Results from the phase III NRG-GY020 trial, presented by Floor Backes at ESMO 2026 during the Oral Session on Gynaecological Cancers (Abstract 1239O), showed that adding pembrolizumab to adjuvant radiotherapy did not significantly improve recurrence-free survival in patients with newly diagnosed, early-stage, high-intermediate-risk (HIR), mismatch repair-deficient (dMMR) endometrioid endometrial cancer.
Despite the established activity of immune checkpoint inhibitors in recurrent dMMR endometrial cancer, the addition of pembrolizumab did not meet the trial’s prespecified criteria for superiority and was associated with increased grade 3–4 toxicity.
Background
Adjuvant vaginal brachytherapy is a recommended treatment for selected patients with high-intermediate-risk early-stage endometrial cancer, reducing the risk of vaginal recurrence following surgery.
Immune checkpoint inhibitors, including pembrolizumab, have demonstrated substantial clinical activity in patients with recurrent or advanced dMMR endometrial cancer. However, whether immunotherapy can improve outcomes when added to adjuvant radiotherapy in early-stage disease remains uncertain.
The NRG-GY020 trial was designed to investigate whether adding pembrolizumab to radiation therapy could reduce recurrence risk in patients with early-stage, HIR, dMMR endometrioid endometrial cancer.
Study Design and Patient Population
NRG-GY020 was a randomized, open-label, phase III superiority trial that enrolled 168 patients between February 2020 and August 2022.
Eligible patients had newly diagnosed, early-stage endometrioid endometrial cancer with mismatch repair deficiency and high-intermediate-risk features defined by GOG-249 criteria.
Patients were randomized 2:1 to receive:
- Radiotherapy plus pembrolizumab: 112 patients.
- Radiotherapy alone: 56 patients.
All patients received vaginal brachytherapy (VBT). External beam pelvic radiotherapy (EBRT) was permitted for patients with stage II disease or stage IB grade 3 tumors.
Pembrolizumab was administered at 400 mg intravenously every six weeks for one year, beginning within seven days of radiotherapy.
The primary endpoint was the proportion of patients achieving recurrence-free survival at three years (RFS3).
Importantly, the primary analysis used a prespecified composite definition in which patients who experienced recurrence, died, or discontinued study participation for another reason before three years were counted as treatment failures.
The median patient age was 66 years. Regarding disease stage, 20% had stage IA, 63% had stage IB, and 17% had stage II disease.
Most patients (82%) had grade 2 or 3 tumors.
Approximately 70% received vaginal brachytherapy alone, while 30% received pelvic EBRT with or without brachytherapy.
Three-Year Recurrence-Free Survival Results
At a median follow-up of 45.5 months, the addition of pembrolizumab did not improve the prespecified primary endpoint.
Three-year recurrence-free survival according to the primary analysis was 70% with radiotherapy plus pembrolizumab versus 73% with radiotherapy alone (P=0.68).
The trial therefore failed to meet its primary superiority endpoint.
However, the individual components of treatment failure differed between groups.
Disease recurrence occurred in:
- 12 of 112 patients (11%) receiving radiotherapy plus pembrolizumab.
- 10 of 56 patients (18%) receiving radiotherapy alone.
Additional treatment failures for reasons other than recurrence or death occurred in 22 patients (20%) in the pembrolizumab group versus five patients (9%) in the radiotherapy-alone group.
These differences are important because the trial’s primary endpoint counted patients who discontinued study participation before three years as treatment failures, even without documented recurrence.
Kaplan–Meier Survival Analysis
In a secondary time-to-event analysis using Kaplan–Meier estimates, three-year recurrence-free survival was numerically higher with pembrolizumab.
The estimated rates were:
- Radiotherapy plus pembrolizumab: 88%.
- Radiotherapy alone: 81%.
The hazard ratio was 0.72 (95% CI, 0.33–1.61; P=0.21).
Although the numerical results favored pembrolizumab, this difference was not statistically significant, and the wide confidence interval reflected uncertainty around the treatment effect.
The discrepancy between the primary RFS3 analysis and the Kaplan–Meier estimates reflects differences in how treatment discontinuation and other study failures were handled. These findings do not establish a recurrence-free survival benefit from adding pembrolizumab.
Overall Survival Outcomes
Overall survival was excellent in both treatment groups. Three-year overall survival was 96% with radiotherapy plus pembrolizumab and 96% with radiotherapy alone.
No overall survival advantage was demonstrated with the addition of immunotherapy. The high survival rates in both groups also highlight the importance of carefully balancing potential treatment benefits against additional toxicity in the early-stage setting.
Safety Findings
Adding pembrolizumab resulted in substantially more severe adverse events. Grade 3–4 toxicity occurred in 43% of patients receiving radiotherapy plus pembrolizumab compared with 15% receiving radiotherapy alone.
The most frequently reported severe toxicities in the combination arm included grade 3 lymphopenia and hypertension. These findings indicate a considerable increase in treatment-related burden without a demonstrated improvement in the primary efficacy endpoint.
Clinical Implications
The NRG-GY020 results provide important evidence regarding the use of immunotherapy in early-stage dMMR endometrial cancer.
Although mismatch repair deficiency is an established predictive biomarker for benefit from immune checkpoint inhibitors in advanced and recurrent disease, this biological characteristic alone did not translate into improved outcomes when pembrolizumab was added to adjuvant radiotherapy in this trial. The findings also emphasize the importance of considering the specific disease setting when integrating immunotherapy into multimodality treatment.
In early-stage endometrial cancer, where survival outcomes are already favorable, the threshold for accepting additional toxicity is particularly important.
While the secondary Kaplan–Meier analysis showed a numerical difference favoring pembrolizumab, the study did not meet its prespecified primary endpoint, and the secondary results were not statistically significant. The trial therefore does not support routine addition of pembrolizumab to adjuvant radiotherapy in patients meeting the NRG-GY020 eligibility criteria.
Conclusion
The phase III NRG-GY020 trial demonstrated that adding one year of pembrolizumab to adjuvant radiotherapy did not significantly improve three-year recurrence-free survival in patients with early-stage, high-intermediate-risk dMMR endometrioid endometrial cancer.
The prespecified three-year RFS rates were 70% with pembrolizumab plus radiotherapy versus 73% with radiotherapy alone, while three-year overall survival reached 96% in both groups.
Importantly, grade 3–4 toxicity was substantially higher with pembrolizumab (43% versus 15%).
The findings confirm that adjuvant radiotherapy alone, using vaginal brachytherapy and/or pelvic external beam radiotherapy as appropriate, remains the standard approach for this trial-defined population, without routine addition of pembrolizumab.
