these clusters were mostly implicated in bacterial or viral infections (AGTRAP, CARD16, CASP1, CD274, CD46, CFD, CLEC2B, DDX60L, HSP90AB1, MAN1A1, PKN2, and SLPI) and cell death or viability (AGTRAP, BAG1, CASP1, CD274, CD46, CFD, HSP90AB1, HSPB1, LRRK2, PKN2, RBM3, SLAMF7, SLC6A8, SLPI, and SNCA). Discussion To our knowledge, this is the first study describing the very early genomic response to septic shock. This study provides an important overview of the genome-wide expression patterns of blood leukocytes over three time points (within 30 min after diagnosis and 24 and 48 h afterwards). Septic shock generates a massive genomic modulation with more than 71% of the host transcriptome altered during the first 48 h after shock. This phenomenon is a very aggressive process as it occurs rapidly and is already present at the very early stage of the syndrome with more than 60% of the modification already in place (Figure 1). As described in trauma patients [14], septic shock appears to produce a global reprogramming of the leukocyte transcriptome affecting multiple cellular and molecular functions and pathways. However, the magnitude of the transcriptome modifications seemed to be earlier, faster, and bigger in septic shock, as regards of the extent and rapidity by which the deregulation occurred. Indeed, more than 71% of the genomic modifications occurred during the first 48 h, whereas most of the alterations described by Xiao et al. in trauma patients were observed over the first 28 days.

o be earlier, faster, and bigger in septic shock, as regards of the extent and rapidity by which the deregulation occurred. Indeed, more than 71% of the genomic modifications occurred during the first 48 h, whereas most of the alterations described by Xiao et al. in trauma patients were observed over the first 28 days. Most of the altered molecular functions were associated to cellular rearrangement (morphology, function, maintenance, and development), therefore suggesting metabolic adaptations (Table 2). Indeed, several studies have shown that metabolic modifications occur during severe injuries and could lead to a metabolic hibernation. This phenomenon plays a role in protecting cells from death leading to multiple organ failures [24,25] and immune-inflammation control [26]. The decrease of leukocyte pathways and functions (Figures 2B and 5A) suggests an anergy of those cells. As an example, SAMSN1, whose expression increased the most (Figure 3B), has been defined as an inhibitor of B cell spreading [27]. Inversely, CD247, whose expression decreased the most, is the CD3ζ-chain, considered as the rate-limiting factor of TCR/CD3 complex formation [28]. This was consistent with the decreased expression of CD3 on circulating lymphocytes observed in septic shock patients [29]. Inversely, the function ‘phagocytosis of cells’ was predicted to be activated during the kinetic. These results could be related to efferocytosis of dying/dead cells triggered by anti-inflammatory signals in a process of resolving inflammation [30].

CD3 on circulating lymphocytes observed in septic shock patients [29]. Inversely, the function ‘phagocytosis of cells’ was predicted to be activated during the kinetic. These results could be related to efferocytosis of dying/dead cells triggered by anti-inflammatory signals in a process of resolving inflammation [30]. Taken together, the patterns of gene expression highlighted that the mechanisms of immune activation and repression are obviously concomitant at the very early phase of septic shock (Figure 3 and Additional file 6: Figure S2), probably before obvious clinical signs. We cannot exclude that earlier modification in gene expression might also have occurred before patients fulfilled the diagnostic criteria of septic shock. Here, we described only the modulation that were observed during the first 48 h after beginning vasopressor therapy. This is in accordance with numerous clinical observations made in the literature showing altered leukocyte functions after septic shock [4,31–33]. This also agrees with previously published data regarding transcription profiles in human sepsis [34] and in trauma patients [14]. Of utmost importance, the early regulation mechanisms observed suggests that the time window to improve the outcome of patient with appropriate therapeutic begins very early after the onset of shock. This observation has been also suggested recently [35].

Taken together, the patterns of gene expression highlighted that the mechanisms of immune activation and repression are obviously concomitant at the very early phase of septic shock (Figure 3 and Additional file 6: Figure S2), probably before obvious clinical signs. We cannot exclude that earlier modification in gene expression might also have occurred before patients fulfilled the diagnostic criteria of septic shock. Here, we described only the modulation that were observed during the first 48 h after beginning vasopressor therapy. This is in accordance with numerous clinical observations made in the literature showing altered leukocyte functions after septic shock [4,31–33]. This also agrees with previously published data regarding transcription profiles in human sepsis [34] and in trauma patients [14]. Of utmost importance, the early regulation mechanisms observed suggests that the time window to improve the outcome of patient with appropriate therapeutic begins very early after the onset of shock. This observation has been also suggested recently [35]. Interestingly, when comparing genomic responses of the two severity groups, we observed that the overall changes (direction of the responses) in gene expression at H0 between these groups were very similar. Importantly, the difference between the two severity groups was particularly noticeable in the degree and the duration of the altered acute inflammatory response.

f the two severity groups, we observed that the overall changes (direction of the responses) in gene expression at H0 between these groups were very similar. Importantly, the difference between the two severity groups was particularly noticeable in the degree and the duration of the altered acute inflammatory response. Most of the activated functions in both groups were related to cell death and apoptosis. However, they were predicted to be activated and trending toward a decrease during the kinetic in SAPSII-low group, whereas these were predicted to be inhibited in SAPSII-high group. These modifications could be related to T cells undergoing apoptosis, therefore limiting self-harmful effects for the host [36]. This observation emphasized that an appropriate response very early in the process might limit or avoid exacerbated inflammation that would lead to increase severity. Inhibited cytotoxicity of cells and ‘viability’ functions, observed in SAPSII-high patient group, again suggests that the SAPSII-low patient group may have developed a better response (or an appropriate response regulation) and a rapid trend toward recovery, while the SAPSII-high group of patients showed a delayed response and an impaired immune dysfunction.

bility’ functions, observed in SAPSII-high patient group, again suggests that the SAPSII-low patient group may have developed a better response (or an appropriate response regulation) and a rapid trend toward recovery, while the SAPSII-high group of patients showed a delayed response and an impaired immune dysfunction. Four distinct patterns associated with the two groups of severity were observed. The two first patterns indicated that SAPSII-low group seemed to have smaller and shorter deregulation of their genomic response as compared with the more severe patients. For example, in clusters 13 and 14 (Figure 6C), OLFM4 (negative regulator of bacterial killing) and RETN and MS4A4A (related to PPARγ that overexpression has been linked to poor outcome [37]) had increased expressions and no return to baseline in SAPSII-high group. This suggested that the bacterial clearance and cell viability were more jeopardized in this group. Once again, these first two patterns indicated that SAPSII-low group seemed to have smaller and shorter deregulation of their genomic response as compared with the more severe patients.

Four distinct patterns associated with the two groups of severity were observed. The two first patterns indicated that SAPSII-low group seemed to have smaller and shorter deregulation of their genomic response as compared with the more severe patients. For example, in clusters 13 and 14 (Figure 6C), OLFM4 (negative regulator of bacterial killing) and RETN and MS4A4A (related to PPARγ that overexpression has been linked to poor outcome [37]) had increased expressions and no return to baseline in SAPSII-high group. This suggested that the bacterial clearance and cell viability were more jeopardized in this group. Once again, these first two patterns indicated that SAPSII-low group seemed to have smaller and shorter deregulation of their genomic response as compared with the more severe patients. In contrast, the two last patterns showed an earlier and more pronounced response in the SAPSII-low group. These results suggest that those genes may play a role in the very early negative feedback mechanisms that limit the process of inflammation during septic shock. Thus, the SAPSII-low group may exhibit a better and rapidly appropriate regulation of the immune and inflammatory response leading to a subsequent better outcome. This agrees with the observation that the number of significant upregulated genes is lower in the SAPSII-high group and increases during the kinetic while decreasing in the SAPSII-low group (Figure 4C).

er and rapidly appropriate regulation of the immune and inflammatory response leading to a subsequent better outcome. This agrees with the observation that the number of significant upregulated genes is lower in the SAPSII-high group and increases during the kinetic while decreasing in the SAPSII-low group (Figure 4C). These results of the utmost importance may suggest that regulatory mechanisms leading to recovery may take place very early after septic shock and thus that the time window to improve the outcome of patients with appropriate therapeutics may begin very early after the onset of shock. Indeed, beyond the fact that differences between the two severity groups were particularly noticeable in the degree and the duration of the altered acute inflammatory response, we observed that the more severe patients did not exhibit the strongest modulation for every process changed (e.g., cell death at H0) in part suggesting that an early and appropriate regulation may be the key for a suitable response. Thus, genes involved in these regulatory processes might be of major interest in additional studies.

at the more severe patients did not exhibit the strongest modulation for every process changed (e.g., cell death at H0) in part suggesting that an early and appropriate regulation may be the key for a suitable response. Thus, genes involved in these regulatory processes might be of major interest in additional studies. There are some limitations of our study. First, our healthy volunteers’ cohort was not age/sex matched with patients (5 men and 20 women in healthy volunteers group). In our opinion, this is not a major issue as important results were mostly obtained from comparison between patients regarding kinetics approach. Second, as the SAPSII-high group tends to be older, the delayed and impaired immune response might have been related to age rather than severity. An additional analysis using ANOVA indicated that the modulation of gene expression was not related to age of patients (data not shown). Third, a reduced number of patients was enrolled and despite a high level of severity (median SAPSII = 45; median SOFA = 10), a low mortality rate was observed. We were therefore not able to study gene expression according to prognosis. Fourth, gene expression analyses were performed in whole blood samples that do not reveal the entire genomic dysfunctions leading to organ failure. Nevertheless, this compartment could indirectly reflect the damages observed in organs and is more easily accessible. In addition, there was a trend toward a lower number of polymorphonuclear cells in high SAPSII patients, but this was not statistically significant. Moreover, we did not found any difference between SAPSII-low and SAPSII-high groups regarding the lymphocyte counts. The differences observed and functions highlighted in our study are unlikely to be due to the differences in cell subpopulations between these groups. Lastly, no functional testing was done to validate that SAPSII-low group has a better integrity of their immune system or a better and earlier recovery. However, the results observed were coherent with findings already published in the literature. Although further studies will be needed to study deeper the mechanisms of systemic inflammatory response regulation (especially the link between genomic disturbances and immune functional testing), this study provides a first overview of the mechanisms involved at the earliest time and at the transcriptomic level after septic shock and may contribute to orientate the next investigations.

of systemic inflammatory response regulation (especially the link between genomic disturbances and immune functional testing), this study provides a first overview of the mechanisms involved at the earliest time and at the transcriptomic level after septic shock and may contribute to orientate the next investigations. Conclusions Pan-genomic approaches, allowing global investigation, should provide crucial keys to understand the much focused orientation of the immune response induced by stress events. These results could contribute to improved management of patients. Indeed, genes implicated very early in the regulatory mechanisms of the response to septic shock could become high medical value biomarkers allowing individualized treatments aimed at restricting the dysfunctions induced by exacerbated or uncontrolled inflammation. Additional files Additional file 1: Table S1. Modulation over the first 48 h of septic shock for selected genes highlighted by the functional analysis. Additional file 2: Table S2. Significant category in septic shock patients over time compared to healthy volunteers. The table shows the predicted activation or inhibition status of the enriched functions. Additional file 3: Figure S1. qRT-PCR correlation with Affymetrix data. Additional file 4: Table S3. Significant ‘cell death’ functions in septic shock patients over time compared to healthy volunteers. Additional file 5: Table S4. Significant categories enriched by the 142 probe sets differentially expressed between.

Additional file 2: Table S2. Significant category in septic shock patients over time compared to healthy volunteers. The table shows the predicted activation or inhibition status of the enriched functions. Additional file 3: Figure S1. qRT-PCR correlation with Affymetrix data. Additional file 4: Table S3. Significant ‘cell death’ functions in septic shock patients over time compared to healthy volunteers. Additional file 5: Table S4. Significant categories enriched by the 142 probe sets differentially expressed between. Additional file 6: Figure S2. Comparison of gene expression patterns of the 142 probe sets differentially expressed (FDR <0.05) between both groups of septic shock patients. Expressions at 0, 24, and 48 h after septic shock of genes in each of the 14 clusters from hierarchical clustering in Figure 6. Competing interests The authors declare that they have no competing interests. Authors’ contributions AL, BM, AP, and GM conceived the study. MAC performed the experiments. MAC, FV, FF, and MP performed the statistical analyses and biological interpretations. All contributed to drafting and critical revision of the manuscript. All authors read and approved the final manuscript. Acknowledgements This project is part of Advanced Diagnostics for New Therapeutic Approaches coordinated by Merieux Alliance and supported by the French public agency OSEO. GM, FV, and AL are supported by funds from the Hospices Civils de Lyon.

Background Acute kidney injury (AKI) is a frequent and costly clinical complication in critically ill patients. Using the RIFLE criteria in 20,126 hospitalized patients, Uchino et al. has found that 20% of the patients had some degrees of acute renal impairment and 3.7% of these patients had AKI [1]. Ischemia/reperfusion (I/R) injury occurring during surgery and shock is one of the major causes of this condition in native and transplanted kidneys [1–5]. The pathogenesis and pathophysiology of I/R-induced AKI is highly complex [4, 5]. In addition to hypoxic hit consequent to ischemia, the reperfusion phase has been associated with additional renal injury. I/R-induced activation of inflammatory pathways has been shown to worsen AKI [6–9]. Furthermore, increased production of radical oxygen species (ROS) and reactive nitrogen species (RNS) [10–13] and a regional imbalance between vasoactive mediators are believed to be of great importance in the development of AKI by leading impaired microcirculatory perfusion and to cellular damage, apoptosis, and irreversible organ failure [4, 5].

tion of radical oxygen species (ROS) and reactive nitrogen species (RNS) [10–13] and a regional imbalance between vasoactive mediators are believed to be of great importance in the development of AKI by leading impaired microcirculatory perfusion and to cellular damage, apoptosis, and irreversible organ failure [4, 5]. In this respect, vanadium compounds are promising in the prevention and treatment of ischemic AKI. Vanadium is an essential trace element in humans and plays various roles through different pathways in metabolism [14, 15]. Recent findings suggest that vanadium may play a pivotal role in the regulation of physiological cell growth, survival, and metabolism. Most biologically active forms of vanadium are the inorganic vanadate salt vanadyl sulfate (VOSO4) and the organic vanadium salt bis oxovanadium (BMOV). Especially, the organic BMOV can be used in vitro and in vivo for its beneficial regulatory metabolic roles without major side effects. BMOV and other vanadium compounds have demonstrated protective against ischemic cascades, apoptosis, and vascular endothelial dysfunction while supporting tissue repair in the heart and the brain [14–16]. In the present study, we therefore aimed to investigate the potential protective effects of BMOV in the acute phase of renal I/R and AKI. To this end, rats received 0 or 15 mg/kg BMOV intravenously 30 min before renal artery clamping, and we measured renal oxygenation and renal function up to 90 min of reperfusion.

In this respect, vanadium compounds are promising in the prevention and treatment of ischemic AKI. Vanadium is an essential trace element in humans and plays various roles through different pathways in metabolism [14, 15]. Recent findings suggest that vanadium may play a pivotal role in the regulation of physiological cell growth, survival, and metabolism. Most biologically active forms of vanadium are the inorganic vanadate salt vanadyl sulfate (VOSO4) and the organic vanadium salt bis oxovanadium (BMOV). Especially, the organic BMOV can be used in vitro and in vivo for its beneficial regulatory metabolic roles without major side effects. BMOV and other vanadium compounds have demonstrated protective against ischemic cascades, apoptosis, and vascular endothelial dysfunction while supporting tissue repair in the heart and the brain [14–16]. In the present study, we therefore aimed to investigate the potential protective effects of BMOV in the acute phase of renal I/R and AKI. To this end, rats received 0 or 15 mg/kg BMOV intravenously 30 min before renal artery clamping, and we measured renal oxygenation and renal function up to 90 min of reperfusion. Methods Animals All experiments in this study were approved by the Institutional Animal Experimentation Committee of the Academic Medical Center of the University of Amsterdam. Care and handling of the animals were in accordance with the EU Directive 2010/63/EU for animal experiments and guidelines for Institutional and Animal Care and Use Committees. The study has been carried out in accordance with the National Institutes of Health (NIH) Guide for the Care and Use of Laboratory Animals. Experiments were performed on 18 Wistar male rats (Harlan, the Netherlands) with mean ± SD body weight of 320 ± 30 g.

idelines for Institutional and Animal Care and Use Committees. The study has been carried out in accordance with the National Institutes of Health (NIH) Guide for the Care and Use of Laboratory Animals. Experiments were performed on 18 Wistar male rats (Harlan, the Netherlands) with mean ± SD body weight of 320 ± 30 g. Surgical preparation The rats were anesthetized with an intraperitoneal injection of a mixture of 100 mg/kg ketamine (Nimatek®, Eurovet, Bladel, the Netherlands), 0.5 mg/kg medetomidine (Domitor, Pfizer, New York, NY, USA), and 0.05 mg/kg atropine sulfate (Centrafarm, Etten-Leur, the Netherlands). After tracheotomy, the animals were mechanically ventilated with a FiO2 of 0.4. Body temperature was maintained at 37°C ± 0.5°C during the entire experiment by external warming. The ventilator settings were adjusted to maintain end-tidal PCO2 between 30 and 35 mmHg and arterial PCO2 between 35 and 40 mmHg. Vessels were cannulated with polyethylene catheters (outer diameter = 0.9 mm; Braun, Melsungen, Germany) for drug and fluid administration and hemodynamic monitoring. A catheter in the right carotid artery was connected to a pressure transducer to monitor mean arterial blood pressure (MAP) and heart rate. The right femoral artery was cannulated for blood sampling. The right femoral vein was cannulated for continuous infusion of Ringer’s lactate (15 mL/kg/h; Baxter, Utrecht, the Netherlands) and ketamine (50 mg/kg/h; Nimatek®, Eurovet, Bladel, the Netherlands).

o monitor mean arterial blood pressure (MAP) and heart rate. The right femoral artery was cannulated for blood sampling. The right femoral vein was cannulated for continuous infusion of Ringer’s lactate (15 mL/kg/h; Baxter, Utrecht, the Netherlands) and ketamine (50 mg/kg/h; Nimatek®, Eurovet, Bladel, the Netherlands). The left kidney was exposed, decapsulated, and immobilized in a Lucite kidney cup (K. Effenberger, Pfaffingen, Germany) via a 4-cm incision in the left flank. Renal vessels were carefully separated under preservation of nerves and adrenal gland. A perivascular ultrasonic transit time flow probe was placed around the left renal artery (type 0.7 RB; Transonic Systems Inc., Ithaca, NY, USA) and connected to a flow meter (T206, Transonic Systems Inc.) to continuously measure renal blood flow (RBF). An estimation of the renal vascular resistance (RVR) was made as RVR [dynes/sec/cm5] = MAP/RBF. The left ureter was isolated, ligated, and cannulated with a polyethylene catheter for urine collection.

, NY, USA) and connected to a flow meter (T206, Transonic Systems Inc.) to continuously measure renal blood flow (RBF). An estimation of the renal vascular resistance (RVR) was made as RVR [dynes/sec/cm5] = MAP/RBF. The left ureter was isolated, ligated, and cannulated with a polyethylene catheter for urine collection. After the surgical protocol (approximately 60 min), one optical fiber was placed 1 mm above the decapsulated kidney and another optical fiber 1 mm above the renal vein to measure oxygenation in the renal microvasculature and renal vein, respectively, using phosphorimetry. A small piece of aluminum foil was placed on the dorsal site of the renal vein to prevent the contribution of underlying tissue to the phosphorescence signal in the venous oxygenation measurement. Oxyphor G2 (a two-layer glutamate dendrimer of tetra-(4-carboxy-phenyl) benzoporphyrin, Oxygen Enterprises Ltd., Philadelphia, PA, USA) was subsequently infused (6 mg/kg IV over 5 min) followed by a 30-min stabilization period. A short description of phosphorimetry is given below, and a more detailed description of the technology has been provided elsewhere [17–20].

etra-(4-carboxy-phenyl) benzoporphyrin, Oxygen Enterprises Ltd., Philadelphia, PA, USA) was subsequently infused (6 mg/kg IV over 5 min) followed by a 30-min stabilization period. A short description of phosphorimetry is given below, and a more detailed description of the technology has been provided elsewhere [17–20]. Experimental protocol After baseline measurements were performed 30 min after Oxyphor G2 infusion, the rats were randomly assigned to one of the following groups: sham-operated time control (n = 6), I/R control (n = 6), and I/R with 15 mg/kg BMOV (n = 6). Considering this is the first study in which BMOV is being utilized in a model of renal I/R injury, we decided to use the recommended dosage (15 mg/kg) by CFM Pharma (Almere, the Netherlands). BMOV solutions were prepared in 2-mL isotonic saline, and infusion was initiated 30 min prior to renal ischemia at an infusion rate of 2 mL/h. Control rats received the same volume of isotonic saline without BMOV. Renal ischemia was created by 30-min clamping of the renal artery, and following the release of the clamp, measurements were continued up to 90 min of reperfusion. The experiments were terminated by infusion of 1 mL of 3 M potassium chloride (KCl).

Control rats received the same volume of isotonic saline without BMOV. Renal ischemia was created by 30-min clamping of the renal artery, and following the release of the clamp, measurements were continued up to 90 min of reperfusion. The experiments were terminated by infusion of 1 mL of 3 M potassium chloride (KCl). Blood variables Arterial blood samples (0.5 mL) were taken from the femoral artery at baseline and 15 and 90 min after reperfusion. The blood samples were replaced by the same volume of Voluven® (Fresenius Kabi Ltd., Runcom, UK). Samples were analyzed for blood gas values (ABL505 blood gas analyzer, Radiometer, Copenhagen, Denmark), hemoglobin concentration, and hemoglobin oxygen saturation (OSM3, Radiometer). Additionally, plasma creatinine concentrations were determined in all samples.

ame volume of Voluven® (Fresenius Kabi Ltd., Runcom, UK). Samples were analyzed for blood gas values (ABL505 blood gas analyzer, Radiometer, Copenhagen, Denmark), hemoglobin concentration, and hemoglobin oxygen saturation (OSM3, Radiometer). Additionally, plasma creatinine concentrations were determined in all samples. Renal microvascular and venous oxygenation Microvascular oxygen tension in the renal cortex (CμPO2), outer medulla (MμPO2), and renal venous oxygen tension (PrvO2) were measured by oxygen-dependent quenching of phosphorescence lifetimes of the systemically infused albumin-targeted (and therefore circulation-confined) phosphorescent dye Oxyphor G2. Oxyphor G2 (a two-layer glutamate dendrimer of tetra-(4-carboxy-phenyl) benzoporphyrin) has two excitation peaks (λexcitation1 = 440 nm, λexcitation2 = 632 nm) and one emission peak (λemission = 800 nm). These optical properties allow (near) simultaneous lifetime measurements in microcirculation of the kidney cortex and the outer medulla due to different optical penetration depths of the excitation light. For the measurement of renal venous PO2 (PrvO2), a mono-wavelength phosphorimeter was used. Oxygen measurements based on phosphorescence lifetime techniques rely on the principle that phosphorescence can be quenched by energy transfer to oxygen resulting in the shortening of the phosphorescence lifetime. A linear relationship between reciprocal phosphorescence lifetime and oxygen tension (given by the Stern-Volmer relation) allows quantitative measurement of PO2[17–20].

es rely on the principle that phosphorescence can be quenched by energy transfer to oxygen resulting in the shortening of the phosphorescence lifetime. A linear relationship between reciprocal phosphorescence lifetime and oxygen tension (given by the Stern-Volmer relation) allows quantitative measurement of PO2[17–20]. Renal oxygen delivery and consumption Arterial oxygen content (AOC) was calculated by (1.31 × hemoglobin × SaO2) + (0.003 × PaO2), where SaO2 is arterial oxygen saturation and PaO2 is arterial partial pressure of oxygen. Renal venous oxygen content (RVOC) was calculated as (1.31 × hemoglobin × SrvO2) + (0.003 × PrvO2), where SrvO2 is venous oxygen saturation and PrvO2 is renal vein partial pressure of oxygen (measured using phosphorimetry). Renal oxygen delivery per gram of renal tissue was calculated as DO2 (mL/min/g) = RBF × AOC. Renal oxygen consumption per gram of renal tissue was calculated as VO2 (mL/min/g) = RBF × (AOC − RVOC). The renal oxygen extraction ratio was calculated as O2 ER (%) = VO2/DO2 × 100.

of oxygen (measured using phosphorimetry). Renal oxygen delivery per gram of renal tissue was calculated as DO2 (mL/min/g) = RBF × AOC. Renal oxygen consumption per gram of renal tissue was calculated as VO2 (mL/min/g) = RBF × (AOC − RVOC). The renal oxygen extraction ratio was calculated as O2 ER (%) = VO2/DO2 × 100. Renal function For the analysis of urine volume, creatinine concentration, and sodium (Na+) concentration at the end of the protocol, urine samples from the left ureter were collected for 10 min. Creatinine clearance rate (CLcrea) per gram of renal tissue was calculated with the standard formula: CLcrea (mL/min/g) = (U × V)/P, where U is the urine creatinine concentration, V is the urine volume per unit time, and P is the plasma creatinine concentration. Renal oxygen consumption efficiency for sodium transport (VO2/TNa+) was assessed as the ratio of the renal VO2 over the total amount of sodium reabsorbed (TNa+, [mmol/min]). TNa+ was calculated according to the following: ((CLcrea × PNa+) − (UNa+ × V)), where PNa+ is the plasma concentration of sodium. Data analysis Statistical analysis was performed using GraphPad Prism version 5.0 for Windows (GraphPad Software, San Diego, CA, USA). Data are presented as mean ± SD unless otherwise stated. Statistical significance of differences between groups was tested using one-way ANOVA with Bonferroni post hoc tests. P values <0.05 were considered significant.

s was performed using GraphPad Prism version 5.0 for Windows (GraphPad Software, San Diego, CA, USA). Data are presented as mean ± SD unless otherwise stated. Statistical significance of differences between groups was tested using one-way ANOVA with Bonferroni post hoc tests. P values <0.05 were considered significant. Results Table 1 shows the systemic and renal hemodynamic variables: MAP, RBF, RVR, DO2, VO2, CμpO2, and MμpO2 at baseline (BL), 15 min after reperfusion (R15), and 90 min after reperfusion (R90). Figure 1 shows the renal DO2 and VO2 and TNa+, renal oxygen handling efficacy (VO2/TNa+), and creatinine clearance rate at the end of the protocol. At baseline, there were no significant differences between groups in any of these variables.Table 1 Systemic and renal hemodynamic variables at BL, R15, and R90

sion (R90). Figure 1 shows the renal DO2 and VO2 and TNa+, renal oxygen handling efficacy (VO2/TNa+), and creatinine clearance rate at the end of the protocol. At baseline, there were no significant differences between groups in any of these variables.Table 1 Systemic and renal hemodynamic variables at BL, R15, and R90 BL (t = 0 min) R15 (t = 45 min) R90 (t = 120 min) MAP [mmHg] Time control 104 ± 4 101 ± 4 99 ± 4 I/R 108 ± 11 104 ± 7 90 ± 18 I/R + BMOV 108 ± 13 102 ± 24 92 ± 12 RBF [mL/min] Time control 4.7 ± 0.5 4.6 ± 0.6 4.6 ± 0.5 I/R 4.6 ± 1.0 2.8 ± 0.5* 3.3 ± 0.1* I/R + BMOV 4.2 ± 1.1 2.7 ± 0.4* 3.4 ± 0.3* RVR [dyn/s/cm5] Time control 1,777 ± 172 1,793 ± 244 1,735 ± 153 I/R 2,004 ± 672 3,097 ± 451* 2,216 ± 502* I/R + BMOV 2,190 ± 723 3,217 ± 1,251* 2,178 ± 240* DO2 [mL O2/min/g] Time control 1.21 ± 0.14 1.13 ± 0.14 1.18 ± 0.11 I/R 1.10 ± 0.24 0.65 ± 0.13* 0.72 ± 0.03* I/R + BMOV 1.05 ± 0.32 0.64 ± 0.14* 0.77 ± 0.10* VO2 [mL O2/min/g] Time control 0.14 ± 0.07 0.13 ± 0.04 0.15 ± 0.02 I/R 0.13 ± 0.10 0.10 ± 0.02* 0.09 ± 0.01* I/R + BMOV 0.13 ± 0.07 0.11 ± 0.05 0.10 ± 0.04* CμpO2 [mmHg] Time control 65 ± 7 62 ± 7 60 ± 6 I/R 63 ± 6 59 ± 9 54 ± 5 I/R + BMOV 66 ± 2 61 ± 9 64 ± 6** MμpO2 [mmHg] Time control 52 ± 6 51 ± 9 49 ± 7 I/R 54 ± 5 50 ± 5 49 ± 2 I/R + BMOV 50 ± 5 50 ± 4 48 ± 3 *p < 0.05 vs time control; **p < 0.05 vs I/R. Figure 1 DO 2 and VO 2 (A), T Na+ (B), VO 2 / T Na+ (C), and CL crea (D) at the end of the protocol.

BL (t = 0 min) R15 (t = 45 min) R90 (t = 120 min) MAP [mmHg] Time control 104 ± 4 101 ± 4 99 ± 4 I/R 108 ± 11 104 ± 7 90 ± 18 I/R + BMOV 108 ± 13 102 ± 24 92 ± 12 RBF [mL/min] Time control 4.7 ± 0.5 4.6 ± 0.6 4.6 ± 0.5 I/R 4.6 ± 1.0 2.8 ± 0.5* 3.3 ± 0.1* I/R + BMOV 4.2 ± 1.1 2.7 ± 0.4* 3.4 ± 0.3* RVR [dyn/s/cm5] Time control 1,777 ± 172 1,793 ± 244 1,735 ± 153 I/R 2,004 ± 672 3,097 ± 451* 2,216 ± 502* I/R + BMOV 2,190 ± 723 3,217 ± 1,251* 2,178 ± 240* DO2 [mL O2/min/g] Time control 1.21 ± 0.14 1.13 ± 0.14 1.18 ± 0.11 I/R 1.10 ± 0.24 0.65 ± 0.13* 0.72 ± 0.03* I/R + BMOV 1.05 ± 0.32 0.64 ± 0.14* 0.77 ± 0.10* VO2 [mL O2/min/g] Time control 0.14 ± 0.07 0.13 ± 0.04 0.15 ± 0.02 I/R 0.13 ± 0.10 0.10 ± 0.02* 0.09 ± 0.01* I/R + BMOV 0.13 ± 0.07 0.11 ± 0.05 0.10 ± 0.04* CμpO2 [mmHg] Time control 65 ± 7 62 ± 7 60 ± 6 I/R 63 ± 6 59 ± 9 54 ± 5 I/R + BMOV 66 ± 2 61 ± 9 64 ± 6** MμpO2 [mmHg] Time control 52 ± 6 51 ± 9 49 ± 7 I/R 54 ± 5 50 ± 5 49 ± 2 I/R + BMOV 50 ± 5 50 ± 4 48 ± 3 *p < 0.05 vs time control; **p < 0.05 vs I/R. Figure 1 DO 2 and VO 2 (A), T Na+ (B), VO 2 / T Na+ (C), and CL crea (D) at the end of the protocol. Systemic and renal hemodynamic and oxygenation variables Renal I/R did not significantly affect MAP, but decreased RBF and increased RVR, which was associated with decreases in both renal DO2 and VO2. BMOV did not affect any of these macrocirculatory hemodynamic and oxygenation variables. CμpO2 and MμpO2 were similar before and after reperfusion. At the end of the protocol, CμpO2 was significantly higher in the BMOV treated group compared to the I/R control group.

associated with decreases in both renal DO2 and VO2. BMOV did not affect any of these macrocirculatory hemodynamic and oxygenation variables. CμpO2 and MμpO2 were similar before and after reperfusion. At the end of the protocol, CμpO2 was significantly higher in the BMOV treated group compared to the I/R control group. Renal function parameters At the end of the protocol, renal DO2 and VO2 were decreased proportionally. Renal TNa+ was significantly reduced in the I/R control group, but in the group receiving BMOV, this decrease was not statistically significant. Renal creatinine clearance rate decreased after I/R, both with and without BMOV administration; however, it did not reach a level of significance in both groups. Similarly renal oxygen handling efficacy (VO2/TNa+) were maintained after I/R in both groups. Discussion The aim of the present study was to test the potential protective effects of BMOV (15 mg/kg) in the acute phase of renal I/R and its effects on renal oxygenation and renal function up to 90 min post-ischemia. The main findings were that (1) BMOV did not significantly affect the systemic or the renal hemodynamic and oxygenation variables and (2) BMOV partially protected TNa+ after I/R. Furthermore, we found that microcirculatory oxygenation in the renal cortex and medulla tended to decrease after I/R. In contrast to the non-treated animals, cortical microcirculatory oxygenation was preserved in the BMOV-treated animals, but no significant differences were seen in the renal medulla.

ed TNa+ after I/R. Furthermore, we found that microcirculatory oxygenation in the renal cortex and medulla tended to decrease after I/R. In contrast to the non-treated animals, cortical microcirculatory oxygenation was preserved in the BMOV-treated animals, but no significant differences were seen in the renal medulla. To our knowledge, this is the first study investigating the effects of organic vanadium compound BMOV in the context of renal I/R injury. Inorganic vanadium compounds are documented to be nephrotoxic, particularly if used chronically [21], but organic vanadium compounds, such as BMOV, are known to have less side effects. In experimental studies, both pre- and post-ischemic administrations of vanadium compounds have been shown to be cytoprotective. We believe that preventive strategies are essential in order to minimize I/R injury. Therefore, we administered BMOV 30 min before the onset of ischemia. The optimal dosage of BMOV in order to prevent renal I/R injury is unknown, but administration of 15 mg/kg is recommended by the manufacturer, which is also comparable to the doses used in earlier studies [22].

sential in order to minimize I/R injury. Therefore, we administered BMOV 30 min before the onset of ischemia. The optimal dosage of BMOV in order to prevent renal I/R injury is unknown, but administration of 15 mg/kg is recommended by the manufacturer, which is also comparable to the doses used in earlier studies [22]. Earlier studies investigating the beneficial effects of vanadium compounds after I/R injury in vivo have mainly been focused on the brain and on the heart. However, the type of vanadium, the type of administration, and the timing of administration varied significantly between studies. In the context of brain I/R, Kawano et al. showed in adult Mongolian gerbils that were subjected to 5-min forebrain ischemia that intraventricular injection of orthovanadate 30 min before ischemia blocked delayed neuronal death [23]. Hasegawa et al. demonstrated the neuroprotective effects of post-ischemic intraperitoneal administration of sodium orthovanadate in rats with transient middle cerebral artery occlusion 1 and 28 days after ischemia [24]. In a subsequent study, the authors determined the therapeutic time window (0, 45, and 90 min post-middle cerebral artery occlusion) and the neuroprotective dose (2 mL/kg and 12.5, 25, 37.5, and 50 mM) of sodium orthovanadate in rats [24]. Later, Shioda et al. found in a mouse model of transient middle cerebral artery occlusion that pre- and post-treatments with bis(1-oxy-2-pyridinethiolato)oxovanadium(IV) significantly reduced infarct volume in a dose-dependent manner and thereby provided neuroprotection in brain I/R injury [25]. The same group also showed that i.p. administration of bis(1-oxy-2-pyridinethiolato)oxovanadium(IV) markedly enhanced brain ischemia-induced neurogenesis in the subgranular zone of the mouse hippocampus [26]. Additionally, they found that amelioration of cognitive dysfunction following brain ischemia was positively correlated with vanadium-induced neurogenesis. Li et al. found that bisperoxovanadium attenuated cellular apoptosis in developing rat brain rescued neurons from hypoxia-ischemia brain damage [27]. Liu et al., furthermore, showed that 4 weeks of administration of sodium orthovanadate in drinking water significantly improved the outcome in rats with streptozotocin-induced diabetes after cerebral ischemia and reperfusion in terms of neurobehavioral function [28]. In the context of myocardial I/R, Geraldes et al.

]. Liu et al., furthermore, showed that 4 weeks of administration of sodium orthovanadate in drinking water significantly improved the outcome in rats with streptozotocin-induced diabetes after cerebral ischemia and reperfusion in terms of neurobehavioral function [28]. In the context of myocardial I/R, Geraldes et al. showed in isolated perfused rat hearts that the presence of vanadate during ischemia resulted in attenuation of acidosis and reduced lactate accumulation [29]. In anesthetized rats, Liem et al. showed that intravenous infusion of BMOV in doses of 3.3, 7.5, and 15 mg/kg i.v. decreased myocardial infarct size dose-dependently when administered before occlusion [22]. Administration of the low dose during ischemia just before reperfusion was ineffective, but administration of the higher doses was equally cardioprotective as compared with administration before occlusion. Bhuiyan et al. showed that post-ischemic treatment with bis(1-oxy-2-pyridinethiolato)oxovanadium(IV) significantly reduced infarct size and improved cardiac function in a dose-dependent manner [16]. That same group also showed that post-treatment with vanadyl sulfate significantly reduced the infarct size and significantly decreased the elevated left ventricular end-diastolic pressure, improved left ventricular developed pressure, and left ventricular contractility in a dose-dependent manner. Keyes et al. showed that the intravenous administration of bisperoxovanadium significantly reduced myocardial infarct size and improved cardiac function [30].

he elevated left ventricular end-diastolic pressure, improved left ventricular developed pressure, and left ventricular contractility in a dose-dependent manner. Keyes et al. showed that the intravenous administration of bisperoxovanadium significantly reduced myocardial infarct size and improved cardiac function [30]. As demonstrated by many of the studies referenced above, vanadium compounds activate protein kinase B (Akt) signaling through inhibition of protein tyrosine phosphatases. Akt is an important signaling molecule that modulates many cellular processes such as cell growth, survival, and metabolism. Hence, by activating Akt signaling, vanadium compounds elicit cytoprotection in brain and myocardial I/R injuries. This would make vanadium also a potential candidate for reducing I/R injury in the kidney.

important signaling molecule that modulates many cellular processes such as cell growth, survival, and metabolism. Hence, by activating Akt signaling, vanadium compounds elicit cytoprotection in brain and myocardial I/R injuries. This would make vanadium also a potential candidate for reducing I/R injury in the kidney. It must be acknowledged that our study has some limitations. Investigation of the dose-dependent response to BMOV, a longer follow-up after I/R, the acting mechanisms of BMOV, and histological evaluation of the effects of I/R and BMOV are the important limitations of our study. Since the potential benefits of BMOV in the context of I/R were never studied before, we chose to perform a relatively simple, short-term study in which we mainly focused on the effects of BMOV on renal oxygenation and function. However, longer studies are required in which different doses of BMOV are given and more detailed analysis of the involved pathways is done, together with histological evaluation. Furthermore, we used creatinine clearance rate and sodium reabsorption as measures of renal function, but we admit that these methods can lead to imprecision due to back leak phenomena and slight tubular creatinine secretion.

ore detailed analysis of the involved pathways is done, together with histological evaluation. Furthermore, we used creatinine clearance rate and sodium reabsorption as measures of renal function, but we admit that these methods can lead to imprecision due to back leak phenomena and slight tubular creatinine secretion. Conclusions In conclusion, pretreatment with the organic vanadium compound BMOV did not significantly affect the systemic or the renal hemodynamic and oxygenation variables and only partially protected renal sodium reabsorption after I/R. However, longer studies are required in which different doses of BMOV are given and more detailed analysis of the involved pathways is done, together with histological evaluation to fully understand the potential protective role of BMOV in the context of renal I/R. Competing interests All authors declare that they have no competing interest regarding this study. Authors’ contributions All authors have contributed significantly to the work and all authors are in agreement with the content of the manuscript. EA, RB, JB and CI participated in research design. EA and UA conducted experiments. Immunohistochemical analysis was performed by AK. EA and RB performed data analysis. EA, RB, DMJM, JB and CI contributed to the writing of the manuscript. RB, CD-T, and CI provided supervision. All authors read and approved the final manuscript.

ood warming by different kinds of pumps (e.g. continuous arteriovenous rewarming, centrifugal vortex blood pump) [19]. In a study of Garraway et al., fast rewarming from 29°C to 37°C (within 85 min versus 217 min) appeared to correct the coagulopathy sooner and consequently may reduce the risk of ongoing bleeding [19]. In conclusion, to the best of our knowledge no studies have investigated the specific effects of isolated accidental hypothermia on haemostasis. At least in part, hypothermia has been induced using cooling blankets, immersion in cool water or evaporation (Additional file 1: Table S1). Therefore, the isolated effects of accidental hypothermia as part of the lethal triad are essentially uncharacterized by previous models [38]. While induced hypothermia might be used to propose hypotheses on possible mechanism of the effects on coagulation, the definite major underlying pathomechanisms can only be assessed if hypothermia occurs spontaneously after trauma-haemorrhage, e.g. in the context of systemic hypoperfusion.

In this first issue of Intensive Care Medicine Experimental, two related studies describe the effects of antibodies directed against adrenomedullin (ADM), a vasoactive peptide that is released from various tissues during systemic inflammation, in murine sepsis. ADM, derived from a larger precursor peptide (pro-ADM), is released into the circulation during systemic inflammation, and the highest plasma concentrations are measured in patients with septic shock [1]. With a plasma half-life of approximately 20 min, it exerts multiple effects of which the most relevant appears to be vasodilation [2]. The consequent clinical sequelae of increased ADM concentrations are far from clear. While elevated ADM levels are associated with impaired outcome in sepsis patients [1] and the measurement of the circulating pro-peptide of ADM might serve as a marker to assess disease severity and predict mortality [3], administration of exogenous ADM actually improved outcome in animal models of septic shock [4].

r. While elevated ADM levels are associated with impaired outcome in sepsis patients [1] and the measurement of the circulating pro-peptide of ADM might serve as a marker to assess disease severity and predict mortality [3], administration of exogenous ADM actually improved outcome in animal models of septic shock [4]. Intriguingly, in this issue of Intensive Care Medicine Experimental, two studies demonstrate beneficial effects of anti-ADM antibodies administered in a murine cecal ligation and puncture (CLP) model [5, 6]. These studies represent the first attempts at antibody targeting of the ADM cascade in vivo. Struck et al. [5] describe the development of monoclonal antibodies directed against various epitopes of ADM and their ability to attenuate sepsis-induced mortality in severe murine sepsis (100% control group mortality within 2 to 3 days). The severity of this model means that any long-term deleterious or beneficial effects of the antibody remain unknown. Notably, all antibodies directed against ADM improved survival. Of interest, the antibody against the N-terminal moiety of ADM prevented mortality most effectively, while it inhibited ADM agonistic activity by only 25%. Wagner et al. [6] performed a more mechanistic study into the effects of HAM1101, the antibody targeted against the N-terminal part of ADM, in a resuscitated murine CLP model. They show that HAM1101 improves vasoactive responsiveness to catecholamines, attenuates systemic inflammation, and improves kidney function. The beneficial effects on the kidneys are likely attributable to reduced iNOS expression and nitrotyrosine formation in HAM1101-treated animals. The effects on inflammatory pmeters are surprising as ADM itself has also been reported to inhibit inflammation [7]. In summary, previous work has established that high endogenous levels of ADM correlate with poor outcomes and that exogenous administration improved outcomes, while now it is reported that antagonism of ADM using antibodies improves outcome in an animal sepsis model.

has also been reported to inhibit inflammation [7]. In summary, previous work has established that high endogenous levels of ADM correlate with poor outcomes and that exogenous administration improved outcomes, while now it is reported that antagonism of ADM using antibodies improves outcome in an animal sepsis model. How can we explain these apparently conflicting results? We naturally tend to prefer simple linear reasoning. However, reality, especially in the case of (patho)physiology, is more complex. ADM-mediated effects appear to represent yet another example of this fact. Timing and the magnitude of increase in ADM may be especially relevant in understanding its clinical effects in sepsis. In this respect, it is important to recognize that Struck et al. found that the least efficacious antibody in terms of antagonistic activity exerted the largest beneficial effect on survival. One could argue that other non-ADM-dependent effects of the antibody may play a role. However, it could also indicate that only very high concentrations of ADM exert deleterious effects, while modest increases of this peptide are beneficial. In addition to concentration-dependent and time-dependent effects, the specific clinical condition of sepsis, e.g., hyper- vs hypodynamic sepsis, is also likely to play a role in the observed discrepancies. It is clear that the path to clinical relevance at the bedside is long and bumpy, and based on the above described results, it is already clear that in sepsis patients, it will be a difficult task to define the therapeutic window of opportunity to modulate the ADM pathway. Nevertheless, further investigation of the effects of modulating this pathway is clearly warranted.

Background Over several decades, lung transplantation has become a consolidated treatment modality. However, the disproportion between organ supply and demand has not been solved. Consequently, donor criteria have been progressively expanded to increase the donor pool [1, 2]. In addition, new techniques such as extracorporeal lung reconditioning [3–5] that allow pharmacological [6, 7], gene [8], or cell [9] therapy have been proposed. As a consequence of these novel opportunities, the concept of organ ‘acceptability’ has been reconsidered: organs that were previously considered unsuitable for transplantation are now well accepted [4, 10–14]. However, while this extended suitability aids solving the organ shortage problem, it also raises several issues, among which the need to implement protocols of treatment in order to preserve over the donation process, if not ameliorate, the function of organs that a few years ago would not have been considered for transplantation.

le this extended suitability aids solving the organ shortage problem, it also raises several issues, among which the need to implement protocols of treatment in order to preserve over the donation process, if not ameliorate, the function of organs that a few years ago would not have been considered for transplantation. The aim of this investigation was to set and characterize a pig model that closely resembles the entire process of lung donation and transplantation. Each phase of the clinical procedure, including the pathophysiological changes induced by brain death, the complexity of donor management, and lung transplantation has been carefully reproduced. Further, we have applied a lung-protective ventilatory strategy in the donor animals and during the reperfusion phase of transplantation and measured gene expression of important cytokines, chemokines and markers of endothelial activation throughout the protocol. Hereafter, we present and discuss the results of our investigation. Methods This experimental study was performed after the Ethics Committee of the Fondazione IRCCS Ca’ Granda - Ospedale Maggiore Policlinico and the Italian Ministry of Health approved the protocol (Permit Number: 05/12). All surgeries were performed under anesthesia, and all efforts were made to minimize suffering. Experiments were performed in conformity to the revised Institute of Laboratory Animal Resources, Commission on Life Sciences, National Research Council ‘Guide for the Care and Use of Laboratory Animals’ National Academy Press, Washington, D.C., 1996 (http://www.nap.edu/catalog/5140.html).

e made to minimize suffering. Experiments were performed in conformity to the revised Institute of Laboratory Animal Resources, Commission on Life Sciences, National Research Council ‘Guide for the Care and Use of Laboratory Animals’ National Academy Press, Washington, D.C., 1996 (http://www.nap.edu/catalog/5140.html). A schematic overview of the protocol is shown in Figure 1. Each experiment was run using two animals (lung donor and transplantation recipient). The donor underwent induction of brain death followed by organ donor management for a total of 9 h after brain death induction; the lungs were then harvested and cold-stored for 8 h. The recipient pig underwent pneumonectomy and transplantation of the left donor lung; post-reperfusion follow-up was carried on for the next 6 h.Figure 1 Schematic overview of the experiment flow. Each experiment was run using two animals (lung donor and transplantation recipient). The donor underwent induction of brain death followed by organ donor management for a total of 9 h after brain death induction (BD); the lungs were then harvested and cold-stored for 8 h (Ischemia). The recipient pig underwent pneumonectomy and transplantation of the left donor lung, post-reperfusion follow-up was carried on for the following 6 hours (Graft). Arrows in the figure represent the following timings: BD refers to the end of donor management; Ischemia indicates the end of 8 h of cold storage and Graft the end of 6 h follow-up after graft reperfusion.

ansplantation of the left donor lung, post-reperfusion follow-up was carried on for the following 6 hours (Graft). Arrows in the figure represent the following timings: BD refers to the end of donor management; Ischemia indicates the end of 8 h of cold storage and Graft the end of 6 h follow-up after graft reperfusion. Anesthesia and monitoring Details of anesthesia and monitoring are described in Additional file 1. Briefly, animals received an intramuscular injection of olanzapine and tiletamine 2 mg (Zoletil, VIRBAC s.r.l., Milan, Italy) and medetomidine 1 mg (Domitor, Pfizer Animal Health, Exton, PA, USA and Div. of Pfizer Inc., New York, NY, USA). A continuous intravenous infusion of propofol (Diprivan, AstraZeneca, Basiglio, Milan, Italy) 10 to 15 mg/Kg/h and medetomidine 3 to 6 μg/Kg/h was then started. A number of catheters were positioned and secured in place to measure arterial, central venous, and pulmonary artery pressure throughout brain death induction, donor treatment, recipient surgery, and follow-up. Cardiac output was measured by the Swan-Ganz catheter. Extravascular lung water (EVLW), global end diastolic volume (GEDV), and stroke volume variation (SVV) were monitored by PiCCO2® (PULSION Medical Systems, AG, Stahlgruberring 28, München, Deutschland, Germany) monitor. Analysis of pO2, pCO2, pH, and derived variables (base excess, HCO3), together with electrolytes (Na+, K+, Ca2+, Cl−), glucose, and lactate concentrations was performed on arterial and central venous samples (Radiometer ABL 800 Flex, Radiometer Medical ApS, Brønshøj, Denmark). Urinary electrolytes concentrations were measured by K.I.N.G.® (Orvim s.r.l., Paderno Dugnano, Milan, Italy). Blood chemistry (blood urea nitrogen (BUN), creatinine, serum glutamic oxaloacetic transaminase (SGOT), serum glutamic-pyruvic transaminase (SGPT), troponin T) was also assessed at baseline and at the end of donor management (BD).

concentrations were measured by K.I.N.G.® (Orvim s.r.l., Paderno Dugnano, Milan, Italy). Blood chemistry (blood urea nitrogen (BUN), creatinine, serum glutamic oxaloacetic transaminase (SGOT), serum glutamic-pyruvic transaminase (SGPT), troponin T) was also assessed at baseline and at the end of donor management (BD). Induction of brain death To induce brain death, a slight modification of the protocol described by Purins et al. was used [15]. Briefly, to rise intracranial pressure (ICP), an 18-Fr Foley catheter (Willy Rush AG, Kernen, Germany) was placed in the epidural space and progressively inflated with saline (1.5 mL every 10 min). ICP was continuously measured by connecting a subdural probe (Integra Neurosciences, TraumaCath, Enterprise Drive, Plainsboro, NJ, USA) to a pressure transducer (TruWave, Edwards Lifesciences LLC, Irvine, CA, USA). The balloon of the Foley catheter was inflated until cerebral perfusion pressure, calculated as mean arterial pressure (MAP) minus ICP (CPP = MAP − ICP), was less than 0 mmHg. At each step of inflation, the microballoon of a 5-Fr pulmonary arterial catheter (Edwards Lifesciences LLC) positioned intraparenchymally was inflated with 1 mL of air; the maneuver was used to calculate intracranial compliance (IC = 1/ΔICP). Once CPP was negative, muscle paralysis, anesthesia, and analgesia were discontinued. Brain death was confirmed at the end of 60 min of CPP < 0 mmHg and before lung retrieval. Clinical signs of brain death included the absence of corneal reflex and the absence of coughing in response to tracheal suctioning. An apnea test was also performed. This was conducted during continuous positive airway pressure verifying the absence of breathing, confirmed by the absence of esophageal pressure deflections (SmartCath Viasys, Palm Springs, CA, USA) while pCO2 was above 60 mmHg (verified by arterial blood gas analysis). In three animals, brain death was also confirmed by electroencephalography (B. E. Light, EBNeuro S.p.A., Florence, Italy) [16]. In these animals, monitoring was extended throughout the protocol procedure, from before brain death induction to the end of brain death donor management.

fied by arterial blood gas analysis). In three animals, brain death was also confirmed by electroencephalography (B. E. Light, EBNeuro S.p.A., Florence, Italy) [16]. In these animals, monitoring was extended throughout the protocol procedure, from before brain death induction to the end of brain death donor management. An exemplificative CT scan shows the position of the intracranial catheters and the cerebral parenchyma deformation before and after the inflation of the epidural Foley catheter (Figure 2). A representative diagram of the site of brain catheter placement (Additional file 2: Figure S1), a radiograph of the epidural Foley catheter once inflated in the pig cranium (Additional file 3: Figure S2), and a photograph showing brain to inflated balloon proportions (Additional file 4: Figure S3) can be found in the Additional files.Figure 2 CT scan analysis of the skull at baseline (A) and after brain death induction (B) in one exemplificative animal. 1, epidural Foley catheter; 2, ICP monitoring catheter; 3, intraparenchymal Swan-Ganz catheter for measuring intracranial compliance.

ditional file 4: Figure S3) can be found in the Additional files.Figure 2 CT scan analysis of the skull at baseline (A) and after brain death induction (B) in one exemplificative animal. 1, epidural Foley catheter; 2, ICP monitoring catheter; 3, intraparenchymal Swan-Ganz catheter for measuring intracranial compliance. Organ donor management After brain death was confirmed, donor animals were treated according to standard physiological targets for the next 6 h [17, 18]. Cardiovascular targets included MAP > 60 mmHg, central venous pressure (CVP) between 5 and 8 mmHg and urine output > 1.5 mL/Kg/h. Hypotension was treated with vasoactive drugs when hemodynamic instability persisted despite adequate volume resuscitation. Norepinephrine was the first choice drug when systemic vascular resistances were below 800 dyn · s · cm−5. 1-Desamino-8-D-arginine vasopressin (0.125 to 0.250 μg endovenous) was given when diabetes insipidus occurred, defined as urine output > 4 mL/Kg/h, urinary specific gravity < 1.005, and blood osmolarity > 300 mOsm/Kg, calculated as [2 × Na] + [glucose/18]. We decided to use methylprednisolone by protocol (15 mg/kg) but not thyroid hormones as their use is suggested only when hemodynamic instability persists despite aggressive treatment [19–22]. After brain death confirmation, mechanical ventilation was set according to a lung-protective ventilatory strategy [18, 23]: tidal volumes of 6 to 8 mL/Kg, positive end-expiratory pressure (PEEP) of 8 to 10 cmH2O, and respiratory rate set to maintain pCO2 lower than 50 mmHg with 7.35 < pH < 7.45. Recruitment maneuvers were performed at the end of each apnea test allowing ten consecutive breaths with an inspiratory pressure target of 40 cmH2O above PEEP of 5 cmH2O.

mL/Kg, positive end-expiratory pressure (PEEP) of 8 to 10 cmH2O, and respiratory rate set to maintain pCO2 lower than 50 mmHg with 7.35 < pH < 7.45. Recruitment maneuvers were performed at the end of each apnea test allowing ten consecutive breaths with an inspiratory pressure target of 40 cmH2O above PEEP of 5 cmH2O. Lung harvest and cold storage A median sternotomy was performed, the thymus removed, the pleura carefully dissected, and the pericardium opened. The superior and inferior cava veins were encircled with silk ties and a bolus of 20,000 U heparin (Pharepa, Pharmatex Italia s.r.l., Milan, Italy) was injected into the jugular vein. Five minutes after the heparin bolus, a cannula was inserted into the main pulmonary artery. A bolus of 250 μg of alprostadil (Prostin, Pfizer Manufacturing Belgium N.V., Puurs, Belgium) was then injected into the main pulmonary artery. The superior and inferior cava veins were then ligated, the ascending aorta clamped, and the left atrial appendage transected. The lungs were then flushed with 60 ml/Kg of cold Perfadex® (Vitrolife Sweden Instruments AB, Billdal, Sweden) at a height of 30 cm above the heart. During the perfusion with the preservation solution, respiratory rate was decreased and FiO2 increased to 100%. Ventilation was discontinued when the heart-lung block was removed. Just before removing the lungs, the trachea was clamped with lungs fully inflated. After removal from the thoracic cavity, the heart-lung block was placed on ice, the heart removed and the lung perfused with Perfadex® in a retrograde manner (i.e., from the left veins to the pulmonary artery). Thereafter, the lungs were placed in a plastic bag (Vitrolife) containing Perfadex® solution and stored on ice for 8 h while continuously monitoring lung surface temperature (Medical Temperature Probes, Siemens S.p.A., Milan, Italy).

Perfadex® in a retrograde manner (i.e., from the left veins to the pulmonary artery). Thereafter, the lungs were placed in a plastic bag (Vitrolife) containing Perfadex® solution and stored on ice for 8 h while continuously monitoring lung surface temperature (Medical Temperature Probes, Siemens S.p.A., Milan, Italy). Lung transplantation A detailed description of the surgical procedure of lung transplantation is reported in Additional file 1. Briefly, the recipient pig was placed in the right lateral decubitus position, a left thoracotomy performed just below the tip of the scapula, and a left pneumonectomy completed. The donor lung was prepared on the back table while on ice then transferred to the thoracic cavity, and bronchial, pulmonary artery, and venous anastomoses were performed. The atrial clamp was then released to de-air the donor lung in a retrograde manner. Thereafter, the artery clamp was opened step by step in 10 min; the bronchial clamp was then removed allowing left lung ventilation.

sferred to the thoracic cavity, and bronchial, pulmonary artery, and venous anastomoses were performed. The atrial clamp was then released to de-air the donor lung in a retrograde manner. Thereafter, the artery clamp was opened step by step in 10 min; the bronchial clamp was then removed allowing left lung ventilation. Ventilation protocol was as follows: during pneumonectomy tidal volume was set at 6 to 8 mL/Kg, PEEP at 5 cmH2O, and FiO2 at 40%. By protocol, increments of PEEP were allowed if PaO2 was <100 mmHg or SpO2 < 95%. After reperfusion, pressure-controlled mode was instituted maintaining the same target volume of 6 to 8 mL/Kg; PEEP was set at 8 cmH2O, and a recruitment maneuver was performed (total target pressure of 45 cmH2O) 45 min after the start of controlled reperfusion. During post-reperfusion follow-up, if SpO2 was <90%, increments of PEEP were allowed up to 15 cmH2O; thereafter, FiO2 had to be increased in case of persistent hypoxia. Respiratory rate was set to maintain pCO2 below 70 mmHg and/or pH > 7.25. In case of persistent hypercapnia, increases of tidal volume were allowed. Cardiovascular targets included MAP > 60 mmHg, CVP between 5 and 8 mmHg, and urine output > 1.5 mL/Kg/h; care was given to administer the least amount of fluid possible. Cardiovascular, respiratory, and metabolic parameters were collected throughout reperfusion and post-reperfusion follow-up.

tidal volume were allowed. Cardiovascular targets included MAP > 60 mmHg, CVP between 5 and 8 mmHg, and urine output > 1.5 mL/Kg/h; care was given to administer the least amount of fluid possible. Cardiovascular, respiratory, and metabolic parameters were collected throughout reperfusion and post-reperfusion follow-up. Assessment of lung function and gene expression Oxygenation was assessed measuring partial pressure of oxygen from peripheral arterial blood samples (PaO2) and calculating the oxygenation index (OI) as (FiO2 × Pawm)/PaO2, where Pawm is mean airway pressure. The physiologic dead space fraction (VD/VT) was computed according to the following formula: VD/VT = (PaCO2 − PECO2)/PaCO2, where PECO2 is the mixed expired carbon dioxide partial pressure obtained by means of expiratory air sampling [24]. EtCO2/PaCO2 was also calculated. Respiratory mechanics was assessed partitioning lung and chest wall components by means of an esophageal balloon catheter, as detailed in the Additional file 1. End-expiratory lung volume (EELV) was measured using the closed circuit helium technique [25]. As index of lung edema, EVLW and wet-to-dry lung ratio (W/D) were measured according to standard procedures (see Additional file 1). Transcriptional expression of tumor necrosis factor alpha (TNFα), interleukin-1 beta (IL-1β), interleukin-6 (IL-6), interferon gamma (IFNγ), high mobility group box-1 (HMGB-1), chemokine CC motif ligand-2 (CCL2-MCP-1), chemokine CXC motif ligand-10 (CXCL-10), interleukin-8 (IL-8), endothelin-1 (EDN-1), intercellular adhesion molecule-1 (ICAM-1), vascular cell adhesion molecule-1 (VCAM-1), and selectin-E (SELE) was evaluated by real-time reverse transcription polymerase chain reaction (PCR) analysis performed on total mRNA isolated from lung tissue samples [26, 27]. A detailed description of the technique of mRNA isolation and gene expression measurement may be found in the Additional file 1.

