When Jonathan Swift wrote of the obstinate conflict which was waged between the little-enders (Lilliputians) and the big-enders (Blefuscudians) (fig 1) over which end of an egg to break before eating it, it was a mirror for the argument between consubstantiation and transubstantiation in the Church of England and the Roman Catholic Church.1 It might equally well have been modelled on the seemingly endless debate which surrounds the classification of dementia with Lewy bodies and Parkinson’s disease with dementia.2, 3 My purpose here is to outline my own understanding of this matter, a subject which the reader may regard as either trivial or critical. I hope to persuade you that it is neither.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1 Gulliver’s Travels by Jonathan Swift.

The general confusion which surrounds this whole area is encapsulated in the following extract from a fictitious referral letter which I have received many times in one version or another: “Dear Psychiatrist, Mr X is a very pleasant retired engineer [neurologists are usually very generous in their assessment of patients’ characters] who has been under my care since 1998 with Parkinson’s disease. He originally responded moderately well to levodopa and his motor symptoms have progressed only slowly, but now he has pronounced postural instability and gait difficulty. Over the last six months his wife reports that he is becoming very forgetful and is experiencing troublesome hallucinations of people whom he sees in his house. I think that he may be developing Lewy body disease and would be grateful for your advice.”

My advice is generally as follows: “Dear Neurologist, Yes, he very probably does have Lewy body disease! Parkinson’s disease, clinically diagnosed according to standard criteria, is associated with Lewy body disease in most cases followed to autopsy. What you probably meant to ask me is not whether Mr X is developing Lewy body disease (which he has probably had for nearly 10 years) but whether his Lewy body disease is now also producing cognitive decline and hallucinations, and to what extent his antiparkinsonian medication is contributing. Implicit in all of this is the possibility that he is progressing along the slippery path to dementia.”

One might regard this response as rather unhelpful, playing with words, a question of semantics—and it certainly isn’t what I actually write in the case notes! But I would argue that the words, and the meanings they carry with them, are critical to understanding what the referrer wants to know about his or her patient. In order to develop the argument further it is helpful to consider the origin and meaning of a variety of terms familiar to most of us.

BASIC TERMINOLOGY

Parkinson’s disease is a term that was first coined by Charcot who had read James Parkinson’s “Essay on the Shaking Palsy”4 in a French translation, and concluded “this is a descriptive and vivid definition that is correct for many cases, most in fact, and will always have the advantage over others of having been the first, but it errs by being too general”. Parkinson’s disease is a clinically defined syndrome. Although histopathological confirmation is said to be the gold standard,5 there are no universally accepted pathological criteria. Most definitions include substantial neuronal loss and gliosis, and the finding of at least one Lewy body in the substantia nigra or locus coeruleus, in the absence of any other disease that could cause parkinsonism.

Parkinson’s disease with dementia is a term that has been in use for many years to describe patients in the clinic, but its emergence as a well-defined syndrome has been slow, and definitions such as that in DSMIVR remain poorly operationalised and imprecise with reference to cognitive and motor slowing, executive dysfunction, impairment in memory retrieval and being frequently exacerbated by depression. The prevalence of dementia in Parkinson’s disease patients was for many years said to be 25–30%, which does probably accurately reflect the situation among people with Parkinson’s disease in the stage when they are attending movement disorder clinics. The recent recognition that dementia becomes increasingly prevalent in Parkinson’s disease populations followed over the course of the disease, with reports of up to 78% period prevalence over eight years,6 has prompted renewed interest in generating more evidence-based criteria for Parkinson’s disease with dementia. A Movement Disorder Society Task Force charged with this responsibility recently reported its findings.7 Many of these patients are very severely motor impaired and the combination of motor, cognitive and neuropsychiatric and behavioural symptoms conspire to move them rapidly into institutional care, a common outcome which partly explains why they do not appear in the movement disorder clinic prevalence estimates.8

