Roles of cardiac sympathetic neuroimaging in autonomic medicine

Abstract

Sympathetic neuroimaging is based on injection of compounds that radiolabel sites of the cell membrane norepinephrine transporter (NET) or that are taken up into sympathetic nerves via the NET and radiolabel intra-neuronal catecholamine storage sites. Detection of the radioactivity is by planar or tomographic radionuclide imaging. The heart stands out among body organs in terms of the intensity of radiolabeling of sympathetic nerves, and virtually all of sympathetic neuroimaging focuses on the left ventricular myocardium. The most common cardiac sympathetic neuroimaging method world-wide is ¹²³I-metaiodobenzylguanidine (¹²³I-MIBG) scanning. ¹²³I-MIBG scanning is used routinely in Europe and East Asia in the diagnostic evaluation of neurogenic orthostatic hypotension (nOH), to distinguish Lewy body diseases (e.g., Parkinson disease with orthostatic hypotension, pure autonomic failure) from non-Lewy body diseases (e.g., multiple system atrophy) and to distinguish dementia with Lewy bodies from Alzheimer’s disease. In the United States ¹²³I-MIBG scanning is FDA-approved for evaluation of pheochromocytoma and some forms of heart failure but not for the above differential diagnoses. Positron emission tomographic methods based on imaging agents such as ¹⁸F-dopamine are categorized as research tools, despite more than a quarter century of clinical experience with them. Considering that ¹²³I-MIBG scanning is available at most academic medical centers in the United States, cardiac sympathetic neuroimaging by this methodology merits consideration as an autonomic test, especially in patients with nOH.

Introduction

Sympathetic neuroimaging is rarely used as part of autonomic function testing in the United States, although this modality is used commonly in Europe and Japan and has been for decades. The discrepancy is due mainly to issues related to approvals by regulatory agencies and third party payors. This review summarizes evidence about the clinical utility of sympathetic neuroimaging, with emphasis on the evaluation of neurogenic orthostatic hypotension (nOH).

Historical Perspective

Sympathetic neuroimaging developed as an offshoot of efforts to visualize and treat endocrine tumors using radioiodinated compounds by William H. Beierwaltes (1917–2005), one of the founders of nuclear medicine [106], at the University of Michigan. He was a co-holder of a patent for radioiodinated Iobenguane, which is synonymous with metaiodobenzylguanidine (MIBG) (Fig. 1). In the early 1980s animal studies showed that after intravenous injection of ¹³¹I- or ¹²³I-MIBG the heart was visible on planar or tomographic images and that reserpine, which blocks vesicular uptake of monoamines, decreased the amount of cardiac radioactivity, indicating cardiac neuronal uptake of the tracer and retention of the radioactivity in storage vesicles in myocardial sympathetic nerves [129]. Beierwaltes’s group began clinical studies involving cardiac sympathetic neuroimaging by ¹²³I-MIBG scanning in the 1980s [107]. Notably, in 1987 they reported, “Generalized autonomic neuropathies were associated with marked diminutions of [¹²³I]MIBG uptake into the heart.”

Fig. 1: Chemical structures of some sympathetic neuroimaging agents.

The agents can be classified in terms of catecholamines and non-catecholamines.

The advent of clinical positron emission tomographic (PET) scanning at about that time led to somewhat parallel efforts to visualize cardiac sympathetic innervation by using ¹⁸F-labelled compounds. Groups at the Brookhaven National Laboratory [22–24] and intramural NIH independently developed ¹⁸F-dopamine [36], and the first clinical ¹⁸F-dopamine scanning studies were reported in the early 1990s [38]. With a few exceptions [17], since then clinical ¹⁸F-dopamine PET scanning has been done solely at the NIH Clinical Center. PET scanning for cardiac sympathetic neuroimaging has continued to be refined at the University of Michigan, with the use in particular of ¹¹C-meta-hydroxyephedrine [96]. A positron-emitting analog of MIBG is under development [95].

