Written by Dr. Roupen Odabashian MD, FRCPC
Oncologist | Founder, MeDucation AI | Updated October 2026
The short answer: RUBY is the trial that moved dostarlimab plus carboplatin and paclitaxel into first-line treatment of primary advanced or recurrent endometrial cancer, and the size of the benefit depends almost entirely on mismatch repair status. In dMMR/MSI-H disease, median PFS was 30.3 vs 7.7 months (HR 0.29)[1]. In the all-comers population, median overall survival was 44.6 vs 28.2 months (HR 0.69)[6]. In MMRp/MSS disease the overall survival signal was exploratory, with an HR of 0.79 whose confidence interval (0.602 to 1.044) includes 1[6].
Disclosure: I am the founder of MeDucation AI, which sells hematology and oncology board preparation. This is a trial explainer and not a product comparison, and every number below links to the source it came from so you can check it yourself.
Ranked Summary: What RUBY Actually Showed, By Population
Population | Headline Finding | Why It Matters for Practice |
|---|---|---|
dMMR/MSI-H | PFS HR 0.29, median PFS 30.3 vs 7.7 months[1]; exploratory OS HR 0.32[6] | Largest, most consistent benefit; near-universal use of dostarlimab-chemo here |
All comers | OS HR 0.69, median OS 44.6 vs 28.2 months[6] | Basis for the August 2024 all-comers FDA approval[3] |
MMRp/MSS | OS HR 0.79 (95% CI 0.602 to 1.044), median OS 34.0 vs 27.0 months, exploratory[6] | Smaller, exploratory benefit; still the majority of patients (70-75%)[3] |
RUBY Part 2 (niraparib added) | PFS HR 0.60 overall, 0.63 in MMRp/MSS[7] | A separate question from Part 1, with more toxicity |
What Did the RUBY Trial Actually Test?
RUBY tested whether adding the PD-1 inhibitor dostarlimab to standard carboplatin-paclitaxel, followed by dostarlimab maintenance, improves outcomes in primary advanced or recurrent endometrial cancer compared to chemotherapy plus placebo[4]. It is a phase 3, randomized, double-blind, multicenter, placebo-controlled trial[4], run as a two-part study: Part 1 compared dostarlimab-chemo to placebo-chemo, and Part 2 added niraparib maintenance on top of dostarlimab[3]. Patients were randomized 1:1 and stratified by MMR/MSI status, prior pelvic radiotherapy, and disease status (recurrent, primary stage III, or primary stage IV)[1]. A prespecified subgroup of patients with dMMR/MSI-H tumors was analyzed, and this subgroup turned out to define the entire narrative of the trial[1].
The primary efficacy endpoint was investigator-assessed progression-free survival by RECIST v1.1[1], with overall survival as a dual primary endpoint in Part 1[3]. The trial is part of an international collaboration between ENGOT and the GOG Foundation[3], and the dMMR/MSI-H efficacy analysis was conducted in a prespecified subgroup of 122 patients[1], with MMR/MSI status determined by local IHC, PCR, or NGS testing, or central IHC testing using the Ventana MMR RxDx Panel when local results were unavailable[1].
Why Does Mismatch Repair Status Matter So Much Here?
Mismatch repair status marks tumors with a fundamentally different immune biology, and RUBY's results track that biology closely. dMMR/MSI-H tumors generate high numbers of neoantigens, which is the mechanistic reason checkpoint inhibition performs so much better in that subgroup[5]. Dr. Mansoor Raza Mirza, the study chair, described the dMMR/MSI-H population as having "hot" tumors and the MMRp/MSS population as having "cold" tumors, noting that "in the patients with 'hot' tumors who had the additional immunotherapy, the 72% benefit in progression-free survival was much larger than the 50% we expected to see, and in those with 'cold' tumors, we saw a moderate effect"[5]. That single distinction is the reason any exam question on this trial is likely to ask you to separate the dMMR/MSI-H result from the overall result rather than quoting one number for the whole trial.
What Were the PFS and OS Results, By Subgroup?
The results split cleanly by mismatch repair status, and every number below comes directly from the published trial reports and company disclosures.
