Clinical characteristics and prognosis of non-APAP drug-induced acute liver failure: a large multicenter cohort study

作者：Han, Lin; Huang, Ang; Chen, Jinjun; Teng, Guangju; Sun, Ying; Chang, Binxia; Liu, Hong-Li; Xu, Manman; Lan, Xiaoqin; Liang, Qingsheng; Zhao, Jun; Tian, Hui; Chen, Songhai; Zhu, Yun; Xie, Huan; Dang, Tong; Wang, Jing; Li, Ning; Wang, Xiaoxia; Chen, Yu^*; Yang, Yong-Feng^*; Ji, Dong^*; Zou, Zhengsheng^*

来源：Hepatology International, 2023.

DOI：10.1007/s12072-023-10541-w

• Acute liver failure
• Drug-induced liver injury
• Transplant-free survival
• Hepatic encephalopathy

BackgroundThere is growing recognition of natural history, complications, and outcomes of patients who develop non-acetaminophen (APAP) drug-induced acute liver failure (ALF). To clarify high-risk factors and develop a nomogram model to predict transplant-free survival (TFS) in patients with non-APAP drug-induced ALF.MethodsPatients with non-APAP drug-induced ALF from 5 participating centers were retrospectively analyzed. The primary endpoint was 21-day TFS. Total sample size was 482 patients.ResultsRegarding causative agents, the most common implicated drugs were herbal and dietary supplements (HDS) (57.0%). The hepatocellular type (R >= 5) was the main liver injury pattern (69.0%). International normalized ratio, hepatic encephalopathy grades, the use of vasopressor, N-acetylcysteine, or artificial liver support system were associated with TFS and incorporated to construct a nomogram model (drug-induced acute liver failure-5, DIALF-5). The AUROC of DIALF-5 for 7-day, 21-day, 60-day, and 90-day TFS in the internal cohort were 0.886, 0.915, 0.920, and 0.912, respectively. Moreover, the AUROC of DIALF-5 for 21-day TFS had the highest AUROC, which was significantly higher than 0.725 of MELD and 0.519 of KCC (p < 0.05), numerically higher than 0.905 of ALFSG-PI but without statistical difference (p > 0.05). These results were successfully validated in the external cohort (147 patients).ConclusionsBased on easily identifiable clinical data, the novel DIALF-5 model was developed to predict transplant-free survival in non-APAP drug-induced ALF, which was superior to KCC, MELD and had a similar prediction performance to ALFSG-PI but is more convenient, which can directly calculate TFS at multiple time points.

• 单位
  5; 1; 北京大学; y; 南方医科大学