BRAF mutation testing has become an essential tool in the diagnostic workup and management of patients with advanced melanoma.
Keith T. Flaherty, MD
About half of all melanomas contain somatic mutations that result in constitutive activation of the BRAF kinase, leading to hyperactive signaling of the growth-promoting mitogen-activated protein kinase (MAPK) pathway. Systemic therapies with selective small-molecule inhibitors of the BRAF kinase, both in monotherapy and in combination with inhibitors of the MEK kinase, can extend survival and reduce recurrence risk in patients withBRAF-mutant tumors.1BRAFmutation testing has therefore become an essential tool in the diagnostic workup and management of patients with advanced melanoma.
70% to 90%
cobas test BRAFV600D & THxID BRAF Kit
7% to 20%
THxID BRAF Kit
1% to 3%
3% to 7%
According to Keith T. Flaherty, MD, with the Massachusetts General Hospital Cancer Center in Boston, “At the moment, the only molecular feature in melanoma that is critical to treatment decision making is the presence or absence of a V600BRAFmutation.” Activating mutations ofBRAFare present in 40% to 60% of all melanoma cases, with more than 90% of mutations found at codon 600 in exon 15. The activating V600E amino acid substitution is by far the most common V600 mutation and accounts for 74% to 95% of these mutations, followed by V600K in 5% to 30%, V600R in 5% to 7%, and V600M in about 4%.2,3More rare mutations include V600D and mutations in other codons with activating, impairing, or unknown effects on kinase activity.3,4Clinical evidence for the activity selective BRAF inhibitors derives from trials that were largely restricted to patients with V600E- and V600K-mutant tumors, as concurrently developedBRAFmutation testing assays were optimized for the detection of these two most common mutations.
Prospective testing of the BRAF inhibitors vemurafenib and dabrafenib as single agents in phase III trials was conducted among patients with V600E-positive metastatic melanoma, demonstrating response rates in the range of 50%, clinical benefit rates close to 90%, and prolongation of progression-free survival (PFS) and overall survival (OS) compared with dacarbazine.5,6
Similarly, phase III studies with combination regimens of dabrafenib and trametinib or vemurafenib with cobimetinib revealed significant PFS and OS benefits with dual BRAF/MEK inhibition over single agent BRAF inhibitors. Patients with V600E- or V600K-mutant melanoma represented the majority of the study population.7-9
Indications for these agents based on phase III trial data therefore incorporate the presence of specific mutations: Combination dabrafenib/trametinib and vemurafenib/cobimetinib therapies and single agent trametinib are indicated for the treatment of patients with V600E or V600K mutations, whereas approved use of single-agent vemurafenib or dabrafenib is limited to the treatment of patients with V600E mutant melanoma.
This rationale is supported by findings thatBRAFmutation type can affect response to treatment with selective BRAF inhibitors. Phase II study findings have shown that although dabrafenib is active in patients with V600K-mutant tumors, ORR and PFS in this subgroup were lower than in patients with V600E-mutant tumors.10,11However, evidence also exists that patients with less common BRAF mutations not represented in phase III trials can benefit from BRAF inhibitor therapy.2,12,13
In a cohort study including patients withBRAFV600R mutations (n = 9), objective responses to monotherapy with selective BRAF inhibitors (dabrafenib or vemurafenib) were seen in 5 of 6 patients evaluable for response assessment. The V600R mutation is common in males and those with ulcerated primary melanomas. Similar to V600K melanoma, V600R is more frequently observed in older individuals.2
According to study author Oliver Klein, MD, with Westmead Hospital, Sydney, Australia, and co-authors, these partial responses occurred rapidly and included “a reduction in the size of lung metastases, liver metastases, stabilization of brain metastases, and a marked reduction in subcutaneous metastases to the scalp. The rapid clinical responses of patients with V600RBRAFmutation-positive melanoma mirror those observed in patients with V600E and V600K mutations, indicating that selective BRAF inhibitors are active against melanoma with this genotype.”
Other reports have shown responses to selective BRAF inhibitors in patients with V600R alterations, including a case report of a patient who responded to vemurafenib treatment12and a phase II study of patients with discrepant genotyping results forBRAFV600E and V600R, who had a confirmed partial response to trametinib with a PFS of at least 57 weeks.13
Information on tumorBRAFmutation status in patients with melanoma is essential for the selection of first-line therapy. Adequate genetic testing relies on a sequence of processes that include selection of tissue source, biopsy approach, specimen preparation and evaluation of tumor content, and choice of test assay.
According to current recommendations, routineBRAFtesting of the primary cutaneous melanoma should not be performed in patients without metastatic disease. In patients with metastatic disease, biopsy of the metastatic lesion, if feasible, is considered the preferred method to obtain sufficient tissue for genetic testing, if systemic therapy is planned, but archival material can also be used.1
According to Flaherty, these recommendations are based on observations thatBRAFmutations arise early during melanocytic proliferation and that up to 80% of benign nevi also harbor V600 mutations, rendering mutation testing of primary tumor tissue non-informative.
“The best available evidence indicates thatBRAFmutations are stably present in all tumor cells in lymph node and visceral metastases, whereas there can be heterogeneity in the primary melanoma. [This supports] the recommended practice of performingBRAFmutation testing in metastatic tumor sites, with previously resected regional lymph nodes (for patients who had stage III disease prior to developing stage IV disease) being a reliable source of tumor material for this testing,” he said.
