Abstract

Key words : Colorectal surgery, Hydroxyproline, Immunosuppression, Mammalian target for rapamycin inhibitor, Transplant, Wound healing

Introduction

Colorectal surgeries constitute 24% to 27% of all general surgery procedures. Anastomotic leakage, which is observed in 5% to 69% of such procedures, often leads to morbidity and mortality.^1-4 The rate of anastomotic leakage due to inhibition of protein synthesis and cell proliferation has been shown to be higher in immunosuppressed patients.^5,6 However, no studies on how to prevent such leakage in these patients have been performed.

Some immunosuppressed patients are solid-organ transplant recipients who may be given mammalian target for rapamycin (mTOR) inhibitors for immunosuppression. However, mTOR inhibitors negatively affect wound healing because they affect the cell cycle and inhibit T-lymphocyte proliferation. Therefore, use of mTOR inhibitors is not recom­mended, especially during the early posttransplant period.^7-11

Mesenchymal stem cells (MSCs), which are produced in many tissues, can be multiplied many times for further use (5 × 10⁴to 2 × 10⁵ stem cells are obtained from 1 g of fat tissue).^12,13 Mesenchymal stem cells primarily originate from bone marrow and adipose tissue, and these increase angiogenesis, local blood flow, and fibroblast and collagen synthesis in tissue.^12,14,15 Adipose-derived stem cells (ADSCs), a type of MSC that occurs via the separation of subcutaneous adipose tissue, have been shown to have a positive effect on wound healing.¹²

The aim of the present study was to determine whether stem cell therapy has a positive effect on wound healing in rats that underwent colon anastomosis and mTOR inhibitor-based immuno­suppression. If stem cell therapy is shown to have a positive effect on anastomotic healing and can reduce anastomotic leaks, then solid-organ transplant recipients may be allowed to receive mTOR inhibitors during early posttransplant, not only in patients with colon anastomoses, but also in patients with vascular, urinary, and biliary anastomoses.

Materials and Methods

The study protocol was approved by the Baskent University School of Medicine Ethics Committee for Animal Experiments (no. DA 17/04), which is in line with the National Institutes of Health Guide for the Care and Use of Laboratory Animals (NIH Publications No. 8023, revised 1978). Animals were obtained from the Baskent University Production and Research Center. The experiments were performed at the Baskent University School of Medicine Research Unit Laboratories. The study included 60 male Sprague-Dawley rats weighing 315 to 460 g (mean: 368 g). All animals were cared for under optimal standard conditions.

Study model

Animals were divided into 4 groups of 14 rats per group, with rats in all groups undergoing colon anastomosis. The 4 remaining rats were used for retrieval of ADSCs. Group 1 (the control group) underwent colon anastomosis only. Group 2 underwent colon anastomosis and then underwent subserosal ADSC injection into each part of the anastomosis. Group 3 was immunosuppressed with everolimus (Certican 0.25 mg water-soluble tablet; Novartis) before undergoing colon anastomosis. Group 4 rats were immunosuppressed with everolimus before colon anastomosis and then underwent subserosal ADSC injection into each part of the anastomosis. Immunosuppression was maintained by administering everolimus, 2.5 mg/kg per day, via oral gavage. Immunosuppression was initiated 7 days before surgery and continued until the day of euthanasia.

All 4 groups were divided into 2 subgroups of 7 each, according to date of death, with rats in the day 4 subgroup euthanized on posttreatment day 4 and rats in the day 7 subgroup euthanized on post­treatment day 7. After euthanasia and relaparotomy, the anastomotic line was located, and then intra­abdominal adhesion was assessed. After assessment, a 60-mm colon segment was resected with surrounding adherent tissue, with the anastomotic line in the middle. Bursting pressure, tissue hydroxyproline level, histopathology (hematoxylin-eosin, trichrome), and identification of labeled stem cells were evaluated on the anastomotic line.

