Presynaptic dopaminergic terminal imaging and non-motor symptoms assessment of Parkinson’s disease: evidence for dopaminergic basis?

Introduction

Contrary to previous perceptions, Parkinson’s disease (PD) is recognised as a multisystem disorder. Besides dopamine (DA), three further key neurotransmitters have been described to be involved in the pathogenesis of PD; namely noradrenaline (NA), acetylcholine (ACh), and serotonin (5HT).^1,2 Consequentially, non-motor symptoms (NMS) in PD can potentially be related to dopaminergic, non-dopaminergic pathogenesis or a combination of both.^1,3 Individual studies indicate that apathy,⁴ anxiety⁵ as well as aspects of sleep disturbances⁶ appear to be linked to striatal dopaminergic deficiency as measured by dopamine transporters (DaT) scans. However, NMS such as depression,⁷ fatigue,⁸ weight changes,⁹ and visual hallucinations (VH)¹⁰ may be driven by deficiency in non-dopaminergic transmitters.

The NMS Scale (NMSS) was validated as the first comprehensive and holistic health-professional completed measure of NMS in PD and has now been used as a primary or secondary outcome measure in a number of clinical trials and epidemiological studies.¹¹ The NMSS allows for calculation and grading of the burden (severity multiplied by frequency) of 30 different NMS, which are covered in nine different domains.^12,13 These are cardiovascular, sleep/fatigue, mood/cognition, perceptual problems/hallucinations, attention/memory, the gastrointestinal tract, urinary system, sexual function, and miscellaneous containing olfactory dysfunction.

In this review we primarily address the relationship of dopaminergic radioligands and the individual NMS covered by NMSS domains to examine a possible underlying dopaminergic basis of these varying NMS (Fig. 1).

To conduct this review, we gathered articles, which used dopaminergic imaging to explore the pathophysiology of different NMS, Fig. 2 summarises our methodology. We had three possible terms; Term A had an asterisk allowing for several terms with the same beginning being considered. Term B covered neuroimaging words, whilst Term C were the possible NMS that could have been used. We initially found 8734 articles, which then left us with 42 studies to include once we removed duplicated and referred to our exclusion criteria.

NMSS domain 1: cardiovascular dysfunction

Cardiovascular dysfunction is a key autonomic feature of PD and patients often present with orthostatic hypotension (OH). There are currently no studies reporting cardiovascular dysfunction in PD to have a dopaminergic basis. However, using the ¹²³I-meta-iodobenzyl guanidine (MIBG) radiotracer, a NA analogue, studies have shown there to be a reduction in the postganglionic presynaptic cardiac sympathetic innervation, suggestive of cardiac sympathetic dysfunction early in PD patients, giving rise to symptoms such as OH.^14–19

Several studies have shown reduction in cardiac uptake regardless of using MIBG or ¹⁸F-DOPA cardiac positron emission tomography (PET) in PD patients with OH when compared to those without.^{14–16,20,21} These results suggest that there might be a decrease in catecholamine uptake (that being DA or NA) in PD patients with OH, but not all studies agree.^22,23

NMSS domain 2: sleep disorders and fatigue

Sleep disturbance occurs in 60–90% of all PD patients³⁰ with symptoms ranging from insomnia, sleep apnoea, restless legs syndrome (RLS), rapid eye movement (REM) behaviour disorder (RBD) to excessive daytime somnolence (EDS). It represents one of the most frequent complaints by the patients.³¹ Two aspects of sleep dysfunction, EDS and RBD, have been studied with dopaminergic imaging and are discussed below.

