Tumour Mutational Burden

by Jason Wasserman MD PhD FRCPC
March 25, 2026

Tumour mutational burden — usually abbreviated as TMB — is a measurement of how many genetic mutations are present in a tumour. The more mutations a cancer carries, the more it tends to look different from normal cells to the immune system. Because immunotherapy drugs work by helping the immune system find and attack cancer, this measurement has become an important biomarker: cancers with a high TMB are more likely to respond to a class of drugs called immune checkpoint inhibitors than cancers with a low TMB. In 2020, the FDA approved the immune checkpoint inhibitor pembrolizumab (Keytruda) for any advanced solid tumour with a high TMB that has progressed after prior treatment — the first time a cancer drug was approved purely based on this measurement, regardless of cancer type. Understanding what a TMB result means — and what it does not mean — is an important part of interpreting a molecular pathology report.

What the test looks for

Every time a cell divides, it must copy its entire DNA. Errors in this copying process, as well as damage caused by environmental exposures such as ultraviolet light from the sun or carcinogens in tobacco smoke, result in mutations — permanent changes in the DNA sequence. Healthy cells have repair systems that catch and fix most of these errors. In cancer cells, those repair systems are often partially or completely broken, allowing mutations to accumulate over time.

TMB is measured as the number of somatic (acquired, non-inherited) mutations per megabase of DNA — a megabase being one million DNA base pairs. A tumour with a high TMB has accumulated many more mutations than a tumour with a low TMB.

Why does this matter for treatment? When a cancer cell carries many mutations, some of those mutations result in abnormal proteins appearing on the cell’s surface. These abnormal proteins, called neoantigens, are foreign-looking to the immune system. T cells — the immune system’s attacking cells — can, in principle, recognize neoantigens and kill the cells that display them. However, cancers have evolved ways to suppress this immune attack, including by producing molecules that act as “off switches” (immune checkpoints) for T cells. Immune checkpoint inhibitors are drugs that block these off-switches, freeing T cells to do their job. Tumours with high TMB present the immune system with more targets to work with, which is why they tend to respond better to checkpoint inhibition than tumours with few mutations.

It is important to understand that TMB is a continuous measurement — there is a spectrum from very low to very high — and the cutoff that separates “TMB-high” from “TMB-low” is somewhat arbitrary. The FDA-approved cutoff is 10 mutations per megabase (mut/Mb), as measured by validated genomic tests such as the FoundationOne CDx assay. This means a result of 10 mut/Mb or higher qualifies as TMB-high, though response rates are meaningfully higher in tumours with 13 mut/Mb or more.

TMB by tumour type: what drives high TMB and what it means for treatment

The biological basis of high TMB varies across cancer types, with important implications for how reliably TMB predicts immunotherapy benefit. Understanding which category your cancer falls into is a useful context for interpreting your result.

Melanoma

Cutaneous melanoma (skin melanoma) consistently has among the highest TMB of any cancer type — typically well above 10 mut/Mb — because it is directly driven by the mutagenic effects of ultraviolet (UV) light from sun exposure. Each time UV radiation damages DNA, and the damage is imperfectly repaired, another mutation accumulates. Over years of sun exposure, this produces a very high mutational load. This biological explanation is directly relevant to treatment: the landmark trials of immune checkpoint inhibitors in melanoma, including early trials of ipilimumab (Yervoy) and subsequent trials of pembrolizumab and nivolumab (Opdivo), showed dramatic and durable responses in a subset of patients — responses that were among the highest seen with any cancer immunotherapy at the time. These responses laid much of the groundwork for TMB as a concept and for the subsequent pan-cancer approval.

In melanoma, however, it is worth knowing that TMB is not the only relevant factor. PD-L1 expression and tumour-infiltrating lymphocyte levels also predict response, and a subset of patients with lower TMB still respond. Conversely, not all high-TMB melanomas respond. For most patients with advanced melanoma, immunotherapy is offered based on the overall profile of the disease, and TMB is one of several factors oncologists consider. The BRAF mutation status of the melanoma is also critical and is discussed in a separate article.