Overview of sympathetic neuroimaging agents

One can classify sympathetic neuroimaging agents in terms of two types of compounds—catecholamines and sympathomimetic amines (Fig. 1). As noted recently in this series [37], catecholamines are catechols (note the “head of the cat”). This review will focus on ¹⁸F-dopamine as the prototype of a catecholamine PET imaging agent. ¹⁸F-Dopamine was the first catecholamine PET imaging agent to be used in clinical autonomic research [38].

Inspection of the chemical structures in Fig. 1 shows that ¹²³I-MIBG is not a catecholamine, a catechol, or a chemical analog of the catecholamine norepinephrine, as sometimes is claimed [11]. Instead, ¹²³I-MIBG is an analog of the sympatholytic agent guanethidine. Catecholamines and sympathomimetic amines are analogous pharmacologically in that they share the characteristic of being substrates for the Uptake-1 process [55] mediated by the cell membrane norepinephrine transporter (NET). Catecholamines and sympathomimetic agents differ importantly, however, in their respective metabolic fates in neuronal and non-neuronal cells (Fig. 2). Catecholamines such as ¹⁸F-dopamine and ¹¹C-epinephrine are substrates for monoamine oxidase (MAO) and catechol-O-methyltransferase (COMT). This offers the opportunity, by simultaneous measurement of the parent catecholamines and their metabolites, to obtain important insights about specific mechanisms underlying catecholamine depletion. In particular, concurrent measurement of myocardial ¹⁸F-dopamine-derived radioactivity and arterial plasma levels of ¹⁸F-dihydroxyphenylacetic acid (¹⁸F-DOPAC) enabled the discovery of an intra-neuronal vesicular storage defect in Lewy body diseases [41]. Combined neuroimaging/neurochemical approaches showed that under resting conditions most of the turnover of catecholamines in sympathetic nerves occurs via net leakage from vesicles, not release with escape from neuronal reuptake [39]. The neuronal uptake of positron-emitting catecholamines occurs extremely rapidly, so that there is little confusion about distinguishing “uptake” and “washout” of the radioactivity and little concern of substantial radioactivity remaining in myocardial cells.

Fig. 2: Concept diagrams for the different fates of sympathetic neuroimaging agents that are catecholamines vs. those that are not catecholamines.

Catecholaminergic sympathetic neuroimaging agents are substrates for monoamine oxidase (MAO) and catechol-O-methyltransferase. ¹²³I-MIBG is not.

¹²³I-MIBG, ¹¹C-meta-hydroxyephedrine, and other radiolabeled sympathomimetic amines are not substrates of MAO or COMT. They are better substrates than catecholamines for extra-neuronal uptake via the Uptake-2 process mediated by organic cation transporters (OCT), and little is known about the metabolic fate of the tracers in myocardial cells. This means that there is more non-specific radioactivity with these tracers than with catecholamines. ¹²³I-MIBG is also not as good a substrate for the NET or vesicular monoamine transporter (VMAT) in sympathetic nerves as are catecholamines. Nevertheless, from the point of view of detecting sympathetic noradrenergic denervation, both types of agents work well because of the shared feature of being substrates for the NET.

¹¹C-Methylreboxetine (¹¹C-MRB) is a NET ligand, and ¹¹C-dihydrotetrabenazine and ¹⁸F-tetrabenazine are ligands for the type 2 VMAT, the form of VMAT in sympathetic nerves. There is no published information about whether these ligands yield satisfactory cardiac sympathetic images. In our experience, ¹⁸F-DOPA, which is used to visualize central dopaminergic innervation, is of no value as a sympathetic neuroimaging agent; this is because ¹⁸F-DOPA is a neutral amino acid and as such is taken up by myocardial cells via the neutral amino acid transporter.

¹²³I-MIBG

Almost all ¹²³I-MIBG scanning studies involve calculation of the heart/mediastinum (H/M) ratio of ¹²³I-MIBG-derived radioactivity (Fig. 3). H/M ratios are measured during an “uptake” phase of about 15–30 minutes and a subsequent “washout” phase of a few hours.

