You are here
Lymph node metastases of melanoma: challenges for BRAF mutation detection
Detection of B-Raf proto-oncogene, serine/threonine kinase (BRAF) mutations is required to predict response toBRAFor mitogen-activated protein kinase kinase 1 and 2 inhibitors in metastatic melanoma. Lymph node (LN) specimens carrying melanoma cells intermingled with abundant lymphocytes often contain low tumor cellularity. This study is aimed to examine challenges in the clinical detection ofBRAFmutations in LN specimens with metastatic melanoma and to illustrate characteristic features of p.V600E and non-p.V600E mutations. In this retrospective study for quality assessment of the pyrosequencing assay, we compared characteristics of 53 LN and 135 non-LN formalin-fixed, paraffin-embedded specimens with metastatic melanoma submitted forBRAFmutation detection over a 40-month period. LN specimens showed a significantly higher incidence of p.V600E mutations than non-LN specimens (49% versus 22%,P< .01) but a significantly lower tumor cellularity, particularly in the case of subcapsular or infiltrative metastases. Mutant allele-specific imbalance of the p.V600E mutation was predominantly present in specimens with distant organ metastases (79% versus 27% in LN metastases versus 13% in primary cutaneous tumors or adjacent soft tissue,P< .001). p.V600K was detected in 23% of men older than 60 years old, compared with 6% in women older than 60 years old and 2% in both men and women younger than 60 years old (P< .001). LN specimens with low tumor cellularity due to numerous adjacent lymphocytes may pose a challenge to clinical detection ofBRAFmutations of melanoma. The higher incidence of p.V600E mutations in LNs may prompt further studies to elucidate if the p.V600E mutation in primary tumors is associated with a higher risk of LN metastasis.
Keywords: Melanoma, BRAF, Lymph node, Tumor cellularity, Mutant allele-specific imbalance.
Mutations of the B-Raf proto-oncogene, serine/threonine kinase (BRAF) gene may lead to constitutive activation of the downstream mitogen-activated protein kinase pathway in a variety of human neoplasms  . Approximately 40% to 60% of the cutaneous melanomas carryBRAFmutations and . Although p.V600E (c.1799T>A) is the most commonBRAFmutation in melanoma, p.V600K (c.1798_1799delinsAA) is more frequent than previously reported, particularly in older patients, , and . Clinically, vemurafenib therapy has shown improved survival in metastatic melanomas for patients with bothBRAFp.V600E mutation and non-p.V600E mutations and . Dabrafenib and trametinib have also been approved for use as monotherapy or in combination inBRAF-mutant metastatic melanoma and . Therefore, it is critical to correctly demonstrateBRAFmutation status to initiate targeted therapy and avoid unnecessary treatment.
Surgical pathologists designate areas for tissue dissection and ensure sufficient tumor cellularity for analytic sensitivity of requested assays. Specimens with metastases within lymph nodes (LNs) may pose a challenge to pathologists in selecting areas for manual macrodissection, particularly areas with microscopic subcapsular metastases or infiltrative metastases that do not form a neoplastic nodule. In specimens with numerous lymphocytes, tumor cellularity may be overestimated because the smaller lymphocytes are easily underestimated, yet contain approximately the same amount of DNA as the larger, more visually impressive cancer cells. This nonneoplastic contamination may jeopardize the detection of mutations in admixed cancer cells.
A previous retrospective review of 463 colorectal cancer specimens submitted to our molecular diagnostics laboratory for detection of Kirsten rat sarcoma viral oncogene homolog (KRAS) mutations included 10 LN specimens  . One of the LNs with less than 20% estimated tumor cellularity showed a focal subcapsular metastasis and also infiltrative parenchymal cancer cells. Pyrosequencing did not detect anyKRASmutations, when DNA was isolated from the whole LN but showed indeterminate pyrograms, when DNA was subsequently isolated from 2 areas with enriched tumor cells. Retrospective analysis by next-generation sequencing (NGS) confirmed a 4% p.G12D mutation.
