TCPR: Dr. Jeste, I know that you were responsible for some of the original research demonstrating that tardive dyskinesia is a possible side effect of antipsychotics. Why were you interested in the topic?
Dr. Jeste: At that time, I was a research fellow at NIMH, and had done a neurology residency at George Washington University. I was interested in studying tardive dyskinesia because it was fascinating to me that a persistent movement disorder could be produced by a drug, which implied that these antipsychotics produced a structural brain disorder. There was no other example of a potentially irreversible movement disorder caused by therapeutic agents in all of medicine. I ended up co-writing a book about it (Dilip Jeste and Richard Wyatt, Understanding and Treating Tardive Dyskinesia, Guilford Press, 1982).
TCPR: I understand that there was quite a bit of resistance in the psychiatric community to the idea that drugs caused TD.
Dr. Jeste: Yes, at that time, people frowned on the idea of TD being caused by drugs; in fact, many people thought it was a symptom of schizophrenia, because some patients with schizophrenia who had not taken antipsychotics developed dyskinesias. And every time a new drug came out, the company said that this drug had not been shown to cause TD. But we know that TD takes months or years to develop.
TCPR: So how did you convince the community that drugs caused TD?
Dr. Jeste: I showed in my literature reviews and through my own research that the incidence of TD was increasing with time and this coincided with the increasing use of antipsychotics.
TCPR: What is the risk of TD with the older agents?
Dr. Jeste: It depends on the age of the patient. In the younger population, the risk is about 5% per year, but in older patients, it is about 30% per year. High potency neuroleptics such as Haldol have a somewhat higher risk than low potency neuroleptics such as Thorazine.
TCPR: What about the TD risk for atypical medications?
Dr. Jeste: The atypicals – even clozapine – are not free of the risk of TD, as all of them are dopamine blockers. At the same time, partly because these newer agents also cause some serotonin blockade, the risk is significantly lower. The risk of TD is generally correlated with the degree to which they produce extra-pyramidal symptoms. Thus, among the atypicals, risperidone may have the highest risk of TD, followed by olanzapine and aripiprazole. Quetiapine, ziprasidone, and clozapine probably have the lowest risk of TD, although they have other side effects such as sedation and postural hypotension with quetiapine. The data on clozapine and ziprasidone in elderly people with Alzheimer’s disease are very limited.
TCPR: Is the risk of EPS and TD related to the dose used?
Dr. Jeste: Yes, but this is especially true for EPS in older people. In the older population, there is a pretty dramatic dose-response curve, and you need to determine the dose that is most effective without leading to excessive EPS. Using these curves, we find that, in patients with psychosis of dementia, optimal dose of risperidone is 1 mg/day, olanzapine 5 mg/day, aripiprazole 10 mg/day, and for quetiapine there is a much broader range from 50-200 mg/day.
TCPR: In your experience, which are the atypicals most likely to cause akathisia?
Dr. Jeste: Aripiprazole is probably the most likely among the atypicals to produce akathisia, followed by risperidone. The risk is, however, much greater with conventional agents like haloperidol.
TCPR: What about the recent black box warnings about mortality caused by antipsychotics in the elderly?
Dr. Jeste: Since it gets confusing to keep track of all these warnings, I like to break it down into two issues: stroke and mortality. Regarding the risk of strokes, this concern first surfaced about risperidone, when a placebo-controlled trial of 300 elderly subjects showed that while risperidone was better than placebo in patients with aggression, agitation, and psychosis of dementia, there were six cases of TIA (transient ischemic attack) or stroke in the risperidone group and none in the placebo group (Brodaty H, et al., J Clin Psychiatry 2003, 64:134-143). At the time, nobody paid much attention because this difference was not statistically significant. But when Janssen combined the data for all placebo-controlled trials they found that risperidone was associated with a significantly higher incidence of stroke than placebo. A year later, the same trend was reported for olanzapine, and a bit later, for aripiprazole. So currently, there are black box warnings regarding a higher risk of stroke associated with risperidone, olanzapine, and aripiprazole in patients with dementia.