Fig. 3: ¹²³I-MIBG SPECT/CT scans scans showing (left) normal left ventricular ¹²³I-MIBG-derived radioactivity, indicated by a high heart (h)/mediastinum (m) ratio and (right) abnormal ¹²³I-MIBG-derived radioactivity with a low h/m ratio.

Figure reproduced with permission of P. Borghammer.

Quantification of radioactivity in regions of interest is difficult because of limited spatial and temporal resolution and lack of attenuation correction. Quantitative methods addressing these limitations are being developed currently [132].

The H/M ratio is a crude parameter influenced importantly by the camera-collimator combinations. One of the reasons for the successful use of ¹²³I-MIBG scanning in Japan is the standardized approach for quantification. Cross-calibration methods have been published for Japan [82] and more recently in Europe [125].

The relatively long scanning intervals in typical ¹²³I-MIBG SPECT scanning render the meaning of “uptake” at best complicated. Because of the extremely rapid exit of injected sympathomimetic amines from the bloodstream, neuronal uptake via the NET, and vesicular uptake via the VMAT, the “uptake” phase probably actually corresponds to a period when not only uptake but also loss of the radioactivity is occurring. Moreover, the loss of radioactivity may reflect not only sympathetically mediated exocytosis but also vesicular leakage with possible backwards exit of the radioactivity via the NET because of the concentration gradient as ¹²³I-MIBG levels in the extracellular space fall.

¹²³I-MIBG scintigraphy vs. PET imaging

No study has compared ¹²³I-MIBG scanning with scanning using a PET imaging agent. Each modality has advantages and disadvantages.

There are practical limits on temporal resolution of ¹²³I-MIBG scintigraphy because of the time required for sufficient radioactivity to accumulate to produce an interpretable image. Attenuation correction, which is needed for quantifying radioactivity within regions of interest, is not done currently, and virtually all literature on the subject is based on H/M ratios. ¹²³I-MIBG scanning rarely is performed along with myocardial perfusion imaging, meaning that separating a sympathetic lesion from a coronary arterial or microvascular lesion can be impossible. This is especially a consideration in the evaluation of diabetic autonomic neuropathy.

PET scanning for cardiac sympathetic neuroimaging requires a system for rapid delivery of the imaging agent from the site of production to the site of administration; personnel carrying out radiochemical syntheses by hand must be monitored carefully to stay within radiation safety guidelines; and there are substantial expenditures for space, equipment, maintenance, and software. Very few centers do cardiac sympathetic neuroimaging by PET scanning, and the procedure is not approved in the United States for any medical indication.

Neurogenic Orthostatic Hypotension (nOH)

Cardiac sympathetic neuroimaging is very useful in the evaluation of patients with neurogenic orthostatic hypotension (nOH). A 4-step algorithmic approach has been proposed based on whether the OH is persistent and consistent, has an identifiable cause, is neurogenic, and is associated with sympathetic noradrenergic denervation [46] (Fig. 4). It is at the last step that cardiac sympathetic neuroimaging comes in.

Fig. 4: Algorithm for the clinical and laboratory evaluation of orthostatic hypotension.

Cardiac sympathetic neuroimaging is used to identify sympathetic noradrenergic denervation. Abbreviations: AAD=autoimmunity-associated autonomic failure with sympathetic denervation; AAG=autoimmune autonomic ganglionopathy; DLB =dementia with Lewy bodies; MSA-C=cerebellar MSA MSA-D=MSA with cardiac sympathetic denervation; MSA-P=parkinsonian MSA; PAF=pure autonomic failure; PD+OH=Parkinson disease with orthostatic hypotension.

