Abstract
Body-first Lewy body disease (LBD) is hypothesized to begin in the peripheral autonomic nervous system years before nigrostriatal involvement. Isolated REM sleep behaviour disorder (iRBD) is considered an early manifestation of body-first LBD but is preceded by an unknown period of peripheral autonomic degeneration. Consequently, the total duration of the prodromal phase remains unknown. We used longitudinal imaging data to model the progression of cardiac sympathetic and nigrostriatal dopaminergic degeneration and employed the resulting curves to estimate the prodromal period of body-first LBD. Cardiac sympathetic trajectories were highly consistent across two cohorts, showing rapid early decline followed by gradual slowing, reaching 50% of the normal value after ~5 years and 25% after ~10 years. Dopaminergic decline was slightly slower, reaching 50% of the normal value after ~8 years and 25% after ~15 years. iRBD patients displayed severe baseline cardiac sympathetic loss (median 21.3% of normal), but mild baseline dopaminergic deficit (median 81.5% of normal), indicating that cardiac sympathetic degeneration may precede dopaminergic nigrostriatal involvement by 8.6 years. Based on baseline dopaminergic deficits in Parkinson’s disease (PD), the combined model estimated that the prodromal phase of body-first LBD, from onset of cardiac sympathetic degeneration to predicted clinical PD diagnosis, exceeds 20 years.
Subject terms: Biomarkers, Diseases, Medical research, Neurology, Neuroscience
Introduction
Lewy body disease (LBD) comprises the clinical phenotypes Parkinson’s disease (PD) and Dementia with Lewy bodies (DLB), characterized by intraneuronal accumulation of α-synuclein. These inclusions are also found in postmortem examinations of patients with isolated REM-sleep behaviour disorder (iRBD)1, and nearly all iRBD patients develop PD or DLB within 15 years2. Therefore, iRBD is also considered an LBD and a prodromal manifestation of the clinical phenotypes3.
LBD patients exhibit neuronal degeneration of multiple neurotransmitter systems in the brain and the peripheral autonomic nervous system. Cardiac sympathetic neurons are among the most vulnerable, but a striking heterogeneity in the timing of this degeneration is seen. At diagnosis, some LBD patients already show severe cardiac sympathetic denervation, whereas others are completely normal - despite overt brain involvement. Among other observations, this has prompted the hypothesis that LBD comprises two main subtypes4,5: a brain-first subtype initially affecting the brain and later the peripheral autonomic nervous system, and a body-first subtype initially affecting the peripheral autonomic nervous system years before the brain. In this framework, cardiac sympathetic denervation is a marker of body-first LBD, which is supported by a recent autopsy study demonstrating that brainstem-predominant Lewy body profiles (i.e., body-first) were associated with severe cardiac sympathetic denervation, whereas amygdala-predominant profiles (i.e., brain-first) showed much more preserved innervation6. This finding suggests that iRBD patients are mainly prodromal body-first LBD patients with brainstem-predominant Lewy pathology, since 90-95% of iRBD cases display severe cardiac sympathetic denervation on imaging7–9.
[123I]meta-iodobenzylguanidine (MIBG) scintigraphy is a robust tool to evaluate cardiac sympathetic innervation and has been validated against postmortem myocardial tyrosine hydroxylase immunostaining10,11. Late static images reflect the uptake and storage of MIBG, i.e., a sympathetic functional entity12. MIBG scintigraphy can distinguish PD from atypical Parkinsonism and serve as a supportive/indicative biomarker in the diagnostic criteria for PD and DLB, respectively13,14.
Nevertheless, while an extensive literature on dopaminergic degeneration exists, the progressive peripheral autonomic degeneration has received limited attention. Therefore, a temporal understanding of the onset and subsequent progressive loss of autonomic neurons is needed. Here, we examine the progressive cardiac denervation using MIBG scintigraphy in large cohorts of iRBD and PD patients. We aimed to model the time course of cardiac sympathetic denervation and estimate how it relates to the onset of nigrostriatal neurodegeneration. This strategy enables an estimate of the total prodromal period in body-first LBD, i.e., from the earliest detectable neurodegeneration until clinical diagnosis of PD or DLB.
Results
Patient cohort characteristics
We collected longitudinal MIBG data from a total of 260 LBD patients from three cohorts: the KPD-PD cohort (n = 195), PACE-PD cohort (n = 44) and PACE-iRBD cohort (n = 21). For comparison, we assembled longitudinal dopaminergic imaging data from the PPMI-PD (n = 426) and PPMI-iRBD (n = 37) cohorts.
Normative MIBG data for the KPD cohort were acquired from 25 healthy control subjects, mean age 67.4 ± 4.8 years, 76% females. For the PACE cohort, normative MIBG data were acquired from 42 healthy control subjects, mean age 66.0 ± 10.6 years, 50% females. Normative [18F]-FE-PE2I PET data in the PACE cohort were acquired from 55 healthy control subjects, mean age 66.2 ± 10.3 years, 52% females. Part of the longitudinal KPD-PD data, baseline PACE-PD and PACE-iRBD data, and parts of the normative imaging data have been published previously5,7,15. Normative dopaminergic imaging data from the PPMI cohort included 184 healthy controls, mean age 60.4 ± 11.4 years, 37% females.
