Welcome international healthcare professionals

This site is no longer supported and will not be updated with new content. You are welcome to browse and download all content already included in the site. Please note you will have to register your email address to access the site.

You are here

BRAF mutation status is an independent prognostic factor for resected stage IIIB and IIIC melanoma: Implications for melanoma staging and adjuvant therapy

European Journal of Cancer



5-year survival for melanoma metastasis to regional lymph nodes (American Joint Committee on Cancer stage III) is <50%. Knowledge of outcomes following therapeutic lymphadenectomy for stage III melanoma related to BRAF status may guide adjuvant use of BRAF/MEK inhibitors along with established and future therapies.


To determine patterns of melanoma recurrence and survival following therapeutic lymph node dissection (TLND) associated with oncogenic mutations.


DNA was obtained from patients who underwent TLND and had ⩾2 positive nodes, largest node >3 cm or extracapsular invasion. Mutations were detected using an extended Sequenom MelaCARTA panel.


Mutations were most commonly detected in BRAF (57/124 [46%] patients) and NRAS (26/124 [21%] patients). Patients with BRAF mutations had higher 3-year recurrence rate (77%) versus 54% for BRAF wild-type patients (hazard ratio (HR) 1.8, p = 0.008). The only prognostically significant mutations occurred in BRAF: median recurrence-free (RFS) and disease-specific survival (DSS) for BRAF mutation patients was 7 months and 16 months, versus 19 months and not reached for BRAF wild-type patients, respectively. Multivariate analysis identified BRAF mutant status and number of positive lymph nodes as the only independent prognostic factors for RFS and DSS.


Patients with BRAF mutations experienced rapid progression of metastatic disease with locoregional recurrence rarely seen in isolation, supporting incorporation of BRAF status into melanoma staging and use of BRAF/MEK inhibitors post-TLND.

Keywords: Melanoma, Molecular diagnostic techniques, Oncogenes, Proto-oncogene proteins B-raf.

1. Introduction

Regional lymph nodes are common first recurrence sites in patients with primary cutaneous melanoma, representing American Joint Committee on Cancer (AJCC) stage III [1], [2], [3], and [4]. Therapeutic lymph node dissection (TLND) is standard, although 5-year survival is <50%. Interferon-2α provides modest improvement in recurrence-free survival, but with significant toxicity [5] . Adjuvant radiation therapy reduces regional recurrence rates without improving overall survival, but is considered relevant for patients at high risk for regional recurrence [6] .

Within AJCC stage III melanoma, survival estimates demonstrate variability in patient outcomes and require further refinement [4] . Recently, molecular and genomic profiling of melanoma by microarray and sequencing studies [7], [8], [9], [10], [11], and [12] has established subtypes and pathways associated with progression [13] .

Activating mutations in BRAF have been reported in 40–50% of primary [13], [14], [15], and [16] and metastatic melanoma [17], [18], [19], [20], [21], and [22] while NRAS mutations were present in 15–20% [17] and [23]. BRAF mutation status is associated with shorter survival for unresectable stage IIIC and stage IV disease [17] . Few studies have reported survival for stage III melanoma patients according to mutation status [24], [25], and [26]. Knowledge of outcomes associated with BRAF mutation is timely, as BRAF/MEK inhibitors improve survival in stage IV melanoma [27], [28], and [29] and are under investigation for stage III melanoma.

No biomarkers have yet been successfully adopted to guide management of stage III melanoma patients, and patterns of recurrence associated with oncogenic mutations have not been well documented. This study reports prevalence of oncogenic mutations in stage III B and C melanoma, recurrence patterns following TLND and prognostic factors for recurrence-free and melanoma-specific survival.

2. Materials and methods

2.1. Patients

A prospective database of stage IIIA-C cutaneous melanoma patients, presenting to the Princess Alexandra Hospital (PAH) Melanoma Unit and affiliated private hospitals, has been maintained since 1997. Permission to collect and use information was approved by the hospital ethics committee. Patients were considered for this study if they underwent TLND surgery at the PAH and had high risk features, defined as either ⩾2 involved lymph nodes; largest lymph node >3 cm; or extracapsular lymph node invasion. Those who had pre-operative therapy or whose tumour tissue was in private pathology labs were excluded. Formalin-fixed paraffin embedded (FFPE) lymph node tumour samples from 134 patients who met these criteria were reviewed according to the AJCC staging system for cutaneous melanoma seventh edition [4] by an experienced melanoma pathologist (DL) and had DNA extracted as described in Supplementary Methods . Demographic, pathologic and treatment-related variables were prospectively recorded.

