We read with interest the paper by Martin and colleagues1 reporting neuropathological findings among 165 cases of Parkinson’s disease without dementia (PDND, N=37), Parkinson’s disease with dementia (PDD, N=114) and Dementia with Lewy Bodies (DLB, N=14). They conduct logistic regression analyses to determine the association of neuropathology scores with a clinical diagnosis of dementia in univariate and multivariate models. They replicate previous findings from our group and others 2,3 that neocortical Lewy Body (LB) pathology is the strongest predictor of dementia in LB disease. They additionally report no association between Alzheimer’s Disease Neuropathological Change (ADNC), β-amyloid (A), neurofibrillary tangle (B), or neuritic plaque (C) scores and dementia. We are surprised to see the conclusions drawn from this work regarding the role of AD co-pathology in PD dementia.
The frequency of comorbid AD (intermediate or high ADNC) is reported to be ~10% in PD, ~35% in PDD and ~70% in DLB.3 We re-examined our pathology cohort of 97 PDND and 175 PDD and show ADNC (Table Top), and A, B, and C scores (Table Bottom) by Lewy body pathology score. Mirroring what is reported by Martin and colleagues, the Penn-based PDD cases have a higher relative frequency of intermediate and high ADNC, which is most evident in cases with diffuse/neocortical Lewy bodies. Furthermore, high A, B, and C is more frequent in PDD compared to PDND in both cohorts. These pathologies co-occur frequently and track together in severity and distribution. Multivariate prediction models examining these pathologies together suffer from high collinearity, which can lead to erratic results when examining individual variables, often attributing effects solely to the strongest variable. Likewise, the issue of collinearity among co-pathologies is affected by the relatively small sample size of the Martin et al study (n=165), further limiting the ability to detect the effects of co-occurring AD pathology. As a consequence, their “negative” finding for AD pathology as a predictor of cognition in PD may be related to sample size or analytical design.
Table.
Top: Alzheimer Disease Neuropathologic Change by Lewy Body Stage (McKeith Criteria). Bottom: Alzheimer’s Disease Neuropathologic Scores by Alzheimer Disease Neuropathologic Change
| Alzheimer’s Disease Neuropathological Change |
|||||
|---|---|---|---|---|---|
| LB | Not | Low | Int | High | |
|
|
|||||
| PDND (n=97) | 1 | 14 (14.4) | 10 (10.3) | 2 (2.1) | 0 (0.0) |
| 2 | 9 (9.3) | 36 (37.1) | 2 (2.1) | 0 (0.0) | |
| 3 | 4 (4.1) | 12 (12.4) | 6 (6.2) | 2 (2.1) | |
|
| |||||
| PDD (n=175) | 1 | 5 (2.9) | 8 (4.6) | 2 (1.1) | 1 (0.6) |
| 2 | 15 (8.6) | 26 (14.9) | 12 (6.9) | 1 (0.6) | |
| 3 | 11 (6.3) | 45 (25.7) | 36 (20.6) | 13 (7.4) | |
| Alzheimer’s pathology scores |
||||
|---|---|---|---|---|
| ADNC | A | B | C | |
|
|
||||
| PDND (n=97) | 0 | 27 (27.8) | 12 (12.4) | 63 (64.9) |
| 1 | 37 (38.1) | 65 (67.0) | 13 (13.4) | |
| 2 | 24 (24.7) | 16 (16.5) | 13 (13.4) | |
| 3 | 9 (9.3) | 4 (4.1) | 8 (8.2) | |
|
| ||||
| PDD (n=175) | 0 | 31 (17.7) | 8 (4.6) | 75 (42.9) |
| 1 | 42 (24.0) | 87 (49.7) | 25 (14.3) | |
| 2 | 49 (28.0) | 60 (34.3) | 33 (18.9) | |
| 3 | 53 (30.3) | 20 (11.4) | 42 (24.0) | |
Numbers represent N (%).
PDND = Parkinson’s disease non-demented; PDD = Parkinson’s disease dementia; LB = Lewy Body Stage (McKeith Criteria). Int= Intermediate. A = β-Amyloid; B = Neurofibrillary Tangles; C= Neuritic Plaques.
With larger studies and better pathological and clinical characterization, the nuanced effects of co-occurring AD pathologies on cognition in PD may become more evident. Co-occurring AD, and in particular cortical tau deposition, likely associate with PDD in a subset of individuals, albeit with smaller effect sizes than neocortical LB.4 Understanding how AD affects cognitive performance in PD remains essential as we move into an era of anti-amyloid therapies in clinical and preclinical AD.
Footnotes
Author Conflicts of Interest:
None to report.
References
- 1.Martin WRW, Younce JR, Campbell MC, et al. Neocortical Lewy Body Pathology Parallels Parkinson’s Dementia, but Not Always. Ann Neurol 2022; Online ahead of print. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Irwin DJ, White MT, Toledo JB, et al. Neuropathologic substrates of Parkinson disease dementia. Ann Neurol 2012;72(4):587–598. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Smith C, Malek N, Grosset K, et al. Neuropathology of dementia in patients with Parkinson’s disease: a systematic review of autopsy studies. J Neurol Neurosurg Psychiatry 2019;90(11):1234–1243. [DOI] [PubMed] [Google Scholar]
- 4.Horvath J, Herrmann FR, Burkhard PR, et al. Neuropathology of dementia in a large cohort of patients with Parkinson’s disease. Parkinsonism Relat Disord 2013;19(10):864–868. [DOI] [PubMed] [Google Scholar]