Frailty and adherence to adjuvant hormonal therapy in older women with breast cancer: CALGB protocol 369901

Vanessa B Sheppard,Leigh Anne Faul,George Luta,Jonathan D Clapp,Rachel L Yung,Judy Huei-Yu Wang,Gretchen Kimmick,Claudine Isaacs,Michelle Tallarico,William T Barry,Brandelyn N Pitcher,Clifford Hudis,Eric P Winer,Harvey J Cohen,Hyman B Muss,Arti Hurria,Jeanne S Mandelblatt
DOI: https://doi.org/10.1200/JCO.2013.51.7367
2014-08-01
Abstract:Purpose: Most patients with breast cancer age ≥ 65 years (ie, older patients) are eligible for adjuvant hormonal therapy, but use is not universal. We examined the influence of frailty on hormonal therapy noninitiation and discontinuation. Patients and methods: A prospective cohort of 1,288 older women diagnosed with invasive, nonmetastatic breast cancer recruited from 78 sites from 2004 to 2011 were included (1,062 had estrogen receptor-positive tumors). Interviews were conducted at baseline, 6 months, and annually for up to 7 years to collect sociodemographic, health care, and psychosocial data. Hormonal initiation was defined from records and discontinuation from self-report. Baseline frailty was measured using a previously validated 35-item scale and grouped as prefrail or frail versus robust. Logistic regression and proportional hazards models were used to assess factors associated with noninitiation and discontinuation, respectively. Results: Most women (76.4%) were robust. Noninitiation of hormonal therapy was low (14%), but in prefrail or frail (v robust) women the odds of noninitiation were 1.63 times as high (95% CI, 1.11 to 2.40; P = .013) after covariate adjustment. Nonwhites (v whites) had higher odds of noninitiation (odds ratio, 1.71; 95% CI, 1.04 to 2.80; P = .033) after covariate adjustment. Among initiators, the 5-year continuation probability was 48.5%. After adjustment, the risk of discontinuation was higher with increasing age (P = .005) and lower for stage ≥ IIB (v stage I) disease (P = .003). Conclusion: Frailty is associated with noninitiation of hormonal therapy, but it does not seem to be a major predictor of early discontinuation in older patients.
What problem does this paper attempt to address?