Component-resolved diagnosis and beyond: Multivariable regression models to predict severity of hazelnut allergy

Area of Science:

• Allergy and Immunology
• Clinical Diagnostics
• Predictive Modeling in Medicine

Background:

• Component-resolved diagnosis (CRD) identifies IgE associations with hazelnut allergy severity.
• Predictive models often overlook combining CRD with clinical factors.
• Accurate prediction of hazelnut allergy severity is crucial for patient management.

Purpose of the Study:

• To analyze associations between hazelnut allergy severity, sensitization patterns, and patient characteristics.
• To develop predictive models integrating clinical history and allergen-specific IgE.
• To enhance the accuracy of predicting severe hazelnut allergy reactions.

Main Methods:

• Recruited 423 hazelnut-allergic patients across 12 European cities.
• Performed IgE testing against individual hazelnut allergens and analyzed clinical data.
• Developed multiple regression models using clinical factors and sensitization patterns (CRD and extract-based) to predict symptom severity during double-blind placebo-controlled food challenges (DBPCFC).

Main Results:

• IgE against Cor a 9 and Cor a 14 were linked to severe symptoms (OR 10.5, 10.1), while Cor a 1 showed a protective association (OR 0.14).
• Combining IgE data for Cor a 1 and Cor a 9 improved predictive accuracy (AUC 0.76).
• A comprehensive model including atopic dermatitis, pollen allergy, IgE to Cor a 14, and walnut sensitization achieved an AUC of 0.91, demonstrating high predictive value.

Conclusions:

• A combined model integrating CRD, clinical background, and extract-based serology surpasses CRD alone in assessing severe hazelnut allergy risk.
• This integrated approach is particularly effective in ruling out severe reactions.
• Clinical factors and specific IgE profiles significantly enhance the prediction of hazelnut allergy severity.