DISCUSSION
We established that coldU in children has a lower resolution rate than CSU. Further, among children with coldU, 17.3% of patients experienced coldA, predominantly cases with elevated eosinophils.
Similar to prior studies, we found that pediatric coldU has no sex predilection12,13, unlike in adults where females make up the majority of cases6,14. The median age of onset of symptoms in our population (9.5 years) was consistent with other reports in children (7-8.5 years)7,12,13.
We found that approximately half of patients had concomitant CU of other types, predominantly CSU, and a minority had cholinergic urticaria. A large retrospective study of coldU in children reported that only 14.2% and 3.6% of participants had concomitant CSU and cholinergic urticaria, respectively12. The elevated proportion of patients with concomitant CSU (42.3%) and cholinergic urticaria (9.6%) in our study may be because patients were assessed for concomitant types of CU at study entry. In line with our results, a study of both adults and children study reported that 8% of patients with cold urticaria had co-existent cholinergic urticaria15.
Other studies suggest that co-existing atopic conditions may be more common. Yee et al. report that among children with coldU, 59.5% had concomitant allergic rhinitis, 46.8% had asthma, and 33.4% had food allergies12. However, the prevalence of atopic dermatitis among children with coldU was slightly lower than in the present study (24.6% versus 30.8%)12. These differences may reflect geographic or ethnic differences between the studied populations16,17.
Relative to CSU, we found an increased proportion of coldU patients with IgE/Th2 mediated diseases (i.e., asthma and atopic dermatitis) and elevated total IgE. Consistent with these findings, the majority of CSU cases are associated with an autoimmune etiology consisting of auto-IgG or auto-IgE antibodies18,19, whereas coldU is predominantly an IgE mediated disease3. A Th2/IgE etiology underlying coldU indicates the potential therapeutic utility of IgE (i.e., omalizumab) and Th2 (i.e., dupilumab) inhibitors. Omalizumab efficacy in treating coldU refractory to antihistamines has been established in adults20, but not yet in children. A case report has documented the resolution of coldU and concomitant atopic dermatitis in an adult taking dupilumab21, and a clinical trial is underway to determine its efficacy in adult coldU (NCT04681729). Dupilumab is already licenced for the treatment of atopic dermatitis in children; therefore, it may be a treatment option in children with both atopic dermatitis and coldU. Further studies are necessary to elucidate the role of IgE and the Th2 response in coldU.
Children in our cohort were predominantly managed with sgAHs, of which 56.8% of patients had well-controlled symptoms. Concomitant CSU was associated with poorly controlled disease on sgAHs. No clinical trials have assessed the efficacy of sgAHs in pediatric coldU22. However, several studies have found that sgAHs significantly diminish symptoms and decrease the minimum temperature at which symptoms develop in adults22-26. Further studies are required to establish the efficacy of sgAHs in the treatment of pediatric coldU, and to determine the usefulness of concomitant CSU as a predictor of sgAH response.
In this study, 17.3% of children had a history of coldA. This is in line with a recent meta-analysis conducted by our team that determined that the prevalence of anaphylaxis among cases of adult and pediatric coldU was 21.5% (95%CI: 15.8%, 28.5%; I2=69%)4. Further, we identified that elevated blood eosinophils were associated with coldA. These results suggest that blood eosinophils may be used identify patients at risk for coldA, which can inform epinephrine autoinjector prescription and counselling of patients to avoid common triggers of coldA. A recent case report recorded the successful treatment of coldU and concomitant CSU with reslizumab, an anti-IL-5 monoclonal antibody, which downregulates eosinophil activation and recruitment27. However, little is known on the role of eosinophils in the pathogenesis of coldU/coldA.
The resolution of coldU in our cohort was fewer than 5 per 100 patient-years, almost half the resolution rate of CSU10. The resolution rate observed was similar to other studies that included both pediatric and adult patients (17.9-26.6% of patients resolved in five years)6,28. Because of insufficient sample size, we were unable to assess factors associated with disease resolution.
Our study has some potential limitations. This study had a relatively small sample size thereby limiting the power of subgroup analyses of participants with coldA and disease resolution. Our study was conducted at a tertiary urticaria referral center, which likely evaluates more severe cases of coldU that are refractory to antihistamines/involve anaphylaxis. However, we also recruited children from clinics outside the hospital and, hence, we believe that this study maintains external validity. Additionally, the wait time to see a specialist at our center is approximately four to eight months; therefore, our results may underestimate disease resolution because more self-limited cases were not recruited. Although we assessed all children with the UCT, it has not been validated for children under 12 years of age. However, the UCT is reported to have high validity and reliability in adults29 and preliminary work by our group supports its validity in children. Finally, we were unable to obtain follow-up data on four (7.7%) of patients and certain lab values were not available in approximately a quarter of patients. However, given that data was available for most patients we do not expect these missing values to bias our results.
Taken together, our findings demonstrated a low resolution rate of pediatric coldU versus CSU. Elevated eosinophils were associated with anaphylaxis, which occurred in approximately one sixth of patients. Future multicentre studies are necessary to determine factors associated with coldU resolution.