Browse the corpus

We are proud to introduce Annals of Surgical Innovation and Research [1], a new forum for cutting-edge findings and innovations in all surgical specialties. This journal is entirely open access, peer-reviewed and online. We are interested in articles from all areas of surgery, minimal invasive techniques and endoscopy, and related bioengineering. We want to promote the exchange of ideas, concepts and findings across all surgical specialties. The following types of articles will be published: research, short communications, hypotheses, commentaries, reviews (as invited articles), case reports, case studies, meeting reports, letters to the editor and book reviews. Open access provides an unparalleled opportunity to present information to specialists and the public. The online appearance of Annals of Surgical Innovation and Research (ASIR) allows the immediate publication of accepted articles and the presentation of large amounts of data, multimedia files and supplemental information. But do we really need another journal in the field of medicine?

Open access provides an unparalleled opportunity to present information to specialists and the public. The online appearance of Annals of Surgical Innovation and Research (ASIR) allows the immediate publication of accepted articles and the presentation of large amounts of data, multimedia files and supplemental information. But do we really need another journal in the field of medicine? Yes we do, because this open access journal will be accessible to everyone, and not just those with access to a library with a subscription. As such, ASIR will help to increase public interest in, and support of, research. Many major founders, such as the NIH and the Wellcome Trust, now have policies to encourage grant holders to release their research to the public as soon as possible after publication [2-4]. These policies are likely to be replicated by other major founders, especially following the recommendations of the UK House of Commons Committee [5]. A country's economy will not influence its researchers' ability to access articles in ASIR, because resource-poor countries (and institutions) will be able to read the same material as wealthier ones [6]).

Yes we do, because this open access journal will be accessible to everyone, and not just those with access to a library with a subscription. As such, ASIR will help to increase public interest in, and support of, research. Many major founders, such as the NIH and the Wellcome Trust, now have policies to encourage grant holders to release their research to the public as soon as possible after publication [2-4]. These policies are likely to be replicated by other major founders, especially following the recommendations of the UK House of Commons Committee [5]. A country's economy will not influence its researchers' ability to access articles in ASIR, because resource-poor countries (and institutions) will be able to read the same material as wealthier ones [6]). Our ambitious aim is to function as an ideal bridge to close the gap between the revolution in surgical innovations and research, especially in minimal invasive surgical technology, and to help surgeons at the front line of patient care. The journal tries to centralize new knowledge in one forum to give a quick and comprehensive review over new innovations and relevant research. The scope of the journal is to advance the quickly developing field of new surgical techniques and innovations in a more and more technically demanding field. It offers an opportunity to publish large datasets, large numbers of illustrations and even movies, and therefore to act as the primary source for surgical innovations to be used by clinicians and researchers.

ly developing field of new surgical techniques and innovations in a more and more technically demanding field. It offers an opportunity to publish large datasets, large numbers of illustrations and even movies, and therefore to act as the primary source for surgical innovations to be used by clinicians and researchers. We recognize the need for a dynamic forum that fosters the exchange of ideas across the entire surgical community, a journal that is open access and online. We now have an opportunity to make Annals of Surgical Innovation and Research just such a journal, and this is an exciting challenge. Assisting me in this new venture is an editorial board, which consist of members out of 14 different countries, many of them true leaders, others very enthusiastic and soon to be. For Annals of Surgical Innovation and Research to become an exciting forum for surgical research and innovation, we need you to contribute your most innovative and significant work in your field. In exchange, Annals of Surgical Innovation and Research will provide a fast, fair, and constructive review process; the highest quality production values; and a dedicated team of professional scientific editors, who are willing to discuss your research. All manuscripts are reviewed on the basis of scientific merit and held to the highest standards of excellence and editorial consistency. Upon acceptance they will be immediately available in PubMed and indexed by several other services.

team of professional scientific editors, who are willing to discuss your research. All manuscripts are reviewed on the basis of scientific merit and held to the highest standards of excellence and editorial consistency. Upon acceptance they will be immediately available in PubMed and indexed by several other services. I would like to take this opportunity to acknowledge our Editorial Board [7] for their commitment to the journal and to express my most sincere gratitude to the entire staff of BioMed Central [8] for their valuable advice and support in launching this journal. Special thanks to Christian Schmidt, Deputy Editor of Molecular Cancer [9], who guided me through the initial steps of this adventure.

Background Laparoscopic approach had a significant impact in the modern treatment of most abdominal pathologies, including gastroesophageal reflux disease (GERD). Multiple well-designed trials have confirmed that laparoscopic Nissen fundoplication (LNF) compares favorably with its open counterpart [1-3]. However, a small but significant number of patients (12% at 5 years in our initial experience) have persisting or recurrent GERD symptoms after antireflux surgery [4,5]. When a surgical complication has been ruled out, poor patient selection, psychological disorders associated or mimicking GERD (i.e. somatization, anxiety or depressive disorders), can play a role. Pre-operative identification of such psychological disorders could help orient patients to a psychological therapy instead of surgery. The aim of this study is to identify psychological factors associated with poor surgical outcomes after antireflux surgery. Materials and methods This is a single institution prospective controlled trial, comparing patients undergoing a laparoscopic Nissen fundoplication for GERD (LNF Group, n = 17, 13 females and 4 males) to patients undergoing an elective laparoscopic cholecystectomy for biliary colic (Control Group, n = 10, 9 females and 1 male). Patients were prospectively recruited between March 2002 and May 2003, from the outpatient consultations of a University Medical Centre (Centre for Minimal Access Surgery, McMaster University, Hamilton, Ontario).

g an elective laparoscopic cholecystectomy for biliary colic (Control Group, n = 10, 9 females and 1 male). Patients were prospectively recruited between March 2002 and May 2003, from the outpatient consultations of a University Medical Centre (Centre for Minimal Access Surgery, McMaster University, Hamilton, Ontario). The LNF Group was recruited among patients who were considering surgical treatment for the management of classical symptoms of GERD. All patients had pre-operative 24-hour pH monitoring, with GERD defined as a percentage time with pH<4 greater than 4% during 24 h, and a Demeester score greater that 14.3. Preoperative workup also included esophageal manometry and esophago-gastro-duodenoscopy. A detailed questionnaire was used to assess GERD specific symptoms (heartburn, pain, regurgitation, dysphagia and fullness) and non-GERD symptoms (vomiting, constipation, diarrhea, loss of appetite and cough). Each symptom was scored as a product of severity (0 = none, 3 = severe) and frequency (0 = none, 4 = daily), for a GERD symptoms score ranging between 0 and 72 and a non-GERD symptoms score ranging between 0 and 48. This symptom score has been validated previously [6].

nstipation, diarrhea, loss of appetite and cough). Each symptom was scored as a product of severity (0 = none, 3 = severe) and frequency (0 = none, 4 = daily), for a GERD symptoms score ranging between 0 and 72 and a non-GERD symptoms score ranging between 0 and 48. This symptom score has been validated previously [6]. Patients with post-operative GERD symptoms score above 12 at 6 months (LNF Poor outcomes Group) were compared to the rest of the LNF Group (LNF Good outcomes Group). A score of 12 was chosen based on our previous experience with the GERD symptoms score (SS)[6]. The SS showed a high responsiveness index, after LNF, which indicate that the instrument is highly sensitive. The minimal clinically important difference is about seven scale points. A 12 points score (on a 0 to 72 points scale) was considered to be a good cut-off point to consider that patients have ongoing or recurrent GERD symptoms. The Control Group consisted in patients with typical biliary pain with radiologic evidence of cholelithiasis. This group represents patients with a specific disease entity that is expected to be cured with a surgical treatment. Patients with known psychiatric or psychologic disorders were excluded, as well as patients with chronic pain syndrome. The two groups had the same psychological assessment one month before surgery, at 3 months and 6 months after surgery. The following tests were administered:

The Control Group consisted in patients with typical biliary pain with radiologic evidence of cholelithiasis. This group represents patients with a specific disease entity that is expected to be cured with a surgical treatment. Patients with known psychiatric or psychologic disorders were excluded, as well as patients with chronic pain syndrome. The two groups had the same psychological assessment one month before surgery, at 3 months and 6 months after surgery. The following tests were administered: - Symptom CheckList-90-R, somatization subset (SCL-90-R): a 12-item questionnaire that measures a current point-in-time somatization symptom complex as outlined in the DSM IV-R [7]. Each item is scored on a 5-point scale (0 to 4). A raw score can be calculated by dividing the summed total score by the number of items. That score is then referred to gender-appropriate norms for conversion to standard t-score, which may alternatively be expressed as a percentile. T-scores equal to or greater than 63 (90th percentile) for the global score index (on the 90 items) or in any two symptom dimensions warrants clinical workup. In this study, the summed total score was used for evaluation of somatization level.

n to standard t-score, which may alternatively be expressed as a percentile. T-scores equal to or greater than 63 (90th percentile) for the global score index (on the 90 items) or in any two symptom dimensions warrants clinical workup. In this study, the summed total score was used for evaluation of somatization level. - Depression Anxiety Stress Scales (DASS): a 21-item instrument for distinguishing between features of depression, physical arousal, and psychological tension and agitation in clinical and non-clinical groups. It has acceptable to excellent internal consistency and concurrent validity [8]. Clinical range, based on normative data, is defined as scores of 16.57 or above for stress, 12.75 or above for anxiety, and 9.26 or above for depression. - Anxiety Sensitivity Index (ASI): a 16-item self-report questionnaire measuring fear of anxiety-related symptoms. Each item is scored using a five point Likert type scale with endpoints of "very little" (0 points) and "very much" (4 points). The ASI yields a total score obtained by summing the rating across all items, with higher scores reflecting higher level of fear and concern regarding physical symptoms of anxiety. Psychometric data is available for community and clinical samples [9];

th endpoints of "very little" (0 points) and "very much" (4 points). The ASI yields a total score obtained by summing the rating across all items, with higher scores reflecting higher level of fear and concern regarding physical symptoms of anxiety. Psychometric data is available for community and clinical samples [9]; - Illness Attitude Scale (IAS): a 29-item questionnaire that measures attitudes, fears and beliefs associated with the psychopathology of hypochondriasis and that of abnormal illness behavior. All items are self-rated on 5 points Likert scales, with endpoints of "no" (zero points) and "most of the times" (four points). The scores on the two subscales (Health Anxiety and Illness Behaviour) appear to differentiate between general medical outpatients (mean of 10.9 and 9.0 respectively), general practice patients (11.3 and 6.8) and subjects from the general population (9.1 and 4.7) [10,11]. - Beck Depression Inventory II (BDI): a 21-item scale designed to measure clinical depression, shown to have excellent psychometric properties. The original authors recommended cut-off scores for non depressed (<10), mildly depressed (10–14), moderately depressed (15–22) and severely depressed (23 +) [12]. Ethic Board approval was obtained to collect these data.

- Beck Depression Inventory II (BDI): a 21-item scale designed to measure clinical depression, shown to have excellent psychometric properties. The original authors recommended cut-off scores for non depressed (<10), mildly depressed (10–14), moderately depressed (15–22) and severely depressed (23 +) [12]. Ethic Board approval was obtained to collect these data. A similar technique of laparoscopic cholecystectomy or laparoscopic Nissen fundoplication was used in all patients. In short, the surgical technique of LNF consisted in a 2 to 3 cm, 360 degree wrap with selective division of short gastric vessels and crura repair [5]. The technique of laparoscopic cholecystectomy included the dissection of Callot's triangle, identification and control of the cystic duct and cystic artery and retrograde dissection of the gallbladder bed. Data were entered in an electronic database and results are expressed as a Mean ± Standard Deviation. Statistical analysis was performed using a Student T tests and statistical significance was set at 0.05. Results The results of the pre-operative and 6-months psychological tests are summarized in Table 1. There was no significant difference between the LNF Group and the Control group. Five patients in the LNF group had abnormal Beck depression inventory's scores that could suggest depressive disorders (mild (2), moderate (2) and severe (1). Three patients normalized their BDI score 6 months after surgery, one patient went from a score of 19 (severe) to 10 (mild) and the last one had an increase from 10 (mild) to 17 (moderate).

up had abnormal Beck depression inventory's scores that could suggest depressive disorders (mild (2), moderate (2) and severe (1). Three patients normalized their BDI score 6 months after surgery, one patient went from a score of 19 (severe) to 10 (mild) and the last one had an increase from 10 (mild) to 17 (moderate). Table 1 Psychological scores before surgery and at 6 months (Mean ± Standard Deviation). LNF Group Control Group Pre-op 6 months Pre-op 6 months SCL-90 12 ± 9 5.5 ± 4.6 7.9 ± 4.6 7.2 ± 5.5 DASS 10.7 ± 8.8 5.3 ± 8.8 5.6 ± 4 10.8 ± 13 Anxiety SI 12.2 ± 8.8 8.8 ± 10.2 8.6 ± 6 10.6 ± 11 Illness AS 30.8 ± 11 24.3 ± 9.9 33.4 ± 10 30.3 ± 16 BDI 8 ± 7 3.7 ± 4.5 6.4 ± 5.2 8.7 ± 11.9 There was also a significant decrease in GERD symptoms score in the LNF Group (39.8 ± 13.2 preop versus 15.7 ± 17 at 6 months, p < 0.05), but also a significant decrease in the non-GERD symptoms score (11.8 ± 10.6 preop versus 4.3 ± 8.4 at 6 months, p < 0.05). This was correlated with a significant drop in the percentage time below pH4 during 24 h and DeMeester score (p < 0.05) (Table 2). Table 2 Evolution in 24 h pH monitoring (Mean ± Standard Deviation). 24 h pH monitoring DeMeester score Pre-op 6 months Pre-Op 6 months

SCL-90 12 ± 9 5.5 ± 4.6 7.9 ± 4.6 7.2 ± 5.5 DASS 10.7 ± 8.8 5.3 ± 8.8 5.6 ± 4 10.8 ± 13 Anxiety SI 12.2 ± 8.8 8.8 ± 10.2 8.6 ± 6 10.6 ± 11 Illness AS 30.8 ± 11 24.3 ± 9.9 33.4 ± 10 30.3 ± 16 BDI 8 ± 7 3.7 ± 4.5 6.4 ± 5.2 8.7 ± 11.9 There was also a significant decrease in GERD symptoms score in the LNF Group (39.8 ± 13.2 preop versus 15.7 ± 17 at 6 months, p < 0.05), but also a significant decrease in the non-GERD symptoms score (11.8 ± 10.6 preop versus 4.3 ± 8.4 at 6 months, p < 0.05). This was correlated with a significant drop in the percentage time below pH4 during 24 h and DeMeester score (p < 0.05) (Table 2). Table 2 Evolution in 24 h pH monitoring (Mean ± Standard Deviation). 24 h pH monitoring DeMeester score Pre-op 6 months Pre-Op 6 months Whole group 7.4 ± 4.6 1.4 ± 2.5 30.4 ± 19.1 6.3 ± 12.1 LNF Poor outcomes 7.3 ± 9.6 0.8 ± 0.8 34.3 ± 40 3 ± 2.8 LNF Good outcomes 7.5 ± 3.3 1.7 ± 2.4 28.7 ± 12.8 7.9 ± 11.8 In the LNF Group, 7 patients had a GERD symptoms score greater than 12 at 6 months (on a 0 to 72 scale) and constitute the LNF-Poor outcomes Group. This group was compared to the other 10 patients in the LNF Group (LNF-Good outcomes Group). Changes in 24 h pH results and DeMeester scores are summarized in Table 2. None of the patients with ongoing GERD symptoms had objective acid reflux, as demonstrated by normal 24 h pH monitoring. The evolution of GERD and non-GERD symptoms score in each group is summarized in Figure 1 and 2 respectively. The difference in non-GERD symptoms scores was not significant. Figure 1 GERD Score, Good versus Poor outcomes Groups (Mean).

Whole group 7.4 ± 4.6 1.4 ± 2.5 30.4 ± 19.1 6.3 ± 12.1 LNF Poor outcomes 7.3 ± 9.6 0.8 ± 0.8 34.3 ± 40 3 ± 2.8 LNF Good outcomes 7.5 ± 3.3 1.7 ± 2.4 28.7 ± 12.8 7.9 ± 11.8 In the LNF Group, 7 patients had a GERD symptoms score greater than 12 at 6 months (on a 0 to 72 scale) and constitute the LNF-Poor outcomes Group. This group was compared to the other 10 patients in the LNF Group (LNF-Good outcomes Group). Changes in 24 h pH results and DeMeester scores are summarized in Table 2. None of the patients with ongoing GERD symptoms had objective acid reflux, as demonstrated by normal 24 h pH monitoring. The evolution of GERD and non-GERD symptoms score in each group is summarized in Figure 1 and 2 respectively. The difference in non-GERD symptoms scores was not significant. Figure 1 GERD Score, Good versus Poor outcomes Groups (Mean). Figure 2 Non-GERD Score, Good versus Poor Outcomes Groups (Mean). The results of psychological tests were also compared between the two groups. A significantly higher preoperative SCL-90-R score was found in patients with ongoing symptoms (18.2 versus 8.3, p < 0.05, Figure 3), reflecting a higher level of somatization in these patients. There was no significant difference in the other psychological tests. Figure 3 SCL-90 Score, Good versus Poor Outcomes Groups (Mean).

The results of psychological tests were also compared between the two groups. A significantly higher preoperative SCL-90-R score was found in patients with ongoing symptoms (18.2 versus 8.3, p < 0.05, Figure 3), reflecting a higher level of somatization in these patients. There was no significant difference in the other psychological tests. Figure 3 SCL-90 Score, Good versus Poor Outcomes Groups (Mean). Discussion Psychological disorders can play a role in a variety of gastrointestinal illness, including esophageal diseases. The influence of psychological factors in gastroesophageal reflux is unknown. Some studies suggest that both acid secretion and gastric motility are affected by emotions or stress, with some personality traits associated with labile basal acid secretion rate [13]. McDonald et al [14] also showed that relaxation training can alter esophageal acid exposure and reflux symptoms reports in patients with GERD.

ies suggest that both acid secretion and gastric motility are affected by emotions or stress, with some personality traits associated with labile basal acid secretion rate [13]. McDonald et al [14] also showed that relaxation training can alter esophageal acid exposure and reflux symptoms reports in patients with GERD. Baker et al [15] did a psychological assessment in 51 patients with GERD and in 43 age-matched controls. His results suggested that although most patients with GERD were psychologically similar to patients without GERD, a subset of psychologically distressed patients (depression, somatization, anxiety and intensity of reporting symptom distress, n = 15) were more likely to be found among patients with GERD. Similarly, Johnson et al[16] reported a greater rate of phobia, obsessionality and somatization disorders in patients who seek medical attention for heartburn, than in those who do not consult. Kamolz et al[17] evaluated the outcomes of laparoscopic antireflux surgery in patients with known anxiety disorders. In that study, surgery improved the quality of life and corrected the acid reflux with however less symptoms relief in some of the patients.

cal attention for heartburn, than in those who do not consult. Kamolz et al[17] evaluated the outcomes of laparoscopic antireflux surgery in patients with known anxiety disorders. In that study, surgery improved the quality of life and corrected the acid reflux with however less symptoms relief in some of the patients. These results suggest that psychological factors can influence the perception and/or the severity of GERD. In this specific study, there was no significant difference in the preoperative psychological assessment between patients suffering from GERD and the control group. However, a subset of patients experienced ongoing GERD symptoms and this specific group had a significantly higher level of somatization, as demonstrated by a higher SCL-90-R, somatization subset score. GERD and somatization disorders Somatoform disorders are defined as psychiatric disorders in which one or more unexplained physical symptoms are the central defining feature of the disease. They frequently have abdominal manifestations, occurring in 17% to 28% of patients with irritable bowel syndrome[18]. They can also influence the surgical outcomes, by altering the patient's perception of pre- and post-operative symptoms and pain[19]. Clouse et al[20] demonstrated that patients with esophageal motility disorders score high on scales of somatic anxiety, depression and somatization. Baker et al[15] also found that somatization disorders are more likely to be found among patients with GERD.

s perception of pre- and post-operative symptoms and pain[19]. Clouse et al[20] demonstrated that patients with esophageal motility disorders score high on scales of somatic anxiety, depression and somatization. Baker et al[15] also found that somatization disorders are more likely to be found among patients with GERD. In this specific study, the SCL-90-R, somatization subset was used to assess somatoform disorders. It is a self-report questionnaire, developed by Clinical Psychometric Research in 1976[21]. It is a reliable tool designed to reflect psychologic symptom patterns of individuals. The somatization subset used in this study measures somatization symptoms complex. Patients who experienced ongoing GERD symptoms after surgery had a significantly higher score on the somatization scale, compared to both the rest of the LNF Group and the Control Group.

flect psychologic symptom patterns of individuals. The somatization subset used in this study measures somatization symptoms complex. Patients who experienced ongoing GERD symptoms after surgery had a significantly higher score on the somatization scale, compared to both the rest of the LNF Group and the Control Group. GERD and depression disorders Depression is one of the most frequent psychological disorder in the general population. As a result, its impact on GERD and surgical outcomes has been evaluated before. Kamolz et al[22] compared GERD symptoms in patients who had both GERD and major depression to a control group, before and after laparoscopic fundoplication. Atypical GERD symptoms (chest pain: 81.6% vs 37.4% and bloating: 92.2% vs 37.4%) were more predominant and graded as much more severe among patients with major depression. Also, all symptoms but dysphagia significantly improved after laparoscopic antireflux surgery in both groups but patients with major depression had a significantly higher rate of chest pain, bloating, and dysphagia. Power et al[23], in a series of 131 patients who had a laparoscopic Nissen fundoplication, identified preoperative psychiatric history (major depression but also any psychogenic stressor) as an accurate predictor of LNF failure. Psychological history, especially with patients who have severe or refractory GERD symptoms, a positive review of systems or disrupted quality of life is thus essential. Patients with refractory GERD symptoms who have psychological issues and impaired quality of life may benefit more form integrated psychological care than referral to procedure oriented specialist[24].

have severe or refractory GERD symptoms, a positive review of systems or disrupted quality of life is thus essential. Patients with refractory GERD symptoms who have psychological issues and impaired quality of life may benefit more form integrated psychological care than referral to procedure oriented specialist[24]. In this specific study, there was no significant difference in the scores of Depression Anxiety Stress Scales and Beck Depression Inventory between any of the groups. However, five patients in the LNF group had abnormal Beck depression inventory's scores that could suggest depressive disorders. Three patients normalized their BDI score 6 months after surgery. In conclusion, the SCL-90-R Somatization Subset, measuring the level of somatization in a patient, may be useful to predict poor outcomes after antireflux surgery. Cognizance of psychological disorders could improve the selection of an optimal treatment for GERD and help reduce the rate of ongoing symptoms after LNF. Larger studies are however required to confirm these results. Acknowledgements The study was presented to the Society of American Gastroenterologist Endoscopic Surgeons, Fort Lauderdale, Florida, in April 2005.

Background New technologies have been proposed and applied in thyroid surgery, such as the mini-invasive video-assisted thyroidectomy (MIVAT) and the intraoperative monitoring of recurrent laryngeal nerve (RLN) [1,2]. In general the essential objectives for thyroidectomy are: conservation of the parathyroid glands, avoidance of injury to RNL, an accurate hemostasis and an excellent cosmesis. The thyroid has a rich blood supply. Each must be securely occluded and divided to perform a safe and expeditious operation [3]. Theodor Kocher is credited with refining the technique of thyroidectomy and reducing the incidence of postoperative hemorrhage [3]. It is difficult to estimate the real impact of bleeding, as main cause of intra-operative accidental lesions of vital structures as RLN [3]. However any surgeon who has routinely been practising thyroid surgery, knows that even minor bleeding may greatly compromise the view of surgical field and lead to severe difficulties in identifying the anatomical structures. Furthermore, management of abnormal bleeding exposes the patient to the morbidity of re-operation. In mini-invasive thyroidectomy intraoperative bleeding is a frequent cause for conversion to open technique [2]. Several devices and techniques, coming from general surgery, are commonly used to control bleeding, during thyroid surgery. Haemostasis in thyroid surgery is achieved by means of conventional clamp-and-tie technique, diathermy, and haemostatic clips and, more recently, by ultrasonic coagulating-dissection and electrothermal bipolar vessel sealing systems. We tested hydroxylated polyvinyl acetal tampons (HPA), their efficacy for small bleeding control and tissue dissection during several thyroid procedures.

clamp-and-tie technique, diathermy, and haemostatic clips and, more recently, by ultrasonic coagulating-dissection and electrothermal bipolar vessel sealing systems. We tested hydroxylated polyvinyl acetal tampons (HPA), their efficacy for small bleeding control and tissue dissection during several thyroid procedures. Materials and methods Tampons We tested the use of hydroxylated polyvinyl acetal (HPA) tampons (Merocel™ – Medtronic Xomed, Jacksonville, Florida, USA) for minor bleeding control, fluid absorption and dissection during thyroid procedures. Hydroxylated polyvinyl acetal tampons are made of synthetic molecule obtained by "foaming at open cell" a fully biocompatible nonirritating polymer. This manufacturing procedure allows the generation of a cellular net made by pores, which are joined together in order to prevent breaks or lost of fibers. They are designed in different shapes, forms and sizes. Specific test demonstrated that HPA tampons are fully biocompatible and able to reduce bacterial grown. Their surface is smooth and they do not stick to tissues. They have an initial, mildly hard, firmness that allows their use as blunt dissection devices. Once in contact with fluids, the polymeric structure provides great absorbing capacity (up to 25 times the original weigh) within a small volume of material. Merocel™ tampons are routinely used for removal of unwanted fluids from the operative field in general, neurosurgical operations and for nasal packing in the treatment of epistaxis [4,5,8-10].

he polymeric structure provides great absorbing capacity (up to 25 times the original weigh) within a small volume of material. Merocel™ tampons are routinely used for removal of unwanted fluids from the operative field in general, neurosurgical operations and for nasal packing in the treatment of epistaxis [4,5,8-10]. Clinical experience From August 2004 till December 2005, HPA tampons have been used, on demand, in 50 different thyroid procedures (Table 1, 2). Table 1 Clinical experience with HPA tampons Thyroid procedure with HPA tampon Number Total thyroidectomy 43 Emithyroidectomy 7 TOTAL 50 Table 2 Final pathology Thyroid procedure with HPA tampon Number Nontoxic Goiter 31 Graves' Disease 15 Differentiated Thyroid Carcinoma 4 TOTAL 50 In 47/50 (94%) cases only one tampon was required, in 3/50 (6%) patients required 2 tampons. This fact is mainly related to the thyroid volume and length of the procedure. The present study did not include cervical lymph node dissections or miniinvasive thyroid procedures.

15 Differentiated Thyroid Carcinoma 4 TOTAL 50 In 47/50 (94%) cases only one tampon was required, in 3/50 (6%) patients required 2 tampons. This fact is mainly related to the thyroid volume and length of the procedure. The present study did not include cervical lymph node dissections or miniinvasive thyroid procedures. Results HPA has been used for both benign and malignant thyroid disease, for hemithyroidectomy as well as for total thyroidectomy (Table 1, 2). HPA tampons have been found extremely useful to absorb blood coming from minor and diffuse losing, caused by dissection of tissues, adhesions and small vessel. Once in contact with the bleeding area, HPA tampons are able to expand quickly, helping to control bleeding by a combined action of fluid absorption and local compression, that facilitates platelets aggregation. Once expanded, the tampon becomes soft and a traumatic and it can be used as gentle retracting and dissecting instrument reducing the risk of accidental damaging.

e able to expand quickly, helping to control bleeding by a combined action of fluid absorption and local compression, that facilitates platelets aggregation. Once expanded, the tampon becomes soft and a traumatic and it can be used as gentle retracting and dissecting instrument reducing the risk of accidental damaging. In a previous laboratory test we found than 1 cm2 of HPA tampon was able to absorb a significantly greater amount of blood than the same surface of standard surgical swab; it has also been found that HPA tampons were able to absorb fluids up to 25 time their initial weight [4]. The porous design of the tampon allows performing an effective aspiration of blood and unwanted fluid, by positioning the tampon directly on the bleeding surface (for example close to RNL) and using the suction device right through the tampon itself (Fig. 1). Thanks to the initial hard consistency, we also used HPA tampons as dissecting instruments during thyroidectomy. This resulted extremely useful for the dissection of the thyroid gland from the trachea and muscles, to identify and isolate the RLN or for blunt dissection and taking down of adhesions. In total thyroidectomy, the tampon was left in situ while performing the controlateral emithyroidectomy. Furthermore, opposite to standard pledgets, HPA tampons are totally fibers free and extremely resistant to strong tractions. Their cost is also fairly low (less than 10€ per tampon).

on and taking down of adhesions. In total thyroidectomy, the tampon was left in situ while performing the controlateral emithyroidectomy. Furthermore, opposite to standard pledgets, HPA tampons are totally fibers free and extremely resistant to strong tractions. Their cost is also fairly low (less than 10€ per tampon). Figure 1 Minor bleeding control using HPA tampon, during thyroidectomy. The porous design of the tampon allows performing an effective aspiration of blood and unwanted fluid, by positioning the tampon directly on the bleeding surface (for example, close to RNL) and using the suction device right through the tampon itself. No postoperative mortality or morbidity in these short series was observed. None of the patients showed evidences of allergic reactions or intolerance; no accidental rupture or intra-abdominal lost of tampons were reported.

Figure 1 Minor bleeding control using HPA tampon, during thyroidectomy. The porous design of the tampon allows performing an effective aspiration of blood and unwanted fluid, by positioning the tampon directly on the bleeding surface (for example, close to RNL) and using the suction device right through the tampon itself. No postoperative mortality or morbidity in these short series was observed. None of the patients showed evidences of allergic reactions or intolerance; no accidental rupture or intra-abdominal lost of tampons were reported. Conclusion The use of local haemostatic devices is spread in general surgery to assist in the control of capillary, venous, small arterial haemorrhage and "oozing" bleeding when legation or other conventional methods of control are impractical or ineffective. New modalities for the haemostasis and dissection in thyroid surgery include clips, tampons, harmonic scalpel, fibrin sealant, electrothermal bipolar vessel sealing system and gauzes. The effect of a combination of new haemostatic devices on the treatment of dissection and bleeding in thyroid surgery aim to improve primary and secondary haemostasis rates and the safety of treatment [6]. There are no data in Literature of the use of HPA in thyroid surgery in particular open thyroidectomies. Their easy application, low cost, and significant blood-loss reduction make these agents attractive also for thyroid surgery. The use of the haemostats has been approved by the U.S. Food and Drug Administration. In our experience HPA have been documented to be a safe device in open surgery.

icular open thyroidectomies. Their easy application, low cost, and significant blood-loss reduction make these agents attractive also for thyroid surgery. The use of the haemostats has been approved by the U.S. Food and Drug Administration. In our experience HPA have been documented to be a safe device in open surgery. During thyroid procedures the tampons are cut to the appropriate size for the placement, with a sterile technique in the begining phases of surgery. The device may be adjusted it in the irregular surfaces of the dissecting area between the trachea and the strap muscles thus protecting the RLN from the suctioning drainage. The haemostat, by swelling, may exert pressure resulting in paralysis and/or nerve damage: the hypothesis that these surgical medicaments placed in close proximity to the RLN could be responsible for some postoperative disturbances and that any alteration in nerve function that occurs postoperatively is usually attributed to the trauma of the surgical procedure rather than the medicament is quite difficult to determine. The surgical sponge has a pH of 6.5 ± 1.0 and should not be placed adjacent to neural tissue. Despite this advice it is often placed in direct contact with neural tissue, particularly in neurosurgery. Loescher and Robinson examined the response of tissue to different haemostatics agents left in situ in particular the neural function was reported [7]. The immediate effect of the haemostat on neural function was rare and variable, responsible for some postoperative disturbances, but the effects of these haemostats on neural function were only temporary. We suggest in patients with haemorrhagic disorders, those requiring oral anticoagulant therapy, with persistent major bleeding after primary haemostatic measures during thyroid surgery, hemostasis can be achieved through these new devices. These new devices achieve rapid haemostasis, they must be used in combination but must be used in conjunction with a meticulous surgical technique. The clips and conventional clamp-and-tie technique ideally are suited to acute bleeding and are most effective when a vessel from a bleeding source can be identified. These surgical sponges are not intended as substitutes of careful surgery, a proper use of sutures and ligatures and must never been used to control haemorrhage from large arteries. Specific experience and training with these new devices is essential for optimal use.

n a vessel from a bleeding source can be identified. These surgical sponges are not intended as substitutes of careful surgery, a proper use of sutures and ligatures and must never been used to control haemorrhage from large arteries. Specific experience and training with these new devices is essential for optimal use. Preliminary experience using HPA has been encouraging, but prospective randomized trials using adequate patient numbers are still needed to validate efficacy and safety. Competing interests The author(s) declare that they have no competing interests. Authors' contributions GD: acquisition of data FR: study conception and design LB: analysis and interpretation of data GD: drafting of manuscript RD: Critical revision and supervision Acknowledgements Supported in part by grants from the Italian Ministry of Education, University and Research (MIUR, Roma, Italy Project "Studies on relationship between fetal microchimerism and thyroid autoimmune disease"). The authors are grateful to Professor L. Bartalena and P. Miccoli, for general support, for excellent technical assistance and helpful discussion.

Background Anonymous reporting of adverse and near-miss events Near miss events occur in every medical facility all over the world. Studies have shown that 4% of hospitalized patients suffer from an adverse event caused by the medical treatment administered [1,2]. These mistakes can result in patient discomfort, irreversible injury, or even death. Although patient safety is a topic of major emphasis for medical facilities, efforts to investigate the causes and prevention of these errors have been insufficient, reflecting a general lack of awareness of the problem [2]. As a consequence of this oversight, the need for effectively collecting and understanding data about these adverse events is paramount in improving patient safety. Some institutions have created systems to encourage medical workers to report these adverse events so that they may be formally documented and reviewed. However, these systems have often proven to be inadequate or ineffective for reviewing the data collected and improving the outcomes in patient safety. Studies on the handling of incident reports indicated that no clear and consistent methodology across a number of organizations were of an appropriate added value to patient safety mechanisms [1].

often proven to be inadequate or ineffective for reviewing the data collected and improving the outcomes in patient safety. Studies on the handling of incident reports indicated that no clear and consistent methodology across a number of organizations were of an appropriate added value to patient safety mechanisms [1]. Existing approaches Although systems have been created for reporting adverse events, there does not exist, to the best of our knowledge, an open source application for reporting patient safety. The open source application is a software system that permits the usage and modification of its source code. In laymen's terms, "open-source," by definition, implies that the software allows anyone to improve, redistribute, and share it with others [3-8]. There are a number of Web-based event reporting systems in health care. US Pharmacopeia's MedMarx system [9] and the Institute for Safe Medication Practices' Medication Errors Reporting Program [10] collect data on medication errors. The Emergency Care Research Institute (ECRI) [11] collects reports of adverse events involving medical products. Mekhijian et al recently implemented a voluntary reporting system that collects data about errors, events, and near misses solely at Ohio State University. All of these systems focus on a specific type of event (e.g., medications) or are localized at one institution. While virtually all medical institutions have some in-hospital mechanism for reporting incidents, typically these systems are mandatory and non-anonymous. This allows blame to be focused on individuals, and, as a result, these systems result in grave underreporting.

g., medications) or are localized at one institution. While virtually all medical institutions have some in-hospital mechanism for reporting incidents, typically these systems are mandatory and non-anonymous. This allows blame to be focused on individuals, and, as a result, these systems result in grave underreporting. A single-institution studyfound that a significant increase in reporting occurred within the first year of a hospital-wide implementation of a Web-based electronic safety event reporting system [12]. Twenty-two percent of the safety events reported resulted in patient harm; 16% of these were near misses. Nurses entered 73% of the events and physicians only 2%. Miscoding of event details and underreporting were common. Most of the events were reported by nurses [12] which highlights the need of a system that allows all staff, including physicians and other team members, involved with patient care to report incidents. At the same time, more than one report may be submitted for the same episode and the system must be prepared to identify such cases. These results suggest that more work is needed to involve care team members in reporting, improve the accuracy of the submitted information, and prioritize and streamline event analysis. Few national surveillance systems [13,14] and other institutional safety reporting systems value the voluntary, anonymous, and confidential aspects of collecting information about adverse events and near misses. Some of these provide feedback in the form of monthly reports, case discussions and/or newsletters [15], rather recognizing the importance of educational interventions such as safety awareness and learning how to use the reporting system [16,12]. Such features are believed to result in great improvement for the reporting systems.

se provide feedback in the form of monthly reports, case discussions and/or newsletters [15], rather recognizing the importance of educational interventions such as safety awareness and learning how to use the reporting system [16,12]. Such features are believed to result in great improvement for the reporting systems. Objective We describe an open-source Web application, Duke Surgery Patient Safety (DSPS), designed at our institution for the anonymous reporting of adverse and near-miss events as well as regular reporting to physicians and hospital administration. This program was designed to streamline this process in the Department of Surgery at our institution; nonetheless, it can be adapted to any patient care department. The primary goal was to design a program that would streamline adverse-event reporting with minimal time required by reporters in the ever-busy and high volume surgical environment. To the best of our knowledge, this is the first ever open-source program created to streamline the reporting of adverse events. More importantly, this program can be freely modified to sync with the existing reporting systems of institutions.

required by reporters in the ever-busy and high volume surgical environment. To the best of our knowledge, this is the first ever open-source program created to streamline the reporting of adverse events. More importantly, this program can be freely modified to sync with the existing reporting systems of institutions. Implementation Goals The primary goal of the DSPS application is to provide a mechanism for anonymous reporting of adverse events. In addition, DSPS should also automate repetitive tasks associated with the administration of patient safety initiatives, such as (1) send regular personalized reminders by e-mail to all potential reporters, (2) allow administrators to properly store, make changes, and reject individual reports as deemed necessary, and (3) generate automated periodic reports for individual practitioners and hospital administrators summarizing adverse report rates in a graphical format. Design objectives The overall objective of the DSPS project was to build a Web-based application that would allow for the anonymous reporting of adverse events. Before designing the application, we analyzed similar tools, including commercially available products, adverse event reporting systems from academic institutions, and recommendations from the Institute of Medicine [17]. Our search resulted in a series of technical requirements that are highly desirable for the application: • User interface should be simple and straightforwardto enhance reporter participation

Design objectives The overall objective of the DSPS project was to build a Web-based application that would allow for the anonymous reporting of adverse events. Before designing the application, we analyzed similar tools, including commercially available products, adverse event reporting systems from academic institutions, and recommendations from the Institute of Medicine [17]. Our search resulted in a series of technical requirements that are highly desirable for the application: • User interface should be simple and straightforwardto enhance reporter participation • The application should run on an Intranet to facilitate access by multiple computers within the same institution • Reporters should have a minimum set of categories to classify their complications according to a standardized coding system determined by the institution, followed by a unique free-text description of the adverse or near-miss event (ANME) and its potential cause • Reporters should be allowed to describe more than one ANME per patient without having to make a second patient entry • Administrators should have the ability to change ANME that are inaccurate and reject others that are not considered to be adverse events • Administrators should have the ability to check the ANME before it is accepted as a true event • In order to incentives reporting in the surgical setting, the application should remind reporters about its existence on a regular basis, also sending them reports about the numbers of complications submitted by their division

• Administrators should have the ability to check the ANME before it is accepted as a true event • In order to incentives reporting in the surgical setting, the application should remind reporters about its existence on a regular basis, also sending them reports about the numbers of complications submitted by their division • Data extraction should be easy, allowing administrators to perform further analyses using external tools The results from the search were implemented in the design of our application thus accomplishing the following design objectives: • Simplifying the reporting process for frontline staff • Eliminating multiple forms used to report adverse events • Increasing the quantity and quality of occurrence data • Improving response time by linking reports to department leadership and key personnel (clinical nurse specialist, unit educator) • Improving evaluation and follow-up through a structured framework • Enhancing the quality and safety of patient care and the employee work DSPS should ideally adhere to national data standards. Although ideal, achieving consensus on such standards remains a controversial matter. A recent effort by the National Surgical Quality Improvement Program [18] has attempted to classify some of the most common postoperative adverse events (Table 1). However, currently this classification is not comprehensive for most surgical subspecialties and, therefore, we have attempted to design a method that will accomplish the following goals: Table 1 List of adverse events as listed in the National Surgical Quality Improvement Program (NSQIP), 2005

DSPS should ideally adhere to national data standards. Although ideal, achieving consensus on such standards remains a controversial matter. A recent effort by the National Surgical Quality Improvement Program [18] has attempted to classify some of the most common postoperative adverse events (Table 1). However, currently this classification is not comprehensive for most surgical subspecialties and, therefore, we have attempted to design a method that will accomplish the following goals: Table 1 List of adverse events as listed in the National Surgical Quality Improvement Program (NSQIP), 2005 Wound Ocurrences Superficial Incisional SSI Deep Incisional SSI Organ/Space SSI Wound Disruption Other (ICD-9) Respiratory Pneumonia Unplanned Intubation Pulmonary Embolism On Ventilator > 48 hours Other (ICD-9) Urinary Tract Occurrences Progressive Renal Insuficiency Acute Renal Failure Urinary Tract Infection Other (ICD-9) CNS Occurences Stroke/CVA Coma >24 hours Peripheral Nerve Injury Other (ICD-9) Cardiac Ocurrences Cardiac Arrest req. CPR Myocardial Infarction Other (ICD-9) Other Ocurrences Bleeding > e units RBCs Graft/Prosthesis/Flap Failure DVT/Throbophebitis Systemic Sepsis SIRS Sepsis Septic Shock Other (ICD-9) • Allow for data exchange across different surgical services and institutions • Be flexible to ensure that specific needs of each surgical division are met

Wound Ocurrences Superficial Incisional SSI Deep Incisional SSI Organ/Space SSI Wound Disruption Other (ICD-9) Respiratory Pneumonia Unplanned Intubation Pulmonary Embolism On Ventilator > 48 hours Other (ICD-9) Urinary Tract Occurrences Progressive Renal Insuficiency Acute Renal Failure Urinary Tract Infection Other (ICD-9) CNS Occurences Stroke/CVA Coma >24 hours Peripheral Nerve Injury Other (ICD-9) Cardiac Ocurrences Cardiac Arrest req. CPR Myocardial Infarction Other (ICD-9) Other Ocurrences Bleeding > e units RBCs Graft/Prosthesis/Flap Failure DVT/Throbophebitis Systemic Sepsis SIRS Sepsis Septic Shock Other (ICD-9) • Allow for data exchange across different surgical services and institutions • Be flexible to ensure that specific needs of each surgical division are met Software architecture The software architecture was developed using a model similar to previous software applications from our group [19-21]. The programming language used was Java and the design model was Model-View-Controller (MVC) (Figure 1). MVC's three components are: a central Model, Views, and Controllers of the Model (Figure 2). The Views and the Controller are the application's interface. They represent the Model to the user. As for the central Model, it displays the logic, as well as the database access, numeric algorithms, and algorithms for data manipulation. The Controller provides information to the Model while the View passes it from the Model to the users. Figure 3 is a diagram of the DSPS architecture. Figure 1 Model-View-Controller (MVC) as its design model. Figure 2 Controllers of the Model.

Software architecture The software architecture was developed using a model similar to previous software applications from our group [19-21]. The programming language used was Java and the design model was Model-View-Controller (MVC) (Figure 1). MVC's three components are: a central Model, Views, and Controllers of the Model (Figure 2). The Views and the Controller are the application's interface. They represent the Model to the user. As for the central Model, it displays the logic, as well as the database access, numeric algorithms, and algorithms for data manipulation. The Controller provides information to the Model while the View passes it from the Model to the users. Figure 3 is a diagram of the DSPS architecture. Figure 1 Model-View-Controller (MVC) as its design model. Figure 2 Controllers of the Model. Figure 3 DSPS architecture. Software interface The Duke Surgery Patient Safety application provides interfaces for three classes of users: ANMEReporters, Administrators, and Report Receivers. Activity flow Before the application was built, we evaluated the usual flow of activities for ANME reporting and how the application would be integrated into this system. This analysis resulted in the following flow of activities in the software architecture (Figure 4): Figure 4 flow of activities in the software architecture. • Weekly e-mails are directly sent to health care professionals working with patients reminding them about the importance of ANME reporting and reminding them of the central Web site.

Activity flow Before the application was built, we evaluated the usual flow of activities for ANME reporting and how the application would be integrated into this system. This analysis resulted in the following flow of activities in the software architecture (Figure 4): Figure 4 flow of activities in the software architecture. • Weekly e-mails are directly sent to health care professionals working with patients reminding them about the importance of ANME reporting and reminding them of the central Web site. • When an ANME is identified, health care professionals go to the Web site and generate the report using a single reporting interface. • Administrators receive an e-mail soon after a new ANME is reported. They access DSPS and receive the information that will allow them to investigate the ANME in further detail. This checking is performed through consultation of medical charts and administrative databases. • Once the ANME is confirmed, the administrator returns to DSPS and confirms its existence. Otherwise, the report can reject by the administrator. • Confirmed ANME are included in a monthly report with information customized to individual practitioners and divisions. ANME reporters The interface for ANME reporters includes functionality that allows reporters to perform the following functions: • Choose the surgical division, subspecialty, and primary surgeon and/or house staffassociated with the ANME. • Provide information that will allow for identification of the patient for whom the ANME is being reported.

ANME reporters The interface for ANME reporters includes functionality that allows reporters to perform the following functions: • Choose the surgical division, subspecialty, and primary surgeon and/or house staffassociated with the ANME. • Provide information that will allow for identification of the patient for whom the ANME is being reported. • Select all pre-defined complication categories that are applicable to this patient, including a text description where reporters can add additional information as needed, including potential cause. Reporters also have the ability to delete complications prior to submission if a category is erroneously selected (Figure 5). Figure 5 Complication Report. • Since at our institution the most frequent users of the system are surgical residents, the reporter interface also has a link to a repository of previous patient safety presentations. Usually residents prepare their presentations based on previous ones, andthe inclusion of this additional resource ensures that they will frequently come to the DSPS site, thus serving as a constant reminder of the importance of patient safety reporting. Report receivers and administrators The administrative interface was designed to allow the administrator complete control over all existing tasks (Figure 6). These include: Figure 6 Report Receivers and Administrators. • Receive e-mail when ANME report is made • Check reliability of ANME report and either accept or reject it

Report receivers and administrators The administrative interface was designed to allow the administrator complete control over all existing tasks (Figure 6). These include: Figure 6 Report Receivers and Administrators. • Receive e-mail when ANME report is made • Check reliability of ANME report and either accept or reject it • Add new administrators. This function is of exclusive use of departmental administrators, so that they can delegate the function of divisional representatives. The administrator can also add new divisions and respective subspecialties, providers, reporters • Add and modify new sets of complication categories that are specific to divisions • Export ANME data in a spreadsheet so that it can be manipulated using data analysis software (e.g., text mining, statistical packages, neural networks, etc) • Define report format and respective release dates for providers. • Generate messages sent to all potential reporters (health care professionals working with patients) on a regular basis. These messages usually contain a message stating the importance of ANME reporting, also providing links to DSPS. • Add educational files to an educational repository. These files function as a library aiding the resident who, at our institution, is in charge of the monthly patient safety case presentation.

• Generate messages sent to all potential reporters (health care professionals working with patients) on a regular basis. These messages usually contain a message stating the importance of ANME reporting, also providing links to DSPS. • Add educational files to an educational repository. These files function as a library aiding the resident who, at our institution, is in charge of the monthly patient safety case presentation. Usability evaluation Comprehensive usability tests were performed by formal usability analysis and a field observation period. Formal usability analysis was conducted in a manner similar to a previous study by the authors' institute [20] In short, twelve physician scientists (n = 10) and nurses (n = 2), from two institutions, were asked to participate to avoid bias from authors participating in the development of the DSPS application. Six users tested the reporter's interface and the other six tested the administrator's interface. Each of these users, by self-report, rated as having average computer literacy with no previous experience with the DSPS application. Formal usability tests followed a protocol where users were observed by two evaluators (who were available to answer any questions from users) and had to complete assigned regulatory tasks. The following DSPS application factors were evaluated: speed, easy of learning, easy of using, understanding of functionality, and navigation. The evaluators who observed the users made detailed notes about number of errors made while using the application. Each participant was instructed to answer a questionnaire at the end of the formal usability analysis with items about interface problems, missing features, and suggestions for overall improvement.

gation. The evaluators who observed the users made detailed notes about number of errors made while using the application. Each participant was instructed to answer a questionnaire at the end of the formal usability analysis with items about interface problems, missing features, and suggestions for overall improvement. As with previous studies by these authors [20], field observations were conducted by one of the co-authors (MM) and consisted of observation of researchers and research administrators during the execution of common regulatory activities using DSPS. Meetings were held with investigators and research administrators in order to obtain detailed feedback concerning any concerns or suggestions for improvement, however we acknowledge that further studies with more users are needed for long-term testing. Results Implementation The current version of DSPS was designed to allow for the anonymous reporting of adverse events, their evaluation and follow-up, while reducing the number of forms used and making the reporting process easier leading to an improvement in the quality and safety of patient care. Its current design incorporates the numerous suggestions received from formal and field usability tests. The DSPS application has been implemented in the Department of Surgery atDuke University Medical Center.

d and making the reporting process easier leading to an improvement in the quality and safety of patient care. Its current design incorporates the numerous suggestions received from formal and field usability tests. The DSPS application has been implemented in the Department of Surgery atDuke University Medical Center. When making an anonymous report, the reporters are asked to enter the division/department name, surgeon and, if appropriate, subspecialty related to the adverse event by choosing from drop down boxes available. Then, they need to provide patient information such as MRN, last name, first name and middle initial. Next, reporters select the date of the complication in a calendar. All complication categories that apply to the specific patient are selected, selecting one at a time from a drop down list. In sequence, reporters can add additional information about the adverse event descriptions and potential cause by typing into the appropriate text areas. At last, they mark the severity of complication based on Complications Grading System: • Grade I – Any deviation from the normal postoperative course without the need for pharmacological treatment or surgical, endoscopic, and radiological interventions. Allowed therapeutic regimens are: drugs as antiemetics, antipyretics, analgesics, diuretics, electrolytes, and physiotherapy. This grade also includes wound infections opened at the bedside. • Grade II – Requiring pharmacological treatment with drugs other than those allowed for grade I complications. Blood transfusions and total parenteral nutrition are also included.

• Grade I – Any deviation from the normal postoperative course without the need for pharmacological treatment or surgical, endoscopic, and radiological interventions. Allowed therapeutic regimens are: drugs as antiemetics, antipyretics, analgesics, diuretics, electrolytes, and physiotherapy. This grade also includes wound infections opened at the bedside. • Grade II – Requiring pharmacological treatment with drugs other than those allowed for grade I complications. Blood transfusions and total parenteral nutrition are also included. • Grade III – Requiring surgical, endoscopic or radiological intervention. • Grade IV – Life-threatening complication (including CNS complications) requiring IC/ICU management. • Grade V – Death of a patient. When the reporters are done, they must click on the "submit report" button. The department administrator will be notified immediately. Each administrator will check the consistency of reports made in their departments and either accept or reject them. Although the adverse event reports are screened by the administrator, one should be aware that researchers, physicians and national organizations interested in patient safety may have different goals. Data standards In DSPS, our approach has been to present surgical specialties with existing standards that are internal to each of the surgical departments using the application. Further modifications are made in the attempt to make the variables compliant with what exists in other divisions, ultimately allowing comparisons across subspecialties.

approach has been to present surgical specialties with existing standards that are internal to each of the surgical departments using the application. Further modifications are made in the attempt to make the variables compliant with what exists in other divisions, ultimately allowing comparisons across subspecialties. Usability 1. Formal usability Results from formal usability testing are summarized in Table 2. Formal usability results demonstrated that, from the perspective of researchers, the DSPS application presented excellent speed (11/12 users strongly agreed), was easy to learn and use (10/12 users strongly agreed), had a functionality that was easily understandable (10/12 users strongly agreed), and a navigation that was intuitive (12/12 users strongly agreed). No users reported any problems in the questionnaire that was presented after formal usability testing. Table 2 Formal usability

Usability 1. Formal usability Results from formal usability testing are summarized in Table 2. Formal usability results demonstrated that, from the perspective of researchers, the DSPS application presented excellent speed (11/12 users strongly agreed), was easy to learn and use (10/12 users strongly agreed), had a functionality that was easily understandable (10/12 users strongly agreed), and a navigation that was intuitive (12/12 users strongly agreed). No users reported any problems in the questionnaire that was presented after formal usability testing. Table 2 Formal usability DSPS speed is excellent. Strongly disagree 0/12 Disagree 0/12 Neutral 0/12 Agree 1/12 Strongly agree 11/12 DSPS is extremely easy to learn Strongly disagree 0/12 Disagree 0/12 Neutral 1/12 Agree 1/12 Strongly agree 10/12 DSPS is extremely easy to use Strongly disagree 0/12 Disagree 0/12 Neutral 0/12 Agree 4/12 Strongly agree 8/12 It is very easy to understand all functionality available within DSPS (e.g., download files, upload files, etc) Strongly disagree 0/12 Disagree 0/12 Neutral 1/12 Agree 1/12 Strongly agree 10/12 The navigation in DSPS is highly intuitive Strongly disagree 0/12 Disagree 0/12 Neutral 0/12 Agree 0/12 Strongly agree 12/12 2. Field usability The first two months of field usability measurement were primarily focused on fixing software bugs, namely non-functional links to other Web pages and click buttons that did not work. DSPS users were encouraged during this time to submit their comments and suggestions directly to one of the co-authors (MM) who investigated any issues with the programmer (HM).

easurement were primarily focused on fixing software bugs, namely non-functional links to other Web pages and click buttons that did not work. DSPS users were encouraged during this time to submit their comments and suggestions directly to one of the co-authors (MM) who investigated any issues with the programmer (HM). During this phase, one issue that was corrected was the automatic email that is sent on a weekly basis to all surgeons and reporters who use DSPS for reporting. Initially, it was not being received by the users on certain weeks. Upon detection of the error, the programmers were able to readily correct the issue. Also during this phase, multiple users identified that the link they were receiving in the weekly automatic email was not working correctly. The users would click on the link and get an error message. However, upon reviewing this issue, it was an error attributable to the email system used at our institution, which frequently can not open links included in an email. To remedy this situation, additional instruction was included in the weekly email to copy and paste the link into a browser window.

et an error message. However, upon reviewing this issue, it was an error attributable to the email system used at our institution, which frequently can not open links included in an email. To remedy this situation, additional instruction was included in the weekly email to copy and paste the link into a browser window. Discussion To our knowledge, this is the first description of an open-source application for reporting patient safety related events. DSPS provides a mechanism for anonymous reporting of adverse events and helps the administration of patient safety initiatives. At the time of this article, Duke Surgery Patient Safety has only been used within a few departments at Duke University, including the Department of Surgery; however, the free distribution of its source code under the GNU Public License is expected to spread its use among other departments and colleague institutions. Although this is possible, we have not tested the cross-application of our software yet, and therefore, we are unable to discuss possible difficulties that might arise. We recommend that potential users consult their institutional IT department for guidance in installation and customization.

and colleague institutions. Although this is possible, we have not tested the cross-application of our software yet, and therefore, we are unable to discuss possible difficulties that might arise. We recommend that potential users consult their institutional IT department for guidance in installation and customization. There are many web-based reporting systems available, but most of them are restricted to one institution or they collect data on just one type of event such as medication errors and adverse events involving medical products. An example is a passive reporting system for adverse effects caused by vaccine. The events are voluntarily reported by the person who experiences them which leads to great underreporting, unconfirmed diagnoses and lack of temporal association [14]. Much of the difficulty with underreporting stems from the fact that many reporting systems focus blame on individuals rather than on the prevention of similar events. The importance of developing consciousness and awareness of patient safety, rather than generating individual blame, has the potential to greatly enhance the patient safety reporting system and the incentive to utilize the system. Ultimately, such system has the potential to improve patient morbidity and mortality in departments where administrative time is limited for members of the patient care team. We are confident that the simplicity and streamlined reporting of DSPS make it a valuable asset for our institution.

ntive to utilize the system. Ultimately, such system has the potential to improve patient morbidity and mortality in departments where administrative time is limited for members of the patient care team. We are confident that the simplicity and streamlined reporting of DSPS make it a valuable asset for our institution. Conclusion Current utilization DSPS is currently used by the Divisions of Orthopaedic Surgery, Neurosurgery, General Surgery, OHN (Otolaryngology, Head and Neck), Pediatric Surgery, Plastic Surgery, Urology and Cardiothoracic Surgery at Duke University Medical Center (DUMC). It is currently being expanded to other divisions within the Department of Surgery at DUMC, as well as other academic and non-academic centers.

osurgery, General Surgery, OHN (Otolaryngology, Head and Neck), Pediatric Surgery, Plastic Surgery, Urology and Cardiothoracic Surgery at Duke University Medical Center (DUMC). It is currently being expanded to other divisions within the Department of Surgery at DUMC, as well as other academic and non-academic centers. Potential uses Although DSPS was originally developed to fit the patient safety needs of the Department of Surgery at DUMC, its modification for use at other institutions is simple and can be easily accomplished by a few modifications in the open-source application. Several features are planned for future versions of DSPS. First, integration with administrative databases would enable our group to flag adverse events in a more automated manner. For example, if hypoglycemia is detected in an inpatient, this might indicate an adverse event (mismanagement of insulin treatment). Second, future versions should have the ability to obtain denominators (total number of cases) from administrative databases in order to generate ANME rates. These rates would be more meaningful that simple frequencies. Implementations of both features require integration of administrative databases with heterogeneous architectures, therefore requiring an interface that may integrate well with XML standards as well as possibly legacy systems.

n order to generate ANME rates. These rates would be more meaningful that simple frequencies. Implementations of both features require integration of administrative databases with heterogeneous architectures, therefore requiring an interface that may integrate well with XML standards as well as possibly legacy systems. Potential reactions According to the Committee on Data Standards for Patient Safety of the Institute of Medicine [22], three main reactions corresponding to the "cycle of fear" can occur once safety data are released. First, "kill the messenger" reactions are triggered and there is an attempt to shift the blame by questioning the measurement system and those conducting the evaluation. Second, "filter the data" starts as it is much easier to achieve better scores by 'gaming the system' than by actually changing and upgrading procedures. Third, "micromanage" response takes place. Reacting to immediate variation in the process data instead of conscientious investigation and remodeling will make the system less efficient. It is the hope of the authors that the release of patient safety reports through DSPS promotes the development of positive learning systems. Education attempts to change the shape of the performance distribution by improving all parts of the process [23]. However, to accomplish this learning system phase, a self-reporting application such as DSPS should not be an isolated solution. A crucial resolution is to give emphasis to safety itself rather than blaming. For example, we hope that ANME case discussions provide opportunity for improvement.

parts of the process [23]. However, to accomplish this learning system phase, a self-reporting application such as DSPS should not be an isolated solution. A crucial resolution is to give emphasis to safety itself rather than blaming. For example, we hope that ANME case discussions provide opportunity for improvement. Availability and requirements • Project name: Duke Surgery Patient Safety (DSPS) • Project home page: (click link for Free software) • Operating system(s): Linux/Windows • Programming language: Java • Other requirements: Java 1.3.1 or higher, Tomcat 5.x or higher • License: GNU General Public License • Any restrictions to use by non-academics: none Abbreviations • DSPS-Duke Surgery Patient Safety • ANME-Adverse and near-miss events • DUMC-Duke University Medical Center Competing interests The author(s) declare that they have no competing interests. Authors' contributions All authors have read and approved the final manuscript. • RP – designed the application, assisted with software testing, conducted the usability tests, and wrote the first draft of the manuscript • RL – assisted with the design of the application, reviewed the manuscript for intellectual content • AS – assisted with the design of the application, reviewed the manuscript for intellectual content • DOJ – assisted with the design of the application, reviewed the manuscript for intellectual content • MH – assisted with the design of the application, reviewed the manuscript for intellectual content

• RL – assisted with the design of the application, reviewed the manuscript for intellectual content • AS – assisted with the design of the application, reviewed the manuscript for intellectual content • DOJ – assisted with the design of the application, reviewed the manuscript for intellectual content • MH – assisted with the design of the application, reviewed the manuscript for intellectual content • MM – assisted with the design of the application, assisted with field testing and software debugging, reviewed the manuscript for intellectual content • HM – assisted with the design of the application, wrote and tested the source code for the application, reviewed the manuscript for intellectual content • AM – assisted with the design of the application, assisted with field testing, reviewed the manuscript for intellectual content • WR – assisted with the design of the application, assisted with field testing, reviewed the manuscript for intellectual content Appendix Programming language and development methodology DSPS was developed primarily using JAVA programming language, operating on a Tomcat server and having a MySQL database as its backend. We used a prototyping development methodology, starting with paper prototype and drawing requirements from that prototype. Once the requirements were reviewed, a first version was developed and tested. This cycle was repeated until the currentversion was obtained.

mcat server and having a MySQL database as its backend. We used a prototyping development methodology, starting with paper prototype and drawing requirements from that prototype. Once the requirements were reviewed, a first version was developed and tested. This cycle was repeated until the currentversion was obtained. Acknowledgements The authors thank Erika Nielsen for assistance with manuscript editing and submission. We also thank the National Center for Research Resources, National Institutes of Health for a Clinical and Translational Science Award (CTSA) in support of this work.

Background Hyperinsulinism is the most common cause of recurrent and persistent hypoglycemia in infancy and childhood [1]. Causes can include nesidioblastosis, pancreatic islet cell tumors such as insulinoma, and associations with multiple endocrine neoplasia syndromes. Pancreatic islet cell tumors have an annual incidence of 1 in 1,000,000, with insulinoma being the most common type, accounting for 70% to 80% of all pancreatic endocrine tumors. Peak incidence is in the third and fifth decades of life, but it can occur in all age groups. These tumors are usually solitary, benign, and diagnosed while still small because of clinical presentation of uncontrollable hypoglycemia [2]. Insulinoma is equally distributed throughout the whole pancreatic gland. Preoperative localization is sometimes challenging, with up to 64% not localized at laparotomy [3]. New improved imaging techniques have allowed for more precise preoperative localization. Noninvasive imaging modalities include transabdominal ultrasound (US), computed tomography (CT), and magnetic resonance imaging (MRI). While transabdominal US shows 30% sensitivity in detecting a tumor, conventional CT provides 73% sensitivity [4,5]. Invasive imaging procedures such as angiography and transhepatic venous hormone sampling are more accurate in detecting tumor, but they are more expensive and depend on the experience of the person performing the study. Sensitivity with angiography is 35% to 65%, and 77% to 100% with transhepatic portal venous sampling [5,6]. Selective arterial calcium stimulation and hepatic venous sampling is another invasive method, with a sensitivity of nearly 100% [6,7]. Endoscopic ultrasound (EUS) is the procedure of choice for tumor localization. With 2–3-mm resolution, EUS can be used with radial (diagnostic) or linear (for fine needle aspiration) echoendoscope and allows continuous imaging technique. Accuracy of EUS is 88% and, in combination with biphasic helical CT, increases to 97%.

scopic ultrasound (EUS) is the procedure of choice for tumor localization. With 2–3-mm resolution, EUS can be used with radial (diagnostic) or linear (for fine needle aspiration) echoendoscope and allows continuous imaging technique. Accuracy of EUS is 88% and, in combination with biphasic helical CT, increases to 97%. Intraoperative EUS can successfully localize tumors in 90% of cases [8] and can determine tumor size and invasion into the duodenum as well as vascular invasion (portal vein, small mesenteric artery, splenic vein) metastasis into the regional lymph nodes. It may distinguish nonmalignant lesions from pancreatic carcinomas. Even with these diagnostic studies, localization of insulinoma can be difficult, and, especially in the child, must be differentiated from nesidioblastosis. This clinical and pathologic disease involves hyperplasia of the beta cells of the pancreas, which leads to persistent hyperinsulinemic hypoglycemia of infancy (PHHI) and can present in adults with organic hyperinsulinism [9]. In infants, the annual incidence is 1 in 50,000 births for the sporadic form, but may be up to 1 in 2,500 births in certain regions [10]. Treatment varies significantly, depending on symptoms, from medical management with diazoxide and chlorothiazide, nifedipine, glucagon, and octreotide to some who require partial or subtotal pancreatectomy [11]. Surgery, however, results in high rates of pancreatic endocrine and exocrine insufficiency, with long-term complications of diabetes mellitus or malnutrition.

oms, from medical management with diazoxide and chlorothiazide, nifedipine, glucagon, and octreotide to some who require partial or subtotal pancreatectomy [11]. Surgery, however, results in high rates of pancreatic endocrine and exocrine insufficiency, with long-term complications of diabetes mellitus or malnutrition. To improve intraoperative localization and confirmation of successful resection, a novel hormonal assay, the rapid intraoperative insulin assay, has been developed. This intraoperative radioimmunoassay for insulin yields results within several minutes and confirms complete resection of insulinoma. It determines the level of active insulin molecule in the serum and is checked before insulinoma enucleation and right after enucleation intraoperatively. Although there is little guidance as to precise criteria for gauging success, one proposed protocol is similar to that used for quick parathyroid hormone (PTH) assay (i.e., a 50% drop). When insulin levels drop more than 50% of baseline within several minutes of insulinoma resection, the surgery is considered successful.

though there is little guidance as to precise criteria for gauging success, one proposed protocol is similar to that used for quick parathyroid hormone (PTH) assay (i.e., a 50% drop). When insulin levels drop more than 50% of baseline within several minutes of insulinoma resection, the surgery is considered successful. Here we present a case of pancreatic insulinoma in a child with symptoms of severe hypoglycemia, causing seizures. The insulinoma was enucleated laparoscopically – a well-accepted technique for appropriately selected lesions [8-10]. Intraoperative insulin measurements confirmed complete tumor resection. We aim to demonstrate that for the pediatric patient, in particular, the rapid intraoperative insulin assay has many advantages for the surgeon in confirming successful resection of insulin-producing tumors.

r appropriately selected lesions [8-10]. Intraoperative insulin measurements confirmed complete tumor resection. We aim to demonstrate that for the pediatric patient, in particular, the rapid intraoperative insulin assay has many advantages for the surgeon in confirming successful resection of insulin-producing tumors. Case presentation The patient is a 13-year-old male with no prior medical history who presented in December of 2004 with a general tonic-clonic seizure after a basketball game. Blood glucose was 33 mg/dL. Workup was delayed. He had a second seizure in late January 2005 associated with a glucose level less than 40 mg/dL. At that time the C-peptide level was 8.1 U/dL. The patient was admitted for a 72-hour inpatient fast in late March 2005, and had a seizure at hour 18 of the fast. His glucose was 33 mg/dL with an insulin level equal to 21 U/dL, confirming the diagnosis of insulinoma. CT scanning and MRI were performed and demonstrated a 1.8 cm mass in the neck of the pancreas. After several additional seizures he was referred to Columbia Presbyterian Hospital for evaluation and admitted from the office for glucose monitoring and preoperative preparation. The patient underwent intraoperative US followed by laparoscopic enucleation of a pancreatic islet cell tumor.

in the neck of the pancreas. After several additional seizures he was referred to Columbia Presbyterian Hospital for evaluation and admitted from the office for glucose monitoring and preoperative preparation. The patient underwent intraoperative US followed by laparoscopic enucleation of a pancreatic islet cell tumor. Operative procedure An arterial line was inserted to allow frequent glucose testing during the operative procedure. The patient's baseline glucose level was 54 mg/dL with a baseline insulin level of 125 U/dL (normal 10–86 U/dL). Glucose solution was administered intraoperatively to maintain a blood glucose level greater than 100 mg/dL. After induction of general anesthesia and positioning in the modified lithotomy position, the abdomen was insufflated to 15 mm Hg with CO2. Two 12 mm trocars and three 5 mm trocars were inserted and the lesser sac entered through the omentum, preserving the right gastroepiploic vessels to expose the anterior surface of the pancreas. The inferior aspect of the pancreas was then mobilized to allow inspection of the posterior surface. The splenic vein was clearly identified and dissected. The pancreas was dissected medially and a space posterior to the pancreas and anterior to the portal vein was developed. Laparoscopic US identified a hypoechoic mass in the neck of the pancreas, which was enucleated using a combination of blunt and sharp dissection techniques. The lesion was placed in a retrieval sac and extracted.

was dissected medially and a space posterior to the pancreas and anterior to the portal vein was developed. Laparoscopic US identified a hypoechoic mass in the neck of the pancreas, which was enucleated using a combination of blunt and sharp dissection techniques. The lesion was placed in a retrieval sac and extracted. Intraoperative insulin levels of 172 U/dL were measured after initial identification of the insulinoma. At 5 and 10 minutes postexcision, values dropped to 87 U/dL and 68 U/dL, respectively, demonstrating complete resection. Total operating time was 120 minutes. Postoperatively, the patient's blood glucose increased to 130 mg/dL. The patient developed a pancreatic fistula postoperatively, which was initially controlled via drainage catheter and conservative measures, but ultimately required re-exploration with a distal pancreatectomy and splenectomy. The patient subsequently recovered without complication and final pathology showed a 1.4 × 1.2 × 0.9 cm, well-differentiated, intermediate grade insulinoma.

hich was initially controlled via drainage catheter and conservative measures, but ultimately required re-exploration with a distal pancreatectomy and splenectomy. The patient subsequently recovered without complication and final pathology showed a 1.4 × 1.2 × 0.9 cm, well-differentiated, intermediate grade insulinoma. Rapid insulin assay To determine insulin levels intraoperatively, we used the STAT-IntraOperative-System (Future Diagnostics, Arlington, MA). This is a chemiluminescence immunoassay for the quantitative determination of insulin levels in human serum and EDTA plasma. The assay uses two monoclonal antibodies against insulin. Patient samples are introduced into wells and incubated for 7 minutes. The washed wells are then placed into the STAT-Read, which calculates insulin levels by measuring emissions of light in relative light units. The time between obtaining a blood sample and reading a result is under 12 minutes.

bodies against insulin. Patient samples are introduced into wells and incubated for 7 minutes. The washed wells are then placed into the STAT-Read, which calculates insulin levels by measuring emissions of light in relative light units. The time between obtaining a blood sample and reading a result is under 12 minutes. Discussion and evaluation This rapid intraoperative test gives us a valuable adjunct for determining complete excision in complicated cases of recurrent or questionable insulinoma. Although not a common problem, for patients with an uncertain diagnosis, this test may provide a novel approach to confirmation. This is especially the case for the pediatric patient. In this report, we highlight the utility of the rapid intraoperative insulin assay as an effective and unique method to confirm curative resection of insulinoma in the pediatric patient, a population where other hyperinsulin-producing conditions, such as nesidioblastosis, must be considered as well. In children, the etiology of hyperinsulinism can be a diagnostic challenge. This assay may represent particular utility intraoperatively to confirm curative resection.

f insulinoma in the pediatric patient, a population where other hyperinsulin-producing conditions, such as nesidioblastosis, must be considered as well. In children, the etiology of hyperinsulinism can be a diagnostic challenge. This assay may represent particular utility intraoperatively to confirm curative resection. Rapid determination of insulin levels with detection of C-peptide is a new modality for diagnosing insulinoma. C-peptide is a 31 amino acid fragment, which is split from the proinsulin molecule to form bioactive insulin in the serum. The analysis is performed by radioimmunoassay and can be done using serum from peripheral blood in a rapid time period. Insulin has a short serum half-life of about 5 minutes; 50% is cleared by the liver and 30% by the kidneys. Several authors have reported the use of intraoperative insulin monitoring during resection of insulinomas in adults, in an attempt to improve intraoperative resection and outcome [15-19]. Although intraoperative assays for parathyroidectomy are well established and available in many centers – they are currently the most widely used intraoperative means for hormone measurement in the United States – the rapid insulin assay is more complicated, is not as widely available, and requires a dedicated team in order to ensure reproducibility and reliability. This assay is a valuable technique that, when performed well, can serve as a useful adjunct to the management and operative resection of patients with known or suspected insulinoma. Our goal is to raise awareness regarding this technique and potential applications that may be useful to the surgeon.

ity and reliability. This assay is a valuable technique that, when performed well, can serve as a useful adjunct to the management and operative resection of patients with known or suspected insulinoma. Our goal is to raise awareness regarding this technique and potential applications that may be useful to the surgeon. Treatment of insulinoma is by surgical resection [2,12]. However, the surgeon may not know preoperatively if the insulinoma is single or multiple or present in a background of islet cell hyperplasia. To answer these questions, the rapid intraoperative serum insulin assay may be useful. In this case report we have presented a child with an insulinoma for whom intraoperative insulin measurements determined completeness of excision with subsequent cure. Postoperatively, the patient remained normoglycemic and all biochemical parameters returned to normal. Conclusion In conclusion, we present a confirmatory case demonstrating the utility of the rapid intraoperative insulin assay in facilitating intraoperative resection and confirmation of complete excision of insulinoma for the first time in a pediatric patient. This population may especially benefit from this novel assay to confirm complete resection and to differentiate multiple etiologies of hyperinsulinism in the pediatric population. Competing interests The author(s) declare that they have no competing interests. Authors' contributions Performed Surgery: VES, WBI Preparation of Manuscript: VES, AS, WBI Critical Review and Corrections: VES, WBI All authors read and approved the final manuscript.

