S901 Group-Based Trajectory Modeling to Identify Response Patterns to Ozanimod Treatment in Patients With Ulcerative Colitis
Bibliographic record
Abstract
Introduction: Ozanimod (OZA) is approved for the treatment (tx) of moderately to severely active ulcerative colitis (UC) in adults. In the phase 3 True North (TN; NCT02435992) study, patient (pt) responses to OZA were heterogenous at Week (W) 52. Methods: This TN post hoc analysis used grouP-based trajectory modeling (GBTM) to identify patterns of OZA response trajectories during the TN maintenance period (MP) in W10 OZA clinical responders who were rerandomized to continue OZA (OZA/OZA arm; n=230) in MP. Change from baseline (BL; TN W0) in partial Mayo score (PMS) was used as a variable for identifying pt clusters with various OZA response patterns. GBTM, determined by maximum likelihood estimation, was used to estimate OZA response trajectories. After modeling, clinical and mucosal outcomes at W52 were compared in the GBTM-identified pt clusters. Results: Of 230 pts in the MP OZA/OZA arm, 1 did not have post-BL PMS data available and was excluded from this analysis; 5 groups of pt-response patterns to OZA therapy were identified during MP. At W10, a PMS response was observed in all 5 trajectory groups (TGs). During the 42-wk MP, TGs diverged into fast sustained improvement (Group [G] 1; n=38 [16.6%, 38/229]); slow sustained improvement (G2; n=85 [37.1%, 85/229]); gradual improvement (G3; n=60 [26.2%, 60/229]); fast rebound (G4; n=25 [10.9%, 25/229]); and slow rebound (G5; n=21 [9.2%, 21/229]). All pts entering the MP were OZA clinical responders, but 20% of pts did not demonstrate sustained improvement from BL in PMS (G4 and G5); 80% demonstrated sustained improvement from BL in PMS (G1–3). From W10 onward, 26.2%, 37.1%, and 16.6% of pts achieved and maintained PMS reductions of ≥3.98, ≥5.27, and ≥6.71 points, respectively. Pts in G4 and G5 were slightly younger at diagnosis and had a higher W10 Geboes histologic index score vs the other TGs; pts in G4 had the most corticosteroid use at screening (Table 1). At W52, more pts in G1–3 combined achieved clinical response (74%) and clinical remission (46%) vs those in G4 and G5 combined (7% and 2%, respectively). G1–3 combined had more pts with endoscopic improvement (56%) and mucosal healing (37%) at W52 vs G4 and G5 combined (7% and 2%, respectively). Conclusion: Five groups of pt-response trajectories to OZA tx were identified; 80% of pts in G1–3 had sustained response to OZA tx. GBTM identified clinically meaningful pt subgroups by pt-response patterns to OZA tx. This approach needs to be validated in prospective studies. Table 1. - Pt characteristics at BL (TN W0) or W10 G1: Fast sustained improvement (n=38; 6.6% [38/229]) G2: Slow sustained improvement (n=85; 37.1% [85/229]) G3: Gradual improvement (n=60; 26.2% [60/229]) G4: Fast rebound (n=25; 10.9% [25/229]) G5: Slow rebound (n=21; 9.2% [21/229]) Sex, male, n (%) 15 (39.5) 43 (50.6) 33 (55.0) 13 (52.0) 12 (57.1) Age at UC diagnosis, y, mean (SD) 34.6 (12.4) 35.7 (14.1) 35.3 (13.0) 29.8 (11.5) 31.5 (11.0) Corticosteroid use at screening, n (%) 11 (28.9) 28 (32.9) 14 (23.3) 11 (44.0) 7 (33.3) Prior use of tumor necrosis factor antagonist, n (%) 9 (23.7) 28 (32.9) 22 (36.7) 9 (36.0) 7 (33.3) Complete Mayo score at W10, mean (SD) 2.8 (1.7) 3.5 (1.8) 3.8 (1.9) 4.8 (2.1) 3.7 (1.6) PMS at W10, mean (SD) 1.3 (1.0) 1.9 (1.2) 2.1 (1.4) 3.0 (1.8) 2.1 (1.2) Mucosal endoscopy subscore of 2 (moderate) at BL, n (%) 14 (36.8) 39 (45.9) 28 (46.7) 9 (36.0) 8 (37.1) Mucosal endoscopy subscore of 3 (severe) at BL, n (%) 24 (63.2) 46 (54.1) 32 (53.3) 16 (64.0) 13 (61.9) Geboes continuous index score at W10, mean (SD) 5.9 (4.3) 5.8 (4.9) 7.0 (5.2) 7.7 (5.1) 7.6 (4.8)
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 | 0.005 | 0.007 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.002 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.000 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.003 | 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".