Diagnosis and staging of melanoma require pathologic examination of a biopsy or surgical specimen.13,14 Quest offers histopathology testing, including gross/microscopic tissue examination and IHC examination for cell markers (Table 1). Additional testing includes molecular testing (single-gene testing and multigene panels) to identify gene variants for targeted therapy or clinical trial eligibility, and gene expression profiling to predict nodal metastasis and long-term outcomes. IHC examination for guiding immunotherapies and serum tests that may be ordered at additional charge by the histopathologist, as discussed below (Table 1).

Staging

The American Joint Committee on Cancer (AJCC) 8th Edition Staging System of Melanoma (2018) is the preferred staging system.2,13 Anatomic staging is based on primary tumor features (T), involvement of regional lymph nodes (N), and whether the cancer has spread to other tissues (metastasis [M]).2,13 In addition to patient age and sex, pathological tumor stage is the most important factor for assessing prognosis and predicting outcomes.15 Primary tumor features important for pathologic staging include the depth of invasion (ie, Breslow thickness), the presence of skin ulceration, and mitotic rate.16,17 Depth of melanoma is an important prognostic factor, and Breslow thickness is now the standard of care because it is more specific than other methods, and the 10-year survival rate decreases with increased depth of the lesion.16,18

Primary cutaneous melanoma is defined as any primary melanoma lesion, regardless of tumor thickness, in patients without clinical or histologic evidence of regional or distant metastatic disease (stages 0-IIC).18 Stage III and IV melanomas include cases with regional disease and distant metastasis.2,13,19,20

Immunohistochemistry

IHC is also important for diagnosis of melanoma and melanoma subtype (eg, p16, SOX10, S100, Melan-A, HMB45, PRAME), lymphovascular invasion (eg, endothelial markers D2-40, CD31), and PTEN expression.21

IHC is also used to assess for metastasis in nodal tissue from a sentinel lymph node biopsy (SLNB, see below) if tumor cells are not evident on hematoxylin and eosin (H&E) staining.22 Overall, 5% to 40% of tumors are up-staged from clinical stage I-II to pathologic stage III based on subclinical micrometastases in an SLNB identified by IHC.13

Chromosomal microarray

Some melanocytic lesions can be difficult to classify as benign or malignant based on histopathological examination and IHC results. The Chromosomal Microarray, Solid Tumor FFPE, Clarisure® Oligo-SNP assay (test code 38842) may help to identify if a lesion is benign or malignant when the results of other pathological examinations are equivocal by providing additional information such as copy number variants (CNV, ie, deletion, duplication, amplification), single-nucleotide polymorphisms (SNPs), and loss of heterozygosity.13 For example, an algorithmic approach that uses >3 CNV combined with histopathological classification has been proposed to help generate an approximate likelihood of malignancy.23

PTEN (phosphatase and tensin homologue) expression

PTEN is involved in melanomagenesis as a tumor suppressor gene.21,22 Germline PTEN variants and deletions/duplications (PTEN Sequencing and Deletion/Duplication; test code 92566) are associated with Cowden syndrome in adults and Bannayan-Riley-Ruvalcaba syndrome in children.24 They are also associated with an increased risk for breast, thyroid, endometrial, kidney, and colorectal cancer, as well as melanoma with autosomal dominant PTEN hamartoma tumor syndrome (PHTS) (includes Cowden syndrome).24

Loss of PTEN expression promotes melanoma cell growth, and melanomas with deficient PTEN expression are associated with higher disease stage and poor survival.21,25 Loss of PTEN expression is defined as <10% of tumor cells showing labeling for PTEN.21,25

While clinical decisions should not be based solely on PTEN expression, loss of PTEN expression may be associated with resistance to BRAF inhibitors.21,25 Reactivating PTEN protein function is a potential therapy for the treatment of PTEN-loss cancers.21,25