Background Laparoscopic Radiofrequency-ablation is a save and effective method for local tumour destruction and offers a combination of minimal invasive surgery and the advantages of laparotomy. However in contrary to the transcutaneously free-hand puncture application of the laparoscopic free-hand puncture is restricted because of the capnoperitoneum and the consecutive fixation of the needle on 2 different points. The use of a laparoscopic ultrasound probe with a canal for puncture can solve this problem and improve the precision of puncture. However, a stiff needle limits the necessary angulation that is needed to reach right-lateral and cranial liver metastases [1].

the consecutive fixation of the needle on 2 different points. The use of a laparoscopic ultrasound probe with a canal for puncture can solve this problem and improve the precision of puncture. However, a stiff needle limits the necessary angulation that is needed to reach right-lateral and cranial liver metastases [1]. An ultrasound-based, laparoscopic online-navigation system, which allows guided out-of-plane needle placement, could significantly increase the accuracy of puncture. The system is aligned to help the surgeon to perform laparoscopic radiofrequency ablations using preoperative planning data and tracked laparoscopic ultrasound (LUS) as well as a tracked RFA needle [2]. The tracking hardware for the prototype of the system consists of an electromagnetical tracking system. In order to connect the tracking system with the RFA needle and the LUS, different adapters have to be used. These adapters should provide a reproducible, sterilisable, easy to handle and save connection between the trackers and the RFA needle and the LUS respectively. Furthermore, the adapters need to be designed in a way that the electromagnetic tracking system is not influenced. In this article, we present our experiences and solutions with the development of electromagnetic adapters.

y to handle and save connection between the trackers and the RFA needle and the LUS respectively. Furthermore, the adapters need to be designed in a way that the electromagnetic tracking system is not influenced. In this article, we present our experiences and solutions with the development of electromagnetic adapters. Methods First experiences Adapter for laparoscopic ultrasound probe In a first step, the electromagnetic tracker of the US-Guide 2000™ system (Ultra Guide®, Tirat Hacarmel, Israel), which is a six-degree-of-freedom electromagnetic tracking system for percutaneous interventions, was connected to the LUS transducer (B-K Medical Denmark, Type 8815) by using adhesive tape. The nearer the adapter could be placed to the tip of the instrument, the higher is the accuracy of the system. The tracking of the LUS was feasible but too inaccurate (Fig 1). Therefore a tighter and more stable adapter was developed in order to solve this problem. This adapter was made of Acrylnitrit/Butadien/Styrol-Copolymer ABS and shaped to fit the convexity of the ultrasound probes head. The LUS with the connected adapter could be placed through a 22 mm trocar without any problems (Fig. 2). However, treatments with this adapter could not be performed under sterile conditions. Therefore the development was focused on sterilisable and biocompatible materials. By doing this, it was possible to design sterilisable adapters for the LUS transducer. However, the adapter with the mounted tracking sensor at the head of the laparoscopic ultrasound probe had a total diameter of 17,5 mm, which still made a 22 mm trocar necessary (Fig. 3) [2-4].

erilisable and biocompatible materials. By doing this, it was possible to design sterilisable adapters for the LUS transducer. However, the adapter with the mounted tracking sensor at the head of the laparoscopic ultrasound probe had a total diameter of 17,5 mm, which still made a 22 mm trocar necessary (Fig. 3) [2-4]. Figure 1 Connection of laparoscopic ultrasound transducers with electromagnetic tracking system US Guide 2000® with adhesive tape. Figure 2 LUS connected to the navigation system US Guide 2000® inserted through a 22 mm trocar in an experimental environment. Figure 3 Adapter with mounted tracking sensor from the US Guide 2000® at the head of the LUS with a total diameter of 17,5 mm. Adapter for laparoscopic RFA needle For the connection of the US-Guide 2000™ navigation system to either expandable RFA needles (RITA Medical System Inc., Mountain View, CA, USA) or internally cooled RFA needles (Radionics Inc., Burlington, MA, USA) the electromagnetic tracker was mounted to the basis of the RFA needle (Fig. 4). We have used the adapter from the US-Guide 2000™ navigation system, which is normally designed for transcutaneous guided punctures and connected with the RFA needle in perpendicular direction. However, rotations of the adapter around the needle could not be compensated by the navigation system. Figure 4 Adapter with mounted tracking sensor from the US Guide 2000® at the needle of the radiofrequency electrode.

Adapter for laparoscopic RFA needle For the connection of the US-Guide 2000™ navigation system to either expandable RFA needles (RITA Medical System Inc., Mountain View, CA, USA) or internally cooled RFA needles (Radionics Inc., Burlington, MA, USA) the electromagnetic tracker was mounted to the basis of the RFA needle (Fig. 4). We have used the adapter from the US-Guide 2000™ navigation system, which is normally designed for transcutaneous guided punctures and connected with the RFA needle in perpendicular direction. However, rotations of the adapter around the needle could not be compensated by the navigation system. Figure 4 Adapter with mounted tracking sensor from the US Guide 2000® at the needle of the radiofrequency electrode. Newly developed adapters All newly developed adapters were constructed using standard 3D CAD software (Solid Edge, UGS). The adapters were build from medical stainless steel alloys (1.4301 or 1.405) and Polyetherketone (PEEK). Prototypes were built in aluminium and polyoxymethilen (POM). Furthermore, new electromagnetic tracking systems (NDI Canada, Type AURORA® and Ascension, Type 3D-Guidance®) with smaller electromagnetic trackers were used.

Newly developed adapters All newly developed adapters were constructed using standard 3D CAD software (Solid Edge, UGS). The adapters were build from medical stainless steel alloys (1.4301 or 1.405) and Polyetherketone (PEEK). Prototypes were built in aluminium and polyoxymethilen (POM). Furthermore, new electromagnetic tracking systems (NDI Canada, Type AURORA® and Ascension, Type 3D-Guidance®) with smaller electromagnetic trackers were used. New adapters for laparoscopic ultrasound probe We first designed two different adapters to combine the LUS transducer with the electromagnetic tracking system. The first LUS adapter is a ring, which consists of two materials (PEEK and stainless steel alloys 1.405), that can be placed directly on the flexible part of the probe. The sensor is placed inside the inner PEEK part and is fixed with the outer steel part. It was evaluated in combination with the image-guided navigation system. The experimental setup consists of a perfused tumor-mimic-model of the porcine liver (Fig. 5) [5,6]. The second adapter, which is made of polyoxymethilen (POM), is pulled over the tip of the probe like a shoe (Fig. 6). The sensor is placed in a drilled hole at the back part of the adapter. Both adapters fit through a 17 mm trocar and were investigated regarding its influence on the electromagnetic trackers. Most recently we were able to design an adapter, which can be used with a 15 mm trocar and the 3D-Guidance tracking system (Fig. 7). This one is also made of PEEK and is mounted to the LUS tip using the puncture channel. Due to sensitivity of the electromagnetic sensors it is fixed inside a PEEK cover that shields the sensor. This Cover is also used to connect the sensor to the back part of the adapter in reproducible way.

tracking system (Fig. 7). This one is also made of PEEK and is mounted to the LUS tip using the puncture channel. Due to sensitivity of the electromagnetic sensors it is fixed inside a PEEK cover that shields the sensor. This Cover is also used to connect the sensor to the back part of the adapter in reproducible way. Figure 5 Connection of LUS transducers with AURORA® electromagnetic tracking by the "ring" adapter. Figure 6 Connection of the LUS transducer with the AURORA® electromagnetic tracking system by using the "shoe" adapter. Figure 7 Connection of the LUS transducer with any electromagnetic tracking system offering sensors with a diameter below 1.8 mm (a) and presentation of the technichal drawings from the 3D CAD software (b) New adapter for laparoscopic RFA needle The newly designed RFA needle adapter can be mounted on the needle itself (Fig. 8). Due to the rotational symmetry of the needle, only the distance from the needle tip must stay constant during calibration and intra-operative use of the tracked needle. The adapter is connected to the RFA-needle by fixing it with a little screw. The two holes of the adapter are used to take up the needle and the sensor. Like the LUS sensor, this sensor is also fixed inside a PEEK cover. Figure 8 Special developed adapter for the laparoscopic RFA-probe (a + b), which is mounted to the basis of the RFA-needle (d) and presentation of the technichal drawings from the 3D CAD software (c).

New adapter for laparoscopic RFA needle The newly designed RFA needle adapter can be mounted on the needle itself (Fig. 8). Due to the rotational symmetry of the needle, only the distance from the needle tip must stay constant during calibration and intra-operative use of the tracked needle. The adapter is connected to the RFA-needle by fixing it with a little screw. The two holes of the adapter are used to take up the needle and the sensor. Like the LUS sensor, this sensor is also fixed inside a PEEK cover. Figure 8 Special developed adapter for the laparoscopic RFA-probe (a + b), which is mounted to the basis of the RFA-needle (d) and presentation of the technichal drawings from the 3D CAD software (c). Results The first experiences with the development of adapters for electromagnetic trackers revealed the feasibility of the combination of an electromagnetic navigation system with a RFA needle and LUS [2-4,6]. However, problems of sterilisation, size of the adapters and fixation of the adapters had to be solved. Using the experiences of the initial versions of the adapters, sterilizable adapters have been developed using biocompatible materials only. The new adapters could be sterilized repeatedly without any problems or surface variations. After a short introduction, a surgical nurse and a surgeon were able to mount the adapters to the LUS probe and the RFA needle under sterile conditions without any difficulties.

been developed using biocompatible materials only. The new adapters could be sterilized repeatedly without any problems or surface variations. After a short introduction, a surgical nurse and a surgeon were able to mount the adapters to the LUS probe and the RFA needle under sterile conditions without any difficulties. The adapters for the LUS can be used to acquire navigated three-dimensional ultrasound even from segments that are difficult to reach (right lateral and cranial segments), because the electromagnetic tracker was attached to the tip of the LUS probe, which allows free angulation of the ultrasound transducer without interference of the navigation system. Initial material testing for the ring adapter showed no detectable interferences of the navigation system [7]. New laboratory tests could confirm that there is no influence of the ring adapter on the tracking system. However, the calibration of the ultrasound probe with the tracking system has to be performed after mounting of the ring adapter, which means under sterile conditions in the operating room. This procedure is time-consuming and is not effective for surgical procedures. The shoe-alike adapter prototype for the laparoscopic probe, which allows a reproducible positioning of the sensor, was designed to solve this problem. The shoe adapter, which consists only of POM has no influence on the tracking sensor, too. The most recently developed adapters for the LUS and laparoscopic RFA-needle have also been tested in organ models. They showed no disturbance of the tracking system as well as reproduceble calibration results due to their defined fixation and combined an ergonomic handling with a sufficient stability even after repeated sterilization.

developed adapters for the LUS and laparoscopic RFA-needle have also been tested in organ models. They showed no disturbance of the tracking system as well as reproduceble calibration results due to their defined fixation and combined an ergonomic handling with a sufficient stability even after repeated sterilization. The new developed adapter for the RFA probe showed neither influence on the NDI Aurora nor on the Ascension 3D-Guidance tracking system. Anatomic landmarks in the perfused porcine liver could be safely reached [6,8]. Accurate measurements are still outstanding. Overall, the newly developed adapter for the LUS and the RFA needle showed good feasibility, ergonomics, sterilizability and stability.

The new developed adapter for the RFA probe showed neither influence on the NDI Aurora nor on the Ascension 3D-Guidance tracking system. Anatomic landmarks in the perfused porcine liver could be safely reached [6,8]. Accurate measurements are still outstanding. Overall, the newly developed adapter for the LUS and the RFA needle showed good feasibility, ergonomics, sterilizability and stability. Discussion The laparoscopic approach for radiofrequency ablation, originally described by Siperstein et al. in 1997, offers a minimally invasive procedure in combination with the advantages of an open procedure [9]. Above that laparoscopic RFA can be performed with the same exactness and effectiveness in comparison to the open approach in well-selected patients [10,11]. Even carrying a higher access trauma laparoscopic RFA shows no significantly increased morbidity- or mortality-rates in comparison to the percutaneous probe application, which means a safe and mild treatment for the patient [10,12]. Unfortunately today there are also limitations to the laparoscopic approach based on general contraindications for laparoscopic procedures and the limitations of laparoscopic needle application. Ultrasound-guided navigation tools for laparoscopic RFA offer a new technique for interventional liver therapy. The major advantage is the possibility of out-of-plane needle placement and the combination of flexibility of free-hand type procedures with the accuracy of a biopsy-transducer. This improves the safety and accuracy of punctures and leads to an improvement of quality of the intervention. However, in order to combine an electromagnetic navigation system with the LUS and the laparoscopic RFA needle, special designed adapters are needed to mount the trackers to the instruments. These adapters must be sterilizable and ergonomic, should provide reproducibility and need to be designed in a way that the navigation system is not disturbed.

n electromagnetic navigation system with the LUS and the laparoscopic RFA needle, special designed adapters are needed to mount the trackers to the instruments. These adapters must be sterilizable and ergonomic, should provide reproducibility and need to be designed in a way that the navigation system is not disturbed. Our evaluation shows the difficulties in designing adapters that combine all required qualifications. Ongoing development is needed to solve existing problems, guarantee steady progress and achieve acceptable results for clinical use. This study presents the feasibility and usability of special designed adapters for electromagnetic trackers to combine an electromagnetic navigation system with the LUS and laparoscopic RFA needle. This is the prerequisite for navigated laparoscopic RFA to improve the accuracy and results of laparoscopic liver interventions. In the next steps we are planning to perform animal studies with the presented adapters. Furthermore, we are working on further miniaturization of the laparoscopic adapters to reduce the used trocar size. Besides that, adapters for hybrid tracking (optical and electromagnetical) of LUS and different RFA needles are already under evaluation, but can not be presented today for patent reasons.

ented adapters. Furthermore, we are working on further miniaturization of the laparoscopic adapters to reduce the used trocar size. Besides that, adapters for hybrid tracking (optical and electromagnetical) of LUS and different RFA needles are already under evaluation, but can not be presented today for patent reasons. Conclusion The design and development of adapters for electromagnetic trackers is the prerequisite for an accurate tracking of the RFA needle for laparoscopic navigation purposes as well as 3D navigated ultrasound data acquisition. We designed, built, tested and used different adapters for the use of a laparoscopic navigation system for the improvement of laparoscopic RFA. Competing interests The author(s) declare that they have no competing interests. Authors' contributions All named authors have made substantial contributions to conception, design and carrying out of the study by participating in the design and development as well as the experimental testing of the electromagnetic adapters. Acknowledgements The authors would like to thank Prof. Dr. S. Klein and his staff from the Laboratory for Devices Engineering, University of Applied Sciences Lübeck and D. Theisen-Kunde and R. Schulz, from Institute for Biomedical Optics Lübeck, University of Lübeck for their support and good collaboration.

Background Malignant brain tumors such as gliomas expand through proliferation and invasion within the confines of the bony skull, hence in a mechanically constraint area. That is, the brain tissue's mechanical 'reserve' rooms, i.e. the interstitial space, the cerebrospinal fluid filled ventricles and the vascular system can only temporarily compensate, through fluid shifts, any sudden or even gradual increase in intracranial pressure (ICP) as it relates to e.g. an intracerebral hemorrhage or a tumor. Once the point of decompensation is reached (Figure 1) any even miniscule increase in volume will trigger a massive, life-threatening increase in ICP [1] that generally leads to surgical intervention in an attempt to resect some of this pressure-raising volume. Figure 1 Intracranial pressure-volume curve. In the case of lower elastance (dotted line), the same increase in volume leads to a lower intracranial pressure increase than in the case of higher elastance (solid line) (from Hase et al. [1], with permission).

Background Malignant brain tumors such as gliomas expand through proliferation and invasion within the confines of the bony skull, hence in a mechanically constraint area. That is, the brain tissue's mechanical 'reserve' rooms, i.e. the interstitial space, the cerebrospinal fluid filled ventricles and the vascular system can only temporarily compensate, through fluid shifts, any sudden or even gradual increase in intracranial pressure (ICP) as it relates to e.g. an intracerebral hemorrhage or a tumor. Once the point of decompensation is reached (Figure 1) any even miniscule increase in volume will trigger a massive, life-threatening increase in ICP [1] that generally leads to surgical intervention in an attempt to resect some of this pressure-raising volume. Figure 1 Intracranial pressure-volume curve. In the case of lower elastance (dotted line), the same increase in volume leads to a lower intracranial pressure increase than in the case of higher elastance (solid line) (from Hase et al. [1], with permission). Standard treatment for highly aggressive gliomas (anaplastic astrocytomas and glioblastomas (GBM)) has therefore changed little over the years and still consists of gross surgical resection followed by adjuvant radiotherapy and chemotherapy (e.g., [2]). The problem is that the already spatially disseminated tumor cannot be visualized at the single cell level with current clinical imaging modalities, let alone be removed completely. Hence, after a period of temporary betterment, the tumor almost always returns, primarily locally, i.e., at the surgical edge, in conjunction with widespread tissue invasion that at some point prohibits further cytoreductive therapy. Therefore, regardless of any treatment, the grim outcome overall did not change in decades with most GBM patients succumbing to the disease within 1–2 years after diagnosis (e.g., [3]). New approaches are therefore desperately needed and, as always, the development of innovative strategies starts with a better understanding of the current limitations. Here, we discuss, from a pure theoretical and simple mechanistic perspective, if the first line approach of surgical debulking has any impact on the characteristic invasive behaviour of these cancer systems post-extirpation of the tumor core.

ive strategies starts with a better understanding of the current limitations. Here, we discuss, from a pure theoretical and simple mechanistic perspective, if the first line approach of surgical debulking has any impact on the characteristic invasive behaviour of these cancer systems post-extirpation of the tumor core. Hypothesis Presentation In vitro, invasive gliomas display finger-like invasive patterns in three-dimensional extracellular matrix environments [4], not unlike the chain-like migration seen from gliomas in vivo by following axonal fiber tracts and blood vessels [5,6]. Such 'fingering' morphology is common in other physical contexts as well: for instance, (i) when a 'crack' occurs in a solid specimen experiencing a point-like internal breakage or, with a growing inclusion inside [7], (ii) when a drop of liquid is injected in a more viscous environment ('Hele-Shaw' effect), or (iii) when a liquid drop impacts on a solid surface, causing the formation of a fluid 'crown' ('Rayleigh' or 'Yarin-Weiss' capillary instability [8]). Detailed investigations of these models predict the number N of fingers which are expected to form depending on the characteristics of the media. Notably, under some very general assumptions, the number of parameters to be accounted for predicting N is very low, and can be reduced to the radius R of the 'drop' or of the inclusion, the external pressure P which acts on it and finally, the surface tension σ. N≈PRσ

Hypothesis Presentation In vitro, invasive gliomas display finger-like invasive patterns in three-dimensional extracellular matrix environments [4], not unlike the chain-like migration seen from gliomas in vivo by following axonal fiber tracts and blood vessels [5,6]. Such 'fingering' morphology is common in other physical contexts as well: for instance, (i) when a 'crack' occurs in a solid specimen experiencing a point-like internal breakage or, with a growing inclusion inside [7], (ii) when a drop of liquid is injected in a more viscous environment ('Hele-Shaw' effect), or (iii) when a liquid drop impacts on a solid surface, causing the formation of a fluid 'crown' ('Rayleigh' or 'Yarin-Weiss' capillary instability [8]). Detailed investigations of these models predict the number N of fingers which are expected to form depending on the characteristics of the media. Notably, under some very general assumptions, the number of parameters to be accounted for predicting N is very low, and can be reduced to the radius R of the 'drop' or of the inclusion, the external pressure P which acts on it and finally, the surface tension σ. N≈PRσ As previously proposed in [9], based on Eq. (1), a dimensionless Invasion Parameter, IP, can be defined assuming that invasion occurs (IP > 1) when N ≥ 1. IP = PR/σ

Hypothesis Presentation In vitro, invasive gliomas display finger-like invasive patterns in three-dimensional extracellular matrix environments [4], not unlike the chain-like migration seen from gliomas in vivo by following axonal fiber tracts and blood vessels [5,6]. Such 'fingering' morphology is common in other physical contexts as well: for instance, (i) when a 'crack' occurs in a solid specimen experiencing a point-like internal breakage or, with a growing inclusion inside [7], (ii) when a drop of liquid is injected in a more viscous environment ('Hele-Shaw' effect), or (iii) when a liquid drop impacts on a solid surface, causing the formation of a fluid 'crown' ('Rayleigh' or 'Yarin-Weiss' capillary instability [8]). Detailed investigations of these models predict the number N of fingers which are expected to form depending on the characteristics of the media. Notably, under some very general assumptions, the number of parameters to be accounted for predicting N is very low, and can be reduced to the radius R of the 'drop' or of the inclusion, the external pressure P which acts on it and finally, the surface tension σ. N≈PRσ As previously proposed in [9], based on Eq. (1), a dimensionless Invasion Parameter, IP, can be defined assuming that invasion occurs (IP > 1) when N ≥ 1. IP = PR/σ Here, P stands for the confining mechanical pressure exerted by the microenvironment on the tumor, R for the radius of the tumor, and σ for the tension at the tumor surface. According to Eq. (2), invasive cell behaviour is to be expected in all but for the case of IP < 1 (which implies large tumor surface tension, small confining pressure and/or small tumor radius values).

erted by the microenvironment on the tumor, R for the radius of the tumor, and σ for the tension at the tumor surface. According to Eq. (2), invasive cell behaviour is to be expected in all but for the case of IP < 1 (which implies large tumor surface tension, small confining pressure and/or small tumor radius values). In the case of malignant gliomas, surgical debulking is geared to reduce the tumor burden, hence R should decrease, which in turn would diminish the product PR and result in an overall IP decline, if σ would to remain constant. To be more specific, according to Schettini and Walsh [10] the reduction of an intracranial balloon volume from 6 ml to 0.5 ml (i.e., of a factor of 12) would produce a reduction of the product PR, being P the intracranial pressure measured in the ipsilateral ventricle of a factor ranging from 24 to 8.8 mmHg, i.e., roughly of an order of magnitude. Such measurements, however, are performed 'acutely', i.e., before compensatory mechanisms, due to both the nonlinear viscoelastic behaviour of the brain and circulatory autoregulation, are elicited. Therefore, on a closer look, there are a few important caveats:

m 24 to 8.8 mmHg, i.e., roughly of an order of magnitude. Such measurements, however, are performed 'acutely', i.e., before compensatory mechanisms, due to both the nonlinear viscoelastic behaviour of the brain and circulatory autoregulation, are elicited. Therefore, on a closer look, there are a few important caveats: • First, this 'volume-pressure' relationship or elastance (and its inverse quantity, compliance) found for the inflated epidural balloon in experimental settings should, in reality, only hold for benign tumors that grow strictly compressive and non-infiltrative in nature such as most meningiomas [11]. In fact, Hase et al. [1] already compared the ICP dynamics post-surgery for such meningiomas versus astrocytomas. The authors reported a higher elastance for meningiomas (comparable to the solid line in Figure 1), as compared to a lower elastance for malignant astrocytomas (suggested as dotted line; see also Table 3 in [1]). That is, in astrocytomas, the same increase in volume leads to a lesser increase in ICP. We argue that the distinct infiltrative nature of these malignant tumors is responsible for causing such 'favourable' mechanical conditions for the neoplasm. (Gliomas for instance release the neutotoxicity-inducing excitatory amino acid glutamate [12] and produce matrix degrading enzymes or metalloproteinases [13] which together can impact and disrupt surrounding tissue architecture, as seen on diffusion tensor NMR [14,15]. None of the complicated biochemical processes driving tumor progression is explicitly included in our hypothesis here; rather, we focus on some simple biomechanical aspects involved). Further, this may delay the onset of purely pressure-related symptoms which in turn can help explain the surprisingly large sizes of some of these malignant tumors at the time of diagnosis. Conversely, surgically reducing tumor volume will also lead to a relatively smaller rate of reduction in P. Returning then to Eq. (2), post-surgical intervention, the PR product may remain elevated and any sub-total resection (with a sizeable portion of R remaining) will only add to this tendency.

he time of diagnosis. Conversely, surgically reducing tumor volume will also lead to a relatively smaller rate of reduction in P. Returning then to Eq. (2), post-surgical intervention, the PR product may remain elevated and any sub-total resection (with a sizeable portion of R remaining) will only add to this tendency. • Secondly, by definition, gross total resection will aim to remove the (on conventional imaging, now generally nuclear magnetic resonance, or NMR, imaging) visible surface of the tumor. The Laplace law should hold for the resulting spherical shell and thus, σ should, if anything, become very small postoperatively which in turn would yield an excessive increase in IP for the remaining, already disseminated cancer cells that remain below the current imaging resolution. It is noteworthy in that context that dexamethasone, a glucocorticosteroid often used peri-operatively with the idea of treating tumoral edema formation, has been shown to also increase the surface tension of the malignant astrocytoma cell lines U-87MG, U-118MG and LN-229 [16]. One could therefore argue that this effect may be one reason why peri-operative steroid therapy is effective to some degree. • And finally, adding the experimentally proven notion that more invasive cells move faster on stiffer substrates [17-19] to the aforementioned lower elastance of astrocytomas post-extirpation also makes for an invasiveness-promoting microenvironment for the remaining tumor cell formations.

• Secondly, by definition, gross total resection will aim to remove the (on conventional imaging, now generally nuclear magnetic resonance, or NMR, imaging) visible surface of the tumor. The Laplace law should hold for the resulting spherical shell and thus, σ should, if anything, become very small postoperatively which in turn would yield an excessive increase in IP for the remaining, already disseminated cancer cells that remain below the current imaging resolution. It is noteworthy in that context that dexamethasone, a glucocorticosteroid often used peri-operatively with the idea of treating tumoral edema formation, has been shown to also increase the surface tension of the malignant astrocytoma cell lines U-87MG, U-118MG and LN-229 [16]. One could therefore argue that this effect may be one reason why peri-operative steroid therapy is effective to some degree. • And finally, adding the experimentally proven notion that more invasive cells move faster on stiffer substrates [17-19] to the aforementioned lower elastance of astrocytomas post-extirpation also makes for an invasiveness-promoting microenvironment for the remaining tumor cell formations. Implications In Guiot et al. [9] we had deduced from Eq. (2) that increasing σ (e.g., through stable E-cadherin transfection as described already for prostate cancer cells in [20]), as well as reducing P and R are the rather obvious, distinct therapeutic goals. While this remains correct, what deserves a more careful evaluation is their relationship particularly within the anatomic confines of a host organ such as the brain. Surgical impact, while reducing R, temporarily relieving P and thus at least subjectively improving the patient's symptoms can, for the reasons stated above, potentially have detrimental inductive effects on tumor invasion for the cells left behind, which are generally believed to significantly contribute to recurrence and treatment failure (e.g., [21]). One could therefore argue, that surgery fails to prevent if not even facilitates the transition of the tumor from a rather volumetric growth process to a more diffusely expanding system, where infiltrating glioma cells can switch to a proliferative phenotype and vice versa [22]; since such single cell dynamics remain hidden largely below the current NMR imaging threshold, this spatio-temporally propagating system now provides the clinician with lesser if any macroscopic targets to go after, up until a sizeable tumor recurs. However, suggesting not to operate would be irresponsible since the severity of clinical symptoms all too often mandates rapid surgical ICP relief and, not surprisingly, numerous studies have shown that gross total resection improves the patient's disease-free interval (e.g., [23]). That is, aside from obtaining tissue for a histo-pathological confirmation of tumor grade, the main biomechanical goal of any surgical intervention is to approximate the flat part of Figure 1. Our conjecture only cautions about the standard approach's potential side effects on the invasive cells left behind and, in doing so, (1) confirms the indication for pharmaceutical peri-operative ICP reduction, and (2) reemphasizes the desperate need for effective anti-invasive targeting modalities that are administered in conjunction with surgery (and conventional adjuvant regimen).

al side effects on the invasive cells left behind and, in doing so, (1) confirms the indication for pharmaceutical peri-operative ICP reduction, and (2) reemphasizes the desperate need for effective anti-invasive targeting modalities that are administered in conjunction with surgery (and conventional adjuvant regimen). (Surgery of solid malignant tumors usually attempts to excise larger margins in healthy tissue which, however, is a limited option in the brain where any even minimal tissue damage can add to the functional deficit already caused by the tumor).

al side effects on the invasive cells left behind and, in doing so, (1) confirms the indication for pharmaceutical peri-operative ICP reduction, and (2) reemphasizes the desperate need for effective anti-invasive targeting modalities that are administered in conjunction with surgery (and conventional adjuvant regimen). (Surgery of solid malignant tumors usually attempts to excise larger margins in healthy tissue which, however, is a limited option in the brain where any even minimal tissue damage can add to the functional deficit already caused by the tumor). The former (1) refers also to avoiding an effective increase in P, known as an 'ICP rebound phenomenon' that can occur e.g. when the osmotic diuretic agent Mannitol (administered to treat elevated ICP) extravasates into the tissue which in turn can cause a reverse osmotic shift and thus a detrimental raise in ICP [24]. Given that glucocorticoids are still the treatment of choice, one would hope for a renaissance of the pharmaceutical field that is geared towards intracranial pressure reduction. The latter (2) supports for instance further development of the recently introduced anti-angiogenesis compounds. They are designed to impact the nutrient supply system (by targeting integrin receptors [25] or related growth factors such as VEGF [26]), thus supposedly help manage the PR dynamics of any residual or recurrent tumor; in doing so, they should also impact structural dissemination routes for invasive glioma cells which are known to move along the perivascular space amongst others [27]. Interestingly, the perivascular space seems to be also a permissive location of cancer stem cells [28] that have been implicated in the development of treatment resistance [29]. Taken together, while recent results raise some caution [30], so far anti-angiogenetic drugs remain a promising strategy as we have argued already in [31] on the basis of organ specific albeit limited carrying capacity.

of cancer stem cells [28] that have been implicated in the development of treatment resistance [29]. Taken together, while recent results raise some caution [30], so far anti-angiogenetic drugs remain a promising strategy as we have argued already in [31] on the basis of organ specific albeit limited carrying capacity. In summary, following a primarily biomechanical conjecture, the impact caused by undoubtedly necessary surgical intervention may have unintended, inductive effects on invasiveness of the glioma cells left behind. This emphasizes the need for pharmaceutical peri-operative ICP reduction, and supports innovative strategies geared directly towards targeting disseminated cells. Competing interests The author(s) declare that they have no competing interests. Authors' contributions TSD developed the hypothesis and drafted the manuscript, with CG being responsible for the mathematical description. Both authors read and approved the final manuscript. Acknowledgements This work has been supported in part by NIH grants CA 085139 and CA 113004 and by the Harvard-MIT (HST) Athinoula A. Martinos Center for Biomedical Imaging and the Department of Radiology at Massachusetts General Hospital. We thank Dr. Roger Haertl (Department of Neurosurgery, New York Hospital & Cornell University) for critical review of the manuscript.

1. Background Esophagogastric anastomotic leakage is a major cause of morbidity and mortality in patients who underwent an esophagectomy [1-3]. Its incidence varies from 5–25%. Many publications have described the cause of this complication, and it has been described that ischemia of the gastric tube is an important cause [4-6]. Scarce information is available about potential solutions to this problem. Research has been done [7] showing a fall in gastric Pt O2 following gastric devascularization, but not after its mobilization. Some articles [8-10] describe pre-operative embolization as a way of increasing the vascularization of the gastric fundus. Others [11,12] hypothesize surgical devascularization of the gastric fundus and delayed anastomosis as a potential solution. A two stage esophagectomy has been described, but transposing the stomach and making the esophagogastric anastomosis after several days. In an attempt to increase gastric vascular flow, we performed a gastric devascularization, as done in a standard esophagectomy. Afterwards, the stomach was partially tubulized, making a greater conditioning. Its vascular flow was measured with a laser doppler (PeriMed, Ohio, US) -endoscopic probe- before and after vascular ligature and partial sectioning and, after 3 weeks, a reoperation allowed us to measure vascular flow once again. Our aim is to evaluate if partial devascularization and transection of the stomach stimulates collateral circulation to develop, and in this way, increase the vascular flow of the gastric fundus.

In an attempt to increase gastric vascular flow, we performed a gastric devascularization, as done in a standard esophagectomy. Afterwards, the stomach was partially tubulized, making a greater conditioning. Its vascular flow was measured with a laser doppler (PeriMed, Ohio, US) -endoscopic probe- before and after vascular ligature and partial sectioning and, after 3 weeks, a reoperation allowed us to measure vascular flow once again. Our aim is to evaluate if partial devascularization and transection of the stomach stimulates collateral circulation to develop, and in this way, increase the vascular flow of the gastric fundus. Additionally, we evaluated the safety of the procedure, in terms of complications, and OR time. The clinical application would be that the risk of esophagogastric anastomotic leakage might be reduced if a better vascularized fundus can be achieved with this technique. 2. Methods Eight animals were used in this study. The pigs were premedicated with IM atropine (0.04 mg/kg) and IM Ketamine (15–25 mg/Kg) or Telazol 5–10 mg/kg IM and received IV Thiopentobarbital (5–11 mg/kg) for induction of anesthesia. Isofluorane (1.5%) was used for maintenance of anesthesia and titrated to effect. For analgesia, 0.01 mg/kg of IV buprenorphine was administered intra-operatively, and a transdermal Fentanyl patch (100 mcg) was placed for 72 hours post-operatively analgesia. Toradol at 0.3–0.7 mg/kg iv/im or Flunixine meglumine at 0.5–2.2 mg/kg iv/im was used if supplemental analgesia was required.

trated to effect. For analgesia, 0.01 mg/kg of IV buprenorphine was administered intra-operatively, and a transdermal Fentanyl patch (100 mcg) was placed for 72 hours post-operatively analgesia. Toradol at 0.3–0.7 mg/kg iv/im or Flunixine meglumine at 0.5–2.2 mg/kg iv/im was used if supplemental analgesia was required. After induction, an open technique was used to place a 10 mm trocar in the abdomen, and pneumoperithoneum was established. A 30 degree laparoscope was used. Five 10 mm trocar and one 5 mm trocar were used. As a first step, the vascular flow at the fundus was measured with a laser doppler. With the Harmonic scalpel (Tyco HealthCare, CT, US), the stomach was mobilized (short gastric vessels sectioned). A vascular stapler was used for the transection of the left gastric pedicle. Following its devascularization, the stomach was be partially tubulized. Beginning from the His's angle, and heading downwards, three sequential cartridges of Endo GIA (Tyco HealthCare, CT, US) were fired [figure 1]. Vascular flow was re-measured at this point. Figure 1 Partial gastric transection after being devascularized. After the procedure, the animals were maintained NPO for the first 12 hs and then gradually advance from water to solid food over the next 24 hs. Three weeks later, the pigs were re-operated. The aim of this second procedure was to measure once again the vascular flow at the fundus, and evaluate macroscopically if devascularization had a positive impact in developing collateral vascularization. A sample of gastric fundus was taken for histopathologic evaluation.

After the procedure, the animals were maintained NPO for the first 12 hs and then gradually advance from water to solid food over the next 24 hs. Three weeks later, the pigs were re-operated. The aim of this second procedure was to measure once again the vascular flow at the fundus, and evaluate macroscopically if devascularization had a positive impact in developing collateral vascularization. A sample of gastric fundus was taken for histopathologic evaluation. 3. Results In order to evaluate our results, macroscopic appearance of the fundus after three weeks of the first operation, microscopic evaluation of the fundus after three weeks of the first operation and laser doppler measurements were taken into account. In our series, all the animals had a "pink" appearance during the last operation. This correlates well with the microscopic evaluation, in which an increased number of blood vessels was present in the fundus when compared with histologically normal areas. For this purpose, the vessels (mean value) enclosed within a uniform rectangular area (as defined by a set of reticles within the microscope) in five random 20× fields in the affected and unaffected areas of tissue were measured. The result showed 10/12/14/11/9 and 4/5/6/6/5 vessels respectively [figure 2]. Figure 2 Histopathologic evaluation of gastric tissue. The graphic shows the increase in the number of vessels in the neovascularized portion of the stomach (most distal part of the fundus).

In our series, all the animals had a "pink" appearance during the last operation. This correlates well with the microscopic evaluation, in which an increased number of blood vessels was present in the fundus when compared with histologically normal areas. For this purpose, the vessels (mean value) enclosed within a uniform rectangular area (as defined by a set of reticles within the microscope) in five random 20× fields in the affected and unaffected areas of tissue were measured. The result showed 10/12/14/11/9 and 4/5/6/6/5 vessels respectively [figure 2]. Figure 2 Histopathologic evaluation of gastric tissue. The graphic shows the increase in the number of vessels in the neovascularized portion of the stomach (most distal part of the fundus). Laser doppler measurements showed an initial drop in the vascular flow (immediately after the devascularization) and an increase when measured three weeks after [figure 3]. Figure 3 Initial decrease in perfusion after devascularization, and its increase after a three period week. The post operative course of all the animals was uneventful, and the mean OR time was 80 min. No leaks of the stapler line were seen. Ulceration and necrosis, two possible complications of the procedure, were not seen.

Figure 3 Initial decrease in perfusion after devascularization, and its increase after a three period week. The post operative course of all the animals was uneventful, and the mean OR time was 80 min. No leaks of the stapler line were seen. Ulceration and necrosis, two possible complications of the procedure, were not seen. 4. Discusion Esophagogastric leakage following an esophagectomy in a common cause of morbidity and mortality. Its incidence varies according to the series, ranging from 5% to 25%. In order to decrease fistula rates, many investigators tried to find the source of this problem in order to avoid it. One of the most important factors leading to fistula formation is a low vascular flow, that interferes with the healing process [4-6]. It has also been described that a drop in the vascular flow occurs immediately after its tubulization, and stays stable after its ascension through the thorax up to the neck [7]. Some investigators have performed a gastric devascularization and delayed for 2–3 weeks the anastomosis, finding a better outcome in terms of fistula formation in these animals [11] while others [9,10] used arterial embolization to create a gastric conditioning and applied it in the clinical setting.

4. Discusion Esophagogastric leakage following an esophagectomy in a common cause of morbidity and mortality. Its incidence varies according to the series, ranging from 5% to 25%. In order to decrease fistula rates, many investigators tried to find the source of this problem in order to avoid it. One of the most important factors leading to fistula formation is a low vascular flow, that interferes with the healing process [4-6]. It has also been described that a drop in the vascular flow occurs immediately after its tubulization, and stays stable after its ascension through the thorax up to the neck [7]. Some investigators have performed a gastric devascularization and delayed for 2–3 weeks the anastomosis, finding a better outcome in terms of fistula formation in these animals [11] while others [9,10] used arterial embolization to create a gastric conditioning and applied it in the clinical setting. The physiologic mechanism of neovascularization development in tissue conditioning has been well described with an analogous procedures. The skin flap transposition for cutaneous reconstruction [13] is an example. In these studies, the flap was partially devascularized and mobilized, and after 2–3 weeks, the skin flap was transposed. Flap necrosis and wound dehiscence were reduced.

issue conditioning has been well described with an analogous procedures. The skin flap transposition for cutaneous reconstruction [13] is an example. In these studies, the flap was partially devascularized and mobilized, and after 2–3 weeks, the skin flap was transposed. Flap necrosis and wound dehiscence were reduced. In an attempt to decrease the vascularization of the fundus at its best, during the first procedure, we partially transected the stomach as it would be done for a gastric tubulization during an esophagectomy in addition to its devascularization (sectioning of the gastroepiploic, short gastric and left gastric pedicle) and mobilization. In this way, the submucosal vascularization would be reduced, producing a greater ischemic conditioning. The macroscopic and microscopic examination of the fundus, as well as the measurement of the vascular flow with a laser doppler (PeriMed, Oh, US) with an endoscopic probe, helped us in assessing the effectiveness of the devascularization and neovascularization process. Potential complications of the procedure we have performed in the series include the ones related to the devascularization and transection processes, and include gastric ulceration and necrosis (devascularization process) and fistula formation secondary to a failure of the stapler line (transection process). None of these complications were present in the ten animals. Besides, though it is not an identical -but similar- procedure, these complications are not widely described in patients operated on for a Collis-Nissen operation.

rocess) and fistula formation secondary to a failure of the stapler line (transection process). None of these complications were present in the ten animals. Besides, though it is not an identical -but similar- procedure, these complications are not widely described in patients operated on for a Collis-Nissen operation. Regarding the time we took to delay the second procedure, we decided to perform it three weeks after the first one. Our explanation to this relies on previous studies [13] and on the potential clinical application of this "two stage" procedure. One can hypothesize that during the first laparoscopic procedure, the surgeon can perform a diagnostic laparoscopy, the devascularization and partial transection procedure and then, the placement of a feeding jejunostomy. If the patient has metastasis or the tumor is unresectable, no further operation is required. On the other hand, if the patient is suitable for an esophageal resection, a second operation will have to be done. A three week period is a reasonable procedure to feed the patient (most patients with esophageal cancer are malnourished) and allow the stomach to develop neovascularization. Here, another question rises. Is a three weeks period, the top of the curve of the neovascularization process? We do not have that answer, but we think that a three week period is transposable to the clinical setting in both oncological and neovascularization process point of view. As stated before, the first articles reporting conditioning of skin flaps used a 2–3 weeks period, reaching good results in terms of wound healing and avoidance of necrosis [13], and Akiyama et al [9] stated that a minimum period of one week is needed to neovascularize the embolized stomach.

tion process point of view. As stated before, the first articles reporting conditioning of skin flaps used a 2–3 weeks period, reaching good results in terms of wound healing and avoidance of necrosis [13], and Akiyama et al [9] stated that a minimum period of one week is needed to neovascularize the embolized stomach. 5. Conclusion We can state the this "two stage" procedure can be applicable for cases of esophageal cancer in which the surgeon can offer the patient the minimally invasive approach. If unresectable, the patient will benefit from avoiding a large incision, but if the tumor is amenable to resection, the stomach will have a three week period increase its vascular flow (decrease incidence of fistula) and the surgeon can use this period to feed the patient through the feeding jejunostomy, which is enough to neovascularize the stomach and feed the patient and avoid leaving the tumor in situ and promote its seeding. Further studies will be needed to confirm this hypothesis.

Background Although normal healing involves the migration of fibroblasts to a wound site to form a matrix of scar tissue, excessive scar tissue may cause impairment in function, either by direct compression or by tethering of nervous tissues to the surrounding muscle or bone. Such epidural fibrosis is estimated as contributing to 60% of all cases of recurring back pain symptoms in the heterogeneous condition known as failed back surgery syndrome (FBSS) which occurs after discectomy or laminectomy [1]. The direct costs of further diagnostic measures, treatments and repeat surgery, as well as additional societal costs, emphasize the importance of preventive measures during surgery. While spinal surgeons have responded to this challenge by improvements in surgical technique, the problem remains and use of an interposition agent to reduce scarring after discectomy is recognized in the recent Cochrane review as likely to be an important strategy [2]. Recently several types of bioresorbable gels applied directly to the organs and acting as a chemical or physical barrier to surgical adhesions have been approved by licensing regulators. As devices, they are subject to different approval criteria than drug products and since the type of procedures which use these devices are often limited in numbers, the initial safety and efficacy data can often be based on small numbers of patient experiences.

cal adhesions have been approved by licensing regulators. As devices, they are subject to different approval criteria than drug products and since the type of procedures which use these devices are often limited in numbers, the initial safety and efficacy data can often be based on small numbers of patient experiences. The importance of close monitoring and the collection of safety data when new devices first receive approval for routine use should not be underestimated. Recently, Adcon-L®, a promising agent licensed for use as an adhesion-reduction device in laminectomy, was withdrawn after approval due to serious adverse events [3-5] which were not apparent in the early clinical trials [6-8]. In light of such events we decided to prospectively evaluate the safety of a new adhesion-reduction device, carboxymethylcellulose (CMC) and polyethylene oxide (PEO) gel, in a consecutive series of patients undergoing spinal microdiscectomy for herniation. Such a study is in accordance with the essential requirements of the Medical Devices Directive [9] and it is good surgical practice to closely monitor use of any new technique, device or other treatment as it is introduced into routine surgery.

cutive series of patients undergoing spinal microdiscectomy for herniation. Such a study is in accordance with the essential requirements of the Medical Devices Directive [9] and it is good surgical practice to closely monitor use of any new technique, device or other treatment as it is introduced into routine surgery. Methods CMC and PEO have separately demonstrated anti-adhesion properties [10-12] and have a good safety record in widespread clinical use. The combination CMC/PEO gel is a 100% synthetic gel (OXIPLEX®/SP Adhesion Barrier Gel, manufactured by FzioMed, Inc. San Luis Obispo, CA, USA and distributed under the trade names OXIPLEX®/SP Adhesion Barrier Gel, DePuy International Ltd, Leeds, UK and MEDISHIELD™ Adhesion Barrier Gel, Medtronic International Trading SARL, Tolochenaz, Switzerland) which received European Class III Device approval and became available for use in Europe in 2002 as an adhesion inhibitor following spinal surgery. Preclinical [13] and initial clinical work [14] with the combination gel showed significant reduction of epidural fibrosis.

rnational Trading SARL, Tolochenaz, Switzerland) which received European Class III Device approval and became available for use in Europe in 2002 as an adhesion inhibitor following spinal surgery. Preclinical [13] and initial clinical work [14] with the combination gel showed significant reduction of epidural fibrosis. In our safety evaluation of this approved agent, a group of 396 patients suffering from one level disc herniation and presenting either with radicular pain resistant to conservative treatment or with radicular pain associated with motor or sensory loss was operated upon by the same surgeon in the same institution (Clinique du Parc Léopold, Brussels, Belgium) between January 1st 2003 and December 31st 2005. Upon completion of a conventional microdiscectomy, in all patients the decompressed nerve root was systematically covered with a thin layer of [CaCL2 + NaCl + Na carboxymethylcellulose (CMC) + polyethylene oxide (PEO)] gel to prevent excessive growth of scar tissue. Because of previous experience with other anti-adhesion gels, all patients were monitored during the application of the gel looking for changes in blood pressure. They were all reviewed 1 and 6 weeks after surgery. The wound was then subjectively assessed for abnormal healing, redness or subcutaneous collections. Finally, all patients that required reoperations for recurrent disc herniation were carefully evaluated perioperatively for scar tissue formation on the incriminated nerve roots as an assessment of efficacy of the CMC/PEO gel in reducing fibrosis formation. The work was reviewed by the local ethics committee and performed in accordance with the ethical standards laid down in the 1964 Declaration of Helsinki with patients giving their informed consent prior to inclusion in the evaluation.

roots as an assessment of efficacy of the CMC/PEO gel in reducing fibrosis formation. The work was reviewed by the local ethics committee and performed in accordance with the ethical standards laid down in the 1964 Declaration of Helsinki with patients giving their informed consent prior to inclusion in the evaluation. Results Handling and use of the gel was simple. No patient presented with any clinically measurable adverse event during surgery at the time of the application of the gel. The mean length of stay after surgery was 5 days, consistent with normal regional practice. Subjective reviews at both week 1 and week 6 post-surgery demonstrated no local skin reaction and no postoperative fluid collection that required drainage. One patient required reoperation after 13 days for an infection of the wound, but we encountered no other abnormalities in wound healing. Five other patients needed reoperations for recurrent herniation, two after less than a week, one after one month, and two within the first year after surgery.

hat required drainage. One patient required reoperation after 13 days for an infection of the wound, but we encountered no other abnormalities in wound healing. Five other patients needed reoperations for recurrent herniation, two after less than a week, one after one month, and two within the first year after surgery. In perioperative assessment of the early reoperations, as expected, there was little or no scar tissue. However, although anecdotal, it is worth mentioning that clinically significant fibrosis reduction was observed in the three patients presenting with delayed recurrences – one patient reoperated after one month and the two patients reoperated within the first year. Figure 1 provides a perioperative view of the operative field in the patient who had recurrent disc herniation one year after initial surgery with CMC/PEO gel. Note the clear limits of the L5 lumbar nerve root (arrow) and the lack of adherent scar tissue. This allowed facilitated dissection and separation of the nerve root from the surrounding tissues. At the end of the reoperation CMC/PEO gel was used again. The herniation did not recur and the clinical evolution of this patient was uneventful. Figure 1 Perioperative view of the operative field for recurrent disc herniation one year after initial surgery with CMC/PEO gel. Note the clear limits of the L5 lumbar nerve root (arrow) and the lack of adherent scar tissue.

In perioperative assessment of the early reoperations, as expected, there was little or no scar tissue. However, although anecdotal, it is worth mentioning that clinically significant fibrosis reduction was observed in the three patients presenting with delayed recurrences – one patient reoperated after one month and the two patients reoperated within the first year. Figure 1 provides a perioperative view of the operative field in the patient who had recurrent disc herniation one year after initial surgery with CMC/PEO gel. Note the clear limits of the L5 lumbar nerve root (arrow) and the lack of adherent scar tissue. This allowed facilitated dissection and separation of the nerve root from the surrounding tissues. At the end of the reoperation CMC/PEO gel was used again. The herniation did not recur and the clinical evolution of this patient was uneventful. Figure 1 Perioperative view of the operative field for recurrent disc herniation one year after initial surgery with CMC/PEO gel. Note the clear limits of the L5 lumbar nerve root (arrow) and the lack of adherent scar tissue. Discussion The first gels to be used for prevention of fibrosis were semi-synthetic carbohydrate polymers such as GT 1587 (Adcon-L®, Gliatech Inc.). Initial animal models proved favorable [15,16] and subsequent clinical research [6-8] indicated the agent was effective in reducing fibrosis. It is our local policy to closely monitor new agents as they are adopted in routine surgery and with the introduction of GT 1587 we commenced evaluation on a consecutive series of patients in whom we used the agent. After treating 46 patients with GT 1587 we identified three patients with painful erysipeloïd-like skin reactions. Two other patients needed reoperations for unexplained painful subcutaneous fluid collections which proved not to be infection related. The possibility of an exsudative reaction to GT 1587 attributed to its animal gelatin a component was put forward. A local complication rate of 10.8% was found unacceptable in our local risk benefit assessment and we stopped using this gel at an early stage. (Personal presentations; Belgian Neurosurgical Society, 2004). Further reports of serious complications associated with the use of GT 1587 began emerging in the literature after the gel had gone into routine clinical use, including per-operative tachycardia and hypotension [3] and increased rate of CSF leakage [4,5]. GT 1587 was subsequently withdrawn from use.

cal Society, 2004). Further reports of serious complications associated with the use of GT 1587 began emerging in the literature after the gel had gone into routine clinical use, including per-operative tachycardia and hypotension [3] and increased rate of CSF leakage [4,5]. GT 1587 was subsequently withdrawn from use. Our personal experiences with GT1587 highlighted the need for close postoperative vigilance with the use of new products even though it was approved for clinical use as having met the essential requirements of the European Medical Devices Directive [9]. A similar issue recently arose with a promising new anti-adhesion gel in gynecological surgery resulting in withdrawal of the agent after both European and USA approval of the device [17]. On the basis of this latter experience, with the European approval of a subsequent adhesion-reduction solution for gynecological and general surgery, a Europe-wide safety registry was established to allow coordinated close monitoring of the device as it went into routine surgical use [18,19].

proval of the device [17]. On the basis of this latter experience, with the European approval of a subsequent adhesion-reduction solution for gynecological and general surgery, a Europe-wide safety registry was established to allow coordinated close monitoring of the device as it went into routine surgical use [18,19]. The results of our clinical safety assessment of CMC/PEO in spinal surgery were positive. While not a randomized controlled clinical efficacy study, this case series represented responsible safety monitoring of an approved agent. In the case series, we experienced no complications related to use of CMC/PEO gel and only five patients needed reoperations for recurrent herniation, which was lower than our previous experience (Personal presentations Société Francophone de Neurochirurgie du Rachis 2006 and European Association of Neurological Societies 2006). It should be noted that, according to communications with the manufacturer of this device (FzioMed, Inc.), that nearly 100,000 units have been distributed worldwide since the product was introduced in 2002 and thus far, there have been no reports of adverse events that were attributable to the device. While perioperative bleeding control, avoidance of excessive nerve root retraction and nucleus remnants during surgery, as well as adequate decompression remain the cornerstones of FBSS prevention – we believe the use of an interposition agent is an important element of an overall anti-adhesion strategy.

The results of our clinical safety assessment of CMC/PEO in spinal surgery were positive. While not a randomized controlled clinical efficacy study, this case series represented responsible safety monitoring of an approved agent. In the case series, we experienced no complications related to use of CMC/PEO gel and only five patients needed reoperations for recurrent herniation, which was lower than our previous experience (Personal presentations Société Francophone de Neurochirurgie du Rachis 2006 and European Association of Neurological Societies 2006). It should be noted that, according to communications with the manufacturer of this device (FzioMed, Inc.), that nearly 100,000 units have been distributed worldwide since the product was introduced in 2002 and thus far, there have been no reports of adverse events that were attributable to the device. While perioperative bleeding control, avoidance of excessive nerve root retraction and nucleus remnants during surgery, as well as adequate decompression remain the cornerstones of FBSS prevention – we believe the use of an interposition agent is an important element of an overall anti-adhesion strategy. Conclusion In close monitoring of use in a large consecutive case series, there were no adverse events related to the use of the CMC/PEO gel. On the basis of this safety evaluation, we are currently using CMC/PEO on a routine basis for all microdiscectomy procedures as part of our overall anti-adhesion strategy and await the finding of prospective control based efficacy studies with interest.

here were no adverse events related to the use of the CMC/PEO gel. On the basis of this safety evaluation, we are currently using CMC/PEO on a routine basis for all microdiscectomy procedures as part of our overall anti-adhesion strategy and await the finding of prospective control based efficacy studies with interest. Competing interests No funding was provided for the conduct of this case review. The work was presented as an oral paper at the 2007 American Association of Neurological Surgeons meeting, April 16th, 2007. The costs of travel and hotel accommodation were reimbursed by FzioMed, Inc. The article-processing charge from the Annals of Surgical Innovation and Research will be covered by a grant from FzioMed, Inc. who have also provided grant funding to facilitate the support of Corvus Communications in the development of this manuscript. Authors' contributions As sole author I conceived and designed the study evaluation, collected and analyzed the findings, developed the core draft manuscript and have approved the final manuscript. Acknowledgements Medical writing support was provided by Alison Crowe and Alastair Knight at Corvus Communications, Buxted, UK with financial assistance from FzioMed, Inc.

Introduction We have the largest reported series of postoperative patients immediately fed ≥ 100 kcal/hr without encountering "feeding intolerance" We re-examined our data, reaching a conclusion that had escaped us at the time. Our patients had been spared the risk of overfeeding because we had avoided the trigger, which was localized distention at the enteral feeding site. Our attention had been focused primarily on efficiently intercepting and removing swallowed air while feeding. We aspirated the proximal duodenum, where the fluid path was narrowed, efficiently removing that gas. Fortuitously, nutrition was introduced into the slightly more distal duodenum. Total inflow that exceeded the impaired peristaltic duodenal outflow refluxed freely within this short, relatively flaccid segment. This refluxing excess also was promptly removed via the more proximal duodenal aspiration orifices. The patients would not have fared as well had we fed into the jejunum. We now postulate that when peristaltic outflow from the feeding site is exceeded by total inflow (feedings superimposed on digestive juices plus swallowed saliva and air), fluid accumulates at this proximal location. The vagally mediated reflex response is slowing of the already sluggish entire gut, leading to a "downhill spiral" of generalized abdominal distention, nausea, and malaise. Respiratory embarrassment develops secondarily. Unexplained bowel necrosis occurs after jejunal feeding, with a reported incidence of 1:1,000. This, too, might similarly follow vagal vascular reflexes in susceptible subjects [1].

, leading to a "downhill spiral" of generalized abdominal distention, nausea, and malaise. Respiratory embarrassment develops secondarily. Unexplained bowel necrosis occurs after jejunal feeding, with a reported incidence of 1:1,000. This, too, might similarly follow vagal vascular reflexes in susceptible subjects [1]. Miedema, et al., at the University of Missouri, prospectively monitored jejunal pressure via the feeding catheter. Postoperative patients that developed "feeding intolerance" did demonstrate elevated feeding site pressure. However, the increased pressure developed too late to guide timely intervention [2]. Monitoring of volume, per se, apparently could be more useful. Why do we want a safer means to feed early and aggressively? Postoperative metabolism resembles that of a growing child. Net protein synthesis is permissible and achievable, but only if adequate nutrition is made available. Trauma produces a hypermetabolic state. Increased energy output usually becomes the determinant of protein anabolism, and protein catabolism has been the rule. However, even net protein synthesis is possible. The major factor limiting adequate enteral nutrition at these times is paralytic ileus, an avoidable G-I complication [3].

uces a hypermetabolic state. Increased energy output usually becomes the determinant of protein anabolism, and protein catabolism has been the rule. However, even net protein synthesis is possible. The major factor limiting adequate enteral nutrition at these times is paralytic ileus, an avoidable G-I complication [3]. No conventionally fed patient fully meets his increased nutritional needs for a variable period after major abdominal operation or comparable trauma. The potentially protective responses (e.g., heightened immune competence and accelerated wound healing) are blunted, limited by the patient's reduced gut function. Everyone's goal is to shorten this vulnerable, uncomfortable, and hospital dependent period. Applying the latest "fast-track" techniques has progressively reduced the duration and severity of postoperative G-I dysfunction. Patients begin eating, reach nutritional goals, and are now discharged sooner, but they still endure days of documented negative protein balances after major abdominal resective surgery. We describe a regimen for achieving the earliest resolution of paralytic ileus and positive protein balances, to date.

I dysfunction. Patients begin eating, reach nutritional goals, and are now discharged sooner, but they still endure days of documented negative protein balances after major abdominal resective surgery. We describe a regimen for achieving the earliest resolution of paralytic ileus and positive protein balances, to date. Materials and methods The subjects were the author's 31 colectomy and 160 consecutive cholecystectomy patients (1962 – 1988), prior to the introduction of laparoscopy into general surgery. All cholecystectomy and nine of the earliest colectomy patients had a Moss® naso-duodenal feeding-decompression device inserted (*Moss Tubes, Inc., West Sand Lake, NY). Subsequent bowel resection patients had the gastrostomy version of this feeding-decompression device placed. An elemental diet (Vivonex, Norwich-Eaton Pharmaceuticals, Norwich, NY) (now Nestle HealthCare Nutrition, Minnetonka, MN) was initiated at 100 – 300 kcal/hour within minutes of surgery. The duodenal aspirate was tested hourly (Clinatest® Tablets, Ames Co., Billerica, MA) for "reducing sugars," as a marker for the removed feeding solution. The "degassed" aspirate was filtered and returned manually via the feeding channel. Hourly nitrogen balances and serum amino acids were determined.

Materials and methods The subjects were the author's 31 colectomy and 160 consecutive cholecystectomy patients (1962 – 1988), prior to the introduction of laparoscopy into general surgery. All cholecystectomy and nine of the earliest colectomy patients had a Moss® naso-duodenal feeding-decompression device inserted (*Moss Tubes, Inc., West Sand Lake, NY). Subsequent bowel resection patients had the gastrostomy version of this feeding-decompression device placed. An elemental diet (Vivonex, Norwich-Eaton Pharmaceuticals, Norwich, NY) (now Nestle HealthCare Nutrition, Minnetonka, MN) was initiated at 100 – 300 kcal/hour within minutes of surgery. The duodenal aspirate was tested hourly (Clinatest® Tablets, Ames Co., Billerica, MA) for "reducing sugars," as a marker for the removed feeding solution. The "degassed" aspirate was filtered and returned manually via the feeding channel. Hourly nitrogen balances and serum amino acids were determined. The nasal tube was removed the morning following surgery. The cholecystectomy patients were discharged after a bowel movement and tolerating a general diet. The colectomy patients transitioned to oral elemental diet and had BaSO4 X-ray motility study over the next 24 hours. They resumed a general diet and met the same discharge criteria. The later colectomy patients, with placement of a gastrostomy version of the feeding-decompression device, had BaSO4 introduced with their feedings 4–6 hours following bowel resection and re-anastomosis. Motility was studied by X-ray for the duration of tube feeding.

The nasal tube was removed the morning following surgery. The cholecystectomy patients were discharged after a bowel movement and tolerating a general diet. The colectomy patients transitioned to oral elemental diet and had BaSO4 X-ray motility study over the next 24 hours. They resumed a general diet and met the same discharge criteria. The later colectomy patients, with placement of a gastrostomy version of the feeding-decompression device, had BaSO4 introduced with their feedings 4–6 hours following bowel resection and re-anastomosis. Motility was studied by X-ray for the duration of tube feeding. Results The aspirate became Clinatest® negative (reducing sugars absent) during the first or second postoperative hour. The total feedings were propelled forward thereafter by peristalsis. A maximum of 200 ml/day of refed gastroduodenal aspirate continued to reflux and was discarded. No patient developed abdominal distention or other signs of paralytic ileus or "feeding intolerance." All patients had achieved positive nitrogen balance within 2–26 hours postoperative (Table 1). Similarly, serum branched chain amino acids rose above basal levels by 2–12 hours following surgery. Table 1 Postoperative Protein Balances

Results The aspirate became Clinatest® negative (reducing sugars absent) during the first or second postoperative hour. The total feedings were propelled forward thereafter by peristalsis. A maximum of 200 ml/day of refed gastroduodenal aspirate continued to reflux and was discarded. No patient developed abdominal distention or other signs of paralytic ileus or "feeding intolerance." All patients had achieved positive nitrogen balance within 2–26 hours postoperative (Table 1). Similarly, serum branched chain amino acids rose above basal levels by 2–12 hours following surgery. Table 1 Postoperative Protein Balances Days Average Daily Protein Hours to Operation Age Fed Catabolized Fed Balance (+) Balance 1. Abdomino-perineal Resection 79 2 88 163 + 75 15 2. Abdomino-perineal Resection 83 4 93 149 + 56 4 3. Abdomino-perineal Resection 29 2 35 41 + 6 24 4. Abdomino-perineal Resection 17 2 101 118 + 17 8 5. Sigmoid Resection 60 2 104 122 + 18 26 6. Sigmoid Resection 69 2 78 100 + 22 7 7. Sigmoid Resection 76 2 42 77 + 35 6 8. Sigmoid Resection 55 2 43 133 + 90 22 9. Sigmoid & Bladder Resection 45 2 85 86 + 1 18 10. Vagotomy 41 1 88 100 + 17 7 11. Vagotomy 64 2 104 154 + 50 6 12. Vagotomy & Pyloroplasty 70 4 68 150 + 82 24 13. Vagotomy & Pyloroplasty 45 2 84 100 + 16 18 14. Vagotomy & Pyloroplasty 60 2 85 136 + 51 6 15. Vagotomy & Pyloroplasty 57 2 106 189 + 83 6 16. V & P + Gastrectomy 38 2 31 85 + 54 20 17. V & P + Gastrectomy + G.B. 60 2 98 104 + 6 10 18. Hysterectomy 45 2 51 74 + 23 6 19. Cholecystectomy 49 2 65 72 + 7 24 20. Cholecystectomy 48 1 49 200 +151 2 21. Cholecystectomy 38 2 82 86 + 4 17 22. Ileostomy Revision 32 1 38 140 +102 7 23. Ovarian Cystectomy 56 2 45 99 + 54 23 24. Colostomy Closure 62 1 54 190 +136 4 (Data presented at the 1963 Surgical Forum – American College of Surgery [6])

6 19. Cholecystectomy 49 2 65 72 + 7 24 20. Cholecystectomy 48 1 49 200 +151 2 21. Cholecystectomy 38 2 82 86 + 4 17 22. Ileostomy Revision 32 1 38 140 +102 7 23. Ovarian Cystectomy 56 2 45 99 + 54 23 24. Colostomy Closure 62 1 54 190 +136 4 (Data presented at the 1963 Surgical Forum – American College of Surgery [6]) X-ray motility studies showed clinically normal peristalsis, with contrast traversing secure, patent, and functional anastomosis, to outline the rectum within 24 hours. Permission was obtained to publish the studies. (Figures 1, 2, 3, 4, 5 and 6). Figure 1 66 y.o. male had sigmoid resection for carcinoma @ 3:00 pm. He was fed immediately for 17 hours via a double lumen, nasoduodenal catheter @ 100 kcal/hr. At 8:00 am the tube was removed, and BaSO4 swallowed. Serial X-rays indicate normal motility. 12:00 noon – 4 hour motility study, 21 hours after surgery. Figure 2 Normal motility. Contrast has reached the transverse colon. The more distal anastomosis is outlined by ring of stainless steel staples. 8:00 pm – 12 hour motility, 29 hours after sigmoid resection. Figure 3 BaSO4 has traversed the secure and patent stapled anastomosis, to outline the rectum. The patient ate his Thanksgiving meal, and was discharged after a bowel movement within 48 hours of surgery. The X-ray shows a normal peristaltic pattern despite surgical trauma to the tissue. 8:00 am – 24 hour motility, 41 hours after sigmoid resection.

the secure and patent stapled anastomosis, to outline the rectum. The patient ate his Thanksgiving meal, and was discharged after a bowel movement within 48 hours of surgery. The X-ray shows a normal peristaltic pattern despite surgical trauma to the tissue. 8:00 am – 24 hour motility, 41 hours after sigmoid resection. Figure 4 54 y.o. woman had a right hemicolectomy for carcinoma @ 11:00 am. Immediately fed with double lumen, feeding-decompression G-tube @ 300 kcal/hr. BaSO4 instilled via the feeding channel at rate exceeding peristaltic outflow. Note that no refluxing BaSO4 escaped aspiration, to enter the stomach. 4:00 pm – 5 minute motility, 5 hours after surgery. Figure 5 Clinically normal motility. Contrast is in the distal small intestine. 8:00 pm – 4 hour motility, 9 hours after a right hemicolectomy. Figure 6 Normal motility. Dilute BaSO4 in the the colon and rectum. She tolerated a general diet, had a bowel movement, and was discharged within 24 hours of resective surgery. Note the laxative action of the excess feedings. 8:00 am – 16 hour motility, 21 hours after a right hemicolectomy. The median time to discharge decreased to 48 hours, with the final four colectomy patients home uneventfully 24 hours after bowel resection and re-anastomosis. The 160:160 consecutive "open" cholecystectomy patients were discharged within 24 hours of surgery.

Figure 6 Normal motility. Dilute BaSO4 in the the colon and rectum. She tolerated a general diet, had a bowel movement, and was discharged within 24 hours of resective surgery. Note the laxative action of the excess feedings. 8:00 am – 16 hour motility, 21 hours after a right hemicolectomy. The median time to discharge decreased to 48 hours, with the final four colectomy patients home uneventfully 24 hours after bowel resection and re-anastomosis. The 160:160 consecutive "open" cholecystectomy patients were discharged within 24 hours of surgery. Discussion Wangensteen studied the factors affecting gastrointestinal function and dysfunction. He reported 70 years ago that even the disruptive consequences of complete small bowel obstruction could be aborted if swallowed air was completely excluded [4]. He over-sewed the dog's terminal ileum, and simultaneously vented its cervical esophagus. His experimental subjects were adequately hydrated by "clysis," but were not otherwise nourished. They survived without intestinal distention for up to two months, ultimately dying of starvation. We developed a labor intensive feeding-decompression regimen that mimicked Wangensteen's animal model, which we utilized experimentally and clinically for immediately postoperative gastric feeding [5] (Figure 7). In 1963 we reported on our first 24 patients at the Surgical Forum of the annual American College of Surgeons meeting [6].

labor intensive feeding-decompression regimen that mimicked Wangensteen's animal model, which we utilized experimentally and clinically for immediately postoperative gastric feeding [5] (Figure 7). In 1963 we reported on our first 24 patients at the Surgical Forum of the annual American College of Surgeons meeting [6]. Figure 7 Progenitor Moss7 Nasal Tube (circa 1961) Reprinted with permission of the American College of Surgeons. Moss G. Nitrogen equilibrium in the early postoperative period. Surg Forum 1963; 14:67–69. The duration of clinically significant paralytic ileus was abbreviated to within two hours postoperatively. Positive protein balances similarly were achieved within hours of surgery. Nutrient absorption and X-ray motility studies documented the more rapid return of normal G-I function. To this day, that regimen allowed the earliest achievement of positive protein balances following colectomy or other trauma. The patient's esophagus was aspirated to remove all swallowed air. Undiluted tracheo-bronchial secretions also were intercepted, demanding continuous irrigation for patency of the suction channel. Unlike Wangensteen's animal model, we provided a separate gastric channel for feeding and manually return of the filtered aspirate. Fear of gastric dilatation further increased the nursing workload, as we were obligated to frequently "check for residual."

anding continuous irrigation for patency of the suction channel. Unlike Wangensteen's animal model, we provided a separate gastric channel for feeding and manually return of the filtered aspirate. Fear of gastric dilatation further increased the nursing workload, as we were obligated to frequently "check for residual." The process was labor intensive, messy, and seldom performed outside of the research environment. The need for herculean nursing attention precluded wider acceptance of the original regimen. None-the-less, our clinical efforts were rewarded with preserved immediately postoperative G-I function. Kehlet's Danish team reports on "fast-track" laparoscopic colectomy [7,8]. They conclude that application of advanced techniques ".... decreased the duration of ileus after colonic surgery to about 2 days, as compared with the usual 3 to 5 days." In Germany, Schwenk, et al., found the time to both bowel movement and tolerance of oral feeding after laparoscopic colo-rectal resection is shortened significantly, but still approximates three days [9]. Hammarqvist, et al., report patients undergoing laparoscopic cholecystectomy develop days of negative protein balance and loss of muscle mass, despite receiving total parenteral nutrition [10].

ce of oral feeding after laparoscopic colo-rectal resection is shortened significantly, but still approximates three days [9]. Hammarqvist, et al., report patients undergoing laparoscopic cholecystectomy develop days of negative protein balance and loss of muscle mass, despite receiving total parenteral nutrition [10]. The major shortcomings of our earliest catheters had been frequent clogging of the esophageal aspiration lumen. As modified since 1985, aspiration was extended distally into the proximal duodenum [11] (Figure 8). Air and liquid were intercepted and removed efficiently within the close confines at this location, confirmed radiographically. A separate duodenal channel fed elemental diet (plus the returned, "degassed" aspirate) just distal to the aspiration orifices. Figure 8 Current Moss Gastrostomy Tube (circa 1985). We reported that decompression efficiency with this approach is more than twelvefold greater than gastric aspiration alone by conventional nasal or gastrostomy tube [12]. Suction applied to the stomach can be expected only to prevent gastric dilatation. The post-pyloric aspiration orifices remove air and digestive secretions that escape proximal removal, efficiently decompressing within this narrow segment of duodenum.

ric aspiration alone by conventional nasal or gastrostomy tube [12]. Suction applied to the stomach can be expected only to prevent gastric dilatation. The post-pyloric aspiration orifices remove air and digestive secretions that escape proximal removal, efficiently decompressing within this narrow segment of duodenum. There is scant duodenal resistance to retrograde flow of excess feedings that outstrip peristaltic outflow. These reflux and also are removed by efficient suction within these close confines, apparently before vagal reflexes can be induced. No liquid or gas can traverse this narrow segment in either direction without being intercepted (Figure 4). Increased volumes of aspirated digestive juices are removed to dilute the swallowed phlegm, and also to remove excess feeding. This "trade-off" to reduce the nursing workload substitutes the equivalent of a high output duodenal fistula, aspiration of 3,000 – 4,000 ml per day of gastric and pancreatic juices, saliva, bile, and succus. Without "refeeding," the patient requires additional intravenous infusions to replace this fluid, as well as laboratory studies to guide restoration of electrolyte balance.

he equivalent of a high output duodenal fistula, aspiration of 3,000 – 4,000 ml per day of gastric and pancreatic juices, saliva, bile, and succus. Without "refeeding," the patient requires additional intravenous infusions to replace this fluid, as well as laboratory studies to guide restoration of electrolyte balance. Aspirate became devoid of elemental diet within two hours of surgery, using Clinatest® tablets as a check for the contained carbohydrate. In my personal experience with hundreds of surgical patients immediately fed @ 100–200 kcal/hour, elemental diet spontaneously propelled by peristalsis beyond the ligament of Treitz continues prograde, where it is totally absorbed. We accidentally (and later deliberately) fed many patients at even higher rates. They suffered no ill effects when their capacity to absorb was exceeded. With air excluded, the unabsorbed elemental diet serves as a mild, self-limited cathartic (Figure 6) (Laxatives currently are an accepted component of "fast-track" regimens.). None of our patients manifested "feeding intolerance," with increasing distention, nausea, etc.

ects when their capacity to absorb was exceeded. With air excluded, the unabsorbed elemental diet serves as a mild, self-limited cathartic (Figure 6) (Laxatives currently are an accepted component of "fast-track" regimens.). None of our patients manifested "feeding intolerance," with increasing distention, nausea, etc. Absorption of a high carbohydrate load is analogous to superimposing a "glucose tolerance test" upon surgical stress, further increasing the patient's insulin requirement. Otherwise healthy young cholecystectomy patients (< 24 years of age) respond by spontaneously raising their serum insulin levels as much as ten-fold above basal [13]. Diabetics and older (> 35 years of age) patients require insulin supplementation for optimum metabolic balance to avoid hyperglycemia during the initial 24 hours postoperatively. The patient reverts to his preoperative insulin requirement thereafter. Close plasma glucose control for seriously stressed patients is becoming a common practice, and is advised with postoperative "enteral hyperalimentation." Clinically normal G-I motility and absorption resumed within hours for our 31 "open" colectomy patients, confirmed by the contrast X-ray studies. They consistently entered positive protein balance after institution of full feeding [14]. Contrast traversed secure and functional anastomosis, to outline the rectum and exit spontaneously in a bowel movement. Their usual course was discharge after 24–48 hours of "open" bowel resection and re-anastomosis [15].

tudies. They consistently entered positive protein balance after institution of full feeding [14]. Contrast traversed secure and functional anastomosis, to outline the rectum and exit spontaneously in a bowel movement. Their usual course was discharge after 24–48 hours of "open" bowel resection and re-anastomosis [15]. For our series of 160 consecutive "open" cholecystectomy patients before laparoscopy was introduced into general surgery, discharge uniformly was within 24 hours [16,17]. Empire (New York State) Blue Cross, jointly with the Visiting Nurse Association of Albany, prospectively studied 19 consecutive patients. They were interviewed preoperatively, the day of surgery, the day after discharge, and again 30 days later. All had continued adequate food intake and fully participated in "activities of normal living" [18]. Our digestive secretions contain large quantities of secretory globulins that are specific for our enteric organisms. Refeeding these secretions, and/or immediately heightened antibody synthesis, may provide sepsis protection. Our single infection (in an acute cholecystitis patient) following 160 consecutive operations, is an apparent 90% reduction in sepsis that supports this hypothesis. Refeeding digestive secretions and immediate postoperative feeding may mimic prophylactic antibiotics, at lower financial and health costs. This requires further clinical study.

