Tumor and Systemic Immunomodulatory Effects of MEK Inhibition

Abstract

Purpose of Review

Mitogen-activated protein kinase (MAPK) kinase (MEK) is an integral component of the RAS signaling pathway, one of the most frequently mutated pathways in cancer biology. MEK inhibitors were initially developed to directly target oncogenic signaling, but are recognized to have pleiotropic effects on both tumor cells and lymphocytes. Here we review the preclinical and clinical evidence that MEK inhibition is immunomodulatory and discuss the potential rationale for combining MEK inhibitors with systemic immunotherapies.

Recent Findings

MEK inhibition may modulate the tumor microenvironment (TME) through direct effects on both tumor cells and immune cells. Despite encouraging evidence that MEK inhibition can reprogram the tumor microenvironment (TME) and augment anti-tumor immunity regardless of KRAS/BRAF status, recent clinical outcome studies combining MEK inhibition with systemic immunotherapy have yielded mixed results. The combination of MEK inhibitors plus systemic immunotherapies has been tolerable, but has thus far failed to demonstrate clear evidence of synergistic clinical activity. These results underscore the need to understand the appropriate therapeutic context for this combination.

Summary

MEK inhibitors have the potential to inhibit oncogenic signaling and reprogram the tumor immune microenvironment, representing an attractive therapy to combine with systemic immunotherapies. Ongoing preclinical and clinical studies will further clarify the immunomodulatory effects of MEK inhibitors to inform the design of rational therapeutic combinations.

Introduction

The MAPK pathway is a highly conserved signal transduction pathway in all eukaryotic cells that regulates a variety of normal cellular functions, including cell proliferation, differentiation, survival, and apoptosis (1). Upon activation of upstream growth factor receptors, RAS, the principal upstream activator of the MAPK pathway, undergoes a conformation change in which guanosine diphosphate (GDP) is exchanged for guanosine triphosphate (GTP)(2)’(3). Genes encoding RAS (KRAS, NRAS, HRAS) are among the most frequently mutated genes in a variety of cancer, including pancreas (90%), colon (50%), thyroid (50%), lung (30%) and melanoma (25%)(4). The conformational change of RAS results in the recruitment and activation of the RAF family of serine-threonine protein kinases (RAF-1, B-RAF, and A-RAF), which activate MEK1/2 (MKK or MAPK kinase) by phosphorylation. MEK1/2 then activates ERK1/2, the canonical mitogen-activated protein kinase (MAPK), via phosphorylation(3). Activated MAPKs can go on to phosphorylate a range of downstream substrates, including protein kinases and transcription factors, which, in turn, lead to the transcription of MAPK-regulated genes. In addition to transcriptional regulation, MAPKs can also regulate gene expression of targets by altering mRNA stability, transport, and translation (Figure 1)(2)’ (5).

Figure 1.

Summary of the immunomodulatory effects of MEK inhibition on tumor cells and immune cell populations.

MEK is considered a central component in the MAPK signaling pathway, and MEK1/2 can be activated by several kinases including Mos, A-Raf, B-Raf, Raf-1, and MEKK (6). However, MEK1 and MEK2 are the only known regulators of ERK1/2, and thus the signals of these upstream activators converge at the level of MEK. Since ERK1/2 have been shown to regulate hundreds of downstream targets, MEK1/2 are seen as “gatekeepers” in the MAPK pathway. Therefore, MEK inhibition (MEKi) represents an attractive drug candidate for tumors that are addicted to the RAS-signaling pathway. Since the first synthetic small-molecule inhibitor of MEKI and MEK2 kinase activity was described in 1995, four small-molecule MEK inhibitors have been approved by the FDA for clinical use in various settings—trametinib, cobimetinib, binimetinib, and selumetinib — and multiple other MEK inhibitors are in clinical or preclinical stages of development (7),(8).

Although MEK inhibitors were developed to target oncogenic signaling, there is an increasing interest in understanding the immune consequences of systemic MEK inhibition. In the current era in which immunotherapy has become firmly established as a pillar of cancer therapy, some patients may receive both targeted inhibition of MEK and immune checkpoint immunotherapy throughout their cancer treatment. In certain tumor types for which MEK inhibition and immune checkpoint inhibitors have substantial efficacy (for example, BRAF mutated melanoma or BRAF mutated colorectal cancer with mismatch repair deficiency), understanding the immune consequences of MEK inhibition may have significant implications for the sequencing of cancer therapies, and also for determining whether these therapies can rationally be combined (9-12). Although targeted inhibition of RAS signaling in such tumors can have rapid and substantial anti-tumor activity, these effects are rarely durable due to the emergence of drug-resistant clones. By contrast, novel immune checkpoint inhibitors have the potential to transform short-lived responses into durable responses, providing a foundation for studying these therapies in combination.