SLNB and gene-expression profiling

Identification of sentinel lymph node involvement (metastasis, micrometastasis) via biopsy is recommended by the NCCN for patients with tumors at stage T2 or higher.13 The NCCN also recommends discussing and considering an SLNB for patients with T1a lesions with Breslow depth >0.5 mm and other adverse features (age ≤42 years, head/neck location, lymphovascular invasion, and/ or mitotic index ≥2/mm2), as the probability of a positive SLNB is 5% to 10%, with additive increased risk when multiple adverse features are present.13 However, about 80% of patients who undergo an SLNB have a negative result (no metastasis), and 11% have a complication of the biopsy such as infection, seroma, hematoma, lymphedema, or nerve injury.26

The MelaNodal Predict test (test code 13222) predicts SLN metastasis in patients who are eligible for an SLNB. The assay is performed using primary tumor tissue from the original diagnostic biopsy specimen and uses the CP-GEP (clinicopathologic-gene expression profile) algorithm developed by SkylineDx in collaboration with the Mayo Clinic.27 The algorithm combines patient age and Breslow depth, and a gene expression profile of 8 target genes (MLANA, GDF15, CXCL8, LOXL4, TGFBR1, ITGB3, PLAT, SERPINE2) to predict the risk of SLN metastasis.27 Patients with a low risk of SLN metastasis may choose to not have an SLNB.27

The CP-GEP algorithm was validated using 754 consecutive primary cutaneous melanoma specimens of US patients with T1 to T3 melanomas (17% prevalence of SLN positivity).27 The results showed the algorithm exhibited a negative predictive value of 96% and a 42% reduction in SLNBs.27 Subsequent studies of patients with T1 to T2 melanoma (8% to 16% prevalence of SLN positivity) reported negative predictive values of >90%, and rates of SLNB reduced by >35%.28–31

MelaNodal Predict test results are reported as “low risk” or “high risk” of SLN metastasis. Low risk indicates a low probability of SLN metastasis, and the patient may elect to not have an SLNB. 28–31 Based on a negative predictive value >90%, a study of patients with T1 to T4 melanoma (10% prevalence of SLN positivity) found that a low-risk result was associated with a 3% risk of SNL metastasis.30 On the other hand, a high-risk result indicates a much higher probability of SLN metastasis and that the patient should be considered for an SLNB.28–31 For example, a study of patients with T1 to T4 melanoma (10% prevalence of SLN positivity) reported a high-risk result was associated with a 24% risk of SLN metastasis.30

Notably, patients with a low-risk result have a low risk of developing recurrence, whereas those with a high-risk result have a higher risk of developing recurrence.32–35 However, the NCCN states that the prognostic utility of CP-GEP tests for SNLB is the subject of on-going studies.13 Thus, the MelaNodal Predict test is intended as an adjunct to the recommended SLNB eligibility to further stratify the risk of SLN metastasis and cannot replace the standard clinical and pathologic recommendations for diagnosis or assessing prognosis.

Genomic landscape

The mutation rate of human cancers has been reported to vary 1,000-fold across different malignancies.36 A study examining the mutation frequencies of various human cancers found that overall, melanoma had the highest mutation frequency of all cancers analyzed.37 The variability in the mutation frequency in melanoma is due to the presence or absence of a known carcinogen, ie, UV exposure. Considering all non-acral cutaneous melanoma, the most frequently mutated genes are BRAF, NRAS, NF1, and KIT (Table 2).2,13,24,38–40 When considering melanoma subtypes, mutant BRAF, CDKN2A, NRAS, and TP53 are most common in cutaneous melanoma; mutant BRAF, NRAS, NF1, and KIT in acral melanoma; and SF3B1 in mucosal melanoma.41 Cutaneous melanoma without BRAF, NRAS, or NF1 mutations are also seen (triple-wild type).2,13 Though variants in many other genes have been associated with melanoma, such as SNX32, STK19, IDH1, MITF, MTOR, TACC1 and others, few have been studied extensively.42,43

Table 2. Frequencies of Common Gene Variants in Non-Acral Cutaneous Melanoma [return to contents]