Our digestive secretions contain large quantities of secretory globulins that are specific for our enteric organisms. Refeeding these secretions, and/or immediately heightened antibody synthesis, may provide sepsis protection. Our single infection (in an acute cholecystitis patient) following 160 consecutive operations, is an apparent 90% reduction in sepsis that supports this hypothesis. Refeeding digestive secretions and immediate postoperative feeding may mimic prophylactic antibiotics, at lower financial and health costs. This requires further clinical study. We consistently noted that elevation of serum amino acids within two hours accompanied positive protein balance and net protein synthesis [19]. Serum branched chain amino acids rose immediately, rather than falling below basal for several days. This was confirmed by a surgical team in Cleveland in 34 consecutive "open" cholecystectomy patients fed with our regimen. Branched chain amino acids, essential for protein preservation and synthesis, rose above basal within one hour, and was sustained for the duration of feeding [20]. Experimental wounds reportedly benefit from this procedure to safely initiate early nutrition. We applied this regimen to dogs subjected to bowel resection. By 72 hours postoperatively for fed versus control beagles, wound DNA synthesis was increased six-fold.

We consistently noted that elevation of serum amino acids within two hours accompanied positive protein balance and net protein synthesis [19]. Serum branched chain amino acids rose immediately, rather than falling below basal for several days. This was confirmed by a surgical team in Cleveland in 34 consecutive "open" cholecystectomy patients fed with our regimen. Branched chain amino acids, essential for protein preservation and synthesis, rose above basal within one hour, and was sustained for the duration of feeding [20]. Experimental wounds reportedly benefit from this procedure to safely initiate early nutrition. We applied this regimen to dogs subjected to bowel resection. By 72 hours postoperatively for fed versus control beagles, wound DNA synthesis was increased six-fold. By 96 hours, large bowel, intestinal, and abdominal wall wounds had bursting strengths doubled, tripled, and quadrupled, respectively. The stronger wounds showed accelerated collagen synthesis, i.e., procollagen content had doubled. Mature wound collagen of control dogs had diminished 50%, while remaining unchanged for fed subjects [21]. A study of colectomy patients using our feeding-decompression regimen in New Zealand confirmed early increased wound procollagen in their experimental, simultaneous forearm wounds [22].

llagen content had doubled. Mature wound collagen of control dogs had diminished 50%, while remaining unchanged for fed subjects [21]. A study of colectomy patients using our feeding-decompression regimen in New Zealand confirmed early increased wound procollagen in their experimental, simultaneous forearm wounds [22]. Immune protein synthesis shared in the accelerated protein synthesis associated with immediate "enteral hyperalimentation." Production of immune globulins doubled experimentally during the initial 24 hours [23]. Our animal findings were reproduced clinically in a controlled study of radical urological surgery at Roswell Park (Buffalo, NY) [24,25]. Their immediately fed patients achieved statistically significant higher levels for both plasma amino acids and immune protein (fibronectin) than unfed controls. With this regimen, secretory globulins in digestive juices that target enteric organisms are neither removed nor sequestered, but carried to the patients' lower G-I tract soon after surgery. These may provide protection against nosocomial infection during the early postoperative period, when the patient is most vulnerable to infection by his own gut organisms. The traumatized and vulnerable patient pays a hidden "price" for delayed or suboptimal enteral nutrition. Härtl, et al., reported a prospective study of 797 severe TBI (traumatic brain injury) patients treated at 22 New York State trauma centers from 2000–2006 [26].

With this regimen, secretory globulins in digestive juices that target enteric organisms are neither removed nor sequestered, but carried to the patients' lower G-I tract soon after surgery. These may provide protection against nosocomial infection during the early postoperative period, when the patient is most vulnerable to infection by his own gut organisms. The traumatized and vulnerable patient pays a hidden "price" for delayed or suboptimal enteral nutrition. Härtl, et al., reported a prospective study of 797 severe TBI (traumatic brain injury) patients treated at 22 New York State trauma centers from 2000–2006 [26]. "Patients who were not fed within 5 and 7 days after TBI had a 2- and 4-fold increased likelihood of death, respectively. The amount of nutrition in the first 5 days was related to death; every 10-kcal/kg decrease in caloric intake was associated with a 30–40% increase in mortality rates." Jeschke, et al., showed that forced feeding lessened the hypermetabolic immune consequences in their scalded rats [27]. Unlike humans, rats have resilient G-I tracts that function well despite such insult. However, humans have a parallel immune response to hypermetabolic stresses and tolerable nutrition changes.

"Patients who were not fed within 5 and 7 days after TBI had a 2- and 4-fold increased likelihood of death, respectively. The amount of nutrition in the first 5 days was related to death; every 10-kcal/kg decrease in caloric intake was associated with a 30–40% increase in mortality rates." Jeschke, et al., showed that forced feeding lessened the hypermetabolic immune consequences in their scalded rats [27]. Unlike humans, rats have resilient G-I tracts that function well despite such insult. However, humans have a parallel immune response to hypermetabolic stresses and tolerable nutrition changes. For our regimen, the feeding solution used must be of low viscosity, and remain completely soluble after contact with acidic or alkaline juices (e.g., casein denatures to becomes "cheese"). Amino acids substitute nutritionally for protein in "elemental diets." This is the only type of feeding solution that we have found consistently suitable for this regimen, totally avoiding potentially obstructing whole protein and fiber for the initial 24 hours [28]. We recently reported a regimen to automate outflow monitoring from the feeding site and "refeed" the aspirate after esophagectomy. The results were identical to our previous manual experience with postoperative immediate feeding, but with a reduced nursing workload [29].

For our regimen, the feeding solution used must be of low viscosity, and remain completely soluble after contact with acidic or alkaline juices (e.g., casein denatures to becomes "cheese"). Amino acids substitute nutritionally for protein in "elemental diets." This is the only type of feeding solution that we have found consistently suitable for this regimen, totally avoiding potentially obstructing whole protein and fiber for the initial 24 hours [28]. We recently reported a regimen to automate outflow monitoring from the feeding site and "refeed" the aspirate after esophagectomy. The results were identical to our previous manual experience with postoperative immediate feeding, but with a reduced nursing workload [29]. "Feeding intolerance" becomes evident only after gut recovery already has been aborted. Because "clinical judgment" alone is unreliable and "overfeeding" so detrimental, clinicians at Memorial – Sloan Kettering Cancer Center (New York, NY) [30] and the University of Ottawa [31] concluded that attempts to enterally feed immediately postoperatively were not warranted. The New York team prospectively studied 100 pairs of matched cancer patients after resective abdominal cancer surgery. Half received closely supervised early jejunal feeding. The fed patients recovered more poorly, including one that developed bowel necrosis. They concluded that immediate enteral feeding after major resective surgery does not warrant the risks.

pairs of matched cancer patients after resective abdominal cancer surgery. Half received closely supervised early jejunal feeding. The fed patients recovered more poorly, including one that developed bowel necrosis. They concluded that immediate enteral feeding after major resective surgery does not warrant the risks. The Ottawa team also found that their early postoperative enteral feeding impaired recovery. They monitored pre- and postoperative mobility and pulmonary mechanics, to gain insight into why feeding proved detrimental. All their postoperative patients initially were bedridden. However, their fed patients were more severely affected, exhibiting poorer respiratory function, increased abdominal distention, and greater immobility. This feeding-decompression regimen safely permits the earliest enteral feeding, despite initially impaired G-I function. More rapid return and utilization of gut function appears to reduce sepsis, mortality, discomfort, and hospital length of stay. Determining optimum nutrition in the face of the heightened metabolic demands of trauma requires further controlled study, but the "feeding intolerance" of paralytic ileus safely can be removed as a confounding factor. Conclusion The symptom complex of "feeding intolerance" most likely is triggered by localized distention at the enteral feeding site. It can be prevented by removing any excess liquid and all swallowed air from this location. A feeding-decompression regimen is reported that: 1. aspirates digestive juices and air from within the stomach and proximal duodenum;

Conclusion The symptom complex of "feeding intolerance" most likely is triggered by localized distention at the enteral feeding site. It can be prevented by removing any excess liquid and all swallowed air from this location. A feeding-decompression regimen is reported that: 1. aspirates digestive juices and air from within the stomach and proximal duodenum; 2. delivers feeding solution into the slightly more distal duodenum; 3. removes excess feedings before localized distention can develop; 4. filters and returns the "degassed" aspirate via the feeding channel. Automation can reduce the mess and nursing work-load of salvaging the aspirated digestive juices. Competing interests The author declares that they have no competing interests. Disclaimer Relating to p.5: *For the past decade, Dr. Moss has had no financial affiliation with Moss Tubes, Inc., which manufactures Moss® medical devices and owns the trademark.

Introduction The rationale of systemic antibiotic prophylaxis is to reduce the incidence of surgical infections in the surgical site [1]. Antibiotic prophylaxis should be instituted only in clean-contaminated surgical procedures or in clean surgical procedures in which an infective complication could represent an occurrence of particular severity, or in clean procedures in which prosthetic implants or other exogenous materials are being used. Antibiotic prophylaxis is not indicated in clean surgical procedures where the potential risk of side effect, bacterial and/or mycotic superinfections and the emergence of bacterial resistant strains outweighs the possible advantages. Surgical procedures are classified into four types in correlation to the increasing risk of bacterial contamination and infection [2]: - clean surgical procedures (incidence of infections < 5%); - clean – contaminated surgical procedures (incidence of infections < 10%); - contaminated surgical procedures (incidence of infections about 20%); - dirty surgical procedures (incidence of infections about 40%).

Introduction The rationale of systemic antibiotic prophylaxis is to reduce the incidence of surgical infections in the surgical site [1]. Antibiotic prophylaxis should be instituted only in clean-contaminated surgical procedures or in clean surgical procedures in which an infective complication could represent an occurrence of particular severity, or in clean procedures in which prosthetic implants or other exogenous materials are being used. Antibiotic prophylaxis is not indicated in clean surgical procedures where the potential risk of side effect, bacterial and/or mycotic superinfections and the emergence of bacterial resistant strains outweighs the possible advantages. Surgical procedures are classified into four types in correlation to the increasing risk of bacterial contamination and infection [2]: - clean surgical procedures (incidence of infections < 5%); - clean – contaminated surgical procedures (incidence of infections < 10%); - contaminated surgical procedures (incidence of infections about 20%); - dirty surgical procedures (incidence of infections about 40%). Thyroidectomy is classified amongst clean surgical procedures, those where there is no intra-operatory bacterial contamination following surgery. Post-operatory wound infections are a very uncommon finding after thyroidectomy. For these reasons international guidelines do not routinely recommend systemic antibiotic prophylaxis [1]. Although many different guidelines of the National Health Service and of Surgical Societies include these recommendations, systemic antibiotic prophylaxis is nonetheless frequently used in thyroid surgery. According to the majority of surgeons, this conduct is justified with the potential risk of infections related to the positioning of drains. The benefits of this antibiotic prophylaxis are not supported by clinical evidence in the literature.

mic antibiotic prophylaxis is nonetheless frequently used in thyroid surgery. According to the majority of surgeons, this conduct is justified with the potential risk of infections related to the positioning of drains. The benefits of this antibiotic prophylaxis are not supported by clinical evidence in the literature. We have conducted a multicentric randomized double-blind study on 500 patients who had undergone thyroidectomy for goitre or thyroid carcinoma. The endpoint of this study was to evaluate the benefits of antibiotic prophylaxis vs. no prophylaxis in patients undergoing thyroid surgery. Materials Between January 2007 and June 2007 a multicentric randomised double blind study was conducted on 500 consecutive patients admitted to our clinical wards. Inclusion criteria were: - men or women between age 16 and 80; - absence of concomitant metabolic (diabetes), infective or hematologic pathologies; - absence of cardiac valvular pathologies; - patients not undergoing corticosteroid or immunosuppressive treatment; - patients in whom an aspirating drain had been positioned. Exclusion criteria were: - patients younger than 16 years and older than 80; - presence of concomitant metabolic (diabetes), infective and hematologic pathologies; - presence of severe obesity; - ongoing corticosteroid or immunosuppressive treatment; - presence of concomitant neoplasms; - patients undergoing thyroidectomy and lymphadenectomy; - patients undergoing secondary surgery in the cervical region; - patients undergoing thyroidectomy for locally advanced tumours;

- presence of concomitant metabolic (diabetes), infective and hematologic pathologies; - presence of severe obesity; - ongoing corticosteroid or immunosuppressive treatment; - presence of concomitant neoplasms; - patients undergoing thyroidectomy and lymphadenectomy; - patients undergoing secondary surgery in the cervical region; - patients undergoing thyroidectomy for locally advanced tumours; - patients undergoing thyroidectomy for autoimmune thyroiditis or Plummer adenoma; - patients with goitres submerged in the thorax; - patients without cervical drain. Eligible patients were proposed to enter the RCT study upon acquisition of adequate informed consent. All patients underwent a minimum ambulatory follow up of 30 days. Methods The 500 patients enrolled in this multicentric trial (mean age 47 years) were randomized in two subgroups of 250 patients. The sequence allocation generation has been formulated by a researcher of another Institution through random numbers generated by a computer; the sequence allocation concealment was actuated by means of an e-mail sent by the researcher right after enrolling the patient in the study. 250 patients were treated with standard antibiotic prophylaxis with Unasyn (sulbactam/ampicillin) 1 fl (3 gr.) 30 min before surgery. No antibiotic prophylaxis was instituted in the remaining 250 patients.

cealment was actuated by means of an e-mail sent by the researcher right after enrolling the patient in the study. 250 patients were treated with standard antibiotic prophylaxis with Unasyn (sulbactam/ampicillin) 1 fl (3 gr.) 30 min before surgery. No antibiotic prophylaxis was instituted in the remaining 250 patients. Antiseptic and aseptic measures were standardized: disinfection by means of Betadine or Hibitane (a disinfectant that does not contain iodine and that is utilized in thyroid neoplasms to avoid false results at the control scintigraphy) followed by the placement of a sterile adhesive transparent cervical film. Hemostasis was very accurate. Absorbable hemostatics (Tabotamp by Ethicon or Floseal by Baxter) were never utilized intentionally for greater homogeneity of the two subgroups. The reconstructive surgical technique was also standardized again with the intent of making the two subgroups as homogeneous as possible. Platisma reconstruction and dermal approximation was obtained with simple interrupted sutures; the skin wound was closed with metallic staples. Wounds were always dressed every third day. As a rule drains were removed on the first postoperative day; when more than 80 cc of serum was collected, the drains were left in and removed in the following days according to the clinical judgement. Within each group of patients, two subgroups were highlighted: the subgroup of patients in whom the drains were removed in the first postoperative day and the subgroup of patients in whom it was necessary to keep the drains for a longer period of time (> 1 day). The statistical analysis of the data was carried out using Review Manager 5 [3].

tients, two subgroups were highlighted: the subgroup of patients in whom the drains were removed in the first postoperative day and the subgroup of patients in whom it was necessary to keep the drains for a longer period of time (> 1 day). The statistical analysis of the data was carried out using Review Manager 5 [3]. Results Only 3 patients, two amongst those who had undergone prophylactic antibiotic treatment and one in the other group, developed an infection of the surgical wound. The infection receded after antibiotic treatment with amoxicillin/clavulanate 1 gr PO tid for 7 d. Slightly better results were obtained in the group no treated with prophylactic antibiotic treatment, although this was not statistically significant. (P = 0,57) (fig. 1). Eight patients (3 pertaining to the group of patients undergoing prophylactic antibiotic treatment and 5 to the group that did not undergo antibiotic prophylaxis) developed oedema and erythema of the surgical wound; there was no pain and hyperthermia together with these signs. No antibiotic treatment was required for their regression. The group with antibiotic prophylaxis showed better results, but this was not statistically significant. (P = 0,48) (fig. 2). Thirty-nine patients (16 pertaining to the group of patients undergoing prophylactic antibiotic treatment and 23 to the group that did not undergo antibiotic prophylaxis) developed oedema of the surgical wound; no pain and hyperthermia was evidenced along with these signs. No antibiotic treatment was required for these signs to recede. The group with antibiotic prophylaxis showed better results, but this was not statistically significant. (P = 0,25) (fig. 3). The comprehensive analysis of all the complications that were observed did not show statistically significant differences between the two groups (odds ratio (OR), 0.84; 95 percent confidence interval (CI), 0.39–1.26; P = 0,23) (fig. 4).

ed better results, but this was not statistically significant. (P = 0,25) (fig. 3). The comprehensive analysis of all the complications that were observed did not show statistically significant differences between the two groups (odds ratio (OR), 0.84; 95 percent confidence interval (CI), 0.39–1.26; P = 0,23) (fig. 4). Figure 1 Wound infection after thyroidectomy. Figure 2 Wound oedema and erytema after thyroidectomy. Figure 3 Wound oedema after thyroidectomy. Figure 4 Comprehensive analysis of all wound complications after thyroidectomy. We performed a subgroup analysis between patients in whom the drain was removed in the first post-operative day and those in whom it was removed in the following days. In 47 patients only it was necessary to keep the drain beyond the first post-operative day. Patients from whom the drain was removed in the first postoperative day showed a slightly inferior incidence of wound complications related to infective occurrence (lower incidence of oedema, erythema or infection of the surgical wound) (P = 0,002) (fig. 5). No significant advantage was evidenced with prophylactic antibiotic treatment in patients in whom the drain was removed on the 1st postoperative day (P = 0,27) (fig. 6). A significative advantage was not substantiated with prophylactic antibiotic treatment even in patients in whom the drain was kept after the 1st postoperative day (P = 0,45) (fig. 7).

enced with prophylactic antibiotic treatment in patients in whom the drain was removed on the 1st postoperative day (P = 0,27) (fig. 6). A significative advantage was not substantiated with prophylactic antibiotic treatment even in patients in whom the drain was kept after the 1st postoperative day (P = 0,45) (fig. 7). Figure 5 Comprehensive analysis of all wound complications after thyroidectomy: patients in whom the drain was removed in the first post-operative day vs those in whom it was removed in the following days. Figure 6 Comprehensive analysis of all wound complications after thyroidectomy in patients in whom the drain was removed in the first post-operative day: antibiotic prophylaxis vs no antibiotic prophylaxis. Figure 7 Comprehensive analysis of all wound complications after thyroidectomy in patients in whom the drain was removed after the first post-operative day: antibiotic prophylaxis vs no antibiotic prophylaxis.

Figure 6 Comprehensive analysis of all wound complications after thyroidectomy in patients in whom the drain was removed in the first post-operative day: antibiotic prophylaxis vs no antibiotic prophylaxis. Figure 7 Comprehensive analysis of all wound complications after thyroidectomy in patients in whom the drain was removed after the first post-operative day: antibiotic prophylaxis vs no antibiotic prophylaxis. Discussion Thyroid surgery is classified as a clean procedure and is associated with a low incidence of wound infections. (0,3%) [4]. Although the majority of international guidelines do not recommend the use of systemic prophylactic antibiotic treatment, this practice is adopted sporadically in some nations and routinely in others [5]. An audit carried out by the British Association of Endocrine Surgery (BAES) in England and Ireland showed that 9% of patients received routine antibiotic prophylaxis, 16% in selected cases and 75% did not receive it [6]. In an Italian retrospective study carried out by Rosato on 14.394 patients, it was evidenced that 50% of surgeons use antibiotic prophylaxis, 17% antibiotic therapy and 33% neither prophylaxis nor therapy [4]. The Scottish Intercollegiate Guidelines network (SIGN) does not advocate antibiotic prophylaxis for benign pathologies regarding it an opportunity to be reserved for selected cases of malignancies [7]. The BAES and the Royal College of Physicians Thyroid Cancer do not list antibiotic prophylaxis amongst the recommendations to follow for thyroidectomy [8,9]. Antibiotic prophylaxis does not ward off the development of infective complications of the surgical wound [4]. The occurrence of severe cervical infections after thyroid surgery is an extremely rare event with an extremely high inherent mortality rate; however the occurrence of these infections, often cellulites with fulminating Streptococcal sepsis cannot be avoided with antibiotic prophylaxis [6].

s of the surgical wound [4]. The occurrence of severe cervical infections after thyroid surgery is an extremely rare event with an extremely high inherent mortality rate; however the occurrence of these infections, often cellulites with fulminating Streptococcal sepsis cannot be avoided with antibiotic prophylaxis [6]. In common clinical practice it is believed that the use of drains after thyroidectomy represents a risk factor for the development of infective complications of the surgical wound. This notion is not supported by clinical data, whereas conflicting evidence was presented recently in a systematic review by Samraj e Gurusamy [10]. This review of the literature has evidenced 5 RCT in which 337 patients with drains were compared with 350 patients without drains following thyroidectomy; the statistical analysis failed to evidence that patients with drains show a significant incidence of infections of the surgical wound (P = 0,41). Therefore antibiotic prophylaxis does not seem to be beneficial in all patients with drains [11].

ins were compared with 350 patients without drains following thyroidectomy; the statistical analysis failed to evidence that patients with drains show a significant incidence of infections of the surgical wound (P = 0,41). Therefore antibiotic prophylaxis does not seem to be beneficial in all patients with drains [11]. Conclusion Our RCT showed that prophylactic antibiotic treatment is not beneficial in patients younger than eighty years old, with no concomitant metabolic, infective or hematologic disease, with no cardiac valvulopathies, not receiving steroidal or immunosuppressive treatment, and not severely obese. Thyroidectomies with associated lymphadenectomy, secondary surgery, excision of locally advanced disease, or of autoimmune thyroiditis and Plummer adenoma were excluded from this analysis. Drains did not represent a risk factor for the development of infections of the surgical wound even in the group of patients not undergoing antibiotic prophylaxis. Our study should be regarded only as a preliminary RCT, and should be followed by a study in which a larger number of patients should be enrolled so that statistically significant data can be obtained. Competing interests The Authors state that none of the authors involved in the manuscript preparation has any conflicts of interest towards the manuscript itself, neither financial nor moral conflicts. Besides none of the authors received support in the form of grants, equipment, and/or pharmaceutical items.

Conclusion Our RCT showed that prophylactic antibiotic treatment is not beneficial in patients younger than eighty years old, with no concomitant metabolic, infective or hematologic disease, with no cardiac valvulopathies, not receiving steroidal or immunosuppressive treatment, and not severely obese. Thyroidectomies with associated lymphadenectomy, secondary surgery, excision of locally advanced disease, or of autoimmune thyroiditis and Plummer adenoma were excluded from this analysis. Drains did not represent a risk factor for the development of infections of the surgical wound even in the group of patients not undergoing antibiotic prophylaxis. Our study should be regarded only as a preliminary RCT, and should be followed by a study in which a larger number of patients should be enrolled so that statistically significant data can be obtained. Competing interests The Authors state that none of the authors involved in the manuscript preparation has any conflicts of interest towards the manuscript itself, neither financial nor moral conflicts. Besides none of the authors received support in the form of grants, equipment, and/or pharmaceutical items. Authors' contributions All authors contributed equally to this work. RC participated in the design of the study and performed the statistical analysis. All authors read and approved the final manuscript.

Background Repair of abdominal wall hernia represents the most common group of operatios performed by general surgeons all around the world. Incisional hernia is a serious complication after abdominal surgery which occurs in 11-23% of laparotomies [1]. In 2003 it was estimated that over 100,000 ventral incisional hernia repairs were performed in the US. Risk factors for incisional hernia formation and preventive strategies are not clearly defined, but according to data from literature, significant demographic factors influencing incisional hernia incidence are age (> 45 years) and male gender. Preoperative anaemia (Hb < 100 g/l) and BMI > 25, associated with previous laparotomies and postoperative catecholamin-therapy also seem to play an important role [2]. The tension-free repair is one of the key concepts in hernia surgery. The use of a mesh prosthesis decreases the recurrence rates, particularly for inguinal and incisional hernias. Recently, the laparoscopic approach extended the options and approaches for repairing the fascial defect.

em to play an important role [2]. The tension-free repair is one of the key concepts in hernia surgery. The use of a mesh prosthesis decreases the recurrence rates, particularly for inguinal and incisional hernias. Recently, the laparoscopic approach extended the options and approaches for repairing the fascial defect. The emergency treatment of incisional hernias is not frequent [3], and its technical approach can be different from the elective one, both for the septic conditions in which the emergency treatment is usually performed and for the patients' age which can lead to several technical difficulties. As an emergency, it often occurs in elderly patients with voluminous hernias complicated with strangulation or obstruction [4]. In some of these cases the surgical approach may also include an intestinal resection, with the possibility of peritoneal contamination [5]. This study aims at reporting our experience in the emergency treatment of complicated incisional hernias, analysing the results obtained with the employment of synthetic prosthesis versus the open surgical repair. Materials and methods We performed a clinical study by revising clinical notes, through which we evaluated the different treatments of patients with complicated incisional hernias. All the patients whose hernia were only an attendant pathology and did not represent itself the cause for an emergency surgical treatment, were excluded fom the trial.

formed a clinical study by revising clinical notes, through which we evaluated the different treatments of patients with complicated incisional hernias. All the patients whose hernia were only an attendant pathology and did not represent itself the cause for an emergency surgical treatment, were excluded fom the trial. Since January 2001 till July 2008, 89 patients (55 males and 34 females) were treated for complicated incisional hernias in emergency. We divided the patients in two different groups. The patients treated with prosthesis apposition (Group I) were 33 (24 males and 9 females) while 56 patients (Group II) (31 males and 25 females) were treated by performing a direct abdominal wall muscles suture. Five of these patients (8,9%) had such voluminous incisional hernias that they could not be treated by carrying out a direct abdominal wall muscles suture (Table 1). Table 1 Emergency treatment of complicated incisional hernias Prosthesis apposition (Group I) Direct abdominal wall muscles suture (Group II) Patients treated 33 56 Omentum resections 14 41

Since January 2001 till July 2008, 89 patients (55 males and 34 females) were treated for complicated incisional hernias in emergency. We divided the patients in two different groups. The patients treated with prosthesis apposition (Group I) were 33 (24 males and 9 females) while 56 patients (Group II) (31 males and 25 females) were treated by performing a direct abdominal wall muscles suture. Five of these patients (8,9%) had such voluminous incisional hernias that they could not be treated by carrying out a direct abdominal wall muscles suture (Table 1). Table 1 Emergency treatment of complicated incisional hernias Prosthesis apposition (Group I) Direct abdominal wall muscles suture (Group II) Patients treated 33 56 Omentum resections 14 41 Intestinal resections 6 15 Results As concerns the group treated with the apposition of prosthesis (Group I), obstruction with no possibility of reduction occurred in 27 cases; in these cases we performed an adhesiolisis and a greater omentum resection (14 patients) which eased the abdominal replacement of the ileal ansae, being the omentum often inflamed, thickened and fibrous. In the remaining 6 cases, strangulation was the reason for an emergency treatment; in these cases we performed a resection of the necrotic ileum and the intestinal continuity was restored by carrying out a mechanic suture. Strangulation occurred in those patients whose incisional hernias had a narrow neck. We registered 7 complications; 4 parcellar cutaneous necrosis, 2 hematomas and 1 wound suppuration (Table 2). Table 2 Complications

Intestinal resections 6 15 Results As concerns the group treated with the apposition of prosthesis (Group I), obstruction with no possibility of reduction occurred in 27 cases; in these cases we performed an adhesiolisis and a greater omentum resection (14 patients) which eased the abdominal replacement of the ileal ansae, being the omentum often inflamed, thickened and fibrous. In the remaining 6 cases, strangulation was the reason for an emergency treatment; in these cases we performed a resection of the necrotic ileum and the intestinal continuity was restored by carrying out a mechanic suture. Strangulation occurred in those patients whose incisional hernias had a narrow neck. We registered 7 complications; 4 parcellar cutaneous necrosis, 2 hematomas and 1 wound suppuration (Table 2). Table 2 Complications Prosthesis apposition (Group I) Direct abdominal wall muscles suture (Group II) Total complications 7 (21%) 24 (43%) Parcellar cutaneous necrosis 4 (57%) 7 (29%) Hematomas 2 (29%) 6 (25%)

Intestinal resections 6 15 Results As concerns the group treated with the apposition of prosthesis (Group I), obstruction with no possibility of reduction occurred in 27 cases; in these cases we performed an adhesiolisis and a greater omentum resection (14 patients) which eased the abdominal replacement of the ileal ansae, being the omentum often inflamed, thickened and fibrous. In the remaining 6 cases, strangulation was the reason for an emergency treatment; in these cases we performed a resection of the necrotic ileum and the intestinal continuity was restored by carrying out a mechanic suture. Strangulation occurred in those patients whose incisional hernias had a narrow neck. We registered 7 complications; 4 parcellar cutaneous necrosis, 2 hematomas and 1 wound suppuration (Table 2). Table 2 Complications Prosthesis apposition (Group I) Direct abdominal wall muscles suture (Group II) Total complications 7 (21%) 24 (43%) Parcellar cutaneous necrosis 4 (57%) 7 (29%) Hematomas 2 (29%) 6 (25%) Wound suppuration 1 (14%) 11 (46%) The surgical technique performed for Group I patients, was carried out through the apposition of polypropylene prosthesis which was placed between the posterior rectus wall and the anterior wall of the rectus sheath. The prosthesis was fixed by non-absorbable interrupted stitches which were sequentially placed through the aponeurosis, the prosthetis, back into the prosthetis and finally again through the aponeurosis, approximately 0.5 cm by the entrance point. The suture must be performed at least 2 cm far from the hernia edge, in a totally sane tissue. On the contrary the patients with voluminous incisional hernias were treated with a PTFE mesh (polytetrafluoroethylene) which was placed intraperitoneally, in contact with the abdominal viscera. The employment of PTFE allows the reduction of viscero-parietal adhesions and the constitution of a stronger abdominal wall.

On the contrary the patients with voluminous incisional hernias were treated with a PTFE mesh (polytetrafluoroethylene) which was placed intraperitoneally, in contact with the abdominal viscera. The employment of PTFE allows the reduction of viscero-parietal adhesions and the constitution of a stronger abdominal wall. In the patients treated with a direct abdominal wall muscles suture (Group II) we carried out an adhesiolysis and a greater omentum resection in 41 patients. In 15 cases we performed a resection of the necrotic ileum followed by a mechanic suture. We registered 24 complications; 7 parcellar cutaneous necrosis, 6 hematomas and 11 wound suppurations (Table 2). All the patients underwent a 6-month follow up; we noticed 9 recurrences (9/56, 16%) in the patients treated with direct abdominal wall muscles suture and 1 recurrence (1/33, 3%) in the group of patients treated with the prostesis apposition (Table 3). Table 3 Recurrences

In the patients treated with a direct abdominal wall muscles suture (Group II) we carried out an adhesiolysis and a greater omentum resection in 41 patients. In 15 cases we performed a resection of the necrotic ileum followed by a mechanic suture. We registered 24 complications; 7 parcellar cutaneous necrosis, 6 hematomas and 11 wound suppurations (Table 2). All the patients underwent a 6-month follow up; we noticed 9 recurrences (9/56, 16%) in the patients treated with direct abdominal wall muscles suture and 1 recurrence (1/33, 3%) in the group of patients treated with the prostesis apposition (Table 3). Table 3 Recurrences Prosthesis apposition (Group I) Direct abdominal wall muscles suture (Group II) Recurrences 1 (3%) 9 (16%) Discussion Incisional hernias can range in size from very small to large and complex ones and appear as a bulge by the area of a previous surgical scar. Nearly any prior abdominal operation can develop an incisional hernia, however the most frequent site is along incisions running down from the breastbone to the pubic area. These hernias may occur after large surgeries such as intestinal or vascular surgery, but also after an appendectomy or even through the small scar of a laparoscopy wound. Surgical repair of incisional hernias is usually recommended, as they can become a medical or surgical emergency. An incisional hernia can be defined as complicated when the involved structures undergo worsening conditions. Particularly, the concerned structures may be described as follows:

Prosthesis apposition (Group I) Direct abdominal wall muscles suture (Group II) Recurrences 1 (3%) 9 (16%) Discussion Incisional hernias can range in size from very small to large and complex ones and appear as a bulge by the area of a previous surgical scar. Nearly any prior abdominal operation can develop an incisional hernia, however the most frequent site is along incisions running down from the breastbone to the pubic area. These hernias may occur after large surgeries such as intestinal or vascular surgery, but also after an appendectomy or even through the small scar of a laparoscopy wound. Surgical repair of incisional hernias is usually recommended, as they can become a medical or surgical emergency. An incisional hernia can be defined as complicated when the involved structures undergo worsening conditions. Particularly, the concerned structures may be described as follows: 1) Cutaneous: large and thin scars, cutaneous atrophy and eczemas, suppurative flogosis, fistulae. 2) Hernial sac: multiple sacs, fibrous septa, sac thickening and adhesions. 3) Visceral: chronic incarceration, obstruction, strangulation, ileum and colonic torsion with progressive damage up to gangrene; greater omentum involvement, mesenteritis and perivisceritis. The frequency of complicated incisional hernias varies from 10 to 40%. The most frequent complications are incarceration, obstruction and strangulation [6].

3) Visceral: chronic incarceration, obstruction, strangulation, ileum and colonic torsion with progressive damage up to gangrene; greater omentum involvement, mesenteritis and perivisceritis. The frequency of complicated incisional hernias varies from 10 to 40%. The most frequent complications are incarceration, obstruction and strangulation [6]. There are two main factors for the pathogeneses of these complications: the hernial orifice rigidity and the presence of tenacious adhesions between the hernial sac and its content or between the sac and the surrounding tissues. The formation of viscero-visceral and viscero-parietal bridles is the necessary condition for the production of strangulation: the intestinal loop contained into the sac is firstly affected by a transit alteration and later on by circulatory disturbance.

There are two main factors for the pathogeneses of these complications: the hernial orifice rigidity and the presence of tenacious adhesions between the hernial sac and its content or between the sac and the surrounding tissues. The formation of viscero-visceral and viscero-parietal bridles is the necessary condition for the production of strangulation: the intestinal loop contained into the sac is firstly affected by a transit alteration and later on by circulatory disturbance. The strangulated intestinal tract rapidly goes towards congestion, edema and turgor caused by a disturbance of venous circulation which is followed by the formation of trasudate into the intestinal loop entrapped into the hernial sac; the intestinal wall goes towards progressive modifications up to necrosis and perforation. The omentum can be involved in the strangulation process; in such case the affected part adheres to the hernial sac and turns into a fibrous tissue. In case of strangulation, the symptoms will be those of a typical intestinal occlusion or subocclusion, depending on the elapsed time and the single material cause, although the synthomatology can sometimes be hard to define; for instance, in obese abdomens it is not easy to recognize the strangulation of small laparoceles.

se of strangulation, the symptoms will be those of a typical intestinal occlusion or subocclusion, depending on the elapsed time and the single material cause, although the synthomatology can sometimes be hard to define; for instance, in obese abdomens it is not easy to recognize the strangulation of small laparoceles. The emergency surgical treatment for complicated incisional hernias, besides the problems given by the lack of intestinal preparation, shows a higher morbidity also due to the development of an acute respiratory failure; this is caused by the abdominal reduction of the herniated viscera which causes an increase of endoabdominal pressure pushing up the diaphragm [7]. Since the newly formed subcutaneous cavity resulting from the hernia reduction can be origin of haematic collections formation associated with a frequent necrosis of the cutaneous rims, we suggest to perform a cutaneous and subcutaneous resection in order to avoid both complications. Furthermore in all the patients we normally place one or two subcutaneous Jackson-Pratt suction drainages. At the moment the most frequent treatment of voluminous incisional hernias is performed through the use of synthetic prosthesis, which allows the abdominal wall reconstruction according to the "tension free" technique.

Since the newly formed subcutaneous cavity resulting from the hernia reduction can be origin of haematic collections formation associated with a frequent necrosis of the cutaneous rims, we suggest to perform a cutaneous and subcutaneous resection in order to avoid both complications. Furthermore in all the patients we normally place one or two subcutaneous Jackson-Pratt suction drainages. At the moment the most frequent treatment of voluminous incisional hernias is performed through the use of synthetic prosthesis, which allows the abdominal wall reconstruction according to the "tension free" technique. The prosthesis is to be placed between the peritoneum and the posterior surface of the rectus abdominis muscles [8,9] or between the posterior surface of the rectus abdominis muscles and their posterior sheath [10]. In both cases the prosthesis apposition allows a strength discharge upon the abdominal wall circumference; moreover, the overlap of the muscle to the prosthesis (properitoneal technique) allows a wider distribution surface of tension strengths. In order to avoid a prosthesis dislocation, it is necessary to fix it to the abdominal wall by non-absorbable interrupted stitches (prolene). The prosthesis must be considerably wider than the parietal breach, so that the endoabdominal pressure might ease its adhesion to the abdominal wall. The anterior rectus muscles sheath is closed by a continuous non-absorbable suture. When it is not possible to suture the abdominal wall because of a massive tissutal loss, we place a PTFE prosthesis in contact with the viscera and we suture it to the muscle sheaths [11,12]. PTFE avoids visceral adhesions, assuring this material to be used in properitoneal locations for voluminous incisional hernias when it is not possible to perform a direct peritoneal suture; it also lowers the infections incidence and causes only a weak foreign body reaction [13-19] (Appendix 1).

muscle sheaths [11,12]. PTFE avoids visceral adhesions, assuring this material to be used in properitoneal locations for voluminous incisional hernias when it is not possible to perform a direct peritoneal suture; it also lowers the infections incidence and causes only a weak foreign body reaction [13-19] (Appendix 1). The described techniques performed in emergency do not show a higher incidence of complications (fistulisation, hematomas and wound dehiscence) compared to elective surgery [20,21]; also according to our experience, the emergency treatment of complicated incisional hernias appears to be feasible, both in terms of post-operative complications and recurrences. Conclusions According to our experience, the emergency treatment of complicated incisional hernias through prosthesis apposition is always feasible and ensures less post-operative complications (21% vs 43%) and recurrences (3% vs 16%) compared to the patients treated with direct muscular suture. The technical approach in emergency is correct when considering the clinical conditions of the patient, the size of the hernia and each eventual complication; such a dealing will minimize the complication rate and make it comparable to the one achieved with elective practice. Furthermore, synthetic prosthesis allows defects of any size to be repaired without tension and with a lower recurrence rate, providing an added support to the weak abdominal wall. Competing interests The authors declare that they have no competing interests.

The technical approach in emergency is correct when considering the clinical conditions of the patient, the size of the hernia and each eventual complication; such a dealing will minimize the complication rate and make it comparable to the one achieved with elective practice. Furthermore, synthetic prosthesis allows defects of any size to be repaired without tension and with a lower recurrence rate, providing an added support to the weak abdominal wall. Competing interests The authors declare that they have no competing interests. Consent section All the patients were informed that their clinical history would be used for a study. All of them signed a personal data treatment consent. No image was used. Authors' contributions FL cooperated in writing the article and translated it into English RC drafted the article EF checked the numbers and percentages UM updated the references VN made the tables LC formatted the article AS searched for the references BR collected patients' data PD chose the most useful and interesting articles in literature about the field CM searched for the references DM collected the patients' consent PC gave some language suggestions CB formatted the references GN supervised the article production FS allowed the collection of the patients' data and supervised the whole work making All the authors read and approved the final version of the manuscript Appendix 1 Features of PTFE prosthesis: Low infections incidence Absorbability by connective tissues Low incidence of adhesions Weak foreign body reaction

Background Surgical resection of colorectal liver metastases offers the best possibility for long-term survival. After complete resection of detectable metastases, about one third of patients survive for at least 5 years [1,2]. The surgical strategy for the treatment of colorectal cancer and synchronous liver metastases remains controversial. Simultaneous colorectal and liver resections might impose an increased risk of postoperative liver dysfunction and septic complications [3-5]. In a previous study, we showed that simultaneous hepatectomy was without effect on the healing of left sided colonic anastomoses, however, there was a tendency towards more postoperative complications [6]. The aim of the present study was to investigate the effects of colonic resection on liver function and regeneration in a rat model. Methods Animals Ninety-six male Sprague-Dawley rats (body weight 200-250 g; Japan SLC, Inc., Hamamatsu, Shizuoka, Japan) were used for the experiments. Animals were acclimatised to our laboratory conditions for at least 7 days before the experiment and housed two or three each in stainless steel, wire-mesh cages. The animal room was maintained at a constant temperature and a 12-hour light-dark cycle. Unlimited access was allowed to standard laboratory food and tap water. This experiment was approved by the Ethics Committee of Jichi Medical School, Tochigi, Japan and the NIH Guide for the Care and Use of Laboratory Animals (Institute of Laboratory Animal Resources, National Research Council).

-hour light-dark cycle. Unlimited access was allowed to standard laboratory food and tap water. This experiment was approved by the Ethics Committee of Jichi Medical School, Tochigi, Japan and the NIH Guide for the Care and Use of Laboratory Animals (Institute of Laboratory Animal Resources, National Research Council). Operative Procedure The rats were block-randomised to one of 6 groups. Group I had a laparotomy performed. Group II had 1 cm of the left sided colon resected and anastomosed. Group III and V had 40% or 70% of the liver resected, respectively. Group IV and VI had 40% or 70% of the liver resected and 1 cm of the left sided colon resected and anastomosed, respectively. Our pilot study showed all rats (n = 14) undergoing over 80% liver resection died within 72 hours and, considering also the data by others, 70% liver resection was chosen as the largest tolerable liver resection. Bowel preparation, prophylactic and therapeutic antibiotics administration were not used. Under the diethyl ether anaesthesia, the abdominal wall hair was shaved with electric clippers, and the field was prepared with an alcohol solution. A laparotomy was performed through a 3.5-cm-long transverse incision. The liver resection was performed according to the principles described by Higgins and Anderson [7]. In rats undergoing 70% liver resection, the median lobe and left lateral lobe were removed. Left lateral lobe was removed in rats undergoing 40% liver resection. The pedicles of lobes were ligated with 5-0 sutures (Prolene®; Ethicon, Sommerville, New Jersey, USA), without portal clamping. Colorectal anastomoses were performed by means of eight interrupted 6-0 sutures (PDSII®; Ethicon). The abdominal wall was closed with 4-0 interrupted sutures in two layers (Maxon®; Davis & Geck, Wayne, New Jersey, USA). Autopsy was performed on all rats that died unexpectedly. Body weight was recorded on postoperative day (POD) 3, 5, and 7, and morbidity and mortality were recorded every day. On POD 7, the rats were anesthetised with diethyl ether, and a laparotomy was performed through the previous incision. The animals were inspected for signs of wound infection, and the abdominal cavity was examined for the presence of bowel obstruction, intraperitoneal abscesses, and anastomotic leakage. The rats were finally sacrificed by rapid collection of blood from the inferior vena cava. All procedures were performed by the same researcher.

s were inspected for signs of wound infection, and the abdominal cavity was examined for the presence of bowel obstruction, intraperitoneal abscesses, and anastomotic leakage. The rats were finally sacrificed by rapid collection of blood from the inferior vena cava. All procedures were performed by the same researcher. Liver Regeneration Rate The preoperative total liver weight was calculated from resected liver weight in Group III, IV, V, and VI [7]. In Group I and II, we used 9.9 g as the preoperative total liver weight that was the mean value of Group III, IV, V, and VI. Postoperative total liver weight was measured at sacrifice. The change in liver weight in Group III, IV, V, and VI was evaluated as hepatic regeneration rate (RR). RR is defined as (liver weight per 100 g of the body weight at sacrifice/preoperative projected liver weight per 100 g of the body weight) × 100. Plasma Endotoxin Level and Branched-Chain Amino Acids to Tyrosine Ratio Blood samples were taken from the inferior vena cava at the time of sacrifice. The concentration of plasma endotoxin was measured by turbidimetric time assay (Endotoxin-Single Test Wako®; Wako chemical, Tokyo, Japan), that measures the turbidity change in a coagulation reaction of a solution containing endotoxin and Limulus amoebocyte lysate [8]. The ratio of branched-chain amino acids to tyrosine (BTR) was measured by enzymatic method (Diacolor-BTR kit®; ONO Pharmaceutical, Osaka, Japan).

e Test Wako®; Wako chemical, Tokyo, Japan), that measures the turbidity change in a coagulation reaction of a solution containing endotoxin and Limulus amoebocyte lysate [8]. The ratio of branched-chain amino acids to tyrosine (BTR) was measured by enzymatic method (Diacolor-BTR kit®; ONO Pharmaceutical, Osaka, Japan). Statistical Analysis The effects of colectomy on plasma endotoxin level, liver regeneration rate, and BTR were analysed using analysis of variance (ANOVA) followed by Scheffé post hoc comparison. The changes in body weight through POD 7 were compared among the groups by repeated measures ANOVA. Differences in the total number of complications, including morbidity and mortality, between groups were tested for significance with the Kruskal-Wallis and multiple comparison tests. Probability values of less than 0.05 were considered to indicate statistical significance. All statistical analyses were performed with the statistical software package StatView® for Windows (Version 5.0, SAS Institute Inc., Cary, North Carolina, USA).

significance with the Kruskal-Wallis and multiple comparison tests. Probability values of less than 0.05 were considered to indicate statistical significance. All statistical analyses were performed with the statistical software package StatView® for Windows (Version 5.0, SAS Institute Inc., Cary, North Carolina, USA). Results Mortality and Morbidity Two rats, one in Group II and one in Group VI, had anastomotic leakage leading to the formation of intraperitoneal abscesses. In Group III, one rat developed intraperitoneal abscesses. In Group VI, one rat had bowel obstruction and wound infection, and another had wound infection only. The total number of complications was significantly higher in Group VI than in Group I, III, IV, and V (p ≤ 0.05, Table 1). Two rats in Group VI died; autopsies revealed no signs of wound infection, bowel obstruction, anastomotic leakage, or intraperitoneal abscesses. Table 1 Influence of Colectomy (Cx) on Mortality and Morbidity in Rats Undergoing Different Extents of Hepatic Resection (Hx) Groups(N = 96) Mortality Leakage Obstruction Abscess Wound Infection Total Hx(-) + Cx(-) (Group I, n = 16) 0 - - 0 0 0 Hx(-) + Cx (+) (Group II, n = 16) 0 1 0 1 0 2 Hx(40) + Cx(-) (Group III, n = 16) 0 - - 1 0 1 Hx(40) + Cx(+) (Group IV, n = 16) 0 0 0 0 0 0 Hx(70) + Cx(-) (Group V, n = 16) 0 - - 0 0 1 Hx(70) + Cx(+) (Group VI, n = 16) 2 1 1 1 2 7* Hx(-): without hepatectomy, Hx(40): 40 percent hepatectomy, Hx(70): 70 percent hepatectomy, Cx(-): without colectomy, Cx(+): 1 cm colectomy. * p < 0.05 compared to groups I, III, IV, and V.

+ Cx(+) (Group IV, n = 16) 0 0 0 0 0 0 Hx(70) + Cx(-) (Group V, n = 16) 0 - - 0 0 1 Hx(70) + Cx(+) (Group VI, n = 16) 2 1 1 1 2 7* Hx(-): without hepatectomy, Hx(40): 40 percent hepatectomy, Hx(70): 70 percent hepatectomy, Cx(-): without colectomy, Cx(+): 1 cm colectomy. * p < 0.05 compared to groups I, III, IV, and V. Body Weight Body weight changes were significantly affected by hepatectomy, Group I and II vs. Group III, IV, V and VI (p ≤ 0.05). In Group I vs. II no significant difference in body weight changes were found, Fig. 1A. Significant differences in body weight were observed on POD5 (p ≤ 0.05) and POD7 (p ≤ 0.01) in Group III vs. IV, Fig. 1B. In Group V vs. VI undergoing 70% hepatectomy, significant differences were found on any postoperative day (p ≤ 0.001), Fig. 1C. Figure 1 Illustrating development in rat body weight compared to postoperative day (POD) 0 in percent. Hx(-): without hepatectomy, Hx(40): 40 percent hepatectomy, Hx(70): 70 percent hepatectomy, Cx(-): without colectomy, Cx(+): 1 cm colectomy. G: Group 1A: No statistical significant difference in weight gain between the 2 groups was observed. 1B: At POD 5 and 7 rats without a simultaneous colectomy gained significantly more weight. 1C: At POD 3, 5, and 7 rats without a simultaneous colectomy gained significantly more weight, (*p ≤ 0.05, **p ≤ 0.01, and ***p ≤ 0.001).

1A: No statistical significant difference in weight gain between the 2 groups was observed. 1B: At POD 5 and 7 rats without a simultaneous colectomy gained significantly more weight. 1C: At POD 3, 5, and 7 rats without a simultaneous colectomy gained significantly more weight, (*p ≤ 0.05, **p ≤ 0.01, and ***p ≤ 0.001). Branched-chain Amino Acids to Tyrosine Ratio Colectomy alone decreased BTR on POD 7 significantly in Group II compared to sham operated animals in Group I (p ≤ 0.001). After a 40% hepatectomy, BTR was significantly higher in rats not subjected to colectomy, Group III vs. IV (p ≤ 0.05). No significant difference was observed in BTR between Group V and VI (Fig. 2). Figure 2 Branched-chain Amino Acids to Tyrosine Ratio (BTR) in vena cava blood seven days after surgery. Hx(-): without hepatectomy, Hx(40): 40 percent hepatectomy, Hx(70): 70 percent hepatectomy, Cx(-): without colectomy, Cx(+): 1 cm colectomy, G: Group, (*p ≤ 0.05 and **p ≤ 0.01). Liver Regeneration Rate RR at POD 7 was significantly higher in rats subjected to simultaneous colectomy and hepatectomy (Group III vs. IV, p ≤ 0.001; Group V vs. VI, p ≤ 0.05). No significant difference was observed in RR between Groups I and II (Fig. 3). Figure 3 Liver regeneration rate seven days after surgery. Hx(-): without hepatectomy, Hx(40): 40 percent hepatectomy, Hx(70): 70 percent hepatectomy, Cx(-): without colectomy, Cx(+): 1 cm colectomy, G: Group, (*p ≤ 0.05 and ***p ≤ 0.001). Serum Endotoxin Levels Serum endotoxin levels at POD 7 were not affected by colectomy (Fig. 4.)

Figure 3 Liver regeneration rate seven days after surgery. Hx(-): without hepatectomy, Hx(40): 40 percent hepatectomy, Hx(70): 70 percent hepatectomy, Cx(-): without colectomy, Cx(+): 1 cm colectomy, G: Group, (*p ≤ 0.05 and ***p ≤ 0.001). Serum Endotoxin Levels Serum endotoxin levels at POD 7 were not affected by colectomy (Fig. 4.) Figure 4 Endotoxin concentration in vena cava blood seven days after surgery. Hx(-): without hepatectomy, Hx(40): 40 percent hepatectomy, Hx(70): 70 percent hepatectomy, Cx(-): without colectomy, Cx(+): 1 cm colectomy. G: Group No statistical significant differences were found between the groups. Discussion The present study showed that rats that had a simultaneous hepatectomy and small colectomy performed had a significantly higher liver RR than rats that had only a hepatectomy carried out. BTR and weight development, on the other hand, were significantly lower in the simultaneous colectomy and hepatectomy groups on POD7. There were no significant differences in serum endotoxin levels between groups on POD 7. Excessive liver resection and colectomy could increase the risk of postoperative complications. Our results suggested simultaneous colon and liver resection has accelerated liver volume, however, liver function has not yet fully recovered at POD 7.

nificant differences in serum endotoxin levels between groups on POD 7. Excessive liver resection and colectomy could increase the risk of postoperative complications. Our results suggested simultaneous colon and liver resection has accelerated liver volume, however, liver function has not yet fully recovered at POD 7. Endotoxin, i.e. the lipopolysaccharide (LPS) portion of the bacterial cell wall, is constantly produced in the gut from the death of Gram-negative bacteria. LPS is regularly absorbed into intestinal capillaries and a low grade of portal venous endotoxinemia is the normal status and an increase is seen during and after colonic surgery [9,10]. Endotoxins seem to be a promoter for liver regeneration [11,12]. Kupffer cells in the liver, representing the largest population of reticuloendothelial system macrophages, normally phagocytize and degrade all the LPS in the portal circulation [13]. Removing a large part of the liver reduces the total number of Kupffer cells and this, together with a higher load of LPS due to a colectomy, may lead to higher endotoxin levels in the systemic circulation [14,15]. We could not demonstrate a difference in endotoxin levels between groups on POD 7, partly due to a large variation. This of course does not rule out that there might have been differences earlier in the postoperative course.

PS due to a colectomy, may lead to higher endotoxin levels in the systemic circulation [14,15]. We could not demonstrate a difference in endotoxin levels between groups on POD 7, partly due to a large variation. This of course does not rule out that there might have been differences earlier in the postoperative course. Rats having a simultaneous hepatectomy and colectomy had a significantly lower weight gain than rats which only had a hepatectomy performed. Even though weight loss is a reliable marker of acute stress in animals, it is hard to accept this as the explanation for the difference we found as only 1 cm of the left colon was resected in the simultaneous groups and this represent only a slight additional trauma [16]. That the small colectomy only represents a slight additional trauma to the animals in the simultaneous groups is supported by the fact that we found no weight difference between rats that had only a laparotomy or colectomy performed. A more plausible explanation for the lower weight gain in the simultaneous groups could be a higher degree of endotoxinemia in the early postoperative course and this in combination with a liver resection might increase the surgical stress put on the animals.

ween rats that had only a laparotomy or colectomy performed. A more plausible explanation for the lower weight gain in the simultaneous groups could be a higher degree of endotoxinemia in the early postoperative course and this in combination with a liver resection might increase the surgical stress put on the animals. The liver synthesizes its own proteins and several protein components of plasma and delivers a balanced mixture of amino acids to other organs via the blood. Removing a large part of the liver may, at least theoretically, impair delivery of amino acids. We choose BTR, a variable known to correlate with the Fisher ratio, serum albumin level, prothrombin time and retention of indocyanine green as a marker of liver function [17,18]. Rats having a simultaneous colectomy performed had a significantly lower BTR compared to rats that only had a hepatectomy carried out. As described above, it is hard to explain this finding with a bigger surgical trauma in the simultaneous groups as only 1 cm of the left colon was removed. A more plausible explanation for the lower BTR in the simultaneous groups could be a higher degree of systemic endotoxinemia in the early postoperative course as this has been shown to pronounce effects on the liver [19]. This is also supported by the fact that we found a significantly higher BTR in rats which had only a laparotomy performed compared to rats which had only a small colectomy carried out, i.e. these two groups are nearly identical with regard to the surgical stress put on them but one would expect a significantly higher degree of systemic endotoxinemia in the latter.

nd a significantly higher BTR in rats which had only a laparotomy performed compared to rats which had only a small colectomy carried out, i.e. these two groups are nearly identical with regard to the surgical stress put on them but one would expect a significantly higher degree of systemic endotoxinemia in the latter. The liver is a crucial organ with high regenerative capacity and many complex functions. The liver is unique with respect to its anatomical position allowing continuous blood flow from the gastrointestinal tract through the sinusoids. The liver is also unique with respect to its cellular composition, comprising metabolically active hepatocytes, non-hepatocytic parenchymal cells, and various immune cells. By its location the liver, among other things, is allowed to function as a biochemical defense against toxic chemicals, hereunder endotoxins, entering from the gastrointestinal tract. Replacement of lost hepatic mass is mediated through proliferation of mature adult hepatocytes and the other hepatic cell types and not by proliferation of a selective subpopulation of stem cells [20]. In the present study we found a higher RR in rats which had a simultaneous hepatectomy and colectomy performed. This is in contrast to two other studies by Hachiya Y. et al. and Miyazaki M. et al. who both found an impaired liver regeneration after simultaneous bowel resection in rats [21,22]. Their set-up differed however from ours in that their bowel resections were bigger, i.e. they exposed the animals to greater surgical stress than we did and this might have implications for liver regeneration. In fact the colon resection in our study was so small; the numbers of complications were smaller, that our model might be regarded as a way to achieve endogenous endotoxinemia without exposing animals to much additional surgical stress. The higher RR in the simultaneous groups could be explained by a higher amount of systemic endotoxinemia as these compounds, as described earlier, have been shown to stimulate liver regeneration [11,12]. In the studies by Hachiya Y. et al. and Miyazaki M. et al. the hepatothrophic effect of the increased amount of endotoxins in the portal blood in their simultaneous groups might have been overwhelmed by a negative hepatothrophic effect of the greater surgical trauma [21,22].

own to stimulate liver regeneration [11,12]. In the studies by Hachiya Y. et al. and Miyazaki M. et al. the hepatothrophic effect of the increased amount of endotoxins in the portal blood in their simultaneous groups might have been overwhelmed by a negative hepatothrophic effect of the greater surgical trauma [21,22]. In conclusion, when planning simultaneous colorectal and extended liver resections one should be aware that this procedure, although stimulating the liver's regenerative capacity judged by RR, seems to restrict liver function judged by BTR. In our model morbidity seems to be related to the extent of hepatic resection. Further studies on humans are needed to investigate the clinical implications of these findings. Abbreviations POD: Postoperative day; RR: Regeneration rate; BTR: Branched-chain Amino Acids to Tyrosine Ratio; ANOVA: Analysed of variance; LPS: Lipopolysaccharide Competing interests The authors declare that they have no competing interests. Authors' contributions HS carried out all rat studies and drafted the manuscript. FVM and ARK participated in the design of the study and performed the statistical analysis. FVM and YY conceived of the study, and participated in its design and coordination and PFJ, MO, and HN helped to draft the manuscript. All authors read and approved the final manuscript.

Background Critical limb ischemia (CLI) has been recently re-defined by the Trans-Atlantic Inter-Society Consensus (TASC II) document [1,2]. Contemporarily, a vivid interest has grown on the attempt to identify biologic markers to be coupled to the usual criteria (clinical, hemodynamic, radiologic) used for diagnosis, treatment and follow-up evaluation of CLI [3]. Microdialysis is a well-known and sensitive method for early detection of tissue ischemia [4,5]. Previous studies have been published concerning microdialitic assessment of CLI in physiologic or stress-induced settings [6-8]. The aim of the present study was to investigate the modifications of some metabolites, representative of the local metabolism, during the perioperative period of peripheral revascularization in patients affected by CLI, in order to assess the potential relation between their trend and surgical outcome. Methods Over a 4-months period, 10 patients (8 males) have been enrolled for this study, all diagnosed with CLI affecting just one limb. Exclusion criteria were: co-existing systemic inflammatory diseases, and previous history of revascularization in the ischemic limb. Mean age was 71 (range, 59-84). Co-morbidities and risk factors included tobacco use (n = 9), hypertension (n = 8), dislypidemia (n = 6), chronic obstructive pulmonary disease (n = 4), ischemic heart disease (n = 2), atrial fibrillation (n = 1). The grade of limb ischemia was classified accordingly to Rutherford [1], and included grade II - category 4 (n = 4) and grade III - category 5 (n = 6).

cco use (n = 9), hypertension (n = 8), dislypidemia (n = 6), chronic obstructive pulmonary disease (n = 4), ischemic heart disease (n = 2), atrial fibrillation (n = 1). The grade of limb ischemia was classified accordingly to Rutherford [1], and included grade II - category 4 (n = 4) and grade III - category 5 (n = 6). All patients underwent lower limb revascularization using endovascular percutaneous transluminal angioplasty (PTA) and/or stenting, under local anesthesia; targeted vessels were external iliac artery (n = 2) and superficial femoral artery (n = 8). During the postoperative course, the patency of the treated vessels, and run-off arteries were confirmed using Doppler-ultrasonography. All patients received hydric diet during preoperative period; low molecular weight heparin was administered during the postoperative hospitalization.

uperficial femoral artery (n = 8). During the postoperative course, the patency of the treated vessels, and run-off arteries were confirmed using Doppler-ultrasonography. All patients received hydric diet during preoperative period; low molecular weight heparin was administered during the postoperative hospitalization. The catheter (CMA 60-CMA®-Solna, Sweden) of the microdialysis consisted of a dialyzing polyamide membrane 30 mm in length, 0.5 mm in diameter, and molecular weight cut-off of 20 kDa. Perfusion was guaranteed by a variable flux pump CMA 106 (CMA®-Solna, Sweden) that administered isotonic saline solution (Na+ 147 mMol, K+ 4 mMol, Ca2+ 2.3 mMol, Cl- 156 mMol, 290 mosm/kg) into the probe inserted in the targeted tissue with stable velocity for the entire procedure. Perfusion fluid flow was maintained at 0.3 μL/min. Ultra-filtrated dialyzed solution has been examined using an enzymatic and colorimetric method for glucose, glycerol, lactate and pyruvate (CMA 600 Microdialysis Analyser-CMA®-Solna, Sweden). Evaluated parameters were intended from the interstitial fluid, and not from peripheral blood.

low was maintained at 0.3 μL/min. Ultra-filtrated dialyzed solution has been examined using an enzymatic and colorimetric method for glucose, glycerol, lactate and pyruvate (CMA 600 Microdialysis Analyser-CMA®-Solna, Sweden). Evaluated parameters were intended from the interstitial fluid, and not from peripheral blood. The catheter was allocated in the subcutaneous tissue into the anterior portion of the foot below the ankle. No stress tests have been performed, and both limbs lied horizontal for the entire procedure. Local anesthesia (Carbocaine® 1%-AstraZeneca-Caponago, Milano-Italy), was used to introduce the catheter into the tissue with a slit cannula introducer, leaving the dialyzing polyamide membrane "in situ" when the samples were obtained; this was performed about 12-hours before revascularization The vials were changed at 3-hours intervals pre- and postoperatively, and on an hourly basis throughout the intraoperative stay. "Positive control" was placed into the controlateral limb in the same position.

e membrane "in situ" when the samples were obtained; this was performed about 12-hours before revascularization The vials were changed at 3-hours intervals pre- and postoperatively, and on an hourly basis throughout the intraoperative stay. "Positive control" was placed into the controlateral limb in the same position. Referral physiologic levels was considered accordingly to what reported by Rosdahl et al. [9] (glucose 4.12 ± 0.31 mMol, of lactate 1.30 ± 0.12 mMol, glycerol 68.7 ± 3.97 μMol). Clinical data were recorded and tabulated with Microsoft Excel® (Microsoft Corp, Redmond, WA, USA). Statistical analysis was computed with SPSS, release 16.0 for Windows (SPSS Inc.®, Chicago, IL, USA). Results were expressed as mean ± SD for continuous variables and frequencies for the categorical ones. A 2-sided p value less than .05 was considered statistically significant. Variables and values were compared between groups with unpaired Student's T test and the Wilcoxon test as appropriate. Results Technical revascularization was successful in all cases. The cannulation was well tolerated in all patients. The site of catheter insertion healed easily in few days without infective complications in any case. Two patients underwent major amputation.

Referral physiologic levels was considered accordingly to what reported by Rosdahl et al. [9] (glucose 4.12 ± 0.31 mMol, of lactate 1.30 ± 0.12 mMol, glycerol 68.7 ± 3.97 μMol). Clinical data were recorded and tabulated with Microsoft Excel® (Microsoft Corp, Redmond, WA, USA). Statistical analysis was computed with SPSS, release 16.0 for Windows (SPSS Inc.®, Chicago, IL, USA). Results were expressed as mean ± SD for continuous variables and frequencies for the categorical ones. A 2-sided p value less than .05 was considered statistically significant. Variables and values were compared between groups with unpaired Student's T test and the Wilcoxon test as appropriate. Results Technical revascularization was successful in all cases. The cannulation was well tolerated in all patients. The site of catheter insertion healed easily in few days without infective complications in any case. Two patients underwent major amputation. During preoperative ischemic phase, glucose concentration lied on a lower baseline in the treated limb if compared to the controlateral (mean, 1.61 ± 1.35 mMol/L vs 4.03 ± 0.63 mMol/L, p = .001). In the ischemic limb, circadian rhythm was not observed. After revascularization, glucose showed a strong increase (mean, 5.86 ± 1.52 mMol/L, p = .008). Circadian rhythm, in relation to diet and morning neuroendocrine stimulation, was restored with a trend and a pace almost identical to the positive control within 24-30 hours after surgery (Figure 1). No restoration of the circadian rhythm was noted in patients who underwent major amputation.

± 1.52 mMol/L, p = .008). Circadian rhythm, in relation to diet and morning neuroendocrine stimulation, was restored with a trend and a pace almost identical to the positive control within 24-30 hours after surgery (Figure 1). No restoration of the circadian rhythm was noted in patients who underwent major amputation. Figure 1 Ischemic vs positive control: comparative analysis of the glucose trend in one patient. The basal level in the ischemic limb was perceptible lower than the positive control. After revascularization (arrow), glucose level in the ischemic limb rapidly decreased and finally re-increased reaching similar trend of the positive control: the circadian rhythm was restored. Note that the important variations in the positive control could be influenced by the neuroendocrine stimulation. Glycerol concentration curves were not deductibles in both the ischemic and the control limbs (mean, 148.43 ± 42.13 mMol/L vs 178.44 ± 75.93 mMol/L, p = .348). It exhibited alternated peaks, either in pre- and postoperative period (Figure 2). Standard deviation expresses a highly disordered tendency; however, after revascularization, every peak paired those of the control limb, demonstrating the restoration of circadian rhythm. Both patients who underwent limb amputation did not show similarity of peaks in both legs within 72-hours from the intervention (Figure 3).

deviation expresses a highly disordered tendency; however, after revascularization, every peak paired those of the control limb, demonstrating the restoration of circadian rhythm. Both patients who underwent limb amputation did not show similarity of peaks in both legs within 72-hours from the intervention (Figure 3). Figure 2 Glycerol analysis in a patient who underwent successful revascularization. As yet described in the text, the circadian rhythm was restored in those patients post-procedurally, and the alternated peaks were similar in the treated limb versus the control limb. Figure 3 Glycerol analysis in a patient whose revascularization failed. In this case it was not evident the restoration of the circadian rhythm and the not coupled curves of the treated limb versus the control limb.

Figure 2 Glycerol analysis in a patient who underwent successful revascularization. As yet described in the text, the circadian rhythm was restored in those patients post-procedurally, and the alternated peaks were similar in the treated limb versus the control limb. Figure 3 Glycerol analysis in a patient whose revascularization failed. In this case it was not evident the restoration of the circadian rhythm and the not coupled curves of the treated limb versus the control limb. During the preoperative ischemic phase, lactate concentration lied on a higher baseline in the ischemic limb if compared to the controlateral (mean, 3.68 ± 1.61 mMol/L vs 1.54 ± 1.36 mMol/L, p = .01). Within the first 24-hours after revascularization, lactate concentration raised strongly (6.58 ± 1.56 mMol/L, p = .002) because of the oxygen restoration and the increased aerobic metabolism, also testified by the decreased of the glucose level (Figure 4). Thereafter, lactate immediately decreased to a concentration similar to the control level (1.71 ± 1.69 mMol/L). In both patients who underwent major amputation, lactate did not show the typical peak of the successful revascularization; in particular, a consistent decrement (< 1.74 mMol/L) was observed early after revascularization, but within 12-hours both patients presented an elevation of lactate comparable to the preoperative ischemic phase (Figure 5).

ent major amputation, lactate did not show the typical peak of the successful revascularization; in particular, a consistent decrement (< 1.74 mMol/L) was observed early after revascularization, but within 12-hours both patients presented an elevation of lactate comparable to the preoperative ischemic phase (Figure 5). Figure 4 Glucose and lactate curves in a patient who had successful revascularization. Note the immediate peak of the lactate post-procedurally (arrow), with the specular reduction of the glucose possibly testifying the restoration of the cellular metabolism. Figure 5 Lactate and pyruvate curves in the ischemic limb of a single patient. After the first attempt of revascularization (arrow), lactate decreased. After a short period, lactate started to rise again (sketched arrow): this patient experienced a recurrent thrombosis of the treated arterial district, and finally underwent major amputation.

ruvate curves in the ischemic limb of a single patient. After the first attempt of revascularization (arrow), lactate decreased. After a short period, lactate started to rise again (sketched arrow): this patient experienced a recurrent thrombosis of the treated arterial district, and finally underwent major amputation. The pyruvate quantification showed an almost equivalent concentration in both legs, even if slightly elevated in the ischemic limb when compared with the positive control (98.4 ± 13.3 vs 96.5 ± 14.22 μMol/L, p = .786). Pyruvate evaluation was important to evaluate the lactate/pyruvate ratio: the trend of this ratio was similar to what observed for lactate, with a rapid decrease after the revascularization procedure. We noted that, after a brief initial decrease of the ratio, it increased again in both the patients who finally underwent amputation (Figure 5, 6). We have not evaluated the pyruvate only, because it suffered the influence of the neuroendocrine activity and the circadian rhythm most importantly than any other substances. Figure 6 Lactate/Pyruvate ratio in the ischemic limb of the patient presented in figure 5. The reduced ratio confirmed the initial success of the surgical procedure. After a short period, lactate/pyruvate ratio started to rise again. Note the persisting high level of the ratio for recurrent ischemia.

