Introduction

The current approach to the classification of neurodegenerative parkinsonism is based principally on the results of the neuropathological examination. Nosology includes Parkinson’s disease (PD), Parkinson’s disease dementia (PDD), dementia with Lewy bodies (DLB), multiple system atrophy (MSA), progressive supranuclear palsy (PSP), corticobasal degeneration (CBD), frontotemporal lobar degeneration (FTLD), and Alzheimer’s disease (AD). All these disorders are clinically characterized by the validated clinical diagnostic criteria. Practically all clinical criteria differentiate “probable”, “possible”, and “definite” degrees of diagnostic certainty.

A more modern classification based on molecular pathology includes categories such as alpha-synucleinopathies, either neuronal (PD, PDD, DLB) or oligodendroglial (MSA), tauopathies (also called FTLD-tau; PSP, CBD, argyrophilic grain disease, AGD, primary age-related tauopathy, PART, and globular glial tauopathy, GGT), FTLD with inclusions of proteins TDP-43 (FTLD-TDP), FUS (FTLD-FUS), or other with immunoreactivity for components of the ubiquitin-proteasome system (FTLD-UPS), and AD.

The trouble is that the pathology is not always “unique” in the sense of typical disease signs and hallmarks. Of course, the neuropathological examination in the typical “sporadic” PD will probably reveal only the pathology typical for the alpha-synucleinopathy (Fig. 1) and in typical PSP, the pathology typical for tauopathy1,2. However, in the majority of cases, the final neuropathological picture is more complicated than the ones just described. It is widely known, that the most frequently observed pathological finding in neurodegenerative disorders (including parkinsonism) is the “double-pathology” (Fig. 2) or even “triple-pathology” (Fig. 3)3,4,5,6,7,8, sometimes called “overlap syndrome”. It means that the characteristic “pure” pathological picture (Fig. 1) is a rather rare case, and that overlaps prevail (Figs. 2 and 3)9,10. Moreover, in many cases, the picture is complicated by concomitant vascular changes.

Fig. 1: Pure alpha-synucleinopathy in typical Parkinson’s disease phenotype.
figure 1

a Classical Lewy bodies in the pigmented neurons of substantia nigra, HE staining, original magnification ×200. b Pathological deposits of α-synuclein in substantia nigra—Lewy bodies, granular cytoplasmic positivity, and dystrophic neurites, stained with a monoclonal antibody against α-synuclein, original magnification ×200.

Full size image
Fig. 2: “Double-pathology“: α-synucleinopathy + tauopathy in Parkinson’s disease dementia (PDD) phenotype.
figure 2

a Lewy body and pale bodies in pontine raphe nucleus, HE staining, original magnification ×200. b Pathological deposits of α-synuclein in pontine raphe nucleus—Lewy bodies, granular cytoplasmic positivity, dystrophic neurites, and dots, stained with a monoclonal antibody against α-synuclein, original magnification ×200. c Pathological deposits of tau protein in pontine raphe nucleus—tangles, pretangles, grains, and threads, stained with a monoclonal antibody against hyperphosphorylated tau, original magnification ×200. d Pathological deposits of tau protein in basal ganglia associated with cribrous state—tau-astrogliopathy (ARTAG), stained with a monoclonal antibody against hyperphosphorylated tau, original magnification ×100.

Full size image
Fig. 3: “Triple-pathology“: α-synucleinopathy + tauopathy + β-amyloid in progressive supranuclear palsy—parkinsonism (PSP-P) phenotype.
figure 3

a Pathological deposits of α-synuclein in the hippocampus—Lewy bodies, dystrophic neurites, and dots, stained with a monoclonal antibody against α-synuclein, original magnification ×100. b Pathological deposits of tau protein in the hippocampus—pretangles, threads, and grains, stained with a monoclonal antibody against hyperphosphorylated tau, original magnification ×100. c Amyloid plaques in hippocampus positive in immunohistochemical reaction with a monoclonal antibody against amyloid-β -peptide, original magnification ×100.