The effects of MAPK signaling and modulation on cancer immunotherapy are complex and context-dependent. In addition to serving as a critical regulator of cancer growth and metastasis, the mitogen-activated protein kinases (MAPK) pathway is also involved in nearly all aspects of immune responses: from the initiation phase of innate immunity, to activation of adaptive immunity, to cell death to prevent uncontrolled immune activation (5). Multiple preclinical studies have found that MEK inhibitors have immunomodulatory effects, including increased tumor immunogenicity and the potential for synergy with systemic immune checkpoint inhibition. However, direct clinical evidence that MEK inhibitors can synergize with systemic immunotherapies is lacking, and the combination of a MEK inhibitor and a PD-L1 inhibitor recently failed to show compelling efficacy in a pivotal clinical trial of colorectal cancer (IMblaze370), a tumor type for which MEK inhibition and immune checkpoint therapy have no meaningful activity in unselected patients (13). Herein we summarize the complex and at times, contradictory effects of MEK inhibitors on the relationship between the tumor and immune environment to better elucidate rational drug combinations.

MAPK Pathway Effects Tumor Cell Immunogenicity

MEK signaling is a key pathway involved in both tumor cell survival and lymphocyte response to antigen stimulation and may modulate the TME through direct effects on both tumor cells and immune cells. Here we summarize the known effects of MEKi on tumor cells that may impact tumor immunogenicity, including alterations in Major histocompatibility complex class I (MHC- I) expression and antigen presentation on target cells, programmed death ligand 1 (PDL-1) expression, and changes in cytokine release.

MHC-I Expression

MHC-I molecules present short peptides from either foreign or native intracellular proteins on the cell surface. The recognition of cancer neoantigens (abnormal proteins not found on normal host cells) in the context of MHC by cytotoxic T cells is a critical means of cancer elimination by the host immune system (14,15). Downregulation of MHC class I molecules by cancer cells results in reduced sensitivity to lysis by cytotoxic T cells and NK cells, providing a mechanism for immune evasion. Consequently, tumors with low-level expression of MHC-I are associated with a poor prognosis (16,17). In order to identify negative regulators of MHC-I, Brea et al. used a pooled human kinome shRNA interference-based approach and identified MAP2K1 (MEK) as a negative regulator of MHC-I expression (18). This observed negative regulation of the MAPK pathway has been confirmed across many studies and cancer cell lines, in which treatment with different MEK inhibitors consistently led to increased MHC-I expression (7),(19),(20), (21),(22). This increase in MHC-I expression is shown to be associated with the upregulation of components of the antigen presentation pathway, such as B₂M, TAP1, and TAP2 at both the mRNA and protein level (18),(21). Importantly, this increase in MHC-I expression on the cell surface enhanced the presentation of antigens, resulting in an increase in specific T-cell activity, confirming the important functional significance (22), (18), (21), (23).

While many have shown that MEK inhibition leads to an increase in MHC-I, it is of interest that the use of an EGFR inhibitor also resulted in the upregulation of surface MHC-I in an NSCLC cell line with activating mutations in EGFR and NRAS, despite no change in pERK levels. This suggests potential activation of a parallel pathway that regulates MHC-I expression in EGFR mutant cancers and implies that the MAPK pathway is not solely responsible for the regulation of surface MHC-I (18).

PD-L1 Expression

PD-L1 (B7-H1, CD274) is a transmembrane protein that has received recent attention for its role in suppressing the immune system. Binding of PD-L1 to its receptor, programmed death receptor (PD-1), found on activated T cells, B cells, and myeloid cells, leads to suppression of T-cell activation and proliferative cytokine release and induction of T-cell apoptosis (24,25). PD-L1 expression is induced on antigen presenting cells by interferons, released by T cells upon recognition of their cognate antigens, and serves as a mechanism to control the level of immune response. However, it is now recognized that tumor cells can co-opt this checkpoint pathway by expressing PD-L1, facilitating immune evasion (26). Unsurprisingly, PD-L1 overexpression is correlated with poor overall survival and progression-free survival in several solid tumor types, including gastric, urothelial, hepatocellular carcinoma, breast, and melanoma. Consequently, there is increased potential for benefit from α-PD1 or PD-L1 immunotherapy (27).