2.2. Follow-up and recurrence

After surgery, patients were followed up every 3 months for 2 years, then every 6 months for 3 years and then annually to 10 years. At follow-up, investigations including imaging were directed at symptoms, unless the patient was on a clinical trial requiring routine radiology. Follow-up was complete on all patients at time of analysis. Recurrence was defined as histological proof or unequivocal radiological evidence as follows: local nodal (within the boundaries of the previous lymphadenectomy); in transit (between primary site and draining lymphatic basins); and distant (all other sites). Recurrence was considered synchronous if detected in two anatomical sites within 30 days of each other. For purposes of analysis, site(s) of first recurrence was/were used.

2.3. Genetic analysis

Mutation detection was performed using the Sequenom MassARRAY® system (Sequenom, San Diego) in conjunction with an extended version of the MelaCARTA panel [30] . Briefly, 20 ng of genomic DNA was amplified and then cleaned with shrimp alkaline phosphatase. Single base pair extension reactions using iPLEX Pro chemistry were performed, resin-treated to remove contaminants and spotted onto SpectroCHIP II arrays. Mutant and wild type alleles were discriminated using the Sequenom MassARRAY® Analyser 4 platform. Mutations were detected using a minimum 10% threshold of the mutant allele peak and were manually reviewed.

2.4. Statistical analysis

Statistical analysis was carried out with SPSS for Windows, version 21.0 (Statistical Package for the Social Sciences, SPSS, Inc., Chicago, IL). Continuous variables were expressed as median (range) and compared using Mann–Whitney U test. Categoric variables were compared using χ2 or Fisher’s exact test. Survival probabilities were estimated by the Kaplan–Meier method, with differences in survival rates assessed using log-rank test to determine univariate significance. Multivariate analyses used Cox regression and included any factor with univariate p value ⩽0.05 for assessment of prognostic factors for recurrence-free (RFS) or disease-specific survival (DSS).

3. Results

3.1. Patients

134 patients with high risk stage IIIB or IIIC melanoma who underwent therapeutic lymph node dissection (TLND) without neoadjuvant therapy were identified. Adjuvant external beam radiation therapy was administered to 66/134 (48%) patients, immunotherapy trials to 13/134 (10%) and no patient received adjuvant systemic therapy. Median follow-up for survivors was 30.0 months (range 1–141 months, standard deviation [SD] 34.91).

3.2. Mutation analysis

We used the extended Sequenom MelaCARTA panel to screen DNA isolated from lymph node tumours for key hotspot mutations known to occur in melanoma, covering 46 mutations in 26 genes [30] . Mutation detection failed in 10/134 patients (7%). 119 mutations were identified in 17 genes in 93/124 (75%) patients, with 31/124 (25%) patients having no mutations detected in this panel ( Table 1 ). Mutations were most common in BRAF (57/124 [46%] patients), predominantly resulting in V600E substitution (39/124 [31%] patients), and NRAS (26/124 [21%] patients). Multiple mutations were identified in 22/124 [18%] patients: 18/124 (15%) patients had 2 genes mutated and 4/124 (3%) patients had mutations in 3 genes ( Supplementary Table 1 ). BRAF and NRAS mutations were mutually exclusive.

Table 1 Prevalence of oncogene mutations detected by MelaCARTA in patients with resected stage IIIB and IIIC cutaneous melanoma (n = 124). Bold type indicates factors that reached statistical significance.

GENE # of samples Percentage (%)
BRAF 57 46
 V600E 39  
 V600K 11  
 V600R 3  
 K601E 3  
 G466E 1  
NRAS 26 21
 Q61Q 11  
 Q61H 9  
 Q61K 4  
 Q61P 1  
 G13R 1  
PPP6C 6 5
 R264C 6  
RAC1 6 5
 P29S 6  
CTNNB1 3 2
 S45F 3  
KIT 3 2
 K642E 2  
 L576P 1  
MEK 2 2
 P124L 1  
 Q56P 1  
MET 3 2
 T992I(T1010I) 3  
PTK2B 1 1
 G414V 1  
 S722F 2  
ERBB4 2 2
 E452K 2  
CDK4 2 2
 R24C 1  
 R24H 1  
IDH1 2 2
 R132C 2  
AKT3 1 1
 E17K 1  
GNAQ 1 1
 R183Q 1  
IDH2 1 1
 R172K 1  
KRAS 1 1
 G12A 1  
No mutations 31 25

3.3. Prognostic factors for recurrence-free and disease specific survival

Median recurrence-free survival (RFS) was 10 months (range 1–141, standard error 2.33). In order to identify mutations associated with RFS, log-rank univariate analyses were undertaken. All mutations within a single gene were considered together. Both RFS and disease-specific survival (DSS) analyses indicated that there were no significant differences in survival between the 39 patients with BRAF V600E mutations and the 18 cases with any other BRAF mutations (data not shown), therefore all BRAF mutations were analysed as a single variable. The only gene in which mutations were found to have prognostic significance for either RFS or DSS was BRAF ( Fig. 1 A and B; overall survival shown in Supplementary Fig. 1 ). There were no significant differences in RFS or DSS between patients with mutations in any other genes and those patients with no mutations detected using the MelaCARTA panel (data not shown). There were no differences in RFS or DSS for patients with one, two or three gene mutations detected (data not shown). Therefore, for remaining analyses, patients were dichotomised according to BRAF mutation status.