Full size image

The question stands whether this observation is reflected intra vitam also in the clinical manifestations and whether the current clinical diagnostic criteria can serve as a valuable tool in the diagnostic process. It has been debated for more than the past 10 years that the “validated” and widely used clinical diagnostic criteria for some of the above-mentioned clinical entities are outdated and based on the state of knowledge at the time of their publication, i.e. the eighties or nineties of the past century11,12,13,14,15,16,17,18,19. Several groups are working hard on the establishment of the new criteria, namely for CBD, PSP, MSA and FTD, but the validation process will certainly take a few years.

The present work aims to re-appraise the issue of Lewy body (LB) diseases, based on past and contemporary nosological and neuropathological correlates and other molecular-biological aspects that may play a role in their pathogenesis16,20,21,22,23,24.

The enigma of classifying PD and other LB disorders

Nowadays, the neurological community is facing a classification problem in the group of intraneuronal synucleinopathies, which covers the proper clinical diagnosis and differential diagnosis among PD, PDD, and DLB5,25,26.

PD and its heterogeneity

Traditionally, sporadic (or idiopathic) PD has been recognized as a unique clinical entity, characterized by the presence of classical clinical signs, and by the typical pathology. The pathological hallmark of PD, the presence of LBs, was described by Friedrich Heinrich Lewy in his chapter of Max Lewandovsky’s neurology textbook, issued in Berlin in 1912. The disease-specific involvement of substantia nigra was then described by Tretiakoff in 1919, who also confirmed the existence of LBs (and named them after Lewy). The fact, that LBs are formed predominantly by pathological alpha-synuclein (αSyn) was revealed by Spillantini and colleagues only 80 years later27,28. The presence of this “alpha-synucleinopathy” was listed as the typical pathological feature of PD, either “sporadic” or “hereditary”29.

“Hereditary PD” is a roofing term, used for the Mendelian forms of PD. However, clinical manifestations of these hereditary disorders are not exactly those of “typical” PD, being in several cases rather suggestive of atypical parkinsonism. Currently, altogether 24 variants which are considered either to be “causal” for the development of parkinsonism or “associated” with its manifestation, were described30,31,32. These variants are listed in Table 1, together with a short description of the parkinsonian phenotype and morphological finding, typical for a given variant (Table 1). One can realize that from this list only 19 are “typical” phenotypes of the disease. The remaining five phenotypes rather resemble different variants of “atypical parkinsonism”33.

Table 1 Genes and cytogenetic locations in hereditary forms of Parkinson’s disease their clinical phenotypes and neuropathological findings.
Full size table

Among these mendelian forms of hereditary parkinsonism is also the parkinsonism with the “typical” phenotype, listed under the gene name LRRK2. Despite the phenotypic homogeneity of LRRK2 parkinsonism, 6 types of pathological findings have been reported in association with this phenotype so far. These included nigral LBs or diffuse LBs, nigral tau without LBs, Alzheimer’s type pathology, axonal spheroids, and degeneration of the zona compacta without LBs, tau or beta-amyloid (Aβ)34. Does it mean, that those people, who suffer from the LRRK2 variant-induced disease, do not suffer from PD? Or, is the definition of PD, as defined by the current diagnostic criteria, outdated?

The fact that the same genetic variation may cause disparate clinical manifestations and pathological findings demonstrate a complex interplay of genetic, environment, and exposures. On the other side, there is a broad spectrum of clinical manifestations of “typical” Lewy-related brain pathology (Lewy pathology, LP) associated with various rare genetic abnormalities35,36,37, where a similar combination of factors can be assumed.

Blurred differences between PDD and DLB

In some PD patients, cognitive impairment may be present. These cases were in the last decade separated into the newly established category of PDD38. Interestingly, these patients who manifested typical signs of PD together with cognitive impairment leading to overt dementia had also a different pathological correlate. This was the “typical” LB pathology accompanied by the presence of Aβ deposits in the limbic system25.