In contrast to colorectal cancer specimens, the specimens submitted for molecular analysis of melanoma are quite frequently LNs. In this study, 28% of the melanoma specimens submitted for detection ofBRAFmutations were in LN tissue. The difficulty of detecting mutations in the colorectal cancer LN metastasis led us to postulate that abundant lymphocytes in melanoma specimens may also be a common problem. This study characterizes the challenges posed by melanoma LN specimens submitted for clinical molecular diagnosis, with the goal of raising quality management considerations.
2. Materials and methods
Between March 2010 and July 2013, 194 formalin-fixed, paraffin-embedded (FFPE) tissues with a diagnosis of primary or metastatic malignant melanoma were submitted for diagnosis ofBRAFmutations by pyrosequencing in the Johns Hopkins Molecular Diagnostics Laboratory. Pyrosequencing failed in 1 of the 194 specimens. Four mucosal melanomas and 1 uveal melanoma were excluded from analysis. The remaining 188 melanoma tissues were obtained from 120 male patients and 67 female patients. Information on age and sex was unknown for 1 specimen. The age of patients ranged from 7 to 88 years with a median of 56 years. There were 146 cutaneous melanomas (49 primary or locally recurrent tumors and 97 metastatic tumors) and 42 metastatic melanomas with an unknown location of the primary tumor. Cutaneous melanoma was located in the head and neck region in 40 specimens and outside the head and neck region in 100 specimens. The location was unspecified in 6 specimens. There were 126 resection specimens (including LN dissection specimens), 54 biopsy specimens, and 8 cell block specimens from fine needle aspiration. There were 53 LN specimens and 143 non-LN specimens. Among the 53 LN specimens, 3 were cell blocks from fine needle aspiration, 7 were core biopsy specimens, and 43 were gross dissection specimens. LN dissection specimens were grouped into those with parenchymal metastases (38 specimens) and those with purely subcapsular metastases (5 specimens) ( Fig. 1 A). Histopathology of 3 specimens with parenchymal metastases showed “infiltrative metastases,” without forming a nodule of more than 2 mm × 2 mm ( Fig. 1 B). The other 33 specimens were classified as nodular parenchymal metastases, containing area(s) that were at least 2 mm × 2 mm and lacking dominant lymphocyte contamination. This retrospective study was conducted under approval by the institutional review board.
DNA was extracted from FFPE neoplastic tissues as described previously  . Hematoxylin and eosin (H&E)--stained slides from 187 specimens were retrospectively reevaluated by a single pathologist (M. T. L.). One non-LN specimen did not have an H&E slide available for reevaluation. Three specimens (1 LN specimen and 2 non-LN specimens) were excluded from analysis of tumor cellularity because of significant crush artifact of the histomorphology. The estimated percentage of tumor nuclei in the remaining 184 specimens was separated into 5 quintiles (1%-20%, 21%-40%, 41%-60%, 61%-80%, and 81%-100%).
Pyrosequencing was performed as described previously  . The limit of detection by pyrosequencing is 5% mutant alleles (or 10% tumor cells for heterozygous mutations). Our initial pyrosequencing assay, performed before November 2011, detected a false-positive signal of p.V600E in specimens with a p.V600K mutation. Our revised assay can distinguish the p.V600K mutation and other non-p.V600E mutations from the p.V600E mutation  . Interpretation of complex pyrogram patterns due to mutations of 2 or more nucleotides on the same allele was resolved by Pyromaker, a free available software program ( http://pyromaker.pathology.jhmi.edu ) and confirmed by Sanger sequencing as described previously, , and .
We previously reported that our older pyrosequencing assay detected p.V600E mutations in 25 of 59 specimens and non-p.V600E mutation in 2 of 59 specimens (excluding 3 specimens with mucosal melanoma)  . A revised assay confirmed p.V600E mutations in 20 of the 25 specimens, whereas the remaining 5 specimens were positive for p.V600K. In combination with the additional 129 specimens (excluding 1 specimen with mucosal melanoma and 1 specimen with uveal melanoma) examined by the revised assay, there were 55 (29.3%) specimens with a p.V600E mutation, 18 (9.6%) specimens with a p.V600K mutation, and 4 (2.1%) specimens with another non-p.V600E mutation.