If there were a suspected secondary cause of nOH such as diabetic autonomic neuropathy, cardiac sympathetic neuroimaging would not be of value diagnostically. Similarly, in nOH associated with peripheral neuropathy (especially “small fiber neuropathy”), cardiac sympathetic neuroimaging has unclear value. In a patient with clinical central neurodegeneration and nOH neuroimaging evidence of cardiac sympathetic denervation argues for PD+OH or DLB+OH, whereas MSA is less likely. On the other hand, the finding of neuroimaging evidence of intact cardiac sympathetic innervation is strong evidence against PD+OH or DLB+OH. In a patient with nOH, no signs of CNS involvement, and neuroimaging evidence of intact cardiac sympathetic innervation, the diagnosis is not necessarily autoimmune autonomic ganglionopathy (AAG), and MSA beginning with autonomic failure is a possibility.

Alternative means to identify sympathetic noradrenergic denervation include plasma levels of catechols during supine rest [44], skin biopsies with a focus on small nerve fibers in arrector pili muscles that receive pure sympathetic noradrenergic innervation [25], and neuropharmacologic probes such as tyramine infusion [53, 103]. It is likely that none of the alternatives is as sensitive as cardiac sympathetic neuroimaging, because in nOH the cardiac noradrenergic lesion is more severe than in other organs or the body as a whole [42, 116].

Lewy body diseases

A substantial and remarkably consistent literature has documented neuroimaging evidence of cardiac sympathetic denervation in Lewy body diseases, including Parkinson disease (PD), dementia with Lewy bodies (DLB), and pure autonomic failure (PAF) [88, 135]. Probably the first report describing neuroimaging evidence of a cardiac sympathetic lesion in PD was by Hakusui et al. in 1994 [49]. There has been post-mortem confirmation of this finding based on profoundly decreased immunoreactive tyrosine hydroxylase in epicardial nerves and myocardial tissue [8, 114] and based on myocardial norepinephrine contents [47, 135].

Since essentially all patients with PD+OH have neuroimaging evidence of cardiac sympathetic denervation [42, 45], the finding of intact innervation in a patient with nOH and parkinsonism excludes PD+OH.

Recent studies have supported the value of combining cardiac sympathetic neuroimaging with central neural biomarkers of Lewy body diseases, such as ¹²³I-FP-CIT SPECT (also called ¹²³I-ioflupane SPECT and DaT scanning [85–87, 123]) and calculation of apparent diffusion coefficients [122]. Another potential approach is to combine cardiac sympathetic neuroimaging with analysis of skin biopsies to identify denervation [33] or synucleinopathy [26] in skin constituents that receive sympathetic noradrenergic innervation such as arrector pili muscles.

Parkinson disease vs. multiple system atrophy

Most patients with MSA, a non-Lewy body form of alpha-synucleinopathy, have neuroimaging evidence of intact cardiac sympathetic innervation [40, 42, 93], although there are exceptions [18, 92, 96].

The vast majority of studies about differentiating PD from MSA by cardiac sympathetic neuroimaging have begun with the movement disorder, not with nOH (Table 1). These studies almost always have not stratified PD in terms of the occurrence of nOH. In Europe ¹²³I-MIBG scintigraphy is recommended by European Federation of Neurological Societies as a diagnostic tool in the evaluation of PD, without regard to OH [9].

Table 1:

Cardiac sympathetic neuroimaging to distinguish PD from MSA

First Author	Year	N PD	N MSA	Sig. diff.?	Citation	Note
Braune	1999	15	5	Yes	[13]
Goldstein	2000	29	24	Yes	[42]
Reinhart	2000	21	7	Yes	[98]
Druschky	2000	10	20	Yes	[27]
Braune	2001	246	45	Yes (meta-analysis)	[12]
Courbon	2003	18	10	Yes	[19]
Saiki	2004	34	7	Yes	[101]
Matsui	2005	17	9	Yes	[76]
Oka	2006	39	17	Yes	[89]
Kashihara	2006	130	11	Yes	[61]
Raffel	2006	9	10	Yes	[96]	Denerv. in MSA too
Rascol	2009	N/A	N/A	(Review)	[97]
Treglia	2011	1226	129	Yes (meta-analysis)	[121]
Orimo	2012	625	N/A	Yes (meta-analysis)	[94]
Umemura	2013	118	20	Yes	[122]
Yang	2017	25	18	Yes	[134]

