Genetic ancestry-associated differences in genomic profiling and treatment patterns in pancreatic ductal adenocarcinoma (PDAC).

Abstract

4138 Background: Prior epidemiological studies have explored race both as a population-level risk factor and as a factor influencing outcomes in PDAC. However, the impact of genetic ancestry on PDAC remains unclear. We assessed the genomic landscape, clinical care patterns, and outcomes across ancestries in a large and diverse PDAC cohort to determine the potential impact of genomics on ancestral disparities. Methods: A total of 24,948 patients diagnosed with PDAC who received tissue-biopsy based comprehensive genomic profiling (CGP; Foundation Medicine, Inc. (FMI)) as part of routine clinical care from Aug 2014-Sep 2023 were examined for genomic differences based on genetic ancestry which was determined using a SNP-based approach. The US-based de-identified Flatiron Health (FH)-FMI clinico-genomic database (CGDB) was used to assess testing and treatment patterns based on ancestry in 3,863 patients of European (EUR) and 414 patients of African (AFR) ancestry diagnosed with PDAC. Results: The genetic ancestry distribution in the overall cohort was 80.9% EUR, 10.0% AFR, 4.9% Admixed American, 3.5% East Asian, and 0.7% South Asian. AFR patients showed a slightly younger median age at biopsy than EUR (64 vs 67, p < .001). KRAS, TP53, CDKN2A/ 2B and SMAD4 were the most commonly altered genes in the cohort, as expected. Compared to EUR, AFR patients were enriched in alterations in TP53 (85% vs 78%, p < .001) and KRAS (95% vs 93%, p < .001). While rare, alterations in GNAS, CHEK2, and BRAF were observed less frequently in AFR (1.7% vs 3.3%, 0.3% vs 1.7%, 0.9% vs 2.2%, respectively, all p < .001). In the CGDB cohort, AFR patients were likely to have lower socioeconomic status (SES) compared to EUR patients (33% of AFR in the lowest SES quintile compared to 9% of EUR; p < .001 across all quintiles). AFR patients were less likely to receive care at academic centers (AFR 14.8% vs EUR 22.2%, p < .001) and less likely to receive an investigational drug during their cancer treatment course (5.9% vs 14.1%, p = .001). Rates of surgery (AFR 30.2% vs EUR 34.6%, p = .08), immunotherapy (3.2% vs 3.3%, p > .99), and targeted therapy (4.1% vs 6.2%, p = .28) were not significantly different. While AFR patients comprised a significantly larger proportion of AFR+EUR CGP volume following the decision to implement Medicare coverage for NGS testing (10.6% after Mar 16, 2018 vs 6.6% prior, p < .001), this remained below the expected based on share of the overall US population. Conclusions: To our knowledge, this is the largest study of genetic ancestry in PDAC to date. There were largely similar patterns of gene alterations across AFR and EUR ancestry groups suggesting non-genomic factors may underlie clinical outcomes disparities which have been reported in other studies. Notably, there were ancestry-based differences associated with SES and both CGP testing and clinical trial enrollment patterns, which may impact patient outcomes.