Fig. 1 Kaplan–Meier plot of (A) recurrence-free survival according to BRAF mutation status for 124 patients with resected stage IIIB–C melanoma, and (B) disease-specific survival according to BRAF mutation status for 120 patients with resected stage IIIB-C melanoma, excluding 4 BRAF positive patients who received vemurafinib for distant recurrence.

3.4. Clinical correlates of BRAF status

Clinicopathological variables according to BRAF mutation status are shown in Table 2 . There were no significant differences between patients with BRAF mutant and BRAF wild-type tumours for age at operation ( Supplementary Fig. 2 , median 52 [range 21–86, SD 19.39] versus 62 [range 24–83, SD 16.65] years, p = 0.15, Mann–Whitney), number of positive nodes removed ( Supplementary Fig. 3 , median 3 [range 1–47, SD 8.75] versus 2 [range 1–17, SD 2.96], p = 0.85, Mann–Whitney), total number of nodes examined (median 20 [range 6–60, SD 13.43] versus 20 [range 4–106, SD 17.90], p = 0.65, Mann–Whitney) or size of the largest lymph node metastasis (median 3.5 [range 1–10, SD 1.73] cm versus 3.0 [range 1–10, SD 2.20] cm, p = 0.15, Mann–Whitney).

Table 2 Clinicopathological factors according to BRAF mutation status. Bold type indicates factors that reached statistical significance.

Clinicopathological factors BRAF mutants (n = 57) BRAF wild-type (n = 67) p value
N % N %
Sex         0.28
 Female 26 46 24 36  
 Male 31 54 43 64  
Primary tumour sites         0.15
 Head & Neck 12 21 13 19  
 Arm 2 4 6 9  
 Leg 11 19 22 33  
 Trunk 23 40 15 22  
 Unknown 9 16 11 16  
Lymph node metastases’ site         0.4
 Neck 16 28 19 28  
 Axilla 23 40 20 30  
 Groin/pelvis 18 32 28 42  
Extracapsular invasion         0.85
 Present 40 70 45 67  
 Absent 17 30 22 33  
AJCC stage          
 IIIB 21 37 22 33 0.71
 IIIC 36 63 45 67  
Lymph node (N) stage         0.85
 N1 (1 positive node) 12 21 17 25  
 N2 (2–3 positive nodes) 21 37 24 36  
 N3 (⩾4 positive nodes) 24 42 26 39  
Post-surgery radiotherapy         0.47
 Yes 31 54 31 46  
 No 26 46 36 54  
Recurrence (any site)         0.006
 Yes 42 74 33 49  
 No 15 26 34 51  

3.5. Recurrence free survival and patterns of recurrence

All recurrences, except for one patient, occurred <3 years after surgery. Therefore we analysed recurrence patterns at 3 years. The 3-year overall RFS for all 124 patients with complete mutation and recurrence data was 32%, with 74 patients developing recurrence. 16 patients developed synchronous recurrence in ⩾2 regions, resulting in a total of 93 recurrence events. Fig. 2 shows patterns of first recurrence according to BRAF status. Comparing BRAF mutant and wild-type groups revealed 3-year overall recurrence rates of 77% for BRAF mutant, and 54% for the BRAF wild-type group (p = 0.008, log rank). 3-year distant recurrence (as part of any recurrence) rate was 73% (37 events) for the BRAF mutant group versus 42% (22 events) for the BRAF wild-type group (p < 0.001 log rank). 3-year local nodal recurrence (as part of any recurrence) rate was 20% (7 events) for the BRAF mutant and 8% (4 events) for BRAF wild-type group (p = 0.15, log rank). The 3-year in transit/subcutaneous recurrence (as part of any recurrence) rate was 33% (13 events) for the BRAF mutant group compared to 22% (11 events) for the BRAF wild-type group (p = 0.19 log rank).


Fig. 2 Venn diagram of 3 year first recurrence pattern of (A) BRAF mutant patients (n = 57) and (B) BRAF wild-type patients (n = 67).

3.6. Predictors of survival

Factors significantly associated with RFS on univariate analysis included BRAF mutation status, number of positive lymph nodes and gender ( Table 3 ). Using multivariate analysis (Cox regression), BRAF mutant status and number of positive lymph nodes were identified as independent prognostic factors for RFS ( Table 3 ).

