Tumor Mutational Burden (TMB)

TMB is defined as the total number of non-synonymous somatic mutations per megabase (Mb) of interrogated tumor genome. It includes single nucleotide variants (SNVs) and small insertions/deletions (indels) within coding regions. The standard unit is mutations per megabase (mut/Mb). Calculation requires subtracting germline polymorphisms through matched-normal filtering or population databases. Synonymous mutations are typically excluded as they do not generate neoantigens. The exonic capture region size directly impacts the denominator, requiring harmonization across assays.

The mechanistic link between TMB and immunotherapy response rests on neoantigen generation. Each somatic coding mutation has the potential to produce a novel peptide sequence presented by major histocompatibility complex (MHC) molecules on the tumor cell surface. Higher TMB increases the probability of generating immunogenic neoantigens recognized as foreign by CD8+ T cells. This rationale is supported by the success of immune checkpoint inhibitors in mismatch repair-deficient (dMMR) and microsatellite instability-high (MSI-H) tumors, which harbor orders of magnitude more mutations than microsatellite-stable counterparts.

TMB can be estimated through multiple sequencing modalities:

• Whole Exome Sequencing (WES): The gold standard, covering ~30-50 Mb of coding sequence. Provides the most comprehensive TMB estimate.
• Targeted Gene Panels: Cover 0.5-2 Mb. Require computational calibration and extrapolation to WES-equivalent values. Accuracy degrades at low TMB levels due to sampling variance.
• Liquid Biopsy (bTMB): Blood-based TMB derived from circulating tumor DNA. Concordance with tissue TMB is moderate and influenced by ctDNA shedding rates and clonal hematopoiesis filtering.

The FDA-approved cutoff of ≥10 mut/Mb (FoundationOne CDx) defines TMB-High status for pembrolizumab eligibility in solid tumors. However, thresholds are assay-specific and tissue-specific:

• TMB-High: Typically ≥10 mut/Mb across most solid tumors.
• TMB-Intermediate: 6-10 mut/Mb, representing a gray zone with variable benefit.
• TMB-Low: <6 mut/Mb, generally associated with lower response rates. Cutaneous squamous cell carcinoma and melanoma exhibit naturally high median TMB (>40 mut/Mb), while gliomas and pediatric tumors show low TMB (<2 mut/Mb).

TMB is an imperfect biomarker with several known limitations:

• Neoantigen quality over quantity: Not all mutations generate immunogenic peptides. MHC binding affinity, clonality, and expression matter.
• Intratumoral heterogeneity: Single-biopsy TMB may not represent the full mutational landscape.
• Technical variability: Panel size, sequencing depth, bioinformatic pipelines, and germline filtering strategies all influence TMB estimates.
• Tissue-specific thresholds: A universal cutoff ignores tumor-type-specific biology.
• Clonal hematopoiesis: Age-related blood cell mutations can falsely elevate bTMB if not computationally filtered.

TMB is increasingly integrated into multifactorial predictive models rather than used in isolation. Key complementary biomarkers include:

• Microsatellite Instability (MSI): Mechanistically related but not perfectly correlated with TMB.
• PD-L1 Immunohistochemistry: Measures target expression; combined with TMB improves predictive accuracy.
• T-cell Inflamed Gene Expression Profile (GEP): Captures the immune microenvironment state.
• HLA Genotype: Determines the repertoire of peptides that can be presented.
• Mutational Signatures: Specific etiologies (e.g., UV, smoking, APOBEC) may generate more immunogenic mutations.