Are <i>HER2</i> and <i>TOP2A</i> Useful As Prognostic or Predictive Biomarkers for Anthracycline-Based Adjuvant Chemotherapy for Breast Cancer?

There is an old maxim that says one should first tell the audience what you are going to tell them, then tell it to them, and then tell them what you have told them. In that vein, I shall begin by saying that the answer to the question in the title of this editorial is no. The reasons are that published data to support the use of human epidermal growth factor receptor 2 (HER2) or topoisomerase (DNA) II alpha 170kDa (TOP2A) as targets are currently contradictory— more so than a year ago—and thatTOP2A measurement is not a standardized technique that is widely available in North American or European pathology laboratories. Going back 10 years or more, a number of investigators published prospectively designed retrospective explorations of the role of HER2 as a predictive factor for differential response of adjuvant therapy of breast cancer with anthracycline-containing compared with non–anthracycline-containing regimens. Although these studies were limited by tissue availability (60% to 95% of cases had tissue available) and by power within individual trials which were originally sized to answer clinical comparisons (but not predictive substudies), results initially seemed consistent. Several of these trials suggested that HER2 protein overexpression or gene amplification was associated with a significantly greater response to anthracyclinecontaining therapy. Indeed, meta-analyses based on all published studies concluded that although many of these substudies did not meet individual significance that meta-analysis technique supported those studies that did in suggesting that HER2 status was a significantly predictive factor for better response to anthracyclinecontaining than to non–anthracycline-containing chemotherapy. Paradoxically, however, a large randomized British trial of adjuvant therapy for breast cancer which showed superiority overall for the anthracycline-containing versus a non–anthracycline-containing regimen was mined to examine the role of HER2 with opposite results. As part of an ongoing project to carry out an individual patient meta-analysis on this subject, we became aware of these contradictory results before they were published, and explored together all possible technical and methodologic reasons for this discrepancy. No explanation could be found for this contradiction, but it did become clear that at least one other published paper supporting the predictive value of TOP2A in this setting also contained data which suggested that HER2 was not predictive in a trial very similar to one which clearly suggested that it did. TOP2A was explored by several of the same clinical trialists who mined their data with regard to the predictive rate of HER2 in trials of anthracycline-containing versus non–anthracycline-containing therapy and in these studies, it seemed also clear and consistent that TOP2A gene amplification was predictive for a benefit of anthracycline-containing versus non–anthracycline-containing therapy. Indeed, this was thought to explain the association of HER2 with this predictive value since TOP2A and HER2 are closely located on chromosome 17 and TOP2A is a known target of the anthracyclines, whereas an association between HER2 amplification and anthracycline response has no obvious direct explanation. However, the same British trial that showed an opposite association with HER2 also failed to confirm the positive value of TOP2A. To further complicate these contradictions, some trials showed a very clear correlation between the predictive results of HER2 and TOP2A alterations (amplification/deletion) while others suggest positive predictive value for TOP2A alterations but not for HER2 amplification. Furthermore, while Slamon and Press reported that only in tumors with HER2 amplification will TOP2A alterations be seen, other studies do not show such a consistent association. The relevance of TOP2A deletion versus amplification is also not clear due to inconsistent results. These differences may relate in part to technical measurement issues. Interest in the role of HER2 and TOP2A as prognostic and particularlyaspredictive factors for the use of anthracycline-containing versus non–anthracycline-containing therapy continues apace. The article published by Tubbs et al in this issue of Journal of Clinical Oncology adds information to the body of knowledge concerning the prognostic value of these biomarkers in women treated with anthracyclinecontaining regimens. But the authors themselves make clear that these data cannot add to any knowledge concerning the predictive value of either of these markers because of the underlying design of the study in which this question was explored. As the senior author of this article has so clearly outlined for us in previous articles, while prognostic factors identify patients whose tumors are associated with differing outcomes, predictive factors are those which can identify different outcomes in association with particular therapies. Predictive factors for a given therapy, anthracyclines in this case, can only be identified in trials in which patients are randomly assigned to at least one arm which does and one arm which does not contain the therapy of JOURNAL OF CLINICAL ONCOLOGY E D I T O R I A L VOLUME 27 NUMBER 24 AUGUST 2

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.