Non-small cell lung cancer

Non-small cell lung cancer (NSCLC) has a high median TMB, particularly in patients with a significant history of tobacco smoking — again a direct consequence of the mutagenic effect of tobacco carcinogens on lung cell DNA. In NSCLC, TMB has been one of the most extensively studied immunotherapy biomarkers, alongside PD-L1 expression. The two measurements provide partially overlapping but distinct information. Some tumours are high in both, some in one but not the other, and the combination is being studied as a more refined predictor of response.

An important caveat in lung cancer is that driver mutation status — particularly EGFR mutations and ALK fusions — is associated with lower TMB and lower immunotherapy response rates, even when overall TMB is above 10 mut/Mb. For patients whose lung cancers carry targetable driver mutations, targeted therapy (rather than immunotherapy) is typically the preferred initial treatment, and TMB plays a less central role in those decisions.

Bladder and urothelial cancer

Bladder and other urothelial cancers have among the highest rates of TMB-high tumours of any common solid cancer — approximately 38% of urothelial cancers have TMB above 10 mut/Mb. This is partly attributable to APOBEC mutagenesis (an internal DNA-editing process that can go awry in cancer cells) and to other mutagenic processes. Immune checkpoint inhibitors are well-established in urothelial cancer, particularly for patients who cannot receive or have progressed on cisplatin-based chemotherapy. TMB is one of several biomarkers used alongside PD-L1 expression to guide immunotherapy decisions in this cancer type.

Endometrial cancer and POLE mutations: the highest TMB subgroup

Endometrial cancer has among the highest rates of TMB-high tumours across all cancer types — approximately 40 to 43% of endometrial cancers have high TMB overall. Within this group, two distinct biological drivers of high TMB are important to understand.

The first is MMR deficiency (dMMR/MSI-H), which is common in endometrial cancer (present in approximately 25 to 30% of cases). MMR deficiency leads to the accumulation of thousands of mutations because the DNA error-correction system is not functioning. These tumours tend to have high TMB and respond well to immunotherapy.

The second, and even more extreme, driver is mutations in the POLE (DNA polymerase epsilon) gene. POLE encodes a protein that is part of the DNA copying machinery. When POLE carries a specific type of inactivating mutation in its “proofreading” domain, the copying machinery becomes extraordinarily error-prone — producing some of the highest TMB values observed in any human cancer, sometimes exceeding 100 or even several hundred mut/Mb. POLE-mutated endometrial cancers represent a distinct molecular subtype with strikingly high immunotherapy response rates. Interestingly, POLE mutations are associated with a paradoxically good prognosis in endometrial cancer despite the tumour appearing high-grade — a feature that pathologists and oncologists are increasingly recognizing as clinically meaningful. POLE testing is now performed as part of the standard molecular classification of endometrial cancer.

Colorectal cancer

In colorectal cancer, TMB distribution is bimodal — most tumours are TMB-low (approximately 85%), while a distinct subgroup is TMB-high. This high-TMB subgroup largely overlaps with, but is not identical to, the dMMR/MSI-H subgroup discussed in the MMR/MSI colorectal cancer article. POLE mutations can also drive ultra-high TMB in a small subset of colorectal cancers. For patients with colorectal cancer and high TMB that is also MSS (microsatellite stable), the relationship between TMB and immunotherapy benefit is less clearly established than in dMMR/MSI-H disease, and TMB alone in MSS colorectal cancer is not currently a reliable standalone predictor of immunotherapy response. The management of colorectal cancer with high TMB is most straightforwardly guided by MMR/MSI status first, with TMB providing additional context.

Other cancers with notable rates of TMB-high tumours

Several other cancer types have meaningful rates of TMB-high tumours and established or emerging immunotherapy activity worth understanding.

