The prognostic role of the
The prognostic role of the ratio between median AF T790M to median AF of was also investigated, confirming the results obtained by other authors, that a higher ratio is associated with deeper tumor shrinkage and better survival [21,23,24]. In fact, the ratio of T790M-to- could serve as quantification of the T790M-positive clones in relation to other resistant clones still carrying the but no T790M, providing a more accurate stratification with respect to absolute level of T790M. Moreover, we observed a trend towards better outcome in patients with clearance of plasmatic mutations during osimertinib treatment, as recently confirmed in literature [25,26]. Only two of the evaluable patients presented both and T790M LLY507 at the first plasma sample collected during osimertinib. Therefore, the trend observed in our population could have been mainly driven by differences between patients with total clearance and patients with persistence of . Similarly to us, Oxnard et al observed that the pattern of tissue resistance could not be predicted by T790M plasmatic behaviour neither at baseline nor during therapy. On the contrary, they showed a different plasmatic behaviour of in the two groups of patients with T790M-loss or persistence at resistance, observing that patients with T790M-loss are more likely to present lower decrease of the , representing sub-clones less sensitive to osimertinib and candidate to drug resistance. Thus, our observation of better outcome in patients with complete clearance finds support in emerging literature even if, due to the small number of patients, we were not able to confirm the predictive role for the tissue resistance mechanisms. To study mechanisms of resistance, tissue re-biopsy at progression to osimertinib was performed in 11 patients. Even though the small number of patients analysed and the limitations due to different platforms used for NGS analyses, qualitative considerations could be taken. Recently, Papadimitrakopoulou et al presented data about resistance mechanisms in patients enrolled in the AURA3 trial well summarizing the complex scenario . In our cohort, 8 out of 11 patients (72%) lost T790M and this observation slightly higher than the percentage described in literature (34–68%) in the most recent cohorts analysed [10,, , , , ]. Loss of T790 M at progression was associated with worse outcome . In our cohort the difference, in terms of PFS, between two subgroups lost vs. maintained T790M was not statistically significant. Resistance to osimertinib by reactivation of EGFR pathway through tertiary mutations was observed in 3 patients that showed the emergence of EGFR C797S (27%). Activation of known by-pass signalling pathways as mechanism of resistance was represented in our cohort by MET (1 patient, 9%) and MYC amplification (1 patient, 9%), and PTEN loss (1 patient, 9%). MET amplification, observed in our study in only one case with loss of T790M, has been described in the literature at frequencies ranging 5–50% [10,, , , , ]. PTEN loss was already mentioned as a possible mechanism of resistance to osimertinib , while MYC amplification is described here, for the first time to best of our knowledge, as potential mechanism of intrinsic resistance. In our cohort, a higher percentage of SCLC was observed (3 patients, 27%) than previously reported in other studies on the same topic (4–15%) [10,, , , , ]. All of SCLC transformed patients were enrolled in the ASTRIS trial only based on liquid biopsy and developed a primary resistance to osimertinib. The plasmatic information of T790M may underestimate the coexistence of different mechanisms of resistance at the moment of progression of disease to first/second-generation EGFR-TKI . After publication of results from the FLAURA trial  and considering the high probability for osimertinib to move into front-line setting, a huge effort is underway to identify possible predictors of treatment efficacy and to stratify patients beyond positivity. Our study, with a limitation due to the small sample size and of some data already published [8,9], confirms the importance of testing plasma at baseline for , as patients with absence or low levels of are significantly associated with better outcome. Moreover, we confirm the tested in plasma as a surrogate biomarker for prognosis and response prediction. Given the heterogeneity of resistance mechanisms to third-generation TKI, tumor re-biopsy after progression on osimertinib remains a crucial issue for understanding tumor biology and to offer patients the best therapeutic options.