3.1. BRAF mutation in LN metastasis
The incidence ofBRAFmutations was significantly higher in the LN specimens (57% versus 35% for non-LN specimens,P< .01), with a higher incidence of p.V600E in LN specimens (49% versus 22%,P< .001) ( Table 1 ). There was no difference in p.V600K incidence between the LN and non-LN specimens (8% versus 10%). Non-p.V600E was observed in 4 (13%) of 30BRAF-positive LN specimens as compared with 18 (38%) of the 47BRAF-positive non-LN specimens (P= .02). There were no significant differences in sex or age distribution between the LN specimens and non-LN specimens (female population, 36% versus 36%; younger than 60-year-old population, 58% versus 54%).
|LN||53||30 (57%)||26 (49%)||4 (8%)||0|
|Non-LN||135||47 (35%)||29 (22%)||14 (10%)||4 (3%)|
|P < .01||P < .001||P = .55|
NOTE. Numbers in parentheses indicate incidence of mutations;Pvalue byχ2test or Fisher exact test.
Abbreviation: LN, lymph node.
3.2. Tumor cellularity in metastatic LN specimens
Tumor cellularity was retrospectively reevaluated ( Table 2 ). With a limit of detection of 10% tumor cells (5% mutant alleles), pyrosequencing was adequate in at least 180 (98%) of 184 specimens with an estimated tumor cellularity of 20% or more. The p.V600E mutation was detected in 3 of the 4 specimens with 1% to 20% tumor cellularity. A false-negative result could not be excluded for the remaining specimen with no detectableBRAFmutation.
|Total (n = 184)||LN (n = 52)||Non-LN (n = 132)|
|1%-20%||4 (2%)||4 (8%)||0 (0%)|
|21%-40%||15 (8%)||8 (15%)||7 (5%)|
|41%-60%||15 (8%)||8 (15%)||7 (5%)|
|61%-80%||55 (30%)||12 (23%)||43 (33%)|
|>80%||95 (52%)||20 (39%)||75 (57%)|
NOTE. Numbers in parentheses indicate the percentage.
Abbreviation: LN, lymph node.
The estimated tumor cellularity was significantly lower in the 52 LN specimens as compared with the 132 non-LN specimens (P< .001, 5 × 2 table). Tumor cellularity was 1% to 20% in 4 (8%) specimens (3 with a subcapsular metastasis and 1 with an infiltrative parenchymal metastasis) and 21% to 40% in 8 specimens (2 with a subcapsular metastasis, 2 with an infiltrative parenchymal metastasis, 2 with a nodular parenchymal metastasis, 1 core biopsy, and 1 fine needle aspiration) ( Table 2 ). The incidence of specimens with an estimated tumor cellularity of less than 40% was significantly higher in the LN specimens (12/52, 23%) as compared with the non-LN specimens (7/132, 5%) (P< .001).
Overall, mutant allele percentages of 20% or less were observed in 8 (31%) of the 26 p.V600E-positive LN specimens as compared with 3 (10%) of the 29 p.V600E-positive non-LN specimens (P= .048 by Fisher exact test), in line with the observation of lower estimated tumor cellularity in LN specimens ( Table 3 ). Among the 8 LN specimens with less than 20% p.V600E mutant alleles, 5 showed infiltrative or subcapsular metastases ( Fig. 1 A and B), 1 (case 3) showed nodule formation with significant lymphocyte contamination, and 2 (cases 26 and 126) showed nodule formation with the observed mutant alleles 10 percentage points less than the predicted mutant allele percentage (half of the estimated tumor cell percentage assuming heterozygosity of mutation) ( Table 4 ). In our previous report, single nucleotide polymorphism (SNP) array analysis of case 126 revealed a gain of chromosome 7 (which carries theBRAFgene), suggesting that gain of the wild-type allele may account for the discrepancy  . The reason for the discrepancy between the observed mutant allele percentage (12%) and the predicted mutant allele percentage (41%-50%) in case 26 is unclear because the SNP array failed. All of the 3 non-LN specimens with p.V600E mutant allele percentages of 20% or less were biopsy specimens with limited neoplastic tissues and significant lymphocyte infiltration adjacent to or surrounding the neoplastic cells.
|Total (n = 55)||LN (n = 26)||Non-LN (n = 29)|
|Case||Specimens||% Observed mutant||% Predicted mutant a|
a Based on the assumption of heterozygosity of mutation with no gain or loss of the BRAF gene.
