The impact of obesity on treatment and outcomes in ovarian cancer by race/ethnicity

e18067 Background: Concern for toxicities in obese ovarian cancer patients may limit weight-based chemotherapy dosing, which has been associated with poor survival. We sought to evaluate the impact of obesity on treatment and outcomes by race/ethnicity at a National Cancer Institute (NCI)-designated Comprehensive Cancer Center (CCC). Methods: 66 minority (non-Hispanic black, Hispanic) epithelial ovarian cancer patients diagnosed 2006 and treated with frontline chemotherapy were frequency matched to two non-Hispanic white (NHW) patients by 5-year age group and stage (n = 198). For frontline carboplatin, relative dose intensity (RDI) was calculated by dividing the dose received by the expected dose, with an RDI < 0.85 indicating carboplatin dose reduction. Patient characteristics and RDI were compared by obesity and race/ethnicity. Cox proportional hazards regression was used to estimate hazard ratios (HR) and 95% confidence intervals (CI) for the association between obesity, race/ethnicity and RDI with mortality, adjusting for confounders. Results: Minorities were more likely to be obese, underinsured, and to experience delays in chemotherapy compared to NHW. Compared to normal/overweight patients, obese patients were more likely to experience carboplatin under dosing in the neoadjuvant (p = 0.004) and adjuvant (p < 0.0001) setting, but there were no differences in RDI by race/ethnicity or toxicities. No association with survival was observed for dose reduction and BMI. Compared to NHW, minorities had similar survival in the first 4 years after diagnosis (HR = 0.99, 95% CI = 0.60-1.63), but an increased risk of mortality > 4 years after diagnosis (HR = 4.07, 95% CI = 1.92, 8.60). Conclusions: Even though minorities were more likely to be obese, there were no differences in dosing by race/ethnicity at an NCI CCC. Minorities had similar survival to NHW in the first 4 years after diagnosis, likely due to similar access to quality care at an NCI CCC. However, > 4 years after diagnosis, the racial/ethnic survival disparity persisted after adjustment for clinical and pathologic characteristics. Expansion of this work to other patient populations may help determine if findings are similar outside of a tertiary care center.