Newer Agents Field Guide
Companion volume — psychiatry care algorithms

Psychiatry Care Algorithms for Newer Antipsychotic Medicines

The last three years didn't just add new molecules to the second-generation shelf — they opened a second axis entirely. This volume is organized the way these drugs actually get chosen in practice: not by receptor class, but by which population and which indication you're treating, since several of the newest approvals sit entirely outside classic schizophrenia pharmacology.

Scope non-D2-centric & newly-indicated agents, 2015–2026
Populations schizophrenia · MDD adjunct · Parkinson's psychosis · Alzheimer's agitation
Companion to the core SGA field guide — receptor pharmacology lives there
01 — The Framing Shift

Why "newer" now means a different axis, not just a newer molecule

Every SGA from risperidone through cariprazine shares the same basic logic: block or partially agonize D2, add 5-HT2A antagonism to soften the EPS/prolactin cost, and differentiate on the rest of the receptor profile. The newest wave breaks that logic in two different directions at once.

◆ Direction one — mechanism

Muscarinic agonism (Cobenfy) and, going back a decade, selective 5-HT2A inverse agonism (pimavanserin) both treat psychotic symptoms with no D2 engagement at all. That's not a refinement of the old algorithm — it's a genuinely separate pathway with its own side-effect logic.

◆ Direction two — indication

Brexpiprazole's 2023 approval for Alzheimer's agitation, pimavanserin's standing indication for Parkinson's disease psychosis, and lumateperone's expansion into MDD adjunct all move established or new SGA chemistry into populations where the old schizophrenia-first algorithm doesn't apply cleanly — elderly patients, neurodegenerative disease, motor-sensitive patients. The "population router" below is built around that reality.

02 — Route by Population

Which newer agent, for whom

Click a population. Each panel gives the practical algorithm shape for that group, not the general schizophrenia algorithm — the sequencing genuinely differs once you're outside working-age schizophrenia.

Schizophrenia
MDD adjunct
Parkinson's disease psychosis
Alzheimer's agitation
Bipolar depression

Schizophrenia — where the newer agents slot into the standard algorithm

The core algorithm (adequate trial → reassess → switch → TRS → clozapine) still governs. Newer agents mainly change what you reach for when D2-blocking side effects, not inefficacy, are the limiting factor.

Poor EPS/metabolic tolerance on D2-blockers
Consider muscarinic-agonist switch or add newer low-metabolic-impact agent

Cobenfy is increasingly discussed as an option here specifically, and lumateperone / milsaperidone extend the low-metabolic-impact roster. None of these currently replace clozapine's role once TRS criteria are met — the evidence base for TRS specifically still belongs to clozapine.

CobenfyBysantiLumateperone

Major depressive disorder — adjunctive to an antidepressant

Atypical antipsychotics as MDD augmentation is an established strategy (aripiprazole, brexpiprazole, quetiapine XR); lumateperone is the newest entrant, positioned partly on a cleaner metabolic/sexual side-effect profile in trial reporting versus older augmentation options.

Inadequate response to antidepressant monotherapy
Augment, don't switch, if partial response
select by side-effect fit: activating vs. sedating vs. metabolic-sensitive patient
LumateperoneBrexpiprazoleAripiprazole (established)

Parkinson's disease psychosis — the D2-blocker-avoidance problem

Nearly every conventional antipsychotic worsens PD motor symptoms because it blocks the same dopaminergic transmission the patient's PD medications are trying to restore. This population needs an entirely different first-line logic than schizophrenia does.

Reduce/simplify dopaminergic PD meds first if feasible
Pimavanserin first-line (no D2 activity)
low-dose quetiapine or clozapine as alternatives if unavailable/ineffective

Pimavanserin remains the only FDA-approved agent specifically for PD psychosis (hallucinations and delusions), with its label clarified to cover patients with or without comorbid dementia. Quetiapine and clozapine are used off-label / evidence-based alternatives specifically because their D2 occupancy is low enough to avoid substantially worsening motor symptoms — most other SGAs are avoided in this population for exactly that reason.

