Handling Incomplete or Late-Onset Toxicities in Early-Phase Dose-Finding Clinical Trials: Current Practice and Future Prospects
Bibliographic record
Abstract
PURPOSE: The way late-onset toxicities are managed can affect trial outcomes and participant safety. Specifically, participants often might not have completed their entire follow-up period to observe any toxicities before new participants would be recruited. We conducted a methodological review of published early-phase dose-finding clinical trials that used designs accounting for partial and complete toxicity information, aiming to understand (1) how such designs were implemented and reported and (2) if sufficient information was provided to enable the replicability of trial results. METHODS: Until March 26, 2023, we identified 141 trials using the rolling 6 design, the time-to-event continuous reassessment method (TITE-CRM), the TITE-CRM with cycle information, the TITE Bayesian optimal interval design, the TITE cumulative cohort design, and the rapid enrollment design. Clinical settings, design parameters, practical considerations, and dose-limiting toxicity (DLT) information were extracted from these published trials. RESULTS: 6.6%). Most trials implementing the TITE-CRM (91.8%, 56 of 61) failed to describe essential parameters in the protocols or the study result papers. Only five TITE-CRM trials (8.2%, 5 of 61) reported sufficient information to enable replication of the final analysis. CONCLUSION: When compared with trials using the rolling 6 design, those implementing the TITE-CRM design exhibited notable deficiencies in reporting essential details necessary for reproducibility. Inadequate reporting quality of advanced model-based trial designs hinders their credibility. We provide recommendations that can improve transparency, reproducibility, and accurate interpretation of the results for such designs.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.055 | 0.464 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.002 | 0.000 |
| Bibliometrics | 0.000 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.001 | 0.002 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; both teacher heads agree on what is shown here.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".