• © 2003 by American Society of Clinical Oncology

In Reply:

In response to the letter by Tartorone et al, we would like to remark that as stated in our article,¹ our data cannot be considered conclusive but they reinforce previous observations on retrospective series suggesting that HER2 positivity is associated with responsiveness to anthracycline-based therapy. Because all of the controlled clinical trials retrospectively analyzed were not originally designed to address the issue of prediction of response to therapy by biomedical markers, and the frequency of the HER2 positivity is relatively low, such studies cannot be considered conclusive.^2–,⁴ Accordingly, no guidelines or recommendations have been drawn on retrospective studies, and HER2 positivity is not considered a predictive factor in selecting a therapeutic regimen. However, the majority of the data are in keeping with an important responsiveness of HER2-positive tumors to anthracyclines, and prospective clinical trials designed to verify the role of HER2 positivity in anthracycline-based regimens are ongoing. The study protocol did not plan the treatment to be administered at disease relapse, leaving the decision at the discretion of the treating physician. In actuality, heterogeneous treatments were applied at relapse, according to patient-tailored clinical considerations. Although relevant, a hypothesis concerning the effect of the timing of the treatment with anthracyclines cannot be evaluated in the present study.

In the meantime, in the attempt to understand the mechanism at the basis of the association between HER2 positivity and response to doxorubicin, in vitro studies have been performed. Our data showed that the major biologic parameter associated with in vitro response to the drug in breast cancer cell lines seems to be the tumor proliferation rate, whereas HER2 expression seems to serve only as a proliferation marker.⁵

The 2000 worldwide overview has confirmed that there is a reduction in the annual risk of disease relapse and death for anthracycline-based regimens. In this worldwide effort, however, no data were available for defining factors able to predict treatment response. New genomics and proteomics technologies will probably help in addressing the issue of prediction of response to therapy,^6,7 thus avoiding the delivery of ineffective drugs and contributing to the improvement of the risk ratio of adjuvant therapies. Indeed, a new subset classification according to gene expression profiles offers an incredibly potent new opportunity to explore predictive factors, considering not a single gene as was done with HER2 but the total genes and pathways expressed by a tumor to define its molecular portrait associated with sensitivity to antitumor therapy.