Background: Desloratadine (DL) is a non-sedating antihistamine first marketed in 2001. Seizure has been reported in children using DL. We conducted a post-authorization safety study in individuals with no history of seizure.
Objectives: To assess the association between DL use and incidence of non-febrile seizure. Methods: We conducted a Nordic registry-based cohort study among individuals with a first prescription redemption of DL in 2001-2015. Individual-level data from health registries in Denmark, Finland, and Sweden were used to derive a common database. DL exposure was a time-varying variable defined for each prescription redemption as days’ supply plus a 4-week grace period. Unexposed periods (remote prior exposure) were defined as starting 26 weeks after date of prescription redemption. The outcome was incident non-febrile seizure. The association between DL exposure and incidence rate (IR) of seizure was analyzed using Poisson regression of number of seizures as outcome and logarithmic transformation of follow-up time as the offset. Potential confounders were identified using causal diagrams and included country, calendar year, sex, age, seasonality, and history of asthma, severe rhinitis and chronic urticaria. Results are presented as IR per 100,000 person-years (PY) and incidence rate ratio with corresponding 95% confidence interval (IRR, 95% CI).
Results: The study included 1,561,410 incident DL users of all ages of which 2,297 individuals were diagnosed with incident seizure in the primary analysis. The IRs of seizure were 39.9 and 30.1 per 100,000 PY for currently DL exposed and unexposed periods. A significant association between current DL exposure and incidence of non-febrile seizure was seen (IRR 1.32 95% CI 1.19; 1.48). The adjusted association between being currently DL exposed and incidence of non-febrile seizure showed an increase of 17% in the IR of non-febrile seizure during the exposed periods (IRR 1.17 95% CI 1.05; 1.31). The association was strongest among children 0-4 years of age (IRR=1.47; 95% CI: 1.16; 1.87).
Conclusion: A 17% higher IR of non-febrile seizure was seen in individuals comparing current DL exposed and unexposed periods. The finding is consistent with pharmacovigilance data. DL should be administered with caution in patients with personal or family history of seizures.
