Parkinson Disease Neuropathological Comorbidities: Prevalences from Younger-Old to Older-Old, With Comparison to Non-Demented, Non-Parkinsonian Subjects

Abstract

Co-existing neuropathological comorbidities have been repeatedly reported to be extremely common in subjects dying with dementia due to Alzheimer disease. As these are likely to be additive to cognitive impairment, and may not be affected by molecularly-specific AD therapeutics, they may cause significant inter-individual response heterogeneity amongst subjects in AD clinical trials. Furthermore, while originally noted for the oldest old, recent reports have now documented high neuropathological comorbidity prevalences in younger old AD subjects, who are more likely to be included in clinical trials. Comorbid neuropathologies in subjects with Parkinson disease have received much less attention. As with AD, comorbidities may interfere with the evaluation of PD clinical trials. We have here examined the decadal-wise presence of multiple co-pathologies, and their clinical effects, in a series of autopsies of PD and control subjects from the Arizona Study of Aging and Neurodegenerative Disorders. Amyloid plaques were present in more than 40% of PD patients in their 60s, and in 85% of those in their 90s. Neurofibrillary tangles were present in PD, as in all elderly humans, from the 60s onwards, while Braak tangle stages of IV or greater, which are strongly associated with severe cognitive impairment, reached 40%, 50% and 60% in those subjects in their 70s, 80s and 90s, respectively. Both plaques and tangles were significant predictors of a lower MMSE score while greater CAA scores had borderline significance. None of these, however, were independently associated with a higher UPDRS score. The ApoE4 allele, while a known predictor of AD pathology, especially amyloid plaques, was not an independent predictor of function with either of these clinical measures, suggesting that its influence is largely mediated by the neuropathologies that it predisposes to. Non-AD tauopathies, including the major conditions of progressive supranuclear palsy (PSP), and corticobasal degeneration (CBD), as well as the microscopic changes of argyrophilic grains (ARG) and aging-related tau astrogliopathy (ARTAG), also co-existed with PD, especially ARG and ARTAG, which ranged between 20% and 40% across decades for the former and up to 80% for the latter; we did not find significant associations of either with final MMSE or UPDRS scores. We failed to find a significant association of limbic TDP-43 histopathology with either final MMSE score or final UPDRS motor score but our subjects were more heavily weighted with PD than those in prior studies and our cognitive correlate, limited to MMSE score, may have missed associations with cognitive domain subsets. We found several cerebrovascular pathologies to be predictors of both cognitive and motor impairment, including brain infarcts, circle of Willis atherosclerosis, and higher white matter rarefaction score. All of the investigated pathology types were common in the non-demented, non-parkinsonian control subjects, and all increased with age in parallel with those co-existing with PD, while with generally lower prevalences. The high concurrence rate of the neurodegenerative protein aggregate diseases is suggestive of either a synergistic co-pathogenesis, where one aggregate type may instigate or accelerate another type, or of one or more underlying predisposing physiological or molecular mechanisms.

Figure 1.

Figure 1 A–D shows graphs depicting percentages of neuropathological comorbidity types in clinicopathologically-diagnosed PD cases, by decade of death age, with Chi-square or Fisher’s Exact Test analysis of differences between decades. Figure 1 E–H show duplicate graphs from A through D with added dashed lines for percentages of the same PD neuropathological comorbidity types that are also present in control subjects. Chi-square or Fisher’s Exact Tests were done to determine significant differences between PD and control cases. For all comparisons: * = p < 0.05; ** = p < 0.01; *** = p < 0.001.