Abbreviations: LN (infil), infiltrative LN metastasis; LN (sub), subcapsular LN metastasis; LN (nod), nodular LN metastasis.
3.3. p.V600E mutation with mutant allele-specific imbalance
Mutant allele-specific imbalance (MASI) indicates the increase of mutant allele percentage through loss of the wild-type allele or gain of the mutant allele. In our previous study using tumor cellularity as a quality assurance measure for accurate clinical detection ofBRAFmutations, 17 of 55 p.V600E-positive specimens showed 53% to 77% mutant alleles, indicating loss of the wild-type allele or gain of the mutant allele because the percentage of a heterozygous mutation should otherwise be below 50%  . In addition, SNP array analysis confirmed the amplification of chromosome 7q involving theBRAFgene in all 3 specimens with an observed mutant allele percentage that was less than 50% but still 10% higher than the predicted mutant allele percentage  . MASI was present in at least 20 (36%) of the 55 p.V600E-positive specimens, including 2 (13%) of the 15 specimens taken from skin or soft tissue on the extremities, face, or trunk; in 7 (27%) of the 26 LN specimens; and in 11 (79%) of the 14 specimens with distant internal organ metastases (5 in brain; 4 in lung; and 1 in liver, gallbladder, stomach, small intestine and vertebrae, respectively). Distant organ metastases were more frequently associated with MASI than skin/soft tissue and LN samples (P< .001, 3 × 2 table).
3.4. p.V600K and other non-p.V600E mutations were predominantly detected in elderly male patients, particularly those with head and neck melanoma
Patients were stratified according to sex and age to determine the correlation withBRAFmutations. Female patients were significantly younger than male patients in this cohort (49/67 or 73% versus 54/120 or 45% of patients aged younger than 60 years,P< .001). The incidence of p.V600E mutation was significantly higher in younger male patients (32% versus 8%), whereas that of p.V600K was significantly higher in older male patients (23% versus 2%) ( Table 5 ). However, the incidences of p.V600E and p.V600K mutations were similar in female patients, regardless of whether they were aged younger or older than 60 years. In males with aBRAFmutation, p.V600K was observed in 1 (6%) of 18 patients aged younger than 60 years in contrast to 15 (63%) of 24 patients aged older than 60 years (P< .0001) ( Fig. 2 ). In females with aBRAFmutation, p.V600K was observed in 1 (4%) of 25 patients aged younger than 60 years and 1 (10%) of 10 patients older than 60 years. All 4 specimens with other non-p.V600E mutations were from male patients aged older than 60 years ( Table 5 ).
|Total||BRAF||p.V600E||p.V600K||Other non p.V600E|
|<60 y||49||25 (51%)||24 (49%)||1 (2%)||0|
|≥60 y||18||10 (56%)||9 (50%)||1 (6%)||0|
|P = .74||P = .94||P = .47|
|<60 y||54||18 (33%)||17 (32%)||1 (2%)||0|
|≥60 y||66||24 (36%)||5 (8%)||15 (23%)||4 (6%)|
|P = .73||P < .001||P < .001|
NOTE. Numbers in parentheses indicate the frequency of mutations;Pvalue byχ2test or Fisher exact test.
TheBRAFmutation was detected in 40 of the 100 non-head and neck specimens (38 p.V600E, 1 p.V600K, and 1 p.K601E) and in 12 of the 39 head and neck specimens (4 p.V600E, 7 p.V600K, and 1 p.V600R). The p.V600K mutation was observed in 7 (58%) of the 12BRAF-positive head and neck specimens, as compared with 1 (3%) of 40BRAF-positive non-head and neck specimens (P< .0001 by Fisher exact test).