Time course of cardiac sympathetic denervation (MIBG)
The median cardiac sympathetic deficit at baseline measured with MIBG was very similar in KPD-PD (33.7% of normal value) and PACE-PD (34.4%) patients and lower in PACE-iRBD patients (21.3%) (Table 1). Cardiac sympathetic denervation progressed very similarly in the KPD and PACE cohorts (Fig. 1). Therefore, the curves were averaged. The average curve exhibited the fastest decline in the upper part, reaching approximately 50% of the normal value after 5 years, followed by a gradually slower decline rate reaching 25% after 10 years, and 0% after 20–25 years. This translates into a median annual decline rate of 10% points in PD patients with baseline MIBG within the normal range (Table 2). With a median MIBG of 21.3% of the normal value in PACE-iRBD patients, the model thereby estimates the resulting period of cardiac sympathetic degeneration to antedate baseline enrollment by at least 11.3 years (Fig. 2).
Table 1.
Demographics and clinical characteristics of the patient cohorts
| PACE | KPD | PPMI | |||
|---|---|---|---|---|---|
| PD | iRBD | PD | PD | iRBD | |
| n | 74 | 54 | 195 | 426 | 37 |
| Sex (F/M) | 20/54 | 8/46 | 97/98 | 154/272 | 6/31 |
| Age at baseline (years) | 67.0 (8.2) | 66.8 (7.6) | 67.7 (8.9) | 61.7 (9.7) | 69.1 (5.3) |
| Time since diagnosis at baseline (years) | 0.59 (0.82) | 1.00 (1.13) | 0 | 0.80 (1.20) | N/A |
| Symptom duration at baseline (years) | 2.13 (1.59) | 6.43 (4.77) | 1.16 (1.08) | 2.18 (2.18) | N/A |
| UPDRS-III at baseline | 19.01 (10.20) | 5.64 (6.87) | 19.06 (9.32) | 21.79 (9.43) | 4.38 (3.82) |
| Hoehn & Yahr stage at baseline (0/1/2/3/4) | 0/28/40/5/0 | N/A | 0/61/120/14/0 | 0/179/245/2/0 | N/A |
| Baseline imaging (n) | MIBG = 73 PE2I = 33 | MIBG = 54 PE2I = 32 | MIBG = 195 | DaT SPECT = 426 | DaT SPECT = 37 |
| Follow-up imaging (n) | MIBG = 44 PE2I = 21 | MIBG = 21 PE2I = 7 | MIBG = 195 | DaT SPECT = 426 | DaT SPECT = 37 |
| Follow-up time (baseline to FU, years) | 3.12 (0.37) | 3.32 (1.18) | 2.27 (0.53) | 3.45 (1.05) | 3.63 (1.05) |
| MIBG at baseline (H/M late, % of normal) | 34.4% [17.9, 67.7] | 21.3% [14.0, 38.9] | 33.7% [19.2, 55.2] | N/A | N/A |
| MIBG at follow-up (H/M late, % of normal) | 26.0% [18.2, 49.2] | 16.7% [7.59, 32.1] | 24.6% [15.2, 50.5] | N/A | N/A |
| Dopamine imaging at baseline (lowest putamen, % of normal) | 32.2% [29.2, 44.8] | 81.5% [70.6, 94.0] | N/A | 31.4% [24.1, 39.3] | 54.8% [45.9, 67.3] |
| Dopamine imaging at follow-up (lowest putamen, % of normal) | 23.8% [16.8, 30.9] | 60.2% [48.2, 86.5] | N/A | 26.1% [19.3, 32.3] | 50.9% [36.8, 72.8] |
Data presented as mean(SD) or median [IQR]. Symptom duration is defined as the time elapsed since self-reported onset of motor symptoms in PD or onset of RBD symptoms in iRBD.
Fig. 1. Cardiac sympathetic and dopaminergic progression curves.
Left: Longitudinal decline of cardiac sympathetic integrity (assessed by MIBG) in two independent cohorts (KPD = red, PACE = green/blue, MEAN = black). Changing color gradient along the curves indicates transitions between successive quartiles. Right: Longitudinal decline of putamen dopamine integrity, assessed by PPMI DaT imaging data (combined PD and prodromal). Notice the similarity between the two curves, although the MIBG curve seems to decline slightly faster. The MIBG curve reach 50% after 5 years and 25% after 10 years, whereas the DaT curve reach 50% of the normal value after 8 years and 25% after 15 years.
Table 2.
