Are surrogate endpoints unbiased metrics compared to hazard ratio for death? An evaluation of clinical benefit scores (CBS) in the American Society of Clinical Oncology (ASCO) value framework.
Bibliographic record
Abstract
6600 Background: Clinical benefit scores (CBS) are a key element of the American Society of Clinical Oncology (ASCO) value framework's Net Health Benefit valuation of cancer therapies. CBS are assigned based on a hierarchy of efficacy endpoints, from hazard ratio for death (HR OS), to median overall survival (mOS), HR for disease progression (HR PFS), median progression-free survival (mPFS), and response rate (RR). When HR OS is unavailable, other endpoints in the hierarchy are used as "surrogates" to calculate CBS via their scaling factors. We aim to examine whether surrogate-derived CBS offer unbiased scoring of clinical benefit compared to HR OS-derived CBS. Methods: CBS for advanced-disease settings were computed for randomized clinical trials (RCTs) of oncology drug approvals by the Food and Drug Administration, European Medicines Agency, and Health Canada, between 2006 and August 2015. Spearman's correlation assessed association between CBS derived from surrogates and HR OS. Mean bias (surrogate-derived CBS minus HR OS-derived CBS) evaluated the tendency for surrogate-derived CBS to over- or under- estimate clinical benefit. Mean absolute error (MAE), a measure of average deviation, assessed precision of surrogate-derived CBS in relation to HR OS-derived CBS. Results: Scored RCTs (n=104) yielded 69, 93, 88, and 89 paired CBS between HR OS and mOS, HR PFS, mPFS, and RR, respectively. See table for results. Restricting to RCTs reporting all endpoints (n=59) and RCTs without OS as primary endpoint (n=68) showed similar results. Conclusions: Findings suggest HR PFS-, mPFS-, and RR-derived CBS are poor "surrogates" as they are imprecise and weakly correlated to HR OS-derived CBS. HR PFS and particularly mPFS exhibit bias to overestimate CBS. [Table: see text]
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 | 0.248 | 0.473 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.002 | 0.003 |
| Bibliometrics | 0.006 | 0.006 |
| Science and technology studies | 0.001 | 0.003 |
| Scholarly communication | 0.003 | 0.003 |
| Open science | 0.001 | 0.004 |
| Research integrity | 0.001 | 0.002 |
| Insufficient payload (model declined to judge) | 0.001 | 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; the direct Gemma label and the distilled Codex classifier 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".