Antipsychotics Ranked in Major Depression

Not all antipsychotics are equal in depression

STUDY: McIntyre RS et al, JAMA Psychiatry 2026;

STUDY TYPE: Systematic review and network meta-analysis

FUNDING: Independent

Background

Five second-generation antipsychotics are FDA-approved as add-ons in major depression. The first three (“ABC”) are dopamine D3 agonists, sometimes called third-generation:

• Aripiprazole
• Brexpiprazole
• Cariprazine

Then comes:

• Lumateperone
• Quetiapine (XR is approved, but IR also worked in industry trials)

But wait, there are more options left out of this study:

• Olanzapine is approved in combination with fluoxetine
• Risperidone is effective but not approved
• Lurasidone is well-studied in depression with mixed features (as monotherapy) but not tested in major depression

The Study

• Network meta-analysis of 22 double-blind, placebo-controlled trials and nearly 11,000 adults with major depressive disorder.
• Primary outcomes were efficacy (at least 50% improvement on the Depression scale [MADRS]) and acceptability (all-cause discontinuation) at six weeks.

Ranking by Efficacy

1. Lumateperone (risk ratio 1.72)
2. Aripiprazole (1.53)
3. Brexpiprazole (1.38)
4. Cariprazine (1.20)
5. Quetiapine XR (1.15, which did not reach statistical significance)

The remission data ranked these the same way.

Ranking by Tolerability

Tolerability, measured as discontinuation due to adverse effects, flipped the order.

1. Aripiprazole
2. Cariprazine
3. Brexpiprazole
4. Lumateperone

Lumateperone had the highest discontinuation rate due to adverse events, despite being the most efficacious. Notably, lumateperone was the only agent not associated with significant weight gain.

Breaking it Down by Dose

The dose breakdown tells us why quetiapine fell hard. An unusual fail in the 225 mg arm, which probably has more to do with the study design than the actual dose.

It also suggests there’s not much benefit, and sometimes more side effects, in raising the dose, although aripiprazole 2 mg looks too low to work, something confirmed by an earlier meta-analysis.

Limitations: Network analyses have built in assumptions that skew the ratings. Almost all trials were industry-sponsored, and no head-to-head trials exist. Placebo response varied widely across trial programs, which likely inflates some efficacy estimates and deflates others — the cariprazine trials had notably high placebo responses.

Practice Implications

1. Don’t take network analyses literally. They are hypothesis-generating. Interpret in light of other research.
2. Aripiprazole ranked #1 for efficacy in past research, but that was before the lumateperone era.
3. The problem with lumateperone’s initial tolerability confirms my experience, and why many of us are starting it low and titrating. Longer term, it has better tolerability.
4. Quetiapine’s low efficacy is the most questionable finding here, and likely due to a single aberrant study. In fact, only olanzapine and quetiapine have good long-term efficacy data in major depression. Brexpiprazole and risperidone failed to prevent depression in industry-sponsored maintenance trials, and the others are untested long-term.