Introduction
The incidence of cutaneous melanoma is increasing throughout most of the Western world [1]. Although cure is achieved with surgery only in the majority of patients with early-stage disease, curative treatment for patients with metastatic melanoma remains elusive. The median survival time for melanoma patients with metastatic disease is 8~9 months, and the 3-year overall survival rate is less than 15% [2]. Melanoma is a heterogeneous disease. As a result of intensive molecular analyses, the influence of several important oncogenes has been discovered. These include oncogenic activation by mutation or amplification in BRAF, KIT, NRAS, cyclin D, cyclin dependent kinase 4, and alterations in the ERBB4 gene [3]. To date, several targeted therapies have been approved for treating melanoma [4].
Mutations in the BRAF gene occur in approximately 50% of cases of cutaneous melanoma [5,6] and the most common BRAF mutation is a T1799A transversion mutation in exon 15 of the gene, which causes V600E (Val600Glu) amino acid substitution in the protein. This results in constitutive activation of BRAF and activation of the MAPK pathway [7]. The presence of mutated NRAS, which has been reported in approximately 15% of cases of cutaneous melanoma [5,6], leads to up-regulation of the MAPK pathway, the phosphatidylinositol 3′ kinase (PI3K) pathway, and the RAL pathway, resulting in inhibition of apoptosis, promotion of cell proliferation, invasion, and anchorage-independent growth [8,9]. The most common NRAS codon 61 mutations in exon 2 in cutaneous melanoma are the Q61R (CAA/CGA) and Q61K (CAA/AAA) changes, which lead to substitutions from glutamine to arginine or to lysine, respectively. Genetic mutations in BRAF and NRAS genes have shown correlation with the clinicopathologic features and progression of melanoma; however, the effect of these mutations on the clinical outcome remains uncertain, and previous studies have reported conflicting results [10-15]. Some investigators have suggested that BRAF mutation may be acquired during development of metastasis and may show correlation with progression rather than initiation [16-18]; however, other investigators have suggested that BRAF mutations are most likely to occur before the development of melanoma metastasis [14,19].
With the advent of BRAF V600E-specific inhibitors, identification of mutations in the BRAF and NRAS genes may be of importance in clinical trials. However, the majority of these studies included patients with recurrent or advanced stage disease and the observations were conducted in Caucasian populations [20-23]. The incidence of KIT gene mutations in Asians (Chinese and Koreans) was lower than that reported in Western patients (7.9~11% vs. 29%) [24-26], indicating that the mutation status of the BRAF and NRAS genes may also differ between Asians and Caucasians. Recently, two large-scale analyses of BRAF and NRAS mutations in Chinese patients demonstrated a lower incidence of BRAF and NRAS mutations in Chinese patients [25,27]. Even though screening for BRAF and NRAS mutations was performed in one Korean study, the incidence of these mutations was extraordinarily low compared to results from Western or other Asian studies [26].
To the best of our knowledge, this is the first study in Korea to investigate the frequency of BRAF and NRAS mutations and to compare genetic differences between primary melanoma and metastatic melanoma tissues. Despite the fact that this was a small series study, findings of this study may be informative because the incidence of BRAF and NRAS mutations among Korean melanoma patients has not been well studied to date.
Methods
We conducted a retrospective review of the clinical records of 130 patients diagnosed as having a malignant melanoma at Gachon University Gil Hospital (GUGH) between 1999 and 2010. Among patients with follow-up information, we selected 10 patients with primary cutaneous melanoma without metastasis during a period of five years and 20 patients with metastatic melanoma. Paraffin blocks or clinical data were available in all cases. Because of technical problems, analysis for the BRAF mutation was performed at least one site in only 22 cases. Only four primary cutaneous melanoma patients without metastasis and 18 patients with metastatic melanoma were included in this study. Clinical characteristics of all 22 patients are shown in Table 1. The Institutional Review Board of GUGH approved the acquisition, analysis, and reporting of patient data and human tissue (approval number: GIRBA 2487).
