SABCS 2017: I-SPY 2 TRIAL predictive biomarker analysis shows promise to improve targeted therapy for high-risk breast cancers

December 7, 2017

New Data Analyses Presented at San Antonio Breast Cancer Symposium Spotlight Predictive Response to Pembrolizumab and MK-2206, and Demonstrate Improved Predictive Performance of Breast MRI.

SAN FRANCISCO, C.A., December 7, 2017– Investigators with the I-SPY TRIALs for Breast Cancer,sponsored and managed by Quantum Leap Healthcare Collaborative, presented analyses today at the 2017 San Antonio Breast Cancer Symposium (SABCS) detailing four promising new or improved pathways for predicting response to therapeutic agents in the neoadjuvant setting. Three of the analyses provide new insights into biomarkers associated with the triple negative breast cancer subset, and a fourth concludes that combining measurement of functional tumor volume (FTV) and apparent diffusion coefficient (ADC) improves the predictive performance of breast MRI.

The I-SPY TRIAL was designed to rapidly screen promising experimental treatments and identify those most effective in specific patient subgroups based on molecular characteristics (biomarker signatures). The multi-center, adaptively randomized I-SPY 2 platform trial uses pathological complete response (pCR) as the primary endpoint to identify investigational agents that will improve outcomes in women with stage II/III breast cancer at high risk for early recurrence. The trial investigates all signatures based on hormone receptor (HR), HER2, and 70-gene (MammaPrint) status.

“In-depth biological and imaging analyses comparing patients who respond to an investigational regimen versus those who do not respond, reveals the mechanisms of response and resistance,” said Dr. Esserman, I-SPY 2 Principal Investigator. “This new knowledge provides critical insights for adapting I-SPY 2 so that we can give the right therapy to the patient.”

In the first analysis, expression of PI3K/AKT pathway genes were evaluated as specific predictive markers of the AKT-inhibitor MK-2206 in the I-SPY 2 TRIAL.  MK-2206 graduated from the trial in 2015 in three biomarker signatures: all hormone receptor (HR)-negative, all HER2-positive, and HR-negative/HER2-positive. While none of the candidate markers tested succeed as specific predictors of response to MK-2206 in the population as a whole, PIK3CA levels within the triple-negative subset associate with pCR specifically in the MK2206 arm, and may merit further evaluation in future trials. Abstract: P2-09-08.

In a second analysis, diffusion-weighted MRI measurements were shown to improve predictive performance of breast MRI for predicting pCR. By combining apparent diffusion coefficient (ADC) with standard functional tumor volume (FTV) MRI measurement, investigators were able to demonstrate a statistically significant improvement over the model using FTV alone. Further improvement can be achieved by adjusting the combined model with breast cancer subtype defined by HR and HER2 status. Abstract: P2-09-23.

Data from a third analysis looked at immune infiltrates (assessed through multiplex fluorescence immunohistochemistry) and immune gene expression signatures as predictors of response to the anti-PD-1 therapy pembrolizumab. Fourteen expression signatures representing different immune cell types (TILs, T cells, CB8 T cells, exhausted T cells, Th1, Tregs, cytotoxic cells, NK, NK CD56dim, dendritic cells, mast cells, B cells, macrophages, and neutrophils) were evaluated. 

Ten out of the 14 were associated with a response in the pembrolizumab arm. Higher expression levels of nine of these cell-type signatures were associated with higher pCR rates (T cells, exhausted T cells, Th1, cytotoxic cells, NK, NK CD56dim, dendritic cells, B cells, and macrophages), whereas higher mast cell signature expression is associated with non-pCR. Of these, three signatures (Th1, B cells and dendritic cells) showed significant interaction with response to pembrolizumab. Of note, dendritic cells and Th1 cells may be specific predictors for Pembro in both the population as a whole, and the triple negative subtype. The presence of mast cells was shown to possibly impede response, especially in HR+HER2- patients. Abstract: PD6-08.

In the fourth analysis, I-SPY investigators hypothesized that a combination of two signatures [MammaPrint High2 (MP2)/PARPi7-high], reflecting DNA damage repair deficiency (deficient mismatch repair or dMMR) and predicting response to veliparib/carboplatin therapy, may also predict response to pembrolizumab. The scientists also tested nine gene expression signatures reflecting different aspects of DNA damage and repair, including homologous recombination, mismatch repair, non-homologous end joining, base excision repair and DNA damage sensing. Of these biomarkers, MP2 status and a DNA damage sensing pathway (which includes the genes ATM, ATR, CHEK1 and CHEK2) —but not the PARPi7 or other repair pathways—was associated with response to pembrolizumab and show promise as predictive biomarkers for immune checkpoint inhibition therapy in breast cancer. Abstract: PD6-14.

About the I-SPY TRIALs

The I-SPY TRIAL (Investigation of Serial studies to Predict Your Therapeutic Response with Imaging And moLecular analysis) is a unique collaborative effort by a consortium that includes the Food and Drug Administration (FDA), industry, patient advocates, philanthropic sponsors, and clinicians from 16 major U.S. cancer research centers. Under the terms of the collaboration agreement, Quantum Leap Healthcare Collaborative is the trial sponsor and manages all study operations.

The trial’s adaptive statistical design was developed by the pioneering principal investigators for the I-SPY TRIAL, Laura J. Esserman, M.D., M.B.A., professor of surgery and radiology and director of the Carol Franc Buck Breast Care Center at UCSF Helen Diller Family Comprehensive Cancer Center(Dr. Esserman is the overall principal investigator for the I-SPY TRIALS), and Donald A. Berry, Ph.D., professor of biostatistics at The University of Texas MD Anderson Cancer Centerand founder of Berry Consultants in collaboration with the FDA, industry, and many leading academic collaboratorsincluding the Biomarker Working Group chair, Laura van ‘t Veer, Ph.D., and Imaging Working Group chair, Nola Hylton, Ph.D, both from the University California San Francisco.The trial is a unique collaborative effort where over 50 clinicians are actively engaged in the conduct of the trial.

The I-SPY 2 TRIAL design is based on Bayesian predictive probability that a biological regimen will be shown to be statistically superior to standard therapy in an equally randomized 300-patient confirmatory trial. Regimens that have a high Bayesian predictive probability of showing superiority in at least one of 10 predefined signatures graduate from the trial. Regimens are dropped for futility if they show a low predictive probability of showing superiority over standard therapy in all 10 signatures. A maximum total of 120 patients can be assigned to each experimental regimen. A regimen can graduate early and at any time after having 60 patients assigned to it.

About Quantum Leap Healthcare Collaborative

Quantum Leap Healthcare Collaborative (Quantum Leap), a non-profit foundation, was established in 2005 as a partnership between medical researchers at University of California at San Francisco and Silicon Valley entrepreneurs. Quantum Leap’s mission is to accelerate transfer of high-impact research in clinical processes and systems technology into widespread adoption so that patients and physicians can benefit from the research as soon as practicable. Quantum Leap provides operational, financial and regulatory oversight to I-SPY. For more information, visit

Pembrolizumab is marketed by Merck & Co., Inc., Kenilworth, N.J., USA under the brand name KEYTRUDA®.  

# # #

For more information, email