Diagnosing Premotor Multiple System Atrophy

Abstract

Background and Objectives

Nonmotor features precede motor symptoms in many patients with multiple system atrophy (MSA). However, little is known about differences between the natural history, progression, and prognostic factors for survival in patients with MSA with nonmotor vs motor presentations. We aimed to compare initial symptoms, disease progression, and clinical features at final evaluation and investigate differences in survival and natural history between patients with MSA with motor and nonmotor presentations.

Methods

Medical records of autopsy-confirmed MSA cases at Queen Square Brain Bank who underwent both clinical examination and cardiovascular autonomic testing were identified. Clinical features, age at onset, sex, time from onset to diagnosis, disease duration, autonomic function tests, and plasma noradrenaline levels were evaluated.

Results

Forty-seven patients with autopsy-confirmed MSA (age 60 ± 8 years; 28 men) were identified. Time from symptom onset to first autonomic evaluation was 4 ± 2 years, and the disease duration was 7.7 ± 2.2 years. Fifteen (32%) patients presented with nonmotor features including genitourinary dysfunction, orthostatic hypotension, or REM sleep behavior disorder before developing motor involvement (median delay 1–6 years). A third (5/15) were initially diagnosed with pure autonomic failure (PAF) before evolving into MSA. All these patients had normal supine plasma noradrenaline levels (332.0 ± 120.3 pg/mL) with no rise on head-up tilt (0.1 ± 0.3 pg/mL). Patients with MSA with early cardiovascular autonomic dysfunction (within 3 years of symptom onset) had shorter survival compared with those with later onset of cardiovascular autonomic impairment (6.8 years [5.6–7.9] vs 8.5 years [7.9–9.2]; p = 0.026). Patients with early urinary catheterization had shorter survival than those requiring catheterization later (6.2 years [4.6–7.8] vs 8.5 years [7.6–9.4]; p = 0.02). The survival of patients with MSA presenting with motor and nonmotor symptoms did not differ (p > 0.05).

Discussion

Almost one-third of patients with MSA presented with nonmotor features, which could predate motor symptoms by up to 6 years. Cardiovascular autonomic failure and early urinary catheterization were predictors of poorer outcomes. A normal supine plasma noradrenaline level in patients presenting with PAF phenotype is a possible autonomic biomarker indicating later conversion to MSA.

Multiple system atrophy (MSA) is a neurodegenerative disorder characterized by autonomic failure, parkinsonism, and cerebellar signs. Nonmotor features, including cardiovascular, respiratory, urogenital, gastrointestinal (GI), sudomotor dysfunction, and REM sleep behavior disorder (RBD), can be presenting symptoms and often precede the motor symptoms in MSA.¹ Patients with MSA presenting with nonmotor features are often misdiagnosed with other conditions. A previous study has shown that more than 40% of patients with MSA who presented with urogenital dysfunction were misdiagnosed with prostatic hypertrophy or bladder dysfunction and underwent urologic surgery with poor outcome,² and patients with MSA initially presenting with orthostatic hypotension (OH) were usually diagnosed with pure autonomic failure (PAF) before the development of other hallmark MSA features.³

PAF is a sporadic α-synucleinopathy disorder characterized by autonomic failure without other neurologic symptoms and signs.⁴ A recent natural history study showed that up to 8% of patients with PAF presenting with isolated cardiovascular autonomic failure eventually evolve into MSA within 4 years after symptom onset.³ Another, retrospective, study demonstrated that approximately 12% of patients with PAF evolve into other neurodegenerative disorders, and more than half (59%) were diagnosed with MSA within the first 3 years after the initial diagnosis of PAF.⁵

Cardiovascular autonomic dysfunction is an integral part of MSA diagnostic criteria⁶ and is also likely to be a key factor influencing survival in MSA.⁷ Two recent autopsy-confirmed studies have identified unfavorable prognostic factors in MSA^7,8 including early development of generalized cardiovascular autonomic dysfunction,⁹ urinary catheterization,⁷ severe autonomic dysfunction,^7,10 later age at onset,¹¹ parkinsonian subtype of MSA,^12,13 and stridor.¹⁴ However, the survival rate and disease progression have not been previously investigated in autopsy-proven cases presenting with motor vs nonmotor features of MSA.

