BOSTON – Pimavanserin, an atypical antipsychotic approved for use in psychosis associated with Parkinson’s disease, was modestly effective in treating psychosis associated with Alzheimer’s dementia in a phase 2 study.
The study of 181 patients showed that pimavanserin (Nuplazid) was associated with a statistically significant 3.76-point improvement on the Neuropsychiatric Inventory–Nursing Home Version (NPI-NH) psychosis score, Clive Ballard, MD, at the Clinical Trials on Alzheimer’s Disease conference. But although pimavanserin was significantly more effective than placebo at 6 weeks, it lost its statistical edge by the trial’s end at 12 weeks, largely because the placebo group improved over the study period.
A key finding was that pimavanserin was more effective in a subset of patients with severe symptoms, reducing those by more than 4 points on the NPI scale, said Dr. Ballard, codirector of the Biomedical Research Unit for Dementia in the Institute for Psychiatry at King’s College London. “A 4-point change is the difference from having moderate symptoms daily to having them weekly. I think this is the most clinically relevant finding.”
The drug seemed to largely spare cognition, which is another notch in its clinical belt, said , professor of neurology at the Mayo Clinic, Scottsdale, Ariz.
“The relative preservation of cognition as seen in an absence of adverse cognitive effects is encouraging, and something pimavanserin may have over its antipsychotic rivals,” Dr. Caselli said in an interview. “The improved scores on the NPI seem modest as does the relative percentage of responders, defined as at least 30% improved NPI-NH score. But at least it is a positive result. One concern is that the company advises it may take 4-6 weeks to see an improvement, which is not the kind of timeline one has with acutely and severely agitated patients. So I suspect antipsychotic drugs, which work more quickly, are likely not going away.”
The 12-week, randomized 181 patients with advanced Alzheimer’s dementia to placebo or 40 mg pimavanserin. They were a mean of 86 years old. The mean baseline NPI-NH psychosis score was 9.8 and the mean Mini–Mental State Exam score was 10.
By 6 weeks, the psychosis score had improved significantly more in the pimavanserin group than in the placebo group (–3.76 vs. –1.93 points; P = .0451). The drug was more effective among patients with severe psychosis at baseline, defined as an NPI-NH psychosis score of at least 12. Among this group, the score improved by 4.43 points. The results were slightly, but not significantly, better in patients who had responded to prior antipsychotic medications and among those who were also taking a selective serotonin reuptake inhibitor. A responder analysis also favored treatment, with 90% of those taking pimavanserin experiencing at least a 30% improvement on the NPI-NH psychosis score, compared with 43% of those taking placebo.
At 12 weeks, however, the overall between-group difference was no longer statistically significant, because the placebo group continued to improve over the treatment period.
Safety and tolerability were important considerations in such an elderly and cognitively compromised group, Dr. Ballard noted. In this respect, pimavanserin performed relatively well. There were more serious adverse events in the treated group (16.7% vs. 11%). These included respiratory infections (5 vs. 2) and urinary tract infection (2 vs. 0). Falls and fractures were similar in both groups. There was one fall in the active group, with one laceration, one hip fracture, and one femoral neck fracture. In the placebo group, there was one fall, one upper limb fracture, one wrist fracture, and one vertebral fracture. Among treated patients, there was also one heart attack and one case of renal failure. Four patients in each group died during the study.
Psychiatric events were more common in the pimavanserin group, most notably agitation (21% vs. 14%). Other psychiatric adverse events included aggression (10% vs. 4%), anxiety (5.6% vs. 2.2%), and dementia-related behavioral symptoms (5.6% vs. 2%). The drug had no effect on Mini–Mental State Exam score.
Pimavanserin was associated with a mean change of 9.4 ms in the heart rate-corrected QT interval, and was more likely to induce a weight loss of 7% or more.
Dr. Ballard had no financial disclosures with regard to pimavanserin or Acadia Pharmaceuticals, which sponsored the trial.