Patterns of endpoint use in oncology pragmatic clinical trials: A review of randomised studies

Pragmatic clinical trials (PCTs) aim to address the effectiveness of interventions when applied under conditions which reflect real-world practice. In oncology, PCTs are increasingly used to address research questions regarding the optimal application of therapeutic interventions. This study aimed to assess patterns of endpoint selection in randomised pragmatic-labelled oncology trials.

We searched PubMed® and ClinicalTrials.gov for randomised oncology trials using pragmatic terms in publications, registry records, or indexing (2014-2025). Source documents for included trials were retrieved and used for data extraction and descriptive analysis.

Our search returned 1342 records, from which 53 unique PCTs were identified following screening. All trials were sponsored by non-commercial entities and nearly half were initiated between 2021 and 2025. Most were phase III, multicentric, superiority trials, while 22/53 trials (41.5%) evaluated a de-escalation strategy. Surrogate primary endpoint use was the most common (43.4% of trials), followed by overall survival (OS) and quality of life (QOL) (34.0% and 17.0% of trials, respectively). Though most trials with surrogate primary outcomes included OS and QOL as secondary endpoints, these were individually specified as key secondary endpoints in only 4.3% of cases.

Surrogate primary endpoints remain more common than OS or QOL in randomised pragmatic-labelled oncology trials. Future pragmatic trials should prioritise patient-centred endpoints where feasible and provide transparent justification for surrogate endpoint selection.