胰十二指肠切除术后晚期出血的诊断与治疗

目的总结胰十二指肠切除术后晚期出血的诊断与治疗经验。方法回顾性分析2002年1月至2013年2月新疆医科大学第一附属医院收治的246例行胰十二指肠切除术患者的临床资料。胰头及壶腹部恶性肿瘤行标准胰十二指肠切除术或联合脏器切除，良性肿瘤及十二指肠乳头肿瘤行保留幽门的胰十二指肠切除术。消化道吻合采用胰肠或胰胃吻合两种方式。患者术后出血时间〉5d定义为晚期出血。消化道出血为消化道出血组，腹腔出血为腹腔出血组。按出血程度分为轻度和重度出血。采取保守治疗和手术治疗（包括介入和开腹手术治疗）两种方法治疗晚期出血。计数资料组间比较采用Fisher确切概率法。结果246例患者中行标准胰十二指肠切除术224例，行保留幽门的胰十二指肠切除术10例，行胰十二指肠切除联合门静脉切除或置换术9例，行胰十二指肠切除联合肠系膜上静脉置换术1例，行胰十二指肠切除联合肝方叶切除术1例，行胰十二指肠切除联合左半肝切除术1例。246例患者中行改良胰肠端侧吻合127例，行胰胃套入吻合53例，行传统胰肠端端套人吻合39例，行胰管空肠黏膜对黏膜吻合27例。患者围手术期死亡15例，病死率为6．10％（15／246）。术后29例患者发生晚期出血，出血发生率为11．79％（29／246）。其中消化道出血14例，腹腔出血15例。29例出血患者中轻度出血9例（消化道出血5例、腹腔出血4例）；重度出血20例（消化道出血9例、腹腔出血11例）。17例患者术后发生先兆出血，其中消化道出血5例、腹腔出血12例。29例患者均经常规保守治疗，消化道出血组患者保守治疗成功率为8／14，腹腔出血组为2／15，两组比较，差异有统计学意义（P〈0．05）。保守治疗失败患者均中转手术治疗。20例重度出血患者中行手术治疗19例，1例经保守治疗成功。9例轻度出血患者全部行保守治疗，1例因肺部感染死亡，其余均获治愈。29例术后晚期出血患者中死亡10例，病死率为34．5％（10／29）。消化道出血组患者病死率为2／14，腹腔出血组为8／15，两组比较，差异无统计学意义（P〉0．05）。结论胰十二指肠切除术后晚期出血常有先兆出血征象，出血程度多为重度。消化道出血经保守治疗多可治愈，腹腔出血需积极手术治疗。

Diagnosis and treatment of delayed postoperative hemorrhage after pancreaticoduodenectomy

Objective To summarize the clinical experience in the diagnosis and treatment of delayed postoperative hemorrhage after pancreatieoduodeneetomy. Methods The clinical data of 246 patients who received pancreatieoduodeneetomy at the First Affiliated Hospital of Xinjiang Medical University from January 2002 to February 2013 were retrospectively analyzed. Patients with pancreatic head carcinoma and ampullary malignan- cies received standard pancreaticoduodeneetomy or combined organ resection, and patients with benign and duode- nal tumor received pylorus-preserving pancreatieoduodenectomy. Pancreatic anastomosis was done using pancreati- eogastrostomy or panereatieojejunostomy. Delayed hemorrhage was defined as bleeding at the operation site after 5 or more postoperative days. Patients were divided into the gastrointestinal hemorrhage group and the abdominal hemorrhage group according to the bleeding site, and the treatment methods included conservative treatment or surgical treatment （including interventional therapy and laparotomy）. The Measurement data were expressed as mean _＋ standard deviation, and the enumeration data were compared using the Fisher exact probability. Results There were 224 patients received panereatieoduodenectomy, 10 received pylorus-preserving panereatieoduodenectomy, 9 received panereaticoduodeneetomy ＋ portal vein resection or replacement, 1 received pancreaticoduodenectomy ＋ superior mesenterie vein replacement, 1 received panereatieoduodenectomy ＋ resection of quadrate lobe of liver and1 received pancreaticeduedenectomy ＋ resection of left semihepatectomy. There were 127 patients received improved end-to-side pancreaticojejunostomy, 53 received pancreaticogastrostomy, 39 received end-to-end pancreati- cojejunostomy and 27 received mucosa-to-mucosa pancreaticojejunostomy. Fifteen patients died perioperatively, with the mortality rate of 6. 10% （15/246）. Twenty-nine patients were complicated with delayed hemorrhage with the rate of 11,79% （29/246）, including 14 with gastrointestinal hemorrhage and 15 with abdominal hemorrhage. There were 9 patients with mild hemorrhage （5 with gastrointestinal hemorrhage and 4 with abdominal hemorrhage） and 20 with severe hemorrhage （9 with gastrointestinal hemorrhage and 11 with abdominal hemorrhage）. Seventeen patients had sentinel bleeding, including 5 with abdominal hemorrhage and 12 with gastrointestinal hemorrhage. Twenty-nine patients received conservative treatment, and the success rates of conservative treatment were 8/14 in the gastrointes- tinal hemorrhage group, which was significantly higher than 2/15 of the abdominal hemorrhage group （ P 〈0.05 ）. Patients who failed in conservative treatment received surgical treatment. Of the 20 patients with severe hemorrhage, 19 were cured by surgical treatment and 1 was cured by conservative treatment. Nine patients with mild hemorrhage received conservative treatment, 8 were cured and 1 died of pulmonary infection. Of the 29 patients with delayed postoperative hemorrhage, 10 died with the mortality rate of 34.5% （10/29）. The morality rate of the gastrointes- tinal hemorrhage group was 2/14, which was lower than 8/15 of the abdominal hemorrhage group, with no signifi- cant difference between the 2 groups （P 〉 0.05 ）. Conclusions Most of the delayed postoperative hemorrhage severe and combined with sentinel bleeding. Gastrointestinal hemorrhage can be cured through conservative treatment, but abdominal hemorrhage need surgical treatment.