An heregulin-EGFR-HER3 autocrine signaling axis can mediate acquired lapatinib resistance in HER2+ breast cancer models

INTRODUCTION: The human epidermal growth factor receptor 2 (HER2) receptor tyrosine kinase (RTK) oncogene is an attractive therapeutic target for the treatment of HER2-addicted tumors. Although lapatinib, an FDA-approved small-molecule HER2 and epidermal growth factor receptor (EGFR) tyrosine kinase inhibitor (TKI), represents a significant therapeutic advancement in the treatment of HER2(+) breast cancers, responses to lapatinib have not been durable. Consequently, elucidation of mechanisms of acquired therapeutic resistance to HER-directed therapies is of critical importance. METHODS: Using a functional protein-pathway activation mapping strategy, along with targeted genomic knockdowns applied to a series of isogenic-matched pairs of lapatinib-sensitive and resistant cell lines, we now report an unexpected mechanism of acquired resistance to lapatinib and similar TKIs. RESULTS: The signaling analysis revealed that whereas HER2 was appropriately inhibited in lapatinib-resistant cells, EGFR tyrosine phosphorylation was incompletely inhibited. Using a targeted molecular knockdown approach to interrogate the causal molecular underpinnings of EGFR-persistent activation, we found that lapatinib-resistant cells were no longer oncogene addicted to HER2-HER3-PI3K signaling, as seen in the parental lapatinib-sensitive cell lines, but instead were dependent on a heregulin (HRG)-driven HER3-EGFR-PI3K-PDK1 signaling axis. Two FDA-approved EGFR TKIs could not overcome HRG-HER3-mediated activation of EGFR, or reverse lapatinib resistance. The ability to overcome EGFR-mediated acquired therapeutic resistance to lapatinib was demonstrated through molecular knockdown of EGFR and treatment with the irreversible pan-HER TKI neratinib, which blocked HRG-dependent phosphorylation of HER3 and EGFR, resulting in apoptosis of resistant cells. In addition, whereas HRG reversed lapatinib-mediated antitumor effects in parental HER2(+) breast cancer cells, neratinib was comparatively resistant to the effects of HRG in parental cells. Finally, we showed that HRG expression is an independent negative predictor of clinical outcome in HER2(+) breast cancers, providing potential clinical relevance to our findings. CONCLUSIONS: Molecular analysis of acquired therapeutic resistance to lapatinib identified a new resistance mechanism based on incomplete and "leaky" inhibition of EGFR by lapatinib. The selective pressure applied by incomplete inhibition of the EGFR drug target resulted in selection of ligand-driven feedback that sustained EGFR activation in the face of constant exposure to the drug. Inadequate target inhibition driven by a ligand-mediated autocrine feedback loop may represent a broader mechanism of therapeutic resistance to HER TKIs and suggests adopting a different strategy for selecting more effective TKIs to advance into the clinic.