Tryptophan metabolites profile predict remission with dietary therapy in pediatric Crohn's disease.

Abstract

Background: Crohn's disease (CD) exclusion diet combined with partial enteral nutrition (CDED + PEN) or exclusive enteral nutrition (EEN) is effective in inducing remission in mild-to-moderate pediatric CD. Although CDED + PEN is better tolerated and has higher compliance compared to EEN, a subset of patients does not achieve remission. Diet-induced remission is shown to be positively associated with specific changes in tryptophan-metabolites.

Objectives: To investigate whether the abundance of baseline fecal tryptophan-metabolites predicts dietary therapy outcomes in pediatric CD.

Design: Diagnostic accuracy study and secondary analysis of previously conducted Randomized Controlled Trial (RCT).

Methods: Twenty-six patients from previously performed RCT of mild-to-moderate pediatric CD were included. The patients were classified as having clinical remission (R) (n = 19 in total; CDED + PEN = 10 and to EEN = 9) or No-Remission (NR) (n = 7 in total; CDED + PEN = 3 and EEN = 4) following 6 weeks of therapy, based on the Pediatric Crohn's Disease Activity Index score (⩽10 = remission). We performed a targeted quantitative analysis of 21 tryptophan-metabolites in baseline (t = 0) fecal samples from both groups, utilizing liquid chromatography coupled with quadrupole mass spectrometry. Receiver operator characteristic curve (ROC) and random forest analysis (RFA) were used to assess the predictive power of fecal tryptophan-metabolites for dietary outcomes at baseline. Ratios of tryptophan-metabolites were compared to investigate different downstream tryptophan pathways.

Results: Baseline fecal kynurenine level was significantly higher in NR compared to R for CDED + PEN (p = 0.02) and EEN (p = 0.04). ROC analysis highlighted the robust predictive power of kynurenine for CDED + PEN (area under the curve (AUC = 0.97)) and EEN (AUC = 0.88)-induced remission. RFA corroborated these observations. The ratio serotonin/kynurenine was the strongest predictor of CDED + PEN-induced remission (AUC = 1). The ratio 5-hydroxytryptophan/kynurenine (AUC = 0.88) predicted EEN-induced remission. By combining data from CDED + PEN and EEN, kynurenine (AUC = 0.91) and ratios of quinolinic acid/kynurenine (AUC = 0.93) and kynurenine/indole-3-acetic acid (AUC = 0.88) demonstrated strong predictive performance for dietary therapy-induced remission.

Conclusion: Baseline tryptophan metabolites have the potential to serve as a biomarker for dietary remission in pediatric CD. Some tryptophan metabolite ratios showed the most promising predictive capabilities. If confirmed in validation studies, baseline fecal tryptophan markers may be able to provide much-needed guidance to personalize dietary intervention within the management of pediatric CD.

Trial registration: NCT01728870.