Endpoint | Population | Dostarlimab-Chemo | Placebo-Chemo | Hazard Ratio |
|---|---|---|---|---|
Median PFS | dMMR/MSI-H | Not reached / 30.3 months (per label)[1][5] | 7.7 months[1] | 0.29 (0.17-0.50)[1] |
PFS at 24 months (KM estimate) | dMMR/MSI-H | 61.4%[4] | 15.7%[4] | 0.28 (0.162-0.495)[4] |
PFS at 24 months (KM estimate) | Overall | 36.1%[4][5] | 18.1%[4][5] | 0.64 (0.507-0.800)[4] |
Median OS | Overall | 44.6 months (32.6-NR)[3][2] | 28.2 months (22.1-35.6)[3][2] | 0.69 (0.539-0.890)[2] |
OS at 2.5-year landmark | Overall | 61% alive[3] | 49% alive[3] | n/a |
At the initial progression-free survival readout reported in The New England Journal of Medicine, median PFS was 11.8 months with dostarlimab and chemotherapy versus 7.9 months with chemotherapy and placebo in the overall trial, with a doubling of the 24-month PFS rate[5]. The dMMR/MSI-H effect was so large that median PFS "could not be estimated" in the dostarlimab arm at that data cut because too few patients had progressed[5]. The later overall survival update, presented with more than three years of median follow-up[3], is what ultimately supported the all-comers FDA approval.
What About RUBY Part 2 and Niraparib?
RUBY Part 2 asked a different question: does adding niraparib maintenance to dostarlimab-chemotherapy improve outcomes further, specifically in the MMRp/MSS population where PARP inhibition might add benefit on top of checkpoint blockade. The combination of dostarlimab plus chemotherapy followed by dostarlimab plus niraparib met its primary endpoint, showing a 40% reduction in risk of disease progression or death (HR 0.60) and a 6.2-month improvement in median PFS (14.5 vs 8.3 months) in the overall population, with a similar 37% reduction (HR 0.63) and 6.0-month PFS improvement in the MMRp/MSS subgroup[2]. This is a distinct regulatory and clinical question from the Part 1 approval, and it came with meaningfully more toxicity, discussed below.
How Does RUBY Compare to NRG-GY018, DUO-E, and AtTEnd?
All four trials added a PD-1 or PD-L1 inhibitor to carboplatin and paclitaxel, and in each one the dMMR hazard ratio was far lower than the pMMR or all-comers hazard ratio. They were separate trials, though, so the numbers below should not be read as a head-to-head comparison[11]. The NCI notes that a direct comparison of the approved drugs is unlikely, and that maintenance duration differed: up to 2 years in NRG-GY018, up to 3 years in RUBY, and for as long as the drug controlled the cancer in DUO-E[11].
Trial (immunotherapy) | Overall PFS HR | dMMR PFS HR | MMRp/pMMR PFS HR |
|---|---|---|---|
RUBY Part 1 (dostarlimab) | 0.64 (0.51 to 0.80)[4] | 0.28 (0.162 to 0.495)[4] | 0.76 (0.59 to 0.98), exploratory[12] |
NRG-GY018 (pembrolizumab) | Not reported in the sources reviewed | 0.30 (0.19 to 0.48)[8] | 0.60 (0.46 to 0.78)[8] |
DUO-E, durvalumab arm | 0.71[9] | 0.42 (0.22 to 0.80)[9] | 0.77 (0.60 to 0.97)[9] |
DUO-E, durvalumab plus olaparib arm | 0.55[9] | 0.41 (0.21 to 0.75)[9] | 0.57 (0.44 to 0.73)[9] |
AtTEnd (atezolizumab) | 0.74 (0.61 to 0.91)[10] | 0.36 (0.23 to 0.57)[10] | Not reported in the source reviewed |
Which Trial Has Overall Survival Data?
RUBY is the one with a statistically supported all-comers overall survival result: median OS 44.6 vs 28.2 months, HR 0.69 (95% CI 0.54 to 0.89)[6]. GSK describes it as the first immuno-oncology-based treatment to show an overall survival benefit in this setting[3]. AtTEnd's interim all-comers OS hazard ratio was 0.82 (95% CI 0.63 to 1.07), and the interval crosses 1[10]. I did not find overall survival results for NRG-GY018 or DUO-E in the sources I reviewed for this post, so I am not comparing them here.
Which Drug Is Approved for Which Patients?
Drug (with carboplatin and paclitaxel) | FDA approval | Population |
|---|---|---|
Dostarlimab | July 31, 2023[1]; expanded August 1, 2024[11] | dMMR/MSI-H first, then all adult patients with primary advanced or recurrent disease[3][11] |
Durvalumab | June 14, 2024[11] | dMMR only[11] |
Pembrolizumab | June 17, 2024[8] | Primary advanced or recurrent endometrial carcinoma, regardless of MMR status[8][11] |
How Much More Toxicity Does Dostarlimab Add, and What Does Niraparib Add on Top?
Dostarlimab roughly doubled treatment discontinuation for adverse events in the initial RUBY report, and adding niraparib in Part 2 pushed discontinuation and immune related events considerably higher. The table collects the absolute numbers that the sources I reviewed actually state.
