Dragonboat's PD-L1 trial cuts enrollment 49.2% to 66 after status lapsed to unknown
NCT03908814 closed out with less than half its original 130-patient target and no posted results, leaving the dose-finding data from LDP undisclosed three years after primary completion.
Executive Summary
- ClinicalTrials.gov now lists NCT03908814 as Completed with enrollment revised down to 66 from a 130-patient target, a 49.2% cut logged on July 6, 2026, alongside the status flip itself Press ReleasePress ReleaseJul 6, 2026.
- The registry's own actual primary completion date is April 10, 2023, meaning the Completed status and enrollment correction arrived more than three years after the trial's stated data cutoff, a gap that raises questions about disclosure discipline rather than trial design.
- Despite the Completed status, ClinicalTrials.gov shows no posted results for either the primary endpoints, maximum tolerable dose and dose-limiting toxicity, or the secondary endpoints including objective response rate.
- AppliedXL's operational risk model scores this trial 95 out of 100, citing a 500-day cumulative primary completion delay and a population shift from combined eligibility and enrollment amendments, the highest risk tier in the model.
- The PD-L1 checkpoint class is dominated by approved agents, durvalumab and atezolizumab among them, with 238 active trials against the target; Dragonboat's early-stage, single-arm cohort sits far outside that registrational frontier.
The registry update
NCT03908814, Dragonboat's Phase 1 dose-escalation study of the anti-PD-L1 antibody LDP in advanced malignant tumors, was updated on July 6, 2026 to show Completed status and an enrollment count of 66, down from an original target of 130 Press ReleasePress ReleaseJul 6, 2026. The two changes, status and enrollment, were logged the same day, years after the trial's actual work had concluded. The registry's stated primary completion date remains April 10, 2023, and the trial ran from April 2019 through that date, a span of 1,463 days.
What the endpoint measures
The trial's primary endpoints are maximum tolerable dose (MTD) and the number of participants with dose-limiting toxicity (DLT), both assessed at the end of the first 28-day treatment cycle. This is a safety and tolerability bar, not an efficacy readout: objective response rate and progression-free survival are listed only as secondary endpoints, evaluated over up to two years of follow-up. No result has been posted for either the primary or secondary endpoints on ClinicalTrials.gov.
The enrollment shortfall
AppliedXL's operational risk assessment for this trial flags a risk score of 95 out of 100, the high end of its scale, driven by a 500-day cumulative delay to the primary completion date and a 49.2% shortfall between the final enrollment of 66 and the original target of 130. The trial also lapsed into Unknown status between April 2022 and its July 2026 reclassification as Completed, a gap of more than four years during which its recruitment state was not current on the registry Press ReleasePress ReleaseJul 6, 2026. AppliedXL's protocol stability tool separately labels the trial Stable, counting only one recorded change event, a reminder that this proxy metric tracks registry change events rather than substantive protocol amendments.
Competitive and field context
LDP enters a PD-L1 checkpoint-inhibitor field with 238 active trials against the target, led by approved agents durvalumab (AstraZeneca) and atezolizumab (Hoffmann-La Roche), both now in Phase 3 studies spanning lung, breast, and bladder cancers. AppliedXL's landscape data shows PD-L1 trial initiation has slowed, with a ratio of 193 recent trials to 1,239 older ones, a signal the mechanism class is decelerating even as it remains structurally crowded. Dragonboat has no other trials on record in AppliedXL's sponsor database, meaning there is no broader pipeline to read through from this cohort's outcome.
This analysis was produced using AI-assisted reporting systems, AppliedXL data, and official public records. These systems undergo editorial review, quality checks, and regular audits by human experts. Errors may still occur, as with any automated system. Always consult the linked primary sources. Read our AI Editorial Policy.