Atypical Antipsychotics Increase Risk of Death for Dementia Patients

October 25, 2005

Pharmaceutical Executive

Pharmaceutical Executive, Pharmaceutical Executive-10-25-2005, Volume 0, Issue 0

A new meta-analysis confirms the risk of atypical antipsychotics noted by FDA earlier this year. But the reasons for this increased incidence of death remain unclear.

A new meta-analysis has brought to light some of the statistical evidence that lead to FDA’s April black-box warning on the dangers of atypical antipsychotics for dementia patients. Published in the Oct. 19 issue of the Journal of the American Medical Association, the study indicates that atypical antipsychotics were associated with a slight increase in death among dementia patients. The analysis of 15 clinical trials found 3.5 percent of dementia patients using these drugs died, compared with 2.3 percent who were on placebos.

    The risks associated of these drugs deserve further analysis, according to experts. But they caution against abandoning use of these drugs in light of these results. “This has to be looked at in perspective,” said lead author Lon Schneider, a professor of psychiatry, neurology, and gerontology at the University of Southern California’s Keck School of Medicine.

    Although not a treatment for dementia itself, antipsychotic drugs are frequently used to treat aggression and delusions associated with the condition, Schneider explained He said these behaviors occur in the majority of dementia patients and are associated with worsening dementia, rapid cognitive decline and earlier mortality.

    According to an accompanying editorial by Peter Rabins and Constantine Lyketsos of Johns Hopkins Medical Institutions, the conditions treated by antipsychotics are present in 60 to 90 percent of dementia patients. In fact, 25 percent of nursing home residents are treated with antipsychotics, they said.

    The trials used in the analysis tested the efficacy of atypical antipsychotics, which were developed in the 1990s. But two included comparisons with an older conventional antipsychotic, haloperidol, which was also associated with an increased risk of death.

Potential Causes

The authors obtained their data from reports of serious adverse events in clinical trials, often unpublished or reported only as posters at conferences, Schneider explained. But individual patient data were not available, the authors said, and data on causes, medical conditions, and other medications, among other things were insufficient for fully understanding the results.

    For example, it is difficult to pinpoint the mechanisms that link antipsychotics to death. Schneider speculated that some patients might have received high doses and become sedated. This can lead to health problems for elderly people, including heart failure, pneumonia, urinary tract infections, edema, deep vein thrombosis, and pulmonary embolism, especially if they are lying down for a long time.

    In the 17 trials examined by FDA, before the agency issued its April advisory, most of the deaths recorded were heart related or caused by infections, primarily pneumonia.

    If sedation proves to be the cause of at least some of these deaths, the risks associated with antipsychotics can be decreased by lowering doses.

    Rabins agreed that sedation could have caused some of the deaths. He indicated that these drugs might also increase the risk of patients falling down, which could lead to death. Another possibility, Rabins noted, is that atypical antipsychotics have been associated with changes in the rate at which the heart conducts electrical impulses.

    The study’s authors expressed hope that the full trial results would be made available for future study, in order to better understand how these patients died.

    “An individual patient meta-analysis might be able to identify characteristics associated with mortality, potentially due to drugs,” they wrote.

Clinical Impact

Schneider and his colleagues recommended in the paper that doctors use a two- to four- week clinical response window for antipsychotics. If a dementia patient is not showing signs of improvement by four weeks, the doctor should discontinue the drug to decrease the risks.

    Rabins responded that four weeks might not be enough time to determine efficacy. The patient might need as much as three months. Further study is necessary to determine the right time period, he said.

    In an interview, Rabins emphasized that these results should not prevent people from prescribing antipsychotics for dementia patients. The modest mortality risk is outweighed in many cases by the benefits if symptoms are causing the patient distress or putting caregivers at risk of harm from aggressive behavior. He said that there is a slight tendency to over-prescribe antipsychotics for insomnia or anxiety, and hopefully these findings may eliminate that practice.

    “These medications clearly have an appropriate use and I think there is adequate data to support that,” said Rabins, a professor of psychiatry and behavioral sciences. “This just puts a little more of a burden on the risks and benefits.”

    Schneider noted that the alternatives to antipsychotics – antidepressants, benzodiazepines and anticonvulsants – all carry risks of their own and do not have documented proof of being effective for dementia patients.

“Antipsychotics have offered the best evidence for efficacy,” he said via email. “There is an absence of evidence for the efficacy of these other drugs.”