Temporal patterns and factors associated with early termination in phase 1 clinical trials for solid tumors

Early termination, defined as the discontinuation of a trial before its planned completion, remains poorly characterized in phase 1 cancer trials. We aimed to (1) identify reasons for early termination in phase 1 trials and compare them with those in late-phase studies and (2) investigate temporal patterns and factors associated with early termination in phase 1 trials.

Trials on solid tumors initiated between 2018 and 2023 were identified using the Aggregate Analysis of ClinicalTrials.gov database. We compared the reasons for early termination between phase 1 and phase 2 to 3 trials using the χ2 test. We estimated the cumulative probability of early termination over time and used a multivariable logistic regression model to investigate determinants of early termination.

We included 2301 studies: 1740 phase 1 and 561 phase 2 to 3 trials. "Sponsor/strategic decisions" was the most common reason for early termination in phase 1 trials (51.1%), while "poor accrual" predominated in phase 2 to 3 trials (35.5%; P < .001). The probability of early termination in phase 1 trials increased rapidly between 10 and 30 months after initiation and plateaued after 60 months. Factors associated with a lower probability of early termination included academic funding and conduct in Europe, Asia, or Oceania. Conversely, trials conducted globally or enrolling patients with thoracic or gastrointestinal cancers were more likely to be discontinued prematurely.

Phase 1 trials are primarily discontinued for "sponsor/strategic decisions." Funding source, tumor type, and geographic location of recruiting centers are independently associated with early termination. The early peak in early termination probability marks a critical window for interventions to improve trial sustainability.