Current Perspectives and Novel Strategies of NRAS-Mutant Melanoma

Introduction

With 324,635 new diagnoses and 57,043 deaths worldwide during 2020, cutaneous melanoma constitutes the 17th most incident and the 22nd most deadly malignancy. These data vary around the continents due to heterogenous frequency of skin phototype, with higher rates of both incidence and mortality in Oceania, Western Europe and Northern America.¹ Although mortality tends to stabilize, melanoma incidence has been rising over the last 40 years.²

An integrative analysis of cutaneous melanomas performed by the TCGA (The Cancer Genome Atlas) network validated four genomic subtypes characterized by BRAF, NRAS, NF1 mutations and a “triple wild-type” subgroup (which includes KIT mutated melanoma).³ Improvement in the understanding of cutaneous melanoma’s biology together with clinical benefit of BRAF-targeted therapy^4–6 and immunotherapy^7–10 over the last decade have changed the therapeutic management of advanced melanoma. However, not many progresses have been made in developing novel therapeutic options for BRAF wild-type melanomas.

NRAS mutant melanoma comprises 20% of all melanoma,³ which appears to confer an aggressive course compared with BRAF-mutant or RAF and NRAS wild-type melanomas.¹¹ In this setting, effective treatment options for this population are required, especially after disease progression on immunotherapy with anti-CTLA4 and/or anti-PD-1 antibodies. MEK inhibitors have been the most exploited treatment option in clinical trials with limited efficacy compared to chemotherapy.¹²

A greater comprehension of the biology and intracellular interactions of NRAS-mutant melanoma will outline novel impactful strategies which could improve prognosis of these subgroup of patients.

Ras as an Oncogene in Melanoma

Activating mutations in the Ras GTPase proteins have been found in one-third of human cancers.¹³ Ras is a superfamily of proteins implicated in cell growth, survival and differentiation, sending intracellular signals from receptor tyrosine kinases (RTK) to the MAPK (mitogen-activated protein kinase) and PI3K (phosphoinositide 3-kinase)-AKT pathways mainly.¹⁴

Three RAS isoforms (NRAS, neuroblastoma ras viral oncogene homolog; KRAS, Kirsten rat sarcoma viral oncogene homolog; and HRAS, Harvey rat sarcoma viral oncogene homolog) have been described and frequently harbor oncogenic mutations in cancer.¹⁵ Regarding cutaneous melanoma, aside from NRAS, we find mutations in KRAS and HRAS accounting for 2% and 1% of the cases, respectively.³

NRAS Melanoma

NRAS was the first oncogene recognized in melanoma¹⁶ and mutations in NRAS account for 20% of all melanomas.³ NRAS and BRAF mutations are usually mutually exclusive.³

NRAS mutations primarily occur at position 61 and involve amino acid change from glutamine (Q) to arginine (R), lysine (K) or leucine (L).¹⁷ These mutations block NRAS into a GTP-bound state impairing its GTPase activity.¹⁸

Mutations at codons 12 or 13 mutations comprise 20% of all NRAS and result in amino acid change from glycine (G) to aspartic acid (D).¹⁷ These mutations prevent the association of GAPase activating proteins (GAP) to NRAS and are more frequent in mucosal melanoma than cutaneous melanoma.¹⁸

NRAS mutations results in an endless activation of the MAPK signaling and hardly ever concur with mutations in the PI3K–AKT pathways, indicating that NRAS could modulate this pathway too.¹⁹ These pathways cause dysregulation of cell-cycle and cellular proliferation signals.¹⁴

NRAS-mutant melanoma differs from BRAF-mutant melanoma from a clinical point of view. Patients are usually older (>55 years) with previous UV exposure.²⁰ Lesions have predisposition to the upper extremities and are thicker with higher Breslow depth.^11,20

Regarding prognosis, similar to BRAF-mutant melanoma, NRAS mutation has been related to aggressive disease traits and with an increased risk of visceral and brain involvement.^21,22

