A23

Although incidence of endometrial cancer is considerably lower in blacks than in whites in the US, mortality is almost double among blacks. Reasons for this include higher prevalence of lethal histologic types, economic factors, and less aggressive treatment. Symptom recognition and access to care might also explain this disparity, but this question has rarely been addressed. In a population based case control study of endometrial cancer in northern New Jersey in 2001-2005, we asked women with endometrial cancer about the symptoms they experienced in the 12 months before diagnosis and the process of obtaining care (time between noticing symptoms and seeing a doctor, type of doctor seen first, type who diagnosed cancer, and whether a gynecologic oncologist was seen). This analysis includes 403 whites and 38 blacks out of the total of 473 cases.
 Nearly all patients reported some symptoms, including 95% of whites and 92% of blacks. Bleeding was by far the most common symptom, reported by 79% of blacks and 72% of whites, followed by spotting (16% and 23%, respectively), abdominal pain (13% and 20%), back pain (16% and 10%), and frequent urination (16% and 10%). Other symptoms were reported by <10% of each group. Among all women, bleeding was associated with BMI >=35, mentioned by 82% of that group vs 68% of other women (p<0.01), with endometrioid histology (mentioned by 75% vs 54% of women with serous or clear cell tumors, p<0.01). Since black women were more likely to have BMI >=35 (47% vs 28% of whites, p<0.05), but were also more likely to have serous or clear cell tumors (24% vs 7% of whites, p<0.001), adjusting for these two factors did not affect the odds ratio comparing bleeding in blacks and whites. The OR was 1.46 (95% CI 0.65-3.28) without adjustment and 1.51 (95% CI 0.64-3.56) with adjustment. Adjustment for other potentially confounding variables did not affect the results. Bleeding was more common in women with local or regional disease (compared to distant or unknown) but stage was similar for blacks and whites. Spotting was not related to BMI or histology.
 The median time between noticing symptoms and seeing a physician was the same for both groups, 30.4 days, and did not vary according to presence or absence of bleeding. Black women were more likely to see a family physician, rather than a gynecologist, first (p<0.05), although the groups were similar in the proportion whose cancer was diagnosed by a gynecologist (84% for blacks and 88% for whites) and in the proportion who ever saw a gynecologic oncologist during the course of their illness (73% and 77% respectively). Among the 69 women of both races who went to a family physician or internist first, the large majority (91%) were diagnosed by a gynecologist or gynecologic oncologist.
 In conclusion, we found no major differences between the racial groups in symptoms, time to seeing a physician, or whether a gynecologic oncologist was seen. The difference noted in the type of physician seen first along with the similarity in proportion diagnosed by a gynecologist may imply a delay in diagnosis for blacks if they were referred to a gynecologist by a family doctor. We do not have data on the length of time from symptom recognition to diagnosis. The study was conducted in a highly urban region, and access to care might be different elsewhere. Although we have not followed these women for survival, survival patterns in New Jersey have, in the past, been similar to those for the US.

First AACR International Conference on the Science of Cancer Health Disparities-- Nov 27-30, 2007; Atlanta, GA