A rapid, non-invasive test to determine HER2-low status in triple negative breast cancer (TNBC) patients

Background: TNBC is defined as a HER2-negative disease, as assessed via immunohistochemistry (IHC). An IHC 3+ score is considered positive and 2+ (with a negative ISH), 1+, and zero scores are considered negative. Recently, HER2-low (1+, 2+) breast cancer has been proposed as a targetable disease distinct from TNBC. Indeed, use of an anti-HER2 antibody drug conjugate (ADC) in HER2-low metastatic patients resulted in improved progression free and overall survival (DESTINY-Breast04), and remarkably generated a response in HER2-zero cases. The potential to expand the current treatment for HER2-low and -zero patients, highlights the limitations of traditional HER2 assessment, including HER2 intra-tumoral heterogeneity, differential IHC pathology staining, and inter-reader variability. We therefore sought to address how to optimally identify HER2-low TNBC patients that may benefit from HER2 ADCs.  


Methods: We developed a rapid, non-invasive test to determine HER2-low status in TNBC tumors. Our test uses radiological imaging (DCE-MRIs) coupled with the biophysical simulation platform TumorScope® and is informed by single cell RNA-sequencing data to connect molecular regulation to the tumor microenvironment in individual patients. We leveraged these relationships to design a rapid, non-invasive test, HER2Scope, that captures the fraction of each TNBC tumor likely to express significant levels of HER2 (a “HER2 score”). We then characterized the HER2 score in a cohort of TNBC patients from our Virtual Tumor Bank, leading us to identify 3 distinct cohorts: HER2-negative, -low, and -moderate tumors.  


Results: Using our platform, we identified that HER2 expression is associated with less proliferation and lower nutrient use. Interestingly, we found that there was significant overlap in HER2 expression between HER2+ and TNBC tumors.  HER2Scope-identified HER2-low and- moderate tumors were associated with a lower likelihood of pCR in response to chemotherapy compared to HER2-zero tumors (25.3% in HER2-moderate vs. 40% in HER2-zero), suggesting a more aggressive and less drug-responsive phenotype in HER2-low and -moderate tumors. We then tested the expected rate of HER2-low and -moderate tumors within TNBC tumors using our Virtual Tumor Bank (n=264) and found a HER2-low and -moderate response rate of 56% (95% CI = 49-62%), which is strikingly similar to the TNBC response rate to trastuzumab deruxtecan (TDXd) reported in the DAISY trial (42%). HER2Scope may therefore be an effective test to identify TNBC patients for HER2 ADC therapies.  


Conclusion: We present a novel test to identify HER2-nonzero TNBC patients rapidly and non-invasively. HER2Scope relies on radiological imaging, and therefore does not depend on an inherently variable score derived from biopsy-driven pathologic staining. Additionally, by evaluating the entire tumor, it addresses the challenge of HER2 intra-tumoral heterogeneity.  

View publication here.