Tumor Mutational Burden and Microsatellite Instability: Biomarkers That Predict Immunotherapy Response

Not every cancer responds to immunotherapy. The transformative results from pembrolizumab, nivolumab, and atezolizumab in subsets of patients with lung cancer, melanoma, and other tumors prompted an urgent question: what distinguishes responders from non-responders? While PD-L1 expression was the first biomarker incorporated into immunotherapy approvals, two additional biomarkers — tumor mutational burden (TMB) and microsatellite instability (MSI) — have emerged as critical tools for identifying patients with "hot" tumors most susceptible to immune checkpoint blockade.

Understanding these biomarkers requires grasping the fundamental biology of how the immune system recognizes cancer, why some tumors evade immune surveillance more effectively than others, and what the practical implications are for testing, drug selection, and predicting outcomes. This guide covers both concepts in depth, their relationship to each other, and the clinical contexts in which each is most meaningful.

The Immune System's Recognition of Cancer: Why Mutations Matter

The immune system's ability to recognize and eliminate cancer cells depends critically on a concept called antigenicity — whether cancer cells display molecular signals that T cells can identify as foreign. This recognition occurs through the presentation of short peptide fragments on MHC class I molecules on the cancer cell surface. When these peptides are derived from mutated proteins — proteins whose sequence differs from the patient's normal germline proteins — they are called neoantigens.

Neoantigens are powerful immunological targets because T cells have not been tolerized against them: the immune system's central tolerance mechanisms eliminate T cells that react against self-proteins (preventing autoimmunity), but neoantigens arising from somatic mutations in tumor cells are not self-proteins, so T cells reactive against them persist in the repertoire. The more neoantigens a tumor generates — which correlates with how many somatic mutations it has accumulated — the more likely that some of those neoantigens will trigger an effective immune response. Checkpoint immunotherapy, by releasing the brakes on T cells (blocking PD-1/PD-L1 or CTLA-4), amplifies this latent anti-tumor immune response.

Tumor Mutational Burden: Counting the Mutations

Tumor mutational burden (TMB) quantifies the total number of somatic mutations per megabase (mut/Mb) of the tumor genome. It is measured by comprehensive genomic profiling panels — typically large next-generation sequencing panels covering 300–500+ genes — which count all non-synonymous coding mutations detected in the tumor DNA compared to matched normal tissue (to exclude germline variants). Whole exome sequencing provides the most comprehensive TMB assessment but is more costly and less standardized than panel-based approaches.

TMB varies dramatically by cancer type. Melanoma and lung cancers driven by UV light and tobacco carcinogens, respectively, have among the highest median TMBs (often >10 mut/Mb) because these environmental mutagens introduce large numbers of point mutations. Pediatric cancers and hematologic malignancies typically have very low TMB. Within any cancer type, there is also wide patient-to-patient variation.

The FDA approved pembrolizumab for TMB-high (≥10 mut/Mb) solid tumors in 2020 — the second tissue-agnostic approval in oncology. However, this approval has been more nuanced in clinical practice than the MSI-H approval, because the predictive value of TMB for immunotherapy response varies significantly by tumor type. In some cancers (NSCLC, urothelial), TMB-high robustly predicts immunotherapy benefit. In others (glioblastoma, soft tissue sarcoma), even TMB-high tumors respond poorly to immunotherapy, suggesting that additional factors — such as the immunosuppressive tumor microenvironment or neoantigen quality — matter independently of mutation count.

Microsatellite Instability: A Defective Proofreading System

Microsatellite instability (MSI) arises from defects in the DNA mismatch repair (MMR) system — the cellular machinery responsible for correcting errors that occur during DNA replication. The MMR system is composed of four key proteins: MLH1, MSH2, MSH6, and PMS2. When any of these proteins is absent or non-functional — due to germline mutations (Lynch syndrome), somatic mutations, or MLH1 promoter methylation — the MMR system fails to correct replication errors.

Microsatellites are short, repetitive DNA sequences scattered throughout the genome (e.g., CACACACACACA). Because these repeated sequences are prone to replication slippage errors, they accumulate insertions or deletions (indels) at high rates in MMR-deficient tumors — a phenomenon detected as altered microsatellite lengths compared to normal tissue. A tumor is classified as MSI-high (MSI-H) when multiple microsatellite loci show instability, MSI-low (MSI-L) when few loci are affected, and microsatellite stable (MSS) when none are.

The consequence of MSI-H is a hypermutated tumor: because frameshift mutations in coding sequences generate highly immunogenic neoantigens (the novel peptide sequences created by reading-frame shifts are very different from normal human proteins), MSI-H tumors are robustly recognized by the immune system. This explains why they respond so dramatically to checkpoint immunotherapy — the immune system already recognizes these tumors as foreign; it just needs the checkpoint brakes released.

Testing for MSI and MMR Status

MSI and MMR deficiency (dMMR) can be detected by complementary methods that generally — but not always — give concordant results. Immunohistochemistry (IHC) for the four MMR proteins (MLH1, MSH2, MSH6, PMS2) is the most widely used approach, testing whether each protein is present (expressed) or absent (lost) in tumor tissue. Loss of expression of any protein indicates dMMR. IHC is inexpensive, rapid, and available at virtually all pathology laboratories, making it the practical first-line test.

PCR-based MSI testing directly measures the lengths of 5 standard microsatellite markers (Bethesda panel) in tumor versus normal DNA. A tumor is MSI-H if ≥2 of 5 (or ≥30% of) markers are unstable. Next-generation sequencing (NGS) panels can simultaneously detect MSI and TMB, making comprehensive genomic profiling an increasingly efficient single test for multiple biomarkers. Liquid biopsy (circulating tumor DNA) MSI testing is also available but with lower sensitivity than tissue-based methods, particularly for early-stage disease.

Clinical Impact: Which Drugs and Which Cancers?

The MSI-H/dMMR biomarker has received FDA approval as a predictive marker for two drugs: pembrolizumab (Keytruda) in 2017 for any MSI-H/dMMR solid tumor — the first tissue-agnostic approval in oncology — and dostarlimab (Jemperli) in 2021, specifically for dMMR solid tumors. Both approvals were based on ORR data across multiple tumor types from basket trials. The response rates in MSI-H tumors are typically 30–40% ORR with chemotherapy-refractory disease, rising to 50–70% in first-line colorectal cancer in the KEYNOTE-177 trial, which established pembrolizumab as first-line standard for MSI-H metastatic colorectal cancer.

MSI-H is most common in colorectal cancer (15% of all CRC, though mostly early-stage; only 4–5% of metastatic CRC), endometrial cancer (25–30%), and gastric cancer (10–15%). It is rare in most other cancer types. Patients with Lynch syndrome — a hereditary condition caused by germline MMR gene mutations — are at markedly elevated risk for multiple MMR-deficient cancers including colon, endometrial, ovarian, urinary tract, and others.