Intraabdominal adhesions were assessed accor­ding to the method of Houston and Rotstein.¹⁶ Adhesions were classified into 3 categories: mild, moderate, and severe (Figure 1; see description in Table 1). To measure intraluminal pressure, the segment from 3 cm distal to 3 cm proximal to the anastomosis was resected and inflated using saline solution pumped through an infusion pump (Figure 2). The highest intraluminal pressure on the monitor was recorded as bursting pressure.¹⁷ Hydroxyproline concentration in tissue samples was determined according to the method of Reddy and Enwemeka.¹⁸ For histopathological assessment, anastomosed colon segments were prepared by a pathologist blinded to the groups. Appropriate cross-sections 5 μm thick were prepared from tissues embedded in paraffin. These were first stained with hematoxylin and then with hematoxylin-eosin.¹⁹ Tissue samples were assessed for mucosal reepithelialization, fibrosis, ischemic necrosis, vascular proliferation, and inflammation in the anastomosis. Labeled stem cells were identified.²⁰ Both the differentiation of stem cells in endothelial cells and the presence of stem cells adjacent to villi were observed (Figure 3).

Adipose-derived stem cell retrieval
The 4 rats used for ADSC retrieval were anesthetized via intraperitoneal injection and then underwent laparotomy via midline incision. Approximately 20 mL of adipose tissue was harvested from the inguinal area, according to the protocol from Ogawa.¹⁹ To enumerate the cells, 0.1 mL of the fraction was taken and stained with methylene blue. Subserosal injection of ADSCs (4 × 10⁵ stem cells per 0.1 mL) into the proximal part (0.1 mL) and distal part (0.1 mL) of each anastomosis in animals in groups 1 to 4 was performed using a purified protein derivative injector (Figure 4).

Surgical procedure
All rats in groups 1 to 4 were anesthetized via intraperitoneal injection and then underwent laparotomy via midline incision. The descending colon was transected full-thickness. A single-layer continuous colon anastomosis was then performed using a single 7/0 Prolene polypropylene suture. Groups 2 and 4 then received ADSC injection. After anastomosis, 0.2 mL of subserosal ADSCs were injected (in equal volume) to the distal and proximal parts of the anastomosis. After the procedure, the laparotomy was closed. All animals were given analgesic and provided with rat feed and water ad libitum postsurgery.

Data analyses
Data were analyzed using SPSS Statistics for Windows version 17.0. For the equality of variance analysis of continuous dependent numeric variables, the Shapiro-Wilk normality test and the Levene test were used. In cases in which parametric test assumptions were true, 2-way analysis of variance (ANOVA) was applied with 95% confidence intervals. In cases in which parametric test assumptions were not true, Bonferroni correction was made on the probability of type I error. To compare the procedure types in terms of day of euthanasia (posttreatment days 4 and 7), if parametric test assumptions were true, one-way ANOVA was used; if not true, the Kruskal-Wallis variance analysis was used. To compare days of euthanasia in terms of the type of procedure, if parametric test assumptions were true, then t tests were used; if not true, then Mann-Whitney U tests were used. For analysis of categorical variables, the Fisher exact test was used for 2 × 2 contingency tables; for higher dimensional contingency tables, the Fisher-Freeman-Halton exact test was used. For all dependency tests, the probability of type I error was set as α = .05.

Results

Assessment of weight loss
There were no significant differences among animals in the day 4 subgroups with regard to weight loss; however, median weight loss in group 3 was approximately 13.5% versus approximately 5.4% in group 1 (P = .013). In the day 7 subgroups, there was a significant difference in weight loss between groups 3 and 1, with median weight loss in group 3 of 13.4% versus approximately 1.3% in group 1 (P = .001) (Table 2). However, when we compared all rats (treatment groups 1-4) in subgroups day 4 versus subgroups day 7, we observed no significant differences in weight loss (P > .05).

Assessment of the severity of intraabdominal adhesion
There was a strong statistical correlation between the type of procedure and adhesion in rats in the day 4 subgroups (P = .036), with lower rate of mild adhesion but higher rates of moderate and severe adhesion in group 1 compared with results shown in groups 2, 3, and 4. Both the extent and severity of adhesion were less in group 4 (which received stem cells) than in group 3. Statistically, the type of procedure also affected the extent of adhesion in all rats in the day 4 subgroups. In the day 7 subgroups, there was no correlation between the type of procedure and adhesion (P = .349); however, in groups 3 and 4 (which received immunosuppression with everolimus), severe adhesion was not observed (Table 3).

Assessment of tissue hydroxyproline level
The effect of the time posttreatment (euthanasia on day 4 versus day 7) on the hydroxyproline level was significant (P < .001), with mean tissue hydroxyproline levels in rats in the day 7 subgroups higher than in rats in the day 4 subgroups. There was a significant difference in the mean hydroxyproline level between groups 1 and 2, between groups 2 and 3, and between groups 2 and 4 (P < .001) (Table 4).