1 (VCAM-1), and selectin-E (SELE) was evaluated by real-time reverse transcription polymerase chain reaction (PCR) analysis performed on total mRNA isolated from lung tissue samples [26, 27]. A detailed description of the technique of mRNA isolation and gene expression measurement may be found in the Additional file 1. Statistical analysis All results are presented as mean ± standard deviation (SD), unless otherwise specified. Continuous variables referring to neurological, cardiovascular, respiratory, and metabolic function were analyzed within donor or recipient animals by ANOVA for repeated measures followed, were appropriate, by Bonferroni test for all pairwise multiple comparisons. Data that were not normally distributed were investigated by ANOVA on ranks followed by Dunn's test for all pairwise comparisons. Parameters of lung function (PaO2/FiO2, OI, VD/VT, EELV, EVLW, EtCO2/PaCO2) taken before (baseline) and after BD and after lung transplantation (Graft) were assessed by means of ANOVA. The gene expression of lung biomarkers and W/D ratio were investigated at the end of BD, after 8 h of cold storage (Ischemia) and at the end of reperfusion follow-up (Graft) by one-way ANOVA and compared to controls (Control). For this purpose, three sham operated pigs were considered. P < 0.05 was accepted as significant. Data were analyzed using Sigma Plot version 11.0 (Systat Software, Inc., GmbH, Munich, Germany).

cold storage (Ischemia) and at the end of reperfusion follow-up (Graft) by one-way ANOVA and compared to controls (Control). For this purpose, three sham operated pigs were considered. P < 0.05 was accepted as significant. Data were analyzed using Sigma Plot version 11.0 (Systat Software, Inc., GmbH, Munich, Germany). Results A total of ten domestic pigs (five donors and five recipients) were consecutively included in the study and investigated as described in Figure 1. Three additional sham-operated animals were included and their lungs used as controls. Induction of brain death and donor management As shown in Figure 2, the inflation of the Foley catheter caused sovratentorial mass expansion that caused an increase in ICP (Figure 3A) and a decrease in CPP (Figure 3B). Intracranial compliance significantly dropped from 0.24 ± 0.13 to 0.02 ± 0.01 mL/mmHg (P < 0.05). When CPP was close to zero, transient hypertension and sustained tachycardia occurred, followed by severe hypotension (Figure 4). In all donor animals, clinical signs confirmed brain death at the end of brain death induction and throughout donor management. A representative pattern of electroencephalogic activity before and after brain death induction is shown in Additional file 5: Figure S4.Figure 3 Intracranial pressure (ICP) and cerebral perfusion pressure (CPP). The rise of ICP (A) and the decrease of CPP (B) during the induction of brain death and during donor management are shown in the figure. The error bars show the standard deviation of the mean.

induction is shown in Additional file 5: Figure S4.Figure 3 Intracranial pressure (ICP) and cerebral perfusion pressure (CPP). The rise of ICP (A) and the decrease of CPP (B) during the induction of brain death and during donor management are shown in the figure. The error bars show the standard deviation of the mean. Figure 4 Hemodynamic response during the induction of brain death and during the following hours of donor management. (A) Mean arterial pressure (MAP). (B) Heart rate (HR). The error bars show the standard deviation of the mean.

induction is shown in Additional file 5: Figure S4.Figure 3 Intracranial pressure (ICP) and cerebral perfusion pressure (CPP). The rise of ICP (A) and the decrease of CPP (B) during the induction of brain death and during donor management are shown in the figure. The error bars show the standard deviation of the mean. Figure 4 Hemodynamic response during the induction of brain death and during the following hours of donor management. (A) Mean arterial pressure (MAP). (B) Heart rate (HR). The error bars show the standard deviation of the mean. Cardiovascular, respiratory, and metabolic parameters collected during the hours of donor management are shown in Table 1. Over time, cardiac output significantly rose (P < 0.05) and (P < 0.05). Volume load (average 8 ± 4 mL/Kg/h) and noradrenaline infusion (average 0.04 ± 0.02 μg/Kg/min) were necessary in all animals. Central venous pressure slightly rose over time (P < 0.05); stroke volume variation was always within normal ranges. Lactate rose by the end of donor management (P < 0.05). Atrial fibrillation occurred in three pigs: infusion of amiodarone (150 mg in 250 mL dextrose 5% over a period of 20 to 30 min through a central venous catheter) decreased ventricular rate but did not revert to sinus rhythm. Clear signs of diabetes insipidus occurred in four out of five cases and were treated with 1-desamino-8-D-arginine vasopressin (1 to 3 μg) [28]. The lung-protective ventilatory strategy set after brain death confirmation resulted in a higher PEEP level (P < 0.05), lower tidal volume (P < 0.05), and higher respiratory rate (P < 0.05); PaCO2 and pH were not significantly different from baseline, while EtCO2 was higher (P < 0.05). Insulin administration was necessary to maintain blood concentrations of glucose to levels that were similar to baseline throughout the donor management protocol. Hypernatremia developed over time (P < 0.05) and was associated with a lower concentration of sodium in the collected urine (P < 0.05).Table 1 Donor parameters

sulin administration was necessary to maintain blood concentrations of glucose to levels that were similar to baseline throughout the donor management protocol. Hypernatremia developed over time (P < 0.05) and was associated with a lower concentration of sodium in the collected urine (P < 0.05).Table 1 Donor parameters Baseline Brain death diagnosis Treatment Procurement P value Temperature, °C 36.3 ± 0.9 37.7 ± 1.3a 37.9 ± 1.2a 38.2 ± 0.7a <0.05 Heart rate, beats/min 100 ± 20 129 ± 38a 145 ± 28a 163 ± 17ab <0.05 Mean arterial pressure, mmHg 75 ± 10 56 ± 8 68 ± 13 75 ± 10 <0.05* Pulmonary artery pressure, mmHg 18 ± 3 21 ± 4 19 ± 4 21 ± 4 0.100 Wedge pressure, mmHg 11 ± 3 13 ± 3 14 ± 1 13 ± 2 0.392 Central venous pressure, mmHg 5 ± 3 7 ± 3 7 ± 3 8 ± 3a <0.05 Cardiac output, L/min 2.9 ± 0.2 3.5 ± 0.6 4.4 ± 1.1a 5.1 ± 0.8ab <0.05 Systemic vascular resistance, dyn · s · cm−5 1,900 ± 269 1,146 ± 187a 1,127 ± 156a 1,051 ± 177a <0.05 Pulmonary vascular resistance, dyn · s · cm−5 195 ± 119 180 ± 65 100 ± 87 134 ± 79 0.103 Stroke volume variation, % 9 ± 2 10 ± 3 12 ± 5 11 ± 4 0.326 Global end diastolic volume, mL 360 ± 53 361 ± 76 394 ± 74 416 ± 108 0.564 Fluid balance, mL −507 ± 1,249 201 ± 1,360 399 ± 1,396 350 ± 1,707 0.136 Urine output, mL/Kg/h 5.2 ± 1.8 3.1 ± 0.6 8.3 ± 12.5 7.8 ± 5.2 0.521 Lactate, mmol/L 1.7 ± 0.5 2.1 ± 0.8 2.5 ± 2.3 4.8 ± 1.8 <0.05*

tion, % 9 ± 2 10 ± 3 12 ± 5 11 ± 4 0.326 Global end diastolic volume, mL 360 ± 53 361 ± 76 394 ± 74 416 ± 108 0.564 Fluid balance, mL −507 ± 1,249 201 ± 1,360 399 ± 1,396 350 ± 1,707 0.136 Urine output, mL/Kg/h 5.2 ± 1.8 3.1 ± 0.6 8.3 ± 12.5 7.8 ± 5.2 0.521 Lactate, mmol/L 1.7 ± 0.5 2.1 ± 0.8 2.5 ± 2.3 4.8 ± 1.8 <0.05* Oxygen delivery, mL/min 282 ± 31 285 ± 65 411 ± 95 483 ± 113ab <0.05 Respiratory rate, breaths/min 14 ± 3 17 ± 2a 19 ± 3a 19 ± 3ab <0.05 Tidal volume, mL 328 ± 50 284 ± 30a 286 ± 23a 288 ± 21a <0.05 Peak airway pressure, cmH2O 17 ± 1 19 ± 1 22 ± 2a 23 ± 2ab <0.05 Plateau airway pressure, cmH2O 12 ± 1 13 ± 1 14 ± 1 14 ± 1 0.061 Mean airway pressure, cmH2O 8 ± 0 11 ± 0a 12 ± 1a 12 ± 2a <0.05 Positive end-expiratory pressure, cmH2O 5 ± 0 7 ± 1a 8 ± 2a 8 ± 2a <0.05 End-tidal CO2, mmHg 43.0 ± 7.6 48.5 ± 6.4 48.9 ± 4.6 51.3 ± 2.8a <0.05 PaCO2, mmHg 40.5 ± 5.7 45.6 ± 3.2 43.6 ± 2.5 45.6 ± 5.2 0.168 pH 7.429 ± 0.036 7.411 ± 0.046 7.402 ± 0.057 7.358 ± 0.057 0.127 Glucose, g/dL 169 ± 56 147 ± 45 160 ± 82 162 ± 55 0.472 Na+, mEq/L 137.0 ± 4.3 137.6 ± 6.1 141.2 ± 4.1 144.2 ± 5.3ab <0.05 K+, mEq/L 4.2 ± 0.5 4.5 ± 0.5 4.2 ± 0.3 4.0 ± 0.1 0.375 Ca++, mEq/L 1.26 ± 0.01 1.18 ± 0.09 1.19 ± 0.06 1.17 ± 0.04a <0.05 Cl−, mEq/L 99.6 ± 2.5 98.6 ± 3.5 101.4 ± 2.9 106.0 ± 5.0abc <0.05 Na+ urinary, mEq/L 64.3 ± 20.4 60.8 ± 31.7 19.7 ± 18.2 7.3 ± 7.3ab <0.05 K+ urinary, mEq/L 29.5 ± 7.3 53.7 ± 23.2 23.2 ± 19.6 6.4 ± 3.3b <0.05 pHurinary 6.610 ± 0.320 6.821 ± 0.132 6.741 ± 0.631 6.229 ± 0.512 0.182 Specific gravityurinary 1.016 ± 2.5 1.011 ± 2.8 1.005 ± 0.0 1.006 ± 7.6a <0.05 Table 1 shows respiratory, hemodynamic, and metabolic parameters and urinary electrolytes of donor pigs collected at the end of surgery (baseline), after the induction and diagnosis of brain death (brain death diagnosis), 3 h after brain death confirmation (treatment), and at procurement (procurement, 9 h after brain death induction). Data are presented as mean ± standard deviation. One-way ANOVA repeated measures. P < 0.05 accepted as significant: avs. Baseline, bvs. Brain Death, and cvs. 3 h. *P < 0.05: When we tested the differences of main arterial pressure between the time points considered, a statistically significant difference was found (P < 0.05). To isolate the group or groups that differ from the others, a multiple comparison procedure was used (Bonferroni t test).

, bvs. Brain Death, and cvs. 3 h. *P < 0.05: When we tested the differences of main arterial pressure between the time points considered, a statistically significant difference was found (P < 0.05). To isolate the group or groups that differ from the others, a multiple comparison procedure was used (Bonferroni t test). The statistical software we used (Sigma Stat) is such that when no significant difference is found between the two groups with the higher difference of mean values, a result of ‘Do Not Test’ is provided by Bonferroni t test for the all other enclosed comparison. A ‘Do Not Test’ should be treated as if there is no significant difference between the means, even though the ANOVA test indicates that this is the case. This apparent discrepancy is due to the fact that differences are close to significance (P = 0.06) but below threshold (set at 0.05). This result may therefore be considered only as a nonsignificant trend. At the end of donor management, hepatic enzymes were not significantly different from baseline (SGOT 41.4 ± 22 vs. 151.2 ± 130.2 U/L, P = 0.145; SGPT 52.9 ± 10.1 vs. 38.2 ± 14.2 U/L, P = 0.124); creatinine and BUN were slightly but significantly higher (creatinine 1.1 ± 0.1 vs. 1.5 ± 0.2 mg/dL, P < 0.05; BUN 14.2 ± 2.2 vs. 36.5 ± 5.26 mg/dL, P < 0.05); there was a trend towards a rise of cardiac troponin T (10.5 ± 10.8 vs. 31.6 ± 15.2 pg/mL, P = 0.082).

0.145; SGPT 52.9 ± 10.1 vs. 38.2 ± 14.2 U/L, P = 0.124); creatinine and BUN were slightly but significantly higher (creatinine 1.1 ± 0.1 vs. 1.5 ± 0.2 mg/dL, P < 0.05; BUN 14.2 ± 2.2 vs. 36.5 ± 5.26 mg/dL, P < 0.05); there was a trend towards a rise of cardiac troponin T (10.5 ± 10.8 vs. 31.6 ± 15.2 pg/mL, P = 0.082). Lung harvest, preservation, and transplantation There were no complications concomitant to lung perfusion, harvest, and back table surgery. Temperature of the graft was always below 8°C during cold preservation. Time from cross clamp to reperfusion was 569 ± 28 min, of which 470 ± 24 min of cold and 98 ± 14 min of warm ischemia, respectively. Surgery was accomplished without major complications. Respiratory and hemodynamic data in the recipient are shown in Table 2; the data referring to the controlled reperfusion are shown in Additional file 6: Table S1 of the supplement material. Positive end-expiratory pressure was increased over time after reperfusion (P < 0.05). Mean arterial pressure and cardiac output progressively dropped (both P < 0.05). Lactate did not change over time, and no signs of inadequate perfusion were present.Table 2 Recipient parameters

6: Table S1 of the supplement material. Positive end-expiratory pressure was increased over time after reperfusion (P < 0.05). Mean arterial pressure and cardiac output progressively dropped (both P < 0.05). Lactate did not change over time, and no signs of inadequate perfusion were present.Table 2 Recipient parameters Baseline Reperfusion 1 h 4 h 6 h P value Temperature, °C 37.2 ± 0.8 37.9 ± 0.6 38.0 ± 0.9 37.9 ± 1.2 38.2 ± 1.2 0.153 Heart rate, beats/min 102 ± 24 103 ± 24 102 ± 28 105 ± 19 104 ± 4 0.891 Mean arterial pressure, mmHg 114 ± 19 99 ± 20 91 ± 21 80 ± 6a 76 ± 10a <0.05 Pulmonary artery pressure, mmHg 23 ± 5 25 ± 4 25 ± 3 24 ± 4 26 ± 4 0.424 Central venous pressure, mmHg 9 ± 4 6 ± 2 8 ± 3 8 ± 1 9 ± 2 0.105 Cardiac output, L/min 3.6 ± 0.4 4.2 ± 1.2 3.3 ± 0.4 3.3 ± 0.7 3.1 ± 0.6b <0.05 Systemic vascular resistance, dyn · s · cm−5 2,433 ± 396 1,877 ± 681 2,101 ± 763 1,787 ± 427 1,766 ± 407 0.183 Stroke volume variation, % 10 ± 4 6 ± 2 6 ± 1 9 ± 5 9 ± 4 0.385 Global end diastolic volume, mL 464 ± 57 436 ± 134 386 ± 51 456 ± 115 414 ± 109 0.890 Urine output, mL · Kg/h 4.3 ± 3.0 3.1 ± 0.7 4.9 ± 3.0 3.6 ± 2.2 2.0 ± 0.7 0.104 Lactate, mmol/L 1.2 ± 0.4 0.7 ± 0.1 1.0 ± 0.6 0.8 ± 0.2 0.8 ± 0.1 0.382 Oxygen delivery, mL/min 473 ± 65 628 ± 199 459 ± 61 428 ± 90b 385 ± 81b <0.05 Respiratory rate, breaths/min 18 ± 0 19 ± 4 19 ± 2 17 ± 4 17 ± 4 0.825 Tidal volume, mL 331 ± 47 295 ± 42 321 ± 99 363 ± 136 343 ± 116 0.393 Peak airway pressure, cmH2O 17 ± 1 20 ± 5 20 ± 3 21 ± 4 22 ± 5 0.157 Plateau airway pressure, cmH2O 11 ± 1 12 ± 2 15 ± 0 17 ± 3a 16 ± 4 <0.05 Mean airway pressure, cmH2O 8 ± 0 9 ± 1 12 ± 2ab 13 ± 2ab 12 ± 2ab <0.05 Positive end-expiratory pressure, cmH2O 5 ± 0 6 ± 2 8 ± 1ab 9 ± 1ab 9 ± 2ab <0.05 End-tidal CO2, mmHg 47.7 ± 5.2 55.4 ± 7.2 56.6 ± 4.5 49.0 ± 8.7 52.5 ± 6.2 0.096 PaCO2, mmHg 46.6 ± 5.2 52.0 ± 5.0 51.6 ± 8.4 41.2 ± 6.7 49.0 ± 11.3 0.053 pH 7.399 ± 0.029 7.349 ± 0.054 7.383 ± 0.057 7.456 ± 0.060b 7.403 ± 0.086 <0.05 Glucose, g/dL 161 ± 52 134 ± 70 112 ± 59 119 ± 32 124 ± 24 0.168 Na+, mEq/L 138.2 ± 0.8 140.3 ± 4.0 141.8 ± 3.5 140.2 ± 4.6 140.7 ± 4.5 0.226 K+, mEq/L 3.8 ± 0.1 4.7 ± 0.3 4.8 ± 0.8a 5.3 ± 0.7a 5.1 ± 0.7a <0.05 Ca++, mEq/L 1.32 ± 0.07 1.20 ± 0.06b 1.23 ± 0.07a 1.16 ± 0.08abc 1.15 ± 0.07abc <0.05 Cl−, mEq/L 98.8 ± 3.6 99.3 ± 2.1 101.0 ± 3.5 101.0 ± 3.7 101.0 ± 5.4 0.829 Na+ urinary, mEq/L 49.4 ± 19.9 - - 60.5 ± 40.3 - 0.475 K+ urinary, mEq/L 38.1 ± 21.4 - - 95.9 ± 46.8 - 0.053 pHurinary 6.801 ± 0.843 - - 6.702 ± 0.978 - 0.875 Table 2 shows respir

.07 1.20 ± 0.06b 1.23 ± 0.07a 1.16 ± 0.08abc 1.15 ± 0.07abc <0.05 Cl−, mEq/L 98.8 ± 3.6 99.3 ± 2.1 101.0 ± 3.5 101.0 ± 3.7 101.0 ± 5.4 0.829 Na+ urinary, mEq/L 49.4 ± 19.9 - - 60.5 ± 40.3 - 0.475 K+ urinary, mEq/L 38.1 ± 21.4 - - 95.9 ± 46.8 - 0.053 pHurinary 6.801 ± 0.843 - - 6.702 ± 0.978 - 0.875 Table 2 shows respir atory, hemodynamic, and metabolic parameters and urinary electrolytes of recipient pigs collected after surgery (Baseline), at the beginning of reperfusion (Reperfusion) and after 1, 4, and 6 h after reperfusion. Data are presented as mean ± standard deviation. One-way ANOVA repeated measures. P < 0.05 accepted as significant: avs. Baseline, bvs. Reperfusion, and cvs. 1 h. Assessment of lung function and gene expression Table 3 shows functional respiratory parameters in the donor and recipient animals. The ratio between EtCO2 and PaCO2, elastance of the respiratory system, EELV and functional dead space were not significantly different at BD or at Graft. There were no signs of lung edema in either BD or Graft, as assessed by EVLW and W/D lung ratio. In three cases, oxygenation of the implanted lung (Graft) was assessed after reperfusion by selective left pulmonary vein and artery blood gas analysis: arterial PaO2/FiO2 was 532 ± 19 mmHg with PvO2 of 34 ± 5 mmHg.Table 3 Lung function parameters

ung edema in either BD or Graft, as assessed by EVLW and W/D lung ratio. In three cases, oxygenation of the implanted lung (Graft) was assessed after reperfusion by selective left pulmonary vein and artery blood gas analysis: arterial PaO2/FiO2 was 532 ± 19 mmHg with PvO2 of 34 ± 5 mmHg.Table 3 Lung function parameters Donor Recipient Baseline Before procurement After reperfusion P value PaO2/FiO2, mmHg 540 ± 32 480 ± 31 532 ± 19 0.564 Elastance of respiratory system, cmH2O/L 20.5 ± 7.9 18.4 ± 7.7 22.6 ± 6.9 0.687 Elastance of lung, cmH2O/L 12.9 ± 5.2 15.3 ± 7.7 15.3 ± 7.0 0.846 End expiratory lung volume, mL 735 ± 187 803 ± 312 749 ± 213 0.900 Physiologic dead space fraction 0.54 ± 0.03 0.51 ± 0.07 0.49 ± 0.08 0.631 End-tidal CO2/PaCO2 1.06 ± 0.04 1.13 ± 0.09 1.19 ± 0.08 0.133 Extravascular lung water, mL 359 ± 79 359 ± 80 366 ± 117 0.949 Wet/dry ratio 6.2 ± 7.0 5.8 ± 0.5 5.6 ± 0.6 0.629 Table 3 shows functional lung parameters in the donor and recipient animals. ‘Before procurement’ refers to data taken at the end the donation process at the time of lung procurement, except for wet-to-dry ratio that was measured in sham-operated animals. BD refers to data taken at the end of donor management; ‘After reperfusion’ refers to data taken at the end of post-reperfusion follow-up. Oxygenation of the implanted lung (After reperfusion, n = 3) was assessed by selective left pulmonary vein blood gas analysis. Data are presented as mean ± standard deviation. One-way ANOVA. P < 0.05 accepted as significant.

management; ‘After reperfusion’ refers to data taken at the end of post-reperfusion follow-up. Oxygenation of the implanted lung (After reperfusion, n = 3) was assessed by selective left pulmonary vein blood gas analysis. Data are presented as mean ± standard deviation. One-way ANOVA. P < 0.05 accepted as significant. Figure 5 shows results of gene expression analysis in lung samples obtained at control, BD, cold ischemia (Ischemia) and after transplantation (Graft). Expression levels of the cytokines TNF α, IL-1β, IL-6, and IFN γ were not significantly different from control expression at any of the examined points (panel A). Increased expression of the chemokines CCL2-MCP-1 at Graft (P < 0.05) and IL-8 at both BD and Graft (P < 0.05) was observed (panel B). CXCL-10 was not altered at either BD, Ischemia, or Graft, whereas expression of HMGB-1 was significantly lower at Graft (P < 0.05, panel B). Adhesion molecules EDN-1, ICAM-1, and VCAM-1 were not affected by either BD, Ischemia, or Graft, while the expression of SELE was significantly higher at Graft (P < 0.05, panel C).Figure 5 Gene expression of lung inflammatory mediators obtained by real-time PCR in lung homogenates. Lung samples for gene expression analysis were obtained at BD (at the end of donor management), after 8 h of cold storage (Ischemia), at the end of reperfusion follow-up (Graft) and compared to controls (Control). Cytokines, interleukin-1 beta (IL-1β), interleukin-6 (IL-6), tumor necrosis factor alpha (TNF α), and interferon gamma (IFN γ) are shown in (A); chemokines, chemokine C-C motif ligand 2 (CCL2-MCP-1), chemokine CXC motif ligand 10 (CXCL-10), interleukin-8 (IL-8), and the oxidative stress index high mobility group box-1 (HMGB-1) are shown in (B); endothelial mediators: endothelin-1 (EDN-1), intercellular adhesion molecule-1 (ICAM-1), vascular cell adhesion molecule-1 (VCAM-1) and selectin-E (SELE) are shown in (C). The error bars show the standard deviation of the mean. One-way ANOVA vs. Control. P < 0.05 accepted as significant: asterisk (*) vs. Control.

(B); endothelial mediators: endothelin-1 (EDN-1), intercellular adhesion molecule-1 (ICAM-1), vascular cell adhesion molecule-1 (VCAM-1) and selectin-E (SELE) are shown in (C). The error bars show the standard deviation of the mean. One-way ANOVA vs. Control. P < 0.05 accepted as significant: asterisk (*) vs. Control. Discussion This study describes a pig model of brain death, donor management, and lung transplantation that closely resembles clinical conditions. The research effort was to reproduce each critical phase of donation and transplantation in a standardized and optimized fashion and to integrate the clinical approach with biofunctional evidence of lung injury. Apart from decapitation [29] and cerebral hemorrhage [30], most of the animal investigations on brain death have utilized models of sovratentorial mass expansion to rise ICP [26, 31–36]. While Neyrinck et al. and Ryan et al. induced brain death by an explosive rise of ICP [34, 35], we elected to rise ICP slowly reproducing the protocol recently described by Purins et al. [15]. The model we used resembles that of a progressively expanding mass that brings to complete ischemia. Despite the specific limitations of an animal model, it could resemble an ICP rise that would occur in case of an untreated trauma or a spontaneous cerebral hemorrhage.

protocol recently described by Purins et al. [15]. The model we used resembles that of a progressively expanding mass that brings to complete ischemia. Despite the specific limitations of an animal model, it could resemble an ICP rise that would occur in case of an untreated trauma or a spontaneous cerebral hemorrhage. Increased intracranial pressure was associated with a decrease in CPP and a drop of intracranial compliance. Similarly to Purins et al., ICP rise was associated with a transient hypertension and tachycardia, followed by severe hypotension. These phenomena occurred at CPP values similar to those recently described in a refinement paper of Purins et al. [33], further confirming the validity of the model. In our model, clinical signs of brain death were evident after 1 h of negative CPP in all animals, including assessment of corneal reflex, coughing in response to tracheal suctioning and execution of an apnea test. Moreover, in a subset of animals, the absence of brain perfusion was confirmed by electroencephalography.

In our model, clinical signs of brain death were evident after 1 h of negative CPP in all animals, including assessment of corneal reflex, coughing in response to tracheal suctioning and execution of an apnea test. Moreover, in a subset of animals, the absence of brain perfusion was confirmed by electroencephalography. Intracranial pressure and CPP were maintained for an extended number of hours to better reproduce the clinical setting. This is a major distinction relative to previous investigations. Indeed, observation times after brain death were 120 min in Purins' research [15], 300 in Neyrinck [34]; 360 in Lyons [31], McLean [32], and Barklin [26], and 480 in Hvas [30]. We prolonged the time from brain death to lung harvested after a total of 540 min after cerebral mass expansion. Even if this time was not as long as that of the investigation of Stieger et al.'s [37], this protracted period allowed to confirm brain death over time and opened to the long-term pathophysiological sequelae of brain death adding to the clinical relevance of the model. In fact, typical cardiovascular, metabolic, and electrolyte derangements of brain death overtly occurred. A distinctive feature of this study is that we treated brain death according to standard clinical practice adopting well-accepted physiological targets [17, 18], whereas others did not [34, 36]. In fact, after brain death induction, there was cardiovascular instability, including a progressive drop of systemic vascular resistance and a rise of cardiac output that required vasoactive drugs and a positive fluid balance for normalization. To optimize ventilatory management, a lung-protective ventilatory bundle of treatment was adopted comprehensive of high PEEP - low tidal volume, apnea test execution during continuous positive airway pressure - CPAP, and lung recruitment maneuvers. As shown by Mascia et al. [23], this strategy likely contributed to prevent lung collapse and allowed to meet standard inclusion criteria for lung donation in all animals: mean PaO2/FiO2 at PEEP 5 cmH2O and FiO2 100% was 480 ± 31 mmHg, well above the conventional threshold for acceptability set at 300 mmHg [2, 38]. Overall, management of pigs after brain death closely mimicked the clinical challenge of treating the multi-organ donors, thus offering a standardized point of reference.

als: mean PaO2/FiO2 at PEEP 5 cmH2O and FiO2 100% was 480 ± 31 mmHg, well above the conventional threshold for acceptability set at 300 mmHg [2, 38]. Overall, management of pigs after brain death closely mimicked the clinical challenge of treating the multi-organ donors, thus offering a standardized point of reference. While previous research has generally focused attention on either the donor or the recipient side, we have accurately reproduced the entire process of organ donation and transplantation. After brain death induction and donor treatment, the lungs were retrieved and cold-stored for 8 h. This time interval represents a realistic frame for potential lung preservation and/or reconditioning strategies. Similar to other investigations, we have then used a single lung transplantation procedure [39, 40]. Lung function early after reperfusion has been carefully monitored, as reperfusion injury is generally considered an outcome measure in lung preservation studies [41]. For this reason, based on the known effects of perfusion and ventilation on ischemia-reperfusion lung injury [42, 43], meticulous attention was given to the reperfusion protocol. The clamp on pulmonary artery was opened stepwise to allow progressive accommodation of blood flow in the newly reperfused lung vasculature and to limit as much as possible shear stress forces. This procedure also contributed to dilute residual unflushed organic acid accumulated within the lung during warm ischemia, possibly dumping the systemic effects of reperfusion. During the first minutes of reperfusion, the ventilatory component of reperfusion injury was completely abolished. Ventilation was in fact resumed only 15 min after reperfusion. At this time, attention was given to avoid ventilator-induced lung injury: a low volume - high PEEP strategy was adopted, recruitment maneuver was postponed, and residual atelectasis initially tolerated to avoid stress load to the endothelial-epithelial barrier likely to occur at high end-inspiratory lung volumes. While tolerating relative hypoxia, oxygen inspiratory fraction was kept in the low range to avoid oxidative stress [44]. These targets were only transiently set, and a full open lung strategy was resumed within 45 min. We adopted this strategy based on the opinion that a slow transition from ischemia to reperfusion is of primary importance to modulate both endothelial and epithelial ischemia-reperfusion injury in the lung.

e stress [44]. These targets were only transiently set, and a full open lung strategy was resumed within 45 min. We adopted this strategy based on the opinion that a slow transition from ischemia to reperfusion is of primary importance to modulate both endothelial and epithelial ischemia-reperfusion injury in the lung. Careful titration of pre-load indexes and ventilation settings was also adopted during post-reperfusion follow-up. This allowed to terminate the experiments with the implanted lung free of edema, as assessed by wet-to-dry lung ratio, and with normal extravascular lung water and oxygenation indexes. While other assessed the function of the implanted lung by blood gas analysis after ligation of the contralateral pulmonary artery, we implemented a different strategy. Indeed, the stress test given by the entire cardiac output flowing through a recently ischemic lung is certainly useful to reveal a possible frailty of the implanted lung. However, such evaluation protocol imposes an innatural hemodynamic challenge to both pulmonary vasculature and right heart that often leads to severe hemodynamic failure. There are in fact authors that report death at reperfusion, early after contralateral pulmonary artery ligation [45–47]. We believe that selective lung arterial and venous blood gas analysis accurately assess lung function without hemodynamic confounding factors. A limitation of this approach is that measures of respiratory mechanics reflect both native lung and graft. However, avoidance of the stress hemodynamic test allowed to better investigate the biology of ischemia-reperfusion lung injury.

s analysis accurately assess lung function without hemodynamic confounding factors. A limitation of this approach is that measures of respiratory mechanics reflect both native lung and graft. However, avoidance of the stress hemodynamic test allowed to better investigate the biology of ischemia-reperfusion lung injury. In this perspective, it is interesting to note that even if all efforts to gently treat the lungs were adopted and physiological endpoints were successfully pursued during donor treatment and after transplantation, there was a clear activation of inflammation in the lung [48, 49]. In a clinical setting, achievement of these physiological parameters at the end of donor or recipient treatment would be certainly satisfactory. However, in spite of this excellent clinical outcome, enhanced expression of the chemokines CCL2-MCP-1 and IL-8 as well as increased transcription of SELE at different phases of the transplantation process suggest the presence of an inflammatory reaction [50]. Of particular interest, increased expression of IL-8 occurred during donor management. This observation confirms the idea that brain death per se induces inflammation in peripheral organs [51, 52]. Indeed, increased production of the neutrophil chemoattractant IL-8 after brain death can facilitate subsequent reperfusion injury. In addition, increased chemokine production at the Graft point can promote rejection. Of note, our reperfusion strategy protected against oxidative injury, as suggested by HMGB-1 downregulation [53, 54].

ed production of the neutrophil chemoattractant IL-8 after brain death can facilitate subsequent reperfusion injury. In addition, increased chemokine production at the Graft point can promote rejection. Of note, our reperfusion strategy protected against oxidative injury, as suggested by HMGB-1 downregulation [53, 54]. Mascia et al. showed that a conventional strategy of lung management during brain death was associated with a rise of plasma cytokines over time, while a protective strategy was not [23]. Here, we show that a lung-protective ventilation throughout brain death donor management and transplantation did not prevent the activation of inflammation in the lung. Both findings underscore the complexity of the interaction between ischemia-reperfusion and mechanical ventilation in lung transplantation and call to the need for reliable models that reproduce each phase of lung donation and transplantation. In fact, an increasing number of transplantations are performed with lungs from marginal donors, and the complex clinical settings often preclude full understanding of new treatment modalities cause-effect relationships.

all to the need for reliable models that reproduce each phase of lung donation and transplantation. In fact, an increasing number of transplantations are performed with lungs from marginal donors, and the complex clinical settings often preclude full understanding of new treatment modalities cause-effect relationships. There are some limitations to the study. In fact, the model resembles that of a progressively expanding mass that brings to complete ischemia, but only under the specific limitations of an animal model. The model is reproducible and with low variablity, yet some differences between animals are present (see atrial fibrillation that occurred in three animals, for instance). The animals investigated are enough to conclude about the validity of the model, but we realize that the absolute number is low, with some reflections to statistical results. Conclusions In conclusion, we have set a pig model that closely resembles the entire process of organ donation and lung transplantation, and we have shown that activation of inflammation in the lung was present despite of an optimized ventilatory management throughout the protocol. The findings of our investigation may represent a starting point for various studies of lung transplantation in a standardized setting. Electronic supplementary material Additional file 1: Supplement. A more detailed description of the section ‘Methods’ of this experimental work may be found in this Additional file. (DOC 54 KB) Additional file 2: Figure S1: Representative diagram of the site of brain catheters placement. (TIF 65 KB)

Conclusions In conclusion, we have set a pig model that closely resembles the entire process of organ donation and lung transplantation, and we have shown that activation of inflammation in the lung was present despite of an optimized ventilatory management throughout the protocol. The findings of our investigation may represent a starting point for various studies of lung transplantation in a standardized setting. Electronic supplementary material Additional file 1: Supplement. A more detailed description of the section ‘Methods’ of this experimental work may be found in this Additional file. (DOC 54 KB) Additional file 2: Figure S1: Representative diagram of the site of brain catheters placement. (TIF 65 KB) Additional file 3: Figure S2: Radiograph of the epidural Foley catheter once inflated in the cranium of the pig. (TIF 542 KB) Additional file 4: Figure S3: Photograph showing brain to inflated balloon proportions. (TIF 612 KB) Additional file 5: Figure S4: Representative pattern of electroencephalogic activity before (upper panel) and after brain death induction (lower panel). (TIF 192 KB) Additional file 6: Table S1: Controlled reperfusion parameters. (DOC 47 KB) Abbreviations BDthe end of donor management after brain death BUNblood urea nitrogen CCL2-MCP-1chemokine CC motif ligand-2 CXCL-10chemokine CXC motif ligand-10 CPPcerebral perfusion pressure CTcomputed tomography CVPcentral venous pressure EDN-1endothelin-1 EELVend-expiratory lung volume EVLWextravascular lung water EtCO2end tidal carbon dioxide FiO2inspiratory fraction of oxygen GEDVglobal end diastolic volume HMGB-1high mobility group box-1

CCL2-MCP-1chemokine CC motif ligand-2 CXCL-10chemokine CXC motif ligand-10 CPPcerebral perfusion pressure CTcomputed tomography CVPcentral venous pressure EDN-1endothelin-1 EELVend-expiratory lung volume EVLWextravascular lung water EtCO2end tidal carbon dioxide FiO2inspiratory fraction of oxygen GEDVglobal end diastolic volume HMGB-1high mobility group box-1 ICintracranial compliance ICAM-1intercellular adhesion molecule-1 ICPintracranial pressure IL-1βinterleukin-1 beta IL-6interleukin-6 IFNγinterferon gamma IL-8interleukin-8 MAPmean arterial pressure OIoxygenation index PaO2partial arterial pressure of oxygen pCO2partial pressure of carbon dioxide PCRpolymerase chain reaction PEEPpositive end-expiratory pressure SpO2peripheral saturation of hemoglobin with oxygen SELEselectin-E SVVstroke volume variation TNFαtumor necrosis factor alpha VCAM-1vascular cell adhesion molecule-1 VD/VTdead volume to tidal volume ratio - physiological dead space W/Dwet-to-dry ratio. Competing interests The authors of this manuscript have no conflict of interest to disclose. Authors' contributions FG and LG conceived and designed the experiments. FV, SC, SF, GMR, JF, AV, LR, PM, GC, and SG performed the experiments. FV, SC, SF, NS, and GMR analyzed the data. CL, AC, and PL contributed to PCR analysis. FV, SC, SF, and LG wrote the paper. All authors read and approved the final manuscript.

The authors of this manuscript have no conflict of interest to disclose. Authors' contributions FG and LG conceived and designed the experiments. FV, SC, SF, GMR, JF, AV, LR, PM, GC, and SG performed the experiments. FV, SC, SF, NS, and GMR analyzed the data. CL, AC, and PL contributed to PCR analysis. FV, SC, SF, and LG wrote the paper. All authors read and approved the final manuscript. The authors thank Fabio Ambrosetti for his valuable technical support, Anna Catania for extensive review of the manuscript, Gianluca Ardolino and Alberto Facchini for assistance on electroencephalographic measurements, and all the nurses of the lung transplantation unit of the Fondazione IRCCS Ca’ Granda who participated in the experiments. The authors also thank Luciano Lombardi for his technical assistance on CT scan brain imaging analysis. This study was funded by Fondazione IRCCS Ca’ Granda - Ospedale Maggiore Policlinico, Milan, Italy.

Background Vitamin C has been shown to have beneficial effects on the microcirculation in moderate sepsis in the rat [1, 2]. On a consensus meeting on vitamin C in acute endothelial pathophysiological conditions in 2006, it was concluded that there were arguments based on experimental studies for the hypothesis that high-dose vitamin C improves microvascular endothelial function in sepsis [3]. This hypothesis was further supported by some recent studies in septic mice. Thus, Fisher et al. showed that vitamin C has positive effects on various pathophysiological changes in sepsis, including the microvasculature of the lung [4, 5], and Zhou et al. showed that vitamin C decreases capillary leakage of different injected tracers [6]. Several experimental studies in different animal models have shown that vitamin C is also beneficial in burns by preventing capillary leakage, lymph flow and resuscitation fluid requirements [7–9]. In a more recent study, vitamin C treatment was shown to reduce the endothelial damage caused by transfusion of plasma from a burned donor rat [10]. Further, a human study showed a reduction in resuscitation volume with vitamin C treatment after severe burn [11]. In contrast, Aliabadi-Wahle et al. found no changes in microvascular permeability or in oedema formation when vitamin C was given after burn in the dog [12].

sfusion of plasma from a burned donor rat [10]. Further, a human study showed a reduction in resuscitation volume with vitamin C treatment after severe burn [11]. In contrast, Aliabadi-Wahle et al. found no changes in microvascular permeability or in oedema formation when vitamin C was given after burn in the dog [12]. Even though not fully understood, suggested mechanisms behind the described beneficial effects of intravenous vitamin C treatment, in both sepsis and burns, are scavenging of reactive oxygen species, reduction of endothelial adhesion molecules and modulation of nitric oxide production [3, 13–15].

sfusion of plasma from a burned donor rat [10]. Further, a human study showed a reduction in resuscitation volume with vitamin C treatment after severe burn [11]. In contrast, Aliabadi-Wahle et al. found no changes in microvascular permeability or in oedema formation when vitamin C was given after burn in the dog [12]. Even though not fully understood, suggested mechanisms behind the described beneficial effects of intravenous vitamin C treatment, in both sepsis and burns, are scavenging of reactive oxygen species, reduction of endothelial adhesion molecules and modulation of nitric oxide production [3, 13–15]. Sepsis, as well as burns, causes transcapillary leakage of plasma, reducing the circulating plasma volume [16, 17]. As discussed above, experimental studies have shown that vitamin C reduces local oedema and leakage of plasma markers. However, these findings do not necessarily reflect a decrease in plasma volume loss, and so far, no study has specifically investigated the effect of vitamin C treatment on plasma volume. In the present study, we therefore tested the hypothesis that vitamin C would reduce the loss of plasma volume in the early stage of sepsis. Treatment was initiated 3 h after induction of sepsis, a more clinically relevant time point than used in most studies found in the current literature, where treatment was started either before or closely after injury (e.g. sepsis, burns). Different dose regimes have been used in previous studies. Beneficial effects on microcirculation have been shown with low-dose treatment [1, 2], but it seems that higher doses are needed to counteract microvascular leakage. We chose to compare two different treatment regimes, both previously shown to be effective to prevent capillary leakage - one with a small bolus dose followed by a continuous infusion [9, 10] and one with a high bolus dose as single treatment [4, 6]. A sham group that underwent the same surgical procedure, but received no treatment, was also included in the study.

gimes, both previously shown to be effective to prevent capillary leakage - one with a small bolus dose followed by a continuous infusion [9, 10] and one with a high bolus dose as single treatment [4, 6]. A sham group that underwent the same surgical procedure, but received no treatment, was also included in the study. Methods Anaesthesia and set-up The study was approved by the Ethical Committee for Animal Research at Lund University, Sweden (application no. M180-10). The animals were treated in accordance with the guidelines of the National Institutes of Health for Care and Use of Laboratory Animals. Male adult Sprague-Dawley rats were used, weighing 337 ± 26 g (mean ± SD). Anaesthesia was induced using a covered glass container with a continuous supply of 5% isoflurane in air (Forene® 100%; Abbot Scandinavia AB, Solna, Sweden), in which the animals were placed. After induction, the animals were removed from the container, and anaesthesia was maintained with 1.5% to 1.8% isoflurane in air using a mask, while tracheostomy was performed. Thereafter, the animals were connected to a ventilator (Ugo Basile; Biological Research Apparatus, Comerio, Italy) and ventilated in a volume-controlled mode with a positive end expiratory pressure of 4 cm H2O. End-tidal PCO2 was continuously monitored (Capstar-100; CWE, Ardmore, PA, USA). Anaesthesia was maintained with 1.5% to 1.8% isoflurane in air throughout the experiment. Body temperature, measured rectally, was kept at 37.1°C to 37.3°C using a feedback-controlled heating pad. The left femoral artery was cannulated to monitor arterial blood pressure and to obtain blood samples for analysis of electrolytes, haematocrit, lactate, arterial blood gases (I-STAT; Abbot Point of Care Inc, Abbot Park, IL, USA) and plasma volumes. The left femoral vein was cannulated and used for infusions, and kept open with a continuous infusion of saline at 0.2 μL/min. The right internal jugular vein was cannulated and used for injection of 125I-albumin for plasma volume measurements. At the end of the experiments, the animals were sacrificed with an intravenous injection of potassium chloride.

and used for infusions, and kept open with a continuous infusion of saline at 0.2 μL/min. The right internal jugular vein was cannulated and used for injection of 125I-albumin for plasma volume measurements. At the end of the experiments, the animals were sacrificed with an intravenous injection of potassium chloride. Experimental procedure A well-established rat model of severe sepsis was used [16, 18]. A longitudinal midline skin incision in the abdominal wall with diathermia was performed, followed by laparotomy by incision along the linea alba. After ligation just below the ileocaecal valve, an incision of 1 cm in length was made in the caecum, allowing leakage of faeces into the abdominal cavity, thereby inducing sepsis/systemic inflammatory response syndrome (SIRS). The abdominal wall and the skin were then closed with clips. There was no bleeding during the experiment.

n just below the ileocaecal valve, an incision of 1 cm in length was made in the caecum, allowing leakage of faeces into the abdominal cavity, thereby inducing sepsis/systemic inflammatory response syndrome (SIRS). The abdominal wall and the skin were then closed with clips. There was no bleeding during the experiment. Plasma volume Plasma volume (PV) was determined with a reliable and established technique, shown to produce reproducible and reliable results [16, 19–21]. As described previously [16], PV was determined by measuring the radioactivity in 100 μL of plasma taken 5 min after an intravenous injection of human 125I-albumin (0.5 mL) with a known amount of activity. The increase in radioactivity was calculated by subtracting the activity in a blood sample taken just before the injection from that taken 5 min after the injection, thereby adjusting for any remaining radioactivity from previous measurements. To calculate the amount of radioactivity given, the remaining activity in the emptied vial, syringe and needle used was measured and subtracted from the total activity in the prepared dose. Sources of error are small with the technique used. Free iodine was measured regularly following precipitation with 10% trichloroacetic acid and was found to be less than 2.0% in the prepared samples. Radioactivity was measured with a gamma counter (Wizard 1480; LKB-Wallac, Turku, Finland).

tivity in the prepared dose. Sources of error are small with the technique used. Free iodine was measured regularly following precipitation with 10% trichloroacetic acid and was found to be less than 2.0% in the prepared samples. Radioactivity was measured with a gamma counter (Wizard 1480; LKB-Wallac, Turku, Finland). Experimental protocol In this study, we evaluated the effect of intravenous vitamin C on plasma volume in the early stage of sepsis in the rat. The septic rats were divided into three groups: a bolus + infusion group (the B + I group, n = 9), a bolus group (the B group, n = 9) and a sham group (the S group, n = 9). Animals that did not show a decrease in PV 3 h after the preparation were considered to be non-septic and were excluded from the study. These animals and animals that died before the end of the experiment were replaced with new animals.

, n = 9), a bolus group (the B group, n = 9) and a sham group (the S group, n = 9). Animals that did not show a decrease in PV 3 h after the preparation were considered to be non-septic and were excluded from the study. These animals and animals that died before the end of the experiment were replaced with new animals. After cannulation and surgical preparation, the animals were left undisturbed for 3 h, a time period previously shown to be sufficient for systemic inflammation and plasma leakage to develop [16]. Three hours after surgical preparation, the treatment was initiated. The B + I group received an intravenous injection of ascorbic acid (2,3-didehydro-l-threo-hexono-1,4-lactone, Askorbinsyra 100 mg/ml, APL, Stockholm, Sweden) of 66 mg/kg, followed by an infusion of 33 mg/kg/h during the rest of the experiment (Figure 1). In previous studies, this dose regime has been shown to be effective in decreasing microvascular permeability after burns in the rat [9, 10]. The B group received a single intravenous bolus injection of 200 mg/kg of ascorbic acid (Figure 1), previously shown to be effective in septic mice [4, 6]. The S group received no treatment, as it was meant to represent a non-treatment situation.Figure 1 Time scale of the experiment. PV1, plasma volume at baseline; PV2, plasma volume 3 h after surgical preparation just before the start of treatment; PV3, plasma volume at the end of the experiment; ABG, arterial blood sample for analysis of blood gases, hematocrit, lactate and electrolytes.

-treatment situation.Figure 1 Time scale of the experiment. PV1, plasma volume at baseline; PV2, plasma volume 3 h after surgical preparation just before the start of treatment; PV3, plasma volume at the end of the experiment; ABG, arterial blood sample for analysis of blood gases, hematocrit, lactate and electrolytes. Plasma volumes were measured at baseline, at 3 h after the end of surgical preparation and at the end of the experiment another 3 h later. Blood samples for measurement of arterial pH, PCO2, PO2, lactate, haematocrit, sodium and potassium were taken at the same time points. Urine was collected in a glass vial placed at the external meatus of the urethra throughout the whole experiment, and the bladder was emptied by external compression at the end of the experiment. Statistical analysis Statistical analyses were performed with GraphPad Prism software version 5.0c for Mac OS X (GraphPad Software, San Diego, CA, USA). Physiological data and plasma volumes were compared using two-way ANOVA for repeated measures followed by Bonferroni post hoc test and unpaired two-tailed Student's t test. Urine productions were compared using unpaired two-tailed Student's t test. Differences were considered significant when p < 0.05. To achieve a statistical power of 90% with a difference in PV (PV3 − PV2) between groups of 4 mL/kg, the calculated sample size for each group was 9. All data were normally distributed. The results are presented as mean ± SD.

g unpaired two-tailed Student's t test. Differences were considered significant when p < 0.05. To achieve a statistical power of 90% with a difference in PV (PV3 − PV2) between groups of 4 mL/kg, the calculated sample size for each group was 9. All data were normally distributed. The results are presented as mean ± SD. Results Five animals died before the end of the experiment, evenly distributed between the groups. Eleven animals were considered to be non-septic and were excluded from the study, as they did not show a decrease in PV 3 h after the preparation. Physiological data Data for sodium (Na+), potassium (K+), haematocrit (Hct), lactate (Lac), pH, PaCO2 and PaO2, from arterial blood samples taken at baseline, at 3 h after the end of surgical preparation and at the end of the experiment 3 h later are summarized in Table 1. Data for arterial blood pressure are presented in Table 2. There were no significant differences between the groups in blood pressure or any of the parameters analysed at any time point. There was a significant increase in potassium and lactate levels, and a decrease in pH and PaCO2 in all groups at the end of the experiments compared to baseline (Table 1).Table 1 Physiological data

were no significant differences between the groups in blood pressure or any of the parameters analysed at any time point. There was a significant increase in potassium and lactate levels, and a decrease in pH and PaCO2 in all groups at the end of the experiments compared to baseline (Table 1).Table 1 Physiological data Na+(mmol/L) K+(mmol/L) Hct (%) Lac (mmol/L) pH PaCo2(kPa) PaO2(kPa) Bolus + infusion group (n = 9) Baseline 136 ± 2 4.9 ± 0.3 42 ± 1 2.4 ± 0.3 7.48 ± 0.04 4.9 ± 0.4 11.0 ± 0.8 3 h after prep 133 ± 3 5.3 ± 0.6 45 ± 3 2.7 ± 0.4 7.44 ± 0.03 4.9 ± 0.5 10.8 ± 0.7 End of experiment 135 ± 2 6.2 ± 0.8*** 50 ± 5** 3.2 ± 0.7** 7.43 ± 0.04* 4.2 ± 0.4** 11.4 ± 0.5 Bolus group (n = 9) Baseline 137 ± 2 4.4 ± 0.3 42 ± 3 2.3 ± 0.5 7.48 ± 0.03 5.0 ± 0.4 11.1 ± 0.6 3 h after prep 133 ± 2 5.2 ± 0.3 45 ± 3 2.9 ± 0.5 7.44 ± 0.02 4.9 ± 0.4 11.4 ± 0.5 End of experiment 135 ± 2 5.8 ± 0.9*** 50 ± 4*** 3.4 ± 0.8** 7.43 ± 0.04** 3.7 ± 0.7*** 12.5 ± 1.3 Sham group (n = 9) Baseline 136 ± 2 4.8 ± 0.6 43 ± 2 2.3 ± 0.2 7.48 ± 0.05 4.9 ± 0.5 11.0 ± 0.9 3 h after prep 134 ± 2 5.1 ± 0.6 45 ± 1 2.6 ± 0.3 7.45 ± 0.03 4.8 ± 0.3 11.0 ± 1.0 End of experiment 134 ± 2 5.9 ± 0.6** 49 ± 2*** 2.9 ± 0.7* 7.44 ± 0.03* 4.1 ± 0.3*** 12.0 ± 0.7 Data (mean ± SD) for sodium (Na+), potassium (K+), haematocrit (Hct), lactate (Lac), pH, arterial partial pressure of carbon dioxide (PaCO2) and arterial partial pressure of oxygen (PaO2). There were no significant differences between any of the parameters analysed, between groups at any time points. Two-way ANOVA for repeated measures followed by Bonferroni post hoc test was used for the statistical analyses. There were significant differences in K+, Hct, Lac, pH and Paco2 between baseline and the end of the experiments in all groups. Unpaired two-tailed Student's t test was used for the analyses. *p < 0.05, **p < 0.01, ***p < 0.001.

peated measures followed by Bonferroni post hoc test was used for the statistical analyses. There were significant differences in K+, Hct, Lac, pH and Paco2 between baseline and the end of the experiments in all groups. Unpaired two-tailed Student's t test was used for the analyses. *p < 0.05, **p < 0.01, ***p < 0.001. Table 2 Blood pressure Baseline 3 h after surg prep 1.5 h after start of treatment 3 h after start of treatment Bolus + infusion group (n = 9) 96 ± 11 92 ± 12 99 ± 14 96 ± 13 Bolus group (n = 9) 88 ± 15 91 ± 12 92 ± 16 90 ± 13 Sham group (n = 9) 95 ± 12 90 ± 11 94 ± 7 93 ± 7 Data (mean ± SD) for mean arterial blood pressure (mmHg) at baseline, at 3 h after the surgical preparation and at 1.5 and 3 h after the start of treatment. There were no differences between any of the groups at any time points. Two-way ANOVA for repeated measures followed by Bonferroni post hoc test was used for the statistical analyses.

mean arterial blood pressure (mmHg) at baseline, at 3 h after the surgical preparation and at 1.5 and 3 h after the start of treatment. There were no differences between any of the groups at any time points. Two-way ANOVA for repeated measures followed by Bonferroni post hoc test was used for the statistical analyses. Plasma volume Plasma volumes at baseline, at 3 h after the end of the surgical preparation and at the end of the experiment 3 h later were 41.2 ± 1.7 mL/kg, 35.2 ± 2.4 mL/kg and 27.4 ± 3.6 in the B + I group; 43.9 ± 3.1 mL/kg, 37.8 ± 3.3 mL/kg and 28.0 ± 5.0 mL/kg in the B group; and 42.4 ± 1.0 mL/kg, 35.9 ± 2.0 mL/kg, and 29.6 ± 2.4 mL/kg in the S group (Figure 2). There were no significant differences in PV between the three groups at any time point. There was a significant reduction in plasma volume in all three groups 3 h after the end of the surgical preparation (PV2) compared to baseline (PV1), and at the end of the experiment (PV3) compared to 3 h after the end of the surgical preparation (PV2) (p < 0.01) (Figure 2).Figure 2 Plasma volumes. Plasma volumes at baseline (PV1), at 3 h after the surgical preparation just before the start of treatment (PV2) and at the end of the experiment (PV3). There was no significant difference between any of the groups at any time points. There was a significant difference between PV1 and PV2, and PV2 and PV3 for all groups. Two-way ANOVA for repeated measures followed by Bonferroni post hoc test was used for the statistical analyses (**p < 0.01).

of the experiment (PV3). There was no significant difference between any of the groups at any time points. There was a significant difference between PV1 and PV2, and PV2 and PV3 for all groups. Two-way ANOVA for repeated measures followed by Bonferroni post hoc test was used for the statistical analyses (**p < 0.01). Urine production Urine production from the end of surgical preparation to the end of the experiment was 6.9 ± 3.4 mL/kg in the B + I group, 8.5 ± 1.4 mL/kg in the B group and 4.7 ± 1.8 mL/kg in the S group. The urine production was significantly larger in the B group than in the S group (p < 0.001) (Figure 3).Figure 3 Urine production. Data for urine production (mL/kg) from the end of surgical preparation to the end of the experiment. There was a significantly larger urine production in the B group compared to the S group. Student's t test was used for the statistical analyses (***p < 0.001).

S group (p < 0.001) (Figure 3).Figure 3 Urine production. Data for urine production (mL/kg) from the end of surgical preparation to the end of the experiment. There was a significantly larger urine production in the B group compared to the S group. Student's t test was used for the statistical analyses (***p < 0.001). Discussion The two investigated treatment regimes of vitamin C had no effect on plasma volume loss or any of the physiological parameters analysed in the early stage of sepsis in the present study in the rat. The larger urine production in the B group was in accordance with previous studies, both in humans and in dogs, showing a diuretic effect of vitamin C [22, 23]. It has been suggested [23] that this effect is due to an increase in glomerular filtration, although the exact mechanism of action is unclear. However, as urine represents loss of fluid from the entire extracellular space, meaning that only 20% to 25% of the volume is lost from the PV, the larger urine production in the B group will only have had a minor effect on the PV (0.8 to 1.0 mL/kg). A previous study on anaesthetised rats with artificial ventilation has shown that perspiration during this time period has no effect on plasma volume [24]. This means that the plasma volume loss in the present study must represent tissue oedema. The higher potassium concentrations and lower PaCO2 and pH at the end of the experiments in all groups are compatible with sepsis/SIRS-induced cell destruction, increased lactate production and subsequent compensatory hyperventilation (Table 1).

entire article was obtained and assessed for suitability by two of the authors (F.H., S.F). Any issue pertaining to eligibility of studies was solved via discussion with the senior author (H-C.P.). Data were extracted according to the information presented in Additional file 1: Table S1 and Additional file 2: Table S2. Accidental hypothermia as part of the lethal triad Beside acidosis and coagulopathy, accidental hypothermia is recognised as one main pillar of the lethal triad after severe trauma [19, 33]. Both acidosis and hypothermia are well known to exert significant effects on the coagulation system, resulting in a significantly higher blood loss due to an increased bleeding time [34] and enhanced mortality rate [34, 35]. Acidosis was proposed to be more detrimental than hypothermia in the development of coagulopathy [34]. However, hypothermia also revealed significant effects on the coagulation system with decreased fibrinogen concentrations (without affecting fibrinolysis) and impaired thrombin generation. In this regard, the onset of thrombin generation is remarkably delayed at a body temperature of 32°C [34]. Hypothermia also induces temporary thrombocytopenia (28% drop at 32°C) and platelet dysfunction [36, 37]. Below a threshold of 33°C, a decrease of body temperature additionally interferes with the plasmatic coagulation system by decreasing the activity of the associated enzymatic reactions [10, 36]. In fact, Martini et al. demonstrated that hypothermia and haemorrhage affected different aspects of the plasmatic coagulation process while both caused coagulation abnormalities: induction of hypothermia (32°C) alone did not result in significant changes in prothrombin time (PT) or activated partial thromboplastin time (aPTT) (assayed at 32°C). In contrast, hypothermia in combination with haemorrhagic shock (loss of 35% of total blood volume) with subsequent haemodilution (by resuscitation with Ringer's solution) led to a prolongation of PT by 40% (assayed at 32°C), without significant effects on aPTT [36]. It was concluded that hypothermia might have differential effects on the intrinsic and extrinsic pathways of the coagulation cascade. The effects of trauma and hypothermia on PT were confirmed by other studies [34, 37]. In contrast to aPTT, the activated clotting time (ACT) was prolonged by either isolated hypothermia or hypothermia in combination with haemorrhage and resuscitation [36].

Discussion The two investigated treatment regimes of vitamin C had no effect on plasma volume loss or any of the physiological parameters analysed in the early stage of sepsis in the present study in the rat. The larger urine production in the B group was in accordance with previous studies, both in humans and in dogs, showing a diuretic effect of vitamin C [22, 23]. It has been suggested [23] that this effect is due to an increase in glomerular filtration, although the exact mechanism of action is unclear. However, as urine represents loss of fluid from the entire extracellular space, meaning that only 20% to 25% of the volume is lost from the PV, the larger urine production in the B group will only have had a minor effect on the PV (0.8 to 1.0 mL/kg). A previous study on anaesthetised rats with artificial ventilation has shown that perspiration during this time period has no effect on plasma volume [24]. This means that the plasma volume loss in the present study must represent tissue oedema. The higher potassium concentrations and lower PaCO2 and pH at the end of the experiments in all groups are compatible with sepsis/SIRS-induced cell destruction, increased lactate production and subsequent compensatory hyperventilation (Table 1). The 125I-dilution technique for measurement of plasma volume is well established and reliable with small potential errors as described in detail previously [16, 19–21]. There might have been overestimation of plasma volume because of transcapillary escape of radioactive albumin during the 5-min period between injection of the tracer and collection of the blood sample. However, errors will have no or minor influence on the conclusions made, as they would have been of the same magnitude in all groups.

e been overestimation of plasma volume because of transcapillary escape of radioactive albumin during the 5-min period between injection of the tracer and collection of the blood sample. However, errors will have no or minor influence on the conclusions made, as they would have been of the same magnitude in all groups. By its vasodilatory effect, isoflurane might have increased the transcapillary plasma leakage by an increase in capillary pressure. This increase, however, must have been of the same magnitude in all groups and will therefore have no influence on the conclusions made. While previous studies have shown beneficial effects of vitamin C, such as reduced transcapillary leakage, reduced lymph flow and reduced oedema formation, using the same treatment regimes that were used in the present study [4, 6, 9, 10], we could not demonstrate any effects on the plasma volume loss.

