Clinical manifestations

Presentation and course

Clinical concepts and problems. Parkinson disease has traditionally been regarded primarily as a motor disorder, but nonmotor symptoms such as cognitive impairment and dementia are common associated features. On sensitive psychometric tests, virtually all patients with Parkinson disease are cognitively impaired relative to age-matched controls, even early in the course of disease. In early stages of Parkinson disease, cognitive deficits are mild and of variable clinical significance. Mild cognitive impairment in Parkinson disease patients is prevalent, and appears to predispose to development of Parkinson disease dementia, and has defined diagnostic criteria (79). Dementia is more likely to appear in Parkinson disease patients with the postural instability-gait disorder (PIGD) phenotype (77). Although classic deficits in Parkinson disease dementia include dysfunction predominantly in executive, attentional, visuospatial, and memory domains, evaluating cognition and interpreting performance in patients with Parkinson disease can be challenging because noncognitive factors may come into play. Depressed mood, confusion, hallucination, and agitation may affect motivation and attention and consequent performance on cognitive testing (135). Speech (hypophonia) and motor disturbances (bradykinesia and tremor) may impair the speed and quality of responses. Cognitive performance may vary in patients who fluctuate on dopaminergic medications and can also be affected directly by polypharmacy. In the "off" state, patients may be severely akinetic, rigid, and hypophonic, whereas at a peak dose in the "on" state, dyskinesias may impede performance, and cognitive performance can be affected directly in both “off” and “on” states (39).

Alzheimer disease, the most common type of dementia, is considered to be the archetypal cortical dementia. Decline in memory acquisition is the earliest and most prominent feature, whereas deficits in language, calculation, constructional tasks, and problem-solving develop later as the illness progresses. In Parkinson disease dementia or dementia with Lewy bodies, brain pathology often includes subcortical (and some cortical) pathology, but may also include Alzheimer disease amyloid plaques and neurofibrillary tangles, producing a more complex or cortical pattern of dementia (40).

Clinical characteristics of Parkinson disease dementia. There are several clinical risk factors for developing Parkinson disease dementia including advanced age and late age at the onset of Parkinson disease (especially after 70 years), mild cognitive impairment (79; 133), nontremor predominant Parkinson disease (37), postural instability and worse gait disturbance (77; 78; 85), early orthostatic hypotension (15), as well as severe motor findings (103), coexisting depression, early executive dysfunction (85), and early onset of hallucinations (01). Mild behavioral impairment has been associated with cognitive decline and brain atrophy in Parkinson patients (135). Sex, race, and smoking history are not consistent risk factors for the development of dementia in Parkinson disease (01; 04). The risk of dementia increases with the duration of Parkinson disease, being 60% after 12 years of follow up in one study and 83% after follow up of 20 years in another study (34; 105). Patients with the postural instability and gait disorder subtype of Parkinson disease have increased risk of dementia and show more rapid cognitive decline when compared to patients with tremor dominant Parkinson disease (33; 11; 77; 78).

Four main categories of symptoms occur in Parkinson disease dementia: cognitive, motor, behavioral, and autonomic. Cognitive decline typically begins and progresses gradually. Episodes of cognitive clouding are common in patients with Parkinson disease dementia and patients may fluctuate in attention and cognition. The episodes of cognitive fluctuations may arise spontaneously, or they can be precipitated by medications, infections, or dehydration. Autonomic symptoms are often comorbid in the form of constipation, urinary incontinence, orthostatic or postprandial hypotension resulting in recurrent episodes of syncope or falls, heart rate variability, and sexual dysfunction (15). In a comparative study, those with Parkinson disease dementia exhibited more frequent cardiovascular autonomic dysfunction compared to those with Lewy body dementia, vascular dementia, and Alzheimer disease (09).

Behavioral symptoms. Neuropsychiatric symptoms are linked to worse cognition and poor quality of life in Parkinson disease (135; 20). These symptoms include hallucinations, visual misperceptions, delusions, depression, apathy, anxiety, and sleep cycle disturbances including REM sleep behavior disorder and insomnia. About 30% to 40% of patients with Parkinson disease experience chronic dysthymia or depression at some point, which can have a major impact on quality of life (102; 135). Features of motor impairment such as monotonous voice, masked facies, and poverty of movement may mimic psychomotor symptoms of depression. More specific symptoms of depression include feelings of worthlessness, guilt, and loss of appetite. Depression in Parkinson disease does not correlate with dopaminergic deficiency; depressive symptoms are only weakly related to the severity of motor dysfunction or to CSF reduced levels of dopamine metabolites. Depression in Parkinson disease may result from impaired meso-cortico-limbic dopaminergic pathways or to serotoninergic pathways, or they may be associated with frontal lobe dysfunction. Agitation is also common in more severe stages of this dementia. Although behavioral symptoms can result from medication side effects (especially anticholinergic agents), or concurrent medical illnesses, they often lack an external cause. Behavioral symptoms can also fluctuate with medication usage, further complicating their interpretation and treatment.

