Clinical manifestations

Presentation and course

Parkinsonism (rigidity, bradykinesia, and postural instability) occurs in at least 90% of all three types of multiple system atrophy (247). Unlike Parkinson disease, however, multiple system atrophy rarely receives sustained benefit from levodopa and exhibits rest tremor in fewer than 10% of cases (compared with 60% to 70% in Parkinson disease). The progression of parkinsonian motor features of multiple system atrophy, as measured by the motor score (Part III) of the Unified Parkinson’s Disability Rating Scale, averages 28% per year, whereas untreated early Parkinson disease progresses less than 4% per year (197).

Autonomic insufficiency is a more prominent part of most cases of multiple system atrophy than of Parkinson disease and has a marked impact on quality of life (30). Cerebellar features are early and important in many patients and predominate in some. Parkinsonism predominates in others. Because there is considerable pathologic and clinical overlap among the three varieties of multiple system atrophy, there is no need for formal criteria differentiating them. A few patients with multiple system atrophy have isolated dysautonomia, cerebellar dysfunction, or parkinsonism (247).

Besides the classic motor signs, multiple system atrophy is associated with a variety of oculomotor abnormalities, including sustained gaze-evoked nystagmus, square wave jerks, slow and hypometric saccades, diminished vestibular-ocular reflex suppression, internuclear ophthalmoparesis, and reduced vertical gaze (53). Rapid eye movement sleep behavior disorder (RBD) occurs in 69% of patients by history and in 90% of patients by polysomnography (169). The condition occurs earlier and more intensely than in Parkinson disease (88), and its violent, even aggressive motor outbursts can injure the patient or bed partner. Using video-polysomnography, RBD was observed in 43 of 49 (88%) patients with multiple system atrophy; in 81%, bed partners reported some improvement in parkinsonian features during RBD episodes (38). The improvement was also noted by the recordings and included improved movement, speech, and facial expression (50% of patients). A study of 55 patients with multiple system atrophy, found that there was no difference in severity of motor or nonmotor symptoms based on the presence or absence of REM behavior disorder (243).

Even more dangerous is obstructive sleep apnea, which occurs in the later stages and may produce sudden death (138). It may be related either to central alveolar hypoventilation caused by loss of brainstem ventilatory drive centers or to obstructive sleep apnea caused by laryngeal dystonia (240) producing upper airway obstruction (88).

Less frequent or severe features of multiple system atrophy include emotional lability, pyramidal signs, contractures, a tendency to lean laterally while seated (the “Pisa syndrome”), supranuclear ophthalmoplegia, antecollis, an irregular or jerky myoclonic action tremor that is of cortical origin (152), large-amplitude myoclonus, respiratory stridor, polyneuropathy, amyotrophy, distal cyanosis, and Raynaud phenomenon. Young-onset multiple system atrophy (onset before age 40), includes dystonia, dyskinesia, and pyramidal signs and is typically more responsive to levodopa than classic multiple system atrophy (16). Dementia occurs no more commonly than in Parkinson disease (175; 229). There is often a mild normocytic/normochromic anemia that may be caused by loss of sympathetic stimulation of renal erythropoietin production (256).

Most prominent among the dysautonomic findings is orthostatic hypotension, but other important defects are postprandial hypotension, supine hypertension, anhidrosis with thermoregulatory disturbance, poor lacrimation and salivation, constipation, and impotence. Bladder emptying function may also be severely impaired. This usually starts with urgency, but later in the course of many cases, simulates prostatic enlargement in men and stress incontinence in women (17). Bladder dysfunction was found to be the presenting symptom in 18.2% of a cohort of 121 patients with multiple system atrophy (184). In a series of 29 autopsy-confirmed cases evaluated at the Mayo Clinic, rapid progression, early postural instability, poor levodopa responsiveness, and symmetric involvement coupled with progressive adrenergic and sudomotor failure, were highly predictive features of multiple system atrophy (87).

According to Xia and Postuma, normal olfaction, normal cognition, urinary retention, stridor, or respiratory insufficiency are useful for separating multiple system atrophy and Parkinson disease or dementia with Lewy bodies (267). Fanciulli and colleagues propose a 4-point scale for diagnosis of MSA-P with 1 point for each of the following: “postural instability with Hoehn and Yahr stage greater than or equal to 3 within 2 years of disease onset, orthostatic hypotension, overactive bladder symptoms, and urinary retention” (54). A score of 2 or greater is provided as leading to 78% sensitivity and 86% specificity for differentiating between Parkinson disease and MSA-P. A separate study by Fanciulli and colleagues shows 95% specificity for urinary retention for differentiation between MSA-P and Parkinson disease as well as 73% negative predictive value for MSA-P diagnosis in the case of absence of both orthostatic hypotension and urinary disturbance (urinary urgency, incontinence, or retention) (55).

