Similar pathobiology of ovarian epithelial cancers in diverse populations with differing incidence rates Free

A18

Aim: Incidence rates of ovarian cancer in minority women are reputedly lower than in White women; nevertheless, survival among Black women appears to be suboptimal. We postulated that a refined stratification of incidence data, with a focus on main histological subtypes of ovarian epithelial cancers (OEC), would help to discern whether any pathobiologic variation underlies outcome disparity.
 Methods: The National Cancer Institute Surveillance, Epidemiology, and End Results (SEER) population database afforded bulk sample size with ethnic/racial diversity. SEER’s 13 Registry Database (calendar period 1992-2004), covers 14% of the US population with 29,842 cases of OEC (92% of total ovarian cancers). This number was sufficient for valid comparisons of self-categorized White, Black and Asian-Pacific Islander (API) women. SEER embedded codes (International Classification of Disease for Oncology-3) provided diagnostic standardization of usual histological subtypes: serous (8441, 8460-1), endometrioid (8380) and mucinous (8470-1, 8480-1). Age-specific incidence rates per 100,000 woman-years (ASIR) were plotted as log-log curves. Age-frequency histograms (age-at-diagnosis density plots) compared age distributions.
 Summary:SEER adjusted incidence rates for all OEC and each histological subtype were highest in White women (p<0.05). Rates ranked in a series (Whites, Blacks, API) for each subtype were: for serous 7.3, 3.9, 3.9, for endometrioid 2.1, 0.9, 1.8, for mucinous 1.4, 1.1, 1.2, and for all OEC 17, 11, 12. Cases of serous histology predominated in all women, but heterogeneity was reflected in the trajectories of ASIR plotted according to pathologic grade: for serous tumors of high grade, curves increased rapidly to ages 45-55 years then slowed, while for low-grade tumors rate curves rose continuously to ages 50-60 years before slowing. The rate curves for high grade and low grade cancers “crossed” between ages 35-45 years. This was consistent with a grade-related qualitative age interaction (or effect modification). Age frequency histograms of cases within the serous subtype also indicated two subpopulations: relatively early-age of onset (low-grade) versus delayed onset (high grade). Analysis of the other OEC subtypes revealed no such dichotomy. Indeed, ASIR plotted for the serous subtype cancers of low grade tracked more closely to curves plotted for either endometrioid or mucinous subtypes/grades than to the curves for serous high grade. Characteristically, ASIR curves for the serous subtype leveled post-menopause (typically ages 50-60 years depending upon the grade); curves for the endometrioid subtype flattened about a decade earlier; and curves for mucinous cancers rose steadily to age 70 years.
 Conclusions:The findings of lower OEC incidence rates in non-White versus White women bolstered earlier but less extensive observations of rate differences. Despite such differences, the consistent results of present data stratification, including a predominance of serous subtype with similar grade heterogeneity, and a congruence of ASIR curves for each of three OEC subtypes, suggest that the age-dependent progression towards each subtype/pathologic grade of OEC runs with generally parallel kinetics in diverse populations. In this context, survival disparity may reflect a complex interplay of environmental factors or host responses.