Therapeutic Strategies in Neurocutaneous Melanocytosis

Molecular Mechanisms and Experimental Models

Experimental research involving drug discovery approaches has been largely impeded in NCM due to the scarcity of human samples and the lack of a faithful experimental model that recapitulates the disease. Historically, lack of interest from funding agencies also accounts for a dearth of support for basic, experimental research in NCM. Notwithstanding, the last decade saw an increase in NCM research initiatives, including the formation of a clinical registry of patients prospectively collecting information about the regular medical care of L/GCMN and NCM patients (ClinicalTrials.gov ID NCT04548817; Sponsor Memorial Sloan Kettering Cancer Center). Identification of a genetic driver of the disease facilitated the development of targeted therapies. Genetic mosaicism and clonal expansion of oncogenically mutant nevocytes have been proposed as the underlying cause of the formation of pigmented lesions, but the cell of origin in NCM remains elusive. It is still open for debate if the oncogenic mutation is acquired in developing melanocytes or even earlier in the developmental timeline at the undifferentiated neural crest cell stage (34). Considering that fully differentiated skin melanocytes in adults respond to oncogenic mutation (BRAF V600E or otherwise) by an oncogene-induced senescent (OIS) benign phenotype comprising of mostly pigmented growth-arrested cells, commonly identified as a small nevus or “mole”, congenital nevi are significantly larger, comprises of cell division capable cells displaying plasticity and, in some cases, proliferative nodules suggest a distinctly different etiology. Additionally, brain lesions observed in NCM patients in conjunction with their skin lesions carrying the same oncogenic mutation point to an embryonic origin, perhaps before the migration of embryonic melanocytes from the neural crest rather than a skin melanocytic origin. A plausible hypothesis is that oncogenic mutations acquired during the embryonic neural crest stage induce the proliferation of melanocytic progenitors. These progenitors, upon differentiating into pigment-producing melanocytes (or nevocytes), will likely form pigmented lesions. The embryonic environment, abundant in growth factors, may facilitate this proliferation by delaying the onset of oncogene-induced senescence, leading to the formation of larger, benign lesions at birth (Figure 1). However, currently, there is no experimental model available to test this hypothesis. Expanding further on this line of thought, undifferentiated nevocytes may persist in specific niches that support their undifferentiated state, thereby acting as a reservoir for mature nevocytes. This hypothesis may account for the slow growth of congenital nevi and NCM compared to acquired nevi (that is, growth-arrested) and aligns with experimental observations of nevocyte behavior both in vitro and in vivo.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1.

Schematic depiction of NCM development hypothesis. Top panel, Neural crest cell differentiation into a fully functional melanocyte. Oncogene activation elicits oncogene-induced senescence, producing acquired nevi. Acquisition of additional genetic alterations gives rise to melanoma. Oncogene activation in a developing neural crest cell/melanocytic progenitor in cooperation with supportive embryonic signaling may allow for larger lesions of congenital origin, including NCM, to develop. Additional genetic events may lead to malignant transformation. NCM: Neurocutaneous melanocytosis. The Figure was created in BioRender.com.