The pyruvate quantification showed an almost equivalent concentration in both legs, even if slightly elevated in the ischemic limb when compared with the positive control (98.4 ± 13.3 vs 96.5 ± 14.22 μMol/L, p = .786). Pyruvate evaluation was important to evaluate the lactate/pyruvate ratio: the trend of this ratio was similar to what observed for lactate, with a rapid decrease after the revascularization procedure. We noted that, after a brief initial decrease of the ratio, it increased again in both the patients who finally underwent amputation (Figure 5, 6). We have not evaluated the pyruvate only, because it suffered the influence of the neuroendocrine activity and the circadian rhythm most importantly than any other substances. Figure 6 Lactate/Pyruvate ratio in the ischemic limb of the patient presented in figure 5. The reduced ratio confirmed the initial success of the surgical procedure. After a short period, lactate/pyruvate ratio started to rise again. Note the persisting high level of the ratio for recurrent ischemia. Discussion Critical limb ischemia is generally considered the end-stage of peripheral arterial obstructive disease, one of the most common manifestations of atherosclerosis; this is a pathological, distinctive condition because of its progressive and persistent lack of blood and oxygen [1]. The consequences of prolonged ischemia, are cellular death and tissue necrosis as a result of hypoxia and energetic substrates depletion, finally leading to limb loss [2,3,10]. Hence, it should be attractive to identify some specific markers that can be useful especially to optimize diagnosis, and to evaluate the outcomes of the revascularization procedures. The main finding of this study was a kind of "metabolic" state characterizing CLI and early postoperative course after endovascular revascularization.

ld be attractive to identify some specific markers that can be useful especially to optimize diagnosis, and to evaluate the outcomes of the revascularization procedures. The main finding of this study was a kind of "metabolic" state characterizing CLI and early postoperative course after endovascular revascularization. First proposed in late '60s for neurological applications, microdialysis has been adapted to test different tissues [11,12]. Microdialysis is a well-established technique for the continuous sampling of small water-soluble molecules within the extracellular space in vivo. Sampling rationale is based on the passive diffusion of substances across a semi-permeable hollow-fiber dialysis membrane driven by a concentration gradient and supported by a continuous flux pump simulating the function of a vein [13]. Substances smaller than the membrane pores can diffuse into the inner lumen of the catheter and be carried to the outlet by the perfusion fluid in relation to membrane pores and velocity. Larger substances will either be completely rejected by the membrane pores or diffuse so slowly through these pores causing their recovery to be negligible. Our choice of 20 kDa pores supported by a 0.3 μL/min fluid flow was dictated by the will to obtain a recovery proximal to 100% of the third space, in order to evaluate the concentration of metabolic markers that directly reflects the effective status of the intracellular metabolism [9,14,15].

recovery to be negligible. Our choice of 20 kDa pores supported by a 0.3 μL/min fluid flow was dictated by the will to obtain a recovery proximal to 100% of the third space, in order to evaluate the concentration of metabolic markers that directly reflects the effective status of the intracellular metabolism [9,14,15]. Similar analysis on CLI metabolism and its modification during revascularization was performed by Metzsch et al. [5]: the authors tested glycerol, glucose and lactate into subcutaneous and muscular tissue of ischemic limb, adopting pectoral subcutaneous tissue as "positive control". Our study aimed to define specifically the clinical condition of these metabolites in CLI, that's why we opted for the non ischemic controlateral limb as control. Indeed, we estimate the subcutaneous tissue of the most distal part of the leg because this peripheral segment is thought to be the most affected, and sensitive to the ischemic condition; in fact, in CLI blood flow is extremely reduced in this area, therefore the metabolic agent derived from the samples are really emblematic of a pathologic condition. In addition, our study significantly differed since we used two identical location than three different sites, we provided a higher number of samples in a more extensive period of time, and most importantly we had amputations that meant a subgroup of control for the different substances.

atic of a pathologic condition. In addition, our study significantly differed since we used two identical location than three different sites, we provided a higher number of samples in a more extensive period of time, and most importantly we had amputations that meant a subgroup of control for the different substances. The lactate in our series presented similar absolute levels and curves to what published in the experience of Metzcsh et al. [5], but glucose and glycerol demonstrated different patterns. In their experience, glucose was influenced by the neuroendocrine stress response that induced a preoperative growth of the baseline level; in contrast to their findings, our glucose levels rapidly decreased in the post-procedural period as a possible consequence of the restoration of an aerobic metabolism. The glycerol in our group showed a lower baseline level. However, it should be taken into account that our cases were affected by critical (and not chronic) limb ischemia.

findings, our glucose levels rapidly decreased in the post-procedural period as a possible consequence of the restoration of an aerobic metabolism. The glycerol in our group showed a lower baseline level. However, it should be taken into account that our cases were affected by critical (and not chronic) limb ischemia. In their study, Rosdahl et al. [9] tested the same substances in physiologic condition, and demonstrated that glucose concentration was equivalent in plasma and in interstitium, whereas plasmatic concentrations of lactate and glycerol were lower than interstitial concentration found in adipose and muscle tissues. In our study we evaluated a chronic disease; we registered an ischemic preoperative condition distinguished by a lower glucose concentration coupled by a concentration higher lactate level than those observed in physiologic condition. Also, in postoperative period we found a tendency of glucose to rise, whereas lactate decreased after surgery with values tending to adapt to physiologic normality. This was supposed to be related to a slow rehabilitation to the newly restored aerobic metabolism. Glucose and glycerol re-establishment of circadian rhythm was detected in all cases with good early outcome. However, we should advise that the influence of adrenergic stimulation cannot be excluded as confounding factor.

This was supposed to be related to a slow rehabilitation to the newly restored aerobic metabolism. Glucose and glycerol re-establishment of circadian rhythm was detected in all cases with good early outcome. However, we should advise that the influence of adrenergic stimulation cannot be excluded as confounding factor. Of interest, both patients who experienced limb loss did not present a decrement of lactate concentration, that tended to persist at high levels if compared to the pyruvate baseline (persistence of ischemia), and did not present an increase of the glucose interstitial concentration. This was hypothesized to be the result of the persistence of an anaerobic metabolism supporting the cellular need in front of the lack of oxygen. The present study deserves a brief final specification: we posed careful attention to avoid additional trauma to the ischemic limb because the system, left in place for several hours even with minimal trauma, would potentially expose to bacterial agents [9,15]. Of note, our procedures healed easily in few days without infective complications, and the catheterization was well tolerated in all cases.

n to avoid additional trauma to the ischemic limb because the system, left in place for several hours even with minimal trauma, would potentially expose to bacterial agents [9,15]. Of note, our procedures healed easily in few days without infective complications, and the catheterization was well tolerated in all cases. Conclusions The main defect of this study was the limited number of patients, along with the potential confounding variables represented by the neuroendocrine response and the short preoperative period of observation. However, in this opening experience microdialysis revealed to be a potential attractive tool to define the metabolic response distinguishing CLI (absence of circadian rhythm of glucose and glycerol, high lactate and lactate/pyruvate ratio), and also to detect quickly the potential persistence or absence of local ischemia after revascularization. Competing interests The authors declare that they have no competing interests. Authors' contributions MT: concept; study design; carried out the analytic procedures; data analysis; writing. EM: data collection; writing. GP: data analysis; writing. PC: concept; study design; data analysis. All authors read and approved the final manuscript. Acknowledgements This Study was financially supported by PRIN 2005 (Projects for Research of National Interest), and by the Italian Ministry of University and Research (MIUR). Presented at: 4th International Conference on Clinical Microdialysis (ICCM), Sep 19th-21st, 2007 Robinson College-Cambridge, UK

Introduction Gastro-oesophageal reflux disease (GORD) is a common problem in the western world with an incidence of 5 per 1000 population per year [1]. The majority of patients have mild symptoms requiring only occasional treatment. However, a minority have persistent symptoms refractory to medical management and may go on to develop significant complications. This latter group initiated the interest in minimal access surgical approaches to treatment pioneered in the early 1990s. Studies have confirmed the strong association between GORD and the risk of oesophageal adenocarcinoma [2]. GORD is also implicated in the aetiology of a number of related disorders [3]. This has sparked a resurgence of interest in methods not merely of symptom control, but of reflux prevention and cure. The proven safety and efficacy of Laparoscopic Anti-Reflux Surgery (LARS) has increased acceptance and demand for this treatment among patients and healthcare providers [4-9]. We have been offering LARS as a modality of treatment for GORD in our district general hospital (DGH) since September 1999. We report on the incidence and presentations of complications arising from this surgery and propose ways of overcoming them.

Studies have confirmed the strong association between GORD and the risk of oesophageal adenocarcinoma [2]. GORD is also implicated in the aetiology of a number of related disorders [3]. This has sparked a resurgence of interest in methods not merely of symptom control, but of reflux prevention and cure. The proven safety and efficacy of Laparoscopic Anti-Reflux Surgery (LARS) has increased acceptance and demand for this treatment among patients and healthcare providers [4-9]. We have been offering LARS as a modality of treatment for GORD in our district general hospital (DGH) since September 1999. We report on the incidence and presentations of complications arising from this surgery and propose ways of overcoming them. Methods The 301 consecutive patients undergoing LARS by a single consultant surgeon from September 1999 were entered into a prospective database. Ethics committee approval was not required for this procedure. Patients referred to our team from primary or secondary care with symptoms suggestive of chronic GORD were evaluated by taking a detailed history of symptoms and risk factors. Most patients had received initial treatment with Proton pump inhibitors (PPI) and pro-kinetic agents in primary care [10]. They were given advice regarding avoidance of risk factors predisposing to GORD and underwent upper Gastro-Intestinal (GI) Endoscopy, pH and manometric studies before a final decision to perform LARS was made. All intraoperative and postoperative complications were entered into the database. Selection criteria for patients undergoing LARS in this study are outlined in Table 1. Postoperative clinic follow up for all patients was at 2 weeks, 3 months, and one year.

tric studies before a final decision to perform LARS was made. All intraoperative and postoperative complications were entered into the database. Selection criteria for patients undergoing LARS in this study are outlined in Table 1. Postoperative clinic follow up for all patients was at 2 weeks, 3 months, and one year. Table 1 Selection Criteria for Patients undergoing LARS (n = 301) ▪ Persistent symptoms of reflux for over a year plus 1 or more of the following: - Symptoms of volume reflux - Patients responding well to PPI, but unwilling to take pills indefinitely - Symptoms incompletely controlled by medications - Respiratory symptoms General anaesthetic technique A single consultant anaesthetist anaesthetized all but 2 patients. This enabled refinement of the technique, particularly with reference to early detection and treatment of pneumothorax, appropriate pain management, and post operative nausea and vomiting (PONV) prophylaxis. Patients were induced and maintained with intravenous (I.V.) Propofol and Remifentanil (Synthetic opioid). The use of a processed EEG monitor was helpful in confirming an adequate depth of anaesthesia. Where PONV occurred, it was treated aggressively with anti-emetics, IV fluids and occasionally by the use of Dexamethasone. Post-operative analgesia consisted of regular mild analgesics; usually Paracetamol and Ibuprofen, supplemented by subcutaneous morphine in small doses. Surgical technique A single consultant surgeon performed all procedures with a standard 5 port incision.

Where PONV occurred, it was treated aggressively with anti-emetics, IV fluids and occasionally by the use of Dexamethasone. Post-operative analgesia consisted of regular mild analgesics; usually Paracetamol and Ibuprofen, supplemented by subcutaneous morphine in small doses. Surgical technique A single consultant surgeon performed all procedures with a standard 5 port incision. The fundus of the stomach was mobilised completely by dividing the Gastro-splenic omentum including the short gastric vessels usually commencing from the splenic hilum to the gastro-oesophageal junction (GOJ). Circumferential dissection of the GOJ was completed preserving both vagi. A rubber sling was used through the retro-oesophageal window to lift the GOJ; further peri-oesophageal dissection, extending into the lower mediastinum, if needed, was done to deliver at least 3 cm of oesophagus into the abdominal cavity without tension. Crural repair was performed with interrupted sutures of 1-0 Ethibond sutures, and care was taken to prevent narrowing of the hiatus. The needle was left in situ in the lower most hiatal stitch to facilitate wrap fixation later. The posterior wall of the fundus was then brought around the GOJ behind the lower end of the oesophagus.

r was performed with interrupted sutures of 1-0 Ethibond sutures, and care was taken to prevent narrowing of the hiatus. The needle was left in situ in the lower most hiatal stitch to facilitate wrap fixation later. The posterior wall of the fundus was then brought around the GOJ behind the lower end of the oesophagus. The possibility of oesophageal rotation was ruled out by the free to and fro passage of the fundus behind the oesophagus (Shoe-shine test). The fundus was brought through the retro-oesophageal window and all traction was released to confirm absence of tension in the wrap (Drop test). One (1-0) Ethibond suture was placed on the posterior 1/3 of each of the crura and left long to facilitate wrap fixation. A 3 cm floppy 360° fundic wrap was now constructed around the lower end of the oesophagus by suturing the anterior and posterior fundic walls around the oesophagus with several interrupted full thickness sutures without including the oesophageal wall within the sutures. The length of the intra-abdominal oesophagus and wrap were accurately measured* intra-operatively. Three-point fixation of the wrap was now carried out by fixing the wrap inferiorly to the lowest stitch of the crural repair and superiorly to the right and left crux respectively using the sutures with the needles left in-situ. All the needles and sling were now withdrawn and haemostasis confirmed. * Method of measurement ❑ Steristrip was applied to a straight 5 mm dissector on which accurate markings were made so that the first mark was 3 cm from the tip of the instrument.

The possibility of oesophageal rotation was ruled out by the free to and fro passage of the fundus behind the oesophagus (Shoe-shine test). The fundus was brought through the retro-oesophageal window and all traction was released to confirm absence of tension in the wrap (Drop test). One (1-0) Ethibond suture was placed on the posterior 1/3 of each of the crura and left long to facilitate wrap fixation. A 3 cm floppy 360° fundic wrap was now constructed around the lower end of the oesophagus by suturing the anterior and posterior fundic walls around the oesophagus with several interrupted full thickness sutures without including the oesophageal wall within the sutures. The length of the intra-abdominal oesophagus and wrap were accurately measured* intra-operatively. Three-point fixation of the wrap was now carried out by fixing the wrap inferiorly to the lowest stitch of the crural repair and superiorly to the right and left crux respectively using the sutures with the needles left in-situ. All the needles and sling were now withdrawn and haemostasis confirmed. * Method of measurement ❑ Steristrip was applied to a straight 5 mm dissector on which accurate markings were made so that the first mark was 3 cm from the tip of the instrument. ❑ Trade-markings near the tip of the ultrasonic dissectors gave an approximation of the length of the wrap. Results 301 patients underwent Laparoscopic anti-reflux surgery at our centre. The average age of the patients was 46 years (Range: 18 – 77 years) with a median age of 47 (Mode 53). Male to female ratio was1.88: 1.

* Method of measurement ❑ Steristrip was applied to a straight 5 mm dissector on which accurate markings were made so that the first mark was 3 cm from the tip of the instrument. ❑ Trade-markings near the tip of the ultrasonic dissectors gave an approximation of the length of the wrap. Results 301 patients underwent Laparoscopic anti-reflux surgery at our centre. The average age of the patients was 46 years (Range: 18 – 77 years) with a median age of 47 (Mode 53). Male to female ratio was1.88: 1. The most common presenting symptoms were heartburn and volume reflux as illustrated in figure 1. Figure 1 Patient distribution by symptoms (n = 301). The results of pH testing and manometry are depicted in figure 2. Figure 2 Findings of pH and manometric study. Surgery was completed laparoscopically in all patients, except in five, where the operation was technically difficult due to pre-existing conditions (Table 2). Table 2 Causes for inability to perform LARS in five patients (n = 301) ▪ Dense fibro-fatty peri-oesophageal and omental tissue ▪ Extreme omental and intra-abdominal adiposity ▪ Large, floppy liver making retraction difficult and risky ▪ Severe epigastric port site bleeding ▪ Dense fibrotic adhesions between stomach, liver and anterior abdominal wall following previous highly selective vagotomy We had a good record of safety with no mortality and low morbidity. All complications, their presentations, and management were entered into a database (Table 3). Table 3 Complications (n = 301)

▪ Dense fibro-fatty peri-oesophageal and omental tissue ▪ Extreme omental and intra-abdominal adiposity ▪ Large, floppy liver making retraction difficult and risky ▪ Severe epigastric port site bleeding ▪ Dense fibrotic adhesions between stomach, liver and anterior abdominal wall following previous highly selective vagotomy We had a good record of safety with no mortality and low morbidity. All complications, their presentations, and management were entered into a database (Table 3). Table 3 Complications (n = 301) Intraoperative bleeding 1 0.33% PONV unresponsive to antiemetics 1 0.33% Tension pneumothorax 2 0.66% Surgical emphysema 8 2.66% Postoperative wound infection 3 0.99% Painful port-sites 4 1.33% Postoperative laparoscopic port site hernia 1 0.33% Postoperative dysphagia 11 3.65% Wrap migration 2 0.66% Wrap ischaemia 1 0.33% Recurrent regurgitation 4 1.33% Recurrent heartburn 29 9.63% Discussion The cardinal principles of the operation are to restore the high-pressure zone and length of lower oesophagus exposed to the abdominal pressure, repair of the hiatal defect, and a short floppy fundoplication around the lower oesophagus [10,11]. This is achieved by hiatal and lower mediastinal oesophageal dissection [12], reduction of the hiatus hernia, and division of short gastric vessels to release the fundus from the spleen. We perform three-point fixation of the wrap to the crura to decrease the incidence of wrap migration and rotation. The oesophagus was not included in the sutures inserted to create the wrap in order to prevent oesophageal injury.

iatus hernia, and division of short gastric vessels to release the fundus from the spleen. We perform three-point fixation of the wrap to the crura to decrease the incidence of wrap migration and rotation. The oesophagus was not included in the sutures inserted to create the wrap in order to prevent oesophageal injury. In most patients with large hiatus hernias, adequate intra-abdominal length can be achieved by a meticulous and higher mediastinal dissection [12]. We encountered failure in two cases where oesophageal shortening had to be treated by a Leigh Collis type gastroplasty. Accurate intraoperative measurements of the oesophageal intra-abdominal length after full mobilisation were made. A shortened oesophagus may predispose to transdiaphragmatic herniation of the wrap; adequate mobilisation of the oesophagus may prevent this complication in patients who have otherwise had a routine repair with uneventful postoperative recovery [12]. Wrap Migration/Transdiaphragmatic Herniation Wrap migration is synonymous with transdiaphragmatic herniation. These patients can present acutely with life threatening complications or may have longstanding herniation with chronic post-operative reflux or retro-sternal pain symptoms. Incidence of wrap migration ranges between 7% – 20% in published literature and accounts for up to 84% of failed laparoscopic repairs [13-16].

on. These patients can present acutely with life threatening complications or may have longstanding herniation with chronic post-operative reflux or retro-sternal pain symptoms. Incidence of wrap migration ranges between 7% – 20% in published literature and accounts for up to 84% of failed laparoscopic repairs [13-16]. Only two patients (0.714%) had acute wrap migration or herniation in our series, and both presented in an acute setting. One of these patients was further complicated by ischaemic perforation of the strangulated gastric greater curvature. One patient had presented after ingestion of more than a litre of beer, though we routinely advise all our patients to avoid carbonated drinks for six weeks after surgery [16].

nted in an acute setting. One of these patients was further complicated by ischaemic perforation of the strangulated gastric greater curvature. One patient had presented after ingestion of more than a litre of beer, though we routinely advise all our patients to avoid carbonated drinks for six weeks after surgery [16]. Some studies have demonstrated an increased incidence of wrap migration following laparoscopic repair of the hiatus rather than open repair. Reasons cited for this include: a tendency to extend the laparoscopic peri-oesophageal dissection high into the mediastinum, fewer adhesions and reduced postoperative pain allowing greater abdominal force to be transmitted to the hiatal region in the initial postoperative period [13]. However, the incidence of acute wrap migration in our series is low compared to others [13-16]. The strategies we employed to reduce this complication were mobilisation of the oesophagus to achieve adequate length of the intra-abdominal oesophagus, routine suture repair of the hiatus (posterior crural repair in particular), three point wrap fixation, good post-operative analgesia and anti-emetics, such as ondansetron [13,14]. Posterior crural repair was snug and made with non-absorbable material like Ethibond that can hold the repair for prolonged periods [14].

, routine suture repair of the hiatus (posterior crural repair in particular), three point wrap fixation, good post-operative analgesia and anti-emetics, such as ondansetron [13,14]. Posterior crural repair was snug and made with non-absorbable material like Ethibond that can hold the repair for prolonged periods [14]. Longstanding wrap herniation or migration may present with symptoms of chronic postoperative reflux with or without dysphagia or odynophagia [14,15]. We investigated such patients with an OGD followed by CT-Scan or Barium meal. However, these investigations may not be accurate enough for small hernias thus the only solution in symptomatic patients is to dismantle the complete repair laparoscopically (small transdiaphragmatic hernias are almost like a Richter's Hernia and may become evident only on taking down the full wrap) and redoing it carefully [14,16]. In such cases laparoscopy will provide diagnostic as well as, therapeutic benefits. Wrap ischaemia with perforation Wrap ischaemia is a distinct possibility in elderly patients with arteriopathy. In such cases, after mobilising the fundus, a partial wrap may be constructed instead of a full wrap which is likely to decrease interstitial resistance to blood flow.

Longstanding wrap herniation or migration may present with symptoms of chronic postoperative reflux with or without dysphagia or odynophagia [14,15]. We investigated such patients with an OGD followed by CT-Scan or Barium meal. However, these investigations may not be accurate enough for small hernias thus the only solution in symptomatic patients is to dismantle the complete repair laparoscopically (small transdiaphragmatic hernias are almost like a Richter's Hernia and may become evident only on taking down the full wrap) and redoing it carefully [14,16]. In such cases laparoscopy will provide diagnostic as well as, therapeutic benefits. Wrap ischaemia with perforation Wrap ischaemia is a distinct possibility in elderly patients with arteriopathy. In such cases, after mobilising the fundus, a partial wrap may be constructed instead of a full wrap which is likely to decrease interstitial resistance to blood flow. Only one patient had wrap ischaemia with perforation, though that was compounded by wrap migration as described previously. This patient presented three months after a successful fundoplication with progressive epigastric pain and vomiting. The symptoms had started after consuming about two litres of beer. Urgent chest x-rays and CT scan showed wrap migration and gastric perforation. Laparotomy through a midline incision showed wrap migration and ischaemic perforation of the strangulated gastric greater curvature. Stapled partial gastrectomy was performed and GOJ was fixed to the crura. The patient remains symptom free for the last 16 months in spite of having no wrap.

tion and gastric perforation. Laparotomy through a midline incision showed wrap migration and ischaemic perforation of the strangulated gastric greater curvature. Stapled partial gastrectomy was performed and GOJ was fixed to the crura. The patient remains symptom free for the last 16 months in spite of having no wrap. Postoperative dysphagia No patients had absolute postoperative dysphagia in our series.11 patients reported dysphagia of varying degrees at 3 months post surgery. One patient had substantial weight loss at four months and underwent revision surgery. 8 patients had persistent dysphagia for solid food at nine months, but had not lost any weight. These patients were investigated with a Barium swallow and an upper GI Endoscopy. They were initially treated with gentle oesophageal dilatation. Three patients who had a variable result with dilatation, but not complete resolution of dysphagia, elected not to have any further surgical procedure. The other five patients with dysphagia not responding to dilatation underwent a re-look operation with treatment of the cause. Two patients had hiatal stenosis due to fibrosis, which was released and the wrap was converted to partial posterior wrap with resolution of dysphagia; however, one patient developed minor reflux after the procedure. In the other three patients, no cause of dysphagia was found and the original wrap was completely dismantled and then converted to a partial posterior wrap, which relieved the symptoms of dysphagia.

to partial posterior wrap with resolution of dysphagia; however, one patient developed minor reflux after the procedure. In the other three patients, no cause of dysphagia was found and the original wrap was completely dismantled and then converted to a partial posterior wrap, which relieved the symptoms of dysphagia. In our experience, balloon dilatation could be tried within the first 3 months to manage early dysphagia; however, if it was not beneficial then there was no benefit in repeating it for significant dysphagia after 3 months. Complete mobilisation of the gastric fundus with division of the short gastric vessels (Classic Nissen's fundoplication) ensures a floppy tension-free fundoplication, thus preventing dysphagia from a tight wrap or twisted oesophagus [11,17-19]. An incompletely mobilised fundus pulls tightly around the oesophagus, thereby twisting the lower oesophagus and the fundoplication valve itself, and leading to postoperative dysphagia. Many studies have failed to reveal any benefit from complete short gastric vessels division [20-22]. This remains an issue under debate.

n incompletely mobilised fundus pulls tightly around the oesophagus, thereby twisting the lower oesophagus and the fundoplication valve itself, and leading to postoperative dysphagia. Many studies have failed to reveal any benefit from complete short gastric vessels division [20-22]. This remains an issue under debate. Dysphagia due to hiatal stenosis following crural repair has been reported [15]. Various groups are currently analysing the benefits and safety of tension free prosthetic mesh repair over sutured repair of hiatus hernia with promising results [23-27]. However, we did not use the mesh to repair hiatus hernias at this stage. Needless to say, whenever these patients have required a revision procedure, due to dense adhesions with the prosthetic mesh, the revision surgery becomes extremely difficult [13,25]. The mesh may also erode into the oesophagus. If after taking posterior crural stitches, the repair appears too tight, an anterior slit could be made on the arch of the crura to relive the constriction around the oesophagus. With a good three point anterior fixation, the wrap is unlikely to herniate anteriorly.

Dysphagia due to hiatal stenosis following crural repair has been reported [15]. Various groups are currently analysing the benefits and safety of tension free prosthetic mesh repair over sutured repair of hiatus hernia with promising results [23-27]. However, we did not use the mesh to repair hiatus hernias at this stage. Needless to say, whenever these patients have required a revision procedure, due to dense adhesions with the prosthetic mesh, the revision surgery becomes extremely difficult [13,25]. The mesh may also erode into the oesophagus. If after taking posterior crural stitches, the repair appears too tight, an anterior slit could be made on the arch of the crura to relive the constriction around the oesophagus. With a good three point anterior fixation, the wrap is unlikely to herniate anteriorly. Recurrent regurgitation Four patients had recurrent volume reflux three months after surgery. Investigations suggested a shortened oesophagus. Two patients underwent revisional surgery with a Leigh-Collis type gastroplasty with anterior fundoplication stitched to the gastric tube. One patient had complete resolution of reflux. Fundoplication could not be completed in the second patient as he developed tension pneumothorax intra-operatively; his regurgitation has persisted. The third patient with a large hiatus hernia underwent distal gastrectomy with Roux-en-Y gastro-jejunostomy as an adequate length of intra-abdominal oesophagus could not be obtained. This cured his symptoms.

d in the second patient as he developed tension pneumothorax intra-operatively; his regurgitation has persisted. The third patient with a large hiatus hernia underwent distal gastrectomy with Roux-en-Y gastro-jejunostomy as an adequate length of intra-abdominal oesophagus could not be obtained. This cured his symptoms. Complete mobilisation of the gastric fundus with floppy 360° wrap with an adequate length is a crucial step in LARS. We routinely divide the short gastric vessels to adequately mobilise the fundus. It has been widely reported that fundoplication operations dramatically reduce the number of transient lower oesophageal sphincter relaxations occurring both in the basal states as well as after artificial gastric distension or meal ingestion. It is accepted that this effect is essential to prevent further reflux. The static mechanical effects of a total fundoplication are the same irrespective of whether all short gastric vessels have been divided or not. However, Engström et al have reported significantly more transient lower oesophageal sphincter relaxations in patients with intact short gastric vessels after a laparoscopic total fundoplication [28]. They believe that this could be explained by the unavoidable partial denervation of the proximal stomach area without completely dividing short gastric vessels. This may affect the triggering of mechano-receptors in the proximal part of the stomach, giving rise to transient lower oesophageal sphincter relaxations [28].

They believe that this could be explained by the unavoidable partial denervation of the proximal stomach area without completely dividing short gastric vessels. This may affect the triggering of mechano-receptors in the proximal part of the stomach, giving rise to transient lower oesophageal sphincter relaxations [28]. Recurrent heartburn Twenty-nine patients suffered recurrent heartburn after surgery. The symptoms were diminished compared to preoperative state, but they continued to need regular or intermittent treatment with PPI. All of these patients appeared to have a small volume wrap when visualised by retroflexion manoeuvre at gastroscopy. All patients were followed up. Six patients had revision surgery. One patient had a short intra-abdominal oesophagus. An oesophageal lengthening procedure was performed with an anterior wrap, and this was successful in relieving the heartburn at 3 years after surgery. In five patients, a longer (5 cm) wrap was fashioned after completely dismantling the previous wrap. This was successful in relieving the heartburn in four of them. We had six patients with mild recurrent symptoms requiring intermittent PPI. No definitive cause for recurrent symptoms could be found in these patients as post-operative OGD and oesophageal pH readings were normal. These patients did not wish to have any further surgery. Twelve patients were satisfied that their symptoms had improved, did not want any further investigations, and continued with intermittent PPI. Two patients are awaiting revision surgery. Three patients were lost to follow up.

We had six patients with mild recurrent symptoms requiring intermittent PPI. No definitive cause for recurrent symptoms could be found in these patients as post-operative OGD and oesophageal pH readings were normal. These patients did not wish to have any further surgery. Twelve patients were satisfied that their symptoms had improved, did not want any further investigations, and continued with intermittent PPI. Two patients are awaiting revision surgery. Three patients were lost to follow up. Recurrent symptomatic heartburn continues to be an issue evading full explanation. There is an emerging belief that there need not be a correlation between position and integrity of the wrap and the results of the surgery [29]. Various authors have seen definitive benefit in revision surgery for recurrent symptoms resulting from anatomical and functional failures recognised post-operatively [30]. Attempting a fundoplication without achieving an adequate intra-abdominal length of oesophagus could be a frequent cause of surgical failure. Actual intra-operative measurement of intra-abdominal oesophageal and wrap length, as we did in our series, may help to recognise inadequate intra-abdominal length of the oesophagus and/or length of the wrap [31,32]. The absence of significant reflux symptoms in our patients who had to have the fundoplication undone or resected post operatively due to intra-operative or post-operative wrap related problems highlights the importance of restoration of adequate intra-abdominal length of oesophagus to the functional end result of the operation [11,12].

gnificant reflux symptoms in our patients who had to have the fundoplication undone or resected post operatively due to intra-operative or post-operative wrap related problems highlights the importance of restoration of adequate intra-abdominal length of oesophagus to the functional end result of the operation [11,12]. Successful LARS leads to a significant improvement in quality of life of patients [4,33,34]. Only 16 patients in our series needed to use proton pump inhibitors post operatively for recurrent symptoms. None of them were on regular proton pump inhibitors. Even these patients feel that their quality of life has improved significantly since they don't need to take daily medications. In terms of therapeutic advantage, LARS achieves symptomatic relief with abolition of mechanical reflux [4,35]. This protects the oesophageal mucosa from injury due to the non-acidic components of the refluxate [12,33,34,36]. This may have an added advantage in promoting mucosal healing, reversal of mucosal changes caused by long standing reflux disease and preventing sequelae of longstanding reflux like Barrett's disease and strictures [37]. This could provide the key to decreasing the incidence of pre-malignant change and possibly lower the incidence of oesophageal cancers, which have been increasing in prevalence over the last decade. Pneumothorax Four patients had a small pneumothorax, and two patients had an inadvertent tension pneumothorax intraoperatively. Tension pneumothorax occurred on the left side in both patients during trans-hiatal dissection for revisional surgery.

In terms of therapeutic advantage, LARS achieves symptomatic relief with abolition of mechanical reflux [4,35]. This protects the oesophageal mucosa from injury due to the non-acidic components of the refluxate [12,33,34,36]. This may have an added advantage in promoting mucosal healing, reversal of mucosal changes caused by long standing reflux disease and preventing sequelae of longstanding reflux like Barrett's disease and strictures [37]. This could provide the key to decreasing the incidence of pre-malignant change and possibly lower the incidence of oesophageal cancers, which have been increasing in prevalence over the last decade. Pneumothorax Four patients had a small pneumothorax, and two patients had an inadvertent tension pneumothorax intraoperatively. Tension pneumothorax occurred on the left side in both patients during trans-hiatal dissection for revisional surgery. In our experience, a small pneumothorax occurring towards the end of the procedure is of little consequence as the CO2 is absorbed very quickly from the pleural cavity [15]. However, a pneumothorax in the middle of the procedure may make the ventilation of the lungs difficult, and could also result in tension pneumothorax with collapse of the ipsilateral lung and compromising cardio-vascular system. Inserting an inter-costal drain intraoperatively would not be useful as the pneumoperitoneum would be lost through the drain.

dle of the procedure may make the ventilation of the lungs difficult, and could also result in tension pneumothorax with collapse of the ipsilateral lung and compromising cardio-vascular system. Inserting an inter-costal drain intraoperatively would not be useful as the pneumoperitoneum would be lost through the drain. Careful, millimetre-by-millimetre dissection and meticulous haemostasis to prevent any deterioration in view of the operative field during trans-hiatal peri-oesophageal dissection is necessary to minimise the risk of this complication [15]. However, should it inadvertently happen, we immediately lower the pneumoperitoneum to 8 mm of Hg, and try to plug the defect in the pleura with omentum or stomach. This strategy has been successfully used in our patients, and we completed all the procedures laparoscopically. We had tried to stitch the pleura intraoperatively in one patient, but did not find this to be an effective method. Surgical emphysema and pneumomediastinum In our series, a degree of surgical emphysema was relatively common. Detection can be difficult as the early signs are subtle and non-specific. It was usually first palpable in the left supra-clavicular area, then becoming bilateral and extending into the neck. Two patients had surgical emphysema involving the eyelids.

Surgical emphysema and pneumomediastinum In our series, a degree of surgical emphysema was relatively common. Detection can be difficult as the early signs are subtle and non-specific. It was usually first palpable in the left supra-clavicular area, then becoming bilateral and extending into the neck. Two patients had surgical emphysema involving the eyelids. The earliest physiological signs were a gradually increasing heart rate, although this was not a specific sign, as we commonly noted tachycardia and hypertension after inducing pneumoperitoneum. It was more common in slim patients and those having more extensive or prolonged mediastinal dissections. Surgical emphysema and pneumomediastinum are inadvertent but unavoidable complications of high mediastinal dissection to achieve adequate intra-abdominal length of oesophagus. All patients need to be forewarned about this complication, as it is often difficult to make accurate predictions preoperatively about the possibility of short oesophagus necessitating high mediastinal dissection. Even in extensive cases, patients usually need only reassurance [15].

ntra-abdominal length of oesophagus. All patients need to be forewarned about this complication, as it is often difficult to make accurate predictions preoperatively about the possibility of short oesophagus necessitating high mediastinal dissection. Even in extensive cases, patients usually need only reassurance [15]. Intra-operative detection of a developing pneumothorax is difficult without a high degree of suspicion as the signs are non-specific. The operator may be unaware of perforating the pleura, as often there are no visual clues; however occasionally, the diaphragm bulges into the peritoneal cavity, which may be the earliest sign of pneumothorax. The early signs of this complication were usually gradual tachycardia and small rises in ventilatory pressures. Tachycardia is not a reliable sign on its own as it occurs commonly (with hypertension) after inducing a pneumoperitoneum. In all cases of tension pneumothorax in our series, it was left sided. Diagnosis was by clinical suspicion and confirmed by careful auscultation of the lung fields. It is important to differentiate this from endobronchial migration of the endotracheal tube, which may easily occur after inducing pneumoperitoneum, as the carina is pushed cephalad. Treatment in all cases was successful deflating the pneumoperitoneum for a brief period until cardio-respiratory parameters normalised, then plugging the defect if one was identifiable, and completing the procedure with the lowest possible intraperitoneal pressures.

In all cases of tension pneumothorax in our series, it was left sided. Diagnosis was by clinical suspicion and confirmed by careful auscultation of the lung fields. It is important to differentiate this from endobronchial migration of the endotracheal tube, which may easily occur after inducing pneumoperitoneum, as the carina is pushed cephalad. Treatment in all cases was successful deflating the pneumoperitoneum for a brief period until cardio-respiratory parameters normalised, then plugging the defect if one was identifiable, and completing the procedure with the lowest possible intraperitoneal pressures. Post operative wound problems Morphine is not an effective treatment for acutely painful port sites, and higher doses may contribute to PONV; injection of local anaesthetics at port sites may be the best option. Three patients with minor wound infections were treated conservatively. Post-operative nausea and vomiting All patients received PONV prophylaxis; where PONV occurred, it was treated aggressively with anti-emetics, IV fluids and occasionally by the use of Dexamethasone. One patient required the insertion of a nasogastric tube few hours postoperatively for persistent nausea and retching, which was not responding to anti-emetics. Conclusion The experience from our series shows that LARS can be provided in a cost effective manner in a DGH NHS setting. This assumes significance considering the prevalence of GORD in the population. LARS can save the community the cost of long-term treatment with PPI and improves the quality of life of patients.

Post-operative nausea and vomiting All patients received PONV prophylaxis; where PONV occurred, it was treated aggressively with anti-emetics, IV fluids and occasionally by the use of Dexamethasone. One patient required the insertion of a nasogastric tube few hours postoperatively for persistent nausea and retching, which was not responding to anti-emetics. Conclusion The experience from our series shows that LARS can be provided in a cost effective manner in a DGH NHS setting. This assumes significance considering the prevalence of GORD in the population. LARS can save the community the cost of long-term treatment with PPI and improves the quality of life of patients. Growing interest in the pathophysiology of GORD has led to technical refinements, which have made LARS effective with very low morbidity. The operation can be performed with minimal pain and postoperative discomfort. This has resulted in growing acceptability of the procedure among the patients and healthcare providers alike. The growing application of LARS to treatment of GORD could have a positive impact on the incidence and prevalence of both direct and indirect sequelae of the disease in the community.

postoperative discomfort. This has resulted in growing acceptability of the procedure among the patients and healthcare providers alike. The growing application of LARS to treatment of GORD could have a positive impact on the incidence and prevalence of both direct and indirect sequelae of the disease in the community. Though LARS can be performed with low morbidity, a high index of suspicion needs to be maintained to detect early postoperative complications. We have seen that favourable outcomes depend on the rapidity of response to the acute complications. Most complications, when detected early, can be rectified laparoscopically avoiding the need for a major laparotomy. There is a need therefore to alert surgeons from other specialities who participate in acute general surgical take to these issues. Competing interests The authors declare that they have no competing interests. Authors' contributions TS wrote the article, participated in the sequence alignment and drafted the manuscript, SB participated in the sequence alignment, formatted the pictures and performed language corrections, AH collected the data and investigation studies, participated in the article design and critically evaluated the article, SGS collected the data and investigation studies, participated in the article design and critically evaluated the article, AP participated in the article design and critically evaluated the article, SEH conceived the study, and participated in its design and coordination, and helped to draft the manuscript. All authors read and approved the final manuscript.

Introduction The use of disposable non-critical items has increased over the years. This practice has been driven in large part by the known risk of fomite infection transmission and a lack of reliable alternatives to standard manual decontamination. A typical disposable item is used once and then discarded, resulting in hundreds of millions of dollars in annual costs. A midsized hospital, for example, may utilize over 30,000 units of disposable pulse oximeter sensors, at $9–$15 per unit [1]. Similar costs incurred across the US for this item only would yield millions of dollars per year (and larger costs for bigger hospitals), resulting in significant annual costs for this single item, disposable pulse oximeter sensors, running into hundreds of millions of dollars per year. Multiply that by the cost of all non-critical disposables and the costs will tend towards the astronomical. Until the cost-effectiveness of the various disposable items has been established, explicit attention should be paid to the cleaning and disinfection of equipment between patients [2]. According to Spaulding, noncritical equipment is defined as those items that come into contact with intact skin but not with mucous membranes [3]. Since intact skin is an effective barrier to most microorganisms, items that contact only the skin do not need to be sterile and may be cleaned where they are used [4]. In 1991, Favero and Bond provided a useful expansion of the Spaulding scheme by dividing the non-critical environmental surfaces into housekeeping surfaces and medical equipment surfaces [5].

r to most microorganisms, items that contact only the skin do not need to be sterile and may be cleaned where they are used [4]. In 1991, Favero and Bond provided a useful expansion of the Spaulding scheme by dividing the non-critical environmental surfaces into housekeeping surfaces and medical equipment surfaces [5]. Non-critical equipment used in the medical environment can, however, serve as fomites harboring microorganisms which can be transmitted [6-10] and contribute to nosocomial infections and hospital outbreaks [7,8]. Microorganisms such as Methicillin-resistant Staphylococcus aureus (MRSA), Vancomycin-resistant enterococcus (VRE), and gram negatives can survive on inanimate objects such as hospital equipment for many months [8,9]. MRSA is now endemic and even epidemic in many U.S. hospitals and long-term care facilities and is of particular importance to ambulatory and ED settings in many communities [11-15]. When clean (washed) hands or gloved hands touch contaminated objects, they become contaminated with similar organisms, which can then potentially be transmitted to other surfaces and people [16,17]. Inadequately disinfected reusable equipment has been reported to be a source of nosocomial infection, particularly with regard to MRSA and VRE. Nosocomial infections [NI], reportedly account for up to 80,000 U.S. deaths each year, according to the Center for Disease Control and Prevention (CDC) [18,19]. Accordingly, the CDC and the Healthcare Infection Control Practices Advisory Committee (HICPAC) recommend routine decontamination of all hospital equipment [20].

socomial infections [NI], reportedly account for up to 80,000 U.S. deaths each year, according to the Center for Disease Control and Prevention (CDC) [18,19]. Accordingly, the CDC and the Healthcare Infection Control Practices Advisory Committee (HICPAC) recommend routine decontamination of all hospital equipment [20]. Manual decontamination is often unsatisfactory due either to ineffectiveness of decontamination protocols or to poor compliance with these protocols, with commonly used equipment in the hospital frequently remaining contaminated after manual decontamination [9]. Particularly problematic have been multi-planar/configurational devices that challenge traditional cleaning modalities. There has been scant attention dedicated to high-volume areas with rapid patient turnover such as the ED (and other ambulatory settings) compared to comparatively low-turnover inpatient settings, e.g., long-term care and intensive care units where most studies to date have been conducted [7,10-12,17,21,22]. Until now, there has not been a method of automated multimodal and multi-configurational image distance-based decontamination of hospital equipment in the clinical area that does not interfere with patient flow. An ideal device would allow for safe, effective, and efficient decontamination of portable hospital equipment of varying sizes in the clinical area. Such a device would be useful in high-volume clinical areas. Looking to the precedent set by operating rooms, exemplified by their ability to effectively sterilize instruments for reuse due to a process- and device-based standard of care, the self-cleaning unit for the decontamination of small objects (SUDS) was developed.

a device would be useful in high-volume clinical areas. Looking to the precedent set by operating rooms, exemplified by their ability to effectively sterilize instruments for reuse due to a process- and device-based standard of care, the self-cleaning unit for the decontamination of small objects (SUDS) was developed. The objective of this study was to develop a device to more effectively decontaminate portable hospital equipment in comparison with standard manual decontamination and to evaluate differences in bacterial contamination rates. Our specific hypothesis is that an automated unit for the decontamination of medical equipment provides a safe, effective, means of decontaminating hospital equipment of various sizes and configurations in high-volume clinical areas without interfering with patient flow. Methods Overview Prospective comparative study of the effectiveness of standard manual and SUDS decontamination of the patient care area in the ED. In brief, results of microbiologic sampling of medical and electronic equipment in patient care areas conducted following standard manual decontamination was compared to microbiologic sampling following use of the SUDS. In this setting, staff members were blinded as to the timing of the proposed research. In the ED setting, instruments were placed in the SUDS immediately after manual decontamination and replaced in the patient care area/bed space for use on subsequent patients (see figure 1). Figure 1 Schematic of Procedure in Emergency Department.

Methods Overview Prospective comparative study of the effectiveness of standard manual and SUDS decontamination of the patient care area in the ED. In brief, results of microbiologic sampling of medical and electronic equipment in patient care areas conducted following standard manual decontamination was compared to microbiologic sampling following use of the SUDS. In this setting, staff members were blinded as to the timing of the proposed research. In the ED setting, instruments were placed in the SUDS immediately after manual decontamination and replaced in the patient care area/bed space for use on subsequent patients (see figure 1). Figure 1 Schematic of Procedure in Emergency Department. Setting The adult ED of Johns Hopkins Hospital (JHH) sees over 58,000 patients annually. The ED has 32 bed spaces, excluding acute specialty care spaces. There are 14 patient care areas, which for the purposes of this study, were defined as enclosed areas that may contain one or more beds having one entrance and exit. Inclusion criteria: all portable equipment (i.e. any detachable, movable, and reusable equipment) as well as all electronic devices utilized in association with patient care areas and bed spaces (e.g., phones and keyboards). Exclusion criteria: Non-detachable and non-movable equipment (e.g. light fixtures). In addition, bed spaces outside the patient care areas, bed spaces in specialty areas such as acute psychiatric care areas, and environmental surfaces were not included in this study. Electronics in shared spaces were excluded, as manual decontamination could not be assured during turnover of individual patients. ED Procedures: The ED component of the study was carried out over a four-day period. Standard practice in the ED is manual decontamination of the clinical area and equipment in the patient care area following each patient discharge and prior to bringing a new patient into the patient care area. Therefore, swabbing is done during turnover, just prior to the admission of a new patient into the patient care area or bed space.

the ED is manual decontamination of the clinical area and equipment in the patient care area following each patient discharge and prior to bringing a new patient into the patient care area. Therefore, swabbing is done during turnover, just prior to the admission of a new patient into the patient care area or bed space. Equipment Equipment was defined as portable if it was detachable, movable, and reusable with patient contact. Electronic equipment in clinical areas was also included in this study as a separate category. In the ED, equipment was classified into three predefined groups: 1) stands, 2) cables, and 3) electronics. For the purposes of this study, portable electronics included either keyboards or phones. Most patient care areas had one keyboard and phone each, both of which were swabbed. In areas where the electronics were a shared resource (e.g., one phone and keyboard serving multiple patient areas), the shared electronic equipment was excluded. Stands included all wheeled, portable equipment (e.g., intravenous poles, sphygnometers). All stands present in the patient care areas during the study period were swabbed. Cables included all portable equipment that was in cable form such as EKG leads and cables, blood pressure cables, pulse oximeter cables, etc., that were used for contact with the patient. Using a random sequence generator, one cable was chosen randomly in each patient care area for sampling.

study period were swabbed. Cables included all portable equipment that was in cable form such as EKG leads and cables, blood pressure cables, pulse oximeter cables, etc., that were used for contact with the patient. Using a random sequence generator, one cable was chosen randomly in each patient care area for sampling. Sampling for microbiologic assessment: For all devices/equipment, the site of culture was predetermined and standardized prior to initial swabbing. Therefore, the same site was chosen for swabbing after both manual and automated decontamination. A 3 cm2 area was marked and swabbed on each instrument. Samples collected were semi-quantitatively assayed for microorganisms. A single investigator swabbed the equipment (see table 1). Table 1 Equipment included in this study [Emergency Setting] Stands Intravenous poles 6 Ultrasound machines 3 Blood pressure stands 9 Pulse oximeter stands 6 Mayo stands 3 Blood pressure monitor 3 Miscellaneous 8 Electronics Wall phones 7 Keyboards 14 Cables

Sampling for microbiologic assessment: For all devices/equipment, the site of culture was predetermined and standardized prior to initial swabbing. Therefore, the same site was chosen for swabbing after both manual and automated decontamination. A 3 cm2 area was marked and swabbed on each instrument. Samples collected were semi-quantitatively assayed for microorganisms. A single investigator swabbed the equipment (see table 1). Table 1 Equipment included in this study [Emergency Setting] Stands Intravenous poles 6 Ultrasound machines 3 Blood pressure stands 9 Pulse oximeter stands 6 Mayo stands 3 Blood pressure monitor 3 Miscellaneous 8 Electronics Wall phones 7 Keyboards 14 Cables Blood pressure 16 Pulse oximeter 8 Ekg cables 8 Intervention The self-cleaning unit for the decontamination of small instruments (SUDS) is a multimodal portable decontamination unit. This unit allows for primary, secondary, and tertiary decontamination mechanisms with aerosolized biocide, ultraviolet light, and dry heat, respectively. Surface and base rotation via a clockwise and counterclockwise mechanism serves to increase the exposure of equipment to the biocide by optimizing air flow directionality. Turbulence generated at the base allows for air flow patterns that increase exposure to the undersurface of the equipment. Image distance-based techniques allow for maximum intervention in specific areas. S-shaped curvatures at the edges of the surface rotatory mechanism allow for the attachment of multiple devices to the surface rotatory mechanism. Air cleaning and filtration modes allow for the expulsion of clean air into the environment; this may be connected to the facility filtration system. Only the primary mode of decontamination (aerosolized biocide) was used in this study.

ism allow for the attachment of multiple devices to the surface rotatory mechanism. Air cleaning and filtration modes allow for the expulsion of clean air into the environment; this may be connected to the facility filtration system. Only the primary mode of decontamination (aerosolized biocide) was used in this study. The advantages of the SUDS are consistency in automated dispensation and the design of rotational and turbulence forces to optimize dispersal. In addition, targeted image distance-based intervention to specific areas allows for maximal exposure of devices of all configurations. Variability in the choice of biocide and the multimodal approach limits the likelihood of the development of biocide resistance. Adjuncts include mechanical, humidity, and pressure-based interventions. Air cleaning allows for use in the clinical area without interfering with patient flow. Cycle time is variable depending upon the biocide used, the targeted and standard dispersal of same, and the ultraviolet light. During this study, cycle time was 30 m for up to 15 instruments of varying configurations in one cycle. The SUDS was derived from the operating room model of instrument care where instruments are sterilized locally and reused in an automated fashion within a culture of staff maintained cleanliness. Since non-critical medical equipment does not need to be sterilized but merely decontaminated, this decontamination can be done on site [23].

ved from the operating room model of instrument care where instruments are sterilized locally and reused in an automated fashion within a culture of staff maintained cleanliness. Since non-critical medical equipment does not need to be sterilized but merely decontaminated, this decontamination can be done on site [23]. Biocide For the purposes of this study, a commonly available biocide, Sporicidin®, was used. Sporicidin® is a 1.56% phenol and 0.06% sodium phenate solution that can be delivered in aerosolized form. Sporicidin® Disinfectant Solution is FDA 510(k)-cleared, EPA-registered for hospital use, and compliant with OSHA Blood-borne Pathogens Standards (29 CFR 1910.1030) [24]. It is used for both direct manual decontamination and aerosolized delivery per manufacturer's directions. SUDS allows for the use of most biocides that can be delivered in an aerosolized form, allowing for versatility of type, concentration, and duration of biocide delivery, which minimizes microbial resistance. Manual cleaning was carried out with AIREX 109 per usual hospital protocol. This is a quaternary ammonium compound. We were unable to get aerosolized versions.

an be delivered in an aerosolized form, allowing for versatility of type, concentration, and duration of biocide delivery, which minimizes microbial resistance. Manual cleaning was carried out with AIREX 109 per usual hospital protocol. This is a quaternary ammonium compound. We were unable to get aerosolized versions. Microbiologic Analysis Rayon swabs (Copan Diagnostics, Corona, CA) were used to obtain cultures. Swab cultures were plated onto trypticase soy agar with 5% sheep blood (SBA) and MacConkey agar, and the swab was then put into a Schaedler's broth to incubate overnight. The SBA and MacConkey plates were plated semi-quantitatively. The broth was included to enrich organisms present in small quantities. If the broth was turbid after 24 hrs of incubation, it was sub-cultured to selective media to evaluate for the presence of methicillin-resistant Staphylococcus aureus (MRSA), vancomycin-resistant Enterococci (VRE), and resistant gram-negative organisms. From the selective media, only potential MRSA, VRE, and resistant gram-negative organisms were completely characterized. All cultures were incubated for 48 h at 35°C under aerobic conditions. Contamination was defined as equipment containing more than 5 colony-forming units (CFUs)/cm2. Clinically significant organisms were defined as those known to cause human disease in a nosocomial setting (MRSA, VRE, resistant gram-negative rods, and fungi).

tures were incubated for 48 h at 35°C under aerobic conditions. Contamination was defined as equipment containing more than 5 colony-forming units (CFUs)/cm2. Clinically significant organisms were defined as those known to cause human disease in a nosocomial setting (MRSA, VRE, resistant gram-negative rods, and fungi). Statistical Analysis The total population of portable items present within a shift that met criteria for this study was defined. This included two cables per bed space, all electronics in the patient care area, and all stands present during the shift. With a 95% confidence interval and 4.5% confidence level, a sample size of 91 instruments was obtained. One cable from each bed space was randomly selected from the minimum 2 cables present using a random sequence generator, while convenience sampling of the stands and electronics was carried out. All electronics as defined in this study from each patient care area and all stands present during the study period were included.

able from each bed space was randomly selected from the minimum 2 cables present using a random sequence generator, while convenience sampling of the stands and electronics was carried out. All electronics as defined in this study from each patient care area and all stands present during the study period were included. Results Baseline level of contamination were measured following standard ED manual decontamination procedures. A total of 182 swabs were taken from predetermined areas on 91 pieces of selected equipment from bed spaces and patient care areas. The various instruments sampled are detailed in Table 1; however, in brief, 14 keyboards, seven wall phones, all 38 stands associated with patient care areas during a shift, and at least one cable from all bed spaces (32) were swabbed. Eighty percent of instruments were contaminated (Figure 2), and 25% (22/91) were contaminated by clinically significant organisms (CSO). It is notable that 75% of equipment was lightly or moderately contaminated (< 50 CFUs/cm2) following manual decontamination; the remaining equipment was > 50 CFUs/cm2. All heavily contaminated equipment had positive cultures. Figure 2 Overall Contamination Levels in the Emergency Department.

Results Baseline level of contamination were measured following standard ED manual decontamination procedures. A total of 182 swabs were taken from predetermined areas on 91 pieces of selected equipment from bed spaces and patient care areas. The various instruments sampled are detailed in Table 1; however, in brief, 14 keyboards, seven wall phones, all 38 stands associated with patient care areas during a shift, and at least one cable from all bed spaces (32) were swabbed. Eighty percent of instruments were contaminated (Figure 2), and 25% (22/91) were contaminated by clinically significant organisms (CSO). It is notable that 75% of equipment was lightly or moderately contaminated (< 50 CFUs/cm2) following manual decontamination; the remaining equipment was > 50 CFUs/cm2. All heavily contaminated equipment had positive cultures. Figure 2 Overall Contamination Levels in the Emergency Department. The most common recovered clinically significant organisms were the gram-negative rods, which were found in 16.5% (15/91) of instruments. Twelve percent (11/91) of instruments had MRSA, 4.4% (4/91) had VRE and fungi, respectively. Fourteen percent (13/91) had more than one organism, and 5.5% (5/91) of instruments had more than 3 organisms. Also, 8.8% (8/91) of instruments had more than one type of GNR. The most common equipment type contaminated with clinically significant organisms was the stands (13/38) 34.2%, while 19% (4/21) of electronics, and 12.5% (4/32) of cables were contaminated (see Figure 3). Of the 13 contaminated stands, 23%(3/13) were contaminated with VRE, 23% (3/13) with MRSA, 84.6%(11/13) with GNRs and 23% (3/13)) with fungi. Fifty-four percent (7/13) contaminated stands had more than one organism (Figure 3).

while 19% (4/21) of electronics, and 12.5% (4/32) of cables were contaminated (see Figure 3). Of the 13 contaminated stands, 23%(3/13) were contaminated with VRE, 23% (3/13) with MRSA, 84.6%(11/13) with GNRs and 23% (3/13)) with fungi. Fifty-four percent (7/13) contaminated stands had more than one organism (Figure 3). Figure 3 Instrument Contamination by Type of Organism: Emergency Department. Note: instruments may be colonized with > 1 clinically significant organism (CSO). For the intervention phase, all instruments were placed in the SUDS for 20 minutes of decontamination and 10 minutes of "stand time." There was no growth of CSOs following SUDS decontamination (p < 0.001). Forty-eight hours after the initial automated decontamination, all previously contaminated emergency room equipment was re-swabbed in the same previously defined area. All of the previously contaminated equipment was found to be free of CSOs 48 hours later. All electronic equipment remained fully functional following SUDS decontamination.

ter the initial automated decontamination, all previously contaminated emergency room equipment was re-swabbed in the same previously defined area. All of the previously contaminated equipment was found to be free of CSOs 48 hours later. All electronic equipment remained fully functional following SUDS decontamination. Discussion Based upon the risk of potential transmission of hospital acquired (nosocomial) pathogens, the CDC and multiple leaders in the field [8,9,20] have suggested that routine cleaning of medical equipment with antimicrobial agents, especially in the immediate vicinity of patients, is recommended. Maintaining the chain of cleanliness, decreasing microorganism load, and targeted eradication of sources of infection have already been shown to decrease hospital acquired infections and eliminate outbreaks [[8,9], and [23]]. Clinically significant organisms have been shown to persist on equipment, sometimes for months [23,25]. A review of the mostly inpatient setting literature reveals that non-critical hospital equipment as well as electronics such as keyboards and wall phones harbor pathogenic and non-pathogenic bacteria. Rates of contamination of up to 100% with varying pathogenic organisms have been described [9]. Boyce et al. reported a contamination rate of 25% with MRSA for an intravenous pole in the inpatient setting (although this was in the vicinity of colonized patients), while Rutela described rates of up to 36% for GNR and 12% VRE on keyboards [10,26,27]. Even when hand hygiene is strictly maintained before and after contact with a patient, washing hands after each piece of medical or electronic equipment is touched would impede patient care to a degree of impossibility. Various researchers have found that hands (even gloved hands) with which one might touch instruments can immediately become contaminated with VRE in a substantial proportion of cases [16,17]. Poor hand hygiene compliance can only potentiate this possibility [28].

would impede patient care to a degree of impossibility. Various researchers have found that hands (even gloved hands) with which one might touch instruments can immediately become contaminated with VRE in a substantial proportion of cases [16,17]. Poor hand hygiene compliance can only potentiate this possibility [28]. The significant levels of contamination with clinically significant organisms in our study (25% of instruments in the emergency room) and others suggest that the possibility of colonization of other personnel, subsequent patients, and other equipment through the hands of healthcare workers and patients is of concern, given the propensity of clean hands to become re-colonized after contact with a contaminated piece of equipment while caring for a particular patient. Given the varied nature of the population that visits emergency departments and clinics, including immune-suppressed patients and patients without intact skin barriers, transmission of such organisms from contaminated equipment ideally should be avoided. Various (and virtually any) organisms can cause infection if they come into contact with a susceptible patient [29] and this is more likely in the emergency or ambulatory setting. The SUDS appears to provide a mechanism for eliminating equipment contamination, thereby preventing non-critical equipment from serving as an incubator and transmitter of clinically significant microorganisms.

e into contact with a susceptible patient [29] and this is more likely in the emergency or ambulatory setting. The SUDS appears to provide a mechanism for eliminating equipment contamination, thereby preventing non-critical equipment from serving as an incubator and transmitter of clinically significant microorganisms. There is scant literature on the incidence/prevalence of equipment contamination in areas where community and hospital mix in a rapid and high-turnover situation with a wide variety of patients of varying clinical complexity and needs that then disperse to a large number of different destinations, as represented by the ED in our study. Most studies to date on contamination rates of hospital equipment have been conducted in the inpatient setting [7,8,11-13,22,24], particularly the intensive care unit (ICU). However, the patient turnover in a single ED bed space (with its associated cables, stands, and electronics) compared to a single ICU bed is over 20 to 1 per year and 10 to 1 in the surgery clinic compared to an ICU bed (based on annual volumes for 2006). Studies evaluating the use of chemical and ultraviolet decontamination of equipment and environmental surfaces have required emptying and sealing off the room during the decontamination process, which is currently prohibitive in high-turnover clinical areas [30-32]. To our knowledge, this is the first study to evaluate the incidence of contamination of hospital equipment and electronics and a targeted automated intervention for same in this high-volume, high-turnover clinical setting without interruptions to patient thoroughfare.

rohibitive in high-turnover clinical areas [30-32]. To our knowledge, this is the first study to evaluate the incidence of contamination of hospital equipment and electronics and a targeted automated intervention for same in this high-volume, high-turnover clinical setting without interruptions to patient thoroughfare. When done in a meticulous fashion, manual decontamination has been shown to significantly decrease colonization rates in parts of certain instruments [9]; however, there are practical limitations to efficient manual decontamination in day-to-day practice. Hospital equipment is of variable configuration, leading to difficulty with manual decontamination. Adequately training personnel to clean specific devices optimally is resource-intensive. Further, some equipment is heavy or awkward; for example, many stands have a center of gravity that predisposes them to tipping over easily, increasing the propensity for accidents involving staff and, potentially, patients when equipment is tipped to the side in order to thoroughly decontaminate it. With the increasing number of patients cared for in ambulatory settings and the high turnover rate, efficient and effective manual decontamination is simply time-challenged. In a global, systemic review of all published and unpublished studies evaluating levels of contamination of healthcare equipment spanning a 32-year period, Schabrun et al. found up to 17 different organisms on a range of medical equipment [9]. This suggests that current decontamination protocols are ineffective or not adhered to by healthcare professionals. The current trend is to utilize more and more non critical disposables. Given all the above, it is clear that alternative approaches to this issue are warranted unless we expect that all hospital equipment, no matter how complex, will become disposable in the future, accepting the exponential costs both in cash and kind (environment).

d is to utilize more and more non critical disposables. Given all the above, it is clear that alternative approaches to this issue are warranted unless we expect that all hospital equipment, no matter how complex, will become disposable in the future, accepting the exponential costs both in cash and kind (environment). The SUDS (figures 4 and 5) provides a multimodal self-cleaning means of decontaminating a wide variety of equipment for staff in high-volume, high-turnover clinical settings. In this study, we showed that the SUDS eliminated clinically significant organisms after a single cycle. The manpower needs were minimal, and the equipment was shown to be free of clinically significant organisms 48 hours later, even after routine use in the emergency department. Automating the process of decontamination of most equipment could potentially allow increased attention to areas closest to patients, such as bedrails, that are likely to be contaminated but are not amenable to detachment and automated decontamination. This in turn could enhance overall compliance and culture change. Automation allows for the maintenance of consistency in the decontamination process, in addition to more effective and targeted dispersal of biocide via rotational, turbulence, and image-based means. Obviously, bacteria cannot be totally eliminated from the environment but equipment in this preliminary study showed that they were free from CSO by culture 48 hours later, having been put through normal use in the emergency room in the 48 hours since commencement of the study.

ional, turbulence, and image-based means. Obviously, bacteria cannot be totally eliminated from the environment but equipment in this preliminary study showed that they were free from CSO by culture 48 hours later, having been put through normal use in the emergency room in the 48 hours since commencement of the study. Figure 4 Photo of the Self-cleaning Unit for the Decontamination of Small instruments. Figure 5 Photo of the Self-cleaning Unit for the Decontamination of Small instruments. Limitations We postulate that the main advantage of the SUDS is the dispersal of the biocide over the entire piece of equipment, irrespective of its type. However, further, more robust studies will need to be carried out to define this further, in view of the differences in the nature of the biocides utilized in this study.

ations We postulate that the main advantage of the SUDS is the dispersal of the biocide over the entire piece of equipment, irrespective of its type. However, further, more robust studies will need to be carried out to define this further, in view of the differences in the nature of the biocides utilized in this study. This study assessed the incidence of colonization of portable and electronic equipment as a group. Further studies would be needed to assess the precise nature and incidence of colonization of individual pieces of equipment within the same ambulatory or emergency setting. The study only assessed re-colonization rates 48 hours after SUDS and therefore cannot comment on resistance of devices to re-colonization beyond 48 hours. Additionally, the 48-hour time frame would be dependent upon the combination and duration of levels of decontamination utilized in a particular cycle. We did not specifically quantify the difference/impact of patient flow between manual decontamination and utilization of the SUDS in this study; however, patient care continued without interruption during the run of this trial in the emergency setting. Only patient care areas and bed spaces were evaluated, and this may lead to an imprecise estimation of the incidence of colonization. Equipment in the central stations in the emergency room was not included in this study, also possibly leading to underestimation. As this study primarily was aimed at evaluating elimination of equipment contamination, neither contamination of environmental surfaces nor levels of hand hygiene were evaluated as part of its scope. However, the tested equipment in this study remained free of clinically significant organisms after 48 hours of subsequent use in a busy emergency area, regardless of the levels of other confounders such as hand hygiene and environmental contamination.

ces nor levels of hand hygiene were evaluated as part of its scope. However, the tested equipment in this study remained free of clinically significant organisms after 48 hours of subsequent use in a busy emergency area, regardless of the levels of other confounders such as hand hygiene and environmental contamination. There are multiple other issues that require further investigation. The precise incidence of infection of patients after contact with contaminated equipment or the colonized hands of healthcare workers is not precisely known. The degree to which the environment (which includes hospital equipment in some studies) is implicated in the direct transmission of organisms resulting in NI (nosocomial infections) has been debated. Hand hygiene compliance and surface environmental decontamination have been confounding variables affecting precise quantification in many studies and are worthy of further study. However, standard of care currently recommends decontamination of non-critical equipment.

ng in NI (nosocomial infections) has been debated. Hand hygiene compliance and surface environmental decontamination have been confounding variables affecting precise quantification in many studies and are worthy of further study. However, standard of care currently recommends decontamination of non-critical equipment. The impact of automated decontamination on overall cost compared with that of disposable equipment should be the subject of future analyses. Frequently in the case of disposable pulse oximeters, the quality is cited as a reason for it to be made as a disposable item. The quality is a distant secondary concern because we can make better reusable equipment: unfortunately, there is no incentive to do so, as detailed by Stewart almost 13 years ago [33]. If millions of dollars in a single hospital are spent on, for example, disposable pulse oximeter sensors, we should be able to quantify the number and types of infections disposable pulse oximeter sensors have prevented. And if we don't know the answer to those questions, shouldn't we be asking ourselves: why not?

33]. If millions of dollars in a single hospital are spent on, for example, disposable pulse oximeter sensors, we should be able to quantify the number and types of infections disposable pulse oximeter sensors have prevented. And if we don't know the answer to those questions, shouldn't we be asking ourselves: why not? Summary There is scant data evaluating contamination of non-critical equipment in the emergency room setting, which differ from the inpatient setting in volume, turnover rate, and variety of clinical conditions seen. Contamination of equipment has been reported in various inpatient settings, and transmission to multiple patients is a possibility even with strict hand hygiene before and after individual patient contact, as evidenced by nosocomial outbreaks linked to contaminated equipment. Our study showed that, with current manual decontamination methods, significant levels of contamination were identified, serving as a potential vector for transfer of organisms to personnel and subsequent patients. The SUDS provides an effective, efficient, and user-friendly multimodal method for decontaminating a wide range of hospital equipment in the clinical area in an automated fashion that is far superior to the current standard of manual decontamination. It can, therefore, potentially be an effective mechanism for decontamination of non-critical disposables and may provide an alternative to the use of disposables, particularly as the SUDS can decontaminate a wide variety multi-modal and multi-configurational equipment.

to the current standard of manual decontamination. It can, therefore, potentially be an effective mechanism for decontamination of non-critical disposables and may provide an alternative to the use of disposables, particularly as the SUDS can decontaminate a wide variety multi-modal and multi-configurational equipment. Competing interests The authors declare that they have no competing interests. Authors' contributions BA conceived of the study, designed and carried out the study, as well as prepared and edited the manuscript. CO performed the cultures and data collection and assisted in drafting the manuscript. TKC contributed to the design of the study, assisted with drafting the manuscript, edited the manuscript and assisted with microbiologic analysis. AA contributed to the design of the study, edited the manuscript, and assisted with microbiologic analysis. TR carried out the microbiologic analysis. BA, WA, RH, REC, and CC all contributed to the engineering works on the device. PH and RR contributed to the design of the study, and edited the manuscript. BS contributed to the design of the study and carried out the statistical analysis. All authors read the final manuscript. Acknowledgements • Dr. Trish Perl, Department of Infectious Diseases, Division of Hospital Epidemiology, Johns Hopkins Medical Institutions • Roy Shipley, Clinical Engineering Services, Johns Hopkins Medical Institutions

Introduction The advances in the experience with endoscopic video-assisted techniques have led to the introduction of dry laboratories by several centres. MIS requires training, crediting and caseload, which are not always available in small centres. Moreover, since it is a technologic surgery, new instrumentation need to be tested in a non-clinical environment and eventually discussed with technicians, particularly with bio-medical engineers. In this field an increase in the use of dry and wet surgical laboratories has been observed. A dry laboratory (dry lab) is specific to work with dry stored materials, electronics and/or large instruments, where, unlike wet laboratories, biological tissues (living or dead) are not utilized. These dry labs have equipped workstations for practicing endoscopic techniques, in a realistic setting, on phantoms and organ models. These environments are considered essential tools to design a minimally invasive surgery training program, gaining, certifying and improving the expertise. Many institutions, universities, centres of excellence are equipped with dry labs in order to provide a teaching, training and evaluation instrument of the fundamental technical skills for surgeons in a laboratory setting [1].

minimally invasive surgery training program, gaining, certifying and improving the expertise. Many institutions, universities, centres of excellence are equipped with dry labs in order to provide a teaching, training and evaluation instrument of the fundamental technical skills for surgeons in a laboratory setting [1]. A dry lab provides laparoscopic simulators such as box trainers and computer-based reality platforms [1]; the objective of the simulator-based curricula is the attainment of basic skills while working in a cheap and pressure-free environment, in order to transfer them from the laboratory to the clinical experience, and to allow the trainees to focus on more complex issues in the operating room. We describe a teaching and training method with objective evaluation to improve the medical engineering students' knowledge on Minimally Invasive Surgery (MIS) instrumentation and techniques through hands-on experience, in order to allow the students to analyze and criticize the devices they tested. Training has been scheduled during a three months course. Methods The course was kept at the "Tor Vergata" University of Rome in a laboratory setting. It involved 20 medical engineering students without any experience in laparoscopy.

We describe a teaching and training method with objective evaluation to improve the medical engineering students' knowledge on Minimally Invasive Surgery (MIS) instrumentation and techniques through hands-on experience, in order to allow the students to analyze and criticize the devices they tested. Training has been scheduled during a three months course. Methods The course was kept at the "Tor Vergata" University of Rome in a laboratory setting. It involved 20 medical engineering students without any experience in laparoscopy. The students were trained during hands-on sessions of the duration of three hours throughout a three-months course. During the experience, the students underwent a theoretical and practical course. The theoretical section consisted of review lectures and discussions about the general background of the procedure, and about selection criteria for the tasks. In addition, an introduction to the instrumentation materials was given. After the theoretical section, the students were allowed to exercise on the box trainer. The laparoscopic simulator consisted of a traditional pelvic trainer. Hands-on exercises were performed on a simulator in order to assess whether specific training exercises were helpful for medical engineering students to familiarize with surgical devices and techniques.