Table 3 Three year recurrence-free survival (RFS), univariate and multivariate analyses of clinicopathologic and molecular factors associated with first recurrence. Bold type indicates factors that reached statistical significance.

Variable Median RFS months 3 y RFS (%) Univariate (p) Multivariate (p) Hazard ratio (95% confidence interval (CI))
Sex     0.033    
 Male (n = 74) 18 43      
 Female (n = 50) 7 24      
Age     0.24    
 ⩽60 (n = 62) 7 32      
 >60 (n = 62) 18 38      
Breslow thickness     0.45    
 <2 mm (n = 53) 7 30      
 ⩾2 mm (n = 48) 10 36      
 Unknown (n = 23)          
Ulceration     0.16    
 No (n = 61) 7 26      
 Yes (n = 41) 15 46      
 Unknown (22)          
Site of surgery     0.17    
 Neck (n = 35) 11 36      
 Axilla (n = 43) 14 48      
 Groin/pelvis (n = 46) 7 24      
Extracapsular invasion     0.15    
 No (n = 39) 10 49      
 Yes (n = 85) 8 29      
Largest tumour     0.18    
 ⩽3 cm (n = 56) 11 39      
 >3 cm (n = 64) 7 29      
 Unknown (n = 4)          
Unknown primary     0.41    
 No (n = 104) 9 34      
 Yes (n = 20) 18 41      
Post-op radiation     0.46    
 No (n = 62) 10 37      
 Yes (n = 62) 10 34      
BRAF     0.008 0.010  
 Wild type (n = 67) 19 46     1
 Mutant (n = 57) 7 23     1.66 (1.04–2.77)
AJCC stage     0.12    
 IIIB (n = 43) 19 41      
 IIIC (n = 81) 8 32      
Number of positive nodes Hazard ratio (HR) 1.04 95% CI 1.01–1.07 0.003 0.017 1.03 (1.01–1.06)

Four patients with BRAF mutations received a BRAF-inhibitor after recurrence and were therefore excluded from DSS analyses. Clinicopathological factors significantly associated with DSS on univariate analysis were BRAF mutation status, extracapsular invasion and number of positive lymph nodes ( Table 4 ). Kaplan–Meier survival curves for DSS in patients with and without BRAF mutations and according to AJCC seventh edition N stage for number of involved lymph nodes are shown in Supplementary Fig. 4 , however we found that number of lymph nodes as a continuous variable carried more significant prognostic information, with a 4% increased hazard for every additional positive lymph node. Multivariate Cox regression analysis identified BRAF mutant status and number of positive lymph nodes as independent prognostic factors for DSS ( Table 4 ).

Table 4 Univariate and multivariate analyses of clinicopathologic and molecular factors associated with disease-specific survival (excluding 4 patients treated with BRAF inhibitors for recurrent disease). Bold type indicates factors that reached statistical significance.

Variable Median DSS months 3 yr DSS (%) Univariate (p) Multivariate (p) Hazard ratio (95% confidence interval (CI))
Sex     0.21    
 Male (n = 73) NR 53      
 Female (n = 47) 19 40      
Age     0.17    
 ⩽60 (n = 59) 18 43      
 >60 (n = 61) NR 53      
Breslow thickness     0.81    
 <2 mm (n = 51) 25 49      
 ⩾2 mm (n = 46) 25 44      
 Unknown (n = 23)          
Ulceration     0.33    
 No (n = 59) 25 42      
 Yes (n = 39) NR 59      
 Unknown (22)          
Site of surgery     0.20    
 Neck (n = 33) NR 51      
 Axila (n = 41) NR 55      
 Groin/pelvis (n = 46) 19 31      
Extracapsular invasion     0.03    
 No (n = 37) 68 NR      
 Yes (n = 83) 39 21      
Largest tumour     0.34    
 ⩽3 cm (n = 55) NR 58      
 >3 cm (n = 61) 24 48      
 Unknown (n = 4)          
Unknown primary     0.70    
 No (n = 100) 25 47      
 Yes (n = 20) 24 49      
Post-op radiation     0.93    
 No (n = 62) 25 48      
 Yes (n = 58) 25 47      
BRAF     0.001 0.004  
 Wild type (n = 67) NR 61     1
 Mutant (n = 53) 16 30     2.29 (1.30–4.05)
AJCC stage     0.06    
 IIIB (n = 43) NR 58      
 IIIC (n = 77) 21 42      
Number of positive nodes Hazard ratio (HR) 1.05 95% CI 1.03–1.08 0.001 0.004 1.04 (1.01–1.07)

4. Discussion

Using the extended MelaCARTA panel, we assessed 46 oncogenic mutations across 134 stage IIIB/C melanoma patients who underwent TLND. The MelaCARTA panel reliably detects mutations in genomic DNA obtained from FFPE tissue [30] and our dataset had a failure rate of 7%. We identified mutations in 17 genes in 75% of patients, in keeping with the highly mutated nature of melanoma genomes [7] and [31].