By its vasodilatory effect, isoflurane might have increased the transcapillary plasma leakage by an increase in capillary pressure. This increase, however, must have been of the same magnitude in all groups and will therefore have no influence on the conclusions made. While previous studies have shown beneficial effects of vitamin C, such as reduced transcapillary leakage, reduced lymph flow and reduced oedema formation, using the same treatment regimes that were used in the present study [4, 6, 9, 10], we could not demonstrate any effects on the plasma volume loss. As mentioned in the ‘Background’, in most of the experimental studies listed above showing beneficial effects of vitamin C, treatment was initiated either prior to injury or shortly thereafter. The fact that we started the treatment 3 h after injury might explain our negative results. Our negative results may also be partly explained by the fact that caecal ligation and incision used in the present study probably resulted in a more severe sepsis than in the caecal ligation and puncture model used in many other studies. We also chose to evaluate the effect on plasma volumes 3 h after initiation of treatment, a time period shorter than in most other previous studies, and it cannot be excluded that this might have contributed to our negative results. In one study [6], treatment with vitamin C (200 mg/kg) was initiated 3 h after injury, as in the present study, and they demonstrated a positive effect, in terms of reduced capillary leakage of Evans blue after 12 h in the septic mouse. This study was performed in rats, and our results might therefore not be directly transferred to man. For example, in contrast to man, rats are able to synthesize vitamin C.

in the present study, and they demonstrated a positive effect, in terms of reduced capillary leakage of Evans blue after 12 h in the septic mouse. This study was performed in rats, and our results might therefore not be directly transferred to man. For example, in contrast to man, rats are able to synthesize vitamin C. Conclusions In conclusion, the present study did not confirm our hypothesis, as intravenous vitamin C treatment initiated 3 h after induction of sepsis had no effect on the loss of plasma volume, or any of the physiological parameters analysed, in the early stage of sepsis in the rat. High-dose vitamin C caused an increase in urine production. Competing interests The authors declare that they have no competing interests. Authors' contribution BB conceived and designed the study, compiled, analysed and interpreted the data, performed the statistical analysis and drafted the manuscript. POG conceived and designed the study, interpreted the data and helped to draft the manuscript. Both authors read and approved the final manuscript. Acknowledgements We thank Helene Axelberg for skilled technical assistance. This study was supported by the Swedish Research Council, Stockholm, Sweden (11581) and Region Skåne (ALF 18401).

Review Introduction Circadian rhythms refer to self-sustained fluctuations with a period of approximately (circa) 1 day (diem) in various physiological processes. Circadian rhythmicity is observed for many hormones in circulation (i.e., corticosteroids) as well as for circulating immune cells and cytokines [1, 2]. Ten circadian clock genes have been identified in human peripheral tissues so far, including Period (Per-1-3), Crypto-chrome (Cry-1 and Cry-2), Clock, and Bmal1, which coordinate with the master clock located in the suprachiasmatic nuclei (SCN) of the anterior hypothalamus [3].

lating immune cells and cytokines [1, 2]. Ten circadian clock genes have been identified in human peripheral tissues so far, including Period (Per-1-3), Crypto-chrome (Cry-1 and Cry-2), Clock, and Bmal1, which coordinate with the master clock located in the suprachiasmatic nuclei (SCN) of the anterior hypothalamus [3]. In mammals, the circadian system is composed of many individual, tissue-specific clocks with their phase being controlled by the master circadian pacemaker of SCN [1]. SCN neurons control clock genes throughout the body by controlling two major communication channels, the endocrine system and the autonomic nervous system (ANS). The recent discovery of a novel third type of retinal photoreceptor, other than rods and cones, provided evidence of a pathway mediating non-visual effects of light [4]. Subsequent signals are directed towards SCN neurons through the retinohypothalamic tract and synchronize them to the day/light cycle. Furthermore, connections of SCN with other hypothalamic structures allow the master clock to synchronize other clock genes in the body [5, 6]. Additionally, through sympathetic nerve projections, SCN output signals induce the release of a major internal synchronizer, the pineal substance melatonin (Figure 1) [5, 7].Figure 1 Melatonin: the ‘master biological clock’. Non-visual effects of light are mediated through specific retinal ganglion cells which subsequently activate SCN neurons. As a result, SCN inhibits the pineal production of melatonin during daytime through a polysynaptic pathway including paraventricular nucleus (PVN), superior cervical ganglia, and preganglionic sympathetic neurons of the lateral horn of the spinal cord. The pineal melatonin is considered the master biological clock that synchronizes the circadian rhythms of different clock genes throughout the body with different external ‘timekeepers’, such as light/dark cycles. Furthermore, the SCN-PVN network is responsible for 24-h period fluctuations of both sympathetic and parasympathetic tone, estimated with heart rate variability analysis, and for circadian oscillations of immunity and endocrine function. During inflammation, circadian rhythms of different hormones are disrupted, whereas immune cells in the periphery suppress melatonin's nocturnal surge through TNF-α and produce melatonin themselves. This extrapineal melatonin acts on a paracrine manner and exhibits both pro- and anti-inflammatory properties, depending on time phase and severity of stress.

s of different hormones are disrupted, whereas immune cells in the periphery suppress melatonin's nocturnal surge through TNF-α and produce melatonin themselves. This extrapineal melatonin acts on a paracrine manner and exhibits both pro- and anti-inflammatory properties, depending on time phase and severity of stress. SCN, suprachiasmatic nucleus; PVN, paraventricular nucleus. Figures are reproduced from the free website: ‘The brain from top to bottom’, according to its copyleft policy (http://thebrain.mcgill.ca/flash/pop/popcopy/popcopy.html).

s of different hormones are disrupted, whereas immune cells in the periphery suppress melatonin's nocturnal surge through TNF-α and produce melatonin themselves. This extrapineal melatonin acts on a paracrine manner and exhibits both pro- and anti-inflammatory properties, depending on time phase and severity of stress. SCN, suprachiasmatic nucleus; PVN, paraventricular nucleus. Figures are reproduced from the free website: ‘The brain from top to bottom’, according to its copyleft policy (http://thebrain.mcgill.ca/flash/pop/popcopy/popcopy.html). Melatonin is synthesized by the pineal gland upon β adrenoreceptor stimulation of pinealocytes, increased during sleepiness, and decreased during wakefulness, and it conveys the information of nighttime to the organism. In healthy humans, melatonin secretion starts between 9:00 p.m. and 11:00 p.m., reaching peak serum levels between 1:00 a.m. and 3:00 a.m. (>40 pg/mL) and then falling to low baseline values between 7:00 a.m. and 9:00 a.m (<7 pg/mL) [8, 9]. It also plays the role of an endogenous synchronizer, which is able to stabilize circadian rhythms and maintain their mutual phase relationships. Furthermore, its rhythm is involved in the regulation of the sleep/wake cycle, sleep structure, and more generally in the temporal organization of immunity [1, 5]. The last is considered as an effective component of ‘predictive homeostasis’ since nearly all organisms have developed mechanisms for anticipating environmental changes to optimize their survival [1, 2]. In this respect, temporal organization of immune response maximizes it at the time of the day that is needed most, since exposure to microbial pathogens depends on intrinsic 24-h rhythms of the host (activity, feeding). Moreover, immune modulation by the ANS, which also displays a diurnal rhythmicity, further supports the notion of immune regulation by light/dark cycle [1, 6]. Melatonin is also considered an active anti-inflammatory molecule due to the inhibition of tumor necrosis factor α (TNF-α) production [9, 10]. In addition, melatonin has an extrapineal source since different gastrointestinal cells synthesize melatonin, which has a peripheral activity (e.g., protection against reperfusion injury in gut mucosa), through its antioxidant properties [11].

due to the inhibition of tumor necrosis factor α (TNF-α) production [9, 10]. In addition, melatonin has an extrapineal source since different gastrointestinal cells synthesize melatonin, which has a peripheral activity (e.g., protection against reperfusion injury in gut mucosa), through its antioxidant properties [11]. Circadian rhythms are also synchronized and maintained by different phase relationships to external factors. These rhythms persist with an identical period (light/dark, sleep/wake) or are different throughout a day. These external factors are also called ‘timekeepers’ and are considered as effective modulators for the circadian oscillator (e.g., light, feeding, ambient temperature, and stress) [12]. Several studies have demonstrated that there is a circadian rhythmicity of different components of the immune system [1, 13–15]. Moreover, it has been suggested that circadian regulation of immunity is necessary for temporal coincidence of all its different molecular steps [13–15]. Thus, circadian oscillations of lymphocyte proliferation, antigen presentation, and cytokine gene expression appear coordinated via SCN output signals. Additionally, the number of most immune cells reaches maximal values during the night and is lowered after arousal [1].

ce of all its different molecular steps [13–15]. Thus, circadian oscillations of lymphocyte proliferation, antigen presentation, and cytokine gene expression appear coordinated via SCN output signals. Additionally, the number of most immune cells reaches maximal values during the night and is lowered after arousal [1]. Circadian physiology and inflammation Experimental data Different pro-inflammatory cytokines, such as TNF-α and interleukin-6 (IL-6), may cross the blood-brain barrier at leaky points (the circumventricular organs (CVO)) and induce a ‘sickness behavior’ , associated with decreased amplitude of circadian rhythmicity, such as loss of sleep/wake cycle [1, 16].

xperimental data Different pro-inflammatory cytokines, such as TNF-α and interleukin-6 (IL-6), may cross the blood-brain barrier at leaky points (the circumventricular organs (CVO)) and induce a ‘sickness behavior’ , associated with decreased amplitude of circadian rhythmicity, such as loss of sleep/wake cycle [1, 16]. Many studies have found that the susceptibility of mice to lipopolysaccharide (LPS) and TNF-α-induced lethality varied significantly throughout the day, depending on the time of administration [17–20]. Moreover, immune response upon LPS challenge, such as cytokine production [21] or toll-like receptor 9 (TLR9) expression [22], has been shown to display circadian rhythmicity, depending on time of LPS administration. Chronic inflammation can also affect SCN output by reducing amplitude and average spiking frequency of SCN neurons [23, 24]. In addition, LPS exposure has been found to suppress mRNA expression levels of different clock genes, in both animal [25] and human studies [26, 27]. However, melatonin and cortisol circadian rhythms were not affected by LPS (Table 1). It has been suggested that centrally regulated hormones' circadian rhythmicity and peripheral clock gene expression are independently regulated during sepsis, reflecting an uncoupling between central and peripheral oscillators during systemic inflammation [28].Table 1 Immune-circadian connection: experimental studies

. It has been suggested that centrally regulated hormones' circadian rhythmicity and peripheral clock gene expression are independently regulated during sepsis, reflecting an uncoupling between central and peripheral oscillators during systemic inflammation [28].Table 1 Immune-circadian connection: experimental studies Author Study design Major outcome Haldberg et al. [17] Susceptibility of mice to Escherichia coli endotoxin-induced lethality Lethality varied significantly throughout the day, depending on the time when mice were challenged Hrushesky et al. [18] Effect of time of TNF-α administration on lethal toxicity in mice Nine-fold variation of lethality being greatest during night and particularly before awakening Keller et al. [21] Splenocytes from mice, isolated at various times of the day, were challenged with LPS Circadian rhythmicity of TNF-α and IL-6 secretion was found. More than 8% of the peritoneal macrophage transcriptome oscillates in a circadian function autonomically and depends on time of LPS challenge Silver et al. [22] Toll-like receptor 9 (TLR9) expressed in peritoneal macrophages were estimated for circadian rhythmicity in a mouse model of sepsis Vaccination with TLR9 ligand as adjuvant at the time of enhanced TLR9 responsiveness induced an improved adaptive immune response many weeks later. Moreover, disease severity was dependent on the timing of sepsis induction, coinciding with daily changes in TLR9 expression Kwak et al. [24] Study of the long-term effects of INF-γ on SCN neurons by treating dispersed rat SCN neurons with INF-γ for a 4-week period Firing of SCN neurons and rhythmic expression of clock gene Per1 exhibited a lower average spiking frequency with reduced amplitude and an irregular firing pattern, in relation with controls Okada et al. [25] LPS effects on mRNA expression of clock genes in rats mRNA expression levels of different clock genes, such as Per 1 and Per 2, both in the liver and SCN neurons on day 1, were suppressed with an expression nadir between 10 and 14 h post-challenge. Subsequently, recovery was noted on day 2, whereas controls exhibited a robust circadian profile Boivin et al. [26] Estimation of clock gene oscillations in human blood mononuclear cells derived from three human volunteers Presence of circadian oscillations of Per 1and Per 2 genes Haimovich et al.

14 h post-challenge. Subsequently, recovery was noted on day 2, whereas controls exhibited a robust circadian profile Boivin et al. [26] Estimation of clock gene oscillations in human blood mononuclear cells derived from three human volunteers Presence of circadian oscillations of Per 1and Per 2 genes Haimovich et al. [27] Assessment of clock gene alterations upon LPS administration in peripheral human blood leucocytes, after challenging them with in vivo endotoxin or saline, either at 09:00 a.m or 09:00 p.m. LPS induced a profound suppression of all clock gene expression by 80% to 90%, between 13 and 17 h post-perfusion, whereas IL-6 and TNF-α returned to baseline within 6 h. However, melatonin and cortisol circadian rhythms were not affected by LPS challenge Pontes et al. [32] Colostrum samples for measuring tumor necrosis factor α (TNF-α) and melatonin content were collected from 18 normal delivered mothers in the morning, and diurnal and nocturnal melatonin levels in colostrum from healthy puerperae and mothers with mastitis were compared Suppression of nocturnal melatonin rise in mothers with mastitis was highly correlated with increased tumor necrosis factor α secretion. On the other hand, stimulated, but not quiescent, immune-competent cells secreted in the colostrum produced melatonin in vitro. In addition, this production ceased after bacteria killing Cruz-Machado et al. [33] Effects of LPS on melatonin production in rat pineal cultures Shutdown of melatonin production through TNF-α induction of NF-kB in pineal microglial cells

nt, immune-competent cells secreted in the colostrum produced melatonin in vitro. In addition, this production ceased after bacteria killing Cruz-Machado et al. [33] Effects of LPS on melatonin production in rat pineal cultures Shutdown of melatonin production through TNF-α induction of NF-kB in pineal microglial cells The immune-pineal axis A continuous communication between the pineal gland and the immune response has been suggested to exist, defining the ‘immune-pineal axis’ [29]. Thus, pineal melatonin nocturnal secretion enhances Th1/Th2 ratio within low ‘chronobiotic’ levels (nM-pM range) and inhibits at the same time both rolling and adherence of leucocytes to the endothelial layer, decreasing unnecessary inflammatory response [29–31]. Furthermore, extrapineal melatonin produced by local immune-competent cells acts in a paracrine manner as anti-inflammatory mediator in higher concentrations (mM range) [32–34]. Thus, it seems that in the early phase of inflammation, the body does not receive circadian information through the hormonal arm.

nse [29–31]. Furthermore, extrapineal melatonin produced by local immune-competent cells acts in a paracrine manner as anti-inflammatory mediator in higher concentrations (mM range) [32–34]. Thus, it seems that in the early phase of inflammation, the body does not receive circadian information through the hormonal arm. Markus et al. [35] have postulated that systemic inflammation activates the nuclear factor kappa B (NF-kB) pathway through LPS/TLR4 signaling at the level of pinealocytes and suppresses central melatonin nocturnal secretion, enhancing migration of immune cells at the site of injury. At the same time, different inflammatory mediators upregulate melatonin production in peripheral macrophages. This extrapineal tissue melatonin has been described as ‘immune buffer’ since it seems to play a dual role [36]. During acute stress, it acts as immunostimulant, improving bacterial phagocytosis, and subsequently, it enhances recovery phase by inducing production of anti-inflammatory cytokines. However, during an exacerbated inflammatory response, melatonin acts mainly as an anti-inflammatory molecule.

s to play a dual role [36]. During acute stress, it acts as immunostimulant, improving bacterial phagocytosis, and subsequently, it enhances recovery phase by inducing production of anti-inflammatory cytokines. However, during an exacerbated inflammatory response, melatonin acts mainly as an anti-inflammatory molecule. Corticosteroids may also affect melatonin pineal production [37, 38]. Thus, by inhibiting the NF-kB pathway in the pineal gland, they can restore its nocturnal rise [37] and enhance its production in a bell-shaped manner [38]. However, they can also decrease the activity of N-acetyltransferase (NAT) which is a key enzyme in the biosynthetic pathway of melatonin and hence inhibit its pineal production [39]. Finally, increased cortisol response to stress has been correlated with decreased amplitude of its own circadian rhythm [40]. In summary, different experimental studies confirm the existence of circadian oscillations of the immune response, which can be significantly suppressed by LPS. In addition, mortality seems to depend on time of LPS administration (Table 1). Circadian rhythm profiles and critical illness Circadian output assessment Periods and modeling variability of different biological time series that reflect circadian output, such as melatonin and cortisol, are assessed via cosinor analysis [41]. Briefly, this technique fits a cosine function of a fixed anticipated period to the data and approximates the following equation to experimental data, using the least squares method for minimization: 1

cal time series that reflect circadian output, such as melatonin and cortisol, are assessed via cosinor analysis [41]. Briefly, this technique fits a cosine function of a fixed anticipated period to the data and approximates the following equation to experimental data, using the least squares method for minimization: 1 where, M is the midline estimating statistic of rhythm (MESOR), the mean level of oscillation; A is the amplitude, the extent of oscillation from the MESOR or half of the total oscillation; π is 3.14159; TAU is the chosen period; t is a temporal fraction of the cycle, an instant of the whole revolution; and Φ (phi) is the acrophase, lag from a defined reference time point (e.g., local midnight when the fitted period is 24 h) of the crest time in the cosine curve fitted to the data (Figure 2).Figure 2 Chronobiologic analysis of a time series through cosinor analysis. Schematic illustration of basic metrics derived from cosinor analysis: This method is applicable to the individual biological time series anticipated to be rhythmic with a given period. The procedure fits a cosine function (blue) to the data (red) by least squares. Midline estimating statistic of rhythm (MESOR) is the mean level of oscillation that is the average value of the rhythmic function (e.g., cosine curve) fitted to the data. Amplitude is the difference between the maximum and the MESOR. Acrophase is the time of occurrence of the maximum value.

ata (red) by least squares. Midline estimating statistic of rhythm (MESOR) is the mean level of oscillation that is the average value of the rhythmic function (e.g., cosine curve) fitted to the data. Amplitude is the difference between the maximum and the MESOR. Acrophase is the time of occurrence of the maximum value. Except for serum melatonin, its urine metabolite 6-sulfatoxymelatonin (6-SMT) [10] and core body temperature (CBT) [42] are accepted biomarkers of circadian rhythm in critically ill patients. Circadian disruption in critically ill patients Circadian rhythms are disrupted by illness and intensive care unit (ICU) environment, associated with patient care interactions and unregulated light/dark patterns. Different clinical studies have demonstrated that a significant proportion of critically ill patients display long-term sleep disturbance and metabolism, suggesting a contribution of biological rhythm alterations [43, 44].

(ICU) environment, associated with patient care interactions and unregulated light/dark patterns. Different clinical studies have demonstrated that a significant proportion of critically ill patients display long-term sleep disturbance and metabolism, suggesting a contribution of biological rhythm alterations [43, 44]. In this respect, many authors have investigated circadian biomarkers in different groups of patients during their ICU stay, in order to assess a potential circadian dysfunction during critical illness (Table 2) [45–51]. Such misalignment occurs when there is an alteration between cycle frequency and phase in two or more rhythms [5, 12]. Its clinical significance has been established in different settings, since it has been shown to induce a prediabetic condition in healthy humans [52] and symptoms associated with heart failure in animal models of cardiovascular disease [53]. Furthermore, misaligning the cortisol rhythm has been shown to induce profound cardiovascular and renal disease sequel, which was subsequently reversed by light exposure therapy in hamsters [54].Table 2 Circadian disruption in critically ill patients: clinical studies

re in animal models of cardiovascular disease [53]. Furthermore, misaligning the cortisol rhythm has been shown to induce profound cardiovascular and renal disease sequel, which was subsequently reversed by light exposure therapy in hamsters [54].Table 2 Circadian disruption in critically ill patients: clinical studies Author Study design Major outcome Tweedie et al. [45] Retrospective study for characterizing core body temperature (CBT) 24-h profiles of 15 ICU patients 80% of all patient days had a significant circadian rhythm with erratic acrophases and normal amplitudes Nuttall et al. [46] Retrospective study assessing clinical significance of circadian rhythms in patients with (≤17) and without (n = 120) ICU psychosis, by comparing for 24 h the time of both temperature and urine output nadir Both groups had altered circadian rhythms, and although all ‘patient days’ had a significant rhythm, 83% of those days had abnormal cosinor-derived parameters Olofsson et al. [47] Study of melatonin levels in both blood and urine in 8 critically ill patients under sedation and mechanical ventilation The circadian rhythm of melatonin release was abolished in all but 1 patient, whereas no correlation was found between melatonin levels and level of sedation Frisk et al. [48] Study of 6-SMT and urine cortisol in 16 patients, treated in the ICU of two regional hospitals Hyposecretion of 6-SMT during mechanical ventilation, increase upon adrenergic stimulation, overall high cortisol excretion and, finally, a disturbed diurnal rhythm of both these hormones in 75% of all patients Paul and Lemmer [49] Measurement of CBT every hour and plasma cortisol and melatonin levels every 2 h for 24 h, in 13 sedated ICU patients following surgery or respiratory failure and 11 patients with brain injury The 24-h circadian profiles of all measured variables were significantly disturbed, with no physiological day-night rhythm in both groups of patients in relation with healthy controls, whereas circadian rhythm alterations were more pronounced in patients with brain injuries Pina et al. [50] Prospective analysis of hourly CBT and 4-h interval urine cortisol, melatonin, and 6-SMT profiles in 8 burn patients and 14 controls for 24 h in three sessions, occurring between ICU days 1 to 3, day 10, and days 20 to 30 Circadian rhythms of all measured variables were abolished in all patients in relation with controls.

spective analysis of hourly CBT and 4-h interval urine cortisol, melatonin, and 6-SMT profiles in 8 burn patients and 14 controls for 24 h in three sessions, occurring between ICU days 1 to 3, day 10, and days 20 to 30 Circadian rhythms of all measured variables were abolished in all patients in relation with controls. Burn ICU patients displayed significantly higher MESORS of CBT, urine melatonin, 6-SMT, and cortisol compared with the control group, during the three sessions of measurements. 24-h circadian profiles were restored within a 30-day period Gazendam et al. [51] Investigation of circadian rhythm disruption in a general ICU population, assessed using CBT profiles over a 48-h period in 21 patients Acrophase shift in all cases. Acute Physiology and Chronic Health Evaluation (APACHE) III score was predictive of circadian misplacement Mudlinger et al. [55] Circadian alterations in 17 septic patients versus 7 non-septic subjects and 21 controls, in the ICU Urinary 6-SMT exhibited circadian rhythmicity in only 1 of 17 septic patients versus 6 of 7 in non-septic patients and 18 of 23 in normal controls. MESORS appeared slightly increased, phase amplitudes were markedly lower, and acrophase occurred later in septic patients. On the contrary, in both non-septic patients and controls, 6-SMT exhibited a circadian rhythm Perras et al. [56] Measurement of single nocturnal melatonin concentration (NMC) in 302 patients during their first night in ICU Analysis of the whole study population did not reveal any correlation between single melatonin measurement and APACHE II score, but in 14 patients with severe sepsis, an inverse correlation was found Bagci et al. [57] Nocturnal plasma melatonin and 6-SMT urine concentrations were measured in 23 septic and 13 non-septic pediatric ICU patients The NMC during septic shock was increased in relation with no shock states. There was no difference for nocturnal and total 6-SMT excretion between septic patients with and without septic shock and non-septic patients. Nocturnal and total 6-SMT excretion was significantly lower in septic patients with than in septic patient without liver dysfunction. Sedation and mechanical ventilation did not affect melatonin excretion Gehlbach et al. [58] Assessment of sleep/wake regulation and circadian rhythmicity for 24 h, through 1-h interval urine measurements of 6-SMT, in 22 mechanically ventilated patients with different diagnoses of ICU admission The 24-h temporal profile of 6-SMT exhibited a phase delay.

n did not affect melatonin excretion Gehlbach et al. [58] Assessment of sleep/wake regulation and circadian rhythmicity for 24 h, through 1-h interval urine measurements of 6-SMT, in 22 mechanically ventilated patients with different diagnoses of ICU admission The 24-h temporal profile of 6-SMT exhibited a phase delay. There was no difference between patients with and without sepsis and no correlation between APACHE II score and 6-SMT amplitude Li et al. [59] 11 septic and 11 non-septic patients in ICU. Peripheral blood was drawn at 4-h intervals during the first day of admission The melatonin secretion acrophase occurred earlier in septic patients compared with non-septic patients. Melatonin MESORS tended to be higher in the septic group. Both Cry-1 and Per-2 expression were decreased, while TNF-α and IL-6 expression were increased in septic patients, reaching a peak at 6:00 p.m, which was consistent with the altered rhythm of melatonin secretion. Suppression of peripheral circadian genes was independent of the melatonin rhythm Plasma levels of melatonin, TNF-α, IL-6, and messenger RNA levels of circadian genes Cry-1 and Per-2 were analyzed Only a few investigators have evaluated circadian alterations during sepsis (Table 2) [55–59]. Mudlinger et al. [55] assessed in ICU circadian disruption in 17 septic patients versus 7 non-septic and 21 controls. Urinary 6-SMT, measured at 4-h intervals over a 24-h period, exhibited significant loss of circadian rhythmicity with no daytime decline in septic versus non-septic patients and controls, respectively.

Mudlinger et al. [55] assessed in ICU circadian disruption in 17 septic patients versus 7 non-septic and 21 controls. Urinary 6-SMT, measured at 4-h intervals over a 24-h period, exhibited significant loss of circadian rhythmicity with no daytime decline in septic versus non-septic patients and controls, respectively. Recently, Li et al. [59] studied for 24 h 11 septic and 11 non-septic ICU patients and measured during the first day of admission plasma levels of melatonin, TNF-α, and IL-6 and messenger RNA levels of circadian genes Cry-1 and Per-2. The authors found altered circadian rhythm of melatonin secretion, decreased expression of both Cry-1 and Per-2, and high levels of TNF-α and IL-6 in septic patients. They also showed that the suppression of peripheral circadian genes was independent of the melatonin rhythm.

he intrinsic and extrinsic pathways of the coagulation cascade. The effects of trauma and hypothermia on PT were confirmed by other studies [34, 37]. In contrast to aPTT, the activated clotting time (ACT) was prolonged by either isolated hypothermia or hypothermia in combination with haemorrhage and resuscitation [36]. As ACT is similar to aPTT except that it includes the interaction of platelets with other clotting components, the observed differences between changes of ACT and aPTT suggest that platelets are an important contributor to the hypothermia-induced effects on coagulation [36]. Using thrombelastography, hypothermia prolonged the initial clotting time and clot formation time and decreased clotting rapidity without affecting clot strength [36, 37]. In contrast, moderate haemorrhage and resuscitation alone had no effect on the initial clot formation or clot rapidity, but resulted in impaired clot strength. Thus, hypothermia and haemorrhage seem to have different effects on the coagulation function. Martini et al. concluded from their studies that the functional changes of the coagulation cascade are due to platelet dysfunction from hypothermia and depletion of fibrinogen and platelets from haemorrhage and resuscitation. When hypothermia and haemorrhage were combined, all abovementioned parameters were impaired [34, 36].

A levels of circadian genes Cry-1 and Per-2. The authors found altered circadian rhythm of melatonin secretion, decreased expression of both Cry-1 and Per-2, and high levels of TNF-α and IL-6 in septic patients. They also showed that the suppression of peripheral circadian genes was independent of the melatonin rhythm. In conclusion, Li et al. [59] confirmed that during acute phase of sepsis in humans, there is an uncoupling of the central master clock and peripheral tissue-specific clock genes, associated with pro-inflammatory cytokine production. Moreover, acrophase shift exhibited an advance rather than a delay in septic patients, contrary to what was found in the study of Mudlinger that included patients with at least 1 week stay in the ICU [55]. We suppose that at the early stages of sepsis, the inverse relation between melatonin and pro-inflammatory cytokines that was clearly shown in different animal models is more evident [35]. However, during the late stages, medications, such as catecholamines and varying levels of sedation [48, 60], could also alter circadian rhythms, since both morphine [60] and benzodiazepines [48] have been shown to induce NAT activity and enhance in a dose-dependent manner daytime production of melatonin [10]. In addition, mechanical ventilation [48] and ICU milieu [61–63] may further disrupt circadian variations, limiting accurate assessment of immune-circadian connectivity.

phine [60] and benzodiazepines [48] have been shown to induce NAT activity and enhance in a dose-dependent manner daytime production of melatonin [10]. In addition, mechanical ventilation [48] and ICU milieu [61–63] may further disrupt circadian variations, limiting accurate assessment of immune-circadian connectivity. ICU environment and circadian output disruption ICU milieu can be considered as a particular stress trigger for the internal circadian clock. Exposure to persistent environmental light has been recognized as a serious concern in the ICU [61–63]. However, different authors have found that light failed to influence circadian rhythms in healthy subjects [64] and in septic critically ill patients under controlled ventilation [65, 66], suggesting that sepsis per se could decrease sensitivity to light exposure. Drugs are potential confounders of immune-circadian connectivity in critically ill patients. In this respect, both opioids [50] and benzodiazepines [67] may alter melatonin production. Additionally, increased sympathetic tone and use of vasopressors during septic shock could theoretically enhance melatonin excretion. However, sympathetic reuptake of norepinephrine [68] and poor responsiveness of human pineal gland to circulating catecholamines [69] protect against the inappropriate increase in pineal melatonin production.

d sympathetic tone and use of vasopressors during septic shock could theoretically enhance melatonin excretion. However, sympathetic reuptake of norepinephrine [68] and poor responsiveness of human pineal gland to circulating catecholamines [69] protect against the inappropriate increase in pineal melatonin production. Another significant stressor of circadian rhythms in sedated patients is the sleep/wake cycle disruption. Different studies have confirmed that the majority of these patients experience either sleep deprivation and/or sleep fragmentation [43, 44, 58, 70]. It has been suggested that the dispersion of episodic ‘sleep-like states’ could be responsible for the reduced amplitude and acrophase delay of urine 6-SMT that was also noticed in healthy subjects [58, 71]. Nevertheless, it seems that sleep per se remains a weak timekeeper in humans without a concomitant change in the light/dark cycle [72]. Finally, delirium has been implicated as a pathologic state modifying melatonin excretion in elderly conscious medical patients [10, 73]. At the same time, melatonin circadian deregulation has been associated with neurotransmitter alterations and subsequent delirium in septic patients [74, 75]. However, it remains unclear if it is the quantity or the rhythm profile of melatonin that is related to delirium occurrence [10].

cious medical patients [10, 73]. At the same time, melatonin circadian deregulation has been associated with neurotransmitter alterations and subsequent delirium in septic patients [74, 75]. However, it remains unclear if it is the quantity or the rhythm profile of melatonin that is related to delirium occurrence [10]. Potential therapeutic implications Duboule [76] and Halberg [77] introduced the term ‘chronomics’, time and rhythm, for describing circadian regulation of animal development and chronotherapy in different disease states. Evidence from observational studies is growing that circadian disruption contributes to the development of cancer [76, 78]. So, it has been suggested that melatonin could be beneficial in cancer treatment when administered at chronobiologically determined optimum times of the day [79].

rapy in different disease states. Evidence from observational studies is growing that circadian disruption contributes to the development of cancer [76, 78]. So, it has been suggested that melatonin could be beneficial in cancer treatment when administered at chronobiologically determined optimum times of the day [79]. Administration of melatonin has been found in both animals [79, 80] and one human study in neonates [81] to reduce hyperinflammatory response during sepsis. In addition, it has been shown that melatonin exhibits an in vitro antimicrobial activity against multi-drug resistant Gram-negative and Gram-positive bacteria due to free iron binding [82] and furthermore can protect kidney grafts from ischemia-reperfusion injury [83]. Moreover, prolonged nighttime melatonin administration lowers blood pressure in hypertensive subjects [84], since SCN neurotransmitter content and transmission are suppressed during hypertension [85]. Finally, in two randomized placebo-controlled trials, melatonin [86] and a synthetic analog [87] were found to decrease incidence of delirium in elderly medical patients, but did not affect its duration or severity. However, intention-to-treat analysis was not possible in the first trial because of lost to follow-up patients.

, in two randomized placebo-controlled trials, melatonin [86] and a synthetic analog [87] were found to decrease incidence of delirium in elderly medical patients, but did not affect its duration or severity. However, intention-to-treat analysis was not possible in the first trial because of lost to follow-up patients. Nevertheless, many aspects remain unresolved. Thus, prior knowledge of the circadian profile of the patient is needed in order to design a personalized melatonin dose and duration of treatment, as well as chronobiologically determined optimum time of administration, since a circadian rhythmicity has been found for both pharmacokinetics and pharmacodynamics of different drugs, such as antibiotics [88]. Furthermore, melatonin excretion can be altered by liver and renal injury or by circadian modulation of hepatic function and glomerular filtration rate [10, 89]. In this respect, different timekeepers, such as light or medications, have been used in cancer or psychiatric disorders, on the right time and order and at a specific phase of the circadian cycle [78, 90]. Similarly, different ‘rhythm therapies’ could be scheduled for ICU patients, following the kairos principle (right time of the day) instead of chronos (time in general) [76, 78]. Moreover, introduction of additional timekeepers and excitation of the biological system with ultradian short-period rhythms, such as light or art therapy, have been found to enhance long-period fluctuations of melatonin by excitation, coupling, and resonance [91]. As a result, a restored circadian rhythmicity has been noticed in patients with sleep disorders and subjects with jet lag [91]. Such effects may also enlarge the circadian cycle of heart rate variability (HRV), which is connected with sleep quality and ANS dysfunction [78, 92].

excitation, coupling, and resonance [91]. As a result, a restored circadian rhythmicity has been noticed in patients with sleep disorders and subjects with jet lag [91]. Such effects may also enlarge the circadian cycle of heart rate variability (HRV), which is connected with sleep quality and ANS dysfunction [78, 92]. It has been postulated that entrained and synchronized circadian rhythms better prepare the physiology of an individual to anticipate normal cycles of energy demand in order to optimize adaptive regulation [93]. This ‘self-adaptation’ behavior is transformed into a ‘self-defense’ response during stress [31], explaining results from different studies. Thus, pro-inflammatory cytokines shut down melatonin's nocturnal surge in the acute phase, whereas exacerbated or chronic inflammation upregulates pineal production through anti-inflammatory mediators, such as corticosteroids [36, 38]. However, there is a lot of heterogeneity in different studies due to interspecies differences or time and severity of inflammatory insult, prompting a standardization of experimental protocols for translating results in the ICU setting.

l production through anti-inflammatory mediators, such as corticosteroids [36, 38]. However, there is a lot of heterogeneity in different studies due to interspecies differences or time and severity of inflammatory insult, prompting a standardization of experimental protocols for translating results in the ICU setting. Since severity of disease varies across the day and night [20, 94] and the temperature curve might exhibit an inverted pattern (febris inversa) in different infections, such as tuberculosis where fever is higher in the morning than in the evening, we suggest that future studies should assess differences in terms of circadian profiles, between patients suffering from an inflammatory episode that occurs at different time points of a 24-h period. Moreover, and since light unresponsiveness of SCN has been found in septic patients [65, 66], we suppose that in this particular group, possible circadian misalignment might reflect mainly individualized immune-circadian connections. In that case, it would be interesting to study if different circadian biomarkers correlate significantly with the Sequential Organ Failure Assessment (SOFA) score of severity of illness and predict mortality better than SOFA. In addition, ICU environmental profiles could be correlated with trajectories of circadian biomarkers, and different environmental approaches to patient care, such as ‘virtual darkness’ by shortening the day length, could be designed and tested to promote more rapid attainment of circadian rhythms [95]. Finally, restoring circadian light/dark cycle might improve immune function through enhanced melatonin production, in the context of reduced energy availability associated with critical illness, as is currently observed in lower mammals during the winter [95].

to promote more rapid attainment of circadian rhythms [95]. Finally, restoring circadian light/dark cycle might improve immune function through enhanced melatonin production, in the context of reduced energy availability associated with critical illness, as is currently observed in lower mammals during the winter [95]. Except for clinical researchers, basic scientists could also benefit from chronobiological analytic tools in order to design experimental studies and assess treatment effects in different septic models. It has been recognized that some of the reasons for negative results in different clinical trials in septic patients [96], despite encouraging results from preclinical studies, are the use of animal models that do not adequately mimic human sepsis [97]. Furthermore, misinterpretation of preclinical data or adoption of different experimental protocols has been considered as a contributing factor for this discrepancy [97]. In this respect, the use of ‘higher fidelity animal models’ has been suggested in order to increase the clinical relevance of experimental research [98]. Nevertheless, we would like to highlight the importance of assessing immune-circadian connectivity as a further step for translating basic science results into successful randomized controlled trials. Thus, different models should evaluate clock gene expression in immune-competent cells upon LPS challenge at standardized time points and in different environmental settings (i.e., light manipulation) [99], whereas clock gene knockout animals could also be used for assessing circadian-immune disconnection. Finally, new statistical methods, such as EUCLIS (EUCLOCK Information System, an EU FP6 project) [100] could be tested for analyzing the genome, the proteome, and the metabolome.

ettings (i.e., light manipulation) [99], whereas clock gene knockout animals could also be used for assessing circadian-immune disconnection. Finally, new statistical methods, such as EUCLIS (EUCLOCK Information System, an EU FP6 project) [100] could be tested for analyzing the genome, the proteome, and the metabolome. Conclusions As was suggested by Haldberg et al. [77], ‘in biologic time series that are dense and sufficient long the characteristics of rhythms and trends can be quantified as elements of structures called chronoms’. ‘Microscopy-in-time’ chronobiology studies cycles in biological time series with mechanisms embedded in living matter, whereas ‘telescopy-in-time’ chronomics assesses their alignment with environmental cues [101]. Thus, chronobiologic surveillance could be implemented in the ICU, serving a better understanding of biologic complexity in critical illness and, subsequently, an individualized optimization of treatment. In this respect, vascular variability anomalies (VVAs) estimated with chronomics, such as heart rate and blood pressure variability, have been recognized as significant risk factors in patients with cardiovascular diseases [102]. Similarly, reduced HRV has been repeatedly demonstrated in patients with sepsis and organ dysfunction [28]; however, chronobiologic analysis has not been performed so far.

et al. concluded from their studies that the functional changes of the coagulation cascade are due to platelet dysfunction from hypothermia and depletion of fibrinogen and platelets from haemorrhage and resuscitation. When hypothermia and haemorrhage were combined, all abovementioned parameters were impaired [34, 36]. As hypothermia-induced coagulopathy cannot be reversed with administration of clotting factors, but can be easily corrected when hypothermia is corrected [10], passive or active rewarming techniques seem to be of major relevance. Both methods can be applied in an external or internal fashion. Besides warmed-air blankets or warmed fluids, the most efficient means of rewarming is extracorporeal blood warming by different kinds of pumps (e.g. continuous arteriovenous rewarming, centrifugal vortex blood pump) [19]. In a study of Garraway et al., fast rewarming from 29°C to 37°C (within 85 min versus 217 min) appeared to correct the coagulopathy sooner and consequently may reduce the risk of ongoing bleeding [19].

such as heart rate and blood pressure variability, have been recognized as significant risk factors in patients with cardiovascular diseases [102]. Similarly, reduced HRV has been repeatedly demonstrated in patients with sepsis and organ dysfunction [28]; however, chronobiologic analysis has not been performed so far. In the context of negative results from different clinical studies in septic patients [96], we suggest that individual rhythm analysis might add significant value to the caring of critically ill. Thus, continuous monitoring of different biosignals, such as electrocardiogram (ECG), could detect diurnal variations in HRV and patterns of change specific for each patient and each pathophysiological state, creating an individual profile of ‘physiomarkers’ that could be used as both a diagnostic and therapeutic monitoring tool in everyday clinical practice. In addition, circadian aspects of pharmacokinetics and both liver and renal function could be considered in daily treatment, in order to increase efficiency and/or reduce adverse effects of medical therapy on a personalized basis. Finally, future clinical trials should assess circadian aspects of immunity and therapeutics for evaluating treatment effects. In this respect, adoption of different modeling techniques and design of in silico studies could be applied towards understanding inflammation and translate computational systems biology approaches in sepsis research to clinical relevance [103]. Abbreviations ANSautonomic nervous system APACHEAcute Physiology and Chronic Health Evaluation CBTcore body temperature

In the context of negative results from different clinical studies in septic patients [96], we suggest that individual rhythm analysis might add significant value to the caring of critically ill. Thus, continuous monitoring of different biosignals, such as electrocardiogram (ECG), could detect diurnal variations in HRV and patterns of change specific for each patient and each pathophysiological state, creating an individual profile of ‘physiomarkers’ that could be used as both a diagnostic and therapeutic monitoring tool in everyday clinical practice. In addition, circadian aspects of pharmacokinetics and both liver and renal function could be considered in daily treatment, in order to increase efficiency and/or reduce adverse effects of medical therapy on a personalized basis. Finally, future clinical trials should assess circadian aspects of immunity and therapeutics for evaluating treatment effects. In this respect, adoption of different modeling techniques and design of in silico studies could be applied towards understanding inflammation and translate computational systems biology approaches in sepsis research to clinical relevance [103]. Abbreviations ANSautonomic nervous system APACHEAcute Physiology and Chronic Health Evaluation CBTcore body temperature CVOcircumventricular organs EUCLOCKEntrainment of the circadian clock ECGelectrocardiograph EU FP6European Union frame project 6 HRVheart rate variability ICUintensive care unit IL-6interleukin-6 INF-γinterferon-γ LPSlipopolysaccharide MESORmidline estimating statistic of rhythm NATN-acetyltransferase NMCnocturnal melatonin concentration

CBTcore body temperature CVOcircumventricular organs EUCLOCKEntrainment of the circadian clock ECGelectrocardiograph EU FP6European Union frame project 6 HRVheart rate variability ICUintensive care unit IL-6interleukin-6 INF-γinterferon-γ LPSlipopolysaccharide MESORmidline estimating statistic of rhythm NATN-acetyltransferase NMCnocturnal melatonin concentration PVNparaventricular nucleus SCNsuprachiasmatic nucleus 6-SMT6-sulfatoxymelatonin SOFASequential Organ Failure Assessment TLRtoll-like receptor TNF-αtumor necrosis factor α VVAvascular variability anomalies. Competing interests The authors declare that they have no competing interests. Authors’ contributions VP conceived and wrote the review. AM, BP, and ML helped with literature research and editing of the manuscript. All authors read and approved the final manuscript.

Review Introduction The origin of hypothermia can differ fundamentally reflected by classifications as endogenous, controlled or accidental hypothermia. Endogenous hypothermia results either from metabolic dysfunctions with decreased heat production (e.g. hypothyroidism) or central nervous system dysfunctions with insufficient thermoregulation (e.g. tumour, trauma). Induced hypothermia, achieved by active cooling, is clinically used after cardiac arrest and in cardiac surgery for its mainly cytoprotective effects. Accidental hypothermia is defined as an unintentional decrease in core temperature during cold exposure in individuals without intrinsic thermoregulatory dysfunction [1].

nduced hypothermia, achieved by active cooling, is clinically used after cardiac arrest and in cardiac surgery for its mainly cytoprotective effects. Accidental hypothermia is defined as an unintentional decrease in core temperature during cold exposure in individuals without intrinsic thermoregulatory dysfunction [1]. A considerable number of patients presenting with accidental hypothermia are trauma victims with an incidence between 12% and 66% at arrival in the emergency room [2–4]. Clinical experience suggests that accidental hypothermia may be a major cause of posttraumatic complications, without being an independent prognostic factor for adverse outcome [5, 6]. The crucial core temperature in trauma patients seems to be approximately 34°C, and mortality rates of up to 100% have been reported in trauma patients with a core temperature <32°C [1, 7]. As pathophysiological reasoning, hypothermia induces platelet dysfunction and impairs the plasmatic coagulation system, especially below a threshold of 33°C. This hypothermia-induced coagulopathy was shown to be associated with significantly increased blood loss [8]. Therefore, the reversal of the ‘lethal triad’ of acidosis, coagulopathy and accidental hypothermia is indispensable for the successful treatment of bleeding trauma patients.

specially below a threshold of 33°C. This hypothermia-induced coagulopathy was shown to be associated with significantly increased blood loss [8]. Therefore, the reversal of the ‘lethal triad’ of acidosis, coagulopathy and accidental hypothermia is indispensable for the successful treatment of bleeding trauma patients. The deleterious effects of accidental hypothermia in multiple trauma patients contrast the beneficial effect of induced hypothermia on organ function during ischemia in elective surgery [9]. Furthermore, experimental studies have highlighted the protective effects of induced hypothermia in the setting of haemorrhagic shock with less pronounced organ dysfunction and reduced histological scores of organ damage, especially when induced rapidly after the traumatic insult [10, 11]. The beneficial effects of induced hypothermia might be explained by ameliorating or preventing the initiation and progression of an overwhelming systemic inflammatory response (‘systemic inflammatory response syndrome’, SIRS) [9–11]. Others have suggested that induction of hypothermia results in a reduction of oxygen demand while maintaining relative oxygen delivery and thereby maintaining levels of energy-rich phosphates, which is advantageous during ischemic periods by prolonging the golden hour of shock [12]. In contrast, accidental hypothermia is associated with reduced concentrations of energy-rich phosphates and higher lactate levels as a result of insufficient tissue perfusion and failure of the organism to maintain normal body temperature. Therefore, accidental hypothermia seems to be a result and driver of severe shock finally leading to adverse outcome [13]. In conclusion, it seems to be essential to differentiate between induced and accidental hypothermia, as these different entities clearly exert divergent effects on the inflammatory response and posttraumatic organ function.

thermia seems to be a result and driver of severe shock finally leading to adverse outcome [13]. In conclusion, it seems to be essential to differentiate between induced and accidental hypothermia, as these different entities clearly exert divergent effects on the inflammatory response and posttraumatic organ function. Recent experimental research has tried to define the role of hypothermia as part of the lethal triad after trauma as well as to balance out potentially beneficial versus adverse aspects of therapeutically induced hypothermia on the posttraumatic course. Furthermore, promising treatment protocols for the induction of hypothermia have been optimized (optimal magnitude, duration, timing, cooling method and cooling or rewarming) and evaluated [14–29].

ance out potentially beneficial versus adverse aspects of therapeutically induced hypothermia on the posttraumatic course. Furthermore, promising treatment protocols for the induction of hypothermia have been optimized (optimal magnitude, duration, timing, cooling method and cooling or rewarming) and evaluated [14–29]. This review aims to differentiate between these two entities of hypothermia and to summarize the current knowledge of the potential therapeutic use of induced hypothermia in the posttraumatic setting. We set the focus on evolutionary highly developed, large-sized species (pigs) as, e.g., the size of the organism (small vs. large animal models) significantly affects the reaction on therapeutically induced hypothermia. Small animals were shown to be very sensitive to hypothermia due to their high surface-to-volume ratio, which has to be taken into account when comparing various studies in this field [30]. Furthermore, small animals present with ‘non-shivering thermogenesis’ [31]. Therefore, porcine models were chosen because their physiological response to trauma-haemorrhage simulates the human response more closely than any other non-primate species. Except for some minor differences, pigs exhibit cardiovascular, haematological, immunological and electrolyte profiles that are almost identical to those in humans [32]. Only the coagulation system differs significantly to humans, with pigs being naturally in a hypercoagulatory state.

han any other non-primate species. Except for some minor differences, pigs exhibit cardiovascular, haematological, immunological and electrolyte profiles that are almost identical to those in humans [32]. Only the coagulation system differs significantly to humans, with pigs being naturally in a hypercoagulatory state. Methods The following criteria were used to determine eligibility of a study to be included in this review. Inclusion criteria were isolated or combined trauma-haemorrhage models, use of pigs as experimental animals, English or German language and the use accidental or induced hypothermia. The use of other large animals or small animals resulted in exclusion of the study. A literature search was carried out on Medline, Embase and Cochrane for studies published until July 2013 on the topic of the effects of accidental and induced hypothermia in porcine models of isolated or combined trauma-haemorrhage. The following key words were used: ‘accidental hypothermia’, ‘spontaneous hypothermia’, ‘induced hypothermia’, ‘therapeutic hypothermia’, ‘pigs’, ‘swine’, ‘trauma’, ‘injury’, ‘hemorrhage’, ‘fracture’ and ‘bleeding’. The references of selected studies were also perused for articles that may have been missed via the electronic search.

key words were used: ‘accidental hypothermia’, ‘spontaneous hypothermia’, ‘induced hypothermia’, ‘therapeutic hypothermia’, ‘pigs’, ‘swine’, ‘trauma’, ‘injury’, ‘hemorrhage’, ‘fracture’ and ‘bleeding’. The references of selected studies were also perused for articles that may have been missed via the electronic search. Study selection The title and abstract of all identified studies were examined by one reviewer (F.H.). Then, the entire article was obtained and assessed for suitability by two of the authors (F.H., S.F). Any issue pertaining to eligibility of studies was solved via discussion with the senior author (H-C.P.). Data were extracted according to the information presented in Additional file 1: Table S1 and Additional file 2: Table S2.

ized by previous models [38]. While induced hypothermia might be used to propose hypotheses on possible mechanism of the effects on coagulation, the definite major underlying pathomechanisms can only be assessed if hypothermia occurs spontaneously after trauma-haemorrhage, e.g. in the context of systemic hypoperfusion. Therapeutically induced hypothermia Considerations on model design and strategies for induction of hypothermia In contrast to elective surgery (e.g. transplantation, neurosurgery, cardiac surgery), induced hypothermia can only be deliberated after the onset of the insult in the setting of severe tissue injuries. Due to the potentially negative side effects of therapeutically induced hypothermia, the impact of a decreased body temperature was investigated in porcine trauma models in order to transfer the results to the clinical setting of multiple trauma patients. The design of these experimental studies differs in diverse aspects (e.g. kind of induced trauma, technique, timing of induced hypothermia and rewarming phase), which can be explained by the different purposes of the study or varying experimental questions.

e results to the clinical setting of multiple trauma patients. The design of these experimental studies differs in diverse aspects (e.g. kind of induced trauma, technique, timing of induced hypothermia and rewarming phase), which can be explained by the different purposes of the study or varying experimental questions. Injury pattern In general, the effects of induced hypothermia have been investigated in models with isolated haemorrhagic shock or in combined models with additional trauma load (Additional file 2: Table S2). Although volume- and pressure-controlled haemorrhage models offer a high level of standardization, the absence of uncontrolled sources of bleeding (e.g. solid organ injuries) has been identified as a limitation of studies with controlled haemorrhagic shock in which no ongoing bleeding is observed. Furthermore, a model of haemorrhage without significant additional tissue trauma is unlikely to validly predict the response to therapy in any clinical setting, as tissue injury alters the response to haemorrhage. The implementation of solid organ injuries seems to be of particular importance in studies investigating the potential clinical application of induced hypothermia after trauma-haemorrhage in order to observe whether bleeding control is worsened when lowering the body temperature [11, 39]. Therefore, models with additional liver, lung or soft tissue trauma (with and without additional fractures) have been developed which allowed investigating whether the injured locus starts to re-bleed after the induction of hypothermia [39, 40].

bleeding control is worsened when lowering the body temperature [11, 39]. Therefore, models with additional liver, lung or soft tissue trauma (with and without additional fractures) have been developed which allowed investigating whether the injured locus starts to re-bleed after the induction of hypothermia [39, 40]. Time point of hypothermia induction In diverse experimental studies, hypothermia was often therapeutically induced in parallel [41], a few minutes after [10] or even before [42, 43] the induction of haemorrhage. Although this strategy seems to be important for mechanistic considerations of hypothermia treatment, it is clearly not compatible to the clinical situation due to several reasons (Additional file 1: Table S1). Firstly, the induction of hypothermia before or in parallel to trauma-haemorrhage is impossible in the setting of multiple trauma. Secondly, a significant number of multiple trauma patients develop accidental hypothermia in the course of trauma-haemorrhage. Therefore, no hypothermia can additionally be induced under these conditions. Lastly, it might be advantageous to induce hypothermia after surgical bleeding control due to the potential coagulopathic effects, particularly in the case of additional organ injuries. Therefore, therapeutic hypothermia was induced after emergency surgery and correction of accidental hypothermia (analogous to the time point of treatment on the intensive care unit) in some combined trauma models [39, 40].

ue to the potential coagulopathic effects, particularly in the case of additional organ injuries. Therefore, therapeutic hypothermia was induced after emergency surgery and correction of accidental hypothermia (analogous to the time point of treatment on the intensive care unit) in some combined trauma models [39, 40]. Velocity and technique of hypothermia induction Numerous studies emphasized that therapeutically induced hypothermia seems to be most effective when induced rapidly after trauma-haemorrhage [10, 11, 41, 44–46]. In this context, surface cooling (evaporation cooling and ice packs), which failed to sufficiently decrease core temperatures, even had adverse effect on the posttraumatic course [47], whereas rapid induction of hypothermia by extracorporeal shunt circulation prolonged short-term survival in the same model. However, it was also supposed that a continuous and prolonged attempt to reduce the core temperature may add additional stress and reduced the survival rate [41].

on the posttraumatic course [47], whereas rapid induction of hypothermia by extracorporeal shunt circulation prolonged short-term survival in the same model. However, it was also supposed that a continuous and prolonged attempt to reduce the core temperature may add additional stress and reduced the survival rate [41]. Besides the rapidity of reducing the body temperature, other factors have also been claimed to have a significant physiological impact on the posttraumatic course after induced hypothermia [48]. It has been suggested that hypothermia should be induced under anaesthesia and muscle relaxation in order to suspend the energy-consuming shivering and the stress response of the sympathetic nervous system to hypothermia [48]. Furthermore, the technique of hypothermia induction seems to be decisive. In this regard, different authors reported that the infusion of 4°C lactated Ringer's solution prolongs survival [44, 45, 48]. However, induction of hypothermia with infusion of saline at room temperature and surface cooling was found to be more effective after haemorrhagic shock [48]. As a possible mechanism, a faster cold response with vasoconstriction and decreased tissue perfusion was suggested, which might have resulted in a worse outcome [48]. In many studies, a rapid decrease of core temperature after severe trauma was reached by extracorporeal cooling while, noteworthy, no adverse effects of fast cooling were reported [10, 39, 40, 43]. For example, some studies used a heparin-free roller pump that has been described to reduce body temperature in a reliable fashion [39, 40]. Others used cardiopulmonary bypass techniques (with heparin) to induce hypothermia, with a rather rapid cooling rate of 2°C/min [10].

fast cooling were reported [10, 39, 40, 43]. For example, some studies used a heparin-free roller pump that has been described to reduce body temperature in a reliable fashion [39, 40]. Others used cardiopulmonary bypass techniques (with heparin) to induce hypothermia, with a rather rapid cooling rate of 2°C/min [10]. Rewarming Rewarming was often performed within the resuscitation process or shortly thereafter (Additional file 1: Table S1). It has been shown that long-term survival after induction of therapeutic hypothermia is also influenced by the rate of rewarming [46]. In this context, a rewarming rate of 0.5°C/h has been proven to be most effective, whereas a faster increase of body temperature (>1.0°C/h) did not significantly alter survival after induced hypothermia compared to normothermic animals in a haemorrhagic shock model [46]. However, rewarming after induced hypothermia should always be evaluated in the context with the duration of hypothermia as rebound effects after too rapid rewarming might depend on the time period between the trauma impact and the rewarming phase.

mia compared to normothermic animals in a haemorrhagic shock model [46]. However, rewarming after induced hypothermia should always be evaluated in the context with the duration of hypothermia as rebound effects after too rapid rewarming might depend on the time period between the trauma impact and the rewarming phase. Effects of induced hypothermia on the posttraumatic course Survival and haemodynamics Various experimental studies demonstrated that therapeutic induction of hypothermia can significantly improve posttraumatic survival [10, 41, 44, 45, 48]. This beneficial effect of a decreased body temperature has been shown for different degrees of hypothermia. Wu et al. found that mild hypothermia (34°C) is effective in a model of haemorrhagic shock that includes resuscitation and intensive care treatment [48]. In another study [49], the induction of moderate hypothermia of 30°C resulted in a reduction of posttraumatic mortality. In a study of Alam et al., the protective effects of profound induced hypothermia (10°C) was highlighted [11], which were confirmed in a model of uncontrolled haemorrhage with major vascular, solid organ and hollow viscus injuries [50]. Furthermore, it seems to be interesting that profound hypothermia (10°C) was found to be more effective than ultraprofound hypothermia of 5°C [51]. It is also to be considered that the duration of hypothermia seems to have an additional impact on the beneficial effects of hypothermia after uncontrolled haemorrhage - at least for profound hypothermia. In this context, 60 min of hypothermic arrest seems to be the upper limit, as a significant decrease in survival as well as an increase of postoperative complications has been described after 120 min [52].

mpact on the beneficial effects of hypothermia after uncontrolled haemorrhage - at least for profound hypothermia. In this context, 60 min of hypothermic arrest seems to be the upper limit, as a significant decrease in survival as well as an increase of postoperative complications has been described after 120 min [52]. Focusing on the association between the technique of hypothermia induction and haemodynamic function as well as survival, it was shown that the administration of cold crystalloid solution (4°C) prolongs survival after uncontrolled haemorrhage [44], which coincided with a lower heart rate, decreased myocardial oxygen requirements and an improved stroke volume index during haemorrhagic shock, probably leading to a better preservation of cardiac function [44]. Wu et al. reported similar findings with best results for induction of hypothermia with infusion of saline at room temperature and surface cooling. In contrast, rapid cooling ice-cold fluids showed a decreased survival rate in the same model. As the infusion of ice-cold fluid was associated with increased mean arterial pressure (MAP) and lactate levels, the authors assumed that vasoconstriction with decreased tissue perfusion played a role in worsening the outcome [48]. The authors postulated that rapid changes in core temperature might trigger a faster cold response with potentially detrimental vasoconstriction and decreased tissue perfusion. Therefore, the authors concluded that it is important to suspend the shivering and the response of the sympathetic nervous system to hypothermia by using anaesthetics and muscle relaxants at the time of hypothermia induction [48].

ld response with potentially detrimental vasoconstriction and decreased tissue perfusion. Therefore, the authors concluded that it is important to suspend the shivering and the response of the sympathetic nervous system to hypothermia by using anaesthetics and muscle relaxants at the time of hypothermia induction [48]. External cooling by a ‘hypothermic bed’ decreased MAP, cardiac output and heart rate after isolated haemorrhagic shock [30, 49, 53]. The cardiovascular adaption to hypothermia after haemorrhage obviously consists of an early heart rate-induced reduction of cardiac output and a later decrease of MAP [30]. Under primary limited (SAP >80 mmHg) and subsequent full (SAP >90 mmHg) resuscitation and deep anaesthesia, surface cooling by means of ice packs to 33°C (0.5°C to 1°C/h) resulted in a significantly improved survival rate compared to normothermia [45] in an isolated haemorrhagic shock model. In contrast, the same group only observed a trend towards improved survival and a transient elevation of markers of organ function after combined trauma and hypothermia induction to 34.5°C. Possible explanations may be the shorter hypothermic period and the milder degree of hypothermia in the latter study [42].

model. In contrast, the same group only observed a trend towards improved survival and a transient elevation of markers of organ function after combined trauma and hypothermia induction to 34.5°C. Possible explanations may be the shorter hypothermic period and the milder degree of hypothermia in the latter study [42]. Besides the technique, degree and duration of induced hypothermia, the velocity of decreasing body temperature has a significant impact as well. Takasu et al. found that under light anaesthesia, surface cooling (evaporation cooling and ice packs) during severe haemorrhagic shock without any fluid resuscitation failed to decrease core temperature and even shortened the survival time in pigs. Moreover, animals in the surface cooling group exhibited a more pronounced decline of MAP, arterial pH and cardiac index than the normothermic group [47]. It was supposed that a continual and prolonged attempt to reduce the core temperature added additional stress and reduced the survival rate [41]. However, rapid extracorporeal cooling to 34.5°C resulted in a significantly increased survival with higher MAP and stroke volume index as well as a lower heart rate compared to normothermic animals in the same model [41]. Accordingly, Alam et al. described hypothermia to be most effective if rapidly induced with a rate of 2°C/min (compared to 1°C or 0.5°C).

o 34.5°C resulted in a significantly increased survival with higher MAP and stroke volume index as well as a lower heart rate compared to normothermic animals in the same model [41]. Accordingly, Alam et al. described hypothermia to be most effective if rapidly induced with a rate of 2°C/min (compared to 1°C or 0.5°C). In summary, the vast majority of experimental studies support a rapid decrease of core temperature as the strategy of choice for hypothermic treatment after severe trauma [10, 41, 43, 44, 51, 54]. Furthermore, the duration and degree of induced hypothermia as well as the deepness of anaesthesia and muscle relaxation seem to have an impact on the therapeutic effects of hypothermia.

rapid decrease of core temperature as the strategy of choice for hypothermic treatment after severe trauma [10, 41, 43, 44, 51, 54]. Furthermore, the duration and degree of induced hypothermia as well as the deepness of anaesthesia and muscle relaxation seem to have an impact on the therapeutic effects of hypothermia. Coagulation system Hypothermia is well known to modify the cellular and plasmatic components of the coagulation system. However, in contrast to the coagulopathic effects of the lethal triad, diverse experimental studies failed to find significant differences in signs of coagulopathy (PTT, platelet count, INR, TEG) between normothermic pigs and after induction of hypothermia (between 33°C and 35°C) [42, 43, 45]. In accordance, therapeutic induction of profound total body hypothermia (10°C) for approximately 60 min as well as induced mild hypothermia (34°C) for about 12 h did not result in a measurable increase of postoperative bleeding in a combined trauma model [10]. Also in other studies, therapeutically induced hypothermia did not result in an enhanced incidence of bleeding complications compared to normothermic animals [44, 48, 49, 53]. This might be explained by the fact that therapeutically induced hypothermia does not seem to affect maximum clot strength [40]. As the maintenance of clot strength has been shown to be crucial for transfusion requirements [36], it is tempting to speculate that the effects of therapeutic hypothermia on the coagulation system are rather limited - at least in porcine models of trauma-haemorrhage [40]. However, the coagulopathic effects of induced hypothermia are discussed controversially at least for body temperatures below 33°C. In this regard, Gröger et al. found a prolonged clotting time and reduced clot firmness at 32°C, whereas 35°C had no effects on these parameters [43]. Furthermore, relevant differences in the coagulation system between pigs and humans have to be considered, when interpreting the effects of hypothermia on coagulation. In pigs, a general hypercoagulatory situation has been described compared to the human situation. Furthermore, it has so far not been possible to induce trauma-haemorrhage-related coagulopathy in pigs [55, 56]. Therefore, further experimental studies (e.g. in other animal species or in pigs with a different genetic background) are needed to define a safe therapeutic window in terms of degree and duration before induced hypothermia might be applicable in the clinical setting.

