Immune-Related Adverse Events Among Patients With Endometrial Cancer Receiving Pembrolizumab Alone or With Other Treatments: A Single-Center Real-World Experience.

Immunotherapy has revolutionized endometrial cancer (EC) treatment; our objectives were to characterize immune-related adverse events (irAE) among EC patients and to identify factors associated with risk for irAE. Patients who received pembrolizumab alone or as part of combination therapy for advanced/recurrent EC from 2014 to 2025 were identified. Baseline demographics and cancer characteristics were collected. Potential immune-related adverse events were collected and graded for the cohort. Multiple logistic regression was then performed. A total of 180 patients were included. The mean age was 63.9 years. Black patients comprised 30.6% of the cohort. Endometrioid was the commonest histologic subtype (38.7%), and most tumors (67.7%) were MMR proficient. Ninety-five patients (52.8%) had any-grade irAE, and 14 (7.8%) experienced serious (grades 3-4) irAE. Hypothyroidism was the most common irAE overall (n = 37; 20.6%,) followed by fatigue (n = 23; 12.8%) and diarrhea (n = 20; 11.1%). The most common serious irAEs were diarrhea (n = 4; 2.2%,) followed by hepatitis, dermatitis, and pneumonitis (each n = 3; 1.7%). On univariate analysis, older age, uninsured status, and receipt of pembrolizumab with lenvatinib were associated with any-grade irAE (all P <0.05). In our multiple logistic regression model, uninsured status was an independent predictor of both any-grade irAE (aOR: 5.49; 95% CI: 1.03-29.2) and serious irAE (aOR: 7.96; 95% CI: 1.44-43.9). Receiving pembrolizumab with lenvatinib was independently associated with any-grade irAE (aOR: 7.13; 95% CI: 2.92-17.4). Uninsured status and coadministration with lenvatinib may increase the risk for pembrolizumab-associated irAE among patients with EC. Further collection of real-world data, including unexplored social and economic factors and novel biomarkers, may predict irAE and ultimately help guide clinical decision-making.