This week, a more than decade-old debate concerning the use of antipsychotic drugs in the elderly with dementia flared up after a study showing that these drugs also increased the risk of death in patients with Parkinson’s disease was published in the journal JAMA Neurology.
Authored by Dr. Daniel Weintraub, an expert on geriatric psychiatry at the Veterans Affairs Parkinson’s Disease Research, Education and Clinical Center in Pennsylvania, the study, “Association of Antipsychotic Use With Mortality Risk in Patients With Parkinson Disease,” echoed earlier findings.
But Weintraub also highlighted the difficult trade-off physicians face when deciding on psychosis treatment in Parkinson’s patients, especially in the face of ambiguous evidence.
Reports published years ago demonstrated that the use of antipsychotic drugs in elderly patients with dementia increased the likelihood of death. This led the U.S. FDA to require a black box warning adjoining the label on antipsychotic drugs. The label states the drugs are not approved for treatment of psychosis in elderly patients with dementia, since such treatment is linked to an increased risk of death.
A 2005 warning including newer generation drugs, generally referred to as atypical antipsychotics, was extended in 2008 to also include older generation, so-called typical drugs.
In Parkinson’s disease, dementia affects up to 80 percent of patients, and 60 percent also experience psychotic symptoms, making the new study especially appropriate.
Although a smaller study published last year reported similar findings, the new report adds significantly to the evidence by its sheer size. Exploring differences in survival in 7,877 treated, and 7,877 untreated, Parkinson’s patients registered in the Veterans Health Administration database from 1999 to 2010, there was plenty of data from which to draw conclusions.
For the retrospective analysis, each patient taking antipsychotic medication was paired with an untreated patient matched by age, sex, and a range of clinical characteristics.
The study showed that antipsychotic use increased the risk of death twice as much after 180 days of treatment, with patients using typical antipsychotics experiencing the highest risk — almost four times more than untreated patients. Those on atypical medication were more than twice at risk (2.3 times) compared to untreated participants.
The findings send a clear message to physicians: Use extreme caution when treating Parkinson’s patients with antipsychotic drugs. The debate the article prompted, however, shows a more nuanced picture of the situation.
In a comment to the article, published by a trio from the University of Florida — Irene Malaty, Michael Okun, and Michael Jaffee, all experts in the field — it became clear the findings warrant additional discussion, particularly of the interpretation of the data.
Unlike the earlier study, the report did not compare untreated psychotic Parkinson’s patients with those receiving antipsychotic medication; there is only a comparison of drug use. Earlier studies, they point out, have shown that psychosis itself is linked to a greater risk of death, independent of drug use.
Parkinson’s is a disease characterized by the death of neurons producing the signaling factor dopamine, and treatment aims to improve dopamine signaling in the brain. Unfortunately, all antipsychotic drugs on the market work in the opposite direction, by decreasing dopamine signaling. This often leads to Parkinson’s-like side effects in psychotic patients without a Parkinson’s diagnosis.
The main cause of death in the study was Parkinson’s disease, and the authors argue that the drugs, by virtue of their actions on dopamine neurons, might have worsened the disease, hastening death.
The idea is supported by the observation that untreated Parkinson’s patients most often died of cardiovascular disease, which is emphasized in an editorial in the same journal that also attempts to bring a more nuanced view to the discussion.
Titled “Antipsychotics and Increased Mortality — Are We Sure?” the editorial, authored by Mark Baron — yet another long-time expert on Parkinson’s disease — points out that the available data can by no means distinguish whether the increased death risk is a result of the underlying psychosis, or caused by its treatment.
Baron, director of the Southeast Veterans Affairs Parkinson’s Disease Research, Education, and Clinical Center in Virginia, also noted that the study likely included a large number of patients where a Parkinson’s diagnosis might have been questionable, since detailed information on diagnostic procedures were lacking. This would make it difficult to conclude if any findings were really influenced by the diagnosis of Parkinson’s disease.
Another study exploring antipsychotic use in elderly Parkinson’s patients and elderly people without a neurological disorder, suggested that such drug use was linked to death independent of diagnosis — a finding supported by other data showing that antipsychotic use in the elderly is a risky practice.
Baron also underscored another crucial point in the editorial — the scarcity of other options. Weighing a possible risk of death against any potential benefits the treatment might have is a dilemma physicians deal with every day, particularly when faced with a severely psychotic patient and the reactions of family members to the black box warning label accompanying such drugs.
Malaty and her colleagues seem to have made up their minds on this point, stating they suspect that a risk-benefit trade-off would likely favor treatment in most patients.
A new drug, Nuplazid (pimavanserin), particularly tailored to treat psychosis in Parkinson’s, was approved earlier this year by the FDA, but its hefty price tag is likely to limit its use. In addition, all studies so far have shown that newer, atypical antipsychotics such as quetiapine pose a smaller risk than older drugs. The drug Clozapin stands out as an option with the least impact on motor symptoms. Less risky options are unavailable, and would need more exploration.
Weintraub, who also serves as an associate professor of psychiatry at Perelman School of Medicine at the University of Pennsylvania, replied swiftly to the comments. He said that while the study could not compare treated and untreated psychotic Parkinson’s patients, statistical tools, adjusting the analysis for the presence of psychosis, did not indicate that it was the psychotic disease itself that was responsible for the increased death rate.
Given that the study suffers an equal lack of certainty of psychiatric diagnosis as Baron pointed out for the diagnosis of Parkinson’s in the first place, this does not seem to be a convincing argument.
Stating that only about half of psychotic patients are treated with antipsychotics, and therefore it is not evident that these patients need antipsychotics, he also rebuked Malaty’s statement of a risk-benefit analysis favoring treatment.
Although Weintraub admits that the study presents no final statement about the risk of antipsychotic drugs in Parkinson’s disease patients, he concurred with the others that more research is needed to understand the risks of the treatment — with clinical trials following patients over time as the most sensible option to get more decisive answers on this sensitive subject.