Journal of Investigative Surgery, 28, 167–172, 2015 C 2015 Informa Healthcare USA, Inc. Copyright ISSN: 0894-1939 print / 1521-0553 online DOI: 10.3109/08941939.2014.995244
ORIGINAL ARTICLE
Effect of HumiraR on Intestinal Anastomotic Response in Rabbits Karen Strebel, MD,1 Susanne R. H. Nielsen, MD,1 Matteo Biagini, MD,2 Niels Qvist, DMSci3
J Invest Surg Downloaded from informahealthcare.com by Nyu Medical Center on 06/20/15 For personal use only.
1
University of Southern Denmark, Odense M, Denmark, 2 Department of Pathology, Roskilde Hospital, Roskilde, Denmark, 3 Department of Surgery, Odense University Hospital, Odense C, Denmark
ABSTRACT Aim: The aim of this study was to compare the strength and degree of inflammation in small intestinal anastoR ), commoses in rabbits after repeated preoperative treatment with the TNF-α antibody, adalimumab (Humira pared to placebo. Method: Thirty-three New Zealand white female rabbits were randomized to three weeks of weekly subcutaneous injections of adalimumab (n = 24) or placebo (n = 9). After this treatment regime, two end to end anastomoses were performed in the ileum. Following euthanasia on postoperative day 5 the anastomoses were evaluated for minimal tensile strength (MITS) and histological parameters of wound healing using a modified Verhofstad Scale. Results: There were no statistically significant differences between the adalimumab and placebo groups in terms of MITS or histological parameters. Multiple regression analyzes revealed that there was no association between MITS and treatment, numbers of sutures, length of surgery, preoperative weight gain, postoperative weight loss or histological score. On the day of surgery the median serum concentration of adalimumab was 5.4 μg/ml (3.4–8.6). Conclusion: Repeated preoperative treatment with adalimumab had no significant influence on MITS or histological score in anastomoses in the small intestine of the rabbits. Keywords: adalimumab; intestinal anastomoses; minimal tensile strength; intestinal healing
INTRODUCTION
healing via T-cell-mediated suppression of the inflammatory reaction. This may lead to impaired collagen synthesis and wound strength [8] and thus increase the risk of anastomotic leakage. Several studies have investigated the influence of R ), on a second TNF-α antibody, infliximab (Remicade postoperative complications. The evidence is ambiguous, with some studies reporting a correlation [9–12], and others not [7, 8, 13]. Few studies have examined the influence of adalimumab on anastomotic leakage and those investigations have not resulted in any firm conclusions [9, 14]. The aim of this study was to compare the strength and degree of inflammation in small intestinal anastomoses in rabbits after repeated preoperative treatment with adalimumab compared with placebo. The primary outcomes were tensile strength and the degree of inflammation.
Inflammatory bowel disease (IBD)—Crohn’s disease and ulcerative colitis—is one of the most prevalent gastrointestinal illnesses in Europe and the United States with increasing incidence [1, 2]. The inflammation in IBD involves increased secretion of pro-inflammatory cytokines, such as tumor necrosis factor-α (TNF- α), which is present in elevated concentrations in the blood, mucosa and stools [1, 3]. Treatment with TNF-α antibodies such as adalimumab, a recombinant human monoclonal antibody, can induce and maintain long-term clinical remission in patients with IBD [4–6]. About two-thirds of patients with Crohn’s disease require surgery for disease control despite receiving standard medical therapy [3, 7]. In these cases the patient is often undergoing intensive medical treatment, including TNF-α antibodies which may impair wound
Received 28 June 2014; accepted 2 December 2014. Address correspondence to Karen Strebel, MD, University of Southern Denmark, Campusvej 55, 5230 Odense M. E-mail: mini [email protected]
167
168
K. Strebel et al.
MATERIALS AND METHODS Sample Size Estimation A pilot study showed that the mean minimal intestinal tensile strength (MITS) in untreated rabbits was 1.89N (SD = 0.36). With a 2:1 (adalimumab:placebo) randomization, a significance level of 5%, a power of 80% and expected perioperative mortality of 25%, a minimum sample size of 32 rabbits was required to detect a 25% difference in MITS, the minimum difference considered clinically relevant. Randomization was done using the online program Research Randomizer [15].
J Invest Surg Downloaded from informahealthcare.com by Nyu Medical Center on 06/20/15 For personal use only.
Animals and Treatment Thirty-six three-month-old New Zealand White female rabbits, with an average weight of 1.5 kg, were housed in standard wire cages with a plateau, wood shavings and free access to hay, food and water. The room was maintained at a temperature of 19–21◦ C and 40–60% humidity on a 12-hour light/dark cycle. The rabbits were acclimatized for at least one week before treatment commenced. A total of 33 rabbits were included, 24 in the adalimumab group and nine in the placebo group. Two rabbits died during the study, both from the adalimumab group. One rabbit died during induction of anaesthesia and one from postoperative intestinal haemorrhage. The adalimumab group received subcutaneous (s.c.) R , Abbott, Illinois, injections of adalimumab (Humira USA) starting three weeks before surgery, and the placebo group received corresponding volumes of sterile water. Adalimumab (40 mg) was dissolved in sterile water to a concentration of 1 mg/ml, and the treatment regimens for both groups were: • • • •
Three weeks before surgery: 1 ml/kg Two weeks before surgery: 0.5 ml/kg One week before surgery: 0.5 ml/kg On the day of surgery: 0.5 ml/kg
The latter injection was given at the end of surgery, after a blood sample for serum-adalimumab had been drawn. The rabbits were weighed before every injection, before surgery, after surgery and before euthanasia.