Baseline MIBG and dopamine imaging values and annualized decline rates
| KPD-PD | n | Median MIBG at baseline (%) | MIBG decline rate (%/year) |
|---|---|---|---|
| Q1, >75% | 20 | 92.2 [83.3, 101.9] | −13.7 [−19.0, −6.4] |
| Q2, 50–75% | 18 | 63.9 [55.0, 70.7] | −6.1 [−9.0, −4.01] |
| Q3, 25–50% | 55 | 34.4 [28.7, 40.0] | −5.8 [−8.5, −3.5] |
| Q4, <25% | 26 | 20.8 [16.5, 22.9] | −2.0 [−3.5, −1.4] |
| PACE-PD + iRBD | n | Median MIBG at baseline (%) | MIBG decline rate (%/year) |
|---|---|---|---|
| Q1, >75% | 10 | 93.3 [86.6, 99.1] | −9.8 [−14.7, −6.6] |
| Q2, 50–75% | 6 | 56.2 [54.0, 58.8] | −6.0 [−9.6, −4.7] |
| Q3, 25–50% | 14 | 34.1 [29.6, 41.4] | −3.0 [−4.9, −2.3] |
| Q4, <25% | 18 | 16.7 [13.9, 19.4] | −1.8 [−2.4, −0.9] |
| PPMI-PD + iRBD | n | Median DaT at baseline (%) | DaT decline rate (%/year) |
|---|---|---|---|
| Q1, >75% | 9 | 89.1 [79.9, 99.8] | −6.8 [−10.5, −3.1] |
| Q2, 50–75% | 38 | 55.5 [52.3, 63.6] | − 4.1 [−5.4, −2.7] |
| Q3, 25–50% | 226 | 34.2 [29.7, 41.0] | −2.4 [−3.5, −1.3] |
| Q4, <25% | 64 | 20.4 [15.8, 22.4] | −1.5 [−2.4, −0.5] |
Cardiac sympathetic (MIBG) and dopaminergic (DaT) imaging values and annualized decline rates by quartile and cohort. Data are presented as median [IQR].
Fig. 2. Baseline cardiac sympathetic imaging data from the PACE and KPD cohorts.
Left: Baseline cardiac sympathetic integrity (MIBG late H/M) data presented as % of normal value. The two PD cohorts show similar cardiac sympathetic degeneration. The PACE-iRBD cohort shows severe cardiac degeneration, median = 21.3%, reflecting a preceding degenerative phase of 11.3 years. Five PACE-HCs with values high in the normal range are not presented in the figure. Right: Three representative MIBG images are shown from each degenerative stage, from normal (100%) to complete denervation (0%). A guideline for qualitative interpretation of MIBG images is provided. White arrows signify the location of the heart.
Time course of nigrostriatal dopaminergic denervation
Median dopaminergic deficits were similar in the PPMI-PD (31.4%) and PACE-PD (32.2%) cohorts, whereas the iRBD patients showed more pronounced dopamine deficits in the PPMI-iRBD cohort (54.8%) than in the PACE-iRBD cohort (81.5%). In a limited subset of PACE-PD and PACE-iRBD participants (not included in the construction of the curve), model-predicted 3-year follow-up [18F]-FE-PE2I values were in close agreement with observed values (Supplementary Fig. 1).
The time course of dopaminergic degeneration, measured from PPMI-PD and PPMI-iRBD data, was slightly slower than seen for MIBG. The curve showed an initial rapid decline, and then slower progression in the lower ranges, approximating 50% of normal value after 8 years, 25% after 15 years, and 0% after 25+ years (Fig. 1). With a median dopaminergic imaging level of 81.5% of normal value in PACE-iRBD patients, the model thereby estimates the average onset of dopaminergic degeneration to antedate baseline imaging by 2.7 years (Fig. 3). The time gap between onset of cardiac sympathetic degeneration and nigrostriatal dopaminergic degeneration was therefore estimated to 8.6 years (Fig. 4). Mean duration of RBD-symptoms in the PACE-iRBD cohort was 6.4 years, suggesting that RBD started after cardiac sympathetic degeneration but before nigrostriatal degeneration. The more pronounced baseline dopaminergic deficit in PPMI-iRBD subjects resulted in an estimated onset of dopaminergic degeneration 6.8 years prior to baseline imaging. As the PPMI does not include MIBG, the time gap between the onset of dopaminergic and cardiac sympathetic degeneration could not be determined for this cohort.
Fig. 3. Baseline dopaminergic imaging data from the PPMI and PACE cohorts.
Left: Baseline dopaminergic imaging data of the most affected putamen presented as % of the normal value. The PACE-iRBD cohort shows mild dopaminergic degeneration, median = 81.5%, reflecting a preceding degenerative phase of 2.7 years. In comparison, PACE-PD dopaminergic degeneration was 32.2% of the normal value. This indicates that the PACE-iRBD cohort will reach the level of PACE-PD in 9.8 years, potentially signifying median time to phenoconversion. One PACE-HC and 23 PPMI-HCs with values high in the normal range are not presented in the figure. Right: A representative [18F]-FE-PE2I PET image is shown from each degenerative stage, from normal (100%) to almost complete denervation (0%). Dopamine transporter imaging was obtained with DaT SPECT in the PPMI data and with [18F]-FE-PE2I PET in the PACE data.
Fig. 4. Combined cardiac sympathetic and dopaminergic progression curves in iRBD.
Left: Simplified model of proposed α-synuclein propagation through the sympathetic pathways in body-first LBD. Note that the parasympathetic, vagal spreading route has received much more attention in body-first LBD but has been purposefully omitted in this illustration to highlight the importance of sympathetic pathways. Right: The cardiac sympathetic (MIBG) and putamen dopamine progression curves with a parallel offset of dopaminergic degeneration based on the baseline PACE-iRBD median MIBG (21.3%) and dopamine imaging (81.5%) results. Notice that the onset of dopaminergic degeneration is estimated to begin 8.6 years later than cardiac sympathetic degeneration, leaving ample time for a body-first α-synuclein pathology to propagate to brainstem structures and initiate the neurodegenerative process.