Therefore, the aim of this study was (1) to characterize the initial symptoms, disease progression, and clinical features at last evaluation in a large UK cohort of autopsy-confirmed MSA and (2) to investigate whether there are any differences in survival, clinical features, and natural history between patients with MSA with motor and nonmotor presentations.

Methods

We retrospectively evaluated autopsy-confirmed MSA cases at the Queen Square Brain Bank, who underwent full autonomic function testing at the Autonomic Unit, National Hospital for Neurology and Neurosurgery (NHHN), Queen Square between 1992 and 2012.

All were reviewed by movement disorder and autonomic specialists. Exclusions included concomitant diseases potentially affecting the autonomic nervous system (such as diabetes mellitus, heart diseases, cancer, and hypothyroidism).

Clinical History, Disease Onset, and Clinical Presentation

Medical records of patients with MSA were systematically reviewed. History and relevant information of each patient was evaluated, including age at evaluation, sex, age at onset, presenting symptoms, time from presenting symptoms to date of autonomic testing, dopaminergic medications, disease duration (time from symptom onset to death), and clinical diagnosis at first and last evaluation at NHNN.

Patients were divided into nonmotor onset and motor onset, based on the first presenting symptoms according to the information recorded by the treating clinician on the medical notes. Nonmotor presenting features included any of the following: cardiovascular autonomic symptoms (e.g., dizziness, light-headedness, visual disturbances, syncope, and coat-hanger pain),¹⁵ bladder dysfunction (urinary incontinence, frequency, urinary retention, and urinary catheterization), erectile dysfunction, GI dysfunction, swallowing difficulties, stridor and RBDs.

Neurogenic bladder was defined as urinary incontinence or retention not attributable to urologic or gynecologic pathology. Motor presenting features comprised parkinsonism and/or cerebellar symptoms.

Cardiovascular Autonomic Testing

Cardiovascular autonomic function tests (AFTs) were performed using Autonomic Unit, Queen Square autonomic protocols.¹⁶ AFT data including blood pressure (BP) and heart rate (HR) responses to head-up tilt (HUT) test, Valsalva maneuver (VM), and HR response to deep breathing were collected. HUT was performed at 60° for 10 minutes. OH was defined as a drop of >30 mm Hg in systolic BP (SBP) or >15 mm Hg in diastolic BP (DBP) on HUT according to the MSA second consensus criteria.⁶ Deep breathing was performed at a rate of 6 breaths per minute for 1 minute, and VM was performed by blowing against a fixed resistance of 40 mm Hg for 15 seconds in the supine position. Supine and tilted plasma noradrenaline (NA) levels were included when available. To define the degree of cardiovascular autonomic dysfunction, we used the following criteria below:

1. Isolated sympathetic dysfunction (OH and abnormal BP [ABP] responses during VM).

2. Isolated parasympathetic dysfunction (abnormal Valsalva ratio and HR responses to deep breathing)

3. Generalized cardiovascular autonomic dysfunction (both sympathetic and parasympathetic autonomic dysfunction; OH with ABP responses during VM, abnormal Valsalva ratio, and HR responses to deep breathing)

4. No cardiovascular autonomic dysfunction

Statistical Methods

Continuous data are presented as mean (±1 SD) and categorical data as number (percentage). Mann-Whitney U tests were used to compare continuous variables between the 2 groups, for example, patients presenting with nonmotor features vs motor (parkinsonian or cerebellar features) for nonnormally distributed data and unpaired t tests for normally distributed data. Chi-square analyses were used for analysis of categorical variables.

Disease duration (time from first symptom to death) was considered as a dependent variable. Other relevant variables, including cardiovascular autonomic measures (such as BP falls during HUT and HR responses to deep breathing), age at onset, sex, MSA subtype (cerebellar subtype; MSA-C and parkinsonian subtype; amd MSA-P), the presence of stridor, respiratory involvement, swallowing difficulties, supine hypertension, early cardiovascular autonomic dysfunction, early neurogenic bladder (≤3 years after onset), urinary catheterization (either indwelling or intermittent), and plasma NA levels, were included as determinants in the model.