PimavanserinQuetiapine (off-label)Clozapine (off-label)

Agitation associated with Alzheimer's dementia

Non-pharmacologic management (environmental modification, caregiver strategies, ruling out pain/ infection/delirium as drivers) remains first-line by every major guideline. Brexpiprazole is the first agent specifically FDA-approved for this indication — not for dementia-related psychosis broadly, and not as a PRN agent.

Rule out reversible causes; non-pharmacologic measures
Brexpiprazole 2–3 mg/day if agitation is severe/persistent
scheduled dosing only, reassess need regularly

Every antipsychotic used in this population, brexpiprazole included, carries the class-wide boxed warning for increased mortality in elderly patients with dementia-related psychosis — see the dedicated warning below. Brexpiprazole's approval is specifically for agitation, and explicitly not for dementia-related psychosis without agitation.

Brexpiprazole

Bipolar I depression

A smaller but real slice of the newer-agent landscape — lumateperone and cariprazine both carry bipolar depression indications with cleaner metabolic profiles than older options like quetiapine or olanzapine-fluoxetine, which remains a consideration when weight/metabolic tolerance is the limiting factor rather than efficacy.

LumateperoneCariprazine
03 — The Approval Arc

Ten years of a widening pipeline

Hover a node for what happened that year. The gap between 2016 and 2019 is real — this is a genuinely recent acceleration, not a steady drip.

04 — Deep Dives

The agents, in detail

Xanomeline–Trospium
Cobenfy (formerly KarXT)
Muscarinic M1/M4 agonistApproved Sep 2024

The first antipsychotic mechanism with zero D2 engagement. Xanomeline is a dual M1/M4-preferring muscarinic agonist doing the antipsychotic work; trospium is a peripherally restricted muscarinic antagonist riding along purely to blunt the cholinergic GI/autonomic side effects that made xanomeline alone intolerable in earlier trials. Because there's no D2 blockade, EPS, prolactin elevation, and much of the classic metabolic signature of D2-blocking agents are largely absent.

52-week pooled EMERGENT data and real-world results presented at APA's 2026 meeting associated it with fewer hospitalizations and reduced overall antipsychotic burden. Switch-study data (SIRS 2026) showed patients could be transitioned directly from oral atypicals while maintaining symptom stability — useful evidence for using it as a switch target, not only a first-line start.

Indication
Schizophrenia, adults
Where it fits
Poor EPS/metabolic tolerance on D2-blockers; increasingly used as a switch target
Main tolerability limiter
Cholinergic GI effects — nausea, dyspepsia, constipation, salivation — despite trospium
Evidence maturity
Young relative to D2-blocking class; real-world data still accumulating
Pimavanserin
Nuplazid
5-HT2A inverse agonist/antagonistApproved Apr 2016

Predates the muscarinic wave by nearly a decade but belongs in the same "no D2 engagement" family — a selective 5-HT2A inverse agonist/antagonist with some 5-HT2C activity and essentially no dopaminergic binding. That absence of D2 activity is the entire point: it's the only agent that treats PD psychosis without also fighting the patient's own dopaminergic PD medications.

A 2026 label update clarified it can be used in PD psychosis patients with or without comorbid dementia. A separate supplemental application seeking a broader "dementia-related psychosis" indication (beyond PD specifically) was not granted by the FDA over safety concerns in the broader Alzheimer's-dementia population — worth knowing, since it's a common point of confusion. The approved indication remains specifically Parkinson's disease psychosis.

Indication
Hallucinations & delusions of Parkinson's disease psychosis
Dose
34 mg once daily, no titration
Motor safety
Does not worsen motor function in pooled trial safety analyses
Not approved for
Dementia-related psychosis broadly (non-PD)
Brexpiprazole
Rexulti — Alzheimer's agitation indication
Agitation approval May 2023

Already established in schizophrenia and MDD adjunct (see the core SGA guide), brexpiprazole became the first agent specifically FDA-approved for agitation associated with Alzheimer's dementia in May 2023, based on 12-week placebo-controlled trials showing meaningful improvement at 2–3 mg/day. Agitation here covers pacing, restlessness, verbal outbursts, and physical aggression — one of the most common and caregiver-burdensome neuropsychiatric symptoms of Alzheimer's disease.

