ESMO Congress 2026

In the adjuvant treatment of endometrial cancer, the established evidence for immune checkpoint inhibitors is concentrated in advanced disease: current guidelines incorporate chemo-immunotherapy for FIGO stage III-IV dMMR patients, but the evidence supporting omission of adjuvant radiotherapy remains insufficient ^[1]. Neither of the two pivotal trials could answer the radiotherapy question—PORTEC-3 included radiotherapy in both arms without immunotherapy, and radiotherapy was not randomized in KEYNOTE-B21, so neither directly tested the safety of omitting radiotherapy ^[1]. In terms of molecular classification, the MMRd group has an intermediate prognosis, adjuvant decisions still rely closely on clinicopathological factors, and the classification has been integrated into ESGO management guidelines ^[6]. There may be synergy between radiotherapy and immune checkpoint blockade, but the optimal sequencing and combination strategy remains unclear in the available evidence ^[5].

NRG-GY020 reported at ESMO26 placed pembrolizumab on the radiotherapy platform in an early-stage dMMR population: patients with stage I/II high-intermediate risk dMMR endometrioid endometrial cancer (n=168, 1:2) were randomized to radiotherapy with or without 1 year of pembrolizumab. The primary endpoint, 3-year recurrence-free status, was negative, with slightly differing numbers across the two posts: one reported 70% for pembrolizumab+RT vs 73% for RT alone (leaving the study counted as failure), the other reported 73% vs 77%, p=0.31. The KM recurrence-free survival curves numerically favored the pembrolizumab arm, but the CI crossed 1; G3/4 toxicity was 43% vs 15%. The posting radiation oncologist stated that RT alone remains the default in this setting.

Relative to existing knowledge, GY020 shifts the question from whether radiotherapy can be omitted in the adjuvant treatment of advanced dMMR disease ^[1] to whether immunotherapy should be added on top of radiotherapy in an earlier, high-intermediate risk population, and returned a negative result. Unlike PORTEC-3 and KEYNOTE-B21, this trial retained radiotherapy in both arms and randomized the addition of immunotherapy ^[1], thus directly answering the question of addition. The posts also raised a second thread of discussion: one commentator questioned whether this is truly negative, or a consequence of an overly aggressive statistical design intended to conserve sample size—meaning the negativity may stem from the design rather than biology.

The limitations are equally clear: the sample size is small (n=168), the primary endpoint counted leaving the study as failure, and the specific numbers differ between the two posts, pending verification against the formal abstract; with the CI crossing 1, the numerical favoring should not be over-interpreted, whereas the 43% vs 15% difference in G3/4 toxicity is a relatively certain signal. This is an interpretation of conference information and does not constitute investment advice.