TCPR: Are quetiapine and ziprasidone safer?
Dr. Jeste: It is hard to be sure. It might be that there are not as many controlled trials on these two drugs (especially ziprasidone) as on the others in this population.
TCPR: What about the risk of higher mortality?
Dr. Jeste: When you look at each drug individually, there is no mortality difference between drug and placebo. But when you combine data for four drugs – risperidone, olanzapine, quetiapine, and aripiprazole – the mortality rate is 1.6 to 1.7 times higher than with placebo. The average death rate in the medication group was 4.5% vs. 2.6% in the placebo group. As this was based on combined data, the FDA felt it merited a black box warning about mortality for all atypical antipsychotics in dementia patients.
TCPR: What was the cause of death in these patients?
Dr. Jeste: The most common causes of death were heart failure and pneumonia, which are also the most common baseline causes of death in the patients with dementia.
TCPR: So how should we use these findings in our treatment decisions with the elderly?
Dr. Jeste: We certainly have to take these black box warnings seriously. On the other hand, psychosis of dementia is common and can be hard to treat. About 15% of people over 65 have dementia, half of them have psychosis, and many have severe agitation. There is no FDA-approved drug for agitation and psychosis in dementia, so any drug use is off label. There are more data on antipsychotics than on any other class of psychotropic drugs in this population, and the studies suggest that they are effective for agitation, though not for psychosis.
TCPR: Really? None of the antipsychotics are effective for dementia-related psychosis? That seems surprising.
Dr. Jeste: Yes it is, and this may imply that not all psychosis is the same. It could well be that the psychosis of dementia is neurobiologically different from the psychosis of schizophrenia. The other issue is that it’s often difficult to measure psychosis in this population. If a patient is already confused and disoriented due to dementia, it may be hard to ascertain the degree of psychosis, because these patients cannot verbalize their symptoms well.
TCPR: If these drugs have been shown effective for agitation in dementia, why are they not approved?
Dr. Jeste: In order to get approval for agitation you have to show that your drug affects agitation in three different conditions. That is because agitation is a symptom and not a disorder per se (unlike schizophrenia or major depression).
TCPR: In your clinical practice, how do you manage agitation or psychosis in dementia?
Dr. Jeste: I don’t think about the drugs first, I think about cause. Agitation in an elderly person may be caused by a delirium due to an infection, such as a urinary tract infection. There are also often psychosocial reasons. For example, you might find out that the patient becomes agitated whenever his favorite daughter leaves town for a conference. I also evaluate whether the agitation or psychosis is really severe enough to merit treatment. A patient might have mild delusions that are not interfering with his or her functioning, in which case there may be no reason to use drugs.
TCPR: Assuming you’ve ruled out or treated other causes of agitation, what meds do you turn to?
Dr. Jeste: My number one pharmacologic choice in patients with severe psychosis and agitation is one of the atypical antipsychotics. The conventional agents are not useful in this population because of the very high risk of EPS and TD. As far as specific choice of antipsychotic, this is usually guided by which side effects I want to avoid in particular patients. For example, I’ll avoid olanzapine in patients with diabetes, I’ll avoid quetiapine in patients who are already sedated, and I’ll avoid risperidone in patients who have Parkinson’s Disease. It is a process of elimination because these drugs are not significantly different from one another in terms of efficacy. I will also follow the usual wisdom in treating the elderly, which is to start low and go slow. For example, I will start with risperidone at 0.25 to 0.5 mg/day. Finally, I will generally try to taper patients off of antipsychotics as soon as possible, generally within a few weeks. I keep in mind the fact that the studies showing increased mortality lasted about 10 weeks. The tapering should be done slowly so as to prevent adverse effects of sudden withdrawal and to assess the need for continued medication at the same dose. My watchwords for pharmacotherapy in elderly patients are: caution, monitoring, and shared decision making (getting the patients and caregivers involved in treatment decisions).