Evaluation of Testing with Baked Milk Muffin to Predict Safe Ingestion of Baked Milk in Unbaked Milk Allergic Subjects

Rationale: Clinical tolerance to baked milk has been associated with small skin prick tests (SPT) to milk extract and low casein ImmunoCAP and, in a retrospective study, a negative SPT to a baked milk muffin slurry.  We hypothesized that a baked milk muffin slurry SPT would have high sensitivity and poor specificity for baked milk reactivity.

Methods: Children aged 2-16 years with a clinical history of milk allergy and a SPT to milk extract 8-14mm were prospectively recruited from office-based allergy practices. Exclusions included uncontrolled asthma and symptomatic accidental exposure to baked milk within 6 months. ImmunoCAP was not required for screening however patients were excluded if ImmunoCAP milk>24.5 or casein>20.2 kIU/ml were obtained clinically. Investigator-blinded SPT to 1g of egg-free baked milk muffin in 10ml saline was performed followed by a graded oral challenge to the same muffins (total 2.6g of milk protein) in hospital. Statistical significance was analyzed with Fisher Exact test. Institutional ethics approval was granted.

Results: Thirty children were enrolled. The prevalence of baked milk reactivity was 40%. Negative (<3mm²) muffin slurry tests were found in 8/30 children (27%) and were associated with non-reactivity to baked milk (p=0.01) with a sensitivity of 1 (0.70-1.00). Specificity was 0.41 (0.19-0.67).

Conclusions: The SPT to muffin slurry had high sensitivity for baked milk reactivity. This readily available, easily performed test may assist the selection of children appropriate for baked milk challenges in the office. Positive SPT to muffin slurry had low specificity and was insufficient to rule in baked milk reactivity.