Olfactory function remains preserved in multiple system atrophy as tested via the University of Pennsylvania Smell Identification Test (UPSIT) (73).

Gastrointestinal symptoms, pain, urinary problems, and postural instability due to orthostatic hypotension are reported more frequently in multiple system atrophy than in Parkinson disease, progressive supranuclear palsy, or corticobasal degeneration (31).

Even though multiple system atrophy has been considered to be a disorder without dementia, neuropathologically confirmed cases of multiple system atrophy with cognitive impairment have been reported (12). MSA-P shows a wider involvement of cognitive dysfunction than MSA-C, possibly associated with prefrontal dysfunction (100). Patients with MSA-P have reduced verbal retrieval, whereas patients with MSA-C have difficulties learning new information (14). Frontal lobe-related executive dysfunction and depression were identified in a study of 61 clinically diagnosed probable multiple system atrophy cases (207). Primarily cortical pathology, as well as damage to cerebellar circuits and frontal lobe input from the striatum, are possible contributors to cognitive impairment (214).

The Movement Disorders Society revised the diagnostic criteria in 2022. Clinically established multiple system atrophy is defined as “autonomic dysfunction, poorly levodopa responsive parkinsonism or cerebellar syndrome, two supportive clinical features, and at least one MRI marker.” Clinically probable multiple system atrophy is defined as two of these three features: "autonomic dysfunction, parkinsonism, and cerebellar dysfunction as well as one supportive clinical feature” (252).

Prognosis and complications

As a heterogeneous group, autopsy-proven cases of multiple system atrophy survived a mean of 8.0 years in an autopsy series (84). A meta-analysis of 433 cases gave a mean survival of only 6.2 years (19), with no difference between the parkinsonian type and cerebellar type. This differs somewhat from the findings of a study of 141 patients with moderately severe multiple system atrophy (mean age at symptom onset 56.2± 8.4 years) who had a median survival of 9.8 years (95% CI 8.1-11.4) (248). Shorter survival was suggested by the parkinsonian variant of multiple system atrophy and incomplete bladder emptying, and shorter symptom duration at baseline and absent levodopa response predicted rapid progression. In a large series, median intervals from onset to the need for gait assistance was 3 years, to wheelchair confinement 5 years, to a bedbound state 8 years, and to death 9 years (245). In the final stage, all three variants will produce a rigid, bradykinetic state leading eventually to a bed-bound existence with aphagia and its complications. Rare cases of young-onset multiple system atrophy have been reported with disease duration of 2 to 10 years at last follow-up visit (16; 104).

Levodopa may induce disabling confusion and psychosis and may cause dyskinesias without concomitant motor benefit (84). Sudden death is over seven times as common in patients with dysautonomic predominance than in other forms of multiple system atrophy (221). Another study found that severe generalized autonomic failure and early requirement for bladder catheterization are associated with shorter survival (57). Similarly, in a natural history study 175 patients with probable MSA-P or MSA-C evaluated every 6 months for 5 years with the Unified Multiple System Atrophy Rating Scale (UMSARS) and the Composite Autonomic Symptoms Scale (COMPASS), the median survival from symptom onset was 9.8 years (95% CI 8.8-10.7), but patients with severe symptomatic autonomic failure at diagnosis progressed more rapidly than those without it (127). The presence of MRI changes specific to multiple system atrophy in early disease was correlated with more rapid disease progression in MSA-P (110).

Cases of multiple system atrophy with prolonged survival, which typically start with parkinsonism and development of autonomic dysfunction with a mean latency of 11 years, after which more rapid clinical worsening begins, have been reported by multiple research centers (103; 167; 26). In addition, a case of multiple system atrophy with onset at the age of 34 has been reported (104).

Clinical vignette

A 57-year-old right-handed bookkeeper, on disability, complained of a 6-year history of autonomic symptoms as exemplified by urinary urgency as well as urge incontinence. About 4 years ago, she developed a staggering “drunk” gait, dysarthria, and difficulty with balance. She had multiple falls due to loss of balance. She had no history of lightheadedness. About a year ago, she developed right upper extremity rest tremor as well as REM behavior disorder. She had been taking carbidopa-levodopa, without any response. Her MRI showed cerebellar atrophy.

