An extracted individual patient data meta-analysis of immunotherapy plus chemotherapy as first-line treatment for endometrial cancer with a PD-L1 and PD-1 inhibitors subanalysis

Background: Immunotherapy (IO) plus chemotherapy (CT) followed by IO has emerged as a new standard first-line treatment for primary advanced or recurrent endometrial cancer (EC) with an unprecedent benefit for deficient DNA mismatch repair (dMMR) patients (pts) and a smaller magnitude of benefit for mismatch repair–proficient (pMMR) disease. We conducted an extracted individual patient data (eIPD) meta-analysis to illustrate the survival curves and provide insights into this new standard.

Methods: We searched PubMed, Embase, Cochrane, and oncology meetings up to April 2024 for randomized phase II or III trials (RCTs) investigating IO plus CT as first-line treatment for EC. IPD was reconstructed from reported Kaplan-Meier plots through WebPlotDigitizer and the R package IPDfromKM and then combined. Additionally, we conducted a trial-level meta-analysis using fixed and random effects models.

Results: Five RCTs were included (NRG-GY018, RUBY, MITO END-3, AtTEnd/ENGOT-en7, DUO-E). 2436 patients were analyzed for progression-free survival (PFS) and 2317 patients for overall survival (OS). The eIPD underscored the significant benefit of adding IO in dMMR patients, with 3-year absolute gains of 36% in PFS (HR 0.36, 95% CI 0.28-0.45) and 28% in OS (HR 0.41, 95% CI 0.30-0.48) with survival curves showing early separation from treatment initiation, reaching a plateau around 12 months, suggesting sustained benefit. For pMMR, a smaller benefit was observed in PFS, with a 3-year absolute gain of 6% (HR 0.78, 95% CI 0.69-0.88). No difference was observed in OS (HR 0.87, 95% CI 0.74-1.01), although the data is immature. Notably, many PFS events occurred early in the pMMR curve, indicating numerous early progressors in both arms. A slight separation of curves began around 10 months, suggesting a subset of pMMR pts may benefit from adding IO to CT. Remarkably, there was a statistically significant benefit for PD-1 inhibitors in the pMMR subgroup (OS PD-1 pMMR: HR 0.78, 95% CI 0.62-0.96) in comparison with PD-L1 inhibitors (OS PD-L1 pMMR HR: 0.93, 95% CI 0.75−1.16).

Conclusions: IO should be part of the standard of care in first-line for dMMR EC. For pMMR tumors, only a small subset of pts appears to benefit from adding IO, and longer follow-up is warranted. Molecular classification exploratory analysis may help us to identify who are the pMMR pts that benefit from IO. This analysis suggests different outcomes between PD-1 and PD-L1 inhibition for pMMR EC, favoring PD-1.

Corresponding author: Maísa Maria Spagnol Trento (e-mail: maisaast@gmail.com).