While is seems undisputed that the MAPK pathway is a negative regulator of MHC-I, the effects of pathway activation and inhibition on PD-L1 expression are less clear and may vary by cancer type. Mechanistically, it was shown that RAS can drive cell-intrinsic PD-L1 expression via a downstream signaling cascade (28). However, the idea that activated RAS leads directly to increased PD-L1 is likely oversimplified as many other mutations and microenvironments can influence MAPK cascade cross-talk (29). In support of this notion, several studies have seen no impact on PD-L1 expression in various cancer cells lines or transient increases that didn’t track with pathway activation upon treatment with MEKis(19), (30). This suggests an uncoupling of MAPK pathway activation levels and PD-L1 expression in cancer cells continually exposed to MEK inhibitors. Other studies have reported variable responses across cell lines, with both increases and decreases seen depending on the cell line and experimental context, though the effect was independent of alterations in the MAPK pathway (30)’(31). These conflictions may explain the lack of consistent significant clinical synergy seen between MEK inhibitors and α-PD1 or PD-L1 immune-checkpoint inhibitors and suggests the need to better understand the impact of the genetic background on tumor PD-L1 regulation (29).

Cytokine release

Tumors secrete various factors such as VEGF, IL-6, and IL-10 that are critical cytokines in the tumor microenvironment, playing a role in immune suppression (32). Collectively, these cytokines have been shown to promote the accumulation of a heterogeneous population of immunosuppressive cells including tumor-associated macrophages (TAMs), myeloid derived suppressor cells (MDSCs) or immature dendritic cells. These cell populations have been shown to inhibit anti-tumor immunity through a variety of mechanisms, such as depletion of arginine, recruitment of T-regulatory cells, and increasing reactive oxygen species (33).

Melanoma cell lines with constitutive MAPK pathway activation showed high levels of VEGF, IL-10, and IL-6 expression (34). Use of the MEKi U0126 resulted in decreased mRNA and protein levels of these soluble immunosuppressive factors. Importantly, this decrease was associated with an increase in the production of the inflammatory cytokines IL-12 and TNFα by dendritic cells. Additionally, KRAS mutated cells were shown to produce increased osteopontin, a cytokine that has been shown to attract TAMs and drive the expansion of the suppressive MDSC population in the spleen (35,36). Treatment of tumor cells with trametinib reduced the production of osteopontin, abrogating the cytokine driven expansion of mMDSCs from myeloid progenitors (37).

MEK Inhibition Effects Immune Cell Populations

T cells

The MAPK axis plays a significant role in T-cell function. Upon engagement of the T-cell receptor, the RAS-MAPK signaling pathway is activated, leading to proliferation and cytotoxic effector functions against tumor cells (38). Consequently, there are concerns that MEK inhibition may dampen T-cell anti-tumor immunity by inhibiting T-cell receptor signaling. In support of this notion, the MEK inhibitor UO126 was shown to have deleterious effects on T-cell proliferation in vitro and IFN- γ production by T cells when antigen-specific T cells were cultured with melanoma cells (39)(40). This negative effect on T-cell proliferation in vitro was seen in multiple T-cell subsets including naïve, effector, and memory (41)(42). The decrease in T- cell proliferation in vitro has been partially attributed to the suppression of the production of growth-promoting cytokines IL-2, TNFα, and IL-3 by stimulated T cells, and exogenous addition of IL-2 reversed the phenotype (43). Importantly, the suppressive effects of MEK inhibitors seen on T-cells were not limited to in vitro models. In an ovarian cancer mouse model, antigen-specific CD8 T cell were shown to have decreased proliferative capacity after treatment with trametinib (41). Additionally, IFN-γ production by CD8⁺ and CD4⁺ T cells isolated from mouse mammary carcinomas treated daily with trametinib was measured and a significant reduction was observed (20).

Other studies found that in vitro results that show decreased functional T-cell activity with MEK inhibitors did not fully translate in vivo. Treatment of established CT26 tumors with trametinib increased tumor CD4⁺ T cell proliferation and infiltration in terms of absolute quantification, although it should be noted that the Treg population was not distinguished from the helper T-cells. Tregs have been shown to have a decreased functional dependence on the MAPK pathway allowing suppressive capabilities to be spared with MEK inhibition (30), (42). Hu-Lieskovan et al. found that while MEKi impaired production of IFN-γ by splenocytes treated in culture, in vivo cytokine production was not effected following isolation from trametinib treated mice and ex vivo re-stimulation (44). In this study, however, trametinib was co-administered with high doses of IL-2. It is possible that this is a confounding variable, as one of the reasons cited for decreased T cell proliferation and IFN-γ production with MEK inhibition was decreased IL-2 production that was reversible upon exogenous IL-2 addition (43). Finally, in a CT26 tumor model, naïve CD8⁺ T-cell priming was impaired in the lymph nodes after treatment, but interestingly, an increase of antigen-specific CD8⁺ T cells was seen in the tumor after initiation of treatment relative to control mice (45). These tumor-infiltrating CD8⁺ T cells were shown to be protected from death driven by chronic TCR stimulation and maintained cytotoxic activity, although once again exogenous IL-2 was added when cytotoxic activity was measured. This apparent increase and improved effector phenotype of CD8⁺ T cells did not have a functional impact however, as depleting these CD8⁺ T cells in MEKi treated tumors did not affect tumor growth (45).

