Abstract A053: Understanding geospatial relationships in ovarian cancer risk

Abstract

Abstract To examine the geographic distribution of ovarian cancer cases in the New England area and identify potential case clusters. Methods: This was a case-control study utilizing 1580 ovarian cancer patients from a single healthcare system (MGB). Using electronic health record (EHR) review, patient primary addresses were geocoded. To account for spatial distribution confounded by access to the MGB health system, 1750 patients from the MGB biobank system were geocoded as controls. Generalized additive models were used to determine the spatial patterning of cases and controls, controlled for age and race. Geospatial analysis was performed using ArcGIS software and R. Results: 3325 patients (1576 ovarian cancer patients, 1749 biobank controls) were successfully geocoded. Distribution of ovarian cancer cases and biobank patient controls were similar (average nearest neighbor distance 6337m vs 5623 m). Among the cohort of patients in New England (defined as MA, ME, NH, VT, RI, CT), there were 1476 ovarian cancer patients and 1663 biobank control patients. Use of a generalized additive model demonstrated higher incidence of ovarian cancer in Western CT, western MA, VT, and southern ME. These data are similar to SEER reported county-level incidence of late-stage ovarian cancer. Conclusions: Understanding the geospatial distribution of ovarian cancer cases can highlight potential clusters of higher incidences of disease. Causal relationships have yet to be established and require future investigation, but these data suggest spatially patterned exposures are worth examination. Citation Format: Victoria Wang, Jaime Hart, Kevin Elias. Understanding geospatial relationships in ovarian cancer risk [abstract]. In: Proceedings of the AACR Special Conference on Ovarian Cancer; 2023 Oct 5-7; Boston, Massachusetts. Philadelphia (PA): AACR; Cancer Res 2024;84(5 Suppl_2):Abstract nr A053.