GME Logo

User login

User menu

Issue 72, Apr 2018
Share This
Do antidepressants work? If so, which ones are better?

Cipriani et al. Comparative efficacy and acceptability of 21 antidepressant drugs for the acute treatment of adults with major depressive disorder: a systematic review and network meta-analysis. Lancet. 2018 Feb 20. pii: S0140-6736(17)32802-7. [Epub ahead of print] PubMed PMID: 29477251.


The findings of this paper have received a lot of attention in the press and on medical websites.

This meta-analysis is an update and expansion of the authors’ previous meta-analysis that aimed to compare and rank antidepressants for the acute treatment of adults with major depressive disorder.


A very thorough search of the literature was done to identify all relevant double-blind, randomized controlled clinical trials of antidepressants.

Both published and unpublished clinical trials were identified.

These clinical trials compared an antidepressant to either a placebo or to another antidepressant.

The trials evaluated the efficacy of a total of 21 different antidepressants.

Trials that included a significant number of patients with bipolar depression, psychotic depression, serious concomitant medical illness, or so-called “treatment-resistant depression” were excluded.

In addition to direct comparisons of one antidepressant to another (called “head-to-head” studies), in order to compare antidepressants that had not been directly compared to each other, a methodology called a network meta-analysis was used.

This study looked at the:

1. Efficacy of the antidepressants (percentages of participants who responded to the treatment), and

2. Acceptability of the antidepressants (percentages of participants who discontinued treatment for any reason).


This study included 522 clinical trials in which a total of about 116,000 persons with major depressive disorder participated.

All studies considered

It found that all 21 antidepressants studied were more effective than placebo.

Before reviewing the main findings regarding discontinuation of treatment, let’s remember that the rates for discontinuation of treatment are affected by many factors including adverse events and lack of efficacy. Here are the three main findings regarding discontinuation:

1. For almost all antidepressants, the rates of discontinuation were not statistically significantly different from placebo.

2. Only two antidepressants (agomelatine and fluoxetine) were associated with a statistically significantly lower discontinuation rate than placebo.

3. Only one antidepressant—clomipramine—was associated with a statistically significantly greater discontinuation rate than placebo.

Only head-to-head studies considered

The paper reported that when only head-to-head studies were considered, “agomelatine, amitriptyline, escitalopram, mirtazapine, paroxetine, venlafaxine, and vortioxetine were more effective than other antidepressants.”

It also reported that, “fluoxetine, fluvoxamine, reboxetine, and trazodone were the least efficacious drugs.”

For discontinuation rates, it reported that agomelatine, citalopram, escitalopram,

fluoxetine, sertraline, and vortioxetine had lower rates of discontinuation than other antidepressants.

And, that amitriptyline, clomipramine, duloxetine, fluvoxamine, reboxetine, trazodone, and venlafaxine had the highest rates of discontinuation.


The authors concluded that:

-All 21 antidepressants in included in this meta-analysis were more efficacious than placebo for the treatment of major depressive disorder in adults.

- There were fewer differences between the antidepressants when comparing them indirectly based on separate trials.

- But the antidepressants varied in their efficacy and efficacy and acceptability in the head-to-head trials, i.e., when one antidepressant was compared to another in the same trial.

Clinical Commentary

The questions of which antidepressants are more efficacious or better tolerated than others are of incredible importance given the high prevalence of major depressive disorder and the very high number of people who take an antidepressant for the treatment of “depression.”

In my opinion, indirect comparisons between antidepressants are not useful because the diagnostic practices, types of patients, characteristics of the illness, etc. vary too much. For example, how can we compare one study done in the US in 1990 with patients recruited through advertisements with another one done with patients coming to a specialty clinic at a European academic medical center 25 years later?

Even among the direct, head-to-head comparisons, I have concerns:

1. Many antidepressants were found to be more efficacious than reboxetine or trazodone. But these comparisons are of little clinical applicability given that reboxetine and trazodone are not the among the more commonly used antidepressants.

2. Many differences between antidepressants are highlighted by the paper even though they were not statistically significant.

So, what head-to-head comparisons were statistically significant?

1. Amitriptyline, escitalopram, mirtazapine, paroxetine, and venlafaxine were all statistically significantly more efficacious than fluoxetine.

2. Escitalopram was also statistically significantly more efficacious than citalopram, clomipramine, and fluvoxamine.

3. Mirtazapine was statistically significantly more efficacious than fluvoxamine.

I think that despite this paper, the choice of antidepressant will continue to be based on a range of factors like the side effect profile, potential for drug interactions, and so on.