In summary, studies tend to agree that MEK inhibition significantly hinders T-cell function and proliferation in vitro. While this finding was translated in vivo in a number of studies, the inconsistencies seen in other studies may be further explained by the time points at which tumor and TILs were analyzed and the treatment schedule, in addition to the differences in tumor models which further influence immune cell populations. This suggests that future studies should focus on finding a treatment schedule that maximizes the positive effects seen.

Immunosuppressive populations

Neoplastic growth can result in a chronic inflammatory response that promotes the expansion and recruitment of myeloid-derived suppressor cells (MDSCs) and tumor associated macrophages (TAM). These heterogeneous and plastic populations hinder anti-tumor immunity by blocking T-cell functions and proliferation through the production of iNOS and arginase and directly drive tumor growth by promoting cancer sternness, angiogenesis, and metastasis formation (46,47). Several studies had shown a reduction in monocytic-MDSCs (mMDSCs) upon treatment with MEK inhibitors (37)(48). This was accompanied by a decrease in arginase production, an indirect explanation for the increase of tumor CD8⁺ T cells seen in some studies, suggesting that the final effect of MEKi on T-cell numbers is dependent on many different pieces (37). In addition to impaired MDSC recruitment, MEK inhibition has been shown to inhibits monocyte to TAM differentiation, thereby preventing the accumulation of TAMs (49).

B cells

While new evidence has pointed to the importance of B cells in modulating the immune response to tumors, very few studies have looked at how targeted therapy, and in particular MEK inhibition, impacts B-cell populations (50). B cells are typically characterized by their positive modulation of immune responses and inflammation achieved through antibody production as well as their ability to promote T-cell activation via antigen presentation (51). Increasingly, more studies are reporting on the existence of discrete subsets of B-regulator cells (52,53). These immunosuppressive cells are able to maintain immune tolerance, suppressing responses to neoplastic growth through the release of anti-inflammatory mediators such as IL-10 and the expression of PD-L1, as well as facilitating the conversion of T cells to T regulatory cells (50). A recent study from our group using a colorectal cancer mouse model found that MEK inhibitors can enhance anti-tumor immunity through the reprogramming of B cells, resulting in a reduced number of Bregs in the tumor and non-tumor draining lymph nodes (54).

The effects of MEK inhibition on immune cell populations are summarized in Figure 1.

MEK Inhibition and Immunotherapy

Blocking antibodies

Immune checkpoint inhibitors represent one of the most promising approaches in the treatment of cancer (55,56). As mentioned earlier, the ability of cancer cells to overcome immunosurveillance is a key factor leading to cancer progression. Drugs interrupting immune checkpoints, such as anti-CTLA-4, anti-PD-1, anti-PD-L1, and others that are in early development, have the potential to unleash anti-tumor immunity, but the percentage of patients that have shown durable responses to immunotherapy is small (26), (57). As a result, there is a growing interest in combining immunotherapy agents with a second therapeutic agent to maximize anti-tumor outcome (58). While immunotherapies and targeted therapies have distinctly different mechanisms of action, it is thought that combinations of both modalities may create synergies with increased benefit for cancer patients.

Consistently across tumor models, the combination of MEK inhibition with immune checkpoint inhibitors resulted in reduced tumor volume and increased overall survival (19),(30),(49),(45). This was frequently accompanied by an increase in tumor CD8⁺ and CD4⁺ T-cell number and proliferative capacity as well as increased expression of IFNγ and Granzyme B (GZMB). MEK inhibition also reversed the increase in Cox2 and Arg1 expression in tumors seen with anti-CTLA4 treatment (49). However, of note the sequence of dosing has been shown to impact in vivo efficacy, with the most potent results observed when the MEK inhibitor was given first (30), (49). This suggests that blocking MAPK signaling is important to prime and synergize tumors in response to immunotherapy through mechanisms explained above such as upregulation of MHC molecules and reduction of immunosuppression factors.

