Comprehensive molecular characterization of endometrial cancer, etiologic heterogeneity, and racial disparities

ABSTRACT Endometrial cancer (EC) is the most common gynecologic cancer in the US. Incidence is increasing, especially for aggressive, understudied tumors that confer poor prognosis and are more often seen in African Americans (AAs). EC has one of the largest survival disparities of all cancers: AAs have >2-fold higher mortality vs. other racial/ethnic groups. The disparity remains after accounting for stage, histology, comorbidities, and treatment. The etiology of aggressive tumors and 2-fold survival disparity are large knowledge gaps in EC. The Cancer Genome Atlas (TCGA) achieved milestones in clarifying endometrial tumor biology. Using exome sequence data, TCGA defined 4 new tumor subtypes with prognostic significance and showed these data can refine subtype classification beyond classic histology. But TCGA used mostly good prognosis endometrioid tumors (>90%) and tumors in white women—with only 46 AAs—to define these subtypes. Our pilot analysis of AA vs. non-AA tumors in these sparse data suggested AAs more often had mutational features suggestive of poor outcomes. We hypothesize somatic differences in AA vs. non-AA tumors may help explain the large survival disparity. Here, we will use the largest, most diverse population to date—including 1,011 AA and 2,043 non-AA cases in the Epidemiology of Endometrial Cancer Consortium (E2C2)—to study genomic variation across the full spectrum of endometrial tumors, distinct risk factor profiles across tumor types, and the role of underlying tumor biology in driving the 2-fold survival disparity. We will: define the mutational landscape and novel tumor subtypes using whole-exome sequence data in 3,054 endometrial tumors and compare these in AA vs. non-AA cases. This will use exhaustive genomic profiling of point mutations, indels, and copy number alterations. Next, we will identify differences in risk factor associations by tumor molecular subtypes in 3,054 cases and 3,054 matched controls. Despite many known EC risk factors, TCGA was not designed to study these in concert with somatic changes. We will combine tumor profiling data in cases with information on known germline genetic and epidemiologic risk factors in cases and controls to study distinct risk factor profiles by tumor subtypes. Finally, we will 3) determine the extent to which tumor molecular subtypes explain the 2-fold survival disparity in AA and non-AA cases: Having characterized tumor genomes, we will use mediation analysis to determine the extent to which tumor molecular profiles in AAs and non-AAs explain the survival disparity. Leveraging E2C2 resources and collaborations, we will characterize the biology and risk profiles of the component subtypes of EC, including aggressive tumors, and somatic differences that drive the survival disparity. Long-term this can lead to refined risk prediction tools, improved targeting of disease prevention and treatment, and strategies to reduce longstanding racial disparities in mortality. Our study will also build a unique platform on which to perform future population-based -omics studies of EC.