Oral Food Challenges to Milk and Egg in Children: Associations with Skin Prick Tests and IgE Sensitization Profiles

Background: Oral food challenge (OFC) remains the gold standard for diagnosing food allergy; however, it is time-consuming and associated with the risk of allergic reactions. The identification of reliable biomarkers capable of predicting OFC outcomes could improve patient selection and reduce the number of OFCs required in clinical practice. Among the most widely used biomarkers are skin prick tests (SPTs) and serum-specific IgE (sIgE) measurements, including component-resolved diagnostics. However, their diagnostic utility, as well as the comparability of different diagnostic platforms, remains incompletely defined in cow’s milk (CM) and hen’s egg (HE) allergy. Methods: We retrospectively analyzed results of diagnostic tests obtained in children undergoing OFCs to baked milk, raw milk, baked egg, and boiled egg. Diagnostic performance of SPTs to either allergen extracts or native food allergens, as well as sIgE measured using ImmunoCAP and ALEX2 platforms, was evaluated using receiver operating characteristic (ROC) analysis and logistic regression models. Results: Diagnostic performance varied across allergy phenotypes. SPTs to native food allergens demonstrated good discriminative capacity for raw milk and boiled egg allergy, whereas SPTs to standardized extracts showed limited diagnostic utility overall. In baked milk allergy, casein-sIgE measured by ImmunoCAP demonstrated the best diagnostic performance. In raw milk allergy, α-lactalbumin-sIgE measured using ALEX2 platform and β-lactoglobulin-sIgE measured by ImmunoCAP showed greater diagnostic relevance than casein-sIgE. In boiled egg allergy, SPT to raw egg yolk demonstrated the highest diagnostic value. In baked egg allergy, ovalbumin-sIgE measured using the ALEX2 platform showed the best performance, while ovomucoid-sIgE did not demonstrate clear superiority. ALEX2 showed good concordance with ImmunoCAP despite systematic differences in absolute sIgE values. Combined multivariable models provided only modest improvements over single predictors and did not achieve sufficient accuracy to replace OFCs. Conclusions: SPTs to either allergen extracts or native food allergens, as well as sIgE measurements, demonstrate moderate and heterogeneous utility in predicting OFC outcomes for CM and HE allergy. Although ALEX2 and ImmunoCAP showed comparable overall performance, platform-specific interpretation may be required. OFC remains indispensable for definitive diagnosis.