Neural crest cells (NCCs) are transient embryonic cells unique to vertebrates with stem cell-like properties having the potential to give rise to neurons and mesenchymal progenitors, including components of the peripheral nervous system and craniofacial osteocytes other than melanocytes (35). The observation that very rarely, ensuant non-melanocytic lesions of neural crest derivatives are also detected in NCM, such as rhabdomyosarcoma, meningioma, and lipoma, lends credence to this possibility (26-28). Experimentally, forced expression of NRAS Q61K under the control of tyrosinase promoter (expression of a mutant oncogene is targeted to melanocytic lineage progenitors) in mice failed to produce brain lesions, although hyperpigmentation of the skin in this model histologically was reminiscent of human congenital nevi (36). Interestingly, the hepatocyte growth factor (HGF), when overexpressed in mice during development, produced both skin hyperpigmentation and pigmented lesions in the brain neonatally (37). HGF plays a crucial role in neural crest delamination and differentiation into several cell types, including melanocytes (38). HGF binding to its receptor MET activates a variety of intracellular signaling. HGF is known to enhance RAS signaling through the MAPK pathway culminating in cell proliferation, activate PI3K-Akt-mTOR signaling to support cell survival and induce motility through p21 activating kinase (PAK) and STAT3 (39). Increased bioavailability of HGF would likely amplify the MAP kinase pathway downstream of RAS, resulting in higher rates of cell proliferation (Figure 2). The HGF-Tg overexpression model, although produced benign leptomeningeal lesions resembling human NCM, this model did not address the role of mutant NRAS Q61K or BRAF V600E as the driver of the disease. It can be surmised that HGF-MET hyperactivity in the same cells affected by NRAS oncogenic activation would likely overamplify the RAS-MAP kinase pathway, leading to a more aggressively proliferating phenotype in addition to supporting cell survival and resisting senescence. Investigations into the possible role of HGF in suppression or delaying OIS are ongoing. Likely, HGF contributes indirectly to senescence prevention by reducing cellular stress and maintaining growth and survival. Takayama et al. reported higher levels of MET protein, a cellular receptor for HGF, in NCM lesions in a case report (40). Unfortunately, the presence of oncogenic mutations in NRAS and/or BRAF was not investigated in that case. Later case reports of NCM patients that detected NRAS Q61K/R or BRAF oncogenic mutations did not investigate the abundance of MET protein in lesional tissues either. Therefore, the molecular events underlying the origin of NCM are still unclear. A pertinent question to ask is if increased HGF activity/bioavailability and oncogenic activation of NRAS or BRAF act synergistically and whether such cooperation is necessary to produce NCM brain lesions. Considering the important role of HGF during fetal development and the possibility of maternal metabolic changes during pregnancy leading to fluctuations of HGF in the fetal microenvironment, this becomes a clinically relevant question (41, 42). Ackerman et al. reported the development of a transgenic mouse where forced expression of NRAS Q61K in developing melanocytes produced only skin hyperpigmentation due to melanocytosis, however, these animals failed to develop leptomeningeal melanocytosis, leaving this question open (36). It remains to be explored if HGF overactivity aids in the development of leptomeningeal lesions in Tyr-NRAS Q61K transgenic mice. On the other hand, Richard Marais’s group reported the development of primary leptomeningeal melanoma when oncogenic NRAS G12D is expressed in transgenic animals under the control of the same tyrosinase promoter (43). This also points to the diversity of outcomes regarding the oncogenicity of different NRAS hotspots in the context of developing melanocytes. Notwithstanding, the majority of NCM patients reported in the literature were positive for either NRAS Q61K or BRAF V600E and not NRAS G12D, the mutations being mutually exclusive (4, 44, 45). Thus, a transgenic mouse model harboring either NRAS Q61K/R or BRAF V600E with pigmented benign leptomeningeal lesions will likely be a faithful model of the disease provided histopathology is comparable with the human disease.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2.

Schematic representation of RAS-MAPK and HGF-MET signaling pathways and possible interactions (shown by arrows) of downstream signaling intermediates. RAS: Rat sarcoma family protein; MAPK: mitogen activated protein kinase; HGF: hepatocyte growth factor; MET: mesenchymal-epithelial transition factor (receptor for HGF). The Figure was created in BioRender.com.

Following in these lines, Pawlikowski et al. reported a transgenic mouse model where Tyr-NRAS Q61K expression was coupled with aberrant Wnt signaling. In this model, brain lesions were produced in transgenic animals in addition to benign melanocytic proliferation in the skin (46, 47). This model remains the only animal model closely resembling NRAS-mutant NCM to date. However, the genetic defect resulting in aberrant Wnt signaling in this model has not been detected in human patients. Defects in any component of the Wnt signaling pathway have not been reported in human patients so far. Some of the polymorphisms in Wnt pathway genes detected by the authors in their cohort of patients can be expected to produce Wnt signaling aberrations based on computer modeling predictions, but actual functional data is still elusive in humans. This remains a serious deficiency of this animal model. Interestingly, this also highlights the role of molecular factors of the embryonic microenvironment in addition to oncogenic NRAS contributing to the development of NCM.

The development of faithful animal models of the disease must rely on the genetic data of human patients, and to that end, more high throughput sequencing and transcriptomic profiling of patients’ lesions are necessary. The information available currently suggests possible cooperation between a genetic mutation in NRAS or BRAF and additional factors present in the embryonic environment during development playing a role. It is not known for certainty what these factors are, but HGF seems to deserve a strong consideration. Therefore, a transgenic mouse model combining these aspects may produce a true recapitulation of NCM. Nonetheless, it is also imperative that the genetic changes created in the animal model remain relevant to human patients. That is yet to be achieved.

Drug Discovery and Targeted Therapies

To develop effective therapies targeting aberrant cells of NCM, it is important to identify the cells of origin and special features of NCM transcriptome to understand the vulnerabilities and drug susceptibilities of these cells. Recently, experimental studies were reported in models derived from patients’ cells in vitro and xenografts. Based on these reports, some new strategies are now being discussed, although their translatability will require rigorous trials.

The majority of NCM cases have NRAS Q61K as the driver mutation, although BRAF mutant NCM cases have also been reported (48). Oncogenic NRAS and BRAF both act through the MAP kinase signaling pathway, which is linked to the cellular proliferation machinery. NRAS being a GTPase, direct targeting of such a molecule has remained challenging for scientists and, in many instances, was deemed undruggable (49). NRAS, when mutated at the 61^st residue to code for Lysine (K) instead of Glutamic acid (Q), becomes constitutively active, bound to GTP, and sequentially activates downstream effectors. RAF kinase (activated by NRAS) phosphorylates MEK (MAPK/ERK kinase), which subsequently activates ERK (extracellular signal-regulated kinase). Expression of cell proliferation, differentiation, and survival genes are influenced by ERK post-activation and upon its translocation to the nucleus (50, 51). Therefore, targeted inhibition of MAP kinase signaling intermediates was considered an attractive strategy to disrupt NRAS oncogenic signaling (52-54). There are ongoing clinical trials that use small molecule inhibitors of the MAP kinase pathway in several NRAS-driven cancers. There are only a few reports in the literature where such potential drugs have been experimentally tested in NCM. Trametinib, or MEK162, which inhibits MEK, a major kinase mediating signaling through the MAP kinase pathway, has been well studied in NRAS-driven cancers and has been approved by the FDA for conducting clinical trials of NRAS mutant melanoma. In preclinical studies, MEK162 was effective in regressing NCM-like CNS lesions in a transgenic mouse (47). However, oncogenic NRAS is also likely to activate redundant signaling parallelly through Akt and continue to function even when MAP kinase signaling is blocked. In line with this idea, MEK inhibition was found to be not as effective alone in a three-dimensional cellular model wherein activated Akt was shown to be functional in NRAS mutant NCM cells (55, 56). Additionally, trametinib may not be useful in NCM cases having wild-type NRAS and BRAF. Therefore, alternative strategies are needed to target cellular mechanisms that are common in most NCM cases. In a report by Ruan et al., 2014, cells isolated from an NRAS and BRAF wild-type NCM patient xenografted in immunocompromised mice showed sensitivity to an IGF receptor blocker, which was proposed as a druggable target in NCM (57). This molecule successfully reduced cell viability also in NRAS mutant NCM cells, which led to the idea that IGF is a common contributor to lesional growth in NCM irrespective of the mutation status (58). Since the IGF receptor predominantly signals through the PI3K – Akt signaling pathway, which is also known to be used by oncogenic NRAS as a redundant signaling alternative, this idea seems tenable. However, PI3K-Akt is also connected through a positive feedback loop to the mTOR signaling pathway that regulates cell survival. Therefore, treatment with Omipalisib, a dual inhibitor of PI3K-mTOR, abrogated cell survival and caused autophagic cell death in the anchorage-independent nevosphere model of NCM patients’ cells (56). These studies point to the importance of targeting multiple signaling intermediates in NCM cells as an effective therapeutic paradigm. Following in these lines, several experimental therapeutic regimens have been tested in the clinic. In most cases, biomarker analysis indicated target validation and initial response indicated tumor regression, but ultimately, the patient did not survive due to advanced disease and/or secondary complications. It would be premature to conclude the efficacy of these experimental treatments without appropriately designed clinical trials and statistically significant datasets. The rarity of the disease poses significant challenges to designing of clinical trials of NCM to test investigational drugs.

Experimental Treatments With Targeted Drugs and Clinical Outcomes

Conventional chemotherapy with Dacarbazine, Vincristine, and Cyclophosphamide has been tried unsuccessfully with NCM patients in the clinic before a driver mutation was identified in NCM and targeted therapy was proposed. Makin et al. reported 5 cases of NCM, including one with primary leptomeningeal melanoma spanning a period of 13 years (1984-1997), who were treated at two hospitals in Northern England. Several combinations of the above-mentioned drugs were used with or without Etoposide. The patients, all children, died from disease complications within 2-8 months of diagnosis without any notable improvement in the clinical presentation of the disease (59). These drugs are classified as broad-spectrum chemotherapeutics with cytotoxic effects on a variety of tissue cells, owing to their inhibitory effect on cellular processes such as replication of DNA and microtubule assembly, etc., during cell division. On the contrary, rational drug discovery approaches showed effective responses when oncogenic NRAS signaling was targeted based on mutation analysis of tumor biopsy. These included MEK162 (known as Trametinib), either alone or in combination with other drugs. Case reports are scarce but provide important insights into the roles of such treatments in regression and disease-free survival of the patients (Table I).

