Probing the Melanoma Genome Reveals New Targets and Challenges

OncologyLive, Vol. 17/No. 12, Volume Vol-17-No-, Issue 12

Next-generation sequencing technology is providing greater insight and has uncovered new and unexpected players in melanoma.

The past 5 years have witnessed a dizzying pace of progress for the treatment of melanoma, with nearly a dozen new molecularly targeted and immune-based therapies transforming the disease from an aggressively lethal malignancy into one that is readily treatable.

Yet, many patients continue to succumb to disease since only some respond to the available drugs and those who do respond often develop resistance. To carve out a path forward requires a broader picture of melanoma development and progression.

Expanding Treatment Options

Next-generation sequencing technology is providing greater insight and has uncovered new and unexpected players in melanoma. It has also revealed one of the highest rates of somatic mutations among all types of cancer, presenting a significant challenge to pick out the true drivers of this disease. Tackling this challenge will likely have important implications for prognosis and therapy in the future.Melanoma is a malignant tumor of the skin pigment—producing melanocytes. As researchers have begun to characterize melanoma at the molecular level, it has become clear that the mitogen- activated protein kinase (MAPK) pathway is a key mediator of growth and proliferation.

The BRAF kinase, a MAPK pathway component that is mutated in around half of all patients with melanoma, offered an obvious target. The development of small-molecule inhibitors of this protein culminated in FDA approval of vemurafenib (Zelboraf) in 2011 and dabrafenib (Tafinlar) in 2013.

UV Damage Leaves Characteristic Marks

1. Lovly C, Pao W, Sosman J. My Cancer Genome website. Updated January 26, 2016. Accessed June 2, 2016.

2. The Cancer Genome Atlas Network. Cell. 2015;161(7):1681-1696.

3. Harbour JW, Onken MD, Roberson ED, et al. Science. 2010;330(6009):1410-1413.

Meanwhile, attempts to target other components of the MAPK pathway led to approval of inhibitors of the kinase immediately downstream of BRAF, mitogen-activated protein kinase kinase (MEK). Melanoma also has been the poster child for immunotherapies that boost the antitumor immune response, particularly immune checkpoint inhibitors. The cytotoxic T-lymphocyte antigen 4 (CTLA-4) inhibitor ipilimumab (Yervoy) launched the field and was followed by drugs targeting programmed cell death-1 (PD-1) receptor and its ligands PD-L1 and PD-L2. Pembrolizumab (Keytruda) and nivolumab (Opdivo) both received regulatory approval in 2014, and were joined by the first immune combination therapy of ipilimumab plus nivolumab in 2015.For the most part, comprehensive genome-wide studies have focused on the most common and aggressive form of melanoma, cutaneous melanoma. Numerous whole-genome and whole-exome studies have been performed, and melanoma was also among the tumor types chosen by The Cancer Genome Atlas (TCGA) as part of its sequencing efforts.Ultraviolet radiation (UVR) is the major environmental risk factor for cutaneous melanoma and this has been reflected in the genomic profiles uncovered in sequencing studies. The vast majority of tumors display a genomic signature indicative of damage caused by UVR, dominated by cytosine-to-thymine (C>T) nucleotide substitutions and CC>TT mutations. In the TCGA’s study, 76% of primary and 84% of metastatic tumors displayed a UVR signature of >60% C>T transitions and >5% CC>TT mutations.

Unveiling Key Players Old and New

Genome sequencing studies also revealed the strikingly high number of somatic mutations in melanoma. According to the TCGA, there were 16.8 mutations/Mb, the highest reported for any cancer evaluated through the program. The high mutational load is thought to result from UV radiation exposure and the high prevalence of C>T substitutions. Furthermore, it’s thought that this is what makes melanomas so immunogenic and thus uniquely susceptible to immunotherapies.Predictably, the standout finding has been the identification of BRAF mutations across all studies, in 50%-60% of cutaneous melanomas. The most common type of BRAF mutation, in up to 90% of cases, results in the substitution of a valine at position 600 to a glutamic acid (V600E).

BRAF mutations are associated with increased sensitivity to BRAF- and MEK-targeted therapies, and determination of BRAF mutation status is a prerequisite for treatment with these drugs. Currently, BRAF and MEK inhibitors are only approved for use in advanced-stage melanomas, but mounting evidence suggests that BRAF mutations can be identified in earlier stages of melanoma.

