Abstract P5-10-12: Adherence, Persistence and Toxicity with Oral CMF in Older Women with Early Stage Breast Cancer in CALGB 49907 (Adherence Companion Study 60104)

Abstract

Abstract Background: Cyclophosphamide-methotrexate-fluorouracil (CMF) is often selected as adjuvant chemotherapy for older patients with early stage breast cancer due to perceived better tolerability compared to other standard regimens such as doxorubicin-cyclophosphamide (AC). We sought to measure adherence and persistence with oral cyclophosphamide and the relationship with toxicity among women who were prescribed CMF in CALGB49907. Patients and Methods: CALGB49907 (NEJM 2009) was a randomized trial that compared standard chemotherapy (CMF or AC) versus capecitabine in patients age ≥65 with stage 1-3 breast cancer. Those randomized to standard chemotherapy were treated with either AC or CMF. For those given CMF, cyclophosphamide was prescribed at 100 mg/m2 daily for 14 consecutive days of each of six 28 day cycles. Persistence was defined as completing all six cycles. Adherence was measured using patient self-report medication calendars, and was calculated as the number of doses that women reported they had taken divided by the doses they were prescribed, taking into account toxicity-related dosing changes. Results: Of the 317 patients randomized to standard chemotherapy, 133 elected to receive CMF. CMF patients were older (p=0.009) and less likely to be Hispanic (P<0.001) than patients who received AC. Median age of CMF patients was 73 (range 65 to 88); only 86 patients (65%) persisted with CMF to completion of planned protocol therapy. In a multivariable logistic regression model, non-persistence was associated with the report of a non-hematologic adverse event (p <0.001), but not with a hematologic adverse event (p = 0.578). Febrile neutropenia occurred in 21% of the non-persistent patients, but only in 1 % of the persistent patients (P<0.001). Fatigue also occurred in 21% of the non-persistent patients, but only in 6% of the persistent patients (p=0.009). Other non-significant covariates included age, race, performance status at enrollment, type of surgery, nodal status, ER status, and PgR status. Ninety-four of 133 (71%) CMF-treated patients submitted medication calendars documenting their use of cyclophosphamide. The only characteristic that was found to differ between women who submitted medication calendars and those who did not was that the former were less likely to be ER-positive (p=0.010). Average adherence across all cycles with available data was 97%, and adherence did not vary by cycle (p=0.46). Five percent of participants reported taking fewer than 80% of expected doses and were deemed non-adherent. In a logistic regression model, there were no significant variables predicting non-adherence. Conclusions: In this substudy of a large clinical trial of older women with early stage breast cancer, self-reported adherence to oral cyclophosphamide as part of standard CMF chemotherapy was high. However, only 65% of older women were persistent with CMF, lower than the 92% observed among patients who elected to receive AC in the trial. Further research is needed to study how non-persistence impacts outcomes in this population and whether persistence can be increased by improving management of toxicities in older women. Citation Information: Cancer Res 2010;70(24 Suppl):Abstract nr P5-10-12.