FDA Narrows Pembrolizumab’s TMB-High Cancer Approval

Pembrolizumab (Keytruda / Keytruda Qlex)What to know

30-second read

For Oncology · Pathology

Who

Adults and pediatric patients with unresectable or metastatic solid tumors with no satisfactory alternative treatment, now defined as TMB–high at 13 or more mutations/megabase.

What changed

FDA converted the June 2020 accelerated approval to traditional (full) approval on October 16, 2026 — but raised the TMB–H cutoff from ≥10 to ≥13 mut/Mb.

Evidence

KEYNOTE-158: in 220 patients with TMB ≥13 mut/Mb, objective response rate was 36% (95% CI, 30–43%); 58% of responses lasted at least 12 months.

Key consideration

Patients with TMB of 10 to 12.9 mut/Mb — eligible under the old criterion — no longer meet the approved indication for this biomarker pathway.

Availability

Applies to both intravenous Keytruda and subcutaneous Keytruda Qlex. No new test is required: FoundationOne CDx has reported both cutpoints since 2020.

Practice impactRecheck TMB value against new cutoff

Before continuing or starting pembrolizumab monotherapy under the TMB-H pathway, pull the quantitative mutations-per-megabase value from the pathology report. A categorical “TMB-High” result is no longer sufficient to establish that a patient meets the label.

Read the full physician analysis

Full analysis · 9 min readFDA approval notice, October 16, 2026 · KEYNOTE-158 (NCT02628067) · Keytruda prescribing information

The U.S. Food and Drug Administration converted pembrolizumab’s tissue-agnostic approval for tumor mutational burden–high (TMB–H) solid tumors from accelerated to traditional approval on October 16, 2026 — six years after the original 2020 clearance made it the second biomarker-defined, tumor-type–independent indication in oncology. The conversion comes with a catch that oncologists need to register immediately: the agency also raised the qualifying TMB threshold from 10 mutations per megabase to 13, which means some patients who qualified under the original label no longer do.

A permanent approval, over a smaller population

Accelerated approval let Merck (known outside the U.S. and Canada as MSD) market pembrolizumab for unresectable or metastatic solid tumors with no satisfactory alternative treatment, selected only by a biomarker — tumor mutational burden — rather than by organ of origin. The condition attached to that approval was that confirmatory evidence had to support the clinical benefit before the FDA would make it permanent. According to the FDA’s October 16 notice, that confirmation held up for patients whose tumors carry 13 or more mutations per megabase, but not for the lower end of the original range: efficacy supporting traditional approval was not confirmed in patients with TMB from 10 to just under 13 mut/Mb. Rather than withdraw the indication, the FDA narrowed it and made the narrower version permanent.

The distinction between accelerated and traditional approval is not a formality for prescribers. Accelerated approval can be withdrawn if confirmatory data disappoint; traditional approval cannot be revoked on that basis. Oncologists now have a durable approval to rely on — for a population defined more narrowly than the one they may have grown used to since 2020.

Where the number 13 comes from

The 13 mut/Mb cutpoint is not new; it was one of two thresholds FDA reviewers examined when they first evaluated pembrolizumab in TMB–H tumors. The pivotal evidence both times is KEYNOTE-158, Merck’s open-label, multi-cohort study of pembrolizumab across advanced solid tumors that had progressed on standard therapy. In 2020, the approval rested on 102 patients with TMB of at least 10 mut/Mb across ten tumor-type cohorts, who had an objective response rate of 29%. Within that same group, the patients whose tumors measured at least 13 mut/Mb already showed a higher response rate on the original analysis. The confirmatory analysis behind this month’s decision formalizes that split, following a larger group of patients selected specifically at the 13 mut/Mb cutpoint.

KEYNOTE-158: the confirmatory population

220Patients with TMB ≥13 mut/Mb across Cohort M (145 patients) and Cohorts A–J (75 patients)

36%Objective response rate (95% CI, 30–43%) in that group, supporting traditional approval

58%Of responders had a response lasting 12 months or longer

29%Objective response rate in the original 2020 cohort (n=102, TMB ≥10 mut/Mb) — the population this approval leaves behind

Figures as reported in the FDA's October 16, 2026 approval notice and corroborating oncology trade coverage of KEYNOTE-158 (NCT02628067); prescribe from the current Keytruda label, not this summary.

The patients caught in between

The practical consequence is a biomarker gap: adults and children whose solid tumors measure 10, 11 or 12 mutations per megabase — TMB–High by the definition oncologists have used for six years — no longer meet the approved indication for pembrolizumab monotherapy on this pathway alone. For a newly presenting patient, that simply means checking the specific number before invoking TMB–H as the basis for treatment; other approved pembrolizumab indications tied to tumor type, PD-L1 expression or MSI–H/dMMR status are unaffected and may still apply. For a patient already being treated under the old criterion whose report falls in the 10–12.9 range, the FDA has not ordered treatment to stop, but continued use no longer has an on-label basis under this indication. That is a conversation to have with the patient directly, and a decision worth documenting, rather than a change to make unilaterally based on a news report.

The test hasn't changed. Reading the result has to.

