On April 12, 2005, the U.S. Food and Drug Administration (FDA) issued a public health advisory on a newly identified concern associated with the off-label use of second-generation (atypical) antipsychotic medications for the treatment of dementia-related behavioral disorders in older adults. This advisory, based on an FDA analysis of data from 17 randomized, controlled trials that enrolled 5377 older adults with dementia-related behavioral disorders, showed that the risk of death in the drug-treated patients was 1.6 to 1.7 compared with that of the placebo group. Treatments consisted of aripiprazole, olanzapine, quetiapine, or risperidone. These trials averaged about 10 weeks. The rate of death was about 4.5% in drug-treated patients and about 2.6% in the placebo group. Most of the deaths appeared to be either cardiovascular (eg, heart failure, sudden death) or infections (eg, pneumonia) in nature.
Several studies, summarized in a Cochrane Review and in a systematic review published in the Journal of the American Medical Association (JAMA) cite the clinically modest but significant reductions in NPS demonstrated among elderly subjects treated with olanzapine or risperidone. To date, while some efficacy has been noted in published placebo-controlled trials conducted with non-antipsychotic medications such as carbamazepine, citalopram, donepezil, galantamine, or memantine, this data is even more limited than that of the second-generation antipsychotics.
There has also been evidence published from an extensive Centers for Medicare and Medicaid (CMS) database that second-generation antipsychotic medications, compared to first-generation (typical) antipsychotics, do not appear to be associated with an increased risk of ventricular arrhythmias or cardiac arrest.
Two observational epidemiological studies were published that examined the risk of death in patients who were treated with first-generation antipsychotic medications:
- Gill, et al, performed a retrospective cohort study in Ontario, Canada of 27,259 adults, 66 years of age or older, with a diagnosis of dementia between April 1997 and March 2002. The investigators compared the risk for death with use of a second-generation antipsychotic versus no antipsychotic and the risk for death with use of a first-generation antipsychotic versus a second-genration antipsychotic. They found that second-generation antipsychotics were associated tih increased mortality as compared to no antipsychotic use as early as 30 days and persisting until study end at 180 days. The investigators found that first-generation antipsychotic use showed a marginally higher risk of death compared with second-generation antipsychotic use. The causes of death were not reported in the study.
- Schneeweiss, et al, performed a retrospective cohort study in British Columbia, Canada of 37,241 adults, 65 years of age or older, who were prescribed first-generation- (12,882) or second-generation (24,359) antipsychotic medications for any reason between January 1996 and December 2004. The investigators compared the 180-day all cause mortality with use of a first-generation antipsychotic versus a second-generation antipsychotic. They found that the risk of death in the first group of patients treated with first-generation antipsychotic medications was comparable to, or possibly greater than, the risk of death in the group of patients treated with second-generation antipsychotic medications. The causes of death with the highest relative risk were cancer and cardiac disease.
The FDA considers that the methodological limitations in these two studies preclude any conclusion that first-generation antipsychotics have a greater risk of death with use than second-generation antipsychotics. The FDA has determined, however, that the overall weight of evidence indicates that first-generation antipsychotics share the increased risk of death in elderly patients with dementia-related psychosis that has been observed for second-generation antipsychotics. Therefore, the prescribing information for all antipsychotic drugs now includes the same information about this risk in a Boxed Warning and the Warnings section.
A recent five-year retrospective study by Rossom, et al, using U.S. Veterans Administration data from more than 89,000 veterans did not find an increased risk of death in veterans with dementia who were prescribed lower doses of olanzapine (<2.5 mg/d), quetiapine (<50 mg/d), or risperidone (<1 mg/d). However, at typically prescribed doses, second-generation antipsychotics (excluding quetiapine) were associated with an increased risk of death. All doses of haloperidol were associated with increased mortality.
The mechanism(s) for increased mortality associated with antipsychotic use remain uncertain and needs careful examination. One study approaching this problem comes from another recent large epidemiological five-year retrospective nested case-control study of primary care patients int eh United Kingdom, by Parker, et al. They found that there was a 32% greater risk for venous thromboembolism (VTE) in those prescribed antipsychotics in the previous 24 months vs nonantipsychotic users. The risk was even greater for new users and for those taking second-generation antipsychotics. The study examined risks by type of antipsychotic, potency and dose, and adjusted for comorbidity and concomitant drug exposer. A total of 25,532 “eligible cases” were selected for this study and a total of 89,491 matched healthy controls were also included. The median age for all participants was 67 years. Understanding mechanisms for adverse events leading to mortality should allow risk profiling of patients and eventually lead to approaches to minimize risk in those patients who otherwise need and benefit from this therapy. (See also References.)