Assessment of bursting pressures
Differences in mean bursting pressure among rats in the different treatment groups at euthanasia day 4 were significant (P < .001). There was a significant difference in mean bursting pressure between groups 1 and 2 and between groups 2 and 3 (P < .001). The mean bursting pressure in group 2 was higher than in groups 1, 3, and 4.
Differences in mean bursting pressure between rats in the different treatment groups at euthanasia day 7 were not significant (P = .074). Intragroup analysis and intergroup analysis demonstrated a significant difference in mean bursting pressure between the rats euthanized on day 4 versus day 7 in treatment groups 1 to 4 (Table 5).

Pathology assessment
Other than vascular proliferation, there were no significant differences in pathological parameters between the groups (Table 6). No significant differences were shown in vascular proliferation among rats euthanized on day 4 (P = .226); however, there was a significant difference between those euthanized on day 7 (P = .003). Marked vascular proliferation was more common in groups 2 and 4 (ADSC injection groups). Moreover, the correlation between the day of euthanasia and vascular proliferation status was significant in group 2 (P = .005) and group 4 (P = .038), and marked proliferation was more common among rats in groups 2 and 4 that were euthanized on day 7 (Table 6).

Discussion

Anastomosis leakage rates have been reported to range from 1.95% to 14.16% following gastroin­testinal surgery in immunosuppressed patients.^5,6 It is also known that everolimus, an mTOR inhibitor, negatively affects wound healing.²¹ The findings in the present study, especially the observed increase in the hydroxyproline level and bursting pressure, showed that subserosal ADSC injection increased anastomosis healing and reduced anastomosis leakage rate in immunosuppressed rats.

After colorectal surgery, the catabolic rate increases, leading to loss of body weight.²² In the present study, animals had significant weight loss on day 7. The greater percentage of weight loss in group 3 was attributed to the antiproliferative and antimitotic effects of the mTOR inhibitor everolimus on cell populations and its inhibitory effect on the proliferation of various cell types, such as smooth muscle cells.²³ Moreover, immunosuppressed rats administered ADSCs (group 4) had less weight loss, indicating that the negative effect of immunosup­pression on body weight can be compensated for by the anabolic effect of stem cells, as reported in other studies.^24,25

Adhesion after intraabdominal surgery leads to intestinal obstruction and intestinal ischemia.²⁶ In the present study, the severity of adhesion was greater in the immunosuppressed rats, whereas injection of ADSCs partially prevented this increase in severity, especially during the early postsurgery period. Pascual and colleagues²⁷ performed colonic anasto­mosis using adipose-derived MSCs in biosutures, similar to the present study, and observed that use of stem cells significantly reduced the adhesion formation rate. The low adhesion rate during the early postsurgery period might be due to an immunomodulatory effect of stem cells, which suppresses in vitro proliferation of activated lymphocytes.²⁸

In tissues, the hydroxyproline level indirectly shows the collagen level and functions as a collagen precursor; mTOR inhibitors decrease protein synthesis, thus leading to a reduction in the tissue hydroxyproline level and, consequently, a reduction in wound healing.^29-32 In rats, hydroxyproline levels in colonic anastomoses were shown to be significantly decreased after treatment with everolimus.^21,33 Similarly, in the present study, we observed this effect in the immunosuppressed rats (groups 3 and 4) during the early postsurgery period (day 4 subgroups). On the other hand, because of the anabolic effect of stem cells and the mechanism of accelerating wound healing, tissue hydroxyproline levels have been noted to be high in groups injected with ADSCs.³⁴ Surprisingly, during the late postsurgery period (day 7 subgroups) in the present study, ADSC administration did not exhibit the expected effect in terms of the tissue hydroxyproline level. Moreover, in the rats immunosuppressed with everolimus, the tissue hydroxyproline level was higher than in the control group during the late postsurgery period. In their study of the effects of rapamycin on wound healing, Ekici and colleagues¹¹ reported that tissue hydroxyproline levels in immunosuppressed rats were not significantly different from levels in rats that were not immunosuppressed. Therefore, we suggest that the mechanisms of action of the mTOR inhibitor and the stem cells on hydroxyproline synthesis are based on different pathways.

