PD-MCI Reversion is Associated with More Favorable CSF Markers Relative to PD-MCI Non-Reverters

Abstract

Objective.

Cognitive impairment is a common non-motor symptom among individuals with Parkinson’s disease (PD). Although the progression of cognitive impairment is generally progressive, individuals with PD-mild cognitive impairment (PD-MCI) may revert to being cognitively normal (CN), which is referred to as PD-MCI reversion. Past studies are inconsistent about whether or not PD-MCI-reverters are at greater risk for re-developing PD-MCI relative to CN individuals. Even less is known about how PD-MCI reverters compare to CN or PD-MCI non-reverters in terms of neurodegenerative biomarkers. The current study examines group differences (CN, PD-MCI-reversion, PD-MCI non-reversion) in CSF markers of amyloid beta (AB), tau (t-tau, p-tau), and alpha synuclein (asyn).

Methods.

The current study used data from the longitudinal international multisite Parkinson Progression Marker Initiative (PPMI). Participants were newly diagnosed with PD and completed a battery of neurocognitive assessments at baseline and subsequent annual follow-ups for up to five years (n=430). Participants were classified as either CN, PD-MCI-reverters, or PD-MCI non-reverters based on the first two neurocognitive assessments.

Results.

The PD-MCI non-reversion group had greater amounts of p-tau/AB and t-tau/AB relative to the PD-MCI reversion group. There were no group differences between the CN and PD-MCI reversion groups in any of the CSF markers.

Conclusions.

PD-MCI reverters may have a more favorable trajectory in terms of Alzheimer’s pathology relative to PD-MCI non-reverters. Future studies are needed to verify if PD-MCI reversion may possibly represent an intermediate prognostic group, between CN and MCI-non-reversion.

Parkinson’s Disease (PD) is a neurodegenerative disease with widespread impact on the body. Hallmark symptoms of PD include tremors, rigidity, bradykinesia and other motor dysfunction; however, nonmotor symptoms, such as cognitive impairment, are becoming increasingly important in the successful management of individuals with PD (1).

Cognitive dysfunction is a major clinical nonmotor symptom of PD (1). Common cognitive difficulties involve problems with visuospatial abilities, executive functioning, language, processing speed, and memory. PD mild cognitive impairment (PD-MCI) refers to a stage of cognitive impairment when cognitive difficulties are detectable on neuropsychological tests, but cognitive difficulties are not severe enough to cause impairment in activities of daily living (ADLs) (2). PD-MCI is considered a transitional stage, between cognitive normal (CN) or having PD dementia (PDD). Indeed, PD-MCI is associated with a greater risk of PDD relative to CN individuals (1).

Although the progression of cognitive impairment is generally considered to be linear (i.e., individuals transition from CN to PD-MCI to PDD), individuals classified with PD-MCI may revert back to CN, a concept referred to as PD-MCI reversion (4,5). A review of reversion rates found 24% of individuals with PD-MCI will revert back to CN within one to six years (6).

This high rate of reversion raises concerns about the utility of PD-MCI as an intermediate diagnosis. For example, if the progression from CN to PD-MCI to PDD is not linear, then clinicians, patients, caregivers and researchers may experience confusion about the prognosis of a PD-MCI diagnosis. Past studies further complicate the significance of PD-MCI reversion. One study found that PD-MCI reverters are at greater risk for future redevelopment of PD-MCI and PDD relative to CN individuals with PD (5). This suggests that a diagnosis of PD-MCI at any time, regardless of reversion, is associated with an increased risk for PDD. Another study found that PD-MCI reverters had a similar risk of cognitive worsening relative to CN individuals (7). Furthermore, this study reported that PD-MCI reverters and CN individuals did not differ in terms of cortical thickness, suggesting that PD-MCI reverters may have preserved structural integrity relative to PD-MCI non-reverters (7). These findings from Jones et al. (1) and Chung et al (7) raise questions about whether PD-MCI reverters should be conceptualized as similar to CN individuals, PD-MCI individuals, or perhaps an intermediate group.

Regarding potential mechanisms of cognitive impairment in PD, the traditional view is that cognitive impairment reflects disruption of frontal-subcortical circuits secondary to alpha synuclein build-up and dopamine depletion (8). However, there is now greater appreciation that cognitive impairment in PD represents heterogeneous mechanisms, including pathologies traditionally associated with Alzheimer’s disease such as tau and amyloid beta (9). Heterogenous cognitive outcomes (i.e. traditional conversion from CN to PD-MCI vs. PD-MCI reversion) may potentially reflect heterogeneous pathologies.