Surgical Procedure The rabbits were pre-anaesthetized (0.3 ml/kg s.c.) with a. mixture of fentanylcitrate (0.315 mg/ml) and R , Vetapharma Ltd, fluanisone (10 mg/ml) (Hypnorm Leeds, United Kingdom) and 2 mg/kg s.c. of midazoR , Roche, Basel, Switzerland). After selam (Dormicum
dation 5 ml of blood was drawn from an ear vein for analysis of serum adalimumab concentration. The rabbits were then intubated with an uncuffed tube with an internal diameter of 2.5–3.5 mm. In the event of airR , B. Braun, way reflexes, 2–8 mg propofol (Propovet Melsungen AG, Germany) was given intravenously (i.v.) through the ear vein. Anaesthesia was maintained R , Abbott Laboratowith 2–3% sevoflurane (Sevorane ries, Illinois, USA) via an MCM 801 ventilator (Dameca, Rødovre, Denmark). The rabbits were ventilated at 18 inhalation–exhalation cycles, tidal volume of 30–35 ml, per minute using a mixture of oxygen and atmospheric air. The mixture was adjusted during surgery to maintain oxygen saturation above 92%. Heart rate and oxygen saturation were monitored with a pulse oximeter placed on the rabbit’s ear. Perioperative analgesia R , Jansen-Cilag, consisted of 50 μg/h fentanyl (Haldid Beerse, Belgium) administered through the ear vein via an infusion pump (Terumo STC-526, New Jersey, USA). During surgery the rabbits were placed on a thermostatically controlled heating pad set to 41◦ C to prevent hypothermia. A 4 cm midline laparotomy was performed under aseptic conditions. Twenty and 40 cm cephalad to the caecal appendix, the small intestine was severed and anastomosed with interrupted serosato-serosa inverted single-layer suturing using nonR 5/0, Johnson & Johnresorbable sutures (Prolene son Nordic, Birkerød, Denmark). The musculofascial layer was closed with running absorbable sutures R 3/0, Johnson & Johnson Nordic, Birkerød, (Vicryl Denmark), and the skin with horizontal mattress R 4/0, Johnson & Johnson Nordic, sutures (Prolene Birkerød, Denmark). Antibiotics (0.4 ml/kg i.v. of a mixture of sulphadoxin 200 mg/ml and trimethoprim R Vet, Sehering-Plough Animal 40 mg/ml (Duoprim health, New Jersey, USA)) were given prophylactically at the end of surgery. After surgery, each rabbit was placed in its cage, on a heating pad, with free access to water and soaked food. R , Buprenorphine (0.125 mg/kg s.c.) (Temgesic Schering-Plough Animal health, New Jersey, USA) was given three times daily for post-operative analgesia. The dose was administered immediately after surgery and then repeated every eight hours until the rabbits no longer showed any signs of pain. On postoperative day five (POD5) the rabbits R 0.3 ml/kg s.c. and were sedated with Hypnorm euthanized with 2 ml pentobarbital (Pentobarbital, 200 mg/ml; KU Life, Copenhagen, Denmark) administered via the ear vein. A re-laparotomy was performed and the two anastomoses were identified and freed from adhesions. The anastomoses were resected with a 5 cm margin proximally and distally. The segments, each 10 cm long with the anastomosis in the middle, were cleaned of faecal contents with sterile saline. The cleaned segments were used in two separate tests, the proximal segment was subjected to histopathological Journal of Investigative Surgery
R Humira and Intestinal Anastomotic Healing 169
TABLE 1 Scoring system for the different histological parameters of anastomotic healing. PMNs = polymorphonuclear cells
Score
Necrosis
0
None
1
Small patches Some patches Massive
2 3
PMNs
Lymphocytes
Macrophages
Normal number Slight increase Marked infiltration Massive infiltration
Normal number Slight increase
Normal number Slight increase
Marked infiltration Massive infiltration
Marked infiltration Massive infiltration
Oedema None Some Marked Severe
J Invest Surg Downloaded from informahealthcare.com by Nyu Medical Center on 06/20/15 For personal use only.
analysis, and tensile strength was tested in the distal segment, the latter within ten minutes of removal.
Histopathological Analysis The samples were fixed in 4% formaldehyde, dehydrated and embedded in paraffin. They were cut in 50 μm slices and stained with haematoxylin and eosin (HE), Sirius, Masson’s Trichroms and ASMA (immune staining for α-smooth muscle actin, to enable evaluation of the submucosal-muscular layer). Histological changes due to anastomotic wound healing, development of granulation tissue and local inflammatory response were assessed according to a modified Verhofstad Scale [16] (Table 1). The examiner was blind to treatment assignment.
Tensile Strength Test The entire distal segment was mounted in the testing machine (LF Plus; Lloyds Instruments, Fareham, UK) equipped with a XLC 10N loadcell (Lloyds Instruments, Fareham, UK). The machine was pre-calibrated by the manufacturer. Initially there was 10 mm separation between the clamps with the anastomosis in the middle, and the segment was stretched at a constant deformation rate of 10 mm/min until rupture occurred. We measured the force applied at two time points, when a tear became visible in the serosa (MITS-1) and when a transmural rupture appeared (MITS-2). A simultaneous drop in the load-strain curve calculated by the software was used to confirm the rupture point. Finally, we noted whether the rupture was located within or outside the anastomosis.
Adalimumab—Blood Concentration Quantification of serum adalimumab was based on use of iLiteTM TNF-alpha reporter cells (Biomonitor, Galway, Ireland), which are human erythroleucaemic K562 C
2015 Informa Healthcare USA, Inc.
Mucosal epithelium
Submucosalmuscular layer
Collagen
Normal glandular Normal cubic
Good bridging
Normal
Average bridging
Incomplete cubic Absent
Poor bridging
Slight increase Marked formation Massive formation
No bridging
Fibroblasts Normal number Slight increase Marked infiltration Massive infiltration
cells transfected with genes from a firefly, to indicate R Light Luthe presence of TNF-α inhibitor via a Victor minescence Counter (Perkin Elmer)[17]. The test has a coefficient of variation of less than 25%, as required. Analysis was carried out by Biomonitor A/S, Symbion Science Park, Copenhagen, Denmark.
Statistical Analysis Weight changes, weight on day of surgery and MITS in the adalimumab and placebo groups were compared using paired Student’s t-tests with unequal variances. The median serum concentration of adalimumab and quartiles were calculated. Two-tailed Fisher’s exact test was used to assess potential group differences in histological parameters. Multiple linear regressions were used to assess whether number of sutures, weight on day of surgery, preoperative weight gain, postoperative weight loss, duration of surgery, histological score or treatment (adalimumab or placebo) had any influence on MITS. Statistical analyses were performed using Stata/IC (version 12.0; Texas, USA).
Ethical Consideration This project was approved by the Danish Experimental Animal Board.