Detection of theBRAFmutation is mandatory to select patients with metastatic melanoma for targeted therapy for this mutation. In this study, pyrosequencing failed in 1 specimen submitted from an outside hospital, with unknown tissue fixation conditions. Given that melanin may inhibit polymerase chain reaction (PCR) efficiency and should be avoided whenever possible  , the accompanying H&E slide was examined; and no melanin was observed. Indeed, tumor pigmentation did not seem to affect our pyrosequencing assay forBRAFmutation detection. Fifty-three (28%) of the 188 melanoma specimens submitted forBRAFtesting were derived from LN specimens. Such specimens may pose challenges to clinical molecular diagnosis due to lower tumor cellularity. This study also found a higher incidence of p.V600E mutations in LN than non-LN specimens, a higher incidence of p.V600K and other non-p.V600E mutations in male patients aged older than 60 years, and a potential association of MASI with distant organ metastasis.
Specimens with low tumor cellularity may lead to false-negative results if assays with low analytic sensitivity are used. Our previous analysis of colorectal specimens showed that adjuvant therapy before resection was the main reason for depleted tumor cellularity in our institute  . LN specimens, though, accounted for only 2% of colorectal specimens submitted forKRASmutation detection in that study. In this work, 28% of specimens analyzed were from LNs, either after resection, biopsy, or fine needle aspiration. Presumably, this relatively high percentage of LN (versus primary tumor or non-LN metastases) specimens submitted is a consequence of the primary tumor resection before the patient's arrival at our tertiary referral center. It is recommended to request the tissue blocks or unstained slides of the primary tumor for potential molecular tests, when the H&E-stained slides of the primary tumor are sent for review.
LN specimens are particularly challenging for molecular testing. In addition to a paucity of tumor cells, numerous surrounding lymphocytes add a layer of difficulty in the selection of target areas, macrodissection of tumor tissue, and accurate estimation of cellularity and . Pathologists tend to overestimate the tumor cellularity because they may base the estimates on the ratio of surface areas between tumor and normal tissue instead of on the ratio of nuclei. Overestimation of tumor cellularity may occur even more frequently in LN specimens. The tumor cells, although visually more impressive, contain the same DNA quantity as the smaller lymphocytes  . This may lead to artifactual discrepancies in reported mutations between the primary tumor and LN metastases, whereas real discrepancies may occur as a result of tumor heterogeneity or treatment-related tumor evolution and .
In addition to selecting adequate specimens, use of molecular assays with higher sensitivity can also help prevent false-negative results from LNs or other specimens with lower tumor cellularity. In addition to pyrosequencing, others assays with 5% or higher analytic sensitivity should be considered over Sanger sequencing, including real-time PCR-based assays, the primer extension MassARRAY system (Sequenom, San Diego, CA), the multiplex SNaPshot assay (Applied Biosystems, Courtaboeuf, France), allele-specific PCR, and NGS, , , and . A p.V600E mutation could have been missed in the 3 LN specimens with 10% or fewer mutant alleles and perhaps also in the 5 LN specimens and 3 non-LN specimens with 20% or less mutant alleles. Recently, we have validated an NGS platform and showed a background noise of less than 0.3% for p.V600E in FFPE-negative control specimens. The p.V600E mutant allele was detected at 1% to 2% in 2 LN specimens with subcapsular metastasis (data not shown). Besides molecular testing, immunohistochemical staining is also a highly sensitive and specific assay for detection ofBRAFmutations  . It is particularly useful when tumor cells are intermingled with nonneoplastic cells, such as in LN fine needle aspiration specimens. However, the currently available antibodies are specific to the p.V600E mutation and therefore may not be able to detect non-p.V600E mutations.
Distinguishing clinicopathological features of patients with p.V600E and p.V600K mutations have been reported, , and . By segregating patients into 4 groups according to age and sex, the current study confirmed a higher frequency of p.V600E mutations in patients aged younger than 60 years and an association of p.V600K with male patients aged older than 60 years. Our observation of more frequent p.V600K mutations in head and neck primary tumors is consistent with the hypothesis that the p.V600K mutation of theBRAFgene in melanomas may be related to chronic sun exposure, , and .
