Early epitope-specific IgE antibodies are predictive of childhood peanut allergy

Abstract

Background: Peanut allergy is characterized by the development of IgE against peanut antigen.

Objective: We sought to evaluate the evolution of epitope-specific (es)IgE and esIgG₄ in a prospective cohort of high-risk infants to determine whether antibody profiles can predict peanut allergy after age 4 years.

Methods: The end point was allergy status at age 4⁺ years; samples from 293 children were collected at age 3 to 15 months and 2 to 3 and 4⁺ years. Levels of specific (s)IgE and sIgG₄ to peanut and component proteins, and 50 esIgE and esIgG₄ were quantified. Changes were analyzed with mixed-effects models. Machine learning algorithms were developed to identify a combination of antigen- and epitope-specific antibodies that using 3- to 15-month or 2- to 3-year samples can predict allergy status at age 4⁺ years.

Results: At age 4⁺ years, 38% of children were Tolerant or 14% had Possible, 8% Convincing, 24% Serologic, and 16% Confirmed allergy. At age 3 to 15 months, esIgE profiles were similar among groups, whereas marked increases were evident at age 2 and 4⁺ years only in Confirmed and Serologic groups. In contrast, peanut sIgE level was significantly lower in the Tolerant group at age 3 to 15 months, increased in Confirmed and Serologic groups but decreased in Convincing and Possibly Allergic groups over time. An algorithm combining esIgEs with peanut sIgE outperformed different clinically relevant IgE cutoffs, predicting allergy status on an "unseen" set of patients with area under the curves of 0.84 at age 3 to 15 months and 0.87 at age 2 to 3 years.

Conclusions: Early epitope-specific plus peanut-specific IgE is predictive of allergy status at age 4⁺ years.

Keywords: Ara h 1; Ara h 2; Ara h 3; Bead-Based Epitope Assay; IgE; IgG(4); Peanut allergy; antibodies; epitopes; machine learning; precision medicine.