Fragility index of trials supporting approval of anti-cancer drugs in common solid tumors.
Bibliographic record
Abstract
2055 Background: The Fragility Index (FI) quantifies the reliability of positive trials by estimating the number of events which would change statistically significant results to non-significant results. Here, we calculate the FI of trials supporting approval of drugs for common solid tumors. Methods: We searched Drugs@FDA to identify randomized trials (RCT) supporting drug approvals by the US Food and Drug Administration between January 2009 and December 2019 in lung, breast, prostate, gastric and colon cancers. We adapted the FI framework (Walsh et al. J Clin Epidemiol 2014) to allow use of time to event data. First, we reconstructed survival tables from reported data using the Parmar Toolkit (Parmar et al. Stat Med 1998) and then calculated the number of events which would result in a non-significant effect for the primary endpoint of each trial. The FI was then compared quantitatively to the number of patients in each trial who withdrew consent or were lost to follow-up. Multivariable linear regression was used to explore association between RCT characteristics and the FI. Results: We identified 69 RCT with a median of 669 patients (range 123-4804) and 358 primary outcome events (range 56-884). The median FI was 26 (range 1-322). The FI was ≤10 in 21 trials (30%) and ≤20 in 31 trials (45%). Among the 69 RCT, the median number of patients who withdrew consent or were lost to follow up was 27 (range, 6-317). The number of patients who withdrew consent or were lost to follow-up was equal or greater than the FI in 42 trials (61%). There was statistically significant inverse association between FI and trial hazard ratio (p0,001) and a positive association with number of patients who were lost to follow-up or withdrew consent (p0,001). There was no association between trial sample size, year of approval or reported p-value and the FI. Conclusions: Statistical significance of trials supporting drug approval in common solid tumors relies often on a small number of events. In most trials the FI was lower than the number of patients lost to follow up or withdrawing consent. Post-approval randomized trials or real-world data analyses should be performed to ensure that effects observed in registration trials are robust.
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 machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.181 | 0.486 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.003 | 0.010 |
| Bibliometrics | 0.023 | 0.011 |
| Science and technology studies | 0.001 | 0.002 |
| Scholarly communication | 0.003 | 0.004 |
| Open science | 0.002 | 0.004 |
| Research integrity | 0.002 | 0.003 |
| Insufficient payload (model declined to judge) | 0.009 | 0.001 |
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; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
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".