Dementia with Lewy bodies is now the preferred term for a series of diagnostic apellations which evolved during the 1980s and 1990s from different research groups.9 These included Lewy body variant of Alzheimer’s disease, Lewy body dementia, senile dementia of Lewy body type, and cortical Lewy body disease. Dementia with Lewy bodies is, like Parkinson’s disease, a clinically defined syndrome and consists of a primary dementia characterised by visuoperceptual and executive dysfunction accompanied by (1) prominent visual hallucinations, (2) fluctuating attention and (3) parkinsonism. The presence of the typical cognitive profile together with two out of three of these core features qualifies for a diagnosis of probable dementia with Lewy bodies which is more than 90% predictive of Lewy body pathology at autopsy. Unfortunately, looking at this the other way round, many Lewy body pathology cases do not present in this “typical” way—indeed they often present with an insidious amnestic syndrome more suggestive of Alzheimer’s disease. This probably reflects the fact that many such cases have substantial additional Alzheimer (neocortical tangle) pathology which colours the clinical picture. The third meeting of the International Dementia with Lewy Bodies Consortium10 addressed this latter issue by producing a two-dimensional approach to the pathological classification of dementia cases with Lewy bodies and Alzheimer pathology, in which the likelihood of clinically presenting as probable dementia with Lewy bodies is directly related to the severity of Lewy body pathology and inversely related to the severity of Alzheimer pathology. The previous notion of being either one thing or the other (dementia with Lewy bodies or Alzheimer’s disease) is now blurred—there are many intermediate cases which are a bit of each.

Lewy body disease is a term which was originally proposed by Professor Kenji Kosaka11 who had been studying the neuropathological changes in patients with dementia for the previous decade and had collected a series of cases with a combination of Lewy bodies in the brainstem, diencephalon, anterior cingulate, amygdala and cerebral cortex. He distinguished “diffuse” Lewy body disease with a high Lewy body burden in most if not all of these areas, from “brainstem predominant” Lewy body disease. A third, “transitional” type of Lewy body disease was intermediate between diffuse and brainstem predominant. Lewy body disease was first conceived therefore as a catch-all term which brought together some rather disparate neuropathological cases. Clinically the patients ranged from typical idiopathic motor Parkinson’s disease, through cases with a mixture of parkinsonism and dementia, to a few cases with dementia but no extrapyramidal motor features (about 25% of the diffuse group). Figure 2 shows the relations between these diagnostic entities.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2 Dementia with Lewy bodies (DLB) and Parkinson’s disease with dementia (PDD) is a not infrequent late complication of PD. By the end stage DLB and PDD patients are often clinically indistinguishable with a combination of extrapyramidal motor features, cognitive impairment, psychiatric symptoms and autonomic dysfunction. End-stage neuropathology is also similar with diffusely distributed Lewy neurites (LN) and Lewy bodies (LB), with beta-amyloid cortical plaques (BAP) also common but neocortical neurofibrillary tangles (NFT) only in a minority.

THE “ONE-YEAR RULE” TO SEPARATE DEMENTIA WITH LEWY BODIES AND PARKINSON’S DISEASE WITH DEMENTIA

The Consensus group which met to delineate the dementia with Lewy bodies syndrome predominantly comprised dementia clinicians rather than movement disorder specialists, and their goal was to produce criteria by which to distinguish dementia with Lewy bodies from Alzheimer’s disease and other primary dementing disorders. They recognised however the symptomatic overlap between dementia with Lewy bodies and patients with Parkinson’s disease who went on to develop dementia as their motor disorder progressed. En passant they therefore generated a definition of Parkinson’s disease with dementia in the form of the infamous one-year rule which stated that if extrapyramidal motor features had been present for 12 months or more before the onset of dementia, the diagnosis should be Parkinson’s disease with dementia, but if dementia occurred within 12 months of the motor features, or indeed preceded the motor features, the diagnosis should be dementia with Lewy bodies.9 It was acknowledged that this boundary in time was entirely arbitrary and existed solely to draw a line between two overlapping clinical syndromes. A return to this boundary issue at the third meeting of the International Dementia with Lewy Bodies Consortium re-stated this basic position as follows “The distinction between dementia with Lewy bodies and Parkinson’s disease dementia as two distinct clinical phenotypes, based solely on the temporal sequence of appearance of symptoms has been criticized by those who regard the different clinical presentations as simply representing different points on a common spectrum of Lewy body disease, itself underpinned by abnormalities in alpha-synuclein metabolism. This unitary approach to classification may be preferable for molecular and genetic studies and for developing therapeutics. Descriptive labels that include consideration of the temporal course are preferred for clinical, operational definitions. Dementia with Lewy bodies should be diagnosed when dementia occurs before or concurrently with parkinsonism, while Parkinson’s disease with dementia should be used to describe dementia that occurs in the context of well-established Parkinson’s disease. The appropriate term will depend on the clinical situation and generic terms such as Lewy body disease are often helpful. In research studies in which distinction is made between dementia with Lewy bodies and Parkinson’s disease with dementia, the one-year rule between the onset of dementia and parkinsonism should be used. Adoption of other time periods will simply confound data pooling or comparison between studies. In other research settings, including pathological studies and clinical trials, both clinical phenotypes may be considered collectively under categories such as Lewy body disease or alpha-synucleinopathy”.10

IS THE DISTINCTION BETWEEN DEMENTIA WITH LEWY BODIES AND PARKINSON’S DISEASE WITH DEMENTIA OF ANY IMPORTANCE?