Three molecular pathways are frequently dysregulated in melanoma: RAS-RAF-MEK-ERK; p16-CDK4-RB; and ARF-p53 [32] . We identified BRAF mutations in 46% of melanoma-positive lymph nodes, consistent with other reports of metastatic melanoma that include stage III disease [8], [17], [23], and [27]. While the majority were BRAF V600E mutations, V600K and other BRAF mutations (19% and 12%, respectively) were present at rates higher than reported by Colombino [23] , but similar to those reported in Australian patients [17] . While previous reports by others have identified differences between BRAF V600E and other BRAF mutations [33] , this did not significantly affect survival in our cohort. We found NRAS mutations in 21% of patients, in keeping with the reported 15–29% [23] and [24]. BRAF and NRAS mutations were mutually exclusive, as reported previously [7] . Consistent with studies in stage III and IV melanoma, patients with BRAF mutations tended to be younger with truncal primaries [17], [24], and [33].

We found both RAC1 and PPP6C mutations in 5% of patients, with other mutations identified in three or less patients. These data are consistent with previous sequencing studies in cutaneous melanoma [7], [9], and [10], excepting ERBB4 (up to 19%) and PTK2B (up to 10%) [34] . Discrepancies are probably explained by our use of the MelaCARTA panel which screens only mutation hotspots, not entire genes. Overall, differences in mutation rates between studies are small and likely reflect variability in methodology, patient population or sample type, as well as chance.

The high 3-year recurrence rate (68%) in our cohort was similar to the 65% recurrence rate reported by the ANZMTG 01.02/TROG 02.01 trial, reflecting our selection of a high-risk patient cohort [6] . The majority (63%) of first recurrence was distant, similar to other studies of stage III melanoma [35] and [36]. Our data, and those of others, found that most recurrences occurred within 2 years of lymphadenectomy, suggesting that our median follow-up of 30 months is unlikely to have underestimated recurrence or survival.

We found that patients whose tumours harbour a BRAF mutation have poor RFS. Regional nodal recurrence was rarely an isolated event with almost all first recurrences involving distant sites, suggesting metastasis occurs early among BRAF mutation positive patients. These findings strongly support the need for adjuvant systemic therapy for patients with BRAF mutations rather than aggressive local management with radiation therapy. In contrast, patients with BRAF wild-type tumours exhibit much better RFS and DSS, yet isolated local or regional recurrence may still be problematic for patients whose tumours are BRAF wild-type but have other high risk features (e.g. extracapsular invasion, >four involved lymph nodes or largest node >3 cm).

RFS and DSS were significantly associated with number of positive nodes and BRAF mutation status. Extracapsular invasion has previously been shown to be an independent predictor of DSS and regional nodal relapse [6] and [37], but was only a univariate prognostic factor for DSS in this study. Unlike other retrospective studies of stage III melanoma, we did not find primary tumour characteristics [36] and [38], age [39] or gender [36], [38], and [40] to be significant prognostic factors. This may be related to study populations: we included higher risk stage IIIB and IIIC melanoma, while other studies included patients with microscopic lymph node disease where primary tumour characteristics may still have prognostic significance. The number of positive lymph nodes has consistently been shown to be the most significant prognostic factor in stage III melanoma [4], [6], [37], and [39]. In the present study we found that the presence of a BRAF mutation in resected lymph node tissue was an independent prognostic factor for both RFS and DSS.

Although the number of published studies is small, the prognostic significance of BRAF mutation status for stage III melanoma has yielded conflicting results [24] and [25]. Two French studies of 105 patients with resected stage III melanoma [25] and 72 sentinel node-positive patients [41] , respectively, found BRAF V600E and V600K mutations, and number of positive nodes, to be independent prognostic factors for overall and distant metastasis-free-survival. In contrast, a Scandinavian study only found a trend towards worse survival for BRAF mutation status from the time of first non-distant metastasis [24] . However, 60% of patients in that study had only one positive node and some patients did not have nodal recurrence, suggesting that different patient populations may account, in part, for the different result. For newly-diagnosed stage IV melanoma patients, Long and colleagues [17] found BRAF mutation to be an adverse prognostic factor, though the interval from diagnosis of first-ever melanoma to distant metastasis was not significantly related to BRAF status. While these findings argue that factors other than BRAF mutations drive progression from primary melanoma, a broader group of stage I-III melanoma was not studied by the authors, making conclusions about the importance of BRAF mutations in early stage melanoma difficult. Of note, NRAS mutations were not prognostic in our study nor that of Ekedahl and colleagues [24] , unlike Jakob et al. who found NRAS mutation had prognostic significance in stage IV melanoma [42] . BRAF mutation status in resected lymph nodes for stage III melanoma was a strong prognostic factor in our study, arguing for consideration of BRAF status in the clinical care of Stage III melanoma patients and its inclusion in future iterations of the AJCC staging system. Stage III patients with BRAF mutations experienced rapid progression with locoregional recurrence rarely seen in isolation, suggesting that strategies aimed at adjuvant systemic therapy, e.g. BRAF/MEK inhibitors, should be considered for such patients before other local treatments.

