Race-specific genomic determinants of therapeutic response in African American NSCLC patients.

e21015 Background: African American (AA) individuals are disproportionately affected by lung cancer in terms of incidence and mortality, despite lower tobacco exposure compared to Caucasian Americans (CA). Since molecular profiling is critically important to guide therapy selection in non-small cell lung cancer (NSCLC), we compared treatment-relevant molecular data in a cohort of AA and CA patients with EGFR or KRAS-driven NSCLC. Methods: This is a retrospective study using a clinicogenomic electronic medical record database from health systems in the United States designed to evaluate outcomes in advanced stage NSCLC. Eligibility criteria for analysis included a diagnosis of NSCLC between 2010 and 2020, stage III or IV, available race/ethnicity data, and identification of common KRAS or EGFR activating variant within 60 days of diagnosis. The primary objective was to define the type of EGFR or KRAS activating variant and determine the frequency of co-occurring alterations in NSCLC-associated genes by cohort (AA and CA). Secondary objectives included examining the timing of genetic testing and use of targeted therapy in these cohorts. Results: A total of 642 NSCLC patients with KRAS (15.0% AA; 81.6% CA) and 348 patients with EGFR oncogenic mutations (13.8% AA; 70.1% CA) met inclusion criteria. Mean time from diagnosis to first molecular test result was 20.9 days for AA vs. 19.9 days for CA (p = 0.5). The EGFR G719S variant was more prevalent in the AA cohort (6.3% AA v 0.4%) with no significant differences observed in the most common KRAS or EGFR variants tested (Table). A total of 100 KRAS-mutant patients (16 AA, 86 CA) had a reported result in at least one co-mutation of interest. The most frequent co-mutations were TP53 (81.3% AA v 65% CA, p = 0.92); KEAP1 (18.8% AA v 8.3% CA, p = 0.52); ATM (18.8% AA v 9.5% CA, p = 0.54); and PTEN (0% AA v 8.3% CA, p = 0.5). A total of 35 EGFR-mutant patients (4 AA, 31 CA) had a co-mutation reported in TP53 (75% AA v 65% CA, p = 0.77). EGFR targeted therapy was reported in 47.9% of AA and 52.9% of CA (p = 0.639). In the KRAS-mutant cohort, delivery of immune checkpoint inhibitors (ICIs) was reported in 29.2% of AA and 39.3% of CA (p = 0.08); chemotherapy was reported in 36.5% AA and 45.2% of CA (p = 0.14). Delivery of both ICI and chemotherapy was reported in 13.5% of AA and 21.9% of CA (p = 0.08). Conclusions: In a real-world cohort of NSCLC patients, we found similar frequencies of common KRAS and EGFR variants in AA and CA, but higher rates of EGFR G719S variant in AA patients. A more comprehensive analysis is needed to further evaluate the trends in co-mutations observed in our analysis, to determine the race-specific impact of these alterations in EGFR and KRAS-driven NSCLC.[Table: see text]