A trend toward a higher incidence of hepatocelullar carcinoma (HCC) in patients with cirrhosis treated with bare-stent transjugular intrahepatic portosystemic shunt (TIPS) has been observed in previous studies. To assess the influence of TIPS as a risk factor for developing HCC, we have compared the incidence of HCC in two retrospective cohorts of patients. The TIPS cohort (n = 138) included patients with cirrhosis who underwent TIPS placement for the treatment of portal hypertension-related complications; the non-TIPS cohort was composed of patients admitted at the hospital at the same time of TIPS insertion who were individually matched 1:1 according to age, sex, Child-Turcotte-Pugh class, and cause of cirrhosis. A stratified Cox model was used to assess risk of HCC development. The median time of follow-up was similar in TIPS and non-TIPS cohorts (30.3 [range, 7.8-119.51 and 31.4 [range, 7.8-110.8] months, respectively). The cumulative probability of developing HCC at 1., 3, and 5 years was 3%, 24%, and 34% for the TIPS cohort and 1%, 6%, and 25%, for the non-TIPS cohort, respectively (Breslow test = 5.23, P =.022). The adjusted hazard ratio was 1.52 (95% confidence interval, 1.06-2.19; P =.02). Hepatitis C virus infection and age were independent predictors of HCC development in patients without TIPS. In conclusion, patients with cirrhosis who are treated with TIPS may have a higher incidence of HCC. This observation suggests the need for a strict HCC surveillance program for these patients, especially if they are not expected to undergo a short- or medium-term liver transplantation.