Combined timeline of neurodegenerative events
Based on our findings, we can construct an estimated timeline of neurodegenerative events in iRBD (with a mean age of 67) (Fig. 5). At an unknown time point before age 56, α-synuclein likely starts to accumulate and spread within the autonomic nervous system. By median age 56, the cardiac sympathetic system begins to degenerate. Four years later, at age 60, RBD symptoms begin. By a median age of 64 years, the dopaminergic degeneration starts. The dopaminergic degeneration will reach the level of de novo PD at age 77, predicting the conversion to PD at this age. As the mean age at enrollment in the PACE-iRBD cohort is 67, the median time to phenoconversion can be estimated from dopaminergic imaging to 9.8 years. To summarize, the total prodromal period in body-first LBD, i.e., from the onset of cardiac sympathetic degeneration to a clinical diagnosis of PD, is therefore estimated to exceed 20 years. Importantly, these estimates are model-based median numbers and not direct observations.
Fig. 5. Estimated timeline of neurodegenerative events in body-first LBD.
Based on progression curves from the present study and baseline median data from the PACE-iRBD cohort, the median age of key events in body-first LBD can be estimated as depicted in the figure.
Discussion
In this study, we determined MIBG progression curves from two independent patient cohorts and found nearly identical trajectories (Fig. 1). The findings suggest that cardiac denervation shows an initial rapid decline (10%-points/year) followed by a slower decline (5%-points/year) towards total denervation. By extrapolating these progression curves, we estimated that cardiac denervation in the PACE-iRBD cohort had started 11.3 years prior to baseline enrollment. The average estimated time to phenoconversion for the PACE-iRBD cohort was predicted to be 9.8 years based on the dopaminergic progression curve. Consequently, our model predicts that the total prodromal period of body-first LBD, where ongoing neurodegeneration can be objectively measured, must in many cases exceed 20 years.
This study is the first to estimate the prodromal period in LBD by modeling the progressive cardiac sympathetic denervation. Yet, the MIBG progression pattern defined here is supported by previous reports. Two studies of more than 200 patients combined found significantly faster annual decline in MIBG uptake in patients with relatively normal baseline uptake compared to those with pathological baseline MIBG15,16. Similar results were reported in a study using the cardiac sympathetic PET tracer [18F]dopamine, where PD patients without orthostatic hypotension had relatively preserved cardiac sympathetic innervation at baseline but showed rapid progression at 3.8-year follow-up17. In contrast, PD with orthostatic hypotension showed cardiac denervation already at baseline with limited progression at follow-up.
To understand how the onset and progression rate of peripheral sympathetic degeneration in LBD relates to CNS neurodegeneration, we included longitudinal dopaminergic imaging data and applied the same approach. The dopaminergic curve showed similarity to the MIBG curve, but with slightly slower progression (Fig. 1), suggesting that cardiac sympathetic neurons exhibit at least similar vulnerability to LBD pathology as nigrostriatal dopaminergic neurons. Dopaminergic degeneration has previously been studied using linear and exponential models based on imaging data, and there are reasons to believe that more flexible, multimodal models are needed to reflect the heterogeneous nature of PD progression18,19. The dopamine trajectory in the present study reproduces previous studies showing a fast initial decline reaching 50% loss in 10–15 years20,21, supporting the robustness of our finding.
Based on the dopaminergic decline rate and baseline dopaminergic difference between PACE-iRBD and PACE-PD patients, the time to phenoconversion was predicted to be 9.8 years. It is well known that a dopaminergic deficit in iRBD signals incipient phenoconversion2,22,23. Here, we suggest that the level of dopaminergic degeneration in iRBD patients may predict when patients convert. For example, an iRBD patient with a normal dopamine system (~100% of normal value) is probably unlikely to convert to PD within the first 10 years. In contrast, a patient with 75% of normal value will most likely convert in 5–10 years, and a patient with 50% of normal value within 5 years (Fig. 3). Importantly, these estimates are model-based predictions and needs validation from longitudinal iRBD cohorts with baseline dopaminergic imaging.
Our extrapolated MIBG curve suggests that peripheral α-synuclein pathology initiates more than 11.3 years before study enrollment of many iRBD patients, and hence when these subjects are in their 50’s. This age group may be an ideal screening population for early body-first LBD detection. Given recent advancements in pathological α-synuclein detection in CSF, skin, and blood using seeding amplification assays (SAA) in iRBD patients24–27, these methods may be able to diagnose LBD before overt peripheral neurodegeneration. However, the amount of pathology is probably sparse in the early phase, and it is unknown if current SAA methods are sufficiently sensitive at this stage of disease, although one autopsy study reported positive skin SAA in 6 of 7 incidental LBD cases, i.e., in the preclinical phase28. In addition, the Parkinson Associated Risk Syndrome Study found 48% positive cases on CSF SAA in older subjects with a family history of PD and hyposmia29.
The recently proposed NSD-ISS system for a biological definition of neuronal α-synuclein disease includes preclinical stages but does not include any markers of peripheral neurodegeneration30. Specifically, stage 1A is defined as positive CSF SAA, and stage 1B additionally includes dopaminergic deficit (but no clinical signs or symptoms)30. Our study suggests that a major lag time between stage 1A and stage 1B exists, estimated here to be 8.6 years for the hypothesized body-first LBD subtype (Figs. 4 and 5). This aligns with our recent postmortem study showing that a “sympathetic body-first LBD subtype” accumulates severe levels of α-synuclein pathology in the sympathetic nervous system, before the brainstem is even affected31. In this context, cardiac sympathetic imaging methods, such as MIBG and [18F]dopamine PET17,32, are undoubtedly powerful tools to detect neurodegeneration much earlier than dopamine scans. Indeed, given the present evidence, we propose adding an additional stage to this framework, i.e., measurable sympathetic neurodegeneration in the absence of dopaminergic degeneration.