Small cell lung cancer (SCLC) has a high overall TMB driven by tobacco carcinogen exposure, similar to NSCLC. Immune checkpoint inhibitors (atezolizumab or durvalumab combined with chemotherapy) are approved as first-line treatment for extensive-stage SCLC, in part reflecting this immunogenic background.

Cervical cancer has a meaningful rate of TMB-high tumours and is one of the cancer types included in the KEYNOTE-158 trial data supporting the pan-cancer TMB approval. Pembrolizumab is approved in cervical cancer based on both PD-L1 expression and, for some indications, regardless of PD-L1 status in recurrent or metastatic disease.

Merkel cell carcinoma is a rare but aggressive skin cancer that responds exceptionally well to immune checkpoint inhibitors. Interestingly, it is an example of a tumour that responds to immunotherapy even when TMB is relatively low — demonstrating that TMB alone does not always predict benefit, particularly when the immune response is being driven by viral antigens (Merkel cell polyomavirus) rather than somatic mutations.

Head and neck squamous cell carcinoma has a variable TMB that is higher in tobacco-related tumours and in a subset of HPV-negative tumours. Pembrolizumab is approved for recurrent or metastatic head and neck cancer based primarily on PD-L1 expression; TMB is an additional factor in some settings.

Oesophageal cancer, gastric cancer, and other gastrointestinal tumours have variable but meaningful TMB rates. Immunotherapy is now part of standard treatment for advanced gastric and oesophageal cancers in specific molecular subgroups, though the role of TMB independent of other markers (particularly MMR status and PD-L1) continues to be defined.

Why is the test done

TMB testing is most commonly ordered in the context of advanced or metastatic cancer, when standard treatment options have been exhausted or when comprehensive molecular profiling is being performed to identify all potential treatment targets. The test is typically not ordered for early-stage cancers where surgery or other local treatments are the primary approach.

Specific reasons your oncologist may have ordered TMB testing include:

To determine eligibility for pembrolizumab under the pan-cancer TMB-high approval. If your tumour is TMB-high (≥10 mut/Mb) and you have progressed after prior treatment and have no other satisfactory options, pembrolizumab may be an option regardless of your cancer type.
As part of comprehensive molecular profiling. Many next-generation sequencing panels automatically calculate TMB alongside all other biomarkers, so a TMB result may appear in your report as a standard component of a panel even if it was not the primary reason for testing.
To complement PD-L1 testing. In some cancer types, TMB and PD-L1 provide overlapping but distinct information about the likelihood of immunotherapy benefit. Some oncologists use both together to make a more informed decision about immunotherapy.
To look for POLE mutations or other causes of ultra-high TMB. In endometrial cancer and some other tumour types, ultra-high TMB may suggest a POLE mutation, which has distinct prognostic and treatment implications.

Who should be tested

Current guidelines recommend TMB testing for patients with advanced or metastatic solid tumours when comprehensive molecular profiling is being performed — particularly when other biomarkers (such as MMR, KRAS, BRAF, and EGFR) are being tested as part of the standard workup for that cancer type. Since most comprehensive next-generation sequencing panels automatically calculate TMB, many patients who undergo NGS panel testing will receive a TMB result without needing to request it separately.

TMB testing is not routinely recommended as a standalone test in early-stage cancers treated with surgery or local therapy. The value of TMB testing is greatest when treatment decisions — particularly about immunotherapy — are actively being made.

How the test is performed

TMB is measured using molecular testing on tumour tissue, most commonly from a biopsy or surgically removed specimen. The most widely used method in clinical practice is targeted next-generation sequencing using a validated gene panel. This test analyses several hundred cancer-relevant genes simultaneously and uses the number and type of mutations detected to estimate the overall mutation rate in the tumour. The FDA-approved companion diagnostic for the pembrolizumab TMB-high approval is the FoundationOne CDx assay, which analyses approximately 1.1 megabases of the tumour genome.

Whole-exome sequencing (WES) — which analyses all protein-coding regions of the genome — gives the most comprehensive TMB measurement, but is more expensive and slower than panel testing and is primarily used in research settings. Panel-based TMB estimates correlate closely with whole-exome TMB and are considered clinically adequate for treatment decisions.

