Background: Hormone receptor positive (HR+), HER2- breast cancer (BC) is a heterogenous disease. We hypothesized that molecular subtypes capturing luminal, basal, and immune biology could predict response for patients (pts) with HR+/HER2- disease in the I-SPY2 trial.

‍Methods: I-SPY2 trial is a phase II, randomized, adaptive study evaluating multiple investigational agents as neoadjuvant BC therapy; the primary endpoint is estimated pCR rate. Investigational agents are given with control weekly paclitaxel x 12, followed by AC x 4. Regimens graduate when the predicted pCR rate in any signature meets the pre-specified threshold of 85% probability of success in a hypothetical 300 pt randomized, phase 3 trial. We analyzed estimated pCR rates for the 1st 7 investigational agents in the HR+/HER- subset, analyzed by clinical/molecular features: BluePrint (BP) Luminal vs. Basal, Mammaprint High1 [MP1] vs. Mammaprint High2 [MP2], MP2 is < -0.57, Responsive Predictive Subtype-5 (RPS-5) (classification based on HR, HER2, immune, DNA-repair, and basal/luminal markers), histology, and stage/nodal status.

‍Results: 38% (379/987) of pts had HR+/HER2- disease. Only pembrolizumab met the pre-specified graduation criteria for HR+/HER2- BC. pCR rates by treatment arm and molecular subtype are described in the Table. 28% were MP2; 72% were MP1. Overall, pCR rates were higher in pts with MP2 vs MP1 disease (30% vs 11%) including with pembrolizumab (55% vs. 21%). 29% were BP Basal, 71% were BP Luminal; BP Basal was more likely to be MP2 than BP Luminal (77% vs 8%). In all arms except MK2206, HR+/HER2- BP Basal pts were more likely to achieve pCR than BP Luminal pts. For MK2206, BP Luminal pts were more likely to achieve pCR. Immune+ by RPS-5 (39% of HR+/HER2-) predicted pCR to pembrolizumab irrespective of BP Basal or Luminal status (11 pCR/16 pts). Results by histology and stage/nodal status will also be reported.

‍Conclusions: Our data suggest that MP2 and BP Basal signatures identify a subset of HR+/HER2- BC more likely to respond to neoadjuvant therapy; and that an immune signature can identify pts more likely to respond to pembrolizumab. These findings will aid in guiding prioritization of targeted agents with the goal to optimize pCR for all pts.

Clinical trial information: NCT01042379