Prognostic factors affecting racial impact in testicular cancer outcomes in the United States.

Abstract

591 Background: We previously reported demographics, race, stage, treatment, and survival patterns in testicular cancer (TC) patients in the U.S. using the National Cancer Database (NCDB) data (Yu et al., JCO:43 (5): Abs 622). We now examine race in a multivariable model to further understand survival outcomes in both seminoma and non-seminoma. Methods: Patient-level data were extracted from NCDB for TC cases diagnosed between 2004 and 2020, including histologic subtypes seminoma (S) and non-seminoma (NS). Statistical analyses involved descriptive statistics, chi-square tests comparing categorical variables, Kaplan-Meier method analyzing overall survival (OS) , and log-rank tests comparing OS across racial/ethnic groups, disease stages, comorbidity scores, facility types, and insurance statuses. Multivariate Cox Proportional Hazards models were used to examine the impact of race, Spanish origin, facility type, Charlson-Deyo Score, Insurance, Stage, Education, and median income on OS. Results: The cohort included 89,550 patients: 81,571 White, 2,860 Black, 3,624 with other races, and 1,495 Unknown. We previously showed that White patients were primarily treated at comprehensive community cancer and academic/research programs, while Black patients were more frequently treated at academic centers. Hispanic patients were underrepresented at academic and comprehensive community cancer programs. Private insurance predominated overall (72%), highest in Whites (72.9%) and lowest in Blacks (54.1%) and Hispanics (48.8%). Medicaid coverage and uninsured status were higher in Black and Hispanic groups. 23589 non-seminoma patients and 18298 seminoma patients were included in the multivariate analyses. Significant factors were worse Charlson-Deyo scores (combined 2&3 vs 0 vs 1) P<0.0001, insurance (combined Medicaid & Medicare & other government vs not insured vs private vs government) P<0.0001, stage (stage I, II, III) P<0.0001, while education (P=0.9044) and median income (P=0.1118) were not significant factors. Conclusions: Race appears to be an important contributing factor to overall survival in non-seminoma patients when accounting for Charlson-Deyo score, stage, and insurance status, but income did not seem to have a significant contributory effect. In multivariate analyses of our TC population overall and specifically in seminoma patients, race appears to have less of an effect on overall survival.