The current study examined group differences (CN, PD-MCI reverters, PD-MCI non-reverters) in biologic markers associated with cognitive decline among patients with PD. Cerebrospinal fluid (CSF) markers of amyloid, tau, and alpha synuclein (asyn) have been shown to be associated with important cognitive outcomes such as longitudinal cognitive decline, PD-MCI and PDD (9). We predict that if PD-MCI reversion is associated with an increased risk of cognitive decline and neurologic compromise then PD-MCI reversion, relative to CN individuals, will be associated with less favorable markers of amyloid, tau and asyn.

METHODS

Study Design

Data was collected from the Parkinson’s Progression Markers Initiative (PPMI), a longitudinal international multisite study. This analysis used data openly available from PPMI, downloaded on 6/12/2023. The sample included 430 newly diagnosed patients with PD, all of whom completed a battery of neurocognitive assessments at baseline and subsequent annual follow-ups for up to 5 years. Participants were excluded from the study if CSF markers and cognition tests were never administered, they were not diagnosed with PD, or they were diagnosed with PDD. Sites secured Institutional Review Board approval and participants provided informed consent. For more details about the sample and longitudinal data see www.ppmi-info.org.

Cognitive Measures

Participants completed a neurocognitive battery at each annual assessment. Letter number sequencing (LNS) from the Wechsler Adult Intelligence Scale, fourth Edition (WAIS-IV) assessed attention and working memory; symbol digit modalities test (SDMT) assessed processing speed; animal fluency assessed language; judgment of line orientation (JLO) assessed visuospatial abilities, and the immediate and delayed recall scores of the Hopkins Verbal Learning Test-Revised (HVLT-R) assessed verbal learning and memory.

Cognitive Status

Individuals were classified as CN or PD-MCI at the baseline and first annual follow-up assessment. PD-MCI criteria was defined by ≥2 scores on neurocognitive measures falling at least 1.5 SD below the mean in accordance with criteria outlined by the Movement Disorder Society (10). Participants were classified as either CN, PD-MCI reverters or PD-MCI non-reverters. Participants were classified as PD-MCI reverters if they met criteria for PD-MCI at baseline and then CN at the first follow-up. Participants were classified as PD-MCI non-reverters if: 1) they met criteria for PD-MCI at both baseline and the first follow-up, or 2) they converted from CN at baseline to PD-MCI at the first annual follow-up. Participants were classified as CN if they met criteria for CN at both baseline and the first annual follow-up.

Motor Symptoms

The Movement Disorder Society Unified Parkinson’s Disease Rating Scale, part III (UPDRS-III) assessed motor symptom severity, on medication. Higher scores indicate greater severity of motor symptoms.

CSF Biomarkers

Total tau (T-tau), phosphorylated tau (p-tau), amyloid beta (AB), and asyn biomarkers were collected from cerebrospinal fluid (CSF) via lumbar puncture. Sites collected between 15–20 mL of CSF at each annual assessment. All sites adhered to the standard operating procedures for ELISA processing of CSF samples outlined in PPMI (www.ppmi.info.org). Biomarker measurements included the following ratios: p-tau/t-tau, t-tau/AB, p-tau/AB, and asyn. Amyloid and tau CSF variables were entered as ratios because past research suggests ratios outperform individual CSF concentrations in terms of identification of cognitive impairment (11). Specifically, larger values of all ratios are typically associated with greater risk of cognitive impairment and dementia (11).

Statistical Analyses

Statistical analyses were conducted using SPSSv.28. Missing data was handled using maximum likelihood estimates. Alpha level of .05 was used for all statistical tests. CSF biomarkers/ratios were entered as the dependent variable for each multilevel linear model (MLM). A separate MLM was conducted for each CSF marker, resulting in a total of four models. Independent variables for each model included cognitive groups (CN stable, PD-MCI non-reverters, and PD-MCI reverters), age, gender, occasion (i.e., baseline, 1st annual follow-up … 5th annual follow-up), and motor severity. Regarding the cognitive grouping variable, the PD-MCI reverters group was entered as the reference category in all analyses (i.e., analyses examined PD-MCI reverters vs. CN, and PD-MCI reverters vs. PD-MCI non-reverters).

Levodopa equivalent daily dose, Geriatric Depression Scale-short form and the Montreal Cognitive Assessment were also used to characterize the sample but not part of the analytic plan.

RESULTS

Sample Characteristics

Baseline demographic and clinical characteristics are in Table 1. A majority of participants (n=324; 75%) were classified as CN, 80 (19%) were classified as PD-MCI non-reverters, and 26 (6%) were classified as PD-MCI reverters. The PD-MCI non-reversion group had a lower representation of whites relative to the other groups. The CN group had less severe motor symptoms and higher scores on the Montreal Cognitive Assessment relative to the two PD-MCI groups. The CN group also had significantly more years of education relative to the PD-MCI non-reversion group.

