Clinical-pathological characteristics and prognosis of elderly primary liver cancer patients

Primary liver cancer (PLC) is a highly malignant disease, with an increasing incidence among the elderly population. However, there is a lack of systematic evaluation regarding its progression, recurrence, and prognostic factors. This study aims to investigate the clinicopathologic characteristics of elderly patients with PLC and to establish predictive models for overall survival and recurrence risk in this population. A retrospective analysis was conducted on 460 elderly PLC patients admitted to Longyou County People's Hospital from January 2018 to December 2019. According to pathological types, patients were classified into hepatocellular carcinoma group (HCC, n=395), intrahepatic cholangiocarcinoma group (ICC, n=49), and combined hepatocellular-cholangiocarcinoma group (CHC, n=16). Based on clinical outcomes, patients were further grouped as survivors (n=228) and non-survivors (n=232), or recurrence (n=223) and non-recurrence (n=237) groups. To validate the predictive models, 210 elderly patients with PLC from the same hospital, admitted between January 2020 and December 2020, were included as a verification cohort. Demographic and clinical data collected included age, sex, alanine aminotransferase (ALT), aspartate aminotransferase (AST), alpha-fetoprotein (AFP), tumor number, and tumor size. Univariate analysis was performed using the chi-square (χ²) test. Then, multivariate logistic regression analysis was performed to identify independent risk factors and establish predictive models for mortality and recurrence. A Random Forest model was also constructed for comparison. Model performance was assessed using the area under the receiver operating characteristic (ROC) curve (AUC) for discrimination, and the Hosmer-Lemeshow test for calibration. HCC predominated among male patients (P<0.05), with significantly higher AFP expression in males (P<0.05). Multivariate analysis identified age ≥80 years, AFP>400 ng/ml, multifocal tumors, tumor size >2 cm as independent predictors of mortality (all P<0.05). Additionally, hypertension, smoking, alcohol consumption, AFP>400 ng/ml, and tumor multiplicity were identified as independent predictors of postoperative recurrence (P<0.05). The constructed predictive models demonstrated good performance in both the training and validation cohorts. Comparative analysis showed similar predictive efficacy between logistic regression and Random Forest models (AUC for mortality prediction: 0.846/0.852 vs. 0.842/0.858, respectively; AUC for recurrence prediction: 0.756/0.762 vs. 0.761/0.764, respectively). The Hosmer-Lemeshow test suggested adequate model calibration (P>0.05). HCC is the most common form of PLC among elderly patients and is closely associated with metabolic and lifestyle factors. Advanced age, elevated AFP, multilocality, and larger tumor size increase mortality risk, while smoking and alcohol consumption further elevate recurrence risk. Early identification of high-risk patients, individualized intervention, and close postoperative follow-up focusing on modifiable risk factors are recommended to improve patient outcomes.