Other than age of onset, temporal course and possibly levodopa-responsiveness, no major differences between dementia with Lewy bodies and Parkinson’s disease with dementia have been found in any variable examined thus far, including cognitive profile, attentional performance, neuropsychiatric features, sleep disorder, autonomic dysfunction, type and severity of parkinsonism, neuroleptic sensitivity and responsiveness to cholinesterase inhibitors.12 There may be some minor pathological differences—for example, people with Parkinson’s disease with dementia have greater nigral neuron loss than dementia with Lewy body patients, which probably reflects disease severity and duration, and there is greater cortical beta-amyloid deposition in dementia with Lewy bodies, particularly compared with late-onset Parkinson’s disease with dementia (more about this later). So is the distinction between dementia with Lewy bodies and Parkinson’s disease with dementia of any practical significance? The answer depends upon who is asking the question. For the patient, family and treating clinician, the diagnostic label generally carries considerable significance because it describes the primary problem (parkinsonism or dementia) and carries with it a particular set of expectations. For the laboratory scientist, diagnostic categories are less important and both syndromes are more easily seen as different expressions of a common underlying disease process. Even in the clinic this unifying approach can be the easiest route—telling a patient that he has Lewy body disease as the explanation for his myriad of symptoms is often more convincing than convoluted discussions about “Parkinson’s plus” or other complex diagnostic epithets. In terms of patient assessment and management, the issues posed by dementia with Lewy bodies and Parkinson’s disease with dementia are essentially the same with a need to treat motor, cognitive, psychiatric and autonomic dysfunction using the same drugs and non-pharmacological strategies.

The relative severity of individual symptoms may vary but in both syndromes there will probably be a need to treat at least two and often more of the categories of symptoms listed above, with the accompanying tension of improving one at the expense of worsening another—for example, aggravating hallucinations and confusion with levodopa. This particular issue is worth commenting on in passing. Most clinicians know from their clinical practice that dopaminergic drugs cause worsening of hallucinations and confusion in an apparently dose-dependent manner. The research literature on the other hand finds hallucinating and non-hallucinating Parkinson’s disease patients not to differ in their mean daily levodopa dose,13 and that high dose levodopa infusion does not produce hallucinations in people with early stage Parkinson’s disease.14 The answer to this conundrum appears to be that confusion and hallucinations are an integral part of Lewy body disease that can occur in the absence of dopaminergic treatment (as in dementia with Lewy bodies), and which seem more likely to occur when there is cognitive impairment and diffuse, cortical pathological involvement. Dopaminergic drugs do not therefore cause the psychiatric symptoms, rather they exacerbate them or bring them forward in time—that is, they would have happened anyway but drug treatment precipitates them at an earlier time. This explains why drug-induced psychotic and confusional symptoms often fail to resolve completely on withdrawal of the offending drug and, even if they do, why they often recur within a fairly short space of time.

WHAT CAUSES DEMENTIA IN PARKINSON’S DISEASE AND DEMENTIA WITH LEWY BODIES?

Since increasing age is a major risk factor for developing dementia in Parkinson’s disease, it was for a long time argued that much if not all dementia in Parkinson’s disease was coincidental and the result of brain ageing or concomitant Alzheimer’s disease. This view is almost tenable if the already quoted prevalence estimates of 25–30% dementia in Parkinson’s disease are used as the basis for calculating risk, but the recent realisation that people with Parkinson’s disease are up to six times more at risk of dementia than age-matched controls makes the coincidental Alzheimer’s disease theory a less than adequate explanation.