Conflict of interest statement

None declared.


This work was supported by the Princess Alexandra Hospital Foundation and the National Health & Medical Research Council (NHMRC). NKH is supported by a Senior Principal Research Fellowship from the NHMRC. We thank Dr. Victoria Atkinson for contribution of clinical data for this study.

Study sponsors had no role in the design, execution or reporting of this study.

Appendix A. Supplementary data


Download file

Supplementary data 1 Supplementary methods.

Download file

Supplementary data 2 Supplementary Tables and figures.


  • [1] A. Calabro, S.E. Singletary, C.M. Balch. Patterns of relapse in 1001 consecutive patients with melanoma nodal metastases. Arch Surg. 1989;124(9):1051-1055 Crossref
  • [2] D.G. Coit, A. Rogatko, M.F. Brennan. Prognostic factors in patients with melanoma metastatic to axillary or inguinal lymph nodes. A multivariate analysis. Ann Surg. 1991;214(5):627-636 Crossref
  • [3] R.J. Lee, J.F. Gibbs, G.M. Proulx, D.R. Kollmorgen, C. Jia, W.G. Kraybill. Nodal basin recurrence following lymph node dissection for melanoma: implications for adjuvant radiotherapy. Int J Radiat Oncol Biol Phys. 2000;46(2):467-474 Crossref
  • [4] C.M. Balch, J.E. Gershenwald, S.J. Soong, J.F. Thompson, M.B. Atkins, D.R. Byrd, et al. Final version of 2009 AJCC melanoma staging and classification. J Clin Oncol. 2009;27(36):6199-6206 Crossref
  • [5] S. Mocellin, S. Pasquali, C.R. Rossi, D. Nitti. Interferon alpha adjuvant therapy in patients with high-risk melanoma: a systematic review and meta-analysis. J Natl Cancer Inst. 2010;102(7):493-501 Crossref
  • [6] B.H. Burmeister, M.A. Henderson, J. Ainslie, R. Fisher, J. Di Iulio, B.M. Smithers, et al. Adjuvant radiotherapy versus observation alone for patients at risk of lymph-node field relapse after therapeutic lymphadenectomy for melanoma: a randomised trial. Lancet Oncol. 2012;13(6):589-597 Crossref
  • [7] E. Hodis, I.R. Watson, G.V. Kryukov, S.T. Arold, M. Imielinski, J.P. Theurillat, et al. A landscape of driver mutations in melanoma. Cell. 2012;150(2):251-263 Crossref
  • [8] X. Wei, V. Walia, J.C. Lin, J.K. Teer, T.D. Prickett, J. Gartner, et al. Exome sequencing identifies GRIN2A as frequently mutated in melanoma. Nat Genet. 2011;43(5):442-446 Crossref
  • [9] S.I. Nikolaev, D. Rimoldi, C. Iseli, A. Valsesia, D. Robyr, C. Gehrig, et al. Exome sequencing identifies recurrent somatic MAP2K1 and MAP2K2 mutations in melanoma. Nat Genet. 2012;44(2):133-139
  • [10] M. Krauthammer, Y. Kong, B.H. Ha, P. Evans, A. Bacchiocchi, J.P. McCusker, et al. Exome sequencing identifies recurrent somatic RAC1 mutations in melanoma. Nat Genet. 2012;44(9):1006-1014 Crossref
  • [11] M.S. Stark, S.L. Woods, M.G. Gartside, V.F. Bonazzi, K. Dutton-Regester, L.G. Aoude, et al. Frequent somatic mutations in MAP3K5 and MAP3K9 in metastatic melanoma identified by exome sequencing. Nat Genet. 2012;44(2):165-169
  • [12] A. Gast, D. Scherer, B. Chen, S. Bloethner, S. Melchert, A. Sucker, et al. Somatic alterations in the melanoma genome: a high-resolution array-based comparative genomic hybridization study. Genes Chromosom Cancer. 2010;49(8):733-745 Crossref
  • [13] J.A. Curtin, J. Fridlyand, T. Kageshita, H.N. Patel, K.J. Busam, H. Kutzner, et al. Distinct sets of genetic alterations in melanoma. N Engl J Med. 2005;353(20):2135-2147 Crossref
  • [14] E. Hacker, N.K. Hayward, T. Dumenil, M.R. James, D.C. Whiteman. The association between MC1R genotype and BRAF mutation status in cutaneous melanoma: findings from an Australian population. J Invest Dermatol. 2010;130(1):241-248 Crossref
