Pathomics signatures and cuproptosis-related genes signatures for prediction of prognosis in patients with hepatocellular carcinoma.

Background: Hepatocellular carcinoma (HCC) is a common malignant tumor with high heterogeneity and poor prognosis, so early prediction and treatment are still difficult. Cuproptosis is a newly discovered type of programmed cell death that has been shown to be closely related to the occurrence and progression of HCC. Cancer morphology is influenced by genetic drivers, and computational pathology methods typically use tissue images such as entire slide images as input to predict clinical or genetic features. Therefore, the comprehensive analysis of pathological features and genomic data provides a feasible way to explore the potential mechanism of the tumor. The objective of this study was to develop a prediction model for HCC prognosis based on the pathomics signatures (PS) and the genomics signatures (GS). Methods: A dataset comprising 315 HCC patients was randomly divided into a training set (n=200) and a validation set (n=115). Prognostic models related to PS and GS were constructed by univariate and multivariate Cox regression analyses and least absolute shrinkage and selection operator (LASSO) regression analysis. Kaplan-Meier survival analysis, receiver operating characteristic (ROC) curve, univariate and multivariate Cox analyses, and nomogram were used to evaluate the predictive performance of the prognostic model. The prognostic value of the model was internally validated. Results: A prognostic model incorporating clinical features, PS, and GS was developed using Cox regression analysis and LASSO regression analyses. Kaplan-Meier survival analysis revealed statistically significant differences in survival time between high-risk and low-risk subgroups in both the training and validation datasets (PS: P=0.003 and <0.001, respectively; GS: P=0.008 and 0.004, respectively). The time-dependent ROC curve showed favorable predictive value for survival in both the training and validation sets. The area under the ROC curves at 1, 3, and 5 years was 0.750, 0.830, and 0.870 in the training set, and 0.780, 0.810, and 0.760 in the validation set, respectively. A nomogram model based on the risk model score could effectively predict the survival probability of HCC patients. The calibration curves further demonstrated the good predictive capability of the nomogram model. Conclusions: The prognostic model incorporating PS and GS could effectively predict the prognosis of HCC patients.