Combined SHPments: an effective therapeutic strategy for MPNST

Abstract

Targeted therapies have provided the foundation for many advances in the treatment options for late-stage cancer patients; however, adaptive and compensatory responses frequently limit their efficacy. Rational combinations of targeted inhibitors are being actively tested in pre-clinical models to form the basis for more durable responses in patients. In this issue, Pratilas and colleagues provide evidence that phosphorylated SHP2 is adaptively upregulated in response to MEK inhibitors in malignant peripheral nerve sheath tumors (MPNSTs) that have lost NF1 expression (1). The authors provide evidence that the combination of SHP2 inhibitors with MEK inhibitors has strong efficacy in pre-clinical MPNST models and propose that this targeted therapy combination should be rapidly translated.

RAS family GTPases are mutated in >30% of human cancers and RAS-driven cancers are in need of more effective treatment options. In addition to RAS mutations, somatic and germline loss of function mutations in the NF1 gene, which encodes a RAS GTPase-activating protein (RAS-GAP), are found in a variety of cancers. Germline loss of NF1 gives rise to the autosomal dominant cancer pre-disposition syndrome neurofibromatosis type 1, associated with neurofibromas and melanocytic cafe-au-lait spots. Approximately ~8–13% of NF1 patients will develop MPNSTs, which most often arise from within plexiform neurofibromas. MPNSTs are soft tissue sarcomas with a proclivity to metastasize and are the major cause of mortality for NF1 patients. These tumors are resistant to standard chemotherapies and radiotherapy; thus, new treatment strategies are much needed (2).

Loss of NF1 expression leads to RAS activation, which activates multiple signaling cascades including the RAF-MEK-ERK1/2, SEK-JNK, PI3K-AKT and RalGDS pathways. Given the importance of ERK1/2 activation for growth and survival, MEK inhibitors have been tested for multiple cancer types. As noted by Wang et al., MPNST cells respond transiently to MEK inhibitors, an outcome that is similar to effects observed in colon cancer and melanoma (references cited in (1)). In these cancer types, adaptive signaling responses involving the up-regulation/activation of multiple RTKs limit the efficacy of BRAF inhibitors and MEK inhibitors. Rational combinatorial strategies have been proposed but a major concern is that targeting one adaptive response will lead to up-regulation of alternative by-pass mechanisms and subvert the effects of the combination approach.

The MEK inhibitor selumetinib is FDA-approved for non-operable plexiform neurofibromas and MEK inhibitor combinations are being tested in MPNST patients (NCT03433183). Based on this knowledge, the authors sought to optimize MEK inhibitor strategies for MPNST patients. They take a rigorous approach utilizing both previously available models as well as new patient-derived MPNST and neurofibroma cell line resources. Not unexpectedly, they detected increased phosphorylation of multiple receptor tyrosine kinases (RTKs) in response to treatment with the MEK inhibitor trametinib. In addition, compensatory phosphorylated/mobility shifted forms of SHP2, indicative of increased SHP2 activity, were detected in the majority of models. SHP2 is a protein tyrosine phosphatase encoded by the gene PTPN11. Binding of SHP2 Src homology 2 domains to phosphorylated tyrosine residues opens up SHP2 and uncovers its phosphatase activity. SHP2 acts on substrates including RasGAP and Sprouty, which are negative regulators of RAS signaling.

The detection of altered SHP2 in response to MEK inhibition led Wang et al. to hypothesize that SHP2 inhibitors would synergize with MEK inhibition. Allosteric SHP2 inhibitors act by retaining SHP2 in a closed inactive state, while SHP2 inhibitors targeting the active site may elicit off target effects (7). In the related article, the authors show that the allosteric SHP2 inhibitor SHP099 enhanced the effects of trametinib to durably inhibit phosphorylated ERK1/2, expression of cyclin D1 and proliferation in vitro. Also, they provide evidence that SHP2 inhibitor maybe useful as a salvage therapy in the MEK inhibitor acquired resistance setting. In vivo, the combination of MEK inhibitor plus SHP2 inhibitor caused tumor regressions and, importantly, no evidence of toxicities was noted. Based on these data, they propose the combination of MEK inhibitors with SHP2 inhibitors for treatment of NF1-associated MPNSTs.

