In a recent systematic review and meta-analysis «Patra S, Nebhinani N, Viswanathan A ym. Atomoxetin...»1, (Patraet al. 2019 Atomoxetine for Attention Deficit Hyperactivity Disorder in Children and Adolescent with Autism: A Systematic Review and Meta-Analysis) efficacy and safety of atomoxetine (ATX) was compared with placebo in the treatment symptoms of ADHD in children and adolescents with autism spectrum disorder. Study inclusion criteria were as follows: randomized placebo-controlled trials comparing ATX with placebo, patient population of children and adolescents of≤18 years with diagnosis of ASD as per DSM5 or Pervasive Developmental Disorder (PDD) as per DSMIV or ICD 10, trials using atomoxetine at standard dose (0.5–1.8 mg/kg/day) and duration 4 weeks, and trials reporting ADHD symptoms as outcome measures.

The primary outcome measures were parent-rated symptoms of ADHD both short-term (≤6 months) and long-term (≥6 months), social behavior, and serious adverse events. ADHD symptoms rated on standard scales like SNAP, DSM-IV rating scale, and ADHD rating scale were included in the analysis. Serious adverse events included: any event of death, life threatening side effects, prolongation of hospital stay, or disability. Secondary outcomes were clinician and teacher-rated ADHD symptoms; overall improvement of ADHD symptoms rated on standard scales like CGI-I; parent stress; quality of life; and non-serious adverse events (nausea and vomiting, decreased appetite, and decreased sleep).

Altogether three RCTs, with 241 participants were included in the analysis. Girls and boys in the age range 5-17 years. The trial duration ranged from 6-10 weeks. One study «Handen BL, Aman MG, Arnold LE ym. Atomoxetine, Par...»2 (Handen et al. 2015) was a four-group study in which participants were randomized into ATX, Placebo, ATX + Parent Training, and Placebo + Parent Training groups. The first two groups on ATX (n= 32) and placebo (n= 32) were included in the analysis because aim of the study was to estimate the effect to ATX.

Only two trials «Handen BL, Aman MG, Arnold LE ym. Atomoxetine, Par...»2, «Arnold LE, Aman MG, Cook AM ym. Atomoxetine for hy...»3 (Arnold et al., 2006; Handen et al., 2015) involving 96 participants provided data on parent rated symptoms of ADHD (inattention and hyperactivity). Results showed beneficial effect of ATX as compared to placebo on parent-rated hyperactivity (SMD−0.73, 95%CI =−1.15 to−0.34, low quality evidence), and parent-rated inattention (SMD−0.53; 95% CI =−0.93 to−0.12, very low-quality evidence). There was no statistically significant improvement in parent-rated oppositional behavior or social behavior. The effects of ATX in clinician- and teacher-rated symptoms remained unclear because there was no statistically significant improvement in clinician-rated and teacher-rated ADHD symptoms. Long-term efficacy of atomoxetine could not be evaluated in the meta-analysis.

Risk of serious adverse events was increased in the atomoxetine group (RR 3, 95% CI0.32 to 27.76, 193 participants low quality evidence); but the increase is not statistically significant (one re-hospitalization due to aggressive behavior which was ascribed to lowering of antipsychotic dose and one hospitalization due to seizure caused by suboptimal antiepileptic concentrations).

As only Harfterkamp et al. «Harfterkamp M, van de Loo-Neus G, Minderaa RB ym. ...»4 had provided clinician-rated scores, pooling of data was not possible and hence, meta-analysis could not be done. None of the studies reported on social behavior or quality of life.

There was a beneficial effect of ATX on overall symptoms of ADHD rated CGI-I as compared to placebo (RR 2.37, 95% CI: 1.38, 4.06), but quality of evidence was low.

All the three trials reported significantly higher rates of non-serious adverse effects like nausea/vomiting, stomach pain, decreased appetite, and decreased sleep in ATX group. In addition, symptoms of fatigue, irritability, tiredness, mood swings, headache, and restlessness were also reported. As compared to placebo, ATX had a higher risk of non-serious side effects. Risk of nausea and vomiting was as follows: RR 1.91, 95% CI 1.24–2.94, 3 trials, 193 participants, I²= 0%, decreased sleep: RR 1.79, 95% CI 1.19–2.70, 3 trials, 193 participants, I²= 0%, and decreased appetite: RR 1.79, 95% CI 1.17 to 2.73, 3 trials, 193 participants, I²= 39%, with ATX; the quality of evidence for all of these outcomes were graded as low.

Risk of bias: Arnold et al. «Arnold LE, Aman MG, Cook AM ym. Atomoxetine for hy...»3 had to be judged as carrying a high risk of detection bias due to the crossover design of the study and lack of blinding of outcome assessment. Arnold et al. «Arnold LE, Aman MG, Cook AM ym. Atomoxetine for hy...»3 and Harfterkamp et al. «Harfterkamp M, van de Loo-Neus G, Minderaa RB ym. ...»4 were sponsored by pharmaceuticals manufacturing ATX.

• Study quality: Moderate
• Applicability: Good for short term treatment of children and adolescents in Finland
• Comment: In this study, atomoxetine was used off-label to treat ADHD symptoms without an ADHD diagnosis. In Finland Atomoxetine is indicated for the treatment of ADHD in children, adolescents and adults aged 6 years and over as part of a comprehensive treatment regimen. Treatment should be initiated by a physician experienced in the treatment of ADHD syndrome. ADHD diagnosis should be made based on the current DSM criteria or ICD classification.

General comments: In addition, Iffland et al. have published a systematic Cochrane review «Iffland M, Livingstone N, Jorgensen M, ym. Pharmac...»5 in 2023 that examined the impact of drug interventions on irritability, aggression and self-harming behaviour in people with autism spectrum disorder. The review compared the effect of placebo and ADHD medications on irritability, aggression and self-harming behaviour. The review did not compare the effect of placebo and ADHD drugs on ADHD symptoms and is therefore not discussed in detail here.