In this section, the characteristic findings of 18F-fluorodeoxyglucose positron emission tomography (FDG-PET) for cerebral metabolism in PET imaging, ^99mTc-ethyl cysteinate dimer (ECD)-SPECT and N-isopropyl-p-123I-iodoamphetamine (IMP)-SPECT imaging for cerebral perfusion in SPECT imaging are described. Lastly, we briefly review the evidence for amyloid deposition finding in DLB. Currently, striatal dopamine transporter (DAT) imaging (e.g., FP-CIT SPECT) is the only neuroimaging tool accepted as the suggestive feature in the revised consensus criteria for DLB [9, 10]. Details of this tool would be described in another chapter. In addition to DAT ligands, several other representative radioligands for molecular and functional imaging are used to support the diagnosis of DLB, as shown in Table 10.1. In Japan, there is no functional cerebral neuroimaging modality approved by the Japanese Pharmaceutical and Medical Devices Agency (PMDA) for diagnosing DLB, other than FP-CIT SPECT.

ECD-/IMP-SPECT and FDG-PET scans are helpful for the clinical diagnosis of DLB [11]. The characteristic occipital hypometabolism in DLB was first reported by Albin et al. [12]. Thereafter, to discriminate DLB from AD, Minoshima et al. clarified that the hypometabolism seen in the occipital lobe, especially in the primary visual cortex (PVC) detected by FDG-PET imaging, is a distinguishing finding in DLB subjects, and there is a high sensitivity and specificity (90 % and 80 %, respectively) [13]. In probable DLB, compared with the PVC, metabolism in the visual association cortex (VAC) is relatively preserved (Fig. 10.2a, b). However, the pathophysiological background of cerebral hypometabolism in the occipital lobe remains unclear. Neuropathologically, in DLB patients, the occipital lobe, which includes both the PVC and the VAC, is not a predominantly vulnerable area for Lewy pathology, such as Lewy bodies and Lewy neurites. Consequently, there is no severe cortical neuronal loss or occipital lobar atrophy. We previously reported that Lewy pathology in DLB brains does not occur in the neuronal cell bodies, including Lewy bodies (LBs), but it occurs in the axonal terminals, including Lewy neurites (LNs) initially [14], and secondary LBs may be formed by transneuronal degeneration in the regions where the degenerative axonal terminals are found [14, 15]. Hence, Lewy pathology in the occipital area may arise along with degeneration of the visuo-amygdaloid pathway [16, 17]. In fact, neuroanatomically, it is known that visual cortex has a connection with the amygdala, one of the most vulnerable areas for Lewy pathology, through the relay of the inferior temporal cortex [18].

A recent neuropathological survey showed that the levels of synaptic proteins such as synaptophysin and syntaxin and the choline acetyltransferase (ChAT) activity in PVC in DLB were significantly lower in those in AD and aged-control groups [19]. Moreover, the densities of LBs were correlated with these synaptic and ChAT changes, implying the association between molecular changes and functional neuroimaging in PVC [19].

In addition to PVC, hypometabolism in the posterior cingulate cortex (PCC) and temporoparietal association area is the frequently accompanying finding in DLB patients [12, 20] (Fig. 10.2b). Minoshima et al. [13] indicated with multivariate analysis that there is a pathophysiological difference between FDG-PET hypometabolism in the temporoparietal association area and that in the occipital lobe area, especially in the PVC, in DLB patients. The detection of Lewy pathology in brains is essential for the pathological diagnosis of DLB. Previous studies, however, have shown that Lewy pathology and AD pathology mostly coexist in DLB brains [21, 22]. One of them reported that, among 42 autopsied DLB brains, about 80 % had a moderate to severe extent of concomitant AD pathology (National Institute on Aging (NIA) criteria: intermediate to high), and the remaining 20 % of brains had only mild AD pathology (NIA: low) [22]. Using postmortem DLB brains, we have previously performed the quantitative assessment regarding α-synuclein, tau, and β-amyloid immunoreactivities, in addition to neuronal number in the occipital, PCC, and temporoparietal association areas, and compared them among the DLB group with/without accompanied Alzheimer-type pathology (ADP) and the AD group [17]. As a result, in the PCC and temporoparietal association area, neuronal loss was milder in the DLB groups than in the AD group, although there were no differences between the two DLB groups. Both tau and β-amyloid immunoreactivities differed between the two DLB groups, being much higher in the DLB group with ADP than in the DLB group without ADP, although they were rather lower in the DLB group with ADP than in the AD group. In the DLB groups, tau and β-amyloid immunoreactivities participated in neuronal loss by multivariate analysis, and the severity of NFT pathology was correlated with the degree of neuronal loss. Similarly, others emphasize that the hypometabolism in the PCC and temporoparietal association area in AD patients may reflect a secondary defect from AD pathology in the medial temporal area [23–25]. These findings suggest that both primary and secondary effects of AD pathology might play an important role in the hypometabolism of the PCC and temporoparietal association areas among DLB patients. Actually, it has been accepted that the PCC is relatively spared from hypometabolism during DLB compared to AD, and so this has been recently called “cingulate island sign (CIS)” [26, 27]. It is intriguing that this characteristic sign is in accordance with our previous findings [17]. Originally, the relative preservation of cingulate cortex was reported by Japanese researchers using the quantitative analysis of FDG-PET data [28]. Its clinical value for differentiating DLB from AD has followed by Lim et al [27]. Through the neuropathological retrospective investigation, Graff-Radford et al. clarified that the CIS ratio did not correlate with amyloid burden, but did correlate with lower tau burden (Braak NFT staging). Then, they concluded that this findings indicate that the CIS in FDG-PET might be a surrogate marker for coexisting tau pathology in DLB [26]. In terms of the clinical relevance of hypometabolism in the temporoparietal and precuneus regions, we recently reported that visual and extracampine hallucinations were more frequently seen in a DLB subgroup with hypometabolism in these regions, than in a subgroup without hypometabolism in these regions [29].