Discussion
In this study, we investigated the natural history of IgE-mediated immediate-type CMA in children aged 6–12 years. Forty-six of the 123 (37%) children with persisting CMA until the age of 6 years acquired CM tolerance by the age of 12 years. Furthermore, we demonstrated that higher CM-sIgE levels, previous anaphylaxis to CM, and complete CM elimination from the diet were independent risk factors for persistent CMA.
It has been reported that most children with CMA acquire tolerance during the pre-school age.1, 8-12, 18-20 Half of Korean children with CMA develop tolerance by the age of 8.7 years.19 A Danish birth-cohort study showed that the tolerance acquisition rates in children with CMA, including non-IgE-mediated CMA, were 92% and 97% at 5 and 15 years of age, respectively.20 However, all of these studies initiated follow-up in infancy, making it difficult to clarify the tolerance acquisition rate in children with CMA persisting beyond school age. Few studies on the natural history of CMA have focused on children beyond school age. Our data demonstrated that the tolerance acquisition rate in children with CMA between the ages of 6 and 12 years was 37%. This rate is obviously higher than that for peanut allergy; however, it seems to be lower than that for hen’s egg allergy (HEA).21, 22 A previous study reported that the rate of tolerance acquisition for peanut allergy in children aged 4–20 years was 21.5%.21 Our institution reported a rate of tolerance acquisition for HEA of 60.5%.22 However, children with OIT were defined as dropouts in this previous study; therefore, their rate may be overestimated. When we defined children with OIT as dropouts, the tolerance acquisition rate was 58% (Supplementary figure 1), which was similar to the results of the previous HEA study.22 When we defined tolerance acquisition as achieving short-term unresponsiveness to 200 mL of unheated CM for children who underwent OIT after the age of 6 years, the tolerance acquisition rate was 50% (Supplementary figure 1); this is lower than the rate of 65.2% reported in the HEA study.22 This difference would result from the poor outcome of OIT to CMA compared to HEA.22 These findings indicate that the “real” tolerance acquisition rate in children aged 6–12 years with CMA might be between 37% and 50%. In many previous studies, CMA persistence was not confirmed in annual OFCs.8, 11, 12, 19 In our institution, children usually undergo blood tests and OFCs annually, providing a more accurate assessment of CMA tolerance.
In the present study, 88% of patients with CMA in the absence of the identified risk factors (higher CM-sIgE levels, previous anaphylaxis to CM, and complete CM elimination) acquired tolerance by the age of 12 years; therefore, these children should be proactively assessed by OFCs for their tolerance acquisition. In contrast, none of the children with CMA presenting all three risk factors acquired tolerance by the age of 12 years (Supplementary figure 3); therefore, these children should be considered for OIT. Previous anaphylaxis and higher CM-sIgE levels were common risk factors identified in the present and previous studies, and both factors contributed to difficulties in acquiring tolerance (Figure 3A and 3C). This is the first report showing that complete CM elimination from the diet is associated with CMA persistence. The immunomodulatory effects of consuming baked milk proteins may accelerate the progression to tolerance of non-baked milk.23 In addition, continuous consumption of small amounts of heated milk has been suggested to increase the safely ingestible amount of CM.24 These results suggest that continuous small-dose intakes of CM may induce immunological benefits for acquiring CM tolerance.
In the present study, CM-sIgE levels in the persistent group were significantly higher than those in the tolerant group at all ages from 6–12 years. Interestingly, CM-sIgE levels decreased annually, not only in the tolerant group, but also in the persistent group. In previous studies on the natural history of CMA, the CM-sIgE level in children with persistent allergy was reported to increase over the first 3–4 years of life, followed by a plateau, and then a gradual decrease until the age of 18 years.9, 10 These findings indicate that decreasing CM-sIgE levels during school age may not be associated with CM tolerance acquisition.
This study has some limitations. First, we defined children undergoing OIT as having persistent CMA, not as dropouts. In children who received OIT, 95% of the children eliminated CM completely, two-thirds had previous anaphylaxis, and the median CM-sIgE level was 58.5 kUA/L at the age of 6; these characteristics were higher than those in the persistent group who did not receive OIT (Supplementary table 1). Since children with OIT should have difficulty acquiring tolerance, we considered that children with OIT should be defined as having persistent CMA rather than as dropouts. Second, this study had a single-center design. Our facility is a tertiary center for food allergies, and children with food allergies tend to be more severely affected than the general population, which may have influenced the study outcomes.
In conclusion, at least one-third of patients who still had IgE-mediated CMA at the age of 6 years acquired tolerance by the age of 12 years. We demonstrated that children with CMA whose CM-sIgE levels were higher than the optimal cutoff value, who previously had anaphylaxis to CM, and who eliminated CM from their diet at the age of 6 years have difficulty acquiring tolerance by the age of 12 years. This study provides important insights into the natural history of CMA persisting into school age and identifies risk factors associated with persistent CMA.