Design: Retrospective institutional cohort with IHC–NGS comparison subset
Population: 5,233 NSCLC with 447-gene NGS; 105 cases tested by both methods
Primary outcome: MTAP deletion frequency and IHC–NGS concordance
Effect: Homozygous deletion 5.6%; concordance 97.3% after excluding tumour content ≤30%

MTAP loss creates a metabolic vulnerability that newer targeted drugs aim to exploit, so laboratories will increasingly be asked to report it. This US institutional study reviewed 5,233 non-small cell lung cancers with NGS covering MTAP, and compared NGS with immunohistochemistry in 105 cases.

Homozygous MTAP deletion was found in 5.6%. It was commoner in adenocarcinomas from never or light smokers and in tumours with EGFR mutations or ALK fusions. After quality control and exclusion of samples with 30% tumour content or less, NGS and IHC agreed in 97.3%. About 13% of deficient cases stained heterogeneously. In the two discordant cases, one NGS call was wrong because of a complex genome, and one tumour had lost protein with no sequencing change.

IHC was the more sensitive test when tumour content was low or loss was subclonal. NGS also identified a partial deletion that could be misread without care.

For laboratories, MTAP IHC is a practical first test, especially on small biopsies, with NGS adding context on co-occurring drivers.

Consider MTAP immunohistochemistry where tumour content is too low for reliable copy-number calls
Treat NGS MTAP calls from samples with 30% tumour or less with caution
Report heterogeneous MTAP staining explicitly; it occurred in about 13% of deficient cases
Expect MTAP deletion alongside EGFR and ALK drivers, not only in smokers

Why it matters

MTAP status is likely to become a treatment-selection biomarker, and the two assays fail in different ways.

The statistics, in plain English

Concordance of 97.3% means the tests matched in all but a few of 105 cases, after excluding low-tumour samples. That figure applies to selected, quality-controlled material, not to every routine specimen.