The advantage of letrozole over tamoxifen in the BIG 1-98 trial is consistent in younger postmenopausal women and in those with chemotherapy-induced menopause

Abstract

Letrozole, an aromatase inhibitor, is ineffective in the presence of ovarian estrogen production. Two subpopulations of apparently postmenopausal women might derive reduced benefit from letrozole due to residual or returning ovarian activity: younger women (who have the potential for residual subclinical ovarian estrogen production), and those with chemotherapy-induced menopause who may experience return of ovarian function. In these situations tamoxifen may be preferable to an aromatase inhibitor. Among 4,922 patients allocated to the monotherapy arms (5 years of letrozole or tamoxifen) in the BIG 1-98 trial we identified two relevant subpopulations: patients with potential residual ovarian function, defined as having natural menopause, treated without adjuvant or neoadjuvant chemotherapy and age ≤ 55 years (n = 641); and those with chemotherapy-induced menopause (n = 105). Neither of the subpopulations examined showed treatment effects differing from the trial population as a whole (interaction P values are 0.23 and 0.62, respectively). Indeed, both among the 641 patients aged ≤ 55 years with natural menopause and no chemotherapy (HR 0.77 [0.51, 1.16]) and among the 105 patients with chemotherapy-induced menopause (HR 0.51 [0.19, 1.39]), the disease-free survival (DFS) point estimate favoring letrozole was marginally more beneficial than in the trial as a whole (HR 0.84 [0.74, 0.95]). Contrary to our initial concern, DFS results for young postmenopausal patients who did not receive chemotherapy and patients with chemotherapy-induced menopause parallel the letrozole benefit seen in the BIG 1-98 population as a whole. These data support the use of letrozole even in such patients.

Trial registration: ClinicalTrials.gov NCT00004205.

Fig. 1

Kaplan–Meier estimates of disease-free survival (DFS) according to treatment group for the 105 patients with chemotherapy-induced menopause at the time of entry into the BIG 1-98 trial

Fig. 2

Univariate Cox model estimates of hazard ratios for disease-free survival (DFS; letrozole vs. tamoxifen) among all monotherapy patients [10], patients with chemotherapy-induced menopause, and patients with age ≤55 years old at randomization with natural menopause who received no chemotherapy. The box size is proportional to the inverse of the standard error of the hazard ratio estimates. The horizontal line gives the 95% confidence interval

Fig. 3

Kaplan–Meier estimates of disease-free survival (DFS) according to treatment group for the 641 patients with potential residual ovarian function the BIG 1-98 trial, defined as ≤55 years old at randomization with natural menopause who received no chemotherapy

Fig. 4

Subpopulation Treatment Effect Pattern Plots (STEPP) of 5-year disease-free survival (DFS) percents for letrozole vs. tamoxifen for overlapping subpopulations defined according to age at randomization for all 3633 patients with natural menopause who received no chemotherapy. The x-axis is the median age for overlapping subpopulations, each including approximately 300 patients. The P values, based on simulations, test for the interaction of age and treatment group