• Excessive daytime somnolence

EDS is the tendency to drift off to sleep quickly and more frequently than usual during the day.³² EDS can be assessed by the Epworth sleepiness scale (ESS).³³ The link between EDS and dopaminergic dysfunction is unclear. Pavese and colleagues conducted a multi-modal PET study, using ¹⁸F-DOPA and ¹¹C-DASB tracers in PD patients with EDS, and reported both dopaminergic and serotonergic dysfunction (Table 1).³⁴ Happe and colleagues had previously proposed dopaminergic dysfunction underlying EDS in PD and additionally, Pavese and colleagues used C-DASB, a serotonin transporter binding ligand, and showed evidence of serotonergic dysfunction in the raphe area of the brain in PD. However, a recent study by Qamhawi and colleagues reported no significant correlation between raphe binding with ¹²³I-FP-CIT-SPECT and EDS.³⁵

• Rapid eye movement sleep Behaviour Disorder (RBD)

REM sleep behaviour disorder (RBD) is characterised by the loss of muscle inhibition during REM sleep, which leads to the physical acting out of violent and dangerous nightmares.^36,37 RBD can entirely be an idiopathic disease (iRBD) or secondary to neurodegenerative conditions such as multiple system atrophy (MSA) or Dementia with Lewy Bodies (DLB). In the field of PD, 60% of PD patients experience RBD³⁸ and 80% of iRBD patients progressing to PD in 10–12 years.³ Hence, RBD is now recognised as the most robust marker of prodromal PD.³⁹ RBD has been shown, at least in part, to be associated with dopaminergic defect, which is consistent with Braak stage 2 pathophysiology.²⁸ Reduction in DaT uptake in iRBD patients has been shown, specifically in the putamen.^40,41 Using ¹²³I-FP-CIT, ¹²³I-IBZM, and ¹¹C-dihydrotetrabenazine (DTBZ) radiotracers, several studies have suggested the nigrostriatal dopaminergic pathway as being implicated in RBD pathogenesis.^42–45 However, Kim and colleagues reported their idiopathic RBD patients had a reduced DaT uptake in the putamen, yet when assessing DaT density in the putamen, the levels remained within normal ranges, leaving them to conclude there may likely be an additional pathogenic pathway implicated in RBD (see Table 1). Studies exploring REM sleep duration, using PET and SPECT imaging, have yielding interesting results whereby the upper brainstem is found to be suppressing REM sleep in early-PD causing uncertainty as to a dopaminergic or non-dopaminergic involvement in the pathophysiology of REM-sleep in PD itself.^46,47 Investigating non-dopaminergic nuclei has led to assumptions of RBD pathophysiology to include the pedunculopontine nucleus and laterodorsal tegmental nuclei (cholinergic nuclei), raphe nucleus (serotoninergic), pre-coeruleus (glutaminergic) and locus coeruleus (noradrenergic).^48,49

• Restless Leg Syndrome (RLS) and Periodic Limb Movements (PLM)

RLS and PLM are common in PD patients.⁵⁰ The precise pathophysiology of both conditions is still unknown and, to our knowledge, there are currently no specific studies investigating RLS-PD pathophysiology using dopaminergic imaging. Nonetheless, studies in idiopathic RLS have suggested a dopaminergic mechanism central to its pathophysiology, which is made evident by the effectiveness of dopaminergic treatment.⁵¹ Studies have shown there to be hypo-dopaminergic activity in idiopathic RLS patients either through reduction in DaT uptake, densities or receptor availability.^52–56 However, others have shown no such change,^57–59 while some have even reported an increase in DaT densities.⁶⁰ Furthermore, it has been hypothesised that compared to PD patients, idiopathic RLS patients may have a mishandling of DA rather than a decrease of dopaminergic cells, as seen in PD.

• Fatigue

Fatigue is a specific NMS in PD with considerable negative impact on the quality of life of patients.⁶¹ Some studies reported fatigue and dopaminergic dysfunction as not being significantly associated when assessed using neuroimaging.^8,62,63 Both, Schifitto and colleagues and Pavese and colleagues found no significant reduction in striatal dopaminergic uptake between fatigued and non-fatigued PD patients (Schifitto et al. 2008, Pavese et al. 2010). However, Pavese also used ¹¹C-DASB PET and reported a significant reduction in serotonergic transporter (SERT) binding particularly in the caudate, putamen, ventral striatum and thalamus (Table 1). Hence, Pavese suggests, not only the involvement of extra-striatal pathways, but also a non-dopaminergic involvement in the form of serotonergic dysfunction underpinning central fatigue.

