With the possible exception of fluoxetine (multiple brands), the vast majority of antidepressants are ineffective, and some may even be unsafe, for use in children and teens with major depressive disorder (MDD), new research shows.
"The only treatment that is evidence-based is fluoxetine," lead researcher Andrea Cipriani, PhD, associate professor, Department of Psychiatry, University of Oxford, United Kingdom, told Medscape Medical News.
However, Dr Cipriani stressed that this applies to children with moderate to severe depression for whom psychotherapy or other nonpharmacologic interventions have been tried without success or in situations in which such interventions are unavailable.
"And it doesn't mean that if I have a patient who is responding to escitalopram [Lexapro, Forest Laboratories, Inc], that I should stop escitalopram and put that patient on fluoxetine, because these are average data for an average patient."
Dr Cipriani also stressed that the use of antidepressants in children is "not cookbook medicine. Everything has to be individualized to the specific patient."
The findings were published online June 8 in the Lancet
One Drug Better Than Placebo
MDD is one of the most common mental disorders in children and adolescents, the investigators note. However, whether to use pharmacologic agents in this population and which drugs are optimal remains controversial.
To compare and rank antidepressants and placebo for the treatment of MDD in young patients, the researchers conducted a meta-analysis to identify both direct and indirect evidence from relevant trials.
They searched for double-blind, randomized controlled trials involving the use of antidepressants for the acute treatment of MDD in children and adolescents through May 31, 2015. The analysis included 34 trials in which 5260 patients were enrolled; 14 antidepressants were assessed for efficacy and tolerability.
The mean study sample size was 159 participants; the mean age of the patients was 13.6 years; and the median duration of acute treatment was 8 weeks. About two thirds of the trials (65%) were funded by pharmaceutical companies.
In terms of study quality, 29% of the trials were rated as having a high risk for bias, 59% as having a moderate risk, and 12% as having a low risk.
The primary outcome was mean overall change in depressive symptoms and the proportion of patients who discontinued treatment due to any adverse events. To assess change in depressive symptoms, the investigators extracted a score from scales used in the studies. These scales included the Children's Depression Rating Scale Revised, the Beck Depression Inventory, and the Children's Depression Inventory.
Secondary outcomes included response rate, all-cause discontinuation, and suicidal behavior and ideation. Response rate was determined on the basis of the proportion of patients who achieved a reduction of 50% or more in depressive symptoms or whose scores on the Clinical Global Impression scale were much improved or very much improved.
The researchers performed a "network" meta-analysis that allowed them to use indirect as well as direct evidence from the relevant trials to calculate comparisons and to rank probabilities for all treatments.
The analysis showed that in terms of efficacy, only fluoxetine was better than placebo (standardized mean difference [SMD], - 0.51; 95% credible interval [CrI], -0.99 to -0.03).
However, the authors point out that "the large credible interval and its upper limit close to the point of no difference raise the question of whether this estimate is robust enough to inform clinical practice."
Nortriptyline (multiple brands) was significantly less effective than seven other antidepressants and placebo.
In terms of tolerability, fluoxetine was significantly better than duloxetine (Cymbalta, Eli Lilly and Company) (odds ratio [OR], 0.13; 95% CrI, 0.13 - 0.95) and imipramine (multiple brands) (OR, 0.23; 95% CrI, 0.04 - 0.78). Citalopram and paroxetine (multiple brands) were significantly better tolerated than imipramine alone (OR, 0.27 and 0.22, respectively). Imipramine was significantly less well tolerated than placebo, as was venlafaxine (multiple brands) and duloxetine.
For the most part, the results for secondary outcomes "were not materially different from, and lent support to, the findings for primary outcomes," the authors write.
Although the current study could not comprehensively assess the risk for suicidality for all drugs, there was robust evidence suggesting a significantly increased risk for suicidality for young people given venlafaxine.
The ranking of treatments, based on cumulative probability plots and surface under the cumulative ranking curve, showed that the most effective treatment was fluoxetine (76.6%) and that the least effective was nortriptyline (3.7%). Fluoxetine also came out on top in terms of tolerability (75.7%), with imipramine coming in last (13.1%).
Advantages of Fluoxetine
Fluoxetine has a number of advantages, said Dr Cipriani. Being the first new-generation antidepressant (selective serotonin reuptake inhibitor), it has been "very well studied" and has many years of "real-world use."
It has a long half-life, requiring a few weeks to metabolize. Because of this, missing a dose may not be a problem.
"With children, there is sometimes a problem with compliance, and if you have a drug with a very short half-life, and you don't take it regularly, you might have withdrawal symptoms," said Dr Cipriani.
Fluoxetine is also widely available. "It's one of the WHO [World Health Organization] essential medicines, so it's available everywhere, including developing countries and where there's a refugee crisis," said Dr Cipriani.
But it is not without disadvantages. For example, switching from fluoxetine to another medication requires a washout period, and fluoxetine can interact with other medications.
"You need to be careful if you are adding other treatments, such as antibiotics," said Dr Cipriani.
