The atypical antipsychotic pimavanserin (Nuplazid) reduced the risk of dementia-related psychosis relapse, according to the phase III HARMONY trial.
Following an open-label treatment phase to determine treatment response, a double-blind phase found that 13% of patients who continued on pimavanserin had a relapse of symptoms versus 28% of those who switched to placebo (HR 0.35, 95% CI 0.17-0.73, P=0.005), reported Erin Foff, MD, PhD, chief medical officer of MapLight Therapeutics and formerly of pimavanserin-maker Acadia Pharmaceuticals, and colleagues.
As described in the New England Journal of Medicine, relapse of psychosis — the primary endpoint of this interim analysis — was assessed in time-to-event analyses as an increase of at least 30% on the Scale for the Assessment of Positive Symptoms-Hallucinations and Delusions (SAPS-H+D) and a score of 6 or 7 on the Clinical Global Impression-Improvement (CGI-I) scale, hospitalization for dementia-related psychosis, stopping the treatment regimen or withdrawal due to lack of efficacy, or the use of other antipsychotics for this indication.
During the initial open-label phase, 61.8% of patients had a sustained response to 34 mg of once-daily pimavanserin and were eligible for randomization.
Although the open-label phase was planned for a 26-week follow-up period, it was stopped early due to efficacy. As a result, the pimavanserin and placebo groups were followed for a median of 18 and 11 weeks, respectively.
“Because the trial was stopped early for efficacy, the ability to assess clinical predictors of relapse is diminished, and it is possible that the active-treatment or placebo group would have had additional relapse events or adverse events if the trial had been longer,” Foff’s group pointed out.
The secondary endpoint — trial discontinuation for any reason — occurred in 22% of patients in the pimavanserin group and 38% of those in the placebo group (HR 0.45, 95% CI 0.26-0.79, P=0.005).
“We may conclude a longer sustained reduction in psychotic symptoms with pimavanserin than with placebo in patients with several types of dementia, no significant liability for extrapyramidal symptoms, and an acceptable side-effect profile,” commented accompanying editorial author Joseph Friedman, MD, of Mount Sinai System of Hospitals in New York City.
Still, he noted, a limitation of the trial was an inability to discern significant antipsychotic effects among different subtypes of dementia, such as Alzheimer’s, dementia with Lewy bodies, or Parkinson’s disease dementia.
This ultimately proved to be a roadblock to approval. The FDA turned down pimavanserin for this indication in April, citing “a lack of statistical significance in some of the subgroups of dementia, and insufficient numbers of patients with certain less common dementia subtypes as lack of substantial evidence of effectiveness to support approval.”
However, developer Acadia Pharmaceuticals responded saying the trial met its prespecified primary and secondary endpoints and that “statistical separation by dementia subgroups and certain minimum numbers of patients with specific subtypes were not among the prespecified requirements.”
Acting as a selective serotonin inverse agonist and antagonist preferentially targeting 5-HT2A receptors, the oral agent was first approved in 2016 for hallucinations and delusions associated with Parkinson’s disease psychosis. It’s currently available as 34 mg capsules and 10 mg tablets for this indication and carries a boxed warning for increased mortality in elderly patients with dementia-related psychosis.
Conducted at 101 global clinical sites, the HARMONY trial included 392 patients with moderate-to-severe psychosis in an initial open-label, single-arm 12-week treatment phase. Preliminary results were first presented at the virtual Psych Congress 2020 meeting.
Among the full cohort, about 66% of the patients had Alzheimer’s disease, 15% had Parkinson’s disease dementia, 10% had vascular dementia, 7% had dementia with Lewy bodies, and 2% had frontotemporal dementia. The average age was 75, the mean duration of cognitive impairment was 4 years, and nearly all participants were white.
After 41 participants were withdrawn for administrative reasons, 217 patients of the remaining patients had a sustained response to 34 mg of once daily pimavanserin. By week 12, there was an average 75.2% reduction from baseline in the SAPS-H+D score.
Among these responders to treatment, 105 were then randomized to continued treatment with pimavanserin and 112 were switched to placebo.
During this double-blind phase, adverse events occurred in 41% of patients on pimavanserin group and in 36.6% of patients on placebo. The most common adverse events reported in the pimavanserin group included headache, constipation, urinary tract infection, and asymptomatic QT prolongation. There was one adverse event resulting in death during the open-label portion and another in the pimavanserin arm after randomization.
Acadia is also testing pimavanserin for other indications, including as an adjunct for major depressive disorder in the CLARITY trial.
The study was funded by Acadia Pharmaceuticals.
Foff was employed by Acadia during the conduct of the study and multiple co-authors were employed by or worked as consultants for the company.