Introduction: Serial monitoring of sensitive and specific blood-based biomarkers of disease burden in cancer patients could enable clinicians to adjust treatments at the earliest signs of disease progression when the likelihood of improved outcomes is probably greatest. Such assays require preanalytical, analytical, and clinical validation prior to use. We conducted analytical validation studies of 7 droplet digital PCR (ddPCR) assays to determine their performance characteristics in detecting mutant TERT, BRAF, and NRAS ctDNA.

Methods: Probe-based ddPCR assays (Bio-Rad) detecting BRAF V600E, K, NRAS Q61K, R, L, and TERT C228T, C250T were studied. Preanalytical validation experiments using plasma collected in EDTA tubes investigated processing variables including plasma volumes, centrifugation speeds, and DNA extraction efficiencies, and evaluated the necessity of internal controls using spike-in experiments. Analytical validation studies assessed accuracy using plasma from healthy donors spiked with known amounts of mutant nucleosomal DNAs from control cell lines. Potentially interfering substances were evaluated using the AcroMetrix inhibition panel (Thermo Scientific). Precision/reproducibility was assessed using combinations of multiple reagent lots and operators. Assay specificity was determined using templates with different mutations to check cross-reactivity. Limits of Blank (LoB) and Limits of Detection (LoD) were calculated for each assay using 8 replicate wells of healthy donor plasma or mutant cell line DNA, respectively. Three-phase inter-laboratory harmonization was conducted using: 1) extracted DNA, 2) plasmas spiked with control DNAs, and 3) melanoma patient samples shipped in PAXgene Blood ccfDNA tubes (Qiagen).

Results: Preanalytical validation yielded a double centrifugation protocol to reduce contaminating large DNA. Assays were highly accurate and precise with 80-90% recovery of spiked-in mutant templates and coefficients of variation (CVs) <10%, respectively. BRAF and NRAS mutation-specific assays also detected alternative specific mutations in their respective targets that were easily discernible from the perfect-match mutation as discrete clusters in the ddPCR output software. The LoBs ranged from 0.24 to 1.68 copies/mL and 0.016% to 0.058% fractional abundance; LoDs ranged from 0.01% to 0.05%. Inter-laboratory harmonization demonstrated CVs ranging from 1.5% to 15%.

Conclusion: This ddPCR-based plasma analysis platform demonstrates low background levels in normal plasma, yielding high sensitivities to detect small amounts of ctDNAs with these mutations. The assays also demonstrate highly reproducible results between laboratories. These assays could be integrated into biomarker-driven clinical trials to evaluate their potential to inform more personalized clinical decision-making.

Citation Format: Mahrukh M. Syeda, Broderick Corless, Iman Osman, Cindy Spittle, George Karlin-Neumann, David Polsky. Analytical validation of 7 droplet digital PCR assays detecting TERT, BRAF, and NRAS hotspot mutations in plasma-derived circulating tumor DNA (ctDNA) [abstract]. In: Proceedings of the AACR Special Conference on Advances in Liquid Biopsies; Jan 13-16, 2020; Miami, FL. Philadelphia (PA): AACR; Clin Cancer Res 2020;26(11_Suppl):Abstract nr A66.