Abstract A111: Race is a predictor for endometrioid endometrial cancer cause-specific survival, but not all-cause survival

Abstract

Abstract Endometrial cancer is the most common gynecologic cancer, with an estimated 42,000 new cases in 2010. The reported difference in overall five year survival between white women and African American (AA) women represents significant survival disparity in cancer. Histologic types of endometrial cancer and stage at diagnosis differ substantially between AAs and whites, but AA women experience poorer survival even after adjusting for these factors. After IRB approval, we performed a case-only retrospective analysis of 362 patients with surgically treated endometrioid endometrial cancer. Data from 120 AA and 242 white women, diagnosed from 1990-2005 at a single institution were analyzed to examine risk factors associated with survival. Age at diagnosis, co-morbid conditions in the 5 years prior to diagnosis, BMI, and smoking history were obtained from medical record review. Tumor histologic type, 1988 FIGO stage, grade, vital status and number of months surviving were obtained from the Metropolitan Detroit Cancer Surveillance System, part of the SEER program. Differences in stage, co-morbid conditions (using a modified Charlson Comorbidity Index, CCI), and BMI were examined by race. Chisquare tests were used to assess differences in distribution of clinical and demographic variables. Log-rank tests and Cox proportional hazards models were used to assess the risk of all-cause and endometrial cancer-related deaths. AA women were more likely than white women to have high grade tumors (p=0.009), hypertension (p=0.0005) and higher mean BMI (p=0.009). No differences by race were identified among the total group for age (p=0.98), stage at diagnosis (p=0.78), categorical modified CCI (p=0.51) or smoking history (p=0.96). When data were stratified by grade, hypertension (p=0.0009) and high BMI (p=0.002) were more frequent among AAs compared to whites with low grade tumors, but not high grade tumors. Among women with high grade tumors, AAs had a greater number of co-morbid conditions (p=0.005) and were more likely to be current or former smokers compared to whites (p=0.003). In univariate survival analysis, age at diagnosis, CCI, stage, and grade were all significantly associated with any cause survival, whereas race was not. When examining deaths due to endometrial cancer, only stage (HR=3.37,95% CI: 2.52-4.51), grade (HR=4.32,95% CI: 2.89-6.47) and race (HR=2.12,95% CI: 1.13-3.97) were predictors of survival. Analysis of endometrial cancer-related deaths by grade suggest this difference in survival by race is seen for high grade (p=0.03) but not low grade (p=0.87) tumors. In this single institution study, race was not a predictor of overall survival, but was associated with endometrial cancer-related survival. Excess of co-morbid conditions was associated with AA race, and overall survival, but not endometrial cancer-related survival in this population. A larger, more comprehensive study, including information regarding socio-economic factors, molecular tumor markers, and treatment modality is necessary to elucidate the causes underlying the large disparity in endometrial cancer survival between AA and white women. Citation Information: Cancer Epidemiol Biomarkers Prev 2010;19(10 Suppl):A111.