It is even more interesting when DLB is also considered. Here the amount of Aβ deposits in the brain should be much higher than in the PDD. The degree of Alzheimer’s pathology, its magnitude within the brain tissue, and the presence of cerebral amyloid angiopathy is probably the most significant pathological difference between these two phenotypes, PDD and DLB5,10,39,40,41. Nevertheless, there is no sharp pathological border between these two pictures, it is a smooth transition from the “pure” PD Lewy pathology to the “mixed” DLB pathology.

This morphological course has also another dimension; the extent of the above-mentioned changes within the brain structures. The concept of the specific spread of LB pathology and its clinical correlates, i.e. the premotor stage of PD, the motor stage of PD, and the subsequent development of cognitive deficit, was described by Braak and confirmed in further clinicopathological studies8,42,43,44,45,46,47,48. However, the results of other studies have shown that the severity of clinical symptoms, the duration of the disease, and the presence of cognitive decline or visual hallucinations do not correlate with the density of LBs49,50. In some patients with PDD, virtually no LBs were seen in the cortical regions or even outside the brainstem45. The question thus remains how the time of development of cognitive deficit and its severity is influenced by the extent and severity of concomitant AD pathology.

Pitfalls of current classification in clinical practice

From the clinician’s point of view, the clinico-pathological correlation has its unique sense in fostering the recognition and supporting further research of that specific personalized (targeted) treatment. Nonetheless, its usefulness in routine clinical practice is rather limited. For the taxonomic classification of the disease or clinical syndrome and the state of the art clinical management, the disease should be coded according to the international standards, either ICD or DSM manuals; the same counts also for the recruitment of the patients into clinical trials.

In many neurological patients, particularly those suffering from neurodegenerative disease, this might represent a serious problem. In the initial (and even in the advanced) stages the typical signs of a given disease may not be present, or vice versa, phenotypic signs of another proteinopathy may be seen34. For this reason, often complex clinical diagnostic criteria are used. Still, their use is suggested in both scientific and clinical communities, and the diagnosis made on criteria is fully accepted in both the scientific and clinical environments.

Are the clinical diagnostic criteria really useful?

Deficiency of Parkinson’s Disease Society Brain Bank Criteria (UK-PDSBB)

The clinical diagnosis of PD has been established in the past 30 years mainly based on the UK-PDSBB, first proposed in 198851. These were later validated in two clinico-pathological studies carried out at Queen Square, in Hughes’ original study in 1992 and its replication published in 2001 and 200211,52,53. The criteria were created using the clinical notes and data retrieved from general practitioner (GP) files and their retrospective correlation with pathological findings. From today’s point of view, it is therefore questionable, whether these criteria will resist the light of today’s molecular genetics and molecular biology state of the art.

The key players in this field are currently the (permanently increasing) numbers of gene variants, causal or associated with the manifestation of “typical“ PD, combined with epigenetic factors, and Braak’s concept formulated almost 20 years ago42,54,55. Nevertheless, the”hotspot“ should be the observations that the pathological neurodegenerative process might manifest in a quite different way than usually described and known (Table 2)15,35,56,57,58,59.

Table 2 Arguments against the current concept of Lewy body diseases.
Full size table

A deeper insight into the structure of UK-PDBB criteria will reveal that they are only the diagnostic criteria of the parkinsonian syndrome (in their Step 1). Parts Step 2 and Step 3 containing exclusion and supportive positive criteria are full of non-specific, frequently obsolete signs based on retrospective data, derived from often incomplete GP clinical files stored in the London Brain Bank together with the fixed brains. However, the American attempt to create an “upgrade” of UK-PDSBB criteria (named NINDS-PD criteria) was only rarely cited and never reached the level of routine clinical use60,61. So, there are enough relevant reasons to put into discussion the reliability of the 34-years old UK-PDSBB clinical diagnostic criteria. They are still used not only for the confirmation of clinical diagnosis when the patients are recruited into the clinical trials but also as a universal tool for any clinical and clinico-pathological research in PD.