On physical examination, there was mild saccadic pursuit with saccadic hypometria. Gait was both broad-based and shuffling. There was mild dysmetria on finger-to-nose testing, which was worse on the left side. The patient also had bradykinesia in the bilateral upper extremities, which was worse on the right side, as well as rigidity, which was worse in the right upper extremity. She was diagnosed with probable MSA-C as the cerebellar features were most prominent.

Biological basis

Etiology and pathogenesis

A novel theory of MSA-P being a prion disorder has been advanced by Dr. Stanley Prusiner (173; 257). In his experiments, human embryonic kidney cells infected with basal ganglia samples from deceased patients with a clinical and pathological diagnosis of multiple system atrophy caused neurodegeneration with alpha-synuclein deposition in mice heterozygously expressing mutant human alpha-synuclein (A53T). Toxicity of alpha-synuclein aggregates in a human neuronal cell model is worsened by increased amounts of alpha-synuclein monomer and the presence of aggregates from human Parkinson disease or multiple system atrophy brain homogenates (225).

One study has shown an association of multiple system atrophy with various exogenous exposures such as solvents, pesticides, metals, and components of plastics (234). Another showed an association with farming experience (235). There appears to be an inverse association with smoking, as occurs in Parkinson disease, which is independent of the farming association (233; 235). In a population of Korean patients, the risk of multiple system atrophy was independent of participation in agriculture (28). A greater than expected exposure of patients to occupational toxins and frequent occurrence of minor dysautonomic symptoms in their relatives suggest that multiple system atrophy may be caused by a hereditary predisposition to an exogenous chemical toxicant (145).

Genetics. Fewer than 1% of patients have a family history of progressive supranuclear palsy (209). A hint of a specific genetic factor is the unconfirmed association of multiple system atrophy with a specific allele in the genes for interleukin 1-beta (147) and alpha-1-antichymotripsin (64). No mutations have been found in LRRK2, the gene presently considered to be the most commonly mutated in Parkinson disease (222), although LRRK2 was found to colocalize with glial cytoplasmic inclusions in laboratory studies (94). Single nucleotide polymorphisms at the SNCA locus, which codes for alpha-synuclein, have been associated with increased risk for developing multiple system atrophy (191). Scholz and colleagues propose that instead of direct amino acid sequence alteration, differences in expression of the SNCA gene may contribute to the development of multiple system atrophy. The G51D mutation of SNCA was identified in a case of young-onset Parkinson disease with both neuronal and glial cytoplasmic inclusions of alpha-synuclein demonstrating features of both Parkinson disease and multiple system atrophy (102). A53E mutation was also associated with multiple system atrophy (133). F-box only protein 7 (FBXO7) mutations are also a cause of young-onset Parkinson disease (43), and this protein was found to colocalize with alpha-synuclein in Lewy bodies of Parkinson disease and glial cytoplasmic inclusions of multiple system atrophy (271). Mutations in COQ2, the gene encoding coenzyme Q10, were found in six families with multiple family members affected by multiple system atrophy (137). Mutations of this gene, such as the V343A variant, may be susceptibility factors for developing sporadic multiple system atrophy. This variant is now referred to as V393A (c.1178T--> C (p.V393A)) given that the incorrect start codon was initially used for reference (95). This variant was not found in a European cohort of 788 patients with multiple system atrophy (198). A Caucasian patient with severe MSA-C pathological changes was reported to have COQ2 p.S146N heterozygous variant (149). Sequencing of COQ2 in a Chinese cohort of 116 patients with MSA-C revealed changes in this gene in three patients: p.L402F and P.R173H missense and p.A32A synonymous mutation (246). In an Italian study of 87 patients with probable multiple system atrophy, homozygous p.A43G mutation was found in a single patient with MSA-C, but according to further analysis, this variant was not pathogenic (180). Decreased levels of coenzyme Q10 were found in postmortem cerebellar specimens of patients with multiple system atrophy (15). In a study of 40 patients with multiple system atrophy, plasma COQ10 levels correlated with motor symptoms in MSA-C only (47). A genome-wide association study by Sailer and colleagues did not find evidence of association of multiple system atrophy with COQ variants; further research is needed regarding this gene in Asian populations (183; 190). Microglial activating gene TREM2 variant p.R47H is suspected of increasing risk of developing multiple system atrophy (150).

A patient with MSA-C diagnosis due to ataxia and autonomic dysfunction with a family history of amyotrophic lateral sclerosis was found to have C9orf72 repeat expansion (71).