Visual hallucinations of either simple or complex form are the most common and debilitating psychotic symptoms in Parkinson disease dementia (74). Visual hallucinations have been shown to be specific for Lewy body pathology (Parkinson disease dementia and dementia with Lewy bodies) and occur rarely in progressive supranuclear palsy and multiple system atrophy (131). The incidence of visual hallucinations in Parkinson disease dementia is reported to be over 50% and is similar to that reported in dementia with Lewy bodies (41). In a 20-year longitudinal follow-up study, 74% of patients with Parkinson disease eventually developed visual hallucinations. A Kaplan-Meier plot showed that visual hallucinations and dementia developed concurrently in one longitudinal cohort (62). Patients hallucinate in both "on" and "off" states, and the hallucinations may recur over days, weeks, or even longer. In Parkinson disease dementia, the visual hallucinations are typically well-formed, vivid images of people or animals, and they may incorporate bright colors and dramatic settings. Patients react variably to the visual hallucinations: some are not disturbed and may recognize the images as being unreal, whereas others find the visions to be threatening or distressing. They may incorporate them into delusional beliefs. Risk factors for minor hallucinations include old age, sleep disturbance, treatment with L-dopa, and cognitive impairment. Age and dementia appear to be the strongest risk factors for visual hallucinations; the dose of L-dopa is not necessarily higher in patients with hallucinations, compared to those without (74). If lowering the dose of L-dopaminergic medication eliminates hallucinations, this usually occurs rapidly over a day or two. Compared to patients without minor hallucinations, those with hallucinations are more likely to have atrophy in cortical areas involved in secondary visual processing (74).

Psychometric features of Parkinson disease dementia. Non-demented patients with Parkinson disease may show subtle deficits on specific cognitive tests relative to age-matched controls. These include tests dependent on psychomotor speed, visuospatial abilities, and executive function. A few studies have found more widespread deficits, also affecting areas such as naming and memory. Age may influence the pattern because studies that reported broader deficits usually assessed older patients. There is a continuum of cognitive impairment in Parkinson disease, and it is difficult to assign absolute cutoff points for dementia versus mild cognitive impairment (109).

Patients with Parkinson disease dementia typically show deficits on tests of executive function, attention, visuospatial abilities, constructional abilities, and some impairment of memory (70; 115; 95). Other aspects of cognition, such as language ability, praxis, orientation, long-term memory, and calculation, are relatively preserved. Brief cognitive tests such as the Mini-Mental State Examination (MMSE) do not adequately assess executive function. More detailed tasks that assess frontal abilities require the patient to sustain attention, strategize in order to solve problems, shift cognitive set, ignore distractions, and respond to feedback. For example, in the Wisconsin Card Sorting Test, a test of frontal lobe function, the subject sorts a deck of cards according to three types of sorting rules. The examiner changes the rule periodically; the patient must work out the new sorting rule and shift set accordingly. Patients with Parkinson disease make sorting errors and perseverate. Although visuospatial processing is impaired in cortical dementia, it is often disproportionally affected in Parkinson disease dementia. Patients with Parkinson disease dementia have deficits on tasks involving drawing figures, arranging objects into spatial patterns, mentally manipulating or rotating images, and visuospatial perception and attention (95). The Montreal Cognitive Assessment (MoCA), which tests executive function, has been validated as more sensitive than the MMSE in detecting and differentiating Parkinson disease dementia (65). Of additional diagnostic value is fluctuation of cognitive function in Parkinson disease dementia. Several scales for assessment of cognitive fluctuation are available, and as fluctuations become more frequent and longer, the degree of Parkinson disease dementia becomes more severe and shows a clinical verisimilitude to that of dementia with Lewy bodies (125).

Memory deficits in Parkinson disease dementia may differ from those in Alzheimer disease. In Alzheimer disease, both initial acquisition and delayed recall of a set of stimuli are impaired. Delayed recall is usually impaired out of proportion to the extent of initial acquisition in Parkinson disease. This can be expressed in terms of the delayed recall score as a percentage of the immediate recall score, which is typically low in Alzheimer disease. Patients with Parkinson disease dementia may have difficulty acquiring and encoding new information (poor immediate recall), but they may retain items that were well-learned, leading to a higher percentage of retention (116). Some studies, however, fail to find differences in memory impairment in Parkinson disease dementia compared to Alzheimer disease (72). A longitudinal study demonstrated that Parkinson patients with an amyloid PET burden that is greater than 1 SD over the average experienced a more rapid decline in memory abilities over 5 to 10 years, compared to those who had an average amyloid burden (95). In general, the following functions are worse in Parkinson disease dementia and dementia with Lewy bodies, compared to Alzheimer disease: attention (greater impairment plus increased fluctuations), executive function, and visuospatial construction. Typically, memory (as above) and language function are not as impaired in Parkinson disease dementia compared to Alzheimer dementia or dementia with Lewy bodies (115; 59).