These considerations may have important implications for patient selection in future clinical trials of very early SAA-positive cases. Such trials should ideally include MIBG (or other markers of peripheral degeneration) to determine which cases eventually develop an LBD, and when this will happen. Also, this approach will limit inclusion of multiple system atrophy cases, who typically show normal cardiac MIBG uptake8,32,33, but may be positive on SAA34. Furthermore, when neuroprotective treatments become available, iRBD patients would be ideal to receive treatment to postpone or even prevent the development of motor and cognitive symptoms. However, iRBD patients still suffer from several non-motor symptoms, including symptoms of orthostatic hypotension, urinary dysfunction, erectile dysfunction, constipation and sleep problems2,7,35,36, emphasizing the future need for even earlier diagnosis, i.e., before significant injury to the autonomic systems and brainstem has occurred.
This study has several limitations. First, the MIBG progression curves is interpreted in a body-first LBD framework but were defined from mixed PD cohorts, which had not been stratified into brain-first and body-first subgroups using polysomnography. However, it is currently not possible to determine an MIBG progression curve using iRBD patients only, as they often display nearly complete cardiac denervation at study inclusion. Thus, inclusion of brain-first (RBD-negative) cases is necessary to model the upper, more normal part of the MIBG progression curve. Still, a key assumption of our main result is that cardiac denervation follows the same trajectory in both subtypes, which might not necessarily be the case. If the sympathetic degeneration in body-first patients progresses faster than brain-first, we may have overestimated the prodromal period in body-first LBD. Second, the two MIBG cohorts did not follow the exact same imaging protocol, as KPD data were obtained 2 h post injection, whereas PACE MIBG data were obtained 3.5 h post injection. The impact of this discrepancy is probably limited, as one study found nearly identical H/M ratios at 2, 3, and 4 h post injection in both LBD subjects and atypical Parkinsonism37. Nevertheless, converting the values to percentage points relative to healthy controls appears to efficiently mitigate these differences, as the distribution of MIBG data was very similar between the two cohorts (Fig. 2). Third, manual delineation of the heart on the 2D scintigraphy may be challenging in the pathological ranges as the heart can be difficult to define and sometimes is completely invisible (Fig. 2). This clearly adds noise to the lower part of the curve where the progression is slow or absent, perhaps also explaining why some of these subjects displayed increasing MIBG H/M values at follow-up. Fourth, differences in medication between baseline and follow-up may also slightly influence progression rates, although the impact of the most common types of medications on MIBG uptake is minimal. Also, subjects on medication with known substantial influence on cardiac MIBG uptake were excluded. However, we cannot rule out that some types of medications have an unknown effect on the cardiac MIBG uptake. Finally, it should be emphasized that the predicted time to phenoconversion was based on extrapolation of the dopaminergic progression curve and not direct observations. Also, a key assumption is that the iRBD patients develop PD (or DLB) symptoms at a similar level of dopaminergic degeneration as observed in our PD cohort, which may not necessarily be true. However, our estimate of 9.8 years is in relatively close agreement with previous reports showing 50% phenoconversion after approximately 8 years2,38.
In summary, we determined MIBG progression curves, reflecting progressive cardiac sympathetic degeneration, from two separate patient cohorts. Thereby, we estimated that the onset of measurable peripheral neurodegeneration in iRBD patients may occur at least 11.3 years prior to study enrollment. By adding the estimated time to phenoconversion of median 9.8 years, our study suggests that the total prodromal/preclinical period in body-first LBD exceeds 20 years. Future studies are needed to determine the prodromal period of brain-first, RBD-negative LBD, which we predict is considerably shorter, as the dopaminergic system is probably among the first to be affected in this subtype.
Methods
Participants
Subject characteristics are presented in Table 1. We included data from three separate cohorts:
-
i.
A Korean Nationwide Parkinson’s Disease cohort39, recruited at movement disorder clinics of KPD St. Mary’s Hospital and Yeouido St. Mary’s Hospital (designated KPD), comprising early PD patients with longitudinal MIBG (KPD-PD).
-
ii.
A Lundbeck Foundation Parkinson’s Disease Research Center cohort, recruited at Aarhus University Hospital, Denmark (designated PACE), comprising de novo PD patients and iRBD patients (PACE-PD and PACE-iRBD) with longitudinal MIBG and baseline dopaminergic imaging ([18F]-FE-PE2I PET).
-
iii.
Data extracted from the Parkinson’s Progression Marker Initiative cohort (designated PPMI), comprising early PD patients (PPMI-PD) and iRBD patients (PPMI-iRBD) with longitudinal dopamine transporter (DaT) imaging. Data for this paper were acquired from the PPMI database on 2025-08-18. To estimate dopamine transporter loss over time, we included patients from the PPMI dataset with baseline DaTscans obtained between 2010-07-01 and 31-12-2019.
