Background: Based on availability of approved agents targeting tumor somatic mutations, current guidelines recommend NGS for Pts with mPC and mUC. However, large cohort studies in mPC and mUC evaluating disparities in NGS testing have not been reported. Herein, we investigated the effect of social determinant of health (SDOH) on disparities in NGS testing in a large real-world dataset of pts with mPC and mUC. Methods: Pts data were extracted from the nationwide Flatiron Electronic Health Record (EHR)-derived de-identified database. Inclusion: new diagnosis of advanced or metastatic (adv/met) PCa and UCa between March 2015 to December 2022 and receipt of cancer therapy (to include only treatment eligible pts). Disparities in incidence of NGS testing after adv/met diagnosis by SES, R/E, region, and insurance were assessed using Fine and Grey’s (FG) modified stratified Cox proportional hazard (PH) model. FG model considered death as a competing risk. Stratification was done by year of adv/met diagnosis (accounting heterogeneity) and R/E (an effect modifier for SES, region, and insurance), assuming different baseline hazards in each stratum. All analysis done using R v.4.2.3. Results: Analytic cohorts comprised 11,936 pts with mPC and 6,490 pts with mUC. Hazard ratios (HR) and 95% confidence intervals (CI) from FG Cox PH models are reported in the table. For both cancer types, pts were more likely to get NGS testing when they were White, non-Hispanic (compared to Black, non-Hispanic), had higher SES, and had commercial insurance. Conclusions: Our data suggest an association between incidence of NGS testing and several SDOH in mPC and mUC. Further studies are needed to elucidate the reasons for these disparities and ways to mitigate them.

^astratified by adv/met diagnosis year. ^bstratified by adv/met diagnosis year and R/E ref: reference NA: Not applicable.