Table 1.

Baseline Demographic and Clinical Information

	CN (n=324)		MCI-NR (n=80)		MCI-R (n=26)		p	Contrast
	N	%	N	%	N	%
% Male	192	59.3%	49	61.3%	15	57.7%	0.930	--
% White	315	97.2%	66	82.5%	25	96.2%	<0.001	MCI-NR < CN & MCI-R
	M	SD	M	SD	M	SD	p	Contrast
Age	63.4	7.5	64.5	7.4	63.6	6.7	0.504	--
Years of Education	15.7	3.3	14.6	4.0	15.0	4.0	0.039	CN > MCI-NR
MDS-UPDRS III Score	21.1	9.5	23.6	9.9	26.4	9.2	0.014	CN < MCI-NR & MCI-R
LEDD mg	158.4	303.9	206.5	363	222.8	363.8	0.338	--
MOCA Score	27.2	2.2	25.1	2.2	25.9	1.9	<0.001	CN > MCI-NR & MCI-R
GDS Score	2.51	2.74	2.80	2.47	2.65	2.08	0.678	--
pTau/Tau Levels	0.08	0.01	0.08	0.01	0.08	0.01	0.942	--
Tau/AB Levels	0.21	0.11	0.24	0.13	0.21	0.10	0.074	--
pTau/AB Levels	0.02	0.01	0.02	0.01	0.02	0.01	0.112	--
Asyn Levels	1479.8	559.8	1432.1	482	1587.9	459.9	0.515	--

CN = cognitively normal; MCI-NR = mild cognitive impairment non-reverters; MCI-R = mild cognitive impairment reverters; MDS-UPDRS III = Movement Disorder Society Unified Parkinson’s Disease Rating Scale- Part III; LEDD = levodopa equivalent daily dose; MOCA = Montreal Cognitive Assessment; AB = amyloid beta; asyn = alpha synuclein; GDS = Geriatric Depression Scale- short form.

PD-MCI Reversion and CSF Markers

Analyses examined the association between CSF markers and cognitive group status (i.e., CN, PD-MCI non-reverters, PD-MCI reverters; Table 2).

Table 2.

Association Between CSF Markers and Cognitive Status Group.

	ptau/AB			tau/AB
Parameter	Estimate	95% CI	p-value	Estimate	95% CI	p-value
CN vs. MCI-Reverters	0.15	−0.06 to 0.36	0.158	0.10	−0.10 to 0.31	0.328
MCI Non-Reverters vs MCI-Reverters	0.27	0.05 to 0.50	0.019	0.25	0.02 to 0.48	0.032
Male Sex	0.09	−0.007 to 0.19	0.070	0.10	−0.01 to 0.20	0.052
Age	0.32	0.28 to 0.37	<0.001	0.31	0.26 to 0.36	<0.001
MDS-UPDRS III	0.05	0.005 to 0.10	0.031	0.06	0.02 to 0.11	0.011
Occasion	0.02	−0.02 to 0.08	0.261	0.04	−0.01 to 0.08	0.140
Pseudo R2	0.11			0.10

Standardized estimates are shown. CI = confidence interval; AB = amyloid beta; CN = cognitively normal, MCI = mild cognitive impairment; MDS-UPDRS III = Movement Disorder Society Unified Parkinson’s Disease Rating Scale- Part III.

In the model with p-tau/AB as the dependent variable, there was a significant difference between the PD-MCI reversion and PD-MCI-non-reversion groups. Specifically, the PD-MCI non-reversion group had greater amounts of p-tau/AB compared to the PD-MCI reversion group regardless of occasion. There was no significant difference between the CN and PD-MCI reversion groups. Greater p-tau/AB load was associated with older age and more severe motor symptoms.

In the model with tau/AB as the dependent variable, the PD-MCI non-reversion had significantly greater amounts relative to the PD-MCI reversion group across all study years. There was no significant difference between the CN and PD-MCI reversion groups. Greater tau/AB load was also associated with older age and greater severity of motor symptoms.

In the models with p-tau/tau and asyn as the dependent variables, there were no significant differences between any of the cognitive groups (Supplemental Table 1) regardless of occasion. Greater patu/tau load was associated with older age, while greater amounts of asyn were associated with both male sex and older age.

DISCUSSION

This study demonstrated that PD-MCI reverters had more favorable markers of tau and amyloid relative to PD-MCI non-reverters. There were no significant differences between the CN and PD-MCI reverter groups in any CSF markers. Findings suggest that PD-MCI reverters may be more similar to CN individuals than PD-MCI non-reverters in terms of neurologic compromise.