The KDP cohort was approved by the institutional review board of Seoul St. Mary’s Hospital, The Catholic University of Korea (KC21OIDI0362), and the PACE cohort by The Central Denmark region Committees on Health Research Ethics (1-10-72-231-19). All participants provided written informed consent in accordance with the Declaration of Helsinki. PD was diagnosed according to the Movement Disorder Society (MDS) diagnostic criteria in the KPD and PACE cohorts14. iRBD was diagnosed according to ICSD-3 criteria, based on video-polysomnography as specified in the AASM manual40. Patients with cardiac diseases with suspected impact on MIBG uptake, including larger myocardial infarctions and diabetic autonomic neuropathy, were excluded. In the PACE-iRBD dataset, patients diagnosed with DLB or PD within 6 months of the baseline scans were excluded from all analyzes (1 DLB, 1 PD). From the PPMI dataset, PD patients with LRRK2 and SNCA mutations were excluded due to their overrepresentation in the dataset relative to the expected background prevalence.
Imaging protocols
KPD
MIBG imaging was performed with a dual-head camera equipped with a low-energy, high-resolution collimator (Siemens, Germany). A late static image was obtained 2 h after injection of 111 MBq.
PACE
MIBG images were obtained 3.5 h after injection of 111 MBq on a dual-head camera equipped with a low-energy, high-resolution collimator (Siemens Symbia SPECT/CT, Germany). [18F]-FE-PE2I PET was obtained on a Siemens Biograph Vision 600 PET/CT (Siemens Healthcare, Erlangen, Germany) from 30 to 40 min after injection of 100 MBq tracer. Dopaminergic imaging data were normalized to MNI space with rigid matching of the subject’s PET to a T1 anatomical MRI performed on either a 3T Siemens SKYRA magnetic resonance system or GE Healthcare SIGNA PET/MR system.
PPMI
The dopamine transporter imaging ([99mTc]-TRODAT-1 or [123I]Ioflupane) protocol can be retrieved here: https://www.ppmi-info.org/study-design/research-documents-and-sops. Healthy controls with visually abnormal DaT-scans or known genetic mutation were excluded. For subjects with multiple available DaT-scans, only the first and last available scans were used for analyzes.
Data analyses
Dopaminergic imaging
[18F]-FE-PE2I PET data from PACE were analyzed in PMOD Biomedical Image Quantification Software (version 4.0, Zurich, Switzerland). Volumes of interest were defined in the putamen and occipital cortex using Hammers N30R83 atlas41. Similarly, putamen and occipital cortex data were extracted from the PPMI database. Specific putamen-to-occipital ratios were calculated as:
| 1 |
For PE2I-PET, putamen-to-occipital ratios (SBRs) were age-adjusted using in-house reference material from healthy controls with a linear model and a reference age of 65 years. A common slope was used for both right and left putamen. The most affected putamen was used for analysis. For DaT scans, the same approach was used based on healthy controls from the PPMI dataset. Only PD patients with a visually abnormal putamen on dopamine imaging were included, resulting in the exclusion of two patients from the PACE dataset and nine patients from the PPMI dataset.
MIBG scintigraphy
Regions of interest of the heart and mediastinum were manually defined, and average pixel counts were extracted to calculate heart-to-mediastinum (H/M) ratios. No age-correction was applied, as MIBG H/M ratios did not correlate with age in healthy control subjects.
Conversion to percentage of normal value
Specific putamen-to-occipital ratios were converted for [18F]-FE-PE2I and DaT SPECT, separately as:
| 2 |
MIBG H/M ratios were converted to percentage of healthy control mean as:
| 3 |
In the following, all percentages denote the percentage of the normal value (healthy control mean), i.e., not percentage loss.
To study the longitudinal dopaminergic and cardiac sympathetic degeneration, data were binned to quartiles based on baseline imaging results as follows: Q1 > 75%, Q2 = 50-75%, Q3 = 25-50%, and Q4 < 25% of the normal value. Annual decline rates were calculated as:
| 4 |
Importantly, this decline rate expresses %-points lost per year relative to the normal value, and not the percental specific-binding-ratio loss as reported in most other studies. For PPMI data, baseline and last visit data were used for analyzes. Baseline results and decline rates for each quartile are shown for dopaminergic imaging and MIBG in Table 2. We excluded patients with increasing imaging values at follow-up for three reasons. First, to exclude patients who have not entered the neurodegenerative phase (for the respective imaging modality). Second, to avoid noise in the lower range of the curve, where progression approaches zero due to a floor effect. Third, to get a conservative estimate of the prodromal LBD period and hence underestimating, rather than overestimating, the true prodromal period. Progression curves including all available data are shown in Supplementary Fig. 2.
Constructing the progression curves
The progression curves and individual data points are presented in Supplementary Fig. 3.
Start of the curve (Q1)
The curve begins at the median baseline value of Q1. The thick segment of the curve shows the median decline rate in Q1 until either (1) the curve reaches the last follow-up measurement (on the time axis), or (2) the curve reaches the median baseline value of Q2.
Connecting the segments
Thin lines connect the thick segments by extrapolation, using the same decline rate observed in the preceding quartile.
Q2 segment
From the median Q2 baseline imaging value, the next thick segment shows the median decline rate for Q2 subjects. This segment continues until it either reaches the last follow-up measurement (on the time axis), or the curve reaches the median baseline value of Q3, i.e., the same process as for Q1.
Q3 → Q4 → zero
The same process is repeated for Q3 and Q4 data and extrapolated to reach 0%.