After decades of hurdles and roadblocks, this is an exciting time for targeting RAS-driven cancers. KRAS G12C selective inhibitors such as Sotorasib/AMG510 and MRTX849 have taken center stage. These inhibitors are effective in pre-clinical models and produce objective responses in subsets of KRAS G12C-driven NSCLC and colorectal cancer patients (3,4). In an alternative strategy, Hofmann et al. developed BI-3406 to inhibit the RAS guanine nucleotide exchange factor (GEF), SOS1, and used BI3406 to enhance sensitivity of KRAS-dependent tumors to MEK inhibitors (5). The current study builds on these reports as an option to target a major RAS effector pathway and broadly inhibit adaptive feedback pathways. Wang et al. data are supported by recent independent studies of SHP2 inhibitors in other RAS-driven cancers. Fedele et al., Ahmed et al., and Lu et al. reported that SHP099 combined with MEK inhibitor suppressed the in vivo tumor growth of KRAS-mutant driven pancreatic ductal adenocarcinoma and non-small cell lung carcinoma (NSCLC) xenografts (references cited in (1)). Nichols et al. used another SHP2 allosteric inhibitor, RMC-4550, and showed monotherapy activity in NF1-mutant melanoma and non-small cell lung adenocarcinoma cell lines. Notably, in the Wang et al. study, MPNST cells were refractory to single agent SHP2 inhibitor in vitro, although delayed tumor growth was observed with SHP2 inhibitor monotherapy in vivo. Other SHP2 inhibitors including TNO155, RMC-4630 and JAB3068, have entered early phase clinical trials as monotherapy and combined with either targeted therapies such as the MEK inhibitor cobimetinib or immunotherapies such as anti-PD-1 pembrolizumab (Clinicaltrials.gov: NCT03634982; NCT03989115; NCT04418661; NCT03114319; NCT04000529; NCT03518554). An early meeting report indicates some evidence of clinical activity in patients (6). Given the lack of effective options for treatment of NF1-associated MPNSTs, the combination of MEK inhibitors with SHP2 inhibitors has high translational potential.

Inhibiting components of the RAS machinery/signaling is not the only approach to broadly target the adaptive RTK responses to targeted therapies. In breast and ovarian cancer models, multiple RTKs are up-regulated in response to targeted inhibitors such as lapatinib. These adaptive kinome responses are effectively targeted epigenetically with BET/BRD inhibitors that competitively displace BRD4 proteins from acetylated histones (8). BET inhibitors regulation of transcriptional programs may be another possible combination with MEK inhibitors to test in MPNST pre-clinical models. A concern with all of these approaches is toxicity. While tolerability was noted in many of the SHP2 inhibitors in pre-clinical in vivo studies, combinations of targeted therapies in patients are often associated with poor tolerability.

An avenue that is not explored by Wang et al., is to examine SHP2 inhibitor effects on the tumor immune microenvironment and possible combination with immune checkpoint therapy. SHP2 binds to phosphorylated tyrosine-based motifs on the cytoplasmic domains of checkpoint receptors such as PD-1 and BTLA and may dephosphorylate costimulatory molecules, CD28 and CD226 to inhibit T cell activation. Immune checkpoint agents (ICi) including anti-PD-1 (pembrolizumab and nivolumab) and anti-CTLA-4 (ipilimumab) now provide new standard of care treatments for NSCLC, cutaneous melanoma, and other cancer types. However, not all patients respond to ICi and the presence of tumor-infiltrating leukocytes (TIL) is a major determinant of the response to ICi. Studies in pancreatic cancer models have already shown effects of SHP2 inhibitors alone and in combination with KRAS G12C inhibitors on CD3+ T cells, CD8+ T cells and CD8+/T reg ratios (9,10). MPNSTs frequently express PDL-1 and display high numbers of infiltrating cytotoxic T cells associated with low abundance of immunosuppressive regulatory T cells. The potential roles of SHP2 inhibitors on immune regulatory cells and the combination of SHP2 inhibitor-based therapies with ICi in MPNSTs represent areas for further exploration.

Overall, Wang et al., bring new resources to the MPNST field and extend our understanding of adaptive feedback mechanisms to targeted therapies. Importantly, they provide a timely pre-clinical study that may direct new combination strategies to optimize MEK inhibitor treatments for NF1-deficient MPNST patients (1).