Agonist antibodies

Costimulatory receptor signaling plays an important role in regulating antigen-specific T-cell proliferation, differentiation, and apoptosis (59,60). OX40 (CD134) and 4-1BB (CD137) belong to the tumor necrosis factor (TNF) receptor (TNFR) family and represent T-cell costimulatory receptors expressed following primary activation via antigen-specific receptors and CD28 costimulatory receptors. The ligation of OX40 with OX40 ligand (OX40L), expressed on antigen presenting cells, leads to improved T-cell mediated anti-tumor immunity, with augmented CD4⁺ and CD8⁺ T-cell clonal expansion, effector differentiation, and survival, and in some cases abrogated suppressive activity of T-regs (61). Activation of the 4-1BB/4-1BBL–mediated signaling cascade has been shown to play a role in T-cell proliferation, activation-induced cell death prevention, eradication of established tumors, enhancement of integrin-mediated cell adherence, and amplification of cytotoxic t-cell responses (62,63). In order to overcome the potentially detrimental effects of MEK inhibition on T-cell function there is a growing interest in combination therapy with agonist antibodies. Importantly, these antibodies have been shown to activate T-cells independently of MEK1/2 signaling, theoretically allowing the restoration of effector function even in the presence of MEK inhibitors (64).

To test this notion, Dushyanthen et al. investigated the benefit of combining agonist therapies with the MEK inhibitor trametinib (20). They found that while MEK inhibition increased tumor immunogenicity, it significantly inhibited early T-cell signaling, reducing both the proliferation and IFN-γ production of CD8⁺ and CD4⁺ T-cells in MAPK/Ras driven TNBC tumors. Both α-OX-40 Ab and α-4-1BB Ab restored the long-term proliferation and cytokine production of T-cell populations both in vitro and in vivo, confirming the ability of these agonists to prevent the impairment of T-cell function. Additionally, triple combination treatment with Trametinib, α-PD-1, and α- OX-40 Ab/α-4-1BB Ab resulted in significantly enhanced inhibition of tumor growth and prolonged survival. Of interest, anti-4-1BB/anti-OX-40 monotherapy enhanced the frequency of MDSCs, which was reduced following the addition of Trametinib (20). While these results are exciting, more studies need to be done in additional models to confirm how broadly these findings can be applied.

Clinical Trials

Over the past decade, immune checkpoint inhibitors have revolutionized the treatment of many cancers and have become firmly established as a pillar of cancer therapy. However, it is currently estimated that only approximately 40% of cancers are candidates for treatment with immune checkpoint inhibitors, and of these tumors only approximately 15% will respond (65). Novel immunotherapy combinations that can extend the benefit of immune checkpoint inhibitors to immune-resistant tumor types remains a critically important area of discovery. In principle, beyond the potential for immunomodulation, the combination of targeted therapies with systemic immunotherapies may be rational. Targeted therapies such as MEK inhibitors can induce rapid and significant tumor regressions, but these responses are often transient because of acquired resistance mutations. The addition of immune checkpoint inhibitors may in theory facilitate tumor elimination, converting a transient clinical response into a more durable responses. The success of combination therapy with a MEK inhibitor and immune checkpoint inhibitors in preclinical models led to the clinical investigation of such combinations in a number of different clinical contexts (Table 1). Despite initial excitement, the results of these trials combining MEK inhibitors with systemic immunotherapies thus far generally failed to show convincing evidence of synergistic antitumor activity.

Table 1.

Clinical trials investigating the combination of MEK inhibitors with Immunomodulatory Agents.

