Comparison of MTAP loss by immunohistochemistry (IHC) and MTAP homozygous deletion by next-generation sequencing (NGS) and DNA fluorescence in-situ hybridization (FISH)

Background: Methylthioadenosine phosphorylase (MTAP) deletion, resulting from genomic alteration, is observed in approximately 10%-15% of solid tumors, including NSCLC, PDAC, and others. Second-generation MTA-cooperative PRMT5 inhibitors have shown encouraging efficacy and safety in patients with advanced solid tumors, emphasizing the need for precise patient selection strategies. DNA NGS, DNA FISH, and IHC are the primary methods of MTAP loss detection in patient tumor samples. FISH is often considered the gold standard for confirming gene deletions due to its high spatial resolution and direct DNA-level detection; however, its limited scalability, longer turnaround, and operational complexity make it less practical for routine screening. NGS enables multiplexed, high-throughput detection of genomic alterations including MTAP deletions with strong analytical performance and broad molecular context. It supports treatment selection, prognostic assessment, and trial eligibility from a single tumor specimen. IHC offers rapid turnaround, preserves tumor morphology, and allows visualization of intratumoral heterogeneity. With the promising results that have been shown so far for MTA-cooperative PRMT5 inhibitors, flexibility in testing methods is important to meet the diverse needs of patients and physicians as the development of these agents moves forward. Here, we present the results of a study assessing the concordance of IHC, NGS, and FISH testing to assess the presence of MTAP loss. Methods: We developed an MTAP IHC assay by screening, optimizing, and biophysically characterizing 13 commercially available antibody clones. We subsequently benchmarked our IHC assay using validated DNA NGS and/or FISH on > 160 solid tumor samples. Results: IHC was highly concordant with NGS (> 90% agreement, Cohen's Kappa > 0.85) and with FISH (> 85% agreement, Cohen's Kappa > 0.70). Similarly, NGS and FISH showed strong concordance (> 90% agreement, Cohen's Kappa > 0.80). Through in-depth assessment of tumor morphology and review of FISH images, we attributed most discordance in the context of apparent MTAP loss on IHC without concomitant MTAP deletion detection on NGS and/or FISH to low/borderline tumor purity. When investigating preliminary patterns of exon-level MTAP homozygous deletion by NGS and FISH in our cohort, we observed some heterogeneity between methods in tumors with partial MTAP deletion. Conclusions: These data demonstrate high overall agreement among IHC, NGS, and FISH for the detection of MTAP loss. IHC offers rapid screening with retained morphologic context, whereas NGS and FISH provide orthogonal molecular resolution. Our findings support IHC as a practical method for MTAP loss detection, along with NGS or FISH.