The students were trained during hands-on sessions of the duration of three hours throughout a three-months course. During the experience, the students underwent a theoretical and practical course. The theoretical section consisted of review lectures and discussions about the general background of the procedure, and about selection criteria for the tasks. In addition, an introduction to the instrumentation materials was given. After the theoretical section, the students were allowed to exercise on the box trainer. The laparoscopic simulator consisted of a traditional pelvic trainer. Hands-on exercises were performed on a simulator in order to assess whether specific training exercises were helpful for medical engineering students to familiarize with surgical devices and techniques. On the simulator, during the practical section of the course, every student was asked to perform a sequence of standardized drills three times in order to improve their dexterity, deep perception, instrument-targeting accuracy, visual and spatial abilities, and hand-eye coordination; time to complete the task was recorded in seconds.

r, during the practical section of the course, every student was asked to perform a sequence of standardized drills three times in order to improve their dexterity, deep perception, instrument-targeting accuracy, visual and spatial abilities, and hand-eye coordination; time to complete the task was recorded in seconds. In more detail, the exercises consisted of: 1) clipping and cutting (a plastic 5 mm diameter tube, marked at its half, was placed in the pelvic trainer; the students had to put a clip before and after the mark and then cut the tube on the mark, so as to simulate clipping and cutting of a vessel); 2) cutting a paper (a square 10 cm × 10 cm paper sheet was placed in the pelvic trainer; the students had to cut it along one of the diagonals, so as to simulate cutting with laparoscopic scissors); 3) positioning a needle (a rubber parallelepiped was placed and fixed in the pelvic trainer; the students had to simulate the positioning of a needle using needle-holders); 4) intra-corporeal knotting (the students had to simulate an intra-corporeal knot on the above rubber parallelepiped); 5) carrying out specimen with EndoBag (the students had to cut a piece of the above rubber parallelepiped and to carry it out using an EndoBag™ (Autosuture™, Norwalk, CT, USA) specimen retrieval system).

intra-corporeal knotting (the students had to simulate an intra-corporeal knot on the above rubber parallelepiped); 5) carrying out specimen with EndoBag (the students had to cut a piece of the above rubber parallelepiped and to carry it out using an EndoBag™ (Autosuture™, Norwalk, CT, USA) specimen retrieval system). After this practical activity, the students were divided into 4 groups and asked to choose an instrument they utilized, to analyze it and to write and discuss a short essay about it, in which they had to describe the instrument (application, purpose, components, functionality, manufacturing techniques and materials) and had to criticize it, pointing out advantages and disadvantages of each instrument and eventually proposing new solutions in order to improve the instrument itself. The devices chosen by the four groups of students were [2]: • Roticulator Endo Mini-Shears™ (Autosuture™, Norwalk, CT, USA) • Blunt tip trocar with balloon tip (Autosuture™, Norwalk, CT, USA) • DST Series™ GIA™ 60 (Autosuture™, Norwalk, CT, USA) • EndoClinch™ II (Autosuture™, Norwalk, CT, USA) Statistics Student t test was used to compare the times required to perform the first vs. the second and the last sequence of drills in order to test the efficacy of training course. All data are exposed in terms of mean ± standard error (SEM). A value of p < .05 was considered statistically significant.

• EndoClinch™ II (Autosuture™, Norwalk, CT, USA) Statistics Student t test was used to compare the times required to perform the first vs. the second and the last sequence of drills in order to test the efficacy of training course. All data are exposed in terms of mean ± standard error (SEM). A value of p < .05 was considered statistically significant. Results Using the simulator, the overall time required to perform the first time the sequence of drills was higher if compared to one required to perform the second and the third ones (clipping and cutting: 65 ± 27 vs. 48 ± 19 vs. 26 ± 8 sec; cutting a paper: 309 ± 158 vs. 152 ± 59 vs. 145 ± 72 sec; positioning a needle: 163 ± 176 vs. 80 ± 32 vs. 75 ± 37 sec; intra-corporeal knotting: 117 ± 97 vs. 77 ± 50 vs. 61 ± 45 sec; carrying out specimen with EndoBag: 48 ± 26 vs. 33 ± 13 vs. 30 ± 15 sec). The difference was statistically significant between the first and the second time performance for every task (p < .05). Moreover, the second performance required a longer time than the third; the difference was close to significance (p = .05) (fig. 1). Figure 1 Time needed to complete the tasks. The comparison between the first and the third performance times was statistically significant for every task (p < .05). The students proposed the following improvements and criticisms about the devices they analyzed: • Roticulator Endo Mini-Shears™ (Autosuture™, Norwalk, CT, USA) Ergonomics 1) The handle causes traumas and local paralysis of fingers. 2) Shaft articulation can be activated only by opposite hand to the hand that holds the device.

The comparison between the first and the third performance times was statistically significant for every task (p < .05). The students proposed the following improvements and criticisms about the devices they analyzed: • Roticulator Endo Mini-Shears™ (Autosuture™, Norwalk, CT, USA) Ergonomics 1) The handle causes traumas and local paralysis of fingers. 2) Shaft articulation can be activated only by opposite hand to the hand that holds the device. Improvements 1) Building of a more ergonomic handle (fig. 2). Figure 2 A more ergonomic handle for the Roticulator Endo Mini-Shears™. 2) Replacement of the grooved collar for articulation with a second scalloped dial located on the handle, to be activated by the same hand that holds the device (fig. 3). Figure 3 Proposed position for the second gear on the Roticulator Endo Mini-Shears™. Single use instrument Problems: short timing life; smaller safety for patient; more difficulty of management (price, stocking and waste disposal); worse services (precision of cut, electronic support). Improvements Use of sterilised handle, in which electronic controls can be located, and single use shaft connecting to handle (fig. 4). Figure 4 Proposal for an electronic handle for the Roticulator Endo Mini-Shears™. Electric conduction Breaking of insulating layer on the shaft; formation of conductive water layer on shaft. Improvements Replacement of the terminal part of insulating sheath with an insulating absorbent protective coating; replacement of the terminal linear shaft with a holding out coating, in order to avoid contact between active blade of scissors and shaft (fig. 5).

Electric conduction Breaking of insulating layer on the shaft; formation of conductive water layer on shaft. Improvements Replacement of the terminal part of insulating sheath with an insulating absorbent protective coating; replacement of the terminal linear shaft with a holding out coating, in order to avoid contact between active blade of scissors and shaft (fig. 5). Figure 5 Improvement proposal for the terminal portion of the Roticulator Endo Mini-Shears™. • Blunt tip trocar with balloon tip (Autosuture™, Norwalk, CT, USA) The built-in adapters could break during their shifting on the entry port; the adapters aren't universal. Improvements Replacement of the built-in adapters system with a more stable and versatile diaphragm-like system, controlled by an external gear. • DST Series™ GIA™ 60 (Autosuture™, Norwalk, CT, USA) 1) The device is heavy, doesn't have an easy handling (first of all in the case of operators with small hands, like women), in particular when positioning the device on the tissue. 2) Firing clips needs too much strength. Improvements 1) Addition of a ring on the external handle for the placement of the operating hand's thumb, in order to give a more correct and firmer handling of the device (fig. 6). Figure 6 Improvement proposal for the DST Series™ GIA™ 60. 2) Reduction of friction and/or addition of a pneumatic firing system. • EndoClinch™ II (Autosuture™, Norwalk, CT, USA) 1) The handle of the instrument is too tough, and a prolonged use may cause little pain to the surgeon's hands. 2) The drive that controls the ratchet is not very handy, either when turning on or turning off.

2) Reduction of friction and/or addition of a pneumatic firing system. • EndoClinch™ II (Autosuture™, Norwalk, CT, USA) 1) The handle of the instrument is too tough, and a prolonged use may cause little pain to the surgeon's hands. 2) The drive that controls the ratchet is not very handy, either when turning on or turning off. Improvements 1) Application of small cushions on the handle itself, in order to give a more comfortable hold to the surgeon. 2) Placement of a sliding drive, similar to the drive that controls the ratchet, on the side of the instrument, near the gear (Fig. 7). Figure 7 Improvement proposal for the EndoClinch™ II.

2) The drive that controls the ratchet is not very handy, either when turning on or turning off. Improvements 1) Application of small cushions on the handle itself, in order to give a more comfortable hold to the surgeon. 2) Placement of a sliding drive, similar to the drive that controls the ratchet, on the side of the instrument, near the gear (Fig. 7). Figure 7 Improvement proposal for the EndoClinch™ II. Discussion The advances in the experience with endoscopic video-assisted techniques have led to the introduction of dry laboratories in several centres. MIS requires training, crediting and case load, not always available in small centres. Therefore dry-lab training courses are necessary for improving the technical skills necessary to apply novel techniques [3-8]. Moreover, surgeons need to create an interface with engineers, in order to better understand the capabilities of a surgical device, enhance existent instrumentation, conceive new and smarter devices and eventually collaborate in the development of new ideas. On February this journal published papers from a panel of expert entitled "Innovation by surgeons" [9-20]. Some of these interesting contributions underline aspects concerning relationship between surgical academic research, technology transfer and business. Particularly, in their contribution, Heller, Michelassi and Schuler [14] give a fundamental contribution on the innovation process in the surgical field, outlining how doing translational research is only possible supporting effective communication between surgeons and biomedical engineers and promoting intercampus initiatives.

rly, in their contribution, Heller, Michelassi and Schuler [14] give a fundamental contribution on the innovation process in the surgical field, outlining how doing translational research is only possible supporting effective communication between surgeons and biomedical engineers and promoting intercampus initiatives. This kind of initiatives allows surgical department to devise a new academic mission: the technology transfer, in addition to teaching, research and clinical activities. This activity brings to production of intellectual property and eventual commercialization. In this field, the interface between surgeons and engineers leads naturally to innovation processes, which mean clinical results, fundraising and prestige for the academic institution. A method has to be identified in order to develop a continuative collaboration: for example, having engineers participate to the operating room activities, analyzing needs, having a systematic approach to solutions, prototyping, running pre-clinical and clinical tests and possibly commercializing the product. Such an activity helps to overcome problems in the interdisciplinary research, letting brilliant ideas become real products. For these reasons, we decided to test a group of medical engineering students in order to improve their practical and theoretical knowledge and their analysis skills about MIS instrumentation, and to analyze them while performing standardized MIS tasks on a pelvic trainer and eventually while presenting their essays about the devices they choose to analyze and criticize.

cal engineering students in order to improve their practical and theoretical knowledge and their analysis skills about MIS instrumentation, and to analyze them while performing standardized MIS tasks on a pelvic trainer and eventually while presenting their essays about the devices they choose to analyze and criticize. We used a simulator since it allows for practice at various levels [4]. Rosser et al [5] and Chung et al [6] were the first authors who devised standardized drills and measured laparoscopic skills using timing alone as an endpoint. In order to standardize our experience, we decided to assess trainees' skills with the same approach. The student's performance was evaluated through the five following tasks: 1) clipping and cutting; 2) cutting a paper; 3) positioning a needle; 4) intra-corporeal knotting; 5) carrying out specimen with EndoBag. During this standardized hands-on experience, the students had the possibility to get comfortable with MIS devices. Such an experience is necessary in order to allow students to understand the advantages and shortcomings of their utilization during MIS procedures. We had the possibility to appreciate how medical engineering students worked on the devices they utilized, producing significant improvement proposals about the devices they tested during the course, together with a significant improvement in performance with increasing skillness, as could be expected. The work in the dry lab seemed to highly stimulate a critical and innovative approach to every single technological devices.

roducing significant improvement proposals about the devices they tested during the course, together with a significant improvement in performance with increasing skillness, as could be expected. The work in the dry lab seemed to highly stimulate a critical and innovative approach to every single technological devices. In conclusion, we believe that devising a dry lab course for medical engineering students may be useful and innovative for teaching and improving analysis and management of laparoscopic devices, allowing identification of problems and "brainstorming" work during the hands-on experience, with the aim of developing better devices [21,22]. Competing interests The authors declare that they have no competing interests. Authors' contributions FR: Designer and supervisor of the project. AM: Medical Engineer, tutor of the medical engineering students. GG: General Surgeon, tutor of the medical engineering students. GO: Responsible for statistical studies. MG: Responsible for statistical studies. AMF: Head of the Department of Surgery and responsible of the project. All authors read and approved the final manuscript. Acknowledgements The Authors thank Silvia Consalvi, Ilaria Di Nucci, Cristina Falcinelli, and Emanuela Girolami, medical engineering students at the University of the Study of Rome "Tor Vergata", for their contribution to this work.

Background The breast is one of the most important female organs. The breast has major implications in sexual arousal, as a result of its visual and sensual properties. Also most important feature of the breast is its capability of milk production. None of the plastic surgery operations put as much a heavy burden on plastic surgeons as reduction mammaplasty. Until today, many reduction mammaplasty techniques were described in the literature but the search for an ideal technique continues. In fact, surgical outcomes should not only fulfill patient expectations for an aesthetic appearance but also provide important breast functions. We believe that the most important philosophy in breast reduction surgery should be preservation of reliable neurovascular and lactational integrity to the nipple. Poor outcomes both affect the women undergoing operation and cause babies to feed on less breast milk, which is not acceptable. Attempts to find the ideal technique for reduction mammaplasty may end only if two secrets are resolved. One is to create the original geometry of the breast and the other is to provide maximum preservation of breast functions. The breast has a very complicated geometry. The complex geometry of the breast has been analysed with a three dimensional laser scanner in many studies [1]. The well-known definition of the breast is a cone horizontally projecting from the anterior thoracic wall [2]. Due to both effects of gravity and the nature of the breast tissue, the superior pole of the breast is a half a cone and the inferior pole is a half a globe (Figure 1). Andrades in a review on reduction mammaplasty techniques emphasized preservation and reconstruction of this cone shape [3]. Functional results of breast reduction are as important as its aesthetic results. The Surgeon General's health goals for 2010 are that 75% of woman initiates breastfeeding and that 50% continue it through 6 months postpartum [4]. Maximum preservation of breast functions depends on exact knowledge of anatomical features of the breast.

lts of breast reduction are as important as its aesthetic results. The Surgeon General's health goals for 2010 are that 75% of woman initiates breastfeeding and that 50% continue it through 6 months postpartum [4]. Maximum preservation of breast functions depends on exact knowledge of anatomical features of the breast. At present, vascularization and innervation of the nipple areola complex (NAC) has been clearly described and the vessels and the nerves have been shown to reach vertically the NAC at the fourth and fifth ribs through a separate fibrous septum [5,6]. Although there is both deep and superficial blood supply system, generally accepted that vascularization and innervation of the NAC is through the central breast parenchyma which can be seen as inferior to the breast shape in standing position. Using central pedicle preserves nerve supply together with the vascular supply to the maximal extent possible in breast reduction surgery [7]. If the glandular tissue is not removed with the central pedicle, then the patient keeps her lactation potential with good nipple sensation [8]. The principle underlying the technique described here is complete preservation of these tissues (Figure 2). Figure 1 Geometry of the breast. Upper part of breast is half cone and lower pole is half a globe. Figure 2 Vascular, neural and glandular anatomy of the breast. Dotted lines indicate the margins of the pedicle in our technique. Major vessels, nerves, and lactipherous ducts to the NAC and mammary gland are preserved in our technique.

Figure 1 Geometry of the breast. Upper part of breast is half cone and lower pole is half a globe. Figure 2 Vascular, neural and glandular anatomy of the breast. Dotted lines indicate the margins of the pedicle in our technique. Major vessels, nerves, and lactipherous ducts to the NAC and mammary gland are preserved in our technique. A larger pedicle does not necessarily achieve better breast functions. Vessels and nerves of the NAC should be completely preserved. As a matter of fact, a large pedicle may cause such complications as displacement and folding of the flaps [9]. So that the breasts look natural after reduction mammaplasty, it can move to all directions and has a soft texture. It has been emphasized that a gland connected to the ducts and the nipple should be preserved for a successful breastfeeding following breast reduction [10]. However, to our knowledge, there have not been any studies showing how much breast tissue should be preserved for sufficient milk production. Maintenance of lactation should never be disregarded. Therefore, a maximum amount of the gland should be preserved. In all techniques except for the central or total posterior pedicle, the pedicle is not based on the gland only. In fact, most of the candidates for reduction mammaplasty have a high BMI[10]; that is, they have fat tissue as much as breast tissue.

sregarded. Therefore, a maximum amount of the gland should be preserved. In all techniques except for the central or total posterior pedicle, the pedicle is not based on the gland only. In fact, most of the candidates for reduction mammaplasty have a high BMI[10]; that is, they have fat tissue as much as breast tissue. Although vertical scar mammaplasty techniques are quite popular now, the most frequently performed technique is inferior pedicled mammaplasty. Critics cite a longer operating time than for vertical mammoplasties, the known scar problems of the Wise inverted T pattern, and the phenomenon of pseudoptosis, often termed "bottoming out"[11]. In our technique, conical plication creates fullness in the superior pole which in turn leads to a juvenile look, the technique does not cause postoperative pseudoptosis which frequently occurs in mammaplasty techniques, central U shaped pedicle allows maximum preservation of functions and the technique is applicable in all breasts irrespective of their sizes. The techniques described in the literature generally were not suitable for very big breasts but our approach is a versatile breast reduction technique. It can be used for all kind of breasts. There is no limitation to the amount of NAC repositioning that can be achieved. In short, maximum preservation of functions and an aesthetic breast with minimum scar are the main goals of the COPCUs mammaplasty.

breasts but our approach is a versatile breast reduction technique. It can be used for all kind of breasts. There is no limitation to the amount of NAC repositioning that can be achieved. In short, maximum preservation of functions and an aesthetic breast with minimum scar are the main goals of the COPCUs mammaplasty. Methods This technique was a modification of the total posterior pedicled mammaplasty described by Moufarrege[12]. The most important feature of the technique was that the central U shaped pedicle was a total posterior pedicle. The "open sky" approach was used and all tissues were easily accessible. Thus, the desirable shape was given and maximum preservation of all anatomical structures was achieved. While central U shaped pedicle was being created, peripheral tissues were resected and posterior and superior connections of the pedicle were preserved completely. The pedicle directly carried the NAC and all vascular and neural connections of the pedicle were preserved.

ation of all anatomical structures was achieved. While central U shaped pedicle was being created, peripheral tissues were resected and posterior and superior connections of the pedicle were preserved completely. The pedicle directly carried the NAC and all vascular and neural connections of the pedicle were preserved. Surgical Technique The first stage of the procedure was marking. A preoperative marking which was quite simple and easy to apply in all patients was developed. As Moufarrege described, the marking was performed when the patient was seated. In order to preserve the axis of each breast crossing the nipple, the vertical axis crossing the nipple and paralleling the margins of the breast was identified and this axis did not have to cross the mid-clavicular line (Figure 3). After the axis of the breast was detected, the inframammarian fold was marked and the upper point of the keyhole pattern was determined. This point was the place where the inframmamarian fold was located (Figure 4). Inframammarian fold was marked (Figure 5). Next, the standard keyhole pattern was marked. Extending arms of the pattern had an angle of 90 degrees and each was 5 cm in length (Figure 6). Moufarrege classified breasts into three based on their size when marking the standard keyhole pattern. We increased the angle between the arms of the keyhole to 135 degrees only in cases of gigantomastia. A larger angle is not more advantageous. In fact, creating a larger angle requires harvesting more skin and causes tension on the suture line, which may lead to difficulties in healing. Arms of the keyhole 5 cm in length formed a curve 3 cm above the inframammarian fold (Figure 7). Then, a vertical pedicle 6 cm in length running the midline of the breast was marked. It extended to 2 cm above the NAC in the superior part and till the end of the marked area in the inferior part. Last, the periareolar area 5 cm in diameter was marked.

gth formed a curve 3 cm above the inframammarian fold (Figure 7). Then, a vertical pedicle 6 cm in length running the midline of the breast was marked. It extended to 2 cm above the NAC in the superior part and till the end of the marked area in the inferior part. Last, the periareolar area 5 cm in diameter was marked. Figure 3 Axis of the breast. Figure 4 Upper point of the keyhole pattern. Figure 5 Inframammarian fold marking. Figure 6 Drawing of the keyhole pattern. Figure 7 Pre-operative markings of the patient on supine position.

gth formed a curve 3 cm above the inframammarian fold (Figure 7). Then, a vertical pedicle 6 cm in length running the midline of the breast was marked. It extended to 2 cm above the NAC in the superior part and till the end of the marked area in the inferior part. Last, the periareolar area 5 cm in diameter was marked. Figure 3 Axis of the breast. Figure 4 Upper point of the keyhole pattern. Figure 5 Inframammarian fold marking. Figure 6 Drawing of the keyhole pattern. Figure 7 Pre-operative markings of the patient on supine position. The second stage was surgery. Patients were in the supine position with a slight flexion in the waist. The tumescent technique was used in all patients. After incisions appropriate for the markings were made, the skin on the pedicle was de-epithelized (Figure 8). Subsequently, skin flaps were undermined, starting in the medial. The breast including dermal fat was undermined from the gland to aponeurosis of the pectoralis major. At the end of undermining, the breast was completely exposed in the front view. Resection of the peripheral tissue started at the medial and continued at the lateral and at the inferior part minimally so as to create a 6 cm-U shaped pedicle in the middle (Figure 9). Resection margins in the inferior did not extend beyond the inframammarian fold and no resection was made in the superior. Unlike the posterior pedicle mammaplasty described by Moufarrege, the technique described here involved minimal resection in the inferior, which prevented excess in the horizontal part, and only one hole was created for drainage. Resection of the external quadrant extending to the subaxiallary region was performed gently and the areolar tissue in this area was preserved especially in cases of gigantomastia and extreme hypertrophy. After the resection was completed, a U shaped total posterior pedicle 6 cm in width remained in the middle (Figure 10). Following resection, conical plication was carried out to achieve superior fullness. Plication was performed in such a way to create a cone at the two o'clock and ten o'clock positions of the NAC with oblique continuous suture with 2/0 PDS (Figure 11 and 12) (Additional file 1). This plication is not dermal suspension, as seen in the video in additional file 1, areolar tissue, fat tissue and glandular tissue of breast are plicated. After conical plication was created, the breast was secured in its new position with temporary sutures running through inferior and superior parts of the NAC. One vertical suture was put 6 cm below the NAC and the area below this point was closed with pursing sutures.

tissue and glandular tissue of breast are plicated. After conical plication was created, the breast was secured in its new position with temporary sutures running through inferior and superior parts of the NAC. One vertical suture was put 6 cm below the NAC and the area below this point was closed with pursing sutures. The subdermis was closed with 3/0 PDS, the vertical incision with 4/0 PDS and the periareolar region with 5/0 PDS without tension. One drainage tube was placed and temporary sutures were removed at the end of the operation (Additional file 2). Only a short vertical scar appeared in all cases and reverse T incision was avoided. Pressure dressing was done at the end of the operation and the drain tubes were removed within two days of the operations. Figure 8 De-epithelisation of the breast. Figure 9 Resection of medical and lateral tissues. Figure 10 U shaped total posterior pedicle 6 cm in width. Figure 11 Plication of the upper part of the pedicle. Figure 12 View of the pedicle pre and post plicated. Superior fullness was created with this plication.

The subdermis was closed with 3/0 PDS, the vertical incision with 4/0 PDS and the periareolar region with 5/0 PDS without tension. One drainage tube was placed and temporary sutures were removed at the end of the operation (Additional file 2). Only a short vertical scar appeared in all cases and reverse T incision was avoided. Pressure dressing was done at the end of the operation and the drain tubes were removed within two days of the operations. Figure 8 De-epithelisation of the breast. Figure 9 Resection of medical and lateral tissues. Figure 10 U shaped total posterior pedicle 6 cm in width. Figure 11 Plication of the upper part of the pedicle. Figure 12 View of the pedicle pre and post plicated. Superior fullness was created with this plication. Methodology A detailed physical examination of the breast includes measurements of breast size, degree of ptosis, masses, superior pole fullness, nipple sternal distance, nipple-inframammarian fold distance were recorded. Anterior, lateral and two oblique photographs were taken to compare preoperative and postoperative superior fullness in controls of patients routinely (figure 13 and 14). Projections of the breasts were evaluated according to the lateral photographs of the patients (Figure 15). Superior fullness was evaluated with measurement of the breast projection. Also, weights of the resected tissue were recorded. Figure 13 Pre and post operative view of the patient operated with COPCUs mammaplasty for reduction of the breast. 360 gr tissues were removed from each breasts.

Methodology A detailed physical examination of the breast includes measurements of breast size, degree of ptosis, masses, superior pole fullness, nipple sternal distance, nipple-inframammarian fold distance were recorded. Anterior, lateral and two oblique photographs were taken to compare preoperative and postoperative superior fullness in controls of patients routinely (figure 13 and 14). Projections of the breasts were evaluated according to the lateral photographs of the patients (Figure 15). Superior fullness was evaluated with measurement of the breast projection. Also, weights of the resected tissue were recorded. Figure 13 Pre and post operative view of the patient operated with COPCUs mammaplasty for reduction of the breast. 360 gr tissues were removed from each breasts. Figure 14 Pre and post operative view of the patient with gigantomastia. 1320 gr tissue were removed from each breasts. Figure 15 Pre and postoperative lateral view of the patients a1-2: Mastopexy b1-2: Moderate breast hypertrophy c1-2: Marked breast hypertrophy d1-2: Gigantomastia. For quality scar evaluation we used visual analog scale. According to scale 0 was unacceptable scar must be corrected surgically and 10 was excellent scar. Patients were evaluated their scars after 6 months post-operatively. The patients were followed minimum six months post-operatively (6 months – 36 months). Written informed consent was obtained from the patient for publication of this case report and accompanying images. A copy of the written consent is available for review by the Editor-in-Chief of this journal

For quality scar evaluation we used visual analog scale. According to scale 0 was unacceptable scar must be corrected surgically and 10 was excellent scar. Patients were evaluated their scars after 6 months post-operatively. The patients were followed minimum six months post-operatively (6 months – 36 months). Written informed consent was obtained from the patient for publication of this case report and accompanying images. A copy of the written consent is available for review by the Editor-in-Chief of this journal Results Our technique was performed in 46 patients. The median age of the patients was 24 years, ranging from 17 years to 63 years. Data of patients are presented in Table 1. The median distance between the jugular notch to nipple was 31 cm, ranging between 22 cm and 48 cm. The mean resected tissue weight per breast was 564 gr, ranging from 273 gr to 1880 gr. Out of 46 patients included in the study, 38 underwent reduction mammaplasty (7 had gigantomastia), four mastopexy and four revision surgery. Out of four patients undergoing revision surgery, one had had inferior pedicle mammaplasty and three had had superior pedicle vertical scar mammaplasty before. All four patients had revision surgery for pseudoptosis. Table 1 Data of the patients: Range Average Age 17–63 24 (median) Jugular notch to nipple distance Pre-operative 22–48 cm. 31 cm. (median) Post-operative 19–23 cm. 21 cm. (median) Nipple to inframammarial crease Pre-operative 8–18 cm. 14.5 cm. Post-operative 7–12 cm. 8 cm. Breast projection Pre-operative 21–54 mm. 31 mm. Post-operative 42–60 mm. 54 mm

Table 1 Data of the patients: Range Average Age 17–63 24 (median) Jugular notch to nipple distance Pre-operative 22–48 cm. 31 cm. (median) Post-operative 19–23 cm. 21 cm. (median) Nipple to inframammarial crease Pre-operative 8–18 cm. 14.5 cm. Post-operative 7–12 cm. 8 cm. Breast projection Pre-operative 21–54 mm. 31 mm. Post-operative 42–60 mm. 54 mm Resection weight(per breast) 273–1880 gr 564 gr Follow up 6–36 months 12 months None of the patients had such major complications as necrosis, partial or total NAC loss. None of them required revision surgery. Two days after removal of the drainage tubes, two patients had minimal hematoma, which was treated conventionally. Two patients had about 1 cm opening on the NAC and suture line, but they healed spontaneously. Quality of the scars were scored by the patients themselves and median was 9, ranging with 7 to 10. All patients were satisfied with aesthetic results. We never performed scar revision surgery. Although we did not make investigational laboratory studies for the sensation of the nipple, none of the patients reported decreased sensual or sexual sensibility in the short term and long term. Six patients gave birth after operation and none of them had lactation difficulties according to their experience. According to their history, they did not need supplementation in first 6 months of breastfeeding.

ple, none of the patients reported decreased sensual or sexual sensibility in the short term and long term. Six patients gave birth after operation and none of them had lactation difficulties according to their experience. According to their history, they did not need supplementation in first 6 months of breastfeeding. Discussion Reduction mammaplasty is one of the most frequently performed operations by plastic surgeons. There have been numerous modifications of reduction mammaplasty [13]. It may be that the breast has a very complex geometric structure and very different anatomical components. Reduction mammaplasty techniques described so far are named after locations of pedicles. Among them are inferior, lateral, media, central, total posterior pedicle and mixed [14-18].

of reduction mammaplasty [13]. It may be that the breast has a very complex geometric structure and very different anatomical components. Reduction mammaplasty techniques described so far are named after locations of pedicles. Among them are inferior, lateral, media, central, total posterior pedicle and mixed [14-18]. The leading cause of ongoing attempts to seek an ideal technique is complications such as failure to achieve the desirable aesthetic result, decreased or lack of lactation, decreased or loss of sensual and erogenous feeling of the nipple, insufficient projection and postoperative pseudoptosis and wound healing problems. Ultimate goal of any pedicle is to provide sufficient blood supply to the nipple areola complex[3]. It has been reported that superior pedicled mammaplasty causes considerable changes in blood circulation due to the transposition of the pedicle and that there is decreased NAC sensation in the superior pedicle in the short term. The nerves innervate the NAC can be easily injured with inferior pole resections with superior pedicle techniques [19,20]. Bottoming out, inferior pole excess or pseudo ptosis, is more frequent in inferior based pedicles [21]. Attempts to seek reduction mammaplasty techniques preserving the NAC emerged from the results of the studies by Bisenberger[22]. However, they revealed considerably high rates of complications. An effective way to prevent complications is to know the breast anatomy well, especially the vascular anatomy of the breast.

. Attempts to seek reduction mammaplasty techniques preserving the NAC emerged from the results of the studies by Bisenberger[22]. However, they revealed considerably high rates of complications. An effective way to prevent complications is to know the breast anatomy well, especially the vascular anatomy of the breast. The classical pattern of blood supply to the breast was first described by Manchot in 1889 and later, Marcus showed that the breast had three different patterns of blood supply [23]. The most recent and extensive study on the issue was performed by Wuringer in 1998[6]. Wuringer examined 28 female breasts and described a horizontal septum extending from fifth rib toward the nipple, dividing the blood supply into a cranial and caudal network. Deventer on 27 cadaver dissections in 2004 demonstrated that the blood flow from the nipple had a quite different pattern and that each breast of the same person might have had differences in blood flow from the nipples[23]. These findings indicated that the tissue below the nipples, especially the horizontal septum should be preserved.

aver dissections in 2004 demonstrated that the blood flow from the nipple had a quite different pattern and that each breast of the same person might have had differences in blood flow from the nipples[23]. These findings indicated that the tissue below the nipples, especially the horizontal septum should be preserved. Nipple necrosis is the most frightening complication of reduction mammaplasty. The rates of nipple necrosis have been reported to be 2.1% in the superodermal pedicle[24], 2.3% in the superolateral[25] and 0.8% in inferior pedicle[26]. The leading cause of nipple necrosis is insufficient arterial blood supply or long-lasting venous congestion; this can be attributed to inadequate knowledge about the vascular anatomy of the NAC and use of long peripheral pedicle and the resultant distortion of the pedicle. However, theoretically total posterior pedicle described by Moufarrege and its modification COPCUs mammaplasty may avoid such complications.

us congestion; this can be attributed to inadequate knowledge about the vascular anatomy of the NAC and use of long peripheral pedicle and the resultant distortion of the pedicle. However, theoretically total posterior pedicle described by Moufarrege and its modification COPCUs mammaplasty may avoid such complications. Although the preoperative marking described in the present study resembles the bipedicle modification described by Khan in 2007, it differs in resection and pedicle considerably from that described by Khan[27]. It was inevitable that the horizontal diameter in bipedicled and total central pedicled techniques was larger than expected; fortunately, U shaped pedicle resolved this issue. In addition, two dead spaces -one in the medial and the other in the lateral- remain in bipedicled techniques and fluid and blood accumulate in these spaces in cases of insufficient drainage; however, in U shaped pedicle all spaces are connected to each other and one effective drainage tube allows collection of fluid and blood. Khan and Moufarrege reported that the biggest advantage of their techniques was preservation of blood supply to the nipple and easy mobilization of the breast towards the superior without loss of viability in the breast tissue. However, U shaped pedicle was more advantageous since it was connected to the NAC both in the superior and in the inferior.

at the biggest advantage of their techniques was preservation of blood supply to the nipple and easy mobilization of the breast towards the superior without loss of viability in the breast tissue. However, U shaped pedicle was more advantageous since it was connected to the NAC both in the superior and in the inferior. At present, there is a general agreement that the most popular techniques are vertical scar mammaplasty and its modifications. Due to high rate of complications (especially wound healing, seroma) in vertical scar mammaplasty, many modifications were presented literature, but use of vertical scar technique for large breasts still is not widely accepted, especially with use of superior pedicle[28]. Rohrich et al presented results of a survey in 2004[29]. According to this report assessing the trends in breast reduction techniques among the members of the American Society for Aesthetic Plastic Surgery found that the most frequent complications for the limited incision technique group were suture spitting, the need for surgical revision and the loss of nipple sensation[28]. Advantage of the superior pedicle technique is an improved superior projection and a stable long-term shape of the breast as compared with the inferior pedicle techniques[30]. But the use of a superior pedicle supposedly increases the rate of areolar necrosis[28]. The techniques providing most effective blood supply to the breast are those with a central pedicle. The central pedicle technique in current use is the end result of serial modifications of Biesenberger's technique[22]. Balch[31] and Hester[32] used the classical T incision, Peixoto[33] and Hagerty [34] used vertical incision, Yousif[35] and Lalonde[36] used horizontal scar and Goes [37] used periareolar technique for central pedicle. Total posterior pedicle was described by Moufarrege[12,18]. It was called total posterior pedicle since the pedicle was just behind the NAC and the whole posterior pedicle was made of the gland. Total posterior pedicle achieved maximum gland and nipple security and Moufarrege reported low rates of complications in more than 10000 patients undergoing reduction mammaplasty. Moreover, none of them were major complications. Although Moufarrege technique was used of the severe gynecomastia[38], Moufarrege did not recommend total posterior pedicle for large breasts.

nipple security and Moufarrege reported low rates of complications in more than 10000 patients undergoing reduction mammaplasty. Moreover, none of them were major complications. Although Moufarrege technique was used of the severe gynecomastia[38], Moufarrege did not recommend total posterior pedicle for large breasts. The complications of the central breast reduction techniques are few and most of these are related to the inverted T-shsped scar, which is often unsatisfactory from the aesthetic point of view[39]. Since we used vertical scar, we did not have scar problem and never revision surgery was performed for the scar. The management of gigantomastia is still debatable. Many authors propose that nipple areola graft can be utilized for the management of gigantomastia. In the present study, none of the patients with gigantomastia required grafting. Location of the NAC on a fully preserved posterior pedicle obviated the use of grafts.

The complications of the central breast reduction techniques are few and most of these are related to the inverted T-shsped scar, which is often unsatisfactory from the aesthetic point of view[39]. Since we used vertical scar, we did not have scar problem and never revision surgery was performed for the scar. The management of gigantomastia is still debatable. Many authors propose that nipple areola graft can be utilized for the management of gigantomastia. In the present study, none of the patients with gigantomastia required grafting. Location of the NAC on a fully preserved posterior pedicle obviated the use of grafts. It is of great importance to maintain the breast shape for a long time after reduction mammaplasty. However, gravity and tissue dynamics make it difficult. Recurrent ptosis may be a problem in all breast reduction techniques. Hammock technique[40], dermis strips[41], synthetic materials[42], fascia lata[43] and internal bra systems [44] have been used to eliminate recurrent ptosis. However, all techniques are based on the idea that the pedicle, like a suspensory ligament, should be suspended from the thorax wall. Around the areola, and especially below the areola, an area of de-epithelialised skin is preserved to be used as a bra-like support of the gland[45]. Dermis suspension gives a well-defined shape intra-operatively, which does not change significantly with time.

pensory ligament, should be suspended from the thorax wall. Around the areola, and especially below the areola, an area of de-epithelialised skin is preserved to be used as a bra-like support of the gland[45]. Dermis suspension gives a well-defined shape intra-operatively, which does not change significantly with time. The philosophy is completely different in the technique presented here. De-epithelized area interacts with the above skin and thus helps to preserve conical plication and decreases effects of gravity. As far as we know, conical plication has not been described in the literature before. The conical plication which we developed is directed towards preservation of the juvenile breast look and superior fullness in the long term.

he above skin and thus helps to preserve conical plication and decreases effects of gravity. As far as we know, conical plication has not been described in the literature before. The conical plication which we developed is directed towards preservation of the juvenile breast look and superior fullness in the long term. In 1985, Pennington performed plication and pedicle suspension in the pectoral fascia to prevent bottoming out, a frequently encountered complication of inferior pedicle, and reported his 20-year experience[11]. Pennington made plication, both superficial and deep, in the inferior pole. According to report plication had the effect of lifting the inframmary crease, narrowing the remaining breast mound, and creating a distinct fold in the pedicle. Unlike the plication by Pennington, plication in our technique is performed in the superior only to create a conical appearance. The suture technique used in our mammaplasty is similar to that described by Toonard for MACS lift[46]. The techniques which buries a lower pole breast parenchymal flap underneath a bipedicled pectoralis major muscle flap have not been supported worldwide because they make breast cancer screening difficult and they violate tissue compartment[47]. However, the technique presented here does not damage the tissue since it only involves plication and no problems due to plication were shown in postoperative mammography in the long term. Hawtof et al. studied 268 patients and concluded that complications were significantly more prevalent in women undergoing reductions of greater than 700 gr per breast[48]. The size of the breast was not associated with complications in the present series. This can be ascribed to safety of the pedicle. We did not observe breast feeding difficulties in our patients after surgery. Because peripheral reduction of the gland does not discontinue the lactiferous ducts and no disturbance of breast feeding is to be expected after this kind of reduction mammaplasty.

esent series. This can be ascribed to safety of the pedicle. We did not observe breast feeding difficulties in our patients after surgery. Because peripheral reduction of the gland does not discontinue the lactiferous ducts and no disturbance of breast feeding is to be expected after this kind of reduction mammaplasty. It is still debatable whether reduction mammaplasty can be used for revision surgery. In this study, four patients underwent our technique for revision. These patients had undergone inferior pedicle and vertical scar mammaoplasty before. The results showed that COPCUs mammaplasty could be used with success regardless of the pedicle created in prior operations. The rate of complications in this study was found to be 9.5%. They were minor complications which did not require surgical treatment. The rate of complications in the literature has been found to range between 3% and 45% [49]. According to the results our experience, advantages of COPCUs mammaplasty provides complete safety of the NAC, creates the most natural breast in terms of tissue consistency and mobility, provides fullness in the superior part of the breast and excellent projection Our technique is very easy to perform and teach since open sky approach is used. It does not increase operation time and does not require liposuction. Reverse T scar is avoided and a very small vertical scar, which can be tolerated by patients, is created.

ess in the superior part of the breast and excellent projection Our technique is very easy to perform and teach since open sky approach is used. It does not increase operation time and does not require liposuction. Reverse T scar is avoided and a very small vertical scar, which can be tolerated by patients, is created. There are not any marked disadvantages of the technique. However, thinning likely to occur in elevation of dermal pedicles may cause skin problems. Although the patients included in this study were heavy smokers, they did not have skin loss. This indicates that dermal flaps have a rich blood supply. Advantages of this technique should be proven with detailed investigational laboratory studies such as sensorial, angiographical and ductal screening tests. But according to our experience, we speculate that our technique can be used safely for breast reduction and mastopexy. Competing interests The author declares that they have no competing interests. Authors' contributions EC designed the study, performed all operations and prepared the manuscript. Supplementary Material Additional file 1 Plication of the breast. The video file provided represents the plication of the upper part of the breast for youthful projection. Click here for file Additional file 2 COPCUs mammaplasty. The video file provided represents the COPCUs mammaplasty in all steps and also pre and post operative views. Click here for file

Background Surgery for left ventricular anterior reconstruction in the setting of ischemic cardiomyopathy has a long history. It has evolved over the years from linear closure of large aneurysms to present geometric reconstruction by means of endoventricular patch [1-4]. Surgical and clinical outcome have extensively been revised in many retrospective studies, highlighting a variability in the results of this surgical treatment [5-10]. The internal border zone between normal and infarcted tissue was usually encircled by the Fontan purse-string, a running circular suture with the aim to approach residual walls and reduce the tissue gap to be closed by the patch. The original circular patch used in the reconstruction of the residual tissue gap has been replaced by an oval-shaped patch [11,12], able to assure a better ventricular shape and volume, which have been pointed out as two of the major determinants of steady long-term results [13-16]. Thanks to the improving outcomes of this approach, a new chapter of left ventricular volume reduction has opened [17-19], supported by new and accurate studies on left ventricular geometry and function [20], with particular interest in myocardial fiber disposition [21,22].

erminants of steady long-term results [13-16]. Thanks to the improving outcomes of this approach, a new chapter of left ventricular volume reduction has opened [17-19], supported by new and accurate studies on left ventricular geometry and function [20], with particular interest in myocardial fiber disposition [21,22]. Methods We here describe a detailed surgical plan of a novel technique aiming at realigning residual myocardial fibers in a physiologic setting. It is a logical consequence of the state of the art of this surgical treatment and derives from studies about the pathological changes that intervene in the myocardium after an acute infarction. The technique has been improved since 2002, until its standard application in 21 patients (2005–2008). As an indirect evidence of restored fiber disposition, we observed in several patients, immediately after the surgical correction in the operating room and during the follow-up, the renewal of apical rotation and, consequently, of left ventricular torsion, a peculiar movement of the normal heart that optimizes ejection efficiency and energy expenditures: it was lost due to the dilation of the apex and to the loss of contractile myocardium and its renewal, although postulated, was never demonstrated before in any technique of ventricular reconstruction [23,24]. Failure to obtain its renewal during a so long clinical history of endoventricular reconstruction testifies that it does not exclusively depend on revascularization and reshaping itself, but on specific solutions addressing fiber rearrangement.

Methods We here describe a detailed surgical plan of a novel technique aiming at realigning residual myocardial fibers in a physiologic setting. It is a logical consequence of the state of the art of this surgical treatment and derives from studies about the pathological changes that intervene in the myocardium after an acute infarction. The technique has been improved since 2002, until its standard application in 21 patients (2005–2008). As an indirect evidence of restored fiber disposition, we observed in several patients, immediately after the surgical correction in the operating room and during the follow-up, the renewal of apical rotation and, consequently, of left ventricular torsion, a peculiar movement of the normal heart that optimizes ejection efficiency and energy expenditures: it was lost due to the dilation of the apex and to the loss of contractile myocardium and its renewal, although postulated, was never demonstrated before in any technique of ventricular reconstruction [23,24]. Failure to obtain its renewal during a so long clinical history of endoventricular reconstruction testifies that it does not exclusively depend on revascularization and reshaping itself, but on specific solutions addressing fiber rearrangement. Surgical procedure We called this surgical procedure "KISS", acrostic for "Keep fIbers orientation with Strip patch reShaping": the name underlines its peculiar action to let normal walls "kiss" together, restoring anatomic fibers' contiguity and orientation. Peculiarities of this reconstruction are:

Failure to obtain its renewal during a so long clinical history of endoventricular reconstruction testifies that it does not exclusively depend on revascularization and reshaping itself, but on specific solutions addressing fiber rearrangement. Surgical procedure We called this surgical procedure "KISS", acrostic for "Keep fIbers orientation with Strip patch reShaping": the name underlines its peculiar action to let normal walls "kiss" together, restoring anatomic fibers' contiguity and orientation. Peculiarities of this reconstruction are: 1. the shape of the patch (reduction of minor axis compared with currently used oval-shaped patch, tailored major axis and arrow-shaped ends); 2. the absence of Fontan purse-string; 3. the asymmetrical way of suturing it inside the ventricle. Preliminary phases are common to the standard operative management: median sternotomy, mild hypothermic (33°C) cardiopulmonary bypass, aortic cross clamping, intermittent cold blood cardioplegia and warm reperfusate. Left ventricle is opened, as usually, through an anterolateral incision two cm aside the left anterior descending artery. Endoventricular reconstruction surgery was performed by means of a very narrow, strip-shaped patch, with arrow-shaped ends, whose minor axis is less than one cm wide and major axis is proportional to the width of infarcted area. The patch used was made either by glutaraldehyde-treated autologous pericardium or by knitted Dacron (Hemashield Platinum Finesse, 0.8 × 7.6 cm, Boston Scientific, Natick, MA, USA).

ed patch, with arrow-shaped ends, whose minor axis is less than one cm wide and major axis is proportional to the width of infarcted area. The patch used was made either by glutaraldehyde-treated autologous pericardium or by knitted Dacron (Hemashield Platinum Finesse, 0.8 × 7.6 cm, Boston Scientific, Natick, MA, USA). Operative steps 1. Once opened the left ventricle, boundaries between normal and fibrotic tissue, if present, are identified; in case of akinetic, non-fibrotic area, thinning of the area and information derived from echocardiographical and/or magnetic resonance studies are used to identify the limits from normal myocardium. 2. The Fontan purse-string is not used, to avoid a circular disposition of the border zone between normal and fibrotic muscle that alters the spatial geometry of myocardial walls and forces myocardial fibers to a non-physiologic orientation; in this way, we let the suture to redirect fibers, without any pre-conditioned deformation. 3. The limit of the fibrosis near the apex is then identified: this is the distal end of the patch and will form the new apex; it corresponds to the angle between the end of necrosis and the septum and is defined as "Apex stitch" (Figure 1). Figure 1 Apex Stitch. Angle between the end of necrosis and the septum: it corresponds to the lower end of the patch and will be the new apex (apex stitch). 4. The limit of fibrotic tissue in the interventricular septum (near the aortic valve) is identified: this is the upper end of the patch and is defined as "High stitch" (Figure 2).

Figure 1 Apex Stitch. Angle between the end of necrosis and the septum: it corresponds to the lower end of the patch and will be the new apex (apex stitch). 4. The limit of fibrotic tissue in the interventricular septum (near the aortic valve) is identified: this is the upper end of the patch and is defined as "High stitch" (Figure 2). Figure 2 High Stitch. Upper limit of fibrotic tissue in the interventricular septum: it is the upper end of the patch (high stitch). 5. The distance between these two points (the high and the apex stitches) along the interventricular septum is the length (major axis) of the patch that is then tailored according to this measure with arrow-shaped ends (Figure 3). Figure 3 Patch measure and tailoring. The distance between these two points (the high and the apex stitches) is the length (major axis) of the patch that is then tailored according to this measure with an arrow-shaped end. 6. The upper end of the patch is secured to the High stitch by a 4.0 polypropylene monofilament (Prolene, Ethicon), tied at its half length with few knots (Figure 2). 7. All the septal rim of the patch is then sutured along the basal border of the fibrotic interventricular septum, from the high to the apex stitch, with a continuous running suture using the 4.0 polypropylene monofilament: in this suture there is almost a patch-tissue equivalence, that is the length of the patch and the length of the myocardial wall are quite equal (Figure 4).

basal border of the fibrotic interventricular septum, from the high to the apex stitch, with a continuous running suture using the 4.0 polypropylene monofilament: in this suture there is almost a patch-tissue equivalence, that is the length of the patch and the length of the myocardial wall are quite equal (Figure 4). Figure 4 Septal Suture. All the inferior (septal) rim of the patch is sutured along the basal border of the fibrotic interventricular septum, from the high to the apex stitch (patch-tissue equivalence). 8. The lateral rim of the patch is then secured to the lateral border zone between fibrotic and normal tissue of the ventricular wall with the other half of 4.0 monofilament: this line is widened by the dilation and the curvature of the left ventricle, so the length of the patch is smaller than the length of the ventricular wall (patch-tissue mismatch) (Figure 5A, B): surgeon must fit the length of lateral wall to the length of the patch, shrinking in this way the ventricular wall to a shorter line, that is, forcing the orientation of residual myocardium (and myofibers' lamina) towards a tighter disposition, compensating the dilation and spacing due to akinetic/dyskinetic portion of the wall.

st fit the length of lateral wall to the length of the patch, shrinking in this way the ventricular wall to a shorter line, that is, forcing the orientation of residual myocardium (and myofibers' lamina) towards a tighter disposition, compensating the dilation and spacing due to akinetic/dyskinetic portion of the wall. Figure 5 Lateral Suture. The lateral border zone between fibrotic and normal tissue of the ventricular wall is secured to the superior (lateral) rim of the patch (patch-tissue mismatch). Panels A and B depict the stretching of the suture that approaches and fits the lateral wall to the patch. The arrow in panel B shows how the suture redirects fiber orientation, approaching a farer point of the myocardial wall to a nearer point on the patch. 9. The two sutures end at the Apex stitch and are tied together (Figure 6). Figure 6 Final appearance. The two sutures end at the Apex stitch and are tied together. 10. Ventriculotomy was closed by overlapping the free edges in a vest-over-pants fashion, in order to occlude the excluded chamber, with or without a felt strip to reinforce the suture (Figure 7). Figure 7 Ventriculotomy closure. Ventriculotomy was closed by overlapping the free edges; note the "goffered" lateral wall, due to length compensation obtained by the suture.

10. Ventriculotomy was closed by overlapping the free edges in a vest-over-pants fashion, in order to occlude the excluded chamber, with or without a felt strip to reinforce the suture (Figure 7). Figure 7 Ventriculotomy closure. Ventriculotomy was closed by overlapping the free edges; note the "goffered" lateral wall, due to length compensation obtained by the suture. Relevant surgical hints • The first running suture ("septal") has quite the same length as the patch: this part of the ventricular wall usually becomes thinner rather than dilated, due to the anatomy of the septum, embedded between the two ventricles, with a dual layers structure vascularized by two coronary systems; in some cases is possible to compensate a greater dilation of this part with the suture, stretching the residual myocardium to a shorter patch, but in most of the cases there is a patch-tissue equivalence. • Position of the new apex is often determined by the end of the necrotic tissue; otherwise, its relocation should be guided by normal geometry; normal values of diastolic left ventricular length are 80 ± 9 mm [25], so this should be the maximum value of the new chamber, adding patch length to the length of the normal upper part of interventricular septum (from aortic valve to akinetic/fibrotic septum); new apex should also be near the level of right ventricular apex, as is in the normal heart.

ntricular length are 80 ± 9 mm [25], so this should be the maximum value of the new chamber, adding patch length to the length of the normal upper part of interventricular septum (from aortic valve to akinetic/fibrotic septum); new apex should also be near the level of right ventricular apex, as is in the normal heart. • Given that usually we can reach the upper septum at 2–3 cm from the aortic valve, it follows that the length of the patch should be 5–6 cm in most cases; the end of right ventricle is another reference point to set the new apex. • The second suture ("lateral") forces the lateral wall to fit to the patch, stretching myocardial fibers towards a shorter line, thus realigning them in a narrower setting, representing the key characteristic of this technique. For this reason the lateral sutured myocardial border appears "goffered" at the end of the procedure, to compensate the discrepancy between patch and myocardial suture rims (Figure 7). • Converting the setting to a geometrical figure, as previously described [26], we have a circular segment in which the chord is the basal line of the fibrotic interventricular septum and the arc is the semicircular line of lateral border zone between fibrotic and normal tissue of the ventricular wall; the surgical suture approaches the arc to the chord, reducing the arc's length of about one half.

e have a circular segment in which the chord is the basal line of the fibrotic interventricular septum and the arc is the semicircular line of lateral border zone between fibrotic and normal tissue of the ventricular wall; the surgical suture approaches the arc to the chord, reducing the arc's length of about one half. • The gross realignment of residual myocardial fibers at the border zone in order to have a recovery in their oblique spatial orientation can only be obtained by a narrow (small minor axis) patch, which forces the realignment, while a wider (great minor axis) patch tends to maintain fibers direction in the setting of a dilated left ventricle, eliminating the length compensation induced by this technique. • The use of a patch is of primary importance for fibers' realignment: despite this technique could seem like a linear suture, only the use of the patch makes to grade the stretching of the two sutures possible, differentiating the fitting of the two rims (septal and lateral) to the relative border of the patch; otherwise, it would be impossible to adjust different fibers' orientation along inferior septum (less dilated = fibers less divergent) and lateral wall (more dilated = fibers more divergent). • The key features of this surgical technique are the small short axis of the patch and the asymmetrical suture at the two sides of the patch: the first one approaches residual myocardium as close as possible; the second differentiates the reconstruction of the dilated border zones to re-establish a physiologic contiguity of myocardial lamina.

s surgical technique are the small short axis of the patch and the asymmetrical suture at the two sides of the patch: the first one approaches residual myocardium as close as possible; the second differentiates the reconstruction of the dilated border zones to re-establish a physiologic contiguity of myocardial lamina. Results Overall Results Twenty-one patients were operated on with this technique. Mean age was 60.8 ± 12.5 years. Eleven of them were male. Eight patients had a mitral regurgitation ≥ grade 3 and were contemporarily treated with mitral ring anuloplasty. All patients received complete coronary revascularization. Patch dimensions were 1.0 ± 0.3 × 5.9 ± 1.0 (short × long axis, cm). No hospital deaths were reported. At a mean follow-up time of 1.95 ± 0.96 years, all patients were alive and in functional NYHA class = 1.3 ± 0.5. Two-dimensional speckle tracking imaging, a reliable ecocardiographical tool for rotation movements of the heart, was used to assess left ventricular torsion in presence of a good acoustic window (16 out of 21 patients). Echocardiographical data are reported in Table 1. Table 1 Overall series data Preop Follow-up data (1.95 ± 0.96 years) EDV 222.8 ± 66.7 136.2 ± 39.3 EDVI 122.0 ± 34.4 73.1 ± 35.2 ESV 159.3 ± 50.7 88.0 ± 31.0 ESVI 87.5 ± 27.9 35.2 ± 16.1 EF 27.9 ± 6.9 36.4 ± 8.5 LV TORSION 2.5 ± 4.6 7.7 ± 3.4 EDV = end-diastolic volume (ml); EDVI = end-diastolic volume index (ml/m2); ESV = end-systolic volume (ml); ESVI = end-systolic volume index (ml/m2); EF = ejection fraction. LV torsion is reported in degrees.

EDVI 122.0 ± 34.4 73.1 ± 35.2 ESV 159.3 ± 50.7 88.0 ± 31.0 ESVI 87.5 ± 27.9 35.2 ± 16.1 EF 27.9 ± 6.9 36.4 ± 8.5 LV TORSION 2.5 ± 4.6 7.7 ± 3.4 EDV = end-diastolic volume (ml); EDVI = end-diastolic volume index (ml/m2); ESV = end-systolic volume (ml); ESVI = end-systolic volume index (ml/m2); EF = ejection fraction. LV torsion is reported in degrees. Exemplary case A 52-years-old male patient, ZD, had the acute episode of myocardial infarction on August 2001, treated with thrombolysis. On September 2001 he performed a coronary angiography showing critical stenoses of left anterior descending (LAD) and right coronary arteries. He then underwent off-pump surgical revascularization by means of three coronary grafts (diagonal, anterior descending and right coronary arteries). The left ventricular function was already compromised (EF = 30%). Postoperative course was uneventful. Since 2004 he suffered from several episodes of heart failure. On July 2004, coronary re-angiography showed a mild stenosis of the graft on LAD (50%) and left ventricular dysfunction (EF = 28%) with anterolateral aneurysm. Grade 4 mitral regurgitation was evident. Ventriculography showed a completely upset LV geometry with consequent loss of ventricular apex and torsion movement (Additional file 1). On May 2005, the patient was reoperated (remodeling time from acute infarction to surgery = 49.7 months). Endoventricular reshaping was performed as described. Patch dimensions were 1 × 6 cm. A 30-mm rigid ring mitral anuloplasty was associated. Left anterior descending artery was grafted by a saphenous vein. Once weaned from extracorporeal circulation, left ventricle showed the renewal of apical rotation and, as a consequence, of left ventricular torsion, as evident in Additional file 2. Echocardiographical data are reported in Table 2. Patient was discharged to a rehabilitation hospital eight days after the operation and is now in first functional NYHA class. A 3D full volume echocardiography was performed at late follow-up (19 months), showing a still present apical rotation at several apical slices (Additional file 3).

data are reported in Table 2. Patient was discharged to a rehabilitation hospital eight days after the operation and is now in first functional NYHA class. A 3D full volume echocardiography was performed at late follow-up (19 months), showing a still present apical rotation at several apical slices (Additional file 3). Table 2 Exemplary case data Preop Postop 7 days Postop 19 months EDD 75 65 68 ESD 59 54 51 APICAL DIAMETER 70 36 35 LLD 82 79 78 LLS 82 76 76 EDV 256 160 136 EDVI 135 84 69 ESV 150 120 95 ESVI 79 62 48 EF 28 35 38 WMSI 2.88 1,88 1,56 MR 4 0 0 NYHA 4 1 1 EDD = end-diastolic diameter (mm); ESD = end-systolic diameter (mm); APICAL DIAMETER = internal diastolic diameter two centimeters above the ventricular apex (mm); LLD = diastolic longitudinal length (mm); LLS = systolic longitudinal length (mm); EDV = end-diastolic volume (ml); EDVI = end-diastolic volume index (ml/m2); ESV = end-systolic volume (ml); ESVI = end-systolic volume index (ml/m2); EF = ejection fraction; WMSI = wall motion score index; MR = mitral regurgitation; NYHA = New York Health Association class. Due to a poor echocardiographical window, this patient was not studied by 2D speckle tracking echocardiography. Some of patients studied by this method were reported as single clinical cases [27,28], demonstrating the renewal and the persistence during time of apical rotation and LV torsion.

NYHA 4 1 1 EDD = end-diastolic diameter (mm); ESD = end-systolic diameter (mm); APICAL DIAMETER = internal diastolic diameter two centimeters above the ventricular apex (mm); LLD = diastolic longitudinal length (mm); LLS = systolic longitudinal length (mm); EDV = end-diastolic volume (ml); EDVI = end-diastolic volume index (ml/m2); ESV = end-systolic volume (ml); ESVI = end-systolic volume index (ml/m2); EF = ejection fraction; WMSI = wall motion score index; MR = mitral regurgitation; NYHA = New York Health Association class. Due to a poor echocardiographical window, this patient was not studied by 2D speckle tracking echocardiography. Some of patients studied by this method were reported as single clinical cases [27,28], demonstrating the renewal and the persistence during time of apical rotation and LV torsion. Discussion Left ventricular structure and alterations induced by myocardial infarction The left ventricle is characterized by a unique myocardial tissue structure [29,30]. The necrosis induced by myocardial infarction alters the myocardial continuum, forcing the whole ventricle to work at a lower efficiency level. Necrotic area spaces out the residual normal tissue and negative remodeling following the acute episode [31] completely alters ventricular geometry and mechanic. Nevertheless, it is demonstrated from in vivo studies that myofibers orientation in the thickness of normal myocardium is not changed after myocardial infarction [32] and that transmural courses of fiber orientation angles near infarct zones were similar to those of normal myocardium [33]. This means that infarct site demarcates at the end of necrotic process, with an extension related to the tributary area of the occluded coronary vessel and to the wavefront phenomenon of ischemic cell death [34]: definitive boundaries are characterized by a still normal myocardium interlaced to the necrotic region. This is shown also in gross histological samples [35]: a normally structured myocardium ends without graduation where starts the fibrotic infarcted tissue. It is also evident that alterations occurring in myocardium adjacent to an infarction consists with myocytes elongation and myofiber rearrangement (slippage), as well as changes in orientation of the laminar structure of the ventricular wall [36].

dium ends without graduation where starts the fibrotic infarcted tissue. It is also evident that alterations occurring in myocardium adjacent to an infarction consists with myocytes elongation and myofiber rearrangement (slippage), as well as changes in orientation of the laminar structure of the ventricular wall [36]. The key point to understand how myocardial structure is altered by coronary vessel occlusion is the relative position of epicardial coronary vessel and fiber orientation. Coronary arteries have an epicardial course not to be squeezed during systole, while fiber orientation is changing throughout myocardium thickness, from epicardium to endocardium, following a continuous angular gradient which is crossed more or less perpendicularly by vessels direction (Figure 8A). The transmural necrosis induced by coronary vessel occlusion concerns a sector of myocardium made by several strata of differently oriented myocardial fibers and ends where the territory of another coronary vessel begins (Figure 8A, shaded area). At the border zone, still normal myocardium with still normal fiber orientation interlaces with the necrotic tissue that lost its 3D fiber structure. Necrotic tissue discontinues continuity and contiguity of myocardial fibers, although they remain normally oriented in the thickness of normal myocardium (Figure 8B, correspondence of points "a" and "a1", "b" and "b1" etc).

l normal fiber orientation interlaces with the necrotic tissue that lost its 3D fiber structure. Necrotic tissue discontinues continuity and contiguity of myocardial fibers, although they remain normally oriented in the thickness of normal myocardium (Figure 8B, correspondence of points "a" and "a1", "b" and "b1" etc). Figure 8 Coronary vessels and fiber orientation. A: Epicardial course of coronary arteries crossing fiber orientation. The shaded area depicts transmural necrosis induced by coronary vessel occlusion which concerns a sector of myocardium made by several strata of differently oriented myocardial fibers and ends where the territory of another coronary vessel begins. B: At the border zone, still normal myocardium with still normal fiber orientation interlaces with the necrotic tissue that lost its 3D fiber structure. Necrotic tissue discontinues contiguity of myocardial fibers, although they remain normally oriented in the thickness of normal myocardium (correspondence of points "a" and "a1", "b" and "b1" etc). Fiber picture obtained by the DTI Track module of MedINRIA software [45]. The infarcted area alters regional and then global function of the left ventricle, expands and dilates the ventricular chamber and dislocates adjacent, still anatomically normal, non-infarcted myocardium. Consequently, fibers in the normal myocardium are displaced by the necrotic, thin, akinetic/dyskinetic portion of ventricular wall and (although normally oriented inside the residual wall) their orientation and spatial disposition is completely upset.

cates adjacent, still anatomically normal, non-infarcted myocardium. Consequently, fibers in the normal myocardium are displaced by the necrotic, thin, akinetic/dyskinetic portion of ventricular wall and (although normally oriented inside the residual wall) their orientation and spatial disposition is completely upset. On the other hand, it has been demonstrated that surgical intervention could effectively alter (positively or negatively) the helix angles in regions adjacent to repair [32]. KISS reconstruction Surgical restoration replaces the infarct scar, and tries to modify ventricular geometry in such a way as to recreate a nearly normal ventricle. The described procedure reduces the distance from the contractile myocardium thanks to the narrow patch that absorbs the stress of the suture, acts as a support for the suture lines and guides the restoration of myocardial wall contiguity and fibers' orientation. A narrow patch also helps to rebuild near-normal ventricular volumes because it does not add any volume itself. Comparing the left ventricle to a prolate ellipsoid, small variations in the minor axis correspond to great changes in ventricular volume (1 cm less in the minor axis produces a reduction in volume of about 30 ml). In this way, major determinants of normal physiology are restored: elliptical shape, near-normal volume, small residual akinesia, contiguity of contractile myocardium, spatial geometry and fibers orientation. Complete revascularization and, when required, the competence of mitral valve crown the physiologic restoration.

n this way, major determinants of normal physiology are restored: elliptical shape, near-normal volume, small residual akinesia, contiguity of contractile myocardium, spatial geometry and fibers orientation. Complete revascularization and, when required, the competence of mitral valve crown the physiologic restoration. Aiming the ventricular restoration at the realignment of residual fibers could overcome the endless contest about akinetic and dyskinetic ventricular walls. We should assume that akinesia and dyskinesia are different stages/states of the same disease and both result in a non-contractile zone of the myocardium. From a strictly functional point of view, both interrupt myocardial continuum, lead to fibers disarrangement and ventricular remodeling, so both are responsible of the same pathophysiology. We should then reconsider that the border zone of myocardial infarction is the limit between fibers that lost and fibers that preserved their anatomical and physiological arrangement. This is the consequence of the ischemic disease, and in case of akinetic areas, in which the border zone is not marked by fibrotic degeneration, it is visually less evident, but functionally has the same effect.

the limit between fibers that lost and fibers that preserved their anatomical and physiological arrangement. This is the consequence of the ischemic disease, and in case of akinetic areas, in which the border zone is not marked by fibrotic degeneration, it is visually less evident, but functionally has the same effect. Finally, also a non-trasmural infarction can damage myocardial continuum in such a way to alter its functionality, though zones of normal fibers remain in the thickness of the infarcted region. These residual fiber islands, embedded in the fibrotic tissue or simply adjacent to it, can not work properly because they are blocked by the non-contractile cells, and the physiologic fiber interlacement that produces the contraction is lost: they do not act on normal adjacent fibers, propagating the contraction, but their action extinguishes against non-contractile tissue. In these cases, function of myocardial wall is the result of the delicate balance between contractile and non-contractile fibers.

ber interlacement that produces the contraction is lost: they do not act on normal adjacent fibers, propagating the contraction, but their action extinguishes against non-contractile tissue. In these cases, function of myocardial wall is the result of the delicate balance between contractile and non-contractile fibers. Left Ventricular Torsion Since 1911 [37] left ventricular torsion was recognized to be essential for ventricular ejection and many studies confirm its importance as an index of normal ventricular function in physiologic and pathologic settings [38-41]. The twisting motion of the left ventricle about its long axis results from the contraction of the opposite, obliquely oriented epicardial and endocardial fibers that produce the contemporary clockwise rotation of the base and counterclockwise rotation of the apex, obtaining the torsion of the heart along its long axis. Loss of myocardial structure in anterolateral region, chamber dilation, volume overload and loss of the ventricular apex, lead to the loss of apical rotation and, consequently, of left ventricular torsion.

ion of the base and counterclockwise rotation of the apex, obtaining the torsion of the heart along its long axis. Loss of myocardial structure in anterolateral region, chamber dilation, volume overload and loss of the ventricular apex, lead to the loss of apical rotation and, consequently, of left ventricular torsion. The evidence of the recovery of left ventricular torsion is an indirect but striking demonstration that fibers should be reset to a physiologic orientation with the described endoventricular reshaping. The evidence of torsion recovery corresponds to a more physiologic rearrangement of myocardial fibers. We observed the renewal of ventricular torsion ever since the end of the surgical procedure in the operating room, immediately after the surgical correction: so it must be due to a mechanic correction of fibers orientation and not only to a progressive positive remodeling after surgery.

rrangement of myocardial fibers. We observed the renewal of ventricular torsion ever since the end of the surgical procedure in the operating room, immediately after the surgical correction: so it must be due to a mechanic correction of fibers orientation and not only to a progressive positive remodeling after surgery. Current Limitations and Future Developments At present, we have only indirect, "gross" evidence of fiber realignment, but we can not correct fiber disposition in order to obtain a "histological" realignment. Fast improvement in medical imaging techniques able to visualize fiber orientation (diffusion tensor magnetic resonance, limited until now by heart movement) will be very useful in this topic and will probably lead to a preoperative assessment full of key data for the surgeon. Some advanced technology laboratories [42,43] are already studying the use of integrated imaging techniques for accurate planning of surgical procedures (for example, tumor removal in the brain with 3D visualization of cerebral fibers, tissue and structures). It would be reasonable to foresee a 3D surgical planning of ventricular reshaping with reference points for fiber disposition that will guide the surgeon during the reconstruction. A so called "fiber-based" therapy could open unexpected improvements in the surgical treatment of the failing heart. For example, the Batista operation and the surgical reshaping in the setting of dilated cardiomyopathy had not the expected diffusion, perhaps because volume reduction did not take care of fiber alignment at the suture site, but caused itself a fiber disarrangement, suturing together non-contiguous parts of ventricular walls with different fiber disposition. A fiber-based physiological view could also help us in the evaluation and/or optimization of medical therapies, mainly in case of new drugs or of non-responder patients.

e suture site, but caused itself a fiber disarrangement, suturing together non-contiguous parts of ventricular walls with different fiber disposition. A fiber-based physiological view could also help us in the evaluation and/or optimization of medical therapies, mainly in case of new drugs or of non-responder patients. Finally, basic science implications of fiber-based left ventricular function could let us use an old knowledge that was forgotten until now, due to the imaging limitations and to the complex 3D anatomy of myocardium, but that could represent the basis for a new approach to the failing heart. Conclusion The knowledge of anatomical structure of left ventricle is very old [44] but it has never been applied in clinical practice. The described endoventricular reconstruction tends to grossly realign fibers in a more physiologic manner, approaching them together and re-establishing their oblique disposition and orientation, compensating in this way the spacing and the anomalies induced by myocardial infarction. The evidence of the renewal of left ventricular torsion could open the way to implement the evaluation of the results of this surgical procedure. Variability in outcome could be overcome with a multifactorial approach to restoration taking care of all aspects of normal physiology of ventricular function, in which fibers-based procedures will have a key role. Future development in medical imaging will probably confirm and support the main role of fibers' disposition in the surgical treatment of failing hearts.

multifactorial approach to restoration taking care of all aspects of normal physiology of ventricular function, in which fibers-based procedures will have a key role. Future development in medical imaging will probably confirm and support the main role of fibers' disposition in the surgical treatment of failing hearts. Competing interests The author declares that they have no competing interests. Authors' contributions MC set the theoretical bases and conceived the surgical procedure, performing all of the operations. The described surgical technique and its peculiar result (the original, never described before renewal of left ventricular apical rotation and torsion) are to be considered his own intellectual property. Supplementary Material Additional file 1 Exemplary Patient: preoperative left ventriculography. Preoperative left ventriculography of the exemplary patient with large dyskinesia: normal geometry and torsion movement are completely lost. Click here for file Additional file 2 Exemplary Patient: renewal of left ventricular torsion at the end of operation. Head view of the heart at the end of reconstruction surgery of the same exemplary patient, normal and slow motion of the same sequence. Cardiopulmonary bypass is off. Ventilation is temporarily off. The counterclockwise torsion movement towards the right ventricle is evident, marked by the displacement of the saphenous graft on left anterior descending artery. Click here for file

Additional file 2 Exemplary Patient: renewal of left ventricular torsion at the end of operation. Head view of the heart at the end of reconstruction surgery of the same exemplary patient, normal and slow motion of the same sequence. Cardiopulmonary bypass is off. Ventilation is temporarily off. The counterclockwise torsion movement towards the right ventricle is evident, marked by the displacement of the saphenous graft on left anterior descending artery. Click here for file Additional file 3 Exemplary Patient: late full volume echocardiography. Simultaneous display of nine short axis views generated from an apical full volume acquisition with 4D mode (Vivid 7, GE Medical Systems, Norway) of the exemplary patient 19 months after the operation. Apical rotation is evident towards the right ventricle. Click here for file

Background The surgical treatment of diaphyseal forearm fracture-nonunions remains a therapeutic challenge for orthopaedic trauma surgeons. Key to success in the management of these demanding conditions is to develop a comprehensive treatment concept which considers the forearm and its adjacent joints, the elbow and wrist, as a complex functional unit [1,2]. Nonunions of the radius and ulna shaft cause a severe anatomic and functional impairment, related to disturbance of the interosseous membrane and dysfunction of the adjacent joints, elbow and wrist [3-6]. These demanding nonunions require the surgical correction to restore the anatomy of the forearm and to improve function [1,7]. New techniques have been recently postulated for the treatment of forearm nonunions, including distraction-compression osteogenesis, locked plating, and locked intramedullary nailing [8-10]. In addition, free fibula transfer flaps have been advocated as a means to restore anatomic length and ensure bony union [11,12]. In the present study, we evaluated the long-term radiological and clinical outcome of 31 consecutive patients treated by autologous bone grafting and compression plating for fracture-nonunions of the forearm. We hypothesized that this "classic" treatment concept would result in excellent clinical outcome and a low incidence of long-term functional impairment.

ng-term radiological and clinical outcome of 31 consecutive patients treated by autologous bone grafting and compression plating for fracture-nonunions of the forearm. We hypothesized that this "classic" treatment concept would result in excellent clinical outcome and a low incidence of long-term functional impairment. Methods A retrospective analysis of a prospective database of all consecutive patients treated for fracture-nonunions of the forearm was performed at a single academic center (Dept. of Orthopaedics and Traumatology, Federal University of Sao Paulo, Brazil). The inclusion criteria consisted of all adult patients (> 18 years) of either gender with posttraumatic nonunions of the radius and/or ulna shaft, in absence of an active infection. Patients with nontraumatic forearm pseudarthrosis, infected nonunions, or with an associated neurological impairment of the ipspilateral upper extremity, which may preclude from an adequate functional assessment, were excluded from analysis. All surgical procedures were performed by a single surgeon (F.B.R.). Nonunions were classified according to the standard classification by Weber and Cech [13]. The pre-operative plan included plain X-rays of the forearm, wrist and elbow, in antero-posterior (a.p.) and lateral views. The presence of infection was excluded by preoperative analysis of systemic infection parameters (WBC, sedimentation rate, CRP) and by intraoperative tissue samples of the resected nonunions which were sent for microbiology cultures and histopathological workup. The standardized treatment concept (Figure 1) consisted of resection of the forearm nonunion, autologous bone grafting with a tricortical graft from the iliac crest with anatomic restoration of the foreram length, as determined by intraoperative fluoroscopy assessment of the adjacent joints, and compression plating using a stainless steel small fragment (3.5 mm) dynamic compression plate (DCP) or limited-contact DCP (LC-DCP). Radiological and functional outcome was determined at a minimum of 2 years follow-up, with standardized intervals at 3, 6, 12, and 24 months after revision surgery. The functional outcome was evaluated according to the criteria defined by Tscherne et al. [14]. These criteria include the range of motion of forearm pro-/supination, wrist and elbow flexion and extension, ulnar shift/deviation, and positive/negative ulnar variant on wrist X-rays.

and 24 months after revision surgery. The functional outcome was evaluated according to the criteria defined by Tscherne et al. [14]. These criteria include the range of motion of forearm pro-/supination, wrist and elbow flexion and extension, ulnar shift/deviation, and positive/negative ulnar variant on wrist X-rays. Bony union was defined in the presence of a minimum of three bridging cortices in plain X-rays of the forearm in a.p. and lateral views, in absence of pain at the nonunion site. Figure 1 Preoperative planning scheme for nonunion resection, placement of corticocancellous iliac crest bone graft, and compression plating.

and 24 months after revision surgery. The functional outcome was evaluated according to the criteria defined by Tscherne et al. [14]. These criteria include the range of motion of forearm pro-/supination, wrist and elbow flexion and extension, ulnar shift/deviation, and positive/negative ulnar variant on wrist X-rays. Bony union was defined in the presence of a minimum of three bridging cortices in plain X-rays of the forearm in a.p. and lateral views, in absence of pain at the nonunion site. Figure 1 Preoperative planning scheme for nonunion resection, placement of corticocancellous iliac crest bone graft, and compression plating. Results and discussion A total of 32 consecutive patients were available for prospective enrollment. One patient died from a cardiovascular condition (myocardial infraction) in the postoperative course and was therefore lost to follow-up. The remaining 31 patients (27 males and 4 females; median age 30 years) presenting with forearm fracture-nonunions were prospectively enrolled into this study. Of these, 26 patients had their dominant arm affected, and 8 patients had a previous history of infection. Eleven patients presented with a both bone fracture-nonunion, and 20 patients had a single bone affected (11 radius, 9 ulna shaft). Eight patients had a history of a previous local infection, which however was completely healed at the time of revisions surgery, as determined by negative intraoperative bone biopsies and tissue cultures. The number of previous surgeries on the affected forearm varied from 1 to 5 (median 1.5). Patients presented with a fracture-nonunion within 5 to 24 months after the initial surgery (median of 7.5 months). According to the Weber and Cech classification, 28 cases (90%) were defined as atrophic, nonviable nonunions. Fourteen of these 28 patients had a segmental bone defect ranging from 1 to 5 cm (median 2.3 cm). Radiographically, a bony union was achieved in 30/31 patients within a mean time of 3.5 months of revision surgery (range 2 to 5 months). Clinically, 29/31 patients showed a good functional outcome, according to the Tscherne criteria, and 26/31 patients were able to resume their previous work. An illustrative case example of uneventful nonunion healing is shown in figure 2.

ients within a mean time of 3.5 months of revision surgery (range 2 to 5 months). Clinically, 29/31 patients showed a good functional outcome, according to the Tscherne criteria, and 26/31 patients were able to resume their previous work. An illustrative case example of uneventful nonunion healing is shown in figure 2. Figure 2 Clinical case example of a patient with a fracture-nonunion of the ulna shaft secondary to an unstable treatment modality using a small fragment tubular plate, lacking adequate stability with interfragmentary compression (A). Revision surgery was performed by corticocancellous bone grafting and compression plating using a more rigid construct with a dynamic compression plate (B). Panel C shows the fully healed radius fracture and ulnar fracture-nonunion after hardware removal.

acking adequate stability with interfragmentary compression (A). Revision surgery was performed by corticocancellous bone grafting and compression plating using a more rigid construct with a dynamic compression plate (B). Panel C shows the fully healed radius fracture and ulnar fracture-nonunion after hardware removal. Two patients developed a posteroperative infection, of which one case was successfully managed by surgical debridement and antibiotics for two weeks. The other patient developed a persistent infected nonunion requiring further revision surgery. This was the only case of the entire cohort of 31 patients which failed to achieve bony union. No postoperative failure of fixation was seen in any patient. One patient was lost to long-term follow-up secondary to death related to a cardiovascular event at 5 months, when the nonunion was fully healed. No association was shown between the time of bony union and the time elapsed between the trauma and the last surgery prior to presentation with a nonunion (data not shown). Residual radiological or clinical deformities were observed in 13 patients (table 1). These radiological findings did not appear to significantly influence the functional outcome, based on the observed clinical outcome in 29/31 patients, as defined by the Tscherne criteria, with less than 10% of restricted range of motion at the wrist and elbow (table 2). Regarding forearm rotation (pro-/supination), 17/31 patients presented losses lower than 20 degrees, and 9/31 presented moderate results with pro/-supination loss between 20–40 degrees. A total of 26/31 patients were able to resume their previous work.