An insufficient mainstay for the formulation of PDD criteria

PDD as a novel “subtype” of the LBD has been gradually recognized in the nineties of the last century. The principal reason which led to its identification was—without any doubt—the introduction of novel drugs into the PD treatment armamentarium, hand-in-hand with the introduction of the treatment of late, advanced, and complicated PD, i.e. deep brain stimulation and subcutaneous apomorphine infusions. Both these approaches led to substantially longer survival of PD patients, so the cases with manifest dementia appeared. In other words, dementia related to PD was unmasked.

When Emre in 2003 discussed the concept of PDD on a more extensive basis, he introduced two most important risk factors: older age as such, and older age at the moment of motor symptoms manifestation62. The detailed description of typical PDD phenotype together with the first suggestion of clinical diagnostic criteria was published in 2007. International experts, led by Emre, indeed performed a critical meta-analysis of published studies. They particularly extracted the neuropsychological manifestations of the typical PDD cognitive disorder, i.e. the progressive executive dysfunction, only later accompanied by the general cognitive dysfunction. They also summarized the results of 24 clinical–pathological studies, published in 1979–200538.

This meta-analysis has two weak points (Table 2). The first is the fact, that the presence of LB pathology has been in the examined studies assessed in three different ways. The second is the fact that all clinical–pathological studies were done retrospectively using brain bank specimens, while the quality of donors’ clinical files was not—and still is not—known (as a model of the accuracy of clinical data may serve for instance the paper by Guo et al.63). Therefore, it is not clear on the basis of which criteria the diagnosis of cognitive impairment was determined. Nevertheless, the whole concept of PDD clinical existence still stands on that paper. The evidence that the neuropsychological profile of DLB and PDD practically does not differ, notably in the initial phase, has become apparent in the past few years (Table 2)61,64.

Another noteworthy aspect of this meta-analysis is the fact, that until the discovery of ubiquitin staining, the dementia was attributed to s.c. “Alzheimer’s pathology” of examined brains. After the introduction of ubiquitin staining, dementia was attributed to the “pathological changes of heterogenous origin”, including “Alzheimer’s and vascular”. Only after 1997, i.e. after the discovery of αSyn and the introduction of routine examination of its presence were the cases of parkinsonism accompanied by cognitive deficit attributed to the progressive “Lewy-body pathology”.

It is also important to mention, that from the strict pathological point of view, there is practically no difference between PD and PDD. Even the experienced neuropathologist is not able to differentiate between these two “disorders”, being able only to recognize the degree of αSyn deposit progression, distribution, and density. Considering all the above-mentioned facts, it is highly questionable for which disorder have Emre and colleagues established their clinical diagnostic criteria in 2007.

The last version of DLB criteria suggests the use of the general term LBD

The next subtype of LBD is the DLB phenotype. In contrast to PDD, this phenotype is a bit better “bordered”, its definition is more intelligible and the pathological finding is unique. This is characterized by diffuse alpha-synucleinopathy, accompanied in most cases by Alzheimer’s changes, especially senile plaques.

The birth of the DLB as a nosological entity was complicated and took many years of scientific debates, consensus meetings, publications, which ran continuously for almost the last decade of the 20th century. Finally, the existence of pathological co-habitation of Alzheimer’s pathological changes together with the diffuse appearance of LBs led the expert panel to the opinion, that the former attempts to name this disease were always the attempts to describe the findings typical for DLB.

The first clinical diagnostic criteria were published by McKeith et al. in 1996; the revised version came in 2005 and the last revision is from 201765,66,67. In the last revision, the nosological entity DLB has been classified rather as “one of the phenotypes in the broader spectrum of LBD”.

