Response: Re: Topoisomerase II Alpha and Responsiveness of Breast Cancer to Adjuvant Chemotherapy

We are pleased to respond to the letter of Oakman, Moretti, Sotiriou, Viale, and Di Leo regarding our publication ( 1 ). We believe, as we stated in the conclusion of this article, that anthracycline benefit is associated with HER2 amplification and with TOP2A alterations. Certainly, our data and those of many others support the relationship between HER2 amplification or overexpression ( 2 ), TOP2A alterations ( 3 , 4 ), and anthracycline benefit. However, we also agree with Oakman et al. that assessment of TOP2A alterations is not ready to be incorporated into routine clinical practice because the assay is neither widely available nor standardized ( 5 , 6 ). Furthermore, we agree that there is still controversy concerning the circumstances in which HER2 serves as a robust biomarker, likely in part because of measurement issues, sample size, and publication bias. However, we believe that Oakman et al. have misunderstood the appropriate interpretation of our data because of a common misunderstanding of the difference between prognostic and predictive biomarkers ( 6–9 ). Yes, our data showed that patients with normal TOP2A gene status, regardless of therapy, tended to do better than patients with TOP2A gene alterations, who tended to have worse outcomes. However, when our data were adjusted for other baseline prognostic factors such as age, tumor size and stage, number of positive nodes, estrogen receptor protein levels, and HER2 amplification, the presence of TOP2A gene alteration was not a prognostic factor, that is, not associated with the clinical course of the disease, regardless of treatment with cyclophosphamide, epirubicin, and 5-fluorouracil (CEF) or with cyclophosphamide, methotrexate, and 5-fluorouracil (CMF). In fact, as these authors themselves describe in their letter, patients with TOP2A gene alterations had statistically significantly better overall survival when treated with CEF than with CMF (hazard ratio [HR] = 0.33, 95% confidence interval [CI] = 0.15 to 0.75, P = .008), whereas there was no statistically significant difference in benefit from CEF vs CMF among women whose tumors had normal TOP2A gene status (HR = 1.09, 95% CI = 0.77 to 1.56, P = .62). The difference between the two hazard ratios was statistically significant, and such a difference is the evidence required for a biomarker to be predictive ( 1 ). We agree with Oakman et al. that an individual patient as reported in our data had either amplification or deletion but not both. However, when the amplification and the deletion groups were combined, that combined group had statistically significantly greater benefit from CEF compared with CMF than those with normal TOP2A status tumors. Others have made this same observation and have combined TOP2A deletion and amplification groups in similar analyses ( 3 ). We agree that the topoisomerase 2α protein, not the gene, is the anthracycline drug target. We have published data showing that whereas HER2 gene amplification does correlate with HER2 protein levels ( 7 ), TOP2A amplification or deletion does not correlate with topoisomerase 2α protein levels ( 10 , 11 ). We also reported ( 11 ) that topoisomerase 2α protein levels may be correlated with additional benefit from CEF compared with CMF just as well as but not better than HER2 amplification or TOP2A gene alterations. We agree that more accurate measurements of topoisomerase 2α protein quantity might show a stronger association with the efficacy of anthracyclines, but there are no strong clinical data to support this hypothesis at present. Of the two references that Oakman et al. quote to support this claim, one describes work in cultured cells ( 12 ) and the other cannot assess the predictive value of any biomarker in relation to anthracycline benefit because virtually all patients in the study were treated with an anthracycline ( 13 ). In summary, we agree with Oakman et al. that there are not fully reliable clinical biomarkers to predict outcome from anthracycline-based chemotherapy. We do not agree, however, with the authors’ misinterpretation of our data as described in paragraph two of their letter or with their confusion between predictive and prognostic factors in clinical trials.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.