Multiple linear regressions were performed to assess the relationship between explanatory variables and disease duration. Kaplan-Meier curves were used to graphically estimate the median time to death, and the log-rank test was used to compare the median time to death among the prognostic factors. Statistical analyses were performed using STATA 11.0 (Stata Corp., College station, TX). All tests were 2 sided, and a p value of ≤0.05 was considered significant.

Standard Protocol Approvals, Registrations, and Patient Consents

The brain donor program and protocols were approved by the NRES Committee London–Central (18/LO/0721), and tissue was stored for research under a license issued by the Human Tissue Authority (No. 12198). Written informed consent was obtained from all donors.

Data Availability

Because of the sensitive nature of the data collected for this study, anonymized data pertaining to the research presented will be made available on reasonable request from external qualified investigators.

Results

Patient Characteristics

Forty-seven patients (male:female, 28:19) with autopsy-confirmed MSA were included in the final analysis. Their mean age at onset was 56 ± 8 years, and their mean disease duration (time from onset to death) was 7.9 ± 2.2 years.

The average time from symptom onset to the first evaluation at the NHNN was 4.2 ± 1.8 years. The final clinical diagnoses included MSA-C (n = 24), MSA-P (n = 21) and progressive supranuclear palsy (PSP, n = 2). There were no significant differences in age, disease duration, and sex between patients with MSA-C and MSA-P.

Patients With MSA With Motor Onset

Thirty-two (68%) patients presented with motor symptoms including parkinsonism or cerebellar features at onset (upper panel, Figure 1). Parkinsonism was more common than cerebellar features among patients with MSA with motor onset. Patients with motor onset reported urinary symptoms within a median of 3 years (range 2–4 years) and cardiovascular autonomic dysfunction within 4 years (range 3–6 years) after symptom onset, respectively (Table 1 and Figure 1). Bladder dysfunction eventually developed at some point during disease progression and was reported in all patients. These patients with MSA presented with motor symptoms including tremor, and movement and balance problems, and were initially diagnosed with Parkinson disease (PD), MSA, atypical parkinsonian disorders, or idiopathic late-onset cerebellar ataxia (ILOCA) at NHNN. These patients subsequently underwent cardiovascular autonomic testing before the diagnosis was subsequently refined to MSA in 96% (45/47) (eFigure 1, links.lww.com/WNL/C160).

Figure 1. Clinical Features From Onset to Last Clinical Evaluation in Patients With MSA With Motor (A) and Nonmotor Onset (B).

*Median, range in years, **p < 0.05. MSA = multiple system atrophy; RBD = REM sleep behavior disorder.

Table 1.

Demographic Data and Clinical Features in Patients With MSA With Motor and Nonmotor Onset

Two patients were diagnosed clinically with PSP with predominant parkinsonism (PSP-P). Both patients presented with asymmetrical parkinsonism in their early 60s and were initially diagnosed with Parkinson disease. Limitation of vertical gaze and slow vertical saccadic eye movements were documented in both patients. Urinary incontinence was reported within the first 2 years in both cases, followed by urinary retention. Urinary catheterization was required at 4–5 years after the first symptom onset. Cardiovascular autonomic testing was performed at 2–4 years after the first symptom onset. There was isolated cardiovascular parasympathetic dysfunction in both cases. The other patients' demographic data and features are presented in Table 1.

Patients With MSA With Nonmotor Onset

Fifteen of 47 (32%) patients initially presented with nonmotor symptoms that preceded motor involvement. Erectile dysfunction and bladder dysfunction were the most 2 common initial symptoms (47% and 40%, respectively) among patients with nonmotor onset.

Symptomatic OH (postural light-headedness and syncope) and RBD occurred as an initial presentation in 1 patient each. Five of 6 (83%) female patients had bladder dysfunction as a presenting feature. The median time to subsequent development of motor features (either parkinsonism or cerebellar features) was 3 years (range 1–5 years) (Figure 1).

Of the patients with nonmotor onset (n = 15), 5 presented with cardiovascular autonomic failure, erectile and bladder dysfunction with no motor features, and were diagnosed with PAF during the first assessment at NHNN (eTable 1, links.lww.com/WNL/C160; patient case numbers 3, 6, 9, 11, and 14). Three patients were initially given a clinical diagnosis of ILOCA and 2 with PD before their diagnosis was changed to MSA. The remaining 5 patients developed other features including poor levodopa response parkinsonism or cerebellar signs and were subsequently diagnosed with MSA at first evaluation at NHNN.

