Baseline Symptoms and Adherence to Oral Breast Cancer Therapy among Older Women

点击翻译按钮获取中文摘要

For the >200,000 older women who develop hormone receptor positive breast cancer yearly, adjuvant endocrine therapy (AET) for at least five years increases overall survival. Yet almost half of patients prescribed AET fail to take all their pills. Education- and reminder-based studies in this and a number of other settings have had disappointingly small effects on adherence, and more complex interventions have proven difficult to sustain. In this proposal we seek a better understanding of the symptoms that women describe as major barriers to AET adherence. We draw on prior research suggesting that preexisting symptoms may play an important role in patients’ experience with AET. Our proposal fills crucial gaps in our understanding by examining older and less selected subjects treated in a usual care practice, and by using a measure that is already routinely collected across the United States. Specifically, we will use the National Comprehensive Cancer Network’s Distress Thermometer (DT) and its accompanying problem list, which have previously been shown to reveal a high burden of symptoms among older women with breast cancer but have not been examined in association with AET or while accounting for the wide spectrum of comorbid conditions of older women. Most cancer centers administer the DT and problem list at diagnosis or earlier in care, increasing their value for informing early intervention to help women continue their AET. We will investigate whether the DT and problem list are predictive of AET discontinuation among older women, with specific aims as follows: Aim 1 To examine the association of patient-reported distress at diagnosis with early discontinuation of AET. We will utilize electronic health record data from a cohort of incident postmenopausal hormone receptor positive breast cancer patients to examine the association of the DT (scored 0 to 10) measured at diagnosis with AET discontinuation in time-to-event analyses, accounting for age, osteoarthritis, and other comorbidities, and other breast cancer treatments. Aim 2. To explore the association of patient-reported symptoms at diagnosis with discontinuation of AET. In the same cohort and using the same general approach, we will explore AET discontinuation and a) patient-reported overall symptom count based on DT problem list) and b) symptom clusters (e.g., fatigue, sleep, concentration). We will use both standard regression and an innovative application of machine learning using Bayesian Additive Regression Trees. With this approach, we will produce immediately actionable findings to improve adherence outcomes, by providing a means of identifying high-risk patients who are the primary targets for the development of new interventions. It could also be an important model for other conditions that are treated with aggressive treatment regimens in older patients, informing care for a large number of older patients.