-haemorrhage-related coagulopathy in pigs [55, 56]. Therefore, further experimental studies (e.g. in other animal species or in pigs with a different genetic background) are needed to define a safe therapeutic window in terms of degree and duration before induced hypothermia might be applicable in the clinical setting. Overall, the secondary induction of hypothermia after primary surgical bleeding control and therapy of trauma-related coagulopathy on the intensive care unit might represent a treatment strategy which can minimize the bleeding risks associated with a therapeutic decrease of body temperature [40].

-haemorrhage-related coagulopathy in pigs [55, 56]. Therefore, further experimental studies (e.g. in other animal species or in pigs with a different genetic background) are needed to define a safe therapeutic window in terms of degree and duration before induced hypothermia might be applicable in the clinical setting. Overall, the secondary induction of hypothermia after primary surgical bleeding control and therapy of trauma-related coagulopathy on the intensive care unit might represent a treatment strategy which can minimize the bleeding risks associated with a therapeutic decrease of body temperature [40]. Metabolism The hypothermia-associated decrease of the metabolic rate in key organs is supposed to be one of the most protective strategies for cellular function during periods of haemorrhage, ischemia and reperfusion. It is well established that alterations in temperature influence biologic reactions with a 50% reduction in metabolism for every 10°C decrease in body temperature (i.e. ‘Q10’ effect) [10]. In addition, a significant reduction in blood catecholamine levels after the induction of hypothermia might additionally diminish the metabolic rate after trauma-haemorrhage [49, 53]. These decreased metabolic and oxygen demands under hypothermic conditions have been associated with reduced oxygen extraction ratio and increased pO2 values [49, 53] as well as with improved oxygen delivery indices in mechanically ventilated pigs [45]. Furthermore, this increased oxygen supply might be beneficial as it counteracts the development of an anaerobic metabolism after trauma-haemorrhage, thereby preserving energy-rich phosphate (e.g. ATP) levels and attenuating a tissue energy depth. Accordingly, decreased posttraumatic lactate levels after induction of hypothermia have been observed [43, 45, 49]. It was also assumed that the maintenance of ATP stores might contribute to a switch of cell death from necrosis to apoptosis, as apoptosis has been associated with higher ATP content compared to necrotic cell death. In accordance, a more pronounced apoptosis after therapeutic induction of hypothermia was observed, whereas cell necrosis was attenuated [43]. This might lead to less local and systemic inflammation as apoptotic events canonically cause much less inflammation in comparison to necrotic events.

ed to necrotic cell death. In accordance, a more pronounced apoptosis after therapeutic induction of hypothermia was observed, whereas cell necrosis was attenuated [43]. This might lead to less local and systemic inflammation as apoptotic events canonically cause much less inflammation in comparison to necrotic events. The effects of hypothermia on the development of metabolic acidosis after trauma-haemorrhage have been controversially discussed in the literature. In a study by Gröger et al. [43], induction of hypothermia resulted in metabolic acidosis which persisted even after rewarming until the end of the experiment. The development of this metabolic acidosis was at least partly explained by an enhanced fat metabolism with increased free fatty acid concentrations and reduced carbohydrate oxidation. A decreased hepatic clearance of acid metabolites under hypothermic conditions might also contribute to this effect. However, others [53] described no effects of moderate hypothermia (30°C) on blood pH and base excess after isolated haemorrhagic shock. Again, the time point of hypothermia induction as well as duration and severity of induced hypothermia might represent significant factors for these differences.

ribute to this effect. However, others [53] described no effects of moderate hypothermia (30°C) on blood pH and base excess after isolated haemorrhagic shock. Again, the time point of hypothermia induction as well as duration and severity of induced hypothermia might represent significant factors for these differences. Besides the effects on metabolism, some studies [43, 49, 53] also found an impact of hypothermia on electrolyte levels. Although a decreased body temperature has been described to alter the function of the sodium-potassium pump resulting in increased potassium levels, induced hypothermia after trauma-haemorrhage maintained normokalemia. In contrast, normothermia was associated with hyperkalemia. This beneficial aspect of hypothermia was explained by a limitation of tissue damage by induced hypothermia. It seems interesting in this study that the higher serum levels of creatine kinase (CK) in hypothermic animals were not a result of muscular tissue destruction but of shivering, as previous blocking of all muscle activity by drugs prevented this CK increase [53].

by a limitation of tissue damage by induced hypothermia. It seems interesting in this study that the higher serum levels of creatine kinase (CK) in hypothermic animals were not a result of muscular tissue destruction but of shivering, as previous blocking of all muscle activity by drugs prevented this CK increase [53]. Inflammatory response The modulation of the inflammatory response after trauma-haemorrhage might be another mechanism by which induced hypothermia exerts protective effects during the posttraumatic course. Induced hypothermia has been shown to attenuate or at least delay the pro-inflammatory and oxidative stress response which was associated with a transitory reduction of kidney and liver dysfunction as well as reduced histological damage (e.g. reduced infiltration of polymorphonuclear granulocytes) in these organs [43, 50, 53]. The augmented liberation of protective heat shock proteins might further contribute to the beneficial effects of hypothermia under those conditions [50].

ion of kidney and liver dysfunction as well as reduced histological damage (e.g. reduced infiltration of polymorphonuclear granulocytes) in these organs [43, 50, 53]. The augmented liberation of protective heat shock proteins might further contribute to the beneficial effects of hypothermia under those conditions [50]. The hypothermia-associated alteration of the T cell cytokine production pattern with a conversion from a Th-1 to a Th-2 cytokine pattern has been shown to result in an anti-inflammatory immunosuppressive profile [1]. It was speculated that hypothermia and trauma might activate the hypothalamic-pituitary-adrenocortical (HPA)-axis, resulting in an increased glucocorticoid secretion. Furthermore, it has been shown that hypothermia preserves plasma glucocorticoid concentration. As glucocorticoid is supposed to be a strong inducer of IL-10 production, these effects of hypothermia might contribute to the hypothermia-associated increase of IL-10 levels. It was therefore suggested that the immunosuppressive profile as well as hypothermia-associated vasoconstriction and local tissue hypoxia may increase the risk of infections and worsen outcome [1, 57]. In accordance, Alam et al. observed an increase in infectious complications after a 2-h period of profound hypothermia (10°C) [52]. However, such effects on the incidence of infectious complications were not described for a duration of 60 min [10, 52], even if an additional colon injury was induced. Therefore, it has to be assumed that the impact of hypothermia on immune function is to some extent time-dependent [10]. Furthermore, for the long-term effects of hypothermia on the pro-inflammatory response, it might be speculated that hypothermia results in a delayed or prolonged release of inflammatory mediators. In this context, Gröger et al. found the highest TNF-α, IL-6 and 8-isoprostane blood levels in hypothermic animals (32°C) at 12 h after resuscitation and at the end of the study period, whereas induced hypothermia of 35°C resulted in an almost similar response as normothermia [43].

ed release of inflammatory mediators. In this context, Gröger et al. found the highest TNF-α, IL-6 and 8-isoprostane blood levels in hypothermic animals (32°C) at 12 h after resuscitation and at the end of the study period, whereas induced hypothermia of 35°C resulted in an almost similar response as normothermia [43]. Limitations Besides the well-described variables which have some influence on the results of experimental studies (anaesthesia before trauma, application of drugs and infusions, small sample sizes) [58], the species-specific differences between humans and pigs deserve particular attention.

ed release of inflammatory mediators. In this context, Gröger et al. found the highest TNF-α, IL-6 and 8-isoprostane blood levels in hypothermic animals (32°C) at 12 h after resuscitation and at the end of the study period, whereas induced hypothermia of 35°C resulted in an almost similar response as normothermia [43]. Limitations Besides the well-described variables which have some influence on the results of experimental studies (anaesthesia before trauma, application of drugs and infusions, small sample sizes) [58], the species-specific differences between humans and pigs deserve particular attention. It is well known that it is difficult to achieve a state of coagulopathy in pigs [23]. In fact, induction of haemorrhagic shock, chest and abdominal trauma did not result in a significant reduction of coagulation activity [39]. This is in line with results of previous studies in which moderate haemorrhage alone and resuscitation with lactated Ringer's solution did not suffice to deplete coagulation substances to alter clotting time [23, 36]. Other authors performed haemodilution before the induction of injury in order to standardize coagulopathy, which certainly fails to mimic the clinical situation [59]. Therefore, data on the effects of therapeutically induced hypothermia on posttraumatic coagulation disorders originating from porcine models and the transferability to the human situation have to be interpreted carefully [23, 36, 38, 44, 58, 60]. In addition, different vasopressin receptors exist in pigs and humans that may result in a different haemodynamic response to exogenously administered vasopressin [61, 62]. Porcine granulocytes should not be considered representative for the human setting due to differences of elastase release and activity. Furthermore, the reticuloendothelial system in swine is located in the pulmonary region, which is in contrast to the situation in humans. This might have significant effects, e.g., on the pulmonary artery pressure in specific situations: acute challenges in swine often result in marked pulmonary arterial hypertension [63].

hermore, the reticuloendothelial system in swine is located in the pulmonary region, which is in contrast to the situation in humans. This might have significant effects, e.g., on the pulmonary artery pressure in specific situations: acute challenges in swine often result in marked pulmonary arterial hypertension [63]. Furthermore, the observation time has to be considered when interpreting the effects of hypothermia found in porcine models of trauma-haemorrhage. In many of the studies, timing was selected to mirror the clinical setting as faithfully as possible, focusing on the first golden hour after trauma [30, 44, 47, 53]. Therefore, it also has to be assumed that a limited observation time might lead to undetected treatment effects (e.g. functional recovery of endothelial or organ function) or compensatory mechanisms of therapeutically induced hypothermia. Furthermore, the incidence of re-bleeding has to be investigated due to the potential negative side effects on the coagulation system. Moreover, immunosuppressive long-term effects with an associated increased incidence of infectious complications have to be taken in account. Finally, potential rebound effects need to be considered. A delayed overwhelming inflammatory response might be postulated after termination of induced hypothermia and rewarming. Accordingly, it might be speculated that a hypothermia-associated improvement of cellular and organ function might deteriorate after rewarming and ultimately result in the same severity of cellular and organ dysfunction as maintenance of normothermia. Hypothermia also seems to affect posttraumatic apoptosis including the inhibition of activation of caspase enzymes and the preservation of mitochondrial function [64]. As the apoptotic process has been described to occur not only early but also relatively late following trauma-haemorrhage and seems to continue for up to 3 days, long-term models are of major importance before implementing therapeutically induced hypothermia in the clinical setting. In order to achieve a longer observation period, animals were extubated after 6 to 24 h in some other experiments and then observed for several days in an awake state [42, 45, 54]. Although this allows the observation of some of the long-term effects of therapeutically induced hypothermia, it is not directly comparable to the clinical situation of multiply injured patients, who are intubated, sedated and treated on the intensive care unit for several days.

ral days in an awake state [42, 45, 54]. Although this allows the observation of some of the long-term effects of therapeutically induced hypothermia, it is not directly comparable to the clinical situation of multiply injured patients, who are intubated, sedated and treated on the intensive care unit for several days. However, the performance of long-term experiments also has major challenges and possible limitations. These include logistic necessities and the associated costs. Furthermore, both personnel with clinical and scientific experience in intensive care medicine and trauma surgery as well as adequately equipped facilities are needed to ensure reliable results. It also has to be taken into account that growth might be an important issue in swine when observing long-term outcomes with observation periods of several weeks. However, this aspect might be of minor importance in models with a study period of several days. It will also gain increasing importance to include potential co-morbidities in the experimental models to resemble the higher number of older trauma patients with a relevant incidence of pre-existing morbidities (e.g. arteriosclerosis, hypertension, diabetes, etc.).

rtance in models with a study period of several days. It will also gain increasing importance to include potential co-morbidities in the experimental models to resemble the higher number of older trauma patients with a relevant incidence of pre-existing morbidities (e.g. arteriosclerosis, hypertension, diabetes, etc.). Conclusion Overall, it remains unclear whether therapeutically induced hypothermia has a promising potential for clinical use in the posttraumatic setting. This presentation of porcine models shows a significant variability of the experimental settings. This is due to the different purposes or the aims of the studies, which result in significant differences of the conditions for induction of trauma (severity and duration of haemorrhage, isolated or combined insults) and hypothermia/rewarming (method, time points, and duration) between porcine models of trauma-haemorrhage. This might partly lead to conflicting study results. However, it is also difficult to prioritize between the significance of the studies as every model has its strengths and weaknesses without an intrinsic value of an experimental model per se. In conclusion, the current state of knowledge of therapeutically induced hypothermia after severe trauma is clearly not sufficient enough to start application in the clinical setting. Induction of hypothermia as a treatment strategy for severely injured patients would require more essential information, such as the optimal rate, time point, and duration of hypothermia induction as well as of rewarming and long-term observational studies.

ufficient enough to start application in the clinical setting. Induction of hypothermia as a treatment strategy for severely injured patients would require more essential information, such as the optimal rate, time point, and duration of hypothermia induction as well as of rewarming and long-term observational studies. Electronic supplementary material Additional file 1: Table S1: Accidental hypothermia [14–29, 33, 35–37, 55, 56, 60, 65–69]. (DOCX 206 KB) Additional file 2: Table S2: Induced hypothermia [10, 11, 30, 40, 45, 47–50]. (DOCX 152 KB) Competing interests The authors declare that they have no competing interests. Authors’ contributions FH identified the studies and wrote the article. FH and SF performed assessment of suitability of full-text articles. PR, SR, MHL, AS, MVG and HA read the selected full-text articles, summarized the content and corrected the manuscript. HCP discussed eligibility of studies in case of concerns of the other authors and corrected the manuscript. All authors read and approved the final manuscript. Acknowledgements We would like to thank Fritz Seidl, M.A. Translating and Interpreting, for proofreading our manuscript.

Background Several molecular mechanisms of inflammation and cellular damage have been implicated in the pathogenesis of sepsis including excessive reactive oxygen species (ROS) generation [1]. Main sources of ROS in the lung during sepsis are inflammatory cells and mitochondria [2, 3]. Production of ROS leads to lipid, protein, and extracellular matrix damage, which increases pulmonary inflammation [4, 5]. Antioxidants are known to counteract the deleterious effects of ROS. Superoxide dismutase (SOD) is a component of antioxidant response and catalyzes the conversion of superoxide anions to hydrogen peroxide [6]. Three SODs are found in mammals and regulate the concentration of superoxide: a cytosolic (SOD1), a mitochondrial (SOD2), and an extracellular (SOD3), which bind to both cell surfaces and extracellular matrices [7]. sod3 gene is highly expressed in the lung, where it plays a major protective role by controlling oxidative stress and inflammation and regulating redox homeostasis of the airways [8]. Localization of SOD3 in the lung depends on its ability to bind to the extracellular matrix by a heparan sulfate domain, which can be fragmented by oxidative damage [5, 9]. Furthermore, extracellular matrix fragments stimulate inflammatory cell migration, which is of concern since matrix components are widely distributed throughout the interstitium [9]. Proteolytic cleavage of SOD3's anchorage domain alters its tissue distribution and, consequently, the oxidant/antioxidant balance [10].

. Furthermore, extracellular matrix fragments stimulate inflammatory cell migration, which is of concern since matrix components are widely distributed throughout the interstitium [9]. Proteolytic cleavage of SOD3's anchorage domain alters its tissue distribution and, consequently, the oxidant/antioxidant balance [10]. In addition, bioactivity of nitric oxide (NO) partially depends on its interaction with ROS, particularly superoxide anions [11]. NO reacts with superoxide to form peroxynitrite (ONOO−), which induces protein oxidation and DNA damage [12]. SOD3 prevents the inactivation of NO by superoxide; therefore, an increase in sod3 expression in blood vessels preserves endothelial function by overcoming oxidative stresses [13]. Furthermore, several evidences suggest that SOD3 may have a protective role against inflammation [10, 14]. Thus, it was hypothesized that endogenous SOD3 could have a major role in lung defenses against oxidative stress during sepsis development. The aims of this study are (1) to determine if there is a relationship between the expression and activity of SOD and the occurrence of oxidative stress and (2) to determine if the administration of a SOD mimetic is able to prevent lung damage in an animal model of sepsis.

st oxidative stress during sepsis development. The aims of this study are (1) to determine if there is a relationship between the expression and activity of SOD and the occurrence of oxidative stress and (2) to determine if the administration of a SOD mimetic is able to prevent lung damage in an animal model of sepsis. Methods Animals Three-month-old male Wistar rats (350 to 400 g) were obtained from our breeding colony. The rats were caged in groups of five, had free access to food and water, and were maintained on a 12-h light–dark cycle (600 to 1800 hours) in a temperature-controlled colony room (22°C ± 1°C). The research protocol was approved by the Ethical Committee for Animal Experimentation of Universidade do Extremo Sul Catarinense under protocol number 21/2011.

ve, had free access to food and water, and were maintained on a 12-h light–dark cycle (600 to 1800 hours) in a temperature-controlled colony room (22°C ± 1°C). The research protocol was approved by the Ethical Committee for Animal Experimentation of Universidade do Extremo Sul Catarinense under protocol number 21/2011. Cecal ligation and perforation surgery The animals were subjected to cecal ligation and perforation (CLP) as previously described [15]. Briefly, the rats were anesthetized with ketamine (80 mg/kg). Under aseptic conditions, a 3-cm midline laparotomy was performed to allow exposure of the cecum with the adjoining intestine. The cecum was tightly ligated with a 3.0 silk suture at its base (below the ileocecal valve) and was then perforated a single time with a 14-gauge needle. The cecum was gently squeezed to extrude a small amount of feces from the perforation site into the peritoneal cavity. The animals were resuscitated with normal saline (50 mL/kg, subcutaneous) immediately and 12 h after CLP. The sham-operated group was submitted to all surgical procedures, but the cecum was neither ligated nor perforated. The animals were killed by decapitation at 3, 6, and 12 h after surgery, and the lung was removed for subsequent analyses. The number of animals in each group per experiment was ten. In some experiments, a SOD mimetic was administered once by intra-tracheal instillation immediately after CLP induction (50 mg/kg), and the lung was removed 24 h after for subsequent analyses. This mimetic has been previously described by Horn et al. [16].

es. The number of animals in each group per experiment was ten. In some experiments, a SOD mimetic was administered once by intra-tracheal instillation immediately after CLP induction (50 mg/kg), and the lung was removed 24 h after for subsequent analyses. This mimetic has been previously described by Horn et al. [16]. Total SOD activity The SOD activity was measured by inhibition of adrenaline auto-oxidation followed by spectrophotometry as previously described [17].

es. The number of animals in each group per experiment was ten. In some experiments, a SOD mimetic was administered once by intra-tracheal instillation immediately after CLP induction (50 mg/kg), and the lung was removed 24 h after for subsequent analyses. This mimetic has been previously described by Horn et al. [16]. Total SOD activity The SOD activity was measured by inhibition of adrenaline auto-oxidation followed by spectrophotometry as previously described [17]. Immunoblotting The lung samples were lysed in Laemmli sample buffer (62.5 mM Tris–HCl, pH 6.8, 1% (w/v) SDS, 10% v/v) glycerol). Protein (30 μg) was fractionated by SDS-polyacrylamide gel electrophoresis and then electroblotted onto nitrocellulose membranes. Protein loading and electroblotting efficiencies were verified by Ponceau S staining. The membrane was blocked in Tween-Tris-buffered saline (TTBS, 100 mM Tris–HCl, pH 7.5, containing 0.9% NaCl and 0.1% Tween-20) containing 5% albumin. The membranes were incubated overnight at 4°C with rabbit polyclonal antibody, targeting SOD1 (Santa Cruz Biotechnology, CA, USA) (dilution range 1:400), SOD2 (Santa Cruz Biotechnology) (dilution range 1:400), SOD3 (Santa Cruz Biotechnology) (dilution range 1:750), iNOS (Santa Cruz Biotechnology) (dilution range 1:400) or anti-β-actin 1:2000, in the presence of 5% milk. Thereafter, the membranes were washed with TTBS. Anti-rabbit immunoglobulin G (IgG) peroxidase-linked secondary antibody was incubated (1:10,000 dilution range), and the immunoreactivity was detected by enhanced chemiluminescence using ECL Plus kit (Thermo Fisher Scientific Inc., Pittsburgh, PA, USA). Densitometric analyses of the films were performed with ImageQuant software (GE Healthcare Life Sciences, Billerica, MA, USA). The blots were developed such that the signals are linear and non-saturating which are required for densitometry. All results were expressed as a relative ratio comparing the immunocontent of SOD1, SOD2, SOD3, and iNOS with that of the β-actin internal control [18].

re (GE Healthcare Life Sciences, Billerica, MA, USA). The blots were developed such that the signals are linear and non-saturating which are required for densitometry. All results were expressed as a relative ratio comparing the immunocontent of SOD1, SOD2, SOD3, and iNOS with that of the β-actin internal control [18]. Enzyme-linked immunosorbent assay to 3-nitrotyrosine contents An indirect enzyme-linked immunosorbent assay (ELISA) was performed to analyze the changes in 3-nitrotyrosine content. Briefly, an anti-3-nitrotyrosine polyclonal rabbit antibody (Santa Cruz Biotechnology) was diluted 2,000-fold in PBS with 5% albumin according to the manufacturer's instructions. Then, microtiter plates (96 wells, with flat bottom) were coated for 24 h with the samples that had been diluted 1:2 in PBS with 5% albumin. The plates were washed four times with wash buffer (PBS with 0.05% Tween-20), and the antibody was added to each plate for 2 h at room temperature. After washing, a second incubation with anti-rabbit antibody peroxidase conjugate (diluted 1:1,000) was performed for 1 h at room temperature. After the addition of substrates, the samples were read in a plate spectrophotometer at 450 nm. The results are expressed as changes in the percentage among the groups [19].

re. After washing, a second incubation with anti-rabbit antibody peroxidase conjugate (diluted 1:1,000) was performed for 1 h at room temperature. After the addition of substrates, the samples were read in a plate spectrophotometer at 450 nm. The results are expressed as changes in the percentage among the groups [19]. Semi-quantitative reverse transcription polymerase chain reaction All transcriptional analyses were performed in the samples in which prior Western blotting experiments revealed differences in the immunocontent. The goal was to evaluate the contribution of each gene to transcriptional changes in the immunocontent of each enzyme. Total RNA was isolated from the rat lung using TRIzol® reagent (Invitrogen, Carlsbad, CA, USA), according to the manufacturer's instructions. Using a spectrophotometer, the purity of the RNA was quantified by calculating the ratio between absorbance values at 260 and 280 nm, and its integrity was confirmed by electrophoresis using a 1.0% agarose gel. Afterward, cDNA species were synthesized using ImProm-II™ Reverse Transcription System (Promega®, Madison, WI, USA), as described by the supplier's instruction. The cDNA products (1 μL) were used as a template for each polymerase chain reaction (PCR) amplification. The PCR parameters were first optimized. Thereafter, the reactions were performed, such that product detection could be performed within the linear phase of messenger ribonucleic acid (mRNA) transcript amplification for each primer pair (Table 1). PCR for the β-actin gene was performed in a total volume of 20 μL using 0.1 μM of each primer, 0.2 μM dNTP, 1.6 mM MgCl2, and 0.2 U Taq platinum DNA polymerase (Invitrogen). For PCR amplification of sod1, sod2, and sod3, the reaction was performed in a total volume of 25 μL using 0.2 μM of each primer, 0.2 μM dNTP, 1.6 mM MgCl2, and 0.25 U Taq platinum DNA polymerase (Invitrogen). The conditions for sod1, sod2, and sod3 PCRs were as follows: initial 1-min denaturation step at 94°C, another 1-min denaturation step at 94°C, 1-min annealing step at 60°C, 1-min extension step at 72°C for 30 cycles, and a final 10-min extension at 72°C. The conditions for the β-actin PCR were as follows: initial 1-min denaturation step at 94°C, another 1-min denaturation step at 94°C, 1-min annealing step at 54°C, 1-min extension step at 72°C for 35 cycles, and a final 10-min extension at 72°C. For each PCR set, a negative control was included.

-min extension at 72°C. The conditions for the β-actin PCR were as follows: initial 1-min denaturation step at 94°C, another 1-min denaturation step at 94°C, 1-min annealing step at 54°C, 1-min extension step at 72°C for 35 cycles, and a final 10-min extension at 72°C. For each PCR set, a negative control was included. The PCR products were then analyzed on a 1% agarose gel containing GelRed® (Biotium, Hayward, CA, USA) and visualized with ultraviolet light. The Low DNA Mass Ladder (Invitrogen) was used as a molecular marker, and the samples were normalized against the constitutively expressed β-actin gene. The band intensities were measured by optical densitometry analysis, and the enzyme/β-actin mRNA ratios were established for each treatment using the freeware Image J 1.37. Each experiment was repeated at least four times using RNA isolated from independent extractions.Table 1 PCR primer design Enzymes Primer sequences (5’-3’) Anneling temperature (°C) PCR product (bp) GenBank accession number (mRNA) sod1 F-TGCGTGCTGAAGGGCGACGGTC 60 438 BC082800 R-AATCCCAATCACACCACAAGCCAAGC sod2 F-CCTACGTGAACAATCTGAACGTCACCGAG 60 373 BC070913.1 R-CCCAGCAGTGGAATAAGGCCTGTGG sod3 F-GCCGAGCAGAACACCTCCAACCACG 60 377 BC061861.1 R-CGCCGCTTCTTGCGCTCCTTTG β-actin F-TATGCCAACACAGTGCTGCTGG 54 210 NP_742006 R-TACTCCTGCTTCCTGATCCACAT

Enzymes Primer sequences (5’-3’) Anneling temperature (°C) PCR product (bp) GenBank accession number (mRNA) sod1 F-TGCGTGCTGAAGGGCGACGGTC 60 438 BC082800 R-AATCCCAATCACACCACAAGCCAAGC sod2 F-CCTACGTGAACAATCTGAACGTCACCGAG 60 373 BC070913.1 R-CCCAGCAGTGGAATAAGGCCTGTGG sod3 F-GCCGAGCAGAACACCTCCAACCACG 60 377 BC061861.1 R-CGCCGCTTCTTGCGCTCCTTTG β-actin F-TATGCCAACACAGTGCTGCTGG 54 210 NP_742006 R-TACTCCTGCTTCCTGATCCACAT Immunohistochemistry Lung sections (sham 0 and 12 and CLP 0 and 12, n = 3 in each group) were deparaffinized and hydrated. To examine their histological features, the lungs were stained with hematoxylin and eosin (H&E). After blocking of endogenous peroxidase, antigen retrieval was performed in a high-temperature Tris-citrate buffer (pH 7.2). The rabbit polyclonal anti-SOD3 (Santa Cruz Biotechnology) (diluted 1:1,600) was used as the primary antibody. The Vectastin ABC Kit (Vector Laboratories, Burlingame, CA, USA) was used as the secondary antibody, and 3,3-diaminobenzidine (DAB, Sigma, St. Louis, MO, USA) was used as the chromogen. Thereafter, the sections were counterstained with Harris hematoxylin (Merck, Darmstadt, Germany). In the negative controls, the first antibody was omitted from the procedure, and the tissues were incubated with bovine serum albumin (BSA) instead. Cytokine determination Lung levels of tumor necrosis factor (TNF)-α, interleukin (IL)-1β, IL-6, and IL-10 were determined by ELISA according to the manufacturer's instructions (PrepoTech, Ribeirão Preto, SP, Brazil). All samples were assayed in duplicate.

Immunohistochemistry Lung sections (sham 0 and 12 and CLP 0 and 12, n = 3 in each group) were deparaffinized and hydrated. To examine their histological features, the lungs were stained with hematoxylin and eosin (H&E). After blocking of endogenous peroxidase, antigen retrieval was performed in a high-temperature Tris-citrate buffer (pH 7.2). The rabbit polyclonal anti-SOD3 (Santa Cruz Biotechnology) (diluted 1:1,600) was used as the primary antibody. The Vectastin ABC Kit (Vector Laboratories, Burlingame, CA, USA) was used as the secondary antibody, and 3,3-diaminobenzidine (DAB, Sigma, St. Louis, MO, USA) was used as the chromogen. Thereafter, the sections were counterstained with Harris hematoxylin (Merck, Darmstadt, Germany). In the negative controls, the first antibody was omitted from the procedure, and the tissues were incubated with bovine serum albumin (BSA) instead. Cytokine determination Lung levels of tumor necrosis factor (TNF)-α, interleukin (IL)-1β, IL-6, and IL-10 were determined by ELISA according to the manufacturer's instructions (PrepoTech, Ribeirão Preto, SP, Brazil). All samples were assayed in duplicate. Lung permeability assay Permeability changes were measured by Evan's blue dye (EBD) leakage from the blood into the airways. EBD (20 mg/kg) was administered by femoral vein injection 1 h before the end of the experiments. One hour later, the mice were bled by cardiac puncture, and the pulmonary vasculature was flushed by right ventricle puncture. The pulmonary vessels were perfused with normal saline to remove EBD from the vascular spaces. The lungs were removed en bloc and dried at 60°C for 24 h. EBD was extracted in formamide at 37°C for 24 h and quantitated by its fluorescence intensity. The extravasated EBD concentration in lung homogenate was calculated against a standard curve.

ere perfused with normal saline to remove EBD from the vascular spaces. The lungs were removed en bloc and dried at 60°C for 24 h. EBD was extracted in formamide at 37°C for 24 h and quantitated by its fluorescence intensity. The extravasated EBD concentration in lung homogenate was calculated against a standard curve. Statistical analysis Data are expressed as mean ± standard deviation. The means for the different groups were compared by t test or one-way or two-way ANOVA followed by Tukey test, depending on the number of experimental groups. Statistical significance was assigned to p < 0.05.

ere perfused with normal saline to remove EBD from the vascular spaces. The lungs were removed en bloc and dried at 60°C for 24 h. EBD was extracted in formamide at 37°C for 24 h and quantitated by its fluorescence intensity. The extravasated EBD concentration in lung homogenate was calculated against a standard curve. Statistical analysis Data are expressed as mean ± standard deviation. The means for the different groups were compared by t test or one-way or two-way ANOVA followed by Tukey test, depending on the number of experimental groups. Statistical significance was assigned to p < 0.05. Results Levels of SOD were determined in the lung 3 to 12 h after sepsis. There was a decrease in the immunocontent of SOD1 at 3 and 12 hours, but not at 6 h, after sepsis induction (Figure 1A,B,C). Furthermore, an increase in the immunocontent of SOD2 was observed at all times (Figure 2A,B,C), and there was an increase in the immunocontent of SOD3 at 6 and 12 h after sepsis induction (Figure 3A,B,C). The SOD activity increased at 6 and 12 h (6 h, 6.03 ± 2.1 vs. 21.1 ± 5.6, p = 0.02; 12 h, 4.9 ± 1.7 vs. 22.3 ± 7.6 U/mg protein, p = 0.01) when compared to the sham and CLP animals, respectively. To determine if these changes of protein content were associated with alterations in gene expression, semi-quantitative reverse transcription polymerase chain reaction (RT-PCR) was performed. There was no consistent pattern of gene expression; while sod1 and sod2 gene expressions increased 12 h after sepsis (Figure 1D,E and Figure 2D,E,F), sod3 gene expression decreased at this time point (Figure 3B).Figure 1 Cytosolic superoxide dismutase (SOD1) alterations early after sepsis induction. The animals were submitted to sepsis or sham-operated, and 3 h (A), 6 h (B), and 12 h (C) after surgery, the immunocontent of SOD1 was evaluated. The sod1 gene expression was evaluated at 3 h (D) and 12 h (E) after surgery. The data are expressed as mean ± SD (n = ten animals/group). p value < 0.05 was considered significant when compared to the control. * denotes p < 0.05.

nd 3 h (A), 6 h (B), and 12 h (C) after surgery, the immunocontent of SOD1 was evaluated. The sod1 gene expression was evaluated at 3 h (D) and 12 h (E) after surgery. The data are expressed as mean ± SD (n = ten animals/group). p value < 0.05 was considered significant when compared to the control. * denotes p < 0.05. Figure 2 Mitochondrial superoxide dismutase (SOD2) alterations early after sepsis induction. The animals were submitted to sepsis or sham-operated, and 3 h (A), 6 h (B), and 12 h (C) after surgery, the immunocontent of SOD2 was evaluated. The sod2 gene expression was evaluated at 3 h (D), 6 h (E), and 12 h (F) after surgery. The data are expressed as mean ± SD (n = ten animals/group). p value < 0.05 was considered significant when compared to the control. Figure 3 Extracellular superoxide dismutase (SOD3) alterations early after sepsis induction. The animals were submitted to sepsis or sham-operated, and 3 h (A), 6 h (B), and 12 h (C) after surgery, the immunocontent of SOD3 was evaluated. The sod3 gene expression was evaluated at 3 h (D) and 12 h (E) after surgery. The data are expressed as mean ± SD (n = ten animals/group). p value < 0.05 was considered significant when compared to the control.

or sham-operated, and 3 h (A), 6 h (B), and 12 h (C) after surgery, the immunocontent of SOD3 was evaluated. The sod3 gene expression was evaluated at 3 h (D) and 12 h (E) after surgery. The data are expressed as mean ± SD (n = ten animals/group). p value < 0.05 was considered significant when compared to the control. Nitrotyrosine levels increased 3 and 12 hours after CLP induction (Figure 4), suggesting that despite the increase of SOD3 content the reaction between superoxide and NO was occurring. Lung content of iNOS did not increase in the lung of septic animals (Figure 5A-C), but it was increased in the pulmonary artery (Figure 5D-E).Figure 4 Nitrotyrosine content early after sepsis induction. The animals were submitted to sepsis or sham-operated, and 3, 6, and 12 h after surgery, the content of nitrotyrosine was evaluated. The data are expressed as mean ± SD (n = ten animals/group). p value < 0.05 was considered significant when compared to the control. Figure 5 Inducible nitric oxide synthase (iNOS) alterations early after sepsis induction. The animals were submitted to sepsis or sham-operated, and 3 h (A), 6 h (B), and 12 h (C) after surgery, the immunocontent of iNOS was evaluated in the lung and in the pulmonary artery (3 h (D), 6 h (E), 12 h (F)). The data are expressed as mean ± SD (n = ten animals/group). p value < 0.05 was considered significant when compared to the control.

to sepsis or sham-operated, and 3 h (A), 6 h (B), and 12 h (C) after surgery, the immunocontent of iNOS was evaluated in the lung and in the pulmonary artery (3 h (D), 6 h (E), 12 h (F)). The data are expressed as mean ± SD (n = ten animals/group). p value < 0.05 was considered significant when compared to the control. These alterations were associated with capillary congestion and infiltration of neutrophils into the alveolar septa (Figure 6). It was possible that the increase in SOD3 protein content did not protect the lung from oxidant injury due to a disorganized distribution resulting from proteolytic cleavage of its heparin-binding domain, but SOD was adequately distributed across the lung (Figure 6). It was expressed mainly in the bronchial and alveolar epithelia of the CLP animals and to a minor extent in the leukocytes and endothelial cells (Figure 6).Figure 6 Distribution of extracellular superoxide dismutase (SOD3) in the lung early after sepsis. The animals were submitted to sepsis or sham-operated, and 12 h after surgery, the distribution of lung SOD3 was determined by immunohistochemistry. Representative photographs of septic (A) and sham (B) animals. Original magnification is × 6,200.

extracellular superoxide dismutase (SOD3) in the lung early after sepsis. The animals were submitted to sepsis or sham-operated, and 12 h after surgery, the distribution of lung SOD3 was determined by immunohistochemistry. Representative photographs of septic (A) and sham (B) animals. Original magnification is × 6,200. It was expected that SOD3 protected the lung from oxidative damage and inflammation, but this was not supported by the presented data. Thus, it was administered a SOD mimetic, and it decreased the lung nitrotyrosine and cytokine levels, as well as preserved lung permeability (Figure 7A,B,C).Figure 7 Effects of the treatment with a SOD mimetic on lung injury induced by sepsis. The animals were submitted to sepsis or sham-operated, and immediately after surgery, a SOD mimetic was administered once by intra-tracheal instillation. Twenty-four hours after surgery, the content of nitrotyrosine (A), lung permeability (B), interleukin-6 (C), and tumor necrosis factor (D) was determined. The data are expressed as mean ± SD (n = ten animals/group). p value < 0.05 was considered significant when compared to the control. Discussion Here, in a CLP model, we demonstrated that during sepsis development, the levels of SOD2 and SOD3 increased in the lung, but these did not prevent nitrosative damage and inflammation. Furthermore, we showed that superoxide-derived lung inflammation could be attenuated by the administration of an exogenous SOD mimetic.

Here, in a CLP model, we demonstrated that during sepsis development, the levels of SOD2 and SOD3 increased in the lung, but these did not prevent nitrosative damage and inflammation. Furthermore, we showed that superoxide-derived lung inflammation could be attenuated by the administration of an exogenous SOD mimetic. An increase in the levels of SOD2 and SOD3 was observed; thus, it was believed that the lung was protected from superoxide-derived oxidative damage. SOD3 is a potent inhibitor of inflammation in lung injury models [20], and mice lacking the sod3 gene are more sensitive to lethal levels of hyperoxia [14]. Furthermore, SOD3 protects against endothelial dysfunction in mice treated with endotoxin [21]. In a murine model of emphysema, the mice that overexpress the sod3 gene or are treated with a SOD mimetic have improved lung compliance, decreased neutrophil influx, release of proinflammatory mediators, and oxidative damage [22]. Here, we show that despite the increase in SOD content, there is an increase in lung nitrotyrosine content.

of emphysema, the mice that overexpress the sod3 gene or are treated with a SOD mimetic have improved lung compliance, decreased neutrophil influx, release of proinflammatory mediators, and oxidative damage [22]. Here, we show that despite the increase in SOD content, there is an increase in lung nitrotyrosine content. In general, inflammatory conditions (including lipopolysaccharide, hypoxia, asbestos exposure, bleomycin, and hyperoxia) induce a decrease in SOD3 content. This observation is most likely a result of proteolysis of its heparin-binding domain rather than alterations in its gene expression [20, 23]. In contrast, at earlier time points after sepsis induction, we demonstrate an increase in SOD3 levels that are properly distributed throughout the lung parenchyma, but this is not sufficient to prevent the formation of peroxynitrite. Increased oxidative damage, despite the presence of higher levels of SOD3, was described in a model of cerebral ischemia [24].

r sepsis induction, we demonstrate an increase in SOD3 levels that are properly distributed throughout the lung parenchyma, but this is not sufficient to prevent the formation of peroxynitrite. Increased oxidative damage, despite the presence of higher levels of SOD3, was described in a model of cerebral ischemia [24]. The reaction of NO with superoxide may lead to an increase in ONOO−, which mediates nitration of tyrosine residues in certain enzymes, including SODs. This modification leads to a reduction in their activity and, as a result, further increase in superoxide levels [25]. Specifically, ONOO− is known to inactivate SOD1 and SOD2 [26, 27]. Due to the structural similarity between SOD1 and SOD3, we speculate that the activity of SOD3 may similarly be decreased by ONOO−. Here, we demonstrate that septic animals have increased levels of nitrotyrosine. This occurs even in the presence of higher SOD3 levels and total SOD activity. This pattern is not expected since in different models, an increase in SOD3 activity is usually associated with a decrease in ONOO−[28]. However, when we used an exogenous SOD mimetic, it is observed that there is a decrease in the nitrotyrosine levels. These observations suggest that the endogenous SOD system is not able to prevent oxidative damage in the CLP model. In addition, a SOD mimetic decreased the markers of lung inflammation and improved lung permeability. Moreover, these findings raise the possibility that SOD mimetics could be used to treat sepsis-related lung injury.

tions suggest that the endogenous SOD system is not able to prevent oxidative damage in the CLP model. In addition, a SOD mimetic decreased the markers of lung inflammation and improved lung permeability. Moreover, these findings raise the possibility that SOD mimetics could be used to treat sepsis-related lung injury. There are some limitations to our study. Firstly, SOD mimetic was administered immediately after CLP, when the septic response was not fully developed and animals did not presented clinical signs of severe infection. Thus, this design did not reflect the clinical scenario, but can be relevant to understand the mechanisms associated with sepsis development. Secondly, nitrotyrosine was not measured by gold-standard techniques, such as tandem mass spectrometry or high-performance liquid chromatography. Since the increase of nitrotyrosine is classically found in sepsis models, the use of semi-quantitative ELISA as a marker of oxidative damage seems to be adequate to the study aim. Conclusions In conclusion, this study demonstrated an increase in SOD3 levels following sepsis; however, this was not able to prevent oxidative stress and inflammation associated with ONOO− production. Nonetheless, the administration of a SOD mimetic had a positive impact, which suggests that administration of exogenous SOD may improve lung function in sepsis. Competing interests The authors declare that they have no competing interests. Authors’ contributions

Conclusions In conclusion, this study demonstrated an increase in SOD3 levels following sepsis; however, this was not able to prevent oxidative stress and inflammation associated with ONOO− production. Nonetheless, the administration of a SOD mimetic had a positive impact, which suggests that administration of exogenous SOD may improve lung function in sepsis. Competing interests The authors declare that they have no competing interests. Authors’ contributions LC, MRB, JCFM, AH, TM, CR, and FD-P conceived this study, participated in the design of the study, and drafted the manuscript. RCG, VRG, CDT, FV, LWK, GMTO, MABP, KVM, and CF participated in the design of the study and in the collection of data. All authors read and approved the final manuscript. Acknowledgements This research was supported by grants from CAPES, FAPESC, UNESC, and CNPq-Instituto Nacional de Ciência e Tecnologia Translacional em Medicina (INCT-TM). Further acknowledgment is expressed to all authors who contributed to this study.

Background Acute respiratory distress syndrome (ARDS) is a common cause of mortality and morbidity in critically ill patients [1]. Although indispensable in the support of ARDS patients, artificial ventilation involves the application of mechanical forces to the lung parenchyma that can further induce injury [2], adding morbidity and mortality [3]. Reducing tidal volumes (VTs) below 6 mL/kg of ideal body weight could potentially decrease the cyclic stretch imposed on the lung [4, 5]. Conversely, excessively low VTs have the potential to lead to clinically significant hypercapnia-related acidosis [6] with harmful side effects [7, 8]. In this scenario, high-frequency oscillatory ventilation (HFOV) has been tested, because of its ability to provide adequate gas exchange even at very low tidal volumes [9–13]. This technique, however, may be cumbersome because it requires a dedicated ventilator and special training. Additionally, the VT delivered can be susceptible to variations in airway resistance such as that which occurs with lung secretions [14]. Last but not least, it requires the use of high airway pressures, which may have deleterious effects, especially on the right ventricle [15]. Recently, two clinical studies in ARDS patients showed neutral [16] or disappointing [17] results in terms of mortality when HFOV was compared to a conventional mechanical ventilation strategy.

ast, it requires the use of high airway pressures, which may have deleterious effects, especially on the right ventricle [15]. Recently, two clinical studies in ARDS patients showed neutral [16] or disappointing [17] results in terms of mortality when HFOV was compared to a conventional mechanical ventilation strategy. An alternative approach could be to apply moderately high frequency positive-pressure ventilation (HFPPV) using conventional mechanical ventilators. A similar strategy was explored in the 1980s [18, 19], but with special ventilators and before the well-established recognition of the importance of lung-protective strategies. HFPPV consists of applying respiratory rates intermediate between those used conventionally (≤ 35 breaths/min) and those used during HFOV (180 to 800 breaths/min). Potential advantages over HFOV would be the possibility to control the VT delivered, the use of conventional mechanical ventilators obviating the need for specialized training, and maintenance of a low mean airway pressure. In this feasibility study, we tested in a swine model of ARDS whether such a strategy could result in VT below 6 mL/kg while avoiding further increases in the partial pressure of arterial carbon dioxide (PaCO2) and maintaining a reasonable mean airway pressure (Pmean).

ng, and maintenance of a low mean airway pressure. In this feasibility study, we tested in a swine model of ARDS whether such a strategy could result in VT below 6 mL/kg while avoiding further increases in the partial pressure of arterial carbon dioxide (PaCO2) and maintaining a reasonable mean airway pressure (Pmean). Methods This study was approved by the Institutional Animal Research Ethics Committees of Hospital Sírio Libanês and of Faculdade de Medicina da Universidade de São Paulo, both in São Paulo, Brazil, and was performed according to the National Institutes of Health (USA) guidelines for the use of experimental animals. The experiments were done in eight previously healthy Agroceres pigs.

h Ethics Committees of Hospital Sírio Libanês and of Faculdade de Medicina da Universidade de São Paulo, both in São Paulo, Brazil, and was performed according to the National Institutes of Health (USA) guidelines for the use of experimental animals. The experiments were done in eight previously healthy Agroceres pigs. Instrumentation The animals were fasted overnight before the experiment with free access to water. They received an intramuscular injection of midazolam (0.3 mg/kg; Dormonid®, Roche, São Paulo, Brazil) and acepromazine (0.5 mg/kg; Acepran®, Andrômaco, São Paulo, Brazil). Through an auricular vein, anesthesia was induced with thionembutal (12 mg/kg; Tiopental®, Abbott, São Paulo, Brazil) and muscular relaxation with pancuronium bromide (0.1 mg/kg; Pavulon®, AKZO Nobel, São Paulo, Brazil). They were then submitted to tracheal intubation (cuffed 7.5-French cannula) and connected to the Servo-300 mechanical ventilator (Maquet, Rastatt, Germany) with the following parameters in a volume-controlled mode: tidal volume of 8 to 10 mL/kg, positive end-expiratory pressure (PEEP) of 5 cmH2O, inspiratory fraction of oxygen (FiO2) adjusted to keep arterial saturation between 94% and 96%, and respiratory rate (RR) necessary to keep PaCO2 between 35 and 45 mmHg. Anesthesia was maintained during the study period with midazolam (0.3 mg/kg/h) and fentanyl citrate (5 μg/kg/h; Fentanyl®, Janssen-Cilag, São Paulo, Brazil) and muscular relaxation with pancuronium bromide (0.2 mg/kg/h). The adequate depth of anesthesia during the surgical period was evaluated with maintenance of physiological variables (heart rate and arterial pressure) and absence of reflexes (corneal and hind limb flexion response), as well as unresponsiveness to stimuli during manipulation. Supplementary boluses of 3 to 5 μg/kg of fentanyl and 0.1 to 0.5 mg/kg of midazolam were administered as necessary. A continuous drip of 1,000 mL/h of Lactated Ringer’s solution was infused until the end of the induction of pulmonary injury, and then a continuous infusion of 5 mL/kg/h of Lactated Ringer was maintained until the end of the study.

to 5 μg/kg of fentanyl and 0.1 to 0.5 mg/kg of midazolam were administered as necessary. A continuous drip of 1,000 mL/h of Lactated Ringer’s solution was infused until the end of the induction of pulmonary injury, and then a continuous infusion of 5 mL/kg/h of Lactated Ringer was maintained until the end of the study. Monitoring with continuous electrocardiography, oxymetry, and blood pressures was done with a multiparametric monitor (Dixtal-Philips DX 2020, São Paulo, Brazil). The left femoral artery was cannulated for blood pressure monitoring and blood sampling. The right internal jugular vein was cannulated with a 9-French introducer sheath (Arrow, Reading, PA, USA) through which a pulmonary artery catheter (Edwards Lifesciences, Irvine, CA, USA) was introduced for monitoring of the mean pulmonary artery pressure (PAPm), cardiac output, central venous pressure (CVP), and mixed venous blood gases (SvO2). A central venous catheter was introduced in the left internal jugular vein. A surgical cystostomy was done to quantify the urine output. The animal was connected to the NICO2 device (Novametrix Medical Systems, Wallingford, CT, USA) for airway end-tidal pressure of carbon dioxide (EtCO2), tidal volume, airway pressures, and airway flow monitoring.

introduced in the left internal jugular vein. A surgical cystostomy was done to quantify the urine output. The animal was connected to the NICO2 device (Novametrix Medical Systems, Wallingford, CT, USA) for airway end-tidal pressure of carbon dioxide (EtCO2), tidal volume, airway pressures, and airway flow monitoring. The regional ventilation was monitored with electrical impedance tomography (EIT; Dixtal-Philips, São Paulo, Brazil) [20, 21]. The lungs were split in sternal and vertebral regions of the same height. The amount of ventilation to the regions studied was reported according to the ventilator settings used. Arterial blood gas analyses were done with the ABL 800 device (Radiometer, Copenhagen, Denmark). After the surgical period, the animals were allowed to rest for 60 min prior to the baseline data acquisition. Measurements In all the steps of the study, the following data were collected:Hemodynamic: heart rate, cardiac output, CVP, mean systemic arterial blood pressure (ABPm), PAPm, pulmonary artery occlusion pressure (PAOP), SvO2, and norepinephrine use and dosage Respiratory: arterial partial pressure of oxygen (PaO2), PaCO2, EtCO2, VT, airway peak pressure (Ppeak), airway plateau pressure (Pplateau) through expiratory valve occlusion after 2 s of inspiratory pause, intrinsic positive end-expiratory pressure (PEEPi) through expiratory valve occlusion after 4 s of expiratory pause, extrinsic positive end-expiratory pressure (PEEPe), mean airway pressure (Pmean), inspiratory flow, inspiratory time (Tinsp), and ventilatory distribution EIT data

clusion after 2 s of inspiratory pause, intrinsic positive end-expiratory pressure (PEEPi) through expiratory valve occlusion after 4 s of expiratory pause, extrinsic positive end-expiratory pressure (PEEPe), mean airway pressure (Pmean), inspiratory flow, inspiratory time (Tinsp), and ventilatory distribution EIT data Metabolic: pH, lactate, temperature, and fluid balance Calculated variables To obtain the calculated variables, we used the following formulas: Cardiac index (CI) = Cardiac output/Weight Systemic vascular resistance index = (ABPm − CVP) × 80/CI Pulmonary vascular resistance index = (PAPm − PAOP) × 80/CI Blood oxygen content (C × O2) = P × O2 × 0.0031 + 1.36 × Hb × S × O2 Minute ventilation = VT × RR PEEPtotal = PEEPi + PEEPe Alveolar oxygen partial pressure (PAO2) = 643 × FiO2/100 − (PaCO2/0.8) Alveolar-arterial oxygen [(A-a)O2] gradient = PAO2 − PaO2 Pulmonary capillary oxygen content (CcO2) = PAO2 × 0.0031 + 1.36 × Hb Pulmonary shunt = (CcO2 − CaO2) × 100/(CcO2 − CvO2) Static compliance (Cstatic) = VT/(Pplateau − PEEPtotal) Dynamic compliance (Cdyn) = VT/(Ppeak − PEEPtotal) Resistance (Ppeak − Pplateau)/Inspiratory flow

Alveolar oxygen partial pressure (PAO2) = 643 × FiO2/100 − (PaCO2/0.8) Alveolar-arterial oxygen [(A-a)O2] gradient = PAO2 − PaO2 Pulmonary capillary oxygen content (CcO2) = PAO2 × 0.0031 + 1.36 × Hb Pulmonary shunt = (CcO2 − CaO2) × 100/(CcO2 − CvO2) Static compliance (Cstatic) = VT/(Pplateau − PEEPtotal) Dynamic compliance (Cdyn) = VT/(Ppeak − PEEPtotal) Resistance (Ppeak − Pplateau)/Inspiratory flow ARDS induction After the baseline data collection, ARDS was induced with repeated whole-lung lavage using 1 L of isotonic saline (37°C) until the PaO2 was below 100 mmHg for at least 10 min. Lung injurious ventilation was then started with the animal ventilated in pressure control mode with PEEP = 3 cmH2O, FiO2 = 1, inspiratory/expiratory time ratio (I/E) = 1:1, Ppeak = 42 cmH2O, and a RR of 20 to 30 breaths/min [22]. Arterial blood gases were obtained every 15 min, and the PEEP could be increased up to 19 cmH2O targeting a PaO2 level between 55 and 80 mmHg, whereas the inspiratory pressure was limited at 48 cmH2O. The injurious ventilation was maintained until one of the following parameters was reached: An interval of 240 min of injurious ventilation A PAPm > 50 mmHg A Cstatic < 10 mL/cmH2O (with a PEEP = 10 cmH2O and VT = 6 mL/kg) A PEEP persistently ≥ 15 cmH2O for at least two consecutive arterial blood sample analyses An ABPm < 70 mmHg in spite of the use of norepinephrine in a dosage higher than 0.5 μg/kg/min

ARDS induction After the baseline data collection, ARDS was induced with repeated whole-lung lavage using 1 L of isotonic saline (37°C) until the PaO2 was below 100 mmHg for at least 10 min. Lung injurious ventilation was then started with the animal ventilated in pressure control mode with PEEP = 3 cmH2O, FiO2 = 1, inspiratory/expiratory time ratio (I/E) = 1:1, Ppeak = 42 cmH2O, and a RR of 20 to 30 breaths/min [22]. Arterial blood gases were obtained every 15 min, and the PEEP could be increased up to 19 cmH2O targeting a PaO2 level between 55 and 80 mmHg, whereas the inspiratory pressure was limited at 48 cmH2O. The injurious ventilation was maintained until one of the following parameters was reached: An interval of 240 min of injurious ventilation A PAPm > 50 mmHg A Cstatic < 10 mL/cmH2O (with a PEEP = 10 cmH2O and VT = 6 mL/kg) A PEEP persistently ≥ 15 cmH2O for at least two consecutive arterial blood sample analyses An ABPm < 70 mmHg in spite of the use of norepinephrine in a dosage higher than 0.5 μg/kg/min After lung injury induction, the stabilization step started. The animal was ventilated according to the recommendations of the interventional group in the "ARMA" study [5] in a volume-controlled mode, with VT = 6 mL/kg, RR = 35 breaths/min (the maximal respiratory rate allowed by the protocol - because of hypercapnia and acidosis), initial PEEP = 10 cmH2O (mean value used in the ARMA study), and FiO2 = 1.

ccording to the recommendations of the interventional group in the "ARMA" study [5] in a volume-controlled mode, with VT = 6 mL/kg, RR = 35 breaths/min (the maximal respiratory rate allowed by the protocol - because of hypercapnia and acidosis), initial PEEP = 10 cmH2O (mean value used in the ARMA study), and FiO2 = 1. An arterial blood sample was obtained every 10 min. Subsequently, PEEP and FiO2 were titrated according to the ARMA study PEEP table (aiming at a PaO2 = 55 to 80 mmHg) [5]. VT and RR were kept constant during the stabilization step with no attempt to correct the PaCO2 level. Experimental protocol After reaching a PaCO2 equilibrium (variation < 5% in three consecutive arterial blood samples), we considered that the stabilization phase was finished. The same PEEP level titrated at this time was used in the following steps of the study. Four sequences of five different RRs were randomly tested, and these sequences were chosen to alternate higher and lower respiratory frequencies (Figure 1). The five RRs (ventilatory modes) randomized were as follows: RR = 30, 60, 90, 120, 150 breaths/min. At each sequence, VT was adjusted to reach a PaCO2 target of 57 to 63 mmHg.Figure 1 Timeline of the study. PEEPT, total end-expiratory positive pressure (intrinsic end-expiratory positive pressure plus extrinsic end-expiratory positive pressure should be the same as the end-expiratory positive pressure titrated following the ARMA PEEP table); V T, tidal volume; FiO2, inspiratory fraction of oxygen; PaCO2, partial arterial carbon dioxide pressure.

tory positive pressure (intrinsic end-expiratory positive pressure plus extrinsic end-expiratory positive pressure should be the same as the end-expiratory positive pressure titrated following the ARMA PEEP table); V T, tidal volume; FiO2, inspiratory fraction of oxygen; PaCO2, partial arterial carbon dioxide pressure. The randomization was done using sealed envelopes containing the proportion of 1:1:1:1 of the following RR sequences: Sequence 1 (60, 150, 90, 120, 30) Sequence 2 (90, 30, 120, 60, 150) Sequence 3 (120, 150, 90, 30, 60) Sequence 4 (150, 90, 30, 120, 60) During this part of the protocol, the animals were ventilated using volume control ventilation, with FiO2 = 1 and square inspiratory flow = 1 L/s. At each step, PEEPi was measured every 10 min, and PEEPe was corrected in order to keep the PEEPtotal equal to the PEEP obtained during the equilibrium step using the ARMA PEEP table. After completion of these randomized sequences, the animals were submitted to HFOV (Figure 1) at 5 Hz (SensorMedics 3100B, Yorba Linda, CA, USA) with a Pmean set at 30 cmH2O, I/E = 1:2, bias flow = 30 L/min, and the initial pressure amplitude = 80 cmH2O [9]. Unlike the other five RRs, of which the sequence was randomized, HFOV was always performed last, because of its higher Pmean, which could induce lung recruitment.