Pathology. The hallmark of multiple system atrophy pathology is the presence of glial cytoplasmic inclusions (231; 266; 94). Glial cytoplasmic inclusions are most consistently present in the external and internal capsules, the central tegmental tract, and the white matter of the cerebellar cortex (10). Alpha-synuclein filaments in multiple system atrophy have a conformation that is different from filaments in diffuse Lewy body disease (195). The glial cytoplasmic inclusion pathology may be caused by internalization of pathological alpha-synuclein into oligodendrocyte precursor cells (96). The mechanism that selects for glial accumulation of alpha-synuclein and preference for the particular alpha-synuclein conformation in glial cytoplasmic inclusions as opposed to the Lewy body conformation has not yet been elucidated (27).

A case with abundant glial cytoplasmic inclusions, but not fulfilling diagnostic criteria for multiple system atrophy, has been reported (161). Of unknown significance is a case of hypoparathyroidism with basal ganglia calcifications in addition to multiple system atrophy pathology and phenotype. Massive lipid-laden macrophage infiltration of the pontocerebellar fibers was described in one case (264).

In a postmortem study of 35 patients with multiple system atrophy, Cykowski and colleagues additionally found neuronal inclusions of alpha-synuclein in the striatum, substantia nigra, anterior cingulate, amygdala, entorhinal cortex, basal forebrain, hypothalamus, and cerebellar roof nuclei (37). Salvesen and colleagues found neuronal loss in frontal and parietal cortex, which may correspond to cognitive impairment in multiple system atrophy (187).

In MSA-P, there is cell type-specific neuronal loss, with calcineurin-positive medium spiny neurons depleted in the posterior putamen as compared to choline acetyltransferase-positive neurons (188). A subsequent study showed that in the caudal and dorsolateral putamen, calbindin-positive neurons are more affected than calcineurin-positive neurons (189). Hyperechogenicity of the lentiform nuclei on transcranial sonography has been found in MSA-P and progressive supranuclear palsy, but it’s unclear what underlies this finding (178).

Lewy body pathology has been found in submandibular glands of patients with Parkinson disease and incidental Lewy body disease, but not in patients with multiple system atrophy (42).

Neuronal cell loss has been reported in the Onufrowicz nucleus and the intermediolateral cell column, as well as a portion of the intermediate reticular formation, dorsal motor nucleus of vagus, nucleus ambiguus, caudal raphe nucleus, arcuate nucleus, pontine micturition center, and locus ceruleus (156). Rostral raphe neurons are relatively preserved in multiple system atrophy, in contrast to diffuse Lewy Body disease (18). The pedunculopontine nucleus has also been implicated in multiple system atrophy (142).

A rare association of multiple system atrophy with Alzheimer disease has been described (181). Alpha-synuclein and phosphorylated tau colocalize in certain brain regions in cases of combined multiple system atrophy and Alzheimer disease (227). Blood-brain barrier dysfunction has been found to correlate with disease severity in multiple system atrophy (211).

Cell biology. Glial cytoplasmic inclusions include hyperphosphorylated alpha-synuclein, ubiquitin, LRRK2, as well as other proteins, including F-box only protein 7 (94; 43; 271). Phosphorylated alpha-synuclein at serine 87 and 129 has been detected in multiple system atrophy brains, but not in controls (158). On the other hand, a series of 58 patients with multiple system atrophy was negative for alpha-synuclein mutations (121). In an animal model of multiple system atrophy, alpha-synuclein knockout mice appeared to be resistant to 3-nitropropionic acid-induced neuronal loss and dendritic pathology (232). In an oligodendroglial cell line, expression of tubulin polymerization promoting protein (TPPP)/p25 triggered formation of alpha-synuclein positive cytoplasmic inclusions and apoptotic cell death (251).

Dysfunction of the ubiquitin-proteasome system has been reported in multiple system atrophy (24). Phosphorylation of DARPP32, a major substrate of cdk5, may be involved in the formation of oligodendrocyte glial cytoplasmic inclusions (82). Reduction in LRRK2 expression in oligodendroglia is associated with increased neuronal loss in multiple system atrophy, but parkin immunoreactivity has been seen in only a small proportion of glial cytoplasmic inclusions (83). An oligodendroglial protein, p25alpha, which interacts with myelin basic protein, migrates to glial cytoplasmic inclusions (212). NUB1, a synphilin-1-interacting protein, is found in the intracytoplasmic inclusions of both neuronal and oligodendroglial cells, neuronal nuclear inclusions, and swollen neurites (224). HtrA2/Omi, a mitochondrial proapoptotic serine protease that is released into the cytosol, has been found in glial and neuronal cytoplasmic inclusions in multiple system atrophy brains (101).

Transcranial magnetic stimulation with triple stimulation technique detects corticospinal tract involvement in multiple system atrophy (52). Median nerve somatosensory evoked potential measurement shows slowing of the central sensory conduction time as the disease progresses (136). Microvessel degeneration has been found in medullary autonomic nuclei (134).