Parkinson disease dementia criteria. Clinical diagnostic criteria for Parkinson disease dementia were established in 2007 (42; 45). Core features include a diagnosis of Parkinson disease and a dementia syndrome leading to impairment in more than one cognitive domain, deficits impairing daily life, and a decline from premorbid functioning. Other cognitive and behavioral features, some of which are described above, are also defined as well as exclusionary criteria and features that lead to diagnostic uncertainty. Criteria for the diagnosis of possible and probable Parkinson disease dementia are also defined. As a point of comparison, consensus criteria for dementia with Lewy bodies were proposed in 1996 and later revised (87; 88). These criteria state that patients must have dementia and at least two of four core clinical features: parkinsonism, cognitive fluctuations, REM sleep behavior disorder, and/or visual hallucinations. Criteria were published for the diagnosis of prodromal dementia with Lewy bodies (89).

Validation of the clinical utility of these diagnostic criteria have been established (16). Differentiation of dementia with Lewy bodies and Parkinson disease dementia remains problematic, as their clinical distinction is based on an arbitrary (1 year) difference between the onset of motor and cognitive symptoms (55). Direct comparisons of cognitive impairment in dementia with Lewy bodies and Parkinson disease dementia have revealed a few differences in most studies (96), even though a tendency toward more pronounced deficits in the executive functions and visual memory have been associated with dementia with Lewy bodies (07; 70). In contrast, a 2-year longitudinal study demonstrated more rapid executive function decline among Parkinson disease dementia patients, compared to those with dementia with Lewy bodies (115).

Prognosis and complications

After dementia develops in a patient with Parkinson disease, the overall clinical course accelerates (84). Patients are vulnerable to complications of severe Parkinson disease, compounded by the presence of dementia. Injury from falls, poor nutrition and aspiration, progressive impairment of gait, poor judgment and agitation due to impaired cognition, susceptibility to infections, and medication toxicity all hastening morbidity and mortality. In a large study of Parkinson disease patients evaluated for 20 years after initial diagnosis, falls were present in 87%, dementia was present in 83%, freezing in 81%, dysarthria in 81%, hallucinations in 74%, choking in 48%, and daytime sleepiness in 20% (62). The presence of dementia in patients with Parkinson disease is associated with earlier nursing home placement (06) and a 2-fold increase of mortality (76). In dementia with Lewy bodies, the average survival from the onset of dementia until death is similar to that reported in Alzheimer disease and ranges from 2 years to 12 years. In a large prospective clinical trial of 72 dementia with Lewy bodies and 44 Parkinson disease dementia patients in the United Kingdom, patients with both diseases deteriorated cognitively (on the MMSE) and motorically (on the total motor score of the UPDRS) to the same extent over a 6-month period of time (83). In the Swedish dementia registry of over 30,000 patients with dementia, those with Parkinson disease dementia and Lewy body dementia were more likely to die of respiratory complications compared to those with Alzheimer disease, and were less likely to have diabetes and ischemic heart disease compared to those with Alzheimer disease. However, those with heart failure and diabetes were associated with higher risk of death (35).

Clinical vignette

A 70-year-old male with idiopathic Parkinson disease diagnosed eight years ago returned to clinic for his annual follow-up. He was accompanied by his wife, who stated that the patient had fluctuating attention and occasionally seemed “spaced-out.” He mentioned a problem with rabbits getting into the house as he had seen a family of them living in the bedroom. He had more difficulty staying engaged and playing the correct cards when he and his wife played euchre with their friends every other week. Finally, he required help with bathing as he had difficulty adjusting the water temperature knobs, as well as finding and applying the soap; this was out of proportion to his motor deficits, which had been relatively stable over the past year. The patient did not have any major changes to his other medical comorbidities and had no changes to his medications. He had imaging of his brain to evaluate for vascular dementia, but the scan did not show significant vascular lesions. Additionally, a geriatric depression scale score did not reveal active depressive symptoms.

After extensive discussion with the patient and his wife, the patient was diagnosed with probable Parkinson disease dementia based on the published diagnostic criteria.

Biological basis

Etiology and pathogenesis

It is not known why specific neurons degenerate in Parkinson disease. Hypotheses concerning etiologic factors in Parkinson disease include interplay between age, genetic and environmental factors (58). The major pathological changes that underlie Parkinson disease dementia may have additional different etiologies.

Several structural changes in the brain may underlie Parkinson disease dementia. They may be divided into separate categories: cellular loss in subcortical structures, specific pattern of cortical thinning, cortical and limbic system Lewy bodies, and coexisting Alzheimer disease pathology (133; 03; 40; 36). Indeed, the coexistence of Lewy body and Alzheimer disease pathologies has been reported to be a better predictor of Parkinson disease dementia than either pathology alone (38; 90). In Parkinson disease dementia, cortical thinning in parietal, temporal, occipital, and frontal cortices have been reported (03). Volume loss in the parahippocampus, insula, and cingulate gyrus are is associated with difficulties in visuospatial processing and episodic memory (03; 134). The pattern of cortical thinning can differentiate Parkinson disease dementia from dementia with Lewy bodies, where predominant frontal cortex thinning is seen in the former, and parietal and occipital cortical thinning is associated with the latter (03). Recent data complicate these findings, however, and suggest that patterned cortical thinning may not robustly distinguish Parkinson disease with and without significant cognitive impairment (56). Cortical thinning and decreased volume in the medial temporal cortex have been associated with Parkinson disease with mild behavioral impairment, relative to healthy controls or those Parkinson patients without mild behavioral impairment (135).