Using the QIAgen Tissue kit (DNeasy Blood & Tissue kit, Cat. No. 69506; Qiagen, Valencia, CA), tumor-rich areas were extracted from five paraffin sections of 10 mm thickness containing a representative portion of each tumor block. For detection of hotspot mutations, we amplified exon 15 of the BRAF gene, and exons 2 and 3 of the NRAS gene, using PCR. PCR primers for BRAF and NRAS mutational analyses were as follows: BRAF exon 15 (forward) 5’-TGCTTGCTCTGATAGGAAAAT-3’ and (reverse) 5’-AGCATCTCAGGGCCAAAAAT-3’; NRAS exon 2 (forward) 5’-GAACCAAATGGAAGGTCACA-3’ and (reverse) 5’-TGGGTAAAGATGATCCGACA-3’, NRAS exon 3 (forward) 5’-AGGCAGAAATGGGCTTGGAT-3’ and (reverse) 5’-CGCCTGTCCTCATGTATTGG-3’. Purified PCR products were sequenced directly on both strands. An ABI PRISM 3100 DNA analyzer (Applied Biosystems, Foster city, CA) was used to perform sequencing reactions. Repeat PCR and sequencing were performed using a different primer for confirmation of the mutations.
Results
Overall, mutations in BRAF exon 15 were detected in six of 22 cases (27.3%) and all BRAF mutations were represented by the valine to glutamic acid substitution at position 600 (V600E). Mutations in NRAS exons 2 and 3 were not detected in any of the cases (Tables 1, 2). BRAF mutations were compared with known prognostic factors in patients with primary and metastatic melanoma. Age, gender, Breslow thickness, and ulceration did not show correlation with the BRAF mutations (Table 1). Among the cases involving metastasis, BRAF mutations were detected in four of 18 cases (22.2%), and in all four cases, BRAF mutations were identified in metastatic lymph nodal tissues. The increase in frequency of BRAF mutation in lymph node metastases, compared with that of other sites, was statistically significant (28.6% vs. 0%, P<0.01).
Table 2 shows the BRAF mutation status of 18 paired primary melanomas and metastatic melanomas. Sites of primary melanoma included the skin in nine cases (there were three cases in the toe, two in the sole and one each in the finger, inguinal area, flank and back), the mucosa in four cases (one case each in the lip, vagina, esophagus and nasal cavity), and the viscera in three cases (two cases were in the small intestine and one case in the ovary), and two cases were of an unknown primary. Sites of metastatic melanoma included 14 cases in the lymph nodes and one in the skin, stomach, small intestine and liver. In six pairs, wild-type genes were detected in both the primary tumors and metastases (cases 1~6), and wild-type genes were detected in the metastases in seven pairs; however, evaluation of the BRAF mutation in the primary tumors failed (cases 7~13). The presence of a BRAF mutation was detected in metastatic tumors in three pairs. Among them, BRAF mutations were detected in both the primary and the metastasis in two pairs (cases 14, 15); however, evaluation of the BRAF mutation in the primary tumor of one pair failed (case 16). Among two cases involving metastatic melanomas of an unknown primary, a BRAF mutation was detected in one case (case 17) and no BRAF mutation was detected in the other case (case 18).
