New Guidance on Antipsychotic Choice

Batya Swift Yasgur, MA, LSW

July 19, 2019

For most antipsychotics used to treat schizophrenia, many differences are relatively small, but there are marked differences in their side effects, new research suggests.

Results of a network meta-analysis of 402 randomized controlled trials (RCTs) that compared 32 older and newer oral antipsychotics either to each other or to placebo showed that with few exceptions, only clozapine, amisulpride, zotepine, olanzapine, and risperidone were significantly more effective for reducing symptoms.

In general, older antipsychotics were more frequently associated with extrapyramidal motor side effects (EPS) and elevations in prolactin levels. Many newer antipsychotics were associated with more weight gain and sedation.

"We hope that doctors can use the efficacy and side effects panels presented in the publication to choose the best-suited drug for the individual patient," Maximilian Huhn, PhD, of the Department of Psychiatry and Psychotherapy, School of Medicine, Technical University of Munich, Germany, told Medscape Medical News.

"The side effect profile should be weighed against the comparably smaller efficacy differences between the individual drugs," he said.

The study was published online July 11 in Lancet.

Newer vs Older Agents

Although antipsychotics are the "mainstay of treatment" for schizophrenia, they are associated with important side effects, the authors write.

For this reason, a clear understanding of the relative risks and benefits of these agents is "essential for informed decision making," the investigators state.

Newer antipsychotics are often the treatment of choice, but older antipsychotics are less expensive and are still used worldwide, especially in low-income countries, the authors note.

In addition, neither the older antipsychotics nor brexpiprazole and cariprazine — both recently approved agents — have been compared in a network meta-analysis.

"There are many antipsychotic treatment options for acute treatment of schizophrenia and many different side effects, so we conducted a network meta-analysis. This is an attempt to summarize the data in a quantitative way," said Huhn.

Of 54,417 studies, the researchers included 402 RCTs of antipsychotics used to treat patients with acute symptoms of schizophrenia or related disorders (n = 53,463 participants; mean [SD] age, 37.40 [5.96] years; 56.02% male; mean duration of illness, 11.9 [5.19] years).

Studies were double-blind and were either placebo controlled or head-to-head comparisons.

Studies of treatment-resistant patients, as well as studies of first episode, predominantly negative or depressive symptoms, concomitant medical illness, and relapse prevention, were excluded, as were trials in which antipsychotics were used as an adjunctive or combination treatment.

The researchers included all oral second-generation ("atypical") antipsychotics available in Europe or the United States, as well as a selection of oral first-generation ("typical" or "conventional") antipsychotics.

The primary outcome was change in overall symptoms of schizophrenia, as measured by published rating scales, such as the Positive and Negative Syndrome Scale or the the Brief Psychiatric Rating Scale.

Secondary outcomes were all-cause discontinuation; discontinuation due to inefficacy; and study-defined responder rates.

Additional secondary outcomes were change in positive, negative, and depressive symptoms; quality of life; and social functioning, as measured by means of published rating scales.

Side effects examined were EPS (assessed on the basis of use of antiparkinson drugs), akathisia, weight gain, prolactin levels, sedation or somnolence, QTc prolongation, and ≥1 anticholinergic side effects.

The network meta-analysis combined direct and indirect comparisons using a Bayesian hierarchical model.

Effect sizes were risk ratios (RRs) for dichotomous outcomes and standardized mean differences (SMDs) for efficacy-related continuous outcomes. These were selected because the study involved use of different rating scales.

Efficacy, Response Rates

All antipsychotics produced greater reductions in overall symptoms than placebo. SMDs ranged from –.89 (95% credible interval [Crl], –1.08 to –.71) for clozapine to –.03 (–.59 to .52) for levomepromazine (n = 218 studies; 40,815 participants; 32 antipsychotics).

Clozapine, amisulpride, zotepine, olanzapine, and risperidone reduced overall symptoms the most. For the remaining drugs, differences were "small or very uncertain."

