Picking the right diet for EoE

A study from Philadelphia/CHOP offers more insight into food selection diets for eosinophilic esophagitis (EoE) (Spergel JM, et al, J Allergy Clin Immunol 2012; 130: 461-7) –thanks to Seth Marcus for forwarding this article to my attention.

For this study, the authors examined their database of 1187 patients. While the data was collected prospectively, this was a retrospective study. Of this 1187, the authors excluded patients with proton pump inhibitor-responsive EoE (n=191) along with patients with more extensive eosinophilic GI diseases (n=55).

Among the remaining 941, the male-to-female ratio was 2.8:1 and the average age was 6.4 years. Concurrent atopic disorders were common: 64% had rhinitis, 50% had asthma, and 24% had atopic dermatitis. Only 18% had no atopic disorders.

The actual number for the study though was 319. Among the 941 noted above, 148 were receiving medications (n=130 for topical steroids), and causative foods were not identified in 474. In some of these patients, families were content to stick with a multiple food elimination without determining with certainty which foods were truly necessary.

In less than 5% of patients, a strict elemental diet was used. In this group, the population was younger (average 2.8 years). Biopsy improvement was noted in “upward of 98%.”

Key findings:

Elimination of foods based on combined skin prick tests (SPT)/atopy patch tests (APT) had an identical response to the six food group diet –53%. The allergy testing group had less eliminated foods (average 3.2 foods) compared with 8 food groups in SFED.
Elimination of milk with SPT/APT testing resulted in 77% response. Authors note that there was a “particularly high false-negative rate (34%)” with milk testing (SPT/APT).
Elimination of top 8 allergens: milk, soy, egg, wheat, and meats [chicken, turkey, pork, beef] had an identical response of 77%.
Elimination of milk, egg, and wheat had a success rate of 48%. Milk only elimination had a 30% response rate.
Most common foods by biopsy: milk (35%), egg (13%), wheat (12%), soy (9%), corn (6%)
Most common foods by symptoms: milk (19%), egg (11%), wheat (9%), soy (10%), beef (8%)
IgE-mediated foods: milk (10%), egg (17%), soy (4%), peanut (22%)

Additional useful information in the addendum of methods notes their technique for APT testing (which is not standardized across centers). The authors use 2 g of dry foods in 2 mL of isotonic saline solution for most foods; for milk, they use 3 g of powdered milk with 1 mL of isotonic saline. Then these mixtures are placed in aluminum cups (6- or 12-mm Finn chambers on Scanpore). These cups are placed on patient’s backs and removed at 48 hours and read at 72 hours.

In addition, for each food, their tables list predictive values (positive predictive value, negative predictive value, sensitivity, and specificity) for SPTs, and for APTs. Overall, the predictive values are quite variable and much different from the general population. For example, in the general population, the negative predictive value is essentially 100% for combination of SPT/APT.

Previous related blog entries: