Risk factors for discontinuing oral immunotherapy in children with persistent cow milk allergy

BACKGROUND: There are no universally accepted criteria for discontinuing milk oral immunotherapy (MOIT) in patients with persistent cow milk allergy (CMA) and little data are available on predictive risk factors for dropping out from oral immunotherapy (OIT), due to allergic reactions or other reasons. METHODS: We retrospectively reviewed clinical records of patients with persistent severe CMA undergoing MOIT in a tertiary care center hospital to investigate risk factors associated with discontinuation of OIT. Persistent and severe allergy was defined as the history of systemic reactions and any milk protein‐specific IgE level >85 kU/ml. All patients were first admitted for an in‐hospital rush phase eventually followed by an at‐home dose increase. We evaluated the effect of various factors on two primary outcomes: the highest dose of milk ingested during the in‐hospital rush phase and during the home OIT phase. RESULTS: We identified 391 patients, of whom 131 met the inclusion criteria for the retrospective study, 54 females and 77 males. Data of the home OIT phase were available for 104 patients (27%). Regarding the home OIT outcome, an association for having a cow milk avoiding diet was found with reaching a dose below 10 ml during the in‐hospital rush phase (relative risks [RR]: 2.33, confidence interval [CI]: 0.85; 6.42), an age above than 10 years from the time of admission (RR: 3.29, CI: 0.85; 12.73), and a higher total number of reactions occurred during the hospitalization (RR: 1.54, CI: 1.02; 2.32), whereas the presence of respiratory reactions with wheezing (RR: 1.93, CI: 0.49; 7.61) and an IM adrenaline use was related to a higher risk of having an OIT still in progress (RR: 5.47, CI: 0.33; 7.73). CONCLUSIONS: In this cohort of children with persistent CMA undergoing OIT who presented with respiratory reactions with wheezing, the development of anaphylaxis with the need for IM adrenaline, and age above 10 years were predictors of poor long‐term outcome.