Risk Stratification Based on Chronic Liver Failure Consortium Acute Decompensation Score in Patients With Child-Pugh B Cirrhosis and Acute Variceal Bleeding

Background and Aims Optimal candidates for early transjugular intrahepatic portosystemic shunt (TIPS) in patients with Child-Pugh B cirrhosis and acute variceal bleeding (AVB) remain unclear. This study aimed to test the hypothesis that risk stratification using the Chronic Liver Failure Consortium Acute Decompensation score (CLIF-C ADs) may be useful to identify a subgroup at high risk of mortality or further bleeding that may benefit from early TIPS in patients with Child-Pugh B cirrhosis and AVB. Approach and Results We analyzed the pooled individual data from two previous studies of 608 patients with Child-Pugh B cirrhosis and AVB who received standard treatment between 2010 and 2017 in China. The concordance index values of CLIF-C ADs for 6-week and 1-year mortality (0.715 and 0.708) were significantly better than those of active bleeding at endoscopy (0.633 [P P < 0.001]) and other prognostic models. With X-tile software identifying an optimal cutoff value, patients were categorized as low risk (CLIF-C ADs <48), intermediate risk (CLIF-C ADs 48-56), and high risk (CLIF-C ADs >56), with a 5.6%, 16.8%, and 25.4% risk of 6-week death, respectively. Nevertheless, the performance of CLIF-C ADs for predicting a composite endpoint of 6-week death or further bleeding was not satisfactory (area under the receiver operating characteristics curve [AUC], 0.588). A nomogram incorporating components of CLIF-C ADs and albumin, platelet, active bleeding, and ascites significantly improved the prediction accuracy (AUC, 0.725). Conclusions In patients with Child-Pugh B cirrhosis and AVB, risk stratification using CLIF-C ADs identifies a subgroup with high risk of death that may derive survival benefit from early TIPS. With improved prediction accuracy for 6-week death or further bleeding, the data-driven nomogram may help to stratify patients in randomized trials. Future external validation of these findings in patients with different etiologies is required.