Plasma Cell-Free DNA Testing of Patients With EGFR Mutant Non-Small-Cell Lung Cancer: Droplet Digital PCR Versus Next-Generation Sequencing Compared With Tissue-Based Results

Purpose: To compare the results of plasma cell-free DNA (cfDNA) droplet digital PCR (ddPCR) and next-generation sequencing (NGS) on detection of epidermal growth factor receptor () primary activating mutations and p.T790M with results of tissue analysis in patients with mutated non-small-cell lung cancer.

Methods: All patients with EGFR mutated non-small cell lung cancer for which a pathology and a plasma specimen were available upon progression between November 2016 and July 2018 were selected. Concordance, Cohen's κ, and intraclass correlation coefficients were calculated.

Results: Plasma cfDNA and pathology specimens of 36 patients were analyzed. Agreement between ddPCR and NGS was 86% (κ = 0.63) for the primary activating mutation and 94% (κ = 0.89) for the p.T790M detection. Allele ratios were comparable, with an intraclass correlation coefficient of 0.992 and 0.997, respectively. Discrepancies of some degree were found in 15 patients (41.7%). In six patients (16.7%), no mutations were detected in cfDNA. In three patients (8.3%), p.T790M was detected in plasma but not in the pathology specimen, whereas in three other patients (8.3%), p.T790M was demonstrated in the pathology specimen but not in plasma. Concordance of cfDNA and pathology for the primary activating mutation was 69% for ddPCR and 83% for NGS. For the detection of p.T790M, this was 75% (κ = 0.49) for ddPCR as well as for NGS.

Conclusion: Mutual agreement is high between NGS and ddPCR in cfDNA on the level of a specific mutation, with comparable ratio results. Plasma testing of primary activating mutations and p.T790M shows high concordance with pathology results, for NGS as well as for ddPCR, depending on the extent of the panel used. In NGS, more genetic aberrations can be investigated at once.