As reported by Hope S. Rugo, MD, and John W. Park, MD, both of the University of California, San Francisco, and colleagues in The New England Journal of Medicine, the multiarm adaptive randomization phase II I-SPY 2 trial has shown that the addition of veliparib/carboplatin and the addition of neratinib to standard neoadjuvant therapy have met criteria predictive of success in a phase III trial of the same comparators in early breast cancer patients with specific biomarker signatures.1,2
I-SPY 2 Details
In the I-SPY 2 trial, multi-arm adaptive randomization is being used to evaluate the effectiveness of multiple new agents added to standard neoadjuvant chemotherapy in producing pathologic complete response in high-risk stage II to III breast cancer based on biomarker subtypes. The aim of the multi-arm adaptive randomization scheme is to more rapidly and accurately identify experimental regimens with a higher likelihood of success in phase II trials that can include smaller numbers of patients.
The breast cancers are categorized into 8 biomarker subtypes based on HER2, hormone receptor, and 70-gene assay status. Patients undergo adaptive randomization within each biomarker subtype to receive regimens that perform better than standard therapy. The regimens are evaluated across 10 biomarker signatures consisting of predefined combinations of biomarker subtypes. A regimen leaves the trial (“graduates”) when it reaches an 85% Bayesian predictive probability of success in a simulated 300-patient equally randomized phase III trial of the same comparators within a biomarker signature.
Veliparib/Carboplatin in Triple-Negative Disease
As reported by Rugo and colleagues,1 veliparib/carboplatin became the first regimen to graduate from the trial. The veliparib/carboplatin regimen was considered only for HER2-negative tumors and was thus evaluated in the three biomarker signatures of HER2-negative, hormone receptor–positive and HER2-negative, and triple-negative disease.
All patients in the trial received standard weekly paclitaxel alone (or with trastuzumab [Herceptin] in HER2-positive disease) or in combination with an experimental regimen, and all subsequently received standard doxorubicin and cyclophosphamide. Overall, 75 patients were randomized to receive veliparib (50 mg twice daily) plus carboplatin (AUC [area under the curve] = 6 on weeks 1, 4, 7, and 10) plus standard treatment; 46 were randomized to receive paclitaxel; and 18 were randomized to receive paclitaxel/trastuzumab as the control treatment. Veliparib/carboplatin was not assigned to patients with HER2-positive tumors. Comparisons were made among 72 patients receiving veliparib/carboplatin and 44 control patients receiving paclitaxel without trastuzumab.Error loading Partial View script (file: ~/Views/MacroPartials/TAP Article Portrait and Quote.cshtml)
The predefined threshold for efficacy for veliparib/carboplatin was achieved among patients with triple-negative disease, with pathologic complete response rates of 51% vs 26%; the probability that veliparib/carboplatin was superior to the control was 99%, and the probability of statistical success in an equally randomized phase III trial including 300 patients was 88%. Although the predefined threshold for efficacy was not met, the addition of veliparib/carboplatin showed evidence of benefit among all patients with HER2-negative tumors (33% vs 22%; 91% probability of superiority, 53% predicted probability of success in a phase III trial). Estimated rates of pathologic complete response were 14% vs 19% in patients with HER2-negative, hormone receptor–positive disease.
The investigators concluded: “The process used in our trial showed that veliparib-carboplatin added to standard therapy resulted in higher rates of pathological complete response than standard therapy alone, specifically in triple-negative breast cancer.”
Neratinib in HER2-Positive, Hormone Receptor–Negative Breast Cancer
As reported by Park and colleagues,2 all patients, again, received standard paclitaxel (with trastuzumab in HER2-positive patients) followed by doxorubicin and cyclophosphamide. Neratinib, a HER2/EGFR inhibitor, was given at 240 mg/d for the first 12 weeks in addition to standard chemotherapy. Overall, 115 patients received neratinib, and 78 received the control treatment (paclitaxel alone in 56 patients and with trastuzumab in 22 patients).
Among the 10 biomarker signatures evaluated, the neratinib group achieved the predefined threshold for efficacy in patients with HER2-positive, hormone receptor–negative disease; in these patients, pathologic complete response was observed in 56% vs 33% of patients. The probability that neratinib was superior to the control treatment was 95%, and the probability of the success of neratinib in a phase III trial in 300 patients was 79% (the 85% threshold had been reached, but the probability was reduced somewhat when all randomized patients had completed neoadjuvant treatment).
Although the prespecified threshold for efficacy was not achieved, the addition of neratinib showed evidence of benefit over control treatment in HER2-positive, hormone receptor–positive disease (30% vs 17%; 91% probability of superiority, 65% predicted probability of success in a phase III trial) and in HER2-positive disease irrespective of hormone receptor status (39% vs 23%; 95% probability of superiority, 73% predicted probability of success in a phase III trial).
The investigators concluded: “Neratinib added to standard therapy was highly likely to result in higher rates of pathological complete response than standard chemotherapy with trastuzumab among patients with HER2-positive, hormone receptor–negative breast cancer.” ■
Disclosure: The study is funded by QuantumLeap Healthcare Collaborative and others. For full disclosures of the study authors, visit www.nejm.org.
Error loading Partial View script (file: ~/Views/MacroPartials/TAP Article Portrait and Quote.cshtml)
With the expansion of our understanding of signaling pathways in normal cells and how they are co-opted or corrupted in malignancy, the number of potential antitumor agents to be tested has...