S. Yauw1, K. Wever1,2, A. Hoesseini1, H. Van Goor1 1Radboud University Medical Center,Department of Surgery,Nijmegen, GELDERLAND, Netherlands 2Radboud University,2Systematic Review Center For Laboratory Animal Experimentation (SYRCLE),Nijmegen, Gelderland, Netherlands
Introduction:
Anastomotic leak rates have failed to diminish despite extensive experimental research in laboratory animals. This warrants a review of animal experiments on anastomotic healing to provide directions for improving animal research.
Methods:
Animal studies on intestinal anastomotic healing were retrieved by a systematic search in Pubmed and Embase databases. Conference abstracts, posters, reviews, studies in human, studies on other types of anastomoses (e.g. esophagojejunal) and intestinal transplants, studies with other outcomes (e.g. on cancer recurrence) and studies using enterotomy instead of anastomosis were excluded. Full text retrieval was attempted by online searches, by consulting Dutch and German university libraries and by E-mail to the corresponding author. All languages but Japanese were included. Study objective, study conclusion, and animal model were recorded. Reporting quality score (adapted from ARRIVE guidelines) and risk of bias (randomization and blinding) were assessed.
Results:
1342 studies were included, of which 245 were conducted in the US. The number of animal experiments on intestinal anastomosis increases, reaching an average of 53 published per year in the last decade. Most studies were therapeutic trials (n=914; 65%) and risk factor identification studies (n=385; 28%). A positive effect on anastomotic healing was reported in 293 of 345 (85%) articles on ‘experimental therapeutics’. Animal models for healing varied widely in terms of animal species, compromise of healing, intestinal segment and outcome measures. The use of rats has increased in contrast to the use of dogs. Methods to compromise healing are used in 32% and vary between different regimens of chemotherapy, ischemia, radiotherapy, immunosuppressants and more. On average 45% relevant study items, in particular anastomotic complications (31%), the use of antibiotics (76%), sterile surgery (83%) and postoperative analgesics (91%), were not reported. Randomization and blinding of primary outcome assessment increased in the last two decades but is still insufficient in the current decade at 62% and 8% respectively.
Conclusion:
Animal experiments on anastomotic leakage are increasing, contrary to the aim of politics, science and the public to reduce animal research. Reporting and methodological quality must improve to refine animal experiments and possibly increase clinical impact. Selecting the most reliable and reproducible models that have translational value is the next step of our research.