Detection of clinical markers of sensitization to profilin in patients allergic to plant-derived foods

Background: A proper classification of patients allergic to plant-derived foods is of pivotal importance because the clinical features of allergic reactions to fruits and vegetables depend on the nature and characteristics of proteins responsible for sensitization. However, in normal clinical settings this is presently impossible. Objective: We sought to detect clinical markers of sensitization to profilin. Methods: Seventy-one patients allergic to fruits and vegetables but not sensitized to lipid transfer protein or natural rubber latex were studied. Food allergy was ascertained on the basis of clinical history and positive skin prick test responses with fresh foods, commercial extracts, or both. Allergies to foods that had caused less than 2 adverse reactions were confirmed by means of open oral challenge. IgE reactivity to rBet v 1/rBet v 2 and to natural Phleum species profilin were detected. Moreover, IgE to the 30- to 40-kd and 60- to 90-kd birch pollen-enriched fractions, which also can be involved in cross-reactivity phenomena, were measured in sera from 52 patients by means of ELISA. Results: On the basis of in vitro tests, 24, 18, and 25 patients turned out to be sensitized to Bet v 1, Bet v 2, or both, respectively. Four patients had negative test results for both allergens. Hypersensitivity to Bet v 2 was strongly associated with clinical allergy to citrus fruits (39% in patients monosensitized to Bet v 2 vs 4% in patients monosensitized to Bet v 1, P < .025), melon or watermelon (67% vs 0%, P < .001), banana (66% vs 8%, P < .001), and tomato (33% vs 0%, P < .05), whereas Bet v 1 sensitivity was associated with clinical allergy to apple (100% vs 39%, P < .001) and hazelnut (56% vs 0%, P < .001). The sensitivity of a history of allergy to gourd fruits, citrus fruits, tomato, banana, or a combination thereof as a means to detect profilin-hypersensitive patients was 85% (41148). The specificity of an allergy to any of these fruits exceeded 85%, with positive predictive values ranging between 68% and 91%. Conclusion: In clinical settings in which laboratory investigations are not easily accessible, allergy to melon, watermelon, citrus fruits, tomato, and banana can be used as a marker of profilin hypersensitivity once a sensitization to natural rubber latex and lipid transfer protein is ruled out.