Preoperative diagnosis of hepatic angiomyolipoma is difficult, and the treatment for it remains controversial. The aim of this study is to review our experience in the treatment of hepatic angiomyolipoma and to propose a treatment strategy for this disease. We retrospectively collected the clinical, imaging, and pathological features of patients with hepatic angiomyolipoma. Immunohistochemical studies with antibodies for HMB-45, actin, S-100, cytokeratin, vimentin, and c-kit were performed. Treatment experience and long-term follow-up results are summarized. During a period of 9 years, 10 patients with hepatic angiomyolipoma were treated at our hospital. There was marked female predominance (nine patients). Nine patients received surgical resection without complications. One patient received nonoperative management with biopsy and follow-up. One patient died 11 months after surgery because of recurrent disease. We propose all symptomatic patients should receive surgical resection for hepatic angiomyolipoma. Conservative management with close follow-up is suggested in patients with asymptomatic tumors and meet the following criteria: (1) tumor size smaller than 5 cm, (2) angiomyolipoma proved through fine needle aspiration biopsy, (3) patients with good compliance, and (4) not a hepatitis virus carrier.

Most pancreatic leaks after pancreaticoduodenectomy were initially treated conservatively, and for the most part, they were self-limiting and closed spontaneously. But in a few patients, hemorrhage occurred 1 to 3weeks postoperatively and remained a fatal complication after pancreaticoduodenectomy. To identify possible risk factors for massive bleeding, the medical records of 402 patients who had pancreaticoduodenectomies were reviewed for leak complications and analyzed for associations between the occurrence of massive bleeding and preoperative, intraoperative, postoperative, and pathologic parameters. In univariate logistic regression analysis, infectious clinical signs (p=0.021, odds ratio [OR]=7.06, 95% CI=1.34, 37.12) and bile in the drainage fluid (p=0.036, OR=5.89, 95% CI=1.13, 30.86) were associated with the risk of developing massive hemorrhage. This combination provided the highest estimate of massive bleeding after pancreatic leak (p=0.005, OR=8.57, 95% CI=1.92, 38.35). Both infectious clinical signs and bile content in the drainage fluid considerably increase the risk of massive bleeding after pancreatic leak. Close observation of patients with these ominous signs is warranted.