High levels of CSF axonal biomarkers are found in MSA-P. Neurofilament light chain, heavy chain, and tau are significantly increased in MSA-P (all p < 0.0001), whereas noradrenergic metabolite 3-methoxy-4-hydroxyphenylethyleneglycol levels are significantly decreased in MSA-P (p < 0.0001) as compared to Parkinson disease (01). No difference in biomarkers has been found between MSA-C and MSA-P (02). CSF hypocretin (orexin) levels are normal (131). Although neurofilament light chain can differentiate between Parkinson disease and atypical parkinsonism, it cannot differentiate between multiple system atrophy, progressive supranuclear palsy, and corticobasal degeneration (33). DJ-1 CSF levels alone or in combination with CSF tau levels can differentiate multiple system atrophy from Parkinson disease (79). A review by the multiple system atrophy biomarkers initiative reveals that DJ-1, alpha-synuclein, total-tau in combination with neurofilament light chain, and catecholamine metabolite levels are the most promising biomarkers so far (116). Singer and colleagues also proposed neurofilament light chain, total alpha-synuclein, and catecholamine metabolite 3,4-dihydroxyphenylacetic acid as possible biomarkers (206). In addition, coenzyme Q10 is being investigated as a peripheral biomarker for multiple system atrophy due to association of the disease with COQ2 mutations as a risk factor for sporadic multiple system atrophy in the Japanese population (112). Measurement of alpha-synuclein in blood exosomes identified by oligodendroglial versus neuronal markers may distinguish multiple system atrophy from Parkinson disease (48).

Pathophysiology. Patients with multiple system atrophy have neurogenic orthostatic hypotension and decreased baroreflex function, with a relatively preserved cardiac sympathetic innervation (86; 220). A relatively high frequency of cardiac ectopy has been found (72). Autonomic dysfunction can be exacerbated by medical treatment of other symptoms (272). Resting baroreflex sensitivity has been shown to be lower in a group of 35 patients with multiple system atrophy (60). Postsynaptic sympathetic nerves in the epicardium are affected in multiple system atrophy (154; 155), but the changes are less prominent than in Parkinson disease or diffuse Lewy Body disease (156). Awake ventilatory response to hypercapnia and hypoxia appears to be affected later in the disease course than baroreflex function (125). In a group of patients with multiple system atrophy who suffered sudden death as compared to patients with multiple system atrophy who died from well-established causes, depletion of serotonergic neurons in the ventrolateral medulla and nucleus raphe obscurus was present (220). Neurons in the intermediolateral cell column, as well as catecholaminergic neurons in the ventrolateral medulla, were affected equally in both groups.

Growth hormone release in response to arginine is impaired in multiple system atrophy, but not in Parkinson disease (156). This response to arginine can also differentiate MSA-P from progressive supranuclear palsy (166). The clonidine stimulation test for growth hormone release has higher sensitivity and specificity for differentiating MSA-P and Parkinson disease. A combination of clonidine and arginine tests could further enhance sensitivity and specificity (270).

In a study, patients with multiple system atrophy were found to have pupillary sympathetic dysfunction from an early stage, whereas in patients with Parkinson disease, it tended to gradually accelerate at a more advanced stage of the disease (263). Pupillography may distinguish multiple system atrophy from Parkinson disease (177). In multiple system atrophy patients, pupillary sympathetic sensitivity has been correlated with the severity of orthostatic hypotension during a head-up tilt test and with the elevation of systolic blood pressure during a noradrenaline infusion test (263).

Blunted heart rate (59) and electroencephalographic changes during period limb movements of sleep have been observed in multiple system atrophy (239). Decreased R-R interval variability and impaired sympathetic skin response correlate more with multiple system atrophy than Parkinson disease (21). Skin sympathetic dysfunction has also been demonstrated via impaired vasodilator response to local heating (262).

Postprandial hypotension correlates with degree of orthostatic hypotension in multiple system atrophy and is ameliorated by acarbose, possibly due to decreased postprandial secretion of glucagon-like peptide 2 (GLP-2) (63).

In an electromyography study, the mean duration of motor unit potentials recorded from the external anal sphincter was significantly longer in patients with multiple system atrophy and progressive supranuclear palsy compared with motor unit potentials recorded from patients with Parkinson disease (P < 0.005 for both) (255).

Chronic orthostatic hypotension in patients with multiple system atrophy leads to structural remodeling of veins and reduced venous compliance, which actually serves to counteract excessive venous pooling in the lower extremities (123).