According to the 2017 clinical diagnostic criteria, for the diagnosis of “probable” DLB is necessary the presence of dementia and other neuropsychological signs as the fluctuations of cognitive dysfunction and fluctuations of awareness and wakefulness, recurrent visual hallucinations, REM sleep behavioural disorder, and one or more spontaneously manifested parkinsonian motor signs. It is known that, while the parkinsonian motor symptoms are present in 25–50% of patients at the moment when the diagnosis of DLB is made, in other cases motor signs developed in the course of the disease. However, almost one-fourth of patients will never manifest any parkinsonian motor signs (Table 2). It is undoubtedly important to highlight this point. In many clinical-pathological studies, the absence of motor signs was the principal cause of diagnostic errors in cases, which were later pathologically diagnosed as DLB68,69,70. The original idea, that the parkinsonian signs appear in DLB patients shortly before the onset of cognitive and psychiatric disorder, and that they are usually rather mild, was substantially revised. Nowadays, the prevailing opinion is that the parkinsonian signs develop only after the manifestation of cognitive disorders and that they can progress into the severity similar to advanced PD. So, the clinical diagnosis of DLB is in majority of cases a diagnosis “per exclusionem”18,70,71,72.

The different phenotypes of LBD are in both clinical routine and research classified based on the mutual relationship between parkinsonian motor signs and the signs of cognitive dysfunction. The “one-year rule” has been established already in the first version of McKeith criteria65. It has been arbitrarily determined, that “if the cognitive disorder appears up to 12 months following the manifestation of parkinsonian signs, the DLB should be considered rather than PDD, no matter what is the character of cognitive disorder”. On the contrary, if the parkinsonian signs are present at the moment of cognitive disorder manifestation for a period longer than 12 months, the diagnosis should strongly incline towards PDD (Table 2). However, as was already mentioned, up to 25% of patients do not manifest motor symptoms68,71.

The presence of at least one parkinsonian sign is among the “core clinical features” of the newly established disorder that has been named “MCI-LB” and that represents the initial phase of DLB19. However, there is no substantial difference between the initial manifestation of cognitive dysfunction in PDD and DLB. So, it may be rather said, that the cognitive dysfunction which appears in a patient suffering from parkinsonian motor signs, and its gradual progression usually lead to the reconsideration of the original PD diagnosis to PDD or DLB. The clinical differentiation between PDD and DLB is possible only gradually (if possible at all), and is based on the appearance of the severity and speed of cognitive dysfunction progression, its fluctuations, and the presence of pathognomonic visual hallucinations. The last version of the DLB clinical diagnostic criteria dealt with a complicated situation by the final statement: “DLB should be diagnosed when dementia occurs before or concurrently with parkinsonism. The term PDD should be used to describe dementia that occurs in the context of well-established PD. In a practice setting the term that is most appropriate to the clinical situation should be used and generic terms such as LBD are often helpful”67.

Pathogenic mechanisms beyond LBs or LBs as an indirect indicator of this disease spectrum

As follows from the previous consideration, the current concept behind classification within this disease spectrum is still based on retrospective clinicopathological studies, which focused exclusively on the presence of LBs and their clinicopathological relevance. However, since the initial description of LBs as a pathological hallmark of PD and the formulation of Braak’s concept of the specific spread of LB pathology, evidence has been accumulating that not only LBs (and their density and distribution) are key players in this group of diseases73. So, what is the true significance of LBs in the pathogenesis of this disease spectrum, and what are the other biological relationships between the entities for which LBs are a common link?

Given the growing knowledge in the field of cell and molecular biology and molecular genetics, it seems that LBs as such do not play a major role in the pathological process and are rather an indirect indicator of these diseases. The spectrum of αSyn accumulations in LB disorders is much broader than the mere presence of LBs and involves also depositions in synapses and neurites74,75,76. The use of modern techniques has revealed further pathological features including the presence of concomitant pathology as such or synergistic relationship between concomitant and αSyn pathology, selective vulnerability due to the anatomical configuration of neurons, synaptic dysfunction, and role of genetic factors77. Unfortunately, documentation of most of these aspects is lacking in the majority of existing clinicopathological studies (Table 2).

Combined pathologies

One of the factors that may be behind the development of cognitive deficit in addition to LBs is the parallel presence of AD pathology. The combination of LBs and AD pathology predicts dementia in PD much better than the severity of any single pathology3. In clinical studies in patients with newly diagnosed PD, the cerebrospinal fluid (CSF) biomarker evidence for Aβ pathology was a significant predictor of subsequent cognitive impairment3,78. Similarly, other studies comparing patterns of CSF biomarkers between patients with DLB and PDD showed that lower levels of Aβ1-42 (combined with higher tau levels) are associated with DLB rather than PDD and are seen particularly in patients with more rapidly progressive dementia79,80.

