Introduction

Neuroimaging evidence indicating cardiac sympathetic denervation in the Lewy body (LB) synucleinopathy Parkinson disease (PD) and generally intact innervation in the non-LB synucleinopathy multiple system atrophy (MSA) was first reported using 18F-dopamine (18F-DA) positron emission tomography (PET) imaging [1]. Numerous studies since then, mainly using 123I-metaiodobenzylguanidine (123I-MIBG) single-photon emission computed tomographic (SPECT) scanning, have confirmed these findings. Of 21 studies in English on this topic involving at least 50 patients, all 21 have reported a significant group difference in the direction of cardiac noradrenergic deficiency in PD (Table 1).

Table 1 Studies using cardiac sympathetic neuroimaging in patients with Parkinson disease (PD) and multiple system atrophy (MSA)
Full size table

The issue of whether cardiac sympathetic neuroimaging can aid the differential diagnosis of PD vs. MSA seems settled, to the extent that it can be settled. 123I-MIBG SPECT scanning for this purpose has been approved for several years in Europe and Asia. In the United States, however, 123I-MIBG SPECT scanning is rarely done for this indication, due to lack of reimbursement by third-party payers. Cardiac PET neuroimaging is conduced only in research settings.

Intra-cardiac regional differences in sympathetic abnormalities

In this issue of Clinical Autonomic Research, Eckhardt and colleagues report a small, cross-sectional study of patients diagnosed clinically with idiopathic PD or the parkinsonian form of MSA (MSA-P) [2]. Anatomic co-registration of 123I-MIBG SPECT images and low-dose chest computed tomography (CT) enabled the examination of regional 123I-MIBG-derived radioactivity within the left-ventricular myocardium. Digital thresholding was applied based on voxel intensity to identify 3 different patterns of radioactivity, homogeneous, non-homogeneously reduced, and absent. Among 10 patients with homogeneous radioactivity, 8 had MSA-P and 2 PD; among 7 with non-homogeneously reduced radioactivity, 3 had MSA-P and 4 PD; and among 9 with absent radioactivity, 1 had MSA-P and 8 PD. Non-homogeneously reduced tracer uptake was most common in the apex and the infero-lateral left-ventricular myocardium.

The same non-homogeneous pattern has been found by 18F-DA [3] and 11C-hydroxyephedrine [4] PET scanning in subgroups of PD patients. The present report confirms these findings using 123I-MIBG SPECT.

Relevance of orthostatic hypotension

In synucleinopathies, a cardinal manifestation of autonomic failure is neurogenic orthostatic hypotension (OH). Abnormal cardiac sympathetic neuroimaging is more consistently found in PD + OH than in PD without OH, by both 123I-MIBG SPECT scanning [5] and 18F-DA PET scanning [6]. In the study by Eckhardt et al., OH was present in only 9/28 (32%) patients (5 PD, 4 MSA-P) [2]. This limits the implication of the study for clinical autonomic research, because in autonomic centers, OH with central neurodegeneration is often the relevant differential diagnostic issue. A recent retrospective analysis of 18F-DA PET imaging separated patients with PD + OH from MSA-P with high sensitivity (92%) and specificity (96%) [7].

Functional correlates of cardiac noradrenergic deficiency

Eckhardt et al. also investigated relationships of 123I-MIBG imaging with physiological indices of autonomic function. An absence of a blood pressure overshoot above baseline after performance of the Valsalva maneuver, a measure of baroreflex-sympathoneural failure, was associated with low myocardial 123I-MIBG-derived radioactivity in PD but not in MSA [2]. This association could reflect postganglionic sympathetic noradrenergic impairment but also central baroreflex-sympathoneural failure.

Eckhardt and colleagues noted a lack of association between cardiac sympathetic innervation and autonomic symptom burden or OH; however, the number of patients with OH may have been too small to conduct meaningful statistical testing.

Clinical significance of cardiac noradrenergic deficiency

Inconsistent results have been reported in the literature regarding the association between motor symptoms and cardiac noradrenergic deficiency in PD. The extent to which cardiac noradrenergic deficiency contributes to cardiovascular dysfunction in synucleinopathies is still unclear and requires further research.

Cardiac sympathetic neuroimaging may provide a better understanding of the pathophysiology of other non-motor symptoms, such as anosmia, rapid eye movement behavior disorder, fatigue, and cognitive dysfunction, all of which have been associated with evidence of myocardial noradrenergic deficiency [8]. Mechanisms for the associations between these non-motor symptoms and cardiac noradrenergic deficiency in PD remain unknown.

Progression biomarkers

There is a push to develop quantifiable biomarkers of disease progression for use in clinical trials for neurological diseases. Longitudinal studies are far less common than cross-sectional studies such as that by Eckhardt et al. Using 18F-DA PET imaging, we have found that 18F-DA-derived radioactivity decreases by a median of 4% per year in PD [9]. Because of individual variability in the rate of decline, a clinical trial of a disease-modifying approach using 18F-DA-derived radioactivity as an outcome measure would require about 70 patients in each of the experimental treatment and standard treatment groups to detect a 50% decrease in the loss of cardiac noradrenergic innervation over a 5-year-period, which would represent a substantial study. Thus, cardiac neuroimaging may be more fitted as a secondary outcome or as part of a composite outcome, in which multiple endpoints are combined. Importantly, there is evidence that decreased vesicular uptake in cardiac sympathetic nerves is present upon initial evaluation of patients with LB synucleinopathies and may provide a biomarker of catecholaminergic dysfunction early in the disease process [9].