Immunotherapy for NRAS Melanoma

Despite all the translational and clinical data gathered since discovery of NRAS mutation as a driver in melanoma, ESMO guidelines recommend treatment with immune checkpoint inhibitors (anti-CTLA4 and/or anti-PD1 inhibitors) as standard of care for the locally advanced unresectable and metastatic staging.²³

However, data from pivotal trials which led to approval of immune checkpoint inhibitors in melanoma did not report outcomes for NRAS subgroup. Data for efficacy of immunotherapy for NRAS melanoma mainly comes from retrospective cohorts. Tables 1 and 2 summarize all reported results.

Table 1 Summary of Efficacy Outcomes of Immunotherapy (Monotherapy and Combinations) in Patients with NRAS-Mutant Melanoma from Retrospectives Cohorts

Table 2 Summary of Efficacy Outcomes of Immunotherapy (Monotherapy and Combination) in Patients with NRAS-Mutant Melanoma Compared to Other Non-NRAS Mutant Melanoma Patients

In a first cohort of 236 NRAS-mutant patients from five skin cancer centers in Germany and Switzerland have been treated with Ipilimumab, anti-PD1 or anti-PD1+Ipilimumab combination. Efficacy outcomes for each treatment group are depicted in Table 1. Comparison to NRAS wild-type subgroup (n=128) evidenced comparable overall response rate (ORR) and disease control rate (DCR) and significantly lower overall survival (OS) (Table 2).²⁴

An independent cohort of 162 patients with NRAS-mutant melanoma from 11 referral centers in Italy treated in first line with immunotherapy reported different results. Cumulative results irrespective of treatment option showed a PFS (progression free survival) of 12 months, OS of 32 months and ORR of 42%. No differences in all efficacy outcomes were seen compared with 169 patients with BRAF/NRAS wild type melanoma (Table 2).²⁵

Finally, a cohort of 60 patients treated with anti-PD1/PD-L1 or Ipilimumab in first or second line also reported cumulative efficacy outcomes. Median PFS was 4.1 months, median OS was 19.5 months and ORR were 64% and 19% for anti-PD1/PD-L1 and Ipilimumab, respectively. Compared to a cohort of 169 patients (53 with BRAF V600 mutation and 116 BRAF/NRAS wild type), immunotherapy for NRAS mutant patients increased DCR, which did not translate into longer PFS and OS.²⁶

Regarding efficacy of anti-PD1+anti-CTLA4 combination compared with anti-PD1 monotherapy in NRAS melanoma, in a cohort of 69 patients treated in first line for unresectable and/or metastatic disease, combination significantly longer PFS (not reached vs 7 months) and a non-significant trend towards prolonged OS (not reached vs 21.9 months).²⁷

Given the different outcomes regarding NRAS-mutant melanoma response to immunotherapy, prospective series are needed to clarify if immune checkpoint inhibitors improve outcomes in this population.

Strategies in Clinical Trials for NRAS Melanoma

Farnesyl-Transferase Inhibitors

NRAS needs to undergo post-translational modifications in order to be active. Farnesylation of a cysteine allows the insertion of NRAS into the cellular membrane and its ulterior activation.²⁸

Farnesyl transferase inhibitors have shown promising preclinical activity.^29,30 Regrettably, these results did not translate into a clinical benefit from lonafarnib³¹ and tipifarnib³² in solid tumors with NRAS and KRAS mutations.