10% of restricted range of motion at the wrist and elbow (table 2). Regarding forearm rotation (pro-/supination), 17/31 patients presented losses lower than 20 degrees, and 9/31 presented moderate results with pro/-supination loss between 20–40 degrees. A total of 26/31 patients were able to resume their previous work. Table 1 Clinical and radiological residual deformities in 13 patients following revision surgery for forearm fracture-nonunions Residual deformity Number of patients (total n = 13) Positive ulnar variant: 1 mm 4 Positive ulnar variant: 2 mm 1 Negative ulnar variant: 1 mm 2 Ulnar head prominence 2 Ulnar head absence 1 Loss of radial bow 3 Table 2 Functional outcome in 30 patients with radiologically healed forearm fracture nonunions, assessed 6 months after revision surgery ROM* wrist Number of patients Normal (identical to contralateral side) 26 Flexion/extension 90°/0°/60° 3 Limited ulnar shift 1 ROM* elbow Number of patients Normal (identical to contralateral side) 27 Extension/flexion 0°/20°/120° 1 Extension/flexion 0°/30°/120° 2 Forearm rotation Number of patients Normal (identical to contralateral side) 17 Pro-/supination 90°/0°/60° 5 Pro-/supination 90°/0°/45° 3 Pro-/supination 80°/0°/70° 1 Pro-/supination 60°/0°/45° 1 Pro-/supination 45°/0°/30° 3 *ROM, range of motion.

Normal (identical to contralateral side) 27 Extension/flexion 0°/20°/120° 1 Extension/flexion 0°/30°/120° 2 Forearm rotation Number of patients Normal (identical to contralateral side) 17 Pro-/supination 90°/0°/60° 5 Pro-/supination 90°/0°/45° 3 Pro-/supination 80°/0°/70° 1 Pro-/supination 60°/0°/45° 1 Pro-/supination 45°/0°/30° 3 *ROM, range of motion. This study demonstrates the efficacy and safety for the treatment concept of autologous bone grafting and compression plating for forearm fracture-nonunions, leading to excellent radiological and functional long-term outcome. Reconstruction of the anatomy of both forearm bones is of crucial importance in the management of the diaphyseal forearm nonunions [1-6]. The concept of corticocancellous iliac crest bone grafting and compression plating for, was previously postulated as an early treatment strategy for traumatic segmental defects of the upper extremity, including forearm fractures [15]. Despite open wounds in some patients which healed by secondary intention, the exposed cortical bone graft was not shown to be prone to infection [15]. More recently, a retrospective analysis of 41 patients with comminuted both bone forearm fractures treated by compression plating with or without primary bone grafting determine a nonunion rate of 12%, and no benefit was revealed for early bone grafting with regard to the rate of union [16]. Barbieri and colleagues reported their experience in a case series of 12 patients treated by iliac crest bone block grafting and compression plating for diaphyseal defects of the forearm, secondary to infection and bone loss [17]. The authors demonstrated a successful union in 10/12 patients, within a mean time period of 17 weeks after the surgical revision [17]. However, a high rate of 30% recurrent infections of was reported in this cohort, which questions the safety of autologous bone grafting in the setting of posttraumatic infection and chronic osteomyelitis. Similarly, Moroni and colleagues reported a high incidence of infection of 12.5% after intercalary bone graft fixation in patients with isolated forearm nonunions [18]. In the present study, the incidence of postoperative infection was much lower (2/31 patients). One case resulted in failure by developing a chronic infected nonunion, while the other case was successfully managed by surgical debridement and antibiotic therapy, resulting in a healed union and a good functional long-term outcome. Bony union were achieved in 96.7% of all cases (30/31) on average time of 3.5 months.

One case resulted in failure by developing a chronic infected nonunion, while the other case was successfully managed by surgical debridement and antibiotic therapy, resulting in a healed union and a good functional long-term outcome. Bony union were achieved in 96.7% of all cases (30/31) on average time of 3.5 months. The functional outcome measured by the Tscherne's criteria showed good results in 26/31 patients. Based on these findings, our data confirm the safety and efficacy of autologous bone grafting and compression plate fixation of fracture nonunions of the forearm [19]. Conclusion Autogenous cortical bone grafts were historically described as a successful modality for the reconstruction of traumatic segmental skeletal defects [20-23]. While the plate fixation of forearm fractures remains the gold standard, complications have been shown to occur in up to 28% of all patients [24]. One of the major challenges of long-term complications are forearm nonunions with bone loss and segmental defects. In the present study, we demonstrate the safety and efficacy of corticocancellous iliac crest bone grafting and compression plating for revision fixation of forearm fracture-nonunions, leading to excellent radiological and functional long-term outcomes. Competing interests The authors declare that they have no competing interests. Authors' contributions FBR, FF, HJAF, and WMA designed the study and were responsible for the clinical care of the patients. FBR and PFS wrote and edited the manuscript. All authors read and approved the final version of this paper.

Background The anal fistula, as a chronic inflammatory process, does not heal spontaneously. Although the conservative management, which consists of antibiotic therapy against the Gram-negative organisms and anaerobic bacteria, may be effective in the acute and early phase of the anal disease, surgery remains the elective treatment. Since 400 BC, when Hippocrates described a fistulotomy and the employment of a cutting seton made of horsehair, the surgical rationale has always been the same. Nowadays the main surgical options are fistulotomy, fistulectomy and loose or cutting seton insertion. Seton insertion is often performed only to prevent further abscess formation, and the laying-open remains the efficacious surgical treatment of the fistula-in-ano. On the other hand the best choice depends on the anatomical characteristics of the fistula, and particularly if the fistulous tract crosses the external sphincter. In this case the surgical dissection could lead to severe sphincter damage with sequent fecal incontinence if the sphincter injury is too extensive. From this point of view, in order to prevent such a sphincter damage causing fecal incontinence, the most fundamental issue is to quantify the remaining functional sphincter. Only when at least 1-2 cm of functional sphincter are saved from surgical dissection the laying-open can be performed without affecting fecal continence. Otherwise the seton drainage with secondary fistulotomy, staged fistulotomy or sliding flap advancement has to be considered in order to reduce the risk of postoperative incontinence. Anyway there are still many patients suffering from postoperative permanent disturbance in anal continence, which is mostly represented by loss of flatus control and soiling, and only seldom by severe fecal incontinence. Nevertheless, a temporary early postoperative incontinence, which improves within 2-3 weeks, is a frequent complication after surgery of any fistulous tract dissecting the sphincter [1-11] (Table 1). Another postoperative complication after surgical treatment of an anal fistula is the recurrence (0-9%) [1-11] (Table 1).

evertheless, a temporary early postoperative incontinence, which improves within 2-3 weeks, is a frequent complication after surgery of any fistulous tract dissecting the sphincter [1-11] (Table 1). Another postoperative complication after surgical treatment of an anal fistula is the recurrence (0-9%) [1-11] (Table 1). Basically it depends on an ineffective surgical treatment but also on the fistula etiology. For these reasons newer sphincter-saving approaches have been applied in the treatment of perianal fistula in order to avoid the risk of fecal incontinence, particularly in patients with high risk. Among these approaches the fibrin glue application is standing out because it is a simple and repeatable technique, whose success rate is improved by repeated injections, and does not interfere or compromise subsequent surgical options. Moreover the prolonged discomfort associated with wound dressing after surgery may be avoided. The first series studies about the fibrin glue treatment of anal fistulas were published by Abel and Hjortrup in the early 90' years [12,13]. Respectively they reported 60% (in 10 patients) and 52% (in 15 patients) healing rates and they both stressed out the importance of thorough curettage to remove all granulation tissue and debris, as well as a wide antibiotic administration. More recent studies do not report the same successful use of this technique, showing low healing rates [14-17]. The aim of this review is to assess the cure and complication rate of fibrin glue application compared to surgery alone for the treatment of perianal fistula.

ris, as well as a wide antibiotic administration. More recent studies do not report the same successful use of this technique, showing low healing rates [14-17]. The aim of this review is to assess the cure and complication rate of fibrin glue application compared to surgery alone for the treatment of perianal fistula. Table 1 Results and complications after surgical treatment of fistula-in-ano.

ris, as well as a wide antibiotic administration. More recent studies do not report the same successful use of this technique, showing low healing rates [14-17]. The aim of this review is to assess the cure and complication rate of fibrin glue application compared to surgery alone for the treatment of perianal fistula. Table 1 Results and complications after surgical treatment of fistula-in-ano. Study Patients (n) Recurrence (%) Disturbance in anal continence (%) Aguilar [1] 189 0.01 0 Bennett [2] 108 2 36 Hill [3] 626 1 4 Koscinski [4] 55 6 0 Khubchandani [5] 137 5.8 - Lilius [6] 150 5.5 13.5 Marks and Ritchie [7] 793 - 17-31 Mazier[8] 1000 3.9 0.01 McElwain [9] 1000 3.6 7-3.2 Parks and Stitz [10] 400 9 - Pearl [11] 1732 1.8 - Methods of meta-analysis Search methods for identification of studies We planned to search for published randomized and controlled clinical trials with no language restrictions, by using the following electronic databases: Cochrane Central Register of Controlled Trials (CENTRAL), MEDLINE (1950 onwards) and EMBASE (1980 onwards). The literature searches were carried out using medical subject headings (MeSH) and free-text word: "rectal fistula"; "anal fistula"; "fibrin adhesive"; "fibrin glue"; "fibrin sealant", "anal fistula plug", "collagen fistula plug". We also checked the reference lists of all the studies identified through the above mentioned methods. The abstracts presented to the following international scientific societies were hand searched: American College of Surgeons (2000 to 2007), American Society of Colon-Rectal Surgeons (1991 to 2007) and Società Italiana di Chirurgia (1985 to 2007), Società Italiana di Chirurgia Colon-Rettale (2006 to 2007), The Courrier de colo-proctologie.

ed to the following international scientific societies were hand searched: American College of Surgeons (2000 to 2007), American Society of Colon-Rectal Surgeons (1991 to 2007) and Società Italiana di Chirurgia (1985 to 2007), Società Italiana di Chirurgia Colon-Rettale (2006 to 2007), The Courrier de colo-proctologie. Data Extraction Two authors (RC, EF) assessed titles or abstracts of all the studies identified by the initial search and excluded irrelevant studies. Full text articles of potentially relevant studies and any studies with unclear methodology were obtained. The two authors assessed all these studies as to whether they met the inclusion criteria for this review, and they evaluated the method of randomization and the adequacy of allocation concealment. Disagreements on inclusion of the studies were solved by discussion and, if necessary, by involving an independent third author (FS). The following information were independently extracted by the two investigators (RC, EF) for each included study: the primary outcome, the number of event of interest, the population included, and information on quality measure including allocation concealment, blinding of outcome evaluators, intention to treat and balance of prognostic factors. Inclusion Criteria To be included in the analysis, studies had to compare conventional surgical treatment versus fibrin glue treatment in patients with anal fistulas.

Data Extraction Two authors (RC, EF) assessed titles or abstracts of all the studies identified by the initial search and excluded irrelevant studies. Full text articles of potentially relevant studies and any studies with unclear methodology were obtained. The two authors assessed all these studies as to whether they met the inclusion criteria for this review, and they evaluated the method of randomization and the adequacy of allocation concealment. Disagreements on inclusion of the studies were solved by discussion and, if necessary, by involving an independent third author (FS). The following information were independently extracted by the two investigators (RC, EF) for each included study: the primary outcome, the number of event of interest, the population included, and information on quality measure including allocation concealment, blinding of outcome evaluators, intention to treat and balance of prognostic factors. Inclusion Criteria To be included in the analysis, studies had to compare conventional surgical treatment versus fibrin glue treatment in patients with anal fistulas. Exclusion Criteria Studies were excluded from the analysis if: 1) the outcomes of interest were not reported for the two techniques, 2) it was impossible to extrapolate or calculate the necessary data from the published results, 3) there was considerable overlap between authors, centres, or patient cohorts evaluated in the published literature. Moreover the studies in which fibrin glue was used in the flap repair of anal fistulas were also excluded from this review.

ble to extrapolate or calculate the necessary data from the published results, 3) there was considerable overlap between authors, centres, or patient cohorts evaluated in the published literature. Moreover the studies in which fibrin glue was used in the flap repair of anal fistulas were also excluded from this review. Outcomes of Interest The following outcomes were used to compare the: 1. Clinical healing of fistula 2. Anal incontinence. Methodological quality EF and RC recorded whether the Authors of the trials used a sample size calculation, or they performed their analysis using an intention-to-treat method. Assessment of the methodological quality of the studies The review authors followed the instructions given in the Cochrane Handbook for Systematic Reviews of Interventions. Measures of treatment effect Dichotomous data were analyzed for relative risk ratio (RR), odds ratio (OR), and the absolute results were measured with the risk differences retrieved to calculate the odds ratio (OR) and 95% confidence intervals (CI) were calculated. The Mantel-Haenszel method was used for the meta-analysis. Results were presented on a forest plot graph. Assessment of heterogeneity Chi-squared test was used for heterogeneity assessment. If different trials used different scales, the results were standardized and then combined (i.e. standardized mean difference). Statistical Analysis We used Review Manager 5 to conduct the review.

Measures of treatment effect Dichotomous data were analyzed for relative risk ratio (RR), odds ratio (OR), and the absolute results were measured with the risk differences retrieved to calculate the odds ratio (OR) and 95% confidence intervals (CI) were calculated. The Mantel-Haenszel method was used for the meta-analysis. Results were presented on a forest plot graph. Assessment of heterogeneity Chi-squared test was used for heterogeneity assessment. If different trials used different scales, the results were standardized and then combined (i.e. standardized mean difference). Statistical Analysis We used Review Manager 5 to conduct the review. Results Eligible Studies There are currently two RCTs on this issue comparing conventional surgical treatment versus fibrin glue treatment in patients with anal fistulas [18,19] (Table 2). The assessment of quality of the studies was evaluated by two assessors (RC and FS) by considering: presence of detailed criteria for assignment of patients to the surgical or fibrin glue treatment group; absence of any difference between the two groups assessing the comparability on the basis of study design or analysis of group differences; adequacy of patients' follow-up (> 6 months). All recurrences after the fibrin glue treatment appear within 3 months and only occasionally they are detectable after 6 months [20]. Table 2 Characteristics of the studies considered.

Results Eligible Studies There are currently two RCTs on this issue comparing conventional surgical treatment versus fibrin glue treatment in patients with anal fistulas [18,19] (Table 2). The assessment of quality of the studies was evaluated by two assessors (RC and FS) by considering: presence of detailed criteria for assignment of patients to the surgical or fibrin glue treatment group; absence of any difference between the two groups assessing the comparability on the basis of study design or analysis of group differences; adequacy of patients' follow-up (> 6 months). All recurrences after the fibrin glue treatment appear within 3 months and only occasionally they are detectable after 6 months [20]. Table 2 Characteristics of the studies considered. Study Patients Treatment Outcomes Greco [18] 77 trans sphincteric anal fistulas 32 surgical treatments (cutting seton) vs. 45 fibrin glue Clinical healing Lindsey [19] 13 simple fistulas (low fistulas) and 29 complex fistulas 23 surgical treatments (fistulotomy) vs. 19 fibrin glue Clinical healing Results from analysis Even if it is not statistically relevant, the healing rate is higher in those patients who underwent conventional surgical treatment (odds ratio OR, 0.50; 95 percent confidence interval CI, 0.02-13.19; P = 0,68) (Figure 1). In this analysis there is a significant heterogeneity (chi-square = 14.14 - I2 = 93%), and therefore for the OR calculation we used the M-H Random test instead of the fixed one. Despite of the lower healing rates in the fibrin glue treatment group, no anal incontinence is noticed (odds ratio OR, 14.37; 95 percent confidence interval CI, 0.75-277.01; P = 0,08) (Figure 2). Also this result is not statistically relevant.

for the OR calculation we used the M-H Random test instead of the fixed one. Despite of the lower healing rates in the fibrin glue treatment group, no anal incontinence is noticed (odds ratio OR, 14.37; 95 percent confidence interval CI, 0.75-277.01; P = 0,08) (Figure 2). Also this result is not statistically relevant. Figure 1 Healing: conventional surgical treatment versus fibrin glue treatment in patients with anal fistulas. Figure 2 Incontinence: conventional surgical treatment versus fibrin glue treatment in patients with anal fistulas. Subgroup analysis We performed two subgroup analyses assessing other sphincter saving approaches for the treatment of fistulas-in-ano in patients with high risk of postoperative disturbance in anal continence: - Fibrin glue in combination with intra-adhesive antibiotics vs. fibrin glue alone. - Anal fistula plug vs. fibrin glue. Fibrin glue in combination with intra-adhesive antibiotics vs. fibrin glue alone Inclusion Criteria of subgroup analysis To be included in this subgroup analysis the studies had to compare the fibrin glue application in combination with intra-adhesive antibiotics treatment vs. a simple fibrin glue treatment of fistulas-in ano. Eligible Studies for subgroup analysis Using the key words listed above, we identified 253 abstracts. The examination of all the abstracts and their references on the basis of the inclusion criteria of this subgroup analysis, only gave us one valid study to be analyzed [21].

Fibrin glue in combination with intra-adhesive antibiotics vs. fibrin glue alone Inclusion Criteria of subgroup analysis To be included in this subgroup analysis the studies had to compare the fibrin glue application in combination with intra-adhesive antibiotics treatment vs. a simple fibrin glue treatment of fistulas-in ano. Eligible Studies for subgroup analysis Using the key words listed above, we identified 253 abstracts. The examination of all the abstracts and their references on the basis of the inclusion criteria of this subgroup analysis, only gave us one valid study to be analyzed [21]. Results from subgroup analysis The analysis of Singer's study did not show differences in healing between treatment with fibrin glue in combination with intra-adhesive antibiotics and fibrin glue alone (odds ratio OR, 1.26; 95 percent confidence interval CI, 0.47-3.36; P = 0,65) [21] (Figure 3). Figure 3 Fibrin glue in combination with intra-adhesive antibiotics vs. fibrin glue alone. Anal fistula plug vs. fibrin glue Inclusion Criteria of subgroup analysis To be included in this subgroup analysis the studies had to compare anal fistula plug treatment versus fibrin glue treatment of fistulas-in ano. Eligible Studies for subgroup analysis The examination of the 253 abstracts and their references previously identified on the basis of the inclusion criteria of this subgroup analysis, allowed us to get only 1 valid study to be analyzed [22].

Anal fistula plug vs. fibrin glue Inclusion Criteria of subgroup analysis To be included in this subgroup analysis the studies had to compare anal fistula plug treatment versus fibrin glue treatment of fistulas-in ano. Eligible Studies for subgroup analysis The examination of the 253 abstracts and their references previously identified on the basis of the inclusion criteria of this subgroup analysis, allowed us to get only 1 valid study to be analyzed [22]. Results from subgroup analysis In Ky's study the healing rate is statistically significant in the patients who underwent the anal fistula plug treatment of fistulas-in ano (odds ratio OR, 9.75; 95 percent confidence interval CI, 1.38-68.78; P = 0,02) [22] (Figure 4). Figure 4 Anal fistula plug vs. fibrin glue. Discussion Fibrin glue treatment of anal fistulas is simple, safe, and painless and the injections can be repeated to increase the healing rate without preventing from other eventual following surgical procedures. For these reasons in the past decade this technique became increasingly popular, but today many doubts about fibrin glue procedure still remain because of its poor long-term results [23-25].

the injections can be repeated to increase the healing rate without preventing from other eventual following surgical procedures. For these reasons in the past decade this technique became increasingly popular, but today many doubts about fibrin glue procedure still remain because of its poor long-term results [23-25]. Our statistical analysis confirms the poor long-term results in the patients who underwent fibrin glue treatment rather than surgery, considering both healing and not-healing after an adequate 6-month follow-up, regardless of the fact that the not-healing rate would be a recurrence or not. We did not calculate any recurrence rate after fibrin glue treatment, as we could not find satisfactory data in literature about accurate recurrence rate in the two included studies. Another important point to consider about fistula healing is that the closure of the external skin wound does not always mean complete healing. Buchanan et al. reported a prospective clinical trial with 22 patients presenting the possibility to establish a complete healing of idiopathic complex anorectal fistulas after fibrin glue treatment. The Authors evaluated the fistula tract healing on the basis of the clinical examination and magnetic resonance imaging (STIR sequence MRI long-term can detect deep persistence). Despite of the skin healing in 77% of the patients at 14 days after treatment, 3 cases (14%) were clinically healed at 16 months and only 2 cases (9%) were radiologically healed at 16 months [25]. In Table 3 we analyzed the different fistulas etiologies in order to estimate the prognostic value of the etiology in the fibrin glue treatment of the perianal fistula [12,15,16,19,20,26-28]. Even though most of the studies in literature include fistulas of mixed etiologies, cryptoglandular fistulas are the major group, whereas no-cryptoglandular fistulas represent only a small amount. The fistula healing rate varies widely (0-100%) in Crohn disease patients. In HIV-associated and rectovaginal fistulas the healing rate appears to be poor, although the number of patients considered is too low to obtain certain results. The number of ileal-pouch anal anastomosis fistula patients treated with fibrin glue is very low as well, but in these patients the results seem to be good. On the other hand it was possible to estimate the prognostic value of the anatomical characteristics of the fistula, particularly comparing simple or complex fistulas (Table 4) [12,15,16,19,20,23-28].

mosis fistula patients treated with fibrin glue is very low as well, but in these patients the results seem to be good. On the other hand it was possible to estimate the prognostic value of the anatomical characteristics of the fistula, particularly comparing simple or complex fistulas (Table 4) [12,15,16,19,20,23-28]. Our analysis shows that simple fistulas have a better healing rate than complex fistulas. Table 3 Healing rates after treatment with fibrin glue for fistula in ano. Cryptoglandular Crohn's disease HIV Rectovaginal Ileal-pouch anal anastomosis Abel [12] 2/3 (66%) 1/3 (33%) Sentovich [15] 25/36 (70%) 4/5 (80%) Loungnarath [16] 5/22 (23%) 4/13 (31%) 1/3 (33%) 3/4 (75%) Lindsey [19] 10/17 (59%) 2/2 (100%) Cintron [20] 44/68 (65%) 2/6 (33%) 1/2 (50%) 1/3 (33%) Venkatesh [26] 12/15 (80%) 0/6 (0%) Patrij [27] 51/69 (74%) Zmora [28] 2/10 (20%) 2/5 2/4 (50%) Total 154/240 (64%) 13/37 (35%) 1/2 (50%) 2/6 (33%) 5/8 (63%) Table 4 Healing rates after treatment with fibrin glue in complex and simplex fistula Complex only Simple and complex Healed Not healed Healed Not healed Abel [12] 6 4 Sentovich [15] 33 15 Loungnarath [16] 12 27 Lindsey [19] 9 13 Cintron [20] 48 31 Zmora [23] 32 28 Tinay [24] 19 25 Buchanan [25] 3 19 Venkatesh [26] 18 12 Patrij [27] 51 18 Zmora [28] 8 16

Total 154/240 (64%) 13/37 (35%) 1/2 (50%) 2/6 (33%) 5/8 (63%) Table 4 Healing rates after treatment with fibrin glue in complex and simplex fistula Complex only Simple and complex Healed Not healed Healed Not healed Abel [12] 6 4 Sentovich [15] 33 15 Loungnarath [16] 12 27 Lindsey [19] 9 13 Cintron [20] 48 31 Zmora [23] 32 28 Tinay [24] 19 25 Buchanan [25] 3 19 Venkatesh [26] 18 12 Patrij [27] 51 18 Zmora [28] 8 16 Total 88 (42.5%) 119 (57.5%) 151 (63%) 89 (37%) Recent articles report encouraging results, which will need further studies to be confirmed, in the repair of anal fistulas by using Surgisis AFP (Anal Fistula Plug: Cook Medical Incorporated, Bloomington, IN), a bioabsorbable plug derived from porcine small submucosa (SIS) [29,30]. The plug is placed into the fistula tract and sutured to the internal opening. SIS promotes tissue remodelling while being slowly incorporated into the body during a 3 to 6-month period. Surgisis AFP long-term closure rate is significantly higher in patients with simple fistulas than complex ones and with non-Crohn disease versus Crohn disease [22]. The major complication of Anal Fistula Plug is a severe perianal sepsis (14.7% Ky 2008 - 29% Lawes 2008) requiring surgical drainage and removal of the plug [22,30].

FP long-term closure rate is significantly higher in patients with simple fistulas than complex ones and with non-Crohn disease versus Crohn disease [22]. The major complication of Anal Fistula Plug is a severe perianal sepsis (14.7% Ky 2008 - 29% Lawes 2008) requiring surgical drainage and removal of the plug [22,30]. Conclusion Nowadays the laying-open, seton insertion, staged fistulotomy and sliding flap advancement are still the main pillar of perianal fistula surgery, while fibrin glue alone or in combination with intra-adhesive antibiotics and the anal fistula plug would rather be used in the patients with high risk of postoperative disturbance of anal continence. In literature we found only two randomized controlled trials comparing conventional surgical management versus fibrin glue treatment in patients with anal fistulas. Although from our statistical analysis we could not find any statistically significant result, the healing rate is higher in the patients who underwent the conventional surgical treatment (P = 0,68), and the anal incontinence rate is very low in the fibrin glue treatment group (P = 0,08). Anyway these limited data do not support the use of fibrin glue. Besides, in our subgroup analysis the use of fibrin glue in combination with intra-adhesive antibiotics does not improve the healing rate (P = 0.65), whereas the anal fistula plug treatment compared to the fibrin glue treatment shows good results (P = 0,02), although the poor number of patients treated does not lead to any significant conclusion. Our systematic review underlines the need of new RCTs upon this issue.

tics does not improve the healing rate (P = 0.65), whereas the anal fistula plug treatment compared to the fibrin glue treatment shows good results (P = 0,02), although the poor number of patients treated does not lead to any significant conclusion. Our systematic review underlines the need of new RCTs upon this issue. Competing interests The Authors state that none of the authors involved in the manuscript preparation has any conflicts of interest towards the manuscript itself, neither financial nor moral conflicts. Besides none of the authors received support in the form of grants, equipment, and/or pharmaceutical items. Authors' contributions All authors contributed equally to this work, read and approved the final manuscript.

Background In the study described herein, a rabbit model of arterial anastomotic bleeding was used to examine the effect of recombinant human thrombin (rThrombin) concentration on time to hemostasis (TTH) under varying conditions of pharmacologic anticoagulation and platelet inhibition.

Background In the study described herein, a rabbit model of arterial anastomotic bleeding was used to examine the effect of recombinant human thrombin (rThrombin) concentration on time to hemostasis (TTH) under varying conditions of pharmacologic anticoagulation and platelet inhibition. Thrombin has been widely used as a topical hemostatic agent during surgical procedures since the 1940s [1,2]. The efficacy of topical thrombin at a concentration of 1000 IU/mL was recently demonstrated in clinical trials, although no clinical trials have compared the effects of different concentrations of topical thrombin on hemostatic efficacy [3-5]. The critical role of thrombin concentration in fibrin clot formation has been demonstrated in a number of in vitro settings; clots formed in the presence of higher concentrations of thrombin had more tightly packed fibrin strands and were more resistant to fibrinolysis, when compared with clots formed in the presence of lower concentrations of thrombin [6-8]. In animal models, TTH was dependent on rThrombin concentration; for example, gauze soaked in rThrombin at concentrations from 500 to 2000 IU/mL stopped bleeding significantly faster in a rabbit model of hepatic bleeding than gauze soaked with rThrombin 100 IU/mL [9,10]. In another study, 125 IU/mL human plasma-derived thrombin did not appear to be an adequate concentration for establishing durable clots in a healthy porcine model of liver injury [11]. In that study, human plasma-derived thrombin (125 IU/mL) with an absorbable gelatin sponge (AGS) improved the incidence of hemostasis compared to saline with AGS. However, rebleeding was also observed at a number of sites during the 12 minute evaluation period, thus thrombin at 125 IU/mL did not appear to be an adequate concentration for establishing durable clots. Although there is substantial inter-species heterogeneity in coagulation capacity, collectively these results suggest that there is an optimum concentration of topical thrombin needed at the wound for rapid TTH and durable clotting.

25 IU/mL did not appear to be an adequate concentration for establishing durable clots. Although there is substantial inter-species heterogeneity in coagulation capacity, collectively these results suggest that there is an optimum concentration of topical thrombin needed at the wound for rapid TTH and durable clotting. Clopidogrel is a potent platelet function inhibitor and is widely prescribed, ranking second in a 2008 list of worldwide pharmaceutical sales [12]. There are a wide variety of opinions regarding the timing for discontinuing clopidogrel before undergoing a surgical procedure. After recent reports of rebound cardiac risk for 90 days following abrupt discontinuation of clopidogrel, it has been argued that stopping clopidogrel in patients with a high risk of perioperative myocardial infarction may affect the incidence of perioperative cardiac events [13]. As a consequence, increasing numbers of patients are presenting for emergent, urgent, or elective surgical procedures with significant clopidogrel platelet inhibition as a result of a reluctance or inability to withhold clopidogrel prior to surgery. The effect of anti-platelet medications on TTH has not been thoroughly evaluated, although recent clinical trial data suggested that TTH was delayed in patients being treated for conditions for which anti-platelet medications are commonly-used when compared with patients who were receiving treatment for conditions for which anti-platelet medications are less-commonly used [14].

thoroughly evaluated, although recent clinical trial data suggested that TTH was delayed in patients being treated for conditions for which anti-platelet medications are commonly-used when compared with patients who were receiving treatment for conditions for which anti-platelet medications are less-commonly used [14]. Based on the observations described above, we hypothesized that adoption of the standard 1000 IU/mL thrombin concentration gained acceptance in clinical practice over the last 50 years because of its observed efficacy over a range of clinical settings. Although thrombin is a remarkably potent effector of hemostasis, the conditions in which exogenous thrombin is applied vary widely, with differences in blood flow, the amount of hemodilution, and the presence or absence of pharmacological anticoagulation agents. Considering these variables, one would predict that topical thrombin applied at lower concentrations may be ineffective in some clinical settings.

exogenous thrombin is applied vary widely, with differences in blood flow, the amount of hemodilution, and the presence or absence of pharmacological anticoagulation agents. Considering these variables, one would predict that topical thrombin applied at lower concentrations may be ineffective in some clinical settings. In this study, the effect of rThrombin concentration on TTH under varying conditions of pharmacologic anticoagulation and platelet inhibition was evaluated by performing in vivo experiments with a range of thrombin concentrations in a model of arterial anastomotic bleeding, in rabbits treated with heparin or with heparin + clopidogrel. Reliability of clot formation was assessed by examining variability in TTH. After hemostasis was achieved, we examined differences in clot durability resulting from various concentrations of rThrombin by evaluating clot burst in animals treated with heparin + clopidogrel. We also performed an in vitro assessment of clot strength and the rate of clot formation by adding different amounts of rThrombin activity to heparin + clopidogrel rabbit blood samples, and evaluating the viscoelastic properties of the clot using modified thromboelastography (TEG) [15].

ated with heparin + clopidogrel. We also performed an in vitro assessment of clot strength and the rate of clot formation by adding different amounts of rThrombin activity to heparin + clopidogrel rabbit blood samples, and evaluating the viscoelastic properties of the clot using modified thromboelastography (TEG) [15]. Methods Animals Female New Zealand White rabbits (Charles River Laboratories, Hollister, CA), weighing 2.0 to 3.8 kg and approximately 12 weeks of age, were used for these studies. Six animals were used in each study. Animals were acclimated to the facility for 7 to 10 days before the surgical procedure and were maintained in good condition. All animal studies were performed under approved protocols in accordance with the Guide for the Care and Use of Laboratory Animals (US DHHS publication NIH 86-23).

imals were used in each study. Animals were acclimated to the facility for 7 to 10 days before the surgical procedure and were maintained in good condition. All animal studies were performed under approved protocols in accordance with the Guide for the Care and Use of Laboratory Animals (US DHHS publication NIH 86-23). In Vivo Study Design Clinical vascular anastomotic bleeding was simulated in a rabbit arteriovenous (AV) shunt model. The experimental design evaluated the dependence of TTH on the applied rThrombin concentration in the presence of heparinization, both with and without clopidogrel platelet inhibition. All rThrombin concentrations were applied using an AGS. Anticoagulation with heparin was used in this model in order to maintain shunt patency. An initial study, evaluating several rThrombin concentrations (31.25, 62.5, 125, and 1000 IU/mL) or placebo, was performed in rabbits treated with heparin. In further studies, hemostasis was evaluated in rabbits treated with heparin + clopidogrel using two concentrations of rThrombin (125 and 1000 IU/mL) or placebo. The investigator was blinded to treatment during the procedure and at the time of hemostasis evaluation, and treatment (placebo, rThrombin) was randomly assigned.

urther studies, hemostasis was evaluated in rabbits treated with heparin + clopidogrel using two concentrations of rThrombin (125 and 1000 IU/mL) or placebo. The investigator was blinded to treatment during the procedure and at the time of hemostasis evaluation, and treatment (placebo, rThrombin) was randomly assigned. Multiple measurements of TTH were performed on each rabbit. Each measurement was performed on a new polytetrafluoroethylene (PTFE) graft. In the initial study of rabbits treated with heparin, 12 grafts were evaluated for placebo and 8 grafts were evaluated for each rThrombin concentration. In the heparin + clopidogrel treated rabbits, 10 grafts were evaluated for placebo and 14 grafts were evaluated for each rThrombin concentration. For each graft, physiological parameters critical to hemostasis (i.e., model control parameters) were measured. These included body weight, activated partial thromboplastin time (aPTT; measured at baseline and after heparin treatment), body temperature, blood flow, and mean arterial pressure (MAP). Comparability of these parameters was used to demonstrate consistency among animals and standardization between grafts, and was necessary in order to consider each measure of TTH in an individual graft as an independent observation.

ter heparin treatment), body temperature, blood flow, and mean arterial pressure (MAP). Comparability of these parameters was used to demonstrate consistency among animals and standardization between grafts, and was necessary in order to consider each measure of TTH in an individual graft as an independent observation. Treatments: rThrombin, Placebo, and Absorbable Gelatin Sponge On the day of the AV graft procedure, a 5000 IU vial of rThrombin (RECOTHROM®, ZymoGenetics, Inc., Seattle, WA) was reconstituted in 5 mL of sterile saline, yielding a 1000 IU/mL rThrombin solution. The 1000 IU/mL solution was diluted with placebo to yield 31.25 IU/mL, 62.5 IU/mL, and 125 IU/mL rThrombin solutions. Placebo solution consisted of the excipients contained in the formulation for rThrombin [16]. AGS (Gelfoam, Pharmacia & Upjohn Company, Kalamazoo, MI, size 100) was cut into 2 × 4 × 1 cm strips and combined with rThrombin or placebo solutions in accordance with the instructions in the rThrombin package insert. The median final volume of rThrombin or placebo mixed per sponge was 0.762 mL (range 0.756-0.775 mL). Clopidogrel Administration Animals received 20 mg/kg oral doses of clopidogrel bisulfate (Plavix®, Bristol-Meyers Squibb/Sanofi Aventis, Bridgewater, NJ) once daily for three days prior to AV shunt placement and hemostasis evaluation. A 75 mg clopidogrel tablet was dissolved in 3 mL of sterile water. Each animal received between 1.6 and 2.2 mL of the 25 mg/mL solution per day by gavage. Surgical procedures were performed 2 hours after the final dose of clopidogrel.

NJ) once daily for three days prior to AV shunt placement and hemostasis evaluation. A 75 mg clopidogrel tablet was dissolved in 3 mL of sterile water. Each animal received between 1.6 and 2.2 mL of the 25 mg/mL solution per day by gavage. Surgical procedures were performed 2 hours after the final dose of clopidogrel. Rabbit AV Shunt Procedure AV shunts were surgically created by using a modification of methods previously described [17]. Briefly, rabbits were immobilized with ketamine hydrochloride (Fort Dodge Animal Health, Ft. Dodge, IA; 50 mg/kg, intramuscular injection), and anesthesia was maintained using inhaled isoflurane (1% to 2%). The animals were placed on a water-jacketed heating pad maintained at 37°C during the experimental period and core body temperature was monitored. The animal's right femoral artery was cannulated for measurement of mean arterial pressure (MAP) and the left femoral vein was cannulated for heparin administration. An AV shunt that linked the blood flow of the left carotid artery and the right jugular vein was created by isolating and cannulating these vessels with a 3 to 4 cm length of Micro-Renathane tubing (MRE 080, 0.080" O.D. × 0.040" I.D., Braintree Scientific, Braintree, MA) connected to a 15 cm length of silicone catheter tubing (7-French, 0.078" I.D. × 0.125" O.D., Access Technologies, Skokie, IL). The catheters were exteriorized and connected with a 2 to 2.5 cm PTFE graft segment (3 mm, SN AFEP 7108, Bard Peripheral Vascular Incorporated, Tempe, AZ). Blood flow through the shunt was measured using a Transonic System Incorporated Flowmeter, model TS410 (Ithaca, NY). Each rabbit received 100 U/kg heparin (Abraxis Pharmaceutical Products, Schaumburg, IL) by intravenous (IV) bolus injection, followed by a continuous 50 U/kg/mL infusion of heparin at a flow rate of 5 mL/hr via the femoral vein; aPTT was measured prior to heparin treatment and 5-10 minutes after IV heparin bolus.

NY). Each rabbit received 100 U/kg heparin (Abraxis Pharmaceutical Products, Schaumburg, IL) by intravenous (IV) bolus injection, followed by a continuous 50 U/kg/mL infusion of heparin at a flow rate of 5 mL/hr via the femoral vein; aPTT was measured prior to heparin treatment and 5-10 minutes after IV heparin bolus. Suture Hole Bleed and Measurement of TTH To achieve consistent conditions between AV grafts, MAP was maintained at approximately 55 mm Hg (+/- 2 mm Hg) by adjusting depth of anesthesia, and aPTT was maintained at a value > 400 seconds by heparin infusion. Suture hole bleeding was assessed by puncturing the center section of the PTFE graft segment with two 18 inch silk suture needles (4-0 needles; reverse cutting needle size P-3, stock number 641G, Ethicon Incorporated, Somerville, NJ), creating 4 needle holes. The AGS containing 0.77 mL of rThrombin or placebo was immediately wrapped around the graft, completely covering the bleeding surfaces (suture holes). The wetted AGS was then covered with gauze sponges, and continuous digital pressure was applied. After 60 seconds the gauze sponges were visually inspected for bleeding. If cessation of bleeding had not occurred, new gauze sponges were applied with digital pressure and this process was repeated until no visible blood was observed on the gauze sponges. TTH was recorded in seconds. If TTH was not achieved within 300 seconds, the experiment was terminated. At the conclusion of the assessment period, the catheters were clamped, flushed with saline, the PTFE segment was removed, and a fresh PTFE segment was inserted. Baseline measurements were obtained for MAP, core temperature, and aPTT. Additional heparin (50 U/kg, IV bolus) was injected as needed to maintain an aPTT > 400 seconds. If a MAP of 55 mm Hg could not be maintained by adjusting the depth of anesthesia, the animal was euthanized.

, and a fresh PTFE segment was inserted. Baseline measurements were obtained for MAP, core temperature, and aPTT. Additional heparin (50 U/kg, IV bolus) was injected as needed to maintain an aPTT > 400 seconds. If a MAP of 55 mm Hg could not be maintained by adjusting the depth of anesthesia, the animal was euthanized. Suture Hole Clot Burst Assessment In the rabbits treated with heparin + clopidogrel, clot burst at the suture needle puncture sites was assessed after the 300 second assessment of TTH, if hemostasis had been achieved. The procedure involved clamping off blood flow to the jugular vein catheter approximately 2-3 cm downstream from the AV shunt graft for a period of 10 seconds. Complete obstruction of blood flow created an increase in blood pressure at the site of the blood clot. Clot burst observation during the 10 second period was recorded as either positive (blood seeped through the AGS onto the gauze sponge) or negative (no seepage was observed). Clot burst was not assessed in rabbits treated with heparin alone. In Vitro Study Design The viscoelastic properties of blood clots were assessed in ex-vivo blood samples. TEG was used to measure the mechanical resistance to clot formation as a function of time and provided an in vitro correlate for the effects of rThrombin concentration on the rate of clot formation and clot strength.

In Vitro Study Design The viscoelastic properties of blood clots were assessed in ex-vivo blood samples. TEG was used to measure the mechanical resistance to clot formation as a function of time and provided an in vitro correlate for the effects of rThrombin concentration on the rate of clot formation and clot strength. TEG Samples from Rabbit AV Shunt Model Blood samples from the central ear artery were collected in citrate. For the 3 rabbits evaluated, a sample was collected 2 hours after clopidogrel dosing and heparin (1 U/mL) was added ex vivo. The citrated rabbit blood was recalcified according to a standard protocol with 0.2 M CaCl2 plus normal saline just prior to the TEG assay (Haemoscope Corporation, Niles, IL). The blood clotted within normal parameters for rabbit blood without exogenous rThrombin. A modified method for rThrombin clot initiation was empirically derived and covered a ten-fold range of rThrombin enzymatic activity (0.76 to 6.1 IU added per mL of rabbit blood). Thromboelastograph Parameters TEG measures clot viscoelastic properties by recording mechanical resistance to clot formation as a function of time. • The rate of clot formation (i.e., development of the fibrin matrix, factor XIII cross-linking, platelet incorporation, and platelet aggregation) is obtained by measuring the alpha angle (α) [15]. The α is derived from the slope of the maximum amplitude (MA) of the measured resistance to clot formation vs. time.

Thromboelastograph Parameters TEG measures clot viscoelastic properties by recording mechanical resistance to clot formation as a function of time. • The rate of clot formation (i.e., development of the fibrin matrix, factor XIII cross-linking, platelet incorporation, and platelet aggregation) is obtained by measuring the alpha angle (α) [15]. The α is derived from the slope of the maximum amplitude (MA) of the measured resistance to clot formation vs. time. • Clot firmness or strength is measured by the MA and is highly influenced by platelet number and function [15]. Clot strength (shear elastic modulus) is represented by G and is calculated from the amplitude. Maximum clot strength is derived as follows: The relationships between both the rate of clot formation (α) and clot strength (Gmax) and the amount of rThrombin activity added for clot initiation were assessed. Statistical Analyses Statistical analyses are described below for each experiment and were performed using SAS software (version 9.2, SAS Institute Inc., Cary, NC). Statistical significance was defined as p-values < 0.05. The number of grafts per treatment group was not based on formal sample size calculations or statistical power.

nalyses Statistical analyses are described below for each experiment and were performed using SAS software (version 9.2, SAS Institute Inc., Cary, NC). Statistical significance was defined as p-values < 0.05. The number of grafts per treatment group was not based on formal sample size calculations or statistical power. Comparison of TTH Each measurement of TTH in an individual graft was considered an independent measurement. That is, the experimental system (rabbit and shunt) was considered uniform except for the presence or absence of clopidogrel. The effect of different rThrombin concentrations on TTH was evaluated with the Kaplan-Meier method. Comparisons of TTH between different concentrations of rThrombin were performed using the log-rank test. TTH times of > 300 seconds were censored. Variability of TTH The reliability of hemostasis onset observed after treatment with 125 IU/mL or 1000 IU/mL rThrombin was compared for rabbits treated with heparin and for rabbits treated with heparin + clopidogrel by analyzing the variability in TTH. Hypotheses of equal variability in TTH for rThrombin concentrations of 125 IU/mL and 1000 IU/mL were tested using the folded F-test. Incidence of Clot Burst After hemostasis had been achieved, we examined potential differences in clot durability induced by different concentrations of rThrombin by evaluating clot burst in rabbits treated with heparin + clopidogrel. Fisher's exact test was used to compare the difference in incidence of clot burst after treatment with 125 IU/mL or 1000 IU/mL rThrombin.

chieved, we examined potential differences in clot durability induced by different concentrations of rThrombin by evaluating clot burst in rabbits treated with heparin + clopidogrel. Fisher's exact test was used to compare the difference in incidence of clot burst after treatment with 125 IU/mL or 1000 IU/mL rThrombin. In Vitro TEG Assays Raw TEG data are presented. Pearson correlation coefficients were calculated to assess the relationship between the amount of rThrombin added and the rate of clot formation and clot strength.

chieved, we examined potential differences in clot durability induced by different concentrations of rThrombin by evaluating clot burst in rabbits treated with heparin + clopidogrel. Fisher's exact test was used to compare the difference in incidence of clot burst after treatment with 125 IU/mL or 1000 IU/mL rThrombin. In Vitro TEG Assays Raw TEG data are presented. Pearson correlation coefficients were calculated to assess the relationship between the amount of rThrombin added and the rate of clot formation and clot strength. Results In Vivo Experiments A strong concentration-dependent effect of rThrombin on TTH was observed in an initial study in rabbits treated with heparin. Kaplan-Meier estimates of median TTH values were 73 seconds (1000 IU/mL rThrombin), 78 seconds (125 IU/mL), 133 seconds (62.5 IU/mL), 214 seconds (31.25 IU/mL), and > 265 seconds (placebo). Median TTH for the placebo group could not be determined because hemostasis was not achieved before the end of the observation period for half of the 12 grafts (i.e., half the TTH values were > 300 seconds). TTH decreased significantly with increasing rThrombin concentration (p < 0.001). A pairwise log-rank test revealed that TTH after treatment with 1000 IU/mL rThrombin (median 73 seconds; 95% CI: 65, 75) was shorter than after treatment with 125 IU/mL rThrombin (median 78 seconds; 95% CI: 70, 95; p = 0.007). In addition, the reliability of hemostasis onset achieved in grafts treated with 1000 IU/mL rThrombin was greater than that observed in grafts treated with 125 IU/mL rThrombin, as indicated by reduced variability in TTH at the 1000 IU/mL rThrombin concentration in heparin treated animals (p < 0.001).

0, 95; p = 0.007). In addition, the reliability of hemostasis onset achieved in grafts treated with 1000 IU/mL rThrombin was greater than that observed in grafts treated with 125 IU/mL rThrombin, as indicated by reduced variability in TTH at the 1000 IU/mL rThrombin concentration in heparin treated animals (p < 0.001). A concentration dependent effect of rThrombin on TTH was also observed in subsequent studies in rabbits treated with heparin + clopidogrel (Figure 1). Kaplan-Meier estimates of median TTH were 71 seconds (1000 IU/mL rThrombin), 177 seconds (125 IU/mL), and 271 seconds (placebo). TTH decreased significantly with increasing rThrombin concentration (p < 0.001). A pairwise log-rank test revealed that TTH was significantly shorter after treatment with 1000 IU/mL (median 71 seconds; 95% CI: 65, 77) than after treatment with 125 IU/mL rThrombin (median 177 seconds; 95% CI: 127, 218; p <  0.001). Figure 1 Kaplan-Meier estimates for the probability of bleeding after treatment with different concentrations of rThrombin or placebo. TTH was measured in the rabbit AV shunt model of anastomotic bleeding after treatment with heparin + clopidogrel. PTFE graft segments were punctured, and AGS containing a standardized volume of 1000 IU/mL or 125 IU/mL rThrombin or placebo was applied to the bleeding surface. The TTH observation period extended from 60 seconds to 300 seconds. 1000 IU/mL rThrombin: N = 14 grafts. 125 IU/mL rThrombin: N = 14 grafts. Placebo: N = 10 grafts.

raft segments were punctured, and AGS containing a standardized volume of 1000 IU/mL or 125 IU/mL rThrombin or placebo was applied to the bleeding surface. The TTH observation period extended from 60 seconds to 300 seconds. 1000 IU/mL rThrombin: N = 14 grafts. 125 IU/mL rThrombin: N = 14 grafts. Placebo: N = 10 grafts. Looking specifically at the response to 1000 IU/mL rThrombin, median TTH was similar for both treatment regimens (heparin: 73 seconds; heparin + clopidogrel: 71 seconds; p = 0.52). In contrast, median TTH in grafts treated with 125 IU/mL rThrombin was prolonged after heparin + clopidogrel treatment (177 seconds) compared with heparin treatment alone (78 seconds; p < 0.001). The reliability of hemostasis onset achieved in grafts treated with 1000 IU/mL rThrombin was greater than that observed in grafts treated with 125 IU/mL rThrombin, as indicated by reduced variability in TTH at the 1000 IU/mL rThrombin concentration in heparin + clopidogrel treated animals (p < 0.001; Figure 2). Variability in the placebo group could not be estimated because hemostasis was not achieved before the end of the observation period for all of the grafts (i.e., some TTH values in the placebo group were > 300 seconds).

e 1000 IU/mL rThrombin concentration in heparin + clopidogrel treated animals (p < 0.001; Figure 2). Variability in the placebo group could not be estimated because hemostasis was not achieved before the end of the observation period for all of the grafts (i.e., some TTH values in the placebo group were > 300 seconds). Figure 2 Variability in time to hemostasis (TTH) after treatment with different concentrations of rThrombin or placebo. TTH was measured in the rabbit AV shunt model of anastomotic bleeding after treatment with heparin + clopidogrel. PTFE graft segments were punctured, and AGS containing 1000 IU/mL or 125 IU/mL of rThrombin or placebo was applied to the bleeding site. The TTH observation period extended from 60 seconds to 300 seconds, as indicated by the dashed lines. Each diamond represents the TTH value for a single graft. Four grafts in the placebo group had not achieved hemostasis at the end of the observation period and are not included in the graph (i.e. 4 TTH values were > 300 seconds). All grafts treated with 125 IU/mL or 1000 IU/mL rThrombin achieved hemostasis within the observation period.

sents the TTH value for a single graft. Four grafts in the placebo group had not achieved hemostasis at the end of the observation period and are not included in the graph (i.e. 4 TTH values were > 300 seconds). All grafts treated with 125 IU/mL or 1000 IU/mL rThrombin achieved hemostasis within the observation period. Clot durability was assessed in animals treated with heparin + clopidogrel. In these studies, multiple consecutive AV shunt grafts were placed and different topical rThrombin or placebo treatments were administered in each animal. The durability of clots formed in grafts treated with 1000 IU/mL rThrombin was greater than that observed in grafts treated with 125 IU/mL rThrombin, as indicated by reduced incidence of clot burst. Clots that formed in the presence of 1000 IU/mL rThrombin were significantly less likely to rupture during clot burst assessment (0% of clots ruptured) than those formed in the presence of 125 IU/mL rThrombin (79% ruptured; p < 0.001). The incidence of clot burst was comparable in the presence of 125 IU/mL rThrombin (79% of clots ruptured) or placebo (80% ruptured). Representative images of clots that ruptured (upper panel, 125 IU/mL rThrombin) and clots that maintained hemostasis (lower panel, 1000 IU/mL rThrombin) are presented in Figure 3. Ruptured clots (top panel) have blood on the gauze sponges and under the grafts that is not present when hemostasis is maintained (bottom panel).

sentative images of clots that ruptured (upper panel, 125 IU/mL rThrombin) and clots that maintained hemostasis (lower panel, 1000 IU/mL rThrombin) are presented in Figure 3. Ruptured clots (top panel) have blood on the gauze sponges and under the grafts that is not present when hemostasis is maintained (bottom panel). Figure 3 Clot burst assessment of AV shunt grafts treated with different concentrations of rThrombin. Rabbits were treated with heparin + clopidogrel. The grafts were wrapped with AGS containing either 125 IU/mL rThrombin (top row) or 1000 IU/mL rThrombin (bottom row). After hemostasis was achieved, blood flow was clamped for 10 seconds, and clot burst was assessed. Images were obtained after the 10 second assessment period. Upper images show sponge saturation and bleed-thru in representative grafts whose hemostasis did not survive clot burst challenge (125 IU/mL rThrombin; 11 of 14 grafts ruptured [79%]). Lower images show representative grafts that maintained hemostasis following clot burst challenge (1000 IU/mL rThrombin; 0 of 14 grafts ruptured [0%]). In the placebo group, most grafts did not maintain hemostasis (8 of 10 grafts ruptured [80%]; data not shown).

ge (125 IU/mL rThrombin; 11 of 14 grafts ruptured [79%]). Lower images show representative grafts that maintained hemostasis following clot burst challenge (1000 IU/mL rThrombin; 0 of 14 grafts ruptured [0%]). In the placebo group, most grafts did not maintain hemostasis (8 of 10 grafts ruptured [80%]; data not shown). Data for body temperature, MAP, and blood flow were collected for each graft. Statistical analysis of these data showed that these parameters were consistently maintained (data not shown). Individual body weights and aPTT values at the beginning and end of the study were similar when compared across animals and treatment groups. Terminal aPTT values in each animal were > 400 seconds. This represented a 2.6-fold or greater increase in aPTT from baseline values, an expected effect of the heparin dosing regimen. In Vitro TEG Assays Increasing amounts of titrated rThrombin activity led to increased rates of clot formation (measured by the alpha angle; Figure 4, top panel; r = 0.72; p = 0.002) and clot strength (shear elastic modulus, defined as Gmax; Figure 4, bottom panel; r = 0.77; p < 0.001) in heparin + clopidogrel blood samples. In these in vitro clotting studies, sequential increases in the amount of rThrombin activity ultimately led to maximum (plateau) values for the rate of clot formation and clot strength.

strength (shear elastic modulus, defined as Gmax; Figure 4, bottom panel; r = 0.77; p < 0.001) in heparin + clopidogrel blood samples. In these in vitro clotting studies, sequential increases in the amount of rThrombin activity ultimately led to maximum (plateau) values for the rate of clot formation and clot strength. Figure 4 Effect of rThrombin concentration on rabbit blood thromboelastograph (TEG) parameters. Blood samples were collected from study animals (N = 3) two hours after they had been treated with clopidogrel, and heparin was added to 1 U/mL. The top panel shows the alpha angle, which is a function of the rate of clot formation, and the bottom panel shows the maximum clot strength (Gmax); alpha and Gmax are shown over a titrated range of rThrombin activity that was added to the TEG reaction.

ad been treated with clopidogrel, and heparin was added to 1 U/mL. The top panel shows the alpha angle, which is a function of the rate of clot formation, and the bottom panel shows the maximum clot strength (Gmax); alpha and Gmax are shown over a titrated range of rThrombin activity that was added to the TEG reaction. Discussion Numerous factors work to reduce both endogenous and exogenous thrombin concentration at the bleeding wound interface, including removal and dilution by hemorrhagic blood flow, rapid binding to inhibitors such as antithrombin III, entrapment in the developing thrombus, or mechanical removal by sponge and/or irrigation. Whether topical thrombin is applied or not, surgeons rely on intraoperative gross evaluations of hemostasis as predictors of whether hemostasis will be durable after wound closure. When hemostasis is delayed, laboratory assessments may be performed to rapidly detect any pathologic derangements in clotting. However, the results do not always correlate with intraoperative bleeding severity, nor do they reliably predict response to surgical intervention because of the many variables influencing clot formation at the wound. We addressed several of these variables in this study by examining the effects of rThrombin concentration on hemostasis in animals treated with commonly-administered anticoagulation agents.

do they reliably predict response to surgical intervention because of the many variables influencing clot formation at the wound. We addressed several of these variables in this study by examining the effects of rThrombin concentration on hemostasis in animals treated with commonly-administered anticoagulation agents. Increasing concentrations of rThrombin decreased TTH in a concentration-dependent manner in a rabbit model of arterial bleeding. Treatment of heparinized animals with clopidogrel delayed the onset of hemostasis. However, treatment with rThrombin at a concentration of 1000 IU/mL mitigated the delay in onset of hemostasis, consistently providing comparable results whether animals were treated with heparin or with heparin + clopidogrel. This suggests that the efficacy of 1000 IU/mL rThrombin as a topical hemostatic agent was unaffected by clopidogrel-induced platelet inhibition. In contrast, the application of 125 IU/mL rThrombin resulted in a more variable and delayed hemostatic response, suggesting that this concentration of rThrombin was too low to reliably overcome clopidogrel-induced platelet inhibition. This is the first report of thrombin concentration-dependent mitigation of the delayed hemostasis resulting from platelet inhibition. This observation lends credence to the widespread use of topical thrombin preparations in vascular and cardiac surgery, where platelet dysfunction is common, and supports the proposition that use of thrombin at 1000 IU/mL evolved because of reliable efficacy over a wide range of clinical settings, including both pathologic and pharmacologically-induced clotting abnormalities.

opical thrombin preparations in vascular and cardiac surgery, where platelet dysfunction is common, and supports the proposition that use of thrombin at 1000 IU/mL evolved because of reliable efficacy over a wide range of clinical settings, including both pathologic and pharmacologically-induced clotting abnormalities. The broad utility of the 1000 IU/mL concentration of rThrombin was further supported by an evaluation of clot durability that was performed by clamping the graft following the achievement of hemostasis (clot burst testing). Clot burst is thought to reflect the adhesiveness of the clot boundary to the PTFE graft material, the resistance of the clot to mechanical disruption, and indirectly, the development of platelet force. Development of platelet force is a process of thrombus maturation in which platelet contraction leads to increased fibrin strand density, thus ensuring that a sudden increase in blood pressure does not lead to clot failure. Normal platelet function is required for physiologic clot initiation [18]. Furthermore, pharmacological or mechanical inhibition of platelet function has been associated with increased bleeding in humans after cardiopulmonary bypass [19]. Persistent hemostasis that resists the stresses of the early recovery period is a desirable outcome for all surgical procedures. We observed that clots formed in the presence of 1000 IU/mL rThrombin were significantly more resistant to disruption than were clots formed in the presence of 125 IU/mL rThrombin in animals treated with heparin + clopidogrel. The disruption of clots formed in the presence of 125 IU/mL rThrombin was indistinguishable from the disruption of clots formed in the presence of placebo. These findings are consistent with those of other investigators; differences in clot structure and maturation at varying thrombin concentrations have been well described [6-8]. The fragile state of hemostasis that was achieved with 125 IU/mL human thrombin and AGS by Adams, et al. can be better understood in light of our observations [11]. The rebleeding wounds observed at 12 minutes in the Adams porcine hepatic bleeding model may have rebled secondary to inadequate clot strength and density and a higher concentration of thrombin might have resulted in more persistent hemostasis.

y Adams, et al. can be better understood in light of our observations [11]. The rebleeding wounds observed at 12 minutes in the Adams porcine hepatic bleeding model may have rebled secondary to inadequate clot strength and density and a higher concentration of thrombin might have resulted in more persistent hemostasis. Finally, results from the in vitro TEG technique are consistent with prior reports that fibrin clot strength is a function of available thrombin concentration, suggesting that there is an optimal concentration range for rThrombin efficacy [6-8]. Although the quantities of rThrombin used to initiate clotting in the TEG experiments cannot be directly compared with that used for hemostasis in the rabbit AV shunt model, increased rates of clot formation and improved clot strength in the TEG experiments was achieved with increasing amounts of rThrombin activity added at the time of clot initiation. These results are also consistent with those reported for human blood samples in which TEG parameters increased across a 10-fold range of rThrombin concentrations [20].

tion and improved clot strength in the TEG experiments was achieved with increasing amounts of rThrombin activity added at the time of clot initiation. These results are also consistent with those reported for human blood samples in which TEG parameters increased across a 10-fold range of rThrombin concentrations [20]. Conclusion In an animal model of vascular bleeding, rThrombin at a concentration of 1000 IU/mL reliably shortened TTH and resulted in more durable hemostasis in the presence of heparin anticoagulation and clopidogrel platelet inhibition when compared with rThrombin at a concentration of 125 IU/mL. This observation has significant implications for current clinical practice. Increasing numbers of patients are presenting for emergent, urgent, or elective surgical procedures with significant clopidogrel platelet inhibition as a result of a reluctance or inability to withhold clopidogrel prior to surgery. Results in this rabbit AV shunt model suggest that topical rThrombin at a concentration of 1000 IU/mL may reliably provide hemostasis in surgical patients who have received clopidogrel, a finding that should prompt further clinical investigation. Competing interests All authors were employees and stockholders of ZymoGenetics, Inc. at the time of this research.

Conclusion In an animal model of vascular bleeding, rThrombin at a concentration of 1000 IU/mL reliably shortened TTH and resulted in more durable hemostasis in the presence of heparin anticoagulation and clopidogrel platelet inhibition when compared with rThrombin at a concentration of 125 IU/mL. This observation has significant implications for current clinical practice. Increasing numbers of patients are presenting for emergent, urgent, or elective surgical procedures with significant clopidogrel platelet inhibition as a result of a reluctance or inability to withhold clopidogrel prior to surgery. Results in this rabbit AV shunt model suggest that topical rThrombin at a concentration of 1000 IU/mL may reliably provide hemostasis in surgical patients who have received clopidogrel, a finding that should prompt further clinical investigation. Competing interests All authors were employees and stockholders of ZymoGenetics, Inc. at the time of this research. Authors' contributions SDH and WAA participated in the design and coordination of the study and statistical analysis. PDB designed the in vitro TEG assays. RG designed and performed the in vivo rabbit assays and clot burst experiments. TZ designed and performed the statistical analyses. All authors were active in drafting the manuscript, and all authors read and approved the final manuscript.

ion of the study and statistical analysis. PDB designed the in vitro TEG assays. RG designed and performed the in vivo rabbit assays and clot burst experiments. TZ designed and performed the statistical analyses. All authors were active in drafting the manuscript, and all authors read and approved the final manuscript. Acknowledgements This study was sponsored by ZymoGenetics, Inc. We would like to thank Gerald Lasser (Gerald.lasser@qsela.com) of Qsela Group (Lynnwood, WA) for his assistance in thromboelastography and data analysis. We also thank Sonia Souza of ZymoGenetics, Inc. for performing some of the statistical analyses and thank Susan Moon of ZymoGenetics, Inc. and Kate Loughney, under the sponsorship of ZymoGenetics, Inc., for their assistance with medical writing.

Background The patients suffering from stenosing colorectal cancer are frequently treated for occlusion (20%) and stenosis (16%) [1]. The elective surgical treatment concerning the stenosis is to be performed after the preoperative assessment of the colonic segments upstream the cancer. If the endoscopic instrument can pass over the lesion, the gold standard is the colonoscopy. On the other hand, when the instrument cannot surmount the stenosis, the proximal segment of colon is to be evaluated through a double contrast barium enema. This approach shows low sensitivity and specificity particularly in the right colon [2]. Threfore when barium enema does not lead to an accurate evaluation, the patient needs an intra-operatory colonoscopy [3]. This investigation implies a few difficulties as the patient needs to be positioned on the operating table. It also requires additional time and can make the abdominal closure more difficult, because of the bowel distension caused by the insufflated air [4,5]. When the endoscopic procedure is incomplete because of the presence of a severe stenosing lesion, a further colonoscopy is required at 3-month follow up [6,7], in order to rule out a missed synchronous lesions. The recent introduction of CTC allows a more adequate preoperative assessment of colon showing higher sensitivity and specificity [8]. The aim of this study is to illustrate our experience with the CTC for the pre-operative assessment of the entire colon in the patients with stenosing colorectal cancer.

When the endoscopic procedure is incomplete because of the presence of a severe stenosing lesion, a further colonoscopy is required at 3-month follow up [6,7], in order to rule out a missed synchronous lesions. The recent introduction of CTC allows a more adequate preoperative assessment of colon showing higher sensitivity and specificity [8]. The aim of this study is to illustrate our experience with the CTC for the pre-operative assessment of the entire colon in the patients with stenosing colorectal cancer. Methods From January 2005 till March 2009, we observed and treated 43 patients with stenosing colorectal neoplastic lesions presenting with changes in the bowel habit, constipation, abdominal pain and lower gastrointestinal bleeding. We observed 28 males and 15 females aged between 47 and 73 years (average age 65 years). All patients did not tolerate the endoscopic procedure. For this reason they underwent a CTC in order to have a complete pre-operative assessment of the entire colon. All 43 patients underwent a pre-operative CTC in order to have a complete assessment of the entire colon. All 43 patients underwent surgical resections and specimen was carefully evaluated to search for any synchronous lesions. Furthermore all patients underwent a follow-up colonoscopy at 3 months from surgery. The results from CTC were compared with both macroscopic examination of specimen and the follow-up colonscopy. Results The CTC showed synchronous lesions in 4 patients (9.3% of the cases) (Table 1): Table 1 Stenosing colorectal cancer and synchronous lesions showed in the pre-operative CTC

Methods From January 2005 till March 2009, we observed and treated 43 patients with stenosing colorectal neoplastic lesions presenting with changes in the bowel habit, constipation, abdominal pain and lower gastrointestinal bleeding. We observed 28 males and 15 females aged between 47 and 73 years (average age 65 years). All patients did not tolerate the endoscopic procedure. For this reason they underwent a CTC in order to have a complete pre-operative assessment of the entire colon. All 43 patients underwent a pre-operative CTC in order to have a complete assessment of the entire colon. All 43 patients underwent surgical resections and specimen was carefully evaluated to search for any synchronous lesions. Furthermore all patients underwent a follow-up colonoscopy at 3 months from surgery. The results from CTC were compared with both macroscopic examination of specimen and the follow-up colonscopy. Results The CTC showed synchronous lesions in 4 patients (9.3% of the cases) (Table 1): Table 1 Stenosing colorectal cancer and synchronous lesions showed in the pre-operative CTC Site of stenosing colorectal cancer Synchronous lesions Site Synchronous lesion Type Diameter of synchronous lesions Surgical Treatment Sigmoid colon Descending colon Pedunculated polyp 8 mm Left laparoscopic hemicolectomy Rectum Ascending colon Pedunculated polyp 12 mm Laparoscopic anterior resection of rectum Sigmoid colon Ascending colon Sessile polyp 18 mm Left laparotomic hemicolectomy and excision of polyp through enterotomy Descending colon Ascending colon Voluminous vegetating cancer 36 mm Subtotal laparotomic colectomy • a pedunculated polyp in two cases (8 and 12 mm of diameter) both in the left colon;

ion of rectum Sigmoid colon Ascending colon Sessile polyp 18 mm Left laparotomic hemicolectomy and excision of polyp through enterotomy Descending colon Ascending colon Voluminous vegetating cancer 36 mm Subtotal laparotomic colectomy • a pedunculated polyp in two cases (8 and 12 mm of diameter) both in the left colon; • a sessile polyp (18 mm of diameter) in ascending colon; • a voluminous vegetating lesion (36 mm of diameter) in the ascending colon. The patients with synchronous colonic lesions are usually treated through a laparotomic access, as identifying laparoscopically synchronous lesions without a tattoo marked preoperatively is difficult. Furthermore, often it is not possible to identify synchronous lesions by intra-operative palpation even if the diameter is larger than one centimetre [9,10]. This supports the need of good preoperative assessment. In our series the patients with synchronous neoplastic lesions underwent the following surgical treatments (Table 1): • Left laparoscopic hemicolectomy (stenosing sigmoid cancer with adjacent peduncalated polyp of descending colon) • Laparoscopic anterior resection of rectum (stenosing rectal cancer with adjacent peduncalated polyp of sigmoid colon). • Left laparotomic hemicolectomy associated with excision of polyp through enterotomy in the ascending colon (stenosing cancer of sigmoid colon associated a sessile polyp of ascending colon).