In PD without OH, only about one-half of the patients have evidence of cardiac sympathetic denervation, a minority have partial denervation, typically in the inferolateral region, with sparing of the anterobasal region (Fig. 5), and a small minority have normal neuroimaging [41, 47, 70, 112, 131]. Without stratification in terms of nOH there can be substantial overlap between PD and MSA (Fig. 6).

Fig. 5: Partial and diffuse cardiac sympathetic noradrenergic lesions indicated by ¹⁸F-dopamine PET/CT scanning.

The cardiac sympathetic lesion can be more prominent in the inferolateral wall than the anteroseptal base of the heart and progress to become diffuse. The subjects gave written informed consent to participate in an IRB-approved protocol before undergoing any research procedures.

Fig. 6: Mean (± SEM) values for interventricular septal myocardial ¹⁸F-derived radioactivity in patients with PD with neurogenic orthostatic hypotension (PD+OH), pure autonomic failure (PAF), PD without neurogenic orthostatic hypotension (PD No OH), multiple system atrophy (MSA), normal volunteers (Normal), and individuals at increased risk for PD (PDRisk [43]).

Numbers in rectangles are numbers of subjects. Mean ages are shown below the group labels. The patients gave written informed consent to participate in an IRB-approved protocol before undergoing any research procedures. Numbers in italics are p values for factorial analyses of variance comparing the PD+OH, PD No OH, and Normal groups. Note lower ¹⁸F-derived radioactivity in PD+OH than in PD No OH, even in groups of similar age.

There seem to be two historical trends that have led to different justifications for including cardiac sympathetic neuroimaging in autonomic medicine. The first is within the domain of the differential diagnosis of chronic autonomic failure and specifically of nOH. The relevant question is: In a patient with nOH and signs of central neurodegeneration, does cardiac sympathetic neuroimaging distinguish PD+OH from MSA? The second trend is within the domain of movement disorders and specifically of parkinsonism. The relevant question is: In a patient with parkinsonism, does cardiac sympathetic neuroimaging distinguish PD from MSA? Answering the first question requires studying patients with PD+OH specifically, whereas the second question can be answered by studying patients with PD regardless of the occurrence of OH. Some studies have reported similarly severely decreased myocardial ¹²³I-MIBG-derived radioactivity in PD independently of OH [48, 75]; others have noted that a sub-population of patients with PD No OH have normal myocardial ¹²³I-MIBG-derived radioactivity [62, 64, 84].

Studies based on ¹⁸F-dopamine PET scanning help resolve this inconsistency. We have found that at the time of initial evaluation patients with PD+OH average about a decade older than do patients with PD No OH [41]. It is possible that patients with PD No OH who are relatively old at the time of motor onset are more likely to have cardiac sympathetic denervation. To take into account the difference in group mean age, from our large ongoing database we deleted data from older PD+OH and younger PD No OH patients to achieve groups matched for mean age (Figure 6). When this was done the PD+OH group still had lower myocardial ¹⁸F-dopamine-derived radioactivity than did the PD No OH group. It is possible that inconsistency in the literature on ¹²³I-MIBG-derived radioactivity in PD+OH vs. PD No OH may reflect insufficient attention to the ages at which the groups were compared or inadequate numbers of subjects to avoid false-negative results. The issue is of some consequence, because if a proportion of PD No OH patients had normal ¹²³I-MIBG-derived radioactivity, this would limit the specificity of the test for differential diagnosis from MSA.

To summarize the matter, ¹²³I-MIBG scanning is valuable for distinguishing PD from MSA, but it is even more valuable for distinguishing PD+OH from MSA. This is because a proportion of PD No OH patients have normal ¹²³I-MIBG data.