Statistics
Descriptive statistics and graphical visualization of imaging data were performed in R (version 4.5.1; R Core Team, 2023) under macOS (version 26.0.1) using the tidyverse framework for data visualization.
Supplementary information
Acknowledgements
The KPD cohort was supported by the Korea National Institute of Health research project (2024ER100201). The Basic Science Research Program through the National Research Foundation of Korea (NRF), funded by the Ministry of Science, ICT, and Future Planning (NRF-2017R1D1A1B06028086). The PACE cohort was funded by the Lundbeck Foundation (grant numbers: R276-2018-294 and R381-2021-1485).
Author contributions
Conceptualization: C.S., O.L.M., P.B., J.H. Data acquisition and analyses: C.S., K.B.A., A.K., K.K., T.L., S.W.Y., D.W.R., Y.O., S.H., J.S.K., J.H. Manuscript preparation: C.S., P.B., J.H. All authors read and approved the final manuscript.
Data availability
The dataset analyzed for this study is not publicly available due to privacy restrictions, but de-identified data are available from the corresponding authors upon reasonable request.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
These authors contributed equally: Per Borghammer, Jacob Horsager.
Contributor Information
Casper Skjærbæk, Email: cas@clin.au.dk.
Per Borghammer, Email: borghammer@clin.au.dk.
Jacob Horsager, Email: jacobnls@rm.dk.
Supplementary information
The online version contains supplementary material available at https://doi.org/10.1038/s41531-026-01455-z.
References
- 1.Maya, G. et al. Post-mortem neuropathology of idiopathic rapid eye movement sleep behaviour disorder: a case series. Lancet Neurol.23, 1238–1251 (2024). [DOI] [PubMed] [Google Scholar]
- 2.Postuma, R. B. et al. Risk and predictors of dementia and Parkinsonism in idiopathic REM sleep behaviour disorder: a multicentre study. Brain142, 744–759 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Heinzel, S. et al. Update of the MDS research criteria for prodromal Parkinson’s disease. Mov. Disord.34, 1464–1470 (2019). [DOI] [PubMed] [Google Scholar]
- 4.Borghammer, P. & Van Den Berge, N. Brain-first versus gut-first Parkinson’s disease: a hypothesis. J. Parkinsons Dis.9, S281–S295 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Horsager, J. et al. Brain-first versus body-first Parkinson’s disease: a multimodal imaging case-control study. Brain143, 3077–3088 (2020). [DOI] [PubMed] [Google Scholar]
- 6.Kivisto, V. et al. Myocardial sympathetic distal axon loss in subjects with Lewy pathology in three autopsy cohorts. Acta Neuropathol.150, 11 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Knudsen, K. et al. In-vivo staging of pathology in REM sleep behaviour disorder: a multimodality imaging case-control study. Lancet Neurol.17, 618–628 (2018). [DOI] [PubMed] [Google Scholar]
- 8.Miyamoto, T. & Miyamoto, M. Reduced cardiac (123)I-MIBG uptake is a robust biomarker of Lewy body disease in isolated rapid eye movement sleep behaviour disorder. Brain Commun.6, fcae148 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Kashihara, K., Imamura, T. & Shinya, T. Cardiac 123I-MIBG uptake is reduced more markedly in patients with REM sleep behavior disorder than in those with early stage Parkinson’s disease. Parkinsonism Relat. Disord.16, 252–255 (2010). [DOI] [PubMed] [Google Scholar]
- 10.Matsubara, T. et al. Autopsy validation of the diagnostic accuracy of (123)I-metaiodobenzylguanidine myocardial scintigraphy for Lewy body disease. Neurology98, e1648–e1659 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Sisson, J. C. et al. Metaiodobenzylguanidine to map scintigraphically the adrenergic nervous system in man. J. Nucl. Med28, 1625–1636 (1987). [PubMed] [Google Scholar]
- 12.Orimo, S., Yogo, M., Nakamura, T., Suzuki, M. & Watanabe, H. 123)I-meta-iodobenzylguanidine (MIBG) cardiac scintigraphy in alpha-synucleinopathies. Ageing Res, Rev.30, 122–133 (2016). [DOI] [PubMed] [Google Scholar]
- 13.McKeith, I. G. et al. Diagnosis and management of dementia with Lewy bodies: fourth consensus report of the DLB Consortium. Neurology89, 88–100 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Postuma, R. B. et al. MDS clinical diagnostic criteria for Parkinson’s disease. Mov. Disord.30, 1591–1601 (2015). [DOI] [PubMed] [Google Scholar]
- 15.Ryu, D. W. et al. Initial versus follow-up sequential myocardial 123I-MIBG scintigraphy to discriminate Parkinson disease from atypical Parkinsonian syndromes. Clin. Nucl. Med.44, 282–288 (2019). [DOI] [PubMed] [Google Scholar]
- 16.Tsujikawa, K. et al. Chronological changes of 123I-MIBG myocardial scintigraphy and clinical features of Parkinson’s disease. J. Neurol. Neurosurg. Psychiatry86, 945–951 (2015). [DOI] [PubMed] [Google Scholar]
