Comparing Two Restrictive Diets for Treating Eosinophilic Esophagitis in Children [Internet]

Excerpt

Background: Eosinophilic esophagitis (EoE) is a chronic esophageal inflammatory disease triggered by exposure to food allergens and effectively treated with dietary therapies. Restrictive diets that eliminate all common food allergens (ie, milk, egg, wheat, soy, fish/shellfish, peanuts/tree nuts) are difficult to implement, especially in children, and pose a significant negative impact on patient quality of life (QoL). Thus, less restrictive—but efficacious—diets represent an option of interest to patients. A prospective evaluation of the efficacy of less restrictive elimination diets in improving histology, symptoms, and QoL outcomes in pediatric patients with EoE is lacking. Our study addressed this notable gap.

Objectives: The aim of this study was to determine and compare the efficacy of less restrictive diets—the 1-food elimination diet (1FED [ie, milk elimination alone]) vs the 4FED (ie, milk, egg, wheat, and soy avoidance) in pediatric patients with EoE (phase 1). The primary efficacy end point was the change in symptom score (Pediatric EoE Symptom Score [PEESS] v2.0). Additional end points included the effect of each therapy on histology and patient-reported outcomes using Pediatric Quality of Life Inventory (PedsQL) and Patient-Reported Outcomes Measurement Information System (PROMIS®) questionnaires. A secondary objective of this study, which constituted phase 2, was to determine the rate of remission after treatment with 4FED (in participants failing 1FED in phase 1) or swallowed glucocorticoids (SGCs; in participants failing 4FED in phase 1).

Methods: This study was a prospective, nonblinded, randomized trial. Children (aged 6-17 years) with active EoE were consecutively recruited by 10 EoE specialty centers in the United States and treated with either 1FED or 4FED for 12 weeks (phase 1). Dietary therapy nonresponders on 1FED in phase 1 continued on to 4FED in phase 2 for 12 weeks. The 4FED nonresponders received SGCs (fluticasone propionate) in phase 2 for 12 weeks. Patients were assessed before initiating therapy and after 12 weeks of therapy in each treatment phase. Differences between baseline and end of treatment scores and differences between groups were evaluated at the end of each phase using a linear mixed-effects model.

Results: Sixty-three patients were randomly assigned 1:1 to the study (1FED, n = 38; 4FED, n = 25). At the end of phase 1, symptom scores (PEESS) were significantly lower (improved) from baseline after 1FED and 4FED treatments. Change in PEESS score was greater in 4FED (−25.0 [95% CI, −33.1 to −16.9]) compared with 1FED (−14.5 [95% CI, −20.4 to −8.6]) with a mean difference of 10.5 (95% CI, 0.5-20.6; P = .041). Histologic remission rates (peak count <15 eosinophils per high-power field) were similar between 1FED (44% [95% CI, 27%-62%]) and 4FED (41% [95% CI, 18%-67%; P = 1.0]) in phase 1. Further, histologic feature scores were significantly improved from baseline after treatment in 1FED and 4FED groups, but the changes were not different between groups (1FED grade: −0.16 vs 4FED grade: −0.14, P = .792; 1FED stage: −0.13 vs 4FED stage: −0.11; P =.752). Clinically important improvements in PedsQL 4.0 psychosocial scores (6.1 points [95% CI, 1.4-10.8; P = .012]) and PROMIS emotional well-being scores (anger, −5.0 points [95% CI, −9.5 to −0.5; P = .026]; anxiety, −5.9 points [95% CI, −9.9 to −2.0; P = .004]) scores were found in participants following 1FED (phase 1). In phase 2, after 12 weeks of therapy, 1 of 8 participants following 4FED and 2 of 4 participants taking SGCs were in histologic remission.

Conclusions: The 4FED is comparatively superior to the 1FED in reducing symptoms. However, a 12-week course of milk elimination achieves improvement in histology (approximately 40% remission rate), reductions in symptom scores, and a generally positive impact on QoL. As such, a 1FED (ie, milk elimination alone) should be considered a reasonable first treatment option.

Limitations: The study included nonblinded interventions, which might have introduced bias into patient-reported outcomes. Due to small sample size, especially the 4FED group, the study had low statistical power to detect some potential treatment differences. The large dropout rate following phase 1 limits the validity of phase 2 results.