Background

Pimavanserin is a serotonergic agent, specifically a selective serotonin (5-HT) 2A receptor inverse agonist. In April 2016, the FDA approved this once-daily atypical antipsychotic to treat the delusions and hallucinations that are a feature of psychosis in Parkinson's disease, and it came on the market in the U.S. in June (Apr 2016 news; prescribing information). Pimavanserin is thought to have fewer adverse effects than older antipsychotic agents used to treat Parkinson's psychosis, but evidence remains limited (e.g., Sarva and Henchcliffe, 2016; Tampi et al., 2019).

The drug carries a black-box warning of increased mortality in elderly people. One large retrospective study reported a higher risk of death in nursing home residents with PD who were prescribed pimavanserin (Hwang et al., 2021), while other analyses found no extra risk (Moreno et al., 2018; FDA analysis). Debate continues regarding these results, and regarding the safety of pimavanserin in real-world use (e.g., see Ganesh and Galetta, 2022; Subbiah et al., 2022; Hwang et al., 2022). A meta-analysis of clinical trials involving 680 patients found significant reduction in hallucinations and delusions, with no change in other symptoms. Adverse events were similar to placebo, except that the pimavanserin group had less orthostatic hypotension (Mansuri et al., 2022). In a small retrospective study of 54 patients, half of patients showed initial improvement in psychosis, but only 15 percent maintained the benefit after a year of treatment (Akbar and Friedman, 2022). An analysis of Medicare beneficiaries with PD who were starting pimavanserin or an atypical antipsychotic indicated lower mortality with pimavanserin, but only during the first 180 days of treatment and only in community-dwelling, not nursing home, patients (Mosholder et al., 2022). Another study found similar reduced mortality for pimavanserin compared to other antipsychotics in the first 180 days, whether patients were at home or in a facility (Layton et al., 2023). A meta-analysis of real-world evidence reported that pimavanserin did not increase the mortality risk in PD psychosis, and performed similarly to other atypical psychotics used in this indication (Isaacson et al., 2024).

Pimavanserin affects heart rhythms and should not be used by people with arrhythmia or those who take drugs that prolong the QT interval. The most common adverse events seen in trials were peripheral edema, nausea, confusion, hallucinations, constipation, and gait disturbance, though Acadia argues that PD with psychoses itself is associated with a higher risk of falls (Forns et al., 2021). In post-marketing reports from the FDA adverse event database, hallucinations and death were most frequently mentioned, while newly identified side effects included sleep disorders and cognitive and behavior changes (Gu et al., 2024). An Acadia-funded analysis of Medicare records claimed that Parkinson’s patients in nursing homes who took pimavanserin had a lower risk of falls and bone fractures, hospitalizations, and emergency room visits, compared to those on other anti-psychotics (Rajagopalan et al., 2024; Rajagopalan et al., 2024). Analysis of a smaller number of Medicare beneficiaries found a lower risk of hospitalization, but similar mortality rates for pimavanserin users compared to quetiapine (Alipour-Haris et al., 2023). Phase 4/real-world trials in Parkinson's psychosis are ongoing (Dashtipour et al., 2021; Cusick and Gupta, 2021).

In efforts to expand the indications for Nuplazid, Acadia initiated trials for psychosis, agitation, and aggression in Alzheimer's disease, for psychosis related to all-cause dementia, and for schizophrenia, insomnia, PTSD, depression, and autism. A different serotonin 2A receptor inverse agonist, nelotanserin, was in Phase 2 development for dementia with Lewy bodies (DLB) but has been discontinued.

Serotonin receptor activation is known to inhibit brain Aβ peptide production by altering the cleavage of amyloid precursor protein (Sheline et al., 2014). In recent preclinical work, pimavanserin or another 5HT2A agonist, M100907, acutely reduced Aβ concentrations in brain interstitial fluid of APP/PS1 transgenic mice. Chronic administration to aged APP/PS1 mice decreased CSF Aβ and brain plaque pathology, and improved cognition (Yuede et al., 2021; Gründer and Cumming, 2021).

Findings

Between November 2013 and November 2016, Acadia conducted a Phase 2 study of pimavanserin in 181 nursing home residents in London, who had psychosis along with their clinical diagnosis of AD. Participants received a 12-week course of 40 mg pimavanserin or placebo, and were evaluated on the nursing home version of the neuropsychiatric inventory (NPI), the short form of the Cohen-Mansfield Agitation Inventory (CMAI-SF), and the Alzheimer’s Disease Cooperative Study-Clinical Global Impression of Change (ADCS-CGIC). The primary outcome was change from baseline to six weeks in the NPI-NH, on which lower scores mean fewer symptoms. By two weeks, both the placebo and pimavanserin groups had improved on the test. At six weeks, those taking the drug had a 39.5 percent reduction in NPI-NH scores, compared to 19.3 percent reduction in the placebo group. However, by 12 weeks, scores were down 40 percent from baseline in both groups. These results were presented at CTAD, and later formally published (Dec 2017 conference news; Ballard et al., 2018). A subsequent subgroup analysis claimed a larger treatment benefit in participants with more severe psychosis (Ballard et al., 2019). Another post hoc analysis reported that people who responded to pimavanserin with a 50 percent or greater reduction in psychotic symptoms also improved on measures of agitation and aggression (Ballard et al., 2020).

