A Novel Machine Learning Algorithm Predicts Dementia With Lewy Bodies Versus Parkinson's Disease Dementia Based on Clinical and Neuropsychological Scores

OBJECTIVE

Our aim was to develop a machine learning algorithm based only on non-invasively clinic collectable predictors, for the accurate diagnosis of these disorders.

METHODS

This is an ongoing prospective cohort study (ClinicalTrials.gov identifier NCT number NCT04448340) of 78 PDD and 62 DLB subjects whose diagnostic follow-up is available for at least 3 years after the baseline assessment. We used predictors such as clinico-demographic characteristics, 6 neuropsychological tests (mini mental, PD Cognitive Rating Scale, Brief Visuospatial Memory test, Symbol digit written, Wechsler adult intelligence scale, trail making A and B). We investigated logistic regression, K-Nearest Neighbors (K-NNs) Support Vector Machine (SVM), Naïve Bayes classifier, and Ensemble Model for their ability to predict successfully PDD or DLB diagnosis.

RESULTS

The K-NN classification model had an accuracy 91.2% of overall cases based on 15 best clinical and cognitive scores achieving 96.42% sensitivity and 81% specificity on discriminating between DLB and PDD. The binomial logistic regression classification model achieved an accuracy of 87.5% based on 15 best features, showing 93.93% sensitivity and 87% specificity. The SVM classification model had an accuracy 84.6% of overall cases based on 15 best features achieving 90.62% sensitivity and 78.58% specificity. A model created on Naïve Bayes classification had 82.05% accuracy, 93.10% sensitivity and 74.41% specificity. Finally, an Ensemble model, synthesized by the individual ones, achieved 89.74% accuracy, 93.75% sensitivity and 85.73% specificity.

CONCLUSION

Machine learning method predicted with high accuracy, sensitivity and specificity PDD or DLB diagnosis based on non-invasively and easily in-the-clinic and neuropsychological tests.

Increased expression of pathological markers in Parkinson's disease dementia post-mortem brains compared to dementia with Lewy bodies

BACKGROUND

Parkinson's disease (PD) and dementia with Lewy bodies (DLB) are common age-related neurodegenerative diseases comprising Lewy body spectrum disorders associated with cortical and subcortical Lewy body pathology. Over 30% of PD patients develop PD dementia (PDD), which describes dementia arising in the context of established idiopathic PD. Furthermore, Lewy bodies frequently accompany the amyloid plaque and neurofibrillary tangle pathology of Alzheimer's disease (AD), where they are observed in the amygdala of approximately 60% of sporadic and familial AD. While PDD and DLB share similar pathological substrates, they differ in the temporal onset of motor and cognitive symptoms; however, protein markers to distinguish them are still lacking.

METHODS

Here, we systematically studied a series of AD and PD pathogenesis markers, as well as mitochondria, mitophagy, and neuroinflammation-related indicators, in the substantia nigra (SN), temporal cortex (TC), and caudate and putamen (CP) regions of human post-mortem brain samples from individuals with PDD and DLB and condition-matched controls.

RESULTS

We found that p-APPT668 (TC), α-synuclein (CP), and LC3II (CP) are all increased while the tyrosine hydroxylase (TH) (CP) is decreased in both PDD and DLB compared to control. Also, the levels of Aβ42 and DD2R, IBA1, and p-LRRK2S935 are all elevated in PDD compared to control. Interestingly, protein levels of p-TauS199/202 in CP and DD2R, DRP1, and VPS35 in TC are all increased in PDD compared to DLB.

CONCLUSIONS

Together, our comprehensive and systematic study identified a set of signature proteins that will help to understand the pathology and etiology of PDD and DLB at the molecular level.

Severe and Regionally Widespread Increases in Tissue Urea in the Human Brain Represent a Novel Finding of Pathogenic Potential in Parkinson's Disease Dementia

Widespread elevations in brain urea have, in recent years, been reported in certain types of age-related dementia, notably Alzheimer’s disease (AD) and Huntington’s disease (HD). Urea increases in these diseases are substantive, and approximate in magnitude to levels present in uraemic encephalopathy. In AD and HD, elevated urea levels are widespread, and not only in regions heavily affected by neurodegeneration. However, measurements of brain urea have not hitherto been reported in Parkinson’s disease dementia (PDD), a condition which shares neuropathological and symptomatic overlap with both AD and HD. Here we report measurements of tissue urea from nine neuropathologically confirmed regions of the brain in PDD and post-mortem delay (PMD)-matched controls, in regions including the cerebellum, motor cortex (MCX), sensory cortex, hippocampus (HP), substantia nigra (SN), middle temporal gyrus (MTG), medulla oblongata (MED), cingulate gyrus, and pons, by applying ultra-high-performance liquid chromatography-tandem mass spectrometry (UHPLC-MS/MS). Urea concentrations were found to be substantively elevated in all nine regions, with average increases of 3–4-fold. Urea concentrations were remarkably consistent across regions in both cases and controls, with no clear distinction between regions heavily affected or less severely affected by neuronal loss in PDD. These urea elevations mirror those found in uraemic encephalopathy, where equivalent levels are generally considered to be pathogenic, and those previously reported in AD and HD. Increased urea is a widespread metabolic perturbation in brain metabolism common to PDD, AD, and HD, at levels equal to those seen in uremic encephalopathy. This presents a novel pathogenic mechanism in PDD, which is shared with two other neurodegenerative diseases.

Apathy as a behavioural marker of cognitive impairment in Parkinson's disease: a longitudinal analysis

Background

Understanding the longitudinal course of non-motor symptoms, and finding markers to predict cognitive decline in Parkinson’s disease (PD), are priorities. Previous work has demonstrated that apathy is one of the only behavioural symptoms that differentiates people with PD and intact cognition from those with mild cognitive impairment (MCI-PD). Other psychiatric symptoms emerge as dementia in PD develops.