RESULTS Tensile Strength Test Data from one rabbit in the adalimumab group was excluded owing to difficulties with mounting the anastomosis, leaving a sample of 30 rabbits for the tensile strength test (adalimumab group: n = 21; placebo group: n = 9). Only two ruptures occurred outside the anastomotic line, one from each group. There was a non-significant difference in MITS-1 between the adalimumab and placebo group of 0.175N (−0.077–0.426)
170
K. Strebel et al.
TABLE 2 Perioperative weight changes and minimal tensile strength (MITS) of the small intestinal anastomoses. MITS-1 = serosal bursting and MITS-2 = transmural rupture. Figures are mean ± SD Adalimumab
Placebo
n = 31 Preoperative weight gain (%) Weight on day of surgery (kg) Postoperative weight loss (%) Tensile strength n = 30 MITS-1 MITS-2
n = 22 38.08 ± 13.16 2.49 ± 0.41 4.70 ± 3.92 n = 21 0.53 ± 0.25N 1.43 ± 0.52N
n=9 38.97 ± 14.66 2.49 ± 0.38 6.10 ± 4.25 n=9 0.71 ± 0.42N 1.49 ± 0.26N
(p = .167). There was a non-significant group difference in MITS-2 of 0.059N(−0.233–0.351) (p = .683) (Table 2). J Invest Surg Downloaded from informahealthcare.com by Nyu Medical Center on 06/20/15 For personal use only.
p-value
Weight
Animal Weight There were no significant group differences in preoperative weight gain, weight on day of surgery or postoperative weight loss (Table 2).
Serum Concentration of Adalimumab Median serum concentration of adalimumab was 5.4 μg/ml (min 3.4, max 8.6) on the day of surgery.
Histological Score Samples from one rabbit in the adalimumab group were excluded due to errors in the embedding process. There were no group differences in any of the histological parameters assessed (Table 3).
Multiple Regression Multiple regression revealed that there was no association between MITS-1 or MITS-2 and treatment, number of sutures, total histological score or length of surgery. There was an association between MITS and weight on day of surgery; MITS-1: = 0.664 (0.245–1.083) (p = .004), MITS-2: = 0.660 (0.149–1.171) (p = .014).
0.88 0.99 0.39 0.17 0.68
DISCUSSION Preoperative adalimumab treatment had no significant influence on minimal tensile strength (MITS) of small intestinal anastomoses when compared with placebo. Furthermore no significant difference was found regarding the histological parameters of wound healing in the adalimumab and placebo groups. Our findings are consistent with Nørgaard et al. [14] who, in a large Danish nationwide register study, found no significant risk of postoperative complications associated with preoperative use of anti-TNF-α agents in Crohn’s patients. A case control study by Canedo et al. [13] reported that the rate of postoperative complications in patients undergoing surgery for Crohn’s disease was similar for a group who underwent preoperative treatment with anti-TNF-α agents and a control group. Previous surgeries and disease severity were important potential confounding factors in these studies. The treatment group had fewer previous surgeries than the controls in the case control study [13]. The Danish register study [14] did not collect data on disease severity at time of surgery. Others have found that preoperative treatment with anti-TNF-α increased risk of anastomotic leakage in patients undergoing surgery for Crohn’s disease and the incidence of postoperative infectious complications [9, 12]. Similar results were reported by Syed et al. [10] but these authors did not find an increase in the risk of anastomotic leakage. Another study on rabbits which used a single bolus injection of adalimumab one week before surgery reported no significant changes in anastomotic strength or degree of inflammation compared to a placebo group [18]; however this treatment regimen is not comparable to clinical practice. We chose to use repeated
TABLE 3 Results of the histological examination (mean ± SD), P-value (Fisher’s exact test) PMNs = polymorphonuclear cells; ME = mucosal epithelium; SML = submucosal-muscular layer
Placebo Adalimumab p-value
Necrosis
PMN
1.2 ± 0.67 1.67 ± 0.91 0.51
1 ± 0.5 1.19 ± 0.68 0.70
Lymphocytes Macrophages 0.56 ± 0.73 0.76 ± 0.62 0.54
0.22 ± 0.44 0.62 ± 0.67 0.29
Oedema
ME
SML
Collagen
Fibroblasts
0.33 ± 0.5 0.62 ± 0.5 0.24
1.1 ± 1.36 2.1 ± 1.18 0.18
1.56 ± 0.73 2.05 ± 0.74 0.31
1.56 ± 0.88 1.71 ± 0.64 0.62
1.44 ± 0.53 1.48 ± 0.68 1.00
Journal of Investigative Surgery
J Invest Surg Downloaded from informahealthcare.com by Nyu Medical Center on 06/20/15 For personal use only.
R Humira and Intestinal Anastomotic Healing 171
adalimumab injections to mimic the clinical regime used in humans; we used a dose corresponding to that recommended for induction and maintenance of disease control in humans [19]. Injections were given weekly rather than fortnightly—the recommended regimen in humans—because rabbits’ metabolise the drug more quickly than humans. In our sample the median serum concentration of adalimumab on the day of surgery was > 5 μg/ml, which corresponds to the recommended maintenance value in humans [20]. An additional dose of adalimumab was administered after surgery, so it is reasonable to assume that all rabbits had a significant serum concentration of adalimumab throughout the healing period. The recommended method for evaluation of anastomotic healing is MITS; it is considered a more valid index of healing than maximal tensile strength, i.e. the maximum force a specimen can withstand without rupturing [18, 21, 22]. MITS is a measure of the strength of the anastomosis rather than the strength of the sutures. Although the test was automated placement of the intestine in the clamps had to be done manually and with precision. In some cases the placement appeared to be inadequate, and may have resulted in a lower MITS value. MITS values may also have been affected by the requirement for anastomoses to be detached from adhesions and mesenteric tissue. However, in practice only a few anastomoses had adhesions, and these were left untouched if they were localized in the anastomotic line to ensure that the anastomosis was not damaged during removal. MITS is not clearly defined in the existing literature so we used two different definitions: (1) a visible tear in the serosa and (2) a transmural rupture. These measures provided a conservative assessment of anastomotic strength, which was considered important as even a small tear can lead to postoperative complications. Despite the small number of methodological problems discussed, the non-significant difference in MITS between the two groups appeared to be valid. Multiple regression revealed an association between MITS and weight on day of surgery in the whole cohort, suggesting that MITS was dependent on the nutritional status and size of the rabbit. There was no group difference in presurgery weight and so it is unsurprising that we found no group difference in MITS. The histological examination was performed by one blinded pathologist using a modification of the widely accepted Verhofstad Scale [16]. Various intestinal suturing techniques are used in clinical practice, such as serosa-to-serosa, mucosa-tomucosa and stapling. There is no evidence to suggest that one method is superior in terms of strength and incidence of postoperative complications [21, 23]. We used the interrupted serosa-to-serosa inverted singlelayer technique to facilitate comparison of our results with other studies. The anastomoses were performed in the ileum rather than the colon because most in C
2015 Informa Healthcare USA, Inc.
testinal surgeries in Crohn’s patients involve the small intestine. The rabbits were euthanized on POD5 because the healing process is well established by this time. Because the “lag” phase, which encompasses the influx of inflammatory cells and macrophages, had occurred shortly before POD5 we were able to evaluate histological parameters of inflammation. On POD5 the proliferation phase is in progress with peak numbers of fibroblasts and collagen density making the anastomosis stronger [24]. Anastomotic leakage and postoperative complications usually occur within the first few postoperative days [24], which also made POD5 an appropriate termination point. In conclusion, we found no significant differences in strength and degree of inflammation in small intestinal anastomoses in rabbits which underwent repeated preoperative treatment with adalimumab or placebo. Further prospective clinical studies are however necessary to exclude preoperative treatment with adalimumab as a risk factor for postoperative complications in patients undergoing intestinal resection for Crohn’s disease.