The degree of α-synuclein phosphorylation due to the synergistic effect of AD pathology

Several in vivo and animal studies have shown a strong correlation between the extent of neurofibrillary tangles, neuritic plaques, and αSyn, suggesting synergistic effects of AD and αSyn pathology81,82.

Phosphorylation is considered as a potential mechanism for this synergy83. In experimental studies, recombinant Aβ can induce phosphorylation of αSyn at Ser129, which is considered to be a major modifier of αSyn in PDD/DLB84,85. Whereas only a small fraction of αSyn (<4%) is phosphorylated in healthy brains, a dramatic accumulation of pS129 (>90%) has been observed within LBs. These findings suggest that this posttranslational modification may play an important role in the regulation of αSyn aggregation, LBs formation, and neuronal degeneration. Higher levels of phosphorylated αSyn are present in the early stages of PDD/DLB before the occurrence of LB pathology, and levels of phosphorylated αSyn correlate with disease severity86,87.

Synuclein oligomers-induced cell loss precedes the formation of LBs

As was already mentioned, LBs considered to be the pathological hallmark of this group of diseases may not even play any causal role in their pathophysiology. As has been shown, the severity of clinical symptoms, disease duration, and presence of cognitive decline or visual hallucinations do not correlate with LBs density49,50. In some patients with PDD, virtually no LBs were seen in cortical regions or even outside the brainstem88. Furthermore, it has been shown, that cell loss can precede LB accumulation, calling into question the hypothesis that LBs are the toxic agents in PD which drive neurodegeneration89,90.

This role should be rather attributed to LBs precursors called oligomers (which cannot be seen in the light microscope) than to the LBs themselves. There is evidence that initial amorphous αSyn deposits known as “pale bodies” and “pale neurites” can mediate cell damage and later lead to a further aggregation91. Thus, neurodegeneration and cell death do not appear to be caused by LBs, but LBs rather protect “toxic” αSyn aggregates. LBs and Lewy neurites are thus more probably an indirect indicator of the disease stage and not a reflection of the whole extent of the neurodegenerative process92,93.

Different conformational properties of α-synuclein

Another aspect supporting the fact that LBs are not the key players in neurodegeneration is the description of different αSyn strains differing in conformational properties. These strains exhibit different cell toxicity and differences in the ability to induce tau protein aggregation. This is again a situation where pathological processes preceding the formation of LBs may affect the course and progression of the disease. Thus, different αSyn strains may also be the factor involved in the phenotypic variability of this group of diseases94,95. It is probable that as yet unknown genetic factors will apply here.

Selective vulnerability due to anatomical configuration of neurons

The selective vulnerability of specific neuronal populations is considered to be one of the factors involved in the specific distribution of pathological changes and the resulting clinical phenotype. The anatomical configuration of neurons (especially those with long hyperbranched axons that project widely to innervate multiple brain regions) is thought to be one of its causes77. More recent neuro-histological studies support the theory that axonal involvement is critical. αSyn aggregation starts in the axonal compartment and progresses back towards the cell body, axons become dystrophic with alterations in axonal transport, and this ultimately leads to cell death96,97.

It has been shown in PD cases that loss of dopamine is more profound at the axon terminals in the caudate and putamen than is the loss of nigral neurons; it suggests that degeneration is greatest in distal parts of the cell. Other neurons preferentially affected in PD, PDD, and DLB also show a similar anatomical configuration. These are mainly cholinergic cells of the nucleus basalis of Meynert that are strongly implicated in the pathogenesis of dementia in PD98 or serotonergic cells of the raphe nucleus, which also have extensive axon projections99. Similarly, the long unmyelinated axons of the peripheral autonomic nervous system may explain the early and prominent involvement of autonomic symptoms in both DLB and PD.

