When to Use, or Stop, Antipsychotics in Alzheimer Disease

Treating psychosis and behavioral problems in dementia patients is problematic because antipsychotic medications have a troublesome risk–benefit ratio (JW Psychiatry Jun 23 2008) and carry a black-box warning; still, many clinicians use them for severe behavioral issues. In this double-blind, placebo-controlled, discontinuation study, researchers assessed relapse rates in patients who had responded to risperidone.

The 180 patients with Alzheimer disease (AD) plus psychosis or agitation–aggression (mean age, 79.6; mean Mini-Mental State Examination score, 13.9) were outpatients or residents of assisted-living or nursing facilities. They received open-label, manufacturer-supplied, flexibly dosed risperidone for 16 weeks (starting dose, 0.25 mg/day; maximum dose, 3.00 mg/day); 112 responded (≥30% improvement in psychosis or agitation–aggression scores and at least much improved on a clinical-impression scale).

Responders were randomized to risperidone continuation for 32 weeks; risperidone, then a switch at 16 weeks to placebo; or placebo for 32 weeks. Sixteen weeks after randomization, rates of relapse (≥30% increase in neuropsychiatric scores and at least much worse on the clinical-impression scale) were 33% with risperidone versus 60% with placebo (P=0.004). During the next 16 weeks, relapse occurred in 15% of risperidone continuers versus 48% of those switched to placebo (P=0.02). Risperidone discontinuation rates were high: 38% in the open phase, 68% with 32 weeks of risperidone, and 29% with the switch to placebo after 16 weeks. Through all study phases, six deaths occurred.