Lewy body disease (LBD) encompasses Parkinson's disease (PD), PD dementia (PDD), and dementia with Lewy bodies (DLBs). The LBDs share pathological hallmarks, Lewy bodies, and similar clinical characteristics, including rapid eye movement sleep behavior disorder (RBD), parkinsonism, fluctuating cognition, and visual hallucinations. The only distinction between PDD and DLB is the time relationship between the onset of dementia and motor symptoms (McKeith et al., 2017); however, cognitive impairment or prognosis and caregiver burden are more severe in DLB than in PDD (Walker et al., 2007; Yoon et al., 2014; Jellinger, 2018).

Along with Lewy bodies, concomitant Alzheimer's disease (AD) pathology is not uncommon and contributes to the clinical characteristics of LBD. Amyloid deposition was related to worse attention-executive and language functions in LBD patients (Biundo et al., 2021). In patients with DLB, a higher amyloid burden is associated with the diagnosis of dementia at a younger age, greater cognitive decline (Sarro et al., 2016), lower cognitive performance (Ferreira et al., 2020), decreased dopamine transporter (DAT) activity (Yoo et al., 2020), and more extensive gray matter atrophy (Sarro et al., 2016; Lee et al., 2018b). Similarly, compared to a lower amyloid burden, a higher amyloid burden is related to reduced cognitive performance (McMillan and Wolk, 2016), faster cognitive decline (Gomperts et al., 2013; Yarnall et al., 2014), shorter time-to-dementia onset (Compta et al., 2011), and increased postural instability and gait difficulty (Muller et al., 2013) in patients with PD.

Pathological and imaging studies have provided ample evidence that patients with DLB have a higher amyloid burden than those with PD (Petrou et al., 2012; Hepp et al., 2016). Additionally, the type and distribution of amyloid in DLB differ from those in PD. For example, the neuritic plaque score is higher in DLB than in PDD, and amyloid is more frequently observed in the entorhinal cortex, amygdala, and putamen in DLB than in PDD (Hepp et al., 2016). Another study showed that patients with DLB had frequent amyloid deposition in the midbrain, while patients with PD had less or scent amyloid deposition in the midbrain (Fujishiro et al., 2010; Sierra et al., 2016; Jellinger, 2023). Moreover, recent studies have reported that the different clinical and pathological features of LBD may be ascribed to the underlying differences in α-synuclein strains (Candelise et al., 2019; Van der Perren et al., 2020), which may also lead to different interactions with other proteins (Sengupta et al., 2020). One autopsy study showed different distribution patterns of α-synuclein pathology in the LBD spectrum depending on the concomitant AD (Toledo et al., 2016). Although the biological effects of mixed pathologies or interactions between α-synuclein and other toxic proteins on neuronal degeneration have been suggested (Sengupta et al., 2020), little is known about how different α-synuclein strains would interact with other proteins.

How does concomitant amyloid accumulation affect the nigrostriatal dopaminergic system in the LBD spectrum and their relationship with parkinsonian motor symptoms is elusive? In this study, we hypothesized that amyloid pathology may interact in a different manner with the nigrostriatal dopaminergic system in DLB and PD; therefore, the patterns of nigrostriatal DAT availability would differ. To test this hypothesis, we performed an interaction analysis in patients with DLB and PD who had undergone ¹⁸F-florbetaben positron emission tomography (¹⁸F-FBB PET) to examine specific ¹⁸F-N-(3-fluoropropyl)-2β-carbon ethoxy-3β-(4-iodophenyl) nortropane (¹⁸F-FP-CIT) PET uptake, which interacts with amyloid depending on the type of LBD.

In this study, we found distinct amyloid-dependent or amyloid-independent ¹⁸F-FP-CIT PET patterns in patients with LBD. First, the presence of amyloid had a detrimental effect on the DAT availability in the anterior and posterior putamen in the PD group only. Second, the presence of amyloid was not associated with DAT availability in any striatal subregion in the DLB group. Third, the presence of amyloid was related to the increased asymmetricity of posterior putaminal DAT availability in both PD and DLB groups. Finally, additional amyloid pathology was not directly related to the parkinsonian motor symptom, but indirectly related to the increased UPDRS Part III score, which was mediated by DAT availability in the anterior and posterior putamen in the PD group only. These data suggest that the interaction between amyloid pathology and the nigrostriatal dopaminergic system is different between DLB and PD groups.