Damage to subcortical cholinergic and aminergic neurons that project to the cortex could contribute to the heterogeneity of cognitive presentations in Parkinson disease dementia (97). These include the medial substantia nigra (a source of dopamine), the locus ceruleus with major noradrenergic projections, the raphe nucleus (point of supply for serotonin), and the nucleus basalis (a key site of origin of cholinergic projections) (97). The nucleus basalis may show damage even in the absence of Alzheimer disease pathology.

Neurochemical deficits have been documented in Parkinson disease dementia and dementia with Lewy bodies (51). Decreased cholinergic input likely results in difficulty with memory and attention; these cholinergic changes may be a necessary step in the development of cognitive decline in Parkinson disease dementia (71). Lower CSF beta-amyloid levels have been observed during the MCI stage in Parkinson disease (133). Lower CSF glucocerebrosidase activity at the time of Parkinson disease diagnosis has been shown to be associated with an increased risk of subsequent development of Parkinson disease dementia (98).

Prominent pathologic changes with widespread Lewy neurites and Lewy bodies have been reported in the fasciculus and central medial nuclei of the rostral intralaminar group in the thalamus in patients with Parkinson disease dementia. These thalamic nuclei are part of the limbic loop, and their degeneration may contribute to the cognitive decline in patients with Parkinson disease (108). The presence of cortical Lewy bodies and Lewy neurites in the frontal cortex is associated with Parkinson disease dementia, independent of the degree of Alzheimer type of pathology (66). Lewy body dementia patients with prominent hallucinations and delusion have a lower density of neurofibrillary tangles than patients with Alzheimer disease (14). Understanding the interaction of synuclein and amyloid in Parkinson disease is further complicated by data suggesting that synuclein directly promotes amyloid formation (107). Furthermore, the density of cortical Lewy bodies does not differ in patients diagnosed with Parkinson disease dementia and patients with dementia with Lewy bodies (61), but patients with hallucinations have significantly higher density of Lewy bodies in the temporal lobes (Harding and Halliday 2002). A study demonstrated higher tau and beta-amyloid levels throughout all brain regions in the Lewy body disorders that have coexisting Alzheimer pathology, compared to those without Alzheimer disease co-pathology (40).

Braak and colleagues described a series of “stages” over which Parkinson disease evolves by reviewing autopsy studies in 168 patients (24). The initial stages primarily involve brainstem and subcortical nuclei changes beginning in the dorsal motor nucleus and reticular zone, progressing to the caudal raphe nuclei, then to the midbrain (substantia nigra), the prosencephalic regions, and finally to the cortex. Subsequent studies by this same group also looked exclusively at the changes related to Parkinson disease dementia and showed that cognitive impairment was directly related to increasing neuropathological stage (26; 25). Sauerbier and colleagues also noted alpha-synuclein pathology spreading from the brainstem or olfactory bulb, or even from the gut or areas innervated by the vagus nucleus, toward the neocortical regions (110). Despite this work, there is little consensus on the pathological cascade that eventually leads to Parkinson disease dementia as there is not a direct correlation between pathology and clinical progression of disease (117). A study demonstrated that disrupted white matter connectivity in frontal and posterior cortical areas were associated with early dementia in Parkinson disease patients (36). Overall, this work and the work of others indicates that Parkinson disease dementia is likely a combination of brainstem changes and diffuse subcortical (and cortical) accumulated pathology that leads to cognitive impairment through multiple mechanisms.

Coexisting Alzheimer disease pathology is common and severe enough to warrant an independent diagnosis of Alzheimer disease in the majority of cases (69). Even when the plaque and tangle burden is below the Alzheimer disease diagnostic threshold, these lesions probably still contribute to dementia. Several studies have shown that a majority with Parkinson disease dementia also have concomitant Alzheimer disease dementia-related changes in the brain (51; 40). Data have shown that synergy between cortical Lewy bodies and amyloid-beta/tau pathologies are associated with Parkinson disease dementia, and the degree of amyloid-beta burden shows a direct relationship with progression to dementia (37). Biomarker data have shown that the amyloid to tau index is lowest amongst Alzheimer disease patients but also decreased in Parkinson disease dementia and dementia with Lewy bodies patients compared to patients with Parkinson disease and controls. Additionally, the biomarker clusterin is elevated to the greatest degree in Parkinson disease dementia patients compared to other pathologies (126). Recent conflicting data, however, do not suggest a relationship between amyloid/tau deposition and impaired cognition in Parkinson disease (132), complicating current understanding of dementia pathogenesis in Parkinson disease.