There are two fundamental approaches when trying to unravel the “what causes dementia in Parkinson’s disease?” question. One is by clinical deduction, the other is by pathological correlation. Clinically it is apparent that dementia in Parkinson’s disease is heterogeneous in its manifestations, suggesting that there is unlikely to be a single underlying cause. The fact that this dementia may feature as a predominantly dysexecutive syndrome (common), as an amnestic “Alzheimer-like” picture (less common), or as dementia with Lewy bodies-like with prominent fluctuations and hallucinations (common) suggests a range of “causes”. The biological candidates are Lewy body formation and synuclein pathology, Alzheimer-type pathology, neuron loss and neurochemical deficits. Studies suggest that Lewy body related pathology is more strongly associated with Parkinson’s disease with dementia than Alzheimer-type changes,15, 16 but the reality is that each individual person who develops dementia in the course of Parkinson’s disease will do so because of the accumulation of a mixture of pathologies. The report of the third Consortium tackled this issue with respect to the pathological “cause” of dementia with Lewy bodies and proposed a new approach that “takes into account both the extent of Lewy-related pathology, and Alzheimer-type pathology, in assessing the degree of certainty that the neuropathological findings explain the dementia with Lewy bodies clinical syndrome.” They concluded that “the likelihood that the observed neuropathology explains the dementia with Lewy bodies clinical syndrome is directly related to the severity of Lewy-related pathology, and inversely related to the severity of concurrent Alzheimer-type pathology.”

This is summarised in figure 3 which represents a probability matrix correlating the clinical picture with the predicted underlying pathology. There is no reason to think that this approach is not equally applicable to Parkinson’s disease with dementia and it certainly offers the clinician a useful way of understanding why the patient is presenting with cognitive decline progressing to dementia in the way that he or she is. At the left-hand side of the matrix are people with Lewy bodies but few Alzheimer lesions. They may simply have Parkinson’s disease and if they become demented they are predicted to be highly likely to have a dementia with Lewy body phenotype with fluctuating confusion and early hallucinations. On the right-hand side are patients with high Alzheimer pathology in addition to their Lewy body disease. These individuals are probably particularly likely to develop dementia which will be of the amnestic type resembling Alzheimer’s disease rather than dementia with Lewy bodies. In the middle are cases with a mixture of pathologies and a corresponding mixture of clinical features. The clinician faced with a patient with Parkinson’s disease with dementia or dementia with Lewy bodies can, with this model, try to determine the underlying pathological cause based on the pattern of clinical features and the results of investigations. Memory deficits, hippocampal atrophy on brain CT/MRI and old age are likely predictors of Alzheimer pathology being the major determinant of dementia. Reduced occipital perfusion (PET or SPECT scanning), prominent attentional and visuo-perceptual impairments are suggestive of Lewy body disease itself being the main contributor.

• Download figure
• Open in new tab
• Download powerpoint

Figure 3 Probability matrix predicting the likelihood of an individual patient having the “typical” dementia with Lewy body clinical syndrome of fluctuating cognitive impairment, recurrent visual hallucinations and parkinsonism, in relation to the relative burdens of Lewy body disease (vertical axis shows increasing pathology: brainstem< limbic, diffuse) and Alzheimer pathology (horizontal axis shows increasing pathology: low, intermediate, high).10 NIA-Reagan, National Institute on Aging-Reagan Institute criteria for the neuropathological diagnosis of Alzheimer’s disease.

WHY DO SOME PEOPLE GET PARKINSON’S DISEASE, SOME DEMENTIA WITH LEWY BODIES, AND SOME PARKINSON’S DISEASE WITH DEMENTIA?

If one accepts the Lewy body disease spectrum model, a question that has to be answered is why different people express the disease in such different clinical ways. A simplistic approach is to assume that the critical factor is the anatomical distribution of pathology/neuronal dysfunction/neurochemical deficits. Substantia nigral loss produces extrapyramidal motor features, cortical involvement increasing the likelihood of dementia. The progressive “spread” of Lewy body pathology from medulla and olfactory bulb through pons, midbrain, basal forebrain and medial temporal cortex to finally involve multimodal association and primary cortices proposed by Braak is compatible with the observation that dementia usually occurs late in the course of Parkinson’s disease—that is, after the motor features.17 It does not however account for dementia with Lewy body cases, which by definition (see the one-year rule above!) start with “cortical” features and 25% of whom will never show extrapyramidal signs. Nor does it adequately account for those who develop dementia early in the course of their Parkinson’s disease, unless it is proposed that they have a particularly aggressive or rapidly progressive form of Lewy body disease. The problem with using pathological studies as the gold standard to define clinical syndromes is that the patterns of pathology generally reflect the sample of patients that were studied and cannot be extrapolated beyond that. In the case of the Braak staging system, the patients studied had originally presented to a movement disorder clinic with motor Parkinson’s disease and some progressed to a state of dementia late in their illness. No dementia with Lewy body cases or patients with clinical presentations other than Parkinson’s disease were included. It is not surprising therefore that Braak reported a caudo-rostral progression of disease and that other distributions of Lewy body pathology (for example, cortical predominant) were not seen.

