A Web-Based Prediction Model for Cancer-Specific Survival of Elderly Patients With Early Hepatocellular Carcinoma: A Study Based on SEER Database.

Primary liver cancer is a common malignant tumor primarily represented by hepatocellular carcinoma (HCC). The number of elderly patients with early HCC is increasing, and older age is related to a worse prognosis. However, an accurate predictive model for the prognosis of these patients is still lacking. Data of eligible elderly patients with early HCC in Surveillance, Epidemiology, and End Results database from 2010 to 2016 were downloaded. Patients from 2010 to 2015 were randomly assigned to the training cohort ( = 1093) and validation cohort ( = 461). Patients' data in 2016 ( = 431) was used for external validation. Independent prognostic factors were obtained using univariate and multivariate analyses. Based on these factors, a cancer-specific survival (CSS) nomogram was constructed. The predictive performance and clinical practicability of our nomogram were validated. According to the risk scores of our nomogram, patients were divided into low-, intermediate-, and high-risk groups. A survival analysis was performed using Kaplan-Meier curves and log-rank tests. Age, race, T stage, histological grade, surgery, radiotherapy, and chemotherapy were independent predictors for CSS and thus were included in our nomogram. In the training cohort and validation cohort, the concordance indices (C-indices) of our nomogram were 0.739 (95% CI: 0.714-0.764) and 0.756 (95% CI: 0.719-0.793), respectively. The 1-, 3-, and 5-year areas under receiver operating characteristic curves (AUCs) showed similar results. Calibration curves revealed high consistency between observations and predictions. In external validation cohort, C-index (0.802, 95%CI: 0.778-0.826) and calibration curves also revealed high consistency between observations and predictions. Compared with the TNM stage, nomogram-related decision curve analysis (DCA) curves indicated better clinical practicability. Kaplan-Meier curves revealed that CSS significantly differed among the three different risk groups. In addition, an online prediction tool for CSS was developed. A web-based prediction model for CSS of elderly patients with early HCC was constructed and validated, and it may be helpful for the prognostic evaluation, therapeutic strategy selection, and follow-up management of these patients.