Genetics. The impact of certain polymorphisms affecting COMT, MAPT, APOE4, and GBA genotypes on cognitive decline in Parkinson disease dementia and dementia with Lewy bodies have been established (97; 27), and sporadic and familial disorders may differ in cognitive profile and dementia risk (53). Apolipoprotein E4 allele has not been shown to predispose to Parkinson disease dementia as it does to Alzheimer disease (51). That being said, several studies have found that ApoE4 does increase the risk of concomitant Alzheimer disease pathology and cognitive changes consistent with Alzheimer disease in Parkinson disease patients (82; 93). One study also found increased odds of hallucinations in Parkinson disease patients with ApoE4 (93). Mutations in the gene encoding alpha-synuclein (SCNA) occur in rare familial cases of early-onset autosomal dominant Parkinson disease patients who have diffusely distributed Lewy bodies in their brains at autopsy (54). SCNA gene duplications can lead to Parkinson disease and also the risk of developing Parkinson disease dementia (51). Patients with the most common form of genetically acquired Parkinson disease – leucine-rich repeat kinase 2 (LRRK2) – do not show an increased risk of Parkinson disease dementia (04). A study showed that common variants near the TMEM106B gene were associated with more rapid cognitive decline in Parkinson disease patients (120). For a list of genes possibly associated with cognitive impairment view Table 1 below.

Table 1. Genetic Predisposition to Parkinson Disease Dementia

Genes associated with cognitive impairment	Familial Parkinson Disease	Sporadic Parkinson Disease
SNCA	Increased	Possible increase with intron 4 haplotype
LRRK2	Decreased	--------
Parkin	Decreased	--------
GBA	--------	Increased
MAPT H1 haplotype	--------	Increased
COMT Met/Met	--------	Mixed findings
APOE ε4	--------	Increased
PINK1	Decreased	--------
DJ-1	Decreased	--------
From: (58).

Epidemiology

Parkinson disease is the second most common neurodegenerative disorder after Alzheimer disease. Parkinson disease rises in prevalence with age and affects 1% to 2% of people 80 years old and over, but with recent data identifying parkinsonian clinical symptoms in approximately 50% of individuals older than 75 years without a diagnosis of Parkinson disease complicating understanding of disease prevalence (111; 29). Cognitive decline is a common and important nonmotor complication of the disease and can significantly affect the quality of life for both the patient and his or her caregivers. The prevalence of Parkinson disease dementia depends critically on methods of assessment and clinical definitions. Based on data, the prevalence of Parkinson disease dementia is approximately 30% in Parkinson disease patients, with 75% or more Parkinson disease patients developing Parkinson disease dementia prior to death. The prevalence of mild cognitive impairment in incident Parkinson disease was 42.5% in one study (133). Longitudinal studies indicate consistently more rapid cognitive decline in older men with Parkinson disease, compared to men without Parkinson disease (21). Mild cognitive impairment in Parkinson disease is a strong predictor of conversion to dementia. A cohort study of patients with mild cognitive impairment in Parkinson disease showed an 11% conversion rate to dementia per year, and 91% had dementia at 16 years (64). Parkinson disease dementia patients are generally older at the time of death (mean, 83.9 years), with longer disease duration (mean, 9.2 years), compared to those who have dementia with Lewy bodies (79.8 years; 6.7 years) (68).

Prevention

Parkinson disease and Parkinson disease dementia cannot be prevented at present. Future studies may identify drugs that will slow the progression of Parkinson disease, delay its onset, or prevent its occurrence. If so, the drugs may have a similar beneficial effect on Parkinson disease dementia. Similarly, strategies that delay the onset or slow the rate of progression of Alzheimer disease may decrease the rate of dementia in Parkinson disease. Regular physical activity and moderate exercise do not affect progression of Parkinson disease itself, but they are associated with slower deterioration of gait and balance, processing speed, and activities of daily living (121).

Differential diagnosis

Evaluation of dementia includes looking for potentially reversible contributing factors such as medications (especially anticholinergic drugs), depression, loss of hearing or vision, metabolic derangement, vitamin deficiency, and infection. Several degenerative disorders may cause dementia and parkinsonism, such as Lewy body dementia, progressive supranuclear palsy, frontotemporal dementia linked to chromosome 17, striatonigral degeneration, corticobasal degeneration, Creutzfeldt-Jakob disease, and Huntington disease. These conditions generally have additional neurologic features of atypical parkinsonism that fail to respond to L-dopa. Normal pressure hydrocephalus patients can have slowed gait, slowed trails, large ventricles, and poor activities of daily living, all features that are predictors of dementia in Parkinson disease (85). More common considerations are Alzheimer disease (particularly in its severe stages), and Alzheimer disease patients with impaired basal ganglia function due to pharmacologic dopaminergic receptor blockade or strategically placed strokes (43). Patients with Alzheimer disease are sensitive to neuroleptics and may develop parkinsonism even after even low doses. Usually this reverses after neuroleptics are discontinued, but reversal may take a month or longer. Persistent parkinsonism after neuroleptic withdrawal in a demented patient suggests dementia with Lewy bodies.

