Association of genomic alterations in circulating tumor DNA (ctDNA) with clinical response to telisotuzumab vedotin (Teliso-V) in 2L+ <i>EGFR</i> wildtype ( <i>EGFR</i> wt) non-squamous non-small cell lung cancer (NSCLC) patients (pts) with c-Met overexpression (OE).

Abstract

3027 Background: Teliso-V is an antibody-drug conjugate comprising the c-Met–targeting antibody telisotuzumab linked to the microtubule inhibitor monomethyl auristatin E. In the LUMINOSITY trial (NCT03539536), Teliso-V monotherapy demonstrated efficacy in EGFR wt pts with c-Met OE (≥25% tumor cells at 3+ intensity by IHC) (Camidge et al. JCO 2024;42:3000-11). We used ctDNA molecular profiling to investigate baseline (BSL) and longitudinal changes in pts’ tumor mutational spectrum, and to identify potential mechanisms of tumor response and drug resistance to Teliso-V. Methods: Pts received 1.9 mg/kg Teliso-V intravenously Q2W. In total, 83 pts with ctDNA data and evaluable tumor assessments in Stage 2 were included in the analysis. Plasma ctDNA was collected at multiple timepoints and analyzed for genomic alterations using the PGDx elio Complete NGS assay (521 genes). Variants with allele frequency (VAF) &lt; 0.3% or from putative clonal hematopoiesis of indeterminate potential genes were removed. High ctDNA levels were defined as having a mean (m)VAF ≥median (2.05%), and low ctDNA levels as mVAF &lt; median. Molecular response (MR) was defined as having ≥50% reduction in the mVAF vs BSL levels without gene amplification. Mutational profiles and their association with RECIST-defined tumor response and/or drug resistance were assessed. Results: Overall, pts with high BSL ctDNA levels had an ORR (28.6%, 12/42) similar to all EGFR wt pts with c-Met OE (28.6%, 46/161) and were not statistically different vs pts with low BSL levels. However, pts with low BSL ctDNA had longer median OS (16.3 vs 8.5 mo) and mPFS (8.1 vs 5.4 mo) vs those with high BSL levels. Although the total pts with genomic alterations (GA) in this analysis was limited, KRAS GA were one of the most common mutations detected at BSL (24%, 20/83 pts). ORR to Teliso-V among pts with the actionable GA (AGA) of KRAS G12C was 100% (5/5). Conversely, among pts with non-AGA KRAS G12V/D/A and Q61H/L, the ORR was 23% (3/13). Additional AGAs were found in BSL ctDNA, including 1 BRAF V600E, 3 MET ex14del, 1 EGFR G719C, and 2 RET1-KIF5B translocations; none had response to Teliso-V. Pts with a MR at week 6 had higher ORR (35% vs 23%), longer median OS (15.5 vs 12.2 mo), and median PFS (8.5 vs 5.7 mo) vs those who did not. One pt with stable disease had several new AGAs detected in circulation at week 6, including activating EGFR ex20ins. At week 24, clinical progression was accompanied by gene amplifications of ERBB2, FGF4, FGFR4, and FGFR3. Other types of pharmacodynamic changes in ctDNA that could predict clinical response or drug resistance will be presented. Conclusions: ctDNA is a promising biomarker in predicting Teliso-V activity. Confirmatory research is planned in larger pt cohorts and/or with tissue-based NGS analyses. Clinical trial information: NCT03539536 .