  • [15] W. Liu, J.W. Kelly, M. Trivett, W.K. Murray, J.P. Dowling, R. Wolfe, et al. Distinct clinical and pathological features are associated with the BRAF(T1799A(V600E)) mutation in primary melanoma. J Invest Dermatol. 2007;127(4):900-905 Crossref
  • [16] A. Viros, J. Fridlyand, J. Bauer, K. Lasithiotakis, C. Garbe, D. Pinkel, et al. Improving melanoma classification by integrating genetic and morphologic features. PLoS Med. 2008;5(6):e120 Crossref
  • [17] G.V. Long, A.M. Menzies, A.M. Nagrial, L.E. Haydu, A.L. Hamilton, G.J. Mann, et al. Prognostic and clinicopathologic associations of oncogenic BRAF in metastatic melanoma. J Clin Oncol. 2011;29(10):1239-1246 Crossref
  • [18] S.J. Gallagher, J.F. Thompson, J. Indsto, L.L. Scurr, M. Lett, B.F. Gao, et al. p16INK4a expression and absence of activated B-RAF are independent predictors of chemosensitivity in melanoma tumors. Neoplasia. 2008;10(11):1231-1239
  • [19] A. Gorden, I. Osman, W. Gai, D. He, W. Huang, A. Davidson, et al. Analysis of BRAF and N-RAS mutations in metastatic melanoma tissues. Cancer Res. 2003;63(14):3955-3957
  • [20] M. Kirschner, B. Helmke, H. Starz, A. Benner, M. Thome, M. Deichmann. Preponderance of the oncogenic V599E and V599K mutations in the B-raf kinase domain is enhanced in melanoma lymph node metastases. Melanoma Res. 2005;15(5):427-434 Crossref
  • [21] R. Kumar, S. Angelini, K. Czene, I. Sauroja, M. Hahka-Kemppinen, S. Pyrhonen, et al. BRAF mutations in metastatic melanoma: a possible association with clinical outcome. Clin Cancer Res. 2003;9(9):3362-3368
  • [22] S. Ugurel, R.K. Thirumaran, S. Bloethner, A. Gast, A. Sucker, J. Mueller-Berghaus, et al. B-RAF and N-RAS mutations are preserved during short time in vitro propagation and differentially impact prognosis. PLoS ONE. 2007;2(2):e236 Crossref
  • [23] M. Colombino, M. Capone, A. Lissia, A. Cossu, C. Rubino, V. De Giorgi, et al. BRAF/NRAS mutation frequencies among primary tumors and metastases in patients with melanoma. J Clin Oncol. 2012;30(20):2522-2529 Crossref
  • [24] H. Ekedahl, H. Cirenajwis, K. Harbst, A. Carneiro, K. Nielsen, H. Olsson, et al. The clinical significance of BRAF and NRAS mutations in a clinic-based metastatic melanoma cohort. Br J Dermatol. 2013;169(5):1049-1055 Crossref
  • [25] S. Moreau, P. Saiag, P. Aegerter, D. Bosset, C. Longvert, Z. Helias-Rodzewicz, et al. Prognostic value of BRAF(V600) mutations in melanoma patients after resection of metastatic lymph nodes. Ann Surg Oncol. 2012;19(13):4314-4321 Crossref
  • [26] G.J. Mann, G.M. Pupo, A.E. Campain, C.D. Carter, S.J. Schramm, S. Pianova, et al. BRAF mutation, NRAS mutation, and the absence of an immune-related expressed gene profile predict poor outcome in patients with stage III melanoma. J Invest Dermatol. 2013;133(2):509-517 Crossref
  • [27] P.B. Chapman, A. Hauschild, C. Robert, J.B. Haanen, P. Ascierto, J. Larkin, et al. Improved survival with vemurafenib in melanoma with BRAF V600E mutation. N Engl J Med. 2011;364(26):2507-2516 Crossref
  • [28] A. Hauschild, J.J. Grob, L.V. Demidov, T. Jouary, R. Gutzmer, M. Millward, et al. Dabrafenib in BRAF-mutated metastatic melanoma: a multicentre, open-label, phase 3 randomised controlled trial. Lancet. 2012;380(9839):358-365 Crossref
  • [29] K.T. Flaherty, J.R. Infante, A. Daud, R. Gonzalez, R.F. Kefford, J. Sosman, et al. Combined BRAF and MEK inhibition in melanoma with BRAF V600 mutations. N Engl J Med. 2012;367(18):1694-1703 Crossref
  • [30] K. Dutton-Regester, H. Kakavand, L.G. Aoude, M.S. Stark, M.G. Gartside, P. Johansson, et al. Melanomas of unknown primary have a mutation profile consistent with cutaneous sun-exposed melanoma. Pigment Cell Melanoma Res. 2013;26(6):852-860 Crossref