Parkinsonism increases as dementia worsens in Alzheimer disease, and is, thus, particularly common in institutionalized patients with severe Alzheimer disease (43). In both Parkinson and Alzheimer diseases, loss of smell has been identified (127). Seventy-two percent of Parkinson patients say that they have normal sense of smell, but when they are tested objectively, olfaction is impaired. The same is true of 74% of Alzheimer patients. Therefore, olfactory tests cannot be used to distinguish Parkinson disease dementia patients from Alzheimer disease patients. In terminal Alzheimer disease, severe immobilization, loss of gait, and paraplegia in flexion often occur. Parkinsonian findings in a patient with Alzheimer disease may not reflect underlying Lewy bodies. As a rough guide, if parkinsonism first appears late in the course of dementia, when the Mini-Mental State Examination score falls below 10/30, Alzheimer disease is the likely diagnosis whereas parkinsonism arising earlier in the course of apparent Alzheimer disease suggests dementia with Lewy bodies. Furthermore, parkinsonism with early visual hallucinations and fluctuating mental status is more indicative of dementia with Lewy bodies, particularly when behavioral/cognitive symptoms develop close in time with parkinsonism (111). In addition, Parkinsonian signs in Alzheimer disease are typically bilateral and symmetrical, and resting tremor is rare. It can be a challenge to conduct and interpret the motor examination in demented patients, however. For example, apathy may lead to flat speech, resembling mild hypophonia and masked facies, apraxia or difficulty following commands may lead to slowed movements, and paratonia (gegenhalten) may be difficult to distinguish from true rigidity.

Diagnostic workup

Patients diagnosed with Parkinson disease and who develop dementia should undergo rigorous systemic and neurologic evaluation to identify potentially reversible or treatable factors mentioned above. If delirium is present, it should be treated, and the patient should be reassessed. Formal cognitive testing is necessary to document dementia, establish its severity, and provide a baseline to follow longitudinally. The Montreal Cognitive Assessment is more sensitive for Parkinson disease dementia than Mini-Mental State Exam due to improved sensitivity for executive deficits in Parkinson disease (30). Information from an informant is important to document functional impairment and clarify details of the history. Serologies should include levels of vitamin B12 and thyroid function. A neuroimaging study, CT or MRI, is obtained in most patients, to search for a structural factor, such as stroke or tumor that could contribute to dementia. Detailed neuropsychological testing is not essential, but may be helpful in cases of mild dementia. EEG changes are nonspecific, but theta slowing in the temporal-occipital regions during rapid eye movement sleep may be predictive of development of dementia in Parkinson disease and in prodromal dementia with Lewy bodies (73; 122). Genetic testing is not recommended in all patients with Parkinson disease dementia, but it can be considered in those with early-onset or in those with a significant family history of Parkinson disease (53; 27).

MRI. There are no specific neuroimaging modalities that can accurately identify Parkinson disease dementia. A host of studies have used multiple brain imaging techniques to look for changes that may help differentiate Parkinson disease dementia from other disorders and control patients (19). In general, Alzheimer dementia patients generally show more atrophy in the mesial temporal lobes and hippocampi as compared to Parkinson disease dementia patients. Parkinson disease dementia patients tend to have more posterior cortical and subcortical changes on MRI. One study found that Parkinson disease dementia patients tended to show more atrophy in the anterior caudate and ventricular enlargement (12). On the other hand, diffusion tensor MRI has shown changes in the white matter tracts of the prefrontal and occipital cortices along with the corpus callosum, internal capsule, and uncinate in cognitively impaired Parkinson disease patients (08). In a MRI study, subcortical white matter changes were shown to be more severe in Parkinson disease dementia patients than in Parkinson patients without dementia or controls (63).

Functional MRI imaging can be used to support the diagnosis of Parkinson disease dementia, but it remains a research, rather than a routine clinical test. Functional imaging explores the functional organization of networks in the brain (03). Functional MRI has revealed disruption of corticostriatal and frontal cortex functional connectivity in Parkinson disease dementia (99); this pattern of functional connectivity disruption can be differentiated from dementia with Lewy bodies, which involves disruption of parietal and occipital cortices (23).

PET and SPECT imaging. These imaging modalities have also been used to study Parkinson disease dementia by assessing the integrity of the nigrostriatal dopaminergic pathways via the use of tracers that bind monoamine transporters (47). Reduced dopamine tracer uptake within the striatum and loss of cholinergic function in the forebrain on PET imaging has been associated with Parkinson disease dementia (113; 47; 67). A pattern of hypometabolism affecting the occipital lobe but not temporoparietal regions may be seen on PET or SPECT studies in dementia with Lewy bodies (94). Dopamine transporter SPECT (DAT) reveals patterns of decreased dopaminergic activity in patients with dementia with Lewy bodies and Parkinson disease dementia that differentiates them from Alzheimer disease (preserved dopaminergic activity) (28). However, it is not possible to distinguish dementia with Lewy bodies and Parkinson disease dementia with Dopamine transporter SPECT due to overlapping patterns (28). Pilotto and others performed a longitudinal FDG-PET study on 54 patients with Parkinson disease (101). They asked expert blinded raters to classify the scans as to whether they showed a typical Parkinson pattern, a dementia with Lewy bodies pattern, or one more like Alzheimer disease, corticobasal syndrome, or frontotemporal dementia. Parkinson patients who were identified at baseline with dementia with Lewy bodies or Alzheimer patterns on FDG-PET were more likely at 4 years to have developed signs of dementia, independent of their baseline cognitive test results.

