Evaluation of a standardized bakery product (SUTMEK) as a potential tool for baked-milk tolerance and immunotherapy research studies

Abstract

Background and Objectives: About 65-80% of children with IgE-mediated cow's milk allergy (CMA) can tolerate extensively heated milk. We have invested in the mass fabrication of a test product containing milk protein baked at 180 degrees C for 30 min (SUTMEK-milk) and a milk-free placebo (SUTMEK-placebo) to carry out a standardised double-blind placebo-controlled food challenge (DBPCFC) test in patients with CMA. Methods: We studied children with IgE-mediated CMA between 13 and 48 months of age. Specific IgEs (spIgE) to milk proteins were quantified. A DBPCFC with our bakery products was performed, and factors determining reactivity to extensively heated milk were evaluated. We also tested the applicability of SUTMEK products in baked-milk oral immunotherapy in a pilot assessment. Results: We studied 15 children (8 girls, 7 boys) with a median age of 26 months (range: 13-48 months). Nine (60%) patients tolerated a challenge with extensively heated milk, while 6 (40%) were found reactive (anaphylaxis: 2, wheezing: 2, urticaria: 2). spIgE to milk, alpha-lactalbumin, and casein, and the wheal diameter on skin prick testing were higher in the reactive group than the tolerant groups (p = 0.001, p = 0.001, p = 0.002, and p = 0.048, respectively). Receiver-operating characteristic curve analyses yielded the following cut-off values for spIgEs that would predict a reactivity to extensively heated milk; milk: 25 kU/L (area under curve, AUC: 0.981), casein: 32 kU/L (AUC: 0.983), and alpha-lactalbumin: 17 kU/L (AUC: 0.981). Nine patients have tolerated well a continued daily consumption of SUTMEK-milk or -placebo for 6 months at the desired doses. Conclusions: Our bakery products were successfully used in DBPCFC studies and qualified as an acceptable tool for use in the research of interventional tolerance induction. Although spIgE appears useful in determining children at high risk of reacting to extensively heated milk, the predictive cut-off values are still far from being perfect.