Detection and Monitoring of BRAF and NRAS Mutant Clones in Myeloma Patients By Digital PCR of Circulating DNA

Introduction
Targeted mutation specific therapy is a promising approach in cancer therapy. However, an obstacle for this approach is the vast heterogeneity of the clonal composition and development. Tumor biopsies represent only a snapshot of the situation. Furthermore, monitoring of the clonal development is difficult because biopsies may not be representative for the whole tumor and availability of repeat biopsies is limited. To meet these difficulties we have established and optimized a method based on Digital PCR (dPCR) for analyses of circulating cell free (cf)DNA from sequential samples of serum and plasma from patients with multiple myeloma.

Methods
We investigated 19 patients for the BRAF V600E mutation. Nine were previously confirmed as mutation positive in bone marrow biopsies/purfied plasma cells by two independent methods (PCR/immunohistochemistry/whole exome sequencing) whereas 10 were mutation negative (Rustad et al Blood Cancer J 2015). Two patients with NRAS Q61K mutation detected in serial bone marrow samples were also included. In total, 67 serum and 21 EDTA-plasma samples were analyzed. Blood samples were taken, processed and frozen at -80⁰ C within 1,5 hour. The samples were stored for a median of 5 years (range 0-23) before DNA isolation and analysis. Mutation detection by dPCR was performed using a droplet-based system and validated primer/probe-sets (BioRad). In-house validation and optimization of the assay was carried out using cancer cell lines OH2 and HT29 with NRAS Q61 and BRAF V600E mutations respectively.  The limit of detection was 1-3 copies of mutated DNA per reaction and no false positives were detected. The threshold of positivity was set to 1 droplet per sample. Experiments were performed in accordance with the Minimum Information for Publication of Digital PCR Experiments (dMIQE) guidelines (Huggett et al Clin Chem 2013).

Results
BRAF or NRAS mutated cfDNA was detected in all patients with a confirmed mutation in tumor tissue, and in none of the mutation-negative controls (p = 0.000003, Fisher's exact test). When looking only at tumor tissue and blood samples obtained at the same time, mutation positivity was confirmed in the blood of 9/10 patients (p = 0.00012). Furthermore, there was a positive correlation between the percentage of mutated plasma cells in bone marrow biopsies and the concentration of mutated cfDNA  (Spearman correlation R = 0.63, p = 0.025).

Serial samples were analyzed from 5 patients and provided information about 3 different aspects:

1. Patients 1 (figure), 2 and 3, had large clones (50-100 %) of BRAF or NRAS mutated cells in diagnostic and relapse bone marrow samples. Mutated cfDNA correlated closely to M-protein levels in these patients as demonstrated in the figure. A corollary of the figure is that the BRAF mutated clone produces M protein and is sensitive to MP.

2. Patient 4 developed a pelvic extra medullary plasmacytoma with 75-100% BRAF mutation positive cells (immunohistochemistry), however, time-matched serum samples showed only a modest peak with 23 mutated copies/ml.

3. Patient 5 had a moderately sized BRAF V600E mutated clone of 50-75 % at diagnosis, which, according to serum levels, persisted through the disease course. However, two months prior to death, the patient rapidly deteriorated and became refractory to treatment. BM aspirate showed 95 % plasma cells with plasmablastic morphology. A serum sample contained > 600 ng/ml of cfDNA, 10-100 fold more than any other sample in our study, and was highly positive for BRAF V600E mutation (59 000 copies/ml). The patient clearly had expansion of an aggressive BRAF mutated clone that could easily be detected by serum analysis.

Figure: Co-variation of M-protein and circulating BRAF V600E mutated DNA in patient 1.

Conclusions
This study demonstrates that mutations such as BRAF V600E and NRAS Q61K can be reliably detected and monitored in sequential serum or plasma samples from myeloma patients.  Quantitative mutation analysis compared to M protein in sequential samples provided significant information with clinical relevance. The great advantage of this approach is the easy access to blood samples compared to bone marrow aspirate/biopsy. This will facilitate studies of clonal development during treatment and detection of druggable mutations.