Detection of mutant BRAF alleles in the plasma of patients with metastatic melanoma

Abstract

Mutations in the BRAF oncogene at amino acid 600 have been reported in 40 to 70% of human metastatic melanoma tissues, and the critical role of BRAF in the biology of melanoma has been established. Sampling the blood compartment to detect the mutational status of a solid tumor represents a highly innovative advance in cancer medicine, and such an approach could have advantages over tissue-based techniques. We report the development of a fluorescence-based polymerase chain reaction (PCR) assay to detect mutant BRAF alleles in plasma. A mutant-specific PCR assay was optimized to specifically amplify the mutant BRAF allele without amplifying the wild-type allele. Experiments mixing DNA from a BRAF mutant melanoma cell line with wild-type human placental DNA in varying proportions were performed to determine the threshold of this assay and to compare it with routine DNA sequencing. The assay was then applied to tissue and plasma specimens from patients with metastatic melanoma. The assay detected 0.1 ng of mutant DNA mixed in 100 ng of wild-type DNA and was 500-fold more sensitive than DNA sequencing. The assay detected mutant BRAF alleles in plasma samples from 14 of 26 (54%) metastatic melanoma patients. These data demonstrate the feasibility of blood-based testing for BRAF mutations in metastatic melanoma patients.

Figure 1

Sensitivity and specificity of the MS-PCR assay. a: Annealing temperature gradient. Annealing temperatures between 35 and 64°C were tested. The amplified products were analyzed using an ABI 310 Genetic Analyzer, and the resulting electropherograms are presented. In the top row, amplification of templates containing the BRAF mutation is markedly reduced or absent at annealing temperatures of 63°C or higher. At all temperatures, the reaction failed to amplify the wild-type control (bottom row). The mutant template consisted of DNA purified from the BRAF mutant melanoma cell line SK-MEL 29; the wild-type template consisted of human placental DNA. The y axis is relative fluorescence units. b: Mixing experiment. The specified quantities of DNA from the BRAF mutant cell line SK-MEL 29 were mixed with human placental DNA (wild-type BRAF) to a sum of 100 ng of DNA. The mixture was used as a template for the MS-PCR. The limit of detection for mutant BRAF alleles using this assay was 0.1 ng of cell line DNA. c: Dilution experiment. Normal blood samples (5 ml each) were spiked with decreasing numbers of SK-MEL 29 cells, mutant for BRAF. Peripheral blood mononuclear cells were isolated, and DNA was extracted. The lowest concentration for which a peak is present is 10 cells/ml. Mutant cells were undetectable at concentrations below this and in the negative control human placental DNA (HPDNA). d: Negative controls. Ten peripheral blood samples from normal volunteers were analyzed by MS-PCR. None of the samples had detectable mutant BRAF alleles (top panel). In the bottom panel, the entire BRAF exon 15 was amplified (as expected) from all volunteer samples and from human placental (HPDNA) and SK-MEL 29 control DNA samples.

Figure 2

Sensitivity of MS-PCR versus sequencing. The sensitivity of the MS-PCR technique is compared with DNA sequencing by testing purified DNA containing 0 to 100% mutant BRAF DNA. In the two top rows, SeqScape software (Applied Biosystems) was used to identify mutant peaks (A) in place of wild type (T) (W denotes the presence of both peaks). BRAF mutations are detected when at least 50% of the sample DNA is derived from mutant tumor cells. At concentrations of 40% or less mutant DNA, conventional sequencing is unable to confidently detect the mutation. In the third row, MS-PCR is able to detect mutations with as little as 10% mutant BRAF DNA. The y axis is relative fluorescence units.

Figure 3

Detection of mutant BRAF alleles in plasma samples from patients with metastatic melanoma. Representative electropherogram results from 10 patients with metastatic melanoma. MS-PCR results are shown in the top panel; control amplification of the entire BRAF exon 15 is shown in the bottom panel. In this group, 7 of 10 (70%) patients had detectable BRAF mutations in their plasma samples. The placental DNA samples were used as controls and were negative for the mutant BRAF and positive for BRAF exon 15, as expected. The y axes are relative fluorescence units.