The approval is deliberately narrow: it is not indicated for dementia-related psychosis without agitation, and it should not be used as an as-needed ("PRN") treatment — it's a scheduled medication with regular reassessment of continued need, not a rescue dose.

Dose
Titrated to 2–3 mg/day target
Common side effects
Headache, dizziness, UTI, nasopharyngitis, somnolence/insomnia
Use pattern
Scheduled only, not PRN; reassess ongoing need regularly
Boxed warning
Class-wide dementia-related psychosis mortality warning applies — see below
Lumateperone
Caplyta
Multi-target, low D2 occupancy

Already approved for schizophrenia and bipolar depression; the newer development is its addition to the roster of atypical antipsychotics used as adjunctive MDD treatment, joining aripiprazole, brexpiprazole, and quetiapine XR. Mechanistically it combines 5-HT2A antagonism with comparatively low striatal D2 occupancy plus serotonin-reuptake and glutamatergic effects described in the literature — trial reporting has emphasized the absence of a mean weight, metabolic, or sexual side-effect signal versus placebo, which is the main selling point over older augmentation agents.

Indications
Schizophrenia, bipolar depression, MDD adjunct
Dose
42 mg once daily, with food
Notable
Minimal mean metabolic/weight/sexual signal in trial reporting
Milsaperidone
Bysanti
Approved 2025

The active metabolite of iloperidone (Fanapt) — Vanda leaned on existing iloperidone trial data in the NDA. Approved for schizophrenia and bipolar I. Its distinct physicochemical properties are seen as favorable for a possible future long-acting injectable formulation, something iloperidone itself never received.

Indications
Schizophrenia, bipolar I
Lineage
Active metabolite of iloperidone
Watch
LAI formulation in future development, not yet available
⚠ Class-wide boxed warning — elderly patients with dementia-related psychosis

Every antipsychotic discussed in the Alzheimer's-agitation panel above — brexpiprazole included — carries the FDA boxed warning that antipsychotic drugs are associated with an increased risk of death in elderly patients treated for dementia-related psychosis, largely from cardiovascular events and infections (pneumonia) in trial data across the class. This warning predates and is separate from any individual drug's own approval studies. It's the reason non-pharmacologic management is first-line, why brexpiprazole's Alzheimer's approval is specifically for agitation rather than psychosis, and why any antipsychotic use in this population should be at the lowest effective dose, for the shortest necessary duration, with regular reassessment.

05 — Watch List

What's coming next

Current to mid-2026. Treat this as a snapshot — pipeline status changes fast, and a compound listed here can be discontinued, redesignated, or approved well before this file is next revised.

Next-gen muscarinic agonists

More than seven companies now have structurally distinct M1-selective, M4-selective, and dual M1/M4-agonist compounds in development within three years of Cobenfy's approval — this looks like a real subclass forming, not a one-off. Several programs are specifically targeting Alzheimer's disease psychosis, a population with no FDA-approved pharmacologic treatment today and where conventional antipsychotics carry the mortality boxed warning discussed above.

Olanzapine LAI, reworked

TEV-'749, a once-monthly subcutaneous olanzapine LAI using Medincell's SteadyTeq delivery, had its NDA accepted by the FDA in February 2026. Phase 3 data through week 56 showed no post-injection delirium/sedation syndrome signal — if it clears review, it would be the first olanzapine LAI without the mandatory 3-hour observation burden that has limited Relprevv's real-world use.

TAAR1 agonism — a cautionary note

Trace amine-associated receptor 1 (TAAR1) agonism was, until recently, the other widely discussed non-dopaminergic mechanism in development. The lead compound in that class failed to separate from placebo in later-stage trials, a useful reminder that "new mechanism" and "will reach approval" are not the same claim — worth tracking status before assuming any pipeline compound will land.