ACKNOWLEDGMENTS Special thanks to veterinarian Henrik Saxtorph, Head of Department, MSc PhD, Peter Bollen, animal technician Anne Mette Durand and the rest of the staff at the Biomedical Laboratory, University of Southern Denmark. Thanks to Claire Gudex for linguistic advice. Declaration of Interest: The authors of this paper have no conflict of interest affecting the research reported here.
FUNDING The project was financed by the research council at Odense University Hospital.
REFERENCES [1] Ahluwalia JP. Immunotherapy in inflammatory bowel disease. Med Clin North Am. 2012;96:525–544. [2] Huang X, Lv B, Jin HF, et al. A meta-analysis of the therapeutic effects of tumor necrosis factor-alpha blockers on ulcerative colitis. Eur J Clin Pharmacol. 2011;67:759–766. [3] Bregnbak D, Mortensen C, Bendtsen F. Infliximab and complications after colectomy in patients with ulcerative colitis. J Crohns Colitis. 2012;6:281–286. [4] Colombel JF, Sandborn WJ, Rutgeerts P, et al. Adalimumab for maintenance of clinical response and remission in patients with Crohn’s disease: the CHARM trial. Gastroenterology. 2007;132:52–65. [5] Hanauer SB, Sandborn WJ, Rutgeerts P, et al. Human antitumor necrosis factor monoclonal antibody (adalimumab)
172
[6]
[7]
[8]
[9]
J Invest Surg Downloaded from informahealthcare.com by Nyu Medical Center on 06/20/15 For personal use only.
[10]
[11]
[12]
[13]
[14]
K. Strebel et al. in Crohn’s disease: the CLASSIC-I trial. Gastroenterology. 2006;130:323–333. Sandborn WJ, Hanauer SB, Rutgeerts P, et al. Adalimumab for maintenance treatment of Crohn’s disease: results of the CLASSIC II trial. Gut. 2007;56:1232–1239. Subramanian V, Pollok RC, Kang JY, et al. Systematic review of postoperative complications in patients with inflammatory bowel disease treated with immunomodulators. Br J Surg. 2006;93:793–799. El-Hussuna A, Andersen J, Bisgaard T, et al. Biologic treatment or immunomodulation is not associated with postoperative anastomotic complications in abdominal surgery for Crohn’s disease. Scand J Gastroenterol. 2012;47:662–668. Samimi R, Flasar MH, Kavic S, et al. Outcome of medical treatment of stricturing and penetrating Crohn’s disease: a retrospective study. Inflamm Bowel Dis. 2010;16:1187–1194. Syed A, Cross RK, Flasar MH. Anti-tumor necrosis factor therapy is associated with infections after abdominal surgery in crohn’s disease patients. Am J Gastroenterol. 2013 Mar 12. Kopylov U, Ben-Horin S, Zmora O, et al. Anti-tumor necrosis factor and postoperative complications in Crohn’s disease: systematic review and meta-analysis. Inflamm Bowel Dis. 2012;18:2404–2413. Appau KA, Fazio VW, Shen B, et al. Use of infliximab within 3 months of ileocolonic resection is associated with adverse postoperative outcomes in Crohn’s patients. J Gastrointest Surg. 2008;12:1738–1744. Canedo J, Lee SH, Pinto R, et al. Surgical resection in Crohn’s disease: is immunosuppressive medication associated with higher postoperative infection rates? Colorectal Dis. 2011;13:1294–1298. Norgard BM, Nielsen J, Qvist N, et al. Pre-operative use of anti-TNF-alpha agents and the risk of post-operative complications in patients with ulcerative colitis - a nationwide cohort study. Aliment Pharmacol Ther. 2012;35:1301–1309.
[15] Urbaniak GC, Plous S. Research Randomizer (version 4.0) [internet]: Social Psychology Network; 2013 [cited 2013 Mar 13]. Available from: www.randomizer.org. [16] Verhofstad MH, Lange WP, van der Laak JA, et al. Microscopic analysis of anastomotic healing in the intestine of normal and diabetic rats. Dis Colon Rectum. 2001;44:423–431. [17] Lallemand C, Kavrochorianou N, Steenholdt C, et al. Reporter gene assay for the quantification of the activity and neutralizing antibody response to TNFalpha antagonists. J Immunol Methods. 2011;28;(373):229–239. [18] Ploug T, Andersen K, Hansen K, et al. Influence of adalimumab treatment on anastomotic strength, degree of inflammation, and collagen formation: an experimental study on the small intestine of rabbits. Inflamm Bowel Dis. 2013;19:254–258. [19] pro.medicin.dk [internet]: Dansk Lægemiddel Information A/S; [updated 2012 Nov 6; cited 2013 Jul 13]. Available from: www.pro.medicin.dk/Medicin/Praeparater/3237. [20] SSI [internet]: Statens Serum Institut; 2013 [updated 2013 May 2; cited 2013 Jul 13]. Available from: www.ssi.dk/ Diagnostik/DiagnostiskHaandbog/400–499/435.aspx. [21] Ikeuchi D, Onodera H, Aung T, et al. Correlation of tensile strength with bursting pressure in the evaluation of intestinal anastomosis. Dig Surg. 1999;16:478– 485. [22] Rossi LF, Ramos RR, Ely JB, et al. Considerations that may influence the result of trials assessing tensile strength in experimental surgery. Acta Cir Bras. 2007;22:499– 502. [23] Ulmer TF, Stumpf M, Rosch R, et al. Suture-free and mesh reinforced small intestinal anstomoses: a feasibility study in rabbits. J Invest Surg. 2013;26:210–216. [24] Thompson SK, Chang EY, Jobe BA. Clinical review: Healing in gastrointestinal anastomoses, part I. Microsurgery. 2006;26:131–136.
