Longitudinal <i>MTAP</i> copy number alteration (CNA) in NSCLC with repeated biopsy samples.

e23348 Background: MTAP homozygous deletion (HomD) is associated with inferior survival outcome and represents an emerging biomarker for PRMT5 and MAT2A inhibitors in non–small cell lung cancer (NSCLC). However, longitudinal transition of MTAP CNA in repeated biopsy samples remains unknown, which may inform patient selection for rebiopsy or potential resistance mechanisms for therapeutics. Methods: We analyzed 30,454 NSCLC patients comprising 36,069 tumor samples from the AACR Project GENIE public v18.0 database. Longitudinal analyses were performed in 2,850 patients with ≥2 sequentially obtained biopsies. CNA was categorized as Intact (No Copy Loss), Heterozygous Loss (HetL, 1-copy loss), and HomD (2-copy loss). Statistical comparisons were performed using contingency analyses, with Fisher’s exact test applied where appropriate. Sensitivity analyses were performed for CNA calling methods. Results: Among patients with ≥2 biopsies, 91.8% (2615), 4.4% (126), and 3.8% (109) were MTAP Intact, HetL, and HomD at baseline, respectively. Among patients with baseline MTAP HetL and Intact, acquired HomD was detected in 4.8% (137) of patients at 2 nd biopsy and more commonly in baseline HetL (13.5% [17/126] vs 4.6% [120/2,615], p &lt; 0.001). Baseline CDKN2A status was associated with a stepwise increase in MTAP HomD acquisition. Among baseline MTAP -intact tumors, rates increased from 3.9% to 7.1% to 13.1% across CDKN2A Intact, HetL, and HomD ( p &lt; 0.001), with a numeric trend among baseline MTAP HetL tumors (0%, 11.9%, 20.8%). Acquisition of MTAP HomD was significantly more frequent in primary-to-metastatic transitions than in metastatic-to-primary pairs (9.1% vs 3.0%, p &lt;0.01). Among tumors with baseline MTAP HetL, MTAP HomD acquisition occurred more frequently in patients with EGFR mutation vs. wildtype (18.2% [10/55] vs 9.9% [7/71]), but not according to KRAS mutation status (0% [0/16] vs 15.5% [17/110]). All major associations were consistent across CNA calling methods. Conclusions: MTAP HomD can be an acquired event in NSCLC patients. Initial MTAP HetL, CDKN2A CNA, and EGFR mutation are predisposed to develop further MTAP copy loss to HomD and may contribute to resistance mechanism and can be prioritized when considering rebiopsy to look for acquired MTAP HomD. Acquired MTAP homozygous deletion at second biopsy. Acquired MTAP HomDel at 2nd Biopsy P value All 137 / 2,741 (5.0%) Baseline MTAP Intact vs. HetL MTAP Intact: 120 / 2,615 (4.6%) MTAP HetL: 17 / 126 (13.5%) &lt; 0.001 Baseline MTAP Intact CDKN2A Intact vs. HetL vs. HomD Intact: 94 / 2,388 (3.9%)HetL: 3 / 42 (7.1%)HomD: 23 / 176 (13.1%) &lt; 0.001 EGFR mutation vs. wildtype EGFR -mut: 55 / 946 (5.8%) EGFR -WT: 65 / 1,765 (3.7%) 0.014 Baseline MTAP HetL CDKN2A Intact vs. HetL vs. HomD Intact: 0 / 1 (0%)HetL: 12 / 101 (11.9%)HomD: 5 / 24 (20.8%) 0.32 EGFR mutation vs. wildtype <jats:td colspan="1"