• Left laparoscopic hemicolectomy (stenosing sigmoid cancer with adjacent peduncalated polyp of descending colon) • Laparoscopic anterior resection of rectum (stenosing rectal cancer with adjacent peduncalated polyp of sigmoid colon). • Left laparotomic hemicolectomy associated with excision of polyp through enterotomy in the ascending colon (stenosing cancer of sigmoid colon associated a sessile polyp of ascending colon). • Subtotal laparotomic colectomy for two synchronous neoplastic lesions located in distant colic segments (descending colon stenosing cancer associated with vegetant neoplastic lesion of ascending colon). Table 1 The macroscopic examination of the specimen revealed three patients with small sessile polyps (3 - 4 mm in diameter) missed in the pre-operative CTC near the stenosing colorectal cancer (Table 2). Table 2 Stenosing colorectal cancers and synchronous lesionsidentified specimen and missed in the pre-operative CTC. Site of stenosing colorectal cancer Site of synchronous lesion Type of synchronous lesions Diameter of synchronous lesions Left colon Left colon Sessile polyp 3 mm Left colon Left colon Sessile polyp 4 mm Left colon Left colon Sessile polyp 3 mm Table 2 The follow-up colonscopy showed four additional endoluminal lesions in three patients: four sessile polyps with diameter of 3, 4, 8 and 11 mm (Table 3) Table 3 Stenosing colorectal cancers and synchronous lesions identified in the follow-up colonoscopy at 3 months after the surgical treatment and missed in the pre-operative CTC

Site of stenosing colorectal cancer Site of synchronous lesion Type of synchronous lesions Diameter of synchronous lesions Left colon Left colon Sessile polyp 3 mm Left colon Left colon Sessile polyp 4 mm Left colon Left colon Sessile polyp 3 mm Table 2 The follow-up colonscopy showed four additional endoluminal lesions in three patients: four sessile polyps with diameter of 3, 4, 8 and 11 mm (Table 3) Table 3 Stenosing colorectal cancers and synchronous lesions identified in the follow-up colonoscopy at 3 months after the surgical treatment and missed in the pre-operative CTC Site of stenosing colorectal cancer Site of synchronous lesion Type of synchronous lesions Diameter of synchronous lesions Left colon Ascending colon Sessile polyp 3 mm Left colon Transverse colon Sessile polyp 4 mm Left colon Transverse colon Sessile polyp 8 mm Rectum Left colon Sessile polyp 11 mm Table 3 Our experience shows that virtual colonoscopy has a sensitivity of 83,7%. Discussion Computed Tomographic Colonography (CTC) is also known as "virtual colonoscopy" and it was introduced in 1994. It is considered a noninvasive method of imaging and it allows to explore the colon by employing a helical CT [11]. Data from literature show that CTC is safer than colonoscopy. Indeed a colonic perforation occurs in 1:1000 patients undergoing conventional colonoscopy and mortality associated with this procedure is 1:5000 [12-17], while the morbidity and mortality associated with CTC are similar to those of the air-contrast barium enema (perforation rate of 1:10000 and mortality rate of 1:50000) [18-20].

lonic perforation occurs in 1:1000 patients undergoing conventional colonoscopy and mortality associated with this procedure is 1:5000 [12-17], while the morbidity and mortality associated with CTC are similar to those of the air-contrast barium enema (perforation rate of 1:10000 and mortality rate of 1:50000) [18-20]. There is only one study in literature showing a perforation rate, which is very low, up to 0.04% (3/7180) [21] and no study showing a case of death after performing a CTC. The CTC has recently been proposed as the standard investigation for the pre-operative assessment of the colonic segments proximal to the obstructive cancer, being well tolerated, less invasive and showing good results [4,22-25]. Usually the colonoscopy is not fully carried out up to the ileocecal valve in 5-10% of the patients; this is mainly due to [26]: • presence of stenosing neoplastic lesion of rectum or distal colon (58,3%): • presence of dolicocolon (33%); • presence of spastic colon (11,7%). • presence of extrinsic compressions (1,6%); • presence of intestinal malrotation (1,6%); • presence of visceral adhesions. In the past in all these cases the alternative investigation was an opaque enema. Nowadays, CTC shows a higher sensitivity making it the choice investigation when the traditional endoscopy cannot be carried out satisfactorily [26]. All CTCs performed in the above mentioned cases were successful (91,7-100%) (table 4) [4,22,23,27]. Table 4 Virtual colonoscopy in stenosing colorectal cancers: sensitivity for synchronous lesions after colonoscopy follow-up at 3 months

In the past in all these cases the alternative investigation was an opaque enema. Nowadays, CTC shows a higher sensitivity making it the choice investigation when the traditional endoscopy cannot be carried out satisfactorily [26]. All CTCs performed in the above mentioned cases were successful (91,7-100%) (table 4) [4,22,23,27]. Table 4 Virtual colonoscopy in stenosing colorectal cancers: sensitivity for synchronous lesions after colonoscopy follow-up at 3 months Preoperative evaluation of the colon proximal to the obstructive colorectal cancer Pts. with colorectal cancers and obstructing colorectal lesions Lesions missing at preoperative virtual colonoscopy Sensitivity Fenlon 1999 CTC 12 pts. 2 83.4% Galia 2001 CTC 19 pts. 3 83.4% Morrin 2005 CTC 17 pts 1 93% Nagata 2009 PET/CT colonography 13 pts. 0 100% Table 4 The cause of failure is frequently based either on neoplastic obstruction or on an inadequate air distension of proximal colon (8,3%) [26]. Data from literature show that CTC allows the diagnosis of synchronous lesions in 27,2%-89,6% of cases [28]. CTC substantially owns the same sensitivity as an optical colonoscopy for polyps > or = 7 mm in diameter [29]. Most of the polyps were observed in ascending colon (31,25%), transverse colon (18,75%), descending colon (18,75%) and sigma (6,25%) [26]. The diameter of the polyps varies between 5 mm (25%), 6-9 mm (62,5%) and 10 mm (12,5%) [26]. Between 6.25% and 8.3% of the observed synchronous lesions resulted to be cancers [22,26].

st of the polyps were observed in ascending colon (31,25%), transverse colon (18,75%), descending colon (18,75%) and sigma (6,25%) [26]. The diameter of the polyps varies between 5 mm (25%), 6-9 mm (62,5%) and 10 mm (12,5%) [26]. Between 6.25% and 8.3% of the observed synchronous lesions resulted to be cancers [22,26]. In those patients with synchronous cancers the most appropriate approach is to remove both lesions at the same time. Fenlon et al. described a subtotal colectomy for sigmoid and trasverse colon cancers and a left hemicolectomy for spleen flexure and sigma lesion [22]. Mingyue et al stated in their study that although the synchronous lesion is a polyp, it should be removed during surgery, otherwise, in the remaining cases, polyps will be removed during the endoscopic follow-up [26]. In the past, a considerable percentage of CTC presented a high incidence of false positives (between 54.6% and 64% for polyps within 6-10 mm of diameter, and 41-51% for polyps larger than 10 mm in diameter) [10,30,31]. Nowadays more modern softwares and new methods for colon cleansing preparation allow a significant reduction of false positives [32], lowering the percentage till 2.1% for polyps and 0.7% for cancers [33] despite the first results appeared in literature [34-36]. Moreover the possibility to perform the CTC even with a mild bowel preparation allows to extend the indication of the cancer screening to those patients considered not enough fit to tolerate a complete bowel preparation [37].

Nowadays more modern softwares and new methods for colon cleansing preparation allow a significant reduction of false positives [32], lowering the percentage till 2.1% for polyps and 0.7% for cancers [33] despite the first results appeared in literature [34-36]. Moreover the possibility to perform the CTC even with a mild bowel preparation allows to extend the indication of the cancer screening to those patients considered not enough fit to tolerate a complete bowel preparation [37]. An accurate evaluation of the colonic segments upstream the neoplastic obstruction is extremely important as the patients with colorectal cancer are often affected by synchronous neoplastic lesions (villous adenomas in 33-55% of the cases and other types of cancers in 3% of the cases) [36,38,39] (figures 1,2,3,4). A surgical re-intervention is required when synchronous neoplastic lesions are detected and this increases morbidity and higher complication rates, especially because of peritoneal adhesions. Figure 1 Cancer of splenic flexure unsurpassed by colonoscopy. Figure 2 3D Reconstruction with iconographic effect similar to double contrast barium enema of splenic flexure's cancer. Figure 3 Downstream polyp of splenic flexure cancer's. Figure 4 Upstream polyp of splenic flexure cancer's. In case of synchronous lesions the prognosis becomes worse than a single cancer (55% survival at 5 years) [40], especially when diagnosis and treatment are not carried out on time.

Figure 2 3D Reconstruction with iconographic effect similar to double contrast barium enema of splenic flexure's cancer. Figure 3 Downstream polyp of splenic flexure cancer's. Figure 4 Upstream polyp of splenic flexure cancer's. In case of synchronous lesions the prognosis becomes worse than a single cancer (55% survival at 5 years) [40], especially when diagnosis and treatment are not carried out on time. Currently CTC is very helpful in colorectal surgery and it is particularly useful during the preoperative study in laparoscopic colorectal surgery; an accurate localization of the lesions not only allows an appropriate trocar insertion, but it also reduces the error due to a lack of tactile sensation [41]. It can be difficult to find out whether synchronous colon cancers are present in an intestinal segment proximal to the obstructive cancer. These patients usually need a further sequent surgical treatment if synchronous colon cancers are detected later [42]. An accurate diagnosis including the evaluation of the colonic segments proximal to the obstruction can avoid additional surgical approaches providing a better quality of life, as well as being cost-effective [9,43-45]. Therefore, one of the important roles of preoperative diagnosis in patients with obstructive cancer is a precise evaluation of the entire colon [4].

of the colonic segments proximal to the obstruction can avoid additional surgical approaches providing a better quality of life, as well as being cost-effective [9,43-45]. Therefore, one of the important roles of preoperative diagnosis in patients with obstructive cancer is a precise evaluation of the entire colon [4]. Conclusion This study shows CTC as a valid pre-operative investigation (sensitivity 83.7%) in order to detect synchronous neoplastic lesions in the patients who do not tolerate the conventional colonoscopy because of stenosing colorectal cancer. The main benefits from from its employment are that the intraoperative colonoscopy is not needed and more synchronous lesions are detected and treated at the time of the elective surgery for the stenosing cancer avoiding further surgery later on. Competing interests The Authors state that none of the authors involved in the manuscript preparation has any conflicts of interest towards the manuscript itself, neither financial nor moral conflicts. Besides none of the authors received support in the form of grants, equipment, and/or pharmaceutical items. Authors' contributions All authors contributed equally to this work. Consent Written informed consent was obtained from the patient for publication of this case report and accompanying images. A copy of the written consent is available for review by the Editor-in-Chief of this journal. Acknowledgements The authors give thanks to Dr Francesco Mancioli, Department of Radiology, St Mary Hospital for the offered photos.

Background Conventional laparoscopic (CL) surgery may offer great advantages to patients but can be demanding for the surgeon because of several technical drawbacks. These limitations include 2-dimensional vision with less than optimal perception of depth, disturbance of the eye-hand-target axis, the fulcrum effect, rigid instruments with limited degrees of freedom and limited tactile feedback. These factors might attribute to the relatively long training period required before reaching a professional level [1,2]. The da Vinci® surgical system from Intuitive Surgical® (Sunnyvale, CA, USA) has been available since 1998 and is still the only robotic surgical system available on the market approved for performing surgical interventions in humans. Several advantages with robot-assisted laparoscopy (RL) over CL have been identified: 3-dimensional visualization of the operative field with depth perception, additional degrees of freedom and downscaling of instrument movements, restoration of the eye-hand-target axis and enhanced stability, elimination of the fulcrum effect and improved ergonomics for the surgeon. One stated consequence of these features is that endoscopic surgical skills are more easily mastered and the learning curve is shortened [2-4]. Some authors have succeeded in performing RL for cases they never tried with CL, possibly indicating that RL is considered easier [5].

The da Vinci® surgical system from Intuitive Surgical® (Sunnyvale, CA, USA) has been available since 1998 and is still the only robotic surgical system available on the market approved for performing surgical interventions in humans. Several advantages with robot-assisted laparoscopy (RL) over CL have been identified: 3-dimensional visualization of the operative field with depth perception, additional degrees of freedom and downscaling of instrument movements, restoration of the eye-hand-target axis and enhanced stability, elimination of the fulcrum effect and improved ergonomics for the surgeon. One stated consequence of these features is that endoscopic surgical skills are more easily mastered and the learning curve is shortened [2-4]. Some authors have succeeded in performing RL for cases they never tried with CL, possibly indicating that RL is considered easier [5]. A definition of the learning curve can be the amount of practice, in terms of time or number of repetitions, needed to reach a certain level of proficiency for completing a specific task. Parameters used when analysing learning curves are time to complete the task, the number of errors made and actions required. Learning curves in daily practice are often defined by operating time, blood loss, morbidity and length of hospital stay [3]. There is only very scant literature on the comparison of learning curves for RL and CL [3,6]. It is also a challenge to interpret the results of earlier studies, one reason being the different levels of previous experience among the participants. However, to know more about the learning curve in minimally invasive surgery and preferably, as a consequence, to be able to shorten the time needed for operative training before reaching a consistent level, is desirable. We therefore decided to test our hypothesis that surgically maiden subjects perform surgical tasks faster with the robot-assisted laparoscopy technique than with the conventional laparoscopy technique.

quence, to be able to shorten the time needed for operative training before reaching a consistent level, is desirable. We therefore decided to test our hypothesis that surgically maiden subjects perform surgical tasks faster with the robot-assisted laparoscopy technique than with the conventional laparoscopy technique. Materials and methods A power calculation, comparing two groups regarding proportions of the specific trait we wanted to study, was performed. At a given power of 80%, an alpha level of 0.05, and a proportion of 0.8 among cases, a sample of at least 12 cases and 12 control subjects is needed to meet a significant effect when the proportion among control subject is 0.2. A sample of 10 cases and 10 control subjects undergoing three repeated trials will be sufficient, although these repeated measures are mutually dependent. From a cohort of approximately 500 medical students at Lund University, volunteers were invited to participate in this project and from these, 20 subjects (10 men and 10 women), were randomly selected. The subjects were between 23 and 30 years old and had no prior practical experience of open surgery, CL or RL. There were no drop-outs or excluded participants.

students at Lund University, volunteers were invited to participate in this project and from these, 20 subjects (10 men and 10 women), were randomly selected. The subjects were between 23 and 30 years old and had no prior practical experience of open surgery, CL or RL. There were no drop-outs or excluded participants. All the subjects were given the same standardized oral and written information by the trial instructor. The tasks the subjects were supposed to perform were also demonstrated once for each of the methods RL and CL before the trial started. One instructor and three evaluators, all working at the Department of Paediatric Surgery were used. The evaluators registered the subjects' performances in accordance with a predetermined template. The subjects were allowed to ask for guidance and the instructor gave them standardized advice along the way. The subjects were not allowed to observe or communicate with each other and were therefore isolated during the study. The primary and secondary end points were time and accuracy when performing the simulated surgical tasks using robotic and conventional laparoscopic instruments. Material and equipment used during the study were: the da Vinci® Surgical System from Intuitive Surgical®, a robotic needle holder, a robotic DeBakey® grasper, laparoscopic optics from KARL STORZ® 0° E-class 26003 AA, a laparoscopic needle holder 26173 KL and a laparoscopic grasper KARL STORZ® 33121, a Skin Pad in Jig® L & T 00131 from: http://www.limbsandthings.com and Vicryl® from Ethicon®, Polyglactin 910, 2-0 (3 Ph. Eur.), CP-1 36 mm 1/2 c 70 cm.

robotic DeBakey® grasper, laparoscopic optics from KARL STORZ® 0° E-class 26003 AA, a laparoscopic needle holder 26173 KL and a laparoscopic grasper KARL STORZ® 33121, a Skin Pad in Jig® L & T 00131 from: http://www.limbsandthings.com and Vicryl® from Ethicon®, Polyglactin 910, 2-0 (3 Ph. Eur.), CP-1 36 mm 1/2 c 70 cm. The workstation was prepared in a standardized way and the participants were allowed to familiarize themselves with the instruments for two minutes before starting the trial. The thread was 20 cm long for both suturing and tying a knot. Each of the 20 students carried out three tasks, grab the needle in a correct way, place three continuous sutures over a rift in the Skin Pad in Jig® and tie a surgical knot. These tasks were done four times with each of RL and CL. The subjects were divided according to gender and half the males and half the females began with RL and CL, respectively. For each set, time and quality indicators (1-6 below) were recorded, giving us a total of 960 sets of data to analyze. 1. Grab the needle - time in seconds 2. Continuous suturing (3 stitches) - time in seconds 3. Tie a knot (double) - time in seconds 4. Damage to the Skin Pad in Jig®? - yes/no 5. Dropped needle? - number of times/set 6. Tearing of the thread? - number of times/set. [see Additional file 1 and 2]

The workstation was prepared in a standardized way and the participants were allowed to familiarize themselves with the instruments for two minutes before starting the trial. The thread was 20 cm long for both suturing and tying a knot. Each of the 20 students carried out three tasks, grab the needle in a correct way, place three continuous sutures over a rift in the Skin Pad in Jig® and tie a surgical knot. These tasks were done four times with each of RL and CL. The subjects were divided according to gender and half the males and half the females began with RL and CL, respectively. For each set, time and quality indicators (1-6 below) were recorded, giving us a total of 960 sets of data to analyze. 1. Grab the needle - time in seconds 2. Continuous suturing (3 stitches) - time in seconds 3. Tie a knot (double) - time in seconds 4. Damage to the Skin Pad in Jig®? - yes/no 5. Dropped needle? - number of times/set 6. Tearing of the thread? - number of times/set. [see Additional file 1 and 2] Each subject's results from the RL and the CL, respectively, were recorded and mean values were calculated for the groups: RL first, CL first, RL last and CL last. The analysis was carried out using Wilcoxon's signed rank test for paired samples and the Mann-Whitney test of two independent samples before and after the cross-over. The transfer effect of difference in experience based on the RL users' prior CL experience and vice versa, was tested regarding the three surgical tasks using the Mann-Whitney test of two independent samples. Friedman's and Wilcoxon's signed rank test were used for analysis of tries of knot tying and suturing. Male and female participants were compared using the Mann-Whitney signed rank-test for paired samples. SPSS statistical software, version 15.0, was used for analysing the data. P < 0.05 was considered significant. This study was approved by the local regional ethical committee March 24th 2009 (Dnr 2009/59).

d suturing. Male and female participants were compared using the Mann-Whitney signed rank-test for paired samples. SPSS statistical software, version 15.0, was used for analysing the data. P < 0.05 was considered significant. This study was approved by the local regional ethical committee March 24th 2009 (Dnr 2009/59). Results The values are summarized in Tables 1 and 2. "Grabbing the needle" and "suturing continuously" were carried out at equal speeds with RL and CL. The transfer effect was seen when performing the continuous suturing for RL but not for CL (Tables 3, 4). The same transfer effect was seen for RL in tying a knot but not for CL. The task of tying a knot was performed faster with RL than with CL regardless of whether RL was performed first or second. A difference was observed in tying the knot when changing to the other operating technique, regardless of which technique the subjects started with. The difference favoured RL and was negative for CL (Table 5). Dropping the needle happened more often during the RL part of the study and tearing the thread only occurred with the RL technique. Damage to the Skin Pad in Jig® was equally common (Table 6). There are learning curves seen for tying a knot (Figure 1) and continuous suturing (Figure 2) for both RL and CL when comparing trials 1 and 4. There was no difference between the male and female subjects' performances regarding any of the three tasks or the three quality indicators included in the study.

on (Table 6). There are learning curves seen for tying a knot (Figure 1) and continuous suturing (Figure 2) for both RL and CL when comparing trials 1 and 4. There was no difference between the male and female subjects' performances regarding any of the three tasks or the three quality indicators included in the study. Figure 1 The mean time for each try to tie a knot for the 20 subjects with CL and RL in seconds. Statistics: Friedman's test: CL knot, sets 1,2, 3 and 4: 0.0001. RL knot, sets 1,2,3 and 4: 0.3 (ns). Wilcoxon Signed Rank test: CL knot, sets 1 and 4: 0.0001. RL knot, sets 1 and 4: 0.009. Table 1 Time in seconds for each group of participants. Grab the needle Continuous suturing Tie a knot MEAN ± SD (RANGE) MEAN ± SD (RANGE) MEAN ± SD (RANGE) CL first 6 ± 4 (2 - 25) 198 ± 82 (108 - 472) 266 ± 173 (68 - 935) RL last 7 ± 5 (2 - 21) 144 ± 70 (57 - 346) 91 ± 35 (45 - 195) RL first 10 ± 8 (1 - 28) 216 ± 123 (76 - 663) 152 ± 94 (48 - 560) CL last 6 ± 4 (2 - 20) 209 ± 96 (86 - 438) 267 ± 186 (70 - 1027) N = 10 in each group Table 2 Qualitative parameters for each group of participants. Damage Drop Break CL first 1 6 0 RL last 5 32 4 RL first 3 47 1 CL last 1 12 0 Number of times damage to pad, dropping the needle or breaking the thread Table 3 Grabbing the needle MEAN ± SD (RANGE) MEAN ± SD (RANGE) P-value RL first vs CL first 10 ± 8 (1 - 28) 6 ± 4 (2 - 25) ns RL last vs CL last 7 ± 5 (2 - 21) 6 ± 4 (2 - 20) ns RL first vs RL last ns CL first vs CL last ns RL first vs CL last ns CL first vs RL last ns Time in seconds for each group of participants Table 4 Continuous suturing.

Table 3 Grabbing the needle MEAN ± SD (RANGE) MEAN ± SD (RANGE) P-value RL first vs CL first 10 ± 8 (1 - 28) 6 ± 4 (2 - 25) ns RL last vs CL last 7 ± 5 (2 - 21) 6 ± 4 (2 - 20) ns RL first vs RL last ns CL first vs CL last ns RL first vs CL last ns CL first vs RL last ns Time in seconds for each group of participants Table 4 Continuous suturing. MEAN ± SD (RANGE) MEAN ± SD (RANGE) P-value RL first vs CL first 216 ± 123 (76 - 663) 198 ± 82 (108 - 472) ns RL last vs CL last 144 ± 70 (57 - 346) 209 ± 96 (86 - 438) ns RL first vs RL last 0.049 CL first vs CL last ns RL first vs CL last ns CL first vs RL last ns Time in seconds for each group of participants and transfer effect Table 5 Tying a knot. MEAN ± SD (RANGE) MEAN ± SD (RANGE) P-value RL first vs CL first 152 ± 94 (48 - 560) 266 ± 173 (68 - 935) 0.005 RL last vs CL last 91 ± 35 (45 - 195) 267 ± 186 (70 - 1027) 0.001 RL first vs RL last 0.004 CL first vs CL last ns RL first vs CL last 0.037 CL first vs RL last 0.005 Time in seconds for each group of participants and transfer effect Table 6 Qualitative parameters for all 20 participants after four repeats with RL and CL expressed as the number of times they occurred (n)

MEAN ± SD (RANGE) MEAN ± SD (RANGE) P-value RL first vs CL first 152 ± 94 (48 - 560) 266 ± 173 (68 - 935) 0.005 RL last vs CL last 91 ± 35 (45 - 195) 267 ± 186 (70 - 1027) 0.001 RL first vs RL last 0.004 CL first vs CL last ns RL first vs CL last 0.037 CL first vs RL last 0.005 Time in seconds for each group of participants and transfer effect Table 6 Qualitative parameters for all 20 participants after four repeats with RL and CL expressed as the number of times they occurred (n) RL CL P-value Damage (n) 8 2 ns Dropped needle (n) 79 18 < 0.05 Torn thread (n) 5 0 < 0.05 Discussion Very few studies comparing the learning curves of RL and CL have been published. In the experimental setting a diversity of parameters, not always well-defined, has been used for analysis of learning curves and only the very beginning of the learning curve is studied. In the clinical setting, an experience bias has been expected due to prior laparoscopic experience of the participating surgeons [2,6-12]. Both experimental and clinical studies show diverging learning curves for robotic surgery. The results of previous studies are not conclusive and to objectively evaluate the learning curve of robotic surgery is difficult.

ias has been expected due to prior laparoscopic experience of the participating surgeons [2,6-12]. Both experimental and clinical studies show diverging learning curves for robotic surgery. The results of previous studies are not conclusive and to objectively evaluate the learning curve of robotic surgery is difficult. Our experimental study included participants without any prior experience of open surgery, RL or CL, making the group homogeneous. The performed tasks, well-defined and described, closely mimicked some of the proper surgical procedures used every day in the operating theatre. We used the only robotic surgical system currently on the market and standard CL instruments. The size of our group of participants and the number of repetitions studied was decided after power calculation.

d and described, closely mimicked some of the proper surgical procedures used every day in the operating theatre. We used the only robotic surgical system currently on the market and standard CL instruments. The size of our group of participants and the number of repetitions studied was decided after power calculation. The task of tying a surgical knot was always faster with RL than with CL, even when the participant had gained no experience by carrying out the first part of the study with CL. There are learning curves seen for tying a knot for both RL and CL when comparing trials 1 and 4. The learning curve is steeper for CL but the curves never cross (Figure 1). These findings differ from most previous studies where the initial performance with RL is often inferior to the performance with CL [3]. In a recent publication by Stefanidis et al. the authors reported that robotic assistance significantly improved intracorporeal suturing performance and shortened the learning curve. They also reported that performance of laparoscopic knot tying without robotic assistance did not improve after three repeats [13]. The first statement is supported by our study but the latter is not since we also saw a significant learning curve for CL. Performing more advanced tasks like tying a knot might be faster for maiden users due to the fact that RL is more "intuitive" with instrument movements mimicking normal hand movements. This is supported by some authors [4]. The fact that RL is performed with 3-dimensional vision instead of the 2-dimensional vision in CL might also improve the performance, as has been suggested by others [7,14].

sers due to the fact that RL is more "intuitive" with instrument movements mimicking normal hand movements. This is supported by some authors [4]. The fact that RL is performed with 3-dimensional vision instead of the 2-dimensional vision in CL might also improve the performance, as has been suggested by others [7,14]. The transfer effect, with a faster performance if the specific method was used as the second part of the study when the tasks had already been tried by the first method, was seen for continuous suturing and tying a knot with RL, but not for CL. This might be interpreted as the RL method being easier to adapt to once acquaintance had been made with the tasks themselves, at least for maiden users. The study by Blavier et al. showed worse performance when shifting from one method to the other in both directions. The shorter learning curve for RL noted by the same authors is supported by our study [7].

od being easier to adapt to once acquaintance had been made with the tasks themselves, at least for maiden users. The study by Blavier et al. showed worse performance when shifting from one method to the other in both directions. The shorter learning curve for RL noted by the same authors is supported by our study [7]. The learning curve consists of an initial steep phase in which performance improves rapidly. When the change in improvement slows down, the learning curve reaches a plateau phase in which variability in performance is small. The number of repetitions reported here are too low to reach consistency, which characterizes the end of the learning curve. The learning curve for CL was steeper, but the number of repetitions too few to disclose a complete learning curve. This was not the aim of the study. We concentrated on the first phase of the learning curve in order to detect even small changes or differences between the two techniques used. From our data, we can therefore only conclude that it is initially easier for novice subjects to use robotic assistance for the specific tasks using the set performance parameters. Whether or not the curves for RL and CL eventually cross after more repetitions, or when the plateau phase of the learning curve for each technique is reached, remains unclear. This could be the aim of another study in the future.

ts to use robotic assistance for the specific tasks using the set performance parameters. Whether or not the curves for RL and CL eventually cross after more repetitions, or when the plateau phase of the learning curve for each technique is reached, remains unclear. This could be the aim of another study in the future. The objective structured assessment of technical skill (OSATS) described by Reznick et al. is a validated tool widely used in the education literature. The OSATS is feasible, reliable and can be used for testing technical competence with high clinical relevance [15]. Since we focused on comparing the different repeats and the transfer effect in all three tasks we did not calculate a total score for the time and accuracy parameters.

ly used in the education literature. The OSATS is feasible, reliable and can be used for testing technical competence with high clinical relevance [15]. Since we focused on comparing the different repeats and the transfer effect in all three tasks we did not calculate a total score for the time and accuracy parameters. Dropping the needle was more common in the RL group. Half of the subjects dropped the needle while performing CL and all but two while performing RL. Furthermore, the thread was only torn when using RL. Tactile feedback is not yet possible in RL, which is the most probable explanation for our findings. In spite of these differences, albeit significant, the performance when using RL was not slower in the task "continuous suturing" compared with CL. Without the dropping of the needle in the task "continuous suturing", RL might have been faster. Learning curves are also seen for a continuous suture for both RL and CL when comparing trials 1 and 4 (Figure 2). The two figures 1 and 2 express the mean time for each try for the specific task but do not consider in what order the task is performed. As already stated, no difference was noted between RL and CL for "continuous suturing". Clinical reports have indicated that the improved vision in RL seems to make up for the lack of tactile feedback for more experienced surgeons [5]. The tearing of threads and dropping of needles is probably a greater challenge to the beginner.

stated, no difference was noted between RL and CL for "continuous suturing". Clinical reports have indicated that the improved vision in RL seems to make up for the lack of tactile feedback for more experienced surgeons [5]. The tearing of threads and dropping of needles is probably a greater challenge to the beginner. Figure 2 The mean time for each try of continuous suturing for the 20 subjects with CL and RL in seconds. Statistics: Friedman's test. CL running suture, sets 1,2,3 and 4: 0,001. RL running suture, sets 1,2,3 and 4: 0,001. Wilcoxon Signed Rank test: CL running suture, sets 1 and 4: 0,001. RL running suture, sets 1 and 4: 0,002. The end points of our study, time and accuracy, may not be the best end points to measure. Length of pathway and economy of movement might be better predictors of learning curve and safe performance of laparoscopic surgery. A further possible limitation of our study is that error reduction, an important goal of training, was not measured. The study of Narazaki et al. suggests that both task completion time and distance travelled is shortened with training in novice users [10].

f learning curve and safe performance of laparoscopic surgery. A further possible limitation of our study is that error reduction, an important goal of training, was not measured. The study of Narazaki et al. suggests that both task completion time and distance travelled is shortened with training in novice users [10]. The suggested advantage of faster laparoscopy in the RL group might not be relevant in clinical surgery since inexperienced users are not supposed to perform advanced laparoscopic surgery. Robotic surgeons today are often senior surgeons and already expert laparoscopists. However, the training to become an expert takes a lot of time and is costly, so learning curves are important also for the future education of young surgeons. If RL is proven easier to master with equal or better results than CL, robotic surgery could be an option for efficient surgical training. The many steps of a surgical intervention each have a learning curve and if learning curves are shorter for RL it may have some clinical relevance even at later stages of training. RL is still in its infancy but offers great opportunities for the future. Major improvements in the availability of tactile feedback and specifically designed instruments are necessary and expected soon. More research needs to be done to define the exact indications for RL to justify the increased costs and the increased time consumption involved, compared with CL. Whether or not these features with improved accuracy, dexterity and visualization enhance surgical performance remains unclear.

RL is still in its infancy but offers great opportunities for the future. Major improvements in the availability of tactile feedback and specifically designed instruments are necessary and expected soon. More research needs to be done to define the exact indications for RL to justify the increased costs and the increased time consumption involved, compared with CL. Whether or not these features with improved accuracy, dexterity and visualization enhance surgical performance remains unclear. In conclusion, we found support for our hypothesis that a surgical task, such as tying a knot, was performed faster using RL than with CL, while easier surgical tasks could be performed equally fast with either technique. The lack of tactile feedback in RL is a factor to consider at least for maiden users. Experience from one technique was transferred to the other. Our data do not support the suggestion that considerable CL experience is important for those starting to use RL. On the other hand, previous experience did matter in our study. No difference between the performances of male and female subjects was noted. Competing interests When performing this work, there were no external influences or conflicts of interest. None of the authors or subjects received fees from the manufacturers of the material or the instruments used in the study reported here.

In conclusion, we found support for our hypothesis that a surgical task, such as tying a knot, was performed faster using RL than with CL, while easier surgical tasks could be performed equally fast with either technique. The lack of tactile feedback in RL is a factor to consider at least for maiden users. Experience from one technique was transferred to the other. Our data do not support the suggestion that considerable CL experience is important for those starting to use RL. On the other hand, previous experience did matter in our study. No difference between the performances of male and female subjects was noted. Competing interests When performing this work, there were no external influences or conflicts of interest. None of the authors or subjects received fees from the manufacturers of the material or the instruments used in the study reported here. Legal requirements The authors guarantee that the manuscript will not be published elsewhere in any language without the consent of the copyright owners, that the rights of third parties will not be violated, and that the publisher will not be held legally responsible if there should be any claims for compensation. This study complies with the current laws of the country in which it was performed. This work was performed in accordance with the rules of the ethical committee at our centre and the ethical standards laid down in the 1964 Declaration of Helsinki.

eld legally responsible if there should be any claims for compensation. This study complies with the current laws of the country in which it was performed. This work was performed in accordance with the rules of the ethical committee at our centre and the ethical standards laid down in the 1964 Declaration of Helsinki. Authors' contributions MA designed the study, coordinated all steps of the study, collected and analyzed the data and wrote the paper. JL gathered all participants of the study, collected the data and took active part in writing of the paper. CCK participated in the design of the study, collected the data and took active part in writing of the paper. EA participated in the design of the study, collected and analyzed the data and took active part in writing of the paper. All authors read and approved the final manuscript. Supplementary Material Additional file 1 The Skin Pad in Jig®. The Skin Pad in Jig® as an *.jpg file, showing the needle, the placed running suture and the tied knot. Click here for file Additional file 2 The tasks of this experimental study performed with robot-assisted laparoscopy. A short video showing the actual tasks in this experimental study performed with robot-assisted laparoscopy. (19 Mb, 130 seconds) Click here for file Acknowledgements • Gillian Sjödahl, Lexis English for Writers, Persikevägen 11, SE-22355 Lund, Sweden, for linguistic revision of the manuscript. • Håkan Lövkvist MSC, biostatistician. Competence Centre for Clinical Research, Skåne University Hospital, SE-205 02 Malmö, Sweden, for statistical advice.

Additional file 2 The tasks of this experimental study performed with robot-assisted laparoscopy. A short video showing the actual tasks in this experimental study performed with robot-assisted laparoscopy. (19 Mb, 130 seconds) Click here for file Acknowledgements • Gillian Sjödahl, Lexis English for Writers, Persikevägen 11, SE-22355 Lund, Sweden, for linguistic revision of the manuscript. • Håkan Lövkvist MSC, biostatistician. Competence Centre for Clinical Research, Skåne University Hospital, SE-205 02 Malmö, Sweden, for statistical advice. • Practicum, Skåne University Hospital, Getingevägen 4, SE-221 85 Lund, Sweden for assistance in performing the study. • The operating unit at the Department of Paediatric Surgery, Children's Hospital Lund for preparing the robot during the study.

Introduction The routine use of video-assisted thoracic surgical (VATS) lobectomy for the treatment of resectable non-small cell lung cancer (NSCLC) remains controversial. Data supporting the use of VATS lobectomy come from randomized trials [1-4], a multicenter phase II study [5], case-control series and large retrospective series [6]. Meta-analyses have recently been published [7,8]. The randomized trials enrolled relatively small numbers of patients and retrospective case series are subject to selection biases. Other recent publications have been case-control series [9,10]. Most case-control techniques attempt to decrease the effect of selection bias when comparing two non-randomized treatment groups by analyzing patients that are "matched" based on preoperative variables that are known to affect the outcomes that are being studied. It can be difficult to find appropriate control patients for a case matching study unless a large control population is available. Even then, cases without an appropriate match in either the control or treatment groups are not usable for analysis. At times, eliminating the cases that cannot be matched may actually result in an increase in selection bias, casting doubt on the conclusion that differences in outcomes between the two groups are due to treatment effects [11].

an appropriate match in either the control or treatment groups are not usable for analysis. At times, eliminating the cases that cannot be matched may actually result in an increase in selection bias, casting doubt on the conclusion that differences in outcomes between the two groups are due to treatment effects [11]. Propensity score based weighting is a rigorous statistical technique for making nonrandomized comparisons that in theory allows all of the patient data to be used from two treatment groups. We used this method to adjust for selection differences between two patient populations with clinical stage I NSCLC who underwent lobectomy performed through either an open or VATS technique. We then compared the perioperative outcomes of the two adjusted groups.

f the patient data to be used from two treatment groups. We used this method to adjust for selection differences between two patient populations with clinical stage I NSCLC who underwent lobectomy performed through either an open or VATS technique. We then compared the perioperative outcomes of the two adjusted groups. Methods Approval for this retrospective data analysis of prospectively collected data was obtained from the Institutional Review Board. Patients with known or suspected clinical stage I NSCLC who underwent surgical resection at Fox Chase Cancer Center (FCCC) during two periods of time, from mid-2003 until November, 2005 and from November 2005-through 2008 were analyzed. Patients undergoing lobectomy, bilobectomy or segmentectomy were included. During the study periods, patients underwent standard preoperative staging with chest computed tomography (CT) and positron emission tomography with CT (PET/CT). At the time of the study, brain MR was performed in patients with neurologic symptoms or for patients with T2 tumors. Cervical mediastinoscopy was performed selectively for patients with abnormal lymph nodes on any imaging study, or for those with T2 or more central tumors.

ositron emission tomography with CT (PET/CT). At the time of the study, brain MR was performed in patients with neurologic symptoms or for patients with T2 tumors. Cervical mediastinoscopy was performed selectively for patients with abnormal lymph nodes on any imaging study, or for those with T2 or more central tumors. All of the operations in this series were performed by a single surgeon (WJS). Standard anesthetic management with single lung ventilation including restriction of intraoperative fluids was used in all patients. Open thoracotomy (THOR) was most commonly performed through a posterolateral incision with entry into the chest through the fifth intercostal space. Resection of a small portion of the 6th rib was routine. Muscle sparing incisions were used occasionally and rib resection was not performed in those instances. VATS lobectomy (VATS) was performed using three incisions, a 5 cm or smaller access thoracotomy and two additional 1 cm incisions. Rib spreading and rib resection were not performed. Visualization was achieved using the video thoracoscope.

s were used occasionally and rib resection was not performed in those instances. VATS lobectomy (VATS) was performed using three incisions, a 5 cm or smaller access thoracotomy and two additional 1 cm incisions. Rib spreading and rib resection were not performed. Visualization was achieved using the video thoracoscope. Lymph node dissection was routinely performed in all cases (THOR or VATS). Postoperative pain control was achieved using patient-controlled analgesia delivered through thoracic epidural catheters (PCEA) or through the use of patient-controlled intravenous narcotics (PCA). Most recently, pain control for VATS lobectomies consists of PCA narcotics supplemented with a continuous local anesthetic infusion (0.2% ropivacaine) though subpleural catheters placed intraoperatively. Patients converted from VATS to thoracotomy were included in the VATS group. All complications were graded according to the National Cancer Institute Common Terminology Criteria for Adverse Events version 3.0 [12].

tinuous local anesthetic infusion (0.2% ropivacaine) though subpleural catheters placed intraoperatively. Patients converted from VATS to thoracotomy were included in the VATS group. All complications were graded according to the National Cancer Institute Common Terminology Criteria for Adverse Events version 3.0 [12]. Patients undergoing thoracotomy (open cases) were the control population and patients undergoing VATS procedures comprised the treatment group. The baseline characteristics of the two groups are shown in table 1 (demographics). The endpoint of the study was postoperative length of stay (LOS), duration of chest tube placement, and overall complication rates. Because this was a nonrandomized comparison of two treatments, the two groups were adjusted for characteristics that were known to influence the main study outcomes. We chose age, sex, the preoperative FEV1 percent predicted and two additional measures, the Charlson Comorbidity Index (CCI) and the American Society of Anesthesiology (ASA) score. (Table 2) The CCI consists of the sum of weighted scores for 19 medical conditions that have been shown to affect mortality. The CCI has been validated in a surgical population of lung cancer patients [13]. The ASA score is a rating scale used by anesthesiologists to estimate the overall condition of a patient in the preoperative period Table 3[14]. Table 1 Charlson Comorbidity Index (CCI)

Patients undergoing thoracotomy (open cases) were the control population and patients undergoing VATS procedures comprised the treatment group. The baseline characteristics of the two groups are shown in table 1 (demographics). The endpoint of the study was postoperative length of stay (LOS), duration of chest tube placement, and overall complication rates. Because this was a nonrandomized comparison of two treatments, the two groups were adjusted for characteristics that were known to influence the main study outcomes. We chose age, sex, the preoperative FEV1 percent predicted and two additional measures, the Charlson Comorbidity Index (CCI) and the American Society of Anesthesiology (ASA) score. (Table 2) The CCI consists of the sum of weighted scores for 19 medical conditions that have been shown to affect mortality. The CCI has been validated in a surgical population of lung cancer patients [13]. The ASA score is a rating scale used by anesthesiologists to estimate the overall condition of a patient in the preoperative period Table 3[14]. Table 1 Charlson Comorbidity Index (CCI) Score Condition 1 Coronary artery diseasea Congestive heart failure Chronic pulmonary disease Peptic ulcer disease Peripheral vascular disease Mild liver disease Cerebrovascular disease Connective tissue disease Diabetes mellitus Dementia 2 Hemiplegia Moderate to severe renal disease Diabetes with end-organ damage Any prior tumor (within 5-years of diagnosis)b Leukemia Lymphoma 3 Moderate to severe liver disease 6 Metastatic solid tumor AIDS (not only HIV positive) Adapted from Birim O, EJCTS 2003 with permission

issue disease Diabetes mellitus Dementia 2 Hemiplegia Moderate to severe renal disease Diabetes with end-organ damage Any prior tumor (within 5-years of diagnosis)b Leukemia Lymphoma 3 Moderate to severe liver disease 6 Metastatic solid tumor AIDS (not only HIV positive) Adapted from Birim O, EJCTS 2003 with permission Table 2 Patient Characteristics, Unadjusted data, % or mean (SD) Thoracotomy n = 62 VATS n = 74 p-value Male 47% 30% 0.054 FEV1% pred 79 (19) 85 (19) 0.040 CCI 1.7 (1.3) 1.7 (1.3) 0.982 Age 65 (10.3) 67 (10.9) 0.489 ASA 2.3 (0.6) 1.8 (0.7) <0.001 Table 3 American Society of Anesthesiologist (ASA) Score

issue disease Diabetes mellitus Dementia 2 Hemiplegia Moderate to severe renal disease Diabetes with end-organ damage Any prior tumor (within 5-years of diagnosis)b Leukemia Lymphoma 3 Moderate to severe liver disease 6 Metastatic solid tumor AIDS (not only HIV positive) Adapted from Birim O, EJCTS 2003 with permission Table 2 Patient Characteristics, Unadjusted data, % or mean (SD) Thoracotomy n = 62 VATS n = 74 p-value Male 47% 30% 0.054 FEV1% pred 79 (19) 85 (19) 0.040 CCI 1.7 (1.3) 1.7 (1.3) 0.982 Age 65 (10.3) 67 (10.9) 0.489 ASA 2.3 (0.6) 1.8 (0.7) <0.001 Table 3 American Society of Anesthesiologist (ASA) Score Class 1 Healthy patient, no medical problems Class 2 Mild systemic disease Class 3 Severe systemic disease, but not incapacitating Class 4 Severe systemic disease that is a constant threat to life Class 5 Moribund, not expected to live 24 hours irrespective of operation An e is added to the status number to designate an emergency operation. An organ donor is usually designated as Class 6 In order to adjust for baseline differences between those patients who did and did not have VATS, we used propensity score based adjustment through propensity score based weighting [15]. Similar propensity score based weighting has been used in a variety of settings to investigate treatment effects using observational data [16,17]. The propensity scores, which are the probabilities of receiving VATS given potential confounders of treatment assignment, were estimated by a multiple logistic regression. The adequacy of the propensity score model was verified by examining adjusted differences in potential confounder variables between the treatment groups. The lack of significant differences in propensity-score adjusted averages of the confounder variables suggested that they could not be confounders after adjustment.

. The adequacy of the propensity score model was verified by examining adjusted differences in potential confounder variables between the treatment groups. The lack of significant differences in propensity-score adjusted averages of the confounder variables suggested that they could not be confounders after adjustment. We used Fisher's exact tests and T-tests to assess unadjusted differences. For inferences concerning length of stay and time until chest tube removal, we used Cox proportional hazards regressions weighted by the inverse of the probability of receiving the treatment actually received. The weight would be the inverse of the propensity score for those in the VATS group and the inverse of one minus the propensity score for those in the thoracotomy group. The bootstrap [18] with 2500 resamples was used to calculate the standard errors. Cox models were used for the time to event variables since some patients were lost to follow-up (for example, discharged with a chest tube) and hence had censored data. We used simple linear regressions similarly weighted by the inverse of the probability of receiving the treatment actually received for adjusted inferences concerning the number of lymph nodes removed and the number of lymph node stations sampled. We used propensity score based weighted logistic models for adjusted inferences concerning complication rate differences.

ted by the inverse of the probability of receiving the treatment actually received for adjusted inferences concerning the number of lymph nodes removed and the number of lymph node stations sampled. We used propensity score based weighted logistic models for adjusted inferences concerning complication rate differences. Results The study population consisted of 136 patients with clinical stage I NSCLC. There were 62 patients from 2003 through late-2005 who underwent lobectomy performed through an open thoracotomy and 74 patients who underwent VATS lobectomy from late 2005 through mid-2008. Patient demographics are shown in Table 1. There were more men in the group undergoing thoracotomy and the VATS group had slightly better pulmonary function based on preoperative FEV1 (% predicted). There were no significant differences between groups with respect to clinical stage, pathologic stage and histology (Table 4). There were no significant differences between groups with respect to the type of lobectomy performed (Table 4). Conversion of a VATS lobectomy to an open procedure occurred in 3/74 (4%) of patients. This generally involved enlargement of the access thoracotomy and placement of a small retractor, and then performance of the lobectomy with standard techniques. These patients were included in the VATS group for the purpose of analysis. Table 4 Details of Surgical Procedures Open lobe (n = 62) VATS lobe (n = 74) P value Histology Adeno 44 71% 50 68% 0.898 Squamous 9 15% 12 16% Other 9 15% 12 16% Clinical stage IA 49 79% 65 88% 0.242 IB 13 21% 9 12% Pathologic stage IA 32 52% 45 61% 0.165 IB 20 32% 12 16%

Results The study population consisted of 136 patients with clinical stage I NSCLC. There were 62 patients from 2003 through late-2005 who underwent lobectomy performed through an open thoracotomy and 74 patients who underwent VATS lobectomy from late 2005 through mid-2008. Patient demographics are shown in Table 1. There were more men in the group undergoing thoracotomy and the VATS group had slightly better pulmonary function based on preoperative FEV1 (% predicted). There were no significant differences between groups with respect to clinical stage, pathologic stage and histology (Table 4). There were no significant differences between groups with respect to the type of lobectomy performed (Table 4). Conversion of a VATS lobectomy to an open procedure occurred in 3/74 (4%) of patients. This generally involved enlargement of the access thoracotomy and placement of a small retractor, and then performance of the lobectomy with standard techniques. These patients were included in the VATS group for the purpose of analysis. Table 4 Details of Surgical Procedures Open lobe (n = 62) VATS lobe (n = 74) P value Histology Adeno 44 71% 50 68% 0.898 Squamous 9 15% 12 16% Other 9 15% 12 16% Clinical stage IA 49 79% 65 88% 0.242 IB 13 21% 9 12% Pathologic stage IA 32 52% 45 61% 0.165 IB 20 32% 12 16% II 5 8% 7 9% III, IV 5 8% 10 14% Type of resection RUL 20 32% 23 31% 0.168 RML 3 5% 10 14% RLL 8 13% 12 16% LUL 18 29% 14 19% LLL 10 16% 10 14% Bilobectomy 3 5% 1 1%

Open lobe (n = 62) VATS lobe (n = 74) P value Histology Adeno 44 71% 50 68% 0.898 Squamous 9 15% 12 16% Other 9 15% 12 16% Clinical stage IA 49 79% 65 88% 0.242 IB 13 21% 9 12% Pathologic stage IA 32 52% 45 61% 0.165 IB 20 32% 12 16% II 5 8% 7 9% III, IV 5 8% 10 14% Type of resection RUL 20 32% 23 31% 0.168 RML 3 5% 10 14% RLL 8 13% 12 16% LUL 18 29% 14 19% LLL 10 16% 10 14% Bilobectomy 3 5% 1 1% Segmentectomy 0 0% 4 5% Adjusting for observable differences between cohorts and balancing important covariates using propensity-score weighting resulted in two well-matched groups for analysis (Table 5). The perioperative results and analyses were based on comparison of data from the adjusted groups. Patients undergoing VATS lobectomy had a shorter length of stay (VATS, 4 days versus THOR, 7 days, p < 0.001, HR = 0.33 in favor of VATS). Patients in the VATS group also had their chest tubes removed earlier (VATS, 3 days versus THOR, 5 days, p < 0.0001, HR = 0.42 in favor of VATS). The number of lymph node stations dissected in the VATS and THOR groups was similar (4.6 stations versus 4.3 stations, respectively, p = 0.31) as was the total number of lymph nodes examined in the specimen (18.1 versus 14.8, p = 0.17). Mortality rates were similar, with 1/74 (1.4%) in the VATS group and 1/62 (1.6%) in the THOR group (p = 1.00). Table 5 Patient Characteristics Adjusted Data, N = 136

Segmentectomy 0 0% 4 5% Adjusting for observable differences between cohorts and balancing important covariates using propensity-score weighting resulted in two well-matched groups for analysis (Table 5). The perioperative results and analyses were based on comparison of data from the adjusted groups. Patients undergoing VATS lobectomy had a shorter length of stay (VATS, 4 days versus THOR, 7 days, p < 0.001, HR = 0.33 in favor of VATS). Patients in the VATS group also had their chest tubes removed earlier (VATS, 3 days versus THOR, 5 days, p < 0.0001, HR = 0.42 in favor of VATS). The number of lymph node stations dissected in the VATS and THOR groups was similar (4.6 stations versus 4.3 stations, respectively, p = 0.31) as was the total number of lymph nodes examined in the specimen (18.1 versus 14.8, p = 0.17). Mortality rates were similar, with 1/74 (1.4%) in the VATS group and 1/62 (1.6%) in the THOR group (p = 1.00). Table 5 Patient Characteristics Adjusted Data, N = 136 Thoracotomy n = 62 VATS n = 74 Male 40% 39% FEV1% pred 80 (19) 81 (20) CCI 1.6 (1.3) 1.7 (1.3) Age 66 (10) 66 (11) ASA 2.1 (0.6) 2.1 (0.7) Perioperative and postoperative complications were classified and graded using the National Cancer Institute Common Terminology Criteria for Adverse Events version 3.0 (Table 6). No differences in complication rates were noted for either cohort (VATS or THOR). No complications were recorded for 40/62 (65%) THOR patients compared to 52/74 (70%) of VATS patients (p = 0.61). There was a trend suggesting a decrease in pulmonary complications in the VATS group 10/74 (14%) compared to the THOR group 20/62 (32%) (p = 0.10). Continuous heart rate monitoring was routine during the second period of the study when VATS lobectomy was initiated. Therefore the low number of arrhythmias in the THOR patients may reflect the fact that only those arrhythmias that were persistent or were associated with symptoms were recorded for those patients, whereas even brief periods of atrial fibrillation were detected and recorded for the VATS group.

lobectomy was initiated. Therefore the low number of arrhythmias in the THOR patients may reflect the fact that only those arrhythmias that were persistent or were associated with symptoms were recorded for those patients, whereas even brief periods of atrial fibrillation were detected and recorded for the VATS group. Table 6 Complication Rates Complication Open N = 62 Unadjusted data VATS N = 74 Unadjusted data P value (unadjusted) P value (adjusted) None 40 (65%) 52 (70%) 0.58 0.61 Pulmonary 20 (32%) 10 (14%) 0.01 0.10 Cardiac-arrhythmia 2 (3%) 10 (14%) 0.07 0.71 Hemorrhage 1 (2%) 4 (5%) 0.38 0.88 Other 1 (2%) 2 (3%) 1.00 0.98 When complications were examined by age (70 years and older){Table 7), the trend toward a decrease in pulmonary complications associated with VATS was present but still did not reach statistical significance, perhaps because of the small number of patients in each group. Table 7 Complication rates in patients aged 70 and older Complication Open N = 22 VATS N = 30 P value (adjusted) None 14 (64%) 21 (70%) 0.77 Pulmonary 7 (32%) 3 (10%) 0.08 Cardiac-arrhythmia 2 (9%) 5 (17%) 0.69 Hemorrhage 0 (0%) 1 (3%) 1.00

Other 1 (2%) 2 (3%) 1.00 0.98 When complications were examined by age (70 years and older){Table 7), the trend toward a decrease in pulmonary complications associated with VATS was present but still did not reach statistical significance, perhaps because of the small number of patients in each group. Table 7 Complication rates in patients aged 70 and older Complication Open N = 22 VATS N = 30 P value (adjusted) None 14 (64%) 21 (70%) 0.77 Pulmonary 7 (32%) 3 (10%) 0.08 Cardiac-arrhythmia 2 (9%) 5 (17%) 0.69 Hemorrhage 0 (0%) 1 (3%) 1.00 Other 0 (0%) 1 (3%) 1.00 Discussion We used propensity score based weighting to perform a non-randomized comparison of patients who underwent wither open thoracotomy and lobectomy or VATS lobectomy for clinical stage I NSCLC. The statistical method that we used generates an adjusted population of patients whose outcomes are slightly different that the observed data. For example, that is the reason that the complication rates in Table 8 (39% for THOR and 34% for VATS) are slightly different than the unadjusted complication rates in Table 6 (35% for THOR and 30% for VATS). The adjusted rates are based on a larger number of patients and therefore should be considered the most generalizable data. After balancing the two cohorts based on preoperative covariates known to influence perioperative outcomes using propensity score based weighting, we found that patients undergoing VATS lobectomy for clinical stage I NSCLC had their chest tubes removed earlier (adjusted median 3 days versus 5 days, p < 0.001) and also had a shorter postoperative length of stay (adjusted median 4 days versus 7 days, p < 0.001)) than patients undergoing open thoracotomy. This is consistent with what others have observed from case control series [9,10] and large retrospective case series [6]. Patients undergoing VATS lobectomy may experience less pain, have improved respiratory mechanics, and develop less of an inflammatory response compared to open thoracotomy, all of which combined may lead to a decreased postoperative length of stay.

rom case control series [9,10] and large retrospective case series [6]. Patients undergoing VATS lobectomy may experience less pain, have improved respiratory mechanics, and develop less of an inflammatory response compared to open thoracotomy, all of which combined may lead to a decreased postoperative length of stay. Table 8 Perioperative Results, Adjusted Data

rom case control series [9,10] and large retrospective case series [6]. Patients undergoing VATS lobectomy may experience less pain, have improved respiratory mechanics, and develop less of an inflammatory response compared to open thoracotomy, all of which combined may lead to a decreased postoperative length of stay. Table 8 Perioperative Results, Adjusted Data Op. N Peri-op mortality (%) Length of Stay (Adjusted median days) Chest tube removal (Adjusted median days) Adverse Events (Adjusted % of patients with at least one) Lymph Node Stations (Adjusted # sampled/patient) Lymph Nodes (Adjusted mean # removed/patient) Thor 62 1 (1.6%) 7 5 39% 4.6 18.1 VATS 74 1 (1.4%) 4 3 34% 4.3 14.8 P = 1.00 HR = 0.33 HR = 0.42 P-value P < 0.0001 P = 0.0001 P = 0.61 P = 0.31 P = 0.17 Despite the marked reduction in postoperative length of stay, we found no statistically significant difference in overall complication rates or specific complication rates between VATS patients and THOR patients. One reason for this is that the number of episodes of atrial fibrillation in the THOR group was probably underestimated since routine use of postoperative heart rhythm monitoring was phased in during the first part of the study when the open lobectomy operations were performed. Therefore, it is likely that episodes of atrial fibrillation in this group were missed and that the true rate of atrial fibrillation in this group was much higher. This would have increased the overall and arrhythmia complication rates in the THOR group. Other studies have suggested that atrial fibrillation rates after VATS lobectomy are not different than after thoracotomy and lobectomy [8,9].

re missed and that the true rate of atrial fibrillation in this group was much higher. This would have increased the overall and arrhythmia complication rates in the THOR group. Other studies have suggested that atrial fibrillation rates after VATS lobectomy are not different than after thoracotomy and lobectomy [8,9]. Mortality rates were not different between groups in our study. We did observe a trend (p = 0.10) toward a decrease in respiratory complication rates in the VATS patients, especially in the 70 year and older group, with the rates of respiratory complications decreasing from 32% in the open group to 10% in the VATS group (p = 0.08). A decrease in respiratory complications in older patients undergoing VATS lobectomy compared to THOR was observed by Cattaneo et al. [10]. It is possible that a statistically significant difference in respiratory complication rates would have been observed in our study if the sample size had been larger. Regarding oncologic efficacy of VATS compared to THOR, we found that all patients underwent compete resections with negative margins. The number of lymph node stations that were dissected and the total number of lymph nodes retrieved were similar for each group. Continued follow up will be required to determine oncologic efficacy of VATS compared to THOR.

compared to THOR, we found that all patients underwent compete resections with negative margins. The number of lymph node stations that were dissected and the total number of lymph nodes retrieved were similar for each group. Continued follow up will be required to determine oncologic efficacy of VATS compared to THOR. The limitations of this study include its retrospective nature, the fact that management changes may have occurred over the study period that were not appreciated or reflected in the analysis (such as the change to continuous heart rate monitoring or other changes in postoperative management), and the sample size. The fact that is not a multicenter study is mitigated by the fact that our conclusions, while analyzed somewhat differently, are similar to those of other authors. The strengths of the study include the fact that all cases were performed by a single surgeon, it is a consecutive case series, and that covariates were balanced between the two cohorts using sophisticated statistical analysis.

our conclusions, while analyzed somewhat differently, are similar to those of other authors. The strengths of the study include the fact that all cases were performed by a single surgeon, it is a consecutive case series, and that covariates were balanced between the two cohorts using sophisticated statistical analysis. Conclusion In summary, we found that patients undergoing VATS lobectomy were able to have their chest tubes removed sooner and were discharged sooner after undergoing lobectomy for clinical stage I NSCLC than patients who underwent thoracotomy and lobectomy for the same indications. A trend toward a decrease in pulmonary complications in the VATS group was also observed, especially in patients 70 years of age or older. The use of propensity score-based weighting in this series and its application to larger databases will allow clinicians to draw stronger conclusions about the relative merits of VATS lobectomy versus open thoracotomy and lobectomy in the treatment of the patient with early stage lung cancer.

70 years of age or older. The use of propensity score-based weighting in this series and its application to larger databases will allow clinicians to draw stronger conclusions about the relative merits of VATS lobectomy versus open thoracotomy and lobectomy in the treatment of the patient with early stage lung cancer. Abbreviations ASA: American Society of Anesthesiologists; CCI: Charlson comorbidity index; CT: computed tomography; CTC: common toxicity criteria; FCCC: Fox Chase Cancer Center; FEV1 forced expiratory volume at 1 second; HR: hazard ratio; LOS: length of stay; MR: magnetic resonance; NCI: national cancer institute; NSCLC: non-small cell lung cancer; PCA: patient-controlled analgesia; PCEA: patient-controlled epidural analgesia; PET/CT Positron emission tomography and computed tomography; THOR: open thoracotomy and lobectomy; VATS: video-assisted thoracic surgery. Competing interests The authors declare that they have no competing interests. Authors' contributions WS performed all surgeries and drafted the manuscript. RM collected data and revised the manuscript in its final version. SO and JFF provided research support and data collection. Oseni presented these data at the American College of Chest Physicians meeting in fall, 2008. BLE is responsible for the statistical analysis of these data. All authors have read and approved the final manuscript. Acknowledgements This Research supported in part by NIH grants P30 CA 06927 and an appropriation from the Commonwealth of Pennsylvania.

Authors' contributions WS performed all surgeries and drafted the manuscript. RM collected data and revised the manuscript in its final version. SO and JFF provided research support and data collection. Oseni presented these data at the American College of Chest Physicians meeting in fall, 2008. BLE is responsible for the statistical analysis of these data. All authors have read and approved the final manuscript. Acknowledgements This Research supported in part by NIH grants P30 CA 06927 and an appropriation from the Commonwealth of Pennsylvania. Portions of this work were presented at the 74th Annual Meeting of the American College of Chest Physicians, Philadelphia, PA, October 29, 2008.

Introduction Gallstones are common in patients throughout the western world and are found in about 10% of the adult population [1]. Gallstone related disease is one of the most frequent medical problems demanding surgical intervention. In Denmark, 130 per 100.000 inhabitants are cholecystectomized each year [2], most often because of biliary attacks of pain. The annual incidence of acute cholecystitis, the second largest group undergoing cholecystectomy, is approximately 20 per 100.000 inhabitants [2]. In a subgroup of patients with acute cholecystitis, surgical intervention is hazardous due to poor performance status. In these patients, the standard treatment options have traditionally been conservative treatment or percutaneous transperitoneal cholecystostomy (PTCS) [3-7]. A significant drawback with both these treatment modalities is a high rate of recurrences, reported to range from 15% to 47% [8,9]. In recent years, an alternative treatment option, endoscopic gallbladder drainage (EGBD), has emerged [10-14]. One theoretical advantage of EGBD compared with PTCS, is that treatment can be sustained for longer periods, because there is not, as in PTCS, an external component of the drain. This possibility for prolonged treatment raises the hope that EGBD may be a more definitive treatment of cholecystitis than PTCS.

ed [10-14]. One theoretical advantage of EGBD compared with PTCS, is that treatment can be sustained for longer periods, because there is not, as in PTCS, an external component of the drain. This possibility for prolonged treatment raises the hope that EGBD may be a more definitive treatment of cholecystitis than PTCS. Calculous cholecystitis is thought to develop when the cystic duct becomes obstructed by an impacting gallstone. Secondary bacterial contamination of the stagnant bile seems an obvious mechanism for exacerbation in the development of acute cholecystitis. Cultures of gallbladder bile however, are only positive in 15% to 30% of cases [15], suggesting that the inflammatory process most often could be of another nature, some suggest inflammation to be caused by a chemical agent [16]. The predominant microorganisms isolated from gallbladder bile in patients with acute cholecystitis are Escherichia coli (60%) and Klebsiella pneumoniae (22%) [15]. Occlusion of the arterial blood supply to the gallbladder may be a fundamental element in the pathogenesis of acute acalculous cholecystitis [17]. Several groups have tried to induce acute cholecystitis in animal models by combining surgical induced cholestasis with either bacterial infection, chemical irritants and/or gallbladder ischaemia[16,18]. We wanted to induce a severe condition of acute cholecystitis in order to demonstrate the effect of the intervention. Therefore we combined the insults, ligating the cystic duct and artery and inoculation of bacteria.

Several groups have tried to induce acute cholecystitis in animal models by combining surgical induced cholestasis with either bacterial infection, chemical irritants and/or gallbladder ischaemia[16,18]. We wanted to induce a severe condition of acute cholecystitis in order to demonstrate the effect of the intervention. Therefore we combined the insults, ligating the cystic duct and artery and inoculation of bacteria. The aim of the present study was to investigate whether internal drainage of the gallbladder could protect against the development of acute cholecystitis in a pig model. Materials and methods The research protocol was approved by the local research committee (registration number: 2008/561-1489) in accordance with the Danish regulations on animal experiments. Twenty female pigs (Danish Landrace/Yorkshire) with a body weight of approximately 65 kg (Research Centre Foulum under the Danish Institute of Agricultural Sciences) were used for the experiment. Animals were randomized in blocks of four. The pigs were divided into two groups and all of the pigs had acute cholecystitis induced as described below. Drained pigs had an internal double pigtail catheter from the gallbladder to the duodenum. In the undrained pigs the catheter was placed as in drained pigs, but was then immediately removed. Acute cholecystitis was induced using a combination of previously described models [16,19-22] with slight modifications, thus comprising: 1) Ligation of the cystic artery. 2) Ligation of the cystic duct. 3) Injection of bacteria into the gallbladder lumen.

Materials and methods The research protocol was approved by the local research committee (registration number: 2008/561-1489) in accordance with the Danish regulations on animal experiments. Twenty female pigs (Danish Landrace/Yorkshire) with a body weight of approximately 65 kg (Research Centre Foulum under the Danish Institute of Agricultural Sciences) were used for the experiment. Animals were randomized in blocks of four. The pigs were divided into two groups and all of the pigs had acute cholecystitis induced as described below. Drained pigs had an internal double pigtail catheter from the gallbladder to the duodenum. In the undrained pigs the catheter was placed as in drained pigs, but was then immediately removed. Acute cholecystitis was induced using a combination of previously described models [16,19-22] with slight modifications, thus comprising: 1) Ligation of the cystic artery. 2) Ligation of the cystic duct. 3) Injection of bacteria into the gallbladder lumen. Surgery day 0 After premedication with an intramuscular injection of Midazolam 0.4 mg/kg and Ketamine 4 mg/kg, the pigs were intubated and mechanically ventilated (Servo 900 ventilator; Siemens-Elema, Solna, Sweden) with a mixture of air, oxygen and 1.5% isoflurane. Fentanyl was given as a continuous intravenous infusion 10-15 ml/h. A midline laparotomy was performed. The infundibulum of the gallbladder was identified and in its proximity the cystic duct and the cystic artery was dissected and exposed. A duodenotomy was made by an anti-mesenterial incision 1-2 cm distal to the pylorus. The papilla Vateri was identified and cannulated with a catheter (Cook; Angiography catheter HNB7.0-NT-100-M-Ns-CN) which was inserted into the gallbladder and through this a guidewire (Boston Scientific; Jagwire 0.035/450) was placed. After removal of the catheter a double pigtail catheter 7F, 9 cm (Cook; Endoprosthesis T7.0-35-9-35-16S-DENDR) was placed guided by the wire with the proximal tip in the lumen of the gallbladder and the distal tip in the lumen of the duodenum. Hereafter, the guidewire was removed in all animals and in half of them the pigtail catheter was removed. In all pigs, the cystic artery and cystic duct were ligated and the duodenotomy was closed with a single layer continuous seromuscular suture. Finally, 2-4 ml of bile was extracted through the wall of the gallbladder, sent for microbiological examination and replaced with a 2 ml suspension in physiological salt water of an overnight broth culture of bacteria (1 ml of Escherichia coli 105/ml and 1 ml of Klebsiella pneumonia 105/ml). The abdomen was closed in two layers with staples in the skin.

ed through the wall of the gallbladder, sent for microbiological examination and replaced with a 2 ml suspension in physiological salt water of an overnight broth culture of bacteria (1 ml of Escherichia coli 105/ml and 1 ml of Klebsiella pneumonia 105/ml). The abdomen was closed in two layers with staples in the skin. Evaluation day 4 The pigs were killed on day four with a captive bolt pistol. A midline laparotomy was performed and the gallbladder was exposed. Bile was extracted for microbiological analysis and the gallbladder was removed and fixed in the physiological expanded state by injecting 10% formaldehyde solution into the lumen of the gallbladder, thus avoiding artifacts such as changes in wall thickness that can be seen in the collapsed state. During the experiment, tests were done according to Table 1. Table 1 Overview of test chronology. Test Day 0 Day 1 Day 2 Day 3 Day 4 Weight X X Temperature X X X X X Blood samples X X X X X Blood culture X Bile culture X X Histology X X indicating a test was done Bloodtests: neutrophils, lymphocytes, eosinophils, monocytes, red blood cells, haemaglobin, haematocrit; liver blood tests: alanin amino transpherase, albumin, alkaline phosphatase, gamma glutamyl transpherase, bilirubin, amylase, acute phase reactants, electrolytes.

Bile culture X X Histology X X indicating a test was done Bloodtests: neutrophils, lymphocytes, eosinophils, monocytes, red blood cells, haemaglobin, haematocrit; liver blood tests: alanin amino transpherase, albumin, alkaline phosphatase, gamma glutamyl transpherase, bilirubin, amylase, acute phase reactants, electrolytes. Culture of bile and blood: The BacT/ALERT(r) blood culture system (bioMerieux) was used for blood culturing. Ten ml of blood was drawn for an aerobic and an anaerobic standard blood culture bottle and incubated for seven days. Bile aspirates were spread on agar plates and cultured aerobically at 36°C and 5% C02 atmosphere (5% blood agar plate and a MacConkey agar plate, Statens Serum Institut, Copenhagen, Denmark) and anaerobically at 36°C for two days (Anaerobic blood agar plates, Statens Serum Institut, Copenhagen, Denmark). All culture positive isolates were subcultured and identified according to conventional microbiological laboratory methods. Histological examination: Following fixation, sections of gallbladder extending from the cystic duct to the fundus were embedded in paraffin using standard methods. Paraffin sections were cut at 5 μm and stained with haematoxylin and eosin, and with Masson's trichrome. Histological examination was performed in a blinded manner by an experienced pathologist using a modification of a generally accepted histological scoring system and a model with reproducible histological changes [21].

affin sections were cut at 5 μm and stained with haematoxylin and eosin, and with Masson's trichrome. Histological examination was performed in a blinded manner by an experienced pathologist using a modification of a generally accepted histological scoring system and a model with reproducible histological changes [21]. Each gallbladder in the present study was graded with this scoring system without knowledge of the intervention. The code was subsequently broken and the data was collated. Briefly, histological score was based on the presence, extent, and severity of haemorrhage, edema, formation of pregranulation tissue (fibroblast proliferation and increased ground substance), lymphatic dilatation, and polymorphonuclear cell (PMN) infiltration in the wall of each gallbladder. Lymphatic dilatation and the presence of PMNs were scored as being either present (1) or absent (0) since their degree of participation in the inflammatory response was essentially all or nothing. Hemorrhage, edema, and pregranulation tissue formation, showed more variation, and were graded accordingly from 0 to 3 (0 = not present, 1 = mild, 2 = moderate, 3 = severe). The individual scores for each histological parameter were added together to give a total possible score of between 0 (histologically normal looking gallbladder, and 11 (the most severely inflamed gallbladder) [21].

ariation, and were graded accordingly from 0 to 3 (0 = not present, 1 = mild, 2 = moderate, 3 = severe). The individual scores for each histological parameter were added together to give a total possible score of between 0 (histologically normal looking gallbladder, and 11 (the most severely inflamed gallbladder) [21]. Statistical analysis Blood, temperature, and weight data from drained and undrained pigs were compared using unpaired parametric (Student's t test) or non-parametric tests (Mann-Whitney U test). Results of blood tests, temperature, and weight were reported as mean and standard error of mean. The histological scores were compared by category and total score using a non-parametric test (Kruskal-Wallis). The assumption of normal distribution was evaluated by visual inspection of QQ-plots and histograms. Non-parametric statistics were applied in the case of skewed distribution. P < 0.05 was considered statistically significant. Results Of the 20 pigs included in the study, 16 survived the full four days after the initial operation leaving eight pigs in each group. Four pigs died, three because of cholascos, one due to gastric retention. There were no significant differences in any blood tests, temperature or weight between the groups. Bile All drained pigs had a yellowish clear transparent bile on day four, whereas undrained pigs had dark green to black non-transparent bile with sludge (Fig. 1, 2 and 3). Figure 1 Bile sample from day 0. Figure 2 Bile sample from gallbladder with drainage catheter on day four.

Results Of the 20 pigs included in the study, 16 survived the full four days after the initial operation leaving eight pigs in each group. Four pigs died, three because of cholascos, one due to gastric retention. There were no significant differences in any blood tests, temperature or weight between the groups. Bile All drained pigs had a yellowish clear transparent bile on day four, whereas undrained pigs had dark green to black non-transparent bile with sludge (Fig. 1, 2 and 3). Figure 1 Bile sample from day 0. Figure 2 Bile sample from gallbladder with drainage catheter on day four. Figure 3 Bile sample from gallbladder without drainage catheter on day four. Bacteria Nine of sixteen bile samples collected on day 0 in the pigs who survived were non-sterile. Escherichia coli alone was found in seven cultures, Escherichia coli and Klebsiella pneumoniae in one culture and finally Non-haemolytic Streptococci in one. On day four a combination of Escherichia coli and Klebsiella pneumoniae were found in all 14 bile cultures, Escherichia coli alone in one and Klebsiella pneumoniae alone in one culture. No other bacteria were cultured from the bile samples. In drained pigs, six and in undrained pigs, eight blood cultures showed bacterial growth on day four; species are shown in Table 2. Table 2 Blood culture results from day four. Name of bacteria Drained Undrained Escherichia coli 5 4 Non-haemolytic Streptococci 2 5 Klebsiella pneumoniae 3 2 Coagulase-negative Staphylococci 1 3 Enterococci 2 3 No bacteria 2 0 Eight pigs in each group.

