Auvelity’s Approval Just Rewrote the Rules for Dementia Agitation Trials

Picture a regulatory affairs director at a mid-size CNS sponsor sitting across from her clinical team in early 2024, staring at the brexpiprazole approval package from May 2023. The FDA had just cleared Rexulti as the first drug ever approved for agitation associated with Alzheimer’s dementia, and her team assumed they now had a template. Run a Phase 3 with a validated agitation scale, hit your primary endpoint, and file. Then the agency went ahead and approved a second drug for the same indication under entirely different evidentiary logic, and the template dissolved.

Auvelity (dextromethorphan hydrobromide and bupropion hydrochloride extended-release tablets) received FDA approval for agitation associated with dementia due to Alzheimer’s diseaseAlzheimer’s disease, becoming the first non-antipsychotic drug cleared for this indication. The agency had granted Auvelity both Breakthrough Therapy Designation and Priority Review. Those two designations sitting together in the same package signal something important: the FDA believed the existing evidence, before the pivotal trial even completed, was already pointing somewhere the agency wanted to follow quickly.

That combination of designations does not happen by accident. It tells you the agency looked at the mechanism, the preliminary data, and the unmet need in a population of patients whose agitation frequently leads to institutionalization, physical restraints, or off-label antipsychotic use with black-box mortality warnings, and concluded that waiting for a slow standard review cycle was the wrong call. What it also tells you, if you are designing a competing trial right now, is that the FDA’s expectations for this indication have just been reset by two approvals that used different pharmacological logic to reach the same label.

Two Approvals, Two Evidence Frameworks

The brexpiprazole approval in May 2023 was built on the architecture of a conventional antipsychotic development program. A Phase 3 clinical trial demonstrated statistically significant reduction in agitation scores, with the Cohen-Mansfield Agitation Inventory serving as the primary outcome measure. The FDA accepted that framework because brexpiprazole’s dopamine D2 partial agonism gave the agency a mechanistic story it recognized from the atypical antipsychotic literature, even if the Alzheimer’s agitation population represented a new indication.

Auvelity operates through a fundamentally different mechanism: NMDA receptor antagonism via dextromethorphan, modulated by bupropion as a CYP2D6 inhibitor that extends dextromethorphan’s half-life. Axsome Therapeutics, the sponsor, was not submitting a “me-too antipsychotic” package. The Breakthrough Therapy Designation the FDA awarded signals the agency was already persuaded by early-phase data that NMDA pathway modulation could address the neurobiological substrate of agitation in ways that dopaminergic agents cannot fully capture.

That mechanistic distinction carries direct trial design consequences. When you submit a Breakthrough Therapy IND for a CNS behavioral indication, the FDA’s guidance on early Alzheimer’s disease drug development, which the agency has updated to address the full spectrum of disease stages, encourages sponsors to engage early on endpoint selection, diagnostic criteria for enrollment, and how the clinical meaningfulness of behavioral improvements will be defined. The guidance is not prescriptive about which agitation scale must be used. That flexibility is real, but so is the agency’s ability to move the goalposts through the intensive FDA guidance meetings that come with Breakthrough status.

Sponsors who treat Breakthrough Therapy Designation as a faster version of standard review are misreading the instrument entirely.

The Endpoint Problem Nobody Is Talking About

Here is where the conventional wisdom about dementia agitation trials breaks down. The prevailing assumption among CNS sponsors is that replicating the endpoint structure of the first approved drug in a new indication is the safest path to approval. If brexpiprazole cleared on the Cohen-Mansfield Agitation Inventory, then every subsequent sponsor should anchor their pivotal trial to the same scale. The logic is understandable: the FDA accepted it once, it will accept it again.

Auvelity’s approval complicates that assumption. The FDA has now cleared two drugs for the same indication through programs with different regulatory designations, different mechanisms, and potentially different endpoint sensitivities. The agency’s guidance on Alzheimer’s drug development explicitly addresses how the appropriate endpoints shift depending on the stage of disease and the nature of the symptom cluster being targeted. Behavioral and neuropsychiatric symptoms like agitation occupy a different evidentiary space than cognitive decline or functional capacity, and the FDA has shown, through these two approvals, that it will evaluate them case by case rather than through a rigid template.

