Individualized targeted exclusion diet based on confocal laser endomicroscopy does not improve irritable bowel syndrome symptoms: a randomized controlled crossover trial.

BACKGROUND & AIMS: In the duodenum of patients with irritable bowel syndrome (IBS), acute food-induced mucosal reactions have been visualized using confocal laser endomicroscopy (CLE). In uncontrolled studies, a diet excluding trigger foods led to symptomatic improvement. Here, we aim to assess the association of food-induced alterations identified using CLE and the efficacy of a targeted diet in a controlled study. METHODS: Double-blind, controlled, crossover study in patients with IBS. Following CLE, patients excluded up to two foods that did (= real diet) and up to two foods that did not (= sham diet) cause alterations in a randomized order. Diets were followed for four weeks each. Clinical response was defined as improving ≥50 points on the IBS symptom severity score (IBS-SSS). Healthy controls underwent the same CLE protocol as patients with IBS. RESULTS: Altered mucosa was visualized in all patients either before (11%) or after food administration. Clinical response rates were not different between real and sham diet (42% vs 36%, OR = 1.33, p = 0.6). Symptoms improved both on the real and the sham diet (-30 vs -20; p = 0.7). Also, in 15 healthy volunteers, altered mucosa was visualized in 100% of the exams either at baseline (13%) or after food administration. CONCLUSIONS: Acute alterations visualized using CLE have poor sensitivity and specificity for dietary responses in IBS patients. Moreover, altered mucosa was present in 100% of healthy volunteers and likely represents a non-specific mucosal reaction to food exposure rather than pathology, challenging the use of CLE in clinical practice to identify culprit foods. (clinicaltrials.gov: NCT05097872).