DOP08 Serum proteomics predict endoscopic remission in patients with Crohn’s Disease
Bibliographic record
Abstract
Abstract Background Recent progress in deciphering the complex pathogenesis of Crohn’s disease (CD) has yielded several effective biologicals. However, ambitious therapeutic goals remain unfulfilled as almost 30% of patients are primary non-responders to a particular biological. This underscores the need for easy-to-implement biomarkers that predict (non-)remission. We aimed to identify serum protein biomarkers that predict endoscopic remission in CD patients. Methods Serum samples from 169 consecutive CD patients with active endoscopic disease (presence of ulcerations) before starting a biological [infliximab (IFX), adalimumab (ADA), vedolizumab (VDZ) or ustekinumab (UST)] to which they were naïve were collected. Patients were prospectively followed with endoscopic re-assessment after 6–12 months. There were 102 patients (Table 1) with endoscopic remission (SES ≤ 2 or disappearance of all ulcers), whereas 67 showed no improvement. Two independent and complementary proteomic platforms were used: 644 proteins belonging to predesigned assays were quantified using Proximity Extension Assay (PEA) technology (Olink Proteomics AB, Sweden). Second, wide protein discovery mass spectrometry (MS)-based technic (Caprion, Canada) was used and quantified another 985 proteins. A multivariate modelling framework was then applied on a randomly selected training sub-cohort (85%). Predictive performance of identified panels was assessed on the remaining test sub-cohort (15%). We sought to implement the same framework on the drug-specific subgroups; however, train/test splitting was not possible in IFX or ADA subgroups due to very few observations in the non-remission arms which diminishes the possibility for reliable predictive modelling. Results Applying the modelling framework on training sets from the general cohort, VDZ subgroup and UST subgroup, proteomic panels were selected and consisted of 26, 6 and 8 proteins, respectively, and showed high performance in the test sets (Table 2). VDZ and UST panels shared only 2 proteins each with the general panel, and had no predictive power (accuracy ≤ 0.5) when used to predict other subgroups, making them specific to their respective drugs. Selected proteins are involved among others in pro-inflammatory, extracellular matrix modelling, coagulation and cellular-vascular interaction pathways (Table 3). Conclusion Applying a multivariate machine learning algorithm on a wide pool of serum proteomics analysed through two discovery technics, we were able to identify 3 proteomic panels that can predict endoscopic (non)remission in patients with CD. Exact implication of these proteins in intestinal inflammation and a validation in an independent cohort is being further investigated.
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 | 0.001 | 0.002 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.000 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.001 | 0.000 |
| 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; 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".