Evaluation of different protocols for classification of pediatric hypersensitivity reactions to nonsteroidal anti-inflammatory drugs: Children with underlying allergic disease should be a separate subgroup

Abstract

Background: Different recommendations for the classification of nonsteroidal anti-inflammatory drug hypersensitivity reactions (NSHSR) in children have been reported but a shortage still exists. Objective: The aim of the present study was to evaluate the inclusivity of two European Academy of Allergy and Clinical Immunology (EAACI) position paper classifications and to characterize the factors that underlie classification discordance in children. Methods: Patients with a history of NSHSR were evaluated with a standardized diagnostic protocol according to EAACI/ European Network for Drug Allergy (ENDA) recommendations. Children were classified and compared according to the EAACI 2013 and the pediatric EAACI/ENDA 2018 classifications. Subjects who were unclassified and those who were classified were compared. Results: Of 232 patients (median [interquartile range] age 6 years (4‐11 years) with a history of NSHSR, 52 (22.4%) were confirmed with diagnostic tests. Thirty-six (69.2%) were classified as having cross-intolerance, whereas 16 patients (30.8%) were classified as selective responders. Eleven of the confirmed cases (21.2%) could not be categorized according to the 2013 EAACI classification, whereas this number was six adolescents (11.5%) when the 2018 EAACI/ENDA pediatric classification was used. Patients who were unclassified and who were all cross-intolerant were more likely to have atopic sensitization (p = 0.001) and asthma as an underlying disease (p = 0.03), higher serum eosinophil count (p = 0.022), and total immunoglobulin E levels (p = 0.007) compared with those who fit well into the classification. In multivariate regression analysis, the presence of atopic sensitization (adjusted odds ratio 20.36 [95% confidence interval, 2.14‐193.48]; p = 0.009) was found to be the only significant underlying factor for an unclassified and/or blended phenotype. Conclusion: The 2013 EAACI classification resulted in a high rate of subjects who were unclassified. Despite better clinical utility, the recent pediatric EAACI/ENDA classification system still has shortcomings in terms of inclusivity for adolescents. Mostly, children with underlying allergic diseases could not be classified by the current guidelines. We propose to classify them as a separate pediatric cross-intolerance subgroup because the underlying mechanism may involve more than cyclooxygenase 1 inhibition.