UCSF

HER2-targeted therapies have been the mainstay of treatment of HER2-positive breast cancer. To date, the selection of patients most likely to respond to HER2-targeted agents is based primarily on HER2 amplification and/or overexpression. However, the correlations among current clinical methods of detecting HER2 amplification and/or overexpression are imperfect with regards to both prognostication and the prediction of drug response to many of the HER2-targeted therapies, and therefore, there is a critical need for the discovery and translation of additional biomarkers that predict patient response to a specific HER2-targeted therapy. Here, we evaluated BluePrint molecular subtypes – a gene expression-based molecular subtype classification – as a predictor of response to HER2-targeted therapies using patient data from the I-SPY 2 TRIAL. We demonstrated the potential clinical utility of BluePrint molecular subtyping in identifying a subset of HER2-positive, estrogen receptor-positive (HER2+/HR+) patients who are less likely to benefit from HER2-targeted therapies. In addition, gene expression analysis of this subset of patients reveal lower immune signaling and higher estrogen receptor expression, and thus may potentially benefit from alternative strategies, such as endocrine therapy or immunotherapy.

In a second study, we evaluated the baseline activation state of 104 key signaling phosphoproteins/ proteins from prosurvival, mitogenic, apoptotic, and growth regulatory pathways as predictors of response to neratinib – an irreversible pan-HER tyrosine kinase inhibitor of EGFR/HER2 – in HER2-positive breast cancer cell line models with differential neratinib sensitivity. We identified 13 phosphoproteins/ proteins, representing a multitude of pathways, in particular the HER family signaling pathway, that are associated with neratinib sensitivity. We also demonstrated in HER2-positive breast cancer cell line models that acquired resistance to neratinib could potentially be mediated through adaptive kinome reprogramming, and that the combination of neratinib and BET bromodomain inhibitor appears to be a promising therapeutic strategy to overcome such resistance.

In conclusion, the work presented here provide insight into mechanisms underlying differential drug responses and resistance to HER2-targeted therapies, and highlight novel genomic and proteomic biomarker candidates that could potentially complement HER2 overexpression and/or amplification in predicting patient response to HER2-targeted therapies.