Experimental evaluation of clinical colon anastomotic leakage

Colorectal anastomotic leakage remains a frequent and serious complication in gastrointestinal surgery. Patient and procedure related risk factors for anastomotic leakage have been identified. However, the responsible pathophysiological mechanisms are still unknown. Among these, ischemia and insufficient surgical technique have been suggested to play a central role. Animal models are valuable means to evaluate pathophysiological mechanisms and may be used to test preventive measures aiming at reducing the risk of anastomotic leakage, such as external anastomotic coating. The aim of this thesis was to: Clarify the best suited animal to model clinical anastomotic leakage in humans; Create animal models mimicking anastomotic leakage in humans induced by insufficient surgical technique and tissue ischemia; Determine the best suited coating materials to prevent anastomotic leakage.

This study is a systematic review using the databases MEDLINE and Rex. MEDLINE was searched up to October 2010 to identify studies on experimental animal models of clinical colon anastomotic leakage. From the Rex database, textbooks on surgical aspects as well as gastrointestinal physiology and anatomy of experimental animals were identified. The results indicated that the mouse and the pig are the best suited animals to evaluate clinical anastomotic leakage. However, the pig model is less validated and more costly to use compared with the mouse. Most frequently, rats are used as models. However, extreme interventions are needed to create clinical leakage in these animals. The knowledge from this study formed the basis for selecting the animal species most suited for the models in the next studies. STUDY 2: In this experimental study, technically insufficient colonic anastomoses were performed in 110 C57BL/6 mice. The number of sutures in the intervention group was reduced to produce a suitable leakage rate. Moreover, the analgesia and suture material were changed in order to optimize the model. In the final experiment, the four-suture anastomoses resulted in a 40% leakage rate in the intervention groups, whereas the eight-suture control anastomoses had a 0% leakage rate. Furthermore, the use of absorbable suture together with voluntarily ingested Temgesic in chocolate spread as analgesic regimen were feasible. This model may be used to test the leakage reducing potential of coating materials. STUDY 3: This experimental study used 53 C57BL/6 mice, in which sufficient eight-suture anastomoses were created. By using bipolar electro-cautery, blood supply was reduced in a stepwise manner to create anastomotic leakage as a result of ischemia. The study showed that reduced blood supply led to large bowel obstruction instead of clinical leakage. However, anastomotic breaking strength was reduced in the ischemic anastomoses. STUDY 4: In this systematic review MEDLINE, Embase and Cinahl were searched up to September 2011 to identify studies evaluating external coating of colonic anastomoses. Most studies were experimental, in which designs were not comparable and many results were contradictory. In a clinical study, a non-significant benefit of fibrin sealant was found. Based on the available clinical and experimental data it was concluded that the fibrin-based sealants, such as Tisseel and Tachosil, and polyethylene glycols may be beneficial. However, further experimental and clinical studies are needed before routine clinical use can be recommended.

The studies in this thesis may be valuable for the experimental research field of clinical anastomotic leakage. The model of technical insufficiency has been improved and is now thoroughly validated. If used by researchers worldwide, comparison of results is possible. Pure ischemia/anoxia may be too simple an approach to create a clinical leakage model. Thus, future models could focus on multiple risk factors. Conclusively, large-scale clinical multicenter studies are needed to definitively evaluate whether coating of colorectal anastomoses may reduce the leakage rate.