(SensorMedics 3100B, Yorba Linda, CA, USA) with a Pmean set at 30 cmH2O, I/E = 1:2, bias flow = 30 L/min, and the initial pressure amplitude = 80 cmH2O [9]. Unlike the other five RRs, of which the sequence was randomized, HFOV was always performed last, because of its higher Pmean, which could induce lung recruitment. An arterial blood sample was obtained each 10 min throughout the remainder of the study. After VT changes or after pressure amplitude changes during HFOV, we waited until there were three consecutive measurements with the PaCO2 levels stable between 57 and 63 mmHg. Data were then collected, and the next step of the sequence was started. Between consecutive steps, a 40-s disconnection from the ventilator was done in order to avoid the carryover of the time-dependent alveolar recruitment. At the end of the experiments, the anesthesia was deepened with propofol overdose, and the animals were euthanized with a bolus of 10 mL of potassium chloride 19.1%.

. Between consecutive steps, a 40-s disconnection from the ventilator was done in order to avoid the carryover of the time-dependent alveolar recruitment. At the end of the experiments, the anesthesia was deepened with propofol overdose, and the animals were euthanized with a bolus of 10 mL of potassium chloride 19.1%. Statistical analysis The Shapiro-Wilk goodness-of-fit model showed a non-parametrical distribution for most variables; therefore, data are reported as median [P25th,P75th]. Wilcoxon’s signed rank test was used to test variables before and after lung injury induction and to compare the upper and lower regional ventilation with the EIT. In order to avoid type I error, a modified Bonferroni’s correction was used to account for the multiple comparisons between upper and lower regions of ventilation. Therefore, the p value considered significant was 0.007 when comparing upper and lower regions during the various frequencies studied and 0.012 when comparing the effects of inspiratory pauses and the alveolar recruitment with a RR of 60 breaths/min. The analysis of variance for repeated measures on ranks (Friedman’s test) was used for analyses during the ventilatory modes tested. The post hoc analyses were done using Student-Newman-Keuls’ test. A p < 0.05 was considered significant. The analyses and graphs were done with the SigmaPlot 12.0 statistical package software (Systat Software, Inc. San Jose, CA, USA).

on ranks (Friedman’s test) was used for analyses during the ventilatory modes tested. The post hoc analyses were done using Student-Newman-Keuls’ test. A p < 0.05 was considered significant. The analyses and graphs were done with the SigmaPlot 12.0 statistical package software (Systat Software, Inc. San Jose, CA, USA). Results Eight pigs weighing 34 [29,39] kg were used. ARDS was induced using 10 [7,16] L of normal saline followed by injurious mechanical ventilation for 210 [40,225] min. The respiratory variables at baseline and after the induction of lung injury are shown in Table 1.Table 1 Respiratory variables at baseline and after the induction of lung injury Variable Baseline After lung injury P/F ratio (mmHg) 427 [368,473] 97 [67,130]* Shunt (%) 13 [12,15] 23 [16,32]* Tidal volume (sternal) 4.8 [3.7,5.9] 4 [4.5,3.5] Tidal volume (ventral) 4.2 [3.0,5.4] 2.5 [2,3] C static (mL/cmH2O) 27 [15,30] 12 [9,14] Resistance (cmH2O/L/s) 8 [7,10] 18 [14,26]* Values are presented as median [P25th,P75th] C static and P/F denote static compliance and the ratio of arterial oxygen concentration to the fraction of inspired oxygen, respectively. *p < 0.05 vs baseline, Wilcoxon’s signed rank test.

C static (mL/cmH2O) 27 [15,30] 12 [9,14] Resistance (cmH2O/L/s) 8 [7,10] 18 [14,26]* Values are presented as median [P25th,P75th] C static and P/F denote static compliance and the ratio of arterial oxygen concentration to the fraction of inspired oxygen, respectively. *p < 0.05 vs baseline, Wilcoxon’s signed rank test. The FIO2 during the stabilization step was 0.7 [0.5,0.9]. The time to PaCO2 equilibrium was similar in the different phases of the experiment and equal to 50 [40,75] min. The most important respiratory data with different RRs are shown in Figure 2 and Table 2. During the stabilization step, PaCO2 was 81 [78,92] mmHg. In all other experimental phases, the PaCO2 was kept in the planned range of 57 to 63 mmHg (Figure 2A). VT could be progressively reduced with increasing RRs (Figure 2B), as did regional ventilation (Figure 3). The ventilation to the dependent parts of the lung reduced to a greater extent leading to an increase in the sternal/vertebral ratio of regional ventilation (Figure 4). Additionally, low values of plateau and driving pressures were maintained at all RRs (Figure 2C,D, respectively). The HFOV led to the highest oxygenation, the lowest VT, and the most homogeneous distribution of ventilation (Table 2, Figures 2B and 4, respectively).Figure 2 Respiratory variables during the ventilatory modes tested. (A) PaCO2 (mmHg; Friedman’s test, p = 0.011). (B) Tidal volume (mL/kg; Friedman’s test, p < 0.001). (C) Plateau pressure (cmH2O; Friedman’s test, p < 0.001). (D) Driving pressure (cmH2O; Friedman’s test, p < 0.001). V T, RR, HFPPV, and HFOV denote tidal volume, respiratory rate, high-frequency positive-pressure ventilation, and high-frequency oscillatory ventilation, respectively. The whiskers denote the P10th and P90th. *Student-Newman-Keuls’ post hoc analysis, p < 0.05 vs stabilization step (V T = 6 mL/kg and RR = 35 breaths/min); #Student-Newman-Keuls’ post hoc analysis, p < 0.05 vs HFPPV = 150.

frequency positive-pressure ventilation, and high-frequency oscillatory ventilation, respectively. The whiskers denote the P10th and P90th. *Student-Newman-Keuls’ post hoc analysis, p < 0.05 vs stabilization step (V T = 6 mL/kg and RR = 35 breaths/min); #Student-Newman-Keuls’ post hoc analysis, p < 0.05 vs HFPPV = 150. Table 2 Respiratory variables through the ventilatory modes tested Variable V T = 6 mL/kg RR = 30 HFPPV = 60 HFPPV = 90 HFPPV = 120 HFPPV = 150 HFOV p value a P/F ratio (mmHg) 95 [87,105] 151 [117,181]b 141 [102,189]b 132 [95,169]b 111 [86,162]b 112 [90,171]b 193 [146,216]b,c P = 0.003 Gradient (A-a)O2 480 [465,493] 396 [383,452]b 427 [378,468]b 427 [394,466] 455 [406,481] 458 [394,478] 365 [350,420]b,c P = 0.014 Minute ventilation (L/min) 6.9 [6.6,8.8] 8.7 [7.4,10.0] 11.1 [11.0,11.5]b 15.6 [14.0,17.5]b 18.6 [13.0,19.2]b 20.2 [19.5,21.3]b 27 [23.5,28.4]b,c P < 0.001 Shunt (%) 43 [41,45] 29 [26,34]b 34 [30,40] 31 [28,44] 34 [31,45] 38 [30,43] 27 [25,32]b,c P = 0.003 EtCO2 (mmHg) 58 [52,60] 43 [32,47]b 41 [37,49]b 40 [30,45]b 40 [31,48]b 34 [28,36]b 27 [23,30]b,c P < 0.001 C static (mL/cmH2O) 12 [10,14] 12 [9,14] 10 [9,13] 10 [9,12]b 10 [7,11]b 9 [8,11]b - P = 0.001

P/F ratio (mmHg) 95 [87,105] 151 [117,181]b 141 [102,189]b 132 [95,169]b 111 [86,162]b 112 [90,171]b 193 [146,216]b,c P = 0.003 Gradient (A-a)O2 480 [465,493] 396 [383,452]b 427 [378,468]b 427 [394,466] 455 [406,481] 458 [394,478] 365 [350,420]b,c P = 0.014 Minute ventilation (L/min) 6.9 [6.6,8.8] 8.7 [7.4,10.0] 11.1 [11.0,11.5]b 15.6 [14.0,17.5]b 18.6 [13.0,19.2]b 20.2 [19.5,21.3]b 27 [23.5,28.4]b,c P < 0.001 Shunt (%) 43 [41,45] 29 [26,34]b 34 [30,40] 31 [28,44] 34 [31,45] 38 [30,43] 27 [25,32]b,c P = 0.003 EtCO2 (mmHg) 58 [52,60] 43 [32,47]b 41 [37,49]b 40 [30,45]b 40 [31,48]b 34 [28,36]b 27 [23,30]b,c P < 0.001 C static (mL/cmH2O) 12 [10,14] 12 [9,14] 10 [9,13] 10 [9,12]b 10 [7,11]b 9 [8,11]b - P = 0.001 C dyn (mL/cmH2O) 8 [7,9] 9 [6,10] 7 [6,9] 7 [6,8] 6 [5,7]b 6 [5,7]b - P < 0.001 Resistance (cmH2O/L/s) 8 [8,10] 10 [9,12]b 9 [8,9]b 9 [8,9]b 8 [8,11]b 9 [8,13]b 17 [13,20]b,c P < 0.001 PEEP total (cmH2O) 14 [11,17] 14 [10,17] 14 [10,17] 13 [10,16] 13 [10,17] 13 [10,17] - P = 0.744 PEEP intrinsic (cmH2O) 0 0 0 0 [0,1] 0 [0,1] 2 [1,3]b - P < 0.001 PEEP extrinsic (cmH2O) 14 [11,16] 14 [11,16] 13 [10,17] 13 [10,16] 13 [10,16] 12 [9,14] - P < 0.001 Peak pressure (cmH2O) 45 [44,48] 54 [47,58]b 44 [42,47] 44 [41,45] 41 [38,44] 41 [38,43] 59 [51,79]b,c P < 0.001 P mean (cmH2O) 17 [15,20] 18 [16,22]b 18 [15,22]b 20 [18,23]b 20 [18,24]b 20 [17,23]b 29 [28,30]b,c P < 0.001 Inspiratory flow (L/s) 1 1 1 1 1 1 - P = 1.000

C dyn (mL/cmH2O) 8 [7,9] 9 [6,10] 7 [6,9] 7 [6,8] 6 [5,7]b 6 [5,7]b - P < 0.001 Resistance (cmH2O/L/s) 8 [8,10] 10 [9,12]b 9 [8,9]b 9 [8,9]b 8 [8,11]b 9 [8,13]b 17 [13,20]b,c P < 0.001 PEEP total (cmH2O) 14 [11,17] 14 [10,17] 14 [10,17] 13 [10,16] 13 [10,17] 13 [10,17] - P = 0.744 PEEP intrinsic (cmH2O) 0 0 0 0 [0,1] 0 [0,1] 2 [1,3]b - P < 0.001 PEEP extrinsic (cmH2O) 14 [11,16] 14 [11,16] 13 [10,17] 13 [10,16] 13 [10,16] 12 [9,14] - P < 0.001 Peak pressure (cmH2O) 45 [44,48] 54 [47,58]b 44 [42,47] 44 [41,45] 41 [38,44] 41 [38,43] 59 [51,79]b,c P < 0.001 P mean (cmH2O) 17 [15,20] 18 [16,22]b 18 [15,22]b 20 [18,23]b 20 [18,24]b 20 [17,23]b 29 [28,30]b,c P < 0.001 Inspiratory flow (L/s) 1 1 1 1 1 1 - P = 1.000 T insp/T tot (%) 15 [14,17] 19 [14,22] 24 [20,28]b 34 [30,37]b 38 [33,44]b 42 [37,49]b 56 [50,67]b,c P < 0.001 Values are presented as median [P25th,P75th]. aThe p value was obtained through Friedman’s test; bStudent-Newman-Keuls’ post hoc analysis, p < 0.05 vs V T = stabilization step (V T = 6 mL/kg and RR = 35 breaths/min); cStudent-Newman-Keuls’ post hoc analysis, p < 0.05 vs HFPPV = 150. Figure 3 Regional ventilation (mL/kg) in the sternal and vertebral portions of the thorax measured through EIT. ©Student-Newman-Keuls’ post hoc analysis, p < 0.05 vs stabilization step (V T = 6 mL/kg and RR = 35 breaths/min) (Friedman’s test, p < 0.001); #Student-Newman-Keuls’ post hoc analysis, p < 0.05 vs HFPPV = 150 (Friedman’s test, p < 0.001); §Wilcoxon’s test, p < 0.007 (Bonferroni’s correction for multiple comparisons) vs the vertebral region (gravitational-dependent region).

tabilization step (V T = 6 mL/kg and RR = 35 breaths/min) (Friedman’s test, p < 0.001); #Student-Newman-Keuls’ post hoc analysis, p < 0.05 vs HFPPV = 150 (Friedman’s test, p < 0.001); §Wilcoxon’s test, p < 0.007 (Bonferroni’s correction for multiple comparisons) vs the vertebral region (gravitational-dependent region). Figure 4 Distribution of regional ventilation (%) in the sternal and vertebral portions of the thorax measured through EIT. ©Student-Newman-Keuls’ post hoc analysis, p < 0.05 vs stabilization step (V T = 6 mL/kg and RR = 35 breaths/min) (Friedman’s test, p < 0.001); #Student-Newman-Keuls’ post hoc analysis, p < 0.05 vs HFPPV = 150 (Friedman’s test, p < 0.001); §Wilcoxon’s test, p < 0.007 (Bonferroni’s correction for multiple comparisons) vs the vertebral region (gravitational-dependent region). Only one animal needed norepinephrine during HFPPV, and the dose varied between 2.4 μg/kg/min (HFPPV = 60) and 3.2 μg/kg/min (HFOV). The hemodynamic and metabolic data with different RRs are shown in Table 3. Of note, the stabilization step with VT = 6 mL/kg and RR = 35 breaths/min was associated with higher pulmonary artery pressures and lower pH.Table 3 Hemodynamic and metabolic variables during the ventilatory modes studied Variable V T = 6 mL/kg RR = 30 HFPPV = 60 HFPPV = 90 HFPPV = 120 HFPPV = 150 HFOV p value a Hemodynamic

Only one animal needed norepinephrine during HFPPV, and the dose varied between 2.4 μg/kg/min (HFPPV = 60) and 3.2 μg/kg/min (HFOV). The hemodynamic and metabolic data with different RRs are shown in Table 3. Of note, the stabilization step with VT = 6 mL/kg and RR = 35 breaths/min was associated with higher pulmonary artery pressures and lower pH.Table 3 Hemodynamic and metabolic variables during the ventilatory modes studied Variable V T = 6 mL/kg RR = 30 HFPPV = 60 HFPPV = 90 HFPPV = 120 HFPPV = 150 HFOV p value a Hemodynamic Heart rate (bpm) 144 [125,165] 165 [124,182] 173 [144,181] 169 [130,189] 164 [128,196] 173 [142,196] 145 [122,155] P = 0.210 Cardiac index (mL/kg/min) 138 [128,153] 126 [121,145] 145 [120,169] 127 [115,158] 141 [118,166] 132 [116,168] 126 [101,142] P = 0.363 SV (mL) 28 [26,42] 27 [24,41] 27 [26,35] 26 [23,44] 27 [26,34] 30 [26,35] 31 [22,44] P = 0.916 ABPm (mmHg) 90 [75,107] 86 [75,112] 84 [72,97] 91 [77,100] 83 [70,112] 78 [69,105] 82 [72,98] P = 0.320 PAPm (mmHg) 43 [38,52] 34 [31,37]b 34 [28,36]b 36 [33,37]b 33 [30,47]b 38 [30,43]b 31 [30,40]b P = 0.018 CVP (mmHg) 9 [9,12] 8 [7,12] 8 [6,10] 9 [6,10] 8 [7,10] 9 [7,10] 11 [10,12]c P = 0.017 PAOP (mmHg) 12 [11,15] 12 [11,15] 12 [9,14] 12 [10,14] 12 [10,15] 12 [10,15] 14 [12,17] P = 0.042 SvO2 (mmHg) 54 [47,70] 70 [49,79] 68 [48,71] 63 [48,66] 64 [43,74] 65 [55,73] 65 [50,75] P = 0.140 SVRI (dynes.s-1(cm5)-1.kg-1 51.8 [41.8,56.4] 47.1 [39.8,65.6] 47.0 [35.0,50.9] 52.7 [36.7,61.4] 41.7 [33.4,63.6] 42.4 [28.8,66.5] 50.4 [31.6,55.8] P = 0.558 PVRI (dynes.s-1.(cm5)-1.kg-1 22.3 [17.5,25.7] 15.7 [13.0,16.8] 15.6 [10.9,17.7] 16.7 [13.9,17.7] 14.2 [12.4,20.8] 16.8 [12.0,20.7] 13.6 [10.9,18.2] P = 0.133

0.140 SVRI (dynes.s-1(cm5)-1.kg-1 51.8 [41.8,56.4] 47.1 [39.8,65.6] 47.0 [35.0,50.9] 52.7 [36.7,61.4] 41.7 [33.4,63.6] 42.4 [28.8,66.5] 50.4 [31.6,55.8] P = 0.558 PVRI (dynes.s-1.(cm5)-1.kg-1 22.3 [17.5,25.7] 15.7 [13.0,16.8] 15.6 [10.9,17.7] 16.7 [13.9,17.7] 14.2 [12.4,20.8] 16.8 [12.0,20.7] 13.6 [10.9,18.2] P = 0.133 Metabolic Lactate (mEq/L) 1.7 [1.1,2.1] 1.3 [0.8,1.7] 1.6 [0.8,2.0] 1.7 [1.1,2.0] 1.4 [1.0,1.9] 1.6 [0.9,2.3] 1.5 [1.1,1.8] P = 0.762 pH 7.13 [7.08,7.2] 7.25 [7.24,7.33]b 7.25 [7.24,7.35]b 7.26 [7.23,7.33]b 7.27 [7.2,7.3]b 7.26 [7.21,7.34]b 7.25 [7.2,7.32]b P = 0.002 Temperature (°C) 38.6 [37.3,39.2] 39.7 [38.0,39.8]b 39.4 [37.6,39.6]b 39.2 [38.1,39.6]b 38.8 [37.6,39.6]b 39.0 [37.8,39.5]b 39.2 [38.1,39.8]b P = 0.007 Fluid balance (mL) −50 [−242,−5] 170 [101,278] 40 [−22,102] 100 [67,110] 60 [5,108] 40 [−58,88] 30 [−21,50] P = 0.044 Values are presented as median [P25th,P75th]. aThe p value was obtained through Friedman’s test; bStudent-Newman-Keuls’ post hoc analysis, p < 0.05 vs V T = stabilization step (V T = 6 mL/kg and RR = 35 breaths/min); cStudent-Newman-Keuls’ post hoc analysis, p < 0.05 vs HFPPV = 150.

8] 40 [−58,88] 30 [−21,50] P = 0.044 Values are presented as median [P25th,P75th]. aThe p value was obtained through Friedman’s test; bStudent-Newman-Keuls’ post hoc analysis, p < 0.05 vs V T = stabilization step (V T = 6 mL/kg and RR = 35 breaths/min); cStudent-Newman-Keuls’ post hoc analysis, p < 0.05 vs HFPPV = 150. Discussion Our main finding was that, during protective mechanical ventilation of a severe ARDS swine model, the use of HFPPV with a conventional ventilator allows further reductions in VT and PaCO2, leading to reductions in driving pressures and plateau pressures without increasing mean airway pressure. We did not identify any significant detrimental effect of the high RRs applied, even after careful assessment of hemodynamics, respiratory system mechanics, and gas exchange. The possibility of further reducing the ventilator-associated lung injury is of utmost importance, with possible implications in terms of reducing death and multiple organ failure in ARDS patients [23]. Ventilation with low VTs (6 mL/kg) is still the standard support for those patients [5], although lower VTs might produce additional protection [6, 24]. Of note, one third of ARDS patients under protective ventilation still have lung hyperdistention, which is associated with increases in systemic inflammatory markers [25]. This subset of patients, usually more severely injured, could possibly benefit from further VT reductions [24].

oduce additional protection [6, 24]. Of note, one third of ARDS patients under protective ventilation still have lung hyperdistention, which is associated with increases in systemic inflammatory markers [25]. This subset of patients, usually more severely injured, could possibly benefit from further VT reductions [24]. Increasing the RR at constant alveolar ventilation, we obtained a progressive decrease in VTs reaching levels below 4 mL/kg. This finding challenges the paradigm - promulgated by the design of many clinical trials that RRs should be kept equal to or less than 35 breaths per minute [5, 26–28]. In our model of severe ARDS, the standard of care [5] settings of VTs at 6 mL/kg and a maximum RR of 35 breaths/min led to a median PaCO2 value of 81 mmHg with a median pH of 7.13. Targeting a PaCO2 of 60 mmHg, we were able to reduce VTs by 36% with a RR of 150 breaths per minute. Other authors have shown, in an experimental model of ARDS, that higher RRs allow for a reduction in VT when associated with a strategy to lower the dead space (aspiration of dead space) [29, 30]. Similarly, a recent study in patients with ARDS showed that protective VT around 4 mL/kg can be achieved with modest increments in RR, provided that care is taken to minimize the circuit dead space [31]. These studies combined increases in RR with other measures to decrease the dead space. Our findings on the isolated effect of RR on the reduction of tidal volume help understand the independent effect of manipulating the RR.

h modest increments in RR, provided that care is taken to minimize the circuit dead space [31]. These studies combined increases in RR with other measures to decrease the dead space. Our findings on the isolated effect of RR on the reduction of tidal volume help understand the independent effect of manipulating the RR. The increases in RR were not associated with significant changes in gas exchange. We did notice a not significant but progressive fall in the median PaO2/FiO2 (P/F) ratio with increases in RR above 30 breaths per minute amounting to a fall of 26% at a RR of 150 breaths per minute (Table 2). Concurrently, the Tinsp/Ttot ratio increased from 19% to 42% when RR increased from 30 to 150 (Table 2), due to the fixed inspiratory flow rate and the need for higher minute ventilation at high RR. These increases in the Tinsp/Ttot ratio would favor a change in the P/F ratio in the opposite direction of the trend we found. These observations emphasize that with our relatively small sample size, we might have been underpowered to detect some differences such as the P/F ratio variation. If such trend proved significant in a larger study, it is possible that the lower tidal volumes at higher RR have favored the development of absorption atelectasis, although we cannot exclude that hemodynamic factors may played a role.

t have been underpowered to detect some differences such as the P/F ratio variation. If such trend proved significant in a larger study, it is possible that the lower tidal volumes at higher RR have favored the development of absorption atelectasis, although we cannot exclude that hemodynamic factors may played a role. HFOV, a more classical strategy than HFPPV to provide adequate gas exchange at very low VTs [32, 33], has been recently shown to provide no benefit or even cause harm to patients with ARDS [16, 17]. Our results showed that HFOV = 5 Hz could stabilize the PaCO2 with VTs 26% lower than HFPPV = 150, however, with a RR twice as high and a Pmean 30% higher [15]. This is illustrative of the disproportionate increases in RR to maintain alveolar ventilation at progressively lower VTs, especially when close to the dead space, and the need to increase Pmean, which may have deleterious hemodynamic effects. The consequence of this ventilation inefficiency might be an increased dissipation of energy in the lungs, potentially leading to more lung injury even at reduced stress and strain per breath. Therefore, reducing VT without increasing mean airway pressure might be of special interest. In that sense, HFPPV might offer a better compromise between VT and RR than HFOV.

y might be an increased dissipation of energy in the lungs, potentially leading to more lung injury even at reduced stress and strain per breath. Therefore, reducing VT without increasing mean airway pressure might be of special interest. In that sense, HFPPV might offer a better compromise between VT and RR than HFOV. Ventilation decreased more in the gravitation-dependent regions, a finding suggestive of reabsorption atelectasis, air trapping, or incomplete filling of those regions due to airway narrowing. Even after taking this ‘functional baby lung’ into account, the net result was likely a lesser degree of tidal lung stretch as suggested by the decrease in driving pressures and plateau pressures. Additionally, despite a preferential reduction in dependent ventilation (Figure 4), HFPPV could result in lower regional VT in non-dependent regions (Figure 3).

by lung’ into account, the net result was likely a lesser degree of tidal lung stretch as suggested by the decrease in driving pressures and plateau pressures. Additionally, despite a preferential reduction in dependent ventilation (Figure 4), HFPPV could result in lower regional VT in non-dependent regions (Figure 3). Limitations Our study has several limitations. First, the arbitrary choice of the target CO2 level during HFPPV can be criticized. The CO2 value can be a confounding factor of the ventilatory settings during ARDS ventilation, with some studies showing a protective [34, 35] and others a potentially deleterious role [36, 37]. We chose a narrow range of 57 to 63 mmHg to avoid such potential confounding effect and to avoid significant acidosis (pH < 7.15), a goal we achieved in all experimental conditions. Likely, the main findings of the study would maintain had a normocapnia target been applied. Second, our study design, with sequential changes in the ventilator settings, was susceptible to carryover phenomena. We tried to avoid that effect through the randomization of sequences, the disconnection from the ventilator between the steps, and through a prolonged wait to the PaCO2 equilibrium. Third, the performance of conventional ventilators declines at very high RRs and low VTs, especially if low-compliance tubing is not employed [38]. Fourth, we did not rule out that histological damage to the lungs might have happened at those very high RRs. Fifth, HFOV was the last step of the study due to logistic issues and at this time the animals had significant positive fluid balances. This could be one explanation why HFOV was not associated with hemodynamic alterations, even with the use of higher Pmean. Finally, we cannot directly extrapolate these experimental findings to patients, who have longer time constants than pigs and might not tolerate RRs as high. Interestingly, those with the most severe lung injury tolerate better very high RR, because of their low time constants. Even so, in our experience, it is difficult to apply RR > 60 breaths per minute to patients without leading to intrinsic PEEP.

have longer time constants than pigs and might not tolerate RRs as high. Interestingly, those with the most severe lung injury tolerate better very high RR, because of their low time constants. Even so, in our experience, it is difficult to apply RR > 60 breaths per minute to patients without leading to intrinsic PEEP. Conclusions In an animal model of severe ARDS, as compared to the standard protective ventilation, high-frequency positive-pressure ventilation delivered by a conventional ventilator allowed further reductions in tidal volume and in inspiratory pressures. As such, HFPPV could be a well-suited alternative in the treatment of severe ARDS with very low lung compliance, although its impact on lung inflammation still awaits evaluation. Abbreviations ABPmmean systemic arterial blood pressure ARDSacute respiratory distress syndrome CcO2pulmonary capillary oxygen content Cdyndynamic compliance CIcardiac index Cstaticstatic compliance CVPcentral venous pressure CxO2blood oxygen content EITelectrical impedance tomography EtCO2airway end-tidal pressure of carbon dioxide FiO2inspiratory fraction of oxygen HFOVhigh-frequency oscillatory ventilation HFPPVhigh-frequency positive-pressure ventilation I/Einspiratory/expiratory time ratio PaCO2partial arterial carbon dioxide pressure PaO2partial pressure of arterial oxygen PAOPpulmonary artery occlusion pressure PAPmmean pulmonary artery pressure PEEPpositive end-expiratory pressure PEEPeextrinsic positive end-expiratory pressure PEEPiintrinsic positive end-expiratory pressure Pmeanmean airway pressure Ppeakairway peak pressure Pplateauairway plateau pressure RRrespiratory rate

PaO2partial pressure of arterial oxygen PAOPpulmonary artery occlusion pressure PAPmmean pulmonary artery pressure PEEPpositive end-expiratory pressure PEEPeextrinsic positive end-expiratory pressure PEEPiintrinsic positive end-expiratory pressure Pmeanmean airway pressure Ppeakairway peak pressure Pplateauairway plateau pressure RRrespiratory rate SvO2mixed venous blood gases Tinspinspiratory time VTtidal volumes. Competing interests The authors declare that they have no competing interests. Authors’ contributions RLC designed the study, participated in the research protocols, analyzed and interpreted the data, and drafted the manuscript. MP designed the study, participated in the research protocols, helped interpret the data, and drafted the manuscript. ELVC helped design the study, participated in the research protocols, analyzed and interpreted the data, and helped draft the manuscript design. SG participated in the research protocols and analyzed the data. LB participated in the discussion, helped interpret the data, and drafted the manuscript. MBPA designed the study, analyzed and interpreted the data, and reviewed the manuscript. LCPA designed and conceived the study, participated in the research protocols, and reviewed the manuscript. All authors read and approved the final manuscript.

the discussion, helped interpret the data, and drafted the manuscript. MBPA designed the study, analyzed and interpreted the data, and reviewed the manuscript. LCPA designed and conceived the study, participated in the research protocols, and reviewed the manuscript. All authors read and approved the final manuscript. Acknowledgements The authors are grateful to the Research and Education Institute, Hospital Sírio-Libanês, São Paulo, Brazil; Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP), São Paulo, Brazil; and Financiadora de Estudos e Projetos (FINEP), Brazil, for the grant offered to the study development. We gratefully thank Marcelo do Amaral Beraldo, Mauro Roberto Tucci, Roberta Ribeiro de Santis Santiago, and Takeshi Yoshida, whose contribution to the animal experiments was very important.

Background Ventilator-induced lung injury (VILI) is a well-recognized complication of mechanical ventilation. The mechanical stress caused by a relatively high tidal volume, applied on an injured lung causes an abnormally high distortion (strain) of lung cells. The cellular responses to increased stress and strain result in alveolar barrier disruption and activation of inflammation, therefore inducing or exacerbating acute lung injury (ALI). Several cell types mediate the effects of strain induced by mechanical ventilation. Increased strain, either in vivo or in vitro, impairs barrier properties of alveolar epithelial and endothelial cells [1–3]. Moreover, both alveolar epithelial and endothelial cells have been shown to produce pro-inflammatory cytokines when subjected to deformation [4, 5]. Alveolar macrophages are also activated by cyclic stretch [6], and have been shown to mediate not only inflammation, but also barrier dysfunction [6–8].

elial cells [1–3]. Moreover, both alveolar epithelial and endothelial cells have been shown to produce pro-inflammatory cytokines when subjected to deformation [4, 5]. Alveolar macrophages are also activated by cyclic stretch [6], and have been shown to mediate not only inflammation, but also barrier dysfunction [6–8]. The extracellular signal-regulated kinase 1/2 (ERK1/2) is a member of the mitogen-activated serine/threonine kinase (MAPK) family, which also includes the kinases p38 and C-Jun N-terminal Kinase (JNK). MAPK are highly conserved enzymes regulating a vast array of cellular functions, including cell survival, proliferation, and differentiation, as well as inflammation and stress responses [9, 10]. Several cells respond to mechanical forces by activating MAPK pathways [11]. Mechanical stress has been shown to stimulate ERK1/2 in pulmonary epithelial cells [12] and endothelial cells [13]. ERK1/2 activation upon cyclic stretch has been reported in human bronchial epithelial cells [4] and primary rat alveolar epithelial cells [14]. In an in vitro model of VILI, cyclic stretch-induced interleukin-8 production from lung epithelial cells was reduced by ERK1/2 inhibition [4]. In vivo, bleomycin plus ventilation-induced lung fibrosis was attenuated in mice with pharmacologic inhibition of ERK1/2 [15].

4] and primary rat alveolar epithelial cells [14]. In an in vitro model of VILI, cyclic stretch-induced interleukin-8 production from lung epithelial cells was reduced by ERK1/2 inhibition [4]. In vivo, bleomycin plus ventilation-induced lung fibrosis was attenuated in mice with pharmacologic inhibition of ERK1/2 [15]. Given the important role of ERK1/2 in cell proliferation, ERK1/2 inhibitors have been investigated as candidate targets in cancer and rheumatoid arthritis [16–18]. Nevertheless, defective ERK1/2 signaling has been associated with autoimmunity [19] raising concerns for the use of ERK1/2 inhibition in inflammatory diseases, emphasizing the need for more selective therapeutic targets [18].

bitors have been investigated as candidate targets in cancer and rheumatoid arthritis [16–18]. Nevertheless, defective ERK1/2 signaling has been associated with autoimmunity [19] raising concerns for the use of ERK1/2 inhibition in inflammatory diseases, emphasizing the need for more selective therapeutic targets [18]. Activation of each of the three MAPKs is controlled by several MAP3Ks ensuring the specificity of signaling responses. Tumor progression locus 2 (Tpl2) is a MAP3kinase that phosphorylates and activates the extracellular signal-regulated kinase (ERK1/2) [20–22]. Tpl2 was originally identified as a proto-oncogene, but is now recognized to play an important role in regulating ERK1/2 signaling in multiple cell types, including T-cells, macrophages, and epithelial cells [22–26]. In vitro studies using Tpl2 overexpressing cells have shown that Tpl2 activates ERK, JNK, p38, and the transcription factors NFAT (nuclear factor of activated T cells) and NF-κB [23, 27]. In vivo, expression of a constitutively active form of Tpl2 under the control of a T cell-specific promoter in mice resulted in development of thymic lymphomas [21]. On the contrary, Tpl2-deficent (Tpl2-/-) mice remain healthy throughout their normal life span [28]. Studies on Tpl2-/-mice have proven that Tpl2 participates in signal transduction of TLR, T-cell receptors, G protein-coupled receptors, tumor necrosis factor (TNF), and CD-40 [25, 28–30]. In these studies, the anti-inflammatory effects of Tpl2 ablation were recognized. Tpl2-/-macrophages produce less inflammatory mediators upon lipopolysacharite (LPS) stimulation [29], and Tpl2-/-mice are resistant to LPS-induced shock [28]. In a model of acute pancreatitis, lung inflammation was less in Tpl2-/-mice [31], and in a model of experimental colitis, Tpl2-/-mice and mice treated with a Tpl2 inhibitor had less bowel inflammation [32].

mediators upon lipopolysacharite (LPS) stimulation [29], and Tpl2-/-mice are resistant to LPS-induced shock [28]. In a model of acute pancreatitis, lung inflammation was less in Tpl2-/-mice [31], and in a model of experimental colitis, Tpl2-/-mice and mice treated with a Tpl2 inhibitor had less bowel inflammation [32]. The present study examined the hypothesis that genetic and pharmacologic inhibition of Tpl2 can ameliorate VILI. We first compared naïve wild type (WT) and Tpl2-/-mice, as well as WT and Tpl2-/-mice ventilated for 4 h with normal tidal volume (VT), to exclude any unexpected effect of Tpl2 deficiency on lung mechanics, response to anesthesia, or normal-VT ventilation. We then subjected WT and Tpl2-/-mice to mechanical ventilation with high VT, under conditions previously shown to induce severe VILI in WT mice [33]. Finally, we examined the effect of pharmacologic inhibition of Tpl2 in WT mice subjected to mechanical ventilation with high VT, when given as a pretreatment, before the initiation of high VT, and, in a more clinically relevant approach, given after initiation of high VT ventilation.

induce severe VILI in WT mice [33]. Finally, we examined the effect of pharmacologic inhibition of Tpl2 in WT mice subjected to mechanical ventilation with high VT, when given as a pretreatment, before the initiation of high VT, and, in a more clinically relevant approach, given after initiation of high VT ventilation. Methods Animal experiments We studied 12 groups of mice, a total of 58 male C57BL6 (WT) and 44 Tpl2-/-mice on C57Bl6 background [28] at 8 to 10 weeks of age (25- to 30-g weight). The experimental groups and number of mice in each group are presented in Table 1, and the experimental procedure is shown in Figure 1. Mice were obtained from the Foundation of Research and Technology Institute Animal Facility. All experiments were approved by the Research Animal Care Committee of University of Crete Medical School and Heraklion Prefecture Veterinary Authority. All experiments were performed at the experimental Intensive Care Medicine laboratory at the Medical School of the University of Crete.Table 1 Experimental groups Strain treatment Sample (N) WT Tpl2 -/-

Methods Animal experiments We studied 12 groups of mice, a total of 58 male C57BL6 (WT) and 44 Tpl2-/-mice on C57Bl6 background [28] at 8 to 10 weeks of age (25- to 30-g weight). The experimental groups and number of mice in each group are presented in Table 1, and the experimental procedure is shown in Figure 1. Mice were obtained from the Foundation of Research and Technology Institute Animal Facility. All experiments were approved by the Research Animal Care Committee of University of Crete Medical School and Heraklion Prefecture Veterinary Authority. All experiments were performed at the experimental Intensive Care Medicine laboratory at the Medical School of the University of Crete.Table 1 Experimental groups Strain treatment Sample (N) WT Tpl2 -/- Control-PV curve and sample collection 10 10 Normal VT ventilation 8 8 High VT ventilation 8 8 High VT ventilation + Tpl2 inhibitor pretreatment 5 High VT ventilation + Tpl2 inhibitor posttreatment 5 Control-no ventilation (ERK1/2 phosphorylation) 6 6 Normal VT 60 min (ERK1/2 phosphorylation) 4 4 Normal VT 30 min + High VT 30 min (ERK1/2 phosphorylation) 8 8 Normal VT 30 min + High VT 30 min (ERK1/2 phosphorylation) + Tpl2 inhibitor pretreatment 4 Figure 1 Experimental procedure. At time 0, animals were anesthetized, tracheostomized, and connected to the ventilator. Through the carotid artery, fluid and anesthesia were administered at a rate of 40 ml/kg/h for 30 min after the initiation of ventilation. At 30 min, fluid infusion rate changed to 15 ml/kg/h in all groups, and ventilator settings were changed as indicated in groups allocated to high VT ventilation. The Tpl2 inhibitor-treated groups received an intraperitoneal injection of Tpl2 inhibitor (thick black arrow), either 5 min prior to the initiation of high VT ventilation (pretreatment) or 30 min after high VT ventilation (posttreatment).

or settings were changed as indicated in groups allocated to high VT ventilation. The Tpl2 inhibitor-treated groups received an intraperitoneal injection of Tpl2 inhibitor (thick black arrow), either 5 min prior to the initiation of high VT ventilation (pretreatment) or 30 min after high VT ventilation (posttreatment). The effects of mechanical ventilation on the development of lung injury were studied in WT and Tpl2-/-mice ventilated with either normal or high tidal volume (VT), as detailed below, and described previously [33]. Mice were anesthetized with intraperitoneal (i.p.) injection of ketamine 100 mcg/g and fentanyl 0.12 mcg/g [33], tracheostomized, and ventilated using SAP830 ventilator (IITC Life Science, Woodland Hills, CA, USA). Ventilation settings for normal VT were VT = 10 ml/kg resulting in peak inspiratory pressure (PIP) = 9 ± 0.5 cmH2O, positive end-expiratory pressure PEEP =1.5 cmH2O, respiratory rate (RR) =130 breaths/min, with recruitment maneuvers performed every 30 min. For high VT, ventilation settings were changed from normal VT after 30 min of hemodynamic stabilization, to high VT = 47 ± 2 ml/kg, targeted to a PIP = 35 ± 0.5 cmH2O, RR = 60 breaths/min, PEEP = 1.5 cmH2O, without recruitment maneuvers. Fraction of inspired oxygen (FiO2) was 30% in all experiments, and inhaled carbon dioxide (CO2) was added in high VT groups to prevent hypocapnia. Arterial blood pressure, PIP, PEEP, and VT were monitored throughout the study. At the end of the 240-min experiment, blood was collected from the arterial line for blood gas analysis, followed by inspiratory pressure volume curve, bronchoalveolar lavage fluid (BALF) and tissue collection. For histological evaluation, lungs from mice not subjected to BALF collection were inflated with 4% paraformaldehyde at a transpulmonary pressure of 25 cmH2O.

ollected from the arterial line for blood gas analysis, followed by inspiratory pressure volume curve, bronchoalveolar lavage fluid (BALF) and tissue collection. For histological evaluation, lungs from mice not subjected to BALF collection were inflated with 4% paraformaldehyde at a transpulmonary pressure of 25 cmH2O. The possible protective effects of pharmacologic Tpl2 inhibition were studied in WT mice ventilated with high VT as described above, and treated with a Tpl2 inhibitor (Calbiochem #616404, USA) 10 mg/kg, 20-μl DMSO in a 0.2-ml normal saline given as a single i.p. injection as per manufacturer's instructions, and previous reports [34]. Two groups of WT mice treated with Tpl2 inhibitor were studied: the pretreatment group, in which the inhibitor was given 5 min prior to the initiation of high VT ventilation, and the posttreatment group, in which the inhibitor was given 30 min after high VT ventilation. As controls served WT and Tpl2-/-mice ventilated briefly (approximately 1 min), until paralyzed, to obtain an inspiratory pressure volume curve and subsequently BALF and tissue samples.

The possible protective effects of pharmacologic Tpl2 inhibition were studied in WT mice ventilated with high VT as described above, and treated with a Tpl2 inhibitor (Calbiochem #616404, USA) 10 mg/kg, 20-μl DMSO in a 0.2-ml normal saline given as a single i.p. injection as per manufacturer's instructions, and previous reports [34]. Two groups of WT mice treated with Tpl2 inhibitor were studied: the pretreatment group, in which the inhibitor was given 5 min prior to the initiation of high VT ventilation, and the posttreatment group, in which the inhibitor was given 30 min after high VT ventilation. As controls served WT and Tpl2-/-mice ventilated briefly (approximately 1 min), until paralyzed, to obtain an inspiratory pressure volume curve and subsequently BALF and tissue samples. The activation of ERK1/2 induced by mechanical ventilation, which is known to occur even after very brief periods of ventilation [14, 15] was studied in WT and Tpl2-/-mice in the same study groups, but total ventilation time was limited to 60 min. Due to the rapid activation of ERK1/2 upon mechanical ventilation different controls were required for this experiment. Specifically, as controls served mice of both genotypes not subjected to mechanical ventilation, euthanized with pentobarbital. In these experiments, only BALF cells and lungs were collected.

60 min. Due to the rapid activation of ERK1/2 upon mechanical ventilation different controls were required for this experiment. Specifically, as controls served mice of both genotypes not subjected to mechanical ventilation, euthanized with pentobarbital. In these experiments, only BALF cells and lungs were collected. Evaluation of lung injury Ventilator-induced lung injury is characterized by high-permeability pulmonary edema and inflammation. The presence of high-permeability pulmonary edema was evaluated using lung compliance, oxygenation, and BALF protein concentration. The presence of inflammation was evaluated by BALF cytokines. Lung histological evaluation also provided information on alveolar membrane integrity and presence of inflammatory cellular infiltration. As an indicator of lung compliance, we used the inspiratory capacity, defined as the volume to inflate the lungs to an airway pressure of 25 cmH2O [33]. Results are expressed as percentage (%) of control and normalized to body weight due to the differences in weight of study mice. Concentration of proteins in BALF was measured using bicinchoninic acid assay (Pierce Chemical Co, Rockford, IL, USA). The levels of the proinflammatory cytokines, interleukin-6 (IL-6), and macrophage inflammatory protein 2 (MIP-2), were measured in BALF ELISAs (R&D Systems Inc, Minneapolis, MN, USA). Paraffin-embedded lung sections, sectioned 6-μm thick and stained with hematoxylin and eosin, were analyzed by a pathologist blinded to the treatment groups. In each group, 20 random high power fields (×400) were scored, for five independent variables: neutrophils in alveolar spaces, neutrophils in interstitial spaces, hyaline membranes, proteinaceous debris filling the airspaces, and alveolar septal thickening, as previously described [35]. The resulting injury score is a continuous value between 0 and 1.

er fields (×400) were scored, for five independent variables: neutrophils in alveolar spaces, neutrophils in interstitial spaces, hyaline membranes, proteinaceous debris filling the airspaces, and alveolar septal thickening, as previously described [35]. The resulting injury score is a continuous value between 0 and 1. Evaluation of ERK1/2 phosphorylation in WT and TPL2-/-mice after high VT ventilation Levels of ERK1/2 phosphorylation were evaluated in BALF cells by flow cytometry and in lung homogenates by Western blot. BALF cells were immediately fixed in 1.5% formaldehyde permeabilized by ice-cold methanol for 10 min and then washed and re-suspended in PBS (Ca2+- and Mg2+-free) containing 0.1 mM EDTA, 5% FBS, and 0.05% NaN3. BALF cells were first incubated with rabbit anti-mouse phospho-p44/42 MAPK (ERK1/2) (Thr202/Tyr204) antibody (Cell Signaling Techn, Danvers, MA, USA) for 1 h at 4°C. Then, cells were incubated for 20 min at 4°C with FITC goat anti-rabbit IgG (BD Biosciences, Franklin Lakes, NJ, USA). To discriminate alveolar macrophages, cell surface staining was carried out by incubation with PerCP-Cy5.5 anti-mouse CD11c (Biolegent, San Diego, CA, USA) for 30 min at 4°C. Appropriate isotype control was also used. The flow cytometry events were acquired in a MoFlo Legacy Cell Sorter (Beckman Coulter, Inc., Fullerton, CA, USA) and analyzed with the use of Summit Software (Summit Software, Inc., Fort Wayne, IN, USA). For Western blot analysis, a 100-μl tissue sample was suspended in 500 μl of lysis buffer containing 50 mM Tris (pH 6.8), 2% sodium dodecyl sulfate, 5 mM EDTA, and protease inhibitors (Complete, Boehringer, Ingelheim, Germany), homogenized, ultrasonicated, and centrifuged. A 0.1* volume of loading buffer (containing 0.3% bromophenol blue, 50% glycerol, 0.3% mercaptoethanol, and 50% [v/v] lysis buffer) was added. Samples were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis and transferred to nitrocellulose membranes. Membranes were labeled with a phospho- and total ERK1/2 primary antibody (Cell Signaling Technology, Danvers, MA, USA). Washed membranes were incubated with goat anti-rabbit anti-serum conjugated with horseradish peroxidase (Amersham International, Amersham, UK). Antigen-antibody complexes on the membranes were detected by enhanced chemiluminescence (Amersham).

total ERK1/2 primary antibody (Cell Signaling Technology, Danvers, MA, USA). Washed membranes were incubated with goat anti-rabbit anti-serum conjugated with horseradish peroxidase (Amersham International, Amersham, UK). Antigen-antibody complexes on the membranes were detected by enhanced chemiluminescence (Amersham). Statistical analysis Data were compared by one-way ANOVA, using the Shapiro-Wilk normality test, and the Kruskal-Wallis test for non-parametric data with Dunn's multiple comparisons posttest, with SigmaStat software. For each of the parameter evaluated, the comparisons made included: comparison between similarly ventilated animals of different genotype or treatment, and of ventilated animals with their genotype-matched controls. All data in text are expressed as means ± SD. Significance was defined as p < 0.05. Results Anesthesia and mechanical ventilation for 240 min were tolerated by WT and TPL2-/-mice Blood pressure averaged 92 ± 20 mmHg after 30 min of MV and 93 ± 23 mmHg at the end of the experiment in WT mice, and 98 ± 21 and 105 ± 24 mmHg, respectively, in Tpl2-/-mice (p > 0.05 for all comparisons). All mice survived the 240 min of mechanical ventilation.

tilation for 240 min were tolerated by WT and TPL2-/-mice Blood pressure averaged 92 ± 20 mmHg after 30 min of MV and 93 ± 23 mmHg at the end of the experiment in WT mice, and 98 ± 21 and 105 ± 24 mmHg, respectively, in Tpl2-/-mice (p > 0.05 for all comparisons). All mice survived the 240 min of mechanical ventilation. WT and TPL2-/-mice do not differ at baseline and after mechanical ventilation with normal VT Control WT and Tpl2-/-mice had no differences on histological appearance, lung compliance, and concentrations of proteins, IL-6 and MIP-2 in BALF. Similarly, after 240 min of mechanical ventilation with normal VT, WT, and Tpl2-/-, the mice had no differences in all indices of lung injury studied, inspiratory capacity (Figure 2), arterial oxygen (Figure 3), and concentration of proteins (Figure 4) and IL-6 (Figure 5) in BALF. Mechanical ventilation with normal VT resulted in a similar increase in BALF IL-6 concentration in both strains.Figure 2 Lung mechanics. Inspiratory capacity, defined as the volume (in ml/kg) to inflate the lungs to an airway pressure of 25 cmH2O, expressed as percentage (%) of control, of wild-type (WT, white boxes), and Tpl2-deficient mice (Tpl2-/-, gray boxes) subjected only to sample collection (control, n = 8 per group), mechanical ventilation with normal tidal volume (VT) for 240 min (n = 7 to 9 per group), or high VT for 210 min (n = 8 to 9 per group), and WT mice subjected to 210 min of high VT ventilation treated with a Tpl2 inhibitor, either prior (Tpl2inh pre, n = 5) or after initiation of high VT ventilation (Tpl2inh post, n = 5). Wild-type mice subjected to high VT ventilation had lower inspiratory capacity compared to control WT mice, as well as compared to similarly ventilated Tpl2-/-mice, and WT mice treated with the Tpl2 inhibitor, *p < 0.01. Data are presented in box plots, where boxes represent 25th to 75th percentile; line represent median and whisker represent min and max, in all figures.

ower inspiratory capacity compared to control WT mice, as well as compared to similarly ventilated Tpl2-/-mice, and WT mice treated with the Tpl2 inhibitor, *p < 0.01. Data are presented in box plots, where boxes represent 25th to 75th percentile; line represent median and whisker represent min and max, in all figures. Figure 3 Oxygenation. Arterial blood gas PaO2 from wild-type (WT, white boxes), and Tpl2-deficient mice (Tpl2-/-, gray boxes) subjected to mechanical ventilation with normal tidal volume (VT) for 240 min, or high VT for 210 min (n = 6 to 8 per group). PaO2 was lower in WT mice subjected to high VT than in similarly ventilated Tpl2-/-mice, *p < 0.05. Figure 4 BALF protein. Bronchoalveolar lavage fluid (BALF) concentration of protein from wild-type (WT, white boxes), and Tpl2-deficient mice (Tpl2-/-, gray boxes) subjected only to sample collection (control, n = 7 to 8 per group), mechanical ventilation with normal tidal volume (VT) for 240 min (n = 6 to 8 per group), or high VT for 210 min (n = 7 to 9 per group), and WT mice subjected to 210 min of high VT ventilation treated with a Tpl2 inhibitor, either prior (Tpl2inh pre, n = 5) or after initiation of high VT ventilation (Tpl2inh post, n = 5). BALF proteins' concentration was higher in WT and Tpl2-/-mice subjected to high VT ventilation, than in control WT and Tpl2-/-mice, respectively, *p < 0.001. BALF proteins' concentration was higher in WT mice subjected to high VT ventilation than in similarly ventilated Tpl2-/-mice, and WT mice treated with the Tpl2 inhibitor, # p < 0.01.

ntration was higher in WT and Tpl2-/-mice subjected to high VT ventilation, than in control WT and Tpl2-/-mice, respectively, *p < 0.001. BALF proteins' concentration was higher in WT mice subjected to high VT ventilation than in similarly ventilated Tpl2-/-mice, and WT mice treated with the Tpl2 inhibitor, # p < 0.01. Figure 5 Bronchoalveolar lavage fluid (BALF) concentration of (A) IL-6 and (B) MIP-2, from wild-type (WT, white boxes), and Tpl2-deficient mice (Tpl2 -/- , gray boxes) subjected only to sample collection (control, n=5-8), mechanical ventilation with normal tidal volume for 240 min, or high tidal volume for 210 min (n=5-7), and WT mice subjected to 210 min of high tidal volume ventilation treated with a Tpl2 inhibitor, either prior (Tpl2inh pre, n=5), or after initiation of high VT ventilation (Tpl2inh post, n=5). BALF IL-6 was lower in genotype-matched control mice than in ventilated mice, *p<0.05. High tidal volume-induced increase in BALF IL-6 was greater in WT untreated mice, than in Tpl2-/- mice, and WT mice treated with the Tpl2 inhibitor, *p<0.05. BALF MIP-2 concentration was higher in WT mice ventilated with high VT than in WT control mice *p<0.001. BALF MIP-2 concentration was also higher in WT mice ventilated with high VT than in similarly ventilated Tpl2-/- mice, and WT mice pre-treated with the Tpl2 inhibitor, *p<0.001.

ted with the Tpl2 inhibitor, *p<0.05. BALF MIP-2 concentration was higher in WT mice ventilated with high VT than in WT control mice *p<0.001. BALF MIP-2 concentration was also higher in WT mice ventilated with high VT than in similarly ventilated Tpl2-/- mice, and WT mice pre-treated with the Tpl2 inhibitor, *p<0.001. Genetic deficiency of Tpl2-/-protects from high VT ventilation-induced lung injury Mechanical ventilation with high VT-induced high-permeability pulmonary edema that was more severe in WT than in Tpl2-/-mice. After 210 min of high VT ventilation, lung compliance decreased in WT mice to 60% of control, but not in Tpl2-/-mice (93% of control; Figure 2). The development of pulmonary edema was associated with impaired oxygenation (Figure 3). Tpl2-/-mice subjected to high VT ventilation had higher PaO2 than WT mice (145 ± 18 vs. 70 ± 13 mmHg, p < 0.05, Figure 2). There were no differences in PaCO2, pH, and lactate between the two genotypes after high VT ventilation (data not shown). The concentration of proteins in BALF increased after high VT ventilation from baseline in WT and Tpl2-/-mice (p < 0.001 vs. corresponding controls). Yet, BALF proteins concentration was lower in Tpl2-/-than in WT mice after high VT ventilation (1,047 ± 385 vs. 2,051 ± 495 μg/ml, p < 0.01, Figure 4).

(data not shown). The concentration of proteins in BALF increased after high VT ventilation from baseline in WT and Tpl2-/-mice (p < 0.001 vs. corresponding controls). Yet, BALF proteins concentration was lower in Tpl2-/-than in WT mice after high VT ventilation (1,047 ± 385 vs. 2,051 ± 495 μg/ml, p < 0.01, Figure 4). Mechanical ventilation with high VT induced an inflammatory response in the lungs of mice of both genotypes, but more severe in WT than in Tpl2-/-mice. BALF concentration of IL-6 increased from control, more in WT than in Tpl2-/-mice, while the concentration of MIP-2 in BALF increased only in WT mice and not in Tpl2-/-mice (Figure 5A,B). In the histological examination, lungs from WT mice subjected to high VT ventilation presented severe injury with the presence of mixed inflammatory infiltrates (neutrophils and alveolar macrophages) in the interstitial and alveolar spaces, edema, and thickening of the alveolar walls, and a lung injury score of 0.7. Lungs from Tpl2-/-mice subjected to high VT ventilation presented only sparse, very mild inflammatory infiltrates, and a lung injury score of 0.33 (Figure 6).Figure 6 Histology. Representative photographs of hematoxylin-eosin-stained lung sections, ×200 magnification, from control wild-type and Tpl2-deficent mice and from wild-type and Tpl2-deficient mice subjected to high tidal volume ventilation for 210 min (n = 4 per group) and calculated lung injury score. Mechanical ventilation with high VT-induced lung injury that was more severe in WT than in Tpl2-/-mice *p < 0.01.

ification, from control wild-type and Tpl2-deficent mice and from wild-type and Tpl2-deficient mice subjected to high tidal volume ventilation for 210 min (n = 4 per group) and calculated lung injury score. Mechanical ventilation with high VT-induced lung injury that was more severe in WT than in Tpl2-/-mice *p < 0.01. Pharmacologic inhibition of Tpl2 protects from high VT ventilation-induced lung injury The observation that Tpl2 deficiency is protective in our VILI model prompted us to examine the potential therapeutic effect of pharmacologic inhibition of Tpl2 in high VT-induced lung injury. First, we examined if pretreatment with Tpl2 inhibitor would ameliorate VILI in WT mice. When WT mice were subjected to high VT ventilation, no differences were observed between untreated mice and mice receiving an i.p. injection of DMSO in normal saline (vehicle) in any of the parameters evaluated (data not shown), and therefore, untreated mice were used as controls. Pretreatment with the Tpl2 inhibitor was effective in preventing high VT-induced decrease in lung compliance (Figure 2). Additionally, high VT-induced increase in the concentrations of protein, IL-6 and MIP-2, observed in ventilated, untreated WT mice, was less in mice pretreated with the Tpl2 inhibitor (Figures 4 and 5).

treatment with the Tpl2 inhibitor was effective in preventing high VT-induced decrease in lung compliance (Figure 2). Additionally, high VT-induced increase in the concentrations of protein, IL-6 and MIP-2, observed in ventilated, untreated WT mice, was less in mice pretreated with the Tpl2 inhibitor (Figures 4 and 5). Although pretreatment was effective, such therapeutic approach is rarely feasible in clinical practice. We, therefore, tested the effects of pharmacological Tpl2 inhibition when the inhibitor was administered 30 min after the initiation of high VT ventilation. We found that posttreatment with the Tpl2 inhibitor was also effective in ameliorating indices of VILI. Lung compliance and concentrations of protein and IL-6 in BALF were similar between WT subjected to high VT ventilation and treated with the Tpl2 inhibitor either before or after initiation of high VT, and always lower than in untreated mice (Figures 2, 3, 4, and 5). Although BALF concentration of MIP-2 was similar in mice treated with the Tpl2 inhibitor before or after high VT, only the pretreated group had significantly lower MIP-2 than the untreated group (Figure 5B).

either before or after initiation of high VT, and always lower than in untreated mice (Figures 2, 3, 4, and 5). Although BALF concentration of MIP-2 was similar in mice treated with the Tpl2 inhibitor before or after high VT, only the pretreated group had significantly lower MIP-2 than the untreated group (Figure 5B). High VT ventilation-induced ERK1/2 phosphorylation is decreased in alveolar macrophages of Tpl2-/-mice and WT mice treated with the Tpl2 inhibitor It is known that ERK1/2 activation contributes to the inflammatory response induced by high VT ventilation, and that Tpl2 is an essential regulator of ERK1/2 activation. We therefore examined ventilation-induced activation of ERK1/2 in alveolar macrophages and total lungs homogenates from WT and Tpl2-/-mice. BALF cells from control, non-ventilated mice and from mice subjected to 60 min of mechanical ventilation consists of >85% alveolar macrophages (data not shown). Mechanical ventilation with high VT but not normal VT increased ERK1/2 phoshorylation in alveolar macrophages of WT mice (Figure 7). This increase in levels of phospho-ERK1/2 after high VT ventilation was not observed in alveolar macrophages from Tpl2-/-mice and in those from WT mice treated with the Tpl2 inhibitor. Levels of phospho-ERK1/2 were similar in lung homogenates from WT and Tpl2-/-mice ventilated on high VT.Figure 7 ERK1/2 phosphorylation. (A) The histograms are representative of p-ERK1/2 fluorescence intensity of BALF CD11c positive cells (alveolar macrophages) from non-ventilated mice (n = 6), and from wild-type and Tpl2-deficient mice exposed to either 60 min of normal tidal volume ventilation (n = 4 per group) or 30 min of normal tidal volume ventilation followed by 30 min of high tidal volume ventilation (n = 4 per group), with or without treatment with a Tpl2 inhibitor given 5 min prior to high tidal ventilation. BALF cells stained with isotype control are also depicted. ERK1/2 phosphorylation, as indicated by mean fluorescent intensity (MFI), was higher in WT mice subjected to high VT ventilation than in WT control mice and WT mice ventilated with normal VT *p < 0.05. ERK1/2 phosphorylation was also higher in WT mice subjected to high VT ventilation than in similarly ventilated Tpl2-/-mice and WT mice treated with the Tpl2 inhibitor *p < 0.05.

ensity (MFI), was higher in WT mice subjected to high VT ventilation than in WT control mice and WT mice ventilated with normal VT *p < 0.05. ERK1/2 phosphorylation was also higher in WT mice subjected to high VT ventilation than in similarly ventilated Tpl2-/-mice and WT mice treated with the Tpl2 inhibitor *p < 0.05. (B) Representative Western blot analysis for phosphorylated and total ERK1/2 of lung homogenates from non-ventilated mice (n = 4 per group) and from WT and Tpl2-/-mice (n = 4 per group) exposed to 30 min of high tidal volume ventilation. Mechanical ventilation was associated with increased ERK1/2 phosphorylation in lungs of both WT and Tpl2-/-mice compared to genotype-matched control mice, *p < 0.05. Discussion The inflammatory response and barrier dysfunction that characterize acute lung injury can be exacerbated by mechanical ventilation, and, apart from low tidal volume ventilation, no effective treatment is available yet. Identifying the intracellular signaling molecules involved in VILI is thus important for designing novel therapeutic approaches. In the present study we show that inhibition of Tpl2, a MAP3K kinase, is protective in a mouse model of VILI. Indices of high-permeability pulmonary edema and lung inflammation induced by injurious ventilation were lower in Tpl2-deficient mice than in WT ones. More importantly, this protective effect was reproduced using pharmacologic inhibition of Tpl2 in WT mice.

