Investigating the risk factors of hepatocellular carcinoma and survival analysis for cirrhosis after transjugular intrahepatic portosystemic shunt in treating portal hypertension.

Objective: This study aimed to explore the risk factors of hepatocellular carcinoma (HCC) and survival analysis for cirrhosis after transjugular intrahepatic portosystemic shunt (TIPS) in treating portal hypertension.

Materials and Methods: A retrospective database review was performed, including 106 patients (33 females and 73 males; aged 26-68 years with mean age of 55.3 ± 9.1 years) who received TIPS for treating recurrent gastroesophageal variceal bleeding or refractory ascites with portal hypertension. All the patients were recruited from the Interventional Oncology Department at Beijing Ditan Hospital between October 2008 and December 2011. The TIPS was successfully performed on all involved patients by puncturing at the right branch of portal vein via right hepatic vein. After TIPS, the patients were consecutively followed up at the outpatient clinic. The patients were examined by contrast-enhanced computed tomography (CT) or magnetic resonance imaging (MRI) of the abdomen every 3 months for 3 years, for detecting the occurrence of malignant nodules and early HCC. The post-TIPS medical database was reviewed with univariate and multivariate analysis to identify the risk factors for new nodules retrospectively.

Results: The overall incidence of HCC was 38.7% (41/106). The multivariate analysis showed that an increased odds ratios (OR) of HCC was observed in the complication of portal hypertension (OR, 0.396; 95% confidence interval [CI], 0.171-0.918; P = 0.031) and preoperative cirrhosis classification (OR, 0.060; 95% CI, 0.021-0.175; P = 0.000). P < 0.05was considered statistically significant. After TIPS, the cumulative probabilities of survival time for patients with cirrhosis at 1, 2, and 3 years were 100%, 68%, and 61%, respectively (log rank test, P = 0.18). The cumulative incidence of new nodules was significantly lower among patients with refractory ascites than those with upper gastrointestinal hemorrhage. Specifically, the survival rates of patients with upper gastrointestinal hemorrhage at years 1, 2, and 3 were 100%, 65%, and 51%, respectively, compared to 100%, 88%, and 85% corresponding to patients with refractory ascites (P = 0.009). The cumulative incidence of HCC was significantly lower in cirrhosis patients with CT identified grade III than those with grade IV. At years 1, 2, and 3, the survival rates of cirrhosis patients with CT identified grade IV were 96%, 22%, and 20%, respectively, compared to 100%, 98%, and 90% in controls (P = 0.012).

Conclusions: The identification of clinical variables associated with increased risks of HCC may be useful for selecting appropriate candidates for TIPS. Results suggested that the patients with cirrhosis of CT identified grade IV and with upper gastrointestinal hemorrhage might be relevant to increased odds of HCC after TIPS.