An elimination diet was effective at inducing symptomatic, endoscopic, and histologic improvement in a study of 50 adults with eosinophilic esophagitis, Dr. Nirmala Gonsalves and her colleagues reported in the June issue of Gastroenterology.
Eliminating the six most common food allergens from the diet for 6 weeks reduced by at least half the peak eosinophil counts in 78% of the study subjects. Moreover, 64% of all subjects achieved a reduction to the target level of 5 or fewer eosinophils per high-powered field.
In the prospective clinical trial, reintroducing the food allergens one at a time allowed the investigators to identify the causative dietary agent(s) in all 20 subjects who chose to participate in a reintroduction experiment.
Eosinophilic esophagitis (EOE) is typically characterized by dysphagia, heartburn, and food impaction, and in this study the proportions of patients reporting these symptoms at enrollment were 96%, 94%, and 74%, respectively. The average duration of symptoms was 6 years.
This is the first study to prospectively examine the effectiveness of an elimination diet in adult EOE, and the first to implicate food allergens in the pathogenesis of the disorder in the adult population, said Dr. Gonsalves and her associates at Northwestern University, Chicago (Gastroenterology 2012 June [doi: 10.1053/j.gastro.2012.03.001]).
Food allergens have been proposed as a primary trigger of EOE in the pediatric population, but an elimination diet was found ineffective in a small series of adult patients, and no further research has examined the issue in adults.
In this study, 25 men and 25 women with EOE consulted with a dietitian specifically trained in allergy diet restriction, who provided them with sample menus and shopping guides. The allergens they avoided were milk, soy, egg, wheat, peanuts/tree nuts, and shellfish/fish.
Before and after the 6-week intervention, all the subjects underwent esophagogastroduodenoscopy with biopsy. Those who achieved histologic remission on the diet were invited to undergo sequential reintroduction of each of the six food allergens, with endoscopies and biopsies to confirm their specific food triggers.
The mean peak eosinophil count at baseline was 53 (range, 17-108) per high-powered field. In the patient population as a whole, this decreased significantly, from 34 to 8 in the proximal esophagus and from 48 to 13 in the distal esophagus.
The primary study end point was histologic improvement in esophageal eosinophilia. A complete response was defined as a peak eosinophil count of 5 or fewer per high-powered field, and 64% of the patients achieved this. A near-complete response was 10 or fewer eosinophils per high-powered field (70% of patients), and a partial response was a 50% or greater reduction in peak eosinophil count (78% of patients).
When the results were broken down by level of treatment response, complete responders had a reduction in mean peak eosinophil count from 32 to 3 in the proximal esophagus and from 48 to 3 in the distal esophagus. Partial responders had a reduction from 59 to 9 in the proximal esophagus and from 50 to 16 in the distal esophagus.
Even "nonresponders" had some reductions, from 42 to 30 in the proximal esophagus and from 62 to 55 in the distal esophagus, the investigators said.
At the same time, dysphagia symptom scores decreased in 94% of the study subjects. The frequency, severity, and duration of dysphagia all declined significantly. Endoscopic features also improved in 78% of patients in response to the elimination diet.
Twenty patients completed the dietary reintroduction process, and symptomatic, endoscopic, and histologic abnormalities recurred in all of them after exposure to the causative food. The median time to symptom recurrence was 3 days.
The most common food triggers of EOE were wheat (60% of cases) and milk (50%), followed by soy (10%), nuts (10%), and egg (5%). Three study subjects had more than one food trigger; milk and wheat were the most common combination. These findings parallel those seen in pediatric studies reported in the literature.
Remarkably, none of the study subjects reported having food allergies or intolerances at baseline. And although all of them underwent skin prick testing for an allergic response to the six foods and to aeroallergens at baseline, the results were not predictive of response to the dietary intervention.
"Skin prick testing accurately predicted only 13% of causal agents, and 67% of patients who had a food trigger identified by the reintroduction process had a negative skin prick test to all foods," Dr. Gonsalves and her associates said.
It remains unclear whether eliminating a food trigger from the diet will prove to be a feasible long-term therapy for adults. "A maintenance arm of this study is ongoing to answer this question," the authors said.