Tpl2, a MAP3K kinase, is protective in a mouse model of VILI. Indices of high-permeability pulmonary edema and lung inflammation induced by injurious ventilation were lower in Tpl2-deficient mice than in WT ones. More importantly, this protective effect was reproduced using pharmacologic inhibition of Tpl2 in WT mice. It is well established that injurious ventilation induces high-permeability pulmonary edema and inflammation. Studies in ARDS patients have shown that even brief periods of injurious ventilation increase BALF inflammatory mediators [36]. In mouse models, high tidal volume ventilation induces lung injury characterized by pulmonary edema with increased concentrations of proteins and cytokines in BALF and inflammatory cell infiltration and diffuse alveolar damage on histology, a picture similar to human ARDS [1]. In this study, high VT ventilation-induced lung injury in WT mice, characterized by deterioration in lung mechanics and oxygenation, increased concentration of proteins and cytokines in BALF, and inflammatory cell infiltration on histology. Tpl2 deficiency was associated with less lung injury upon high VT ventilation. Specifically, Tpl2-deficient mice showed no deterioration of lung mechanics and oxygenation, less injury on histology, and lower levels of BALF proteins and cytokines than WT mice, suggesting a protective role of Tpl2 inhibition in VILI.

gy. Tpl2 deficiency was associated with less lung injury upon high VT ventilation. Specifically, Tpl2-deficient mice showed no deterioration of lung mechanics and oxygenation, less injury on histology, and lower levels of BALF proteins and cytokines than WT mice, suggesting a protective role of Tpl2 inhibition in VILI. The ERK1/2 pathway is a highly conserved signaling pathway involved in fundamental cellular processes such as growth, differentiation, and survival. ERK1/2 is activated by a wide variety of receptors including G protein-coupled receptors (GPCR), tyrosine kinase receptors, TLRs, ion channels, and others [9, 10]. Studies have shown that ERK1/2 is also involved in mechanotransduction [1], and that ERK1/2 activation by mechanical stretch contributes to the inflammatory response induced by injurious ventilation [4, 15].

including G protein-coupled receptors (GPCR), tyrosine kinase receptors, TLRs, ion channels, and others [9, 10]. Studies have shown that ERK1/2 is also involved in mechanotransduction [1], and that ERK1/2 activation by mechanical stretch contributes to the inflammatory response induced by injurious ventilation [4, 15]. Tpl2 is an essential regulator of ERK1/2 activation, mediating signals initiated by cytokine or Toll-like receptors to induce pro-inflammatory cytokine production [22, 28]. In disease models, Tpl2-/-mice were initially found to be resistant to LPS/d-galactosamine-induced endotoxin shock [28]. Recent studies showed that Tpl2 deficiency was associated with reduced adipose tissue inflammation in diet-induced obesity, and reduced acetaminophen-induced liver injury [37, 38]. In an experimental colitis model, pharmacologic as well as genetic inhibition of Tpl2 was found effective in reducing bowel inflammation [32]. Yet, the role of Tpl2 in inflammatory lung diseases has not been investigated. Involvement of Tpl2 in lung inflammation has been only indirectly demonstrated, since Tpl2 was found up-regulated in a proteome analysis of lung tissues from rats exposed to cigarette smoke [39], and mice lacking Tpl2 had reduced lung inflammation in a model of acute pancreatitis [31]. This study showed that Tpl2 is involved in barrier dysfunction and inflammation triggered by injurious ventilation, as inhibition of Tpl2 ameliorated lung inflammation and high-permeability pulmonary edema.

cigarette smoke [39], and mice lacking Tpl2 had reduced lung inflammation in a model of acute pancreatitis [31]. This study showed that Tpl2 is involved in barrier dysfunction and inflammation triggered by injurious ventilation, as inhibition of Tpl2 ameliorated lung inflammation and high-permeability pulmonary edema. The pathway consistently shown to be impaired in the absence or inhibition of Tpl2 is the ERK1/2 pathway [25, 30–32, 37]. Also in this study, high tidal volume-induced ERK1/2 activation was lower in alveolar macrophages from Tpl2-/-mice and from WT mice treated with the Tpl2 inhibitor than from WT untreated mice. The observation that ERK1/2 activation was similar in lung homogenates of WT and Tpl2-/-mice could be explained by the presence of other ERK1/2 activating signals, not mediated by Tpl2, such as TGFβ or other growth factors. The possible inhibition of other kinases was not examined and cannot be ruled out.

treated mice. The observation that ERK1/2 activation was similar in lung homogenates of WT and Tpl2-/-mice could be explained by the presence of other ERK1/2 activating signals, not mediated by Tpl2, such as TGFβ or other growth factors. The possible inhibition of other kinases was not examined and cannot be ruled out. An important finding of this study is that pharmacologic inhibition of Tpl2, and not only genetic, was effective in ameliorating VILI. Systemic administration of a Tpl2 inhibitor protected from VILI, both when given as pretreatment and when given after initiation of injurious ventilation. A time point well after establishing high VT (30 min) was chosen as posttreatment, reasoning that ERK1/2 activation occurs sooner, even within 5 min after high stretch [14, 15], and that a deterioration of lung injury resulting in high airway pressures would be recognized by a clinician within 30 min and prompt therapeutic interventions. WT mice receiving the Tpl2 inhibitor after being on high VT ventilation for 30 min had similar lung mechanics, and BALF protein, IL-6 and MIP-2 concentrations as Tpl2-/-mice at the end of the experiment, and lower than untreated, ventilated on high VT, WT mice. The only observed difference was that the delayed treatment with the Tpl2 inhibitor was not able to prevent the increase in BALF concentration of MIP2 induced by injurious ventilation. These findings suggest that the protective effect of Tpl2 inhibition is quite well maintained when the inhibitor is administered after an increase in airway pressure is observed, as it would happen in clinical practice.

was not able to prevent the increase in BALF concentration of MIP2 induced by injurious ventilation. These findings suggest that the protective effect of Tpl2 inhibition is quite well maintained when the inhibitor is administered after an increase in airway pressure is observed, as it would happen in clinical practice. The potential advantage of Tpl2 as therapeutic target over ERK1/2 is its selective activation by inflammatory stimuli with a consequent reduction in side effects. Thus, inhibition of Tpl2 will not affect activation of ERK1/2 by other agonists, like growth factors, which has been proven protective in several diseases, such as myocardial and cerebral ischemia-reperfusion injury [40, 41]. Small molecule inhibitors designed to suppress Tpl2 were able to inhibit pro-inflammatory cytokine production from LPS-treated human primary macrophages [42], supporting the therapeutic potential of Tpl2 inhibition. Additionally, inhibition of Tpl2 has been a promising target for inflammatory diseases including inflammatory bowel disease, rheumatoid arthritis and liver disease [42–44], and Tpl2 inhibitors are among the ones to be tested in clinical trials [18].

ophages [42], supporting the therapeutic potential of Tpl2 inhibition. Additionally, inhibition of Tpl2 has been a promising target for inflammatory diseases including inflammatory bowel disease, rheumatoid arthritis and liver disease [42–44], and Tpl2 inhibitors are among the ones to be tested in clinical trials [18]. Of course the potential adverse effects of Tpl2 inhibition have been evaluated neither in clinical practice nor under experimental conditions. Although Tpl2-/-mice appear to be prone to chemically induced carcinogenesis in several animal models [22], they do not develop spontaneous cancers, and the clinical importance of Tpl2-deficiency-mediated carcinogenesis in acute inflammatory conditions, such as VILI is probably small; as such treatments would be given for only brief periods of time. Data on toxicity of Tpl2 inhibitors are lacking, but an inhibitor of MEK1/2 (mitogen-activated protein kinase kinase), a kinase that is upstream of ERK1/2 in the Raf-MEK-ERK pathway, has been investigated in phase I and phase II clinical trials in patients with advanced malignancies and proved to have a good safety profile [45–48]. A more important concern of Tpl2 inhibition is the potential immune suppression. Indeed, Tpl2 ablation ameliorated macrophages response to LPS [28, 49], increased susceptibility of mice to Listeria monocytogenes infection [50], and decreased clearance of Toxoplasma gondii [51]. It is possible that Tpl2 inhibition could adversely affect critically ill patients with ongoing infection. This is a common limitation of all anti-inflammatory therapeutic interventions, such as corticosteroids in ALI-VILI.

ice to Listeria monocytogenes infection [50], and decreased clearance of Toxoplasma gondii [51]. It is possible that Tpl2 inhibition could adversely affect critically ill patients with ongoing infection. This is a common limitation of all anti-inflammatory therapeutic interventions, such as corticosteroids in ALI-VILI. In this study, we use our established model of aseptic, ‘one-hit’ VILI [33], to clarify the role of Tpl2. This ‘one-hit’ model is widely used in studies to investigate the pathogenesis of VILI. Using this model, we were able to confirm that inhibition of Tpl2 ameliorates indices of high-permeability pulmonary edema and inflammation induced by injurious ventilation. Nonetheless, this ‘one-hit’ model has inherent limitations. Specifically, a very high, not clinically relevant tidal volume has to be used. We and others have shown [33, 52, 53] that only such high volume can result in distortion of lung units required to induce VILI in healthy mice. Yet, it is now clear that in patients with ARDS some alveoli are exposed to such high distending pressures [54]. Additionally, to compensate for the hemodynamic effects of high intrathoracic pressures, mice receive fluid loading, that may contribute to the development of pulmonary edema. As all groups of ventilated mice received the same amount of fluids, but only the group of WT mice ventilated on high tidal volume developed significant edema, it is reasonable to assume that fluid loading may only exacerbate the high tidal volume-induced alteration in alveolar permeability.

lopment of pulmonary edema. As all groups of ventilated mice received the same amount of fluids, but only the group of WT mice ventilated on high tidal volume developed significant edema, it is reasonable to assume that fluid loading may only exacerbate the high tidal volume-induced alteration in alveolar permeability. Conclusion In conclusion, Tpl2 contributes to the pathogenesis of high-permeability pulmonary edema and inflammation induced by high tidal volume ventilation, as genetic deficiency of Tpl2 appears to be protective in this model of murine VILI. Additionally, pharmacologic inhibition of Tpl2 is effective in ameliorating indices of VILI, even when given after establishing injurious ventilation, suggesting a potential therapeutic role for Tpl2 inhibitors in VILI.

me ventilation, as genetic deficiency of Tpl2 appears to be protective in this model of murine VILI. Additionally, pharmacologic inhibition of Tpl2 is effective in ameliorating indices of VILI, even when given after establishing injurious ventilation, suggesting a potential therapeutic role for Tpl2 inhibitors in VILI. Authors’ information E Kaniaris, MD, is a fellow in pulmonary medicine and graduate student at the Department of Intensive Care Medicine, Experimental Intensive Care Medicine Laboratory, KV, MD, PhD, is an assistant professor in Intensive Care Medicine, PI at the Experimental Intensive Care Medicine Laboratory, EV, MD, is a fellow in pediatrics and graduate student at the department of Clinical Chemistry, ET, MD, is a graduate student at the Department of Intensive Care Medicine. E Kondili, MD, PhD, is an assistant professor in the Department of Intensive Care Medicine; EL, MD, is a pathologist at the Department of Pathology. CT, PhD, is associate professor in Clinical Chemistry, head of the Laboratory of Clinical Chemistry, and member of the research team who generated the Tpl2-deficient mice, and DG, MD, PhD, is a professor and head of the Department of Intensive Care Medicine. All departments and authors are affiliated to the Medical School of the University of Crete in Greece. Abbreviations ALIacute lung injury BALFbronchoalveolar lavage ERK1/2extracellular signal-regulated kinase 1/2 GPCRG protein-coupled receptors IL-6interleukin-6 JNKC-Jun N-terminal kinase MAPKmitogen-activated serine/threonine kinase MEK1/2 (or MAP2K)mitogen-activated protein kinase kinase

Authors’ information E Kaniaris, MD, is a fellow in pulmonary medicine and graduate student at the Department of Intensive Care Medicine, Experimental Intensive Care Medicine Laboratory, KV, MD, PhD, is an assistant professor in Intensive Care Medicine, PI at the Experimental Intensive Care Medicine Laboratory, EV, MD, is a fellow in pediatrics and graduate student at the department of Clinical Chemistry, ET, MD, is a graduate student at the Department of Intensive Care Medicine. E Kondili, MD, PhD, is an assistant professor in the Department of Intensive Care Medicine; EL, MD, is a pathologist at the Department of Pathology. CT, PhD, is associate professor in Clinical Chemistry, head of the Laboratory of Clinical Chemistry, and member of the research team who generated the Tpl2-deficient mice, and DG, MD, PhD, is a professor and head of the Department of Intensive Care Medicine. All departments and authors are affiliated to the Medical School of the University of Crete in Greece. Abbreviations ALIacute lung injury BALFbronchoalveolar lavage ERK1/2extracellular signal-regulated kinase 1/2 GPCRG protein-coupled receptors IL-6interleukin-6 JNKC-Jun N-terminal kinase MAPKmitogen-activated serine/threonine kinase MEK1/2 (or MAP2K)mitogen-activated protein kinase kinase MAP3Kmitogen-activated protein kinase kinase kinase MIP-2macrophage inflammatory protein 2 PIPpeak inspiratory pressure Tpl2tumor progression locus 2 Tpl2-/-Tpl2 deficient VILIventilator-induced lung injury VTtidal volume WTwild-type. Competing interests The authors of this study have no competing interests to declare. Authors’ contributions

MEK1/2 (or MAP2K)mitogen-activated protein kinase kinase MAP3Kmitogen-activated protein kinase kinase kinase MIP-2macrophage inflammatory protein 2 PIPpeak inspiratory pressure Tpl2tumor progression locus 2 Tpl2-/-Tpl2 deficient VILIventilator-induced lung injury VTtidal volume WTwild-type. Competing interests The authors of this study have no competing interests to declare. Authors’ contributions E Kaniaris performed in vivo and in vitro experiments and drafted the manuscript, KV conceived the study, performed in vivo experiments, analyzed data, drafted, and edited the manuscript. EV and ET performed and analyzed in vitro experiments. E Kondili contributed to study design and edited the manuscript; EL performed and evaluated the histology. CT contributed to the study design, and data analysis and drafted the manuscript. DG contributed to the study design and coordination and edited the manuscript. All authors read and approved the final manuscript. Acknowledgements This study has been supported by the Cretan Society for Research in Intensive Care Medicine.

Background Extracorporeal membrane oxygenation (ECMO) is increasingly being utilized as a viable supportive treatment modality for acute severe refractory cardiorespiratory failure [1-3]. It can be used short term where recovery of the lungs or heart is anticipated, or as a bridge to long-term mechanical cardiac support or lung transplantation. Patients supported on ECMO are not only critically unwell but are often at cross roads while they await organ recovery or further definitive therapy [1]. Despite controversies [4] around what constitutes optimal nutritional support in an intensive care unit (ICU), preserving nutritional status and overall physical condition is an important aspect of their management [5]. There is growing evidence that malnutrition in the critically ill influences morbidity and mortality [6-8]. Patients may receive ECMO support for weeks or may have had chronic disease which impacts their nutritional status pre-morbidly. Available data indicates that feed intolerance may occur in up to 38% of the ECMO patients [5] and that patients may be receiving as little as 50% of their nutritional requirements due to enteral feed intolerance [9] resulting in large caloric deficit. In this setting, sequestration if any of vital nutrients in ECMO circuits may lead to further nutritional debilitation.

ccur in up to 38% of the ECMO patients [5] and that patients may be receiving as little as 50% of their nutritional requirements due to enteral feed intolerance [9] resulting in large caloric deficit. In this setting, sequestration if any of vital nutrients in ECMO circuits may lead to further nutritional debilitation. The ECMO circuit is not, however, a passive conduit for the blood and may sequester a variety of circulating compounds such as drugs [10-12] and possibly nutrients, effectively reducing the bioavailability of these compounds. Other ex vivo studies in cardiopulmonary bypass circuits have demonstrated that these circuits can sequester trace elements [13]. Hence, it is therefore plausible that there may be clinically relevant interactions between the ECMO circuit and macro-/micronutrients. However, to date, there is no data pertaining to nutrient disposition in the ECMO circuit. The aim of this 24-h single-dose study was to identify any significant reduction in circulating concentrations of macronutrients (glucose, amino acids and fatty acids) or micronutrients (fat soluble vitamins, water soluble vitamins, selenium, copper, zinc and manganese) in an ex vivo ECMO circuit model primed with the fresh human whole blood. Methods Ethics approval was obtained from the local Human Research Ethics Committee (HREC/12/QPCH/90).

The aim of this 24-h single-dose study was to identify any significant reduction in circulating concentrations of macronutrients (glucose, amino acids and fatty acids) or micronutrients (fat soluble vitamins, water soluble vitamins, selenium, copper, zinc and manganese) in an ex vivo ECMO circuit model primed with the fresh human whole blood. Methods Ethics approval was obtained from the local Human Research Ethics Committee (HREC/12/QPCH/90). ECMO circuits The methods for the development of the ex vivo model of ECMO used in the experiment have been published previously [14]. Briefly, three new ECMO circuits were used (Maquet Cardiopulmonary AG, Hechinger Strabe, Germany). Each circuit had an identical composition: Bioline tubing, a PLS Quadrox D oxygenator and a Rotaflow pump head. A reservoir bladder (R-38; Medtronic Pty Ltd., Minneapolis, MN, USA) was added to contribute compliance to the system and thus, allow blood sampling from a closed circuit. The three circuits were initially primed with 0.9% saline (Baxter Healthcare, Toongabbie, NSW, Australia). Five thousand international units of Porcine mucous heparin (Pfizer Pty Ltd., Perth, Australia) was added to the circuits. The fresh human whole blood (6 days old) was then added to the circuit in exchange for the saline prime. Circuits were volume pressurized to obtain post-oxygenator pressures of 230 to 250 mmHg.

a). Five thousand international units of Porcine mucous heparin (Pfizer Pty Ltd., Perth, Australia) was added to the circuits. The fresh human whole blood (6 days old) was then added to the circuit in exchange for the saline prime. Circuits were volume pressurized to obtain post-oxygenator pressures of 230 to 250 mmHg. The final volume of the pressurized circuit was 819 ± 82 mL with a mean haemoglobin value of 70 ± 2 g/L. Oxygen tension in the circuit blood was maintained between 150 to 200 mmHg (mean 175 ± 8 mmHg). Circuit temperature was maintained at 37°C. The pH of the circulating blood was maintained in the range of 7.25 to 7.55 by use of carbon dioxide gas or sodium bicarbonate solution added to the circuit. The pH, pO2, pCO2, potassium and lactate were checked at baseline, 6 h and at the conclusion of the experiment (24 h) to ensure consistency between all three circuits. Flow rates in each circuit were maintained at 4 to 5 L/min. Ambient temperature and light conditions were identical for all three circuits during the experiment.

O2, pCO2, potassium and lactate were checked at baseline, 6 h and at the conclusion of the experiment (24 h) to ensure consistency between all three circuits. Flow rates in each circuit were maintained at 4 to 5 L/min. Ambient temperature and light conditions were identical for all three circuits during the experiment. Preparation and injection of nutrient solutions Standard parenteral nutrition (PN) solution was prepared from a ready-to-mix triple phase bag containing amino acids, glucose, lipids and electrolytes (4QH.01TC, Baxter Healthcare, Toongabbie, NSW, Australia) as per manufacturer's recommendations. One dose each of trace element solution (10 mL Baxter MTEFE) and vitamin powder (Baxter Cernevit reconstituted in 5 mL water) was added to the PN bag through a sterile port. Each PN bag was agitated for 3 min to ensure thorough distribution of contents while being covered by a light impermeable bag provided by the manufacturer. After reconstitution, 10 mL of this solution was extracted in a clean 10-mL syringe and then injected into the ex vivo circuit at the post-oxygenator site. This dose was selected to emulate typical concentrations observed in an ECMO patient receiving total PN (70 mL/h of 3:1 TPN premix solution) over a 24-h period. This was necessary to exclude any concentration-dependent sequestration.

10-mL syringe and then injected into the ex vivo circuit at the post-oxygenator site. This dose was selected to emulate typical concentrations observed in an ECMO patient receiving total PN (70 mL/h of 3:1 TPN premix solution) over a 24-h period. This was necessary to exclude any concentration-dependent sequestration. Blood sample collection and nutrient assays A 3-min period of equilibration was allowed for each circuit at a flow rate of 4 L/min prior to injection of PN solution. Blood samples (10 mL each) were collected from each circuit from the post-oxygenator site at time-points: baseline (after equilibration), 1 min (after addition of PN), 30 and 60 min and 6, 12 and 24 h. Figure 1 details the collection of samples using light protective precautions to prevent photodegradation. Samples were analysed using nutrient assays as outlined in Table 1.Figure 1 Sample collection schedule. Table 1 Nutrient assay methods Analyte Methodology

Blood sample collection and nutrient assays A 3-min period of equilibration was allowed for each circuit at a flow rate of 4 L/min prior to injection of PN solution. Blood samples (10 mL each) were collected from each circuit from the post-oxygenator site at time-points: baseline (after equilibration), 1 min (after addition of PN), 30 and 60 min and 6, 12 and 24 h. Figure 1 details the collection of samples using light protective precautions to prevent photodegradation. Samples were analysed using nutrient assays as outlined in Table 1.Figure 1 Sample collection schedule. Table 1 Nutrient assay methods Analyte Methodology Zinc (serum) The Wako Zinc (Zn) test WAKO Germany Selenium Atomic absorption spectrometry Varian Model 280Z Zeeman GFAAS Manganese (whole blood Mn) Atomic absorption spectrometry Varian Model 280Z Zeeman GFAAS Copper Atomic absorption spectrometry Varian Model 280Z Zeeman GFAAS Vitamin E (α-tocopherol) Reversed-phase HPLC with multiwavelength detection Waters™ 2690/5 HPLC Alliance Separations System Vitamin A (retinol and retinol esters) Reversed-phase HPLC with multiwavelength detection Waters™ 2690/5 HPLC Alliance Separations System Plasma amino acids Reverse phase UPLC and TUV detector MassTrak™ Amino Acid Analysis Kit Waters™ Acquity Ultra Performance LC System Vitamin C (ascorbic acid) Spectrophotometric endpoint assay Horiba ABX Diagnostics Pentra 400 Vitamin D (25-OH vitamin D2 and D3) Liquid chromatography-tandem mass spectrometry TECAN EVO2. Waters™ Acquity Ultra Performance LC system Haptoglobin Turbidimetric method Beckman Coulter Synchron Clinical System Triglycerides Timed-endpoint method Beckman Coulter Synchron Clinical System Cholesterol Timed-endpoint method Beckman Coulter Synchron Clinical System Protein (total protein) Timed-endpoint biuret method Beckman Coulter Synchron Clinical System Glucose (serum glucose) Glucose oxidase method Beckman Coulter Synchron Clinical System

Timed-endpoint method Beckman Coulter Synchron Clinical System Cholesterol Timed-endpoint method Beckman Coulter Synchron Clinical System Protein (total protein) Timed-endpoint biuret method Beckman Coulter Synchron Clinical System Glucose (serum glucose) Glucose oxidase method Beckman Coulter Synchron Clinical System Statistical analysis We used a mixed model with a random intercept for each circuit to control for the nonindependence of results from the same circuit. The nutrient concentrations in samples drawn at 1-min interval following the injection of PN solution into the circuit were included in the model as a baseline. The key predictor variable was time in hours which we assumed to be linear. All analyses were made using R version 3.0.2. This model accounts for the repeated responses from the same circuit using a random intercept. The concentration of nutrient versus time curves (mean ± SEM) were plotted using GraphPad Prism Version 5.03. Results There were no circuit complications. There were no significant differences between oxygen or carbon dioxide tensions, pH, temperature, electrolyte composition and degree of hemolysis between circuits. Circuit parameters are described in Table 2.Table 2 Circuit parameters Circuit 1 2 3 Total circuit volume (mL) 857 738 861 Crystalloid volume (mL) 325 203 374 Whole blood volume (mL) 532 535 487 Age of whole blood (days) 6 6 5 Temperature (C) 36.5 36.7 36.6 Heparin dose (IU) 5,000 5,000 5,000 Calcium chloride dose (mL) 10 10 10 Calcium dose (mmol) 3.4 3.4 3.4 Oxygen (L/min) 6 6 6

Results There were no circuit complications. There were no significant differences between oxygen or carbon dioxide tensions, pH, temperature, electrolyte composition and degree of hemolysis between circuits. Circuit parameters are described in Table 2.Table 2 Circuit parameters Circuit 1 2 3 Total circuit volume (mL) 857 738 861 Crystalloid volume (mL) 325 203 374 Whole blood volume (mL) 532 535 487 Age of whole blood (days) 6 6 5 Temperature (C) 36.5 36.7 36.6 Heparin dose (IU) 5,000 5,000 5,000 Calcium chloride dose (mL) 10 10 10 Calcium dose (mmol) 3.4 3.4 3.4 Oxygen (L/min) 6 6 6 Twenty-one samples were analysed for thirty-one nutrient compounds. Linear changes in hourly nutrient concentrations in the ECMO circuit are presented in Table 3. The 95% CIs for mean hourly changes for a majority of nutrients indicate minimal variability in nutrient concentrations between circuits. Concentration versus time curves for the test compounds are graphed in Figures 2,3,4. There were significant reductions (p < 0.05) in circuit concentrations of some amino acids [alanine (10%), arginine (95%), cysteine (14%), glutamine (25%) and isoleucine (7%)], vitamins [A (42%) and E (6%)] and glucose (42%) over 24 h. Significant increases in circuit concentrations (p < 0.05) were observed over time for most amino acids, zinc and vitamin C. There were no statistically significant reductions in total proteins, triglycerides, total cholesterol, selenium, copper, manganese and vitamin D concentrations within the ECMO circuit over a 24-h period. While the reductions in glucose concentrations and an increase in other macro- and micronutrient concentrations probably reflect cellular metabolism and breakdown, the decrement in arginine and glutamine concentrations may be attributed to their enzymatic conversion to ornithine and glutamate, respectively. There was no significant correlation between physicochemical properties such as molecular weight, polarity and solubility on circuit disposition.Table 3 Linear changes over time in nutrient levels

nine and glutamine concentrations may be attributed to their enzymatic conversion to ornithine and glutamate, respectively. There was no significant correlation between physicochemical properties such as molecular weight, polarity and solubility on circuit disposition.Table 3 Linear changes over time in nutrient levels Nutrient Units Mean Lower Upper p value

nine and glutamine concentrations may be attributed to their enzymatic conversion to ornithine and glutamate, respectively. There was no significant correlation between physicochemical properties such as molecular weight, polarity and solubility on circuit disposition.Table 3 Linear changes over time in nutrient levels Nutrient Units Mean Lower Upper p value Macronutrients Glucose* mmol/L −0.44 −0.47 −0.41 <0.001 Total protein g/L −0.02 −0.08 0.04 0.54 Total cholesterol mmol/L 0.00 −0.00 0.00 0.60 Triglycerides mmol/L −0.00 −0.02 0.00 0.45 Alanine* μmol/L −4.87 −7.78 −1.95 0.008 Arginine* μmol/L −19.6 −25.4 −13.83 <0.001 Citrulline μmol/L 0.10 0.04 0.15 <0.003 Cystine* μmol/L −0.07 −0.16 0.01 0.12 Glutamate μmol/L 3.86 3.23 4.48 <0.001 Glutamine* μmol/L −2.34 −3.29 −1.38 <0.001 Glycine μmol/L −0.04 −3.23 3.14 0.97 Histidine μmol/L 1.14 0.78 1.51 <0.001 Isoleucine* μmol/L −0.61 −1.08 −0.13 0.02 Leucine μmol/L 2.43 1.69 3.17 <0.001 Lysine μmol/L 0.8 0.34 1.410 0.006 Ornithine μmol/L 20.64 15.42 25.85 <0.001 Phenylalanine μmol/L 1.04 0.57 1.51 0.001 Proline μmol/L 0.24 −0.60 1.08 0.57 Serine μmol/L 0.32 −0.20 0.83 0.23 Taurine μmol/L 0.90 0.43 1.37 0.002 Threonine μmol/L 0.83 0.35 1.31 0.007 Tyrosine μmol/L 0.62 0.46 0.77 <0.001 Valine μmol/L 1.86 1.01 2.71 <0.001 Trace elements Copper μmol/L −0.00 −0.02 0.02 1 Manganese nmol/L −0.20 −0.51 0.1 0.23 Selenium μmol/L 0.00 −0.00 0.01 0.21 Zinc μmol/L 0.06 0.04 0.09 <0.001 Vitamins Vit A* nmol/L −0.01 −0.01 −0.01 <0.001 Vit C nmol/L 1.53 1.01 2.06 <0.001 Vit D nmol/L −0.05 −0.22 0.13 0.62 Vit E* nmol/L −0.03 −0.04 −0.01 0.003 Using a mixed model with post-parenteral nutrition solution injection plasma nutrient concentrations as the baseline and a random intercept for each circuit. Mean changes and 95% confidence intervals. The estimates in the table are per hour. For example, for alanine, the mean loss in 4.87 units per hour with a 95% confidence interval of −7.78 to −1.95. Nutrients with significant reductions in concentrations over 24 h are marked in asterisk (*).

m intercept for each circuit. Mean changes and 95% confidence intervals. The estimates in the table are per hour. For example, for alanine, the mean loss in 4.87 units per hour with a 95% confidence interval of −7.78 to −1.95. Nutrients with significant reductions in concentrations over 24 h are marked in asterisk (*). Figure 2 Glucose, lipid and total protein concentrations in the ex vivo ECMO circuit over the 24-h study period. Figure 3 Amino acid concentrations in the ex vivo ECMO circuit over the 24-h study period (a-c). Figure 4 Circulating trace element and vitamin concentrations in the ex vivo ECMO circuit over the 24-h study period. Discussion To our knowledge, this study represents the first investigation of macro- and micronutrient disposition in ex vivo ECMO circuits. Significant alterations in macro- and micronutrient concentrations were observed in this study. Most significantly, there is potential for circuit loss of essential amino acid isoleucine and lipid soluble vitamins (A and E) in the ECMO circuit, and the mechanisms for this need further exploration. While the results are generally reassuring from a macronutrient perspective, prospective studies in clinical subjects are indicated to further evaluate the influence of ECMO circuit on micronutrient concentrations and clinical outcomes.

A and E) in the ECMO circuit, and the mechanisms for this need further exploration. While the results are generally reassuring from a macronutrient perspective, prospective studies in clinical subjects are indicated to further evaluate the influence of ECMO circuit on micronutrient concentrations and clinical outcomes. The findings of this preliminary study that identifies nutrients that risk being lost in ECMO circuits are significant for the following reasons. In the absence of circuit sequestration of most macro- and micronutrients, it can be assumed that the nutrient bioavailability from enteral feeds should be almost entirely related to feed tolerance and absorption and clinicians can focus on improving either of these as feasible. Equally, when feed intolerance is intractable, PN in standard doses may be a viable option. Most importantly, identifying the macro- and micronutrients that are most affected by the addition of an extracorporeal circuit to a critically ill patient may allow better nutritional surveillance and/or supplementation as appropriate. While the tested nutrients play an important role in health and disease, it is unlikely that the adverse clinical outcomes if any of circuit nutrient loss and benefits if any of supplementing at risk nutrients will ever be tested in a randomised trial design and is probably not of sufficient priority. However, minimising the biological burden from mechanical assist device therapy has the potential to improve patient outcomes, and this study is one step in that direction.

and benefits if any of supplementing at risk nutrients will ever be tested in a randomised trial design and is probably not of sufficient priority. However, minimising the biological burden from mechanical assist device therapy has the potential to improve patient outcomes, and this study is one step in that direction. Nutrient compounds possess unique physicochemical properties - size, charge and hydrophobicity vary significantly within and between each nutrient group [15]. Glucose is polar, ionic and hydrophilic; the amphipathic fatty acids possess both hydrophilic and lipophilic properties; and amino acids can be measured on a hydrophobicity index due to the varying solubilities of different amino acids in water and in varying pH conditions [15-17]. The majority of vitamins are hydrophilic with the exception of vitamins A, D, E and K, which are lipophilic; trace elements such as copper, manganese and zinc are positively charged cations; and selenium is a negatively charged anion [15]. No clear correlation could be established between physicochemical properties such as polarity, solubility or molecular weight and circuit behaviour of nutrients.

nd K, which are lipophilic; trace elements such as copper, manganese and zinc are positively charged cations; and selenium is a negatively charged anion [15]. No clear correlation could be established between physicochemical properties such as polarity, solubility or molecular weight and circuit behaviour of nutrients. In this setting, interpretation of the alterations in individual nutrient concentrations is challenging as several physical, chemical, environmental and other unknown factors may be at play. The decrement in arginine and glutamine concentrations may be attributed to their enzymatic conversion to ornithine and glutamate, respectively [18-21]. Similarly, the decrease in glucose concentrations over time may be explained in part by cellular metabolism. The reductions in concentrations of amino acids such as cysteine, isoleucine and alanine and vitamins (A and E) may be explained by instability, degradation, circuit sequestration or oxidative mechanisms [22]. It is possible that lipophilicity of vitamins A and E may affect their circuit disposition; however, this was not a consistent trend as another lipophilic vitamin D was relatively unaffected. We also hypothesise that an increase in concentrations of many other nutrients in a closed circuit probably represents breakdown of cellular and noncellular components of the whole blood and detailed evaluation of mechanisms responsible for alterations in individual nutrient concentrations is beyond the scope of this experiment.

pothesise that an increase in concentrations of many other nutrients in a closed circuit probably represents breakdown of cellular and noncellular components of the whole blood and detailed evaluation of mechanisms responsible for alterations in individual nutrient concentrations is beyond the scope of this experiment. Regardless of the cause, the net result is that some nutrients are significantly lost in the ECMO circuit and the clinical relevance of this remains unclear. While saturation of circuit binding sites over time may partly alleviate circuit loss of nutrients, the circuit loss from other mechanisms such as chemical and physical degradations, oxidation etc. may continue over the duration of ECMO, as up to second/third of patient's cardiac output may be exteriorised every minute for gas exchange and mechanical circulatory support [23]. Even though enteral or parenteral nutritional support and endogenous synthetic/metabolic mechanisms ensure an ongoing supply of macro- and micronutrients into the blood stream during ECMO, its unclear how much of their bioavailability is affected by ongoing losses in the circuit. In addition, the circuit loss of some nutrients can further be compounded by critical illness factors [24-26] such as inflammation, multiple organ dysfunction, fluid shifts and capillary leaks that may further affect their bioavailability. Interestingly, this may be the fate of many circulating endogenous compounds such as hormones and many other biologically active substances which are neither monitored nor replaced during ECMO. This calls for further clinical studies to further elucidate the biological impact of the ECMO circuit on circulating nutrients.

terestingly, this may be the fate of many circulating endogenous compounds such as hormones and many other biologically active substances which are neither monitored nor replaced during ECMO. This calls for further clinical studies to further elucidate the biological impact of the ECMO circuit on circulating nutrients. This study has limitations. We attempted to replicate the clinical scenario ex vivo, which allowed us to specifically study the interactions between the nutrients and the ECMO circuit under physiologic conditions in the absence of patient and pathological factors. This single-dose study, however, contrasts with the clinical scenario wherein there is ongoing delivery of nutrients into the blood stream and ECMO support may be continued for days to weeks. A continuous infusion of PN solution into the fully primed, noncompliant circuit was not feasible. Although a reservoir bladder which is seldom used in adult patients may have added compliance to the circuit, it could have further confounded our results by sequestering test compounds or creating areas of stasis, both of which are undesirable. The absence of circulating drugs, other blood components and metabolites in our experiment may also influence nutrient interactions with the components of the circuit. We did not test nutrient losses in physiologic controls (fresh human whole spiked with equivalent dose of PN solution, stored at 37°C over 24 h) to measure nutrient losses if any over time, independent of the ECMO circuit which to an extent limited the interpretation of our results.

with the components of the circuit. We did not test nutrient losses in physiologic controls (fresh human whole spiked with equivalent dose of PN solution, stored at 37°C over 24 h) to measure nutrient losses if any over time, independent of the ECMO circuit which to an extent limited the interpretation of our results. However, this study delineates the circuit behaviour of thirty-one vital nutrients and serves as a necessary first step in identifying at risk nutrients and identifies areas for further research. Conclusions Except for essential amino acid isoleucine and vitamins (A and E), most other macro- and micronutrients tested in this study were stable in the ECMO circuit over 24 h. Exteriorisation of large amounts of patient blood volume for ECMO is life-saving but may lead to circuit loss of vital nutrients. Critical illness may further exacerbate these circuit interactions of nutrients, and future studies should further evaluate altered nutrient disposition during ECMO and their potential impact on clinical outcomes. Abbreviations ECMOextracorporeal membrane oxygenation ENenteral nutrition MTEFEmicronutrient and trace elements with iron PNparenteral nutrition SEMstandard error of the mean TPNtotal parenteral nutrition Competing interests The authors declare that they have no competing interests. Authors’ contributions

Conclusions Except for essential amino acid isoleucine and vitamins (A and E), most other macro- and micronutrients tested in this study were stable in the ECMO circuit over 24 h. Exteriorisation of large amounts of patient blood volume for ECMO is life-saving but may lead to circuit loss of vital nutrients. Critical illness may further exacerbate these circuit interactions of nutrients, and future studies should further evaluate altered nutrient disposition during ECMO and their potential impact on clinical outcomes. Abbreviations ECMOextracorporeal membrane oxygenation ENenteral nutrition MTEFEmicronutrient and trace elements with iron PNparenteral nutrition SEMstandard error of the mean TPNtotal parenteral nutrition Competing interests The authors declare that they have no competing interests. Authors’ contributions KS, KE, ER and JF designed the study, obtained grant funding and ethics approval. KE, ER, CM and KS conducted the ex vivo experiment and collected blood samples. AB performed statistical analysis. KS, KE and ER analysed the data prepared the draft manuscript. All authors critically evaluated the manuscript and approved the final draft prior to submission.

obtained grant funding and ethics approval. KE, ER, CM and KS conducted the ex vivo experiment and collected blood samples. AB performed statistical analysis. KS, KE and ER analysed the data prepared the draft manuscript. All authors critically evaluated the manuscript and approved the final draft prior to submission. Acknowledgements This study was in part funded by a New Investigator Grant from The Prince Charles Hospital (TPCH) Foundation and the Critical Care Research Group. The Australian Red Cross Blood Service and Australian governments that fully funded the blood service for the provision of blood products and services to the Australian community. We thank the blood donors of Australia. We gratefully acknowledge the detailed assistance and expertise with the nutrient assays from Ms. Brooke Berry from Queensland Pathology Statewide Services (QPSS) at the Royal Brisbane and Women's Hospital. We thank TPCH pathology staff for their facilitation of sample storage and handling. We acknowledge TPCH Critical Care Services for usage of the simulation area in the ICU for the experiments. JFF acknowledges the Queensland Health Medical Research fellowship.

Background Acute intracranial pathophysiological events occurring at the time of acute subarachnoid hemorrhage (SAH) and their sequelae during the first days thereafter have recently received increased attention. Clinical and experimental work has long demonstrated that early brain injury (EBI) after SAH plays an important role in the disease pathophysiology [1-3]. A large clinical trial showed that reduction of delayed cerebral vasospasm failed to improve clinical outcomes [4]. Hence, interventions acting early in the disease course continued to gain in importance [5-7]. Despite increased research efforts in recent years, there is still relatively little known about what triggers pathophysiological mechanisms that result in EBI after SAH [8,9]. EBI is an umbrella term that embraces consequences of complex pathophysiological mechanisms that occur as a result of the initial bleed [8] and are unlikely to be solely responsible for early ischemic damage. It is evident that co-factors such as microvascular filling defects [10], breakdown of ionic homeostasis [11], blood brain barrier disruption, microarterial narrowing [12], and decreased bilateral regional cerebral blood flow (rCBF) worsen ischemia. To date, there has been little agreement on what triggers these processes which ultimately result in EBI after SAH.

scular filling defects [10], breakdown of ionic homeostasis [11], blood brain barrier disruption, microarterial narrowing [12], and decreased bilateral regional cerebral blood flow (rCBF) worsen ischemia. To date, there has been little agreement on what triggers these processes which ultimately result in EBI after SAH. It has been demonstrated that rapid and large increase in intracranial pressure (ICP) leads to more severe acute pathophysiologic (greater rCBF reduction) and histological changes (increased in Fluoro-jade B (FJB)- and terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL)-positive cells) after experimental SAH [13]. However, in this experimental setting, the extent of SAH was macroscopically more pronounced (reflected in a nearly four times higher hemoglobin concentration in the subarachnoid space basal brain areas) in animals with larger increase in ICP. And since subarachnoid blood per se is well known to cause direct brain damage, late rCBF reduction, and neuronal and astrocytic apoptosis independent of initial ICP increase [14-18], it still remains a matter of debate whether ICP increase or the extent of subarachnoid blood represents one of the main causes for increased EBI after SAH. In order to investigate how peak ICP, extent of subarachnoid blood, and hyperacute depletion of cerebral perfusion pressure (CPP) may correlate with the onset of EBI, we used a blood shunt SAH model to control and simulate various degrees of ICP increase.

It has been demonstrated that rapid and large increase in intracranial pressure (ICP) leads to more severe acute pathophysiologic (greater rCBF reduction) and histological changes (increased in Fluoro-jade B (FJB)- and terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL)-positive cells) after experimental SAH [13]. However, in this experimental setting, the extent of SAH was macroscopically more pronounced (reflected in a nearly four times higher hemoglobin concentration in the subarachnoid space basal brain areas) in animals with larger increase in ICP. And since subarachnoid blood per se is well known to cause direct brain damage, late rCBF reduction, and neuronal and astrocytic apoptosis independent of initial ICP increase [14-18], it still remains a matter of debate whether ICP increase or the extent of subarachnoid blood represents one of the main causes for increased EBI after SAH. In order to investigate how peak ICP, extent of subarachnoid blood, and hyperacute depletion of cerebral perfusion pressure (CPP) may correlate with the onset of EBI, we used a blood shunt SAH model to control and simulate various degrees of ICP increase. Methods A total of 21 3-month-old female New Zealand White rabbits weighing 2.4 to 4.3 kg were used in this study. The study was incorporated as a subproject of ongoing experimental studies and performed in accordance with the National Institutes of Health guidelines for the care and use of experimental animals and with the approval of the Animal Care Committee of the Canton of Bern, Switzerland (approval #107/09) [19]. The animals were housed in groups (two to four animals per cage) at 22°C to 24°C under a 12-h light-dark cycle with free access to food and tap water.

th guidelines for the care and use of experimental animals and with the approval of the Animal Care Committee of the Canton of Bern, Switzerland (approval #107/09) [19]. The animals were housed in groups (two to four animals per cage) at 22°C to 24°C under a 12-h light-dark cycle with free access to food and tap water. Study design Sixteen rabbits underwent various degrees of ICP-controlled (range 40 to 120 mmHg) SAH to generate a spectrum of ICP values as described in more detail below. Five animals served as sham-operated controls. All surgical procedures were performed under sterile conditions at the Experimental Surgical Institute, Department of Clinical Research, Bern University Hospital, Bern, Switzerland. A veterinary anesthesiologist monitored the animals during surgery and throughout anesthetic recovery.

s served as sham-operated controls. All surgical procedures were performed under sterile conditions at the Experimental Surgical Institute, Department of Clinical Research, Bern University Hospital, Bern, Switzerland. A veterinary anesthesiologist monitored the animals during surgery and throughout anesthetic recovery. Anesthesia, clinical observation, and sacrifice Induction of general anesthesia was performed by subcutaneous administration of ketamine (30 mg/kg; Ketalar, 50 mg/ml, Pfizer AG, Zurich, Switzerland) and xylazine (6 mg/kg; Xylapan, 20 mg/ml, Vetoquinol AG, Bern, Switzerland) and continued intravenously. Room air-enriched oxygen was provided to the spontaneously breathing animals. The animals underwent clinical observation during anesthetic recovery (first 3 h) and from then on every 6 h. Neurological status was graded at 3, 6, 12, 18, and 24 h post-SAH according to a four-point grading system [20]: grade 1, no neurological deficit; grade 2, minimal or suspected neurological deficit; grade 3, mild neurological deficit without abnormal movement; and grade 4, severe neurological deficit with abnormal movement. Euthanasia was performed 24 h post-SAH induction under the same anesthesia as previously described, by intra-arterial bolus injection of sodium thiopental (40 mg/kg) (Pentothal®, Ospedalia AG, Hünenberg, Switzerland).

al deficit without abnormal movement; and grade 4, severe neurological deficit with abnormal movement. Euthanasia was performed 24 h post-SAH induction under the same anesthesia as previously described, by intra-arterial bolus injection of sodium thiopental (40 mg/kg) (Pentothal®, Ospedalia AG, Hünenberg, Switzerland). SAH induction, instruments, and data acquisition Since our primary research question required an entire spectrum of various degrees of SAH, measured as ICP increase, we were dependent on tight control of ICP. Hence, we ultimately have chosen the blood shunt model which has been validated for that purpose in various species [21-24]. The model and techniques were used to induce SAH in rabbits as described previously [19]. Briefly, on day 0, the cisterna magna was punctured with a pediatric spinal access needle (22 G × 40 mm) and connected via a pressure tube and interposed three-way stopcock to the subclavian artery. The three-way stopcock was used for blood pressure measurement, as a blood sampling port, and to allow regulation of the bleeding. Neuromonitoring including an ICP monitor catheter tip (OLM Intracranial Pressure Monitoring Kit, Camino, Model 110-4B, Camino Laboratories, San Diego, CA, USA) and two laser-Doppler flowmetry fine needle probes (MNP110XP, 0.48-mm diameter, Oxford Optronix Ltd., Oxford, UK) which were positioned in the olfactory bulb and bilateral frontal lobe according to outer skull landmarks was done [25].

essure Monitoring Kit, Camino, Model 110-4B, Camino Laboratories, San Diego, CA, USA) and two laser-Doppler flowmetry fine needle probes (MNP110XP, 0.48-mm diameter, Oxford Optronix Ltd., Oxford, UK) which were positioned in the olfactory bulb and bilateral frontal lobe according to outer skull landmarks was done [25]. Standard cardiovascular monitoring (mean arterial blood pressure (MABP), heart rate, electrocardiogram, end-tidal CO2, and SaO2) was performed at a sampling rate of 100 Hz (Datex S5 Monitor, GE Medical Systems CH, Glattbrugg, Switzerland), and the data were transferred via the analog output interface to an analog-digital converter/data logger, stored (Biopac MP100 and acqKnowledge version 3.8.1; BIOPAC Systems, Inc., Goleta, CA, USA), and processed for pre-analysis using scripting software matlab (Mathworks 130 Inc., Natick, MA, USA). Pressures were zeroed to the level of the heart before and after each session, and pressure calibration of the AD converter and data-logging system was done once before the series started. Arterial blood gas status was analyzed (ABL 725, Radiometer, Copenhagen, Denmark) before SAH induction.

130 Inc., Natick, MA, USA). Pressures were zeroed to the level of the heart before and after each session, and pressure calibration of the AD converter and data-logging system was done once before the series started. Arterial blood gas status was analyzed (ABL 725, Radiometer, Copenhagen, Denmark) before SAH induction. Baseline values were measured during a time period of 6 min. SAH was initiated by opening the blood shunt to let blood stream into the atlanto-occipital cistern under arterial pressure. MABP and bilateral rCBF were recorded for 15 min after initiation of SAH. Closure of the stopcock interposed in the shunt allowed creation of various degrees of ICP increase (40 to 120 mmHg) and subsequent CPP depletion. Most of the animals were exposed sequentially to ‘Spontaneous SAH’ (n = 10). After opening of the shunt, ICP increases without any intervention until reaching a plateau. If this plateau phase was maintained for more than 10 s, the shunt was closed. The shunt was also closed if ICP decrease occurred spontaneously (no later than 30 s from start of the plateau phase; we therefore did not allow for potential rebleeding). Most of the animals with spontaneous SAH thus suffered severe SAH with high ICP values. In a minority of cases, spontaneous plateaus were reached at rather low ICP levels. These mild cases of SAH probably occurred due to premature thrombosis within the shunt system.

we therefore did not allow for potential rebleeding). Most of the animals with spontaneous SAH thus suffered severe SAH with high ICP values. In a minority of cases, spontaneous plateaus were reached at rather low ICP levels. These mild cases of SAH probably occurred due to premature thrombosis within the shunt system. However, to test our hypothesis, we needed the entire spectrum of different ICP values. Therefore, we performed ‘Controlled SAH’ to add missing SAH severities by means of ICP increase (n = 6). These animals were randomized to planed maximal ICP values. Target peak values were achieved by closure of the shunt during ICP increase within the range of minimal (40 mmHg) and most severe (120 mmHg) spontaneous ICP increase. An overview of SAH start and various stop procedures of shunt-induced SAH are given in Figure 1.Figure 1 Spontaneous and ICP controlled SAH. Scheme depicts start (down arrow) and stop (red X) procedures of shunt-induced SAH. (A) ‘Spontaneous SAH’. The shunt is closed after ICP either reached a plateau (10 s; short horizontal arrow) or started to decrease (30 s; long horizontal arrow). (B) ‘Controlled SAH’. The shunt is closed at any level prior to development of ICP plateau.

rt (down arrow) and stop (red X) procedures of shunt-induced SAH. (A) ‘Spontaneous SAH’. The shunt is closed after ICP either reached a plateau (10 s; short horizontal arrow) or started to decrease (30 s; long horizontal arrow). (B) ‘Controlled SAH’. The shunt is closed at any level prior to development of ICP plateau. Tissue processing, histology, and histochemistry Under general anesthesia, intracardiac perfusion-fixation was carried out on day 1 after SAH at room temperature with 400 ml of 0.1 M phosphate-buffered solution (PBS) followed by 400 ml fixative (4% paraformaldehyde in 0.1 M PBS, pH 7.3). The brains were removed from the skull, and the basal and hemispheric surfaces were analyzed to identify accumulated blood clots and distribution of subarachnoid blood. The severity of SAH was categorized as reported previously with slight modifications [26] as follows: 0: no blood; 1: minimal blood; 2: moderate blood clot (basal arteries visible); and 3: massive blood clot (visual obliteration of basal arteries). The summed score of each cistern (range: 0 to 12) determined the final grade of controls (0), mild (1 to 4), moderate (5 to 8), or severe (9 to 12) SAH (Additional file 1: Figure S1).

d; 1: minimal blood; 2: moderate blood clot (basal arteries visible); and 3: massive blood clot (visual obliteration of basal arteries). The summed score of each cistern (range: 0 to 12) determined the final grade of controls (0), mild (1 to 4), moderate (5 to 8), or severe (9 to 12) SAH (Additional file 1: Figure S1). The brains were then immersed in the fixative overnight and cryoprotected at 4°C followed by immersion in 15% sucrose in 0.1 M PBS. The brains were cut into four blocks between the forebrain (olfactory bulb) and cerebellum, embedded in paraffin, and cut into consecutive 7-μm sections. The cut surface of block one was placed through the cortical punch defect of the ICP and rCBF probes. The first section of blocks two to four was stained with hematoxylin and eosin, and the most representative fields containing the hippocampus and basal cortex (BC) were selected for additional cuts of nine consecutive sections used for histochemical analysis. Cells with damaged deoxyribonucleic acid (DNA) and neurodegeneration secondary to ischemia were detected using the TUNEL method (Roche Diagnostics AG, Rotkreuz, Switzerland) and FJB (Millipore AG, Zug, Switzerland). Nuclei were counterstained with 4',6-diamidino-2-phenylindole (DAPI) (Roche Diagnostics AG, Basel, Switzerland). An observer (L.A.) blind to sample identity counted the number of co-localized cells in regions of interest defined on coronal sections for each hemisphere in the basal cortex (0.9 mm2) and along the hippocampal cornu ammonis regions CA1 and CA3 (each 0.9 mm).

ylindole (DAPI) (Roche Diagnostics AG, Basel, Switzerland). An observer (L.A.) blind to sample identity counted the number of co-localized cells in regions of interest defined on coronal sections for each hemisphere in the basal cortex (0.9 mm2) and along the hippocampal cornu ammonis regions CA1 and CA3 (each 0.9 mm). Data analysis and statistical methods Statistical analysis was performed using IBM SPSS statistical software version 20.0 (IBM Corp., New York, NY, USA) and processed for pre-analysis using Matlab scripting software (Mathworks Inc., Natick, MA, USA). Areas under the curve (AUC) were calculated based on the trapezoidal rule on the 100-Hz acquisition data set before any further post-processing. ANOVA regression analysis was used for calculation of correlations between effects of SAH on hemodynamics, rCBF, ICP, CPP, number of FJB- and TUNEL-positive cells, and the amount of subarachnoid blood (SAH blood score), respectively. The time buckets chosen for analyses (1.5, 3, and 6 min after SAH) were predetermined (a priori) based on the pathophysiology of the rabbit shunt model (peak ICP increase within 1 to 2 min, maximal CPP depletion within the first 3 min, ICP steady-state values within 5 to 10 min) [19]. Based on the relative CPP depletion during the first 3 min of SAH, the animals were assigned post hoc to one of three groups of hemorrhage severity: mild (n = 5; relative CPP < 0.25), moderate (n = 3; relative CPP > 0.25 but < 0.4), and severe (n = 4; relative CPP > 0.4). The groups' mean cell counts of FJB- and TUNEL-positive cells of both hemispheres were then compared among those and to the control group using one-way ANOVA and Bonferroni post hoc testing. Values are expressed as means of each group ± SD. Data from neurological deficit scores and subarachnoid blood scores are given as median and range. A probability value of <0.05 was considered statistically significant. The strength of linear correlations between the variables was expressed by the linear regression coefficient (r) and its squared value r2.

group ± SD. Data from neurological deficit scores and subarachnoid blood scores are given as median and range. A probability value of <0.05 was considered statistically significant. The strength of linear correlations between the variables was expressed by the linear regression coefficient (r) and its squared value r2. Results Mortality, morbidity, and neurological status The mortality of SAH animals was 25% (4 out of 16 rabbits). All of the animals that died were allocated to spontaneous SAH and died shortly after initiation of the bleeding due to respiratory arrest and severe bradycardia. Surviving animals (total n = 12; n = 6 ‘Spontaneous SAH’ and n = 6 ‘Controlled SAH’) demonstrated a uniform early post-SAH clinical course, with slow recovery within the first 3 h (median neurological score: 3; range 2 to 4). Most SAH animals (n = 7) recovered completely within 24 h post-surgery (median neurological score: 1; range 1 to 4). Control animals showed uneventful recovery within 3 to 6 h.

’) demonstrated a uniform early post-SAH clinical course, with slow recovery within the first 3 h (median neurological score: 3; range 2 to 4). Most SAH animals (n = 7) recovered completely within 24 h post-surgery (median neurological score: 1; range 1 to 4). Control animals showed uneventful recovery within 3 to 6 h. Interplay between MABP, ICP, CPP, and rCBF All SAH animals demonstrated a rapid increase in ICP with a corresponding marked decrease in CPP and rCBF. The increase of MABP during the peak phase of ICP (Cushing reflex) was more pronounced than the further increase in ICP, resulting in rapid recovery of CPP to 81.5% ± 12% of baseline values within 15 min. The mean relative rCBF depletion of both hemispheres was significantly correlated to reduction in CPP during the initial 15 min after SAH in a linear regression pattern (reg coeff r = 0.82, r2 = 0.68, p < 0.001). Time course and individual values of ICP, MABP, CPP, and local rCBF are given in Figure 2.Figure 2 ICP, MABP, CPP and rCBF time-course and values of all SAH animals. All SAH animals demonstrated a rapid increase in ICP (A). Marked increase of MABP (B) occurred during the peak phase of CPP (C) and rCBF (D) depletion. Mean values ±SD are presented in (E). ICP, MABP, and CPP values are displayed in mmHg. rCBF of the right hemisphere is given as mean relative baseline values within 15 min.