Fluid biomarkers. No single clinical laboratory biomarkers (cerebrospinal fluid, serum, or urine) are available to help diagnose Parkinson disease dementia. However, biomarker analysis of cerebrospinal fluid may have the ability to discriminate Parkinson disease dementia patients from other conditions. Three key pathologies (alpha-synuclein, tau, and amyloid) appear to be related to cerebral volume loss in Parkinson disease dementia and dementia with Lewy bodies (03; 40). Studies measuring cerebrospinal fluid total alpha-synuclein have been inconsistent, but it may be a potential biomarker to differentiate Parkinson disease from Alzheimer disease and other forms of parkinsonism (92; 53; 124; 27). A panel of nine spinal fluid biomarkers (total tau, phosphorylated tau, beta-amyloid 1-42, neurofilament light chain, alpha-synuclein, amyloid precursor protein soluble metabolites alpha and beta, and two neuroinflammatory markers [monocyte chemoattractant protein-1 and YKL-40]), was able to differentiate Parkinson disease dementia from progressive supranuclear palsy, multiple system atrophy, and corticobasal syndrome patients (80). Additionally, a set of five spinal fluid biomarkers (total tau, phosphorylated tau, amyloid beta 42, APOE genotype, and SPARE-AD imaging score) was able to identify patients with Parkinson disease dementia from patients with Parkinson disease without dementia with an 80% accuracy (18). Ultimately, panels including multiple laboratory biomarkers may improve diagnostic accuracy in Parkinson disease dementia.

Management

No medical treatment has been proven to delay, stabilize, or reverse the biological processes underlying progressive dementia in Parkinson disease. However, treatment with cholinesterase inhibitors can lead to improvements in cognition, behavior, and quality of life for both the patient and the caregiver, as well as allow for possible prolongation of home-based care (112). Other interventions and medications are aimed at control of target symptoms (32). Comprehensive reviews of literature have been done to examine the available evidence (13). Palliative care has recently become incorporated in the management of Parkinson disease, especially since the central principles of palliative care promote support for care partners and emphasize the importance of narrative medicine in improving communication with patients (119).

Cognitive symptoms. Given that there are marked cholinergic deficits in the brains of patients with Parkinson disease dementia, the cholinergic system has been the most explored in regard to management and evidence suggests that cholinesterase inhibitors are efficacious (123). Both placebo-controlled and open-label studies of cholinesterase inhibitors have shown improvement of cognitive fluctuations, apathy, and visual hallucinations. Open-label studies with rivastigmine (104; 44) and donepezil (17; 49) have suggested improvement of both cognitive and noncognitive symptoms without worsening extrapyramidal signs.

A large double-blind, placebo-controlled study of rivastigmine in 541 patients with Parkinson disease dementia found that those on rivastigmine were significantly improved in measures of cognition, global functioning, activities of daily living, and in behavior, particularly with regards to reduction in visual hallucinations (44). Adverse effects primarily involved gastrointestinal symptoms such as nausea, vomiting, and diarrhea. Most adverse events occurred during the initial titration phase and were usually transient. There was increased tremor in 10% of treated patients, but this was usually mild and did not lead to discontinuing the medication. This study used capsules whereas the rivastigmine patch, which has fewer adverse effects, is most commonly used in clinical practice. Further analysis in this trial showed that the subgroup with hallucinations had the most profound improvement in cognitive testing (31).

Increased MMSE and Clinician’s Global Impression of Change scores, but no change in the Neuropsychiatric Inventory score, have been reported in a placebo-controlled trial with donepezil (05). The incidence of significant side effects caused by peripheral cholinergic stimulation may be higher in patients with Parkinson disease, and patients need to be monitored for orthostatic hypotension and diarrhea. Possible worsening of parkinsonism and tremor has been occasionally noted with cholinesterase inhibitors (57; 10). In general, rivastigmine or the other cholinesterase inhibitors should not be discontinued without gradual taper as sudden decline in cognition has been reported in these patients (91).

Memantine has been studied in a number of small and intermediate sized trials with dementia with Parkinson disease. Results to date have been equivocal regarding cognitive or behavioral benefits in these populations, and it remains unclear whether benefits outweigh potential adverse effects in these populations. A small 72 patient double-blind, placebo-controlled trial showed improved clinical global impression of change at 24 weeks in Parkinson disease dementia and dementia with Lewy bodies patients (02). A subsequent 199 patient randomized, double-blind, placebo-controlled trial showed only a benefit in dementia with Lewy bodies patients, but not in Parkinson disease dementia patients (46). A meta-analysis of both memantine and cholinesterase inhibitors showed that only cholinesterase inhibitors improved cognition, but both cholinesterase inhibitors and memantine slightly improved clinicians’ global impression of change (129).

Ambroxol, which raises levels of glucocerebroside and subsequently lowers alpha-synuclein, is under active investigation for potential disease slowing effect in mild to moderate Parkinson disease dementia. Although there are no specific trials looking at Parkinson disease dementia patients, cognitive training and physical exercise may be beneficial for improving cognitive function in those with Parkinson disease (75; 114). For review of clinical trials in this area see: (128).

