Should Adjuvant Weekly Paclitaxel Be Considered Less Efficacious Than Anthracyclines Plus Cyclophosphamide For Lower-Risk Patients With Early-Stage Breast Cancer?

TO THE EDITOR: The report of the Cancer and Leukemia Group B (CALGB) 40101 trial has shown that doxorubicin plus cyclophosphamide (AC) is potentially superior to weekly paclitaxel (T) suggesting that anthracyclines should not be abandoned in the management of early-stage breast cancer, even in subsets with a low/moderate risk of relapse as seen in those included in this study. Besides the methodological issues of all trials with a factorial design discussed in the accompanying editorial, we have noticed that patients with either human epidermal growth factor receptor 2 (HER2) -positive or HER2-negative breast cancer were recruited in the CALGB 40101 trial without prior stratification. In our opinion, HER2 status may be a potential confounder in the analysis testing the noninferiority of T compared with AC. The prognosis of HER2-positive breast cancers in the absence of trastuzumab is notoriously worse compared with HER2-negative tumors, but HER2 positivity may also have a predictive value of efficacy of anthracyclines. A published meta-analysis of adjuvant trials with individual patient data demonstrated a quantitative interaction between anthracycline activity and HER2 status. Furthermore, in relation to taxanes, a retrospective analysis of the CALGB 9344/INT0148 trial showed a significant benefit with the addition of paclitaxel after adjuvant therapy with doxorubicin and cyclophosphamide only in patients with tumors that had HER2 overexpression or amplification. To our knowledge, no data on the relative efficacy of anthracyclines over paclitaxel according to HER2 status are available up to now. The CALGB 40101 trial initiated in the early 2000s, and at that time HER2 status was not considered a stratification variable. As a consequence, only 48% of the 3,871 patients enrolled onto this study had their HER2 status available. This limitation notwithstanding, we believe that it would be clinically relevant to explore the hypothesis that the relative efficacy of AC compared with T might be different according to HER2 status in the CALGB 40101 data set. This single subgroup-specific analysis would involve approximately 1,800 women and might potentially have adequate statistical power. Treatment-associated toxicities are an important concern, particularly when adjuvant therapy is prescribed in a patient with breast cancer with low risk of relapse, as Shulman et al pointed out in their article. If the advantage of AC over T is mainly confined to HER2positive tumors, then there would be a rationale for reconsidering the less toxic regimen with single-agent T as a possible option for HER2negative patients, which should be tested in a future prospective randomized trial.