Distinct patterns of cow's milk allergy in infancy defined by prolonged, two-stage double-blind, placebo-controlled...

The clinical manifestations of cow's milk allergy (CMA) are highly variable, and challenges usually identify only immediate, IgE mediated reactions.

To clearly identify CMA of immediate and delayed types using a two-stage, double-blind, placebo-controlled food challenge (DBPCFC), and to prospectively compare the clinical history and analyses of specific IgE antibodies to milk in predicting outcome of DBPCFC.

A total of 69 patients (33 girls, 36 boys) were recruited for study based on a history highly suggestive of CMA and resolution of symptoms on a bovine protein-free diet. After skin-prick tests (SPTs) and search for allergen-specific serum IgE antibodies by enzyme allergosorbent test (EAST), a two-stage DBPCFC was performed over several days.

Of 16 patients (mean age 36.9 months) classified as probable immediate reactors based on the history, 10 (62.5%) had a positive DBPCFC with similar patterns to historical adverse reactions (< or = 2 h after milk exposure). The other 53 (77%) patients (17.3 months) had a history of probable delayed type CMA presenting with predominantly gastrointestimal symptoms from 2 h and up to 6 days after milk exposure. Of these, 15 (28.8%) had a positive DBPCFC, again with a symptom pattern similar to the history. Sensitivity/specificity of SPT was similar to that of EAST for both immediate (70/83% and 62/83% respectively, NS) or delayed (0/97% and 0/97%) CMA confirmed by DBPCFC.

Using our two-stage, prolonged DBPCFC, we clearly identified two groups of children with CMA, reflecting different pathogenesis of either immediate-type IgE-dependent, or delayed-type IgE-independent allergy. Although useful in immediate reactors, IgE antibody determination cannot predict the outcome of DBPCFC in delayed reactors. A thorough clinical history was the most helpful tool to predict the type of response in challenge positive patients.