- 17.Lamotte, G., Holmes, C., Wu, T. & Goldstein, D. S. Long-term trends in myocardial sympathetic innervation and function in synucleinopathies. Parkinsonism Relat. Disord.67, 27–33 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Ikeda, K., Ebina, J., Kawabe, K. & Iwasaki, Y. Dopamine transporter imaging in Parkinson disease: progressive changes and therapeutic modification after anti-Parkinsonian medications. Intern Med58, 1665–1672 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Kaasinen, V. & van Eimeren, T. Balancing practicality and complexity in neuroimaging models of Parkinson’s disease progression. NPJ Parkinsons Dis.11, 262 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Kim, H. K. et al. Temporal trajectory model for dopaminergic input to the striatal subregions in Parkinson’s disease. Parkinsonism Relat. Disord.103, 42–49 (2022). [DOI] [PubMed] [Google Scholar]
- 21.Sakakibara, S. et al. Longitudinal change of DAT SPECT in Parkinson’s disease and multiple system atrophy. J. Parkinsons Dis.10, 123–130 (2020). [DOI] [PubMed] [Google Scholar]
- 22.Arnaldi, D. et al. Presynaptic dopaminergic imaging characterizes patients with REM sleep behavior disorder due to synucleinopathy. Ann. Neurol.95, 1178–1192 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Iranzo, A. et al. Dopamine transporter imaging deficit predicts early transition to synucleinopathy in idiopathic rapid eye movement sleep behavior disorder. Ann. Neurol.82, 419–428 (2017). [DOI] [PubMed] [Google Scholar]
- 24.Kuzkina, A. et al. Dermal real-time quaking-induced conversion is a sensitive marker to confirm isolated rapid eye movement sleep behavior disorder as an early alpha-synucleinopathy. Mov. Disord.38, 1077–1082 (2023). [DOI] [PubMed] [Google Scholar]
- 25.Siderowf, A. et al. Assessment of heterogeneity among participants in the Parkinson’s Progression Markers Initiative cohort using alpha-synuclein seed amplification: a cross-sectional study. Lancet Neurol.22, 407–417 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Liguori, R. et al. A comparative blind study between skin biopsy and seed amplification assay to disclose pathological alpha-synuclein in RBD. npj Parkinsons Dis.9, 34 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Okuzumi, A. et al. Propagative alpha-synuclein seeds as serum biomarkers for synucleinopathies. Nat. Med.29, 1448–1455 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Mammana, A. et al. RT-QuIC detection of pathological alpha-Synuclein in skin punches of patients with Lewy body disease. Mov. Disord.36, 2173–2177 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Marek, K. et al. Evidence for alpha-synuclein aggregation in older individuals with hyposmia: a cross-sectional study. eBioMedicine112, 105567 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Hoglinger, G. U. et al. A biological classification of Parkinson’s disease: the SynNeurGe research diagnostic criteria. Lancet Neurol.23, 191–204 (2024). [DOI] [PubMed] [Google Scholar]
- 31.Andersen, K. B. et al. Sympathetic and parasympathetic subtypes of body-first Lewy body disease observed in postmortem tissue from prediagnostic individuals. Nat. Neurosci. 10.1038/s41593-025-01910-9 (2025). [DOI] [PMC free article] [PubMed]
- 32.Lenka, A., Isonaka, R., Holmes, C. & Goldstein, D. S. Cardiac (18)F-dopamine positron emission tomography predicts the type of phenoconversion of pure autonomic failure. Clin. Auton. Res.33, 737–747 (2023). [DOI] [PubMed] [Google Scholar]
- 33.Nagayama, H., Hamamoto, M., Ueda, M., Nagashima, J. & Katayama, Y. Reliability of MIBG myocardial scintigraphy in the diagnosis of Parkinson’s disease. J. Neurol. Neurosurg. Psychiatry76, 249–251 (2005). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Andreasson, M. et al. Cutaneous nerve fiber pathology and function in Parkinson’s disease and atypical Parkinsonism - a cohort study. npj Parkinsons Dis.11, 170 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Barber, T. R. et al. Prodromal Parkinsonism and neurodegenerative risk stratification in REM sleep behavior disorder. Sleep40, 10.1093/sleep/zsx071 (2017). [DOI] [PMC free article] [PubMed]
- 36.Fedorova, T. D. et al. Imaging progressive peripheral and central dysfunction in isolated REM sleep behaviour disorder after 3 years of follow-up. Parkinsonism Relat. Disord.101, 99–104 (2022). [DOI] [PubMed] [Google Scholar]
- 37.Chun, I. K. Optimal Protocol and Clinical Usefulness of (123)I-MIBG cardiac scintigraphy for differentiation of Parkinson’s disease and dementia with Lewy body from non-Parkinson’s diseases. Nucl. Med. Mol. Imaging57, 145–154 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Iranzo, A. et al. Neurodegenerative disorder risk in idiopathic REM sleep behavior disorder: study in 174 patients. PLoS ONE9, e89741 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Oh, S. et al. Profile for brain disease research infrastructure for data gathering and exploration (BRIDGE) platform. Aging Dis. 10.14336/AD.2024.1432 (2025). [DOI] [PMC free article] [PubMed]
- 40.Sateia, M. J. International classifi cation of sleep disorders-third edition: highlights and modifi cations. Chest146, 1387–1394 (2014). [DOI] [PubMed]
- 41.Hammers, A. et al. Three-dimensional maximum probability atlas of the human brain, with particular reference to the temporal lobe. Hum. Brain Mapp.19, 224–247 (2003). [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The dataset analyzed for this study is not publicly available due to privacy restrictions, but de-identified data are available from the corresponding authors upon reasonable request.