Motor vs Nonmotor Onset: Differences in Autonomic and Clinical Features

There were no differences in age at onset (p = 0.88) and disease duration (p = 0.67) between patients with motor and nonmotor onset. Patients with nonmotor onset had a significant delay in clinical diagnosis compared with those with motor onset (4.8 ± 2.0 years vs 3.5 ± 1.8 years; p = 0.03) (Table 1).

At last evaluation, there were no differences in reported symptoms including bladder dysfunction, requirement for urinary catheterization, stridor, swallowing difficulty, and constipation (all p > 0.05) between patients with motor and nonmotor onset. OH was present in all patients with nonmotor onset and 75% (24/32) of patients with motor onset (p = 0.03). In contrast, RBD was more commonly reported in patients with motor-onset MSA in comparison to those with nonmotor onset (odds ratio 4.3; 95% CI 1.00–18.36, p = 0.05) (Figure 1).

Cardiovascular Autonomic Testing

All patients underwent comprehensive cardiovascular autonomic testing, which showed (1) generalized autonomic failure (both sympathetic and parasympathetic autonomic dysfunction) in 20 (43%) patients; (2) isolated sympathetic dysfunction (OH and ABP responses during VM) in 19 (40%) patients; (3) isolated parasympathetic dysfunction in 5 (11%); and (4) no cardiovascular autonomic dysfunction at the first evaluation in 3 (6%) patients (Table 2).

Table 2.

Cardiovascular Autonomic Measures in MSA According to the Severity of CV Autonomic Dysfunction at First Evaluation

Patients with MSA with isolated sympathetic impairment were significantly younger and had an earlier onset than those with generalized cardiovascular autonomic dysfunction (p < 0.05; Table 2). There were no differences in supine SBP between patients with MSA with isolated sympathetic impairment compared with those with generalized cardiovascular autonomic dysfunction.

OH was documented in 83% (39/47) on HUT. Supine hypertension, defined as supine SBP ≥150 mm Hg or DBP ≥90 mm Hg, was present in 21% (10/47 patients) at first evaluation. The degree of OH (∆supine − HUT BP) was not significantly different in patients with MSA with isolated sympathetic impairment compared with those with both sympathetic and parasympathetic impairment (p > 0.05).

Twenty of 47 patients (43%) had plasma NA measurements both on supine and HUT. Mean supine plasma NA was 345.0 ± 145.6 pg/mL with a minimal rise on HUT (increase of 34.7 ± 51.0 pg/mL). There were no differences in supine plasma NA levels between patients with motor and nonmotor presenting MSA (motor vs nonmotor; 350 ± 156 pg/mL vs 332 ± 120 pg/mL, p = 0.8). All patients with MSA with initial PAF phenotype had normal supine plasma NA levels with no rise on HUT (mean supine plasma NA 357.6 ± 165.4 with a change of −11.8 ± 27.8 pg/mL on HUT; normal supine plasma NA 200–500 pg/mL).

Prognostic Factors for Survival

There was no difference in survival between MSA presenting with motor and those with nonmotor symptoms (p > 0.05; Table 1). Patients with MSA with early cardiovascular autonomic dysfunction (within 3 years of symptom onset) had shorter disease survival compared with those with later onset of cardiovascular autonomic impairment (6.8 years [5.6–7.9] vs 8.5 years [7.9–9.2]; p = 0.026) (Figure 2).

Figure 2. Kaplan-Meier Curves for Survival Outcome.

(A) Symptom onset to death comparing between patients who developed early cardiovascular autonomic dysfunction (within 3 years of disease onset; p = 0.026); (B) symptom onset to death comparing between patients who developed early urinary catheterization (within 3 years of disease onset; p = 0.019); and (C) symptom onset to death comparing between patients who developed early bladder dysfunction (within 3 years of disease onset; p = 0.67).

Thirty-two of 47 (68%) reported urinary symptoms (urinary frequency, incontinence, or retention), and 10/30 (33%) had severe bladder dysfunction requiring indwelling or intermittent catheterization within 3 years of symptom onset. Patients with MSA with early urinary catheterization (within 3 years of symptom onset) had shorter survival in comparison to patients who had catheterization later in disease progression (6.2 years [4.6–7.8] vs 8.5 years [7.6–9.4]; p = 0.019) (Figure 2).

