直肠癌前切除术后吻合口漏危险因素分析及CT诊断的作用

Risk factors of anastomotic leakage after anterior resection for rectal cancer and the diagnostic value of computed tomography signs in anastomotic leakage

目的探讨直肠癌前切除（Dixon术）术后吻合口漏的危险因素以及CT诊断吻合口漏的准确性。方法采用回顾性病例对照研究方法，分析2013年1月至2015年6月间北京大学第一医院普通外科收治的452例直肠癌患者Dixon术后的临床及病理资料。分析患者性别、年龄、体质指数（BMI）、有无糖尿病、术前有无低血红蛋白血症（Hb〈90g／L）、术前有无低白蛋白血症（Alb〈35g／L）、肿瘤下缘距肛缘距离、肿瘤直径、肿瘤分化程度、肿瘤分期、有无新辅助治疗、手术方式、术中是否保留左结肠动脉以及是否行预防性性造口对Dixon术后发生吻合口漏的影响。单因素分析采用x。检验或Fisher确切概率法检验，多因素分析采用有序分类自变量logistic回归。结果全部患者中，男性281例，女性171例，中位年龄64（18～88）岁。其中有47例（10．4％）术后发生吻合口漏，诊断吻合口漏的中位时间是术后6．5（3—31）d。出现吻合口漏后，1例因呼吸衰竭，在术后1个月内死亡；11例行二次挽救性造口手术治疗，第2次保护性造口手术距离第1次手术的中位时间为7（2～34）d，术后均恢复顺利，并均在术后2年内完成造口还纳术；其余35例予以抗感染、腹腔冲洗等保守治疗并获得痊愈。无吻合口漏组的术后住院时间为（8．4±2．4）d。明显短于术后吻合口漏组（34．6±15．7）d，差异有统计学意义（t=24．127，P=0．008）。单因素分析结果显示：BMI≥28kg／m^2（χ^2=7．550，P=0．000）、糖尿病（χ^2=5．055，P=0．025）、术前Hb〈90g／L（χ^2=5．718，P=0．017）、术前Alb〈35g／L（χ^2=8．096，P=0．004）、肿瘤下缘距肛缘距离〈6cm（χ^2=8．205，P=0．004）及术中结扎左结肠动脉（χ^2=16．540，P=0．000）与Dixon术后吻合口漏发生有关。多因素分析显示：BMI≥28kg／m^2（OR=1．758，95％CI：1．265～2．454，P=0．021）、肿瘤下缘距肛缘距离〈6cm（OR=1．530，95％CI：1．035～2．117，P=0．037）和术中结扎左结肠动脉（OR=1．551，95％CI：1．035～2．131，P=0．042）是影响Dixon术后发生吻合口漏的独立危险因素。CT检查Dixon术后吻合口漏的灵敏度为91．2％（31／34），术后7dCT检查吻合口漏的假阳性率为0。结论BMI≥28kg／m^2,肿瘤下缘距肛缘距离〈6cm以及结扎左结肠动脉的直肠癌患者Dixon术后易出现吻合口漏；推荐存在上述危险因素的患者于术后7d行CT检查确认吻合口愈合情况。

Objective To investigate the risk factors and computed tomography （CT） diagnostic accuracy of anastomotic leakage after resection of rectal cancer （Dixon）. Methods This retrospective study was conducted in Peking University First Hospital from January 2013 to June 2015. A cohort of 452 patients with rectal cancer was enrolled in the study. All the patients underwent anterior resection. The relationship between clinical-pathological data （including sex, age, body mass index （BMI）, presence of diabetes, hypohemoglobin （Hb 〈 90 g/L）, hypoalbuminemia （Alb 〈 35 g/L）, the distance from the lower edge of the tumors to the anus, tumor diameter, tumor differentiation, tumor TNM stage, neoadjuvant therapy status, ligation of the left colonic artery （LCA）, preventive colostomy, and anastomotic leakage was analyzed retrospectively. Univariate analysis using χ^2 test and multivariate analysis by using the Ordered Classification Arguments Logistic regression model. Results Of all the cases, 281 and 171 patients were men and women, respectively. The median age was 64 years （range, 18-88 years）. Forty-seven patients （10.4%） were diagnosed with anastomotic leakage, and the median diagnostic time of anastomotic leakage was 6.5 days （range, 3-31 days）. One patient with anastomotic leakage died because of respiratory failure within 1 month postoperatively ; 11 patients underwent salvage colostomy performed 2-34 days （median, 7 days） after the first surgery. All the 11 patients underwent colostomy closure within 2 years. The other 35 patients recovered by antibiotic and peritoneal lavage treatment. The mean length of postoperative hospital stay in patients without anastomotic leakage was 8.4±2.4 days, which was significantly shorter than that in patients with anastomotic leakage （34.6±15.7 days）, and the difference was statistically significant （t = 24.127, P = 0.008）. The results of the univariate analysis showed that BMI≥28 kg/m^2（χ^2 = 7.550, P = 0.000）, diabetes mellitus （χ^2= 5.055, P = 0.025）, Hb 〈 90 g/Lpreoperatively （χ^2= 5.718, P = 0.017）, Alb 〈 35 g/L preoperatively （χ^2= 8.096, P = 0.004）, distance of 〈 6 cm from the lower edge of the tumors to the anus （χ^2= 8.205, P= 0.004） and LCA ligation （χ^2 = 16.540, P = 0.000） were risk factors for the occurrence of anastomotic leakage. Multivariate analysis showed that BMI≥28 kg/m^2 （OR = 1.758, 95%CI： 1.265- 2.454, P = 0.021）, distance of 〈 6 cm from the lower edge of the tumors to the anus （OR=1.530, 95%CI： 1.035-2.117, P= 0.037）, LCA ligation （OR = 1.551, 95%CI： 1.035-2.131, P= 0.042） were independent risk factors for anastomotic leakage. The CT diagnostic sensitivity of anastomotic leakage was 91.2%（31/34）. The false positive rate of CT for diagnosing anastomotic leakage was zero 7 days after the Dixon procedure. Conclusion Important factors, including BMI of patients, LCA ligation, and the distance from the lower edge of the tumors to the anus are related with anastomotic leakage. The individual treatments should be considered based on the patient＇s clinical condition. CT was recommended 7 days postoperatively when anastomotic leakage was highly suspected.