In November 2016, Acadia started SERENE, a Phase 2 study comparing the effect of 34 mg and 20 mg pimavanserin to placebo in people with agitation and aggression as part of their Alzheimer's disease, on the CMAI-SF as primary outcome. This trial anticipated enrolling 432 patients at 56 sites, but stopped randomizing patients in November 2017 and was marked as complete in March 2018, with 111 patients enrolled and 83 of them having completed treatment. In March 2019, the trial's status was further updated to "terminated." Results, as well as sponsor rules restricting principal investigators from discussing and publishing results, are published in the clinicaltrials.gov History of Changes section of this trial. At 111 actual participants, the trial was underpowered to detect a treatment effect. The SERENE trial had a one-year open-label extension study, which recruited 79 patients and ended in February 2019.

From 2017 to 2019, Acadia conducted a Phase 3 study at 101 study locations in the U.S., Europe, and Chile, to evaluate pimavanserin’s efficacy in preventing psychosis relapse in people who had previously responded to the drug. The HARMONY trial enrolled 392 participants with all-cause dementia who had had symptoms of psychosis for at least two months. In an initial, 12-week open-label phase, all participants took 34 mg pimavanserin with the option to drop to 20 mg. The 217 who responded with a decrease in psychotic symptoms continued into a 26-week double-blind period, in which they were randomized to pimavanserin or placebo. The primary endpoint was delay in psychosis relapse with a p value below 0.0033. Relapse was defined as hospitalization due to dementia-related psychosis, significant deterioration of dementia-related symptoms on clinical scales, withdrawal from the study due to lack of efficacy, or use of an off-label antipsychotic medication. Patients were followed until relapse occurred (Cummings et al., 2018). According to a company press release, an interim analysis indicated that this trial met its primary endpoint early, and the trial was stopped (Sep 2019 news). Trial results published after peer review reported that 13 percent of the pimavanserin group relapsed, compared to 28 percent in the placebo group (Tariot et al., 2021). Of the participants, 66 percent had Alzheimer’s, 15 percent had Parkinson’s, and the remainder had vascular dementia, dementia with Lewy bodies, or frontotemporal dementia. Common treatment-related adverse events were urinary tract infection, headache, and constipation. QT interval prolongation occurred in 1.3 percent of pimavanserin-treated people. The types of side effect commonly associated with antipsychotics were infrequent (see also Tariot et al., 2022).

In May 2018, Acadia started enrolling 300 people who had neuropsychiatric symptoms along with a neurodegenerative disease into a Phase 3 safety study, conducted at 116 sites in Europe, Asia, and North and South America. It enrolled patients living at home or in assisted living. The study compared the effect of an eight-week course of 34 mg pimavanserin to placebo on treatment-emergent adverse events, extrapyramidal symptoms, and the MMSE. In October 2019, enrollment was increased to 750. The study was completed in May 2022 with 784 participants. A one-year open-label extension with 595 participants ran through May 2023. According to published results, the incidence of adverse events was similar between pimavanserin and placebo, and the drug did not affect motor or cognitive function (Alva et al., 2024).

In June 2020, based on data from the HARMONY dementia trial, Acadia submitted a supplemental new drug application to the FDA for pimavanserin to treat hallucinations and delusions in people with dementia-related psychosis (press release). On April 5, 2021, the company announced that the agency had declined to approve, citing a lack of substantial evidence of effectiveness. The FDA reportedly noted the absence of statistically significant changes in some of the dementia subgroups, and low numbers of patients with less common dementia subtypes (press release). According to Acadia, the decision was unexpected, given that the FDA had previously approved the study design, which analyzed people with different types of dementia as a single group, and met its primary endpoints. The FDA also rejected results from the SERENE Alzheimer’s trial as supporting evidence, saying the single-center study was not adequate and well-controlled. In February 2022, Acadia resubmitted the application to seek approval for a narrower indication of Alzheimer’s disease psychosis (company release). On June 17, an FDA advisory committee voted 9-3 that the drug appeared to be ineffective for this indication (press release). On August 5, Acadia announced that the FDA had declined to approve pimavanserin (press release). The agency stated that the positive results in the Harmony study appeared to be driven by the robust response among the Parkinson’s dementia subgroup (e.g. see Weintraub et al., 2024). The FDA recommended that Acadia perform another trial for psychosis in Alzheimer’s dementia.

A meta-analysis of safety outcomes in four placebo-controlled trials, presented at CTAD 2020, indicated that the drug caused no cognitive decline. In people treated with pimavanserin for up to nine months, changes in MMSE scores were small and similar to placebo. Of 14 cognition-related adverse events, confusion and memory impairment were reported in all four studies and occurred in fewer than 2 percent of patients. Pimavanserin did not affect motor function.

In October 2020, a 130-person randomized controlled trial started evaluating pimavanserin for a treatment benefit on impulse control disorder in Parkinson's disease. It is projected to run until summer 2023, at 16 sites in France.

In June 2022, a PET study was planned to begin assessing 5HT2A receptor density and occupancy at baseline and after six weeks pimavanserin treatment in people with PD, using the ^18FMH.MZ tracer (Kramer et al., 2020). The study began in January 2023 to enroll 75 patients at Vanderbilt University Medical Center, and will run until June 2025.

In March 2024, Acadia stopped all development of pimavanserin in non-Parkinson’s indications after the drug failed a Phase 3 trial for negative symptoms of schizophrenia (press release).

Several investigator-initiated studies are continuing to evaluate pimavanserin for psychosis in people with Lewy body dementia, for sleep and insomnia in people with PD psychosis, and for PTSD, aggression, and autism spectrum disorder symptoms.

For all trials of pimavanserin, see clinicaltrials.gov.