Journal of Investigative Surgery
ORIGINAL ARTICLE
Effect of HumiraR on Intestinal Anastomotic Response in Rabbits Karen Strebel, MD,1 Susanne R. H. Nielsen, MD,1 Matteo Biagini, MD,2 Niels Qvist, DMSci3
J Invest Surg Downloaded from informahealthcare.com by Nyu Medical Center on 06/20/15 For personal use only.
1
University of Southern Denmark, Odense M, Denmark, 2 Department of Pathology, Roskilde Hospital, Roskilde, Denmark, 3 Department of Surgery, Odense University Hospital, Odense C, Denmark
ABSTRACT Aim: The aim of this study was to compare the strength and degree of inflammation in small intestinal anastoR ), commoses in rabbits after repeated preoperative treatment with the TNF-α antibody, adalimumab (Humira pared to placebo. Method: Thirty-three New Zealand white female rabbits were randomized to three weeks of weekly subcutaneous injections of adalimumab (n = 24) or placebo (n = 9). After this treatment regime, two end to end anastomoses were performed in the ileum. Following euthanasia on postoperative day 5 the anastomoses were evaluated for minimal tensile strength (MITS) and histological parameters of wound healing using a modified Verhofstad Scale. Results: There were no statistically significant differences between the adalimumab and placebo groups in terms of MITS or histological parameters. Multiple regression analyzes revealed that there was no association between MITS and treatment, numbers of sutures, length of surgery, preoperative weight gain, postoperative weight loss or histological score. On the day of surgery the median serum concentration of adalimumab was 5.4 μg/ml (3.4–8.6). Conclusion: Repeated preoperative treatment with adalimumab had no significant influence on MITS or histological score in anastomoses in the small intestine of the rabbits. Keywords: adalimumab; intestinal anastomoses; minimal tensile strength; intestinal healing
INTRODUCTION
healing via T-cell-mediated suppression of the inflammatory reaction. This may lead to impaired collagen synthesis and wound strength [8] and thus increase the risk of anastomotic leakage. Several studies have investigated the influence of R ), on a second TNF-α antibody, infliximab (Remicade postoperative complications. The evidence is ambiguous, with some studies reporting a correlation [9–12], and others not [7, 8, 13]. Few studies have examined the influence of adalimumab on anastomotic leakage and those investigations have not resulted in any firm conclusions [9, 14]. The aim of this study was to compare the strength and degree of inflammation in small intestinal anastomoses in rabbits after repeated preoperative treatment with adalimumab compared with placebo. The primary outcomes were tensile strength and the degree of inflammation.
Inflammatory bowel disease (IBD)—Crohn’s disease and ulcerative colitis—is one of the most prevalent gastrointestinal illnesses in Europe and the United States with increasing incidence [1, 2]. The inflammation in IBD involves increased secretion of pro-inflammatory cytokines, such as tumor necrosis factor-α (TNF- α), which is present in elevated concentrations in the blood, mucosa and stools [1, 3]. Treatment with TNF-α antibodies such as adalimumab, a recombinant human monoclonal antibody, can induce and maintain long-term clinical remission in patients with IBD [4–6]. About two-thirds of patients with Crohn’s disease require surgery for disease control despite receiving standard medical therapy [3, 7]. In these cases the patient is often undergoing intensive medical treatment, including TNF-α antibodies which may impair wound
Received 28 June 2014; accepted 2 December 2014. Address correspondence to Karen Strebel, MD, University of Southern Denmark, Campusvej 55, 5230 Odense M. E-mail: mini [email protected]
167
168
K. Strebel et al.
MATERIALS AND METHODS Sample Size Estimation A pilot study showed that the mean minimal intestinal tensile strength (MITS) in untreated rabbits was 1.89N (SD = 0.36). With a 2:1 (adalimumab:placebo) randomization, a significance level of 5%, a power of 80% and expected perioperative mortality of 25%, a minimum sample size of 32 rabbits was required to detect a 25% difference in MITS, the minimum difference considered clinically relevant. Randomization was done using the online program Research Randomizer [15].
J Invest Surg Downloaded from informahealthcare.com by Nyu Medical Center on 06/20/15 For personal use only.
Animals and Treatment Thirty-six three-month-old New Zealand White female rabbits, with an average weight of 1.5 kg, were housed in standard wire cages with a plateau, wood shavings and free access to hay, food and water. The room was maintained at a temperature of 19–21◦ C and 40–60% humidity on a 12-hour light/dark cycle. The rabbits were acclimatized for at least one week before treatment commenced. A total of 33 rabbits were included, 24 in the adalimumab group and nine in the placebo group. Two rabbits died during the study, both from the adalimumab group. One rabbit died during induction of anaesthesia and one from postoperative intestinal haemorrhage. The adalimumab group received subcutaneous (s.c.) R , Abbott, Illinois, injections of adalimumab (Humira USA) starting three weeks before surgery, and the placebo group received corresponding volumes of sterile water. Adalimumab (40 mg) was dissolved in sterile water to a concentration of 1 mg/ml, and the treatment regimens for both groups were: • • • •
Three weeks before surgery: 1 ml/kg Two weeks before surgery: 0.5 ml/kg One week before surgery: 0.5 ml/kg On the day of surgery: 0.5 ml/kg
The latter injection was given at the end of surgery, after a blood sample for serum-adalimumab had been drawn. The rabbits were weighed before every injection, before surgery, after surgery and before euthanasia.