SAH animals demonstrated a rapid increase in ICP (A). Marked increase of MABP (B) occurred during the peak phase of CPP (C) and rCBF (D) depletion. Mean values ±SD are presented in (E). ICP, MABP, and CPP values are displayed in mmHg. rCBF of the right hemisphere is given as mean relative baseline values within 15 min. Gross examination of brain and histology There were no complications related to wound healing, cerebrospinal fluid leakage, or infections along the frontal osteotomy sites, subclavian skin incision, or nuchal cisternal injection point. Twenty-four hours post-SAH, all surviving rabbits (n = 12) demonstrated extensive coagulated diffuse subarachnoid blood in the chiasmatic and pre-chiasmatic cisterns (I: 2; 1 to 3), basal cistern (II: 3; 2 to 3), prepontine and interpeduncular cisterns (III: 3; 1 to 3), and cistern magna (IV: 2; 1 to 3), resulting in moderate (n = 2, 7; 7), and severe (n = 10, 10 to 12) grades of SAH. No subarachnoid blood was observed in control animals (n = 5). Cells with DNA damage and neurodegeneration by means of TUNEL- and FJB-positive cells were detected in the basal cortex regions and the hippocampus (CA1 and CA3) of both hemispheres in all animals. Merged co-localization with DAPI confirmed that TUNEL- and FJB-positive staining was generally located in the nucleus (Figures 3 and 4). Animals exposed to mild CPP depletion showed no differences in the number of cells with DNA damage and neurodegeneration irrespective of location when compared with the control animals (Figures 3 and 4). However, those animals exposed to moderate CPP depletion demonstrated significantly more TUNEL- and FJB-positive stained cells in the hippocampus formation as well as in the BC region than the mild SAH group or the control animals. The mean cell counts of the animals that suffered severe SAH did not statistically differ from those animals exposed to moderate CPP depletion, irrespective of location (Additional file 2: Tables S1 and S2).Figure 3 Cells with DNA damage in hippocampus and basal cortex. Bar graphs demonstrate mean TUNEL-positive cell count of both hemispheres (±SD) subdivided into controls (n = 5) and different degrees of CPP depletion (mild, n = 5; moderate, n = 3; severe, n = 4). Quantification was performed by counting TUNEL-positive cells co-localizing with DAPI (blue) in regions of interest in the hippocampus (A) and basal cortex (B).

ive cell count of both hemispheres (±SD) subdivided into controls (n = 5) and different degrees of CPP depletion (mild, n = 5; moderate, n = 3; severe, n = 4). Quantification was performed by counting TUNEL-positive cells co-localizing with DAPI (blue) in regions of interest in the hippocampus (A) and basal cortex (B). Histochemistry (C): Co-localization with DAPI (left column) confirmed that TUNEL-positive staining (middle column) was generally in the nucleus (right column). Hollow arrows show DAPI-positive nuclear staining (blue). Filled arrows identify TUNEL-positive cells. Scale bars = 50 μm. *p < 0.05. **p < 0.01. Figure 4 Neurodegeneration in hippocampus and basal cortex . Bar graphs demonstrate mean FJB-positive cell count of both hemispheres (±SD) subdivided into controls (n = 5) and different degrees of CPP depletion (mild, n = 5; moderate, n = 3; severe, n = 4). Quantification was performed by counting FJB-positive cells co-localizing with DAPI (blue) in regions of interest in the hippocampus (A) and basal cortex (B). Histochemistry (C): Co-localization with DAPI (left column) confirmed that FJB-positive staining (middle column) was generally in the nucleus (right column). Hollow arrows show DAPI-positive nuclear staining (blue). Filled arrows identify FJB-positive cells. Scale bars = 50 μm. *p < 0.05, **p < 0.01.

basal cortex (B). Histochemistry (C): Co-localization with DAPI (left column) confirmed that FJB-positive staining (middle column) was generally in the nucleus (right column). Hollow arrows show DAPI-positive nuclear staining (blue). Filled arrows identify FJB-positive cells. Scale bars = 50 μm. *p < 0.05, **p < 0.01. Correlation between CCP depletion, SAH blood score, peak ICP, and EBI A significant positive linear correlation between CPP reduction within the first 3 min after SAH and the total number of TUNEL- and FJB-positive cells (means of left and right hemispheres) was found in CA1 and CA3 regions (r2 = 0.51, p < 0.01 for the FJB-positive cells and r2 = 0.35, p < 0.05 for the TUNEL-positive cells, respectively) as well as in the basal cortex region for the TUNEL-positive cells (r2 = 0.58, p < 0.01). There was no linear correlation, however, between the relative CPP's area under the curve within the first 3 min and FJB-positively stained cells in the basal cortex region (r2 = 0.24, p > 0.1). The more severe the temporary CPP reduction, the more pronounced seemed the neuronal cell death and neurodegeneration (Figure 5).Figure 5 Correlation between hyperacute CPP reduction/peak ICP and early brain injury . A significant positive linear correlation between CPP reduction within the first 3 min after SAH and the total number of the mean of left and right hemispheres TUNEL-positive cells was found in both the basal cortex region (A) as well as the hippocampus formation (B), revealing that neuronal cell death and neurodegeneration are linked to the severity of temporary CPP shortage during the first 3 min of SAH. Peak ICP did not significantly correlate with the number of cells with DNA neither in the basal cortex (C) nor in the hippocampal (D) regions. Graphs include the 95% confidence intervals.

g that neuronal cell death and neurodegeneration are linked to the severity of temporary CPP shortage during the first 3 min of SAH. Peak ICP did not significantly correlate with the number of cells with DNA neither in the basal cortex (C) nor in the hippocampal (D) regions. Graphs include the 95% confidence intervals. There were no linear correlations between the SAH blood score and the relative CPP depletion within the first 3 min and between the SAH blood score and relative CPP at the time of maximal depletion (Additional file 1: Figure S2). There was also no linear correlation between SAH blood score and the total amount of TUNEL- or FJB-positive cells (means of both hemispheres). These findings hold true for both the hippocampus formation and basal cortex region (Additional file 1: Figure S3). Peak ICP did not significantly correlate with any of our neuronal cell damage parameters (p > 0.05). Discussion The results of this study demonstrate that the more severe the CPP shortage during the hyperacute phase of SAH, the more pronounced was the number of cells with DNA damage and the degree of neurodegeneration in the hippocampal region and the number of cells with DNA damage in the basal cortex 24 h after experimental SAH. The findings of early neuronal damage were independent of peak ICP and the amount of subarachnoid blood. Using the ICP-controlled shunt model, the results extend prior findings [13-18] and suggest that CPP depletion at the time of SAH potentially triggers processes that eventually result in EBI after SAH.

rimental SAH. The findings of early neuronal damage were independent of peak ICP and the amount of subarachnoid blood. Using the ICP-controlled shunt model, the results extend prior findings [13-18] and suggest that CPP depletion at the time of SAH potentially triggers processes that eventually result in EBI after SAH. Although significant, the correlation between hippocampal TUNEL- and FJB-positive cells (means of left and right hemispheres) and hyperacute CPP depletion was weak (r2 = 0.35 and r2 = 0.51). In the basal cortex region, the correlation between CPP shortage within the first 3 min and TUNEL-positive cells was slightly stronger (r2 = 0.58) but did not reach statistical significance for FJB-positive cells. One explanation for the absent correlation of FJB-positive cells could be a mismatch of TUNEL and FJB staining of neurons that underwent oxidative stress [13,27]. Another explanation can be the lack of specificity of FJB [28]. Additional detection of apoptosis by measurement of apoptosis-related proteins would have improved the quantification of cell death, and double labeling of FJB with NeuN would have increased the specificity for detection of neuronal degeneration.

[13,27]. Another explanation can be the lack of specificity of FJB [28]. Additional detection of apoptosis by measurement of apoptosis-related proteins would have improved the quantification of cell death, and double labeling of FJB with NeuN would have increased the specificity for detection of neuronal degeneration. Due to animal welfare regulations, the study was performed with adult female rabbits only. Using female animals carries the risk that estrogens may attenuate SAH-induced apoptosis [29]. The amount of neuronal cell death may also depend on gender and chosen injection anesthesia. It remains unknown whether and to what extent the use of ketamine compromised pathophysiological parameters and neuronal cell death after SAH. Nevertheless, the impact of moderate CPP compromise on cells with DNA damage is highlighted by the differences observed between SAH animals and controls. Contribution of acute global cerebral ischemia to EBI after SAH Clinical observations have long emphasized the important relationship between increase in ICP within the first minutes after aneurysm rupture and occurrence of cerebral ischemia [30]. Although we found a positive correlation between CPP reduction and EBI, it is unlikely that the temporary perfusion shortage is solely responsible for the early neuronal damage detected.

ortant relationship between increase in ICP within the first minutes after aneurysm rupture and occurrence of cerebral ischemia [30]. Although we found a positive correlation between CPP reduction and EBI, it is unlikely that the temporary perfusion shortage is solely responsible for the early neuronal damage detected. The present data demonstrate that neurodegeneration and neuronal cell death occurred not only in animals with severe acute global cerebral ischemia but also in subjects with moderate CPP reduction. Furthermore, neuronal cell death was equally detected in regions that are less susceptible to ischemia (basal cortex) and regions most vulnerable to ischemic stress (hippocampus). Thus, it can be hypothesized that in addition to temporary CPP reduction during hyperacute SAH, other mechanisms - probably triggered by initial ICP increase and subsequent CPP depletion - are likely responsible for focal and global perfusion deficits and subsequent ischemic damage in the early phase after SAH.

ampus). Thus, it can be hypothesized that in addition to temporary CPP reduction during hyperacute SAH, other mechanisms - probably triggered by initial ICP increase and subsequent CPP depletion - are likely responsible for focal and global perfusion deficits and subsequent ischemic damage in the early phase after SAH. Processes such as acute vasoconstriction (large and small parenchymal vessels) [12,31], perivascular swelling (intra- and extracellular edema) [32], microvascular filling defects (detachment of endothelial cells, platelet aggregation, and microthrombosis) [33], breakdown of ionic homeostasis (increase in extracellular glutamate and cortical spreading depression) [11], and decreased cerebral blood flow (CBF) (probably as secondary response to decreased cerebral metabolic rate and decreased spontaneous electrical activity) are all likely to worsen ischemia. It remains largely unexplored to what extent these mechanisms contribute to EBI and how they are connected among each other. There were no differences in the number of cells with DNA damage and neurodegeneration between animals that suffered mild CPP challenge (relative CPP < 0.25) and control animals. Although the number of animals in the post hoc stratified groups is small, the analysis suggests that the threshold for triggering early brain injury lies between 25% and 40% of CPP depletion during the hyperacute phase of SAH.

between animals that suffered mild CPP challenge (relative CPP < 0.25) and control animals. Although the number of animals in the post hoc stratified groups is small, the analysis suggests that the threshold for triggering early brain injury lies between 25% and 40% of CPP depletion during the hyperacute phase of SAH. Main triggers of mechanisms that result in EBI after SAH To date, there has been little discussion about the triggering event for the mechanisms eventually leading to EBI after SAH. The mechanisms that are made responsible for EBI are believed to be activated as early as the aneurysm rupture. Blood streams under arterial pressure into a closed cranium causing rapid rise in ICP and marked CPP depletion which in turn reduces CBF. Experimental clarification of whether an initial global increase in ICP (respectively CPP reduction) or extravasated blood triggers the early pathophysiological sequelae causing EBI was inherently complicated by the lack of models that allow precise control of ICP, or CPP reduction, during acute SAH. In a comparison of two different SAH rat models, a rapid and large increase in ICP led to more severe acute pathophysiologic (decrease in rCBF) and histological changes (increased in FJB- and TUNEL-positive cells) than minor changes in ICP [13]. However, macroscopic examination also demonstrated significantly higher amounts of subarachnoid hemoglobin in the group with the greater ICP increase and therefore did not confirm whether ICP increase or the amount of subarachnoid blood represents the main cause of acute SAH sequelae.

s) than minor changes in ICP [13]. However, macroscopic examination also demonstrated significantly higher amounts of subarachnoid hemoglobin in the group with the greater ICP increase and therefore did not confirm whether ICP increase or the amount of subarachnoid blood represents the main cause of acute SAH sequelae. It has been demonstrated that subarachnoid blood can cause direct brain damage, late rCBF reduction, and neuronal and astrocytic apoptosis independent of initial ICP increase [14-18]. However, we could not establish any relationship between the amount of subarachnoid blood and the degree of early (24 h) neuronal cell damage, either in close proximity to the brain surface (basal cortex) or in deep brain regions (hippocampus). A potential explanation could be that even SAH of minor extent can cause acute vasoconstriction, marked CPP, and subsequent rCBF decline [31]. Initial increase in ICP is considered to play an important role in the pathophysiology of EBI [5]. Prior studies have noted that a rapid increase in ICP triggers sympathetic nerve activity [34] decreases CBF, and upregulates contractile receptors in cerebral arteries [31] and therefore potentially can cause brain damage [14,34]. Despite these findings, we were not able to demonstrate a relationship between the peak ICP and the degree of neurodegeneration and number of cells with DNA damage found 24 h after SAH.

creases CBF, and upregulates contractile receptors in cerebral arteries [31] and therefore potentially can cause brain damage [14,34]. Despite these findings, we were not able to demonstrate a relationship between the peak ICP and the degree of neurodegeneration and number of cells with DNA damage found 24 h after SAH. CPP depletion as parameter for the severity of EBI after SAH A possible explanation for the missing correlation between peak ICP and brain damage might be that peak ICP does not reflect the actual perfusion shortage during the hyperacute phase of SAH. Individual variations in baseline MABP and intensity of Cushing reflex at the time of bleeding influenced CPP depletion during SAH. CPP challenge was best represented by relative CPP reduction during the first minutes of SAH but not by peak ICP.

reflect the actual perfusion shortage during the hyperacute phase of SAH. Individual variations in baseline MABP and intensity of Cushing reflex at the time of bleeding influenced CPP depletion during SAH. CPP challenge was best represented by relative CPP reduction during the first minutes of SAH but not by peak ICP. The amount of subarachnoid blood also seems not to be reliable for assessment of severity of EBI after SAH. Within the very first phase after the onset of experimental SAH, large blood volumes stream into the subarachnoid space [22]. After a phase of compensatory mechanisms, any further small change in intracranial volume then causes rapid increase in ICP, as well as a large CPP reduction, dependent on the degree of the Cushing reflex. This could explain why we recorded moderate and severe SAH blood scores even in animals with only minor CPP challenges. In addition, one needs to keep in mind that the degree of cisternal SAH is performed by visual inspection. It would have been favorable to either quantify the amount of SAH by determining the hemoglobin concentration in various cisterns [13] or to directly measure the blood flow in the shunt during acute SAH [22].

ges. In addition, one needs to keep in mind that the degree of cisternal SAH is performed by visual inspection. It would have been favorable to either quantify the amount of SAH by determining the hemoglobin concentration in various cisterns [13] or to directly measure the blood flow in the shunt during acute SAH [22]. It also has to be acknowledged that the significant correlation between hemodynamic events and signs of neuronal damage in our study does not allow drawing causal conclusion. Our study has rather hypothesis-generating character. In this respect, we believe that also the lack of correlation between peak ICP/amount of subarachnoid blood and degree of early neuronal cell damage is an important information. It would be of much interest to verify the presented results by a modified study design in which different ICP values are generated with the same amount of blood. The blood injection technique could generate various degrees of peak ICP (by varying the injection time) with the same amount of blood; however, the ICP profiles would differ significantly among animals. Conclusions The severity of EBI in terms of neuronal cell death and neurodegeneration correlates with the degree of hyperacute CPP challenge. Initial global ischemia, however, is not solely responsible for EBI. These results suggest that other processes, potentially triggered by hyperacute CPP depletion, play a major role in the onset of EBI. The total amount of subarachnoid blood and peak ICP failed as a surrogate marker for the severity of EBI.

challenge. Initial global ischemia, however, is not solely responsible for EBI. These results suggest that other processes, potentially triggered by hyperacute CPP depletion, play a major role in the onset of EBI. The total amount of subarachnoid blood and peak ICP failed as a surrogate marker for the severity of EBI. Additional files Additional file 1 This file contains supplementary figures. Additional file 2 This file contains supplementary tables. Serge Marbacher and Volker Neuschmelting contributed equally to the study. Competing interests The authors declare that they have no competing interests. Authors' contribution Author contributions to the study and manuscript preparation include the following. Conception and design of the study were done by SM, VN, and HRW. Acquisition of data was performed by LA, VN, and SM. Analysis and interpretation of data were carried out by SM, VN, and SMJ. Statistical analysis was done by VN and SM. The draft of the article was performed by SM and VN. Critical revision of the article was done by SMJ, JF, MvG, HRW, and JT. Administrative and technical/material support was provided by HRW and MvG. Study supervision was done by JF. All authors read and approved the final manuscript.

al analysis was done by VN and SM. The draft of the article was performed by SM and VN. Critical revision of the article was done by SMJ, JF, MvG, HRW, and JT. Administrative and technical/material support was provided by HRW and MvG. Study supervision was done by JF. All authors read and approved the final manuscript. Acknowledgements We appreciate the skillful management of animal care, anesthesia, and operative assistance of Daniel Mettler, DVM; Max Müller, DVM; Daniel Zalokar; and Olgica Beslac, Experimental Surgical Institute, Department of Clinical Research, Bern University Hospital, Bern, Switzerland. We would like to thank Michael Lensch, Head Research Nurse, Department of Intensive Care Medicine, Bern University Hospital and University of Bern, Bern, Switzerland, for real-time data monitoring and extensive post-processing of the physiological parameters. We express our gratitude to Andreas Raabe, MD, PhD, and Angelique Ducray, PhD, Department of Neurosurgery, Neurosurgical Research Institute, Bern University Hospital and University of Bern, Bern, Switzerland, for their technical laboratory support. We are indebted to Regula Markwalder, MD; Karin Portmann; and Mengia Berthold, Institute of Pathology, Länggasse, Bern, Switzerland, for their advice and support in histopathology and histochemistry. This work was supported by the following: for manpower (neuromonitoring) and technical laboratory support (kits for FJB and TUNEL staining): the Department of Intensive Care Medicine, Bern University Hospital and University of Bern, Bern, Switzerland and Department of Clinical Research, University of Bern, Bern, Switzerland; for histopathological and histochemical advice and support: Institute of Pathology, Länggasse, Bern; and for animals/keeping, OP team/room: The Research Fund from the Cantonal Hospital Aarau, Aarau, Switzerland.

Background Intravascular catheters are widely used and are indispensable for proper patient management. However, its use implies several risks, which are mainly infectious. Catheter-related bloodstream infection (C-RBSI) is the more important entity which is associated to high rates of morbidity, mortality, and sanitary costs [1]. In general, in those catheters that are inserted for more than 8 days, infection is mainly acquired by an intra-luminal route (66%) because of hub contamination. Hub colonization after manipulation is responsible for 29% to 60% of catheter-related infections (CRI) [2]. Recently, it has been shown that the use of closed needleless connectors is effective against microorganism penetration by hub contamination. Some studies tested the in vitro efficacy of different needleless access devices to prevent the ingress of microorganisms [3-8]. However, these studies have heterogeneous designs, and they are all performed by instilling high concentrations of microorganisms through the catheter. Our study purpose was to create an in vitro model lasting long enough be used to compare various connectors simulating the real daily handling of these devices based on blood culture bottles. Methods Setting This in vitro study has being carried out in the Laboratory of the Clinical Microbiology and Infectious Disease Department of the Hospital General Universitario Gregorio Marañón.

Our study purpose was to create an in vitro model lasting long enough be used to compare various connectors simulating the real daily handling of these devices based on blood culture bottles. Methods Setting This in vitro study has being carried out in the Laboratory of the Clinical Microbiology and Infectious Disease Department of the Hospital General Universitario Gregorio Marañón. Study design The model consisted of 40 blood culture bottles with a peripheral venous cannula inserted under sterile conditions and left in place (Figure 1). In each bottle, a disinfectable needle-free closed connector (CLAVE™, CareFusion, Spain) was used to close the cannulas. Twice a day, each cannula was manipulated while instilling 1 mL of either sterile saline or propofol. Manipulation of the bottles was divided into four models: ten bottles (five saline, five propofol) were manipulated with clean gloves and disinfecting the connector with alcohol (controls), ten bottles (five saline, five propofol) were manipulated without gloves (hands), ten bottles (five saline, five propofol) were manipulated with gloves impregnated with a 0.5 McFarland ATCC 29213 Staphylococcus aureus solution, and ten bottles (five saline, five propofol) were manipulated with gloves impregnated with a 0.05 McFarland ATCC 29213 S. aureus solution. Additional photograph files show this in more detail (see Additional files 1 and 2).Figure 1 Experimental model of the blood culture with the cannula and the CLAVE™.

ureus solution, and ten bottles (five saline, five propofol) were manipulated with gloves impregnated with a 0.05 McFarland ATCC 29213 S. aureus solution. Additional photograph files show this in more detail (see Additional files 1 and 2).Figure 1 Experimental model of the blood culture with the cannula and the CLAVE™. Before handling the connectors in the hand model, a base count and phenotypic identification of the colonizing microorganisms on the manipulators' hands were performed until the bottles became positive. Laboratory procedure The bottles were incubated in a BACTEC 9120 System (Becton Dickinson Microbiology Systems, Maryland, DE, USA) up to 10 days at 37°C under continuous agitation. Each day, it was observed whether there was positivity in the bottle fluid (alert from the BACTEC 9120), and in case it occurred, 100 μL of the fluid was cultured into the blood, MacConkey, and Brucella agar and incubated under aerobic and anaerobic conditions for 48 h at 37°C. Microorganism identification was performed using standard procedures [9]. At the end of the incubation time (10 days, 20 mL), the negative bottles were also tested by culturing 100 μL of the fluid into the blood, MacConkey, and Brucella agar and incubated under aerobic and anaerobic conditions for 48 h at 37°C. The different study variables were annotated in a data collection form.

Laboratory procedure The bottles were incubated in a BACTEC 9120 System (Becton Dickinson Microbiology Systems, Maryland, DE, USA) up to 10 days at 37°C under continuous agitation. Each day, it was observed whether there was positivity in the bottle fluid (alert from the BACTEC 9120), and in case it occurred, 100 μL of the fluid was cultured into the blood, MacConkey, and Brucella agar and incubated under aerobic and anaerobic conditions for 48 h at 37°C. Microorganism identification was performed using standard procedures [9]. At the end of the incubation time (10 days, 20 mL), the negative bottles were also tested by culturing 100 μL of the fluid into the blood, MacConkey, and Brucella agar and incubated under aerobic and anaerobic conditions for 48 h at 37°C. The different study variables were annotated in a data collection form. Statistical analysis The qualitative variables appear with their frequency distribution. The quantitative variables are summarized as the median with interquartile range (IQR). Continuous variables were compared using the median test for non-normally distributed variables. The chi-square or Fisher exact test was used to compare categorical variables. The Kaplan-Meier survival curve was used to compare the time to positivity with the degree of contamination. All statistical tests were two-tailed. Statistical significance was set at p < 0.05 for all the tests. The statistical analysis was performed with SPSS 18.0.

Statistical analysis The qualitative variables appear with their frequency distribution. The quantitative variables are summarized as the median with interquartile range (IQR). Continuous variables were compared using the median test for non-normally distributed variables. The chi-square or Fisher exact test was used to compare categorical variables. The Kaplan-Meier survival curve was used to compare the time to positivity with the degree of contamination. All statistical tests were two-tailed. Statistical significance was set at p < 0.05 for all the tests. The statistical analysis was performed with SPSS 18.0. Ethics This experimental design does not include human subjects and does not use human tissue or samples, so it was exempt from approval of the local ethics committee.

All statistical tests were two-tailed. Statistical significance was set at p < 0.05 for all the tests. The statistical analysis was performed with SPSS 18.0. Ethics This experimental design does not include human subjects and does not use human tissue or samples, so it was exempt from approval of the local ethics committee. Results and discussion Overall, all bottles in the control group were negative at the end of the incubation time. In contrast, almost all bottles (38/40) in the three contamination experiments were positive during the incubation time. In the hand model with saline, we recovered the same Staphylococcus epidermidis strain in four out of the five bottles, which was phenotypically coincident with that isolated from the manipulators' hands. In the remaining bottle, we recovered a Staphylococcus warneri strain which was also the same strain we isolated from the manipulators' hands. In the hand model with propofol, all bottles recovered a Micrococcus sp., which was phenotypically coincident with that isolated from the manipulators' hands. In the remaining models with S. aureus with both saline and propofol, the ATCC 29213 S. aureus was recovered in 18 out of the 20 bottles (Table 1).Table 1 Characteristics of the experiment regarding contamination model and type of instillation fluid Contamination degree Instillation fluid Colonizing microorganisms MTP (IQR), h Manipulator hand flora

Results and discussion Overall, all bottles in the control group were negative at the end of the incubation time. In contrast, almost all bottles (38/40) in the three contamination experiments were positive during the incubation time. In the hand model with saline, we recovered the same Staphylococcus epidermidis strain in four out of the five bottles, which was phenotypically coincident with that isolated from the manipulators' hands. In the remaining bottle, we recovered a Staphylococcus warneri strain which was also the same strain we isolated from the manipulators' hands. In the hand model with propofol, all bottles recovered a Micrococcus sp., which was phenotypically coincident with that isolated from the manipulators' hands. In the remaining models with S. aureus with both saline and propofol, the ATCC 29213 S. aureus was recovered in 18 out of the 20 bottles (Table 1).Table 1 Characteristics of the experiment regarding contamination model and type of instillation fluid Contamination degree Instillation fluid Colonizing microorganisms MTP (IQR), h Manipulator hand flora Handsa Saline S. epidermidis, S. warnerii 76.25 (65.92 to 125.01) S. epidermidis, S. warnerii, CoNS, SV, P. acnes, Micrococcus sp. Propofol Micrococcus sp. 74.91 (51.45 to 99.00) S. epidermidis, S. warnerii, CoNS, SV, P. acnes, Micrococcus sp. 0.5 MF SAb Saline MSSA (ATCC 29213) 104.92 (45.97 to 239.94) NA Propofol MSSA (ATCC 29213) 147.89 (66.04 to 236.58) NA 0.05 MF SAc Saline MSSA (ATCC 29213) 240.20 (154.82 to 360.00) NA Propofol MSSA (ATCC 29213) 66.11 (58.01 to 69.11) NA Controld Saline NA NA NA Propofol NA NA NA

ii, CoNS, SV, P. acnes, Micrococcus sp. 0.5 MF SAb Saline MSSA (ATCC 29213) 104.92 (45.97 to 239.94) NA Propofol MSSA (ATCC 29213) 147.89 (66.04 to 236.58) NA 0.05 MF SAc Saline MSSA (ATCC 29213) 240.20 (154.82 to 360.00) NA Propofol MSSA (ATCC 29213) 66.11 (58.01 to 69.11) NA Controld Saline NA NA NA Propofol NA NA NA aManipulation of the connector without gloves. bManipulation of the connector with gloves impregnated with a 0.5 McFarland solution of Staphylococcus aureus ATCC 29213. cManipulation of the connector with gloves impregnated with a 0.05 McFarland solution of Staphylococcus aureus ATCC 29213. dStandard sterile manipulation with clean gloves and with disinfection of connectors using alcohol. MTP, median time to positivity; IQR, interquartile range; SA, Staphylococcus aureus; MF, McFarland; CoNS, coagulase-negative staphylococci.

impregnated with a 0.05 McFarland solution of Staphylococcus aureus ATCC 29213. dStandard sterile manipulation with clean gloves and with disinfection of connectors using alcohol. MTP, median time to positivity; IQR, interquartile range; SA, Staphylococcus aureus; MF, McFarland; CoNS, coagulase-negative staphylococci. Regarding the median (IQR) time to positivity, there were no differences in either the hand model (saline 76 h (65.92 to 125.01) vs. propofol 75 h (51.45 to 99.00), p > 0.05) or in the 0.5 McFarland S. aureus model (saline 105 h (45.97 to 239.94) vs. propofol 148 h (66.04 to 236.58), p > 0.05). In contrast, in the 0.05 McFarland S. aureus model, the median (IQR) time to positivity was significantly higher when instilling saline than when instilling propofol: 240 h (154.82 to 360.00) vs. 66 h (58.01 to 69.11), p = 0.008. Moreover, when comparing the median (IQR) time to positivity between the three models with saline, the 0.05 McFarland S. aureus model was also significantly higher than the hands model: 240 h (154.82 to 360.00) vs. 76 h (65.95 to 125.00), p = 0.016 (Table 2, Figure 2).Table 2 Median time to positivity of the bottles comparing different contamination models and different fluids Fluid MTP (IQR), h p Hands 0.5 MF SA 0.05 MF SA Saline 76.25 (65.95 to 125.00) 104.92 (45.96 to 239.94) 240.20 (154.82 to 360.00) 0.041 a Propofol 74.91 (51.44 to 98.99) 147.89 (66.03 to 236.58) 66.11 (58.01 to 69.11) 0.24 p 0.69 1.00 0.008

Regarding the median (IQR) time to positivity, there were no differences in either the hand model (saline 76 h (65.92 to 125.01) vs. propofol 75 h (51.45 to 99.00), p > 0.05) or in the 0.5 McFarland S. aureus model (saline 105 h (45.97 to 239.94) vs. propofol 148 h (66.04 to 236.58), p > 0.05). In contrast, in the 0.05 McFarland S. aureus model, the median (IQR) time to positivity was significantly higher when instilling saline than when instilling propofol: 240 h (154.82 to 360.00) vs. 66 h (58.01 to 69.11), p = 0.008. Moreover, when comparing the median (IQR) time to positivity between the three models with saline, the 0.05 McFarland S. aureus model was also significantly higher than the hands model: 240 h (154.82 to 360.00) vs. 76 h (65.95 to 125.00), p = 0.016 (Table 2, Figure 2).Table 2 Median time to positivity of the bottles comparing different contamination models and different fluids Fluid MTP (IQR), h p Hands 0.5 MF SA 0.05 MF SA Saline 76.25 (65.95 to 125.00) 104.92 (45.96 to 239.94) 240.20 (154.82 to 360.00) 0.041 a Propofol 74.91 (51.44 to 98.99) 147.89 (66.03 to 236.58) 66.11 (58.01 to 69.11) 0.24 p 0.69 1.00 0.008 aWe found statistically significant differences between the 0.05 McFarland S. aureus model and the hand model. MTP, median time to positivity; SA, Staphylococcus aureus; MF, McFarland. Figure 2 Main results. MF, McFarland; SA, Staphylococcus aureus; MTP, median time to positivity.

Propofol 74.91 (51.44 to 98.99) 147.89 (66.03 to 236.58) 66.11 (58.01 to 69.11) 0.24 p 0.69 1.00 0.008 aWe found statistically significant differences between the 0.05 McFarland S. aureus model and the hand model. MTP, median time to positivity; SA, Staphylococcus aureus; MF, McFarland. Figure 2 Main results. MF, McFarland; SA, Staphylococcus aureus; MTP, median time to positivity. We have created an in vitro model of connector handling with gloves impregnated with a 0.05 McFarland S. aureus solution while instilling saline, lasting long enough to be used as a useful model to test the efficacy of closed needleless connectors. Some recent in vitro studies used standard models of contamination to compare different needleless connectors against the capacity of microorganisms to ingress through the catheter lumen. However, some of them were performed using a two-phase method, in which, first, the connectors were contaminated with a single known microorganism (such as S. aureus or S. epidermidis, which is commonly associated to catheter infection) at different concentrations and, second, they were manipulated to instill contaminated infusion fluids [3,4,6,7,10,11]. But this methodology did not simulate the real practice of connector handling.

th a single known microorganism (such as S. aureus or S. epidermidis, which is commonly associated to catheter infection) at different concentrations and, second, they were manipulated to instill contaminated infusion fluids [3,4,6,7,10,11]. But this methodology did not simulate the real practice of connector handling. We compared two solutions of S. aureus of different concentrations by colonizing the surface of a blood culture bottle (while instilling sterile fluids), which simulates better the real daily practice, instead of instilling a contaminated fluid directly through the catheter. Besides, we tested not only a known concentration of a single known microorganism, but also the microorganisms from the flora of the manipulators' hands, as it has been demonstrated that these are highly colonized and could be a potential source of contamination [12]. With this model based on a simple daily handling, we have demonstrated that manipulation of connectors using gloves impregnated with a 0.05 McFarland solution of S. aureus while instilling sterile saline had longer times of positivity, allowing us to prove this model in the comparison of other types of connectors with time enough to detect differences between them regarding their efficacy to prevent the entry of microorganisms through the catheter. Our study also supports the importance of disinfection before handling a connector, as in our model, all bottles manipulated without gloves turned positive up to 3 to 4 days after catheter insertion.

We compared two solutions of S. aureus of different concentrations by colonizing the surface of a blood culture bottle (while instilling sterile fluids), which simulates better the real daily practice, instead of instilling a contaminated fluid directly through the catheter. Besides, we tested not only a known concentration of a single known microorganism, but also the microorganisms from the flora of the manipulators' hands, as it has been demonstrated that these are highly colonized and could be a potential source of contamination [12]. With this model based on a simple daily handling, we have demonstrated that manipulation of connectors using gloves impregnated with a 0.05 McFarland solution of S. aureus while instilling sterile saline had longer times of positivity, allowing us to prove this model in the comparison of other types of connectors with time enough to detect differences between them regarding their efficacy to prevent the entry of microorganisms through the catheter. Our study also supports the importance of disinfection before handling a connector, as in our model, all bottles manipulated without gloves turned positive up to 3 to 4 days after catheter insertion. Conclusions A daily connector handling with a 0.05 McFarland S. aureus solution while instilling saline proved to be a useful model to test the efficacy of closed needleless connectors. Future studies must be performed with a larger sample and comparing other types of connectors.

Our study also supports the importance of disinfection before handling a connector, as in our model, all bottles manipulated without gloves turned positive up to 3 to 4 days after catheter insertion. Conclusions A daily connector handling with a 0.05 McFarland S. aureus solution while instilling saline proved to be a useful model to test the efficacy of closed needleless connectors. Future studies must be performed with a larger sample and comparing other types of connectors. Additional files Additional file 1: Manipulation with gloves. The manipulation model showing the handling of the bottles while instilling the fluids using gloves impregnated with the S. aureus solution. Additional file 2: Manipulation without gloves (hands). The manipulation model showing the handling of the bottles while instilling the fluids without gloves. Abbreviations C-RBSIcatheter-related bloodstream infections CRIcatheter-related infections IQRinterquartile range Competing interests This work was supported by grants from CareFusion. MG (CP13/00268) was supported by the Fondo de Investigación Sanitaria. Authors' contributions MG performed the study design and drafted the manuscript. MP participated in the design of the study, carried out the manipulation model, and participated in the analysis of the data. LA has been involved in the design of the study and participated in the statistical analysis. PM made substantial contributions to the acquisition of data and its interpretation. EB participated in the study design and gave the final approval of the version to be published. All authors read and approved the final manuscript.

been involved in the design of the study and participated in the statistical analysis. PM made substantial contributions to the acquisition of data and its interpretation. EB participated in the study design and gave the final approval of the version to be published. All authors read and approved the final manuscript. Authors' information MG is a young researcher at the Microbiology Department, whose main project is based on the prevention, diagnosis, and treatment of C-RBSI. MP is a research nurse from the Cardiac Surgery Postoperative Care Unit, who daily manipulates central venous catheters. LA is responsible for the BACTEC area and is a specialist in statistics. PM is responsible for the diagnosis area of catheter colonization and C-RBSI. EB is the head of the Clinical Microbiology and Infectious Diseases Department and an expert on C-RBSI.

Background Renal ischemia-reperfusion injury (IRI) is a common cause of acute kidney injury (AKI) in critically ill patients. Critically ill patients with AKI generally receive continuous renal replacement therapy, but the effects are insufficient to spare them from high mortality [1,2]. The high risk of death from AKI stems from extrarenal complications resulting from inter-organ crosstalk and multiple organ dysfunction syndrome [2,3]. Renal IRI exemplifies the pathophysiological significance of increased cytokine levels and enhanced inflammatory responses [4,5] that injure and inflame remote organs such as the lung [6] and heart [7].

ems from extrarenal complications resulting from inter-organ crosstalk and multiple organ dysfunction syndrome [2,3]. Renal IRI exemplifies the pathophysiological significance of increased cytokine levels and enhanced inflammatory responses [4,5] that injure and inflame remote organs such as the lung [6] and heart [7]. Atrial natriuretic peptide (ANP) has natriuretic, diuretic, and vasodilating properties and serves important functions as a regulator of blood pressure and fluid volume homeostasis [8]. ANP increases the glomerular filtration rate (GFR) by dilating afferent arterioles and constricting efferent arterioles to increase the glomerular capillary hydraulic pressure [9]. It has also been found to enhance recovery from renal IRI by increasing the renal medullary blood flow in rats [10]. In a clinical setting, ANP infusion improves pulmonary capillary wedge pressure and cardiac index in patients with acute heart failure [11] and preserves renal function after cardiovascular surgery [12-14]. ANP has also been found to confer anti-inflammatory effects by inhibiting nuclear factor (NF)-κB activation and cytokine production [15-17]. In a recent study by our group, ANP pre-treatment prevented kidney-lung crosstalk in a rat model of renal IRI [18]. Yet it remains unclear whether ANP post-treatment protects the heart as well as lung after renal IRI. We hypothesized that the post-treatment might benefit the kidney, lung, and heart in a general fashion by attenuating inflammation. We divided the experiments into two parts, I and II. Our hypothesis of experiment I is that unilateral renal IRI induces inflammation on the contralateral kidney as well as remote organs and ANP post-treatment attenuates kidney-lung crosstalk by inhibiting expanding inflammation. Therefore, we examined the effects of IRI-induced inflammation on the contralateral kidney, lung, and heart in a rat model of unilateral renal IRI with mechanical ventilation and elucidated whether ANP post-treatment attenuates inter-organ crosstalk among the kidney, lung, and heart by inhibiting inflammation. Further, in experiment II, we adopted a rat model of bilateral renal IRI to bring our model somewhat closer to clinical reality. Our hypothesis of experiment II is that bilateral renal IRI induces kidney injury accompanied by increase in circulating cytokines and ANP post-treatment attenuates release of cytokines from the kidney into circulation.

e adopted a rat model of bilateral renal IRI to bring our model somewhat closer to clinical reality. Our hypothesis of experiment II is that bilateral renal IRI induces kidney injury accompanied by increase in circulating cytokines and ANP post-treatment attenuates release of cytokines from the kidney into circulation. Therefore, we determined plasma cytokine concentration in the rat model of bilateral renal IRI excluding the effects of mechanical ventilation and saline and elucidated the inhibitory effect of ANP post-treatment on spreading inflammation. Methods All the protocols in this study were approved by the Institutional Animal Care Committee of Tokyo Medical and Dental University (0140245A).

Therefore, we determined plasma cytokine concentration in the rat model of bilateral renal IRI excluding the effects of mechanical ventilation and saline and elucidated the inhibitory effect of ANP post-treatment on spreading inflammation. Methods All the protocols in this study were approved by the Institutional Animal Care Committee of Tokyo Medical and Dental University (0140245A). Experiment I Animal preparation The animals were handled and cared for in accordance with the National Institutes of Health guidelines. Thirty-four male Sprague-Dawley rats (body weight 254 to 311 g) were anesthetized with an intraperitoneal injection of pentobarbital sodium (5 mg/100 g body weight). Each animal underwent a tracheostomy and intratracheal cannulation and was mechanically ventilated (Respirator Model SN-480-7, Shinano Ltd., Tokyo, Japan) under the following conditions: FIO2 0.21, tidal volume of 10 ml/kg with 5 cmH2O positive end-expiratory pressure, respiratory rate of 30 to 40 cycles/min. The right carotid artery was cannulated with a catheter for continuous measurement of the arterial pressure and heart rate and for intermittent arterial blood sampling. The arterial pressure was measured with a blood pressure amplifier (AP-641G, SEN-6102M, Nihon Kohden, Tokyo, Japan) and data acquisition system (PowerLab2/26, ML826, ADInstruments, Bella Vista, Australia) by connecting the catheter to a transducer and calibrating at zero at the midchest. The right femoral vein was cannulated with a catheter for infusion of saline or ANP. The ANP was a generous gift from the Daiichi Sankyo Company (Tokyo, Japan).

isition system (PowerLab2/26, ML826, ADInstruments, Bella Vista, Australia) by connecting the catheter to a transducer and calibrating at zero at the midchest. The right femoral vein was cannulated with a catheter for infusion of saline or ANP. The ANP was a generous gift from the Daiichi Sankyo Company (Tokyo, Japan). Renal ischemia-reperfusion The left renal pedicle was exposed via a midline incision, clamped with a vascular clip for 30 min, and released. Occlusion was verified visually by the change in the color of the kidney to a paler hue. After clamp removal, the restoration of the blood flow to the kidney was confirmed upon the return of the original color. The abdomen was closed in one layer. The sham surgery consisted of the same procedure, but with no clamping of the left renal pedicle. This renal ischemia-reperfusion injury is a model of AKI.

hue. After clamp removal, the restoration of the blood flow to the kidney was confirmed upon the return of the original color. The abdomen was closed in one layer. The sham surgery consisted of the same procedure, but with no clamping of the left renal pedicle. This renal ischemia-reperfusion injury is a model of AKI. Experimental protocol The rats were randomized to four experimental groups: an 1) IRI + saline group (n = 10), 2) IRI + ANP group (n = 10), 3) sham + saline group (n = 6), and 4) sham + ANP group (n = 8). All of the animals were mechanically ventilated. From 5 min after clamping of the left renal pedicle, the IRI + saline and sham + saline groups were infused with saline for 3 h 25 min at a rate of 6 ml/kg/h. The ANP infusion in the IRI + ANP and saline + ANP groups was started at the same time point (from 5 min after clamping of the left renal pedicle) and administered at the same rate and duration (0.2 μg/kg/min for 3 h 25 min) using saline mixed with ANP dissolved in 2-ml portions of distilled water. The heart rate, mean arterial pressure, arterial blood gases, and plasma concentrations of lactate, creatinine, and potassium were measured at baseline and at 1, 2, and 3 h after declamping. Blood gas analysis was performed on a blood gas analyzer (Radiometer ABL 837, Radiometer Medical ApS, Copenhagen, Denmark). At the completion of the experiment, all of the animals were killed with overdose of pentobarbital. The kidney, lung, and heart were harvested and either preserved at −80°C until use for the cytokine mRNA analysis or preserved in formalin until the histologic examination.

, Radiometer Medical ApS, Copenhagen, Denmark). At the completion of the experiment, all of the animals were killed with overdose of pentobarbital. The kidney, lung, and heart were harvested and either preserved at −80°C until use for the cytokine mRNA analysis or preserved in formalin until the histologic examination. Wet/dry ratio of the lung The wet/dry ratio of the lung is a gravimetric measure of pulmonary edema and an accurate gauge of changes in the lung dry mass [19]. We measured the wet/dry ratio by the same method reported by Heremans et al. [20] by desiccating the lung at 80°C until a constant weight was obtained. The ratio was calculated as a parameter of lung edema.

the lung is a gravimetric measure of pulmonary edema and an accurate gauge of changes in the lung dry mass [19]. We measured the wet/dry ratio by the same method reported by Heremans et al. [20] by desiccating the lung at 80°C until a constant weight was obtained. The ratio was calculated as a parameter of lung edema. RNA extraction and TaqMan real-time PCR Total RNA was extracted from the kidney, lung, and heart with TRIzol reagent (Invitrogen, Carlsbad, CA, USA) according to the manufacturer's instructions. The RNA concentration was determined by the absorbance read at 260 nm (GeneQuant 100, GE Healthcare UK Ltd, Chalfont St Giles, Buckinghamshire, UK). The primers and TaqMan probes for tumor necrosis factor (TNF)-α, interleukin (IL)-1β, IL-6, and glutaraldehyde-3-phosphate dehydrogenase (GAPDH) mRNA were purchased from a commercial laboratory (Applied Biosystems, Foster City, CA, USA). The mRNA expressions of TNF-α, IL-1β, and IL-6 were determined by real-time polymerase chain reaction (PCR). cDNA was synthesized using TaqMan reverse transcription reagents (Applied Biosystems, Roche Molecular Systems, Inc., Branchburg, NJ, USA) and quantified using a thermal cycler (PC707, ASTEC Co., Ltd., Minato-ku, Japan). TaqMan real-time PCR was performed using an ABI 7900HT (Applied Biosystems, Foster City, CA, USA). TaqMan rat GAPDH was used as an internal control and relative gene expression values were determined using the 2−ΔΔCT method [21].

NJ, USA) and quantified using a thermal cycler (PC707, ASTEC Co., Ltd., Minato-ku, Japan). TaqMan real-time PCR was performed using an ABI 7900HT (Applied Biosystems, Foster City, CA, USA). TaqMan rat GAPDH was used as an internal control and relative gene expression values were determined using the 2−ΔΔCT method [21]. TNF-α, IL-6, and NF-κB immunostaining and scoring in the kidney, lung, and heart Five rats from each group were used for the immunohistochemical examination. The kidney, lung, and heart were resected, embedded in paraffin, sliced into thin sections, and immunostained. Anti-TNF-α goat polyclonal antibody (SC-1348, diluted 1:20) and anti-IL-6 rabbit polyclonal antibody (SC-1265-R, diluted 1:200) were purchased from Santa Cruz Biotechnology (Dallas, TX, USA). Anti-NF-κB rabbit monoclonal antibody (1559-1, clone E381, diluted 1:200) was purchased from Epitomics (Burlingame, CA, USA). The sections were deparaffinized with xylene. For the IL-6 immunostaining, the sections were heat-treated in a microwave oven in citric acid buffer at pH 6.0 for 20 min and then air-cooled for 20 min. For the TNF-α and NF-κB immunostaining, the heat treatment was omitted. The subsequent immunostaining procedure was commenced by rehydrating the sections with an alcohol series and then treating them for 10 min with dH2O and H2O2 to inactivate the endogenous peroxidase. The antibodies were then added to the sections in a moisture chamber and reacted at RT for 3 h. After washing in phosphate buffer solution with Tween20 (PBST) for 30 min, the TNF-α samples were reacted for 30 min by indirect immunostaining using anti-goat antibody conjugated with horseradish peroxidase (P0449, diluted 1:30, DAKO, Tokyo, Japan). The IL-6 and NF-κB samples were visualized using a Novo Link Polymer Kit (RE7280-K, Leica Microsystems, Tokyo, Japan). After reacting the linker and polymer in the kit for 30 min each, the slides were visualized with diaminobenzidine, counterstained with hematoxylin, dehydrated, and cover-slipped. TNF-α, IL-6, and NF-κB expressions were evaluated semi-quantitatively by randomly choosing five areas in each slide and having them uniformly evaluated in a high-power field (×200) by a pathologist who had no knowledge of the experimental conditions (one of the authors). Scores of 3, 2, 1, and 0 were respectively assigned to fields with strong, moderate, weak, and negligible staining for each immunostaining.

five areas in each slide and having them uniformly evaluated in a high-power field (×200) by a pathologist who had no knowledge of the experimental conditions (one of the authors). Scores of 3, 2, 1, and 0 were respectively assigned to fields with strong, moderate, weak, and negligible staining for each immunostaining. The level of expression was the mean value of five fields (TNF-α, IL-6, and NF-κB expression score, respectively).

five areas in each slide and having them uniformly evaluated in a high-power field (×200) by a pathologist who had no knowledge of the experimental conditions (one of the authors). Scores of 3, 2, 1, and 0 were respectively assigned to fields with strong, moderate, weak, and negligible staining for each immunostaining. The level of expression was the mean value of five fields (TNF-α, IL-6, and NF-κB expression score, respectively). Experiment II Thirteen male Sprague-Dawley rats (body weight 327 to 376 g) were anesthetized with an intraperitoneal injection of pentobarbital sodium (5 mg/100 g body weight). Each animal was allowed to breathe spontaneously, without mechanical ventilation. The right carotid artery was cannulated with a catheter and the arterial pressure was measured with a blood pressure amplifier as stated above. The right femoral vein was cannulated with a catheter for infusion of lactated Ringer's (LR) solution or ANP. The bilateral renal pedicles were clamped with vascular clips for 30 min and released. The rats were randomized to three groups: 1) IRI + LR group (n = 5), 2) IRI + ANP group (n = 5), and 3) sham + LR group (n = 3). From 30 min after clamping, the IRI + LR and sham + LR groups were infused with LR for 3 h at a rate of 6 ml/kg/h. The ANP infusion in the IRI + ANP group was started at the same time point and administered at the same rate and for the same duration (0.2 μg/kg/min for 3 h) using LR mixed with ANP dissolved in 2-ml portions of distilled water. The heart rate, mean arterial pressure, arterial blood gases, and plasma concentrations of lactate, creatinine, and potassium were measured at baseline and at 1, 2, and 3 h after declamping. The plasma concentrations of TNF-α, IL-1β, and IL-6 were determined at baseline and at 3 h using the rat ELISA kit (Quantikinetm TM, R&D Systems, Minneapolis, MN, USA) following the manufacturer's instruction. The samples were tested in duplicate.

tassium were measured at baseline and at 1, 2, and 3 h after declamping. The plasma concentrations of TNF-α, IL-1β, and IL-6 were determined at baseline and at 3 h using the rat ELISA kit (Quantikinetm TM, R&D Systems, Minneapolis, MN, USA) following the manufacturer's instruction. The samples were tested in duplicate. Statistical analysis All data are shown as median and interquartile range (IQR). The hemodynamics; plasma concentrations of creatinine, lactate, potassium, and cytokines; and blood gas variables were all analyzed by the Kruskal-Wallis test at a fixed time point (3 h after declamping), as notable time-dependent changes in these parameters were found in repeated measures ANOVA for all three groups. The Kruskal-Wallis test was used to compare the wet/dry ratio, cytokine mRNA expression, and TNF-α, IL-6, and NF-κB scoring among the four groups. If the result from the Kruskal-Wallis test was significant, then the Mann-Whitney U test was similarly applied to analyze each pairing of groups. A p value of less than 0.05 was considered statistically significant.

e the wet/dry ratio, cytokine mRNA expression, and TNF-α, IL-6, and NF-κB scoring among the four groups. If the result from the Kruskal-Wallis test was significant, then the Mann-Whitney U test was similarly applied to analyze each pairing of groups. A p value of less than 0.05 was considered statistically significant. Results Experiment I Changes in hemodynamic variables and plasma concentrations of creatinine and potassium IRI did not induce any significant change in the heart rate or mean arterial pressure; however, IRI significantly (p < 0.05) increased the plasma concentrations of creatinine and potassium at 3 h. Post-IRI treatment by ANP prevented these changes in the variables related to renal function caused by IRI, and the values at 3 h in the IRI + ANP group were significantly lower than those in the IRI + saline group (Figure 1).Figure 1 Changes in the HR, MAP, and plasma concentrations of creatinine and potassium during mechanical ventilation. Values are expressed as median. Vertical lines indicate the interquartile range (IQR). ANP, atrial natriuretic peptide; HR, heart rate; IRI, ischemia-reperfusion injury (unilateral); MAP, mean arterial pressure. *p < 0.05, **p < 0.01 vs. the IRI + ANP group; † p < 0.05, †† p < 0.01 vs. the sham + ANP group; # p < 0.05, ## p < 0.01 vs. the sham + saline group.

lines indicate the interquartile range (IQR). ANP, atrial natriuretic peptide; HR, heart rate; IRI, ischemia-reperfusion injury (unilateral); MAP, mean arterial pressure. *p < 0.05, **p < 0.01 vs. the IRI + ANP group; † p < 0.05, †† p < 0.01 vs. the sham + ANP group; # p < 0.05, ## p < 0.01 vs. the sham + saline group. Changes in arterial blood gas variables, plasma lactate concentration, and lung wet/dry ratio IRI induced significant metabolic acidosis at 3 h (p < 0.01, Figure 2) with significantly elevated levels of plasma lactate (p < 0.05, Figure 2). Post-IRI treatment by ANP prevented IRI-induced metabolic acidosis and plasma lactate elevation. IRI significantly (p < 0.01) increased the lung wet/dry ratio, and ANP prevented this increase, as well (Figure 2).Figure 2 Changes in arterial blood gas variables, lactate concentration, and lung wet/dry ratio during mechanical ventilation. In blood gas variables and lactate concentration, value are expressed as median. Vertical lines indicate the interquartile range (IQR). In lung wet/dry ratio, boxes extend from the 25th to 75th percentile; the horizontal line shows the median. Error bars show the minimum and maximum. ANP, atrial natriuretic peptide; IRI, ischemia-reperfusion injury (unilateral). **p < 0.01 vs. the IRI + ANP group, †† p < 0.01 vs. the sham + ANP group, ## p < 0.01 vs. the sham + saline group.

boxes extend from the 25th to 75th percentile; the horizontal line shows the median. Error bars show the minimum and maximum. ANP, atrial natriuretic peptide; IRI, ischemia-reperfusion injury (unilateral). **p < 0.01 vs. the IRI + ANP group, †† p < 0.01 vs. the sham + ANP group, ## p < 0.01 vs. the sham + saline group. Cytokine mRNA expression in the kidney, lung, and heart Unilateral IRI significantly increased the mRNA expressions of TNF-α, IL-6, and IL-1β in both the ipsilateral kidney (p < 0.05 for IL-6; p < 0.01 for TNF-α and IL-1β) and the contralateral (right) kidney (p < 0.01 for TNF-α, IL-6, and IL-1β). Post-IRI treatment by ANP prevented the elevation in all these proinflammatory cytokines at 3 h after IRI (Figure 3). Furthermore, IRI significantly increased the mRNA expressions of TNF-α, IL-1β, and IL-6 in the lung (p < 0.01, Figure 4) and those of IL-1β and IL-6 in the heart (p < 0.01, Figure 4), and ANP prevented these elevations in the expression of the transcripts of proinflammatory cytokines in these remote organs (Figure 4).Figure 3 Comparison of the mRNA expression of cytokines by TaqMan real-time PCR in the kidneys. ANP, atrial natriuretic peptide; IL-1β, interleukin 1-β; IL-6, interleukin 6; IRI, ischemia-reperfusion injury; LK, left kidney; RK, right kidney; TNF-α, tumor necrosis factor-α. Boxes extend from the 25th to 75th percentile; the horizontal line shows the median. Error bars show the minimum and maximum. *p < 0.05, **p < 0.01 vs. the IRI + ANP group; †† p < 0.01 vs. the sham + ANP group; # p < 0.05, ## p < 0.01 vs. the sham + saline group.

idney; RK, right kidney; TNF-α, tumor necrosis factor-α. Boxes extend from the 25th to 75th percentile; the horizontal line shows the median. Error bars show the minimum and maximum. *p < 0.05, **p < 0.01 vs. the IRI + ANP group; †† p < 0.01 vs. the sham + ANP group; # p < 0.05, ## p < 0.01 vs. the sham + saline group. Figure 4 Comparison of the mRNA expression of cytokines by TaqMan PCR in the lung and heart. ANP, atrial natriuretic peptide; IL-1β, interleukin 1-β; IL-6, interleukin 6; IRI, ischemia-reperfusion injury; L, lung; H, heart; TNF-α, tumor necrosis factor-α. Boxes extend from the 25th to 75th percentile; the horizontal line shows the median. Error bars show the minimum and maximum. **p < 0.01 vs. the IRI + ANP group, †† p < 0.01 vs. the sham + ANP group, ## p < 0.01 vs. the sham + saline group.

ischemia-reperfusion injury; L, lung; H, heart; TNF-α, tumor necrosis factor-α. Boxes extend from the 25th to 75th percentile; the horizontal line shows the median. Error bars show the minimum and maximum. **p < 0.01 vs. the IRI + ANP group, †† p < 0.01 vs. the sham + ANP group, ## p < 0.01 vs. the sham + saline group. Histological detection and localization of TNF-α, IL-6, and NF-κB in the kidney, lung, and heart TNF-α was detected and localized in the vascular endothelial cells of the kidney, in the bronchial epithelial cells of the lung, and in the vascular endothelial cells of the heart. IL-6 was detected and localized in most vascular endothelial cells, in a few proximal convoluted tubules of the kidney, and in the columnar epithelial cells of the bronchioles of the lung and the vascular endothelial cells of the heart. NF-κB was detected and localized in the proximal convoluted tubules of the kidney, bronchioles of the lung, and myocardium of the heart. IRI significantly increased the TNF-α, IL-6, and NF-κB expression scores of the left (ipsilateral) kidney (p < 0.01) and the TNF-α and NF-κB expression scores of the right (contralateral) kidney (p < 0.01). Post-IRI ANP treatment prevented all these elevations (Figure 5). IRI significantly (p < 0.05) increased the TNF-α, IL-6, and NF-κB expression scores of the lung, and ANP prevented these increases (Figure 6), whereas IRI did not induce significant changes in the TNF-α, IL-6, and NF-κB expression scores in the heart.Figure 5 Evaluation of TNF, IL-6, and NF-κB expressions in the kidney. Upper figures: The evaluation of tumor necrosis factor (TNF)-α expression in the vascular endothelial cells of the kidney. TNF-α protein was stained in brown, and the level of TNF-α expression was scored: score 0 (hardly stained), score 1 (weakly stained), score 2 (moderately stained), score 3 (strongly stained). Comparison of the TNF-α expression scores in the left kidney (LK) and right kidney (RK). ANP, atrial natriuretic peptide; IRI, ischemia-reperfusion injury. Middle figures: The evaluation of interleukin (IL)-6 expression in the vascular endothelial cells and proximal convoluted tubules of the kidney. IL-6 protein was stained in brown, and the level of IL-6 expression was scored: score 0 (hardly stained), score 1 (weakly stained), score 2 (moderately stained), score 3 (strongly stained). Comparison of the IL-6 expression scores in the left kidney (LK) and right kidney (RK).

proximal convoluted tubules of the kidney. IL-6 protein was stained in brown, and the level of IL-6 expression was scored: score 0 (hardly stained), score 1 (weakly stained), score 2 (moderately stained), score 3 (strongly stained). Comparison of the IL-6 expression scores in the left kidney (LK) and right kidney (RK). Lower figures: The evaluation of nuclear factor (NF)-κB expression in the proximal convoluted tubules of the kidney. NF-κB protein was stained in brown, and the level of NF-κB expression was scored: score 0 (hardly stained), score 1 (weakly stained), score 2 (moderately stained), score 3 (strongly stained). Comparison of the NF-κB expression scores in the left kidney (LK) and right kidney (RK).

ximal convoluted tubules of the kidney. NF-κB protein was stained in brown, and the level of NF-κB expression was scored: score 0 (hardly stained), score 1 (weakly stained), score 2 (moderately stained), score 3 (strongly stained). Comparison of the NF-κB expression scores in the left kidney (LK) and right kidney (RK). Figure 6 Evaluation of TNF, IL-6, and NF-κB expressions in the lung. Upper figures: The evaluation of tumor necrosis factor (TNF)-α expression in the bronchial epithelial cells of the lung. TNF-α protein was stained in brown, and the level of TNF-α expression was scored: score 0 (hardly stained), score 1 (weakly stained), score 2 (moderately stained), score 3 (strongly stained). Comparison of the TNF-α expression scores in the lung. ANP, atrial natriuretic peptide; IRI, ischemia-reperfusion injury. Middle figures: The evaluation of IL-6 in the columnar epithelial cells of the bronchioles of the lung. IL-6 protein was stained in brown, and the level of IL-6 expression was scored: score 0 (hardly stained), score 1 (weakly stained), score 2 (moderately stained), score 3 (strongly stained). Comparison of IL-6 expression score in the lung. Lower figures: The evaluation of nuclear factor (NF)-κB expression in the bronchioles of the lung. NF-κB protein was stained in brown, and the level of NF-κB expression was scored: score 0 (hardly stained), score 1 (weakly stained), score 2 (moderately stained), score 3 (strongly stained). Comparison of NF-κB expression score in the lung.

valuation of nuclear factor (NF)-κB expression in the bronchioles of the lung. NF-κB protein was stained in brown, and the level of NF-κB expression was scored: score 0 (hardly stained), score 1 (weakly stained), score 2 (moderately stained), score 3 (strongly stained). Comparison of NF-κB expression score in the lung. Experiment II Bilateral IRI procedures in the kidney did not change the heart rate, but significantly (p < 0.05) decreased the mean arterial pressure. IRI increased the plasma concentrations of creatinine and potassium, and ANP prevented the increase in the latter (Figure 7). IRI was not found to elicit acidosis by respiratory compensation and no significant change in arterial blood gas bicarbonate was observed (Figure 8). IRI also left the plasma lactate concentration unchanged. IRI significantly (p < 0.05) increased the plasma concentrations of IL-1β and IL-6, but not the concentration of TNF-α, and ANP attenuated the increases in IL-1β and IL-6 in 13 rats (Figure 9).Figure 7 Changes in the HR, MAP, and plasma concentrations of creatinine and potassium during spontaneous breathing. Values are expressed as median. Vertical lines indicate the interquartile range (IQR). ANP, atrial natriuretic peptide; HR, heart rate; IRI, ischemia-reperfusion injury (bilateral); LR, lactated Ringer's solution; MAP, mean arterial pressure. *p < 0.05 vs. the IRI + ANP group, # p < 0.05 vs. the sham + LR group.

thing. Values are expressed as median. Vertical lines indicate the interquartile range (IQR). ANP, atrial natriuretic peptide; HR, heart rate; IRI, ischemia-reperfusion injury (bilateral); LR, lactated Ringer's solution; MAP, mean arterial pressure. *p < 0.05 vs. the IRI + ANP group, # p < 0.05 vs. the sham + LR group. Figure 8 Changes in arterial blood gas variables during spontaneous breathing. Values are expressed as median. Vertical lines indicate the interquartile range (IQR). ANP, atrial natriuretic peptide; IRI, ischemia-reperfusion injury (bilateral). Figure 9 Changes in plasma concentrations of IL-1β and IL-6 during spontaneous breathing. ANP, atrial natriuretic peptide; IL-1β, interleukin 1-β; IL-6, interleukin 6; IRI, ischemia-reperfusion injury (bilateral); LR, lactated Ringer's solution.

