肝癌伴胆管癌栓术后肝功能衰竭的危险因素分析

Risk factors for postoperative liver failure of patients with hepatocellular carcinoma and bile duct tumorthrombus

目的探讨肝癌伴胆管癌栓术后肝功能衰竭的危险因素，建立术后肝功能衰竭的风险评估模型。方法回顾性分析第二军医大学附属东方肝胆外科医院2002年3月至2011年2月收治的107例接受肝癌切除术的肝癌伴胆管癌栓患者的临床资料。根据术后是否发生肝功能衰竭，将患者分为无肝功能衰竭组（98例）和肝功能衰竭组（9例）进行队列研究。对围手术期可能与肝功能衰竭发生相关的多种因素进行分析，筛选肝癌伴胆管癌栓术后肝功能衰竭的危险因素，并建立肝功能衰竭的风险预测模型。单因素分析采用Logistic二元回归模型，筛选获得有统计学意义的指标纳入Logistic多元回归模型进行多因素分析。结果107例患者中105例行肝癌切除＋胆总管切开取栓术，2例行肝癌切除＋肝外胆管切除＋胆肠吻合术；手术时间为2．0～5．5h；术中出血量为200～3500ml。无肝功能衰竭组患者中，胸、腹腔积液5例，胆道出血3例，切口感染2例，胆道感染、胆汁漏、上消化道应激性溃疡、胸椎硬膜外血肿各1例。胸椎硬膜外血肿患者经胸椎减压止血治疗后出血停止，但遗留截瘫；其余患者经过对症、支持治疗后痊愈。肝功能衰竭组患者中，2例因术后急性肝功能衰竭抢救无效死亡，7例因术后亚急性肝功能衰竭死亡（排除因肿瘤复发或药物因素死亡）。单因素分析结果表明：术前TBil、Alb、Pre-Alb、白球比值（A／G），癌栓分布及术中出血量和术后剩余肝脏体积占全肝体积比与肝癌伴胆管癌栓患者术后发生肝功能衰竭相关（OR=3．017，0．191，0．248，2．681，9．048，4．759，13．714，P〈0．05）。多因素分析结果显示：术前TBil〉256．5μmoL／L、术前A／G≤1．3和术后剩余肝脏体积占全肝体积比〈50％是肝癌伴胆管癌栓患者术后发生肝功能衰竭的独立危险因素（OR=5．537，11．107，172．450，P〈0．05）。术后肝功能衰竭风险预测模型为Z=1．711×（术前TBil）＋2．408x（术前A／G）＋5．150x（术后剩余肝脏体积占全肝体积比）一17．288，Z值越大，术后发生肝功能衰竭的预期风险越高；Z值〉0时，术后发生肝功能衰竭的预期风险〉50％。结论术前TBil〉256．5μmol／L、术前A／G≤1．3、术后剩余肝脏体积占全肝体积比〈50％是肝癌伴胆管癌栓患者术后发生肝功能衰竭的独立危险因素。采用肝功能衰竭风险预测模型对肝癌伴胆管癌栓患者进行有效的筛选，可降低术后肝功能衰竭的发生率。

Objective To investigate the risk factors for postoperative liver failure of patients with hepato- cellular carcinoma （ HCC ） and bile duct tumor thrombus through a risk evaluation model. Methods The clinical data of 107 patients with HCC and bile duct tumor thromhus who received hepatic resection at the Eastern Hepato- biliary Surgery Hospital from March 2002 to February 2011 were retrospectively analyzed. All patients were divided into the non-liver failure group （98 patients ） and liver failure group （9 patients ）. Risk factors associated with liver failure were analyzed and a risk evaluation model was established. All data were analyzed using the bivariate regression model, and factors with significance were further analyzed using the multivariate regression model. Results Of the 107 patients, 105 received hepatic resection ＋ choledochotomy ＋ thrombectomy and 2 received hepatic resection ＋ extrahepatie bile duct resection ＋ cholangiojejunostomy. The operation time was 2.0-5.5 hours, and the intraoperative blood loss was 200-3500 ml. In the non-fiver failure group, 5 patients had pleural andperitoneal effusion, 3 had biliary bleeding, 2 had incisional infection, 1 had biliary infection, 1 had bile leakage, 1 had stress-induced ulcer of upper digestive tract and 1 had thoracic epidural hematoma. The bleeding of the patients with thoracic epidural hematoma was stopped after thoracic spinal decompression, but subsequent paraplegia occurred. In the liver failure group, 2 patients died of postoperative acute liver failure, and 7 patients died of postoperative subacute liver failure （ death caused by tumor recurrence or medicine was excluded）. The results of univariate analysis showed that preoperative total bilirubin, albumin, pre-albumin, albumin/globulin ratio, distribution of tumor thrombus, operative blood loss and ratio of postoperative residual liver volume to the total liver volume were correlated with the postoperative liver failure in patients with HCC and bile duct tumor thrombus （ OR = 3. 017, 0. 191,0. 248, 2. 681, 9. 048, 4. 759, 13. 714, P 〈 0.05）. The results of multivariate analysis showed that preoperative total bilirubin 〉 256.5 μmol/L, albumin/globulin ratio≤ 1.3 and postoperative residual liver volume 〈 50% were the independent risk factors of postoperative liver failure （ OR = 5. 537, 11. 107, 172. 450, P 〈 0.05 ）. The risk evaluation model was Z = 1.77 x preoperative total bilirubin ＋ 2.408 x preoperative albumin/globulin ratio ＋5. 150 x ratio of postoperative residual liver volume to the total liver volume-17. 288. The risk of postoperative liver failure increased as the increase of Z value. The risk of postoperative liver failure 〉 50% when the Z value 〉 0. Conclusions Preoperative total bilirubin 〉 256.5μmol/L, albumin/globulin ratio ≤ 1.3 and postoperative residual liver volume 〈 50% were the independent risk factors of postoperative liver failure. Risk evaluation model is helpful in screening the risk factors so as to decrease the incidence of postoperative liver failure.