Acute in-vivo evaluation of bleeding with Gelfoam^TM plus saline and Gelfoam plus human thrombin using a liver square lesion model in swine

Background Management of post-operative bleeding has historically used topical bovine thrombin. However, possible harm through activation of coagulation inhibitors has encouraged investigation with other hemostatic agents. This study utilized a novel ordinal bleeding model to test whether a Gelfoam + human thrombin solution is superior at controlling bleeding when compared to Gelfoam + saline solution at different time intervals. Study design Four swine underwent open laparotomy after receiving unfractionated heparin. Twenty open liver biopsies were performed in each swine; ten biopsies treated with Gelfoam + human thrombin solution and 10 biopsies treated with Gelfoam + saline solution. Three, 6 min, and 12 min after the procedure, bleeding was objectively graded by a four-point model. Results There was a significant (P < 0.017), treatment effect on each success/failure outcome (success = bleeding score ≤ 1; failure = bleeding score > 1) at 3 (P < 0.001), 6 (P < 0.001), and 12 (P = 0.003) min, based on a 2 × 2 Fisher’s exact test. Similarly, there was a significant treatment effect on each success/failure outcome and four-point bleeding score based on a multiple logistic regression analysis controlling for pig, lesion weight, and initial bleeding taking into consideration repeated measures at three time points. Conclusions The results demonstrate a superior treatment effect for control of bleeding using human thrombin compared to a saline solution. Future studies should compare bovine thrombin versus human thrombins ability to control bleeding as well as the hazard of each in activating coagulation inhibitors.