Variant gene (%)	Clinical correlations	Relative prognosis	Treatment options
BRAF (40-80)24	Younger age Intermittent sunlight exposure	Neutral	Combination dabrafenib and trametinib or immune checkpoint inhibitors Ongoing clinical trials
CDKN2A (13-30)39	Early age of onset Multiple primary lesions Familial inheritance Associated with pancreatic cancer	Poor	No specific targeted therapy
NRAS (28)24,38,40	Older age Non-sun-damaged skin	Poor	Clinical trials of MEK1/2 inhibitors, CDK4/6 inhibitors
NF1 (14)24,39	Older age Males Sun-exposed skin	Poor	No ongoing trials or specific treatment options
KIT (1-3)24,39,a	Chronic sun exposure	Poor	KIT inhibitors (eg, imatinib, sunitinib, nilotinib)b
Triple wild-type (15)24,c	Males	Poor	No ongoing trials or specific treatment options

Non-acral includes superficial spreading, nodular, and lentigo melanoma.
a	KIT variants are relatively uncommon; however, the frequency is approximately 22% in patients with triple wild-type melanoma (wild-type for BRAF, NRAS, NF1).13 KIT variants are most common in acral melanoma.39
b	KIT exon 11 and 13 variants exhibit a relatively high sensitivity to KIT inhibitors. KIT exon 17 variants and KIT amplifications have minimal or no sensitivity to KIT inhibitors.13
c	Triple wild-type is not a gene variant per se; it defines melanoma without BRAF, NRAS, or NF1 variants.13

 

Targeted therapies based on specific variants have emerged as an important treatment for patients with advanced melanoma.13,19,20 However, while many variants have been associated with melanoma, most have not been effectively targeted.13,19,20 While multigene testing may identify variants that are not actionable, they can identify patients eligible for clinical trials.12,13 Actionable and nonactionable variants are discussed below.

BRAF variants

BRAF is a serine threonine kinase that activates the mitogen-activated protein kinase (MAPK) pathway; BRAF variants are associated with unrestrained cell growth and proliferation.13 BRAF variants are most common in the 600th codon (V600); V600E is the most frequent V600 variant (80%), followed by V600K (15%) and V600R/M/D/G (5%).13,44 Intermittent sun exposure, younger age, and trunk location are associated with a higher frequency of BRAF variants.24 BRAF variants are seen in approximately 40% to 80% of cutaneous melanomas.40

Identification of BRAF variants is important for guiding treatment. Melanomas with BRAF V600 variants are sensitive to BRAF inhibitors, which should not be used in patients without BRAF V600 variants.45 BRAF V600 variants are also sensitive to mitogen-activated protein kinase kinase (MEK) inhibitors (MEK1 and/or MEK2 inhibitors).13 Clinical trials have shown that the combination of BRAF and MEK inhibitors (eg, dabrafenib/trametinib) are superior to either agent alone in patients with BRAF V600 variants.13,14 However, resistance to BRAF inhibitors typically develops, and clinical trials are investigating if the addition of immunomodulatory agents to dabrafenib/trametinib delays or prevents the development of resistance.13,24

National Comprehensive Cancer Network (NCCN) and European Society for Medical Oncology (ESMO) guidelines recommend BRAF varianttesting to inform treatment decisions for patients with high-risk stage III and IV cutaneous melanomas.2,13 Next-generation sequencing (NGS), single-gene assays, and small multigene panels can be used for molecular testing.13 IHC staining, typically used as a rapid screening test, can detect aberrations in proteins consistent with gene variants such as BRAF V600E; gene variants are then confirmed with molecular testing.13

The Cobas® 4800 BRAF V600 Mutation Test (test code 90956) detects the BRAF codon 600 variant V600E.46 The test is an FDA-approved companion diagnostic used to determine eligibility of patients with advanced melanoma for treatment with vemurafenib (targeting BRAF V600E variant) and cobimetinib (targeting MEK) in combination with vemurafenib.46,47

Quest offers NGS testing to identify these BRAF variants through single-gene BRAF analysis (test code 16767) and multigene NGS panels including the core (test code 93234) and expanded solid tumor panels (93233), which include full exonic coverage of the BRAF gene as part of next generation NGS comprehensive profiling (see Table 1 and “Other variants” below).41

CDKN2A variants

CDKN2A is a tumor suppressor gene, and germline and somatic variants are associated with melanoma.

