2024 Volume 18 Issue 6 Pages 567-583
The prognosis for patients with hepatocellular carcinoma (HCC) depends on tumor stage and remnant liver function. However, it often includes tumor morphology, which is usually assessed with imaging studies or pathologic analysis, leading to limited predictive performance. Therefore, the aim of this study was to develop a simple and low-cost prognostic score for HCC based on serum biomarkers in routine clinical practice. A total of 3,100 patients were recruited. The least absolute shrinkage and selector operation (LASSO) algorithm was used to select the significant factors for overall survival. The prognostic score was devised based on multivariate Cox regression of the training cohort. Model performance was assessed by discrimination and calibration. Albumin (ALB), alkaline phosphatase (ALP), and alpha-fetoprotein (AFP) were selected by the LASSO algorithm. The three variables were incorporated into multivariate Cox regression to create the risk score (APP score = 0.390* ln (ALP) + 0.063* ln(AFP) - 0.033*ALB). The C-index, K-index, and time-dependent AUC of the score displayed significantly better predictive performance than 5 other models and 5 other staging systems. The model was able to stratify patients into three different risk groups. In conclusion, the APP score was developed to estimate survival probability and was used to stratify three strata with significantly different outcomes, outperforming other models in training and validation cohorts as well as different subgroups. This simple and low-cost model could help guide individualized follow-up.