Combining cardiac sympathetic neuroimaging with plasma levels of catechols adds relatively little compared to cardiac sympathetic neuroimaging alone in differentiating PD from MSA. In our experience, low plasma levels of 3,4-dihydroxyphenylglycol (DHPG, the main neuronal metabolite of norepinephrine) occurs in some patients with PD+OH but is normal in PD without OH.

Amyloidosis

In most cases amyloidosis is acquired. Amyloidosis has been classified into several forms, including immunoglobulin light chain diseases, chronic inflammatory disorders, dialysis, old age, and hereditary. When presenting mainly as sensory, autonomic, and motor polyneuropathy, the term autonomic polyneuropathy has been used, and when presenting mainly as cardiomyopathy, amyloid cardiomyopathy has been used. A mixed form also can occur.

In contrast with alpha-synuclein, intra-neuronal accumulation of which can explain autonomic neuropathy in Lewy body diseases, amyloid accumulates extracellularly. Exactly how amyloidosis leads to sympathetic denervation is poorly understood. Decreased H/M ratios of ¹²³I-MIBG-derived radioactivity in amyloid autonomic neuropathy are associated with poor prognosis [108].

Hereditary transthyretin amyloidosis is a rare but clinically and scientifically important disorder that is transmitted as an autosomal dominant trait [35]. The disease results from extracellular deposition of the mutated protein, transthyretin. In hereditary transthyretin amyloidosis cardiac dysautonomia assessed by ¹²³I-MIBG scanning predicts long-term survival after liver transplantation [7].

Dementia with Lewy bodies

Neuroimaging evidence of cardiac sympathetic denervation characterizes dementia with Lewy bodies (DLB) [66, 88, 113] and predicts DLB in individuals with amnestic mild cognitive impairment [30], although it should be noted that to date cardiac sympathetic neuroimaging studies have not been done in groups of patients with subsequently pathologically confirmed diagnoses [111].

This modality is useful for differentiating DLB from Alzheimer’s disease and other forms of dementia (Table 2) [61, 111, 137]. The distinction is clinically relevant in allowing for early initiation of effective treatment with acetylcholinesterase inhibitors and avoiding adverse effects from antipsychotic medications that are dopamine receptor blockers. Dopamine receptor blockade in DLB could precipitate overt parkinsonism or worsen OH.

Table 2:

Cardiac sympathetic neuroimaging to distinguish DLB from Alzheimer’s disease

First Author	Year	N DLB	N AD	Sig. diff.?	Citation	Note
Yoshita	2001	14	14	Yes	[136]
Taki	2004	N/A	N/A	Yes (Review)	[115]
Orimo	2005	7	10	Yes	[91]
Hanyu	2006	32	40	Yes	[50]
Hanyu	2006	19	39	Yes	[50]
Yoshita	2006	37	42	Yes	[137]
Nakajima	2008	N/A	N/A	N/A (Review)	[83]
Rascol	2009	N/A	N/A	N/A (Review)	[97]
Treglia	2010	N/A	N/A	N/A (Review)	[119]
Kim	2015	22	37	Yes	[63]
Yoshita	2015	61	46	Yes	[135]	Multi-center study
Manabe	2017	79	19	Yes	[72]
Abbasi	2017	7	10	Yes	[1]
McKeith	2017	N/A	N/A	N/A	[79]	Consensus statement
Sonni	2017	N/A	N/A	N/A (Review)	[111]
Toru	2018	56	59	Yes	[117]
Nuvoli	2018	N/A	N/A	N/A (Review)	[86]

As in the diagnostic evaluation of parkinsonism, combining cardiac sympathetic neuroimaging with biomarkers of central dopaminergic deficiency may improve the accuracy of the testing for DLB [111, 120].