Although the location and mechanism of the interactions between toxic protein pathologies are yet to be elucidated in neurodegenerative diseases, several clinicopathological studies have provided evidence of a close intergroup association between toxic proteins. In patients with LBD including PD or DLB, a pathological study reported that the deposition of amyloid detected by [¹¹C] Pittsburgh compound B had a significant correlation with Lewy body score and the severity of the stage of neurofibrillary tangle (Colom-Cadena et al., 2013; Shirvan et al., 2019). Furthermore, the presence of amyloid was associated with greater α-synuclein accumulation in LBD (Miller et al., 2022). In addition, Lewy body counts were significantly correlated with senile plaque counts and, to a lesser extent, neurofibrillary tangles in patients with PDD (Apaydin et al., 2002). Functional imaging studies using multiple radiotracers reported similar results that the burden of amyloid and tau proteins was closely correlated with each other in patients with DLB (Lee et al., 2018a). Importantly, the pattern of tau accumulation in DLB patients was distinctive compared to patients with AD.

Interestingly, we found that the presence of amyloid pathology in LBD had a distinct effect on the patterns of DAT availability depending on the specific LBD disease. Considering that subregional DAT availability in the striatum corresponds to dopaminergic neuronal projections from the substantia nigra (SN) (Fearnley and Lees, 1991), striatal DAT patterns in this study may represent region-specific dopaminergic neuronal loss in the SN. This study showed that the detrimental effect of amyloid on the lateral tier of nigra dopaminergic neurons that project to the putamen was dependent on the specific disease type of LBD. This means that the presence of amyloid led to significantly decreased DAT availability in the putamen, respective of the disease groups. In contrast, amyloid-independent DAT availability in the caudate and striatum was not specific to the type of LBD, such that the presence of amyloid had no significant effects on dopaminergic neurons outside the lateral portion of the SN in both groups. In addition, the detrimental effect of the presence of amyloid on the AR between the bilateral posterior putamen was significant in the PD group. This suggests that concomitant amyloid pathology influences more on the more affected side of the substantia nigra and that the interaction between amyloid and α-synuclein could have a synergistic effect rather than an additive effect. However, the presence of amyloid was associated with increased AR, without having a significant effect on the putaminal DAT availability in the DLB group. This suggests that the interaction between amyloid and α-synuclein in DLB patients may be present, at least on the more affected side of the substantia nigra, but overall, the interaction is less prominent than that in the PD group. These results may provide indirect evidence for the region- and disease-dependent effects of amyloid on the nigra dopaminergic neurons.

Recently, accumulating evidence has suggested that various clinicopathological traits in α-synucleinopathies may be ascribed to different α-synuclein strains (Shahnawaz et al., 2020; Holec and Woerman, 2021; Ayers et al., 2022). In terms of α-synuclein characteristics in patients with DLB, toxicity to the dopaminergic neurons in the SN appears to be less potent compared to α-synuclein in PD (Van der Perren et al., 2020). Moreover, Toledo et al. showed that demented subjects with AD and Lewy body-related pathology (Dem-AD-LB, which could partially overlap DLB with an amyloid group in our study) had less reduced substantia nigra neurons than PD with AD or PD without AD, and DLB without AD. Dem-AD-LB patients were less likely to show involvement in Lewy body-related pathology in the brainstem (Toledo et al., 2016). Therefore, disease-specific characteristics of α-synuclein strains would explain the more detrimental effects of concomitant amyloid on the putaminal DAT availability observed in the PD group compared to the DLB group. This result implies that the α-synuclein strain interacts differently with β-amyloid, depending on the specific type of α-synucleinopathies, as well as on specific regions of nigra dopaminergic neurons. Further clinicopathological data are needed to uncover the relationship between α-synuclein strains and other toxic proteins.

