Microsatellite instability and mismatch repair
In one sentence
Microsatellite instability is alteration of short repeated DNA sequences; mismatch-repair deficiency is impaired correction of DNA copying errors that can produce that pattern.
The intuition
Repeated letters are easy to miscopy: a run of six can accidentally become five or seven. Cells have proofreading machinery that catches certain copying errors. The analogy has limits: a repeat-length test measures the resulting DNA pattern, while a protein stain examines components of the repair machinery. They do not inspect the same thing.
How it works
Microsatellites are short repeated DNA sequences. Microsatellite instability (MSI) describes altered repeats in tumor DNA. Mismatch repair (MMR) corrects certain base-pairing errors and small copying insertions or deletions. Deficient mismatch repair (dMMR) can allow repeat changes to accumulate. NCI: MSI.
Two common approaches ask different questions:
| Approach | Direct observation | Interpretation boundary |
|---|---|---|
| DNA repeat testing by polymerase chain reaction (PCR), which amplifies selected DNA, or validated sequencing | Instability across specified repeat sites | A method-defined DNA pattern, not identification of every possible repair defect |
| Immunohistochemistry (IHC) | Presence or loss of staining for repair proteins, commonly MLH1, PMS2, MSH2 and MSH6 | Protein staining pattern; requires internal controls and appropriate tumor evaluation |
MSI-high (MSI-H) means the assay's instability criteria are met. Microsatellite stable (MSS) means its stability criteria are met. An inadequate or indeterminate specimen should keep that label rather than silently become MSS. Likewise, visible repair-protein staining does not demonstrate every aspect of functional repair.
The College of American Pathologists (CAP) guideline recommends different testing approaches across cancer types. For cancers outside its better-established settings, it calls for validation in the specific cancer type; a sequencing assay validated for colorectal cancer is not automatically validated for breast cancer. It also expressly rejects tumor mutational burden as a surrogate for MMR status. CAP recommendations.
Why it matters in cancer
MMR/MSI testing can support certain treatment and inherited-risk questions, but each needs its own interpretation. A tumor finding does not alone diagnose Lynch syndrome, an inherited cancer-predisposition condition. Some colorectal tumors acquire reduced MLH1 through DNA methylation; clinical hereditary-risk evaluation distinguishes possible causes. NCI genetics PDQ.
The U.S. pembrolizumab label includes a specific MSI-H/dMMR indication with disease-setting, prior-treatment and test requirements. That connection is a reason to measure accurately, not a promise that every MSI-H cancer responds or that an exploratory result establishes eligibility. FDA label, section 1.8.
Assay card
| Field | What to retain |
|---|---|
| Measures and method | Repeat instability by DNA testing, or repair-protein staining by IHC; record which question was tested |
| Input and tissue cost | Validated tumor specimen; IHC uses sections and DNA extraction consumes material. Requirements and controls differ |
| Output and units | MSI-H/MSS/indeterminate and sometimes an assay-defined unstable-site fraction; IHC patterns by named protein |
| Thresholds | Method- and cancer-specific criteria; no universal instability percentage or staining shortcut applies to every assay |
| Failure modes | Low tumor content, degraded material, insufficient informative repeats, failed controls or misleading staining |
| Validation limits | Discordance requires review of specimens, controls and methods. Neither assay alone is a germline diagnosis or response guarantee |
Common confusions
- MSI versus dMMR: a DNA consequence and a repair defect are related, but distinct observations.
- MMR versus homologous recombination deficiency (HRD): these concern different repair processes.
- Stable versus mutation-free: MSS does not mean there are no somatic mutations.
- Tumor finding versus inherited syndrome: acquired and inherited causes need separation.
Try it
A fictional sample has intact IHC controls and retained repair-protein staining. Its DNA MSI test is indeterminate because few usable tumor molecules were recovered. Can the report conclude “confirmed MSS”?
Answer: No. Retained staining and an indeterminate DNA test must retain their separate results. The cause of the inadequate DNA result and the validated follow-up workflow need review.
Explain it back
“An MSI test examines ______; MMR IHC examines ______.”
One answer: “repeat-sequence patterns; staining for specified repair proteins in the sampled tissue.”
Takeaway
Keep the DNA pattern, repair-protein evidence and clinical question attached to their own validated methods.
Related concepts
Sources and scope
Source check: October 10, 2026. Assay comparison and clinical boundaries; the example is fictional. Expert and learner review remain pending.
- NCI: microsatellite instability.
- CAP 2022: MMR/MSI testing recommendations, including cancer-specific validation and the TMB boundary.
- NCI: colorectal cancer genetics PDQ, MMR and MLH1 methylation sections.
- FDA: 2026 pembrolizumab label, section 1.8.