General issues in treatment of Parkinson disease dementia. The management of Parkinson disease dementia should include consideration of the patient, caregiver, and social and physical environments. Issues such as competence to work and drive, home safety, and the need for assistance at home should be evaluated. The caregiver should be educated about dementia and the strategies to support and maintain the patient. The help of social workers and community resources such as day care centers and support groups can provide substantial assistance. The patient's social and physical environment must be adapted to his or her disabilities and the daily routine should be kept simple and predictable. Factors such as inappropriate levels of stimulation or poor sleep may worsen cognitive functioning. The treating physician should look out for treatable or reversible causes of symptoms. The sudden onset or worsening of hallucinations, confusion, somnolence and agitation, or rapid cognitive or motor decline should prompt a thorough search for an underlying medical illness, metabolic or electrolyte derangement, or side effects of medications.

Motor dysfunction. In Parkinson disease dementia, there is risk that dopaminergic replacement therapies used as a standard of care may cause or worsen visual hallucinations or agitation, whereas medications with anticholinergic actions may cause confusion. Such medications, which may be necessary for the practical management of patients, should be closely monitored and titrated to maintain an appropriate balance between motoric efficacy and cognitive/behavioral side effects. When L-dopa is titrated carefully, it may well continue to improve motor function in Parkinson disease dementia with minimal adverse effects. It does not improve cognitive function per se, although some patients may show increased speed of thought and increased attention. In untreated Parkinson patients with worsening motor symptoms, a trial of low-dose L-dopa is warranted. L-dopa is less likely than the newer dopamine receptor agonists to provoke hallucinations.

Depression. Persistent depression in patients with Parkinson disease dementia may benefit from treatment with psychotherapy, antidepressant medications, and occasionally electroconvulsive therapy. Double-blind, placebo-controlled trials of various drugs in Parkinson disease with or without dementia have generally had large placebo effects, and efficacy has been equivocal (130). Tricyclic antidepressants may be efficacious in patients with Parkinson disease, but anticholinergic side effects and orthostatic hypotension may limit their use in Parkinson disease dementia. In patients with agitated depression, the sedation of tricyclic antidepressants can be useful in promoting sleep. Newer antidepressants in the selective serotonin reuptake inhibitor and selective serotonin-norepinephrine inhibitor classes may be effective in treating depression, especially when withdrawal and apathy are prominent. Of these medications, paroxetine and venlafaxine have arguably the strongest data to support their efficacy in Parkinson disease depression (106). Electroconvulsive therapy may be helpful when depression persists, especially when severe apathy is present (50). The AMANDYSK Trial assessed the long-term effect of chronic treatment of Parkinson disease patients who have levodopa-induced dyskinesia that is treated with amantadine. When amantadine was randomly withdrawn, apathy and fatigue worsened in those who were assigned to placebo, compared to those who remained on active treatment (100).

Psychosis. Before resorting to antipsychotic medications for behavioral control, environmental or interpersonal factors should be sought. Adjuncts to behavior management include regular scheduling of activities and sleep, or using techniques such as distraction. There is no single type of medication that consistently reduces delusions, hallucinations, agitation, insomnia, or other behavioral changes associated with dementia. Because of the possibility of adverse side effects, medication should be initiated at low doses, increased slowly as tolerated, and decreased or discontinued if the behavior abates. For example, quetiapine can be started at 25 mg qhs, then increased over several weeks to 100 mg qhs if the delusional ideation persists. A low dose of a long-acting benzodiazepine such as clonazepam is particularly effective if the patient has REM sleep behavior disorder. More chronic use of benzodiazepines increases the risk for sedation and falls, a particular danger in this population with impaired righting reflexes. Short-acting benzodiazepines should be discouraged because they can cause paradoxical confusion in patients with dementia.

High potency neuroleptics such as haloperidol often worsen the symptoms of parkinsonism, and they can lead to marked motor deterioration in patients with dementia with Lewy bodies. This is why atypical neuroleptics are preferable, but are not without risk.

Quetiapine, an atypical antipsychotic agent, has virtually no anticholinergic activity, making it an excellent theoretical choice (52; 118).

In 2016, the FDA approved pimavanserin, a 5-HT2A inverse agonist without dopaminergic, adrenergic, histaminergic, or muscarinic affinity, for use in the treatment of hallucinations and delusions in patients with Parkinson disease psychosis. Its antipsychotic effects are thought to be more helpful in those with than without cognitive impairment and may be enhanced for those already on cholinesterase inhibitors (48; 81).

Cholinesterase inhibitors have been shown to improve hallucinations and psychosis in demented patients with Parkinson disease. Several clinical trials have reported the efficacy of this class of medication.

REM sleep behavior disorder. Movements of the extremities associated with vivid dreams may be fluctuating and severe, and they may cause insomnia in bed partners or lead to significant injuries of the patient or bed partner. REM sleep behavior disorder in Parkinson disease dementia is treated similarly to those who have REM behavior disorder in dementia with Lewy bodies, utilizing melatonin 3 to 9 mg and/or clonazepam 0.5 to 2 mg in the evening before bedtime (22).

Special considerations

Pregnancy

Dementia in Parkinson disease almost always occurs beyond childbearing age.

Anesthesia

There are no studies specifically addressing the impact of anesthesia in patients with Parkinson disease dementia. Demented patients can experience prolonged confusion after emerging from general anesthesia. If possible, tranquilizers and sedatives should be minimized during the post-anesthesia recovery period.