Epidermal growth factor receptor (EGFR) inhibitors are highly effective in treating non-small cell lung cancers (NSCLC) expressing activated EGFR, particularly those harboring EGFR mutations. However, most patients who benefit from EGFR inhibitors achieve only partial responses or stable disease, facilitating the emergence of resistance. Thus, progression-free survival advantages in responding patients are modest. Combination therapy, preferably using agents with synergistic activity, could both improve responses and reduce acquired resistance rates. We hypothesized that combining MEK inhibitors with EGFR inhibitors could result in such a benefit. The MAPK pathway lies downstream of EGFR and transduces both proliferative and survival signals in a variety of cancer types. Inhibitors of this pathway are currently in clinical trials, but little evidence exists supporting the use of these agents as monotherapy in EGFR-dependent NSCLC. In this study, we find EGFR-dependent NSCLC cell lines are moderately sensitive to loss of ERK1/2 activity, either by small molecule inhibition or by siRNA knockdown. The consequence of inhibition is dependent upon the trophic content of the culture media, primarily anti-proliferative in serum-rich conditions and pro-apoptotic in serum-poor conditions. However, when MEK inhibition was combined with EGFR inhibitors, cytotoxic synergy was observed for all EGFR-dependent cell lines tested in serum-containing media. Enhanced cytotoxicity is demonstrated in cell lines with and without EGFR mutations, including those harboring the T790M escape mutation. These findings support future clinical studies that combine EGFR-and MEK1/2-targeted agents to investigate whether improved outcomes can be achieved in clinically screened EGFR-dependent NSCLC.

Purpose: The intent of the study described herein is to develop and characterize a microRNA gene expression signature of response to EGFR inhibition portable to solid tumors for which treatment with an EGFR inhibitor is indicated. Experimental Design: Treatment with inhibitors of the epidermal growth factor receptor (EGFR) can result in clinical response in both non-small cell lung cancer (NSCLC) and pancreatic ductal adenocarcinoma (PDAC), but only in a minority of unselected patients. Identification of biomarkers that can direct treatment to the responsive population is essential and here we describe generation of a microRNA gene signature of response to EGFR inhibition. Using a Taqman MicroRNA Array of 381 probes, we demonstrate that a 20-gene microRNA signature derived from NSCLC cell lines can predict sensitivity of both lung and pancreatic cancer cell lines to erlotinib, a small molecule inhibitor of EGFR. Overlap of independently generated microRNA expression signatures of NSCLC and pancreatic cell lines produces a fourgene signature. We characterize a target of the reduced signature, ZEB1, in NSCLC and pancreatic cancer cell lines by evaluation of mRNA and protein expression. We evaluated ZEB1 protein and the ZEB1 target, E-cadherin, after transfection of mir-200c into inhibitor-resistant NSCLC cell line, A549. Results: We have identified a 20-gene microRNA signature of response to EGFR inhibition that predicts both inhibitor-sensitive NSCLC and pancreatic cell lines. By overlapping independent signatures of response in NSCLC and pancreatic cancer cell lines, a four gene microRNA signature is identified that targets proteins responsible for epithelial to mesenchymal transition (EMT). Of those proteins, ZEB1 is a transcription factor responsible for repression of E-cadherin expression. Both ZEB1 mRNA and protein expression are down-regulated in erlotinib-sensitive NSCLC in which the microRNA genes 200b, 200c, and 141 are up-regulated. Ectopic expression of mir-200c in inhibitor-resistant lines induces cell death, suggesting that down-regulation of this microRNA is essential for viability in these NSCLC lines. Conclusions: The 20-gene miRNA signature of sensitivity provides a means for defining response that can be generated from FFPE samples of tumor rather than fresh tumor samples. Whereas mRNA is difficult to isolate from FFPE samples due to degradation, microRNA may survive the fixation process. Thus, response to EGFR inhibition in the second- or third-line can be predicted from fixed tumor samples collected early in the treatment of an individual patient, eliminating the need for acquisition of fresh tumor. Further, the microRNA signature identifies the well-studied EMT pathway indicative of EGFR-inhibitor sensitivity and provides biological significance for the members of the signature. Citation Information: Cancer Res 2009;69(23 Suppl):B74.