Tumor molecular profiling is usually conducted on DNA extracted from formalin-fixed paraffin-embedded (FFPE) tissue specimens after pathological examination has confirmed the presence of tumor cells and determined tumor content. Factors that may affect testing results include tissue integrity, presence of high levels of pigmentation, or lymphocyte infiltration.14,15
According to Monika Jurkowska, PhD, with Genomed Health Center in Warsaw, Poland, and co-authors, key steps for reliable testing involves the selection of “block surface with sufficient tumor cellularity (>5-20% depending of the analytical sensitivity of the applied method and feasibility of subsequent dissection) but also without extensive necrosis, blood or melanin pigments content.”16
Validated approaches for the detection ofBRAFmutations in FFPE tumor material tissue include polymerase chain reaction (PCR)-based assays, sequencing technologies (Sanger, pyrosequencing, next generation sequencing [NGS]) or immunohistochemistry, among others. Methods differ in sensitivity, specificity, and cost. Depending on scope, detection methods can be broadly grouped into targeted assays that only identify a genetic alteration of interest, such as allele-specific PCR tests, and screening assays that can simultaneously identify all mutations in a larger genomic region, such as bidirectional direct Sanger sequencing.
Direct sequencing is highly specific but has a relatively low sensitivity (80%-93.4%), requiring a relatively high proportion of mutant tumor cells within the sample.17High Resolution Melt analysis (HRM) and Sanger sequencing technologies have proven highly specific and relatively sensitive (100% specific and up to a 6.6% allele frequency sensitivity), whereas pyrosequencing is sensitive (5% allele frequency) but less specific (90%).18
PCR-based tests have a higher sensitivity than Sanger sequencing (97.5%) and require only small amounts of DNA. The FDA approved Cobas 4800 BRAF V600 and THxID BRAF tests were developed concurrently with vemurafenib and dabrafenib, respectively, and were optimized to detect the two most common V600 mutations, V600E and V600K, with differences in specificity.
The Cobas 4800 BRAF V600 mutation test received approval along with vemurafenib as a companion diagnostic.15This targeted assay combines allele-specific real-time PCR and TaqMelt technology to determineBRAFV600 mutation status in DNA isolated from FFPE tumor tissue and has an approximate sensitivity of 5% mutation-bearing cells in a mixed sample.19The assay is highly specific for the V600E mutation (98%) but also has cross-reactivity with V600K (approximately 50%), and with V600D and V600E2 mutations.15,16,19The test only identifies the presence of a mutation without differentiation between mutations and does not reliably detect V600 mutations other than V600E.20
The THxID BRAF test was developed to identify melanoma patients withBRAFV600E or V600K mutations for treatment with dabrafenib and trametinib.14This PCR-based assay incorporates an internal control to validate adequate DNA integrity and relies on primers that are specific for each of the two mutations. The assay is largely non-cross reactive with other V600 mutants and identifies V600E and V600K mutations with high concordance with high-resolution melt (HRM) analysis and Sanger sequencing results (95.9%).14,21
Based on studies comparing PCR-based assays with sequencing approaches, the exclusive use of PCR-based assays for the detection ofBRAFmutations will fail to identify patients with more rare mutations, such as V600R and D, who may also benefit from targeted therapy.
Comparison of PCR/Sanger sequencing assays with the Cobas 4800 BRAF V600 test in archival material (236 FFPE samples, cutaneous melanoma lymph node metastases) found that both methods detectedBRAFmutations at a similar rate (60.9% vs. 61.0%), with a concordance of 95.2%.16Sanger sequencing failed to produce results in 2.5% of samples but was highly specific and reproducible in identifying V600E, K, and D mutations, whereas the Cobas assay produced results for all samples but false negative results in 2.5% of samples that contained V600E, K, and D mutations.16
Similarly, a study comparing differentBRAFmutation detection methods in 295 melanoma FFPE samples found that relative to Sanger sequencing, the Cobas 4800 BRAF V600 test had only 80.5% sensitivity (95% CI, 72.4%-86.6%; 118 vs 96 of 275 samples).22False negative samples per Cobas assay included tumors in which sequencing identified V600E, K, and R mutations.22
Kevin Qu, with Quest Diagnostics Nichols Institute, and coauthors explained that the presence of more than one mutant nucleotide (dinucleotide) could affect test outcomes: “The cobas test is a single-oligonucleotide probe-based test that has greater analytical sensitivity than Sanger sequencing. However, as our findings suggest, its specificity for a single-nucleotide mutation means that it may miss relevant dinucleotide mutations in some cases. Sanger sequencing can detect single-nucleotide and dinucleotide mutations in the same region, which accounts for its higher detection rate in this study.”
To address this, an ideal testing strategy may call for samples that test negative according to PCR assays to be additionally validated by a second sequencing approach that reliably identifies non-V600E mutations, or confirm their absence. This may represent a practical and comprehensive strategy to identify patients withBRAFmutant melanoma. Kevin Qu and coauthors conclude: “Use of Sanger sequencing as a second-line test for samples that are negative on the cobas assay could be a rational approach to maximizing the potential benefit of vemurafenib.”
Similarly, Jurkowska and coauthors argued that, “By definition, Cobas cannot detect activatingBRAFmutation located outside codon V600. For those variants, detection based on sequence analysis (Sanger/ NGS seems to be [the] method of choice at the current state.” According to the authors, once sufficient knowledge about the significance of otherBRAFmutations for melanoma treatment decisions has accumulated, efforts will also need to focus on the “optimal mode of detection, since targeted genotyping would not cover all possibilities.”16