WP 2.1 - Predictive values of Component Resolved Diagnostics in food allergy


A food allergic diagnosis is today based on case history, relevant sensitization measured with specific IgE (s-IgE) antibodies and confirmed with an oral challenge. Food items consist of a variety of different proteins, some with high allergenic potential, other harmless. Measurement of S?IgE includes the sum of all antibodies directed towards any protein found in the food investigated. Component Resolved Diagnostic (CRD) utilize, that relevant allergenic proteins are isolated (either purified or recombinant), and IgE antibodies therefore binds to only one unique protein. Specific IgE antibodies (s?IgE) have previously been correlated to clinical outcome, but impact remains low. Protein components used in CRD provide high diagnostic and prognostic potential, but data on longitudinal correlations to the individual course of allergy are sparse. A mathematical model to describe patient’s individual longitudinal course, based on their relevant serology, have never been developed, but would be advantageous and a valuable tool for both patient and clinician.



Levels of specific peanut components have been shown to correlate partially to the outcome of challenges . A unique longitudinal relationship between peanut components has been identified (figure). It is assumed that sensitization to components affects the course, e.g. persistence or tolerance development, however the role of components is complex and not comparable between components or types of allergens; therefore multiple allergens should be approached statistically differently. By considering information from irregularly sampled serology components we wish to predict the longitudinal course of each individual. In particular we wish to identify the important components and the time period for measurement in order to provide the highest predictability of tolerance status.



Blood samples routinely obtained during diagnostic evaluation of patients suspected to all food allergens have been stored at the Allergy Centre since 1998, and has now be re?analyzed for newly developed specific protein components. Pairing these novel serological data with already existing data from the same patients’ oral challenges will create the largest available database on multiple foods, following more than 1000 patients longitudinally. Measurements of the different components occur, however, at irregular time points for each individual and independent of the time points where the actual allergy tolerance is determined. When predicting each course, we need to account for the inherent censoring in the single patient, as well as the available cumulative exposure history of each individual. This requires development of an extension of the classical Aalen model such that it can accommodate censored event data together with time?varying influence of the individual component values.



Webdesign af mira.dk