Bursting pressure is an indicator of anastomotic healing and the collagen concentration in the colon. In their study of the effect of ADSCs on healing after ischemic colonic anastomosis, Jong Han Yoo and colleagues³⁵ reported that bursting pressure and collagen concentration were directly proportional. Willems and colleagues³⁶ examined the permanent effects of everolimus against experimental injury to the colon and fascia in rats and observed no significant difference in the collagen concentration between their groups on day 7, whereas bursting pressure was directly correlated with the dose of everolimus. Similar to results reported in the literature, we found that everolimus decreased the bursting pressure, whereas ADSC injection completely reversed this effect during the early postsurgery period. During the early postsurgery period in particular, this effect can be considered to play a very effective role in reducing anastomosis leakage. Although the hydroxyproline level decreased despite ADSC administration in group 4, the bursting pressure increased, suggesting that the anastomotic resistance-increasing effect of stem cell treatment is based on parameters other than the hydroxyproline level.

Our histopathological assessment showed no significant differences in the pathological parameters among the 4 groups, other than vascular prolife­ration, which was attributed to the hypothesis that immunosuppression and/or stem cell injection does not affect the cell proliferation aspect of repair.³⁷ Furthermore, based on the present findings, we suggest that the cause of increased vascular prolife­ration in the groups that received ADSCs was due to the increase in the vascular endothelial growth factor level caused by the stem cells and the ability of stem cells to differentiate into such cell types as endo­thelial cells, as reported by Caziuc and colleagues.¹²

Our present study has some limitations, including the lack of both measurement of the tissue collagen level and measurement of molecular parameters after stem cell injection. In addition, although the experimental animals were immunosuppressed for 7 days before surgery, which may appear to be sufficient for preclinical work, it may not be sufficient to be representative of patients receiving long-term immunosuppressive therapy.

Conclusions

Injection of ADSCs into the anastomosis site following gastrointestinal surgery and solid-organ transplant in immunosuppressed patients may be helpful to reverse the negative effects of immuno­suppression on anastomosis healing. Of note, most of the positive effects of ADSC injection on anastomosis occurred only during the early postsurgery period (the day 4 subgroups); however, considering that most instances of anastomosis leakage develop during the early postoperative period, ADSC injection could be used to reduce the rate of leakage, thus decreasing patient morbidity and mortality.

References:

1. 1. Schrock TR, Deveney CW, Dunphy JE. Factor contributing to leakage of colonic anastomoses. Ann Surg 1973;177(5):513-518. doi:10.1097/00000658-197305000-00002
   CrossRef - PubMed
   
2. Liu JH, Etzioni DA, O’Connell JB, Maggard MA, Ko CY. The increasing workload of general surgery. Arch Surg 2004;139(4):423-428. doi:10.1001/archsurg.139.4.423
   CrossRef - PubMed
   
3. Nursal TZ, Anarat R, Bircan S, Yildirim S, Tarim A, Haberal M. The effect of tissue adhesive, octyl-cyanoacrylate, on the healing of experimental high-risk and normal colonic anastomoses. Am J Surg 2004;187(1):28-32. doi:10.1016/j.amjsurg.2003.02.007
   CrossRef - PubMed
   
4. Nursal TZ, Bal N, Anarat R, et al. Effects of a static magnetic field on wound healing: results in experimental rat colon anastomoses. Am J Surg 2006;192(1):76-81. doi:10.1016/j.amjsurg.2006.01.024
   CrossRef - PubMed
   
5. Eriksen TF, Lassen CB, Gogenur I. Treatment with corticosteroids and the risk of anastomotic leakage following lower gastrointestinal surgery: a literature survey. Colorectal Dis 2014;16(5):O154-O160. doi:10.1111/codi.12490
   CrossRef - PubMed
   
6. Sims SM, Kao AM, Spaniolas K, et al. Chronic immunosuppressant use in colorectal cancer patients worsens postoperative morbidity and mortality through septic complications in a propensity-matched analysis. Colorectal Dis 2019;21(2):156-163. doi:10.1111/codi.14432
   CrossRef - PubMed
   
7. Laplante M, Sabatini DM. mTOR signaling in growth control and disease. Cell 2012;149(2):274-293. doi:10.1016/j.cell.2012.03.017
   CrossRef - PubMed
   
8. Ueno P, Felipe C, Ferreira A, et al. Wound healing complications in kidney transplant recipients receiving everolimus. Transplantation 2017;101(4):844-850. doi:10.1097/TP.0000000000001392
   CrossRef - PubMed
   
