Evidence from a clinical perspective

According to the most recently revised criteria for the clinical diagnosis of DLB (McKeith et al., 2005), dementia (defined as a progressive cognitive decline with functional impairment) is a requirement for the diagnosis. Core features of this clinical entity include: fluctuations (for which specific measures have now been designed to improve the specificity of this core feature), visual hallucinations (which should be formed, complex and recurrent), and Parkinsonism. As stated, suggestive features which add to the diagnostic accuracy of the criteria include REM sleep behavior disorder (RBD), neuroleptic sensitivity, and low dopamine uptake in the basal ganglia as demonstrated by functional imaging (PET or SPECT). In addition, a number of supportive features are included in the criteria (these do not add to the sensitivity or specificity of the criteria), which include: repeated falls, autonomic dysfunction, other forms of hallucinations, loss of consciousness, delusions, imaging and electroencephalogram (EEG) findings. Some indicators that the diagnosis is not DLB were also provided, and these include evidence of another condition, in particular cerebrovascular disease, that may explain the clinical syndrome, as well as the occurrence of Parkinsonism only in late stages of the disease (McKeith et al., 2005). The temporal sequence of symptoms appears to be de-emphasized in the revised criteria. Strict adherence to the “one year rule” is not a requirement for the diagnosis. In fact, concurrent presentation of cognitive and motor signs are acceptable for the diagnosis of DLB; however, the diagnosis of PDD should be made only when the diagnosis of PD has been established. The “one year rule” is still being recommended for differentiation between DLB and PDD in the context of clinical research (McKeith et al., 2005).

Clinical diagnostic criteria for PDD are available; however, it has not been validated with pathological correlation studies (Emre et al., 2007). Current criteria rest on the fundamental notion that a diagnosis of idiopathic PD must be established prior to the onset of cognitive dysfunction. The cognitive deficit must present insidiously, progress slowly, and must involve more than one cognitive domain, representing a decline from baseline with significant cognitive impairment. A series of associated cognitive features are provided and include impaired attention, impairment in executive function, impaired visuo-spatial function, memory and language. Associated behavioral features which were supportive of the diagnosis of PDD included: apathy, personality changes, hallucinations, delusions, and excessive daytime sleepiness. Features which make the diagnosis of PDD less likely include the existence of co-morbidities and a dubious time course. Finally, a series of conditions were provided which, when present, make it impossible to diagnose PDD accurately, and these include: acute confusional states which may be secondary to systemic illness, major depression, and features highly compatible with the diagnosis of a vascular dementia. Based on these features, the designations of “probable” and “possible” PDD are made when core features were present, and supporting associated features were also present with few or no “red flags” to cast doubt on the diagnosis (Emre et al., 2007). Further guidelines to aid in the diagnosis of PDD based on these criteria were provided by the Movement Disorder Society Task Force, where specific algorithms were provided, including particular measures to assess cognitive dysfunction for both experienced and non-experienced clinicians who may be asked to make the diagnosis of PDD in clinical practice (Dubois et al., 2007).

Criteria aside, there are differences across the clinical spectrum between DLB and PDD. Apart from an earlier presentation of dementia in DLB, there is, at least, a quantitative difference in terms of severity of executive dysfunction; where it is more pronounced in DLB than in PDD. Therefore, the presence of a dysexecutive syndrome may not differentiate PDD from DLB, but, in addition to Parkinsonism, fluctuations and hallucinations, it does help to differentiate LBDs from early AD (which shows stronger early amnestic features). Relative to PDD, DLB patients show greater problems with conceptual and attentional tasks (Downes et al., 1998; Aarsland et al., 2003). Fluctuations and visuo-spatial impairment have been investigated and no significant qualitative or quantitative difference has been found between these two entities, although this may be related to our ability to assess this issue objectively in casual questioning of the patient and caregiver. Some differences have also been reported in terms of psychiatric features with a significantly increased frequency of depression, hallucinations and delusions in DLB compared with PDD (Aarsland et al., 2004).

In terms of motor symptoms, DLB patients tend to present with more ambulatory dysfunction, symmetric motor symptoms, and less severe Parkinsonian features in general (Burn et al., 2006a), compared with PDD patients. Furthermore, neuroleptic sensitivity has been shown to be not only more frequent in DLB (Aarsland et al., 2004) but also more severe when present, with a lower mortality for neuroleptic sensitivity reactions in PDD compared to DLB (Ferman et al., 2007).