BRAFmutation status provides several clinical standard-of-care applications, such as prediction of responsiveness to BRAF inhibitors in melanoma patients and anti-epidermal growth factor receptor therapy in colorectal cancer patients, , , , and .BRAFmutation status is also part of the algorithm to distinguish sporadic colorectal cancers with microsatellite instability from hereditary nonpolyposis colorectal cancers  . The p.V600E mutation has been associated with central LN metastasis of papillary thyroid carcinoma  . A recent larger cohort study confirmed this association in the overall studied population but not the patient population with the classic variant of papillary thyroid carcinoma  . However, the association ofBRAFmutation status and primary melanoma with LN metastases remains unclear. In a study of 105 patients with stage III melanoma,BRAFmutations were examined in LN specimens from 100 patients by Sanger and/or pyrosequencing  . The presence of aBRAFmutation and the number of involved LNs were significantly associated with overall survival. However,BRAFmutation status did not correlate with a higher incidence of sentinel node involvement or a higher number of LN metastases. In another study of 291 specimens from 108 patients by Sanger sequencing, 84 LN specimens did not show a higher incidence ofBRAFmutations  . In this study, we showed a higher incidence of p.V600E in LN specimens (26/53, 49%) as compared with non-LN specimens (29/135, 22%). Examination of theBRAFmutations in the primary tumor and prospective follow-up of LN involvement are needed to elucidate if melanoma with p.V600E mutation may be associated with more frequent LN metastases.
MASI is a common phenomenon in cancer and has been shown to carry an adverse prognosis not only in hematologic malignancies but also in solid tumors and . MASI results from either amplification of mutant alleles or deletion of wild-type alleles and results in a higher than expected mutant allele frequency based on the assumption of a heterozygous mutation. We confirmed that MASI of theBRAFmutation was present in melanomas, especially in specimens with distant organ metastases. However, our approaches may not be able to detect MASI in specimens with low tumor cellularity. The incidence of MASI is likely higher than we observed, particularly in LN specimens.BRAFamplification was reported in melanoma before the era of targeted therapy and has recently been shown to mediateBRAFinhibitor resistance, , and . Further investigation is warranted to elucidate if pretreatment MASI predicts a poor response toBRAFinhibitor therapy.
In summary, LNs were common specimens submitted for molecular diagnosis of patients with melanoma. LN specimens, especially those with microscopic subcapsular or infiltrative metastases, may pose challenges for molecular diagnosis because heavy lymphocyte contamination within the area(s) designated for manual macrodissection often causes both low tumor cellularity and difficulty in the estimation of tumor cellularity. Molecular diagnostic assays with higher analytic sensitivity are needed to minimize false-negative results. Further studies are warranted to investigate if the p.V600E mutation in primary tumors is associated with a higher risk for LN metastasis and if MASI ofBRAFmutations beforeBRAFinhibitor treatment predicts a poor response to targeted therapy.