No new companion diagnostic requirement comes with this decision. FoundationOne CDx, the next-generation sequencing assay FDA approved alongside the original 2020 indication, has reported both the 10 mut/Mb and 13 mut/Mb cutpoints on its validated TMB output since that approval, because the agency examined both thresholds from the start. The practical task for oncology practices is not re-testing patients or revalidating a new assay; it is making sure the ordering physician, not just the pathology report, is reading the number rather than the category. A chart or an electronic order set that still fires on a binary “TMB–High: Yes” flag will now overstate eligibility for roughly the bottom third of patients who would have qualified a month ago.

The TMB-H indication, before and after October 16

	2020 accelerated approval	2026 traditional approval
Approval status	Accelerated — contingent on confirmatory evidence	Traditional — permanent, not contingent on further trials
Qualifying TMB	≥10 mutations/megabase	≥13 mutations/megabase
Supporting population	102 patients, KEYNOTE-158	220 patients, KEYNOTE-158 (a subset selected at the higher cutpoint)
Objective response rate	29% (95% CI, 21–39%)	36% (95% CI, 30–43%)
Products covered	Keytruda (IV) only — Keytruda Qlex did not yet exist	Keytruda (IV) and Keytruda Qlex (subcutaneous)
Companion diagnostic	FoundationOne CDx, approved concurrently	FoundationOne CDx, unchanged — already reports both cutpoints

Prescribe from the current Keytruda and Keytruda Qlex labels, not this summary; this table compares the regulatory record, not clinical guidance.

Reimbursement follows the label, and the label just moved

Regulatory authorization and payer coverage are not the same decision, and nothing about billing mechanics changes here: pembrolizumab is billed under its existing J-code as a medical-benefit drug regardless of which indication supports its use, and that does not change with this label update. What can change is medical necessity review. Commercial payers and Medicare Administrative Contractors that cover pembrolizumab for the TMB–H indication typically tie coverage to the FDA label or to compendia such as the NCCN Drugs & Biologics Compendium, both of which will reflect the narrower 13 mut/Mb cutoff going forward. A prior authorization or claim built on a 10–12.9 mut/Mb result is now harder to defend on this pathway specifically, even though it may have been approved without question a month ago. This story does not establish how quickly individual payers will update their policies; practices billing under this indication should confirm current payer policy rather than assume continuity.

What this means at the point of care

• New patients. Confirm the quantitative TMB value, not just a categorical result, before citing TMB–H as the basis for pembrolizumab monotherapy when no other approved indication applies.
• Patients already on treatment. Pull the original TMB value for anyone started under this indication; a result of 10 to 12.9 mut/Mb is now outside the label, which calls for a documented discussion with the patient rather than an immediate stop.
• Order sets and pathways. Update any internal decision support, clinical pathway or order set that still triggers on a binary TMB–High flag rather than the numeric cutpoint.
• Billing and prior authorization. Re-verify payer medical-necessity criteria for this indication; claims built on the 10 mut/Mb threshold are more exposed going forward.
• No change to testing. FoundationOne CDx and other validated TMB assays already report the 13 mut/Mb cutpoint; this is a label change, not a new diagnostic requirement.
• Other indications unaffected. Pembrolizumab's tumor-specific, PD-L1–based and MSI–H/dMMR indications are governed by their own criteria and are not touched by this action.

The headline on this approval is permanence, but the practical story is a narrower population carrying it. Oncologists who built workflows around “TMB–H means 10 or above” have one number to update — and one conversation to have with any patient whose report sits in the gap that just opened between the old criterion and the new one.

From a biomarker signal to a permanent, narrower label

1. 2014

   KEYNOTE-158 opens

   Merck begins the open-label, multi-cohort study of pembrolizumab across advanced solid tumors that would become the basis for both TMB-H decisions.

2. Jun 2020

   Original accelerated approval

   FDA clears pembrolizumab for TMB-H solid tumors (≥10 mut/Mb) with no satisfactory alternative, the second tissue-agnostic biomarker indication after MSI-H/dMMR.

3. Sep 2025

   Keytruda Qlex approved

   The subcutaneous co-formulation with berahyaluronidase alfa-pmph is cleared across most Keytruda solid-tumor indications.

4. Oct 16, 2026

   Traditional approval, narrowed

   FDA converts the TMB-H indication to full approval, raising the qualifying threshold to ≥13 mut/Mb for both Keytruda and Keytruda Qlex.

5. Ahead

   Payer and pathway updates

   Compendia listings, payer medical-necessity policy and institutional order sets are expected to catch up to the narrower cutoff on their own timelines.

Frequently asked questions

My patient's pathology report says only "TMB-High," not a number. What should I do?

Ask the lab for the quantitative value in mutations per megabase. Since 2020, FoundationOne CDx and other validated assays have reported both the 10 mut/Mb and 13 mut/Mb cutpoints; a categorical “TMB-H” flag alone no longer tells you whether a patient meets the current labeled indication.

Does this affect a patient who already started pembrolizumab monotherapy under the old TMB-H criterion?

It can. If that patient's tumor measured 10 to 12.9 mut/Mb, they no longer fall within the approved indication, even though nothing about their tumor or their response has changed. There is no FDA directive to stop treatment in a patient already responding, but continuing is now off-label for this indication, which can affect coverage and warrants a documented discussion with the patient.

Does the new 13 mut/Mb cutoff apply to the subcutaneous formulation, Keytruda Qlex, too?

Yes. The October 16 action covers both pembrolizumab (Keytruda, intravenous) and pembrolizumab and berahyaluronidase alfa-pmph (Keytruda Qlex, subcutaneous) for this indication; the two products carry the same TMB-H criteria and the same approval status.