First-line pembrolizumab versus carboplatin plus paclitaxel in participants with mismatch repair-deficient advanced or recurrent endometrial cancer (KEYNOTE-C93/GOG-3064/ENGOT-en15): interim analysis of a multicentre, open-label, randomised, active-comparator controlled, phase 3 trial.

Endometrial cancer is the most common gynecologic malignancy in high-income countries, with incidence continuing to rise alongside the obesity epidemic, and patients with advanced disease have a poor prognosis^[1]. The mismatch repair–deficient (dMMR) subtype is sensitive to immune checkpoint inhibitors, a property already validated in the first-line treatment of colorectal cancer: pembrolizumab monotherapy significantly prolonged progression-free survival compared with chemotherapy^[2]. In endometrial cancer, multiple phase 3 studies have established immunotherapy combined with chemotherapy as the standard of care; dostarlimab^[3], pembrolizumab^[4], durvalumab^[5], and atezolizumab^[6] have all demonstrated significant benefit in dMMR populations, but all require concomitant platinum–paclitaxel chemotherapy, which brings typical toxicities such as alopecia, peripheral neuropathy, and myelosuppression, along with fixed treatment cycles and substantial impact on quality of life. There is an urgent clinical need to clarify: for dMMR advanced or recurrent endometrial cancer, can immunotherapy monotherapy replace chemotherapy in the first-line setting to achieve "chemotherapy-free" treatment?

KEYNOTE-C93/GOG-3064/ENGOT-en15 is a multicenter, open-label, actively controlled phase 3 trial designed precisely for this purpose, enrolling 299 patients with unresectable stage III/IV or persistent/recurrent dMMR endometrial cancer at 122 centers across 24 countries, randomized 1:1 to receive pembrolizumab 400 mg intravenously every 6 weeks (up to 18 cycles) or carboplatin AUC 5-6 plus paclitaxel 175 mg/m² intravenously every 3 weeks (6 cycles)^[7]. The prespecified interim analysis (data cutoff April 24, 2026) showed that in the intention-to-treat population, median progression-free survival reached 38.2 months (95% CI 17.9-not reached) in the pembrolizumab group versus only 8.4 months (6.8-10.4) in the control group, with a hazard ratio of 0.57 (95% CI 0.42-0.79; p=0.0002). Progression-free survival rates at 6, 12, and 36 months were 75% vs 76%, 67% vs 40%, and 53% vs 30%, respectively, with the separation of the curves widening over time, suggesting long-term benefit potential for immunotherapy monotherapy. Overall survival showed a numerical advantage (HR 0.77, 95% CI 0.52-1.15) but did not cross the prespecified statistical significance boundary (p=0.10), and the medians were not reached in either group.

The safety differences were striking: grade 3 or higher treatment-related adverse events occurred in 18% (27/149) of the pembrolizumab group versus 51% (73/144) of the chemotherapy group. Chemotherapy-specific alopecia (50% vs 2%), anemia (49% vs 5%), neutropenia (28% vs 3%), and peripheral neuropathy (26% vs 0%) nearly disappeared in the immunotherapy monotherapy group; fatigue was seen in both groups (26% vs 13%). Discontinuation due to adverse events was 9% (14/149) in the immunotherapy group, lower than 15% (21/144) in the chemotherapy group; treatment-related deaths numbered 2 in the immunotherapy group (1 hepatic failure and 1 bacterial meningitis) and 1 in the chemotherapy group (pulmonary embolism). Although quality-of-life data were not detailed in the abstract, the differences in toxicity profile strongly suggest that immunotherapy monotherapy can substantially improve patients' treatment experience.

This study demonstrated at the phase 3 level that first-line immunotherapy monotherapy in dMMR advanced endometrial cancer delivers both statistically and clinically significant PFS benefit compared with standard chemotherapy, with treatment completed with markedly better tolerability, providing high-level evidence for a "chemotherapy-free" strategy. However, the authors explicitly note that the study did not directly compare with the current standard of care (immunotherapy plus chemotherapy), and OS data remain immature, with long-term follow-up ongoing (NCT05173987)^[8]. Going forward, attention should be paid to the final OS results, the impact of subsequent therapies on the survival curves, and further biomarker selection of the patients most likely to benefit, in order to determine the precise positioning of immunotherapy monotherapy in the first-line treatment algorithm.