Numerous other driver genes have now been identified in melanoma, including other components of the MAPK pathway. After BRAF, the most frequently somatically altered gene in melanoma is NRAS, a member of the RAS GTPase family of proteins, which sits upstream of BRAF in the MAPK pathway, recruiting it to the cell membrane to be activated.

NRAS mutations occurred in 28% of tumor samples studied by the TCGA, which analyzed the genomes of 333 cutaneous primary and/or metastatic melanomas and matched peripheral blood from 331 adult patients from 14 tissue source sites. The majority of NRAS mutations occur in codon 61 and prevent the protein from being inactivated, thus continually stimulating the MAPK and other pathways in which the RAS protein is involved, including the phosphatidylinositol-3-kinase (PI3K) pathway.

Overall, the TCGA study identified 13 significantly mutated genes by whole-exome sequencing. In addition to BRAF and NRAS, other MAPK pathway genes that were significantly mutated included NF1, a GTPase-activating protein that serves as a negative regulator of RAS activity, terminating MAPK signaling. NF1 was mutated in 14% of tumor samples, more than half of which were predicted to result in a loss of function of the protein. Though fairly rare, there were also mutations in the MEK1 gene, also called MAP2K1, which have been noted in previous studies.

Cutaneous Melanoma

BRAF Subtype

Genes that have been implicated in melanoma but were missing from the TCGA’s list included PREX2, GRIN2A, ERBB4, MITF and KIT. PREX2 encodes a phosphatidylinositol-3,4,5-trisphosphate (PIP3)-dependent Rac exchange factor, activating RAC1. It has recently been shown that PREX2 is a negative regulator of PTEN. Mutations in GRIN2A have also been described in numerous whole-exome sequencing studies of melanoma, with an overall frequency of more than 20%. The GRIN2A protein is a member of the ionotropic glutamate receptors.The TCGA investigators used the significantly mutated genes that they identified to classify cutaneous melanomas into four subtypes with distinct genomic profiles and clinical characteristics.The largest was the BRAF subtype, defined by the presence of BRAF hotspot mutations in 52% of tumors, 87% of which were V600 mutations. The second most common BRAF mutation was in the K601 residue and both these and V600 mutations were mutually exclusive with NRAS mutations.

RAS Subtype

PTEN mutations and deletions were more frequent in these BRAF-mutant melanomas. There were also frequent focal amplifications in the BRAF, MITF, and PD-L1 genes in this subgroup, the latter of particular interest since PD-L1 expression can predict response to PD-1 inhibitors. Patients in this subgroup tended to be younger.The second subtype was distinguished by the presence of hotspot mutations in the RAS genes. All three members of the family were mutated, but NRAS most frequently. HRAS and KRAS mutations were mutually exclusive with NRAS and BRAF mutations. No direct inhibitors of NRAS are clinically available and RAS has proved a particularly challenging target.

NF1 Subtype

Triple Wild-Type

Since NRAS activates the MAPK and PI3K pathways, ongoing research is examining the possibility of blocking oncogenic NRAS signaling in melanoma by inhibiting downstream components of these pathways. Although single-agent PI3K inhibitors have proved ineffective in preclinical studies, MEK inhibitors have had significantly greater success in early clinical development and numerous combination strategies are being evaluated in clinical trials.A third subgroup was characterized by NF1 mutations and had the highest mutational load, with 39 mutations/Mb, more than twice that of the other subtypes. NF1 mutations tended to be negatively correlated with BRAF but not RAS mutations. Patients in this group were significantly older.The final group was classed as triple wild-type, lacking hotspot mutations in any of the three key drivers found in the other subgroups. It was heterogeneous in nature, displaying several other types of mutation at a low frequency, including GNAQ, GNA11, KIT, CTNNB1, and EZH2.

Other Genomic Alterations

Timing of Mutations

This subgroup had fewer TP53 mutations, and a much less distinct UV signature, with only 30% of tumors showing genomic indications of UVR-induced DNA damage. There were frequent focal amplifications in genes encoding known oncogenes, particularly tyrosine kinase receptors such as KIT.Other types of genomic alterations besides mutations have been examined in melanoma. Oncogenic gene fusions appear to be relatively rare and mostly involve members of the MAPK pathway. They may represent an alternative method of MAPK pathway activation since, for the most part, they are not found in conjunction with BRAF mutations or other mutations that activate the MAPK. The TCGA study identified several novel recurrent fusion events involving the AKT3, MITF, and HMGA2 genes; fusions and other complex structural genomic variations were more frequently observed in the triple wild-type subgroup. It remains to be seen whether these events are oncogenic drivers or passenger events.Numerous genome sequencing studies have shed light on the question of which genes are altered, but researchers now are also trying to understand when these genes are altered—which genes are important early on in melanoma initiation and which control later progression to more advanced stages of disease.