In drained pigs, six and in undrained pigs, eight blood cultures showed bacterial growth on day four; species are shown in Table 2. Table 2 Blood culture results from day four. Name of bacteria Drained Undrained Escherichia coli 5 4 Non-haemolytic Streptococci 2 5 Klebsiella pneumoniae 3 2 Coagulase-negative Staphylococci 1 3 Enterococci 2 3 No bacteria 2 0 Eight pigs in each group. Histopathology The inflammation scoring system demonstrated a highly statistically significant difference in all examined parameters. In all instances, scores were higher in gallbladders from undrained pigs (Table 3). Table 3 Histopathological scoring system. Lymphatic dilatation (0-1) Polymorphnuclear cells (0-1) Hemorrhage (0-3) Edema (0-3) Pregranulation tissue (0-3) Sum (0-11) Drained 0.25(-0.14; 0.64) 0.25(-0.14; 0.64) 0 (0; 0) 0.25 (-0.14; 0.64) 0.38(-0.06; 0.81) 1.13(-0.39;2.64) Undrained 1(1;1) 1(1;1) 0.75(0.36;1.14) 1.50 (0.87; 2.13) 2(01.55;2.45) 6.25(5.38;7.12) p-value 0.0027 0.0027 0.0027 0.0023 0.0008 0.0006 The range of the possible categorical score is indicated in the paragraph in the top line. The scores for drained and undrained are demonstrated as mean (confidence interval). In the bottom line the p-values are presented. The histological differences between drained an undrained gallbladders are illustrated in Figure 4, 5 and 6. Figure 4 Gallbladder from drained pig. The mucosa is intact with a normal number of mononuclear cells (mostly lymphocytes). The muscle and subserosal layers are unremarkable without significant edema, lymphatic dilatation or inflammation. Hematoxylin & eosin.

The histological differences between drained an undrained gallbladders are illustrated in Figure 4, 5 and 6. Figure 4 Gallbladder from drained pig. The mucosa is intact with a normal number of mononuclear cells (mostly lymphocytes). The muscle and subserosal layers are unremarkable without significant edema, lymphatic dilatation or inflammation. Hematoxylin & eosin. Figure 5 Gallbladder from undrained pig four days after ligation. The mucosa, which can just be seen in the upper left-hand part of the figure, shows partial loss. There is inflammation of the underlying muscle layer, whilst the subserosal region is markedly expanded with edema, lymphatic dilatation, formation of pre-granulation tissue and extensive acute inflammation. Hematoxylin & eosin. Figure 6 Higher power picture of Figure 5 of the subserosa showing focal haemorrhage with pre-granulation tissue and marked acute inflammation. Hematoxylin & eosin. Discussion The present study in a pig model shows that internal drainage from the gallbladder to the duodenum protects against development of acute cholecystitis judged by a histopathological scoring system. Internal drainage could not prevent the bile from still being infected with bacteria inoculated into the gallbladder lumen four days earlier. There was, however, a distinct difference in the macroscopic appearance in the sense that the bile in the internal drainage group was yellow and transparent whereas it was dark and non-transparent in the undrained pigs.

ile from still being infected with bacteria inoculated into the gallbladder lumen four days earlier. There was, however, a distinct difference in the macroscopic appearance in the sense that the bile in the internal drainage group was yellow and transparent whereas it was dark and non-transparent in the undrained pigs. Multiple factors have been hypothesized to be involved in the development of acute cholecystitis and, therefore, have been used in different animal models. Obstruction of the cystic duct by an impacting gallstone has been the primary causative factor to be proposed, whilst other factors to be implicated include impairment of the gallbladder blood supply, secondary bacterial infection, and abnormal concentrations of bile constituents leading to a chemical inflammation [16,19-22]. In the actual study, acute cholecystitis was induced combining previous described models with modifications comprising ligation of the cystic artery, ligation of the cystic duct and injection of a 2 ml suspension of bacteria (1 ml of Escherichia coli 105/ml and 1 ml of Klebsiella pneumonia 105/ml) [16,19-22]. The severeness of the combined insults might explain the high mortality. Sections from the gallbladders were scored using a previously described histopathological scoring system and showed pronounced signs of inflammation in the group without internal drainage, indicating that the model used for inducing acute cholecystitis in the actual study was adequate [21].

Multiple factors have been hypothesized to be involved in the development of acute cholecystitis and, therefore, have been used in different animal models. Obstruction of the cystic duct by an impacting gallstone has been the primary causative factor to be proposed, whilst other factors to be implicated include impairment of the gallbladder blood supply, secondary bacterial infection, and abnormal concentrations of bile constituents leading to a chemical inflammation [16,19-22]. In the actual study, acute cholecystitis was induced combining previous described models with modifications comprising ligation of the cystic artery, ligation of the cystic duct and injection of a 2 ml suspension of bacteria (1 ml of Escherichia coli 105/ml and 1 ml of Klebsiella pneumonia 105/ml) [16,19-22]. The severeness of the combined insults might explain the high mortality. Sections from the gallbladders were scored using a previously described histopathological scoring system and showed pronounced signs of inflammation in the group without internal drainage, indicating that the model used for inducing acute cholecystitis in the actual study was adequate [21]. Using a histopathological scoring system based on the presence, extent and severity of hemorrhage, edema, formation of pregranulation tissue, lymphatic dilatation, and PMC infiltration in the gallbladder wall, we found a highly significant protective effect of internal drainage from the gallbladder to the duodenum in the development of acute cholecystitis. These findings support the theory that a gallstone impacting the cystic duct plays a crucial role as a pathogenetic mechanism in the development of acute cholecystitis and it may be an important factor in the maintenance of this disease state. There were only mild degrees of inflammation identifiable microscopically in the internal drainage group, in spite of the fact, that these animals had the cystic artery ligated and had bacteria inoculated into the gallbladder. This suggests that the latter factors are of relatively less importance as single pathogenetic insults compared to outflow impediment in the development of acute cholecystitis. The findings in the actual study also allows us to speculate that internal drainage might be a way to treat already evolved acute cholecystitis.

his suggests that the latter factors are of relatively less importance as single pathogenetic insults compared to outflow impediment in the development of acute cholecystitis. The findings in the actual study also allows us to speculate that internal drainage might be a way to treat already evolved acute cholecystitis. Surprisingly, bile cultures sampled day 0 showed bacterial growth in nine of 16 cases. This could not be due to contamination from the inoculate, since extraction of bile was done prior to injection of bacteria, but we offer two other explanations: 1. Retrograde contamination from the duodenum while placing the stent, since this was done prior to extraction of bile or 2. In a high proportion of normal pigs, bile is colonized with bacteria from the intestinal tract. On day four, all bile cultures were contaminated with bacteria in both of the groups, suggesting that infected bile as a sole factor is insufficient to produce acute cholecystitis. These findings also support the above mentioned theory that gallstone impaction in the cystic duct is the key pathogenetical mechanism in the development acute cholecystitis, i.e. acute cholecystitis will not develop if the bile drainage through the cystic duct is preserved in spite of bacterial infection.

olecystitis. These findings also support the above mentioned theory that gallstone impaction in the cystic duct is the key pathogenetical mechanism in the development acute cholecystitis, i.e. acute cholecystitis will not develop if the bile drainage through the cystic duct is preserved in spite of bacterial infection. Fourteen of 16 pigs showed bacterial growth in blood cultures on day four. Some of the strains cultured (i.e. Staphylococci and Steptococci) could be due to contamination from the skin. However, Escherichia coli and Klebsiella pneumonia demonstrated in blood cultures most probably originated from the gallbladder as these were the predominant bacteria cultured from bile. It is not possible to determine whether the positive blood cultures had any effect on the well-being of the animals in the present study as we had no control animals with negative blood cultures. In conclusion, the present study demonstrated that internal drainage from the gallbladder to the duodenum prevents the development of acute cholecystitis. Since development of acute cholecystitis can be prevented by internal drainage, future studies must demonstrate whether internal drainage can be applicable as a therapeutic intervention in patients suffering from accumulated biliary attacks of pain or in patients already suffering from acute cholecystitis. Competing interests The authors declare that they have no competing interests.

In conclusion, the present study demonstrated that internal drainage from the gallbladder to the duodenum prevents the development of acute cholecystitis. Since development of acute cholecystitis can be prevented by internal drainage, future studies must demonstrate whether internal drainage can be applicable as a therapeutic intervention in patients suffering from accumulated biliary attacks of pain or in patients already suffering from acute cholecystitis. Competing interests The authors declare that they have no competing interests. Authors' contributions DWK participated in the design of the study, carried out the surgery, drafted the manuscript and performed the statistical analysis. FVM participated in the design of the study and helped to draft the manuscript. JKM was responsible for analysis of cultured bacteria. SJHD did the histological examination and helped to draft the manuscript. PFJ conceived of the study, and participated in its design and coordination and helped to draft the manuscript. All authors read and approved the final manuscript. Acknowledgements Mette Jensen is thanked for excellent laboratory assistance.

Background Oncologic outcomes after surgical treatment of rectal cancer have been improved by techniques such as the total mesorectal excision (TME) [1,2]. The division of the distal rectum with adequate tumor clearance is a critical step in a successful TME, and this is most commonly achieved using a stapling device. The ability to place a stapling device deep in the pelvis with good visualization could determine whether a sphincter-preserving operation is performed, or whether an abdominoperineal resection would be necessary. A distal resection margin of 2 cm has been generally accepted as oncologically sound [3,4], although recent studies suggest that a resection margin as short as 1 cm is safe, especially after neoadjuvant chemoradiotherapy [5-7]. Even this 1 cm distal margin, however, could be jeopardized by patient factors and ergonomically incorrect stapling devices. Appropriate distal rectal stapling is of vital importance since there is a correlation between close distal rectal margins and rectal cancer recurrence [8]. Furthermore, struggling to place a rectal stapler at the pelvic floor could lead to unwarranted trauma to a malignant tumor or even to perforation of the rectum, risking local tumor recurrence.

ling is of vital importance since there is a correlation between close distal rectal margins and rectal cancer recurrence [8]. Furthermore, struggling to place a rectal stapler at the pelvic floor could lead to unwarranted trauma to a malignant tumor or even to perforation of the rectum, risking local tumor recurrence. Surgeons specifically trained in deep pelvic surgery may deal with a close distal margin utilizing a hand-sewn coloanal anastomosis, where the rectum is amputated from the anus through a transanal approach, and then an anastomosis is established between the colon and anal canal using manually placed sutures. This approach remains the gold standard in cases of a threatened distal margin. In reality, however, surgeons are often not trained to perform this procedure or find it challenging. This point is illustrated in a study of a nationwide database in the US, where patients treated by surgeons with a high volume of rectal cancer surgery (≥ 10 per year) were five times more likely to undergo a sphincter-saving operation for rectal cancer compared with low volume surgeons (1-3 cases per year) [9]. A dependable, user-friendly stapler designed for distal rectal stapling could potentially have allowed more of these patients to undergo a restorative operation. A stapling device should not only be easy to apply, but must also be reliable since failure to form a proper staple line could lead to serious adverse outcomes such as an anastomotic leak.

apler designed for distal rectal stapling could potentially have allowed more of these patients to undergo a restorative operation. A stapling device should not only be easy to apply, but must also be reliable since failure to form a proper staple line could lead to serious adverse outcomes such as an anastomotic leak. Distal rectal stapling remains one of the most difficult challenges in surgery. The pelvis is restricted by its bony confines, and within this space exist other structures such as the bladder, prostate, uterus, and vagina. This complexity is further magnified in a narrow male pelvis. A proper stapling device must have a small profile, small enough to be passed around a bulky rectal tumor within a fixed space, yet be large enough to occlude the entire rectum and divide it accurately at a right angle. A stapler that reliably and uniformly reaches the most distal aspect of the rectum will have major impact on surgical outcome and surgeon satisfaction, reducing the level of frustration associated with a critical part of a difficult operation. Thus far, advancements in stapling technology have allowed for more sphincter-saving procedures to be performed for distal rectal tumors, but despite this, the current stapling tools remain far from perfect.

urgeon satisfaction, reducing the level of frustration associated with a critical part of a difficult operation. Thus far, advancements in stapling technology have allowed for more sphincter-saving procedures to be performed for distal rectal tumors, but despite this, the current stapling tools remain far from perfect. The CONTOUR® curved cutter (CC) (Ethicon Endo-Surgery, Cincinnati, OH) is a single-patient-use stapler that was designed with a curved head that cuts and staples. The device delivers four staggered rows of titanium staples, with a knife between the second and third row. When applied, it occludes the tissue on both sides of a 40-mm transection. The CC is available with "green" and "blue" staple cartridges that compress tissue to approximately 2 mm and 1.5 mm, respectively. The iDrive™ powered handle (Covidien, New Haven, CT) utilizes a microprocessor-controlled hand-held unit that controls the functions of stapler closing, firing, and cutting with the push of a button. The iDrive™ powered handle is compatible with a low-profile 45-mm right angle linear cutter (RALC) single use reload, which delivers four staggered rows of titanium staples, with a knife between the second and third row of staples. When applied, it occludes the tissue on both sides of a 45-mm transection. The RALC single use reload attaches to the iDrive™ powered handle such that the jaws are perpendicular to the shaft of the handle. In addition, the surgeon has the ability to select whether to use the device in a "green" (compresses tissue to 2.0 mm) or "blue" mode (compresses tissue from 1.5 mm to 2.0 mm) without having to change the staple reload cartridge.

e iDrive™ powered handle such that the jaws are perpendicular to the shaft of the handle. In addition, the surgeon has the ability to select whether to use the device in a "green" (compresses tissue to 2.0 mm) or "blue" mode (compresses tissue from 1.5 mm to 2.0 mm) without having to change the staple reload cartridge. A comparative study of these two innovative stapling devices was undertaken in a human cadaver model to evaluate stapler access and visibility in the deep pelvis. Furthermore, we aimed to study what anatomic factors present the biggest impediments to distal rectal stapling using the current stapling devices. Methods Twelve male cadavers underwent pelvic dissection by four surgeons (all certified by the American Board of Surgery, three with board specialty in Colon and Rectal Surgery). A low midline incision was created, and the rectum was mobilized circumferentially to the pelvic floor as in a TME. The anatomic landmarks of each pelvis were carefully studied. Distances were measured and recorded between: 1. the symphysis pubis and umbilicus, 2. the right and left anterior superior iliac spines, 3. the symphysis pubis and the sacral promontory, 4. the pelvic floor and the sacral promontory, 5. the tip of the coccyx and the symphysis pubis, 6. the right and left pelvic sidewalls (i.e., transverse diameter of the pelvic inlet), and 7. the anal verge and the pelvic floor.

and left anterior superior iliac spines, 3. the symphysis pubis and the sacral promontory, 4. the pelvic floor and the sacral promontory, 5. the tip of the coccyx and the symphysis pubis, 6. the right and left pelvic sidewalls (i.e., transverse diameter of the pelvic inlet), and 7. the anal verge and the pelvic floor. Both the CC and RALC 45-mm stapling devices were applied to the rectum, advanced as deep as possible in the pelvis in both coronal and sagittal configurations (see Figure 1 and 2), and engaged. All four surgeons operated independently on all twelve cadavers, for a total of 48 applications for each stapler per position. To increase objectivity, randomization was used to determine which stapling device was used before the other. After the staplers were placed as distally as possible on the rectum (see Figure 3 for RALC depiction), the distance of the stapling device from the pelvic floor was measured. This was calculated from an initial suture that was placed on the anterior rectum 5 cm from the pelvic floor, prior to any engagement of stapling devices. The surgeons were then asked to answer a questionnaire that rated various aspects of access, visibility, and ease of placement of the stapling devices. Figure 1 Illustration of coronal placement of the RALC. Figure 2 Illustration of sagittal placement of the RALC. Figure 3 Distal placement of the RALC on the rectum.

Both the CC and RALC 45-mm stapling devices were applied to the rectum, advanced as deep as possible in the pelvis in both coronal and sagittal configurations (see Figure 1 and 2), and engaged. All four surgeons operated independently on all twelve cadavers, for a total of 48 applications for each stapler per position. To increase objectivity, randomization was used to determine which stapling device was used before the other. After the staplers were placed as distally as possible on the rectum (see Figure 3 for RALC depiction), the distance of the stapling device from the pelvic floor was measured. This was calculated from an initial suture that was placed on the anterior rectum 5 cm from the pelvic floor, prior to any engagement of stapling devices. The surgeons were then asked to answer a questionnaire that rated various aspects of access, visibility, and ease of placement of the stapling devices. Figure 1 Illustration of coronal placement of the RALC. Figure 2 Illustration of sagittal placement of the RALC. Figure 3 Distal placement of the RALC on the rectum. Statistical Methods To account for potential investigator and cadaver differences, regression models were used for statistical analysis. Linear regression was performed for continuous outcomes and logistic regression for binary and ordinal outcomes. Other than the main variable of interest, i.e. the RALC versus CC device effect, investigator and cadaver indicator variables were included in the model as multicollinearity constraints allowed. A p value of < 0.05 was regarded as statistically significant.

comes and logistic regression for binary and ordinal outcomes. Other than the main variable of interest, i.e. the RALC versus CC device effect, investigator and cadaver indicator variables were included in the model as multicollinearity constraints allowed. A p value of < 0.05 was regarded as statistically significant. Results The median distance of the stapler from the pelvic floor in the coronal position for the RALC was 1.0 cm (range, 0-4.0 cm), compared to 2.0 cm (0-5.0 cm) for the CC, p = 0.003. In the sagittal position, the median distance from the pelvic floor was 1.6 cm (0-3.5 cm) for the RALC versus 3.3 cm (0-5.0 cm) for the CC, p < 0.0001. (Table 1) Table 1 Distance of stapler from pelvic floor Stapler placement Stapler used Distance (cm) from PF median (range) Distance (cm) from PF mean (std dev) Mean Effect P-value Coronal RALC CC 1.0 (0-4.0) 2.0 (0-5.0) 1.4 (1.3) 2.0 (1.2) -0.54 0.003 Sagittal RALC CC 1.6 (0-3.5) 3.3 (0-5.0) 1.5 (1.2) 3.2 (1.6) -1.70 < 0.0001 PF = pelvic floor RALC = Right Angle Linear Cutter CC = Curved Cutter Covariates included investigator and cadaver indicators.

Stapler placement Stapler used Distance (cm) from PF median (range) Distance (cm) from PF mean (std dev) Mean Effect P-value Coronal RALC CC 1.0 (0-4.0) 2.0 (0-5.0) 1.4 (1.3) 2.0 (1.2) -0.54 0.003 Sagittal RALC CC 1.6 (0-3.5) 3.3 (0-5.0) 1.5 (1.2) 3.2 (1.6) -1.70 < 0.0001 PF = pelvic floor RALC = Right Angle Linear Cutter CC = Curved Cutter Covariates included investigator and cadaver indicators. Applied in the coronal position, the RALC was superior to the CC in respect to: 1. the incidence of interference by the symphysis pubis, 2. the number of readjustments of the stapler, 3. ease of placement in the pelvis, 4. impediment of visibility, and 5. access in the pelvis. (Tables 2, 3, 4) Applied in the sagittal position, the RALC was superior to the CC in respect to: 1. the incidence of interference by the symphysis pubis, 2. the number of readjustments of the stapler, 3. ease of placement in the pelvis, 4. ease of placement around the colon and rectum, 5. containment of the entire rectum in the stapler when clamping, 6. ability to hold and retain tissue, 7. visibility rating, and 8. access in the pelvis. (Tables 2, 5, 6) The RALC scored consistently higher than the CC in all questions. Table 2 How many times did you have to readjust the stapler? Stapler placement Stapler used 0 1 2 3 or more Odds Ratio (for fewer times) P-value Coronal RALC CC 18 11 23 20 4 14 2 3 2.75 0.01 Sagittal RALC CC 19 11 21 17 6 9 1 11 3.53 0.002 RALC = Right Angle Linear Cutter CC = Curved Cutter Table 3 Stapler Performance: Yes/No questions, Coronal placement

Table 2 How many times did you have to readjust the stapler? Stapler placement Stapler used 0 1 2 3 or more Odds Ratio (for fewer times) P-value Coronal RALC CC 18 11 23 20 4 14 2 3 2.75 0.01 Sagittal RALC CC 19 11 21 17 6 9 1 11 3.53 0.002 RALC = Right Angle Linear Cutter CC = Curved Cutter Table 3 Stapler Performance: Yes/No questions, Coronal placement Question Stapler used YES NO Odds Ratio (for yes) P Value Was there interference of the pubic symphysis limiting placement of the device? RALC CC 1 (2%) 13 (27%) 47 35 0.054 0.006 Was there readjusting of the instrument for optimal placement? RALC CC 30 (63%) 37 (77%) 18 11 0.47 0.11 Did the instrument impede visibility RALC CC 1(2%) 23 (48%) 47 25 0.02 0.0003 Was the whole rectum contained in the device after clamping RALC CC 45 (94%) 46 (96%) 3 2 1.00 0.96 RALC = Right Angle Linear Cutter CC = Curved Cutter Table 4 Stapler Performance: Rating questions, Coronal placement Rating Stapler used Excellent Adequate Poor Odds Ratio (for poorer results) P Value Ease of placement in the pelvis RALC CC 39 (81%) 21 (45%) 9 (19%) 26 (55%) 0 0 0.18 0.0007 Ease of placement around the colon and rectum RALC CC 36 (75%) 28 (60%) 11 (23%) 17 (36%) 1 (2%) 2 (4%) 0.43 0.08 Ability to hold and retain tissue without slippage RALC CC 44 (94%) 45 (96%) 3 (6%) 1 (2%) 0 1 (2%) 1.49 0.67 Visibility RALC CC 38 (79%) 30 (63%) 10 (21%) 14 (29%) 0 4 (8%) 0.39 0.46 Pelvic access RALC CC 38 (79%) 18 (38%) 10 (21%) 28 (58%) 0 2 (4%) 0.14 < .0001 RALC = Right Angle Linear Cutter CC = Curved Cutter Table 5 Stapler Performance: Yes/No questions, Sagittal placement

Rating Stapler used Excellent Adequate Poor Odds Ratio (for poorer results) P Value Ease of placement in the pelvis RALC CC 39 (81%) 21 (45%) 9 (19%) 26 (55%) 0 0 0.18 0.0007 Ease of placement around the colon and rectum RALC CC 36 (75%) 28 (60%) 11 (23%) 17 (36%) 1 (2%) 2 (4%) 0.43 0.08 Ability to hold and retain tissue without slippage RALC CC 44 (94%) 45 (96%) 3 (6%) 1 (2%) 0 1 (2%) 1.49 0.67 Visibility RALC CC 38 (79%) 30 (63%) 10 (21%) 14 (29%) 0 4 (8%) 0.39 0.46 Pelvic access RALC CC 38 (79%) 18 (38%) 10 (21%) 28 (58%) 0 2 (4%) 0.14 < .0001 RALC = Right Angle Linear Cutter CC = Curved Cutter Table 5 Stapler Performance: Yes/No questions, Sagittal placement Question Stapler used YES NO Odds Ratio (for yes) P Value Was there interference of the pubic symphysis limiting placement of the device? RALC CC 2 (4%) 18 (38%) 46 30 0.064 0.0006 Was there readjusting of the instrument for optimal placement? RALC CC 29 (60%) 37 (77%) 19 11 0.43 0.07 Did the instrument impede visibility RALC CC 3 (6%) 6 (13%) 45 42 0.32 0.13 Was the whole rectum contained in the device after clamping RALC CC 45 (94%) 34 (71%) 3 14 18.08 0.007 RALC = Right Angle Linear Cutter CC = Curved Cutter Table 6 Stapler Performance: Rating questions, Sagittal placement

Question Stapler used YES NO Odds Ratio (for yes) P Value Was there interference of the pubic symphysis limiting placement of the device? RALC CC 2 (4%) 18 (38%) 46 30 0.064 0.0006 Was there readjusting of the instrument for optimal placement? RALC CC 29 (60%) 37 (77%) 19 11 0.43 0.07 Did the instrument impede visibility RALC CC 3 (6%) 6 (13%) 45 42 0.32 0.13 Was the whole rectum contained in the device after clamping RALC CC 45 (94%) 34 (71%) 3 14 18.08 0.007 RALC = Right Angle Linear Cutter CC = Curved Cutter Table 6 Stapler Performance: Rating questions, Sagittal placement Rating Stapler used Excellent Adequate Poor Odds Ratio (for poorer results) P Value Ease of placement in the pelvis RALC CC 38 (79%) 13 (27%) 10 (21%) 21 (44%) 0 14 (29%) 0.08 < .0001 Ease of placement around the colon and rectum RALC CC 36 (75%) 27 (56%) 11 (23%) 8 (17%) 1 (2%) 13 (27%) 0.29 0.007 Ability to hold and retain tissue without slippage RALC CC 46 (98%) 32 (68%) 0 2 (4%) 1 (2%) 13 (28%) 0.04 0.002 Visibility RALC CC 38 (79%) 21 (44%) 10 (21%) 14 (29%) 0 13 (27%) 0.16 0.0001 Pelvic access RALC CC 37 (77%) 15 (31%) 11 (23%) 20 (42%) 0 13 (27%) 0.10 < .0001 RALC = Right Angle Linear Cutter CC = Curved Cutter

Rating Stapler used Excellent Adequate Poor Odds Ratio (for poorer results) P Value Ease of placement in the pelvis RALC CC 38 (79%) 13 (27%) 10 (21%) 21 (44%) 0 14 (29%) 0.08 < .0001 Ease of placement around the colon and rectum RALC CC 36 (75%) 27 (56%) 11 (23%) 8 (17%) 1 (2%) 13 (27%) 0.29 0.007 Ability to hold and retain tissue without slippage RALC CC 46 (98%) 32 (68%) 0 2 (4%) 1 (2%) 13 (28%) 0.04 0.002 Visibility RALC CC 38 (79%) 21 (44%) 10 (21%) 14 (29%) 0 13 (27%) 0.16 0.0001 Pelvic access RALC CC 37 (77%) 15 (31%) 11 (23%) 20 (42%) 0 13 (27%) 0.10 < .0001 RALC = Right Angle Linear Cutter CC = Curved Cutter Placed coronally, the stapler rarely obstructed visibility while using the RALC (1/48 applications), while visual impediment was present in 23/48 (48%) using the CC, p = 0.0003. (Table 3) This involved anterior structures, including the prostate, in 95% of the case. Visual impediment was rare with both stapling devices placed sagittally. (Table 5) Superior visibility was experienced with the stapler in the coronal compared to sagittal position with the RALC in 67% of the applications (32/48), and in 96% (43/45) of the applications with the CC. (Table 7) Overall, the surgeons rated the best position for the RALC as coronal in 73% (35/48), and in 98% (44/45) with the CC. (Table 7) Table 7 Preference questions: Coronal and sagittal placement Question Stapler used CORONAL SAGITTAL Visually which is best? RALC CC 32 (67%) 43 (96%) 16 (33%) 2 (4%) What is the best placement of the device? RALC CC 35 (73%) 44 (98%) 13 (27%) 1 (2%) RALC = Right Angle Linear Cutter CC = Curved Cutter

Placed coronally, the stapler rarely obstructed visibility while using the RALC (1/48 applications), while visual impediment was present in 23/48 (48%) using the CC, p = 0.0003. (Table 3) This involved anterior structures, including the prostate, in 95% of the case. Visual impediment was rare with both stapling devices placed sagittally. (Table 5) Superior visibility was experienced with the stapler in the coronal compared to sagittal position with the RALC in 67% of the applications (32/48), and in 96% (43/45) of the applications with the CC. (Table 7) Overall, the surgeons rated the best position for the RALC as coronal in 73% (35/48), and in 98% (44/45) with the CC. (Table 7) Table 7 Preference questions: Coronal and sagittal placement Question Stapler used CORONAL SAGITTAL Visually which is best? RALC CC 32 (67%) 43 (96%) 16 (33%) 2 (4%) What is the best placement of the device? RALC CC 35 (73%) 44 (98%) 13 (27%) 1 (2%) RALC = Right Angle Linear Cutter CC = Curved Cutter The measurements of pelvic anatomic factors in the twelve cadavers are listed in Table 8. The mean distance from the pelvic floor to the stapling device by cadaver are depicted in Figure 4 and 5, for the coronal and sagittal positions, respectively. Cadavers #2, 3, 6, and 9 demonstrated the longest average stapler height from the pelvic floor, and were deemed the "difficult access" pelvises. Using regression analysis, a longer distance of the stapler from the pelvic floor was found to correlate with a shorter distance between the tip of the coccyx and pubic symphysis (p = 0.002).

6, and 9 demonstrated the longest average stapler height from the pelvic floor, and were deemed the "difficult access" pelvises. Using regression analysis, a longer distance of the stapler from the pelvic floor was found to correlate with a shorter distance between the tip of the coccyx and pubic symphysis (p = 0.002). Table 8 Measurement of anatomic distances Question Median* (range) Umbilicus to pubic symphysis 15 (11-18) Anterior superior iliac spine (ASIS) to ASIS 24 (20-33) Pubic symphysis to sacral promontory 12.25 (8-13.5) Pelvic floor to sacral promontory 12.5 (9-18) Tip of coccyx to pubis symphysis 11.8 (8-15) Pelvic inlet** 9.0 (7-13) Anal verge to pelvic floor 6 (5-10) Measurements are in cm *12 cadavers **Pelvic inlet measured as transverse diameter between right and left pelvic sidewall Figure 4 Mean distance from the pelvic floor to the device using coronal placement. Figure 5 Mean distance from the pelvic floor to the device using sagittal placement.

Question Median* (range) Umbilicus to pubic symphysis 15 (11-18) Anterior superior iliac spine (ASIS) to ASIS 24 (20-33) Pubic symphysis to sacral promontory 12.25 (8-13.5) Pelvic floor to sacral promontory 12.5 (9-18) Tip of coccyx to pubis symphysis 11.8 (8-15) Pelvic inlet** 9.0 (7-13) Anal verge to pelvic floor 6 (5-10) Measurements are in cm *12 cadavers **Pelvic inlet measured as transverse diameter between right and left pelvic sidewall Figure 4 Mean distance from the pelvic floor to the device using coronal placement. Figure 5 Mean distance from the pelvic floor to the device using sagittal placement. Discussion Despite recent advances in surgical stapling technology, the development of the ideal stapling device for the distal rectum has been slow. The bony confines of the pelvis, especially android pelvises, cause impediments to visibility and access in the distal rectum. The challenges are more pronounced if the dissected rectum is fatty and bulky, and in the presence of a large rectal tumor. Previous studies have revealed that the male gender, shorter distance of the tumor from the anal verge, and narrower pelvic dimensions are associated with longer operative times and worse outcomes in rectal surgery, including anastomotic leak and positive tumor margins [10-13].

in the presence of a large rectal tumor. Previous studies have revealed that the male gender, shorter distance of the tumor from the anal verge, and narrower pelvic dimensions are associated with longer operative times and worse outcomes in rectal surgery, including anastomotic leak and positive tumor margins [10-13]. An ideal rectal stapler should be easily advanced to the pelvic floor, have excellent ability to hold tissue, accept a wide range of tissue thickness, seal and divide the rectum reliably and safely, and occlude both the proximal and distal ends of the cut bowel. The challenge in designing such a stapler is that a low-profile stapler may have difficulty incorporating the entire rectum within its jaws, or may require several applications resulting in intersecting staple lines susceptible to ischemia. The rectal transection should ideally be perpendicular to the pelvic floor, but to design a stapler that deflects nearly 90 degrees limits its jaw strength and the deployment of a cutting knife.

e rectum within its jaws, or may require several applications resulting in intersecting staple lines susceptible to ischemia. The rectal transection should ideally be perpendicular to the pelvic floor, but to design a stapler that deflects nearly 90 degrees limits its jaw strength and the deployment of a cutting knife. In this cadaver study, we compared two stapling devices for the rectum that have similar mechanisms in that they have the ability to occlude both ends of the staple line. This is accomplished by the placement of four rows of staples, with a cutting knife in between the second and third rows. This improves on the traditional method of distal rectal division that involves occluding the bowel distal to the tumor with an atraumatic clamp (to limit spillage of stool and liberated tumor cells), stapling the rectum distal to the clamp, and using a long-handled knife deep in the pelvis to divide the bowel under restricted visual access.

aditional method of distal rectal division that involves occluding the bowel distal to the tumor with an atraumatic clamp (to limit spillage of stool and liberated tumor cells), stapling the rectum distal to the clamp, and using a long-handled knife deep in the pelvis to divide the bowel under restricted visual access. We found that the RALC, compared to the CC, could be placed deeper in the pelvis, has improved visualization, is obstructed less by surrounding tissues, and has pelvic access that was superior (rated "excellent" in 38/48 [79%] as opposed to 18/48 [38%] for the CC). This is likely explained by the smaller profile of the RALC 45-mm single-use load and the perpendicular orientation of the jaws in relation to the shaft of the handle. The RALC Single Use Reload's dimensions are 14.6 mm in width, 40.1 mm in height (closed), and 66.5 mm in length, compared with the CC's dimensions of 25.4 mm in width, 95.0 mm in height (closed), and 62.2 mm in length. We also found that the RALC had excellent and equal access to the deep pelvis whether placed coronally or sagittally, but that the CC was best used coronally. The sagittal placement of the CC resulted in poorer ratings for pelvic access compared with its coronal application. In addition, the CC placed in a sagittal configuration clamped the rectum 1.3 cm more proximally than coronally (3.3 versus 2.0 cm from the pelvic floor).

ttally, but that the CC was best used coronally. The sagittal placement of the CC resulted in poorer ratings for pelvic access compared with its coronal application. In addition, the CC placed in a sagittal configuration clamped the rectum 1.3 cm more proximally than coronally (3.3 versus 2.0 cm from the pelvic floor). By performing anatomic measurements in the pelvis, we hoped to evaluate what anatomic factor posed the greatest limitation to distal rectal stapling. A prior study by Gu, et al. that utilized magnetic resonance (MR) pelvimetry in rectal cancer surgery demonstrated that failure of a sphincter-preserving procedure was predicted by a shorter distance from the upper pubic bone to the sacrococcyx, longer distance of the sacrococcyx, and excessive curvature of the sacrum [14]. Another study correlating MR pelvimetry with difficulty of a laparoscopic proctectomy found that a large sagittal pelvic outlet coupled with a narrow transverse intertuberous distance predicted longer operative time [12]. In our current cadaver study, we found that the stapler distance from the pelvic floor was influenced significantly by the distance between the symphysis pubis and the tip of the coccyx. This anterior-posterior restriction of the pelvic outlet seemed to pose more difficulty than any anatomic measurement in the transverse direction or at the pelvic inlet. This knowledge, we hope, could help in the development of improved stapling devices in the future.

ween the symphysis pubis and the tip of the coccyx. This anterior-posterior restriction of the pelvic outlet seemed to pose more difficulty than any anatomic measurement in the transverse direction or at the pelvic inlet. This knowledge, we hope, could help in the development of improved stapling devices in the future. There are limitations of this study. First, there is contraction of the mesorectum in a cadaver compared to a live body, and this simplifies placement of the stapling device both around the rectum and into the pelvis. How these stapling devices would perform in the setting of a bulky mesorectum or large rectal tumor remains unanswered by this study. It is possible that the improved access in the deep pelvis with the RALC would be magnified in live surgery; however, it is also possible that the advantages seen in a cadaver model with the RALC will no longer be appreciated in a more technically difficult setting. Secondly, this study only investigated pelvic access and reach. The devices were clamped onto the distal rectum but not fired in order to maximize the number of applications per cadaver. Thus, one critical aspect of stapler performance-how the devices stapled-could not be assessed. Thirdly, investigator bias could not be eliminated, as it was impossible to blind the surgeon to the technique. Furthermore, this research project was funded by one company and not the other, and the participating surgeons received honoraria for their time. Lastly, it would have been interesting to study the influence of body mass index (BMI) on staple placement. This analysis was not performed, however, since the range of the cadavers' BMI (measured by the supplying company) was narrow; the median BMI (and range) of the cadavers was 24.7 (23.7-25.8) kg/m2.

or their time. Lastly, it would have been interesting to study the influence of body mass index (BMI) on staple placement. This analysis was not performed, however, since the range of the cadavers' BMI (measured by the supplying company) was narrow; the median BMI (and range) of the cadavers was 24.7 (23.7-25.8) kg/m2. Conclusion A significant improvement in access and visibility was seen with the RALC compared to the CC in cadavers, and the RALC was the preferred stapler in the majority of applications in the deep pelvis. This could provide important clinical benefit to both the patient and surgeon during difficult rectal surgery. Competing interests Toyooki Sonoda, MD: Received honoraria from Covidien for speaking events, lectures at laparoscopic educational courses, and hourly compensation for work on this cadaver project and manuscript preparation. There is no consultation agreement, stock or stock option ownership, patent-licensing agreements, or research support. There is no financial association with Ethicon. Juan-Carlos Verdeja, MD: No financial interests, consultant agreements, or speaker bureau agreements with Covidien and Ethicon. Received hourly compensation from Covidien for work on this cadaver project and manuscript review.

Competing interests Toyooki Sonoda, MD: Received honoraria from Covidien for speaking events, lectures at laparoscopic educational courses, and hourly compensation for work on this cadaver project and manuscript preparation. There is no consultation agreement, stock or stock option ownership, patent-licensing agreements, or research support. There is no financial association with Ethicon. Juan-Carlos Verdeja, MD: No financial interests, consultant agreements, or speaker bureau agreements with Covidien and Ethicon. Received hourly compensation from Covidien for work on this cadaver project and manuscript review. David E. Rivadeneira, MD: Received compensation as a consultant and has a consultant agreement with both Covidien and Ethicon. Received honoraria from Covidien for speaking events, laparoscopic educational courses, and hourly compensation for work on this cadaver project and manuscript review. There is no ownership of stocks, stock options, or equity interests, patent-licensing agreements, or research support. As a group of four participating surgeons, hourly compensation for work on the cadaver laboratory and manuscript preparation totaled $42,900. Authors' contributions TS was involved in the study design, data acquisition, analysis, interpretation of data, and drafting of the manuscript. JV and DR were involved in the study design, data acquisition, analysis, interpretation of data, and critical revision of the manuscript. All authors read and approved the final manuscript.

Authors' contributions TS was involved in the study design, data acquisition, analysis, interpretation of data, and drafting of the manuscript. JV and DR were involved in the study design, data acquisition, analysis, interpretation of data, and critical revision of the manuscript. All authors read and approved the final manuscript. Acknowledgements and funding The authors acknowledge Jessica Chowaniec, clinical research associate at Covidien, who was indispensible in the conception, design, acquisition of data, and analysis/interpretation of data. The authors thank Ping-Yu Liu, PhD for statistical analysis. The acquisition of cadavers, surgical instruments, and stapling instruments were funded by Covidien. All authors were reimbursed by Covidien for time spent conducting the research project, and for preparation and review of the manuscript.

Background Pain is an unpleasant sensory and emotional experience associated with actual or potential tissue damage, or described in terms of such damage [1]. Pain management, also called algiatry, employs an interdisciplinary approach for easing the suffering and improving the quality of life of those experiencing pain [2]. Because of the fear of pain in dental injections, some people avoid, cancel or do not appear for dental appointments [3]. Pain and anxiety control is one of the most important aspects in administration of local anesthetic in dental practice. Administration of local anesthetic produces pain and anxiety that may cause subsequent unfavorable behavior [4]. The levels of, and relationships between dental fear, general fears and phobias were studied by Berggren in 109 adult patients at a specialized dental fear clinic using two dental fear scales [5]. The results indicated that a large proportion of these dentally fearful individuals were prone to injection fear. The pain induced by infiltration of local anesthetic agents can be reduced in a number of complementary methods which include application of topical analgesics such as methocaine [6], suggestion [7], distraction techniques [8], counter irritation [9,10], varying the rates of infiltration [11], buffering the local anesthesia [12-15], reduced speed of injection [16,17] and use of vibration [18-23]. Melzack and Schecter [24] showed that itch can be reduced by vibration of stimulated area. They concluded that their results may be attributed to physiological activities occurring at the early stages of information transmission. Vibratory stimulation is a potential method for the treatment of pain. It is one of several non-pharmacological techniques used to reduce pain [25]. The effects of vibration on pain have been reported in both clinical [21,26] and experimental settings [18,27]. Vibration activates both superficial and deeply located receptors [18,22,28]. Lundeberg et al proved reduction of pain during vibratory stimulation in patients suffering acute or chronic musculoskeletal pain of different origin.

tion on pain have been reported in both clinical [21,26] and experimental settings [18,27]. Vibration activates both superficial and deeply located receptors [18,22,28]. Lundeberg et al proved reduction of pain during vibratory stimulation in patients suffering acute or chronic musculoskeletal pain of different origin. They noticed that sixty-nine per cent of the patients reported a reduction of pain during vibratory stimulation [19,20,27]. Nanitsos et al investigated the effect of vibration on pain during local anesthesia injections. Their results indicated that compared to no vibration-stimulus injections, injections with vibration resulted in less pain and lower pain rating by studied patients [19]. Based on gate control theory [29,30], mechanisms of pain relief induced by vibration are elaborated by several workers suggesting pain can be reduced by simultaneous activation of nerve fibers that conduct non-noxious stimuli [1,19,20]. In this regard Longe et al [31] and Aminabadi [9] indicated that counter stimulation reduces pain perception. They concluded that when vibration is applied as a counter stimulation to an anesthetic injection, it will reach the brain before the pain sensation does. The brain can perceive only one sensation at a time; therefore, the sensation that arrives at the brain first is the one that will be felt. Syringe Micro Vibrator is a new design being introduced for the first time in the field of Dentistry. As a vibration stimulus, structural constituents, analogy and the role of SMV to alleviate patient pain and anxiety during dental anesthesia injection would be discussed hereafter.

n first is the one that will be felt. Syringe Micro Vibrator is a new design being introduced for the first time in the field of Dentistry. As a vibration stimulus, structural constituents, analogy and the role of SMV to alleviate patient pain and anxiety during dental anesthesia injection would be discussed hereafter. Presentation of the hypothesis SMV was designed to provide feasibility to alleviate injection pain and anxiety in clinical practice. Upon mounting on a conventional dental anesthesia injection syringe, its motor is switched on and the clinician then uses normal injection technique to administer the anesthetic. Its parallel mounting on the syringe allows clinician to rotate the syringe while in the mouth, if necessary. Its main structural parts consist of a) stainless steel shell bearing four flexible attachment arms, b) eccentrically weighted plate and motor, and c) two button batteries. SMV is registered as an invention in the field of dentistry and received Iran National Patent number of 63765. Schematic mounting position of SMV on injection syringe is indicated in Figure 1.

less steel shell bearing four flexible attachment arms, b) eccentrically weighted plate and motor, and c) two button batteries. SMV is registered as an invention in the field of dentistry and received Iran National Patent number of 63765. Schematic mounting position of SMV on injection syringe is indicated in Figure 1. Figure 1 Structural components of Syringe Micro Vibrator (SMV) and its mounting position on dental syringe barrel. A) Posterior-anterior view of SMV, structural components consist of: a) stainless steel shell containing motor and eccentrically weighted plate, b) power switch, c) stainless steel cap, d) four flexible attachment arms for firm attachment and e) shell concavity for well adaptation on syringe barrel. B) Anterior-posterior view of SMV and stainless steel cap bearing button cell batteries, C) SMV is applicable to most standard conventional syringes which compensates minute variations of different barrel diameters through four flexible grasping positioning arms and shell concavity, D) SMV mounted on syringe barrel and E) Lateral syringe view indicates that mounting of SMV causes no restriction for the replacement of cartridge.

o most standard conventional syringes which compensates minute variations of different barrel diameters through four flexible grasping positioning arms and shell concavity, D) SMV mounted on syringe barrel and E) Lateral syringe view indicates that mounting of SMV causes no restriction for the replacement of cartridge. Testing the hypothesis Micro vibration of injection needle with ultra high frequency and ultra low altitude will alleviate the pain and anxiety during intraoral injections. SMV bears an off-set rotating micro vibration motor and can be easily mounted on any standard dental syringe. Accordingly, "Pain gate" would be shut by stimulating nerves providing reduction in patient pain and discomfort during injection period. SMV makes nerve endings sense micro vibrations at the very early stage, so will remarkably reduce the pain transmission. Implication of SMV would have advantages of: 1) From the aspect of the patient pain management, SMV contributes both physiologically (based on Gate Control Theory of pain) and psychologically (based on the device function as will be explained by dentist to the patient as a modern pain reducing technology), 2) from the aspect of clinician, SMV motor provides vibrations with ultra high frequency to alleviate pain, but since it has ultra low vibration altitude, it has no adverse effect on the clinician dexterity and accuracy during injection and it does not interfere with pin point localization of injection site, 3) it alleviates two types of pain injection including both needle insertion pain and balloon effect due to forceful penetration of anesthetic into the surrounding tissue.

no adverse effect on the clinician dexterity and accuracy during injection and it does not interfere with pin point localization of injection site, 3) it alleviates two types of pain injection including both needle insertion pain and balloon effect due to forceful penetration of anesthetic into the surrounding tissue. The micro vibration slowly reduces such balloon and enhances tissue infiltration of injected anesthesia, 4) it is easy to use and does not provide any inconvenience for the clinician during injection operation due to a- low weight that does not affect the accuracy of clinician, b- small size that keeps well visibility, c- battery powered and lack of wire or hose attachment, 5) firm grasp by four flexible attachment arms it provides: a- convey of efficient vibration to syringe barrel and consequently to needle, b- applicable to all standard conventional syringes and compensates minute variations of different barrel diameters, c- no screw or spare appliance needed for its mounting to or removal from the syringe barrel, so its application is fast and easy, d- it can be attached anywhere along the syringe barrel while it does not cover or mask the cartridge, 6) no need to replace the existing syringes or purchase further spare parts, 7) it has a detachable motor, a stainless steel shell and four attachment arms which are autoclavable and are the only parts in contact with patient during injection. The motor should not be heat sterilized but for disinfestations, it can be easily removed from its shell socket before autoclaving or can be wiped off with a surface disinfestant and 8) use of SMV would save time since it eliminates the period needed for application and onset of topical anesthetic. It also shortens period of anesthesia onset with increase of anesthetic diffusion pace to surrounding tissues since the ultra vibration enhances tissue infiltration of injected anesthetic.

sinfestant and 8) use of SMV would save time since it eliminates the period needed for application and onset of topical anesthetic. It also shortens period of anesthesia onset with increase of anesthetic diffusion pace to surrounding tissues since the ultra vibration enhances tissue infiltration of injected anesthetic. Simplicity of application Implication of SMV requires no topical anesthetic or special technique. SMV would weigh approximately 30-50 grams and would be easily mounted on dental injection syringes. The SMV package would bear motor assembly, autoclavable stainless steel shell bearing flexible attachment arms, two button batteries, a recharging unit and power cord. SMV will bring more comfort for both patient and the physician during the process of injection. This device by creating micro vibration would be effective in reducing the pain and anxiety confronted with most types of intraoral injections as palatal, mandibular block, intraligamental and local infiltration.

power cord. SMV will bring more comfort for both patient and the physician during the process of injection. This device by creating micro vibration would be effective in reducing the pain and anxiety confronted with most types of intraoral injections as palatal, mandibular block, intraligamental and local infiltration. Discussion Most patients feel physically and psychologically uncomfortable about penetration of injecting needle into their oral tissues. Some of them do not convey this feeling to the clinician, so the number of patients fearful of the dental pain experience and feeling discomfort is more than what is seen in clinic, however, in many patients fear of injection contributes significantly to postponement of dental treatment however, less painful injection experience with SMV lowers the future fear of injection in such patients. The "Gate-control" theory [29,30] reveals that using appropriate pressure or vibration can close the neural "gate" so that the central perception of itch and pain is reduced. In other words, the "Pain gate" can be shut by stimulating nerves responsible for carrying the signals which enable the relief of pain through massage techniques, rubbing, pressure, ice packs, acupuncture, electrical analgesia and the application of vibration. Examples of less pain experiences based on Gate theory may be seen in: a- Incision in bare hands while playing in snow feels less pain than normal, b- use of simultaneous cold and vibration in ladies epilating devices as accessory stimuli reduces transmission of main stimulus by local nerve endings in hair removal leading to reduced prick and c- some dentists have developed pain reducing techniques like shaking the syringe in their grasp while dispensing the anesthetic [25].

imultaneous cold and vibration in ladies epilating devices as accessory stimuli reduces transmission of main stimulus by local nerve endings in hair removal leading to reduced prick and c- some dentists have developed pain reducing techniques like shaking the syringe in their grasp while dispensing the anesthetic [25]. Comparative study A similar device reported in the literature, VibraJect®, has controversial performance. Blair [32] recommended the use of VibraJect® for painless injection. In contrast, Yoshikawa et al [33] found no significant pain reduction when VibraJect® was applied with a conventional dental syringe. Saijo et al [34] evaluated the effectiveness of VibraJect® in combination with an electrical injection device. Injections were given into the alveolar mucosa adjacent to the root apex of the maxillary lateral incisor in 10 volunteers. VibraJect® was randomly applied to either the left or right side of the injection. They found no statistically supports use of VibraJect® but expressed that it offers a simple and easy-to-use solution that can anesthetize patients quickly in a more comfortable manner. They also point out that VibraJect® enables a less painful palatal injection because it delivers small amounts of anesthetic solution over a period of time. Another supporting result was statistically performed by Purray et al [35] at Queens University with conclusion that the vibrating syringe attachment resulted in reduced pain levels on receiving intraoral injections. The study performed on 400 patients and showed that VibraJect® statistically reduced the amount of pain score from 4.6 to 1.7 which has never been statistically achieved before. Accordingly it is conclusive that VibraJect® effectiveness is approved, however as indicated in Table 1 SMV has improvements in several structural features including vibration mode bringing enhanced effectiveness to the method of using vibration in dental anesthesia injection. In the table, technical features, performance and a brief summary for specification differences between SMV and VibraJect® are indicated correspondingly. At last, it should be pointed out that SMV is not only a useful accessory device for ordinary patients, but also more useful for pediatric patients and those who have a phobia of intraoral injection or pain. The author believes SMV will bring more comfort for both patient and the physician during the process of injection.

last, it should be pointed out that SMV is not only a useful accessory device for ordinary patients, but also more useful for pediatric patients and those who have a phobia of intraoral injection or pain. The author believes SMV will bring more comfort for both patient and the physician during the process of injection. Table 1 Comparative study indicating specification differences between Syringe Micro Vibrator (SMV) and VibraJect®

last, it should be pointed out that SMV is not only a useful accessory device for ordinary patients, but also more useful for pediatric patients and those who have a phobia of intraoral injection or pain. The author believes SMV will bring more comfort for both patient and the physician during the process of injection. Table 1 Comparative study indicating specification differences between Syringe Micro Vibrator (SMV) and VibraJect® Specifications VibraJect® SMV Advantage of SMV Attachment interface Clip Four flexible grasping positioning arms and concave shell contact surface Firm grasping, efficient vibration conveyance, causes no restriction for the replacement of cartridge, applicable to all standard conventional syringes which compensates minute variations of different barrel diameters Mounting angle on syringe barrel Angular Parallel Most efficient contact, least clinician vision masking, least patient discomfort, allowing the clinician to rotate the syringe while in the mouth, if necessary Vibration device Eccentrically weighted shaft Eccentrically weighted plate Yield of uniform micro vibration, more precise localization of injection point Vibration mode High frequency Ultra high frequency and ultra low altitude Efficiently enhance patient pain reduction, and ease of clinician maneuver and accuracy during injection Accordingly, what expressed in this article about using SMV is based upon potential usefulness of vibratory stimulation for pain since vibration would increase the threshold for dental injection pain as indicated in earlier findings of other researchers [18,23,36]. The author expresses that although SMV is designed to alleviate the injection pain during anesthesia injections, there may be no significant difference in pain perception in patients with a higher pain tolerance. Since pain tolerance cannot be estimated ahead of injection [37], it is recommended to use SMV for most patients.

thor expresses that although SMV is designed to alleviate the injection pain during anesthesia injections, there may be no significant difference in pain perception in patients with a higher pain tolerance. Since pain tolerance cannot be estimated ahead of injection [37], it is recommended to use SMV for most patients. Implications of the hypothesis SMV is a useful accessory device adaptable for dental injection syringe and conventional intra muscular injections to alleviate pain and stress of injection. Abbreviations SMV: Syringe Micro Vibrator. Competing interests The authors declares that he has no competing interests. Authors' contributions AHSB carried out the entire design of the study and draft the manuscript. He read and approved the final manuscript. Acknowledgements I like to express my deepest thanks for help, scientific advice and support to: Professor M. Yaghmaei, Oral and Maxillofacial Surgeon and Head of the Oral and Maxillofacial Surgery Department; Dr A. Khojasteh, Oral and Maxillofacial Surgeon and Head of Exceptional Talent Center of the College of Dentistry; Dr M. Nouri, Orthodontist and Head of Communication and Information Technology Center; Dr F. Poordanesh, Oral and Maxillofacial Surgeon; Dr O. Diyanat, Endodontist; Dental School, Shahid Beheshti University of Medical Sciences, Tehran, Iran.

Introduction The typology of coagulation disorders is extremely heterogeneous. Today, we frequently treat patients taking oral anticoagulant therapy (OAT), a prophylaxis against the occurrence of thromboembolic events [1,2]. Recent studies on the Italian population indicate an estimated 250.000 patients taking OAT to date. Indications for use of this pharmacological prophylaxis [3,4] are confirmed by cardiac diseases, vasculopathies with thromboembolic etiology, a few enzymatic-receptorial anomalies, as well as qualitative/quantitative alterations of some cellular hematic components [5,6]. Today, we deal with patients undergoing dicoumarolics therapy, which interfere with the metabolism of vitamin K and therefore with the synthesis of coagulation factors-- II, VII, IX, and X [7,8]. These patients often take the most frequent OAT, such as: • acenocoumarol: intermediate action, which begins within 24-48 h and ends within 2-4 days. • warfarin: long and delayed action from 3 to 7 days. Warfarin has a more permanent and long-lasting effect, because of its longer biological half-life (about 36 h). The aim of the present study was to apply a protocol, which could provide a safe intra- and postoperative management of patients on OAT. Materials and methods In the present multicentric study, we focused on a cohort of 193 patients (119 men aged between 46 and 82 years and 74 women aged between 54 and 76 years) undergoing OAT for more than 5 years.

The aim of the present study was to apply a protocol, which could provide a safe intra- and postoperative management of patients on OAT. Materials and methods In the present multicentric study, we focused on a cohort of 193 patients (119 men aged between 46 and 82 years and 74 women aged between 54 and 76 years) undergoing OAT for more than 5 years. This study is carried out in compliance with the Helsinki Declaration; it is divided into a preoperative phase, an intraoperative phase, and a postoperative phase including a 2-months follow-up. a) Preoperative Phase Medical history was carried out through specific questionnaires for the diagnosis of hemorrhagic disorders [9]. We asked the patients if they ever had hemorrhagic diathesis, its type (spontaneous or post-traumatic/post-surgical), and outcome. It is important to ask the patients if they drink alcohol or take unnecessary substances [10]. A careful assessment of the ongoing pharmacological therapy should be carried out. Some currently used drugs (such as penicillin and cephalosporin, trimetoprim, gentamicin, rifampicin, analgesics and non-steroidal anti-inflammatory drugs (NSAIDS), cardiovascular and anti-diabetic drugs, diuretics, platelet anti-aggregants, heparins, gold salts, and anti H2) can produce thrombocytopenia [5]. • On the day of the surgery, the dental surgeon has to investigate the patient's hemocoagulative structure through a series of hematoclinical examinations.

a) Preoperative Phase Medical history was carried out through specific questionnaires for the diagnosis of hemorrhagic disorders [9]. We asked the patients if they ever had hemorrhagic diathesis, its type (spontaneous or post-traumatic/post-surgical), and outcome. It is important to ask the patients if they drink alcohol or take unnecessary substances [10]. A careful assessment of the ongoing pharmacological therapy should be carried out. Some currently used drugs (such as penicillin and cephalosporin, trimetoprim, gentamicin, rifampicin, analgesics and non-steroidal anti-inflammatory drugs (NSAIDS), cardiovascular and anti-diabetic drugs, diuretics, platelet anti-aggregants, heparins, gold salts, and anti H2) can produce thrombocytopenia [5]. • On the day of the surgery, the dental surgeon has to investigate the patient's hemocoagulative structure through a series of hematoclinical examinations. Counseling between the dental surgeon, patient, and hygienist is necessary to make patients aware of the importance of dental hygiene in preventing dental and periodontal pathologies, which could later require more destructive and invasive therapies [11]. In addition to this, inflammatory or degenerative pathologies of the oral cavity play a crucial role in the genesis of bacterial endocarditis, a non-rare phenomenon among chronic valvulopathy patients and among those with valvular prostheses [12]. Bacterial Endocarditis (BE) is a potentially lethal disease, even if patients are on antibiotics. Mortality can reach up to 30% of cases.

In addition to this, inflammatory or degenerative pathologies of the oral cavity play a crucial role in the genesis of bacterial endocarditis, a non-rare phenomenon among chronic valvulopathy patients and among those with valvular prostheses [12]. Bacterial Endocarditis (BE) is a potentially lethal disease, even if patients are on antibiotics. Mortality can reach up to 30% of cases. The British Society Antimicrobial Chemotherapy (BSAC) has revised the cardiac table of patients risking BE. According to these new indications, there are 3 types of patients risking BE: 1) patients with previous BE; 2) patients with cardiac valve replacement; and 3) patients who underwent surgical shunt creation, especially in the cardiopulmonary area. These new guidelines are no longer ambiguous about performing an antibiotic prophylaxis (table 1) apart from the routinely removal of infections during the days preceding the oral surgery. Table 1 Pattern of Antibiotic Prophylaxis DRUGS DOSAGE FIRST CHOICE Amoxicillin Adults: 2 gm per os, 1 hour before surgery 1,5 gm per os, 6 hours after initial dose Children: 50 mg/Kg per os, 1 hour before surgery PATIENTS UNABLE TO TAKE DRUGS PER OS Ampicillin Adults: 2 gm i.m., before surgery Children: 50 mg/Kg ev or im, 30 min before surgery IN CASE OF ALLERGY TO PENICILLIN

DRUGS DOSAGE FIRST CHOICE Amoxicillin Adults: 2 gm per os, 1 hour before surgery 1,5 gm per os, 6 hours after initial dose Children: 50 mg/Kg per os, 1 hour before surgery PATIENTS UNABLE TO TAKE DRUGS PER OS Ampicillin Adults: 2 gm i.m., before surgery Children: 50 mg/Kg ev or im, 30 min before surgery IN CASE OF ALLERGY TO PENICILLIN Clindamycin Cephalexin* or Cephadroxil* Azithromycin or Clarithromycin Adults: 600 mg per os, 1 hour before surgery Children: 20 mg/Kg per os, 1 hour before surgery Adults: 2 gr per os, 1 hour before surgery Children: 50 mg/Kg per os, 1 hour before surgery Adults: 500 mg × os 1 hour before surgery Children: 15 mg/Kg per os, 1 hour before surgery ALLERGIC PATIENTS UNABLE TO TAKE DRUGS PER OS Clindamycin Cephazolin* Adults: 600 mg ev, 30 min before surgery Children: 20 mg/Kg ev, 30 min before surgery Adults: 1 gr im, 30 min before surgery Children: 25 mg/Kg im or ev, 30 min before surgery *provided that there are no serious allergic reactions to penicillin The aim of antibiotic prophylaxis is to have an effective serum concentration of the active principle during the whole perioperative period. The antibiotic therapy should be given shortly after the beginning of the treatment (1-2 h) and within 6-8 h from surgery; a longer therapy is justified only in case of infected tissues or delayed wound healing.

The aim of antibiotic prophylaxis is to have an effective serum concentration of the active principle during the whole perioperative period. The antibiotic therapy should be given shortly after the beginning of the treatment (1-2 h) and within 6-8 h from surgery; a longer therapy is justified only in case of infected tissues or delayed wound healing. b) Intra-operative phase Fifteen days before surgery, periodontal preparation of patients through tartar ablation by means of an ultrasonic equipment. To avoid periodontal hemorrhages, it is necessary to invite patients (before ablation) to do oral rinses for 1 min, with 10 ml of 4.8% solution of tranexamic acid. Moreover, it is important to prescribe to patients oral rinses with chlorhexidine digluconate 0.12%, 2 times a day for 14 days before surgery. The day of surgery, it is necessary to assess the hemocoagulative parameters by taking a blood sample, which will allow to evaluate PT, PTT, and especially the INR measured 24 h before surgery. In the last 3 years, we treated 193 patients on OAT with a standardized management and a 2-months follow-up. Although we evaluated every single clinical case in its uniqueness, we wanted to apply a protocol providing for a safe intra- and postoperative management of patients receiving OAT.

The day of surgery, it is necessary to assess the hemocoagulative parameters by taking a blood sample, which will allow to evaluate PT, PTT, and especially the INR measured 24 h before surgery. In the last 3 years, we treated 193 patients on OAT with a standardized management and a 2-months follow-up. Although we evaluated every single clinical case in its uniqueness, we wanted to apply a protocol providing for a safe intra- and postoperative management of patients receiving OAT. There are not any contraindications to the use of local anesthetics: mepivacaine, lidocaine, or articaine are recommended (articaine is needed when an intense ischemia is required and when surgery lasts more than 1 h, thanks to its longer action than mepivacaine and to its good effectiveness in controlling postoperative pain). The literature does not recommend the use of vasoconstrictors as they could cause tachycardia episodes [13]; nevertheless, we think that a safe use of a vasoconstrictor is possible by means of an aspirating syringe (carpule) and a good anesthesia technique, except for patients who recently had a myocardial infarction or coronary artery bypass surgery (≤ 8 months). This will bring 2 advantages: 1) The vasoconstrictor will guarantee an optimal zonal ischemia and a longer anesthetic effect, which is very useful when treating hemorrhagic patients [14]; 2) The vasoconstrictor (red phials) allows for an effective anesthesia, using a smaller quantity of anesthetic compared to green phials (anesthetic without vasoconstrictor).

1) The vasoconstrictor will guarantee an optimal zonal ischemia and a longer anesthetic effect, which is very useful when treating hemorrhagic patients [14]; 2) The vasoconstrictor (red phials) allows for an effective anesthesia, using a smaller quantity of anesthetic compared to green phials (anesthetic without vasoconstrictor). Among the hemostatic agents that a dental surgeon can use, there are ferric sulfate, oxidized cellulose, gelatin sponges, and anti-fibrinolytics. Oxidized regenerated cellulose [15] (Surgicel, Tabotamp) consists of a sterile gauze and oxidized cellulose. First, it acts as a mechanical barrier and then as a viscous mass, which works as an artificial clot. Antifibrinolytics (Tranex, Ugurol) perform a good local hemostatic action. Local use of tranexamic acid to prevent and treat hemorrhage in the oral cavity is recommended. The use of this active principle encourages clot formation and reduces the bleeding time [16]. Before surgery, we perform a standardized preoperative procedure. We recommend a pre-anesthesia by administering benzodiazepines (15 gtts, 20 min. before surgery). As a matter of fact, premedication allows for greater patient compliance during surgery, and the adrenalin plateau caused by intraoperative stress will be avoided. We suggest to our patients oral rinses with chlorhexidine digluconate 0.2% solution [17]. A careful disinfection of perioral and intraoral mucus-dental tissues by using an iodized solution is also necessary. Rinsing the operating field with tranexamic acid solution (4.8%) allows to control hemostasis since initial dieresis.

We suggest to our patients oral rinses with chlorhexidine digluconate 0.2% solution [17]. A careful disinfection of perioral and intraoral mucus-dental tissues by using an iodized solution is also necessary. Rinsing the operating field with tranexamic acid solution (4.8%) allows to control hemostasis since initial dieresis. Performing a venous access allows to operate quickly in case of hemorrhagic emergency. In this case, patients could have hypotensive phenomena, which involve difficult identification of the venous access, because vessels could be partially collapsed. In major oral surgery, 2 g of tranexamic acid are administered intravenously (IV) 1 h before surgery (divided into 4 × 0.5 g doses every 15 min) and after preparing the venous access [16]. The cardiac activity and the pressure status have to be steadily monitored using specific equipment (CardiocapII, Datex). The surgical procedure is performed following a standardized controlled protocol.

In major oral surgery, 2 g of tranexamic acid are administered intravenously (IV) 1 h before surgery (divided into 4 × 0.5 g doses every 15 min) and after preparing the venous access [16]. The cardiac activity and the pressure status have to be steadily monitored using specific equipment (CardiocapII, Datex). The surgical procedure is performed following a standardized controlled protocol. A plexus, intraligamentous, or locoregional anesthesia is performed using a slow infusion of mepivacaine 3%; we have described above what we think about the use of vasoconstrictors, which will bring more benefits than objective risks. Following a periotome syndesmotomy, a luxation of the dental-alveolar gomphosis will be performed. We suggest to reduce to a minimum the use of surgical elevators, because the moments of force of these instruments cause severe trauma to soft tissues along with microlesions to the alveolar ridge. After avulsion, manual reduction of the "greenstick" fracture [2] of the post-extractive alveolus will be performed by two-finger compression of the vestibular and lingual-palatal alveolus portions, in order to reduce the fracture occurred during avulsion. Accurate alveolar bone cleaning is to perform: by means of an alveolar spoon, we will remove all the possible granulation tissues representing the primum movens in the residual cyst formation, namely post-extractive alveolites [1,2].

A plexus, intraligamentous, or locoregional anesthesia is performed using a slow infusion of mepivacaine 3%; we have described above what we think about the use of vasoconstrictors, which will bring more benefits than objective risks. Following a periotome syndesmotomy, a luxation of the dental-alveolar gomphosis will be performed. We suggest to reduce to a minimum the use of surgical elevators, because the moments of force of these instruments cause severe trauma to soft tissues along with microlesions to the alveolar ridge. After avulsion, manual reduction of the "greenstick" fracture [2] of the post-extractive alveolus will be performed by two-finger compression of the vestibular and lingual-palatal alveolus portions, in order to reduce the fracture occurred during avulsion. Accurate alveolar bone cleaning is to perform: by means of an alveolar spoon, we will remove all the possible granulation tissues representing the primum movens in the residual cyst formation, namely post-extractive alveolites [1,2]. A bone rongeur or a bone mill with a turbine motor (continuous irrigation) allows to eliminate the bone roughness, which could damage the soft tissues of the antagonist arch. We compressed the post-extraction alveolus with oxidized regenerated cellulose.

Accurate alveolar bone cleaning is to perform: by means of an alveolar spoon, we will remove all the possible granulation tissues representing the primum movens in the residual cyst formation, namely post-extractive alveolites [1,2]. A bone rongeur or a bone mill with a turbine motor (continuous irrigation) allows to eliminate the bone roughness, which could damage the soft tissues of the antagonist arch. We compressed the post-extraction alveolus with oxidized regenerated cellulose. we suggest an absorbable suture (Vicryl 3/0 or 4/0 to avoid trauma of suture removal) using a "simple running stitch," with a further safety suture using a "simple interrupted stitch"; then, there will be a 30 minute compression of the surgical area, followed by sterile gauze application filled with tranexamic acid. If bleeding stops about 30 min later, it is the clinical demonstration of the technical success of surgery. c) Post-operative phase Patients should be discharged at least 60 minutes after surgery (in order to verify the absence of hemorrhagy); we advise a cold and liquid diet during the next 3 days and a tepid and semiliquid diet during the following 3 days. Patients should perform oral rinses for 2 minutes with 10 ml of tranexamic acid, 4 times a day for 7 days [16]. After rinsing, patients should not drink or eat for about an hour.

c) Post-operative phase Patients should be discharged at least 60 minutes after surgery (in order to verify the absence of hemorrhagy); we advise a cold and liquid diet during the next 3 days and a tepid and semiliquid diet during the following 3 days. Patients should perform oral rinses for 2 minutes with 10 ml of tranexamic acid, 4 times a day for 7 days [16]. After rinsing, patients should not drink or eat for about an hour. During the first 30 days after surgery, we suggest patients to rinse their mouth every 12 h with chlorhexidine digluconate 0.12% solution; anti-inflammatory therapy should be prescribed only in case of complex surgeries or if patients experience postoperative pain: 500 mg of paracetamol every 8 hours will be a precaution against hemorrhagic problems caused by the NSAIDs. The highest risk of postoperative hemorrhagic diathesis is during the first 12 h after the surgery; our follow-ups allow us to exclude hemorrhagic relapse from the 6th day after the surgery. In case of postoperative hemorrhage, a buffer imbibed in tranexamic acid should be applied together with compression of the area [18]; if bleeding persists, up to 2 g of tranexamic acid divided into 4 doses of 0.5 g has to be administered. In incoercible cases, it is indispensable to consult the hematologist in order to reduce OAT and develop a strategy to stop hemorrhagic diathesis [19,20].

hould be applied together with compression of the area [18]; if bleeding persists, up to 2 g of tranexamic acid divided into 4 doses of 0.5 g has to be administered. In incoercible cases, it is indispensable to consult the hematologist in order to reduce OAT and develop a strategy to stop hemorrhagic diathesis [19,20]. We recommend the use of absorbable sutures (Vicryl 3/0 or 4/0) to avoid trauma caused by suture removal. Non-absorbable sutures are particularly indicated for areas presenting aponeurosis, mimic muscles, frenula, etc. Removal should nonetheless be performed on the 8th day and in the less traumatic way. In case of slight hemorrhages, it is enough to apply a tampon of tranexamic acid and repeat oral rinses for 1 or 2 days. Regular follow-ups should be planned (7, 15, 30, and 60 days after surgery). They include intra- and extraoral examinations, a measurement of symptoms of pain by visual analogue scale (VAS), the motivation of patients to oral hygiene, and the possibility of further dental treatments.

We recommend the use of absorbable sutures (Vicryl 3/0 or 4/0) to avoid trauma caused by suture removal. Non-absorbable sutures are particularly indicated for areas presenting aponeurosis, mimic muscles, frenula, etc. Removal should nonetheless be performed on the 8th day and in the less traumatic way. In case of slight hemorrhages, it is enough to apply a tampon of tranexamic acid and repeat oral rinses for 1 or 2 days. Regular follow-ups should be planned (7, 15, 30, and 60 days after surgery). They include intra- and extraoral examinations, a measurement of symptoms of pain by visual analogue scale (VAS), the motivation of patients to oral hygiene, and the possibility of further dental treatments. Results Out of 193 treated patients, just 2 of them (only 1.03% of cases) had problems related to surgery: the 1st one had a slight hematoma in the area of labial commissure, due to the trauma caused by the Farabeuf retractor, but it regressed in a few days. The 2nd one had pains and did not present clot formation in the post-extractive site. Intraoral examination confirmed the diagnosis of post-extractive alveolitis, while the forthcoming pathological medical history indicated that the patient did not follow hygiene rules (smoker) in the immediate postoperative period; the alveolus was cleaned and the symptoms significantly reduced in the follow-up period. The results of this study show a protocol which allows a safe treatment of patients on OAT.

Results Out of 193 treated patients, just 2 of them (only 1.03% of cases) had problems related to surgery: the 1st one had a slight hematoma in the area of labial commissure, due to the trauma caused by the Farabeuf retractor, but it regressed in a few days. The 2nd one had pains and did not present clot formation in the post-extractive site. Intraoral examination confirmed the diagnosis of post-extractive alveolitis, while the forthcoming pathological medical history indicated that the patient did not follow hygiene rules (smoker) in the immediate postoperative period; the alveolus was cleaned and the symptoms significantly reduced in the follow-up period. The results of this study show a protocol which allows a safe treatment of patients on OAT. The problems that may principally arise in this type of surgical patients are related to a deficient or excessive clotting, we may also have problems of excessive susceptibility to bacterial infections. With this protocol we have treated 193 patients and we have had none of these problems, except in only 2 cases. These data are significant and show that the management of these patients is possible if we follow a proper protocol such as this one here described.

blems of excessive susceptibility to bacterial infections. With this protocol we have treated 193 patients and we have had none of these problems, except in only 2 cases. These data are significant and show that the management of these patients is possible if we follow a proper protocol such as this one here described. Discussion Dental procedures represent a particularly common intervention for patients receiving anticoagulant therapy. The most recent literature indicated that in most cases, no change in the intensity of anticoagulation is needed. As we can highlight in the literature, there are no well-documented cases of severe bleeding in this setting, but there are many documented cases of embolic events in patients whose warfarin therapy was discontinued for dental treatment. If there is a need to control local bleeding, tranexamic acid administration has been used successfully without interrupting anticoagulant therapy [18,19].