Trial design and temporal trends of radiotherapy combined with immune checkpoint inhibitors in head and neck cancer

Abstract Radiotherapy combined with immune checkpoint inhibitors (RT + IO) is a rapidly expanding treatment strategy in head and neck cancer, but the design of the registered trial portfolio has not been systematically quantified. Using the Aggregate Analysis of ClinicalTrials.gov (AACT) database (snapshot 3 September 2026), we identified all head and neck cancer trials and removed label-based (MeSH) over-annotation to define a site-focused universe of 5,709 trials. Trials were classified as RT + IO, RT-only, IO-only or neither using explicit structured dictionaries, and co-administration was assessed at arm level. Classification was validated by two independent reviewers (Cohen's kappa = 0.918). Of the 5,709 trials, 237 (4.2%) were RT + IO and 233 were interventional. The proportion of RT trials that were RT + IO rose over time (Cochran-Armitage z = 3.39, P = 0.0007), and RT + IO trials were more frequent among trials from China (adjusted OR 2.75, 95% CI 1.85–4.08) and industry-sponsored trials (OR 5.66, 3.83–8.39). Same-arm co-administration was present in at least 121 of 237 trials (51.1%), and results had been posted for 38 of 233 (16.3%). The registered RT + IO portfolio in head and neck cancer is substantial, geographically concentrated, and dominated by design features that may limit direct attribution of combined-modality effects.