E. Roozen1, R. Lomme1, H. Van Goor1 1RadboudUMC,General Surgery,Nijmegen, GELDERLAND, Netherlands
Introduction:
Intra-operative or traumatic bleeding are common problems in surgery causing significant morbidity and mortality. Hemostatic products are increasingly recognized as an important measure to control bleeding. There is a need for safe, synthetic, cheap and effective alternatives for the already available hemostatic products. Preclinical testing of these products is predominantly performed in pig liver injury models. We develop hepatic injury models in small animals in order to replace the pig model. These models are our first step to ultimately create ex vivo perfusion models in order to reduce, refine and replace (3R’s) animals. Aim of this study is to explore the consistency in creating a significant bleeding defect of the rat liver to discriminate between different hemostatic patches in terms of efficacy.
Methods:
Two models were evaluated, a liver biopsy punch model and a partial liver resection model . It is known that punch biopsies are reproducible because the surface area and depth have the same dimensions and can be controlled for. Resection models vary more, but have more clinical relevance. Thirty rats were used, in each rat 2 defects were created. In experiment 1, different punch diameters and depths were used to obtain a reproducible and significant bleeding (punch model). In another set of animals partial-lobe resection was performed (resection model). Rats were randomized for either model, receiving an active patch (A1) or a control non-active patch (P1) on both defects. In experiment 2 the feasibility of folding the patch around the edges of the resected area was explored using 2 active patches (A2 and A3).
Outcome for consistency were the severity of bleeding, amount of bloodloss (BL,weight absorbed blood), the surface area of the defect (photodigital planimetry), and the weight of the resected specimen. Outcome for efficacy were time to hemostasis (TTH) and BL after patch application, prior to hemostasis.
Results:
A punch diameter of 8x3mm and the resection created consistent bleedings. 37/60 defects were evaluable for efficacy. There was no difference in TTH (p=0,715) and BL (p=0,440) between the A1 and P1 in the punch model, whereas A1 significantly decreased TTH (p=0,003) and BL (p=0,001) compared to P1 in the resection model. Non-folding was better feasible than folding the patches without differences in TTH and BL. A2 and A3 patches differed in TTH (p:0,02) and BL in both the folding (TTH: p=0,02; BL: p=0.049) and non-folding (TTH: p=0,000 ; BL: p=0.005) application.
Conclusion:
In rats a consistent and easy to operate, partial liver resection model can be created that discriminates between well and worse performing hemostatic products using easy to measure and clinically relevant outcomes.
