Risperidone for Behavioral Disturbance Associated with Dementia

Abstract & commentary

Source: De Deyn PP, et al. A randomized trial of risperidone, placebo, and haloperidol for behavioral symptoms of dementia. Neurology 1999;53:946-955.

Aggression and other behavioral symptoms of dementia (e.g., agitation, pacing) psychosis create stress for caregivers and often adversely effect the patient’s quality of life. Many patients with dementia ultimately require pharmacologic intervention to manage disruptive behaviors. Neuroleptics are often effective, but conventional neuroleptics are associated with extrapyramidal symptoms (EPS) and anticholinergic effects, particularly in older patients. Open trials of atypical antipsychotics (e.g., risperidone, olanzapine) appear promising in this population.

Participants in this randomized double-blind study were 55 years or older, institutionalized, and diagnosed with dementia. Patients were excluded if they had other psychiatric disorders, ECG or lab abnormalities, or clinically relevant neurologic disease. After a one-week washout period, patients were randomized in a double-blind fashion to placebo, risperidone, or haloperidol. Medication was started at 0.25 mg bid, increased every four days if indicated to 1 mg bid; if the patient suffered no side effects and symptoms persisted, the dose could be increased to 2 mg bid. Lorazepam was allowed concurrently. Patients were evaluated at the time of screening, baseline, and weeks 1, 2, 4, 6, 8, 10, and 12. Efficacy measurements included the BEHAVE-AD (25 items; symptoms clusters of psychosis, activity, aggression, diurnal disruption, mood, and anxiety/phobias; 4-point severity scale) and the Cohen-Mansfield Agitation Inventory (CMAI; 29 items; 7-point severity scale). The primary endpoint was defined a priori as the percentage of patients with 30% reduction on the BEHAVE-AD at week 12. EPS was measured by structured examination and Mini Mental Status Examination (MMSE) scores were also serially measured.

A total of 371 patients (56% women) with dementia of the Alzheimer type (67%), vascular dementia (26%), or mixed dementia (7%) were enrolled. The participants had a median age of 81 years, a median duration of institutionalization of four months, and a median time of 4.3 years between onset of dementia and trial entry. Baseline BEHAVE-AD, CMAI, and MMSE scores were 16, 26, and 8, respectively, indicating severe dementia and agitation. A total of 223 (68 risperidone, 81 haloperidol, and 74 placebo) completed the trial, with 70 days as the mean duration of treatment, and 1.2 mg as the mean dose of medication; there were no significant between-group differences. Adverse effects were reported by 76.5%, 80%, and 72.8% of patients in the risperidone, haloperidol, and placebo groups; only somnolence was higher in the treatment groups compared to the placebo group (18% greater with haloperidol and 12.2% greater with risperidone).

The percentage of patients who achieved 30% reduction in the BEHAVE-AD in the risperidone, haloperidol, and placebo groups was 55%, 63%, and 47%, respectively, at endpoint and 72%, 69%, and 61% at week 12. The risperidone group showed significantly greater improvement than placebo in the mean BEHAVE-AD total score at week 12, with the effect seen as early as week two and sustained for the duration; likewise, CMAI scores were significantly reduced. The treatment effect with risperidone was seen for patients with Alzheimer and vascular types of dementia. The haloperidol group showed significant improvement in the BEHAVE-AD aggression score, but not in the total score for completers, compared to placebo. Interestingly, the BEHAVE-AD total score was significantly lower for the haloperidol group compared to placebo when the data were analyzed with respect to all endpoints (i.e., treatment completers and non-completers). The haloperidol group, but not the risperidone group, had significantly more EPS than the placebo group. Lorazepam use in the three groups was similar. An ANOVA model revealed the change in aggression was explained partially by a direct treatment effect and partially by a change in psychosis. Limitations included the arbitrary criteria used to judge response, the concurrent use of lorazepam, and the high response rate in the placebo group.

Comment by Donald M. Hilty, MD

Demented patients with behavioral disturbances are difficult to treat. Psychosocial interventions include ongoing orientation to the environment, minimizing changes in routine, minimizing interruptions of sleep, and staff education. Medications are often necessary, but the response to trials is modest (e.g., 30%-40%, in contrast to a rate of 60+% for antidepressants for patients with major depression). Medications currently under evaluation for behavioral disturbance associated with dementia include divalproex, buspirone, and atypical antipsychotics. These options appear to be advantageous in terms of side effect profile compared to typical neuroleptics. This study indicates that risperidone is well-tolerated and has a direct treatment effect for these patients (and not just an effect via reduced psychosis). It is unclear why the conflicting results occurred with the haloperidol group. More data are needed to confirm the results of this study.