Prognostic nomogram predicting survival of patients with unresectable hepatocellular carcinoma after hepatic arterial infusion chemotherapy

BACKGROUND AND AIM Hepatic arterial infusion chemotherapy (HAIC) has shown encouraging efficacy in the treatment of hepatocellular carcinoma (HCC). This study aims to establish and validate a novel nomogram to predict individualized survival outcomes for patients with unresectable HCC after HAIC. METHODS Between January 2016 and December 2018, 463 patients diagnosed with HCC who initially received HAIC were included in this study (training cohort: n = 308; validation cohort: n = 153). The prognostic nomogram was constructed based on the training cohort using the independent predictors assessed by the multivariate Cox proportional hazards model. The predictive accuracy and discriminative ability of the model were evaluated by the concordance index (C-index), calibration curve and area under the time-dependent receiver operating characteristic (tdAUC) curve. RESULTS After a median follow-up of 35.4 months, 358 patients had died. Six factors, including C-reactive protein, albumin-bilirubin grade, alpha fetoprotein, extrahepatic metastasis, portal vein invasion and tumor size, were selected to establish the nomogram. In the training cohort, the C-index of the nomogram was 0.710, which was significantly better than that of six conventional staging systems (P < 0.001), and the nomogram had a higher tdAUC over time. The calibration curve showed good agreement between the predicted probability and actual outcome. According to specified values, the nomogram stratified patients into three or four risk groups (P < 0.001). Similar findings could be observed in the validation cohort. CONCLUSION The nomogram in this study accurately predicted the OS of patients with unresectable HCC after HAIC.