Heart failure

Patients with childhood dilated cardiomyopathy [78], hypertrophic cardiomyopathy [90], acute myocarditis [2], and heart failure [51, 69] have been reported to have abnormal ¹²³I-MIBG cardiac scanning. The severity of the scanning abnormality is associated with the functional classification of the heart failure [51]. A low H/M ratio of ¹²³I-MIBG-derived radioactivity is associated with poor outcome in idiopathic dilated cardiomyopathy[32, 58, 77, 81], ischemic cardiomyopathy [128] and congestive heart failure [60].

Treatment of heart failure is associated with improvement in the H/M ratio [3, 15, 20, 59, 104, 109, 110, 118, 124].

In patients with chronic heart failure, low H/M ratios of ¹²³I-MIBG-derived radioactivity are associated with an increased risk of cardiac events [57] including sudden death [65]. In patients with an implanted cardiac defibrillator and well-compensated heart failure, ¹²³I-MIBG washout is positively related to fast ventricular arrhythmic episodes [68] and other ventricular arrhythmias [10].

In the United States, ¹²³I-MIBG, marketed as AdreView by GE Healthcare, was approved by the FDA in 2013 for cardiac sympathetic neuroimaging in heart failure (NYHA class 2–3 heart failure with a left ventricular ejection fraction of 35% or less). This was based on data from the ADMIRE-HF trial. In this observational study, 25% of 961 patients with systolic heart failure who underwent ¹²³I-MIBG scanning had a primary event (NYHA functional class progression, potentially life-threatening arrhythmia, or cardiac death). The two-year event rate was 15% for those who had an H/M ratio > 1.60 and 37% for those with an H/M ratio < 1.60, for a hazard ratio of 0.40 [57]. Whether ¹²³I-MIBG scanning helps identify patients likely to benefit from implantation of a cardiac defibrillator is under investigation.

There are several potential explanations for low H/M ratios of ¹²³I-MIBG-derived radioactivity in heart failure. Cardiac sympathetic denervation cannot be sufficient because of the relatively rapid improvement by essentially all treatments that reduce the severity of the heart failure. Neuronal uptake of norepinephrine in the heart has been reported to be normal in congestive heart failure [80]. It seems likely that the low H/M ratios of ¹²³I-MIBG-derived radioactivity reflect markedly increased exocytotic release of norepinephrine [28]. Treatment of heart failure reduces the sympathetic hyperactivity [99] and can therefore explain the improvement in H/M ratios. Locally high norepinephrine concentrations at neuronal uptake sites could also compete with ¹²³I-MIBG for uptake into sympathetic nerves, giving the appearance of decreased NET activity.

Takotsubo cardiopathy

Takotsubo cardiopathy is an acute, reversible form of heart failure brought on by emotional distress or other factors that cause acute release of catecholamines [4, 133]. A takotsubo is a Japanese ceramic pot for catching octopi, and patients with takotsubo cardiopathy have a lack of contraction or even a ballooning out of the left ventricular apex and normal basal contraction during systole, giving the heart the appearance of a takotsubo. Mid-ventricular and even reverse patterns of myocardial akinesia have also been described. The condition is especially common in post-menopausal women.

Many studies have agreed on low H/M ratios of ¹²³I-MIBG-derived radioactivity in takotsubo cardiopathy [5, 14, 54, 126]. There are probably several determinants of the abnormality. First, plasma levels of norepinephrine and especially epinephrine are markedly increased in takotsubo cardiopathy [6, 130], and high circulating catecholamine levels can compete with cardiac sympathetic neuroimaging agents at neuronal uptake sites [29]. Second, all forms of heart failure are associated with increased sympathetic noradrenergic outflow to the heart [28, 80, 100], and high local interstitial fluid concentrations of norepinephrine released from cardiac sympathetic nerves could compete with ¹²³I-MIBG for the NET. Third, because of the elevated rate of exocytotic release, ¹²³I-MIBG taken up into intra-neuronal vesicles can be released at an increased rate; this would be especially manifested by increased “washout” of ¹²³I-MIBG-derived radioactivity.