In melanoma, R115777 was tested in a Phase II trial enrolling 14 patients. No data about their NRAS status was reported. Unfortunately, no response was evidenced to the farnesyl transferase inhibitor.³³

The lack of success observed appears to be related to alternative prenylation by geranylgeranyltransferase I (GGTase I) in the alternative prenylation.³⁴ Farnesyl transferase and GGTase I inhibitors combinations have been evaluated in clinical trials with concerns regarding toxicity.³⁵

MEK Inhibitors

Due to the absence of direct NRAS inhibitors, focus moved towards targeting downstream effectors of the MAPK pathway, with MEK 1/2 inhibitors (MEKi).³⁶ The use of MEKi is to date the most investigated approach. These drugs are oral, competitive or non-competitive, allosteric inhibitors of MEK.³⁶

Third-generation MEKi development have led to evaluation of drugs such as trametinib (GSK1120212)³⁷ and binimetinib (MEK 162)^38,39 in NRAS-mutant melanoma patients.

The Phase I study of trametinib included seven patients with NRAS-mutant melanoma. Stable disease in two patients (29%) was the best response achieved.³⁷ No further development was done for trametinib in NRAS-mutant melanoma. Combination strategies of trametinib with novel agents could play a role in future clinical trials for NRAS-mutant patients.

On the other hand, the phase II trial of binimetinib showed encouraging activity in 30 NRAS-mutant melanoma patients with an ORR of 20% and a median PFS of 3.7 months.³⁸ Taking into account these results, a randomized Phase III trial (NEMO trial) was undergone in 402 NRAS mutated melanoma patients. Binimetinib significantly increased ORR (7% vs 15%) and PFS (1.5 vs 2.8 months; Hazard Ratio 0.62) compared to dacarbazine. However, no differences in terms of OS were observed.³⁹ Based on these results, binimetinib monotherapy did not received approval for the treatment of NRAS-mutant melanoma.

Other MEKi such as pimasertib (AS703026)⁴⁰ or RO4987655⁴¹ have also been tested for efficacy. Pimasertib was evaluated in a randomized phase II trial with 194 NRAS-mutant patients versus Dacarbazine. A significant benefit for pimasertib was observed, with a mPFS of 13.0 weeks (vs 6.9 weeks) and a DCR of 37.7% (vs 26.6%). Once more, no differences in OS were observed (8.9 vs 10.6 months).⁴⁰ RO4987655 (CH4987655) has also been tested in a phase I trial enrolling 8 patients with NRAS-mutant melanoma. ORR was 13% with partial response (PR) as best responses.⁴¹

RAF Inhibitors

RAF isoforms (ARAF, BRAF and CRAF) are the downstream effector after RAS proteins in the MAPK pathway. NRAS-mutant melanoma is refractory to BRAF inhibitors. BRAF inhibitors induce paradoxical activation of the MAPK pathway due to CRAF-mediated ERK phosphorylation.^42,43

Pan-RAF inhibitors have emerged as an option for RAS or RAF mutant tumors. LY3009120, a pan-RAS inhibitor, has been assessed in the setting of a phase I clinical trial which enrolled a total of 51 patients. Regarding molecular subgroups, only 5 patients with NRAS-mutant tumors (4 melanoma and 1 breast cancer patients) were included, with only 1 SD as best response. Given the lack of efficacy and the unfavorable toxicity profile, LY3009120 further development has been discontinued.⁴⁴ Belvarafenib, a novel pan-RAF inhibitor, is under phase I evaluation and its results are awaited.⁴⁵

In order to increase pan-RAF efficacy, preclinical combination with MEK inhibitors has been investigated. Pan-RAF inhibitor (Amgen Compd A) and Trametinib combination was found to significantly enhance cell growth inhibition and suppress MAPK activation (evaluated by p-ERK Western blotting) compared to monotherapy in NRAS-mutant melanoma cell lines. Efficacy seem to be associated with MAPK dependency and the presence of MAPK-independent pro-survival and anti-apoptotic signals.⁴⁶

PI3K-AKT-mTOR Pathway Inhibitors

PI3K-AKT-mTOR pathway activation is commonly found in both BRAF and NRAS mutant melanoma. AKT3 amplification and mRNA overexpression are frequent in RAS mutant melanomas (around 40% of the cases), whereas PTEN mutations and deletions are enriched in BRAF mutant melanomas (around 20% of the cases). PIK3CA mutations have been observed in 3% of all melanomas.³ Combination PI3K or AKT inhibitors with MEK inhibitors has shown synergic growth inhibition in melanoma cell lines with activating NRAS mutations.⁴⁷