Supplementary to the notion of a non-dopaminergic involvement, Chou and colleagues recently hypothesised cholinergic dysfunction to also be involved by using ¹¹C-methyl-4-piperidinyl propionate (PMP) acetylcholinesterase (AChE) and ¹¹C-DTBZ monoaminergic PET imaging.⁶³ However, their results found no significant evidence to support this hypothesis. Nonetheless, clinical experience dictates that dopaminergic therapies can be effective in treating fatigue-PD patient groups, which has led many to conclude that dopaminergic dysfunction might have a partial role.^62,63

NMSS domain 3: mood and apathy

Neuropsychiatric problems are a common manifestation in PD⁶⁴ with depression being the most prevalent with up to 45% of patients affected.⁶⁵ Here, we explore the most commonly discussed mood and apathetic problems PD patients face and examine its potential pathophysiology using different radiotracers with a focus on DA.

• Depression

The pathophysiology of depression has been associated with dopaminergic defect by several studies (Table 2) reporting an inverse correlation of depression with dopaminergic availability.^66–70 However, as Braak hypothesis suggests, early lower brainstem pathophysiology may involve several other nuclei and thus other neurotransmitters,²⁸ specifically serotonin.⁷¹ When investigating this association, studies have found there to be an inverse correlation between SERT binding within areas such as the dorsal midbrain, suggestive of a serotonergic dysfunction.^72,73 Politis and colleagues conducted a large in vivo study, using ¹¹C-DASB PET in antidepressant-naïve PD patients. Their results show an altered serotonergic function associated with higher depression levels in these patients, which suggests abnormal serotonergic neurotransmission in PD depression pathophysiology. Further probing into alternative neurotransmitter involvement has found there to also be noradrenergic dysfunction in the locus coeruleus.^69,74

• Anxiety

Anxiety composes a range of disorders, which can be classified into three categories (anxiety disorder, obsessive-compulsive disorder, and trauma and stressor-related disorder).⁷⁵ Being a disorder, which can coexist with depression,⁷⁶ the pathophysiology of anxiety is thought to be dopaminergic in part. Anxiety has also been shown to be a dopaminergic medication-related phenomenon evident in the dominant relationship with non-motor fluctuations in PD.⁷⁷ Studies from Weintraub and colleagues, Erro and colleagues, and others (Table 2) have reported a reduction in dopaminergic uptake in the right striatum of anxious PD patients. However, when assessing different forms of anxiety, studies have not reported the same trend. Moriyama and colleagues, and Kaasinen and colleagues reported a positive correlation between dopaminergic DaT uptake within the striatum and social anxiety or personality traits and anxiety in patients, respectively.^78,79 The variation found here, suggests anxiety to be heterogeneous in origin with a partial dopaminergic basis. Remy and colleagues demonstrated a negative correlation between the severity of anxiety with binding at the locus coeruleus and bilateral amygdala using ¹¹C-RTI32, forging the concept of a DA-noradrenergic system involvement in PD-anxiety.⁶⁹ This is supported by the understanding that both noradrenergic and dopaminergic pathways project from the locus coeruleus to sites including the amygdala and striatum.^80,81