Safety measure | Dostarlimab arm | Control arm |
|---|---|---|
Part 1: discontinuation due to a treatment-emergent adverse event | 17.4%[5] | 9.3%[5] |
Part 2: dostarlimab or placebo discontinued for an adverse event | 24.1%[7] | 5.2%[7] |
Part 2: niraparib or placebo discontinued for an adverse event | 15.7%[7] | 4.2%[7] |
Part 2: immune related adverse events | 36.6%[7] | 6.3%[7] |
In Part 2, grade 3 or higher treatment-emergent events were about 36% higher and serious events about 24% higher in the dostarlimab plus niraparib arm, reported as relative increases rather than absolute rates[7]. No MDS or AML cases were reported[7]. The FDA lists rash, diarrhea, hypothyroidism and hypertension as the most common adverse reactions with dostarlimab plus chemotherapy, along with immune mediated events such as pneumonitis, colitis, hepatitis, endocrinopathies, nephritis and skin reactions[1].
The other trials show the same pattern. In DUO-E, grade 3 or higher events during maintenance were 41.1% with durvalumab plus olaparib vs 16.6% with chemotherapy alone[9], and in AtTEnd grade 3 or higher events were 25.8% with atezolizumab vs 14.1% with placebo[10]. Across the three trials the NCI summarizes, more than half of immunotherapy participants had at least one serious side effect and 13% to 20% stopped treatment because of side effects[11]. Definitions and reporting windows differ between trials, so treat these as context and not as a ranking.
What Should You Remember About RUBY for the Boards?
These are my own study priorities, not an official blueprint. If you can do these five things, you can handle most questions built on this trial:
Name the population before you quote a number. dMMR/MSI-H: PFS HR 0.29[1]. All comers: OS HR 0.69[6]. MMRp/MSS: OS HR 0.79 with a confidence interval of 0.602 to 1.044, exploratory[6].
Know that approval came in two steps: dMMR/MSI-H on July 31, 2023[1], then all adult patients on August 1, 2024[11].
Know how MMR status was determined: local IHC, PCR or NGS, or central IHC with the Ventana MMR RxDx Panel when local results were unavailable[1].
Keep Part 2 separate. It tested adding niraparib maintenance, with PFS HR 0.60 overall and 0.63 in MMRp/MSS[7], and with more toxicity.
Do not carry RUBY's hazard ratios onto NRG-GY018, DUO-E or AtTEnd. Each trial has its own control arm, maintenance duration and endpoints[11].
Frequently Asked Questions About the RUBY Trial
What is the RUBY trial?
RUBY is a phase 3, randomized, double-blind, placebo-controlled trial of dostarlimab plus carboplatin and paclitaxel, followed by dostarlimab maintenance, in primary advanced or recurrent endometrial cancer[1].
Does dostarlimab improve overall survival in MMRp/MSS endometrial cancer?
The signal is favorable but exploratory. Median OS was 34.0 vs 27.0 months with an HR of 0.79 (95% CI 0.602 to 1.044), and the interval includes 1[6].
When did the FDA approve dostarlimab for endometrial cancer?
The FDA approved it for dMMR/MSI-H disease on July 31, 2023[1], and the approval expanded to all adult patients with primary advanced or recurrent disease on August 1, 2024[11].
How does RUBY compare with NRG-GY018 and DUO-E?
All three show a much larger PFS benefit in dMMR than in pMMR disease (dMMR hazard ratios of 0.28, 0.30 and 0.42 for the dostarlimab, pembrolizumab and durvalumab arms)[4][8][9]. They are separate trials with different maintenance durations, so they cannot be ranked against each other[11].
What did RUBY Part 2 test?
Part 2 tested whether adding niraparib maintenance to dostarlimab plus chemotherapy improved outcomes further. PFS HR was 0.60 overall (median 14.5 vs 8.3 months) and 0.63 in MMRp/MSS (median 14.3 vs 8.3 months), with higher discontinuation and immune related event rates[7].
How Do You Keep Trials Like RUBY Straight?
RUBY is easy to misquote because it has two parts, three populations and two approval dates. If you are building a system for retaining trial data, my posts on how to keep up with oncology on a real schedule and on whether spaced repetition is worth building into your exam prep cover how I think about it, and 5 free hematology board review questions shows the format of a MeDucation question. MeDucation's question bank is written and reviewed by practicing, board certified hematologist-oncologists and is not AI generated. This article is educational and is not treatment advice.
References
FDA approves dostarlimab-gxly with chemotherapy for endometrial cancer | FDA
RUBY trial results: The new standard of care in advanced/recurrent endometrial cancer | ecancer
FDA Expands Dostarlimab Approval for All Advanced Endometrial Cancers | Targeted Oncology
Dostarlimab-Based Combinations in Advanced Endometrial Cancer | The ASCO Post
Endometrial cancer: durvalumab and pembrolizumab with chemotherapy | NCI Cancer Currents
RUBY Trial Supports Addition of Dostarlimab to Chemo as SOC in Endometrial Cancer | CancerNetwork
