The safety and efficacy of sublingual and oral immunotherapy for milk allergy

Abstract

Background: Oral immunotherapy (OIT) and sublingual immunotherapy (SLIT) are potential therapies for food allergy, but the optimal method of administration, mechanism of action, and duration of response remain unknown.

Objective: We sought to explore the safety and efficacy of OIT and SLIT for the treatment of cow's milk (CM) allergy.

Methods: We randomized children with CM allergy to SLIT alone or SLIT followed by OIT. After screening double-blind, placebo-controlled food challenges and initial SLIT escalation, subjects either continued SLIT escalation to 7 mg daily or began OIT to either 1000 mg (the OITB group) or 2000 mg (the OITA group) of milk protein. They were challenged with 8 g of milk protein after 12 and 60 weeks of maintenance. If they passed the 60-week challenge, therapy was withdrawn, with challenges repeated 1 and 6 weeks later. Mechanistic correlates included end point titration skin prick testing and measurement of CM-specific IgE and IgG(4) levels, basophil histamine release, constitutive CD63 expression, CD203c expression, and intracellular spleen tyrosine kinase levels.

Results: Thirty subjects with CM allergy aged 6 to 17 years were enrolled. After therapy, 1 of 10 subjects in the SLIT group, 6 of 10 subjects in the SLIT/OITB group, and 8 of 10 subjects in the OITA group passed the 8-g challenge (P = .002, SLIT vs OIT). After avoidance, 6 of 15 subjects (3 of 6 subjects in the OITB group and 3 of 8 subjects in the OITA group) regained reactivity, 2 after only 1 week. Although the overall reaction rate was similar, systemic reactions were more common during OIT than during SLIT. By the end of therapy, titrated CM skin prick test results and CD63 and CD203c expression decreased and CM-specific IgG(4) levels increased in all groups, whereas CM-specific IgE and spontaneous histamine release values decreased in only the OIT group.

Conclusion: OIT was more efficacious for desensitization to CM than SLIT alone but was accompanied by more systemic side effects. Clinical desensitization was lost in some cases within 1 week off therapy.

FIG 1

Study timeline with key features highlighted. Subjects were randomized at T2 to either SLIT (goal dose, 7 mg), OITB (goal dose, 1000 mg), or OITA (goal dose, 2000 mg). CM-specific IgE, CM-specific IgG₄, and CM-titrated skin prick test (SPT) results were obtained in all subjects at T1, T3 to T5, and T7. Basophil studies were done at T1, T2 to T5, and T7.

FIG 2

Food challenge outcome. The CM protein threshold is shown for baseline (T1), after 12 weeks of maintenance (T4), after 60 weeks of maintenance (T5), and 1 week (T6) and 6 weeks (T7) off therapy. A, SLIT/SLIT group. B, SLIT/OITB group. C, SLIT/OITC groups. Bars represent medians. *P < .05 and **P < .01 compared with T1.

FIG 3

Mechanistic changes. Individual line plots showing changes with treatment for the 3 randomization groups (A, SLIT/SLIT group; B, SLIT/OITB group; and C, SLIT/OITA group) in CM-specific IgE level (1), CM-specific IgG₄ levels (2), and end point titration skin prick test responses. Shown is the average wheal size over 5 CM dilution. Bars represent medians. The shaded blue area represents the period of withdrawal of therapy. *P < .05 and **P < .01 compared with T1.

FIG 4

Basophil activity. A, HR in media alone (SHR) or after stimulation with CM extract, goat anti-human IgE antibody (IgE), peanut extract (PN), or N-formylmethionine (f-met) at T1 to T5 and T7. B, Constitutive surface expression of CD63 and CD203c and intracellular Syk by basophils at T1 to T5 and T7. A box-and-whisker plot is shown; outliers are not shown. *P < .05 and **P < .01 compared with T1.