Selective serotonin reuptake inhibitors (SSRIs) can't be recommended for treating autism in children or adults at this point, according to a Cochrane review.
The antidepressants have shown emerging evidence of harm without any benefit in pediatric cases, reported Katrina Williams, PhD, of the University of New South Wales and Sydney Children's Hospital in Sydney, Australia, and colleagues in the Cochrane Library.
In adult cases, two small studies have shown improvements in overall symptoms and obsessive-compulsive behavior along with aggression in one and anxiety in the other.
Action Points
- Note that the authors conclude that off-label use of SSRIs cannot be recommended for adults due to small positive effects and small sample size of the trials.
- Note that the authors state that due to emerging evidence that SSRIs are not effective and can cause harm in children, they cannot be recommended as a treatment for children with autism at this time.
However, this limited evidence in adults is not yet sufficiently robust for clinicians to rely on, Williams' group cautioned.
"Decisions about the use of SSRIs for established clinical indications that may co-occur with autism, such as obsessive-compulsive disorder and depression in adults or children, and anxiety in adults, should be made on a case-by-case basis," they recommended in the paper.
Physicians need to be explicit with parents and patients about the limited evidence, risks of SSRI treatment, and other options, Williams' group added.
Antidepressants have become one of the most commonly prescribed psychotropic medications for autism spectrum disorders and the class for which prescribing has risen most, Williams' group noted.
The researchers systematically searched for randomized controlled trials that compared any dose of oral SSRI with placebo in autism spectrum disorders.
They found five studies that included only children and two that included only adults, for a total of 271 participants. Aside from the one study done in France and one done in Japan, the trials were conducted in the U.S.
The drugs used in these seven trials were fluoxetine (Prozac) in two, fluvoxamine (Luvox) in two, fenfluramine (Pondimin, removed from the U.S. market in 1997) in two, and citalopram (Celexa) in one.
Duration of treatment ranged from five to 12 weeks. Inclusion criteria also varied based on diagnostic criteria and the intelligence level of the participants.
Because each study used different tools to measure outcomes or evaluated different components of the outcomes, Williams and colleagues determined that meta-analysis wouldn't work.
In children, findings included:
- With citalopram, no significant effect on any component of the parent-rated Repetitive Behavior Scale-Revised (P>0.36 for all), on the composite of clinical global impression and obsessive-compulsive behavior (P=0.28), proportion of responders on the clinical global impression (P=0.99), or obsessive and compulsive score alone (P=0.85)
- With fenfluramine, no significant change from baseline or compared with placebo on core features of autism on the Behavior Summarized Evaluation
- With fluvoxamine, no improvement in core features of autism, although one study suggested that 10 of 18 patients "responded" to the drug
- With fluoxetine, a trend for a reduction in the composite of clinical global impression and repetitive behavior compared with placebo (P=0.056), but no significant benefit for clinical global impression alone or compulsion score alone
In adults, both studies were small. Thus, one study of fluoxetine showed improvement in behavior change on the clinical global impression but based on benefit in three of only six patients.
Fluvoxamine in adults yielded improvement in clinical global impression in 53% of participants compared with placebo's 0%.
With regard to other non-core features of autism in adult cases, fluoxetine reduced obsessions (P=0.03) and anxiety (P=0.03) with trends favoring composite obsession and compulsion (P=0.06) and depression (P=0.06).
Fluvoxamine in adults improved both obsession and compulsion scores (P<0.02 and P<0.001, respectively, at 12 weeks) and aggression (P<0.03).
Adverse events in children were significantly elevated with citalopram and fluvoxamine (in one of three trials). Citalopram was also linked to a severe case of seizures that persisted after withdrawal from the drug.
Fenfluramine in children was associated with weight loss overall, two cases of increased withdrawal and sadness, and two cases of increased stereotypies.
Adverse events in adults were not reported in the small fluoxetine study, while fluvoxamine appeared well tolerated.
The reviewers cautioned about the lack of medium- and longer-term follow-up and the absence of trials on the other SSRIs commonly used in clinical practice, particularly sertraline (Zoloft).
Disclosures
The review was supported by the Children's Hospital at Westmead and by Financial Markets Foundation for Children and the Australian Department of Health and Aging.
Williams reported no conflicts of interest to disclose.
A co-author reported having consulted for Eli Lilly and Janssen; having had research grants from Eli Lilly and Celltech; being on the advisory board for Eli Lilly, Janssen, Novartis, and Shire; and having given presentations for Eli Lilly, Pfizer, Janssen, and Sanofi.
Primary Source
Cochrane Database of Systematic Reviews
Williams K, et al "Selective serotonin reuptake inhibitors (SSRIs) for autism spectrum disorders" Cochrane Database of Syst Rev 2010; 8: CD004677.