A multisite blinded study for the detection of BRAF mutations in formalin-fixed, paraffin-embedded malignant melanoma

Abstract

Melanoma patients with BRAF mutations respond to treatment with vemurafenib, thus creating a need for accurate testing of BRAF mutation status. We carried out a blinded study to evaluate various BRAF mutation testing methodologies in the clinical setting. Formalin-fixed, paraffin-embedded melanoma samples were macrodissected before screening for mutations using Sanger sequencing, single-strand conformation analysis (SSCA), high resolution melting analysis (HRM) and competitive allele-specific TaqMan® PCR (CAST-PCR). Concordance of 100% was observed between the Sanger sequencing, SSCA and HRM techniques. CAST-PCR gave rapid and accurate results for the common V600E and V600K mutations, however additional assays are required to detect rarer BRAF mutation types found in 3-4% of melanomas. HRM and SSCA followed by Sanger sequencing are effective two-step strategies for the detection of BRAF mutations in the clinical setting. CAST-PCR was useful for samples with low tumour purity and may also be a cost-effective and robust method for routine diagnostics.

Figure 1. Representative results for BRAF mutation screening using Sanger sequencing (A–D), single strand conformation analysis (SSCA; E–H), high resolution melting analysis (HRM; I–L) and competitive allele-specific TaqMan (CAST-PCR; M-P).

The first column is an example of wildtype BRAF, the second column of V600E, the third column of V600K and the fourth of K601E. Blue arrows indicate the wildtype profile, while red arrows indicate mutant profiles. For SSCA, E represents the V600E mutation and K the V600K mutation. In HRM graphs (difference plots), red lines indicate positive controls, blue lines negative controls, green lines the sample and black lines a sample/wildtype mix. In CAST PCR, green lines indicate the reference assay, blue lines the V600E assay and orange lines the V600K assay. Slight cross reactivity of V600E positive samples with the V600K assay (orange line) can be seen (N), but not the reverse (O). CAST-PCR using probes for V600E and V600K did not detect the K601E mutation (P).

Figure 2. Results for sample P46 containing the double mutation c.1799_1801delinsAGG (p.Val600_Lys601delinsGluGlu).

This mutation was detected by CAST-PCR with the V600E probe (D), as well as by SSCA (B) and HRM (C).

Figure 3. Results for sample P33 showing cross reactivity between the V600K probe (orange line) and a V600E mutation (blue) in CAST-PCR (D).

In this case the dCt value for the V600K assay was lower than the threshold value of 9.96, however Sanger sequencing (A) clearly showed this to be a V600E mutation. The results for SSCA (B) and HRM (C) also suggest a single V600E mutation.

Figure 4. Results for sample P08 showing wildtype profiles (blue arrows) by sequencing (A), SSCA (B) and HRM (C).

Two separate runs of CAST-PCR (D, E) revealed a V600E mutation for this sample, with dCt values in each case of <9.96.

Figure 5. Two samples (P8 and P14) were positive for the BRAF V600E mutation using CAST-PCR but negative using SSCA, HRM and Sanger sequencing.

Shown are representative sequencing traces from one of the positive replicates for each of P8 and P14 using LCN-HRM. Red arrows indicate the BRAF V600E mutation.