Predicting decompensation in unresectable hepatocellular carcinoma on atezolizumab plus bevacizumab: the ARTE Score

Background & Aims Hepatic decompensation is a major event in patients with unresectable hepatocellular carcinoma treated with systemic therapy. Early identification of high-risk patients is essential. The aim of the study was to develop a simple score to predict decompensation in patients treated with atezolizumab plus bevacizumab. Methods This study enrolled 453 patients from the Atezolizumab-Bevacizumab Real-Life Experience for Treatment of Hepatocellular Carcinoma database (atezolizumab plus bevacizumab first-line, Child-Pugh A). External validation cohort included 292 patients from the atezolizumab plus bevacizumab–real database. Independent predictors derived from Cox regression were combined into a weighted score and operationalized as a point-based algorithm. Patients were then stratified into low-, intermediate-, and high-risk groups. Results In the Atezolizumab-Bevacizumab Real-Life Experience for Treatment of Hepatocellular Carcinoma (ARTE) database, 74 (16.3%) patients developed hepatic decompensation (median follow-up, 14 months). Neoplastic portal vein thrombosis (hazard ratio, 1.97; 95% confidence interval, 1.20–3.23; P = .007), elevated bilirubin (hazard ratio, 2.61; 95% confidence interval, 1.52–4.47; P < .001), and low platelets (hazard ratio, 1.82; 95% confidence interval, 1.07–3.10; P = .026) were independent predictors, and they were incorporated into the ARTE score. Patients were categorized as low- (0–1 points; n = 360), intermediate- (2 points; n = 49), or high-risk (3–4 points; n = 44). Intermediate-risk patients had a 1.96-fold decompensation risk ( P = .038), and high-risk patients a 4.28-fold risk ( P < .001), compared with low-risk patients. Significant separation of decompensation-free survival across groups was observed ( P < .001), with good discrimination (Harrell's C = 0.7022). Decompensation-free survival at 12 and 24 months was 87% and 83% for low-risk, 79% and 64% for intermediate-risk, and 57% and 56% for high-risk patients, respectively. The ARTE score retained prognostic value in the atezolizumab plus bevacizumab–real cohort (hazard ratio, 1.78; P = .009). Conclusions The ARTE score is an easy-to-use tool to predict hepatic decompensation in unresectable hepatocellular carcinoma, aiding clinical decision-making.