LAUSR

Objectives: Clinical trials are the backbone in the investigation and approval of novel cancer treatments. In recent trials on immunotherapy that led to regulatory approvals, the elderly and minority populations have been underrepresented. We sought to compare the characteristics of ovarian cancer patients enrolled in a recent NRG/GOG trial with those of a corresponding real-world population. Methods: Data from NRG/GOG262 trial were compared to advanced ovarian cancer patients in Flatiron - a longitudinal, demographically and geographically diverse database comprised of health records for >2 million US cancer patients, obtained from >280 cancer clinics. We extracted data for all patients in Flatiron with stage III/IV ovarian cancer initially diagnosed between 1/1/2011 and 12/31/2019. Demographics (age and race), clinical characteristics (tumor stage and histology) and treatment patterns (receipt of neoadjuvant therapy and bevacizumab) were compared against those of the NRG/GOG262 trial, using chi-square tests to assess differences. Results: Of 3,439 patients, 692 were from NRG/GOG262 and the remaining 2,747 were from Flatiron electronic health records. Relative to NRG/GOG262, real-world patients were older (>60 years old) (71% vs. 54%; p Conclusions: Relative to patients enrolled in a clinical trial, real-world patients are older, more racially diverse, more likely to undergo neoadjuvant treatment, and less likely to utilize bevacizumab. Our data suggest a need to modify the design of future clinical trials in order to enhance the generalizability of results to real-world patients.