Why Microsatellite Instability Matters: Microsatellite instability (MSI) is a genomic phenotype caused by defects in DNA mismatch repair (MMR) machinery. In cancers like colorectal and endometrial cancer, MSI-high (MSI-H) status identifies patients who respond dramatically to immune checkpoint inhibitors such as pembrolizumab. The FDA has approved pembrolizumab for any MSI-H solid tumor, making MSI testing relevant across cancer types.
MSI in Lung Cancer is Poorly Characterized: Unlike colorectal cancer where MSI occurs in about 15% of cases, lung cancer has historically been considered MSI-rare. Most large lung cancer studies did not systematically assess MSI, leaving the true prevalence, molecular characteristics, and clinical significance poorly understood.
Study Scope: This large French study applied an AI-driven bioinformatics tool called MIAmS to analyze next-generation sequencing (NGS) data from 1,547 lung cancer patients. The goal was to establish the true prevalence of MSI in lung cancer and characterize the genetic and clinical features of MSI-positive patients.
MIAmS Technology: MIAmS (Microsatellite Instability Analysis using Machine learning on Sequencing) is a bioinformatics pipeline that uses machine learning to classify MSI status directly from NGS panel data - the same sequencing already performed for EGFR, KRAS, and other driver mutations - without requiring a separate dedicated MSI test.
MIAmS Bioinformatics Tool: MIAmS analyzes the distribution of reads spanning microsatellite loci from standard NGS panel BAM files. Using a machine learning classifier trained on known MSI-H and MSS (microsatellite stable) samples, it computes an MSI score for each tumor without requiring matched normal tissue. The algorithm considers the cumulative instability across multiple microsatellite markers.
Patient Cohort: 1,547 consecutive lung cancer patients who underwent routine molecular profiling by NGS were included. The cohort comprised adenocarcinoma (the most common subtype), squamous cell carcinoma, small cell lung cancer, and other subtypes. This represents one of the largest systematic MSI studies in lung cancer.
Confirmatory Testing for MSI-H Cases: All cases flagged as MSI-H by MIAmS underwent confirmatory testing using immunohistochemistry (IHC) for MMR protein expression (MLH1, PMS2, MSH2, MSH6) and standard PCR-based fragment analysis for direct MSI confirmation at five canonical microsatellite loci.
Molecular Characterization: Confirmed MSI-H cases were analyzed for tumor mutational burden (TMB), co-occurring driver mutations, smoking history, and overall mutational signatures to build a comprehensive molecular portrait of lung cancer MSI.
MSI Prevalence - Rare but Detectable: Of 1,547 patients, MIAmS identified 8 cases as MSI-H, representing a prevalence of 0.52%. This is substantially lower than colorectal cancer (approximately 15%) but confirms that MSI does occur in lung cancer. All 8 flagged cases were confirmed by follow-up IHC and PCR testing, and 6 had confirmed dMMR (deficient mismatch repair) by IHC loss of MLH1 or PMS2.
Smoker-Only Association: A striking finding was that all 8 confirmed MSI-H lung cancer patients were current or former smokers. This association between smoking and lung cancer MSI contrasts with colorectal MSI, which is predominantly a non-smoker/sporadic or Lynch syndrome phenomenon, and suggests a distinct etiopathogenesis for lung cancer MSI.
Extremely High Tumor Mutational Burden: MSI-H lung cancers had a median TMB of 21.4 mutations per megabase, compared to a median of approximately 4-6 mutations per megabase in typical lung cancers. This hypermutated phenotype is expected with MMR deficiency and suggests these tumors would be strong candidates for immunotherapy.
Predominant MLH1/PMS2 Loss: Among the 6 cases with IHC-confirmed dMMR, loss of MLH1 and PMS2 expression was the most common pattern, similar to sporadic colorectal MSI where MLH1 promoter methylation silences the gene. MSH2 and MSH6 loss was less frequent in this cohort.
Actionable FDA-Approved Therapy: Pembrolizumab has tumor-agnostic FDA approval for MSI-H cancers, meaning the 0.52% of lung cancer patients with MSI-H status qualify for this targeted immunotherapy. While the absolute percentage is small, given that lung cancer affects millions globally, this represents thousands of potentially treatable patients annually.
MIAmS Enables No-Cost MSI Detection: Because MIAmS analyzes existing NGS panel data that lung cancer patients already receive for EGFR/KRAS/ALK testing, incorporating MSI profiling adds no additional cost or tissue requirement. This argues strongly for universal MSI assessment in all lung cancer NGS panels.
High TMB as Co-biomarker: The finding that MSI-H lung cancers have very high TMB (21.4 Mut/Mb) means that TMB-high thresholds used for immunotherapy selection would capture most of these patients even without dedicated MSI testing. However, MSI-specific pembrolizumab approval provides a more direct regulatory pathway.
Smoking as a Clinical Clue: The exclusive association with smokers may help clinical teams prioritize MSI testing or have a higher index of suspicion in heavily smoking patients, though the rarity means systematic testing across all patients remains preferable to selective testing.
Outcomes Data Needed: While this study establishes MSI prevalence and molecular features, clinical outcomes data on MSI-H lung cancer patients treated with checkpoint inhibitors is very limited due to the rarity of this subtype. Multicenter registries collecting treatment responses in MSI-H lung cancer patients are urgently needed.
Comparison with Other Solid Tumors: The observation that lung cancer MSI shares some but not all features with colorectal MSI (similar MLH1/PMS2 loss but different smoking association and histological context) warrants further comparative molecular analysis. Understanding these differences could inform organ-specific immunotherapy response predictions.
MIAmS Validation on Other Platforms: MIAmS was validated on one institutional NGS pipeline. Extension to other commercially available NGS panels (Foundation One, Guardant, etc.) and different sequencing platforms would broaden its applicability and allow multicenter implementation.
Lynch Syndrome Screening: In colorectal cancer, early-onset MSI-H often indicates Lynch syndrome (hereditary MMR gene mutations). Whether lung cancer MSI-H patients should also be referred for germline testing and Lynch syndrome evaluation remains an open question that future work should address.