Patients with MSA with older age at onset ≥51 years had shorter disease duration compared with those with a younger age at onset (7.3 years [6.5–8.0] vs 9.1 years [8.2–9.9]; p = 0.018). The later age of onset was well correlated with shorter disease duration in patients with generalized cardiovascular autonomic failure (sympathetic and parasympathetic cardiovascular autonomic dysfunction) (r = −0.71, p < 0.01), but not in other groups with less severe autonomic dysfunction (all p > 0.05) (Figure 3).

Figure 3. Scatter Plot Showing the Correlation Between Time From Onset to Death (Disease Duration: Y Axis) and Age at Onset (X Axis) in Patients With MSA With Generalized Cardiovascular Autonomic Dysfunction (Both Sympathetic and Parasympathetic Impairment) on Autonomic Testing.

MSA = multiple system atrophy.

Among the minority of patients with MSA with no cardiovascular autonomic dysfunction (n = 3) at first evaluation, the disease duration ranged from 9 to 10 years, which appeared to be longer than those with cardiovascular autonomic dysfunction (Table 2). However, statistical analysis was not performed due to limited subgroup sample size. Balance problems and parkinsonism were presenting features in these patients, and the diagnosis of MSA was made at a time when patients presented with a combination of cerebellar features/parkinsonism, erectile dysfunction, severe bladder dysfunction leading to urinary catheterization, or other features according to the MSA consensus criteria (stridor, sleep apnea, or severe dysphagia). These patients had no cardiovascular dysfunction on repeat testing up to even 6 years after the first autonomic evaluation.

There was no association between MSA subtype, the presence of stridor, respiratory involvement, swallowing difficulties, supine hypertension, plasma NA levels, and early neurogenic bladder (within 3 years after disease onset), excluding severe neurogenic bladder needing catheterization (all p > 0.05). There was also no association between nonmotor presenting features, including erectile dysfunction, RBD, bladder dysfunction, or OH, and disease duration (p > 0.05).

Discussion

The main finding of this study is that almost a third of our patients with MSA presented with nonmotor features at the onset. Among these patients, all but one who reported RBD as their initial symptom presented with autonomic features, including erectile failure, bladder dysfunction, and/or OH. These findings are consistent with a recent study looking at initial symptoms in non–autopsy-confirmed MSA.¹⁷ The patients developed other features including poorly levodopa responsive parkinsonism, cerebellar signs, OH, or severe neurogenic bladder dysfunction during disease progression, and subsequently, their diagnoses were changed to MSA.

In our cohort, patients with MSA with motor onset usually developed bladder and cardiovascular autonomic dysfunction within 3 and 4 years, respectively. The mean interval from the onset of bladder symptoms to motor features in our study is in concordance with a large prospective study in patients with MSA who initially presented with bladder dysfunction.¹⁸

Two patients were initially diagnosed with PD and 3 with ILOCA before being rediagnosed with MSA. A previous study has demonstrated that approximately a quarter of patients with a diagnosis of olivopontocerebellar atrophy will evolve to MSA within 5 years.¹⁹

Our findings demonstrated that nonmotor features can predate motor symptoms by up to 6 years. These figures are comparable to previous studies^1,20,21 and emphasize the importance of regular monitoring for emerging features in these patients. We found that the final diagnosis of MSA was delayed by over a year in nonmotor onset, compared with patients presenting with motor onset.

RBD was more frequently reported at final evaluation in patients with motor than nonmotor onset MSA. It is a well-known feature that can predate other neurologic symptoms in patients with synucleinopathies.²² Its higher occurrence in patients with motor-onset MSA may reflect the selective vulnerability and differential involvement of key networks between these subgroups.²³ The anatomic substrate and pathophysiology of RBD is not completely understood. Several anatomic structures including dorsal midbrain, perilocus coeruleus, sublaterodorsal nucleus as well as basal ganglia, substantia nigra, and frontal cortex, have been proposed to play a crucial role in RBD.^24,25 Of interest, imaging studies have linked idiopathic RBD with locus coeruleus-noradrenergic dysfunction.²⁶ Quantitative pathologic study in this area could shed some light on the differences in progression of autonomic pathology between these subgroups of MSA.

