Sir,

We read with interest the recent publication by Heinzerling et al (2013), demonstrating intra-patient heterogeneity of BRAF mutation status between tumours in 10 of 53 (18.9%) patients. However, we have great concern that the results of the study may reflect the (less than 100%) sensitivity of the molecular techniques employed and/or an incorrect assumption that the primary melanoma was the source of the metastatic disease rather than true intra-patient BRAF heterogeneity.

Potentially, the results of the study by Heinzerling et al could have tremendous clinical importance, as accurate determination of a patient’s melanoma BRAF status is critical when planning treatment for melanoma patients with advanced stage disease. Targeting the mitogen-activated protein kinase (MAPK) pathway in patients with BRAF-mutant metastatic melanoma has vastly improved clinical outcomes; however, BRAF inhibitors may paradoxically activate the MAPK pathway in wild-type BRAF melanomas and therefore adversely affect survival if such patients are treated with BRAF inhibitors. Thus, if intra-patient melanoma BRAF heterogeneity exists and treatment decisions are made on the basis of mutation assessment of a single tumour, potentially effective treatment may not be offered in a significant proportion of patients, or alternatively, treatment may be administered that is potentially detrimental.

Although the results of the study by Heinzerling et al are in keeping with other recent reports of heterogeneity in 15% and 13.5% of patients (Colombino et al, 2012; Saint-Jean et al, 2013), two recent studies (Boursault et al, 2013; Menzies et al, 2013) demonstrated very little heterogeneity of BRAF status within metastatic melanoma patients. Several factors may have influenced the results of these studies. First, the techniques used to determine BRAF status were different in the ‘higher’ and ‘lower’ discordance studies. The latter studies used a highly sensitive and specific immunohistochemical technique (the anti-BRAFV600E VE1 antibody) that enables determination of the BRAF status in all individual cells by direct visualisation and at the same time confirmation that they are in fact tumour cells. This technique is not reliant on a certain percentage of tumour cells being present. In contrast, the former studies used molecular methods such as pyrosequencing, allele-specific PCR, and Sanger sequencing, all of which may have false-negative results when samples contain low tumour content. A recent study highlighted the problem of false-negative mutation tests by molecular techniques. Discordant BRAFV600E status was identified in 5 of 97 specimens; subsequent molecular retesting both confirmed an initial molecular misdiagnosis in 4 of the 5 cases and the greater accuracy of BRAF protein immunohistochemistry (Long et al, 2013).

Another factor that may have resulted in heterogeneity is the assumption that any given primary melanoma is the culprit tumour from which the metastatic disease was derived. Ten per cent of patients with metastatic melanoma have a history of multiple primary melanomas (Murali et al, 2012). Even in patients with a history of only a single known primary melanoma, sometimes the site of locoregional metastasis is not in keeping with the T-stage or site of the presumed primary melanoma, or it does not occur within a plausible time period, suggesting that an occult primary melanoma may have led to the metastatic disease. In this situation, close scrutiny of a patient’s clinical history is required to ensure accurate assignment of the ‘culprit’ primary melanoma (Murali et al, 2012).

Clinical responses observed in patients treated with BRAF inhibitors do not support the suggestion of intra-patient BRAF heterogeneity as all metastases have a uniform initial metabolic response to BRAF inhibition assessed using FDG-PET imaging (McArthur et al, 2012), and all resistant lesions resected from patients still contain mutant BRAF (McArthur et al, 2011; Poulikakos et al, 2011; Van Allen et al, 2013).

Further clinical studies are required to examine the issue of intra-patient discordance of BRAF. Carefully assigning primary melanomas as culprit lesions, and using accurate BRAF testing methods with adequate tumour cell content would be the requirements to underpin the data.