A Risk Scoring System for Predicting Overall Survival in Patients with Liver Cancer

Background: Conventional strategies for assessing clinical and pathological risk factors have been adopted to predict clinical outcomes in patients with hepatocellular carcinoma (HCC). We hypothesized that an ensemble learning approach that incorporates multidimensional features would enable the construction of an effective prediction model for patient-specific risk profiles. Methods: We analyzed data from 222 stage II-III HCC patients who underwent surgical resection at Eastern Hepatobiliary Surgery Hospital (Shanghai, China). We developed five machine learning-based estimation models for patient overall survival. The models were trained on 155 patients with 48 features (included clinical and pathological risk factors, laboratory tests and in situ immunological profiles), and validated on the remaining 67 patients. Findings: A risk scoring system incorporating multiple machine learning methods was developed for survival prediction. For all models tested, immune features such as CD68+ and CD8+ immune cell infiltration played a crucial role in predicting patient overall survival. The risk scoring system stratified patients into high-risk and low-risk subgroups (validation cohort: HR, 6.5; 95% CI, 2.4-18; p = 3.4e-05). In the validation set, the scoring system predicted the half-year mortality of patients with an AUC of 0.9 (95% CI, 0.771-1.029) and 1-year mortality with an AUC of 0.897 (95% CI, 0.816-0.978). The system was also predictive of the time to recurrence (p <0.0001). Interpretation: The machine learning-based risk scoring system offers a novel strategy that incorporates multidimensional risk factors to predict clinical outcomes and may help medical practitioners optimize clinical follow-up or therapeutic interventions. Funding Statement: This study was supported by grants from The National Key Sci-Tech Special Project of China (No. 2018ZX10302204003, No. 2018ZX10302207), the Shanghai Education Committee of Shuguang Plan (No. 18SG32), the National Key Basic Research Program “973 project” (No. 2015CB555400), and the National Natural Science Foundation of China (No. 81472282, No. 81672899). Declaration of Interests: The authors declare that they have no conflict of interest. Ethics Approval Statement: This study was performed in accordance with the Declaration of Helsinki and was approved by the ethics committee of Eastern Hepatobiliary Surgery Hospital.