-  H. Davies, G.R. Bignell, C. Cox, et al. Mutations of the BRAF gene in human cancer. Nature. 2002;417:949-954 Crossref
-  G.V. Long, A.M. Menzies, A.M. Nagrial, et al. Prognostic and clinicopathologic associations of oncogenic BRAF in metastatic melanoma. J Clin Oncol. 2011;29:1239-1246 Crossref
-  A.M. Menzies, L.E. Haydu, L. Visintin, et al. Distinguishing clinicopathologic features of patients with V600E and V600K BRAF-mutant metastatic melanoma. Clin Cancer Res. 2012;18:3242-3249 Crossref
-  W.O. Greaves, S. Verma, K.P. Patel, et al. Frequency and spectrum of BRAF mutations in a retrospective, single-institution study of 1112 cases of melanoma. J Mol Diagn. 2013;15:220-226 Crossref
-  J.A. Sosman, K.B. Kim, L. Schuchter, et al. Survival in BRAF V600-mutant advanced melanoma treated with vemurafenib. N Engl J Med. 2012;366:707-714 Crossref
-  G.A. McArthur, P.B. Chapman, C. Robert, et al. Safety and efficacy of vemurafenib in BRAF(V600E) and BRAF(V600K) mutation-positive melanoma (BRIM-3): extended follow-up of a phase 3, randomised, open-label study. Lancet Oncol. 2014;15:323-332 Crossref
-  K.T. Flaherty, C. Robert, P. Hersey, et al. Improved survival with MEK inhibition in BRAF-mutated melanoma. N Engl J Med. 2012;367:107-114 Crossref
-  A. Hauschild, J.J. Grob, L.V. Demidov, et al. Dabrafenib in BRAF-mutated metastatic melanoma: a multicentre, open-label, phase 3 randomised controlled trial. Lancet. 2012;380:358-365 Crossref
-  J. Dudley, L.H. Tseng, L. Rooper, et al. Challenges posed to pathologists in the detection of KRAS mutations in colorectal cancers. Arch Pathol Lab Med. 2014; [in press]
-  J.C. Dudley, G.T. Gurda, L.H. Tseng, et al. Tumor cellularity as a quality assurance measure for accurate clinical detection of BRAF mutations in melanoma. Mol Diagn Ther. 2014;18:409-418
-  M.T. Lin, L.H. Tseng, R.G. Rich, et al. Δ-PCR, a simple method to detect translocations and insertion/deletion mutations. J Mol Diagn. 2011;13:85-92 Crossref
-  G. Chen, M.T. Olson, A. O'Neill, et al. A virtual pyrogram generator to resolve complex pyrosequencing results. J Mol Diagn. 2012;14:149-159 Crossref
-  M.T. Olson, C. Harrington, K. Beierl, et al. BRAF pyrosequencing analysis aided by a lookup table. Am J Clin Pathol. 2014;141:639-647 Crossref
-  L. Eckhart, J. Bach, J. Ban, E. Tschachler. Melanin binds reversibly to thermostable DNA polymerase and inhibits its activity. Biochem Biophys Res Commun. 2000;271:726-730 Crossref
-  H. Viray, K. Li, T.A. Long, et al. A prospective, multi-institutional diagnostic trial to determine pathologist accuracy in estimation of percentage of malignant cells. Arch Pathol Lab Med. 2013;137:1545-1549 Crossref
-  A.J. Smits, J.A. Kummer, P.C. de Bruin, et al. The estimation of tumor cell percentage for molecular testing by pathologists is not accurate. Mod Pathol. 2014;27:168-174 Crossref
-  K. Schmid, N. Oehl, F. Wrba, R. Pirker, C. Pirker, M. Filipits. EGFR/KRAS/BRAF mutations in primary lung adenocarcinomas and corresponding locoregional lymph node metastases. Clin Cancer Res. 2009;15:4554-4560 Crossref
-  C.B. Han, J.T. Ma, F. Li, et al. EGFR and KRAS mutations and altered c-Met gene copy numbers in primary non-small cell lung cancer and associated stage N2 lymph node-metastasis. Cancer Lett. 2012;314:63-72
-  S. Magnin, E. Viel, A. Baraquin, et al. A multiplex SNaPshot assay as a rapid method for detecting KRAS and BRAF mutations in advanced colorectal cancers. J Mol Diagn. 2011;13:485-492 Crossref
-  H. Halait, K. Demartin, S. Shah, et al. Analytical performance of a real-time PCR-based assay for V600 mutations in the BRAF gene, used as the companion diagnostic test for the novel BRAF inhibitor vemurafenib in metastatic melanoma. Diagn Mol Pathol. 2012;21:1-8 Crossref