Epidemiology

At a large referral center, 100 patients with multiple system atrophy reported a median age at onset of 53 years (range 33 to 76 years) (247), whereas another registry reported a mean age at onset of 60 years (SD=9; range 34 to 83 years) (215). In a study of the natural history of multiple system atrophy in the United States by Low and colleagues, median survival from symptom onset was 8.9 years, with presence of severe autonomic failure signaling worse prognosis (median survival of 8 versus 10.3 years) (127). Degree of autonomic failure, urinary catheterization within 3 years of onset, bladder symptoms within 3 years of onset, older age of onset, and falls within 3 years of onset were found to be associated with worse prognosis in a study by Coon and colleagues (34). In Jellinger's unpublished series of 600 autopsied brains from patients with parkinsonian disorders, multiple system atrophy occurred in 5.1% (175). The sexes appear to be affected equally (215). The incidence in the 50- to 99-year age group has been measured at 3.0 new cases per 100,000 person-years (22), and the prevalence in a medically based series in London was 4.4 per 100,000 population (192), about 3% of that of Parkinson disease. The annual incidence of new cases is about 0.6 per 100,000 population (234; 215). The crude incidence of MSA-P in northern Sweden is 2.1 per 100,000 (122). In Iceland, the incidence was 0.7 per 100,000 and the prevalence was 3.4 per 100,000 (20). Multiple system atrophy of the cerebellar type comprises 29% of the cases of sporadic, adult-onset ataxia (03). Rare cases of multiple system atrophy with onset prior to the age of 40 have been reported (104).

Also see the article on Epidemiology of movement disorders.

Differential diagnosis

Clinical and imaging findings of sporadic adult-onset ataxia of unknown etiology have some overlap with multiple system atrophy (04). Although multiple system atrophy is considered a sporadic disease, four families, one of which had a consanguineous marriage, were reported with familial cases of multiple system atrophy phenotype with autonomic failure, parkinsonism, and cerebellar ataxia (77). These patients were negative for spinocerebellar ataxia types 1, 2, 3, 6, 7, 12, 17, dentatorubral-pallidoluysian atrophy, and alpha-synuclein mutations. A clinical review of 249 sporadic adult-onset degenerative ataxia cases found higher scores on the Scale for the Assessment and Rating of Ataxia (SARA) in cases that met MSA-C criteria compared to those that did not (70).

Spinocerebellar ataxia type 2 may present with a parkinsonian or cerebellar phenotype and should be considered in the differential diagnosis, especially if an autosomal dominant inheritance pattern is established (115; 265). Spinocerebellar ataxia type 3 has also been reported with a multiple system atrophy phenotype (261). Aprataxin mutations leading to ataxia with oculomotor apraxia can also have an overlapping phenotype with MSA-C (13).

Autonomic dysfunction in late-onset cerebellar ataxia complicates differential diagnosis with MSA-C (119). An SCA17 case with progressive ataxia, autonomic dysfunction, parkinsonism, supranuclear gaze palsy, and cognitive impairment has been reported (120). A case of SCA17 with reduced penetrance given 41 CAG repeats (44 repeats are needed for full penetrance) was reported. This patient had lack of putaminal hyperintense rim, but did have urinary incontinence and erectile dysfunction. Although he fulfilled diagnostic criteria for probable multiple system atrophy, the overall clinical picture was more consistent with SCA17 (46). In contrast, a series of 199 patients with a clinical diagnosis of multiple system atrophy was negative for SCA12 (29). In addition, hereditary spastic paraplegias, fragile-X tremor ataxia syndrome, mitochondrial disease, Gaucher disease, Perry syndrome, and cerebrotendinous xanthomatosis can potentially mimic multiple system atrophy (213).

“Red flags” with greater than 95% specificity for MSA-P as opposed to Parkinson disease include postural instability within 3 years of disease onset and resulting recurrent falls, wheelchair dependency within 10 years of onset, Pisa syndrome (prolonged episodes of lateral trunk flexion), disproportionate antecollis (237), contractures of hands or feet excluding Dupuytren disease, diurnal inspiratory stridor, nocturnal inspiratory stridor, inspiratory sighs, severe dysphonia, severe dysarthria, severe dysphagia, and emotional incontinence (108). Laryngeal abductor paralysis (141), generally occurring in advanced stages of multiple system atrophy, has also been reported in the early stages (40). It may lead to nocturnal sudden death (50). Stridor is a high-pitched inspiratory sound associated with laryngeal dysfunction (35). Laryngoscopy and video polysomnography may be useful for diagnostic confirmation of stridor (35). Stridor may have different patterns based on multiple system atrophy phenotype. Todisco and colleagues found that MSA-P has a dystonic pattern, whereas MSA-C had a dystonic pattern with additional neurogenic findings on vocal cord EMG (230).

