The I-SPY 2 trial identified a neo-adjuvant regimen containing the investigational drug neratinib and standard chemotherapy to be beneficial for hormone receptor (HR)-negative, human epidermal growth factor 2 (HER2)-positive primary (nonmetastatic) breast cancer patients. This data was presented at the American Association for Cancer Research 2014 Annual Meeting, in San Diego, California.
“The I-SPY 2 trial is a randomized phase 2 clinical trial for women with newly diagnosed stage II breast cancer,” said John W. Park, MD, professor of medicine at the University of California San Francisco (UCSF) Helen Diller Family Comprehensive Cancer Center in San Francisco. “The trial tests whether adding investigational drugs to standard chemotherapy in the neoadjuvant setting is better than standard chemotherapy alone, and the goal is to match investigational regimens with patient subsets on the basis of molecular characteristics [referred to as biomarker signatures] that benefit from the regimen.
“We were pleased that the algorithm used for the predictions functioned so well and that it actually stopped assigning neratinib to patient subgroups which were not benefiting from the drug while at the same time increasing its assignment to patient subgroups which were benefiting from the drug,” added Park.
A Bayesian algorithm used in this trial identified that the neratinib-containing regimen had accrued sufficiently to graduate based on its prediction that this regimen would be highly likely to succeed in a phase 3 trial in this same subset of patients.
The trial is conducted using adaptive randomization: A patient with a particular subtype of breast cancer who enters the trial is preferentially assigned to treatment regimens that are performing better in patients with the same subtype of breast cancer.
This trial enrolls patients who have stage II breast cancer with a tumor size at least 2.5 cm and are considered to be at high risk for early breast cancer recurrence when evaluated by MammaPrint. The primary end point of this study is pathological complete response (pCR) in breast and lymph nodes at the time of surgery.
The algorithm randomly assigned 115 patients to the arm of the trial that contained neratinib, a pan-HER inhibitor. The rates of pCR on the neratinib arm were compared with those of 78 patients who were concurrently randomized to the control arm containing standard chemotherapy. These comparisons were made for 10 biomarker signatures prospectively defined by categories of HR, HER2, and MammaPrint.
The investigators concluded that the probability that the neratinib-containing regimen has a higher rate of pCR than control therapy in HR-negative, HER2-positive breast cancer (one of the 10 biomarker signatures) is 95% and its predictive probability of success in a future, randomized, 300-patient phase 3 trial is 79%. The algorithm also predicted this drug combination is likely to be beneficial for all HER2-positive breast cancer patients (a second of the 10 biomarker signatures), with the probability of superiority over standard therapy and the probability of success in a phase 3 trial being 95% and 73%, respectively.