Surgical Procedure The rabbits were pre-anaesthetized (0.3 ml/kg s.c.) with a. mixture of fentanylcitrate (0.315 mg/ml) and R , Vetapharma Ltd, fluanisone (10 mg/ml) (Hypnorm Leeds, United Kingdom) and 2 mg/kg s.c. of midazoR , Roche, Basel, Switzerland). After selam (Dormicum
dation 5 ml of blood was drawn from an ear vein for analysis of serum adalimumab concentration. The rabbits were then intubated with an uncuffed tube with an internal diameter of 2.5–3.5 mm. In the event of airR , B. Braun, way reflexes, 2–8 mg propofol (Propovet Melsungen AG, Germany) was given intravenously (i.v.) through the ear vein. Anaesthesia was maintained R , Abbott Laboratowith 2–3% sevoflurane (Sevorane ries, Illinois, USA) via an MCM 801 ventilator (Dameca, Rødovre, Denmark). The rabbits were ventilated at 18 inhalation–exhalation cycles, tidal volume of 30–35 ml, per minute using a mixture of oxygen and atmospheric air. The mixture was adjusted during surgery to maintain oxygen saturation above 92%. Heart rate and oxygen saturation were monitored with a pulse oximeter placed on the rabbit’s ear. Perioperative analgesia R , Jansen-Cilag, consisted of 50 μg/h fentanyl (Haldid Beerse, Belgium) administered through the ear vein via an infusion pump (Terumo STC-526, New Jersey, USA). During surgery the rabbits were placed on a thermostatically controlled heating pad set to 41◦ C to prevent hypothermia. A 4 cm midline laparotomy was performed under aseptic conditions. Twenty and 40 cm cephalad to the caecal appendix, the small intestine was severed and anastomosed with interrupted serosato-serosa inverted single-layer suturing using nonR 5/0, Johnson & Johnresorbable sutures (Prolene son Nordic, Birkerød, Denmark). The musculofascial layer was closed with running absorbable sutures R 3/0, Johnson & Johnson Nordic, Birkerød, (Vicryl Denmark), and the skin with horizontal mattress R 4/0, Johnson & Johnson Nordic, sutures (Prolene Birkerød, Denmark). Antibiotics (0.4 ml/kg i.v. of a mixture of sulphadoxin 200 mg/ml and trimethoprim R Vet, Sehering-Plough Animal 40 mg/ml (Duoprim health, New Jersey, USA)) were given prophylactically at the end of surgery. After surgery, each rabbit was placed in its cage, on a heating pad, with free access to water and soaked food. R , Buprenorphine (0.125 mg/kg s.c.) (Temgesic Schering-Plough Animal health, New Jersey, USA) was given three times daily for post-operative analgesia. The dose was administered immediately after surgery and then repeated every eight hours until the rabbits no longer showed any signs of pain. On postoperative day five (POD5) the rabbits R 0.3 ml/kg s.c. and were sedated with Hypnorm euthanized with 2 ml pentobarbital (Pentobarbital, 200 mg/ml; KU Life, Copenhagen, Denmark) administered via the ear vein. A re-laparotomy was performed and the two anastomoses were identified and freed from adhesions. The anastomoses were resected with a 5 cm margin proximally and distally. The segments, each 10 cm long with the anastomosis in the middle, were cleaned of faecal contents with sterile saline. The cleaned segments were used in two separate tests, the proximal segment was subjected to histopathological Journal of Investigative Surgery
R Humira and Intestinal Anastomotic Healing 169
TABLE 1 Scoring system for the different histological parameters of anastomotic healing. PMNs = polymorphonuclear cells
Score
Necrosis
0
None
1
Small patches Some patches Massive
2 3
PMNs
Lymphocytes
Macrophages
Normal number Slight increase Marked infiltration Massive infiltration
Normal number Slight increase
Normal number Slight increase
Marked infiltration Massive infiltration
Marked infiltration Massive infiltration
Oedema None Some Marked Severe
J Invest Surg Downloaded from informahealthcare.com by Nyu Medical Center on 06/20/15 For personal use only.
analysis, and tensile strength was tested in the distal segment, the latter within ten minutes of removal.
Histopathological Analysis The samples were fixed in 4% formaldehyde, dehydrated and embedded in paraffin. They were cut in 50 μm slices and stained with haematoxylin and eosin (HE), Sirius, Masson’s Trichroms and ASMA (immune staining for α-smooth muscle actin, to enable evaluation of the submucosal-muscular layer). Histological changes due to anastomotic wound healing, development of granulation tissue and local inflammatory response were assessed according to a modified Verhofstad Scale [16] (Table 1). The examiner was blind to treatment assignment.
Tensile Strength Test The entire distal segment was mounted in the testing machine (LF Plus; Lloyds Instruments, Fareham, UK) equipped with a XLC 10N loadcell (Lloyds Instruments, Fareham, UK). The machine was pre-calibrated by the manufacturer. Initially there was 10 mm separation between the clamps with the anastomosis in the middle, and the segment was stretched at a constant deformation rate of 10 mm/min until rupture occurred. We measured the force applied at two time points, when a tear became visible in the serosa (MITS-1) and when a transmural rupture appeared (MITS-2). A simultaneous drop in the load-strain curve calculated by the software was used to confirm the rupture point. Finally, we noted whether the rupture was located within or outside the anastomosis.
Adalimumab—Blood Concentration Quantification of serum adalimumab was based on use of iLiteTM TNF-alpha reporter cells (Biomonitor, Galway, Ireland), which are human erythroleucaemic K562 C
2015 Informa Healthcare USA, Inc.
Mucosal epithelium
Submucosalmuscular layer
Collagen
Normal glandular Normal cubic
Good bridging
Normal
Average bridging
Incomplete cubic Absent
Poor bridging
Slight increase Marked formation Massive formation
No bridging
Fibroblasts Normal number Slight increase Marked infiltration Massive infiltration
cells transfected with genes from a firefly, to indicate R Light Luthe presence of TNF-α inhibitor via a Victor minescence Counter (Perkin Elmer)[17]. The test has a coefficient of variation of less than 25%, as required. Analysis was carried out by Biomonitor A/S, Symbion Science Park, Copenhagen, Denmark.
Statistical Analysis Weight changes, weight on day of surgery and MITS in the adalimumab and placebo groups were compared using paired Student’s t-tests with unequal variances. The median serum concentration of adalimumab and quartiles were calculated. Two-tailed Fisher’s exact test was used to assess potential group differences in histological parameters. Multiple linear regressions were used to assess whether number of sutures, weight on day of surgery, preoperative weight gain, postoperative weight loss, duration of surgery, histological score or treatment (adalimumab or placebo) had any influence on MITS. Statistical analyses were performed using Stata/IC (version 12.0; Texas, USA).
Ethical Consideration This project was approved by the Danish Experimental Animal Board.