As with most situations when the theoretical model fails to fully explain the facts, the Braak staging system may need some extension to account for the full clinical variation of Lewy body disease. For example, Japanese scientists have reported three pathological subtypes of Lewy body disease: type 1, brainstem predominant; type 2, “almost equal” for brainstem and cerebral cortex; and type 3 with a cerebral cortex-predominant pattern.18 This challenges the notion that Lewy body disease has to begin in subcortical regions and spread upwards. Another recent study found late-onset Parkinson’s disease with dementia (more than 10 years between the onset of motor symptoms and dementia) was more highly correlated with cortical cholinergic deficiency and less with cortical pathological lesion burden (both plaques and Lewy pathology), compared to Parkinson’s disease with dementia cases with earlier onset of dementia, who more closely resembled the pathology of dementia with Lewy bodies with higher burdens of cortical disease.19 The authors concluded that “While there is a clear relationship between the duration of Parkinson’s disease prior to the onset of dementia and key neuropathologic and neurochemical characteristics, there is a gradation of these differences across the dementia with Lewy bodies/Parkinson’s disease dementia spectrum and the findings do not support an arbitrary cut-off between the two disorders.”

CONCLUSIONS

As with the disagreement between the islanders of Lilliput and Blefuscu over which end of the egg to break, there can ultimately be no right or wrong about what one calls Parkinson’s disease with dementia or dementia with Lewy bodies. It is a matter of opinion and without stretching the analogy too far, the egg will taste pretty much the same wherever it is cracked, much as patient management should follow the same general principles, whatever the diagnostic label. However, patients and their families will expect us to be consistent in our approach and this is where the importance of this issue lies. Neurologists, old age psychiatrists, geriatricians, general practitioners and any other health professionals who come into contact with what is an increasingly prevalent disease would do well to agree on the best way to describe it.20

EDITOR’S NOTE

As Editor I can but stand back from the debate between the reviewer and the author; they seem to know their Gulliver’s Travels rather well:

First Dr Corey-Bloom: “In light of the Jonathan Swift reference, it might well be worth pointing out that although Gulliver helped the Lilliputians defeat the Blefuscudian fleet, he eventually received a warm welcome in the court of Blefuscu, a satirical allusion to the arbitrariness of positions. It may, however, be too early to conclude that there are no differences in clinical features between dementia with Lewy bodies and Parkinson’s disease with dementia. Subtle cognitive differences have been described, with dementia with Lewy bodies patients making more conceptual and attentional errors, and adverse reactions to neuroleptics may also be more frequent. Dementia with Lewy bodies patients tend to have fewer signs of parkinsonism and never prominent tremor; Parkinson’s disease with dementia patients, on the other hand, at least initially, have more asymmetry in their motor features. Pathologically, the brain regions that might aid in distinguishing dementia with Lewy bodies from Parkinson’s disease with dementia include the substantia nigra, where neuronal loss is greater in Parkinson’s disease with dementia, and the striatum, where alpha-synuclein pathology may be greater in dementia with Lewy bodies. Although most would agree that dementia with Lewy bodies and Parkinson’s disease with dementia are more similar than different, and that the third report of the Dementia with Lewy Bodies Consortium was an important first step, large, prospective clinicopathological correlation studies are still necessary to validate and refine their findings.”

And then Dr McKeith: “I think this analogy might stand up better than Dr Corey-Bloom credits. Perhaps my plan has been (like Gulliver) to get into the governing bodies of two cultures (psychiatry and movement disorder) in order to help negotiate a settlement between them. I won’t be drawn further on the point but it is a possibility! I agree that there are subtle clinical and pathological differences between dementia with Lewy bodies and Parkinson’s disease with dementia, and I have championed keeping the distinction between them for clinical use. My position is that the argument over such similarities and differences is pointless—hence my attempts at intervention and peacemaking.”