Psychosis in Parkinson’s disease: unexplained observations in a seemingly simple model

KEYWORDS:

• Parkinson’s disease
• Parkinson’s disease psychosis
• dementia with Lewy bodies
• hallucinations
• delusions
• pimavanserin
• clozapine
• atypical antipsychotic drugs

It is difficult to imagine a more interesting disorder than Parkinson’s disease (PD).[1] There are motor problems, some of which are generally dopamine responsive (bradykinesia and rigidity), others which are not (posture, speech, balance, freezing of gait) and tremor, whose response is more varied. There is autonomic dysfunction, which can range in severity from being subclinical to debilitating, and is not dopamine related.

And then there is a wide variety of phenomenologically fascinating behavioral problems, both intrinsic and iatrogenic, of which psychosis may be the most fascinating.[2] In addition, it holds keys, as yet not understood, to the better understanding of psychotic symptoms in general, and the role that dopamine plays. The recent introduction of pimavanserin, a serotonin inverse agonist, has added another layer of complexity to our currently limited understanding. In this editorial, we highlight the fascinating questions that remain unanswered about the pathophysiology of PD in general and PD psychosis (PDP) in particular. Our expert opinion about management of PDP has been published previously.[3]

Psychotic symptoms, hallucinations and delusions, are the hallmark features of PD psychosis (PDP),[2,4] as well as dementia with Lewy bodies (DLB).[5] About 30% of drug-treated PD patients have visual hallucinations, and a much smaller number of untreated patients may have them as well.[6] Auditory and other hallucinations occur less commonly.[4] The lifetime prevalence of visual hallucinations is approximately 50%.[6] The hallucinations are quite stereotypic, usually involving people or pet animals, ignoring the patient, and usually the same people or animals appear each time.[7] These hallucinations are typically visual and have no emotional content, whereas those in primary psychiatric disorders are almost entirely auditory and generally ego syntonic, reflecting the patient’s underlying affect at the moment. Delusions affect about 5% of drug-treated patients and these are usually paranoid in nature, frequently of spousal infidelity or abandonment.[4] These problems are more stressful for caregivers than motor dysfunction and are the main precipitants for nursing home placement in the U.S.[8] The phenomenology of psychotic symptoms seems to be exactly the same as that seen in dementia with Lewy bodies (DLB)[2,5] and the sensitivity to drugs that block dopamine receptors appears also to be about the same as in DLB. The major risk factors for PDP symptoms are dementia and PD medications.

What is interesting about PDP? While schizophrenia has often been hypothesized to be a dopamine-excess disorder, based on its response to anti-dopaminergic drugs, the psychosis in PD induced by dopamine agents is quite different than schizophrenia. PD patients are a testing ground – ‘canary in the mine shaft’ [9] – for parkinsonian side effects of antipsychotic drugs that block dopamine receptors. If a drug causes PD patients to worsen, it will undoubtedly cause some neurologically normal patients taking the drug to develop parkinsonism, which should be a warning to doctors who prescribe antipsychotic drugs for depression. Clozapine, at doses of one to two orders of magnitude less than used in schizophrenia have not only been found to improve PDP without worsening motor function,[10,11] but is far more effective for this than it is for schizophrenia or bipolar disease, and without the metabolic effects. And, for unknown reasons, not likely related to anticholinergic effects, improves tremor, often dramatically.[12] More puzzling is the observation that in several double blind, placebo-controlled trials, quetiapine does not have any effects on motor function but does not improve the psychosis,[13,14] and olanzapine worsens motor function and does not improve psychosis.[15] Melperone, like quetiapine, does not improve psychosis although is well tolerated.[16]

The exciting development in this field was the positive trial of pimavanserin,[17] a 5HT2A inverse agonist. It improved psychosis and had a side effect profile similar to placebo without motor side effects. This drug was developed based on the observation that clozapine, and the other second generation antipsychotic drugs blocked both dopamine D2 receptors and serotonin 5HT2A receptors. Of course, clozapine has effects on multiple neurotransmitters. Previous trials of pimavanserin for PDP yielded suggestive benefits, but failed to meet defined criteria for benefit.[18] Unlike the clozapine trials, where the benefits were almost immediate, pimavanserin required over two weeks to work. Whether it works as well as clozapine remains to be seen. Currently it is under fast track review by the FDA.

The interesting questions that remain are why clozapine improves PDP but not olanzapine, quetiapine, or melperone; why are clozapine, quetiapine, and melperone free from worsening parkinsonism, but not other atypical antipsychotics; what is the relationship between dopaminergic agents that cause psychosis and pimavanserin which does not affect dopamine; and why should pimavanserin take so much longer than clozapine to work? Other questions to consider are why the hallucinations in DLB patients who are not taking psychoactive medication are identical to the hallucinations that occur with dopaminergic agents in people with PD? Why is the dopamine-induced psychotic syndrome the same as that occurring in DLB, a presumably dopamine deficiency state? Why are the hallucinations induced by non-dopamine medication like amantadine and anticholinergics the same as seen with dopamine-stimulating drugs?[19] And why is the occurrence of dopamine-induced hallucinations more closely related to level of attention and stimulation rather than the serum drug level?[20] Why does effectiveness differ for some atypical antipsychotics (e.g. clozapine is effective; quetiapine and olanzapine are not) when they all share an affinity for the 5HT2A receptors?

We think that the answers to these questions are unknown and that no theory adequately explains almost any of it. The truth, we believe, is that we know a lot less about brain function than we think we do and that our current conception of dopamine’s involvement in psychosis is extremely oversimplified.

Declaration of Interest

J.H. Friedman: Abbvie; Avid, Acadia, Teva, Demos Press; Cambridge University Press. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.