Effectiveness of a thrombin-gelatin flowable for treating severe liver bleeding: an experimental study

Abstract

Abstract Background Current scientific evidence has pointed out the relevance of hemostatic products for improving clinical outcomes in liver trauma, including increased survival rates and reductions in bleeding-related complications. The purpose of this study was to compare the use of the gelatin-thrombin flowable (Flowable) versus the standard technique of Packing in a new experimental liver injury model. Methods Twenty-four swine were prospectively randomized to receive either Flowable or standard packing technique. We used a novel severe liver injury model, in which the middle and left suprahepatic veins were selectively injured, causing an exsanguinating hemorrhage. The main outcome measure was the percentage of lost blood volume. Results The median total percentage of total blood volume per animal lost, from injury to minute 120, was significantly lower in the Flowable group (15.2%; interquartile range: 10.7–46.7%) than in the Packing group (64.9%; Interquartile range: 53.4–73.0%) (Hodges-Lehmann median difference: 41.1%; 95% CI: 18.9–58.0%, p = 0.0034). The 24-hour survival rate was significantly higher in the Flowable group (87.0%) than in the Packing group (0.0%) (Hazard ratio (HR) 0.08; 95% confidence interval 0.102 to 0.27; p < 0.0001). Mean-arterial pressure was significantly lower at minute 60 and 120 in the Flowable group than in the packing group (p = 0.0258 and p = 0.0272, respectively). At minute 120, hematocrit was higher in the Flowable than in the packing group (Hodges-Lehmann median difference: 5.5%; 95%CI: 1.0 to11.0, p = 0.0267). Finally, the overall-surgical-procedure was significantly shorter with Flowable than with Packing (Hodges-Lehmann median difference: 39.5 s, 95% CI: 25.0 to 54.0 s, p = 0.0004). Conclusions The use of the Flowable was more effective in achieving hemostasis, reducing blood loss, and improving survival rates than standard packing in a severe porcine-liver bleeding model.