合并门静脉高压症的原发性肝细胞癌发生上消化道出血的风险因素分析

目的 探讨合并门静脉高压症(portal hypertension,PH)的肝细胞癌(hepatocellular carcinoma,HCC)发生上消化道出血的风险因素.方法 回顾性分析2005年1月1日至2009年8月1日收治的231例HCC-PH临床资料.二分类Logistic回归模型行上消化道出血的风险多因素分析.Kaplan-Meier法计算总体生存时间,作Log-rank检验.ROC曲线评估风险因素预测能力.结果 在231例HCC-PH治疗随访共有28例发生上消化道出血,发生后中位生存时间0.8个月(0.10～2.40个月).发生上消化道出血与无上消化道出血者1、2、3年生存率分别为3.57%(1/28)、0%(0/28)、0%(0/28)和21.18%(43/203)、14.29%(29/203)、4.43%(9/203)(P=0.022,0.031,0.605).多因素分析显示AFP界值(P=0.026)和AST≥2N(P=0.004)是HCC-PH发生上消化道出血的风险因素.ROC曲线显示PI≥7.242时,预测HCC-PH发生上消化道的出血灵敏度为81.0%,特异度81.0%,曲线下面积(Area Under the ROC Curve,AUC)为0.828(95%CI,0.698～0.957).不同风险等级病例之间上消化道出血的发生率差异明显(4/151比24/80,P=0.000).结论 上消化道出血是HCC-PH的终末期事件之一,预后不良.AFP界值和AST≥2N是HCC-PH发生上消化道出血的风险因素.风险因素等级划分有助于针对性的筛查和治疗.

Abstract：

Objective To explore the risk factors for upper gastrointestinal haemorrhage (UGH) in hepatocellular carcinoma (HCC) with portal hypertension (PH). Methods We retrospectively reviewed the medical records of 231 patients with HCC-PH treated in our Department from 1st January 2005 to 1st August 2009. The clinicopathologic factors were evaluated for their possible association with UGH in univariate analysis followed by multivariate analysis using Logistic regression model. The overall survival (OS) was calculated by the Kaplan-Meier method. Receiver operating characteristics (ROC) analysis with calculation of the area under the curve (AUC), sensitivity, and specificity were carried out to assess the predictive ability of the independent risk factors. Results Among 247 patients diagnosed with HCC-PH, 231 patients met the inclusion criteria and were entered into this study. UGH occurred in 28 patients (12.12 %, 28/231). Patients suffering from UGH had a higher 30-and 60-d mortality when compared with the non UGH group (53.57% vs. 4.43%, 96.43%vs. 10.34%, P＜0. 001, 0. 001). The 1-,2-and 3-year overall survival (OS) rates in the non-UGH and the UGH groups were 3. 57% (1/28), 0% (0/28), 0% (0/28) and 21.18% (43/203), 14.29% (29/203), 4.43% (9/203), respectively. There was a trend towards a non-significantly statistical difference in long-term (≥3 yr) survival (P=0. 605). UGH had a dismal prognosis with a median OS of 0. 8 months (0. 10-2. 40 months). Multivariate analysis of the risk factors showed elevated alpha-fetoprotein (AFP) (P = 0. 026) and aspartate aminotransferase (AST) more than twice normal (2N)(P=0. 004) were predictive factors, in particular, AST≥2N. A cutoff value (PI≥7. 242) predicted UGH with an AUC of 0.828 (95%CI, 0.698-0.957), sensitivity of 81.0% and a specificity of 81.0%, as calculated from the ROC. Risk score stratification predicted UGH to show a statistically significant difference (P＜0. 001). Conclusions UGH, as one of the end-stage incidents of HCC-PH,had a dismal prognosis. Patients with elevated AFP levels and AST levels above 2N were associated with high risks for UGH and should be monitored carefully or offered prophylactic treatments. Risk score stratification was useful for prediction of UGH.

Analysis of risk factors for upper gastrointestinal haemorrhage in hepatocellular carcinoma with concurrent portal hypertension

Objective To explore the risk factors for upper gastrointestinal haemorrhage (UGH) in hepatocellular carcinoma (HCC) with portal hypertension (PH). Methods We retrospectively reviewed the medical records of 231 patients with HCC-PH treated in our Department from 1st January 2005 to 1st August 2009. The clinicopathologic factors were evaluated for their possible association with UGH in univariate analysis followed by multivariate analysis using Logistic regression model. The overall survival (OS) was calculated by the Kaplan-Meier method. Receiver operating characteristics (ROC) analysis with calculation of the area under the curve (AUC), sensitivity, and specificity were carried out to assess the predictive ability of the independent risk factors. Results Among 247 patients diagnosed with HCC-PH, 231 patients met the inclusion criteria and were entered into this study. UGH occurred in 28 patients (12.12 %, 28/231). Patients suffering from UGH had a higher 30-and 60-d mortality when compared with the non UGH group (53.57% vs. 4.43%, 96.43%vs. 10.34%, P＜0. 001, 0. 001). The 1-,2-and 3-year overall survival (OS) rates in the non-UGH and the UGH groups were 3. 57% (1/28), 0% (0/28), 0% (0/28) and 21.18% (43/203), 14.29% (29/203), 4.43% (9/203), respectively. There was a trend towards a non-significantly statistical difference in long-term (≥3 yr) survival (P=0. 605). UGH had a dismal prognosis with a median OS of 0. 8 months (0. 10-2. 40 months). Multivariate analysis of the risk factors showed elevated alpha-fetoprotein (AFP) (P = 0. 026) and aspartate aminotransferase (AST) more than twice normal (2N)(P=0. 004) were predictive factors, in particular, AST≥2N. A cutoff value (PI≥7. 242) predicted UGH with an AUC of 0.828 (95%CI, 0.698-0.957), sensitivity of 81.0% and a specificity of 81.0%, as calculated from the ROC. Risk score stratification predicted UGH to show a statistically significant difference (P＜0. 001). Conclusions UGH, as one of the end-stage incidents of HCC-PH,had a dismal prognosis. Patients with elevated AFP levels and AST levels above 2N were associated with high risks for UGH and should be monitored carefully or offered prophylactic treatments. Risk score stratification was useful for prediction of UGH.