Characteristics and Outcomes of Patients With Hepatocellular Carcinoma Diagnosed at John Peter Smith Hospital

Purpose: The incidence of hepatocellular carcinoma (HCC) is increasing in the US, particularly in individuals infected with hepatitis C (HCV). Although early detection is crucial for better outcomes, at present, there is conflicting evidence regarding HCC screening and its reduction on cancer-related mortality. This study aimed to determine overall survival, prognostic factors influencing survival, and the effects of screening at-risk patients on HCC-related mortality. Methods: A retrospective chart review of patients diagnosed with HCC from 1/1/2018 to 6/4/2021 for the one-year survival analysis and 6/4/2019 for the three-year analysis. Person-time was calculated as the days from the date of diagnosis until the last known encounter or death. The primary exposure of interest was screening within two years prior to the diagnosis date via ultrasound, MRI, and/or CT. Potential covariates were age at diagnosis, race/ethnicity, gender, insurance status, alcohol use disorder, HCV, HBV, and cirrhosis. Kaplan Meier, log rank test, and Cox proportional hazards (CPH) model were used to assess survival curves, survival distributions across screening status, and the effects of additional covariates on prognosis at one and three years, respectively. A backwards stepwise regression was used on the covariates identified via a pre-univariate filtering to construct a multivariable model. Results: There were 165 and 71 patients who met the one- and three- year inclusion criteria, respectively. Survival at one and three years was 38% and 14%, respectively. Median survival for the 165 patients was 265 days (95% CI: 166, 337). Overall, 36% (n=59/163; 2 missing) and 27% (n=19/70; 1 missing) were screened prior to diagnosis. The CPH model showed a statistically significant difference in hazard ratio of death in the first year for those without screening compared to patients with screening (HR: 1.9; 95% CI: 1.2, 3.0; p-value: 0.005). After adjusting for race/ethnicity and insurance type, the CPH model yielded similar results (HR: 2.2; 95% CI: 1.3, 3.6; p-value: 0.002). The CPH model showed a statistically significant difference in hazard ratio of death in three years for those without screening compared to patients with screening (HR: 3.4; 95% CI: 1.7, 7.1; p-value: 0.001). After adjusting for gender, HBV, and race/ethnicity, the CPH model yielded similar results (HR: 2.2; 95% CI: 1.2, 4.0; p-value: 0.009). Conclusion: Overall survival in patients diagnosed with HCC at JPS, a safety-net hospital, is similar to national statistics. Screening in patients at-risk for HCC shows improved survival at one and three years. Further evaluation based on the extent of disease at the time of diagnosis, treatment decisions, and type/timing of screening could be beneficial in determining the outcomes in HCC patients.