Background:
The prevalence of food allergies in children is estimated to be 5.8%.The mechanism underlying food allergies are either IgE mediated, non-IgE mediated or both. Rath et al demonstrated in-vivo barrier dysfunction via confocal laser endomicroscopy (CLE) in 96% of adult patients with food allergies, with a negative predictive value of 95.7%. Similarly, Fritscher-Ravens et al demonstrated endomicroscopy findings of epithelial breaks and widening of intervillous spaces when suspected food allergens were administered directly to the duodenal mucosa of adults presenting with Irritable Bowel Syndrome (IBS) with negative serum IgE food testing results. Administering exclusion diets to these patients dramatically improved symptoms. The ability to observe these dynamic changes in real-time is promising in IBS. However, the utility of CLE in visualizing the integrity of the intestinal barrier in children with atypical presentation of food allergies is unknown.

Methods:
This was a single-center retrospective review of 19 pediatric patients between the ages of 2-19 years who underwent EGD and CLE with Food Allergy Sensitivity Testing (FAST) from September 1, 2019- June 17, 2023. Diagnosis of food allergy was determined by serologic testing for food specific IgE. Patients who were negative for food allergy testing and negative for celiac disease testing, were asked to follow a hypoallergenic exclusion diet, eliminating milk, soy, peanut, egg, seafood, and wheat, 1-2 weeks prior to the procedure. Patients also maintained a food diary.

The FAST procedure initiated with an EGD to ensure intact mucosa, following which 1-2.5 ml of 10% fluorescein administered intravenously. Duodenal mucosa was flushed with normal saline. The first potential allergen was applied to the duodenum and inspected with the CLE probe after two minutes. A positive CLE reaction consisted of 1) fluorescein leakage, 2) cell shedding (Fig1). This was repeated for three potential allergens per patient.

Results:
Mean age was 11.57 years. All 19 patients underwent food allergy testing. Two patients did not have documentation of celiac disease testing (Table 1). A negative food allergy as well as negative serological testing for celiac disease was observed in 12 (63.1%) patients. Of the 12, 8 (66.7%) had a positive CLE reaction. CLE reaction to wheat was the most common. On subsequent follow up visits, seven patients with positive CLE/FAST, reported resolution of symptoms once the offending agent was removed from their diet.

Discussion & Conclusion:
Detection of potential food allergens using CLE with FAST is a promising screening tool to detect atypical presentation of food allergies in patients with (1) negative celiac and food allergy testing and (2) food allergy testing showing very low/low IgE response with questionable symptomatology to food allergens.

Fig1. CLE image showing fluorescein leakage and cell-shedding post allergen exposure.

Table1. Patients stratified by food allergy and celiac disease testing.

Clifton Huang

Appearances