After takotsubo cardiopathy, decreased ¹²³I-MIBG-derived radioactivity can persist even after return of cardiac function to normal [16, 102].

Pheochromocytoma

Pheochromocytomas are rare tumors of catecholamine-synthesizing cells, usually in the adrenal gland, although they can be located elsewhere, and are usually non-malignant. Pheochromocytomas are important clinically because they constitute a potentially surgically curable form of hypertension and because, if left undetected, they can produce sudden, unexpected hypertensive paroxysms. Since the tumors usually express the NET, ¹²³I-MIBG scanning can visualize them [71, 105], and ¹²³I-MIBG scanning is approved in the United States for the diagnostic evaluation and localization of pheochromocytoma.

In this review pheochromocytoma is mentioned because of the potential effects on cardiac sympathetic noradrenergic functions. High circulating catecholamine levels in patients with pheochromocytoma can result in dilated cardiomyopathy that is reversible upon removal of the tumor [52]. This effect is reminiscent of the situation in takotsubo cardiopathy noted above, and indeed takotsubo cardiopathy can result from pheochromocytoma [31, 67].

Probably because of competition between endogenous catecholamines and ¹²³I-MIBG for neuronal uptake sites in pheochromocytoma, patients with pheochromocytoma can have decreased H/M ratios of ¹²³I-MIBG-derived radioactivity, with reversal of the abnormality after tumor resection [56].

Diabetic Autonomic Neuropathy

The presence of cardiovascular autonomic neuropathy in diabetes mellitus has been associated with about a doubled risk of silent myocardial ischemia and mortality [74]. Diagnosis usually rests on a battery of tests, although approaches vary by center [127]. Cardiac sympathetic neuroimaging studies have found decreased ¹²³I-MIBG-derived radioactivity in diabetic patients compared to controls, even in the absence of other signs and symptoms of autonomic neuropathy [21, 34, 73]. Further studies are needed to determine the clinical utility of ¹²³I-MIBG scanning in the evaluation of diabetic autonomic neuropathy, especially since coronary microvascular lesions could decrease ¹²³I-MIBG-derived radioactivity.

Conclusions

Cardiac sympathetic neuroimaging by ¹²³I-MIBG scanning is currently approved in the United States to help assign prognosis in patients with heart failure. ¹²³I-MIBG scanning focusing on the adrenal glands is also approved in the diagnostic evaluation of pheochromocytoma. Based on extensive and long-standing experience with ¹²³I-MIBG scanning in Europe and the Far East and on compelling literature, we propose that indications for cardiac sympathetic neuroimaging by ¹²³I-MIBG scanning should be expanded to the diagnostic evaluation of nOH in the setting of parkinsonism to diagnose PD+OH differentially from MSA, and of cognitive dysfunction to diagnose DLB differentially from other forms of dementia. ¹⁸F-Dopamine PET scanning for cardiac sympathetic neuroimaging offers advantages over ¹²³I-MIBG SPECT scanning but is available only at the NIH Clinical Center.

Abbreviations:

AAD

autoimmunity-associated autonomic failure with sympathetic denervation

AAG

autoimmune autonomic ganglionopathy

ANS

autonomic nervous system

COI

chronic orthostatic intolerance

DLB

dementia with Lewy bodies

¹⁸F-DA

¹⁸F-dopamine

H/M ratio

heart/mediastinum ratio

¹²³I-MIBG

¹²³I-metaiodobenzylguanidine

MSA

multiple system atrophy

MSA-P

parkinsonian form of multiple system atrophy

NET

cell membrane norepinephrine transporter

nOH

neurogenic orthostatic hypotension

PAF

pure autonomic failure

PET

positron emission tomography

PD+OH

Parkinson disease with orthostatic hypotension

POTS

postural tachycardia syndrome

SNS

sympathetic noradrenergic system

SPECT

single photon emission computed tomography

VMAT

vesicular monoamine transporter