Comparison of the Amplification Refractory Mutation System, Super Amplification Refractory Mutation System, and Droplet Digital PCR for T790 M Mutation Detection in Non-small Cell Lung Cancer after Failure of Tyrosine Kinase Inhibitor Treatment.

Plasma mutation detection has the advantages of non-invasiveness and accessibility. Here, we evaluated three methods, the amplification refractory mutation system (ARMS), second-generation ARMS (SuperARMS), and droplet digital PCR (ddPCR), to assess their concordance and feasibility for the detection of mutations in plasma samples. Non-small lung cancer patients with stage IIIB/IV that were resistant to epidermal growth factor receptor-tyrosine kinase inhibitor (EGFR-TKI) treatment were enrolled. Blood samples were collected within 14 days after TKI resistance. Each sample was simultaneously assessed by the three methods. In total, 169 patients were enrolled; 54.4% were female, 72.2% were diagnosed with stage IV disease; and 97.6% had adenocarcinoma. T790 M mutations were detected in 42 (24.8%) of the 169 samples using ARMS, one of which carried the T790 M alone, 22 that also encoded exon 19 deletions, and 19 with L858R mutations. For the SuperARMS assay, 59 (34.9%) samples exhibited the T790 M mutation, and 110 (65.1%) showed no detectable T790 M mutation. ddPCR showed that 61 (36.1%) samples contained the T790 M mutation, whereas 108 (63.9%) were not positive. T790 M abundance ranged from 0.04% to 38.2%. The median T790 M abundance was 0.15% for total samples and 2.98% for T790 M mutation samples. The overall concordance was 78.7% (133/169) among ARMS, SuperARMS, and ddPCR. Compared with patients with stage III disease, patients with stage IV disease exhibited a higher T790 M mutation detection rate (28.7% vs. 14.9% by ARMS; 37.7% vs. 27.7% by SuperARMS; and 41.8% vs. 21.3% by ddPCR). Liquid biopsy showed promise and has the advantages of non-invasiveness and accessibility. T790 M detection based on circulating tumor DNA showed high concordance. Compared with non-digital platforms, ddPCR showed higher sensitivity and provided both frequency and abundance information, which might be important for treatment decisions.