MET-FISH Evaluation Algorithm: Proposal of a Simplified Method.

Abstract

MET amplifications (METamp) occur in 5% of NSCLC and represent in most case mechanisms of resistance to ALK and/or EGFR-targeted therapies. METamp detection can be performed using different techniques, although Fluorescence In-Situ Hybridization (FISH) remains the gold-standard, especially in the context of subclonality. To date current evaluation algorithms of MET amplifications are time consuming. Aim of the study was to identify a faster, equally reliable diagnostic algorithm for the detection of METamp, which is currently classified in negativity and low/intermediate/high-level amplification. N=497 NSCLC cases with available MET-FISH data had been selected. The results based on the first evaluated 20 cells had been re-calculated and compared with the definitive results based on 60 cells. For n=464 (93.4%) identical results had been obtained when counting 20 cells instead of 60 cells. Thirty-three cases (5.6%) showed a discrepancy, leading to an incorrect upgrade to a higher diagnostic category (n=25) and to an incorrect downgrade (n=8). We propose a simplified, yet equally reliable MET FISH-algorithm: after accurate screening of the whole tumor slide, twenty tumor cells have to be evaluated and results calculated: If the result is negative, or if all criteria of high-level METamp are fulfilled, the case can be signed out as such. All other cases should be considered as equivocal and additional 40 cells have to be counted. Given that, reliable results can be obtained by counting 20 cells only and an "equivocal" category for cases that need further investigation have been clearly defined.