A Simulation Analysis to Evaluate the Effect of Prospective Biomarker Testing on Progression-Free Survival (PFS) in DLBCL
Bibliographic record
Abstract
Abstract Introduction. Novel treatment regimens that combine chemotherapy with targeted agents are being developed for DLBCL. Evaluation of these targeted agents in clinical trials may require prospective biomarker testing, such as immunohistochemistry (IHC), to select patients based on relevant molecular features. However, prospective biomarker testing takes days and there is concern that the delay in treatment may prevent enrollment of patients with the most aggressive disease, biasing the trial population. Here, using data from the Phase 3 GOYA trial (NCT01287741) in first-line DLBCL, we evaluate the impact on PFS of a treatment delay that simulates the additional screening time required for prospective IHC testing. Methods. In GOYA, previously untreated DLBCL patients were randomized 1:1 to obinutuzumab or rituximab plus 6 or 8 cycles of CHOP. Median times from clinical diagnosis to initiation of screening (“DTIS interval”), and from initiation of screening to randomization (“screening time”) were determined in patients with IPI 2-5 (n=1124 patients with timeline information available, no prospective testing). A two-step analysis of PFS was conducted by dividing the evaluable patient population into three groups of varying DTIS intervals (Figure 1, Step 1), and then into additional groups of varying screening times (Figure 1, Step 2). Groups for screening times were 1) Results. Median time from diagnosis to randomization was 24 days (range 2 to 1106 days, 95th percentile of 66 days). Patients with times from diagnosis to randomization below 15 days (n=250) had significantly worse outcome than patients that were randomized in 15 or more days (n=413 for 15-28 days and n=466 for >28 days; p 28 days from diagnosis to randomization. In Step 1 of the two-step analysis, DTIS times of 14 days were observed for 190, 228 and 706 patients, respectively. Patients with a DTIS interval of 14 days from diagnosis to screening; no additional difference in 3-year PFS was observed in subgroups beyond 14 days. In Step 2, we asked if a longer screening interval (screening to randomization in Figure 1), simulating additional time for prospective testing, affected patient outcome. Of the 190 patients with the shortest DTIS interval ( 14 days (not shown). Conclusion. In GOYA, short PFS was associated with Download : Download high-res image (175KB) Download : Download full-size image Disclosures Szafer-Glusman: Genentech: Employment. Liu: F. Hoffmann-La Roche: Employment. Peale: Genentech / Roche: Employment, Equity Ownership. Farazi: Genentech: Employment. Ray: Genentech/Roche: Employment. Horn: F. Hoffmann-La Roche Ltd: Employment. Oestergaard: F. Hoffmann-La Roche Ltd: Employment. Kornacker: F. Hoffmann-La Roche: Employment; Roche: Equity Ownership. Sehn: Janssen: Consultancy, Honoraria; Celgene: Consultancy, Honoraria; Abbvie: Consultancy, Honoraria; Amgen: Consultancy, Honoraria; Seattle Genetics: Consultancy, Honoraria; Roche/Genentech: Consultancy, Honoraria. Fingerle-Rowson: F. Hoffmann-La Roche Ltd: Employment, Equity Ownership. Venstrom: Genentech, Inc.: Employment. Byrtek: Genentech: Employment, Equity Ownership. Punnoose: Genentech: Employment.
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 | 0.009 | 0.024 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.002 | 0.002 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.001 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.002 |
| Insufficient payload (model declined to judge) | 0.004 | 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; 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".