The Safety of Atomoxetine for the Treatment of Children and Adolescents with Attention-Deficit/Hyperactivity Disorder: A Comprehensive Review of Over a Decade of Research

Atomoxetine is a noradrenergic reuptake inhibitor prescribed for attention-deficit/hyperactivity disorder (ADHD) that first gained approval in the USA in 2002 and has been authorized in 97 countries worldwide. The aim of this paper is to comprehensively review publications that addressed one or more of seven major safety topics relevant to atomoxetine treatment of children and adolescents (aged >/=6 years) diagnosed with ADHD. While the review focuses on children and adolescents, publications in which data from patients aged >18 years and from 6 to 18 years were analyzed in the same dataset were included. Using a predefined search strategy, including agreement of two reviewers when selecting papers, reduced the potential for bias. Using this process, we identified 70 eligible papers (clinical trials, epidemiological studies, and case reports) across the seven topics. We also referred to the European Summary of Product Characteristics (SPC) and US label. We found 15 papers about suicidality, three about aggression/hostility, seven about psychosis/mania, six about seizures, seven about hepatic effects, 29 about cardiovascular effects, and 28 about growth and development. The main findings (i.e., those from the largest and most well-conducted studies/analyses) are as follows. A large register-based study of pediatric and adult patients (6818 received atomoxetine) calculated a hazard ratio of 0.96 for suicide-related events during treatment with atomoxetine, and a meta-analysis of 23 placebo-controlled studies (N = 3883), published in 2014, found no completed suicides and no statistically significant association between atomoxetine and suicidality. The frequency of aggression/hostility was not statistically significantly higher with atomoxetine, e.g., experienced by 1.6 % (N = 21/1308) of atomoxetine-treated patients versus 1.1 % (N = 9/806) of placebo-treated patients in one meta-analysis. Symptoms of psychosis and mania were mainly observed in patients with comorbid bipolar disorder/depression. Based on spontaneous reports, during a 2-year period when 2.233 million adult and pediatric patients were exposed to atomoxetine, the reporting rate for seizures was 8 per 100,000 patients. In the manufacturer's database, atomoxetine was a "probable cause" of three hepatic adverse events (AEs) (all reversible hepatitis), and 133 hepatic AEs had possible confounding factors and were "possibly related" to atomoxetine, during 4 years when atomoxetine exposure had reached about 4.3 million patients. Rare cases of severe liver injury are described in the US label and European SPC; a case requiring liver transplantation is described in the US label. In a comprehensive review of a clinical trials database (N = 8417 received atomoxetine), most pediatric patients experienced modest increases in heart rate and blood pressure, and 8-12 % experienced more pronounced changes (>/=20 bpm, >/=15 to 20 mmHg). However, in three long-term analyses (>/=2 years), blood pressure was within age norms, and few patients discontinued due to cardiovascular AEs. As described in the European SPC, QT interval prolongation is uncommon, e.g., in an open-label study, 1.4 % of 711 children and adolescents had prolonged QTc intervals (>/=450 ms in males, >/=470 ms in females) that were not clinically significant at >/=3 years of treatment with atomoxetine. The European SPC warns about potential QT interval prolongation in patients with a personal or family history, or if atomoxetine is administered with other drugs that potentially affect the QT interval. Decreases in growth (weight and height gain) occurred and were greatest in patients of above average weight and height, but appeared to recover over 2-5 years of atomoxetine treatment. In conclusion, suicidality, aggression/hostility, psychosis, seizures, liver injuries, and prolonged QT interval are uncommon or rare in children and adolescents treated with atomoxetine, based on data from the predefined search and from the European SPC. Overall, the data that we assessed from our search do not suggest that associations exist between atomoxetine and suicidality or seizures. The data also suggest that an association may not exist between atomoxetine and aggression/hostility. While atomoxetine may affect the cardiovascular system, the data suggest these effects are not clinically significant in most patients. Reductions in growth appear to be reversible in the long term