Independent features associated with germline CDKN2A variants include multiple primary melanomas, high number of family members with melanoma, Breslow thickness >0.4 mm, and early age of melanoma onset.11 Persons with a germline CDKN2A variant have an increased risk for melanoma (28% to 76% risk to age 80) and pancreatic cancer.13 Multigene panel testing that includes CDKN2A is recommended for patients with invasive cutaneous melanoma who have a first-degree relative diagnosed with pancreatic cancer.13 Quest offers both comprehensive (test code 38600) and guideline-based (test code 38611) multigene hereditary cancer panels that detect germline CDKN2A variants (Table 1).

Approximately 13% of melanomas have somatic CDKN2A mutations, around 30% have CKDN2A deletions,36 and some harbor CDKN2A duplications.48 CDKN2A Sequencing and Deletion/Duplication (test code 93939) can identify CDKN2A variants as well as deletions/duplications.49 Somatic testing for CDKN2A variants is included in the solid tumor core and expanded panels (Table 1). There is currently no specific targeted therapy for melanoma with CDKN2A variants.

KIT variants

KIT is a proto-oncogene receptor tyrosine kinase present on cell membranes. Binding to stem cell factor activates the KIT protein and subsequently signaling pathways associated with cell growth, proliferation, and migration.13,24 KIT variants occur in 2% to 8% of melanomas; they are most common in acral melanoma and in melanoma on skin with chronic sun exposure.13,39

No effective targeted therapy has been developed for melanoma with KIT variants.24 Around 30% to 50% of melanomas with KIT exon 11 variants respond to tyrosine kinase inhibitors; however, resistance to treatment typically occurs within 1 year.18,50 Patients with KIT variants (c-KIT Mutation Analysis, Cell-based; test code 19961) may be eligible to participate in clinical trials.48,51

NF1 variants

NF1 is a tumor suppressor gene that encodes the protein neurofibromin 1. NF1 variants are present in about 12% of cutaneous melanomas and typically occur in sun-exposed skin of older males.40 These melanomas are aggressive and associated with poor survival; no targeted treatments have been developed.36,52 NF1 variants can be detected by NGS and may identify patients eligible for clinical trials.41

NRAS variants

NRAS is a GTPase that activates MAPK signaling and other signaling pathways, leading to cell growth and proliferation.13 NRAS variants are present in approximately 15% to 25% of cutaneous melanomas.40 They are more common in non-sun exposed skin but may occur in skin with chronic and intermittent sun exposure as well as acral and mucosal melanomas.36,40 NRAS variants are associated with aggressive disease and poor prognosis.36 No targeted treatments are available for melanoma with NRAS variants; however, findings from clinical trial suggested that binimetinib (a MEK1/2 inhibitor) provided a survival benefit as compared to dacarbazine chemotherapy.36 Detection of NRAS variants by NGS (see “Other variants” below) can help identify patients eligible for clinical trials.41

Triple-wild type

Melanoma without BRAF, NRAS, and NF1 variants is defined as triple-wild type and typically occurs in males 60 to 70 years of age.24 Triple-wild type melanomas are heterogeneous and may contain GNA11, GNAQ, SF3B1, and KIT variants.24 No targeted agents have been developed, and the prognosis is poor. Triple-wild type melanoma is diagnosed by exclusion of BRAF, NRAS, and NF1 variants by NGS (see “Other variants” below), which may identify patients eligible for clinical trials.41