In terms of risk for future cognitive decline, at least two studies in PD (5,12) and multiple studies in healthy aging populations (13,14) have shown that MCI-reverters are at increased risk for future cognitive decline. Among individuals with PD, we previously demonstrated that PD-MCI reverters had 3.2 to 7.7 increased odds, relative to CN individuals, of either re-converting to PD-MCI or developing PDD in the subsequent 3 years (5). However, this risk was qualitatively lower than PD-MCI non-reverters, who had 5.2 to 23.3 increased odds of meeting criteria for PD-MCI or PDD in the subsequent 3 years. This may suggest that PD-MCI reverters should be conceptualized as an intermediate group whose risk of cognitive decline is greater than CN individuals but lower than PD-MCI non-reverters.

In PD, the clinical implications of comorbid Alzheimer’s pathology among PD patients has primarily focused on increased risk of cognitive impairment and faster time to PDD (15). Additionally, there are at least four studies documenting an association between amyloid pathology and the postural instability gait difficulty subtype (16–19). However, the clinical relevance of comorbid Alzheimer’s pathology to other meaningful areas, such as treatment of motor/non-motor symptoms, cognitive profiles (i.e. executive dysfunction, amnestic subtypes, etc.), risk of other non-motor complications and outcomes following deep brain stimulation surgery are less elucidated in literature (15,20).

Among non-PD aging populations, altered neuroimaging and CSF ratios are consistently associated with MCI, but these same markers are inconsistently associated with MCI reverters. Moradi and colleagues (2021) compared 68 AD-MCI reverters to 337 AD-MCI stable patients on brain atrophy and CSF biomarkers (t-tau, p-tau, and AB) (14). Brain atrophy and altered CSF levels were altered in the AD-MCI stable group, but the MCI reversion group had more favorable levels of t-tau, p-tau and AB relative to the AD-MCI stable group. Similarly, Thomas et al. found MCI reverters were similar to CN but significantly different than the MCI non-reversion group in levels of t-tau, p-tau and AB (21,22). If these biomarkers are signs of future cognitive decline and neurologic compromise then findings suggest that MCI reverts may have a more favorable prognosis relative to MCI non-reverters.

Although we are unaware of another study investigating differences in asyn, tau and amyloid between PD-MCI-reverters and PD-MCI non-reverters, our current findings are consistent with a past structural and functional imaging study of PD-MCI-reversion (7). Specifically, this study found that while PD-MCI-reverters and non-reverters differed in terms of cortical thickness and functional connectivity in default mode and executive control networks; they generally did not find differences between the CN and PD-MCI-reversion groups. Indeed, the authors concluded that structural and functional integrity may be relatively preserved among individuals with PD-MCI-reversion.

We did not find significant group differences in CSF markers of asyn. Similarly, other past PPMI studies of PD-MCI have failed to find group differences between CN and PD-MCI, or found that CSF-asyn is not associated with risk of incident PD-MCI (23,24). In contrast, markers of amyloid burden are more consistently associated with PD-MCI status in the PPMI cohort, suggesting that amyloid burden, rather than asyn, may be a significant driver of cognitive impairment early in PD (15,16).

Limitations in the current study included that participants consisted of individuals newly diagnosed with PD. Findings may not generalize to later stages of PD or samples where cognitive impairment is more common. Similarly, PD-MCI reversion was defined by the baseline and first annual follow-up assessment. While this approach is consistent with past papers of MCI-reversion both in AD (21,22) and PD (5,7) future studies should investigate PD-MCI reversion in later years. Ultimately, while the fact that participants were followed for up to five years is impressive and should be considered a strength of the study, we cannot rule out that group differences would emerge later in the disease course. The current study utilized CSF markers of asyn, tau and amyloid. There are several additional potential mechanisms of cognitive impairment and markers of neurodegeneration that could potentially differ between PD-MCI reverters and PD-MCI non-reverters. This includes but is not limited to differences in genetics, epigenetics, other comorbidities (e.g. cerebrovascular disease), structural brain imaging, neuroinflammation, cholinergic degeneration, and dopamine availability.

Overall, PD-MCI reverters may have a more favorable outcome in terms of Alzheimer’s pathology relative to PD-MCI non-reverters. This seems to be counterintuitive to past PD and healthy aging studies that find a diagnosis of PD-MCI, regardless of reversion status, is associated with future cognitive decline. Future studies are needed to verify if PD-MCI reversion may possibly represent an intermediate prognostic group between CN and MCI-non-reversion.