Biomarker and pathology studies in neurodegenerative cognitive impairment

Abstract

Background: Dementia is a major cause of functional impairment and early death in older age groups. Neurodegenerative disorders are the most common cause of dementia. The most frequent neuropathological lesions include neurofibrillary tangles and senile plaques, hallmark lesions for Alzheimer´s disease (AD), and Lewy body pathology, which characterize Lewy body disease (LBD). Clinically, the neuropathological entity LBD can present as either Parkinson´s disease (PD) or dementia with Lewy bodies (DLB), differentiated on the basis of the presenting symptoms being either motor or cognitive. While the majority of LBD patients develop both motor symptoms and cognitive impairment, some patients with clinical PD will never experience cognitive impairment and likewise some patients with DLB will never develop motor symptoms. Similarly the clinical presentation of AD is also heterogeneous, for instance, the highly variable occurrence of neuropsychiatric symptoms and rate of progression. These differences have a major impact on quality of life for patients and carers, as well as health care costs, but their mechanisms and neuropathological underpinnings are poorly understood. Furthermore the correlation between clinical diagnosis and neuropathological findings is relatively low, and LBD patients presenting with cognitive impairment particularly risk being misclassified as AD. This highlight the need for more precise biomarkers for these clinical syndromes that can be implemented at the start of and during the course of the disease. Biomarkers may inform about disease pathology, thus paving the way for new treatment, they increase diagnostic accuracy and aid in setting a prognosis. Biomarkers are needed in the selection of patients for treatment studies and to identify which patients should benefit from new treatment when available. The cerebrospinal fluid (CSF) biomarkers beta-amyloid 42 (abeta42), total tau (t-tau) and tau protein phosphorylated at amino acid 181 (p-tau181) reflect key AD pathologies. The Lewy bodies found in LBD are composed mainly of the protein !-synuclein. !-synuclein is reduced in CSF in LBD, but with considerable overlap between LBD, controls and other disease groups. Aim: The main aim of this thesis was to increase understanding of pathological mechanisms underlying important clinical features in neurodegenerative cognitive impairment, by exploring the associations between clinical presentation and biomarkers and pathology. The first objective was to explore the association between AD pathology CSF markers and neuropsychiatric symptoms in newly diagnosed AD patients; secondly to assess the association between CSF markers of AD and LBD pathology and early cognitive impairment in PD; thirdly to examine the correlation between clinical diagnosis of DLB and Lewy body pathology at autopsy. Methods: This is a clinical translational neuroscience project based on two clinical cohort studies. The dementia Study of Western Norway (Demvest) included newly diagnosed dementia patients from specialist clinics in geriatric medicine and old age psychiatry in Western Norway. The Parkinson´s Progression Markers Initiative (PPMI) is an international multicentre study, including newly diagnosed PD patients and healthy controls. A comprehensive battery of neuropsychological tests, a structured neuropsychiatric evaluation, clinical examination, and imaging were part of both studies. CSF sampling was done according to standardized protocols and CSF was analysed using commercially available immunoassays. In the Demvest study, participants were recruited for brain donation, and autopsy results were obtained applying commonly used neuropathological protocols and diagnostic criteria. Results: We undertook three specific studies to investigate objective I, II and III. In study I, apathy in patients with early Alzheimer´s disease correlated with t-tau and ptau181 concentrations in CSF, higher values being associated with more severe apathy. There were no associations between depression or psychosis and agitation and CSF markers. In study II, decreased CSF !-synuclein in newly diagnosed PD-patients without dementia correlated with impaired global cognition and impairment of executive functions and attention. CSF abeta42 was decreased in PD with mild cognitive impairment compared with controls after adjusting for covariates. No correlations were found between memory or visuospatial functions and CSF markers. Study III examined autopsy results of 56 patients followed from dementia diagnosis to death. 20 patients received a pathological diagnosis of LBD; the corresponding clinical diagnosis were probable DLB (n=11), Parkinson´s disease with dementia (PDD) (n=5) and probable or possible AD (n=4). Of the 56, 14 patients received a clinical diagnosis of probable DLB, 11 of these had pathological LBD and three AD. Sensitivity, specificity, positive and negative predictive values of a clinical DLB diagnosis were 73%, 93%, 70%, and 90% respectively. Conclusions and implications: We have reported a novel association between neuropsychiatric symptoms and CSF biomarkers reflecting core AD pathology. The relationship between t-tau and p-tau181 and apathy may reflect an association between neurofibrillary tangle pathology and apathy in early AD. Cognitive impairment in early PD was associated with biomarkers of both Lewy body and AD pathology. 18 of 20 LBD patients in the Demvest study had Braak neurofibrillary tangle stage IV or higher, representing severe AD pathology at autopsy. Thus our findings suggest a role for AD pathology in both early and established LBD. Accurate diagnosis is crucial for clinical practice and research. With a sensitivity of 73%, the clinical 2005 DLB criteria are not sensitive enough. More than one in four DLB patients were not identified even when structured rating scales for core DLB symptoms were applied. We regard a specificity of 93% as satisfactory. Our results illustrate that not all DLB patients fulfil the 2005 DLB criteria at disease presentation, highlighting the need for re-evaluation of the diagnosis if new symptoms appear. Studies applying the most recent 2017 DLB criteria will show if this revision has increased sensitivity without decreasing specificity.