Early prediction of resistance to tyrosine kinase inhibitors by plasma monitoring of EGFR mutations in NSCLC: a new algorithm for patient selection and personalized treatment.

In Non-Small-Cell Lung Cancer (NSCLC) patients treated with Tyrosine Kinase-Inhibitors (TKIs) therapy, the emergence of acquired resistance can be investigated by plasma monitoring of circulating tumor DNA (ctDNA). A series of 116 patients with EGFR-positive lung adenocarcinomas were treated with first/second generation EGFR TKIs. At clinical progression, 64 (55%) EGFR T790M plasma positive patients were subjected to second line-treatment with osimertinib and strictly monitored during the first month of therapy. Plasma analysis by the EGFR Cobas test showed in 57 (89%) cases a substantial decrease in the levels of the sensitizing EGFR mutant allele (sEGFRma), down to a not detectable value. These patients were defined as plasmatic good responders (PGR). In 7 (11%) patients, the sEGFRma did not drop to zero (plasmatic poor responders, PPR). In these latter cases, Massive Parallel Sequencing (MPS) analysis at the end of the first month and at clinical progression showed the presence of resistant-inducing mutations, including MET and HER2 gene amplification, KRAS and PIK3CA gene mutations. PPR showed disease progression in 5 (71%) cases, stable disease in 2 (29%) cases, and a shorter median Progression-free survival (PFS) (4.3 ± 1.1 months) than that observed in PGR (13.3 ± 1.2 months) (P < 0.0001). Our data indicate that plasma monitoring by a simple RT-PCR-based EGFR mutation test in the first month of treatment may be useful for a rapid identification of patients to be subjected to further characterization by MPS. A diagnostic algorithm for an early detection of resistance-inducing mutations and patient management is reported.