The authors caution about clinical interpretation of the findings, owing to the uncertainty of estimates and potential bias due to selective reporting. As well, unpublished studies were not included in the analysis, and published reports may overestimate the efficacy of treatments.
In 2003, the US Food and Drug Administration (FDA) and other regulatory agencies added a black box warning to the labeling of prescription antidepressants, highlighting the possibility of suicidality in some children with MDD.
According to Dr Cipriani, there is an increased risk for suicidality, but the risk is not the same for all drugs.
Although the data gathered by the FDA are not powered to provide "clear answers," there is "a trend" of some drugs being worse than others in terms of suicidality risk, said Dr Cipriani. "The one-size-fits-all approach is misleading." The evidence suggests that sertraline and fluoxetine are not associated with an increased risk for suicidality, she added.
Children Not Small Adults
What is becoming clear from the research, said Dr Cipriani, is that children with MDD significantly differ from adults with depression. In children, the brain is still developing, and the clinical features of depression in children are different from those in adults.
"In adults, mainly in the elderly, depression is a lot about mood, while in young people, it's a lot about irritability and difficulty with concentration," said Dr Cipriani. "We tend to call it all depression and assume that the same drugs work for young people."
A main problem is lack of knowledge about the pathophysiology of the disorder and identifying markers. "We don't have hard outcomes to test a treatment."
A question plaguing trials of antidepressants ― among adults as well as children ― is the high placebo response rate. According to Dr Cipriani, 50% to 60% of patients respond to an antidepressant, and about 40% respond to placebo, so the "added value" is about 10% to 15%.
The placebo response has increased over the years ― it was about 20% in the 1980s ― probably because of methodologic differences in trials. One example would be differences regarding the inclusion of less severe patients, who tend to respond more to placebo.
"But in real practice, we don't use a placebo," said Dr Cipriani.
Childhood depression is "a huge problem" that is growing, according to Dr Cipriani. About 3% of children younger than 12 years suffer major depression, as do 6% of persons aged 13 to 18 years.
Adverse Events Underreported
In an accompanying editorial, Jon Jureidini, MD, a child psychiatrist at the University of Adelaide and Women's and Children's Hospital, in Australia, writes that the evidence for the use of antidepressants in children is even weaker than the current study suggests.
"The data that the authors were working from was what were available in published articles and clinical study reports," Dr Jureidini told Medscape Medical News. "Unless you have access to individual patient-level data, you can't be confident that what's in those tables and spreadsheets that these authors had to work from is actually an accurate representation of what was found."
From his own research, Dr Jureidini has learned that "adverse events are underreported, and some outcome measures are distorted."
One problem is that some of the clinical trials are carried out by pharmaceutical companies, which "control the flow of information," he said.
Anther problem is that doctors are often convinced that these medicines do work, he added.
"They don't do it deliberately, but they look at things with more of a positive lens, because they want to show that these drugs they regard as positive and valuable are in fact positive and valuable."
What is particularly bothersome, said Dr Jureidini, is that even though these drugs are not first-line therapy, "the idea seems to be that you give the antidepressants anyway," he said. "That's just bad medicine; if they're not good drugs, then you should avoid giving them."
The use of an antidepressant in a child should usually be done only in the inpatient setting, said Dr Jureidini.
He himself has not initiated antidepressant therapy in a child more than once or twice in the past 5 years. However, it is not uncommon for him assume care for a child who is already receiving an antidepressant, "and I wouldn't necessarily stop that."
In the absence of available psychotherapy, there are other nondrug approaches clinicians can take to treat a child with depressive symptoms.
One is to look for a better explanation for the symptoms than to just call it major depression. "Very often, symptoms are caused by life circumstances, and sometimes those life circumstances can be addressed in a way that resolves the symptoms."
In situations in which there is no apparent explanation for the symptoms or in which nothing can be done about them directly, Dr Jureidini suggests taking a "watchful waiting" approach, even in cases of moderate and severe depression.
It might be a matter of helping the child make sense of what they are feeling. For example, he or she might still be grieving from the loss of a beloved grandmother.
Dr Jureidini pointed out that depression usually lasts weeks, not months.
Does he think the use of antidepressants in children will disiminish significantly in response to this new study? "I hope so, but probably not."
He sees two factors standing in the way of getting the message across ― pharmaceutical marketing, and key opinion leaders.
"In any community, particularly in North America, you find senior, often academic child psychiatrists and others who are enthusiastic prescribers of antidepressants and who will influence the prescribing pattern of many other doctors."
He anticipates that "prominent figures" will "stand up and say either publicly or within their academic communities that this paper is misleading, that antidepressants save lives, and that we should all continue prescribing them."
Dr Cipriani has received personal fees from Accord Healthcare as an expert witness for a patent issue about quetiapine extended release. In the past 3 years, the University of Adelaide has been paid by Baum, Hedlund, Aristei and Goldman, Los Angeles, for Dr Jureidini to provide expert analysis and opinion about documents obtained from Forest Pharmaceuticals in relation to pediatric antidepressant randomized controlled trials.
Lancet. Published online June 8, 2016. Abstract, Editorial