Confocal laser endomicroscopy (CLE) enables real-time visualization of epithelial barrier responses following mucosal exposure to food antigens. While CLE has demonstrated value in detecting non–immunoglobulin E (IgE)–mediated food allergy in adults, evidence in pediatrics remains limited. This study aimed to evaluate the feasibility, diagnostic yield, and clinical impact of CLE-guided duodenal food challenge in children with food-related gastrointestinal symptoms and negative conventional allergy testing.

We prospectively enrolled children (2–18 years) with chronic gastrointestinal symptoms temporally associated with food ingestion and negative serum IgE and skin-prick tests. During upper endoscopy, standardized food antigens (cow’s milk, wheat, soy, egg, yeast) were sequentially applied to the duodenal mucosa. CLE was performed immediately after each challenge. A CLE-positive reaction was defined by the presence of at least two features within 5 minutes: epithelial fluorescein leakage, widening of the intervillous space, fluorescent signal between enterocytes, or an increase in intraepithelial lymphocytes. Following CLE, targeted elimination diets were prescribed based on the identified offending antigen. Clinical outcomes were assessed at 8–12 weeks.

Thirty-six children were enrolled. CLE was feasible and safe in all children. A positive reaction occurred in 24/36 patients (66.7%). The most frequent triggers were cow’s milk (41.7%), wheat (29.2%), soy (16.7%), egg (8.3%), and yeast (4.1%). CLE-positive patients consistently demonstrated rapid epithelial leakage and intervillous widening, indicating acute barrier dysfunction. At follow-up, 20/24 CLE-positive children (83.3%) achieved significant symptom improvement (>60%) after targeted dietary elimination, whereas CLE-negative children showed minimal benefit from empirical diet modification.

In this pediatric cohort, CLE revealed a high prevalence of non–IgE-mediated food allergy not identified by standard allergy tests. CLE-guided food elimination led to substantial clinical improvement, underscoring CLE’s diagnostic value as a real-time functional tool for evaluating food-induced mucosal responses in children and for guiding personalized dietary therapy.