Pembrolizumab Gains Traditional Approval in TMB-High Tumors

What the approval changes

The October 16, 2026, action places pembrolizumab under traditional approval for a tumor-agnostic indication defined by tumor mutational burden, or TMB, rather than the organ where the cancer originated. The indication applies across eligible solid tumors, but only when every label condition is met.

The threshold is at least 13 mutations per megabase. Patients must have unresectable or metastatic disease that has progressed following prior treatment, and they must lack satisfactory alternative treatment options. The approval includes adults and children, making age alone insufficient to determine eligibility or exclusion.

Traditional approval is a regulatory conclusion that the statutory standard for effectiveness has been met. It differs from accelerated approval, which permits earlier authorization based on a surrogate or intermediate endpoint considered reasonably likely to predict clinical benefit and ordinarily requires confirmatory evidence. The conversion does not mean every TMB-high tumor will respond, nor does it make pembrolizumab a preferred first-line option across cancers.

The revised 13-mutations/Mb cutoff is clinically important. The original 2020 accelerated approval used a threshold of at least 10 mutations/Mb, as determined by an FDA-approved test. Clinicians should therefore not assume that documentation supporting the earlier cutoff automatically establishes eligibility under the traditional approval.

Evidence supporting a biomarker-defined indication

The foundational evidence came from KEYNOTE-158, a multicohort, open-label phase 2 study without a randomized comparator. The analysis included previously treated patients with advanced solid tumors and prospectively evaluated outcomes by tissue TMB. Because the study was nonrandomized, differences between biomarker groups cannot establish that TMB itself caused the observed difference in response.

In the FDA efficacy population supporting the original accelerated approval, 102 patients had tumors measuring at least 10 mutations/Mb. The objective response rate by blinded independent central review was 29%, with a 95% confidence interval of 21% to 39%. Responses were often durable: 57% lasted at least 12 months and 50% lasted at least 24 months.

An exploratory breakdown reported a higher response rate among tumors with at least 13 mutations/Mb than among tumors between 10 and less than 13 mutations/Mb. These historical findings provide context for the narrower threshold, although the FDA’s 2026 regulatory determination, rather than this subgroup analysis alone, defines the current indication.

These are single-arm response data, not comparative estimates of survival benefit. They also combine several tumor types, each represented by a relatively small number of patients. A pooled response rate can therefore obscure substantial variation among cancers.

Applying the eligibility criteria

Testing comes first. TMB is an assay-derived count of qualifying somatic mutations normalized to the genomic territory analyzed and reported as mutations per megabase. Results can vary with panel size, sequencing depth, bioinformatic filtering, specimen quality and whether tumor-only or matched normal sequencing is used. A value from one platform should not automatically be treated as interchangeable with the same numeric value from another.

The operative result should come from testing that satisfies the current FDA label and companion-diagnostic requirements. Reports should identify the assay, specimen and numeric result rather than state only that a tumor is “TMB-high.” An older report using a 10-mutations/Mb cutoff may require review or repeat testing if it does not provide enough information to determine whether the tumor reaches 13 mutations/Mb.

TMB is not interchangeable with microsatellite instability, mismatch-repair deficiency or PD-L1 expression. These biomarkers may coexist, but each has its own assay methods, evidence base and labeled indications. A tumor that is microsatellite instability-high may qualify for pembrolizumab through a separate indication even when a usable TMB result is unavailable.

The treatment-history language also matters. The tumor must have progressed following prior treatment. This makes the indication a later-line option rather than a general authorization to substitute pembrolizumab for established initial therapy solely because TMB is elevated.

“No satisfactory alternative treatment options” requires clinical judgment grounded in the tumor type, molecular profile, prior therapies and the patient’s condition. Relevant alternatives may include approved systemic regimens, another biomarker-directed treatment, surgery or radiation when feasible, and an appropriate clinical trial. An alternative is not necessarily satisfactory merely because it exists, but the rationale for concluding that alternatives are unavailable or unsuitable should be documented.

Including pediatric patients does not eliminate the need to consider the limited representation of children in tumor-agnostic datasets. Tumor biology, mutation burden distributions and available standard therapies differ by age and cancer type. Safety considerations also remain those of immune checkpoint inhibition, including potentially serious immune-mediated toxicities described in the prescribing information.

Important uncertainties

The historical efficacy dataset was small, nonrandomized and heterogeneous. It was designed to evaluate antitumor activity, not to compare pembrolizumab with tumor-specific standard therapy or demonstrate an overall-survival advantage. Objective response is clinically meaningful when durable, but it does not capture stable disease, symptom effects or comparative quality of life.

Generalizability is another limitation. Some tumor types were not represented in the foundational analysis, and response may differ according to cancer lineage, the mutations contributing to TMB and features of the tumor microenvironment. A threshold creates a practical regulatory boundary, not a biological discontinuity between 12.9 and 13 mutations/Mb.

Assay variability remains consequential near that boundary. Low tumor purity, limited tissue and differences in mutation filtering can alter the reported value. Clinicians and laboratories should resolve ambiguous or borderline results before using TMB as the sole basis for treatment eligibility.

Finally, traditional approval should not be interpreted as evidence that TMB supersedes tumor-specific guidelines. The approval supports pembrolizumab as an option for a defined population after progression and when satisfactory alternatives are lacking. It does not establish superiority over an available standard regimen.

Questions clinicians ask

Does a result of 13 mutations/Mb establish eligibility by itself?

No. The numeric threshold is only one requirement. The patient must also be an adult or child with an unresectable or metastatic solid tumor that progressed after prior treatment, with no satisfactory alternative options, and the result must meet the testing provisions of the current FDA label.

Can a report labeled TMB-high at 10 mutations/Mb still be used?

Not automatically. The traditional approval described here uses at least 13 mutations/Mb. If an older report gives the exact value and the assay meets current requirements, it may be interpretable; a categorical result based only on the former 10-mutations/Mb cutoff may not be sufficient.

How should “no satisfactory alternatives” be interpreted?

It is a patient- and tumor-specific assessment, not a declaration that no other treatment exists. Clinicians should consider approved tumor-specific regimens, actionable biomarkers, local treatment, prior intolerance, contraindications and clinical trials, then document why remaining options are unavailable or not satisfactory.

What does the pediatric inclusion establish?

It permits use within the labeled pediatric population when all disease, testing and treatment-history criteria are met. It does not imply that children were extensively represented in each tumor cohort, so tumor-specific evidence, developmental considerations and the known risks of immune checkpoint blockade remain important.

References

1. FDA Grants Traditional Approval for Pembrolizumab in TMB-High Solid Tumors — U.S. Food and Drug Administration, 2026 2. FDA Approval Summary: Pembrolizumab for the Treatment of Tumor Mutational Burden-High Solid Tumors — Clinical Cancer Research, 2021 3. Study of Pembrolizumab in Participants With Advanced Solid Tumors (KEYNOTE-158) — ClinicalTrials.gov, 2015 4. FDA Approves Pembrolizumab for Adults and Children With TMB-H Solid Tumors — U.S. Food and Drug Administration, 2020