MEK Inhibitor	Indication	Design	Phase	# of patients	Results	NCT	Reference
Cobimetinib	Advanced BRAFV600 Wild-type Melanoma	Cobimetinib and Atezolizumab vs Pembrolizumab; Randomized	III	450	No improvement in Progression Free Survival	NCT03273153	Miller WH, Kim TM, Lee CB, et al. J Clin Oncol. 2017;35(suppl; abstr 3057). doi: 10.1200/JCO.2017.35.15_suppl.3057
	Gallbladder Carcinoma, Stage IV intrahepatic Cholangiocarcinom a, unresectable Cholangiocarcinom a	Atezolizumab vs Atezolizumab and Cobimetinib; Randomized	II	82	Improvement in Progression Free Survival and Disease Control Rate.	NCT03201458	N/A
	Breast Cancer	Cobimetinib+Paclitaxel vs Cobimetinib+Atezolizumab+Paclitaxel vs Cobimetinib+Atezolizumab+Nab-Paclitaxel; Randomized	II	169	ORRs were similar between the A + C + P arm and A + C + nP arm. Numerically higher ORR and PFS were observed in pts with PD-L1+ disease	NCT02322814	Brufesky A, Kim SB, Zvirbule Z, et al. Journal of Clinical Oncology 37, no 15_suppl (May 20, 2019)1013-1013. doi:10.1200/JCO.2019.37.15_suppl.1013
	Colorectal Cancer	·Atezolizumab vs Cobimetinib+Atezolizumab vs Regorafenib; Randomized	II	363	Atezolizumab plus cobimetinib did not prolong overall survival (OS) vs. regorafenib	NCT02788279	Bendell J, Ciardiello F, Tabernero J, et al. Annals of Oncology 2018; Volume 29, Issue suppl_5. mdy208.003, https://doi.org/10.1093/annonc/mdy208.003
	Melanoma	Atezolizumab+Cobimetinib+Vemur afenib vs Cobimetinib+Vemurafenib; Randomized	III	513	Increase in Progression Free Survival and Duration of Response	NCT02908672	N/A
	Malignant Melanoma	Sequencing schedule of Cobimetinib+Vemurafenib+Atezolizumab	II	176	N/A	NCT02902029	N/A
	Multiple Myeloma	Cobimetinib vs Cobimetinib+Venetoclax vs Cobimetinib+Venetoclax+Atezoliuzmab: Randomized	I/II	72	N/A	NCT03312530	N/A
	Breast Cancer	Atezolizumab+Cobimetinib vs Atezoluzimab+Idasanutlin; Randomized	I/II	80	N/A	NCT03566485	N/A
	Advanced or Metastatic Solid Tumors	Nivolumab vs Nivolumab+Ipilimumab vs Nivolumab+Ipilimumab+Cobimetinib (Pancreatic adenocarcinoma)	I/II	162	N/A	NCT01928394	N/A
	Breast Cancer, Estrogen Receptor positive Breast Cancer	Pre-operative window of opportunity study (Atezolizumab vs Atezolizumab+Cobimetinib vs Atezolizumab+Ipatasertib vs Atezolizumab+Ipatasertib+Bevacizumab)	II	160	N/A	NCT03395899	N/A
Selumetinib	Advanced/Metastatic Solid tumors	Selumetinib+Pembrolizumab dose escalation; non randomized	I	50	N/A	NCT03833427	N/A
	Lung Cancer	Intermittent selumetinib+durvalumab+Tremelimumab vs continuous selumetinib+Durvalumab+Tramelimumab; Randomized	I/II	40	N/A	NCT03581487	N/A
	Advanced Solid Tumors	Ascending doses Selumetinib+MEDI4736 vs Selumetinib+MEDI4736+Tremelimumab; Non-randomized	I	58	N/A	NCT02586987	N/A
	Locally Advanced or Metastatic Non-Small-Cell Lung Cancer	Gefitinib, AZD9291, or Selumetinib+Docetaxel + MEDI4736 vs Tremelimumab+MEDI4736; Randomized	II	32	N/A	NCT02179671	N/A
Binimetinib	Colorectal Cancer	Encorafenib, Binimetinib, Nivolumab; Single Group	I/II	38	N/A	NCT04044430	N/A
	MSS, RAS-mutant Colorectal Cancer	Binimetinib+Nivolumab vs Binimetinib+Nivolumab+Ipilimumab; Non-randomized	Ib/II	90	N/A	NCT03271047	N/A
	Pancreatic Cancer, Non-Small Cell Lung Cancer	Avelumab+Binimetinib vs Avelumab+Binimetinib+Talazoparib; Randomized	II	127	N/A	NCT03637491	N/A
	Metastatic Colorectal Cancer	Pembrolizumab+Binimetinib or Pembrolizumab+chemo vs Pembrolizumab+chemo+Binimetinib; Non-Randomized	I	220	N/A	NCT03374254	N/A
	Breast Cancer	Binimetinib+Avelumab vs anti-Ox40+Avelumab vs Utomilumab+Avelumab; Randomized	II	150	N/A	NCT03971409	N/A
Trametinib	Colorectal Cancer, Malignant Neoplams of Digestive Organs	Durvalumab+Trametinib; single group	II	56	N/A	NCT03428126	N/A
	Non-Small Cell Lung Cancer, KRAS Gene Mutation	Trametinib+Pembrolizumab vs Pembrolizumab+Trametinib; Non-Randomized	I	42	N/A	NCT03299088	N/A
	Metastatic Melanoma	Dabrafenib+Ipilimumab vs Dabrafenib+Trametinib+Ipilimumab; Randomized	I	38	N/A	NCT01767454	N/A
	Melanoma	·Spartalizumab+Dabrafneib+Trametinib vs Dabrafenib+Trametinib; Randomized	III	538	N/A	NCT02967692	N/A
	Melanoma	Dose escalation; Durvalumab+Dabrafenib+Trametinib vs Durvalumab+Trametinib; Non-Randomized	I	68	N/A	NCT02027961	N/A
	Melanoma	BGB324+Pembrolizumab vs BGB324+Dabrafenib+Trametinib vs Pembrolizumab vs Dabrafenib+Trametinib; Randomized	I/II	92	N/A	NCT02872259	N/A
	Melanoma	Sequential Dabrafenib+Trametinib then Pembrolizumab vs Concurrent Dabrafenib+Trametnib and Pembrolizumab vs Pembrolizumab; Randomized	II	60	N/A	NCT02858921	N/A
	Colorectal Cancer	Nivolumab+Trametinib vs Nivolumab+Trametinib+Ipilimumab; Non-Randomized	I/II	345	N/A	NCT03377361	N/A
	Melanoma	Safety of Pembrolizumab+Trametinib+Dabrafenib; Randomized	I/II	190	N/A	NCT02130466	N/A
	Metastatic Melanoma	BRAF600 mutant Penbrolizumab+Dabrafenib+Trametinib; BRAFV600 WT Pembrolizumab+Trametinib; Non-Randomized	II	50	N/A	NCT03149029	N/A

