Antipsychotics Ranked in Schizophrenia

The Antipsychotic Hierarchy Is Real — and Guidelines Should Reflect That

REVIEW OF: Schneider-Thoma J et al, Lancet 2026;407:876–891

STUDY TYPE: Network meta-analysis of randomized trials

Which is the best antipsychotic to start with in schizophrenia? Guidelines rarely choose favorites, but some stand out. This massive network meta-analysis of 438 randomized controlled trials and nearly 80,000 participants across 24 antipsychotics makes the efficacy differences hard to ignore.

The Top

Clozapine led in efficacy (effect size 0.90), followed by amisulpride, olanzapine, and risperidone — each superior to at least three other agents. At the bottom sat lumateperone, brexpiprazole, iloperidone, cariprazine, and lurasidone, which were less effective than at least three other drugs.

Xanomeline–trospium (Cobenfy), the first non-dopaminergic antipsychotic, ranked in the top third for efficacy and spared patients the usual side effects that cause long-term problems. No weight gain, no EPS, no prolactin elevation. The tradeoff is cholinergic and anticholinergic side effects that cause short-term discontinuation. Xanomeline–trospium had the highest discontinuation rate in the analysis.

The Middle

Partial agonists (aripiprazole, brexpiprazole, cariprazine) showed better overall tolerability but middling efficacy — aripiprazole is the best bet in this class given its middle-of-the-pack efficacy ranking.

The Rest

The tables above detail efficacy for positive (psychotic) and negative (functional impairment) symptoms. At the end are rankings of side effects to help personalize the selection.

Limitations

These rankings rests entirely on placebo-controlled trials, so head-to-head data are still needed. The “network” method bakes in a lot of assumptions, but studies using different techniques have confirmed the superior efficacy of clozapine, amisulpride, olanzapine, and risperidone.

Clinical Implications

One guideline is in line with these rankings, the Psychopharmacology Algorithm Project. They suggest…

1. Start with a well-tolerated generic antipsychotic (or amisulpride outside the US)
2. If a full trial doesn’t bring meaningful results, start olanzapine
3. If a full trial doesn’t bring meaningful results, start clozapine

I agree, though I might start with cariprazine (Vraylar) for its benefits in negative symptoms, or xanomeline–trospium (Cobenfy) for its long-term safety (the algorithm takes cost into account, focusing on generics).

So what does this analysis add? Aripiprazole is also a good first-line choice, given its balance of efficacy and tolerability.

Share Your Input in Comments

1. Which antipsychotic do you recommend first line?

Side Effects

Weight Gain

No surprise to see xanomeline-trospium, ziprasidone, lurasidone, and lumateperone as the most favorable here. Aripiprazole is also low, but in reality is hit or miss (some lose, some gain). Interesting that samidorphan did not significantly distinguish it from pure olanzapine, but those are not head to head trials.

Sedation

Clozapine is the most sedating here, and that problem was not relieved by modafinil in a controlled trial. Xanomeline-trospium has a wide spread on both sides – likely because it contains both a cholinergic (activating) and an anticholinergic (sedating).

Extrapyramidal Symptoms (EPS, Stiffness)

Those with low dopamine blockade rise to the top.

Prolactin

Aripiprazole is ideal here. It actually lowers prolactin and is used as an antidote for hyperprolactinemia. Risperidone is among the most likely to raise prolactin (with rates up to 70%).

Cholinergic and Anticholinergic

Xanomeline-trospium touches both of these, but may be worse on the cholinergic side (nausea, diarrhea, salivation, sweats, etc).

QTc Prolongation (Cardiac)

Sertindole is off the charts here, but pimozide may be the riskiest and was not included in the analysis. Because of this risk, the FDA requires testing of CYP 2D6 before dosing pimozide above 4 mg/day (or 0.05 mg/kg/day in children).

Stopping med for any reason

This is where xanomeline-trospium stands out. Patients are more likely to stop it early because of nausea or other GI side effects, which is unfortunate as it has the best long-term tolerability. What strategies are you using to improve its tolerability?

—Chris Aiken, MD
Director, Psych Partners
Editor in Chief, Carlat Psychiatry Report