Factors Associated with Aromatase Inhibitor Use in a Large Cohort of Women Cared for in an Integrated Health Plan.

Abstract

Listen

Translate article

Abstract Background: Third generation aromatase inhibitors (AI) have been well tolerated in clinical trials and may have fewer side effects than tamoxifen. While national treatment guidelines recommend AI for postmenopausal women with ER+ tumors, the extent to which women are prescribed AIs in community settings in the U.S., and as front-line or extended therapy, remains unknown. Using data from a large non-profit health plan that serves over 3 million residents in Southern California, we report results examining the characteristics of breast cancer (BCa) survivors who were ever prescribed AIs.Methods: This cross-sectional analysis examined a cohort of 25,577 BCa (in situ and invasive cases) diagnosed from 1996-2006. We collected demographic, tumor, and cancer treatment information from automated health plan databases and SEER-affiliated cancer registry files. We identified patient and clinical factors associated with AI use. P-values (2-sided) were based on the chi-square distribution. Adjusted odds ratios (for age and year at diagnosis) and 95% confidence intervals were estimated using logistic regression.Results: In this cohort, nearly 37% ever used AIs (P<0.0001). Use of AIs increased by year (P<0.0001). Compared to non-Hispanic whites, minority women were less likely prescribed AIs (AOR=0.75 for blacks, 95% CI: 0.69-0.81). Although most women who used AI were first diagnosed with early stage disease (2 cm vs. <1cm). Women with positive lymph nodes had twice the odds of AI use (OR=2.27, 95% CI: 2.13-2.42) than those with negative nodes. Women with ER+ or PR+ lesions were more likely to have ever used AI (P<0.0001) as were those with HER- tumors (AOR=1.67, 95% CI: 1.48-1.89). Women who underwent breast conserving surgery without radiation were 36% less likely to use AI (AOR=0.64 95% CI: 0.59-0.69) than those with radiation. Of the total cohort, nearly 64% received adjuvant treatment. Among the 6716 women who received tamoxifen, about 43% (2918/6716) also used AI. Of note, during the 2005-2006 diagnosis years, the percentage treated with tamoxifen dropped among women ever prescribed AI. Of the 6627 women treated with chemotherapy only, 36% also received AI (2354/6627).Conclusions: Odds of AI use were greater among older women (age 60+), white race, higher stage at diagnosis, worse histology, positive lymph nodes, and those with ER+, PR+, or HER- tumors. However, grade and tumor size did not influence AI use. More work is needed to determine how patterns of AI use alone or following tamoxifen, medication changes, and tumor characteristics at initial diagnosis impact long-term breast cancer recurrence in this diverse group of patients in southern California. Citation Information: Cancer Res 2009;69(24 Suppl):Abstract nr 2079.