Methacholine-induced cutaneous flare response: bivariate analysis of responsiveness and sensitivity.

Cutaneous reactions to allergens exhibit a sigmoid dose-response relationship. Available methods for evaluating the allergen skin-test response do not adequately account for the sigmoid curve. Methodologic factors handicap quantitative studies of allergens based on skin-test reactivity. This problem was evaluated with a pharmacologic agonist that mimics cutaneous reactivity. Epicutaneous tests with appropriate concentrations of methacholine were used to provoke flare responses in 84 healthy subjects. A novel hyperbolic tangent model of the sigmoid dose-response curve was used to estimate responsiveness (R) as the midpoint of the sigmoid curve. Sensitivity (C) was estimated as the agonist concentration yielding a flare response equivalent to R. Estimates of sensitivity were independent of estimates of responsiveness (r=-0.0565, p=0.6642). The geometric mean methacholine sensitivity among health subjects was 287 mM/L, and average methacholine responsiveness was 4.9 mm. The mathematic model used in these studies fitted observations surprisingly well (X2(84)=37.044, p greater than 0.95). Differences in methacholine sensitivity and responsiveness related to race, sex, and allergic and/or vasomotor tendencies were detected but were subtle and did not account for a significant portion of the variation among healthy subjects. This model may provide a useful method for quantifying cutaneous immediate hypersensitivity reactions in patients.

Duke Scholars

Author Kerry L. Lee Biostatistics & Bioinformatics, Division of Biostatistics