A novel detection strategy for an emerging biomarker: MTAP loss detection with circulating tumor DNA (ctDNA)

Background: Methylthioadenosine Phosphorylase ( MTAP ) is a tumor suppressor gene frequently deleted in cancer, most often through loss of the 9p21 locus ( MTAP , CDKN2A , CDKN2B ). MTAP loss is an emerging biomarker with investigational therapies, yet detection has relied on tissue analysis, with prevalence estimates of 10-18% in advanced (adv) non-small cell lung cancer (NSCLC). Detection of large gene deletions by ctDNA is technically difficult and MTAP loss is not reported on most commercial ctDNA panels. Given the convenience of liquid biopsy and potential to expand trial eligibility, we analyzed the prevalence and genomic landscape of MTAP loss in lung cancer using a commercial ctDNA assay. Methods: Patients (pts) with adv lung cancer who had Guardant360® Liquid testing between 7/2024-6/2025 were identified from the Guardant Health database. The start date aligned with assay expansion (>740 genes/epigenomics) and inclusion of MTAP loss reporting. MTAP loss was defined as bi-allelic loss by homozygous deletion (HomDel) or single copy deletion (del) with a co-occurring loss of function variant; prevalence and co-alterations (co-alt) were assessed by histology and tumor fraction (TF). Co-alt comparisons used two-sided Fisher's Exact Test (p<0.05 significant). Results: MTAP loss was detected in 1.9% (330/17,333) of all lung cancer pts, 2.1% (55/2591) squamous, and 1.8% (244/13,745) non-squamous (nsq), regardless of TF; 94% (310/330) were NSCLC. Median (m) age was 69; 54.2% were male. For adv NSCLC, prevalence increased with TF, reaching 10.9% (294/2706) in pts with TF > 10%, consistent with tissue-based rates. In nsqNSCLC, frequently co-altered genes included CDKN2A (96%), TP53 (73%), EGFR (43%), KRAS (28%), KEAP1 (27%), STK11 (26%), RB1 (25%), and CCNE1 (24%). Compared to nsqNSCLC without MTAP loss, CDKN2A ( mostly HomDel), CCNE1 , EGFR , BRCA2 , KEAP1 , STK11 , RB1 , and FGFR1 were enriched (p<1×10⁻10). mTF and bTMB were higher in the MTAP loss cohort (30.9% vs 0.7%; 15.1 vs 12.5 mut/Mb). Oncogenic EGFR mutations (muts) occurred in 16.4% (54/330), including exon 19 del (46%), L858R (23%), exon 20 insertion (11%), and other muts (20%); mTF was 28.1% (range 7.7-90%). Oncogenic KRAS muts were present in 19.1% (63/330), most commonly G12X (70%), with mTF of 26.3% (range 1.6-80.7%). Overall prevalence of MTAP loss in EGFR and KRAS mutated NSCLC was 2.3% (54/2304) and 2.0% (63/3172), respectively. Conclusion: This first ctDNA analysis of MTAP loss prevalence and co-mutational landscape in NSCLC demonstrates that MTAP loss can be detected by ctDNA across histologies when TF is sufficient, offering a practical alternative when tissue is unavailable or rapid results are needed. ctDNA detection of MTAP loss may be further enhanced by methylation approaches for low-shed cases. Broader use of ctDNA testing could expand MTAP loss detection, increasing clinical trial eligibility and access to new therapies.