Pembrolizumab alone beat chemotherapy as first-line treatment for dMMR endometrial cancer

For years, the first treatment for advanced endometrial cancer has been chemotherapy — carboplatin and paclitaxel — whatever the tumour's biology. A phase 3 trial published in The Lancet on 26 October asks whether a subset of these cancers would do better starting on immunotherapy alone, skipping chemotherapy entirely [s1].

Why this subset is different

Roughly a quarter to a third of endometrial cancers are mismatch repair-deficient, written dMMR. The cell's system for correcting copying errors in DNA is broken, so the tumour accumulates large numbers of mutations. That makes it conspicuous to the immune system — and unusually responsive to drugs that take the brakes off immune cells. Pembrolizumab is one of those drugs, blocking the PD-1 checkpoint [s1].

Pembrolizumab already had a role in endometrial cancer, but the question KEYNOTE-C93 set out to answer was more pointed: in the first-line setting, for dMMR disease specifically, is the antibody on its own better than the chemotherapy that has long been standard?

How the trial was built

KEYNOTE-C93, also known as GOG-3064 and ENGOT-en15, was a multicentre, open-label, randomised, active-comparator phase 3 trial run at 122 centres across 24 countries, registered as NCT05173987 [s1][s2]. It enrolled 299 participants with dMMR advanced or recurrent endometrial cancer who had not had systemic treatment for advanced disease [s1].

They were randomly assigned to one of two arms: 149 to pembrolizumab, given at 400 mg intravenously every six weeks for up to 18 cycles, and 150 to carboplatin plus paclitaxel, the standard chemotherapy doublet given every three weeks for six cycles [s1]. Enrolment ran from February 2022 to December 2023 [s1].

The trial had two primary endpoints — progression-free survival, read by a blinded independent central review panel that did not know which treatment a patient was on, and overall survival [s1]. Testing both matters: a drug can delay what shows up on a scan without helping people live longer, and endometrial cancer trials have been caught out by that gap before. This is an interim analysis, not the final read; data were cut on 24 April 2026 [s1].

What it found

On progression-free survival, the gap was large. Median progression-free survival was 38.2 months with pembrolizumab against 8.4 months with chemotherapy, a hazard ratio for progression or death of 0.57 (95% CI 0.42–0.79; p=0.0002) [s1]. The shape of the benefit is worth reading closely. At six months the two groups looked almost identical — 75% of the pembrolizumab group and 76% of the chemotherapy group were alive without progression — because chemotherapy works briskly at first [s1]. From then on the curves diverged sharply: at 12 months, 67% of the pembrolizumab group were alive without progression versus 40% on chemotherapy, and at 36 months, 53% versus 30% [s1]. The advantage of the immunotherapy is durability, not a fast start.

Overall survival, the other primary endpoint, tells a more cautious story. The hazard ratio for death was 0.77 (95% CI 0.52–1.15; p=0.10), with median survival not reached in either arm — a difference that did not cross the prespecified significance threshold at this interim look [s1]. In plain terms: the trial has clearly shown that pembrolizumab delays progression, but it has not yet shown that it helps people live longer.

The safety contrast favoured the antibody. Grade 3 or higher treatment-related adverse events occurred in 18% of the pembrolizumab group versus 51% of the chemotherapy group [s1]. For a first-line treatment given to people who may stay on it for a long time, the lower burden of severe side effects is itself a meaningful finding.

What is, and is not, established

What KEYNOTE-C93 supports is that, for dMMR advanced or recurrent endometrial cancer specifically, starting with pembrolizumab alone produced far longer progression-free survival and markedly fewer severe treatment-related side effects than the standard chemotherapy doublet [s1]. That is a strong case for using the immunotherapy first in this biologically defined group.

The qualifications matter. This is an interim analysis, and overall survival — the endpoint that most directly measures benefit to patients — has not yet reached statistical significance; longer follow-up is needed to see whether the progression advantage translates into longer life [s1]. The trial was open-label, which is precisely why the blinded central review of scans is the part of the design that the lack of masking could not distort [s1]. And the findings apply only to the dMMR subset — they say nothing about the larger group of endometrial cancers whose mismatch repair machinery is intact [s1].

Red flags: anyone on a checkpoint inhibitor such as pembrolizumab who develops new breathlessness, persistent diarrhoea, a rash, or unusual fatigue should contact their oncology team promptly, as these can signal immune-related side effects. This article describes research and is not medical advice. First-line treatment choices in endometrial cancer are for patients and their treating oncologists.