Dual inhibition with alpelisib (a selective PI3Kα inhibitor) and binimetinib has been assessed in a phase Ib trial enrolling 58 patients with BRAF or RAS mutated advanced solid tumors. Five patients with NRAS-mutant melanoma were enrolled achieving an ORR of 20%.⁴⁸

The combination of pimasertib and voxtalisib (a dual PI3K/mTOR inhibitor) was tested in a phase Ib trial including 146 patients. Twenty patients with BRAF V600-mutant melanoma (progressing on BRAF inhibitors) were included (3 patients in escalation cohorts and 17 patients in disease specific expansion cohort). Response rate was 14% in the melanoma cohort (1 complete response and 1 PR). Absence of clinical efficacy and tolerability concerns led to premature trial termination.⁴⁹

Trametinib combination with an AKT inhibitor (GSK2141795) has also been evaluated in a phase II clinical trial including 10 NRAS mutant patients with melanoma. No objective responses were observed and the PFS was 2.3 months.⁵⁰ Even though preclinical evidence supported dual MAPK-PI3K/AKT/mTOR inhibition, clinical trials show absence of efficacy and poor drug combination tolerability.

CDK4/6 Inhibitors – Cell Cycle

Cell cycle is deregulated in BRAF and NRAS-mutant melanomas. NRAS mutant melanomas harbor CDKN2A alterations (including mutation, deletion or promoter hypermethylation) and CCND1 amplifications in 70% and 10% of cases, respectively.³

Moreover, NRAS activation causes an increase on cyclin D1 expression which regulates cyclin-dependent kinase 4/6 (CDK 4/6) involved in G1/S cell-cycle checkpoint.⁵¹

Based on these observations, combination of ribociclib (a CDK 4/6 inhibitor) and binimetinib has been evaluated in 63 NRAS mutant advanced or metastatic melanoma patients under a phase Ib/II trial. Results from the phase II dose expansion showed an ORR 19.5% (n=41) with a PFS of 3.7 months.⁵²

Ongoing trial in NRAS mutant melanoma patients involving the combination of LXH254 (an RAF inhibitor) with LTT462 (an ERK 1/2 inhibitor), Trametinib or Ribociclib is still recruiting.⁵³

Epigenetics

NRAS-mutant melanoma is more associated with CpG Island methylation pattern (CIMP) than BRAF-mutant melanoma. Mutations in the chromatin remodeler ARID2 gene and the epigenetic regulator IDH1 were also found in 16% and 9% of NRAS-mutant samples, respectively.³ Combination strategies with epigenetic modulators could increase efficacy of MEK inhibitors.

Data from de TCGA revealed that high BRD4 mRNA expression was associated with poor outcomes. Bromodomain and Extraterminal Domain (BET) proteins, such as BRD4, read acetylated lysine residues in histones and non-histone proteins promoting gene expression. In vitro (spheroids) and in vivo (xenografts) treatment combining JQ-1 (a BET inhibitor) and Mirdametinib (MEKi) decreased tumor growth rate and induced apoptosis in NRAS-mutant melanoma models. Efficacy was associated with downregulation of the TCF19 transcription factor and E2F1/3-dependent targets, involved in cell cycle G1 to S transition.⁵⁴

Tyrosine Kinase Inhibitors

NRAS melanoma cell lines, unlike BRAF-mutant melanoma, harbor constitutive phosphorylation of receptors with tyrosine kinase activity, such as Axl, ERBB2, c-MET or EGFR.⁵⁵ However, activity of tyrosine kinase inhibitors have been examined in melanoma patients with limited benefit as monotherapy.

