MSI-H by PCR, dMMR by IHC, MSI-H by NGS — are they the same thing?

All three capture the same biology (failure of mismatch repair) but they're not identical assays. Concordance is generally >90% but discrepancies happen, especially in non-colorectal tumors. If results disagree, reflex orthogonal testing (e.g., NGS-MSI plus IHC for MMR proteins). All three qualify for tumor-agnostic pembrolizumab.

When is Lynch syndrome workup required?

Always, when MSI-H or dMMR is found in colorectal or endometrial cancer (and increasingly other tumors). NCCN-equivalent guidance (paraphrased without quoting) recommends germline panel testing in this scenario. For MLH1 loss specifically, BRAF V600E and MLH1 promoter methylation reflex to distinguish sporadic from Lynch first.

Does TMB-high overlap with MSI-H?

MSI-H tumors are typically TMB-high (>20-30 mut/Mb). But TMB-H without MSI is also actionable (pembrolizumab has a separate tumor-agnostic indication for TMB ≥10 mut/Mb in solid tumors after standard therapy). The two indications cover overlapping but distinct populations.

Variant cheat sheet · any solid tumor (tumor-agnostic)

MSI-H / mismatch repair-deficient solid tumors (tumor-agnostic)

Free clinical reference. Educational use only — not a substitute for professional medical advice.

TL;DR

MSI-H / dMMR solid tumors are AMP/ASCO/CAP Tier I-A for checkpoint inhibitor therapy. Pembrolizumab has a tumor-agnostic FDA indication (KEYNOTE-158). Dostarlimab covers dMMR endometrial. Nivolumab + ipilimumab for MSI-H colorectal. Always pair with Lynch syndrome workup in eligible patients — germline testing, counseling, family cascade testing.

Biology

MSI-H (microsatellite-instability-high) and dMMR (mismatch repair-deficient) are equivalent phenotypes capturing failure of the DNA mismatch repair machinery. The resulting tumors accumulate frameshifts and a high neoantigen load, making them uniquely sensitive to PD-1 / PD-L1 blockade. Causes: germline MLH1/MSH2/MSH6/PMS2/EPCAM mutations (Lynch syndrome), MLH1 promoter hypermethylation (sporadic), biallelic somatic inactivation.

Epidemiology

MSI-H prevalence varies by tumor type: ~15% of colorectal, ~25-30% of endometrial, lower in most other solid tumors. Roughly 3-4% of all solid tumors pan-tumor are MSI-H. Lynch syndrome accounts for ~3% of CRC and a slightly smaller fraction of endometrial cancer.

Detection

IHC for MMR proteins (MLH1, MSH2, MSH6, PMS2): loss = dMMR.
PCR for microsatellite instability (Bethesda panel or equivalent): MSI-H if ≥2 of 5 markers unstable.
NGS-based MSI: assessed by tumor-only or tumor-normal pipelines on comprehensive panels.
If MLH1 loss on IHC: BRAF V600E and MLH1 promoter methylation reflex to distinguish sporadic from Lynch.
Lynch syndrome workup: germline panel testing for MLH1/MSH2/MSH6/PMS2/EPCAM.

FDA-approved targeted therapies

Drug	Indication	Line	Tier
Pembrolizumab	Tumor-agnostic MSI-H / dMMR after standard therapy (KEYNOTE-158); 1L MSI-H mCRC (KEYNOTE-177); dMMR endometrial 1L	1L+ (tumor-dependent)	I-A
Dostarlimab	dMMR endometrial cancer (GARNET); dMMR recurrent/advanced solid tumor	1L+ (tumor-dependent)	I-A
Nivolumab + ipilimumab	MSI-H mCRC (CheckMate 8HW); other MSI-H solid tumors	1L (CRC)	I-A
Nivolumab monotherapy	MSI-H mCRC after prior treatment	Later-line	I-A

Tier = AMP/ASCO/CAP 2017 (Li MM et al., J Mol Diagn 2017). Drug names link to openFDA labels where available. Synthetic data; consult primary sources before treatment.

Resistance pathways

Loss of HLA class I expression (antigen presentation).
JAK1/JAK2 mutations disabling interferon signaling.
Acquired T-cell exhaustion phenotype.
B2M loss — frequently seen in MSI-H tumors that progress on immunotherapy.

Caveats

MSI-H is the indication, but tumor heterogeneity matters: BRAF V600E mCRC is enriched for MSI-H (sporadic CIMP-high), and immunotherapy 1L may be preferred over BRAF-targeted combination depending on disease tempo and AE considerations. Always check for Lynch syndrome: it has implications for the patient and family cascade testing well beyond the immediate cancer.

Frequently asked questions

MSI-H by PCR, dMMR by IHC, MSI-H by NGS — are they the same thing?: All three capture the same biology (failure of mismatch repair) but they're not identical assays. Concordance is generally >90% but discrepancies happen, especially in non-colorectal tumors. If results disagree, reflex orthogonal testing (e.g., NGS-MSI plus IHC for MMR proteins). All three qualify for tumor-agnostic pembrolizumab.
When is Lynch syndrome workup required?: Always, when MSI-H or dMMR is found in colorectal or endometrial cancer (and increasingly other tumors). NCCN-equivalent guidance (paraphrased without quoting) recommends germline panel testing in this scenario. For MLH1 loss specifically, BRAF V600E and MLH1 promoter methylation reflex to distinguish sporadic from Lynch first.
Does TMB-high overlap with MSI-H?: MSI-H tumors are typically TMB-high (>20-30 mut/Mb). But TMB-H without MSI is also actionable (pembrolizumab has a separate tumor-agnostic indication for TMB ≥10 mut/Mb in solid tumors after standard therapy). The two indications cover overlapping but distinct populations.

Citations

Marabelle A et al. Pembrolizumab in MSI-H non-CRC tumors (KEYNOTE-158). J Clin Oncol 2020; 38:1-10.
André T et al. Pembrolizumab vs chemotherapy in MSI-H mCRC (KEYNOTE-177). NEJM 2020; 383:2207-2218.
Mirza MR et al. Dostarlimab in dMMR advanced endometrial (RUBY). NEJM 2023; 388:2145-2158.
Lenz HJ et al. Nivolumab + ipilimumab in MSI-H mCRC (CheckMate 8HW). NEJM 2024.