Background: Emerging data suggest that treatment outcomes with aromatase inhibitors (AIs) and/or tamoxifen may differ for tumors that express both the estrogen receptor (ER) and the progesterone receptor (PR) (ER+/PR+) compared with those that lack PR expression (ER+/PR-). However, the optimal sequencing of AIs and tamoxifen as adjuvant therapy is not known and may differ for biologic subsets of cancers.
Methods: Markov models were used to simulate disease-free survival (DFS) separately among postmenopausal women with ER+/PR+ cancers and women with ER+/PR- cancers. By using risk estimates reported from randomized clinical trials, treatment with 5 years of an AI alone with sequential treatment consisting of tamoxifen with crossover to an AI at 2 years was compared.
Results: For women with ER+/PR+ cancers, sequential therapy with tamoxifen followed by crossover to an AI at 2 years yielded modest improvements in 10-year DFS estimates compared with planned AI monotherapy (84.3% vs. 82.2% and 68.8% vs. 64.8% for lymph node-negative and lymph node-positive patients, respectively). However, for women with ER+/PR- cancers, upfront treatment with an AI yielded improved outcomes with 10-year DFS rates of 90.5% and 80.1% for the lymph node-negative and node-positive groups, respectively, compared with 88.2% and 76.1%, respectively, for sequential treatment with tamoxifen followed by an AI.
Conclusions: Modeling estimates suggested that the optimal endocrine treatment strategy may differ based on the biologic features of breast cancer tumors. Patients with ER+/PR+ tumors achieved optimal 10-year DFS estimates with tamoxifen followed by a crossover to AI therapy, whereas patients with ER+/PR- tumors fared best when they initiated treatment with AI.
Copyright 2006 American Cancer Society.