As an example, lenvatinib achieved an ORR of 9% and a mPFS 3.7 months in patients with BRAF wild type advanced melanomas.⁵⁶ Axitinib yielded an ORR 18.8% and a 6-months PFS 33.9% in molecular unselected population.⁵² Patients were less pre-treated (maximum of 1 previous line) and only 25% of patients received immunotherapy in the axitinib trial compared to lenvatinib trial.^56,57

Sorafenib and tivantinib (a MET inhibitor) combination have been evaluated in eight patients with NRAS-mutant melanoma. PR was objectified in two patients and another two stable disease (SD) as best response.⁵⁸

Axitinib has been also combined with chemotherapy. Combination with Carboplatin/Paclitaxel (n=38 patients) increased ORR to 22% and PFS to 8.7 months in unselected population, with better outcomes for non-BRAF V600E/K population. Regarding NRAS-mutant patients (n=8), 2 patients achieved PR and 6 patients SD as best response.⁵⁹

Combination of TKI with MAPK or PI3K-AKT-mTOR inhibitors might decrease acquired resistance to targeted therapy and compensatory feedback of the intracellular pathway.⁶⁰ Combination of pazopanib and trametinib has been performed under a phase I trial. Four patients with melanoma were enrolled, with SD as best response in 3 patients. Unfortunately, no information about RAF/RAS status was given.⁶¹

ERK Inhibitors

ERK is the final kinase in the MAPK signaling pathway. Therefore, ERK inhibitors (ERKis) could be an interesting treatment target for NRAS-mutant melanoma patients.^13,14 Preclinical data supports this concept.

SCH772984 has proved efficacy in NRAS, KRAS, and BRAF-mutant cell lines as in melanoma animal resistant to BRAF inhibitors.⁶² In NRAS-Q61H mutant cell line, ERK inhibition with VX-11e displayed greater proliferative arrest than MEKi. Treatment with MEKi did not suppressed ERK-dependent phosphorylation of downstream targets. RTK activation and cross-talk of other signaling pathways with ERK pathway were the explanation.⁶³

Ulixertinib is a reversible, ATP-competitive ERK1/2 inhibitor. Safety and preliminary efficacy have been assessed in a phase-I clinical trial enrolling patients with NRAS-mutant and BRAF-mutant melanomas. Preliminary results from one of the expansion cohorts for NRAS-mutant melanoma patients (n=19), not previously treated with BRAF and/or MEK inhibitors, showed PR in 13.5% and DCR in 52.6% of the patients.^64,65

Further development of novel ERKi in phase I clinical trials as single agents or in combination with MEKis, chemotherapy or targeted agents are awaited.

Table 3 summarizes the efficacy outcomes of targeted therapies (monotherapy and combinations) in patients with NRAS-mutant melanoma.

Table 3 Summary of Efficacy Outcomes of Targeted Therapies (Monotherapy and Combinations) in Patients with NRAS-Mutant Melanoma. “Sample Size” Refers to Number of Patients Included with NRAS-Mutant Melanoma in the Clinical Trial

Future Strategies for NRAS-Mutant Melanoma

The KRAS Situation

To date, targeting directly on GTP binding pocket in RAS protein has been difficult due to the strong affinity between GTP and RAS.¹⁸ Nevertheless, recently, a small molecule that specifically and irreversibly inhibits KRAS G12C (AMG 510 or Sotorasib) has shown antitumor activity in patients with KRAS G12C-mutant advanced solid tumors.⁶⁶

Sotorasib forms an irreversible covalent bond to the sulfur atom in the cysteine residue that is present in the mutated form of KRAS G12C, but not in the normal form. The inhibitor traps KRAS G12C in the inactive GDP-bound state, inhibiting its phosphorylation activity.⁶⁷

New RAS targeting strategies still need further development. For example, instead of targeting RAS protein activity we could block protein translation by RNA interference. Challenges regarding RNA delivery, intravascular degradation, intracellular trafficking, and potential off target effects are delaying its clinical implementation.⁶⁸

