Dostarlimab Improves Advanced Endometrial Cancer Survival Dostarlimab plus chemotherapy produced durable survival gains in biomarker-selected advanced endometrial cancer.
*Phase 3 randomized controlled trial subgroup analysis; Level 2.

Citation

Powell MA, Roed H, Willmott LJ, et al. Four-year survival outcomes with dostarlimab plus chemotherapy in mismatch repair deficient/microsatellite instability-high primary advanced or recurrent endometrial cancer in the RUBY trial. Gynecologic Oncology. 2026;212:207-217. doi:10.1016/j.ygyno.2026.05.008

Background

Chemotherapy alone was long the usual first treatment for advanced or recurrent endometrial cancer, but survival was poor. This update tests whether adding dostarlimab provides durable benefit in a biomarker-defined group.

Patients

118 adults with primary stage III/IV or recurrent dMMR/MSI-H endometrial cancer with low chance of cure by surgery or radiation. Patients needed adequate tumor tissue for biomarker testing.

Intervention

Dostarlimab plus carboplatin and paclitaxel for 6 cycles, then dostarlimab maintenance for up to 3 years.

Control

Placebo plus the same chemotherapy, then placebo maintenance.

Outcome

Progression-free survival, overall survival, response duration, and safety.

Follow-up Period

Median 55.6 months.

Results

Outcome Dostarlimab benefit 4-year rates NNT
Progression-free survival HR 0.30 (95% CI 0.17 to 0.52) 57.9% vs 15.7% 3
Overall survival HR 0.34 (95% CI 0.19 to 0.63) 72.8% vs 40.3% 4
dMMR/MSI-H: mismatch repair deficient or microsatellite instability high; HR: hazard ratio; NNT: number needed to treat.
Median progression-free and overall survival were not reached with dostarlimab. A model estimated a 54% cure-potential rate with dostarlimab versus 14% with placebo. No new safety signals emerged; serious treatment-related events were more common with dostarlimab, and 2 treatment-related deaths occurred. Efficacy was analyzed by randomized assignment within this biomarker subgroup. Control was chemotherapy alone, the prior standard.

Limitations

Small subgroup size limits precision. Overall survival in this biomarker group was exploratory. Conditional survival and cure estimates were post hoc models, not proof of cure. GSK sponsorship and employee authorship may introduce bias. Results apply to advanced or recurrent biomarker-selected disease, not lower-risk endometrial cancer.

Funding

Funded by GSK; employees coauthored, creating industry-bias concern.

Clinical Application

Use dostarlimab with chemotherapy for eligible mismatch-repair-deficient advanced or recurrent disease; do not generalize to lower-risk endometrial cancer.