Compared with placebo, amisulpride, risperidone, olanzapine, paliperidone, and haloperidol were found to be more effective than other drugs in reducing positive symptoms (n = 117 studies; 31,179 participants; 21 antipsychotic). Clozapine, amisulpride, olanzapine, and (to a lesser extent) zotepine and risperidone reduced negative symptoms significantly more than many other drugs (n = 132 studies; 32,015 participants; 21 antipsychotics).

Sulpiride, clozapine, amisulpride, and olanzapine were associated with significantly greater reduction of depressive symptoms; aripiprazole was associated with the greatest improvement in quality of life, and thioridazine with improvement in social functioning (n = 89 studies, 19,683 participants, n = 10 studies, 3341 participants; and n = 16 studies, 4370 participants, respectively).

Although 192 studies (n = 35,115 participants) used very different cutoffs for defining response rates, 29 of 31 antipsychotics included were found to have significantly higher response rates compared with placebo, with RRs ranging from 2.16 (95% CrI, 1.53 – 3.55) for thioridazine to 1.11 (95% CrI, 1.01 – 1.19) for brexpiprazole.

Individualized Treatment

When compared to placebo, clopenthixol was associated with the lowest discontinuation rates, and haloperidol with the highest.

  • Twelve antipsychotics, with zotepine, olanzapine, and sertindole topping the list, were associated with the most weight gain, compared to placebo (n = 116 studies; 28,317 participants).

  • Clozapine was associated with less risk for use of antiparkinson medications, compared with haloperidol; the highest risk was found with chlorpromazine (n = 136 studies; 24,911 participants). There was a similar hierarchy for akathisia (n = 116 studies, 25,783 patients).

  • The drugs most implicated in elevations in prolactin level were olanzapine, asenapine, lurasidone, sertindole, haloperidol, amisulpride, risperidone, and paliperidone (n = 90 studies; 21,569 patients).

  • Quetiapine was least implicated in causing QTc prolongation, followed by olanzapine and risperidone (n = 51 studies; 15,467 participants).

  • Of 32 antipsychotics examined in 162 studies (n = 30,770 participants), all were associated with some degree of sedation, but 18 were associated with significantly more sedation than placebo, with RRs ranging from 1.33 (95% CrI, 1.00 – 1.68) for paliperidone to 10.20 (95% CrI, 4.72 – 29.41) for zuclopenthixol.

  • Risk for side effects related to anticholinergics (which are often used to treat EPS) was most prominent with risperidone and haloperidol (n = 134 studies; 26,904 participants).

"All antipsychotics have their individual advantages and disadvantages, and none is optimal. This shows that the ideal drug has not been found yet," said Huhn.

He suggested that physicians tailor drug choice to each patient on the basis of efficacy and side effects presented in their findings. For example, "if a patient already has a metabolic problem, the doctor could choose an antipsychotic drug with an appropriate side effect profile (low risk for metabolic side effects)," he said.

On the other hand, for a patient with a history of cardiac problems, an agent that has a low risk for QTc prolongation may be a better choice.

Better-Informed Decisions

Commenting on the study for Medscape Medical News, Mark Olfson, MD, MPH, Elizabeth K. Dollard Professor of psychiatry, medicine, and law, Columbia University, New York City, who was not involved with the research, called the study "the largest of its kind." He noted that it offered "easy-to-interpret rankings of individual antipsychotic medications, not just on efficacy but on a host of side effects."

The study will "help psychiatrists and other physicians to make better-informed decisions," Olfson added.

The authors caution clinicians to "remember that reported results are averages and that response and side effects might vary considerably in individual patients."

The study was funded by the German Ministry of Education and Research. Huhn has received speaker's honoraria from Janssen and Lundbeck. The other authors' disclosures are listed in the original article. Olfson has disclosed no relevant financial relationships.

Lancet. Published online July 11, 2019. Abstract

For more Medscape Psychiatry news, join us on Facebook and Twitter


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.