Blood-based TMB assessment, using circulating tumour DNA from a blood sample (liquid biopsy), is an active area of research but is not yet validated for routine clinical use in most settings.

How results are reported

TMB results are typically reported in the molecular testing or genomic profiling section of your pathology report. They will usually appear in one or both of the following formats:

A numerical score in mutations per megabase (mut/Mb). For example, “TMB: 14 mut/Mb” or “TMB: 4.2 mut/Mb.” This is the raw measurement. Higher numbers indicate more mutations per unit of DNA analysed.
A categorical label: TMB-high or TMB-low. Most reports also interpret the numerical score using the FDA-approved cutpoint. A score of 10 mut/Mb or higher is reported as TMB-high (TMB-H); a score below 10 mut/Mb is reported as TMB-low (TMB-L). Some laboratories also include an intermediate category.

Some reports will include a brief comment explaining the clinical significance of the TMB result—for example, whether it suggests eligibility for pembrolizumab or recommends discussing it with an oncologist. Others will state the score and label without further interpretation.

What the result means

TMB-low (below 10 mut/Mb)

A TMB-low result indicates the tumour has a relatively low number of somatic mutations and is less likely to respond to immune checkpoint inhibitor therapy based solely on TMB. This does not mean immunotherapy is completely ruled out — eligibility for immunotherapy in most cancer types is determined primarily by other biomarkers such as PD-L1 expression and MMR/MSI status, which can indicate immunotherapy benefit even at low TMB. A TMB-low result means that TMB is not adding evidence in favour of immunotherapy for your tumour.

TMB-high (10 mut/Mb or above)

A TMB-high result means the tumour carries an above-threshold number of somatic mutations and may be more likely to respond to immune checkpoint inhibitors. Specifically, it may make your tumour eligible for pembrolizumab under the pan-cancer TMB-high approval if you have progressed after prior treatment and have no other satisfactory treatment options.

Several important caveats about a TMB-high result are worth understanding:

TMB-high does not guarantee an immunotherapy response. In the KEYNOTE-158 trial that supported the FDA approval, the overall response rate among TMB-high patients was 29% — meaning roughly 7 in 10 did not respond. TMB enriches for response; it does not predict it with certainty.
Cancer type matters. TMB-high is a more reliable predictor of immunotherapy benefit in some cancers (melanoma, lung, bladder, endometrial) than in others (colorectal MSS disease, prostate cancer, glioma). Response rates in KEYNOTE-158 ranged widely — from around 47% in endometrial cancer to lower rates in other tumour types included in the trial. Discussing what TMB-high means specifically for your cancer type with your oncologist is important.
TMB-high and dMMR/MSI-H are related but not the same. Many dMMR/MSI-H tumours are also TMB-high because the MMR deficiency allows mutations to accumulate. However, only about 16% of TMB-high tumours are also MSI-H — meaning most TMB-high tumours have their high mutation count from other causes. If your tumour is both TMB-high and dMMR/MSI-H, your oncologist will consider both findings; the dMMR/MSI-H status is typically the more reliable predictor of immunotherapy benefit in such cases.
The 10 mut/Mb threshold is not absolute. The cutoff of 10 mut/Mb was chosen based on data from a companion diagnostic assay and is a regulatory designation, not a biological bright line. Response rates are meaningfully higher in tumours with TMB above 13 mut/Mb than in those with TMB between 10 and 13 mut/Mb. Your oncologist may consider the actual numerical value, not just the high/low categorization, when weighing treatment options.

TMB and the pan-cancer pembrolizumab approval

The FDA’s June 2020 accelerated approval of pembrolizumab for TMB-high (≥10 mut/Mb) solid tumours was a significant milestone — the fourth time the FDA approved a cancer drug based on a molecular feature of the tumour rather than where in the body the cancer originated. The approval was based on data from KEYNOTE-158, a large multi-cohort clinical trial that enrolled patients with ten different types of advanced cancer. Among the 102 patients whose tumours were TMB-high, the overall response rate was 29%, and 57% of responders maintained their response for at least 12 months — a signal of durability that is particularly meaningful in advanced cancer.