Other variants

Certain variants in a number of other genes, including BAP1, CDK4, MITF, POT1, and TP53, have been associated with cutaneous melanomas and other malignancies (eg, CDK4 variants are associated with cutaneous melanoma and pancreatic cancer).11,12,53 Variants in ACD, ATM, BAP1, BRCA1/2, CDK4, MITF, POT1, PTEN, TERF2IP, and TERT may indicate a heritable predisposition to melanoma, although CDKN2A is most common.11,12,53 In some cases, gene variants associated with certain cancer syndromes confer an increased, but poorly defined, risk of melanoma.53 For example, variants in TP53 (the gene that encodes tumor protein p53) are associated with pancreatic, breast, prostate, colon, and ovarian cancer and are also associated with increased risk of melanoma.53 The association of TP53 variants with a number of different malignancies may reflect the important role p53 plays as a tumor suppressor protein (ie, maintaining genetic integrity and regulating the expression of target genes involved in DNA repair, apoptosis, the cell cycle, and differentiation).54 Overexpression of p53 is associated with tumorigenesis and can be used as a surrogate marker for assessing whether tumor cells contain p53 variants.54

In addition to the variants discussed above, Quest offers testing for other variants as part of large NGS panels for solid tumors spanning either 49 genes (test code 93234) or 522 genes and the TERT promotor (test code 93233). In the larger panel, 55 common acceptor genes are also sequenced from RNA to detect fusions (Table 1 and Appendix) and splice variants. Reports from variant panel testing include the clinical significance, prognosis, and predicted response to therapy for the variant. The variants are classified into 4 tiers based on the strength of the current evidence for their clinical significance (Table 3).55 Some variants are detected only within targeted regions of the selected genes but not in the promoter and intronic variant regions (except for the TERT promoter, fusions, and splice site variants).

Table 3. Variant Classification Tiers [return to contents]

Tier50	Strength of significance	Type of evidence
1	Strong clinical significance	Actionability supported by large studies with expert consensus Included in professional guidelines to guide clinical decision-making for the given tumor type
2	Potential clinical significance	Actionability supported by multiple small or preclinical studies or case reports, with or without expert consensus Included in professional guidelines to guide therapy selection for a different tumor type Fulfills criteria for clinical trial inclusion
3	Uncertain clinical significance	No known actionability or significance in current literature Not found in the general population
4a	Benign or likely benign	No known actionability or significance in current literature Found in the general population

a Tier 4 variants are not reported.

 

Large NGS panels can also be used to simultaneously detect fusions for which there is limited evidence of the effectiveness of certain therapies (eg, NTRK fusions) and to evaluate tumor mutational burden (TMB) and microsatellite instability (MSI).13 These are gene-agnostic measures of hypermutation and defective DNA repair mechanisms within tumor cells that can also be used to assess eligibility for some therapies. See the “Tumor mutation burden and microsatellite instability” section for more information.

Immune checkpoint immunohistochemistry

Immune checkpoints refer to the interactions between receptors on activated T cells and ligands that stop normal cells from being targeted for destruction (eg, during an infection). Some tumor cells, including melanoma, take advantage of the protective role of ligands by expressing them at high levels to evade the immune system. Immune checkpoint inhibitors block ligand-T-cell receptor binding, enabling the immune system to attack cells expressing these ligands. Blocking the immune checkpoint has been shown to be effective in the treatment of advanced melanoma and other malignancies.56–58

IHC is used to measure expression of immune checkpoint components to determine tumor sensitivity to immune checkpoint inhibitors. For some types of tumors IHC results can be used to determine potential patient eligibility for checkpoint inhibitor treatment, although this is not currently the case for melanoma.21,25

Available therapeutic monoclonal antibodies for melanoma target 1 of 3 checkpoints: cytotoxic T lymphocyte-associated antigen 4 (CTLA-4) and its receptor; programmed death ligand 1/programmed cell death protein 1 (PD-L1/PD-1); and lymphocyte-activation gene 3 (LAG-3).59 IHC testing plays a role in assessing the likelihood of response to PD-L1/PD-1 inhibitors but not LAG-3 inhibitors nor CTLA-4 inhibitors.

