I-SPY 2 Graduates its First Experimental Regimen

Designing resource efficient development programs to identify drug candidates that should be advanced to late stage development and to determine the most useful trial design and/or target patient population for a Phase III trial are ongoing challenges for drug manufacturers. The I-SPY 2 trial is an innovative “find a winner” trial that aims to address some of these challenges by screening a number of different novel experimental agents in the neoadjuvant setting in combination with a standard chemotherapy regimen of paclitaxel followed by AC compared to standard chemotherapy alone. In this novel trial design, the randomization of patients to treatment arms is adaptive, with randomization probabilities continuing to be updated as the trial progressed. When an experimental treatment arm has enrolled 60-120 patients, an algorithm is used to decide whether that experimental regimen should “graduate.”  Graduating from the trial is defined as having at least an 85% likelihood of success in a randomized Phase III trial in the neoadjuvant setting with a target enrollment of 300 patients and an endpoint of pathological complete response (pCR).  Based on biomarker/subtype testing of patients after enrollment, experimental agents can graduate in a variety of patient segments, such as HER2+/HR-, triple negative disease, and HER2-/HR+.

Multiple pharmaceutical companies are participating in this trial, which to date has been investigating seven different experimental regimens.  Results from the first experimental regimen to graduate, the combination of the PARP inhibitor veliparib (AbbVie) and carboplatin, were presented at 2013 San Antonio Breast Cancer Symposium (SABCS) annual meeting on Friday, December 13.1 Enrollment to the veliparib plus carboplatin experimental arm was only allowed for HER2- patients, making the regimen eligible to graduate in three different patient populations: all HER2- patients, HER2-/HR+ patients, and triple negative patients. A total of 72 patients were randomized to receive veliparib plus carboplatin in combination with standard neoadjuvant chemotherapy, and 44 HER2- patients who were concurrently randomized to receive standard neoadjuvant chemotherapy alone were used as a control.

The addition of veliparib and carboplatin to standard neoadjuvant chemotherapy improved the estimated pCR in triple negative patients, but not in HER2-/HR+ patients. Although actual pCR rates in the study arms were not calculated because of a concern over bias due to the adaptive randomization in the trial design, a Bayesian model was used to provide estimated pCR rates. In HER2-/HR+ patients, the estimated pCR rate was 14% in the veliparib/carboplatin arm and 19% in the control arm, with a 28% probability that the combination is superior to control. In this patient population, the veliparib/carboplatin arm met the criteria for futility, with a 9% predictive probability of success in a Phase III trial. In contrast, the estimated pCR rate in triple negative patients was 52% in the veliparib/carboplatin arm compared to 26% in the control arm, with a 99% probability that the combination is superior to the control. In this patient population, the veliparib/carboplatin arm met the trial’s criteria for graduation, with a 90% predictive probability of success in a Phase III trial.

While the utility of the I-SPY 2 trial design will not be able to be fully evaluated until after the drug regimens in multiple experimental arms either meet criteria for futility or graduate and are subsequently advanced through successful Phase III programs, these data provide an early signal of the benefits of this novel trial design. With the efficient, collaborative design of this trial, data were able to be provided on the efficacy and potential Phase III success of the veliparib/carboplatin regimen in multiple patient segments based on enrollment of fewer than 100 patients into the experimental arm. Of course, these data are not without caveats; for example, the trial design does not allow for a comparison that could evaluate the relative contributions of the veliparib portion of the regimen and the carboplatin portion of the regimen to the efficacy of the combination. Theoretically, it could be possible that the enhanced efficacy of the regimen is due primarily to the effect of the added carboplatin, and results from an additional trial would be necessary to determine whether or not that is the case. However, for an agent like veliparib, which, if it does enter Phase III development, would be competing with other PARP inhibitors like BMN 673 (BioMarin) and olaparib (AstraZeneca) in a race to be the first agent in this drug class to enter the market for breast cancer, the benefits of this type of collaborative, relatively high throughput type of screening in early stage clinical development will likely provide an important benefit of identifying an appropriate patient population for late stage development, with potentially less time and cost. The results of additional I-SPY 2 experimental arms, as well as the results of late stage development of I-SPY 2 graduates (Puma Biotechnology recently announced the graduation of neratinib in HER2+/HR- patients; Puma press release, December 4, 2013), will be eagerly awaited to see if the potential benefits of this type of innovative trial will be borne out in practice.


  1. Rugo et al., SABCS 2013, Abstract S5-02

By: Stephanie Hawthorne, PhD, Director, Clinical and Scientific Assessment, Kantar Health and Cory Blaiss, PhD, Analyst, Clinical and Scientific Assessment, Kantar Health

article register

Recent Posts

Recent Comments



Post a Comment

You must be logged in to post a comment