RESULTS Tensile Strength Test Data from one rabbit in the adalimumab group was excluded owing to difficulties with mounting the anastomosis, leaving a sample of 30 rabbits for the tensile strength test (adalimumab group: n = 21; placebo group: n = 9). Only two ruptures occurred outside the anastomotic line, one from each group. There was a non-significant difference in MITS-1 between the adalimumab and placebo group of 0.175N (−0.077–0.426)
170
K. Strebel et al.
TABLE 2 Perioperative weight changes and minimal tensile strength (MITS) of the small intestinal anastomoses. MITS-1 = serosal bursting and MITS-2 = transmural rupture. Figures are mean ± SD Adalimumab
Placebo
n = 31 Preoperative weight gain (%) Weight on day of surgery (kg) Postoperative weight loss (%) Tensile strength n = 30 MITS-1 MITS-2
n = 22 38.08 ± 13.16 2.49 ± 0.41 4.70 ± 3.92 n = 21 0.53 ± 0.25N 1.43 ± 0.52N
n=9 38.97 ± 14.66 2.49 ± 0.38 6.10 ± 4.25 n=9 0.71 ± 0.42N 1.49 ± 0.26N
(p = .167). There was a non-significant group difference in MITS-2 of 0.059N(−0.233–0.351) (p = .683) (Table 2). J Invest Surg Downloaded from informahealthcare.com by Nyu Medical Center on 06/20/15 For personal use only.
p-value
Weight
Animal Weight There were no significant group differences in preoperative weight gain, weight on day of surgery or postoperative weight loss (Table 2).
Serum Concentration of Adalimumab Median serum concentration of adalimumab was 5.4 μg/ml (min 3.4, max 8.6) on the day of surgery.
Histological Score Samples from one rabbit in the adalimumab group were excluded due to errors in the embedding process. There were no group differences in any of the histological parameters assessed (Table 3).
Multiple Regression Multiple regression revealed that there was no association between MITS-1 or MITS-2 and treatment, number of sutures, total histological score or length of surgery. There was an association between MITS and weight on day of surgery; MITS-1: = 0.664 (0.245–1.083) (p = .004), MITS-2: = 0.660 (0.149–1.171) (p = .014).
0.88 0.99 0.39 0.17 0.68
DISCUSSION Preoperative adalimumab treatment had no significant influence on minimal tensile strength (MITS) of small intestinal anastomoses when compared with placebo. Furthermore no significant difference was found regarding the histological parameters of wound healing in the adalimumab and placebo groups. Our findings are consistent with Nørgaard et al. [14] who, in a large Danish nationwide register study, found no significant risk of postoperative complications associated with preoperative use of anti-TNF-α agents in Crohn’s patients. A case control study by Canedo et al. [13] reported that the rate of postoperative complications in patients undergoing surgery for Crohn’s disease was similar for a group who underwent preoperative treatment with anti-TNF-α agents and a control group. Previous surgeries and disease severity were important potential confounding factors in these studies. The treatment group had fewer previous surgeries than the controls in the case control study [13]. The Danish register study [14] did not collect data on disease severity at time of surgery. Others have found that preoperative treatment with anti-TNF-α increased risk of anastomotic leakage in patients undergoing surgery for Crohn’s disease and the incidence of postoperative infectious complications [9, 12]. Similar results were reported by Syed et al. [10] but these authors did not find an increase in the risk of anastomotic leakage. Another study on rabbits which used a single bolus injection of adalimumab one week before surgery reported no significant changes in anastomotic strength or degree of inflammation compared to a placebo group [18]; however this treatment regimen is not comparable to clinical practice. We chose to use repeated
TABLE 3 Results of the histological examination (mean ± SD), P-value (Fisher’s exact test) PMNs = polymorphonuclear cells; ME = mucosal epithelium; SML = submucosal-muscular layer
Placebo Adalimumab p-value
Necrosis
PMN
1.2 ± 0.67 1.67 ± 0.91 0.51
1 ± 0.5 1.19 ± 0.68 0.70
Lymphocytes Macrophages 0.56 ± 0.73 0.76 ± 0.62 0.54
0.22 ± 0.44 0.62 ± 0.67 0.29
Oedema
ME
SML
Collagen
Fibroblasts
0.33 ± 0.5 0.62 ± 0.5 0.24
1.1 ± 1.36 2.1 ± 1.18 0.18
1.56 ± 0.73 2.05 ± 0.74 0.31
1.56 ± 0.88 1.71 ± 0.64 0.62
1.44 ± 0.53 1.48 ± 0.68 1.00
Journal of Investigative Surgery
J Invest Surg Downloaded from informahealthcare.com by Nyu Medical Center on 06/20/15 For personal use only.
R Humira and Intestinal Anastomotic Healing 171
adalimumab injections to mimic the clinical regime used in humans; we used a dose corresponding to that recommended for induction and maintenance of disease control in humans [19]. Injections were given weekly rather than fortnightly—the recommended regimen in humans—because rabbits’ metabolise the drug more quickly than humans. In our sample the median serum concentration of adalimumab on the day of surgery was > 5 μg/ml, which corresponds to the recommended maintenance value in humans [20]. An additional dose of adalimumab was administered after surgery, so it is reasonable to assume that all rabbits had a significant serum concentration of adalimumab throughout the healing period. The recommended method for evaluation of anastomotic healing is MITS; it is considered a more valid index of healing than maximal tensile strength, i.e. the maximum force a specimen can withstand without rupturing [18, 21, 22]. MITS is a measure of the strength of the anastomosis rather than the strength of the sutures. Although the test was automated placement of the intestine in the clamps had to be done manually and with precision. In some cases the placement appeared to be inadequate, and may have resulted in a lower MITS value. MITS values may also have been affected by the requirement for anastomoses to be detached from adhesions and mesenteric tissue. However, in practice only a few anastomoses had adhesions, and these were left untouched if they were localized in the anastomotic line to ensure that the anastomosis was not damaged during removal. MITS is not clearly defined in the existing literature so we used two different definitions: (1) a visible tear in the serosa and (2) a transmural rupture. These measures provided a conservative assessment of anastomotic strength, which was considered important as even a small tear can lead to postoperative complications. Despite the small number of methodological problems discussed, the non-significant difference in MITS between the two groups appeared to be valid. Multiple regression revealed an association between MITS and weight on day of surgery in the whole cohort, suggesting that MITS was dependent on the nutritional status and size of the rabbit. There was no group difference in presurgery weight and so it is unsurprising that we found no group difference in MITS. The histological examination was performed by one blinded pathologist using a modification of the widely accepted Verhofstad Scale [16]. Various intestinal suturing techniques are used in clinical practice, such as serosa-to-serosa, mucosa-tomucosa and stapling. There is no evidence to suggest that one method is superior in terms of strength and incidence of postoperative complications [21, 23]. We used the interrupted serosa-to-serosa inverted singlelayer technique to facilitate comparison of our results with other studies. The anastomoses were performed in the ileum rather than the colon because most in C
2015 Informa Healthcare USA, Inc.
testinal surgeries in Crohn’s patients involve the small intestine. The rabbits were euthanized on POD5 because the healing process is well established by this time. Because the “lag” phase, which encompasses the influx of inflammatory cells and macrophages, had occurred shortly before POD5 we were able to evaluate histological parameters of inflammation. On POD5 the proliferation phase is in progress with peak numbers of fibroblasts and collagen density making the anastomosis stronger [24]. Anastomotic leakage and postoperative complications usually occur within the first few postoperative days [24], which also made POD5 an appropriate termination point. In conclusion, we found no significant differences in strength and degree of inflammation in small intestinal anastomoses in rabbits which underwent repeated preoperative treatment with adalimumab or placebo. Further prospective clinical studies are however necessary to exclude preoperative treatment with adalimumab as a risk factor for postoperative complications in patients undergoing intestinal resection for Crohn’s disease.
