Comparison of the clinical and genomic profiles of endometrial cancer in Appalachian and non-Appalachian patients.

5591Background: Kentucky has the highest incidence and mortality from gynecologic cancer in the United States, and the incidence of endometrial cancer (EC) in the Kentucky Appalachian (AP) population is even higher than non-Appalachian (non-AP). The purpose of this study is to compare demographics, molecular profiles, and outcomes between AP and non-AP populations. Methods: Using the Total Cancer Care prospective cohort study, we evaluated 836 women with EC, 40 AP (4.8%), 700 non-AP (83.7%), and 96 with unknown residence (11.5%). We used descriptive statistics and univariate analyses to compare AP vs non-AP and MutSig to identify significantly mutated genes. RNASeq data was trimmed and processed with STAR. Differential expression and pathway enrichment analyses were processed with EdgeR and GSEA, respectively. We grouped patients by POLE mutation first, then microsatellite instability (MSI) high. We identified CN clusters with hierarchical clustering and grouped the remainder by CN status. We compared overall survival (OS) among molecular subgroups using Kaplan-Meier curves and log-rank tests. Results: Most participants were white (90.2%), had a median age of 62.7 years, median BMI of 34.0 mg/m², had stage I disease (64.6%), and endometrioid histology (83.3%). AP women were diagnosed at a higher stage (stage IV 25.8% vs 7.7%, p=0.0008) and were more likely to have serous carcinoma (33.3% vs 9.0%, p<0.0001). There was no difference in age, race, or BMI between AP and non-AP. The AP group had significantly more EVI5 mutations than the non-AP group (17.5% vs 6.0%, p=0.012), as well as significant differences in RNA expression, specifically upregulation in oxidative phosphorylation pathways. Cluster (c) analysis identified 6 CN clusters; c 2 corresponded to the TCGA CN high group, and we divided the TCGA CN low group into two distinct molecular subgroups: CN moderate (c 1, 3, 4, 6) and CN low (c 5) as there was significant differences in CN alteration between these clusters. CN high had worse OS, CN moderate had improved OS, and CN low had intermediate OS (p<0.0001); significance persisted when controlling for covariates. Most patients were in CN moderate (n=245, 29.52%), followed by MSI High (n=210, 25.3%), CN low (n=175, 21.1%), CN high (n=160, 19.3%), and POLEmutated (n=40, 4.8%). Conclusions: This is a larger independent replicate of TCGA molecular subtyping, which includes a previously unevaluated AP population. We identified additional CN clusters and a new CN group with significantly different survival. AP patients were more likely to have advanced stage disease, serous histology, and EVI5 mutations, which is an oncogene involved in cell cycle regulation not previously implicated in EC. AP patients also have significant differences in RNA expression. These molecular differences may reveal opportunities for treatment and reduce cancer disparities.