“Cold hand sign,” with poor circulatory return after blanching, is a clue for multiple system atrophy with poorly understood pathophysiology (168). Lower limb tremor, although a possible presentation of Parkinson disease, may be suspicious for multiple system atrophy (78). Symmetric symptoms, which were reported in 48% of a cohort of patients with MSA-P, should also increase suspicion for this disorder when evaluating patients with Parkinson disease (74). Emotional incontinence exhibited by inappropriate laughter or crying has also been reported in 36% of a series of patients with MSA-C (162). “Cold hand sign” with palmar skin temperature less than 28°C was reported in 6% of multiple system atrophy cases and no Parkinson disease or normal control cases (11).

The presence of excessive square-wave jerks, mild to moderate hypometria of saccades, impaired vestibulo-ocular reflex suppression, spontaneous nystagmus, or positioning downbeat nystagmus may be oculomotor clues to the presence of multiple system atrophy, whereas the presence of clinically slow saccades or moderate to severe gaze restriction argues against multiple system atrophy (08). Of note, hypometric saccades have been reported in Parkinson disease (09). Ocular jaw synkinesia, described in Parkinson disease, is not seen in MSA-P (186).

Differentiation of MSA-P from the subset of patients with Parkinson disease and diffuse Lewy body disease with prominent autonomic failure is a profound diagnostic challenge. In a retrospective study of 134 patients with clinical diagnosis of multiple system atrophy on whom subsequent pathological evaluation was performed, the diagnostic accuracy of multiple system atrophy versus diffuse Lewy body disease, Parkinson disease, and progressive supranuclear palsy was 64%, with autonomic failure in patients with diffuse Lewy body disease and Parkinson disease as the main cause of misdiagnosis of multiple system atrophy (107). In a prospective study of patients with multiple system atrophy and Parkinson disease, scores on self-report questionnaires of autonomic symptoms as well as on clinical assessments of autonomic function were similar in the two groups (124). The pattern of anhidrosis is different in the two groups. Intact quantitative sudomotor axon testing in anhidrotic areas confirmed a preganglionic lesion in multiple system atrophy patients, whereas patients with Parkinson disease had anhidrosis due to a peripheral lesion (124). Postganglionic sudomotor denervation reported in a study confounds this diagnostic method (172).

Alpha-synuclein pathology was seen on colonic biopsy of the enteric nervous system in one out of six patients with multiple system atrophy and in five out of nine patients with Parkinson disease (171). Thus, although synuclein pathology is less common in multiple system atrophy than in Parkinson disease, it may involve the entire enteric nervous system.

Although progressive supranuclear palsy parkinsonism (PSP-P) shares many features with multiple system atrophy, autonomic dysfunction is uncommon in this disorder and can help distinguish it from multiple system atrophy (254). In a study by Koga and colleagues, the prominent cerebellar features of the cerebellar subtype of progressive supranuclear palsy (PSP-C) led to misdiagnosis as MSA-C (107).

Corticobasal degeneration may be confused with multiple system atrophy, but the former is highly asymmetric and always shows important apraxia and cortical sensory loss. The multiple system atrophy with a predominance of dysautonomia may be confused with progressive autonomic failure. However, the dysautonomic features in multiple system atrophy typically start with bladder and erectile dysfunction, proceeding to thermoregulatory difficulty and orthostatic hypotension, and finally to ventilatory dysrhythmias. Progressive autonomic failure, on the other hand, begins with thermoregulatory and blood pressure dysregulation; bladder dysfunction occurs last, and ventilatory arrhythmias do not occur (129).

In a large series of patients carrying a diagnosis of multiple system atrophy of the cerebellar type, 4% (3 of 76 patients) had a mutation in the FMR1 gene, suggesting that fragile-X-associated tremor-ataxia syndrome should be considered in the differential diagnosis (97).

Multi-infarct states may mimic multiple system atrophy and many other degenerative disorders but are usually identified by vascular lesions on MRI scan, arteriosclerotic risk factors, and a history of stepwise progression. Additionally, a rare case of multiple system atrophy masking symptoms of multiple sclerosis has been reported (58). Cases of pure autonomic failure do not have the additional parkinsonian or cerebellar signs of multiple system atrophy on physical examination.

Diagnostic workup

MRI will help rule out a multi-infarct state or normal-pressure hydrocephalus as a cause of dopa-unresponsive parkinsonism and will also rule out an anatomical lesion of the cerebellum.