In a phase 1 clinical trial across multiple tumor types, the combination of cobimetinib plus the PD-L1 inhibitor atezolizumab had manageable safety and demonstrated multiple objective responses in mismatch-repair proficient colorectal cancer (66). These results were particularly striking because MEK inhibitors and systemic immunotherapies have historically shown little or no activity in mismatch-repair proficient colorectal cancer as single agents. In this phase 1 study, there was no evident enrichment in of response in patients with KRAS/BRAF-mutant tumors versus wild-type tumors. These clinical findings provided support for the idea that MEK inhibitors are immunomodulatory and can sensitize immune excluded-tumors or immune-desert tumor types to respond to inhibitors of the PD-L1 pathway. Consistent with preclinical observations, paired pretreatment and cobimetinib on-treatment biopsies from this phase 1 study revealed an increase in CD8⁺ T-cell infiltration, suggesting successful translation of a preclinical finding into clinical development.

Based on these encouraging results, the combination of cobimetinib and atezolizumab were subsequently tested in the phase 3 colon cancer IMblaze370 study. In this study there was a trend towards longer overall survival for patients treated with cobimetinib plus atezolizumab over atezolizumab monotherapy, but the combination failed to meet its primary endpoint of superiority over a standard of care in this setting, regorafenib (13). Additionally, a phase III randomized study (IMspire170) investigating combination treatment with cobimetinib and atezolizumab (PD-L1 inhibitor) versus pembrolizumab (PD-1 inhibitor) in patients with BRAFV600 wild-type advanced melanoma failed to meet its primary endpoint of improved progression free survival with the combination (67). These two negative trials, one in a classically immune excluded-tumor tumor type (IMblaze370), and the other in a classically immune-inflamed tumor type (IMspire170), diminished initial hopes that this combination would be paradigm changing despite favorable effects of MEKi on the tumor immune microenvironment.

Other subsequent clinical trials of MEKi in combination with systemic immunotherapies did meet their primary outcomes, although none of these studies conclusively showed evidence of synergistic clinical activity. A smaller signal finding phase 2 trial in cholangiocarcinoma showed that the combination of cobimetinib plus atezolizumab significantly prolonged survival over atezolizumab monotherapy, but response rates in the combination treatment arm were low (68). In a separate phase 3 trial in BRAFV600E mutated melanoma (IMspire150) the combination of vemurafenib (B-Raf inhibitor), cobimetinib, and atezolizumab increased the median progression free survival from 10.6 months to 15.1 months when compared to treatment with vemurafenib, cobimetinib, and placebo (69). Intriguingly, overall response rates in the atezolizumab and control groups were very similar, and thus the progression free survival benefit of atezolizumab versus control was largely driven by the prolonged duration of response among the responders (21·0 vs 12·6 months). These results were similar to the phase 2 KEYNOTE-022 study, in which the combination of pembrolizumab plus dabrafenib and trametinib had longer median progression-free survival than placebo plus dabrafenib and trametinib. Together, these two studies indicate that checkpoint inhibition may deepen responses with targeted therapy in BRAFV600E melanoma. Although a late separation of overall survival curves favoring combination therapy was evident in both studies, the appropriate sequencing of immunotherapy and targeted therapy remains undetermined. Ongoing melanoma studies such as ImmunoCobiVem (NCT02902029), SECOMBIT (NCT02631447), DREAMseq (NCT02224781), COMBI-i (NCT02967692) may help to address clinical questions regarding the optimal sequence or combination of BRAF/MEK inhibition and immune checkpoint inhibitor therapy in BRAFV600E mutated melanoma. However, none of these studies will fully answer whether MEK inhibition and systemic immunotherapies have additive or synergistic activity.