  • [31] L.B. Alexandrov, S. Nik-Zainal, D.C. Wedge, S.A. Aparicio, S. Behjati, A.V. Biankin, et al. Signatures of mutational processes in human cancer. Nature. 2013;500(7463):415-421 Crossref
  • [32] K. Dutton-Regester, N.K. Hayward. Reviewing the somatic genetics of melanoma: from current to future analytical approaches. Pigment Cell Melanoma Res. 2012;25(2):144-154 Crossref
  • [33] A.M. Menzies, L.E. Haydu, L. Visintin, M.S. Carlino, J.R. Howle, J.F. Thompson, et al. Distinguishing clinicopathologic features of patients with V600E and V600K BRAF-mutant metastatic melanoma. Clin Cancer Res. 2012;18(12):3242-3249 Crossref
  • [34] T.D. Prickett, N.S. Agrawal, X. Wei, K.E. Yates, J.C. Lin, J.R. Wunderlich, et al. Analysis of the tyrosine kinome in melanoma reveals recurrent mutations in ERBB4. Nat Genet. 2009;41(10):1127-1132 Crossref
  • [35] E. Romano, M. Scordo, S.W. Dusza, D.G. Coit, P.B. Chapman. Site and timing of first relapse in stage III melanoma patients: implications for follow-up guidelines. J Clin Oncol. 2010;28(18):3042-3047 Crossref
  • [36] R.R. White, W.E. Stanley, J.L. Johnson, D.S. Tyler, H.F. Seigler. Long-term survival in 2,505 patients with melanoma with regional lymph node metastasis. Ann Surg. 2002;235(6):879-887 Crossref
  • [37] S.P. Prens, A.P. van der Ploeg, A.C. van Akkooi, C.A. van Montfort, A.N. van Geel, J.H. de Wilt, et al. Outcome after therapeutic lymph node dissection in patients with unknown primary melanoma site. Ann Surg Oncol. 2011;18(13):3586-3592
  • [38] C.I. Amos, L.E. Wang, J.E. Lee, J.E. Gershenwald, W.V. Chen, S. Fang, et al. Genome-wide association study identifies novel loci predisposing to cutaneous melanoma. Hum Mol Genet. 2011;20(24):5012-5023 Crossref
  • [39] S.E. Young, S.R. Martinez, M.B. Faries, R. Essner, L.A. Wanek, D.L. Morton. Can surgical therapy alone achieve long-term cure of melanoma metastatic to regional nodes?. Cancer J (Sudbury, Mass). 2006;12(3):207-211 Crossref
  • [40] A. Joosse, S. Collette, S. Suciu, T. Nijsten, P.M. Patel, U. Keilholz, et al. Sex is an independent prognostic indicator for survival and relapse/progression-free survival in metastasized stage III to IV melanoma: a pooled analysis of five European organisation for research and treatment of cancer randomized controlled trials. J Clin Oncol. 2013;31(18):2337-2346 Crossref
  • [41] M. Picard, N. Pham Dang, M. D’Incan, S. Mansard, P. Dechelotte, B. Pereira, et al. Is BRAF a prognostic factor in stage III skin melanoma? A retrospective study of 72 patients after positive sentinel lymph node dissection. Br J Dermatol. 2014;10.1111/bjd.12939 [Epub ahead of print]
  • [42] J.A. Jakob, R.L. Bassett Jr., C.S. Ng, J.L. Curry, R.W. Joseph, G.C. Alvarado, et al. NRAS mutation status is an independent prognostic factor in metastatic melanoma. Cancer. 2012;118(16):4014-4023 Crossref


a Surgical Oncology Group, School of Medicine, The University of Queensland, Translational Research Institute, Woolloongabba, QLD, Australia

b Queensland Melanoma Project, Discipline of Surgery, The University of Queensland, Princess Alexandra Hospital, Woolloongabba, QLD, Australia

c QIMR Berghofer Medical Research Institute, Oncogenomics Laboratory, Brisbane, QLD, Australia

d Department of Pathology, Princess Alexandra Hospital, Woolloongabba, QLD, Australia

lowast Corresponding author at: School of Medicine, The University of Queensland, Translational Research Institute, Woolloongabba, QLD 4102, Australia. Tel: +61 738477133; fax: +61 733465598.