Predicting MEK Inhibitor Response in Lung Cancer: A Proper Signature Is Required

In this issue of Clinical Cancer Research, Brant and colleagues (1) report a novel assay to measure ERK signaling output in tumor specimens and suggest a potential utility in predicting response to MEK inhibitor treatment.

The ERK signaling pathway plays a crucial role in the regulation of growth and survival. Under physiologic conditions, this pathway is activated when extracellular growth factors bind to and activate receptor tyrosine kinases (RTKs), which leads to GTP loading of the small GTPase RAS. In turn, GTP-bound RAS recruits and activates RAF kinases in a complicated process involving dimerization and phosphorylation. RAF subsequently phosphorylates MEK, a dual-specificity kinase, which then phosphorylates and activates its substrate ERK. Once activated, ERK phosphorylates a number of cytosolic and nuclear substrates. The latter include members of the ETS family of transcription factors, which control the expression of genes (e.g., CCND1 and MYC) responsible for cell-cycle progression or other cellular functions (Fig. 1). In addition, ERK exerts feedback regulation, both by directly phosphorylating upstream signaling intermediates and by indirectly inducing the expression of feedback proteins, including Sprouty, which block the activation of RAS by RTKs, and dual-specificity phosphatases (DUSPs), which dephosphorylate ERK.

A number of commonly occurring genomic alterations deregulate ERK signaling in cancer. In lung adenocarcinoma, the most common subtype of lung cancer, alterations that activate ERK signaling are found in approximately 75% of patients (2). These include KRAS and BRAF mutations, as well as NF1 loss and alterations that activate RTKs. KRAS mutations are the most frequent activating alterations found in lung cancer. Despite this, there are currently no therapies that directly target this oncoprotein available for use in the clinic. Because these mutations result in activated ERK signaling, treatment with a MEK inhibitor ought to inhibit the growth of tumors harboring them. Although MEK inhibitor monotherapy has clinical activity against BRAF V600E–mutant melanomas (3), this approach is less effective in patients with KRAS-mutant lung cancers (4). This is thought to occur because tumors harboring KRAS mutations have a more pronounced adaptive response to MEK inhibitor treatment (5–7) and because MEK inhibitors have a narrow therapeutic index, which limits the total amount of drug that can be safely administered to patients. A recent phase II trial of previously treated patients with advanced KRAS-mutant non–small cell lung cancer (NSCLC) showed statistically significant improvement in progression-free survival in the trametinib and docetaxel arm compared with the docetaxel-only arm, though the effect was small (8). These studies have led to the hypothesis that the effect of MEK inhibitor treatment may be more pronounced if it were feasible to select those patients most likely to benefit from the MEK inhibitor.

The NSCLC-specific ERK signaling output signature (or MEK signature) that is described in the article by Brant and colleagues (1) was derived from a previously reported 18-gene signature (9) and was further refined to six genes that are downstream of ERK: PHLDA1, SPRY2, DUSP6, DUSP4, ETV4, and ETV5. The expression level of these genes was captured using the NanoString nCounter system. This approach should allow improved sensitivity compared with microarrays and compatibility with the widely available formalin-fixed paraffin-embedded tissue samples. Indeed, when tested in patient-derived biopsy specimens, the assay was able to detect ERK signaling output in all samples, even with suboptimal amounts of sample RNA. Robust reference genes were selected to ensure consistent normalization of sample data for scoring. The assay was validated by measuring the effect of MEK inhibitor treatment and siRNA-mediated knockdown of KRAS in lung cancer cell lines. All tested cell lines showed a decrease in the MEK signature score after MEK inhibitor treatment, but only those harboring a KRAS mutation had a response to the KRAS siRNA. As expected, the authors noted a significant overlap between MEK signature scores and the presence of a KRAS mutation, yet they also identified a distinct population where a high MEK signature score was observed in the absence of a detectable KRAS mutation. This is not surprising given that lung adenocarcinomas often harbor other mutations that activate ERK. Finally, it was shown that the MEK signature score was mostly conserved between primary and metastatic lesions, and this variation was 3-fold smaller than the variability between patients.

The most intuitive approach to determine the activation of ERK signaling in cancer is to measure ERK phosphorylation. The level of pERK, however, is the product of the kinase activity of MEK and the phosphatase activity of DUSPs. As described above, DUSPs are transcriptionally induced and stabilized by active ERK (10). When ERK is inhibited, expression of DUSPs decreases, leading to restored levels of pERK (11, 12). Thus, in some cellular contexts, a relatively small change in pERK may be coupled with a significant suppression of ERK signaling output (Fig. 1, inset). In addition, the level of pERK in patient specimens is predominantly detected by IHC, which lacks robust quantification and is subject to interpreter bias. With these in mind, assays measuring the expression of ERK-dependent genes, such as the approach described in the study by Brant and colleagues (1), are expected to provide a quantitative, and perhaps more accurate, measurement of pathway activity. In this regard, such assays hold the promise of being important pharmacodynamic markers of ERK signaling inhibitor treatment in the clinic.

A more comprehensive assessment of pathway activation by the signatures described in the study can also help the identification of patients most likely to benefit from MEK inhibitor treatment. Identifying which patients are most likely to respond to such treatment may be an important step in overcoming the narrow therapeutic index of these drugs. Nevertheless, further testing is needed to determine whether the MEK signature score is associated with a response to MEK inhibitor treatment in patients with KRAS-mutant cancer and to determine its applicability in the context of different driver mutations and/or other tumor lineages.

In summary, the work by Brant and colleagues (1) is an important step in developing a point-of-care assay to optimally quantify ERK signaling in patients. It has the potential to aid the selection of patients for targeted therapy, which in turn may lead to improved outcomes, particularly in patients whose tumors harbor KRAS mutations. Additional clinical testing is needed, however, before it can become a standard of care in the clinic.

No potential conflicts of interest were disclosed.

Conception and design: Y. Xue, P. Lito

Writing, review, and/or revision of the manuscript: Y. Xue, P. Lito

This work was supported by the NIHK08CA191082-01A1 and the Druckenmiller Center for Lung Cancer Research at Memorial Sloan Kettering Cancer Center (to P. Lito). Y. Xue was supported by a Medical Scientist Training Program grant from the National Institute of General Medical Sciences of the NIH under award number T32GM007739 (to the Weill Cornell/Rockefeller/Sloan Kettering Tri-Institutional MD-PhD Program).