Comparison of immunohistochemistry, PCR, and NGS for the evaluation of mismatch repair deficiency and microsatellite instability in colorectal cancer: a retrospective study

Microsatellite instability (MSI) and mismatch repair (MMR) deficiency serves as a critical predictive biomarker for the efficacy of immune checkpoint inhibitors (ICIs) in colorectal cancer (CRC). However, discrepancies among detection methods may lead to inconsistent results, which may contribute to ICIs resistance. This study aimed to evaluate the concordance of MMR status and MSI assessment across three commonly used methods and to explore factors underlying discordant cases. We retrospectively analyzed MSI status in 534 CRC patients using three detection methods: immunohistochemistry (IHC) for MMR protein expression, and polymerase chain reaction (PCR) and next-generation sequencing (NGS) for MSI status assessment. Concordance between methods was measured using Kappa statistics, and discordant cases were further investigated. Among the 534 cases, 7.3% (39/534) were classified as deficient mismatch repair (dMMR) by IHC, 7.5% (40/534) as MSI-H by PCR, and 7.3% (39/534) as MSI-H by NGS. The highest concordance was observed between PCR and NGS (Kappa = 0.986), followed by NGS and IHC (Kappa = 0.943), and PCR and IHC (Kappa = 0.930). Discordant results occurred in five cases (0.9%), primarily attributable to assay-specific limitations, tumor heterogeneity, technical or pre-analytical issues, and biological variability. IHC, PCR, and NGS showed high overall concordance in MSI and MMR assessment. The small number of discordant cases highlights the inherent limitations of individual methods. While single-method testing is generally reliable, multi-platform evaluation may provide additional insights in diagnostically challenging or borderline cases.