Objective

We explored statistical models of longitudinal change to detect apathy as a behavioural predictor of cognitive decline in PD.

Methods

We followed 104 people with PD intermittently over 2 years, undertaking a variety of motor, behavioural and cognitive measures. We applied a linear mixed effects model to explore behavioural factors associated with cognitive change over time. Our approach goes beyond conventional modelling based on a random-intercept and slope approach, and can be used to examine the variability in measures within individuals over time.

Results

Global cognitive scores worsened during the two-year follow-up, whereas the longitudinal evolution of self-rated apathy scores and other behavioural measures was negligible. Level of apathy was negatively (− 0.598) correlated with level of cognitive impairment and participants with higher than average apathy scores at baseline also had poorer cognition. The model indicated that departure from the mean apathy score at any point in time was mirrored by a corresponding departure from average global cognitive score.

Conclusion

High levels of apathy are predictive of negative cognitive and behavioural outcomes over time, suggesting that apathy may be a behavioural indicator of early cognitive decline. This has clinical and prognostic implications.

Electronic supplementary material

The online version of this article (10.1007/s00415-019-09538-z) contains supplementary material, which is available to authorized users.

Parkinson's-adapted cognitive stimulation therapy: a pilot randomized controlled clinical trial

Cognitive stimulation therapy (CST) is widely used with people with dementia, but there is no evidence of its efficacy in mild cognitive impairment or dementia in Parkinson's disease (PD-MCI; PDD) or dementia with Lewy bodies (DLB). We aimed to explore the impact of 'CST-PD', which is home-based, individualized CST adapted for this population. In a single-blind, randomized controlled exploratory pilot trial (RCT), we randomized 76 participant-dyads [PD-MCI (n = 15), PDD (n = 40), DLB (n = 21) and their care partners] to CST-PD or treatment as usual (TAU). CST-PD involves home-based cognitively stimulating and engaging activities delivered by a trained care partner. Exploratory outcomes at 12 weeks included cognition (Addenbrooke's Cognitive Evaluation; ACE-III), neuropsychiatric symptoms and function. In care partners, we assessed burden, stress and general health status. Relationship quality and quality of life were assessed in both dyad members. At 12 weeks, the ACE-III showed a nonstatistically significant improvement in the CST-PD group compared with the TAU group, although neuropsychiatric symptoms increased significantly in the former. In contrast, care partners' quality of life (d = 0.16) and relationship quality ('satisfaction', d = 0.01; 'positive interaction', d = 0.55) improved significantly in the CST-PD group, and care burden (d = 0.16) and stress (d = 0.05) were significantly lower. Qualitative findings in the CST-PD recipients revealed positive 'in the moment' responses to the intervention, supporting the quantitative results. In conclusion, care-partner-delivered CST-PD may improve a range of care-partner outcomes that are important in supporting home-based care. A full-scale follow-up RCT to evaluate clinical and cost effectiveness is warranted.

Dementia-Related Functional Disability in Moderate to Advanced Parkinson's Disease: Assessment Using the World Health Organization Disability Assessment Schedule 2.0

Dementia is a common nonmotor condition among people with moderate or advanced Parkinson’s disease (PD). Undoubtedly, profound motor symptoms cause remarkable impairment in daily activities; however, dementia-related disabilities have not been thoroughly investigated, especially not with consideration of differences according to sex. The present study used the World Health Organization Disability Assessment Schedule 2.0 (WHODAS 2.0) to compare the functional disability between men and women with PD (PwP) with and without dementia. This study employed a registry of disability evaluation and functional assessment using the Taiwan Data Bank of Persons with Disability between July 2012 and October 2018. To investigate dementia-related disability in PwP, 1:1 matching by age and Hoehn-Yahr stage was conducted, which resulted in the inclusion of 1605 study participants in each group. The present study demonstrated that among the six major domains of WHODAS 2.0, the section of “Getting alone with others” was significantly worse in both genders of PwP with dementia; however, a greater disability in fulfilling activities of daily living was only noted in male PwP with dementia but not in their female counterparts. Neither the inability to provide self-care nor participation were significantly different between the sexes. Our findings suggested that deteriorating social relationships were a dementia-related disability in all PwP at the moderate and advanced disease stages. Regarding the performance of activities of daily living, deterioration was related to dementia only in male PwP. Such disabilities could indicate cognitive impairment in people with moderate or advanced PD and could be used as an indicator for the early detection of dementia in PwP by healthcare professionals through the easier functional assessment of the WHODAS 2.0.

Laminar degeneration of frontal and temporal cortex in Parkinson disease dementia

To investigate cortical laminar degeneration in Parkinson's disease (PD) with dementia (PDD). Changes in density of α-synuclein-immunoreactive Lewy bodies (LB), Lewy neurites (LN), and Lewy grains (LG) together with surviving neurons, abnormally enlarged neurons (EN), vacuoles, and glial cell nuclei were measured across cortical laminae of frontal and temporal cortex in fifteen cases of PDD using quantitative methods and polynomial curve-fitting. Most frequently, LB and LN were distributed across all laminae, while LG were distributed in upper cortical laminae. Low densities of EN were present in most cases distributed across all cortical laminae. Densities of vacuoles and glia were greatest in upper and lower cortical laminae, respectively. In most gyri, there were no spatial correlations between the densities of LB, LN, and LG. Cortical degeneration of frontal and temporal lobes in PDD affects all cortical laminae. Laminar distributions may result from the spread of α-synuclein pathology from subcortical regions and subsequent spread via the cortico-cortical pathways. This spread may be a major factor in the development of dementia in PD.