The component-specific to total IgE ratios do not improve peanut and hazelnut allergy diagnoses

Background: Specific IgE measurement predicts the outcome of oral food challenges with considerable uncertainty when evaluating food allergy. Objective: Our aim was to assess whether accounting for the ratio of component- or allergen-specific to total IgE can improve this prediction. Methods: This multicenter study collected blood samples from children with suspected peanut or hazelnut allergy referred to allergy specialist clinics for food challenges. Specific IgE to peanuts, hazelnuts, and their components (Ara h 1, Ara h 2, Ara h 3, Ara h 8, Cor a 1, Cor a 8, Cor a 9, and Cor a 14) and total IgE levels were determined by using the ImmunoCAP-FEIA. Specific to total IgE ratios were compared with raw IgE levels in terms of discrimination and prediction. Results: Eighty-eight (43%) of 207 children with suspected peanut allergy and 44 (31%) of 142 children with suspected hazelnut allergy had symptoms during food challenge. Discrimination was similar for raw and ratio measures: areas under the curve of 0.93 for Ara h 2-specific IgE versus 0.92 for the Ara h 2-specific/total IgE ratio and 0.89 for Cor a 14-specific IgE versus 0.87 for the Cor a 14-specific/total IgE ratio. The probability for a positive peanut challenge with 0.35 kU/L Ara h 2-specific IgE was 16% when the total IgE level was greater than 500 kU/L compared with 51%/48% for low/medium total IgE levels (<100/100-500 kU/L). A positive hazelnut challenge with 0.35 kU/L Cor a 14-specific IgE was estimated in 7% when total IgE levels were high compared with 34%/32% with low/medium total IgE levels. Conclusions: Raw Ara h 2- and Cor a 14-specific IgE levels were the best single predictors for pediatric peanut and hazelnut allergies, suggesting the omission of challenges at very high levels. Calculating ratio measures did not improve prediction in this population. However, estimation of individual probabilities for challenge outcomes could be supported by total IgE levels because high levels might indicate lower probabilities at a given component-specific IgE level.