EP358/#948 Evaluating the racial/ethnic representation of targeted therapy clinical trials in gynecologic oncology: who is being left behind?

Abstract

Methods Using the Florida Cancer Data System (FCDS), women with EC diagnosed from 1981-2016 were identified. Demographic and clinical information were abstracted. Women who self-identified as Black and born in the US, Jamaica, or Haiti were included. Statistical analyses were performed using chi-square, Cox proportional hazards models, and the Kaplan-Meier methods with significance set at p<0.05. Results 3434 women met inclusion criteria. Compared to USB, JB and HB had more high-grade histologies, more advanced disease, were uninsured or government-insured, and received more chemotherapy (all p<0.05) (table 1). In multivariate analyses, age (HR 1.02, p=0.008), distant disease (HR 2.32, p<0.001), high-grade histology (HR 2.15, p<0.001), surgery (HR 0.23, p<0.001), and chemotherapy (HR 0.67, p=0.01) were independently associated with OS. In addition, relative to HB women, USB (HR 0.59, p=0.004), and JB (HR 0.54, p=0.026) had improved OS. Among patients with serous EC, HB women had markedly worse median OS (18.5 months) compared to USB (32.2 months) and JB (41 months) (p=0.027) (figure 1). Conclusions Country of birth affects EC survival, with Haitian women demonstrating worse outcomes. Understanding the biologic and social etiologies of these findings is necessary.