Dependency of tissue necrosis on gelatin sponge particle size after canine hepatic artery embolization

Purpose: To determine the optimal size of gelatin sponge particles (GSPs) to produce maximum tumor necrosis with minimum side effects after canine hepatic artery embolization (HAE). Methods: GSPs were separated into four size ranges: A, up to 200 mu m (mean 152) as Gelfoam powder; B, 200-500 mu m (mean 336) as Gelfoam powder; C 500-1000 mu m (mean 649) as Spongel; and D, 1000-2000 mu m (mean 1382) as Spongel, Three mongrel dogs were assigned randomly to HAE with each particle size. On day 7 after HAE, the livers were removed and subjected to pathological examination. Results: The mean volume of liver necrosis was 11% after embolization, with particle size A, 36.3% with B, 0% with C, and 1% with D, Coagulation necrosis was found in all livers with particles of sizes A and B, and in 1 of 6 with sizes C and D. Bile duct injury was found in five of six dogs with sizes A and B and in none with sizes C and D. Gallbladder necrosis was found in one dog with size B and pancreas necrosis in one with size A. Conclusion: GSPs of 500 mu m are considered optimally effective for tissue necrosis according to this model.