A recent study sequenced almost 300 genes in 150 samples from 37 primary melanomas and adjacent precursor lesions. In addition to further ing our understanding of the genomic evolution of melanoma, the study provided evidence for an intermediate type of melanoma lesion that falls between benign nevi and unequivocal melanoma. Histologically, there are some tumors that are not readily classified as either benign or malignant and this novel grouping could help to remedy that situation.

Other Melanoma Subtypes

BRAF V600E mutations were found in benign lesions, reinforcing the suggestion from other studies that these may be an early event, while NRAS and other driver mutations were more heavily enriched in the intermediate lesions. More advanced melanomas subsequently accumulate additional mutations in genes such as ARID1A, TERT, and CDKN2A. While the mutational load increased as tumors progressed, there was a strong UVR signature across all stages.The extremely high number of mutations found in cutaneous melanomas presents a significant challenge, making it difficult to distinguish true driver mutations from the background noise. Thus far, the relatively small sample sizes examined in genome sequencing studies mean that there are likely to be many more driver genes that have yet to be identified.

Another significant challenge is the underrepresentation of different kinds of melanoma in these sequencing studies. Melanomas arising on chronically sun-damaged skin, such as cutaneous melanoma, differ from those arising on skin without chronic sun damage, in terms of the site of the primary tumor, the age of onset, how quickly they develop, and their clinical presentation. The limited sequencing studies that have been performed so far in these types of melanoma, which include acral, mucosal, and uveal melanomas, have provided tantalizing evidence that these types of tumors are also distinct in their genomic background. As expected, they lack the UVR signature seen in a majority of cutaneous melanomas, but they also have significantly lower mutational loads. The number of somatic mutations in acral melanoma, for example, is 1.02—3.68 mutations/Mb, according to a recent study.

The types of mutations observed in these melanoma subtypes are also distinct. Uveal melanomas lack both BRAF and NRAS mutations and instead are associated with recurrent mutations in the guanine nucleotide- binding protein subunit alpha-11 (GNA11) and its paralog, GNAQ.

The proteins encoded by these genes function as a subunit of Gq proteins, which activate phospholipase C and coordinate a number of different signaling pathways, including MAPK. Uveal melanomas also have a high frequency of mutations in the BRCA-associated 1 (BAP1) gene. Acral and mucosal melanoma appear quite similar on a genomic level.

The two subtypes both exhibit mutations in BRAF and NRAS, though at a considerably lower frequency than cutaneous melanomas, and also share mutations in the KIT gene, in both the juxtamembrane and kinase domains, which may make them sensitive to KIT inhibition.

Jane de Lartigue, PhD, is a freelance medical writer and editor based in New Haven, Connecticut.

Key Research

  1. Berger MF, Hodis E, Heffernan TP, et al. Melanoma genome sequencing reveals frequent PREX2 mutations. Nature. 2012;485(7399):502-506.
  2. Furney SJ, Turajlic S, Stamp G, et al. The mutational burden of acral melanoma revealed by whole-genome sequencing. Pigment Cell Melanoma Res. 2014;27(5):835-838.
  3. Hill VK, Gartner JJ, Samuels Y, Goldstein AM. The genetics of melanoma: recent advances. Annu Rev Genomics Hum Genet. 2013;14:257-279.
  4. Hodis E, Watson IR, Kryukov GV, et al. A landscape of driver mutations in melanoma. Cell 2012;150(2):251-263.
  5. Shain AH, Yeh I, Kovalyshyn I, et al. The genetic evolution of melanoma from precursor lesions. N Engl J Med. 2015;373(20):1926-1936.
  6. The Cancer Genome Atlas Network. Genomic classification of cutaneous melanoma. Cell. 2015;161(7):1681-1696.
  7. Zhang T, Dutton-Regester K, Brown KM, Hayward NK. The genomic landscape of cutaneous melanoma. Pigment Cell Melanoma Res. 2016;29(3):266-283.