- For children with newly diagnosed ulcerative colitis (UC), early mild disease and remission by week 4 predict long-term remission.
- Genetic, microbiome data may help predict escalation to anti-TNFα therapy.
Why this matters
- Few robust data guide prognostication or treatment of newly diagnosed UC in children.
- 400 participants evaluable at week 52.
- 38% (150/400) achieved remission.
- Remission predictors: mild clinical severity, week-4 remission, baseline hemoglobin ≥10 g/dL in those not achieving week-4 remission.
- Model area under curve (AUC), 0.70 (95% CI, 0.65-0.75).
- Specificity, 77%.
- Validation in separate cohort: mild clinical severity, week-4 remission were predictive: AUC, 0.65.
- Adding genetic and microbiome data: AUC, 0.75.
- Escalation predictors: Mayo score ≥11; lack of week-4 remission; low baseline hemoglobin, serum 25(OH)D, rectal eosinophils.
- Model: AUC, 0.78.
- Adding genetic and microbiome data: AUC, 0.88.
- Prospective multicenter inception cohort PROTECT (n=428).
- Children ages 4-17 years with newly diagnosed UC underwent guideline-based treatment with mesalazine (Pentasa; Shire US Inc.) or corticosteroids.
- Using baseline and 4-week data, including pretreatment rectal gene expression and gut microbiome, authors derived models to predict remission, escalation.
- Outcome: week-52 corticosteroid-free remission with mesalazine alone.
- Funding: NIH.
- Insufficient data to validate escalation model.