In contrast, OH was more common in patients with MSA with nonmotor onset than those with motor onset. Previous studies have reported a correlation between the loss of preganglionic neurons in the intermediolateral column and the degree of OH,²⁷ and the depletion of C1 catecholaminergic neurons in the rostral ventrolateral medulla is thought to be partly responsible for the pathogenesis of OH in MSA.²⁸ Certain areas of the central autonomic network may be more vulnerable in patients with nonmotor onset MSA.

These differences warrant further pathologic studies on the distinct initial loci of α-synuclein pathology deposition influencing the evolution of clinical features and disease progression between patients with MSA with motor and nonmotor onset. A recent study has reported a subgroup of patients with MSA who presented with isolated urinary retention requiring urinary catheterization, sexual and bowel dysfunction, and abnormal anal sphincter EMG even before developing motor signs.²⁹ These findings raise the possibility that the pathologic substrate for nonmotor onset patients may start from the spinal cord before involving other regions of the CNS.²⁹

It is well recognized that a proportion of patients with MSA may initially present with a PAF phenotype before motor features emerge. Our study has demonstrated that at least 10% of patients with MSA start with the PAF phenotype before evolving into classical MSA features, with an average conversion time of 3 years. This is consistent with a recent study looking at the predictors of conversion from PAF to either MSA or PD/dementia with Lewy bodies.⁵ Previous case reports have shown that autonomic failure can predate motor features for more than a decade in a subgroup of patients with MSA.^8,30 Our findings revealed that patients with MSA who presented with PAF features had normal supine plasma noradrenaline levels (200–500 pg/mL) with no significant rise on HUT, which is in line with a prospective cohort study on the natural history of PAF.³ A normal supine plasma noradrenaline level in patients with PAF may indicate the likelihood of MSA conversion. This raises the possibility that autonomic dysfunction and RBD is a prodomal phase of MSA.³¹

Two patients with a clinical diagnosis of PSP at final evaluation were found to have MSA at postmortem. They presented with asymmetrical parkinsonism with poor responses to levodopa, bladder symptoms, limitation of vertical gaze with slow vertical saccadic eye movements, and isolated parasympathetic dysfunction on cardiovascular autonomic testing. Recent studies have also described some patients with PSP-like presentation, who were subsequently found to have MSA at postmortem,⁸ and almost 6% of patients with autopsy-confirmed PSP were misdiagnosed with MSA in life.³²

Mild autonomic impairment in PSP is thought to be linked with tau deposition in selected brainstem nuclei involved in the regulation of cardiovascular function and micturition.³³ Onuf's nucleus was reported to be involved in autopsy-confirmed PSP cases.³⁴ Pathologic involvement of key autonomic structures in both MSA and PSP may give rise to a substantial overlap in the clinical presentation of some patients with MSA-P and PSP-P. However, the presence of severe cardiovascular autonomic and bladder dysfunction (indicated by time from onset to urinary catheterization) and early autonomic dysfunction are more suggestive of MSA, as opposed to PSP.³⁵

Our study found no difference in survival between patients with MSA presenting with motor and nonmotor features. Nevertheless, selected characteristics and nonmotor features including early urinary catheteriztion (within 3 years of disease onset) and later age at onset (over 51 years) were associated with poor survival in patients with MSA. These findings are consistent with both previous autopsy-confirmed^7,11 and nonautopsy MSA studies.^9,20,30 Our study confirmed that patients with MSA with cardiovascular autonomic dysfunction had shorter survival than those without, as previously reported in another autopsy-confirmed study.⁷

Cardiovascular autonomic testing plays an important role in diagnosing MSA. Forty-five of 47 (96%) fulfilled Gilman diagnostic criteria for probable MSA after the autonomic testing at NHNN. Some patients with MSA may only have isolated parasympathetic dysfunction (i.e., abnormal HR responses to deep breathing without OH on HUT) on cardiovascular autonomic testing. We propose that formal cardiovascular autonomic assessments should be performed in all patients with suspected MSA, as isolated parasympathetic deficits are unlikely to be detected by bedside cardiovascular autonomic testing. In our series, approximately 6% of patients with MSA had normal cardiovascular autonomic function on testing up to 6 years after symptom onset, suggesting that although cardiovascular autonomic dysfunction is a key feature in patients with MSA, normal cardiovascular function does not necessarily exclude a diagnosis of MSA.