9. Sandrini S, Setti G, Bossini N, et al. Steroid withdrawal five days after renal transplantation allows for the prevention of wound-healing complications associated with sirolimus therapy. Clin Transplant 2009;23(1):16-22. doi:10.1111/j.1399-0012.2008.00890.x
   CrossRef - PubMed
   
10. Dean PG, Lund WJ, Larson TS, et al. Wound-healing complications after kidney transplantation: a prospective, randomized comparison of sirolimus and tacrolimus. Transplantation 2004;77(10):1555-1561. doi:10.1097/01.tp.0000123082.31092.53
    CrossRef - PubMed
    
11. Ekici Y, Emiroglu R, Ozdemir H, Aldemir D, Karakayali H, Haberal M. Effect of rapamycin on wound healing: an experimental study. Transplant Proc 2007;39(4):1201-1203. doi:10.1016/j.transproceed.2007.04.013
    CrossRef - PubMed
    
12. Caziuc A, Calin Dindelegan G, Pall E, Mironiuc A. Stem cells improve the quality of colonic anastomoses - A systematic review. J BUON 2015;20(6):1624-1629.
    CrossRef - PubMed
    
13. Volz AC, Huber B, Kluger PJ. Adipose-derived stem cell differentiation as a basic tool for vascularized adipose tissue engineering. Differentiation 2016;92(1-2):52-64. doi:10.1016/j.diff.2016.02.003
    CrossRef - PubMed
    
14. Adas G, Kemik O, Eryasar B, Okcu A, Adas M, Arikan S, et al. Treatment of ischemic colonic anastomoses with systemic transplanted bone marrow derived mesenchymal stem cells. Eur Rev Med Pharmacol Sci 2013;17:2275-2285.
    CrossRef - PubMed
    
15. Gong W, Guo M, Han Z, et al. Mesenchymal stem cells stimulate intestinal stem cells to repair radiation-induced intestinal injury. Cell Death Dis 2016;7(9):e2387. doi:10.1038/cddis.2016.276
    CrossRef - PubMed
    
16. Houston KA, Rotstein OD. Fibrin sealant in high-risk colonic anastomoses. Arch Surg 1988;123(2):230-234. doi:10.1001/archsurg.1988.01400260118015
    CrossRef - PubMed
    
17. Shikata J, Shida T. Effects of tension on local blood flow in experimental intestinal anastomoses. J Surg Res 1986;40(2):105-111. doi:10.1016/0022-4804(86)90110-1
    CrossRef - PubMed
    
18. Reddy GK, Enwemeka CS. A simplified method for the analysis of hydroxyproline in biological tissues. Clin Biochem 1996;29(3):225-229. doi:10.1016/0009-9120(96)00003-6
    CrossRef - PubMed
    
19. Ogawa R. The importance of adipose-derived stem cells and vascularized tissue regeneration in the field of tissue transplantation. Curr Stem Cell Res Ther 2006;1(1):13-20. doi:10.2174/157488806775269043
    CrossRef - PubMed
    
20. Ogawa R, Mizuno H, Hyakusoku H, Watanabe A, Migita M, Shimada T. Chondrogenic and osteogenic differentiation of adipose-derived stem cells isolated from GFP transgenic mice. J Nippon Med Sch 2004;71(4):240-241. doi:10.1272/jnms.71.240
    CrossRef - PubMed
    
21. van der Vliet JA, Willems MC, de Man BM, Lomme RM, Hendriks T. Everolimus interferes with healing of experimental intestinal anastomoses. Transplantation 2006;82(11):1477-1483. doi:10.1097/01.tp.0000246078.09845.9c
    CrossRef - PubMed
    
22. Ryan NT. Metabolic adaptations for energy production during trauma and sepsis. Surg Clin North Am 1976;56(5):1073-1090. doi:10.1016/s0039-6109(16)41032-7
    CrossRef - PubMed
    
23. Sehgal SN. Sirolimus: its discovery, biological properties, and mechanism of action. Transplant Proc. 2003;35(3 Suppl):7S-14S. doi:10.1016/s0041-1345(03)00211-2
    CrossRef - PubMed
    