Figure 8 Changes in arterial blood gas variables during spontaneous breathing. Values are expressed as median. Vertical lines indicate the interquartile range (IQR). ANP, atrial natriuretic peptide; IRI, ischemia-reperfusion injury (bilateral). Figure 9 Changes in plasma concentrations of IL-1β and IL-6 during spontaneous breathing. ANP, atrial natriuretic peptide; IL-1β, interleukin 1-β; IL-6, interleukin 6; IRI, ischemia-reperfusion injury (bilateral); LR, lactated Ringer's solution. Discussion The most compelling findings observed in this rat model of unilateral renal IRI with mechanical ventilation (experiment I) were that unilateral renal IRI induced inflammation not only in the ipsilateral kidney but also in remote organs including the contralateral kidney, lung, and heart and ANP post-treatment inhibited inflammation of these organs. In addition, ANP post-treatment inhibited the renal IRI-induced metabolic acidosis, pulmonary edema, and increases in the plasma concentrations of lactate, creatinine, and potassium. Although unilateral renal IRI is not the main cause of AKI in critically ill patients, unilateral renal IRI remains a major problem in surgeries, such as renal transplantation [22] and juxtarenal and suprarenal abdominal aortic aneurysm repair [23]. The renal function of these patients must be preserved during the perioperative period. We therefore tried to investigate the effects of unilateral renal IRI on the non-ischemic contralateral kidney, as well as the lung and heart. Renal IRI augmented the mRNA expressions of TNF-α, IL-1β, and IL-6 in the kidney and lung, and this effect was inhibited by the ANP post-treatment. Renal IRI also augmented the mRNA expression of IL-1β and IL-6 in the heart, and the ANP post-treatment again inhibited the augmenting action. The ANP post-treatment prevented the renal IRI-induced localization of TNF-α, IL-6, and NF-κB in the kidney and IL-6 and NF-κB in the lung. Furthermore, in experiment II, the bilateral renal IRI increased the plasma concentrations of IL-1β and IL-6, but not the concentration of TNF-α, and the ANP treatment 30 min after clamping attenuated the increases in IL-1β and IL-6. The plasma TNF-α concentration might have passed its peak at 3 h. These results corroborate earlier evidence of an ANP-conferred enhancement of recovery from renal IRI in rats [10] and strengthen evidence favoring ANP as a possible treatment for AKI. In an earlier study by our group, ANP preserved renal function after suprarenal abdominal aortic cross-clamping in a dog model approximating ischemic AKI following abdominal aortic aneurysm repair [24].

ncement of recovery from renal IRI in rats [10] and strengthen evidence favoring ANP as a possible treatment for AKI. In an earlier study by our group, ANP preserved renal function after suprarenal abdominal aortic cross-clamping in a dog model approximating ischemic AKI following abdominal aortic aneurysm repair [24]. In clinical studies, low-dose ANP infusion after cardiovascular surgery enhanced the renal excretory function, decreased the probability of dialysis, and improved the dialysis-free survival in ischemic acute renal failure [12-14]. The induction of IRI in the present study led to reduction in MAP in the saline group and LR group, but not in the sham group or ANP group. The plasma lactate concentration was also found to increase sharply in the IRI + saline group, but no such increase was observed after ANP treatment. These findings suggest that the administration of ANP might contribute to prevention of extravasation of the fluid, which resulted in maintenance of peripheral circulation. Our results showed successful results of ANP post-treatment in attenuating renal IRI and reducing cytokine mRNA expression in the kidney, lung, and heart. ANP post-treatment also reduced plasma cytokine (IL-1β and IL-6) concentrations, and this might be one of the mechanisms explaining the therapeutic effect of ANP on remote organ inflammation.

ful results of ANP post-treatment in attenuating renal IRI and reducing cytokine mRNA expression in the kidney, lung, and heart. ANP post-treatment also reduced plasma cytokine (IL-1β and IL-6) concentrations, and this might be one of the mechanisms explaining the therapeutic effect of ANP on remote organ inflammation. Renal IRI generally leads to outer medullary congestion and hypoxia, conditions that predispose patients to ischemic injury in the S3 segment of the proximal tubule [25]. ANP increases GFR by dilating the afferent arterioles and constricting the efferent arterioles to increase glomerular capillary hydraulic pressure [9]. The reno-protective effect of ANP may derive from protection against medullary ischemia via ANP-induced increases in the medullary vasa recta blood flow [10,26]. Renal IRI and inflammation of remote organs Renal IRI engages the innate and adaptive immune responses and works in conjunction with cytokine generation within the kidney [27]. Once this process starts, cellular and soluble mediators injure remote organs such as the lung and heart via organ crosstalk. Kidney-lung interaction Renal IRI induced lung inflammation in a mouse model with systemic inflammatory syndrome and upregulated IL-6 mRNA expression in both the kidney and lung [6]. These findings are consistent with our present results. After renal IRI, we also detected and localized IL-6 in the columnar epithelial cells of the bronchioles of the lung and the vascular endothelial cells and proximal convoluted tubules of the kidney in our animals.

6 mRNA expression in both the kidney and lung [6]. These findings are consistent with our present results. After renal IRI, we also detected and localized IL-6 in the columnar epithelial cells of the bronchioles of the lung and the vascular endothelial cells and proximal convoluted tubules of the kidney in our animals. The alveolar epithelium has features in common with the renal tubular epithelium, such as localization of water channels and ion transporters [28]. Mechanisms of renal IRI-induced lung injury are assumed to include a dysregulation of water clearance, inflammation, an innate immune response, proinflammatory cytokines, oxidative stress, and apoptosis [29]. Studies confirming the expression of ANP and its receptors and their variable modes of regulation in the immune system [30] support the notion that ANP has immunomodulatory potency. ANP inhibits the activation of NF-κB production in both mouse macrophages and endothelial cells [15-17]. ANP post-treatment of our experimental animals inhibited the mRNA expression of TNF-α, IL-1β, and IL-6 in the kidney. The ability of ANP to suppress the induction of proinflammatory cytokines such as TNF-α, IL-1β, and IL-6 may signify a substantive anti-inflammatory action on the kidney and lung. ANP post-treatment was also found to inhibit the activation of NF-κB production in the kidney and lung in our study. These findings suggest that ANP has anti-inflammatory effects on both organs.

lammatory cytokines such as TNF-α, IL-1β, and IL-6 may signify a substantive anti-inflammatory action on the kidney and lung. ANP post-treatment was also found to inhibit the activation of NF-κB production in the kidney and lung in our study. These findings suggest that ANP has anti-inflammatory effects on both organs. Increased capillary endothelial permeability is a major pathologic mechanism of pulmonary edema in acute lung injury and acute respiratory distress syndrome. Our group previously reported that ANP improved pulmonary gas exchange by reducing extravascular lung water in patients with acute lung injury [31] and in a canine model with oleic acid-induced pulmonary edema [32]. This finding is corroborated by reports that ANP knockout in mice increases the severity of lung inflammation and vascular barrier dysfunction caused by bacterial pathogens [33,34]. Increased ANP levels in patients with acute lung injury [35] may represent an important compensatory mechanism aimed at attenuation of injury and lung barrier dysfunction. Tian et al. [36] recently demonstrated a novel protective mechanism of ANP against pathologic hyper-permeability and suggested a pharmacological intervention for the prevention of increased vascular leak via PAK1-dependent modulation of guanine nucleotide exchange factor H1 activity. The activity of renal NF-κB appears to increase in the absence of the functional guanylyl cyclase/natriuretic peptide receptor-A (GC-A/NPRA) gene 1 and to elicit abnormalities by stimulating the synthesis of proinflammatory cytokines [37]. These findings, taken together, show that ANP protects the kidney by preventing proinflammatory cytokines via conterregulatory effects on NF-κB signaling.

tional guanylyl cyclase/natriuretic peptide receptor-A (GC-A/NPRA) gene 1 and to elicit abnormalities by stimulating the synthesis of proinflammatory cytokines [37]. These findings, taken together, show that ANP protects the kidney by preventing proinflammatory cytokines via conterregulatory effects on NF-κB signaling. Kidney-heart interaction Several inflammatory mediators participate in the pathophysiologic process of cardiorenal syndrome [38]. Increased production of inflammatory cytokines may adversely affect myocardial function. Elevated levels of circulating TNF-α and IL-6 are associated with the development of congestive heart failure and mortality in congestive heart failure patients [39,40]. Several different pathways, most notably the activation of inflammatory transcription factors and the induction of inflammatory genes and cytokines, may contribute to heart injury following renal IRI.

associated with the development of congestive heart failure and mortality in congestive heart failure patients [39,40]. Several different pathways, most notably the activation of inflammatory transcription factors and the induction of inflammatory genes and cytokines, may contribute to heart injury following renal IRI. Renal IRI induced the mRNA expressions of TNF-α, IL-1β, and IL-6 in the heart, in our experiments, and ANP post-treatment attenuated the mRNA expressions of the latter two, IL-1β and IL-6. Yet the localizations of TNF-α, IL-6, and NF-κB in the heart at 3 h in the IRI + saline group were not significantly increased or significantly different from the localizations in other groups. These findings suggest that the cardiac levels of TNF-α, IL-6, and NF-κB were not increased in the heart 3 h after renal IRI. It may take more than 3 h to increase the cardiac levels of cytokines and NF-κB. In experiments with a rat model of renal IRI, Kelly [7] demonstrated an increase of circulating TNF-α by 1 h post renal ischemia, a further increase at 2 h, and steady elevation of the cytokine for 24 h. The cardiac levels of immnoreactive IL-1 and TNF-α in the same animals were elevated at 6, 24, and 48 h after renal ischemia, and echocardiography revealed left ventricular dysfunction, a likely sign of heart failure, at 48 h after renal IRI. It may take a longer time for renal IRI to induce hear failure.

ine for 24 h. The cardiac levels of immnoreactive IL-1 and TNF-α in the same animals were elevated at 6, 24, and 48 h after renal ischemia, and echocardiography revealed left ventricular dysfunction, a likely sign of heart failure, at 48 h after renal IRI. It may take a longer time for renal IRI to induce hear failure. Inhibition of inter-organ crosstalk by ANP It may be difficult to differentiate between inhibition of inter-organ crosstalk and direct organ protection, given that ANP has now been shown to confer protective effects on other organs in addition to the established anti-inflammatory effects. We know, however, that inter-organ crosstalk develops via cellular mediators such as neutrophils, macrophages, and lymphocytes, and inflammatory cytokines [29]. Matsumura et al. [41] have reported that ANP modulates the neutrophil functions and exerts protective effects against the neutrophil-induced endothelial cytotoxity. Chujo et al. [42] have also shown that ANP significantly inhibits IRI-induced increases in renal cytokine-induced neutrophil chemoattractant-1, a chemokine responsible for the activation of neutrophils and for neutrophil chemotaxis to sites of injury. Regarding plasma cytokine concentration which would be the main route of expansion of inflammation, ANP post-treatment attenuated IRI-induced elevation of the plasma concentrations of IL-1β and IL-6. We therefore suppose that ANP may both directly and indirectly disrupt the inter-organ crosstalk following renal IRI.

y. Regarding plasma cytokine concentration which would be the main route of expansion of inflammation, ANP post-treatment attenuated IRI-induced elevation of the plasma concentrations of IL-1β and IL-6. We therefore suppose that ANP may both directly and indirectly disrupt the inter-organ crosstalk following renal IRI. Limitations of this study There are some limitations to this study. First, in the present study, we have evaluated the effects of ANP on only renal IRI. Considering that there are multiple causes of AKI in critically ill patients (e.g., sepsis, nephrotoxic agents, hypoperfusion, and their combination) other than IRI, we cannot refer to the effects of ANP on AKI caused by other mechanisms. However, because it was notable that ANP post-treatment was effective to reduce tissue injury in the lung and kidney both in unilateral and in bilateral renal IRI, further study is needed to elucidate whether this beneficial effect might be observed in other pathophysiologic conditions. Second, saline infusion and mechanical ventilation might be an aggravating factor for organ injury in the present study. Unilateral renal IRI (experiment I) with mechanical ventilation induced significant metabolic acidosis. This result may have been due to a decrease in the renal blood flow by the positive pressure ventilation. Further, the applied tidal volume of 10 ml/kg might be a little too high to protect against lung injury. Bilateral renal IRI without mechanical ventilation (experiment II) did not induce acidosis by respiratory compensation, and the arterial blood gas bicarbonate was maintained by the infusion of LR. These findings suggest that the mechanical ventilation and saline both worsened the arterial blood gas parameters after the renal IRI. A recent clinical study has actually shown chloride-restrictive fluid infusion to be significantly associated with a significant decreased incidence of AKI and a significantly decreased use of renal replacement therapy in critically ill patients [43]. Therefore, we should consider the possibility that mechanical ventilation per se, ventilator setting, and type of infusion become exacerbation factors for organ dysfunction after the renal IRI. Nevertheless, it is noteworthy that ANP post-treatment has clearly prevented IRI-induced remote organ inflammation even in the condition with these kinds of aggravating factors.

mechanical ventilation per se, ventilator setting, and type of infusion become exacerbation factors for organ dysfunction after the renal IRI. Nevertheless, it is noteworthy that ANP post-treatment has clearly prevented IRI-induced remote organ inflammation even in the condition with these kinds of aggravating factors. Conclusions Unilateral renal IRI with mechanical ventilation induced inflammation not only in the ipsilateral kidney but also in remote organs including the contralateral kidney, lung, and heart. ANP post-treatment inhibited renal IRI-induced metabolic acidosis and the mRNA expression of TNF-α, IL-1β, and IL-6 in the kidney and lung and IL-1β and IL-6 in the heart. In addition, ANP post-treatment attenuated the IRI-induced increases in the plasma concentrations of IL-1β and IL-6, as well as the IRI-induced histological localization of TNF-α, IL-6, and NF-κB in the kidney and lung. These findings show that ANP conferred a reno-protective effect and anti-inflammatory effect both on the kidney and on the lung in the rat model of renal IRI. The cardiac levels of TNF-α, IL-6, and NF-κB were not significantly increased at 3 h after renal IRI, suggesting that renal IRI-induced heart injury may occur later than lung injury. Further studies are needed to elucidate the anti-inflammatory effects of ANP on the heart. Abbreviations AKIacute kidney injury ANPatrial natriuretic peptide GFRglomerular filtration rate ILinterleukin IRIischemia-reperfusion injury NF-κBnuclear factor-κB TNF-αtumor necrosis factor-α Competing interests The authors declare that they have no competing interests.

Conclusions Unilateral renal IRI with mechanical ventilation induced inflammation not only in the ipsilateral kidney but also in remote organs including the contralateral kidney, lung, and heart. ANP post-treatment inhibited renal IRI-induced metabolic acidosis and the mRNA expression of TNF-α, IL-1β, and IL-6 in the kidney and lung and IL-1β and IL-6 in the heart. In addition, ANP post-treatment attenuated the IRI-induced increases in the plasma concentrations of IL-1β and IL-6, as well as the IRI-induced histological localization of TNF-α, IL-6, and NF-κB in the kidney and lung. These findings show that ANP conferred a reno-protective effect and anti-inflammatory effect both on the kidney and on the lung in the rat model of renal IRI. The cardiac levels of TNF-α, IL-6, and NF-κB were not significantly increased at 3 h after renal IRI, suggesting that renal IRI-induced heart injury may occur later than lung injury. Further studies are needed to elucidate the anti-inflammatory effects of ANP on the heart. Abbreviations AKIacute kidney injury ANPatrial natriuretic peptide GFRglomerular filtration rate ILinterleukin IRIischemia-reperfusion injury NF-κBnuclear factor-κB TNF-αtumor necrosis factor-α Competing interests The authors declare that they have no competing interests. Authors' contributions

Conclusions Unilateral renal IRI with mechanical ventilation induced inflammation not only in the ipsilateral kidney but also in remote organs including the contralateral kidney, lung, and heart. ANP post-treatment inhibited renal IRI-induced metabolic acidosis and the mRNA expression of TNF-α, IL-1β, and IL-6 in the kidney and lung and IL-1β and IL-6 in the heart. In addition, ANP post-treatment attenuated the IRI-induced increases in the plasma concentrations of IL-1β and IL-6, as well as the IRI-induced histological localization of TNF-α, IL-6, and NF-κB in the kidney and lung. These findings show that ANP conferred a reno-protective effect and anti-inflammatory effect both on the kidney and on the lung in the rat model of renal IRI. The cardiac levels of TNF-α, IL-6, and NF-κB were not significantly increased at 3 h after renal IRI, suggesting that renal IRI-induced heart injury may occur later than lung injury. Further studies are needed to elucidate the anti-inflammatory effects of ANP on the heart. Abbreviations AKIacute kidney injury ANPatrial natriuretic peptide GFRglomerular filtration rate ILinterleukin IRIischemia-reperfusion injury NF-κBnuclear factor-κB TNF-αtumor necrosis factor-α Competing interests The authors declare that they have no competing interests. Authors' contributions CM conceived and designed the experiments and wrote the paper; MH, MT, and CM performed the experiments; QY and TU determined the plasma cytokine concentrations; SA and MK helped with the real-time PCR; SI and YE participated in the immunohistochemical examination; and MT analyzed the data. All authors read and approved the final manuscript.

ed the experiments and wrote the paper; MH, MT, and CM performed the experiments; QY and TU determined the plasma cytokine concentrations; SA and MK helped with the real-time PCR; SI and YE participated in the immunohistochemical examination; and MT analyzed the data. All authors read and approved the final manuscript. Acknowledgements We thank Dr. Koji Kido for the technical assistance. This study was supported by a Grant-in-Aid for Scientific Research from the Ministry of Education, Culture, Sports, Science and Technology of Japan (22592010).

Background Compartment syndrome is the final result of a process that begins with the persistent increase in pressure within a tissue or parenchyma to such an extent that it is capable of altering the regional vascular inflow. Compartment syndrome may culminate in local organ failure, and if the damage persists, multiple organ failure may occur, resulting in the death of the patient. This is true for any body compartment surrounded by a rigid or semi-rigid structure, a situation often found in intracranial hypertension syndrome or abdominal compartment syndrome [1,2]. From a hydraulic point of view, the ‘renal compartment’, whose content and structure are the parenchyma and renal capsule, respectively, should not be different from other body compartments. In analogy to the intracranial hypertension syndrome, a sudden increase in fluid volume inside the renal parenchyma can result in a substantial intrarenal pressure and, as a consequence, a decrease in the renal perfusion pressure [3]. In effect, acutely injured kidneys where edema is present typically involve ischemic regions of the outer medulla due to a reduction in the vascular inflow [4,5]. From a clinical perspective, the venous congestion due to an increase of the vascular permeability to proteins, as well as the decrease of the intrarenal perfusion pressure increase the risk of developing a new or persistent septic acute kidney injury (AKI) [6].

medulla due to a reduction in the vascular inflow [4,5]. From a clinical perspective, the venous congestion due to an increase of the vascular permeability to proteins, as well as the decrease of the intrarenal perfusion pressure increase the risk of developing a new or persistent septic acute kidney injury (AKI) [6]. The present study aims to understand, both from experimental and biomechanical approaches, the functional dependence between the renal compartment volume increments and changes in the intrarenal pressure mediated by the renal capsule. We hypothesize that pressure and volume in the renal compartment are not linearly related, and may present a functional dependence similar to those found in other compartmental syndromes. Methods Animal preparation This study used anesthetized domestic large white piglets purchased from a local vivarium specialized in this species. The Universidad Andrés Bello Ethics Committee approved the experimental protocol. All experimental procedures were in accordance with the Guiding Principles in the Care and Use of Laboratory Animals adopted by the American Physiological Society. The study was powered to detect an increase in intrarenal pressure. Sample size needed to achieve an 80% study power was six, with a 0.05 one-sided significance level and a standard deviation of 33% [7].

with the Guiding Principles in the Care and Use of Laboratory Animals adopted by the American Physiological Society. The study was powered to detect an increase in intrarenal pressure. Sample size needed to achieve an 80% study power was six, with a 0.05 one-sided significance level and a standard deviation of 33% [7]. Surgical preparation and anesthesia Animals were premedicated with intramuscular midazolam (0.5 mg/kg), methadone (0.5 mg/kg), and ketamine (15 mg/kg), followed by induction with intravenous propofol (3 mg/kg). Tracheal intubation was performed with a cuffed tracheal tube (5.0-mm internal diameter; Mallinckrodt Shiley, St. Louis, MO, USA) for inhalation anesthesia with isoflurane 1.5%. An adequate level of anesthesia is assumed if reflexes are absent. Anesthesia and neuromuscular blockade were maintained by continuous infusion of propofol (10 mg/kg/h), fentanyl (5 μg/kg/h), and vecuronium (0.3 mg/kg/h) throughout the all experiments which lasted for less than 1 h. Heart rate, mean arterial pressure, and temperature were continuously monitored during the whole duration of the experiment. Before laparotomy, PaO2, pH, PaCO2, serum creatinine, and hemoglobin were assessed with an i-STAT® (Abbott Laboratories, Princeton, NJ, USA) in blood samples from the arterial catheter.

than 1 h. Heart rate, mean arterial pressure, and temperature were continuously monitored during the whole duration of the experiment. Before laparotomy, PaO2, pH, PaCO2, serum creatinine, and hemoglobin were assessed with an i-STAT® (Abbott Laboratories, Princeton, NJ, USA) in blood samples from the arterial catheter. Mechanical ventilation Animals were ventilated with anesthesia workstation Fabius GS® premium (Dräger Medical, Lübeck, Germany) using the volume control mode. Initial settings were: VT = 10 mL/kg, PEEP = 5 cmH2O, fraction of inspired oxygen = 0.4, inspiratory time = 1.0 s, and respiratory rate (RR) = 20 breaths/min. RR was adjusted to achieve a partial pressure of carbon dioxide (PaCO2) 40 ± 10 Torr.

S® premium (Dräger Medical, Lübeck, Germany) using the volume control mode. Initial settings were: VT = 10 mL/kg, PEEP = 5 cmH2O, fraction of inspired oxygen = 0.4, inspiratory time = 1.0 s, and respiratory rate (RR) = 20 breaths/min. RR was adjusted to achieve a partial pressure of carbon dioxide (PaCO2) 40 ± 10 Torr. Pressure-volume curve protocol After surgical preparation, each animal was placed in supine decubitus and a midline laparotomy was performed. Left kidneys were identified and dissected free of surrounding tissue. Since the kidney can be considered as relatively non-compressive and predominantly fluid in character, subject to Pascal's law, the intrarenal pressure can be measured in nearly every part of the kidney. Intrarenal pressure was measured using a catheter Camino 4B® inserted 1 cm in the left lower renal pole and connected to Camino Single Parameter Monitor Model SPM1 (Integra NeuroSciences, New Jersey, USA). Ureters were sectioned and a 10-Fr Foley catheter into the renal pelvis through the ureter (Figure 1). The abdomen was closed temporarily to maintain thermoregulation. The volume-pressure curve for each animal was obtained by first removing the urine inside the renal pelvis, after which a controlled volume of normal saline was injected and the peak value of the time course of pressure was recorded, to finally draw all the fluid inside the pelvis. This cycle was repeated in time intervals of approximately 3 min. The total injected fluid volume was increased by 1 mL for each subsequent cycle, starting with an injected volume of 1 mL. Following the rapid rise in pressure, the renal pressure was permitted to return to the initial control level before the next injection sequence. In the case of decapsulated kidneys, we followed the decapsulation technique described in [8], where the capsule is first incised and elevated at its lateral margin, then cut from the superior to the inferior pole and finally stripped apart in the medial plane.Figure 1 Schematic of the experimental setup design for measuring the intrarenal pressure. Fluid is injected through a Foley catheter into the renal pelvis, and intrarenal pressure is measured using a Camino catheter located in the lower renal pole.

or to the inferior pole and finally stripped apart in the medial plane.Figure 1 Schematic of the experimental setup design for measuring the intrarenal pressure. Fluid is injected through a Foley catheter into the renal pelvis, and intrarenal pressure is measured using a Camino catheter located in the lower renal pole. While under anesthesia, the animals were euthanized by 10% potassium chloride infusion until the detection of ventricular fibrillation or asystole. Statistical analysis Data are expressed as mean values ± SEM. Normality was assessed with the Anderson-Darling test. The Wilcoxon signed-rank test and the Friedman test with Bonferroni correction were conducted to compare consecutive measurements of studied variables. Significance was set at P < 0.05. All statistical analyses were performed using SPSS 20.0 (SPSS Inc., Chicago, IL, USA). Results Six anesthetized and mechanically ventilated piglets were studied (12.3 ± 2.4 kg). The baseline characteristics are described in Table 1 . All of the animals completed the experimental protocol.Table 1 Baseline characteristics of the piglets included in the study Baseline HR (bpm) 97 ± 11 MAP (Torr) 85.5 ± 5.7 Temperature (°C) 37.6 ± 0.2 pH 7.43 ± 0.03 PaO2 (Torr) 171.2 ± 6.2 PaCO2 (Torr) 43.1 ± 3.1 Serum creatinine (mg/dL) 1.13 ± 0.05 Hemoglobin (g/dL) 9.0 ± 0.3 Data is expressed as mean value ± SEM. HR, heart rate; MAP, mean arterial pressure.

Results Six anesthetized and mechanically ventilated piglets were studied (12.3 ± 2.4 kg). The baseline characteristics are described in Table 1 . All of the animals completed the experimental protocol.Table 1 Baseline characteristics of the piglets included in the study Baseline HR (bpm) 97 ± 11 MAP (Torr) 85.5 ± 5.7 Temperature (°C) 37.6 ± 0.2 pH 7.43 ± 0.03 PaO2 (Torr) 171.2 ± 6.2 PaCO2 (Torr) 43.1 ± 3.1 Serum creatinine (mg/dL) 1.13 ± 0.05 Hemoglobin (g/dL) 9.0 ± 0.3 Data is expressed as mean value ± SEM. HR, heart rate; MAP, mean arterial pressure. The pressure-volume experimental data are shown in Figure 2 and summarized in Table 2. The measured intrarenal pressure increased progressively as successive volume increments of normal saline were injected into the renal pelvis. The pressure-volume curve in normal kidneys exhibited a marked nonlinear behavior. In particular, it can be observed from Figure 2 that the pressure is not proportional to the volume injected. An exponential expression was found to give the best fit to the data, resulting in the following mathematical relation:Figure 2 Pressure-volume curve in the renal compartment. Data is expressed as mean value ± SEM. Solid line represents the exponential fit to the data. p, intrarenal pressure; V, volume injected to the renal pelvis. Table 2 Pressure-volume data in the renal compartment V (mL) 0 1 2 3 4 Intrarenal pressure (Torr) 12.0 ± 2.1 17.5 ± 2.9 20.2 ± 3.4 31.2 ± 6.8 70.8 ± 16.7 Data is expressed as mean value ± SEM. V, volume injected to the renal pelvis. p=10.854e0.413ΔVTorr

The pressure-volume experimental data are shown in Figure 2 and summarized in Table 2. The measured intrarenal pressure increased progressively as successive volume increments of normal saline were injected into the renal pelvis. The pressure-volume curve in normal kidneys exhibited a marked nonlinear behavior. In particular, it can be observed from Figure 2 that the pressure is not proportional to the volume injected. An exponential expression was found to give the best fit to the data, resulting in the following mathematical relation:Figure 2 Pressure-volume curve in the renal compartment. Data is expressed as mean value ± SEM. Solid line represents the exponential fit to the data. p, intrarenal pressure; V, volume injected to the renal pelvis. Table 2 Pressure-volume data in the renal compartment V (mL) 0 1 2 3 4 Intrarenal pressure (Torr) 12.0 ± 2.1 17.5 ± 2.9 20.2 ± 3.4 31.2 ± 6.8 70.8 ± 16.7 Data is expressed as mean value ± SEM. V, volume injected to the renal pelvis. p=10.854e0.413ΔVTorr where p is the intrarenal pressure and ΔV is the volume of the normal saline injected to the renal pelvis.

Table 2 Pressure-volume data in the renal compartment V (mL) 0 1 2 3 4 Intrarenal pressure (Torr) 12.0 ± 2.1 17.5 ± 2.9 20.2 ± 3.4 31.2 ± 6.8 70.8 ± 16.7 Data is expressed as mean value ± SEM. V, volume injected to the renal pelvis. p=10.854e0.413ΔVTorr where p is the intrarenal pressure and ΔV is the volume of the normal saline injected to the renal pelvis. Biomechanical model We model a portion of the renal capsule as a membrane with spherical-cap shape subject to an internal pressure exerted by the renal parenchyma acting as a fluid (Figure 3). Let p be the fluid pressure, σ be the capsule stress (force per unit area), and h and r be the capsule thickness and the radius of the capsule in the deformed state, respectively. From Laplace's law, the intrarenal pressure and capsule stress are related throughFigure 3 Schematic of the biomechanical model for the kidney capsule. p, intrarenal pressure; σ, capsule stress; h, capsule thickness; r, capsule radius. 1 p=2hrσ The capsule stress is related to its deformation through a constitutive law. Nedeker et al. [9] performed extensive mechanical testing of porcine kidney capsules, proposing an incompressible hyperelastic Ogden model [10]. For a capsule element under equi-biaxial loading subject to internal pressure [11], the constitutive law reads 2 σ=μ1λ15−λ−30+μ2λ7.5−λ−30

formation through a constitutive law. Nedeker et al. [9] performed extensive mechanical testing of porcine kidney capsules, proposing an incompressible hyperelastic Ogden model [10]. For a capsule element under equi-biaxial loading subject to internal pressure [11], the constitutive law reads 2 σ=μ1λ15−λ−30+μ2λ7.5−λ−30 where μ1 = 0.0015 Torr (0.2 MPa) and μ2 = 0.0315 Torr (4.2 MPa) are material constants determined from experiments [6], and λ is the stretch ratio, defined as the ratio between the length of a segment after and before deformation. Let H and R be the undeformed capsule thickness and radius, respectively. The incompressibility of the tissue implies 3 hr=HR×1λ3 The increase in volume ΔV due to fluid injection relates to the stretch by 4 ΔV=λΔ3−1V0 where λΔ is the incremental stretch ratio. Let λ0 be the stretch ratio of the capsule before the fluid injection. Then, the total capsule stretch is 5 λ=λ0×λΔ By combining Equations 1 to 5, we obtain the following nonlinear expression for the renal pressure as a function of the injected fluid volume 6 p=2HRμ1λ0121+ΔvV04−λ0−331+ΔvV0−11+μ2λ04.51+ΔvV01.5−λ0−181+ΔvV0−6

where λΔ is the incremental stretch ratio. Let λ0 be the stretch ratio of the capsule before the fluid injection. Then, the total capsule stretch is 5 λ=λ0×λΔ By combining Equations 1 to 5, we obtain the following nonlinear expression for the renal pressure as a function of the injected fluid volume 6 p=2HRμ1λ0121+ΔvV04−λ0−331+ΔvV0−11+μ2λ04.51+ΔvV01.5−λ0−181+ΔvV0−6 The thickness of excised renal capsule has been previously reported [9], from which we set H = 100 μm. The parameters R = 6 × 105 μm, V0 = 8 mL, and λ0 = 1.07 were determined from a curve fit to the experimental results, as shown in Figure 4. It is important to remark that the value of the stretch ratio λ0 is within the range of residual strain found in many biological tissues [12].Figure 4 Comparison between the biomechanical model of the renal capsule, the capsulated and decapsulated kidneys. Data is expressed as mean value ± SEM.

s, as shown in Figure 4. It is important to remark that the value of the stretch ratio λ0 is within the range of residual strain found in many biological tissues [12].Figure 4 Comparison between the biomechanical model of the renal capsule, the capsulated and decapsulated kidneys. Data is expressed as mean value ± SEM. Discussion A major result of this work is the nonlinear relation between intrarenal pressure and volume in the intact kidney, which is well described by an exponential relation. This finding is indicative of a mechanical behavior commonly observed in organs confined by a rigid or semi-rigid continent, like the cranial and intra-abdominal compartments. As a proof of concept, the pressure-volume curve protocol was performed on two kidneys previously decapsulated. The resulting pressure-volume curve is well explained by a linear relation (R2 = 0.95), in contrast to nonlinear dependence found in intact kidneys. The proposed biomechanical model was able to explain the increase in intrarenal pressure for volume increments up to 3mL in the experiments performed. However, for higher increments of volume, there is an abrupt increase in pressure that cannot be explained by the mechanical confinement provided by the renal capsule alone. From a purely mechanical point of view, an abrupt increase of the pressure, and therefore of the membrane tension, can be explained by an increase of the membrane thickness supporting the applied pressure. This can be inferred from Equation 1, by noting that, for a given volume, stress σ and radius r are fixed, and thus the only way to increase the intrarenal pressure p is by increasing the continent thickness h. Since the capsule thickness cannot grow spontaneously and, in fact can only shrink as the capsule stress increases, an adjacent layer of parenchymal tissue must carry the additional overpressure, a phenomenon we term here as tissue recruitment (Figure 5).Figure 5 Comparative plot of pressure-volume curves for the intact kidney, the decapsultaed kidney, and the biomechanical model. The capsule confinement can explain the nonlinear relation between pressure and volume for small amounts of fluid. Tissue recruitment provides additional stiffness to the renal continent in order to bear the intrarenal overpressure.

ume curves for the intact kidney, the decapsultaed kidney, and the biomechanical model. The capsule confinement can explain the nonlinear relation between pressure and volume for small amounts of fluid. Tissue recruitment provides additional stiffness to the renal continent in order to bear the intrarenal overpressure. In AKI, a frequent finding is the increase in the kidney volume due to edema [4]. Further, hypoperfusion of the outer medulla is common in many forms of AKI [13-15]. In view of our results, hypoperfusion may be explained by a reduction of the renal perfusion pressure ostensibly caused by the increase of intrarenal pressure due to the volume increment. This idea is supported by the fact that the use of vasodilators to revert renal hypoperfusion has been ineffective to restore blood flow [14-18], indicating that renal perfusion is not controlled by vasomotor tone but rather by renal parenchymal pressure. Concordantly, recent studies in an ischemia-reperfusion murine model demonstrate that preventing the rise in intrarenal pressure caused by interstitial edema by making a small incision in renal capsule attenuates the risk of functional renal impairment. These findings suggest that a rise in parenchymal pressure may be contributed to the acute kidney injury caused by ischemic insult [7].

demonstrate that preventing the rise in intrarenal pressure caused by interstitial edema by making a small incision in renal capsule attenuates the risk of functional renal impairment. These findings suggest that a rise in parenchymal pressure may be contributed to the acute kidney injury caused by ischemic insult [7]. Conclusions We have studied the dependence of the intrarenal pressure on the fluid volume in the porcine intact kidney. A highly nonlinear relation between the intrarenal pressure and the injected volume was found, which confirms the existence of a mechanical behavior commonly observed in organs confined by a rigid or semi-rigid continent, which we refer to here as the renal compartment. In contrast, decapsulted kidneys present a pressure-volume linear relation, thus confirming the role of the renal capsule as a continent. From a biomechanical analysis, it can be concluded that the observed nonlinear pressure-volume relation cannot be solely explained by the confinement conferred by the renal capsule, suggesting that above a certain level of intrarenal pressure, tissue recruitment at the kidney periphery occurs in order to sustain higher levels of intrarenal pressure. The mechanical role of the renal capsule investigated in this work may have important implications in elucidating the role of decompressive capsulotomy in preventing the rapid intrarenal pressure increase in acutely injured kidneys (e.g., kidney transplantation). Future studies could assess the effect of renal decapsulation on renal blood flow, renal oxygenation and perfusion, microcirculation, and renal function in acutely injured kidneys.

f decompressive capsulotomy in preventing the rapid intrarenal pressure increase in acutely injured kidneys (e.g., kidney transplantation). Future studies could assess the effect of renal decapsulation on renal blood flow, renal oxygenation and perfusion, microcirculation, and renal function in acutely injured kidneys. Abbreviations AKIacute kidney injury RRrespiratory rate PaCO2partial pressure of carbon dioxide PaO2partial pressure of oxygen Competing interests The authors declare that they have no competing interests. Authors' contributions PC developed the study design, collected the data, helped with the statistical analysis and interpretation of data, and drafted and revised the manuscript. CS, PL, TS, and FL helped to design the study, contributed to data collection, and revised the manuscript. DH helped to design the study, performed the statistical analysis and biomechanical model, interpreted the data, and drafted and revised the manuscript. All authors read and approved the final manuscript. Acknowledgements The authors wish to thank Dr. Felipe Cavagnaro (Chairman of Pediatric Nephrology, Clínica Alemana de Santiago, Chile) and Dr. Francisco Cano (Chairman of Pediatric Nephrology, Hospital Luis Calvo Mackena, Chile) for their feedback and comments in the preparation of this manuscript. This work was supported by grants SOCHIPE 2012001 (PC), FIAC UAB 1102 (PL and FL), and Fondecyt 11121224 (DH).

Background Chronic obstructive pulmonary disease (COPD) is a leading cause of mortality and disability internationally and is predicted by 2020 to be the third most likely cause of death [1,2]. COPD is a progressive disease and is associated with increasing frequency and severity of exacerbations. Mechanical ventilation (MV), either invasive or non-invasive, may be a life-saving measure in managing respiratory failure due to an acute exacerbation of COPD [3]. However, mechanical ventilation is also associated with significant rates of morbidity and mortality. An improved understanding of the underlying pathophysiologic mechanisms of COPD is essential for the development of more effective and more individualized ventilation strategies for COPD patients.

erbation of COPD [3]. However, mechanical ventilation is also associated with significant rates of morbidity and mortality. An improved understanding of the underlying pathophysiologic mechanisms of COPD is essential for the development of more effective and more individualized ventilation strategies for COPD patients. Computer simulators that can accurately represent the particular disease state of an individual COPD patient could be an extremely valuable research tool for investigating the respiratory pathophysiology of COPD and predicting the effects of specific MV settings on the patient. Many researchers have worked on the development of physiological simulators, and various types of mathematical models have been proposed in the literature (e.g., [4–11]). However, these models generally employed only a very small number of compartments for gas exchange - in this paper, we show that such model cannot provide an accurate representation of the particular heterogeneous effects of COPD on alveolar mechanics. Also, most previous efforts to match physiological simulators to specific disease states have relied on manually manipulating the parameters in the simulator. Loeppky and co-workers [7] used a two-compartment model to match the COPD patient data generated from the multiple inert gas elimination technique (MIGET). Besides the use of a rather simplistic model, the study suffered from the use of only two parameters to represent V̇/Q mismatching; this is a significant limitation, since gas exchange in COPD patients is often characterized by different V̇/Q patterns with two or three modes [12]. In more sophisticated multi-compartmental models, matching has been achieved with some success by adjusting the resistance for 100 compartments manually [13], a challenging and time-consuming task. Indeed, it is obvious that manual matching is only practical for relatively simple models with a small set of adjustable parameters and a limited set of patient data for matching.

g has been achieved with some success by adjusting the resistance for 100 compartments manually [13], a challenging and time-consuming task. Indeed, it is obvious that manual matching is only practical for relatively simple models with a small set of adjustable parameters and a limited set of patient data for matching. A small number of previous studies have investigated the use of numerical optimisation approaches for the model matching task; a tidally breathing model has been matched to MIGET measurements from emphysema and embolism patients using the interior-reflective Newton algorithm [14], and two and three compartmental models were matched to data from intensive care patients with acute lung injury (ALI) using Brent's method [15]. These studies again used rather simple models with only a few compartments, and both studies employed very simple optimization algorithms, for which the quality of the matching achieved depended completely on the initial values chosen for the optimisation parameters. Since there is very little information available about how to choose these initial estimates, the likelihood of such approaches finding the best possible match to patient data is very small. Moreover, due to the simplified nature of the models used, continuous V̇/Q distributions could not be produced, and thus, the model outputs can only be compared to patient data at two or three isolated points on the V̇/Q curve.

likelihood of such approaches finding the best possible match to patient data is very small. Moreover, due to the simplified nature of the models used, continuous V̇/Q distributions could not be produced, and thus, the model outputs can only be compared to patient data at two or three isolated points on the V̇/Q curve. In this paper, we investigate the minimum level of complexity required in a computer simulation model in order to provide a representation of COPD pathophysiology that is accurate enough to allow the simulator to be used for studies on the design of novel therapeutic strategies for individual patients. In order to address this question, we employ a sophisticated simulation model of lung physiology that incorporates tidal ventilation, pulsatile pulmonary blood flow, hypoxic pulmonary vasoconstriction (HPV), a realistic and validated oxygen-hemoglobin model, and up to 200 individually configurable alveolar compartments. Our simulator has been developed over the past decade and has been used and validated in a number of previous studies [16,17,13]. In the present study, we couple this simulator to software implementing a global optimization algorithm [18], based on evolutionary principles (a genetic algorithm) which allows large numbers of model parameters to be simultaneously optimized in order to match the model outputs to detailed COPD patient data. An attractive feature of our approach from the point of view of clinical researchers and practitioners is that it can be highly automated, so that the user does not need to have detailed knowledge of the underlying algorithms but can use the software to fit models of varying complexity to different patient data sets. To illustrate how the proposed simulator could be used to investigate questions of clinical relevance, we present the results of an investigation into computing ventilator settings that optimally manage the trade-off between ensuring adequate gas exchange and minimizing the risk of ventilator-associated lung injury for an individual COPD patient.

ed simulator could be used to investigate questions of clinical relevance, we present the results of an investigation into computing ventilator settings that optimally manage the trade-off between ensuring adequate gas exchange and minimizing the risk of ventilator-associated lung injury for an individual COPD patient. Methods The computational simulator The simulation used in this study is a multi-compartmental computational model that uses an iterative technique to simulate integrated respiratory and cardiovascular pathophysiological scenarios [17,19,20]. A detailed description of the principles and mathematical equations underlying the computational model implemented in our simulator is provided in Additional file 1. In contrast to previous models of COPD pathophysiology that included only two or three alveolar compartments, our model allows the user to define the number of compartments (each with its own individual mechanical characteristics) to be implemented in the simulation. This allowed us to investigate in detail the relationship between the number of compartments in the model and its ability to match individual patient data. Each ith alveolar compartment has a unique and configurable bronchiolar resistance RB,i [cmH2O∙s/l], pulmonary vascular resistance RV,i[dyn∙s/cm5], stiffness index Si [cmH2O/ml2], and extrinsic pressure Pext,i [cmH2O](giving 4×N adjustable parameters in total, where N is the number of compartments). The ability to adjust these parameters individually across up to 200 alveolar compartments allows the model to recreate the heterogeneous effects of COPD on the overall physiology of the lung. The model also includes specific equations to represent the effects of alveolar collapse, threshold opening pressure, alveolar stiffening, and airway obstruction. The net effect of these components of the simulation is that the defining, clinical features of COPD may be observed in the model: alveolar gas trapping (with intrinsic positive end-expiratory pressure (PEEP)), collapse-reopening of alveoli (with gradual reabsorption of trapped gas if reopening does not occur), limitation of expiratory flow, and increased functional residual capacity - see Additional file 1 for further details.

erved in the model: alveolar gas trapping (with intrinsic positive end-expiratory pressure (PEEP)), collapse-reopening of alveoli (with gradual reabsorption of trapped gas if reopening does not occur), limitation of expiratory flow, and increased functional residual capacity - see Additional file 1 for further details. Patient data Two different sets of patient data from the literature are used in this study. The first dataset is from [21], where a 55-year-old patient is sedated and paralyzed, and relevant data for configuring the model and ventilator settings are shown in Table 1. For this case, we attempt to match the outputs of our simulator to the data reported for the following patient variables - PO2, PCO2, deadspace, shunt, mean and standard deviation of V̇ and Q, and ventilation-perfusion distribution across a number of ranges (see Table 2). The second set of patient data is from [12]. The 64-year-old patient is reported to be in a stable condition, and relevant model and ventilator configuration parameters are shown in Table 3. For this case, we attempt to match our model directly to patient V̇/Q curves generated via MIGET measurements.Table 1 Model configuration for the first patient dataset Parameter Value Respiratory frequency [bpm] 13 Tidal volume [ml] 590 FIO2 0.4 Inspiratory flow pattern Constant flow Cardiac output [l/min] 5.0 PEEP [cmH2O] 0 IE 1:3 RQ 0.8 Table 2 Matched parameter values and the reference data for the first patient dataset No. Parameter Data Model outputs ( N = number of compartments) N = 10 N = 25 N = 50 N = 100

Parameter Value Respiratory frequency [bpm] 13 Tidal volume [ml] 590 FIO2 0.4 Inspiratory flow pattern Constant flow Cardiac output [l/min] 5.0 PEEP [cmH2O] 0 IE 1:3 RQ 0.8 Table 2 Matched parameter values and the reference data for the first patient dataset No. Parameter Data Model outputs ( N = number of compartments) N = 10 N = 25 N = 50 N = 100 1 PaO2 [mm Hg] 125.2 165.29 143.18 133.73 127.92 2 PaCO2 [mm Hg] 46 43.08 44.73 43.62 45.76 3 Dead space fraction 64.9 61.24 62.76 61.28 65.28 4 Shunt fraction 6.8 12.9 9.02 8.07 7.2 5 mean_ V̇ 0.99 1 1 1 1 6 mean_Q 0.27 0.2 0.2 0.25 0.2 7 sd_V̇ 1 0.92 1.15 0.92 1.15 8 sd_Q 1.34 1.38 1.38 1.38 1.38 9 0.1 < V̇/Q < 1, V̇ 19.9 20.43 20.06 22.68 19.94 10 1 < V̇/Q < 10, V̇ 14.3 19.69 16.57 16.47 14.48 11 0.01 < V̇/Q < 0.1, P 15.8 22.63 18.86 14.64 16.16 12 0.1< V̇/Q < 1, P 62.1 55.21 58.12 65.57 64.92 13 1 < V̇/Q < 10, P 10.9 11.05 11.58 10.63 10.56 Total matching error 1.33 0.46 0.11 0.10 Simulation time [h] 11 32 41 67 Table 3 Model configuration for the second patient dataset Parameter Value Respiratory frequency [bpm] 16 Tidal volume [ml] 410 FiO2 0.21 Inspiratory flow pattern Constant flow Cardiac output [l/min] 3.4 PEEP [cmH2O] 0 IE 1:3 RQ 0.8

1 PaO2 [mm Hg] 125.2 165.29 143.18 133.73 127.92 2 PaCO2 [mm Hg] 46 43.08 44.73 43.62 45.76 3 Dead space fraction 64.9 61.24 62.76 61.28 65.28 4 Shunt fraction 6.8 12.9 9.02 8.07 7.2 5 mean_ V̇ 0.99 1 1 1 1 6 mean_Q 0.27 0.2 0.2 0.25 0.2 7 sd_V̇ 1 0.92 1.15 0.92 1.15 8 sd_Q 1.34 1.38 1.38 1.38 1.38 9 0.1 < V̇/Q < 1, V̇ 19.9 20.43 20.06 22.68 19.94 10 1 < V̇/Q < 10, V̇ 14.3 19.69 16.57 16.47 14.48 11 0.01 < V̇/Q < 0.1, P 15.8 22.63 18.86 14.64 16.16 12 0.1< V̇/Q < 1, P 62.1 55.21 58.12 65.57 64.92 13 1 < V̇/Q < 10, P 10.9 11.05 11.58 10.63 10.56 Total matching error 1.33 0.46 0.11 0.10 Simulation time [h] 11 32 41 67 Table 3 Model configuration for the second patient dataset Parameter Value Respiratory frequency [bpm] 16 Tidal volume [ml] 410 FiO2 0.21 Inspiratory flow pattern Constant flow Cardiac output [l/min] 3.4 PEEP [cmH2O] 0 IE 1:3 RQ 0.8 Automated matching to patient data Exacerbations of COPD are frequently associated with deterioration in gas exchange and associated hypoxemia. Unsurprisingly, increased inequality in V̇/Q relationships appears to be the major determinant of these changes [1]. Therefore, a key requirement for the simulation of COPD pathophysiology is the ability to accurately match the V̇/Q distributions seen in patient data. In our simulator, the V̇/Q distribution can be manipulated by adjusting the bronchiolar resistance and pulmonary vascular resistance, stiffness, and extrinsic pressure for each compartment in the model. For example, by increasing vascular resistance in a region of the lung, an area of relative dead space can be created. While all parameters can be manually adjusted, this becomes impractical as the number of compartments N in the model increases, since the total number of parameters is 4×N.

ch compartment in the model. For example, by increasing vascular resistance in a region of the lung, an area of relative dead space can be created. While all parameters can be manually adjusted, this becomes impractical as the number of compartments N in the model increases, since the total number of parameters is 4×N. To address this issue, we formulate the model-matching problem as an optimization problem, where the difference between the model outputs and the data is captured in a cost function, and the model parameters that can be varied are the variables for the optimization problem. As mentioned above, two sets of patient data are considered in this study. For both cases, a primary focus is on accurately representing the imbalance in the V̇/Q distribution caused by the disease state. In the first dataset [21], the horizontal axis of the V̇/Q diagram is divided into several segments on which the percentage of V̇/Q indicated in the data will be matched. Other parameters that also need to be matched include PaO2, PaCO2, and mean and standard deviation of ventilation and perfusion. For the second case, the whole V̇/Q curve is considered (i.e., every point on the curve is to be matched). The matching error can then be defined based on these data, which is given as: ET=∑i=1nEi2 where ET is the total residual error representing the matching accuracy, Ei=xi−xidxid is the error for parameter i, xi is the model output value for parameter i, and xid is the value of the data for parameter i.

To address this issue, we formulate the model-matching problem as an optimization problem, where the difference between the model outputs and the data is captured in a cost function, and the model parameters that can be varied are the variables for the optimization problem. As mentioned above, two sets of patient data are considered in this study. For both cases, a primary focus is on accurately representing the imbalance in the V̇/Q distribution caused by the disease state. In the first dataset [21], the horizontal axis of the V̇/Q diagram is divided into several segments on which the percentage of V̇/Q indicated in the data will be matched. Other parameters that also need to be matched include PaO2, PaCO2, and mean and standard deviation of ventilation and perfusion. For the second case, the whole V̇/Q curve is considered (i.e., every point on the curve is to be matched). The matching error can then be defined based on these data, which is given as: ET=∑i=1nEi2 where ET is the total residual error representing the matching accuracy, Ei=xi−xidxid is the error for parameter i, xi is the model output value for parameter i, and xid is the value of the data for parameter i. Global optimization algorithms can then be used to find model parameter values that minimize the value of ET, i.e., minimize the difference between the model outputs and the data. The procedure is illustrated in Figure 1 - in each iteration, a set of parameter combinations are sent to the simulator, and the outputs from the simulator are evaluated by the optimization algorithm which then generates the updated parameter values for the next iteration until the termination criterion is reached (i.e., the condition to get a best matching is found).Figure 1 Conceptual representation of the model matching process.

ator, and the outputs from the simulator are evaluated by the optimization algorithm which then generates the updated parameter values for the next iteration until the termination criterion is reached (i.e., the condition to get a best matching is found).Figure 1 Conceptual representation of the model matching process. In the model-matching procedure, the model parameters are allowed to vary continuously between physiologically realistic upper and lower bounds - see Table 4. COPD is characterized by restrictions to flow in airways (bronchi and bronchioles) due to increased excretion of mucus and inflammation of the bronchiolar walls, effectively reducing the radius of the tube through which the airways ventilate the alveoli. Based on the Hagen-Pouiselle relationship:Table 4 Model parameters – nominal values and allowable ranges (for N = 100) Parameters Nominal value Variation ranges Bronchiolar resistance R Bi [cmH2O∙s/l] 600 [300, 3 × 105] Vascular resistance R Vi [dyn∙s/cm5] 1.6 × 104 [8 × 103, 8 × 106] Stiffness coefficient S i [cmH2O/ml2] 0.05 [0.025, 0.15] Extrinsic pressure P ext,i [cm H2O] 28.8 [−20, 28.8] R∝ηLr4

In the model-matching procedure, the model parameters are allowed to vary continuously between physiologically realistic upper and lower bounds - see Table 4. COPD is characterized by restrictions to flow in airways (bronchi and bronchioles) due to increased excretion of mucus and inflammation of the bronchiolar walls, effectively reducing the radius of the tube through which the airways ventilate the alveoli. Based on the Hagen-Pouiselle relationship:Table 4 Model parameters – nominal values and allowable ranges (for N = 100) Parameters Nominal value Variation ranges Bronchiolar resistance R Bi [cmH2O∙s/l] 600 [300, 3 × 105] Vascular resistance R Vi [dyn∙s/cm5] 1.6 × 104 [8 × 103, 8 × 106] Stiffness coefficient S i [cmH2O/ml2] 0.05 [0.025, 0.15] Extrinsic pressure P ext,i [cm H2O] 28.8 [−20, 28.8] R∝ηLr4 the resistance (R) in a single tube is directly proportional to the length (L) of the tube and the viscosity (η) and inversely proportional to the fourth power of the radius (r4). To achieve the required precision in matching the extreme ranges of V̇/Q distributions for the second dataset, we have investigated the effect of increasing the upper bound on compartmental resistances from 500 to 2,000 times their nominal values, i.e., the radius of the airways can be reduced by as much as 85%. To ensure that this does not result in model configurations that produce unrealistic values for total lung resistance, we incorporated a constraint function into the optimization algorithm that limits the increase in the total lung inflow resistance to ten times its basic (healthy) value, as expected in patients with COPD [22].

that this does not result in model configurations that produce unrealistic values for total lung resistance, we incorporated a constraint function into the optimization algorithm that limits the increase in the total lung inflow resistance to ten times its basic (healthy) value, as expected in patients with COPD [22]. Previous attempts to use optimization for matching models of pulmonary disease states to patient data have used rather simple algorithms that require good initial ‘guesses’ for the parameters in order to be effective. In the case of COPD, however, current understanding of the associated pathophysiology provides little guidance into how to choose initial values for the model parameters across large numbers of alveolar compartments. In this study, we therefore employed an advanced global optimisation algorithm known as a genetic algorithm. This general purpose stochastic search and optimization procedure, based on genetic and evolutionary principles [23], has been shown to have a much higher chance of finding optimal solutions for difficult problems with large numbers of variables. Full details of the particular optimization algorithm used in this study and how it was implemented with the model are provided in Additional file 1.

genetic and evolutionary principles [23], has been shown to have a much higher chance of finding optimal solutions for difficult problems with large numbers of variables. Full details of the particular optimization algorithm used in this study and how it was implemented with the model are provided in Additional file 1. Using the simulator for clinical investigations To illustrate how the simulator can be used for clinical investigations, we consider the problem of identifying ventilator settings that minimize the risk of ventilator-associated lung injury (VALI). We consider the following five key ventilator settings as variable parameters that may be adjusted to optimize the trade-off between effective gas exchange and minimizing the risk of VALI: (1) tidal volume (Vtidal, [ml]) - the volume of air traveling in or out of the patient's lungs during every breath; (2) ventilation rate (VentRate, [breaths/min]) - the number of breaths per minute; (3) duty cycle (I:E) - the ratio of inspiratory time to total ventilatory cycle duration; (4) PEEP, [cmH2O] - the positive pressure in the lungs at the end of exhalation; and (5) fraction of inspired oxygen (FIO2) - the fraction of oxygen constituting the inhaled volume of gas as provided by the mechanical ventilator.

(3) duty cycle (I:E) - the ratio of inspiratory time to total ventilatory cycle duration; (4) PEEP, [cmH2O] - the positive pressure in the lungs at the end of exhalation; and (5) fraction of inspired oxygen (FIO2) - the fraction of oxygen constituting the inhaled volume of gas as provided by the mechanical ventilator. The maximum allowable ranges of variation for the values of these parameters have been defined based on current clinical practice and to be consistent with data available from clinical trials [24,25]. Vtidal is allowed to vary within a range from 390 to 650 ml, corresponding to 6 to 10 ml/kg for a body weight of 65 kg. VentRate is bounded within the range 9 to 16 breaths/min, I:E is limited to the interval 0.25 to 0.5 (i.e., a ratio between 1:4 and 1:2), PEEP is constrained within 0 to 5 cmH2O, and FIO2 is bounded within 0.21 to 1. A summary is provided in Table 5.Table 5 MV setting parameter variation bounds and desired model outputs Variation ranges Acceptable values Desired MV setting parameters Vt [ml] [390, 650] VentRate [bpm] [9,16] I:E [0.25, 0.5] PEEP [cmH2O] [0, 5] FiO2 [0.21, 1] Model outputs PO2 [kPa] >8 12 PCO2 [kPa] >4, <8 5.3 Palv [kPa] <4 -

The maximum allowable ranges of variation for the values of these parameters have been defined based on current clinical practice and to be consistent with data available from clinical trials [24,25]. Vtidal is allowed to vary within a range from 390 to 650 ml, corresponding to 6 to 10 ml/kg for a body weight of 65 kg. VentRate is bounded within the range 9 to 16 breaths/min, I:E is limited to the interval 0.25 to 0.5 (i.e., a ratio between 1:4 and 1:2), PEEP is constrained within 0 to 5 cmH2O, and FIO2 is bounded within 0.21 to 1. A summary is provided in Table 5.Table 5 MV setting parameter variation bounds and desired model outputs Variation ranges Acceptable values Desired MV setting parameters Vt [ml] [390, 650] VentRate [bpm] [9,16] I:E [0.25, 0.5] PEEP [cmH2O] [0, 5] FiO2 [0.21, 1] Model outputs PO2 [kPa] >8 12 PCO2 [kPa] >4, <8 5.3 Palv [kPa] <4 - Three key physiological indicators are also defined. To monitor effective arterial oxygenation, partial pressure of oxygen, PaO2, needs to be considered. In order to maintain effective arterial oxygenation, PaO2 is constrained to be higher than 8 kPa, with a desired value of 12 kPa. Arterial partial pressure of carbon dioxide, PaCO2 is another key indicator of alveolar ventilation that also indirectly reflects acid-base balance. PaCO2 is constrained to be between 4 and 8 kPa with a desired value of 5.3 kPa. The risk of barotrauma is proportional to the peak alveolar pressure, (Palv, kPa above atmospheric pressure), and Palv is limited to 4 kPa, where Palv is calculated as the average of the peak pressure in the most highly pressurized 25% of all alveoli.

nstrained to be between 4 and 8 kPa with a desired value of 5.3 kPa. The risk of barotrauma is proportional to the peak alveolar pressure, (Palv, kPa above atmospheric pressure), and Palv is limited to 4 kPa, where Palv is calculated as the average of the peak pressure in the most highly pressurized 25% of all alveoli. Requirements on the above physiological indicators can be captured as an optimization problem and formulated mathematically as: minJ1J2 where J1=w1PaO2−12+w2|PaCO2−5.3| J2=w3Palv Large values of J1 will be produced by ventilator settings that provide poor gas exchange, while large values of J2 will be produced by combinations of ventilator settings that cause high peak alveolar pressures (and hence increase the risk of VALI). By requiring both J1 and J2 to be minimized simultaneously, we can search for combinations of ventilator settings that optimally manage the trade-off between effective gas exchange and minimizing the risk of VALI. w1,w2, and w3 are weighting functions that are used in the optimization process to ensure that equal priority is given to each of the different objectives. Since we are trying to minimize two objectives J1 and J2 at the same time, a multi-objective optimization algorithm called non-dominated sorting genetic algorithm II (NSGA-II) was used here [26].

ctions that are used in the optimization process to ensure that equal priority is given to each of the different objectives. Since we are trying to minimize two objectives J1 and J2 at the same time, a multi-objective optimization algorithm called non-dominated sorting genetic algorithm II (NSGA-II) was used here [26]. Results Matching results for the first dataset For the first dataset, the ventilation-perfusion distribution is characterized by several segments based on the V̇/Q ratio, which includes the amount of ventilation in the ranges 0.1< V̇/Q < 1 and 1< V̇/Q < 10, and the amount of perfusion in the ranges 0.01< V̇/Q < 0.1, 0.1< V̇/Q < 1, and 1< V̇/Q < 10 together with the shunt and dead space. Other data considered are the PO2, PCO2, mean V̇/Q ratio of the pulmonary blood flow, and ventilation distribution as well as standard deviations (dispersion) of pulmonary blood flow and ventilation. The simulation was performed separately for four cases with the total number of compartments in the model being varied between 10, 25, 50, and 100 to represent different levels of model complexity.