Background: In pollen-related food allergy, extracts for skin prick tests (SPTs) are often not standardized, and the test reliability is affected by false-negative reactions.
Objective: We sought to evaluate a panel of recombinant allergens (RAs) derived from one allergenic food for use in component-resolved in vivo diagnosis, taking cherry as a model food.
Methods: Seventy-nine subjects were included in the study: 24 Swiss patients (group 1) with a positive double-blind placebo-controlled food challenge result to cherries, 23 patients with birch pollen allergy but without cherry allergy (group 2), 23 nonatopic subjects (group 3), and 9 Spanish patients with a history of a cherry allergy (group 4). SPTs were performed in duplicate by using recombinant cherry allergens (Bet v 1-related allergen: recombinant (r) Pru av 1; profilin: rPru av 4; and lipid transfer protein: rPru av 3) in concentrations of 10, 50, and 100 microg/mL. Furthermore, IgE reactivity to rPru av 1, rPru av 4, and rPru av 3 was assessed by means of immunoblot analysis.
Results: SPT responses with rPru av 1, rPru av 4, and rPru av 3 were positive in 92%, 17%, and 4% of the patients in group 1; in 74%, 30%, and 0% of the patients in group 2; in 0%, 22%, and 89% of the patients in group 4; and negative for all nonatopic subjects (group 3). Thus the sensitivity of a positive SPT response to at least one of the 3 RAs was 96%. The specificities, negative predictive values, and positive predictive values with the 3 RAs were 100%, 96%, and 100% if calculated in relation to the nonatopic control group but 17%, 79%, and 60% when calculated in relation to the control group with birch pollen allergy. The correlation between SPT and immunoblotting results was excellent. Sensitization to rPru av 3 was associated with more severe symptoms than sensitization to rPru av 1.
Conclusions: SPTs with RAs proved to be highly sensitive for diagnosis of cherry allergy. Component-resolved in vivo diagnosis with standardized amounts of stable RAs allows us to determine sensitization patterns directly, to correlate them with severity of clinical symptoms, and to analyze geographic differences.