Expand Content
Share this Section:
A new treatment for bipolar I depression in children and adolescents

DelBello et al. Efficacy and Safety of Lurasidone in Children and Adolescents With Bipolar I Depression: A Double-Blind, Placebo-Controlled Study. J Am Acad Child Adolesc Psychiatry. 2017 Dec;56(12):1015-1025. PubMed PMID: 29173735.


There is very little data on the treatment of bipolar depression in children and adolescents.

Lurasidone has been approved by the FDA for both monotherapy and adjunctive therapy of bipolar depression in adults.

This clinical trial aimed to evaluate the efficacy and safety of lurasidone in children and

adolescents with bipolar depression.


This was a multicenter, six-week randomized, double-blind, placebo-controlled clinical trial. The study was conducted 64 different sites around the world.

Children and adolescents 10 to 17 years old with a diagnosis of bipolar I depression were enrolled. The diagnosis was based on a semi-structured interview, the Schedule for Affective Disorders and Schizophrenia for School-age Children-Present and Lifetime (K-SADS-PL) that was administered by a “trained clinician.”

Participants were randomly allocated to receive double-blind treatment with lurasidone (flexible dose 20 to 80 mg/day) or placebo for six weeks.

Many rating scales were used, but the primary outcome measure was change in the severity of depression, as measured by the Children's Depression Rating Scale-Revised

(CDRS-R) total score, from baseline to week six (i.e., at the end of six weeks).


A total of 347 patients were randomized and received at least one dose of study drug.

Their mean age was about 14 years.

After six weeks of receiving the study drug, the end of the study, participants who received lurasidone had statistically significantly greater improvement than those who received placebo.

The effect size was 0.45 (moderate). Note: this effect size is similar to that for lurasidone as a treatment for bipolar depression in adults (0.51).

Anxiety, quality of life, and global functioning also showed greater improvement in the lurasidone group.

Rates of discontinuation from the study were similar for the lurasidone (92%) and placebo (90%) groups.

What about weight gain? The average weight gain is, in my opinion, not useful and potentially misleading. The percentages of participants who had significant weight gain over up to six weeks of treatment (defined by convention as gaining 7% or more of the baseline weight) were 4% on lurasidone and 5% on placebo. But the percentages of participants for whom increased weight was reported as an adverse event were 6.9% on lurasidone and 1.7% on placebo.

There was a small but statistically significant average increase in serum prolactin in participants who received lurasidone. The percentages of participants in whom there was a clinically relevant increase in serum prolactin was not stated in the paper. But, there were no prolactin-related adverse events.


Lurasidone monotherapy (20 to 80 mg/day) was efficacious for treating bipolar I depression in children and adolescents.

Clinical Commentary

On March 6, 2018, the FDA approved lurasidone as monotherapy for bipolar I depression in pediatric patients (10 to 17 years old).

This study is important because of the previously limited data on the treatment of bipolar I depression in children and adolescents.

One fact that is buried in the paper but I think could be important is that the effect size for lurasidone was much greater for participants who were 15 to 17 years old than for those 10 to 14 years old (0.7 versus 0.1). That’s a huge difference in effect size. This difference in effect size was because the drug-placebo difference was much less in the younger participants. Could this be because bipolar disorder in children is biologically different from that in adolescents?

Expand Content
Share this Section:
This genetic test may help guide escitalopram treatment

Jukić et al. Impact of CYP2C19 Genotype on Escitalopram Exposure and Therapeutic Failure: A Retrospective Study Based on 2,087 Patients. Am J Psychiatry. 2018 Jan 12:appiajp201717050550. [Epub ahead of print] PubMed PMID: 29325448.


Escitalopram is mainly metabolized by the cytochrome P450 2C19 enzyme (CYP2C19).

This study evaluated whether variations (polymorphisms) in the gene that codes for the CYP2C19 enzyme is associated with differences in serum levels of escitalopram and with lack of benefit from escitalopram treatment.

The variations in the CYP2C19 gene are known to be associated with function of the enzyme. The gene can code for low or absent activity, average activity, or increased activity.

Based on the nature of two genes in the pair, each person can be considered to have gene variations that predict that person with be one of the following with regard to CYP2C19 activity (in order of increasing activity of the enzyme):

1. Poor metabolizer

2. Intermediate metabolizer (less enzyme activity than Extensive Metabolizers but more activity than Poor Metabolizers)

3. Extensive (“normal”) metabolizer

4. Ultrarapid metabolizer


Data on about 2000 patients were obtained for this study from a drug monitoring database in Norway.