There are some limitations of our study. First, the clinical history and data obtained retrospectively from the medical notes with lack of standardized autonomic questionnaires may result in an incomplete record of certain clinical features, such as RBD, olfactory function, GI symptoms (i.e., constipation) and respiratory sighs/stridor in some patients, or these symptoms being underreported. RBD was less frequently reported in the clinical history before knowledge of its importance in synucleinopathies, and the frequency of RBD in our patients is also based on clinically suspected but not polysomnography-confirmed cases. Furthermore, we have no quantitative sudomotor function data in these patients, as this was not routinely performed in our department.

Moreover, we have limited information regarding the actual causes of death, including sudden unexpected death or unrelated concurrent medical conditions. These are often reported in patients with MSA³⁶ and may have influenced the disease duration in our patients. Finally, the patients in our cohort were all referred to a tertiary neurosciences unit and may represent more severe cases of MSA leading to poorer survival rate. The strength of our study is a large cohort of neuropathologically confirmed cases. Furthermore, these patients were referred from local neurologists, neurogeneticists (ataxia clinic), movement disorder, and autonomic specialists at NHNN. Also, in all our patients, comprehensive cardiovascular autonomic testing and detailed autonomic symptom profile were recorded during their disease course by specialists and clinical autonomic scientists.

In summary, nonmotor features, particularly genitourinary and cardiovascular autonomic dysfunction, are common presenting symptoms in MSA. Almost a third of patients with MSA in our cohort presented with nonmotor features before developing motor involvement within 6 years. Patients with MSA with later age at onset and early impairment of cardiovascular autonomic function or bladder dysfunction requiring urinary catheterization have shorter survival. A proportion of patients with MSA start with PAF phenotype before developing motor features. Normal supine plasma noradrenaline levels in patients presenting with PAF are a red flag that may raise the possibility of subsequent conversion to MSA. The discrepancy in frequencies between RBD and OH in motor-onset and nonmotor onset MSA may indicate impairment of different brain areas. Understanding these differences may shed light on the selective vulnerability of brain and spinal cord nuclei in subtypes of MSA and aid potential means for intervention or treatment at an earlier and thus modifiable stage of the disease.

Glossary

ABP

abnormal BP

BP

blood pressure

AFT

autonomic function test

DBP

diastolic BP

GI

gastrointestinal

HR

heart rate

HUT

head-up tilt

ILOCA

idiopathic late-onset cerebellar ataxia

MSA

multiple system atrophy

MSA-C

MSA cerebellar subtype

MSA-P

MSA parkinsonian subtype

NHHN

National Hospital for Neurology and Neurosurgery

OH

orthostatic hypotension

PAF

pure autonomic failure

PD

Parkinson disease

PSP

progressive supranuclear palsy

PSP-P

PSP with predominant parkinsonism

RBD

REM sleep behavior disorder

SBP

systolic BP

VM

Valsalva maneuver

Appendix. Authors

Study Funding

V. Iodice was supported by the National Institute for Health Research University College London Hospitals Biomedical Research Centre. J.N. Panicker undertook this work at UCLH/UCL Institute of Neurology and is supported in part by funding from the United Kingdom Department of Health NIHR Biomedical Research Centres funding scheme. The Queen Square Brain Bank for Neurological Disorders receives support from the Reta Lila Weston Institute of Neurological Studies and the Medical Research Council. T.T.Warner is supported by the Reta Lila Weston Trust and the MRC (N013255/1). J.L.Holton is supported by the Multiple System Atrophy Trust; the Multiple System Atrophy Coalition; Fund Sophia, managed by the King Baudouin Foundation and Karin & Sten Mortstedt CBD Solutions. V.Chelban is supported by the Association of British Neurologists/MSA Trust Clinical Research Training fellowship (F84 ABN 540868). S. Koay was supported by the Guarantors of Brain Entry Fellowship.

Disclosure

The authors report no relevant disclosures. Go to Neurology.org/N for full disclosures.