Background: A milk-free diet with substitute formula should be established when immediate symptomatic hypersensitivity to cows' milk protein (CMP) is diagnosed, and therefore an accurate diagnosis is very important.
Objective: This study aims to find the optimal cutoff values for specific IgE antibody levels that discriminate between allergic and tolerant infants by using cows' milk and its principal proteins as allergens.
Methods: A prospective study was carried out on 170 patients under 1 year old (mean, 4.8 months). These patients were seen consecutively over a 4-year period in our outpatient clinic and for the first time because of a reaction suggesting immediate hypersensitivity after ingestion of cows' milk formula. A clinical history, prick test with cows' milk and its proteins (alpha-lact-albumin, beta-lactoglobulin, and casein), determination of specific IgE antibodies with the CAP system FEIA for the same allergens as for the prick test, and a challenge test according to the diagnostic protocol were performed in all of the children. A study of validity of the prick test (cutoff point, 3 mm) and CAP system by using different cutoff points in the specific IgE values for cows' milk and its proteins were also analyzed.
Results: Prevalence of immediate symptomatic hypersensitivity to CMP in this study was 44%. When both the whole milk and its principal milk proteins were used in the prick test, the negative predictive value was very high, and a negative value excluded allergy in 97% of the patients. When the different cutoff points of the specific IgE for milk were analyzed, 2.5 KU(A)/L had a positive predictive value of 90% and 5 KU(A)/L had a positive predictive value of 95%.
Conclusions: When diagnosing immediate hypersensitivity to CMP in infants, negative skin test responses exclude allergy in most of the patients. If the prick test response is positive, specific IgE levels for cows' milk may be helpful. If these values are 2.5 KU(A)/L or greater, the challenge test should not be performed because of its high positive predictive value (90%).