-  J. Lade-Keller, K.M. Rømer, P. Guldberg, et al. Evaluation of BRAF mutation testing methodologies in formalin-fixed, paraffin-embedded cutaneous melanomas. J Mol Diagn. 2013;15:70-80 Crossref
-  G.V. Long, J.S. Wilmott, D. Capper, et al. Immunohistochemistry is highly sensitive and specific for the detection of V600E BRAF mutation in melanoma. Am J Surg Pathol. 2013;37:61-65 Crossref
-  R. Jewell, P. Chambers, M. Harland, et al. Clinicopathologic features of V600E and V600K melanoma—letter. Clin Cancer Res. 2012;18:6792 Crossref
-  A.D. Bucheit, E. Syklawer, J.A. Jakob, et al. Clinical characteristics and outcomes with specific BRAF and NRAS mutations in patients with metastatic melanoma. Cancer. 2013;119:3821-3829 Crossref
-  F. Di Nicolantonio, M. Martini, F. Molinari, et al. Wild-type BRAF is required for response to panitumumab or cetuximab in metastatic colorectal cancer. J Clin Oncol. 2008;26:5705-5712 Crossref
-  W.K. Funkhouser Jr., I.M. Lubin, F.A. Monzon, et al. Relevance, pathogenesis, and testing algorithm for mismatch repair-defective colorectal carcinomas: a report of the association for molecular pathology. J Mol Diagn. 2012;14:91-103 Crossref
-  R.P. Tufano, G.V. Teixeira, J. Bishop, K.A. Carson, M. Xing. BRAF mutation in papillary thyroid cancer and its value in tailoring initial treatment: a systematic review and meta-analysis. Medicine (Baltimore). 2012;91:274-286
-  C. Li, P. Aragon Han, K.C. Lee, et al. Does BRAF V600E mutation predict aggressive features in papillary thyroid cancer? Results from four endocrine surgery centers. J Clin Endocrinol Metab. 2013;98:3702-3712 Crossref
-  S. Moreau, P. Saiag, P. Aegerter, et al. Prognostic value of BRAF(V600) mutations in melanoma patients after resection of metastatic lymph nodes. Ann Surg Oncol. 2012;19:4314-4321 Crossref
-  M. Colombino, M. Capone, A. Lissia, et al. BRAF/NRAS mutation frequencies among primary tumors and metastases in patients with melanoma. J Clin Oncol. 2012;30:2522-2529 Crossref
-  J. Soh, N. Okumura, W.W. Lockwood, et al. Oncogene mutations, copy number gains and mutant allele specific imbalance (MASI) frequently occur together in tumor cells. PLoS One. 2009;4:e7464 Crossref
-  D.J. Hartman, J.M. Davison, T.J. Foxwell, M.N. Nikiforova, S.I. Chiosea. Mutant allele-specific imbalance modulates prognostic impact of KRAS mutations in colorectal adenocarcinoma and is associated with worse overall survival. Int J Cancer. 2012;131:1810-1817 Crossref
-  C. Spittle, M.R. Ward, K.L. Nathanson, et al. Application of a BRAF pyrosequencing assay for mutation detection and copy number analysis in malignant melanoma. J Mol Diagn. 2007;9:464-471 Crossref
-  H. Shi, G. Moriceau, X. Kong, et al. Melanoma whole-exome sequencing identifies (V600E)B-RAF amplification-mediated acquired B-RAF inhibitor resistance. Nat Commun. 2012;3:724 Crossref
-  J. Villanueva, J.R. Infante, C. Krepler, et al. Concurrent MEK2 mutation and BRAF amplification confer resistance to BRAF and MEK inhibitors in melanoma. Cell Reprogram. 2013;4:1090-1099 Crossref
a Department of Pathology, Johns Hopkins University School of Medicine, Baltimore, MD, 21287 USA
b Department of Pathology, Penn State Hershey Medical Center, Hershey, PA, 17033 USA
c Departments of Pathology, Massachusetts General Hospital, Boston, MA, 02114 USA
d Department of Medical Genetics, National Taiwan University Hospital, Taipei, 100, Taiwan
e Department of Oncology, Johns Hopkins University School of Medicine, Baltimore, MD, 21287 USA
☆ Competing interests: All authors confirm no conflict of interest. This retrospective analysis was not funded.
© 2014 Elsevier Inc., All rights reserved.