Data on the CYP2C19 genotype and serum escitalopram levels had been collected from these patients in the past. That is, the study was retrospective.

For data analysis, the serum concentration was adjusted for the dose of escitalopram.


Compared with Extensive Metabolizers, serum concentrations of escitalopram were three times greater in Poor Metabolizers and about 1.5 times greater in Intermediate Metabolizers. This is to be expected due to lowered metabolism of the drug by the CYP2C19 enzyme in both Poor Metabolizers and Intermediate Metabolizers.

On the other hand, compared with Extensive Metabolizers, serum concentrations of escitalopram were 10 to 20% less in Ultrarapid Metabolizers, which is also to be expected.

So, the genetic variations were associated with serum concentrations of escitalopram in a way that makes sense.

But, did this affect clinical outcomes?

The study looked at the percentages of patients who switched to another antidepressant within one year after the last measurement of serum escitalopram levels. These percentages varied from 12% in Extensive Metabolizers, 14% in Intermediate Metabolizers, 18% in Ultrarapid Metabolizers who were heterozygous, 29% in Ultrarapid Metabolizers who were homozygous, and 31% in Poor Metabolizers. To me, the clear stepwise relationship—increasing risk of switching with step-wise increasing metabolism is important to note.

Note that in this study we don’t know why the participants switched from escitalopram to another antidepressant. Reasons may have included non-response, inadequate response, side effects, etc.


The CYP2C19 genotype had a significant effect on both serum escitalopram levels and the probability of switching from escitalopram to another antidepressant.

The authors concluded that these findings support the potential clinical utility of CYP2C19 genotyping for individualization of treatment with escitalopram.

Clinical Commentary

Given that escitalopram is a very commonly used antidepressant, the findings of this study are of considerable clinical importance. Note that Ultrarapid and Poor metabolizers together were 33% of the patients in this study, i.e., they are not rare.

Note that, presumably, switching from escitalopram may be due to either side effects or a lack of an adequate response. That is why it should not be surprising that both Poor Metabolizers and Ultrarapid Metabolizers have increased rates of switching from escitalopram.

The current clinical utility of pharmacogenomic testing in clinical practice is hotly debated and I cannot go into the pros and cons here. But, I do want to point out that whether genotyping should be done routinely or only in cases of non-response or unexpected side effects is not simply a scientific question. It is also an issue of cost-effectiveness. The cost of testing may vary widely depending on what genes and polymorphisms are being tested, insurance contracts, etc.

The authors argue that it may be especially important to get to an appropriate serum escitalopram concentration quickly in certain types of patients, e.g., those with high rates of nonadherence or with suicidal ideation.

Expand Content
Share this Section:
Should we add or remove an anticholinergic in persons with tardive dyskinesia?

Bergman and Soares-Weiser. Anticholinergic medication for antipsychotic-induced tardive dyskinesia. Cochrane Database Syst Rev. 2018 Jan 17;1:CD000204. Review. PubMed PMID:



Anticholinergic medications are often used along with antipsychotics in order to treat extrapyramidal side effects like parkinsonism and dystonia.

But, it is not clear whether anticholinergic medications also work for tardive dyskinesia.

Also, there is some animal data to suggest that long-term use of anticholinergic medications may be associated an increased risk of tardive dyskinesia. And, it is believed that stopping anticholinergic medication may reduce the symptoms of tardive dyskinesia.

This review from the Cochrane Collaboration aimed to assess whether either the use of or discontinuation of anticholinergic medication reduces antipsychotic-induced tardive dyskinesia.


A thorough search of the literature was done to identify all relevant controlled clinical trials.

Clinical trials were included if the participants had antipsychotic-induced tardive dyskinesia and schizophrenia or other chronic mental illness, and had been randomly allocated to either:

1. An anticholinergic medication versus placebo, another intervention, or no intervention.

2. Discontinuation versus continuation of an anticholinergic medication.


Only two relevant clinical trials were identified and even combining the two, only a total of 30 patients were randomized.

The participants had been diagnosed with schizophrenia.

One study (n=20) compared patients randomized to receive either procyclidine

(an anticholinergic) or isocarboxazid (an MAO inhibitor) for up to 40 weeks. This study found that those randomized to receive the anticholinergic were statistically significantly much less likely to have clinically relevant improvement than those randomized to receive isocarboxazid (relative risk 0.24).

The other study compared discontinuation versus continuation of anticholinergic medication.

