Mismatch Repair Status and Microsatellite Instability in Ovarian Cancer: What the Test Tells Us

What the test measures

The DNA mismatch repair (MMR) system is the cell's proofreading crew. As DNA polymerase copies the genome, it occasionally slips — inserting or deleting a base, particularly in the repetitive stretches we call microsatellites. Four proteins do most of the corrective work, and they work in pairs: MLH1 with PMS2, and MSH2 with MSH6. When one partner is missing or nonfunctional, the pair destabilizes, and errors go uncorrected.

Over many cell divisions those uncorrected errors accumulate, and microsatellite lengths drift. That drift is what we call microsatellite instability (MSI). So deficient mismatch repair (dMMR) and MSI-high (MSI-H) aren't two separate findings — they're two readouts of the same underlying biology, one measured at the protein level and one at the DNA level [2]. That's why the two assays are usually concordant.

In ovarian cancer, this biology matters most in the endometrioid and clear cell histotypes. High-grade serous carcinoma, the most common ovarian cancer, rarely shows dMMR, so testing is most informative when the tumor's histology points toward the endometrioid or clear cell categories [1].

How it's tested

Two assay approaches are in routine use, both performed on formalin-fixed, paraffin-embedded (FFPE) tissue.

Immunohistochemistry (IHC) stains for the four MMR proteins. A normal, or proficient, result shows retained nuclear staining in tumor cells for all four. Loss of nuclear staining in one or more proteins — with intact staining in adjacent normal cells acting as an internal control — indicates dMMR. That internal control is essential; without it, you can't distinguish true protein loss from a failed stain.

PCR- or NGS-based MSI testing interrogates a panel of microsatellite loci, comparing tumor to normal DNA (or using a reference distribution) to detect instability. A tumor scored as MSI-H shows instability at a defined proportion of loci [2].

Preanalytic quality drives both. Cold ischemia time, fixation duration, and block age all affect antigen preservation and DNA integrity. Poorly fixed or over-decalcified tissue can yield weak or patchy staining that mimics loss. Interpret every stain against its internal control, and treat equivocal cases as equivocal.

What each result state means

dMMR / MSI-H. The tumor has lost functional mismatch repair. This carries two implications: a hereditary one and a therapeutic one. It flags the tumor as hypermutated, which is the biological rationale behind checkpoint-inhibitor eligibility, and it raises the possibility of an underlying germline cause.

pMMR / MSS. Mismatch repair is intact. The tumor is microsatellite stable. This result doesn't trigger the same downstream considerations. It's a negative screen, not a guarantee against every hereditary syndrome, but for MMR-related pathways it's reassuring.

One important nuance sits between these states. Isolated loss of MLH1/PMS2 is frequently caused by somatic MLH1 promoter hypermethylation — an acquired, non-hereditary event. That's why MLH1 promoter methylation testing is commonly performed as a reflex when this pattern appears.

What it determines for treatment eligibility and Lynch screening

A dMMR/MSI-H result serves two distinct downstream functions, and the pathology report should address both clearly.

For Lynch syndrome screening, dMMR raises the possibility of a germline mutation in one of the MMR genes. Because ovarian cancer can be a Lynch-associated malignancy, a dMMR finding in an endometrioid or clear cell tumor is a meaningful signal. The report should communicate the specific pattern of protein loss, since that pattern narrows the candidate gene. When MLH1/PMS2 loss is seen, the report should state whether reflex MLH1 promoter methylation testing was performed and its result, because a methylated tumor is far more likely sporadic. When methylation is absent, or when the loss pattern points to MSH2, MSH6, or PMS2, the report should recommend germline evaluation and genetic counseling. That documentation matters — it's often the trigger that gets a patient and family the referral they need. dMMR IHC and MSI testing are screening tools, not germline diagnoses; confirmation requires dedicated germline testing.

For therapy, a dMMR/MSI-H result informs eligibility for the immune checkpoint inhibitor drug class under tumor-agnostic criteria, which apply across solid tumors regardless of primary site [2]. This is an eligibility determination made by the treating clinical team — the pathology finding opens a door; it doesn't prescribe what walks through it. Nothing here constitutes treatment advice for any individual patient.

Caveats and what's evolving

Several practical points deserve emphasis. IHC and MSI testing are highly concordant but not perfectly so; rare discordant cases occur, and in those situations the two assays can be used to cross-check one another [2]. MSH6 loss in particular can be subtle or heterogeneous on IHC, and prior therapy can occasionally alter staining. When clinical suspicion is high but IHC is ambiguous, orthogonal MSI testing is reasonable. Histotype context remains central: applying these tests most deliberately in endometrioid and clear cell carcinomas reflects where the biology concentrates [1].

In practice, a dMMR result sets a small workflow in motion. Isolated MLH1/PMS2 loss reflexes to MLH1 promoter methylation testing. The report then documents the loss pattern, the methylation result, and — where indicated — a note recommending germline evaluation and genetic counseling. That short chain of steps is where a stain on a slide becomes actionable information for the patient and their family.

References

1. WHO Classification of Tumours Editorial Board. Female Genital Tumours, 5th ed. Lyon: IARC; 2020. ISBN 978-92-832-4504-9.
2. Deficient Mismatch Repair and Microsatellite Instability in Solid Tumors. PMCID: PMC12072705.

Note: Reference 2 is cited as provided in the source record; author names, journal title, volume, issue, and page numbers were not available in the supplied record and could not be completed without introducing unverified information.