Abstract 1644: Progression-free survival (PFS) as a surrogate for overall survival (OS)
Bibliographic record
Abstract
Abstract Background: PFS is criticized as being not reliably predictive of OS, and statistically significant gains may be clinically insignificant. However, PFS has the advantage as an endpoint of needing shorter follow up/fewer patients and is not rendered falsely negative by crossover (CO) and long post progression survival (PPS). Absolute PFS gain may be a surrogate for absolute OS gain (Ann Oncol 2016;27:373). Method: We extracted PFS/OS data from randomized drug comparisons with >200 incurable solid tumor patients published 01/2007-06/12/2017 in J Clinical Oncology and New England J Medicine. For comparisons with CO unspecified or in <20% of patients, and p<0.05 for PFS or OS, we used nonlinear regression analysis (NLRA) of PFS/OS curves (omitting curve sections with <~10 residual patients) to determine PFS/OS half-lives (t1/2s: time to progression/death in half the remaining patients). We calculated ΔmPFS and ΔmOS (experimental minus control arm medians) and ΔPFSt1/2 and ΔOSt1/2 (experimental minus control arm t1/2s). Results: 249 trials and 287 across-arm comparisons were evaluable. The table presents ΔPFS/ΔOS associations for studies with PFS p<0.05 and CO<20% or unspecified. PFS changenNo. (%) of comparisons with associated change in OS (ΔmOS for ΔmPFS or ΔOSt1/2 for ΔPFSt1/2)ΔOS <2 monthsΔOS > 2 monthsΔmPFS: <1.5 months3224 (75%)8 (25%)>1.5 months4722 (47%)25 (53%)ΔPFSt1/2: <1.5 months3531 (89%)4 (11%)>1.5 months6210 (16%)52 (84%)>2 months414 (10%)37 (90%)ΔOS <3 monthsΔOS >3 monthsΔPFSt1/2: >3 months263 (12%)23 (88%)ΔOS <4 monthsΔOS >4 monthsΔPFSt1/2: >4 months131 (8%)12 (92%)Concordance of ΔOS with ΔPFS for ΔPFS < vs >1.5 months better for t1/2s than medians, Fisher's exact test p=0.0004, and correlation of ΔOS with ΔPFS was also better for t1/2s (R2=0.60, p<0.0001, slope = 1.51) than for medians (R2=0.37, p<0.0001, slope = 0.88) To assess the risk of ΔPFS under-predicting ΔOS, we then assessed all comparisons, irrespective of CO and PFS significance. Of 287 comparisons, 13 (5%) had ΔOS >2m with p<0.05 but ΔPFS <1.5m or ΔPFS p>0.05. In 7/13, PFS curves fit NLRA 2-phase decay models (suggesting 2 distinct subpopulations), vs 25/100 with PFS 2-phase decay in other comparisons (Fisher's exact test p=0.047). Conclusions: Demonstrating ΔPFSt1/2 >1.5-2.0 months with p<.05 indicates an 84-90% probability of a ΔOSt1/2 gain >2 months, and could be a useful ΔOSt1/2 surrogate. This endpoint could mitigate impact of CO and PPS while making trials faster/cheaper. T1/2s are superior to medians. If PFS curves fit NLRA 2 phase models in phase II trials one should collect full phase III OS data (censored at CO) to reduce risk of a false negative. Citation Format: David J. Stewart, Dominick Bosse, John Hilton, Glenwood Goss, Michael Fung-Kee-Fung, Derek Jonker. Progression-free survival (PFS) as a surrogate for overall survival (OS) [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2018; 2018 Apr 14-18; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2018;78(13 Suppl):Abstract nr 1644.
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 | 0.026 | 0.047 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.004 |
| Bibliometrics | 0.003 | 0.005 |
| Science and technology studies | 0.000 | 0.001 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.000 | 0.001 |
| Insufficient payload (model declined to judge) | 0.007 | 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".