Discussion
The occurrence of BRAF mutations in patients with cutaneous melanoma ranges from 22% to 72% [28]. Approximately 90% of the reported BRAF mutations occur at residue 600, located in the activation domain of this kinase [29,30]. In the Catalogue of Somatic Mutations in Cancer (COSMIC) database, the occurrence of a number of less common variant codon 600 mutations has been reported in patients with malignant melanoma; these non-V600E mutations comprise approximately 10% of all codon 600 mutations reported in patients with malignant melanoma (MM) [31]. However, these studies have been conducted mainly in Caucasian populations, and the mutational status of the BRAF gene and the clinical significance of this mutation have not been well studied in Asian populations. Recently, findings from two large-scale studies examining BRAF V600E mutations in the Chinese population demonstrated the presence of BRAF mutations in 15% and 25.5% of cases of MM, respectively [25,27]. In addition, one Japanese study reported the detection of the BRAF mutation in only nine of 35 patients (26%) with MM [32]. According to the results of one previous Korean study, the BRAF mutation was detected in only one of 49 patients (2%) with MM [26]. However, compared to results from Western or other Asian studies, this incidence was very low. This study has a limitation in that the study samples were taken from acral and mucosal melanomas only. In our study, 27.3% of the patients with MM harbored BRAF mutations. This result was similar to the results of other Asian studies and this also suggests that the incidence of BRAF mutations in the Asian population is lower than that reported by Western studies. The incidence of BRAF mutation varies according to histologic subtype and tumor location [28]. In addition, Caucasians often show superficial spread of MM and nodular MM [33], whereas Asians present with acral lentigenous MM [32]. Findings of one Korean study demonstrated that 65% of the melanomas were located on the hands and feet, with acral lentiginous melanoma being the most common histologic subtype [34]. Therefore, racial, histological, and locational differences may be one explanation for the lower rate of BRAF mutations in Koreans and Asian patients with MM.
BRAF mutation was detected in four out of 14 cases (28.6%) of lymph node metastases; however, no mutations in the metastases of other sites were observed, which is statistically significant (P<0.01). However, the size of the study sample may be too small to draw an accurate conclusion. One study reported a somewhat higher frequency of BRAF mutations in primary melanomas associated with concurrent lymph node metastasis; however, the difference was not statistically significant [18]. The association of BRAF mutation with lymph node metastasis remains unclear.
In our study, unfortunately, eight primary tumors failed to demonstrate their BRAF mutational status. Traditional direct (Sanger) sequencing has been widely used in clinical laboratories for mutation testing, including BRAF mutations [35]; however, this method suffers from limited sensitivity in the detection of mutations that are present in low percentages in a specimen [36]. In one study, the failure rate of Sanger sequencing was reported as 9.2% [37]. These eight cases were punch-biopsied specimens from the skin or mucosa, with a very low tumor content. Efforts to secure specimens with a high tumor content and to enrich the tumor by macro- or microdissection may result in the enhanced accuracy of Sanger sequencing [37]. In addition, new methods, such as pyrosequencing assays, should be applied for accurate and quantitative identification of BRAF mutations [38].
Unfortunately, NRAS mutation was not identified in this study. The occurrence of NRAS mutations in patients with cutaneous melanoma ranges from 0% to 50%, and NRAS mutations have been reported to show a significant association with nodular melanoma [28]. In one large-scale Chinese study, the incidence of NRAS mutations was reported as 7.2% [25]. In a Korean study, NRAS mutation was detected in only one of 47 patients (2.1%) with mucosal or acral melanoma [26]. Further studies are needed in order to determine the incidence of BRAF and NRAS mutations in Korean populations according to the histologic subtype and to determine whether BRAF or NRAS mutations in conjunction with other genetic aberrations found in MMs, such as epigenetic changes of tumor suppressor genes and/or allelic imbalance, play a role in tumor progression or metastasis.
For comparison of the mutational status between primary and metastatic MM, we assessed 18 pairs of primary MM tumors and their nodal or extranodal metastases. In six pairs, the wild-type gene was detected in both the primary tumors and metastases, and among six pairs, the wild-type gene was detected at three sites (primary tumor, lymph node metastasis, and brain metastasis) in one pair (case 5). In two pairs, BRAF mutations were detected in both the primary tumors and their respective metastases.
In conclusion, the incidence of BRAF mutations in the Korean population is low as in the Japanese and the Chinese reports and NRAS mutation is a very rare event. Although the number of tumors included in this study was too small to be conclusive, the results of our study indicated that the BRAF mutation may be an early event that occurs before metastasis, and that BRAF mutation did not appear to contribute to disease progression or metastasis. Larger studies will be needed in order to determine whether the BRAF mutation plays a significant role in lymph node metastasis.
