Reply to T. Shimoi et al and Y. Shimanuki et al

We appreciate the interest expressed by Shimoi et al¹ and Shimanuki et al² in our recent manuscript.³ They raise thoughtful issues around two main questions. First, might there be a subset of patients who benefit from platinum therapy? Second, why was disease-free survival in capecitabine-treated patients in EA1131 so much worse than in capecitabine-treated patients in the CREATE-X⁴ trial? In both cases, we stand by our reported conclusions.

Negative clinical trial results often spawn the search for subsets of patients who might benefit from the therapies in question. Shimoi et al¹ wonder specifically about the accuracy of subtyping used in EA1131, and whether we may not have focused on the optimal subtype. In addition, they correctly point out that patients who harbor deleterious BRCA mutations may be uniquely sensitive to DNA-damaging agents such as the platinums.⁵

Molecular classification schemes have differed in how they define basal subtype. However, there is substantial overlap, and nearly 40% of basal-like breast cancers as determined by the PAM50 classification used in EA1131 would be classified as either basal-like 1 or 2 in the Lehmann classification.⁶ At the time of protocol development in 2013, germline BRCA mutation testing was not yet recommended for all patients with triple-negative breast cancer (TNBC). However, we suspect that patients with BRCA-mutated cancers are under-represented in EA1131, as they (1) are more likely to achieve a pathologic complete response to anthracycline- or alkylating-based neoadjuvant chemotherapy, (2) are more likely to have been treated with a platinum agent in the neoadjuvant setting, and (3) are likely to have preferentially participated in the OlympiA trial⁷ of adjuvant olaparib versus observation. All these factors (alone or in combination) would have made them ineligible for EA1131. Regardless, analyses of tumor specimens banked as part of EA1131 are underway and will explore the Lehmann classification, BRCA mutation status, and other potential markers of platinum sensitivity.

Cross-trial comparisons are also fraught with danger and require careful analyses of differences in the patient population and/or treatment that may contribute to differences in outcome. Many factors undoubtedly contribute to the poor invasive disease-free survival in EA1131 compared with CREATE-X,⁴ such as the purposeful selection of higher-risk patients in EA1131 (ie, 100% of EA1131 participants had residual disease in the surgical specimen corresponding to Residual Cancer Burden [RCB] II-III, whereas only 58% of CREATE-X participants had similar amounts of residual disease) and the greater patient diversity in trial EA1131 (eg, 100% of CREATE-X participants were Asian, whereas 20% of EA1131 participants were Black, who are 30% more likely to die from breast cancer than their Non-Hispanic White counterparts⁸ even after controlling for socioeconomic status and access to care).

In the CREATE-X trial, 80% patients were able to complete capecitabine at 1,250 mg/m² twice daily × eight cycles. In the EA1131 trial, we used a dose of capecitabine that is commonplace in the United States (1,000 mg/m² twice daily × six cycles), but despite these modifications, about 20% of patients were unable to complete treatment without dose delays or adjustments. More than 85% of patients in EA1131 received neoadjuvant anthracyclines, not dissimilar to the 94% figure observed in CREATE-X, but we recognize that median time from surgery to initiation of adjuvant therapy was relatively long in EA1131 (120 days), although such data were not reported in CREATE-X (albeit patients were allowed to initiate treatment up to 120 days from surgery). Although these issues in total compound on the challenges of doing cross-trial comparisons, they have no impact on our ability to compare platinum to capecitabine when administered in the postsurgical post neoadjuvant chemo setting as in EA1131. As such, the findings from EA1131 are unequivocal.

In summary, we do not believe that the factors raised by Shimoi et al¹ explain the negative results reported at ASCO 2021 and published in Journal of Clinical Oncology, and we affirm our conclusion that platinum agents should not be administered in the postneoadjuvant chemo setting for patients with residual TNBC identified at surgery. The use of platinum agents as a component of neoadjuvant or adjuvant therapy (without prior exposure to neoadjuvant chemo) remains investigational, and clinical trials addressing these questions are ongoing. Most importantly, patients with larger volumes of residual disease postneoadjuvant chemotherapy (eg, RCB II-III) have a much worse long-term outcome than originally perceived, despite the benefit of adjuvant capecitabine first observed in the CREATE-X trial, which reminds us of the critical need for more active and effective therapies for patients with high-risk TNBC.

AUTHORS' DISCLOSURES OF POTENTIAL CONFLICTS OF INTEREST

Reply to T. Shimoi et al and Y. Shimanuki et al

The following represents disclosure information provided by authors of this manuscript. All relationships are considered compensated unless otherwise noted. Relationships are self-held unless noted. I = Immediate Family Member, Inst = My Institution. Relationships may not relate to the subject matter of this manuscript. For more information about ASCO's conflict of interest policy, please refer to www.asco.org/rwc or ascopubs.org/jco/authors/author-center.

Open Payments is a public database containing information reported by companies about payments made to US-licensed physicians (Open Payments).