Tumor expression of PD-L1 is inadequately predictive of response to immune-checkpoint blockade to determine potential eligibility for checkpoint inhibitor treatment in melanoma.13 NCCN guidelines do not recommend that clinical and treatment decisions be based on IHC analyses of PD-L1 expression.13

For melanoma, tumor expression of PD-L1 is a suggestive biomarker—tumors expressing little or no PD-L1 are less likely to respond to PD-1 pathway blockade.60 Quest offers PD-L1, IHC with Interpretation (test code 94007) as a generic laboratory developed test (LDT) and a non-FDA approved IHC-based test (clone SP263) to assess PD-L1 protein expression. Results are based on %TC and %IC, which are the percentages of the tumor area covered by PD-L1–staining tumor cells or tumor-infiltrating immune cells (ICs), respectively. Below are examples of how IHC testing has been employed in clinical trials supporting 3 PD-L1 inhibitors that are used in patients with advanced (unresectable/metastatic) melanoma.

Atezolizumab

Atezolizumab is a therapeutic monoclonal antibody targeting PD-L1 that is used for the treatment of patients with unresectable or metastatic BRAF V600 positive melanoma (stage IIIC-IV) in combination with cobimetinib and vemurafenib (MEK and BRAF kinase inhibitors, respectively).61 Patients treated with atezolizumab, vemurafenib, and cobimetinib (atezolizumab group) had significantly greater progression-free survival (PFS;15.1 months) than those treated with atezolizumab placebo, vemurafenib, and cobimetinib (10.6 months; P=.025).61 Notably, the frequency of PD-L1 positivity, based on ≥1% tumor infiltrating ICs was similar between both groups.61 A study of patients with BRAF V600 positive advanced or metastatic melanoma treated with atezolizumab alone showed an objective response rate (ORR) of 35%, which was similar between PD-L1 positive and PD-L1 negative patients; PD-L1 (clone SP142) positivity was defined as tumor infiltrating IC1/2/3 ≥1%.62

Nivolumab

Nivolumab is a therapeutic monoclonal antibody targeting PD-1 that is used as monotherapy or in combination with ipilimumab, an antibody that targets the CTLA-4 checkpoint, or relatlimab, an antibody that targets LAG-3.57,63,64

In patients with unresectable stage III or IV melanoma, treatment with nivolumab plus ipilimumab resulted in a median PFS of 11.5 months, compared to 2.9 months for ipilimumab alone and 6.9 months for nivolumab alone.65 PD-L1 positivity (clone 28-9) was defined as at least 5% of tumor cells showing cell surface PD-L1 staining of any intensity in a section containing at least 100 tumor cells that could be evaluated65:

• In PD-L1-positive patients, median PFS was 14.0 months in the nivolumab plus ipilimumab group and nivolumab-alone group and 3.9 months in the ipilimumab-alone group.
• In PD-L1-negative patients, PFS was 11.2 months in patients treated with nivolumab plus ipilimumab, compared to 5.3 months for those treated with nivolumab alone and 2.8 months for those treated with ipilimumab alone.

The combination of nivolumab and relatlimab has been approved for the treatment of unresectable or metastatic melanoma.64 The RELATIVITY-047 trial compared nivolumab and the combination of nivolumab and relatlimab in patients with previously untreated metastatic or unresectable melanoma.64 The median PFS with relatlimab-nivolumab was 10.1 months compared with 4.6 months with nivolumab alone. Patients were stratified by PD-L1 (clone 28-9) and LAG-3 expression (<1% and ≥1% for both)64:

• PD-L1 expression ≥1% corresponded to similar PFS in both treatment groups (15 to 16 months).
• PD-L1 expression <1% corresponded to a longer PFS of 6.4 months for the relatlimab-nivolumab combination compared to 2.9 months for nivolumab alone.
• Benefit was greatest with LAG-3 expression ≥1% for both treatment groups, but the relatlimab-nivolumab combination had significantly longer PFS than nivolumab.

Another study of patients with disease refractory to PD-1/PD-L1 treatment reported objective response rates of 9% to 12% for the relatlimab-nivolumab combination.63 PD-L1 and LAG-3 expression ≥1% in tumors appeared to result in an enriched response to therapy, but responses were observed regardless of expression.