• Apathy

One third of PD patients experience apathy,⁸² characterized by a state of emotion with reduced motivation and a sense of reduced goal-directed behaviour, and ambition.⁸³ Several types of apathy have been described and there is recognition that at least in part, apathy has a dopaminergic origin possibly through the involvement of the mesocorticolimbic circuit.^84–86 Thobois and colleagues used [¹¹C]-raclopride, a DA D2/D3 receptor ligand, and reported that PD-apathy patients had reduced synaptic DA release in the mesocorticolimbic system.⁸⁷ In pure apathy (non-demented and non-depressed) PD patients, there has been a demonstration of reduced dopaminergic uptake in the striatum.^4,69 However recently, Chung and colleagues have demonstrated there to be no association of striatal dopaminergic binding in early PD with apathy.⁸⁸ Unfortunately, currently there is little evidence exploring the association of PD-apathy with dopaminergic dysfunction independent of other neuropsychiatric conditions. However, the reduced caudate uptake is mirrored in other neurodegenerative studies such as Dementia with Lewy bodies,⁸⁹ Alzheimer’s disease,⁹⁰ and frontotemporal dementia.⁹¹ The use of subthalamic nucleus deep brain stimulation (STN-DBS) in PD has itself been identified as being associated with inducing postoperative apathy.⁹² However, evidence exists showing there to be some predisposition to this risk factor in these patients including DA agonist withdrawal syndrome.^93–95 Studies have explored mesolimbic dopaminergic dysfunction in STN-DBS induced apathy in PD patients^54,96,97 finding there to be different mechanisms at play between early and late PD. Therefore, initially the dopaminergic mesocortical system is involved due to the relative sparing of the nigrostriatal dopaminergic system.

• Cognitive Impairment (CI)

CI and dementia have been associated with PD. Around 40% of patients have CI⁹⁸ at early-stage of PD,⁹⁹ and around 80% of patients may experience PD-related dementia (PDD) at a late-stage.^100,101 Mild CI in PD (MCI) is also somewhat prevalent, where a recent review from the Movement Disorders Society (MDS) task force reports a prevalence of 27% (range 19–38%).¹⁰² A dopaminergic basis of cognitive impairment is possible and using ¹⁸F-DOPA PET, studies have demonstrated reduced dopaminergic uptake in PD patients at different stages of their condition with CI or PDD^99,103–109 (Table 3), especially in the caudate nucleus.^110–112 PDD is known to be a late-manifestation in PD.¹¹³ However, conversely some patients have been shown to present with early dopaminergic uptake changes within frontal structures critical to cognitive and executive function⁹⁹; thus cognitive impairment can be an early-manifestation in PD. Three SPECT studies have shown there to be an association between reduced dopaminergic uptake and cognitive impairment in PD and CI as well as PD and MCI,^114–116 however the authors do not use a uniform cognitive function test making their clinical definition of CI vary slightly. Nonetheless, they all report a significant correlation between striatal dopaminergic defect (more commonly unilateral and contralateral to the most affected side) and cognitive impairment existing in these PD patients (Table 3).

The pathophysiology of cognitive impairment in PD may also involve the brainstem and corticostriatal pathway with cholinergic dysfunction.^117–121 Using 2-¹⁸F-FA-85380 PET, studies have shown there to be cholinergic dysfunction of not just the striatum, but also the cerebellum, pons, and thalamus.^122,123 In PDD, studies using N-¹¹C-methyl-4-piperidyl-acetate (¹¹C-PMP) acetylcholinesterase (AChE) PET have reported cholinergic degeneration.^124,125 Bohnen and colleagues used ¹¹C-PMP AChE PET on PDD patients finding a strong correlation of reduced radiotracer uptake with performances on working memory, attentional, and executive function tests suggesting a dominant cholinergic basis to these functions.^126,127

NMSS domain 4: perceptional disorders

Perceptional disorders in PD, ranging from VH to delusions, are particularly prevalent with one-in-four PD patients experiencing VH.^128–132 A dopaminergic basis of VH and other perceptional disorders is being researched. A cortico-striato-thalamocortical dysfunction has been suggested,^89,133,134 however at present, work is required to determine and distinguish the differing forms of hallucinations. Recent data provides evidence of a dopaminergic basis in VH¹³⁵ (Table 4), but not auditory hallucinations.¹³⁶ Several studies also support non-dopaminergic involvement through both serotonergic¹⁰ and cholinergic^{129,130,137–140} means.

