Microsatellite Instability and Mismatch Repair

Mismatch repair is the system that corrects copying errors in DNA, mostly in short repeated sequences called microsatellites. When it fails (mismatch repair deficiency), errors build up, and microsatellite lengths change from one cell to the next, which is called microsatellite instability (MSI). The tumour has many mutations and is unusually sensitive to immune checkpoint treatment.

Testing has three goals. It finds tumours that may arise from Lynch syndrome, the commonest inherited bowel and endometrial cancer syndrome. It gives prognostic information in colorectal cancer, particularly in early stage. It identifies patients whose tumours are likely to respond to immune checkpoint inhibitors, in several cancer types.

Two methods are used, which agree in over 90% of cases. Immunohistochemistry stains for the four mismatch repair proteins (MLH1, MSH2, MSH6 and PMS2) in tumour sections, and loss of nuclear staining means deficiency. PCR (or next generation sequencing) compares microsatellite lengths in tumour and normal tissue.

Results take 2 to 5 days for immunohistochemistry and 3 to 7 days for PCR. Mismatch repair deficiency is found in about 15% of colorectal cancers (about 12% sporadic and 3% Lynch syndrome), 25 to 30% of endometrial cancers, 10 to 20% of gastric cancers, and a small percentage of other tumours.

Indications.

• All newly diagnosed colorectal and endometrial cancers, as universal screening for Lynch syndrome and for treatment selection.
• Cancers in people with a personal or family history suggesting Lynch syndrome (gastric, small bowel, ovarian, urothelial, pancreatic, biliary, brain and sebaceous tumours).
• Colorectal cancer in patients under 50 years, or with several related cancers.
• Selection for immune checkpoint inhibitor treatment in advanced solid tumours.
• Prognosis in stage II colon cancer, since deficiency carries a better outcome and less benefit from fluoropyrimidine alone.
• Interpretation of tumour mutation burden and of results from sequencing panels.

Not indicated for tumours with no link to Lynch syndrome or immunotherapy, and not as a replacement for germline testing, which is the test that confirms Lynch syndrome.

Specimen requirements.

• Formalin fixed, paraffin embedded tumour sections, with a good fixation time and fresh sections. Normal tissue (adjacent mucosa, lymphocytes, or blood) is needed as an internal control for immunohistochemistry and as a comparison for PCR.
• Sufficient tumour content (at least 20% for PCR).
• Pre-treatment biopsy or the resection specimen. Tissue after neoadjuvant chemoradiotherapy can show reduced staining.
• Clinical details, family history and previous tests.

Normal appearances.

Proficient mismatch repair shows nuclear staining for all four proteins in tumour cells, and staining of normal cells (lymphocytes, stroma, normal glands) as an internal control. By PCR, the tumour has stable microsatellites (MSS), or low instability, with no change in lengths between tumour and normal.

Interpretation.

• Deficient mismatch repair (dMMR) or high microsatellite instability (MSI-H): loss of one or more proteins, or instability in two or more markers of a five marker panel (or a high score by sequencing).
• Loss of MLH1 and PMS2: most often sporadic through hypermethylation of the MLH1 promoter, which is confirmed by BRAF V600E mutation (in colorectal cancer) or MLH1 promoter methylation testing. If both are absent, germline testing of MLH1 is needed.
• Loss of MSH2 and MSH6: strongly suggests Lynch syndrome, with germline testing of MSH2 and EPCAM.
• Isolated loss of MSH6 or PMS2: Lynch syndrome is possible, with germline testing of the respective gene.
• Normal staining with a MSI-H result by PCR, or loss with MSS by PCR: discordant results are reviewed, and testing with the other method is done.
• Sporadic dMMR: usually in older women, with right sided colon tumours, a poor differentiation, and BRAF V600E.
• Immunotherapy: dMMR or MSI-H solid tumours respond to checkpoint inhibitors at higher rates, with durable responses in a considerable fraction.
• Stage II colon cancer: dMMR carries a lower risk of recurrence, and fluoropyrimidine alone is not helpful.

Critical values and findings needing immediate action.

• Loss of mismatch repair proteins suggesting Lynch syndrome: referral to genetic counselling and germline testing, with screening plans for the patient and relatives.
• dMMR or MSI-H in advanced cancer: urgent discussion about immune checkpoint treatment.
• Discordant or unexpected results: review by the pathologist and repeat testing.
• A tumour in a family with a known pathogenic mismatch repair variant: testing for the familial variant.

Limitations and artefacts.

• Poor fixation, long cold ischaemia or old blocks cause weak or absent staining, including in normal cells, so a loss without internal control staining is not valid.
• Neoadjuvant chemoradiotherapy can reduce staining, so pre-treatment biopsy tissue is used when possible.
• Heterogeneous or subclonal loss: staining is lost in part of the tumour, and the whole section is examined.
• Weak or patchy staining, and missense mutations that give a non-functional protein with retained staining, cause false normal results by immunohistochemistry (about 5 to 10% of Lynch cases).
• PCR panels based on mononucleotide repeats can miss MSH6-related instability, which is often low or absent.
• Low tumour content gives false MSS results on PCR.
• Some tumours (for example, endometrial) have loss from somatic mutations, so germline testing is needed to find Lynch syndrome.
• Instability alone does not diagnose Lynch syndrome. It occurs in sporadic tumours.
• Interpretation of MSI differs in other tumour types, and validated cut offs are needed.

Special groups.

• Young patients and people with multiple primary cancers: higher chance of Lynch syndrome, and germline testing even if tumour tests are normal when suspicion is high.
• Women with endometrial cancer: universal tumour testing, and referral for genetic testing when loss of MSH2, MSH6 or PMS2, or MLH1 loss without methylation.
• Families with Lynch syndrome: cascade testing of relatives, with surveillance (colonoscopy every 1 to 2 years from the age of 20 to 25 years) and consideration of risk reducing surgery.
• Patients with advanced cancers other than colorectal: testing identifies candidates for immunotherapy.
• Constitutional mismatch repair deficiency in children: biallelic germline variants, with early onset tumours.
• Regions: testing is available in few centres. Immunohistochemistry is cheaper and more widely available than PCR and sequencing, and a four antibody panel is a practical first test. Cost of immunotherapy, and limited genetic counselling services, restrict the clinical value.

Follow up tests.

• BRAF V600E testing and MLH1 promoter methylation testing.
• Germline testing for MLH1, MSH2, MSH6, PMS2 and EPCAM (gene panel or single gene).
• Genetic counselling and cascade testing of relatives.
• Surveillance colonoscopy, gynaecological and other screening for carriers.
• Tumour mutation burden and PD-L1 testing.
• Multidisciplinary meeting for treatment planning.
• Staging imaging.
• Repeat testing on another block when results are discordant.