Abstract S3-04: The change from brand-name to generic aromatase inhibitors and hormone therapy adherence for early stage breast cancer

Abstract

Abstract Background: Non-adherence to adjuvant hormonal therapy is common and has been associated with increased copayment amount. We investigated the change in adherence patterns before and after the introduction of generic aromatase inhibitors in 2009. Patients and Methods: Using de-identified, integrated pharmacy and medical claims data from OptumInsight, we identified women >50 years old on hormonal therapy for early breast cancer with at least 2 mail order prescription refills between 1/1/07 and 12/31/11. We categorized hormonal therapy as tamoxifen (TAM), aromatase inhibitor (AI) brand and AI-generic. Variables that were evaluated included demographic and clinical information, household income, number of prescriptions, diagnostic codes, prescription refill dates, provider specialty, coverage type, deductible amount and co-payment amounts. Monthly copayment amount was categorized as <$10, $10-$20, and ≥$20. Non-adherence was defined as a medication possession ratio <80% of eligible days during the first year after initiation. We compared patients who completed their AI prior to July 2009 (pre-generic) to those who initiated after July 2009 (post-generic). Results:We identified 3,351 subjects (1,658 pre-generic; 1,693 post-generic).In the pre-generic cohort, 28.5% took tam and 65.3% took AI-brand for at least 1 year. In the post-generic cohort, there was minimal change in TAM (26.5%); however, only 7.3% remained AI-brand while 67.3% became AI-generic (13.2% switched from brand to generic). Non-adherence increased from 32.7% with copayment <$10/month compared to 44.4% with copayment >$20/month (P<0.0001). For TAM, 62% had a copay <$10 while <1% had a copay >$20; however, for AI brand 75.5% had a copay >$20. There was a non-significant increase in non-adherence of TAM and AI-brand over time. However, non-adherence to AI-brand was 40.3% and AI-generic was lower at 31.4% (P<0.001). In an unadjusted model, AI-brand was associated with decreased adherence (OR = 0.66, P<0.001) and AI-generic was associated with increased adherence ((OR 1.66, P<0.001) compared to TAM. However, in a multivariate model controlling for copayment, only the association between increased adherence with AI-generic (OR 1.65, P<0.001) remained significant. Adherence was also associated with middle-household income (OR 1.37, P<0.01) and high-household income (OR 1.39, P = 0.03) compared to low income. Monthly copayment amount was inversely associated with adherence with $10-$20 (OR = 0.70; P<0.001) and >$20 (OR = 0.49; P<0.001) compared to <$10. In this cohort, adherence was not associated with race, insurance type, age or comorbidity. Conclusions:We found that a higher prescription copayment amount was associated with non-adherence to AI. After controlling for copayment and time, adherence was similar between AI-brand and TAM, but adherence was significantly improved for AI-generic. In addition, other financial factors, such as copayment amount and income, were associated with hormone therapy adherence. Because hormone therapy non-adherence is associated with worse survival outcomes, future public policy efforts should be directed towards reducing financial constraints as a means of increasing the proper use of these life-saving medications. Citation Information: Cancer Res 2013;73(24 Suppl): Abstract nr S3-04.