Lower DaT binding in the striatum in early PD measured by [¹²³I] [ß]-CIT tracer (DaTscan) is associated with increased prospective risk of psychosis spectrum at 5 years.¹³⁴ It is unclear whether this binding reduction is the underlying mechanism of the psychosis spectrum or, whether an indirect association, for example reflecting more extensive neurodegenerative involvement in the psychosis spectrum, is present. A serotonergic imaging study using the 5HT_2A receptor ligand setoperone-F¹⁸ identified increased binding in patients with VH in ventral occipito-temporal regions and bilateral frontal cortex.¹⁰ In contrast, a 5HT_1A receptor binding study in post-mortem tissue found no association with psychosis spectrum, although 5HT_1A binding was elevated in PD irrespective of hallucination status in sublayers of orbito-frontal, ventral temporal, and motor cortex.¹⁴¹

NMSS domain 5: attention defects

PD patients have been noted to experience various attentional function deficits, including visuospatial¹⁴² and during performance of tasks requiring a switch of behaviour.^143,144 The evidence thus far supports a dopaminergic pathology underlying attention deficits in PD. Work from Rinne and colleagues (Table 5) find correlation between reduced ¹⁸F-DOPA uptake in the caudate and frontal cortex with attentional and working memory deficit. Further evidence supports dopaminergic deficit within the frontal cortex (more specifically the medial portion), alongside the anterior cingulate and the dorsolateral prefrontal cortex.⁹⁹ However, Bohnen and collegues showed a robust correlation with cortical AChE activity with attention and working memory; suggestive of cholinergic involvement.¹²⁴

NMSS domain 6: gastrointestinal tract

Gastric dysfunction in PD is a prevalent issue with symptoms ranging from drooling, dysphagia, and constipation to gastroparesis^145,146 with constipation being suggested as a pre-motor marker in a recent Danish study.¹⁴⁷ There is currently no evidence that supports a dominant dopaminergic pathogenesis for gastric symptoms, whereas cholinergic dysfunction has been suggested by Gjerloff and colleagues who investigated the parasympathetic involvement of AChE binding using ¹¹C-donepezil PET. They report a significant decrease in ¹¹C-donepezil uptake in the small intestines and pancreas, proposing cholinergic dysfunction of the enteric nervous system in PD.¹⁴⁸

NMSS domain 7: urinary dysfunction

Referred to collectively as lower urinary tract symptoms (LUTS), one of the key and most frequent autonomic problems in PD is bladder dysfunction.¹⁴⁹ PD patients experience elevated urinary frequency, urgency, nocturia, incontinence, and voiding.¹⁵⁰ There is evidence for an underlying dopaminergic basis in the pathophysiology of LUTS (Table 6). Dopaminergic influences in the micturition reflex are present in inhibitory pathways arising from dopaminergic Substantia Nigra pars Compacta fibres, whilst the stimulatory affect arises from dopaminergic ventral tegmental area (VTA) fibres.^151,152 Work from Winge and colleagues further supports a dopaminergic dysfunction (Table 6).²⁰¹ However, the micturition reflex and LUTS are controlled by a number of neurotransmitters including DA, serotonin, NA, and ACh.¹⁵³ Moreover, PD-LUTS patients do not necessarily respond to levodopa or dopaminergic treatment but may instead require anticholinergic treatment suggesting a dominant underlying cholinergic basis.^152,154

NMSS domain 8: sexual dysfunction

Sexual dysfunction is a common problem for many PD patients^155,156 and erectile dysfunction (ED), hyper-sexuality, loss of lubrication, loss of libido, and involuntary urination during sex are just some such symptoms.^{149,157–161} There is poor evidence that dopaminergic dysfunction may underlie sexual dysfunction. ED can be caused by both vascular and hormonal, as well as neurological pathologies¹⁶² and using PET and fMRI studies, there has been identification of dopaminergic and serotonergic structures, such as the insula, caudate nucleus, putamen, thalamus, and nucleus accumbens as likely being involved in ED pathogenesis.^163–167 However, sexual issues in PD could also be drug induced manifestations of impulse control disorders (ICD).^168–172 Discussion of functional imaging based studies of ICD is beyond the scope of this review.

