Herrmann_2007_Can.J.Psychiatry_52_630

OBJECTIVE: To systematically review published clinical trials of the pharmacotherapy of neuropsychiatric symptoms of Alzheimer disease (AD). METHOD: We searched MEDLINE and EMBASE for published English-language medical literature. Our review focused on randomized controlled trials (RCTs) and corresponding metaanalyses.
RESULTS: The pharmacotherapy of neuropsychiatric symptoms of AD has been studied with numerous RCTs. The largest number of studies has focused on antipsychotics. Data are of reasonably high quality and indicate that risperidone and olanzapine are more effective than placebo for institutionalized patients with severe agitation, aggression, and psychosis. The efficacy of antipsychotics is counterbalanced by safety concerns that include cerebrovascular adverse events and mortality. Cholinesterase inhibitors and memantine appear to have modest benefits for patients with mildly to moderately severe symptoms. Antidepressants are effective for treating depression in AD, but more data are required to determine the efficacy of trazodone and citalopram for agitation and aggression. Carbamazepine appears to be efficacious, although side effects and concerns about drug-drug interactions limit its use. The data do not support the use of valproate. Benzodiazepines should only be used for short-term, as-needed use. There are insufficient data on other pharmacologic interventions, such as beta blockers, buspirone, and estrogen preparations.
CONCLUSIONS: Although there have been numerous well-designed studies of the pharmacotherapy of neuropsychiatric symptoms in AD, safer and more effective treatments are urgently needed.