Nipun Verma, Tingting Qi, Pratibha Garg, Arun Valsan, Gowripriya Nair, Julian Pohl, Arathi Venu, Xianbo Wang, Xin Zheng, Guohong Deng, Yan Huang, 孟忠吉, Yanhang Gao, Zhiping Qian, Xiaobo Lu, Feng Liu, Yu Shi, Shang Jia, Yubao Zheng, Yingli He, Meiqian Hu, Changze Hong, Sina Jameie‐Oskooei, R MOOKERJEE, Mohammad Mahdi Saeidinejad, Sunil Taneja, Jinjun Chen, Hai Li, Ajay Duseja, Cornelius Engelmann, Rajiv Jalan
Background and Aims: Acute-on-chronic liver failure (ACLF) lacks a universally accepted definition, and recent efforts have proposed consensus organ failure criteria. In this study, we aimed to compare the clinical validity of a recently proposed consensus ACLF framework with the outcome-calibrated A-TANGO classification. Methods: We performed a multinational cohort study including 2,398 patients from the TIH cohort (India) and 2,568 from the CATCH-LIFE cohort (China) who were hospitalized with acute decompensation of cirrhosis. ACLF was defined using A-TANGO and an operationalized version of the 2025 consensus framework. Outcomes were 28- and 90-day mortality. Analyses assessed case capture, overlap, mortality risk, sensitivity, specificity, and net reclassification improvement (NRI). Results: ACLF prevalence differed substantially by definition. In TIH, A-TANGO classified 79.2% as ACLF versus 42.3% by the consensus definition; in CATCH-LIFE, the corresponding values were 31.4% versus 5.8%, respectively. Most consensus ACLF cases were captured by A-TANGO, which additionally classified 26%-37% of patients as having ACLF. These patients had substantial mortality (28-day: 18.1%-26.9%; 90-day: 33.2%-37.9%), significantly higher than those negative by both frameworks and comparable to established ACLF risk thresholds. A-TANGO showed higher sensitivity for 28-day mortality (TIH: 94.1% vs. 67.8%; CATCH-LIFE: 76.1% vs. 25.6%), whereas consensus criteria were more specific. Reclassification analyses showed improved discrimination with A-TANGO (NRI: 17.1% in TIH; 27.4% in CATCH-LIFE). Within the consensus non-ACLF group, A-TANGO further stratified patients into distinct risk groups with stepwise increase in mortality. Conclusions: In conclusion, the two frameworks identify fundamentally different populations. The consensus definition significantly reduces sensitivity and under-recognizes high-risk patients. Compared with consensus definitions, the outcome-calibrated framework better supports diagnosis, clinical decision-making, risk stratification, and trial design in ACLF.