ACKNOWLEDGMENTS Special thanks to veterinarian Henrik Saxtorph, Head of Department, MSc PhD, Peter Bollen, animal technician Anne Mette Durand and the rest of the staff at the Biomedical Laboratory, University of Southern Denmark. Thanks to Claire Gudex for linguistic advice. Declaration of Interest: The authors of this paper have no conflict of interest affecting the research reported here.
FUNDING The project was financed by the research council at Odense University Hospital.
REFERENCES [1] Ahluwalia JP. Immunotherapy in inflammatory bowel disease. Med Clin North Am. 2012;96:525–544. [2] Huang X, Lv B, Jin HF, et al. A meta-analysis of the therapeutic effects of tumor necrosis factor-alpha blockers on ulcerative colitis. Eur J Clin Pharmacol. 2011;67:759–766. [3] Bregnbak D, Mortensen C, Bendtsen F. Infliximab and complications after colectomy in patients with ulcerative colitis. J Crohns Colitis. 2012;6:281–286. [4] Colombel JF, Sandborn WJ, Rutgeerts P, et al. Adalimumab for maintenance of clinical response and remission in patients with Crohn’s disease: the CHARM trial. Gastroenterology. 2007;132:52–65. [5] Hanauer SB, Sandborn WJ, Rutgeerts P, et al. Human antitumor necrosis factor monoclonal antibody (adalimumab)
172
[6]
[7]
[8]
[9]
J Invest Surg Downloaded from informahealthcare.com by Nyu Medical Center on 06/20/15 For personal use only.
[10]
[11]
[12]
[13]
[14]
K. Strebel et al. in Crohn’s disease: the CLASSIC-I trial. Gastroenterology. 2006;130:323–333. Sandborn WJ, Hanauer SB, Rutgeerts P, et al. Adalimumab for maintenance treatment of Crohn’s disease: results of the CLASSIC II trial. Gut. 2007;56:1232–1239. Subramanian V, Pollok RC, Kang JY, et al. Systematic review of postoperative complications in patients with inflammatory bowel disease treated with immunomodulators. Br J Surg. 2006;93:793–799. El-Hussuna A, Andersen J, Bisgaard T, et al. Biologic treatment or immunomodulation is not associated with postoperative anastomotic complications in abdominal surgery for Crohn’s disease. Scand J Gastroenterol. 2012;47:662–668. Samimi R, Flasar MH, Kavic S, et al. Outcome of medical treatment of stricturing and penetrating Crohn’s disease: a retrospective study. Inflamm Bowel Dis. 2010;16:1187–1194. Syed A, Cross RK, Flasar MH. Anti-tumor necrosis factor therapy is associated with infections after abdominal surgery in crohn’s disease patients. Am J Gastroenterol. 2013 Mar 12. Kopylov U, Ben-Horin S, Zmora O, et al. Anti-tumor necrosis factor and postoperative complications in Crohn’s disease: systematic review and meta-analysis. Inflamm Bowel Dis. 2012;18:2404–2413. Appau KA, Fazio VW, Shen B, et al. Use of infliximab within 3 months of ileocolonic resection is associated with adverse postoperative outcomes in Crohn’s patients. J Gastrointest Surg. 2008;12:1738–1744. Canedo J, Lee SH, Pinto R, et al. Surgical resection in Crohn’s disease: is immunosuppressive medication associated with higher postoperative infection rates? Colorectal Dis. 2011;13:1294–1298. Norgard BM, Nielsen J, Qvist N, et al. Pre-operative use of anti-TNF-alpha agents and the risk of post-operative complications in patients with ulcerative colitis - a nationwide cohort study. Aliment Pharmacol Ther. 2012;35:1301–1309.
[15] Urbaniak GC, Plous S. Research Randomizer (version 4.0) [internet]: Social Psychology Network; 2013 [cited 2013 Mar 13]. Available from: www.randomizer.org. [16] Verhofstad MH, Lange WP, van der Laak JA, et al. Microscopic analysis of anastomotic healing in the intestine of normal and diabetic rats. Dis Colon Rectum. 2001;44:423–431. [17] Lallemand C, Kavrochorianou N, Steenholdt C, et al. Reporter gene assay for the quantification of the activity and neutralizing antibody response to TNFalpha antagonists. J Immunol Methods. 2011;28;(373):229–239. [18] Ploug T, Andersen K, Hansen K, et al. Influence of adalimumab treatment on anastomotic strength, degree of inflammation, and collagen formation: an experimental study on the small intestine of rabbits. Inflamm Bowel Dis. 2013;19:254–258. [19] pro.medicin.dk [internet]: Dansk Lægemiddel Information A/S; [updated 2012 Nov 6; cited 2013 Jul 13]. Available from: www.pro.medicin.dk/Medicin/Praeparater/3237. [20] SSI [internet]: Statens Serum Institut; 2013 [updated 2013 May 2; cited 2013 Jul 13]. Available from: www.ssi.dk/ Diagnostik/DiagnostiskHaandbog/400–499/435.aspx. [21] Ikeuchi D, Onodera H, Aung T, et al. Correlation of tensile strength with bursting pressure in the evaluation of intestinal anastomosis. Dig Surg. 1999;16:478– 485. [22] Rossi LF, Ramos RR, Ely JB, et al. Considerations that may influence the result of trials assessing tensile strength in experimental surgery. Acta Cir Bras. 2007;22:499– 502. [23] Ulmer TF, Stumpf M, Rosch R, et al. Suture-free and mesh reinforced small intestinal anstomoses: a feasibility study in rabbits. J Invest Surg. 2013;26:210–216. [24] Thompson SK, Chang EY, Jobe BA. Clinical review: Healing in gastrointestinal anastomoses, part I. Microsurgery. 2006;26:131–136.
Journal of Investigative Surgery
