In early-stage Alzheimer’s disease, Mrs. P enters a nursing home because her daughter, who usually takes care of her, is hospitalized for cancer chemotherapy. Mrs. P promptly develops paranoid delusions and refuses her medications for high blood pressure and high cholesterol. What are the treatment options? Is any one antipsychotic safer than another?
Writing an antipsychotic prescription for patients such as Mrs. P is no longer a quick scribble. First we learned that atypicals may alter glucose and lipid metabolism in clinically troublesome ways.1,2 Then we learned that antipsychotics can triple the risk for sudden death in older patients with dementia. (See FDA advisory, Related resources.)
How great are the risks, who is at risk, and how strong is the evidence for these new risks? The debate is not yet settled, but the boundaries of good practice for antipsychotic use in older patients are being redrawn. This is particularly true for those with dementia, in whom antipsychotics’ risk/benefit ratio is higher than for older patients with schizophrenia or bipolar disorder.
Unproven effectiveness
Antipsychotics are not FDA-approved for treating dementia-related psychosis. Though antipsychotics are commonly used off-label to treat behavioral disturbances in the elderly with dementia, no standard of care exists for managing these symptoms with drugs. So far, the evidence for antipsychotics’ effectiveness for dementia’s behavioral and psychological symptoms is spotty at best.
Sink et al (Table 1)3 systematically reviewed 14 randomized, controlled trials and concluded “there is no clear evidence that typical antipsychotics are useful for treating neuropsychiatric symptoms [of dementia].” They concluded from 6 studies of atypicals that only olanzapine and risperidone had shown efficacy, but the effects were “modest and further complicated by risk of stroke.” When the benefits are modest, the risks are more difficult to justify.
When medication is necessary, on the other hand, the Expert Consensus Panel for Using Antipsychotic Drugs in Older Patients reported in 2004 that “for agitated dementia with delusions, the experts’ first-line recommendation is an antipsychotic drug alone…. Risperidone (0.5 to 2.0 mg/day) was first line, followed by quetiapine (50 to 150 mg/day) and olanzapine (5.0 to 7.5 mg/day) as high second-line options.”4
CATIE studies. The definitive prospective study of antipsychotics’ effectiveness in dementia has not been completed. The National Institute of Mental Health is sponsoring CATIE—Clinical Antipsychotic Trials of Intervention Effectiveness—a multi-site research program comparing the effectiveness and outcomes of antipsychotics in treating schizophrenia and Alzheimer’s disease. Results of the Alzheimer disease arm5 are expected next year.
The schizophrenia arm comparing four atypical antipsychotics (quetiapine, risperidone, ziprasidone, and olanzapine) and one typical antipsychotic (perphenazine) found:
- Typical and atypical antipsychotics were similarly effective in 1,493 patients with chronic schizophrenia
- 74% of patients discontinued assigned medications before 18 months for lack of efficacy, intolerable side effects, or other reasons.6
Table 1
Timeline: Evidence on risks, efficacy of atypical antipsychotics
| Year | Summary of study findings, FDA warnings |
|---|---|
| 2002 | Higher incidence of stroke seen with risperidone than with placebo in two of four clinical trials (Wooltorton7) Health Canada and Janssen-Ortho warn Canadian physicians of possible link between risperidone and stroke |
| 2003 | FDA warns of increased risk of stroke with risperidone |
| 2004 | Threefold increased risk of sudden cardiac death associated with antipsychotic use in patients age >65, most without dementia (Straus et al11) |
| 2005 | Stroke risk reported no greater in older patients who took atypicals than in those who took typical antipsychotics (Gill et al10) Analysis of 14 controlled trials finds “no clear evidence” that typical antipsychotics are effective in dementia; atypicals’ effects seen as “modest” (Sink et al3) FDA issues warning after finding 60% increase in risk of sudden death in review of 17 trials in older patients receiving atypical antipsychotics for dementia Efficacy of conventional and atypical antipsychotics found similar in patients with chronic schizophrenia (Lieberman et al6) |
Stroke risk
How strong is the evidence for stroke risk among older patients who take antipsychotics? Wooltorton7 first raised concern about increased risk of stroke with risperidone in 2002 in a summary of four clinical trials. Though none was designed to examine stroke risk as the primary outcome, two showed significantly higher incidence of cerebrovascular events with risperidone than with placebo. The stroke rate with risperidone was double that with placebo (4% vs 2%) across the total 1,200 subjects in the four studies.
This preliminary report led to an FDA advisory but surprisingly no definitive studies or systematic reviews. No epidemiologic studies have examined stroke rates in those who take antipsychotics compared with those who don’t, while controlling for common stroke predictors. So we have a warning based on post hoc analyses in two positive and two negative studies, but no sound estimate of how much antipsychotic use in general adds to the risk of stroke.