FDA Converts Pembrolizumab's Tumor-Agnostic Approval to Full Status, Confirming Durable Responses in High-Mutation-Burden Cancers

On October 16, 2026, the FDA converted pembrolizumab's (Keytruda) 2020 accelerated approval for tumor mutational burden-high (TMB-H) solid tumors into full, traditional approval — a milestone that confirms, with mature follow-up data, that a meaningful share of patients with highly mutated cancers get durable benefit from checkpoint immunotherapy regardless of where their tumor started. The upgrade is genuinely good news for patients who respond, but it comes with an important and under-reported catch: the FDA simultaneously narrowed who qualifies, raising the TMB threshold from 10 to 13 mutations per megabase after post-approval data failed to confirm benefit in the lower range. This article explains what changed, what the data actually show, and who may now lose eligibility they previously had.

What Happened

Pembrolizumab first received accelerated approval for TMB-H solid tumors in June 2020 — one of oncology's earliest "tumor-agnostic" approvals, meaning eligibility was based on a tumor's genetic signature rather than its organ of origin. Accelerated approval is conditional: it allows a drug onto the market based on a response-rate signal, with the requirement that a confirmatory trial later verify real clinical benefit.

That confirmation has now arrived. On October 16, 2026, the FDA converted the indication to traditional approval for adult and pediatric patients with unresectable or metastatic TMB-H solid tumors who have progressed after prior treatment and have no satisfactory alternative options, using an FDA-authorized test to determine TMB status. The action was reported by OncLive, Targeted Oncology, The ASCO Post, CURE Today, Oncology Nursing News, and Cancer Network, and it also appears on the FDA's own oncology approval notifications list for October 2026.

The Data Behind the Conversion

The traditional approval rests on pooled data from KEYNOTE-158 (NCT02628067), an open-label, non-randomized trial:

• Cohort M prospectively enrolled 145 patients with TMB-H, non-MSI-H tumors.
• Cohorts A–J contributed a prospectively planned retrospective analysis of 75 additional TMB-H patients drawn from ten cohorts of previously treated solid tumors (melanoma and non-small cell lung cancer were excluded from the trial).
• Across all 220 patients with TMB of at least 13 mutations per megabase, the objective response rate was 36% (95% CI, 30%–43%).
• Notably, 58% of those who responded had a response lasting at least 12 months or longer — a sign that when pembrolizumab works in this setting, the benefit tends to be durable rather than fleeting.

That compares to the original 2020 accelerated-approval dataset of 102 patients, which showed a 29% ORR and 57% of responses lasting at least a year. The new analysis, with more than double the patient numbers and longer follow-up, essentially reproduces — and modestly strengthens — the original signal.

The Catch: A Narrower Definition of "High" Mutational Burden

This is where the story requires real nuance rather than celebration alone. The accelerated approval in 2020 defined TMB-H as 10 or more mutations per megabase. In reviewing the confirmatory data, the FDA concluded that efficacy was not confirmed for tumors in the 10-to-under-13 mutations-per-megabase range. As a result, traditional approval applies only to patients at 13 mut/Mb or higher.

Practically, this means some patients who previously qualified for pembrolizumab under the TMB-H label — those sitting in that 10–12 mut/Mb band — no longer meet the criteria for this specific indication going forward. That is not a safety reversal and does not mean pembrolizumab stopped working for anyone who already benefited from it; it means the agency is holding the tumor-agnostic label to the same evidentiary bar used everywhere else in oncology, and the data simply didn't support extending that specific approval pathway to the lower-burden group. Patients and oncologists will need to confirm TMB status under the updated, FDA-authorized testing threshold.

What It Means for Patients

For patients whose tumors carry a TMB of 13 mutations per megabase or higher and who have exhausted standard options, this conversion is a reassuring development: it means the benefit seen six years ago wasn't a statistical fluke, and clinicians can now rely on a confirmed, non-conditional approval when considering pembrolizumab for genetically complex, treatment-resistant cancers regardless of tumor site. A roughly one-in-three response rate, with more than half of responders maintaining benefit past a year, is a meaningful option for people who have already tried other therapies and have "no satisfactory alternative."

For patients in the 10–12 mut/Mb range, the practical effect is less certain and worth discussing directly with an oncologist — eligibility under this specific label has narrowed, though other pembrolizumab indications or clinical trial options may still be relevant depending on the specific cancer type.

What's Still Unknown or Unresolved

Several caveats are worth keeping in view:

• The underlying trial (KEYNOTE-158) was single-arm and non-randomized, not a head-to-head comparison against another active treatment, so the 36% ORR describes response to pembrolizumab alone rather than a direct superiority claim.
• Cohorts A–J were analyzed retrospectively, even though the overall approval strategy was prospectively planned — a real but secondary limitation compared with the prospectively enrolled Cohort M.
• The article coverage reviewed for this piece did not break down response rates by specific tumor type within the TMB-H population, so it is unclear whether certain cancers respond more or less reliably than others within this biomarker-defined group.
• Known checkpoint-inhibitor risks remain unchanged: labeling continues to warn of immune-mediated adverse reactions, infusion-related reactions, complications related to allogeneic stem cell transplantation, and embryo-fetal toxicity. The overall safety profile was described as consistent with pembrolizumab's established monotherapy data, not safer or riskier than before.

Bottom Line

This is a genuine, evidence-based step forward for a biomarker-driven approach to cancer immunotherapy: six years and hundreds of additional patients after its conditional launch, pembrolizumab's TMB-H indication has now cleared the FDA's higher bar for full approval, with durable responses in roughly a third of eligible patients. But the honest version of this story isn't unambiguous good news for everyone who previously qualified — the agency tightened the mutational-burden cutoff because the data didn't support the lower range, which is exactly how evidence-based regulation is supposed to work, even when it narrows access for some. Patients with highly mutated, treatment-resistant solid tumors and their oncologists should revisit TMB testing results against the new 13 mut/Mb threshold to understand where they stand.