Conclusions and Future Directions

Multiple preclinical studies have investigated the immunomodulatory effects of MEK inhibitors with, at times, contradictory conclusions (See Table 2). While there is agreement that MEKi can enhance the expression of MHC-I and antigen presentation, the mechanism(s) through which this occurs remains unclear. The effects of MEK inhibition on T-cell function is controversial and may be context-dependent, and the effects of MEKi on other immune populations remains largely unknown. The combination of MEKi with systemic immunotherapies shows potentially synergistic activity in preclinical models, but clinical studies have not demonstrated clear evidence that these agents are synergistic and MEKi has thus far failed in the clinic as a therapeutic strategy for transforming immune-resistant tumors into immune responsive tumors.

Table 2.

Summary of immunomodulatory effects, highlighting degree of confidence in the finding.

Category of Change	Increased			Decreased			No Change
	Finding	Reference	Type Of Study	Finding	Reference	Type Of Study	Finding	Reference	Type Of Study
MHCI Expression	Increased IFNg lead to increased MHC I expression	Loi et al, 2015(19)	In Vivo and In Vitro
		Brea et al, 2016(18)	In Vitro
		Angell et al, 2014(22)	In Vitro
		Dushyanthen et al, 2017(20)	In Vivo and In Vitro
	MEK inhibition leads to intrinsic upregulation of MHC I	Brea et al, 2016(18)	In Vitro
		Angell et al, 2014(22)	In Vitro
	MEK inhibition leads to pAKT mediated MHC I upregulation	Mimura et al, 2013(21)	In Vitro
PDL1 Expression	Increased PD-L1 Expression	Loi et al, 2015(19)	In Vivo and In Vitro	Decreased PD-L1 Expression	Jiang et al, 2013(70)	In Vitro	No Persistent Change in PD-L1 Expression	Angell et al, 2014(22)	In vitro
	Cell Line Dependent Increased PD-L1 Expression	Atefi et al, 2014(31)	In Vitro		Sumimoto et al, 2016(71)	In Vitro		Liu et al, 2015(30)	In vitro
				Cell Line Dependent Decreased PD-L1 Expression	Atefi et al, 2014(31)	In Vitro
T Cell Infiltration	Increased T Cell Homing	Liu et at, 2015(30)	In Vivo
		Ebert et al, 2016(45)	In Vivo
T Cell Activation	Increased CD4 T Cell Proliferation	Liu et al, 2015(30)	In Vivo	Decreased T Cell Proliferation	DeSilva et al, 1998(72)	In Vitro	In Vivo Cytokine Production Is Not Affected	Hu-Lieskovan et al, 2015(44)	In Vivo
	Decreased exhaustion and apoptosis	Ebert et al, 2016(45)	In vitro and in vivo		Dushyanthen et al, 2017(20)	In Vivo and In Vitro	No Affect On CD8+ T cells	Ebert et al, 2016(45)	In Vivo
					Boni et al, 2010(39)	In Vitro
					Vella et al, 2013(40)	In Vitro
					Allegrezza et al, 2016(41)	In Vitro and In Vivo
					Zwang et al. 2016(42)	In Vitro
				Decreased T Cell IFNg Production	Boni et al, 2010(39)	In Vitro
					Dushyanthen et al, 2017(20)	In Vivo and In Vitro
					Hu-Lieskovan et al, 2015(44)	In Vitro
				Decreased Activation of Antigen-specific T cells	Vella et al, 2013(40)	In Vitro
				Decreased T Cell IL2, TNFa, IL3 Production	Dumont et al, 1998(43)	In Vitro
				Impaired tumor specific CD8+ T cell expansion and priming	Ebert et al, 2016(45)	In vivo
Tumor Cytokine Production				Decreased Tumor IL10, IL6, VEGF Expression	Sumimoto et al, 2006(34)	In Vitro
				Decreased Tumor Osteopotin Production	Kim et al, 2014(35)	In Vivo
					Lin et al, 2015(36)	In Vivo

More research is needed to understand why MEKi has broadly failed as an immunomodulatory strategy in the clinic despite success in multiple preclinical models. One possibility is that systemic MEKi has some potentially detrimental effects on anti-tumor immunity (eg a decrease PD-L1 expression thereby negating the benefit of αPD1/PD-L1 therapy, impaired T cell activation and priming), that can outweigh the positive effects of MEKi on the TME in some contexts. This highlights the need to further understand the therapeutic context(s) in which synergy is possible. Alternatively, MEKi plus PD-L1 may be insufficient to overcome the immunosuppressive phenotype of the tumors in which this strategy has been tested. More research is needed to determine mechanisms of resistance to MEKi plus αPD1/PD-L1 therapy, and to explore additional combinatorial strategies. Ongoing studies may further clarify the full effects of MEK inhibitors on immune cells, aiding in the design of more effective drug combinations that better harness the potential of MEK inhibitors.