Neuroimaging evidence of cardiac noradrenergic deficiency can precede striatal dopamine deficiency and the onset of motor signs in PD [10]. In the prospective, longitudinal intramural NINDS PDRisk study, in individuals at high risk for developing PD (at least 3 of the following risk factors: genetic susceptibility, olfactory dysfunction, dream enactment, and OH), low cardiac 18F-DA-derived radioactivity has been reported to predict a diagnosis of PD by 3 years of follow-up [11]. The findings indicate the potential of cardiac sympathetic neuroimaging for providing a preclinical biomarker of PD; however, the generalizability to the broader population of individuals with fewer risk factors is unknown.

Dysfunction vs. denervation

PET imaging agents such as 18F-DA, 11C-epinephrine, and 11C-phenylephrine are substrates for metabolism by monoamine oxidase (MAO). After uptake into the neuronal cytoplasm, these agents have two alternative fates. The main fate is vesicular uptake; oxidative deamination via MAO is a minor fate. By measuring the deaminated metabolite of 18F-DA simultaneously with 18F-DA, one can examine the efficiency of vesicular sequestration. It was by applying this approach that decreased intra-neuronal vesicular uptake was discovered in LB diseases [6]. Confirmation of this finding has come from tracking the radioactivity during the testing session. PD + OH is associated with accelerated loss of 18F-DA-derived radioactivity [12]. The implications for treatment of this functional abnormality are substantial. Put simply, you cannot treat dead neurons, whereas neurons that are dysfunctional but extant (“sick-but-not-dead”) might be salvaged.

This functional abnormality cannot be quantified by 123I-MIBG SPECT or by 18F-hydroxyephedrine PET, since neither imaging agent is a substrate for MAO. On the other hand, accelerated loss (“washout”) of 123I-MIBG-derived radioactivity might provide a biomarker of increased sympathetically mediated exocytosis. This possibility can be tested by determining effects of sympatholysis on 123I-MIBG-derived radioactivity.

PAF as a prototype of “body-first” PD

A model of brain-first versus body-first subtypes of PD has been proposed by Borghammer and colleagues [13]. In the brain-first (top–down) type, αS pathology initially occurs in the brain, with secondary spreading to the peripheral autonomic nervous system; whereas in the body-first (bottom–up) type, the pathology originates in the enteric or peripheral autonomic nervous system and then spreads to the brain. The Lewy body form of PAF can evolve to PD or dementia with Lewy bodies (DLB) [14]. Measuring CSF levels of catecholamines and analyzing post-mortem data in autonomic synucleinopathy patients and controls, we reported recently that the synucleinopathies have in common evidence of central noradrenergic deficiency but differ in extents of central dopaminergic deficiency—prominent in PD and MSA, less pronounced in PAF [15]. Therefore, PAF may be the prototype for peripheral-to-central disease progression in LB diseases. Only a minority of PAF patients seem to develop symptomatic central synucleinopathy; however, these patients might be identifiable based on biomarkers of central nervous system involvement, such as by 18F-DOPA PET scanning [10] or by brainstem neuromelanin magnetic resonance imaging [16].

PET vs. SPECT imaging

PET scanning offers several advantages over SPECT scanning. These include better spatial resolution, a smaller amount of injected radioactivity, the possibility of measuring radioactivity concentrations in tissues in absolute terms, and, importantly, the analysis of curves relating tissue radioactivity vs. time (time–activity curves), which provides valuable information about not only innervation but also functional abnormalities in residual nerves, such as a vesicular storage defect [6]. The time frames involved with PET and SPECT scanning differ. Injected cardiac sympathetic neuroimaging agents exit the bloodstream almost instantly and are taken up by sympathetic nerves extremely rapidly. Although the initial 123I-MIBG SPECT scan, at about 15–30 min, is called “uptake,” peak 18F-DA-derived radioactivity normally occurs by about 5 min after tracer injection.

Comparing 18F-DA PET with 123I-MIBG SPECT scanning in the same patients with autonomic synucleinopathies would provide valuable information, and we would welcome investigators outside the NIH to join in the IND covering 18F-DA PET scanning, with a view toward a multi-center collaborative study.

Implications

Research on cardiac sympathetic neuroimaging in synucleinopathies is not dead. Early involvement of the autonomic nervous system should be the focus of future studies. In patients for whom the pathophysiological process starts in the periphery and propagates to the brain, understanding and modulating the peripheral process might not only help alleviate the associated non-motor symptoms but also prevent the central degeneration that leads to the motor signs and symptoms. Cardiac sympathetic neuroimaging combined with other measurable and robust biomarkers of preclinical disease may serve as a tool to target catecholaminergic neurodegeneration at the earliest stages of LB diseases.