MEK162 has been reported to produce favorable results in clinical trials with cancers that involved dysregulated MAP kinase or ERK (60, 61), suggesting it could be tested as a therapeutic candidate for NRAS mutant NCM. A 13-year-old patient with NCM positive for somatic NRAS Q61K mutation was treated with MEK162 on a compassionate use basis after repeated surgery failed to prevent disease progression. The patient died five days after the therapy was started. Leptomeningeal tumor samples were studied for the effect of the drug, which showed a marked decrease in MIB1 and phosphorylated ERK (target of MEK162) proteins, suggesting a positive biomarker response (62). Important to note here is the histological finding that the lesion was comprised of cells ovoid and monotonous and exhibited low mitotic activity and no necrosis, which differs from typical melanoma histological features, although the lesion was extensive and covered significant areas of the brain and spinal cord.

Following this report, there was another report of four patients with NRAS-mutant brain melanoma that arose in conjunction with CMN and associated CNS abnormalities who received Trametinib as treatment. All these patients were diagnosed with diffuse leptomeningeal melanoma that harbored NRAS Q61K mutation and were symptomatic at the time of diagnosis. After they were started on Trametinib treatment, mild improvement was noted in all patients, and the symptom-free period varied from one month to 9 months; however, discontinuation of therapy resulted in rapid deterioration and death. The authors noted that the underlying cause of disease progression could be attributed to the presence of additional mutations in the melanoma tumor (63).

In 2018, Sharouf et al. described a case diagnosed with giant congenital melanocytic nevus, hydrocephalus, and leptomeningeal melanocytosis at five years of age. The patient was treated with Trametinib for seven months in addition to a ventriculoperitoneal shunt placement. The patient experienced symptomatic relief during this period, followed by detection of status epilepticus at 30 months of age. NRAS mutant melanoma was detected at 21 months of age from leptomeningeal and skin biopsies. At this stage, the patient’s condition started deteriorating rapidly and finally succumbed to death at 32 months of age (64). The authors did not present data on biomarker evaluation of drug response in this study. Although Trametinib was not as effective in controlling melanocytic tumor progression in the CNS context, interestingly, trametinib treatment of skin lesions of L/GCMN was found to be extremely effective in a patient whose lesional cells carried an AKAP9-BRAF fusion and, hence, should have activated the MAP kinase pathway (65).

Trametinib, when combined with other drugs, has shown improved treatment outcomes. In 2020, Vanood et al. reported NCM diagnosis in a 13-year-old patient with multiple congenital melanocytic nevi and clinical symptoms of seizure. The patient was found positive for NRAS Q61K mutation, and a diagnosis of NCM was made from MRI analysis. The patient was initially treated with Levetiracetam to manage seizures, the dosage of which was later increased and coupled with Trametinib. The authors report that the patient had no seizure episode reported at the 10-month follow-up, and at the one-year follow-up, no further clinical progression of neurological symptoms was noted (66). The authors describe the veracity of the disease as not so severe, offering hope that drug administration at the early onset of symptoms could offer a favorable prognosis.

Another interesting approach was followed by Hanft et al., 2022 to treat a patient with diffuse leptomeningeal melanosis and CNS melanoma. In this case, the patient was treated with azacitidine (DNA methyltransferase inhibitor) in combination with trametinib, and a durable positive clinical response was noted. The authors also showed that azacitidine acted synergistically with Trametinib in trametinib-resistant melanoma cells in vitro (67). More work needs to be done to understand the underlying molecular mechanism.

In a comprehensive analysis reported by Abele et al. on the treatment outcome of CNS melanoma patients, documented in the “German Registry for Rare Pediatric Tumors (STEP registry)”, aged 0-18 years, whose melanoma arose from NCM, nearly 100% fatality was noted. 85% of patients carried NRAS oncogenic mutation, and n=9 patients were treated with Trametinib. The authors noted short-term clinical improvement in some patients from treatment with Trametinib or Temozolomide, adding that the regimens failed to alter the overall survival rates. They infer that the overall survival rate of patients with CNS melanoma arising in NCM, even after targeted chemotherapy, was found to be worse in the pediatric group when compared to adults with primary CNS melanoma, whose one-year overall survival was around 50% (68).

