Tetsuya Hasegawa1, Hideki Ota2, Tomonori Matsuura2, Kazumasa Seiji2, Shunji Mugikura2, Fuyuhiko Motoi2, Michiaki Unno2, Kei Takase2. 1. Departments of Diagnostic Radiology and Surgery, Tohoku University Graduate School of Medicine, 1-1 Seiryo-machi, Aoba-ku, Sendai, Miyagi, 980-8574, Japan. Electronic address: te2ya.hase@gmail.com. 2. Departments of Diagnostic Radiology and Surgery, Tohoku University Graduate School of Medicine, 1-1 Seiryo-machi, Aoba-ku, Sendai, Miyagi, 980-8574, Japan.
Abstract
PURPOSE: To evaluate risk factors predicting death and complications of primary therapy for hepatic and gastric duodenal artery pseudoaneurysms following endovascular treatment (EVT) after pancreaticoduodenectomy (PD). MATERIALS AND METHODS: Between April 2004 and December 2014, 28 patients (mean age, 64.7 y) with post-PD hemorrhage underwent EVT. Prevention of hepatic artery blockage via stents or side-holed catheter grafts was stratified in cases without a replaced hepatic artery. Mortality and major hepatic complications following EVT were evaluated according to age; sex; surgery-EVT interval; presence of portal vein stenosis, shock, and coagulopathy at EVT onset; and post-EVT angiographic findings. RESULTS: All hemorrhages were successfully treated with microcoils (n = 17; 61%), covered stents (n = 1; 3%), bare stent-assisted coil embolization (n = 5; 18%), or catheter grafts with coil embolization (n = 5; 18%). Hepatic arterial flow was observed after EVT in 18 patients (64%). Mortality and major hepatic complication rates were 28.6% and 32.1%, respectively. Hemorrhagic shock and coagulopathy at EVT onset (n = 8 each; odds ratio [OR], 27; 95% confidence interval [CI], 3.1-235.7; P < .01) were significantly associated with mortality. Coagulopathy at EVT onset (adjusted OR [aOR], 48.1; 95% CI, 3.2-2,931), portal vein stenosis (n = 16; aOR, 16.9; 95% CI, 1.3-721.9), and no visualization of hepatopetal flow through the hepatic arteries (n = 10; aOR, 29.5; 95% CI, 2.1-1,477) were significantly associated with major hepatic complications. CONCLUSIONS: EVT should be performed as soon as possible before the development of shock or coagulopathy. Hepatic arterial flow visualization decreases major hepatic complications.
PURPOSE: To evaluate risk factors predicting death and complications of primary therapy for hepatic and gastric duodenal artery pseudoaneurysms following endovascular treatment (EVT) after pancreaticoduodenectomy (PD). MATERIALS AND METHODS: Between April 2004 and December 2014, 28 patients (mean age, 64.7 y) with post-PDhemorrhage underwent EVT. Prevention of hepatic artery blockage via stents or side-holed catheter grafts was stratified in cases without a replaced hepatic artery. Mortality and major hepatic complications following EVT were evaluated according to age; sex; surgery-EVT interval; presence of portal vein stenosis, shock, and coagulopathy at EVT onset; and post-EVT angiographic findings. RESULTS: All hemorrhages were successfully treated with microcoils (n = 17; 61%), covered stents (n = 1; 3%), bare stent-assisted coil embolization (n = 5; 18%), or catheter grafts with coil embolization (n = 5; 18%). Hepatic arterial flow was observed after EVT in 18 patients (64%). Mortality and major hepatic complication rates were 28.6% and 32.1%, respectively. Hemorrhagic shock and coagulopathy at EVT onset (n = 8 each; odds ratio [OR], 27; 95% confidence interval [CI], 3.1-235.7; P < .01) were significantly associated with mortality. Coagulopathy at EVT onset (adjusted OR [aOR], 48.1; 95% CI, 3.2-2,931), portal vein stenosis (n = 16; aOR, 16.9; 95% CI, 1.3-721.9), and no visualization of hepatopetal flow through the hepatic arteries (n = 10; aOR, 29.5; 95% CI, 2.1-1,477) were significantly associated with major hepatic complications. CONCLUSIONS:EVT should be performed as soon as possible before the development of shock or coagulopathy. Hepatic arterial flow visualization decreases major hepatic complications.