When Alzheimer’s disease patients who took risperidone for psychoses discontinued the drug, they were more than four times more likely to relapse than were those who continued treatment in a randomized, placebo-controlled trial.

The ADAD (Antipsychotic Discontinuation in Alzheimer’s Disease) trial suggests that at least some patients might benefit from a longer course of antipsychotic treatment, Dr. Davangere Devanand said at the Alzheimer’s Association International Conference 2012.

The ADAD trial began with a 16-week, open label phase in which 180 patients with dementia and psychotic symptoms or agitation and aggression received a flexible dose of risperidone. At the end of that period, 110 responders were randomized to one of three 32-week arms: continued risperidone treatment; placebo for 32 weeks; or risperidone for 16 weeks followed by 16 weeks of placebo.

The mean daily dose was small, at 0.97 mg. The primary outcome was relapse to psychotic symptoms. At the end of the run-in period, the average total Neuropsychiatric Inventory score had decreased from 36 to 9.

In the first half of the randomization period, 60% of those on placebo relapsed, compared with 33% of those who continued the drug (hazard ratio, 1.97).

The second 16-week period evaluated the change in those who switched from risperidone to placebo, compared with the continuation group. Relapse occurred in 48% of the switched patients, compared with 15% of those who continued risperidone (HR, 4.33)

“There were no significant interactions predicting who would relapse, even in the baseline measures of psychosis or cognition,” said Dr. Devanand, codirector of the memory disorders center and codirector of the late life depression clinic at the New York State Psychiatric Institute, New York.

There were no significant between-group differences in adverse events, he said. There were three deaths in the run-in phase and three in the randomization phase (two on risperidone and one in placebo), but Dr. Devanand said that the sample size wasn’t large enough to draw any conclusions about mortality risks.

He pointed out several factors that could have affected the study’s outcome. “This is a discontinuation trial, so the sample was an enriched population,” he said. “This might mean that relapse is more likely in patients who have previously responded to a drug.”

Dr. Devanand disclosed financial relationships with Eli Lilly, Novartis, Bristol-Myers Squibb, and Sanofi Aventis. Janssen Pharmaceuticals supplied the risperidone used in the study but had no input in the data compilation. The study was supported by the National Institutes of Health and the National Institute on Aging.