Synaptic dysfunction

The synapse is another potential location for early involvement. It seems, that presynaptic involvement is an event that precedes neuronal death. Presynaptic αSyn microaggregates can easily impact post-synaptic dendritic spines. Almost complete loss of the dendritic spines in frontal cortical neurons has been found in patients with DLB compared with age-matched controls using visualization of silver impregnation technique. A similar loss of dendritic spines was seen in the striatum in PD100,101. αSyn aggregation starts at either the synapse or axon branch points that subsequently affect vesicle trafficking and impair neurotransmitter release. This causes postsynaptic dendritic spines degeneration with loss of synaptic connections93.

Role of genetic factors

In contrast to PD (and therefore also PDD) in which numerous disease-related gene loci have been described, it has long appeared in DLB that genetic factors play here virtually no role. Only in the last decade, some variants have been identified in DLB. One of the first hints that genetics plays the same role in PD and DLB came from the studies of glucocerebrosidase (GBA). Homozygous GBA mutations are known to cause a lysosomal storage disorder (Gaucher disease), while heterozygous mutations are considered a risk factor for PD102. In DLB, a similar effect of these mutations was identified suggesting that there is an identical underlying lysosomal dysfunction present in both diseases103.

An association study showed that common variability is also involved in DLB. Variants at the APOE, SNCA, and SCARB2 loci were shown to be associated with DLB cases104. While the association of APOE variants was identical to that observed in AD, the SNCA and SCARB2 variants have different association profiles than the associations reported for the same loci in PD. Since DLB is not only characterized by LBs but also by the presence of Aβ, the association of DLB with the ε4 allele of APOE is likely driven by the Aβ pathology-promoting effect of this particular variant. Regarding the SNCA gene, the haplotype conferring risk is different for PD and DLB; in PD having an association with 3′ of gene and DLB appearing to occur 5′ of the gene. Although it is not clear at this stage what are the implications of this difference, it may influence the distribution of the LBs in the brain tissue, presumably through differential expression of the gene. The SCARB2 gene encodes lysosomal protein that is associated with PD, but unlike PD where it is not considered a major risk factor, with DLB its risk seems relatively high104. These data indicate that DLB has not only a genetic component but also that this component has a unique architecture (when compared to PD and AD) leading to the specific phenotype of DLB.

The genetic differences between PDD and DLB have, so far, not been studied in detail. Some factors predisposing to the development of earlier dementia in PDD cases have a genetic basis. For example, rapid eye movement sleep behaviour disorder, which is predictive of cognitive involvement when it occurs in patients with PD105, is more common in patients carrying GBA and SNCA mutations and less common in patients carrying LRRK2 mutations73.

Conclusions

In our opinion, the current pieces of knowledge suggest that PD, PDD, and DLB represent closely related but different, heterogeneous subtypes of an α-synuclein-associated disease spectrum. Given the controversies about the nosology of these disorders, continuous effort is necessary to distinguish among them more clearly and to clarify the underlying pathogenic mechanisms to enable effective mechanistic-based treatment, considering that no disease-modifying therapies are currently available. Further elucidation of the relations between PD, PDD, and DLB including better insight into common genetic and epigenetic risk factors and pathogenetic molecular pathways responsible for the clinical manifestations of these disorders will be necessary as the basis for future preventive and symptomatic treatment options.

Clarification and understanding of biological factors will most likely lead to a shift in the concepts of these diseases and to thinking about their natural course from a pathobiological point of view. In the era of personalized medicine, the genetic risk may be used in early recognition to predict the risk of developing cognitive deficit in patients with PD pathology and may replace the current arbitrary clinical criteria still used to distinguish DLB from the questionable concept of PDD. The understanding of synucleinopathies nosology may be the best we have to proceed with clinical trials. Nevertheless, it must be kept in mind that the artificial differentiation instead of “aggregation” of these entities may be the reason why these trials frequently fail.

Returning to the original question of whether to clinically label a group of these diseases as “Lewy body diseases” or “diseases with Lewy bodies”, the latter seems more accurate because of the current state of knowledge. Nevertheless, there should be only two: PD and DLB.