These studies indicate that Trametinib, despite its targeted inhibition of the MAP kinase pathway, may not be sufficient as a monotherapy to produce an effective, durable clinical response in NCM in the context of CNS lesions. NCM, with all its similarities with other melanocytic malignancies, is also different in its response to chemotherapy targeting oncogenic NRAS signaling alone. In conclusion, more research needs to be done to understand the cellular and molecular characteristics of NRAS/BRAF mutant NCM, identify drug susceptibilities of mutant NCM cells, delineate the origin and progression of the disease and the role of microenvironment in it towards developing a specific therapeutic regimen for NCM.

Future Directions

Future research investigations into the therapeutic landscape of NCM would benefit from a multidisciplinary approach drawing from advances in genetics, molecular biology, pharmacology, and improved patient care strategies. Several key areas of investigation could be:

Targeted therapies. Investigations aimed at understanding the molecular etiology of NCM need to be undertaken to thoroughly understand the drug susceptibilities of NCM cells. High throughput sequencing studies at the level of whole genome and/or transcriptome would reveal new insights into specific mutations and associations between other genetic and/or non-genetic factors contributing to altered cell signaling, which in turn could contribute to the development and progression of the disease. With significant technological advances in the field of transcriptomics, gene expression patterns can now be analyzed at single-cell resolution, allowing for a deeper analysis of genes, gene networks, and cellular pathways underlying a disease phenotype. This approach could lead to the identification of druggable targets within NCM cells and/or the tumor microenvironment that more precisely addresses the underlying pathology.

Novel drug combination therapies. New drug development approaches may be followed that target multiple cellular pathways in addition to the major driver mutation(s) to provide for a more robust and durable response. Some of the probable targets could be epigenetic modifiers [such as azacytidine (67) or Vorinostat (69)], drugs targeting growth factor receptors (such as IGF1 receptor blocker or MET inhibitors), drugs targeting cell survival mechanisms (such as mToR inhibitors) in addition to targeting the oncogenic signaling through the MAP kinase pathway. Combination therapies may also be tested for more effective management options for patients depending on the severity of disease progression and clinical symptoms.

Gene therapy. This approach involves gene editing technology to correct or compensate for a defective gene that is linked to the development of disease (in this case, the mutant oncogene that drives NCM cell proliferation). Although the technology is still in its early stages of development, it represents a promising new strategy to be tested in NCM since the underlying pathogenic driver of NCM is genetic mutations in NRAS or BRAF.

Immunotherapy. Although published reports using PD1 blockers and interferons did not improve overall survival in NCM, and immunotherapy using monoclonal antibodies could be challenging considering their inability to effectively cross the blood-brain barrier (70), more studies need to be undertaken to investigate the overall role of the immune system in NCM. Inflammation and pruritus have been observed in patients with L/GCMN and cutaneous lesions of NCM patients. It remains to be explored if targeting inflammatory pathways produces any beneficial outcome in the context of CNS lesions.

Developing a relevant animal model of NCM. Efficient animal models are needed to study in vivo drug efficacies observed in vitro for translating those findings into effective treatments in the clinic. Previous studies by our group and others have suggested that NRAS Q61K mutant cells isolated from L/GCMN and NCM lesions may be responsive to dual PI3K-mTOR inhibitors owing to the redundant signaling of mutant NRAS through the MAP kinase and Akt pathways in vitro (56, 71, 72). A faithful animal model of NCM will facilitate experimental approaches testing gene therapy, and immunotherapy approaches in addition to targeted and combinatorial chemotherapies in vivo. Combination therapies aimed at alleviating neurological symptoms may also be tested in animal models that can recapitulate them along with targeted inhibition of the mutagenic signaling.

Advanced imaging techniques. With more and more advancements in the field of imaging technology, the ability to monitor and evaluate the progression of NCM will be improved. This will not only empower diagnostic capabilities but also will enable more specific and timely interventions.

Patient registries and data collection. Considering the rarity of the disease, data on the patient’s disease pathology, clinical outcomes concerning treatment responses, and overall clinical and molecular parameters are extremely valuable to devise future directions of research. Increased interaction and partnerships between clinicians, researchers, and patient advocacy groups are vital to developing collaborations facilitating such collection of data.

Clinical trials. Participation of patients in ongoing and future clinical trials is key to the development of novel therapies and the improvement of current therapies. Being a rare disease, conducting clinical trials for NCM could be a challenging task in terms of patient enrolment and participation. Currently, NCM patients participate in trials designed for other closely related diseases, such as primary CNS melanoma and/or other cancers of CNS.