Although mutation of NF1 has been described in non–small cell lung cancer (NSCLC), co-mutation with RASA1, another Ras-GTPase activating protein (RasGAP), defines a novel genetically defined subclass of NSCLC. RASA1/NF1-mutant cell lines are highly sensitive to MEK inhibitors, warranting clinical evaluation of MAPK inhibition in this subclass of patients. Clin Cancer Res; 24(6); 1243–5. ©2018 AACR.
See related article by Hayashi et al., p. 1436
In this issue of Clinical Cancer Research, Hayashi and colleagues report that loss-of-function mutation of RASA1, a Ras-GTPase activating protein (RasGAP), is significantly enriched in patients with NF1-mutated non–small cell lung cancer (NSCLC; ref. 1). Co-occurring mutation of RASA1 and NF1 shows mutual exclusivity with other oncogenic drivers, such as KRAS and EGFR, unveiling RASA1/NF1 co-mutation as a new, potentially targetable subclass of NSCLC.
Approximately two thirds of patients with NSCLC harbor genetic alterations that activate receptor tyrosine kinase/RAS pathway–mediated downstream mitogenic signaling, including the MAPK and PI3K pathways. The clinical importance of genomically subclassifying lung adenocarcinoma based upon these drivers has been proven by the development of effective targeted therapies directed against EGFR, ALK, ROS1, BRAF, MET, RET, and NTRK, although limited by acquired resistance. Notably, KRAS, the most frequently mutated oncogene in patients with NSCLC, has remained a refractory target despite extensive efforts. Although targeting downstream mitogenic signaling effectors, such as MAPK and PI3K, has been one approach to subvert oncogenic KRAS, clinical trials combining inhibitors of these pathways have been ineffective. Because KRAS activates multiple other effector pathways, including those that co-opt innate immune signaling, it is possible that alternate approaches to combination therapy will yield better results. Direct targeting of specific KRAS mutations (e.g., the common smoking-associated G12C mutation) and anti–PD-1–directed or adoptive cell immunotherapy strategies have also shown recent promise, although clearly more work is needed to benefit the majority of KRAS-mutant patients in the clinic (2).
With broad application of next-generation sequencing data analysis, and increased power as sample sizes grow, the spectrum of clinically relevant, novel lung adenocarcinoma alterations continues to expand, shrinking the group previously classified as lacking major oncogenic drivers. Loss-of-function mutation of the NF1 gene, a canonical member of the RasGAP family, has been well described in lung cancer. Inactivation of NF1 function results in hyperactivation of wild-type RAS proteins by rendering them stuck in the GTP-bound form. Although the detailed mechanistic consequences following NF1 depletion are not fully understood—for example, which wild-type RAS isoforms, including KRAS, NRAS, and HRAS, mainly contribute to oncogenic activity—NF1 inactivating mutation by itself shows only partial mutual exclusivity with KRAS and EGFR mutations, suggesting additional tumor-suppressive functions of NF1. In addition, how mutation in other members of the RasGAP family impacts RAS signaling, alone or in concert with NF1 mutation, has been incompletely characterized.
In their article, Hayashi and colleagues demonstrate that a RASA1 inactivating mutation is observed in 1% to 3% of patients with NSCLC through analysis of genomic data from the MSK-IMPACT clinical cohort and The Cancer Genome Atlas (TCGA; ref. 1). Interestingly, the RASA1 mutation was highly enriched in patients with NF1-mutant NSCLC, and co-mutation of RASA1/NF1 exhibited complete mutual exclusivity with KRAS and EGFR mutations compared with single mutation of either the RASA1 or NF1 gene. Moreover, although NF1-only mutation was found in lung adenocarcinoma (81%) with higher frequency, RASA1/NF1 co-mutation was distributed evenly in adenocarcinoma and squamous cell carcinoma. Similar to KRAS mutation itself, and in contrast to alterations in other oncogenic drivers, such as EGFR and ALK, RASA1/NF1 co-mutation was enriched in patients who were smokers. Importantly, the authors further demonstrated that targeting downstream MAPK signaling with MEK inhibition in vitro was significantly more potent in NSCLC cells with RASA1/NF1 co-mutation compared with single mutation of either the RASA1 or NF1 gene, or with cells that harbor oncogenic KRAS or EGFR mutations. Together, these findings provide a novel druggable subset of NSCLC, including patients with both lung adenocarcinoma and squamous cell carcinoma.
