P1.14-21 Propensity-Matched Genomic Profiling Comparison of Lung Adenocarcinomas Among 3 Races, a Multicenter Study of 4655 Patients

The mutation rates of multiple genes are reported to differ among lung adenocarcinoma samples from different races. However, the baseline heterogeneity among different race cohorts may confound the interpretations of the results. This study is the first propensity-matched study designed to determine the genomic and therapeutic actionable disparities among the samples of Asian vs White and Black vs White races. Sequencing data from 4 centers with primary races documented as “Asian”, “Black”, or “White” was accessed from GENIE database. Propensity score analysis (nearest neighbor matching method) was applied to Asian vs White and Black vs White cohorts using age, sex, sample type (primary/metastasis) and enrolling center. Mutation rates, pathway alterations, and therapeutic actionability were compared among the cohorts. Therapeutic actionability was annotated using OncoKB annotator. Fisher’s exact test and FDR adjust method were used for comparison. A total of 4655 patients were included before matching (Figure A). 2 pairs of cohorts, 350 Asian vs 1400 White patients and 208 Black vs 832 White patients were identified for study after propensity matching. 11 genes, including EGFR, KRAS, STK11, KEAP, BRAF, SMARCA4, PIK3CA et.al. were significantly differentially mutated among Asian and White patients (adjusted P value<0.05), while only EGFR mutation rate was significantly different between Black and White cohorts (adjusted P value=0.04). In the analysis of 10 oncogenic pathways, RTK/RAS, cell cycle, hippo, notch, nrf2, wnt pathway alteration rates are different among Asian and White samples (adjusted P value<0.05). The comparison between matched Black and White samples observed no significant difference in pathway alterations. In the therapeutic actionability analysis, 4663 mutations, 974 copy number alterations, 196 fusions were annotated with at least one actionable drug in the whole cohort. 67% of samples from Asian patients embraced at least one genomic alteration corresponding to level 1 drugs, compared to 30% in matched White samples (P<0.001) (Figure B). In the Black vs White group, the rate of corresponding level 1 drugs per sample is 39% vs 29% (P=0.003) (Figure C). In this propensity-matched study, samples from Asian LUAD patients displayed significantly different genomic mutation rate, pathway alteration rate, and therapeutic actionability characteristics compared to White patients. Samples from Black patients have a higher rate of EGFR mutations and level 1 therapeutic actionability than white patients. But the difference is more prominent in the Asian vs White group compared to Black vs White group.