NCCN has suggested that high PD-L1 expression (>5%) may be a marker for equivalent outcomes with nivolumab monotherapy versus combination ipilimumab and nivolumab in patients with unresectable or metastatic melanoma.13 Low PD-L1 expression may be a marker for worse outcomes with nivolumab monotherapy compared to ipilimumab/nivolumab combination therapy.13 NCCN considers the combination of nivolumab and relatlimab as treatment for recurrent disease and as second-line treatment for unresectable or metastatic disease.13

Pembrolizumab

Pembrolizumab is a therapeutic monoclonal antibody targeting PD-1 that is used for the treatment of metastatic melanoma with a reported response rate of 38% to 52%, a 12-month OS rate of 74%, and a 5-year survival rate of 41%.22,66 Monotreatment with pembrolizumab in patients with stage IIIA-IIIC disease resulted in a relapse-free survival (RFS) rate of 64%, compared to 44% in the placebo group.33 The improvement in RFS was not related to BRAF mutation status or PD-L1 expression33; PD-L1 positivity was defined as a melanoma score ≥2 (≥1% to <10% membrane staining) in tumor and tumor-associated ICs (clone 22C3).67

Tumor mutation burden and microsatellite instability

TMB is defined as the number of mutations found per Mb. In melanoma as well as other cancers, TMB has been shown to correlate with response to immune checkpoint inhibitors (single anti-PD-L1/PD-1 blocking agents and ipilimumab/nivolumab combination therapy).13,25,68

MSI is genetic instability in short nucleotide repeats (microsatellites) as a result of abnormal DNA mismatch repair69 (not to be confused with microsatellitosis, which represents microscopically identified lymphatic metastasis13). DNA mismatch errors occur spontaneously during DNA replication; however, cells that are mismatch repair deficient (dMMR) accumulate mismatch repair errors.69 Many malignancies, including melanoma, can have high MSI/dMMR.69 MSI has traditionally been evaluated using PCR methods; however, this allowed interrogation of only 5 to 7 loci.70

NGS panels can be used to simultaneously evaluate TMB and MSI (and the interrogation of thousands of microsatellite loci),70 and high TMB (TMB-H) and MSI (MSI-H) are predictive biomarkers of response to treatment with immune checkpoint inhibitors.13,21 However, NCCN guidelines indicate that the use of TMB to guide treatment decisions for patients with advanced melanoma is currently investigational.13 Furthermore, although MSI is a frequent event, there is no consensus on the definition of MSI-H in melanoma; consequently, the clinical utility of MSI in melanoma needs to be more fully explored.21

Serum tests

Cell-free circulating DNA (plasma) testing

Testing blood plasma for circulating tumor cells (CTCs) and circulating tumor DNA (ctDNA, a subset of cell-free DNA [cfDNA]) is used in the management of a number of malignancies, including melanoma.39

In patients with advanced melanoma, the BRAF V600E variant may be detected via ctDNA. High ctDNA levels prior to beginning treatment are associated with a lower response rate to BRAF kinase inhibitors and shorter PFS.39 In addition, an increase of ctDNA during treatment has been associated with acquired resistance and disease progression.39 A study also found that undetectable ctDNA at baseline or within 8 weeks of beginning therapy was an independent predictor of response and survival in PD-1 antibody–treated patients with melanoma.71 A pilot study reported that PD-L1 expression on CTCs may be predictive of response to pembrolizumab and prolonged PFS.72 Another study reported that ctDNA is an independent prognostic biomarker of survival in patients with metastatic melanoma with BRAF or NRAS variants.73

However, NCCN cautions that given the possibility of a false negative, a negative ctDNA result should prompt tissue testing.13

Lactate dehydrogenase

Lactate dehydrogenase (LDH) catalyzes the conversion of pyruvate into lactate in anoxic conditions, such as in the oxygen-deficient tumor microenvironment. The AJCC 8th edition melanoma staging system considers elevated LDH as an adverse prognostic indicator for patients with stage IV melanoma and indicates a lower chance of survival.36 LDH level may also be useful for predicting response to therapy.36 Elevated levels of LDH have been associated with a poor response to PD-L1 blockade.74 Patients with high LDH levels may also achieve the greatest benefit from combination immunotherapy.57