Detection of MET Amplification (metamp) in Patients with EGFR Mutant (m) NSCLC after First-Line (1L) Osimertinib.

9074 Background: METamp is a common resistance mechanism to osimertinib in patients with EGFRm NSCLC and is associated with sensitivity to MET TKIs. Reported METamp rates vary considerably depending on the type of biopsy and assay used. Here, we report a large comprehensive analysis of METamp detected by TBx FISH (FISH+) and/or LBx NGS (L+) after 1L osimertinib. Methods: During prescreening of the INSIGHT 2 study (NCT03940703) of tepotinib + osimertinib in post 1L osimertinib patients (pts) with EGFRm NSCLC, METamp was centrally assessed by TBx FISH ( MET GCN ≥5 and/or MET/ CEP7 ≥2) and/or by LBx NGS ( MET plasma GCN ≥2.3; Archer). Results: Of 472 pre-screened pts (64% female, 57% Asian, 97% adenocarcinoma), 350 provided TBx and 443 provided LBx. After excluding 36 TBx and 7 LBx ‘not analyzed/not evaluable’ samples, there were 314 TBx (163 from primary tumors, 151 from metastases) and 436 LBx samples with METamp results (positive or negative). Overall, FISH+ METamp was detected in 159/314 pts (50.6%), with similar FISH+ rates in the primary tumor (77/163, 47.2%) and metastases (82/151, 54.3%). FISH+ METamp detection by region was: 43.6% (82/188) in Asia, 61.5% (72/117) in Europe, and 55.6% (5/9) in the US. When limiting analysis to sites without reported local prescreening, the overall FISH+ METamp rate was 46.5% (93/200) and by region was: 43.5% (60/138) in Asia, 52.7% (29/55) in Europe, and 57.1% (4/7) in the US. In 159 FISH+ pts, median GCN was 11.8 (range 5.0–50.6) and median MET/CEP7 ratio was 2.3 (range 0.8–12.7). Mean TBx turn-around time (TAT) from shipment to results was 6.8 days. Overall, L+ METamp was detected in 51/436 pts (11.7%). In 35 L+ pts who were also FISH+, the median GCN was 16.6 (range 5.3–45.3) and median MET/CEP7 ratio was 4.5 (range 1.0–12.7). 299 pts had both central TBx FISH and LBx NGS results (Table), of whom 152 (50.8%) were FISH+ and 38 (12.7%) were L+. LBx NGS identified METamp with high specificity (negative percentage agreement [NPA] 98.0%) but low sensitivity (positive percentage agreement [PPA] 23.0%) compared with TBx FISH: only 3/38 L+ samples were FISH− but 117/152 FISH+ samples (77.0%) were L−. Clinical activity of tepotinib + osimertinib was comparable in FISH+ and L+ pts. Conclusions: In this large comprehensive analysis, FISH+ METamp was detected in ~50% of pts progressing on 1L osimertinib, while L+ METamp was detected in only 11.7% of pts. METamp, which is the most common mechanism of resistance post 1L osimertinib, can frequently be undetected by using LBx NGS only. Given high specificity but low sensitivity of LBx, L+ pts can receive a MET TKI while a negative METamp result by LBx should be confirmed by FISH. FISH allows optimal identification of METamp post 1L osimertinib, can be delivered in a clinically meaningful timeframe, and may allow more pts to benefit from an oral MET TKI. Clinical trial information: NCT03940703 . [Table: see text]