The approval applies to adult and paediatric patients with any unresectable or metastatic TMB-high solid tumour, as measured by an FDA-approved test, who have progressed after prior treatment and have no other satisfactory options. It is an accelerated approval — meaning it was granted based on response rate as a surrogate marker and may be subject to confirmation in later trials. The requirement that the test be performed using an FDA-approved companion diagnostic (the FoundationOne CDx assay) is clinically important: not all TMB tests use the same methods or cutoffs, and results from different platforms may not be directly interchangeable. If your TMB result was measured on a different platform, your oncologist may consider whether the result is validated for treatment decisions.

The limitations of TMB as a biomarker

TMB is a useful but imperfect tool, and this should be understood before interpreting a result. Several important limitations are worth knowing:

TMB does not work equally well in all cancer types. In certain cancers — including prostate cancer, glioma, and some breast cancers — high TMB does not reliably predict immunotherapy benefit, possibly because the immune microenvironment of those tumours does not support the T-cell killing response even when neoantigens are present. Treating a TMB-high prostate cancer or glioma the same as a TMB-high melanoma would be an oversimplification.
Not all mutations produce useful neoantigens. TMB counts all mutations, but only a fraction of those mutations result in proteins that the immune system can effectively recognise. The quantity of mutations is a proxy for immunogenicity, not a direct measure of it.
Assay variability is a real issue. Different NGS panels cover different numbers and types of genes, use different computational methods, and may report different TMB scores for the same tumour. This means TMB values are not always directly comparable across laboratories or platforms.
High TMB from DNA repair treatment damage is different from naturally occurring high TMB. In rare cases, treatment-related mutagenesis — for example, after prior chemotherapy with certain alkylating agents — can elevate TMB in a way that does not reflect the same biology that drives immunotherapy response in treatment-naive tumours. Your oncologist will take your treatment history into account when interpreting a TMB result.

These limitations do not diminish the value of TMB as a biomarker. Still, they do mean that TMB results should always be interpreted in the context of your specific cancer type, other biomarker results, and clinical situation — not in isolation.

What happens next

If your pathology report includes a TMB result, the next steps depend on where you are in your treatment:

If your result is TMB-low, immunotherapy is less likely to be recommended based on TMB alone, but other biomarkers (PD-L1, MMR/MSI status) will still be considered. Your oncologist will review the full molecular profile and recommend treatment accordingly.
If your result is TMB-high and you are being treated for advanced or metastatic cancer, your oncologist will discuss whether pembrolizumab is an appropriate option for your situation, taking into account your cancer type, prior treatments, and other biomarker results. If pembrolizumab or another checkpoint inhibitor is already part of your planned treatment based on another biomarker (such as dMMR/MSI-H), the TMB result provides additional supporting information.
If your result shows extremely high TMB (well above 10 mut/Mb), your oncologist may discuss whether a POLE mutation or other specific cause has been identified and what this means for treatment options and prognosis.

If you are uncertain about what a TMB result means for your specific cancer, asking your oncologist to walk through the result in the context of all your other molecular findings is the most useful approach.

Questions to ask your doctor

What was my TMB score — and does it qualify as TMB-high by the FDA-approved definition?
What does a TMB-high result mean for my specific cancer type? Is TMB a reliable predictor of immunotherapy benefit in my cancer?
Was my TMB measured using the FoundationOne CDx assay or another validated platform?
Is my tumour also dMMR/MSI-H — and if so, how do those two findings interact?
Does my TMB result make me eligible for pembrolizumab under the pan-cancer approval?
Are there clinical trials relevant to my TMB result that I should consider?
If my TMB is very high, could a POLE mutation be the cause — and if so, what does that mean for my prognosis and treatment?
Are there other biomarkers in my report that I should be understanding alongside my TMB result?