Consider what that means operationally for a Phase 3 trial enrolling patients with moderate-to-severe Alzheimer’s dementia and clinically significant agitation. Your enrollment criteria must distinguish agitation from other neuropsychiatric inventory domains, particularly anxiety and psychosis, because the FDA will scrutinize whether your enrolled population actually has the condition your drug targets. Your rater training program must demonstrate cross-site reliability on your chosen scale. And your statistical analysis plan must pre-specify how you handle the floor effects and dropout patterns that are endemic to late-stage dementia trials, where patients deteriorate cognitively even during a 12-week efficacy window.

Two approvals do not give you two templates. They give you two data points that the FDA will use to triangulate where your evidence package lands.

What Sponsors Must Build Into Their Strategy Now

The practical implication of having both Rexulti and Auvelity on the market for the same indication is that the FDA’s internal reviewers now have a comparative benchmark that did not exist in 2022. A third sponsor filing a New Drug Application for dementia agitation will face a division that can point to brexpiprazole’s Phase 3 results and to Auvelity’s Breakthrough-accelerated package and ask: where does your evidence sit relative to either of these? That is a higher implicit evidentiary bar, even if the formal regulatory standard has not changed.

Sponsors who plan to enter this space need to do three things their predecessors did not have to do. First, they need a pre-IND or Type B meeting specifically scoped to endpoint alignment, not just trial design feasibility. The question to put to the FDA directly: will the agency accept the same primary endpoint used in the brexpiprazole program, given that Auvelity’s mechanism may have produced a different effect profile on the same scale? Second, they need a rater qualification protocol that is documented, auditable, and defensible against a Complete Response Letter that challenges inter-rater variability in a cognitively impaired population. Third, they need a caregiver burden component in their protocol, not as a secondary endpoint but as a patient experience outcome that maps to clinical meaningfulness, because the FDA’s guidance framework increasingly requires sponsors to demonstrate that a statistically significant change in a rating scale translates into something a patient or caregiver can actually perceive.

That last point is where most CNS sponsors underinvest. A 12-point reduction on the Cohen-Mansfield Agitation Inventory is a number. What it represents in the lived experience of a family caregiver managing a parent with Alzheimer’s who is striking out at night-shift aides is the clinical story the FDA needs to tell when it writes the approval press release. Build that story into your trial design before the agency asks for it in a late-cycle review meeting.

The Phase 3 infrastructure for dementia agitation trials has also become more technically demanding in the period between brexpiprazole’s development program and Auvelity’s approval. Decentralized data collection for behavioral endpoints in cognitively impaired populations carries real data integrity risks: caregiver-reported outcomes collected remotely need the same validation rigor as in-clinic assessments, and any deviation from your eCOA validation plan will surface in a GCP inspection as a potential source of endpoint contamination. The FDA does not distinguish between a small trial and a large one when it comes to data provenance on primary endpoints.

Back to that regulatory affairs director staring at the brexpiprazole approval package in early 2024. She thought she had a roadmap. What she actually had was the first coordinate on a map the FDA is still drawing. Auvelity’s approval added the second coordinate, and the line between them points somewhere more demanding than either approval alone would suggest: a regulatory environment where two approvals in a novel indication raise the implicit standard for every sponsor who follows, where Breakthrough Therapy Designation is a conversation, not a guarantee, and where the evidence package that cleared in 2023 may not be sufficient in 2026. The third drug to seek approval for Alzheimer’s agitation will be judged against both predecessors simultaneously, and the sponsors who are designing that trial today need to be building for that comparison, not for the label the first approval created.

References

1. JAMA — “FDA Approves New Medication for Agitation Associated With Dementia”
2. FDA Press Announcement — “FDA Approves First Drug to Treat Agitation Symptoms Associated with Dementia Due to Alzheimer’s Disease” (Rexulti/brexpiprazole, May 2023)
3. FDA Press Announcement — “FDA Approves First Non-Antipsychotic Drug to Treat Agitation Associated with Dementia” (Auvelity)
4. FDA — “Early Alzheimer’s Disease: Developing Drugs for Treatment Guidance for Industry”
5. First Word Pharma — Brexpiprazole Phase 3 Clinical Trial Results