The other trial (n=10) compared anticholinergic withdrawal with anticholinergic continuation. There was no statistically significant difference in participants leaving the study early. But, it was not reported whether or not there was a clinically relevant improvement in the tardive dyskinesia.


The authors concluded that no confident statement can be made about whether or not anticholinergic medications or their discontinuation are efficacious for the treatment of antipsychotic-induced tardive dyskinesia.

Clinical Commentary

I have frequently seen patients with tardive dyskinesia who were treated with anticholinergic medications based on the false rationale that  “tardive dyskinesia is an extrapyramidal side effect and anticholinergic medications work for extrapyramidal side effects.”

Clinicians should note that anticholinergics have not been shown to work for tardive dyskinesia and clinical experience agrees with this. The limited available data, in fact, suggest that they may worsen tardive dyskinesia.

Tardive dyskinesia continues to be commonly missed in clinical practice until it has progressed considerably. Better efforts at early detection should be made.

Two medications are currently FDA-approved for the treatment of tardive dyskinesia. In addition, the efficacy of older strategies also needs to be further evaluated.

Expand Content
Share this Section:
One more way in which medications for ADHD may be helpful

Man et al. Effectiveness of Pharmacological Treatment for Attention-Deficit/Hyperactivity

Disorder on Physical Injuries: A Systematic Review and Meta-Analysis of Observational Studies. CNS Drugs. 2017 Dec;31(12):1043-1055. Review. PubMed PMID: 29255995.


Persons with attention-deficit/hyperactivity disorder (ADHD) are known to be more prone than others to a variety of physical injuries, e.g., motor vehicle accidents, fractures, and brain injuries.

Several previously published observational studies have looked at whether or not pharmacological treatment of ADHD reduces the incidence of physical injuries in these patients. Presumably, such reduction in risk is mediated by reduction in inattention and impulsivity.

The paper describes a systematic review and meta-analysis of the literature on this topic.


Physical injury was defined for this analysis as physical injuries that required medical attention, e.g., hospitalization, emergency department visits, or general practitioner visits.

Relevant studies were identified by a search of the literature.

The search included studies in both children and adults and of both stimulant and non-stimulant ADHD medications.


Ten observational studies were included in the analysis.

The studies either used the individual patient as his or her own control or compared two different groups of individuals.

Treatment with ADHD medications was associated with a statistically significant decrease in physical injuries.

For the studies that compared individuals to themselves, the adjusted rate ratio for physical injuries was 0.76, i.e., a 24% reduction in physical injuries. In subgroup analyses, the ratio was lower for adults than for children (0.60 versus 0.86).

For studies comparing one group of individuals to another, the adjusted rate ratio was 0.88, i.e., a 12% reduction in physical injuries. In subgroup analyses, the reduction in risk was statistically significant only in adults (rate ratio 0.86) and not in children (rate ratio 0.93).


This meta-analysis found a decreased risk of injuries in persons with ADHD who were treated with ADHD medications.

Clinical Commentary

An important fact that is often forgotten in debates about the pros and cons of medications for ADHD is that persons with ADHD are at increased risk of medically significant physical injuries and even death.

Given the large number of persons who have ADHD and a 12 to 24% reduction in risk, medication treatment of ADHD may be associated with a large reduction in clinically significant physical injuries.

But it appears that, for whatever reason, this benefit of ADHD medications in reducing physical injuries might be greater in older adolescents and adults than in children.

Expand Content
Share this Section:

Rajnish Mago, MD
Medical Editor, GME Research Review

GME Research Review is a monthly newsletter edited by Rajnish Mago, MD, who is author of "The Latest Antidepressants" and "Side Effects of Psychiatric Medications: Prevention, Assessment, and Management." Dr. Mago selects, summarizes, and provides a clinical commentary on the latest published research in psychiatry. 

We are always carefully evaluating which research papers to discuss in GME Research Review. Have come across a research paper published in the last 6 months that you thought is clinically relevant? Do you want me to analyze it for you and for the benefit of others? Please email Dr. Mago the citation at [email protected].

To contact GME, email us at [email protected]

GME does not provide medical advice. The website and articles are intended for informational purposes only. They are not a substitute for professional medical advice, diagnosis or treatment. Never ignore professional medical advice in seeking treatment because of something you have read on the GME Website. If you think you may have a medical emergency, immediately call your doctor or dial 911.

Log In

Join GME For Free

Become a GME subscriber and gain full access to our extensive library of 700+ psychiatric medical education videos, free CME webcasts, latest research updates, and more. To sample our content, watch the featured videos below.
Join Today!