NMSS domain 9: miscellaneous

• Olfactory changes

Braak and colleagues proposed the idea that PD-pathology begins in extra-nigral structures, hence why olfactory dysfunction is a common initial prodromal symptom for many PD patients.²⁸ There is some evidence that dopaminergic dysfunction is responsible for olfactory symptoms (Table 7). Using SPECT imaging, several studies now have found there to be supporting evidence of DaT uptake reduction in hyposmic patients.^173–180 Scherfler and colleagues offer concordant evidence that both nigral and olfactory tract degeneration parallels that of putaminal dopaminergic dysfunction in PD patients,²⁰² although not all studies agree.^181–183 In contrast, two studies present evidence that DA agonists are ineffective in treating hyposmic symptoms however, this may be because the damage is simply too excessive.^184,185

• Weight changes

PD patients characteristically undergo diet/metabolism-unassociated weight loss starting early in the course of the disease.^186,187 Based on available evidence, a dopaminergic basis for weight change in PD patients is not unexpected, however only one study by Lee and colleagues explores this (Table 7). DA is involved in modulating the reward and motivational properties of food intake¹⁸⁸ causing problems in weight gain and loss.¹⁸⁹ Weight gain is commonly associated with DA agonist treatment due to the side effects of ICD, specifically compulsive binge eating.^190–193 However, evidence suggests a prominent serotonergic involvement, and weak potential for noradrenergic action¹⁹⁴ as serotonin is thought to play a crucial role in modulation of appetite.¹⁹⁵ In a study using ¹¹C-DASB, a marker for SERT, Politis and colleagues demonstrated increased tracer binding in the rostral raphe nuclei, hypothalamus, caudate nucleus, and ventral striatum in PD patients with abnormal BMI changes.⁹ Interestingly, gain in BMI was associated with raised ¹¹C-DASB binding in the anterior cingulate cortex when compared to those with reduced BMI. These findings imply that decreased levels of serotonin, due to elevated clearance, could lead to abnormal BMI changes. Furthermore, Sharma and colleagues propose introduction of a Park-weight subtype following observation, using standardized olfactory assessments, reporting PD patients with severe olfactory dysfunction correlate with having an increased risk of weight loss.^196,197

Conclusion

To our knowledge, this is the first review that has summarised available evidence exploring the possible dopaminergic basis of NMS pathophysiology using the domains integral to the NMSS, a widely used validated measure of holistic NMS assessment. Hence, it also addresses an unmet need in this regard.

We have found there to be 12 NMS with imaging based evidence for at least in part, a dopaminergic pathophysiological basis (Table 8). The use of radiotracers has certainly evolved and as such tracers such as, ^99mTc-TRODAT-1 SPECT is not regularly used in research now due to its low unreliability and low specificity in comparison to other imaging modalities. Furthermore, we have highlighted key NMS which have non-dopaminergic pathophysiologic involvement (Table 9).

Our review findings are summarised in Table 10, where we have classified evidence for the neurotransmitters involved in the pathophysiology of NMS into four arbitrary categories (strong, moderate, weak, and conflicting evidence). We have catogorised stronger evidence as having open-label trials, or more than 3 publications demonstrating neuroimaging based evidence. Moderate evidence; as having clinical trials, or 2–3 studies presenting neuroimaging based evidence, while weak evidence is defined as having some clinical cases, or 1 study reporting neuroimaging based evidence. Finally, conflicting evidence is defined having 2 or more conflicting studies. This table presents the first summary of NMS in relation to the potential neurotransmitters involved in their pathology. It also shows how some NMS such as insomnia, anhedonia, and delusions have had no or very little exploratory research conducted.

Competing interests

The authors declare no conflict of interests.