Background
Phase III RCT (GOG 258/NRG GOG 0258). 813 patients with stage III or IVA endometrial carcinoma of any histology (including carcinosarcoma). Randomized to volume-directed RT (pelvic EBRT 45 Gy ± paraaortic extension per nodal status) + concurrent cisplatin 50 mg/m² (2 cycles during RT) then carboplatin (AUC 5) + paclitaxel (175 mg/m²) × 4 cycles vs carboplatin + paclitaxel × 6 cycles alone (no RT). Designed to determine whether adding RT to carboplatin + paclitaxel improves survival in stage III-IVA endometrial cancer.
Interventions and follow up
Arm A: Volume-directed EBRT 45 Gy (pelvic ± paraaortic) + concurrent cisplatin × 2, then carboplatin + paclitaxel × 4
Arm B: Carboplatin AUC 5 + paclitaxel 175 mg/m² × 6 cycles alone (no RT)
Primary endpoint: Recurrence-free survival (RFS)
mFollow up: 47 month
Arm B: Carboplatin AUC 5 + paclitaxel 175 mg/m² × 6 cycles alone (no RT)
Primary endpoint: Recurrence-free survival (RFS)
mFollow up: 47 month
Results
60-month RFS: 58.7% (RT+CT) vs 58.3% (CT alone), HR 0.90 (95% CI 0.74–1.10), P=.30 — equivalent
60-month OS: 70.3% (RT+CT) vs 73.5% (CT alone) — not significantly different
Vaginal recurrence: 2% (RT+CT) vs 7% (CT alone), P<.001 — RT significantly reduces vaginal recurrence
Pelvic/paraaortic recurrence: 11% (RT+CT) vs 20% (CT alone), P<.001 — RT reduces locoregional recurrence
Distant recurrence: 27% (RT+CT) vs 21% (CT alone), P=.02 — more distant recurrence with RT+CT
60-month OS: 70.3% (RT+CT) vs 73.5% (CT alone) — not significantly different
Vaginal recurrence: 2% (RT+CT) vs 7% (CT alone), P<.001 — RT significantly reduces vaginal recurrence
Pelvic/paraaortic recurrence: 11% (RT+CT) vs 20% (CT alone), P<.001 — RT reduces locoregional recurrence
Distant recurrence: 27% (RT+CT) vs 21% (CT alone), P=.02 — more distant recurrence with RT+CT
Adverse events
Main adverse events: Grade ≥3 hematologic toxicity: 58% (RT+CT) vs 63% (CT alone) — similar. Grade ≥3 GI toxicity: 11% (RT+CT) vs 5% (CT alone) — more with RT. Grade ≥3 neurologic: higher with CT alone (more cycles of paclitaxel). Treatment completion: 74% vs 80%.
Conclusions
Combined RT + chemotherapy did not improve RFS or OS vs chemotherapy alone for stage III-IVA endometrial cancer, while reducing locoregional recurrence with RT at the cost of more distant recurrence with RT+CT (fewer systemic therapy cycles), establishing chemotherapy alone (carboplatin + paclitaxel × 6) as an acceptable treatment for stage III disease.
Key Limitations
Key Limitations: The RT+CT arm had fewer chemotherapy cycles (4 vs 6 with CT alone), potentially impairing systemic disease control — this may explain the higher distant recurrence with RT+CT. The two arms differ in both RT delivery AND chemotherapy intensity, making direct comparison of RT vs no-RT difficult. OS was not significantly different but numerically favored CT alone — raises concern about locoregional benefit being offset by distant failure increase. No molecular biomarker stratification.
Clinical Context
GOG 258 and PORTEC-3 provide complementary guidance: PORTEC-3 showed RT+CT improves FFS (vs RT alone); GOG 258 showed CT alone and RT+CT have similar RFS/OS. Together, they suggest that for stage III endometrial cancer, pelvic RT provides locoregional control but may not improve OS when added to full-course carboplatin + paclitaxel (6 cycles). The optimal integration of RT + CT vs CT alone remains debated, especially for IIIC disease. Adding immunotherapy (pembrolizumab: NRG GY018, RUBY) to chemotherapy is now standard for MMRd/MSI-high stage III-IV endometrial cancer.
References
References: Matei D et al, N Engl J Med 2019 (GOG 258)