Background: Previous economic evaluations compared specific chemotherapy agents using input parameters from clinical trials and resource utilization costs. Cost-effectiveness of treatment groups (drug classes) using community-level effectiveness and cost data, however, has not been assessed for elderly patients with breast cancer.
Objective: To assess the cost-effectiveness of chemotherapy regimens by age and disease stage under “real-world” conditions for patients with breast cancer.
Methods: The Surveillance Epidemiology and End Results-Medicare data were used to identify patients with breast cancer with American Joint Committee on Cancer stage I/II/IIIa, hormone receptor–negative (estrogen receptor–negative and progesterone receptor–negative) patients from 1992 to 2009. Patients were categorized into three adjuvant treatment groups: 1) no chemotherapy, 2) anthracycline, and 3) non–anthracycline-based chemotherapy. Median life-years and quality-adjusted life-years (QALYs) were measured using Kaplan-Meier analysis and were evaluated against average total health care costs (2013 US dollars).
Results: A total of 4575 patients (propensity score–matched) were included for the primary analysis. The anthracycline group experienced 12.05 QALYs and mean total health care costs of $119,055, resulting in an incremental cost-effectiveness ratio of $7,688 per QALY gained as compared with the no chemotherapy group (QALYs 7.81; average health care cost $86,383). The non–anthracycline-based group was dominated by the anthracycline group with lower QALYs (9.56) and higher health care costs ($122,791). Base-case results were found to be consistent with the best-case and worst-case scenarios for utility assignments. Incremental cost-effectiveness ratios varied by age group (range $3,790–$90,405 per QALY gained).
Conclusions: Anthracycline-based chemotherapy was found cost-effective for elderly patients with early stage (stage I, II, IIIa) breast cancer considering the US threshold of $100,000 per QALY. Further research may be needed to characterize differential effects across age groups.