NF1

NF1 (neurofibromin 1) acts as a tumor suppressor through GAP activity which turns the active RAS-GTP to RAS-GDP.⁶⁹ Calpain 1 (CAPN1) is a calcium-dependent neutral cysteine protease that regulates NF1 degradation in melanoma cell lines.⁷⁰

Combination of Calpain 1 inhibitor with Trametinib added antiproliferative activity in terms of cell growth reduction compared to Trametinib alone in melanoma cell lines.⁷⁰ This combination may increase MEK inhibitors’ efficacy in NRAS-mutant melanoma in the future.

PPP6C

PPP6C (Protein Phosphatase 6 Catalytic Subunit) encodes a Serine/Threonine phosphatase involved in cell cycle regulation and progression.⁷¹ Among the different substrates of PPP6C we find MEK⁷² and Aurora-A.⁷³

Loss of PPP6C promotes MEK hyperphosphorylation (at activating and crosstalk phosphorylation sites) inducing ERK pathway signaling and resistance to MEK inhibitors in vitro.⁷²

Loss of its activity is also linked to elevated Aurora-A kinase activity, leading to mis-segregation of chromosomes during mitosis leading to chromosome instability. Aurora-A inhibitors may be a treatment strategy to treat melanomas harboring PPP6C inactivation.⁷³

STK19

STK19 (serine threonine kinase 19) is a serine/threonine kinase which regulates NRAS signaling.⁷⁴ It phosphorylates NRAS S89 residue, activating its signaling via the MAPK and PI3K pathways in human melanocyte cell lines.^74,75 STK19 activating mutations are mutually exclusive with BRAF in melanoma.⁷⁵

In NRAS Q61R transgenic mice, the STK19 D89N mutant promoted oncogenic NRAS driven melanomagenesis. ZT-12-037-01 (an STK19 inhibitor) reduced cell division and induced apoptosis in NRAS-STK19 mutant mice xenografts.⁷⁵ Consistent with these observations, STK19 inhibition could be a potential therapeutic strategy for NRAS-mutant melanoma.

PTPN11

PTPN11 (protein tyrosine phosphatase non-receptor type 11) gene encodes SH2, a phosphatase which contains two tandem Src homology-2 domains which interact with phospho-tyrosines in transmembrane cellular receptors. SHP2 activity is involved in RAS signaling activation.⁷⁶

Activating mutations of PTPN11 appear to be oncogenic in melanoma. Conversely, SHP2 inhibitor SHP099 promotes antitumoral response of NRAS Q61K-mutant melanomas in mice models.⁷⁷

OTHER

Other targets not directly related to RAS signaling which have shown early preclinical evidence in combination with MEKi involve inhibition of the ROCK 1/2 (GTPase-activated serine/threonine kinases),⁷⁸ ER (estrogen receptor)⁷⁹ or HSP90 (Heat shock protein-90).⁸⁰

Clinical Trials Ongoing

Table 4 summarizes ongoing clinical trials with potential beneficial treatment strategies recruiting patients with NRAS-mutant melanoma either as the main target population or included in unselected solid tumor cohort.

Table 4 Ongoing Clinical Trials Recruiting Patients with NRAS-Mutant Melanoma

Conclusion

NRAS-mutant melanoma constitutes around 15–20% of patients with melanoma and is known to be associated with poorer prognosis.

Clinical trial experience over the last years with different MEK inhibitors have shown limited efficacy in monotherapy. Treatment strategies in the near future will include combination of MEK inhibitors with inhibitors of the main intracellular pathways (either involved in RAS signaling or not) or other drugs disrupting cell-cycle checkpoints or epigenetics. Currently, little effort has been done towards mutant NRAS specific treatment, either with direct inhibitors or post-traductional interactions.

To date, checkpoint inhibitors seem to be as effective or even more effective in patients with NRAS-mutant melanoma compared to distinct melanoma subtypes.