Evidence from a pathological perspective

The sensitivity and specificity of clinical diagnostic criteria for DLB and PDD depend not only on correct pathological correlation, but also on accurate pathological diagnostic criteria. The revised pathological criteria for the diagnosis of DLB takes into account not only LB pathology and pattern of distribution, but also AD pathology. The recommendations propose a specific method for quantifying LB pathology, and recommend NIA-Reagan criteria for assessing amyloid β (Aβ) plaques and Braak staging for assessing neurofibrillary tangles (NFTs). A probability that the findings are consistent with the DLB clinical syndrome is then assigned to each case depending on the pathological findings (McKeith et al., 2005). No criteria for the pathologic diagnosis of PDD have currently been proposed. Issues at the core of the pathological distinction of DLB and PDD include: type of pathology (PD type: LBs, LNs; or AD type: Aβ plaques, NFTs), correlation between pathologic change and clinical symptomatology, topographic distribution of pathology and timing of pathologic change. Aβ burden is not only abundant in DLB but has been found to correlate with severity of dementia in this patient population (McKeel et al., 2004). Tau positive neuritic pathology correlates not only with synaptic loss in DLB but also strongly with clinical phenotype (Revuelta et al., 2008; Merdes et al., 2003). Aβ plaques and NFTs, although present in PDD, are not as prominent and do not generally correlate well with dementia severity, although some studies have found an association (SantaCruz et al., 1999; Harding and Halliday, 2001; Jellinger et al., 2002; Mastaglia et al., 2003).

Cortical LBs and LNs are the principal pathologic entity responsible for the clinical dementia syndrome in both PDD and DLB, and correlate well with severity of dementia in both entities (Hurtig et al., 2000; Aarsland et al., 2005b), however, it appears that the type and degree of AD pathology present is what determines which specific clinical syndrome develops.

Another significant difference between these two entities is the distribution of pathology; where temporal lobe structures are affected disproportionately in DLB and limbic structures are affected with higher frequency in PDD (Kovari et al., 2003). Furthermore, the degree of LB pathology in temporal lobe structures correlates well with well-formed visual hallucinations (a characteristic clinical feature) in DLB (Harding et al., 2002).

Finally, the timing in which these pathological changes arise needs further consideration, particularly in distinguishing these two clinical syndromes that have classically been defined by onset of cognitive dysfunction. Although LB pathology has been correlated with clinical symptoms in LBDs, multiple studies have failed to correlate the amount of LB pathology and early onset of cognitive dysfunction, as in DLB. These findings are consistent with the idea that DLB is a more pathologically heterogeneous disorder, and earlier symptom onset is likely to be related to concurrent pathology (Aarsland et al., 2004).

Evidence from a neurochemical perspective

Differences in the chemical physiology can be divided into dopaminergic and cholinergic dysfunctions. Degree of cell loss in the caudate and substantia nigra is comparable in PD and DLB, with marked involvement in both. However, cell loss in the substantia nigra correlates with duration of illness, which is, by definition, longer in PD. It follows that there is more significant striatal dopamine loss in PD than in DLB. D2 receptor upregulation and dopamine turnover is also greater in PD and PDD than in DLB (Piggott et al., 1999).

Involvement of the basal forebrain results in cortical cholinergic deficits in DLB and PDD which are not present in PD without dementia. However, several studies have noted differences in cholinergic deficits between these two disorders, including: striatal cholinergic deficits present only in DLB, thalamic cholinergic deficits found only in DLB and PDD with Parkinsonism, and cholinergic deficits which correlate with visual hallucinations in DLB but not in PDD. Furthermore, M1 cholinergic receptors were found to be significantly reduced in DLB, but not in PD, with PDD patients showing intermediate levels (Aarsland et al., 2004).

Evidence from an imaging perspective

Imaging studies can be characterized as those which investigate structural, metabolic, perfusion or neurotransmitter changes. Several studies have been performed using MRI for structural changes, PET for glucose metabolism, cholinergic transmission, acetylcholinesterase activity and 18F-dopa uptake as well as SPECT to document differences in perfusion. Although differences between Parkinsonian patients with and without dementia and differences between LBDs and AD have been found, no consistent evidence to differentiate DLB from PDD has been uncovered. Future studies with Pittsburg Compound B (PIB) PET may be useful in determining Aβ burden and henceforth in differentiating DLB from PDD (Lippa et al., 2007; Emre et al, 2007).