This study also demonstrated that additional amyloid pathology was indirectly related to the increased UPDRS Part III score. In addition, the detrimental effect of amyloid on parkinsonian motor symptoms in patients with LBD was disease-specific. The indirect effect of amyloid was present in the PD group only. In terms of neuroanatomical correlates of parkinsonian motor symptoms, it has been postulated that parkinsonism in patients with PD is based on neuronal loss in the SN, whereas parkinsonism in DLB is attributed to widespread anatomical structures beyond the SN (Postuma et al., 2012). Given that the different neural correlates of parkinsonism exist in LBD, it is speculated that the amyloid pathology may exert effects differently in the distinct anatomical area responsible for parkinsonism, depending on the type of LBD. The mediation analysis revealed that the effect of amyloid on motor deficits was mediated by DAT availability in the anterior and posterior putamen in the PD group. Considering that the effect of amyloid on DAT availability in the anterior and posterior putamen was LBD type-specific, additional neurotoxic effects of amyloid on nigra dopaminergic neurons projecting into the putamen that exhibits a disease-specific pattern may result in additive motor deficits in the PD group. To date, the location and mechanisms of the interaction between amyloid and synuclein pathologies remain to be elucidated. In our study, the presence of amyloid was associated with putaminal dopaminergic depletion, further correlating with motor dysfunction. Therefore, among striatal subregions in patients with PD, the putamen might be the primary region where the interaction between amyloid and synuclein pathologies occurs. This result may provide a clue for clinical relevance regarding the interaction of mixed pathologies in neurodegenerative diseases.

In terms of spatial or asymmetric patterns, the AR and ISR did not differ between the DLB and PD groups. Thus, there was no difference between the two groups in the asymmetry or anterior–posterior gradient of substantia nigra degeneration evaluated by DAT imaging. However, when considering the presence of amyloid, the anterior–posterior gradient of DAT availability in the putamen was more prominent in the PD group relative to the DLB group. There has been controversy over differences in patterns of striatal DAT availability between DLB and PD patients. The patterns of striatal DAT availability were compatible between two groups (Colloby et al., 2004; Marquie et al., 2014; Gomperts et al., 2016; Pilotto et al., 2019), whereas others reported lower caudate DAT availability in DLB patients and greater asymmetricity or antero-posterior gradient of DAT availability in PD patients (Ransmayr et al., 2001; O'Brien et al., 2004; Walker et al., 2004). Our results imply that the presence of amyloid may influence the spatial or asymmetric distribution of DAT availability in the PD and DLD groups, and thus, the prevalence of amyloid positivity may contribute to the different results in previous studies.

Our study has some limitations. First, the diagnosis of LBD was not pathologically confirmed; thus, an alternative diagnosis of DLB may be possible. However, all DLB patients in this study met the probable DLB criteria from the 2017 revised criteria for DLB. In addition, we used additional supportive imaging biomarkers to diagnose patients more accurately with DLB using ¹⁸F-FP-CIT PET. Second, there was a significant difference in the proportion of patients with dementia between the two groups. Although we tried to overcome this difference by using cognitive status as covariates, the results should be interpreted with caution. Third, we used a dichotomized approach to amyloid rather than global SUVR as a continuous variable. Traditionally, quantitative cutoffs for amyloid PET positivity have been defined to discriminate AD patients from non-AD patients. However, there have been some studies dealing with the subthreshold amyloid accumulation which has clinical significance for a worse prognosis (Bischof and Jacobs, 2019). Therefore, we performed additional analysis with global SUVR as a continuous variable. However, there was no significant interaction effect or independent effect of amyloid when we treated the amyloid uptake as a continuous variable (Supplementary Table 1). This might suggest that a substantial accumulation of amyloid is needed to have significant interaction with PD. Finally, the patients in our study had a relatively older onset age. Older onset age could affect the prevalence of amyloid positivity or be associated with aggravated nigra neuronal cell death. However, there was no difference in the onset age between the two groups (74.07 ± 6.47 in the DLB group and 73.81 ± 6.56 in the PD group).

In conclusion, DLB and PD interact with amyloid pathology differently with regard to nigrostriatal dopaminergic dysfunction, which could provide indirect evidence of the presence of different α-synuclein strains in synucleinopathies and their distinct interaction with amyloid pathology.

The data and code used in this work are available from the corresponding author upon reasonable request.

The studies involving human participants were reviewed and approved by Yonsei University Severance Hospital Institutional Review Board. Written informed consent for participation was not required for this study in accordance with the national legislation and the institutional requirements.

KB, JC, and PL contributed to the conception and design of the manuscript. KB, SC, HY, and YS organized the database. KB, MP, YL, and JC performed the statistical analysis. KB wrote the first draft of the manuscript. JC, MP, and PL wrote sections of the manuscript. All authors contributed to the manuscript revision, read, and approved the submitted version.