24. Zeng Q, Li X, Beck G, Balian G, Shen FH. Growth and differentiation factor-5 (GDF-5) stimulates osteogenic differentiation and increases vascular endothelial growth factor (VEGF) levels in fat-derived stromal cells in vitro. Bone. 2007;40(2):374-381. doi:10.1016/j.bone.2006.09.022
    CrossRef - PubMed
    
25. Cao Y, Sun Z, Liao L, Meng Y, Han Q, Zhao RC. Human adipose tissue-derived stem cells differentiate into endothelial cells in vitro and improve postnatal neovascularization in vivo. Biochem Biophys Res Commun. 2005;332(2):370-379. doi:10.1016/j.bbrc.2005.04.135
    CrossRef - PubMed
    
26. Stommel MWJ, Ten Broek RPG, Strik C, et al. Multicenter observational study of adhesion formation after open-and laparoscopic surgery for colorectal cancer. Ann Surg. 2018;267(4):743-748. doi:10.1097/SLA.0000000000002175
    CrossRef - PubMed
    
27. Pascual I, de Miguel GF, Gomez-Pinedo UA, de Miguel F, Arranz MG, Garcia-Olmo D. Adipose-derived mesenchymal stem cells in biosutures do not improve healing of experimental colonic anastomoses. Br J Surg. 2008;95(9):1180-1184. doi:10.1002/bjs.6242
    CrossRef - PubMed
    
28. Puissant B, Barreau C, Bourin P, et al. Immunomodulatory effect of human adipose tissue-derived adult stem cells: comparison with bone marrow mesenchymal stem cells. Br J Haematol. 2005;129(1):118-129. doi:10.1111/j.1365-2141.2005.05409.x
    CrossRef - PubMed
    
29. Kuper MA, Scholzl N, Traub F, et al. Everolimus interferes with the inflammatory phase of healing in experimental colonic anastomoses. J Surg Res. 2011;167(1):158-165. doi:10.1016/j.jss.2009.07.013
    CrossRef - PubMed
    
30. Poffenbarger PL, Haberal MA. Role of serum nonsuppressible insulin-like activity (NSILA) in wound healing. I. Influence of thyroparathyroidectomy on serum NSILA and wound healing in the rat. Surgery. 1976;80(5):608-616.
    CrossRef - PubMed
    
31. Akinbingol G, Borman H, Maral T, et al. Wound healing at adaptation zones of skin flaps harvested from acute burned skin. Burns. 2013;39(6):1206-1211. doi:10.1016/j.burns.2013.01.011
    CrossRef - PubMed
    
32. Ozcelik U, Ekici Y, Bircan HY, et al. Effect of topical platelet-rich plasma on burn healing after partial-thickness burn injury. Med Sci Monit. 2016;22:1903-1909. doi:10.12659/msm.895395
    CrossRef - PubMed
    
33. Kuper MA, Trutschel S, Weinreich J, Konigsrainer A, Beckert S. Growth hormone abolishes the negative effects of everolimus on intestinal wound healing. World J Gastroenterol 2016;22(17):4321-4329. doi:10.3748/wjg.v22.i17.4321
    CrossRef - PubMed
    
34. Uysal AC, Mizuno H, Tobita M, Ogawa R, Hyakusoku H. The effect of adipose-derived stem cells on ischemia-reperfusion injury: immunohistochemical and ultrastructural evaluation. Plast Reconstr Surg 2009;124(3):804-815. doi:10.1097/PRS.0b013e3181b17bb4
    CrossRef - PubMed
    
35. Yoo JH, Shin JH, An MS, et al. Adipose-tissue-derived stem cells enhance the healing of ischemic colonic anastomoses: an experimental study in rats. J Korean Soc Coloproctol 2012;28(3):132-139. doi:10.3393/jksc.2012.28.3.132
    CrossRef - PubMed
    
36. Willems MC, van der Vliet JA, de Man BM, van der Laak JA, Lomme RM, Hendriks T. Persistent effects of everolimus on strength of experimental wounds in intestine and fascia. Wound Repair Regen 2010;18(1):98-104. doi:10.1111/j.1524-475X.2009.00558.x
    CrossRef - PubMed
    
37. Wagner OJ, Inglin RA, Bisch-Knaden S, et al. Sirolimus and intraoperative hyperthermic peritoneal chemoperfusion with mitomycin-C do not impair healing of bowel anastomoses. Transpl Int 2008;21(6):554-563. doi:10.1111/j.1432-2277.2007.00635.x
    CrossRef - PubMed