Several questions remain regarding the translation of these findings to the clinic. First, although targeting the MAPK pathway directly downstream of oncogenic BRAF mutation in lung adenocarcinoma has been effective clinically (3), MEK inhibition in KRAS-mutant NSCLC has negligible activity (Fig. 1; ref. 4). Most likely, this is due to the fact that, as mentioned, KRAS activates multiple parallel, downstream signaling pathways. Although RASA1/NF1 mutation should theoretically activate the RAS pathway upstream and behave similarly to oncogenic KRAS, might co-mutation of these GAP proteins indeed create a unique dependency on MAPK pathway signaling, as the authors demonstrate in vitro? Furthermore, if this is the case, would combined RAF/MEK inhibition with dabrafenib and trametinib, which shows greater efficacy in BRAF-mutant lung adenocarcinoma and amelioration of some toxicities (3), also be more effective in RASA1/NF1-mutated NSCLC? Another possibility is that the behavior of RASA1/NF1-mutated NSCLC may differ from KRAS-mutant NSCLC due to different co-mutated tumor suppressors. For example, the authors show that the RASA1/NF1 alterations are significantly associated with TP53 mutation, but not STK11, which is commonly inactivated together with KRAS and portends poor outcome. Regardless, the discovery of this subclass and the profound sensitivity in vitro to MEK inhibition warrants clinical evaluation of this hypothesis in treatment-refractory patients.
Another interesting issue raised by this study involves the potential to repurpose FDA-approved drugs for off-label usage in scenarios with strong preclinical data. Because dabrafenib and trametinib are FDA approved for BRAF-mutant lung adenocarcinoma, in the absence of an available clinical trial, is it reasonable for a clinician to seek off-label approval and treat a patient with RASA1/NF1-mutated lung cancer who has no other viable therapeutic options with this combination therapy? There is a precedent for doing this with crizotinib, originally designed as a potent MET inhibitor. Although at the time only FDA approved for ALK-rearranged lung adenocarcinoma, based upon strong preclinical data supporting sensitivity of the MET exon skipping mutation to MET inhibition in vitro, we treated such a patient off label with crizotinib and observed a clinical response (5), which was later validated by larger case series and clinical trials. Although treating patients in the context of a clinical trial is clearly the preferred route, documenting results of off-label therapy and publishing case reports or case series of results are likely to catalyze industry efforts to invest in larger scale clinical trials. Furthermore, access to clinical trials geographically and based upon slot availability remains a major barrier for the majority patients. In the era of expanded genomic testing, as these low-frequency subclasses of NSCLC and other cancers continue to be identified, what is the best route not only to ensure early access to potentially effective targeted therapies but also to collect the data?
Regardless, this initial description of RASA1/NF1 co-mutated NSCLC and the strong preclinical data warrant clinical evaluation of strategies that incorporate MAPK pathway inhibition. Hopefully, the results observed will side with the efficacy observed in BRAF-mutated lung adenocarcinoma and not with those seen with oncogenic KRAS, which remains a major hurdle.
Disclosure of Potential Conflicts of Interest
D.A. Barbie is a consultant/advisory board member for N-of-One. No potential conflicts of interest were disclosed by the other author.
Writing, review, and/or revision of the manuscript: S. Kitajima, D.A. Barbie
This work was supported by NIHR01CA190394 (D.A. Barbie).