The “hot cross bun” sign is a hyperintensity in the pons on T2 MRI images of MSA-C (07; 202). Of note, the “hot cross bun” sign has also been observed in spinocerebellar ataxia types 2 and 3, parkinsonism secondary to vasculitis, and variant Creutzfeldt-Jacob disease (25; 139; 23; 208). This sign may be clearer on T2* imaging as compared to T2 (182).

Putaminal atrophy can differentiate multiple system atrophy from Parkinson disease (23; 196). Meta-analysis of six studies confirmed a statistically significant reduction in putamen volume in multiple system atrophy as opposed to Parkinson disease (185). Signal loss in dorsolateral putamen on T2 MRI sequences with the presence of a hyperintense lateral rim in fluid-attenuated inversion recovery (FLAIR) sequences has a specificity of 0.97 for discriminating between multiple system atrophy and Parkinson disease (241). Although specificity is high, sensitivity of these signs on 1.5 Tesla MRI in early stages of the disease is suboptimal (23; 196). Hyperintense putaminal rim on three Tesla T2 images by itself may be a nonspecific finding (23; 196). Presence MRI changes specific for multiple system atrophy were associated with faster clinical progression in a post-hoc analysis of the multiple system atrophy–rasagiline study (110).

“Pointers” raising the possibility of multiple system atrophy also include hyperintensity of the middle cerebellar peduncle, pons, and cerebellum (196). Susceptibility-weighted MRI (76), as well as magnetic resonance spectroscopy (75), may help in differentiating between multiple system atrophy and Parkinson disease. The European Multiple System Atrophy Study Group found that severity of orthostatic hypotension was not associated with multiple system atrophy subtype and that sensitivity to pick up orthostatic hypotension increases from 54% at 3 minutes to 72% with a prolonged 10-minute orthostatic challenge (165).

A set of diagnostic criteria proposed by Quinn (175) was modified and operationalized by a Consensus Conference in 1999. The criteria from the Second Consensus Conference are outlined here (69).

A validated disability rating scale is available (253).

Table 1. Multiple System Atrophy Diagnostic Criteria

Table 2. Additional Diagnostic Criteria for Possible Multiple System Atrophy

Investigational diagnostic methods. Numerous studies evaluating possible additional diagnostic methods to separate multiple system atrophy from clinically similar disorders have been performed. Given the rarity of multiple system atrophy, these usually involve small numbers of subjects. Many involve advanced imaging acquisition and analysis methods that may not be readily available in daily clinical practice.

Fractional anisotropy and apparent diffusion coefficient changes are present prior to hyperintense putaminal rim on MRI sign and may be used to differentiate multiple system atrophy from Parkinson disease (91). A scoring system of putaminal atrophy, signal hypointensity, and abnormal disruption of the hyperintense lateral rim of the putamen on 3T MRI had sensitivity of 70.6% and specificity of 93% for distinguishing between Parkinson disease and early MSA-P as shown in a 3-year follow-up study (117). Correlating with putaminal atrophy on imaging studies, on gross specimens, there is dorsal atrophy and darkening of the putamen due to lipofuscin, neuromelanin, and increased iron (93).

Hypometabolism on fluorodeoxyglucose positron emission tomography (FDG-PET) in the brainstem, cerebellum, or putamen, as well as presynaptic nigrostriatal dopaminergic denervation in SPECT or PET, are included in the diagnostic criteria (69). A variety of investigational techniques have potential for the differentiation of multiple system atrophy and Parkinson disease, as well as between the subtypes of multiple system atrophy. In PET testing, putamen to substantia nigra ratios and putamen to caudate ratios of 6-[18F]fluorodopa-derived activity as well as cardiac 6-[18F]fluorodopamine-derived activity separate Parkinson disease from multiple system atrophy (73; 113). Also, 123I-Ioflupane scanning can be used to differentiate multiple system atrophy from Parkinson disease (174). Reduction in AChE activity as measured by [(11)C] N-methylpiperidin-4-yl propionate PET has been seen in the thalamus (-27%) and the posterior lobe of cerebellar cortex (-36%) in patients with MSA-C (80). A fully automated, voxel-based FDG-PET method has been shown to differentiate multiple system atrophy from Parkinson disease and controls (49). Analysis of metabolic patterns visualized via [18F]FDG PET by Kim and colleagues classified multiple system atrophy into three metabolic categories: striatal, cerebellar, and mixed hypometabolism, with striatal and mixed groups showing corresponding striatal dopamine transporter loss on ([18F]FP-CIT) PET imaging (105). Substantia nigra hyperechogenicity on transcranial sonography, seen in Parkinson disease, was observed less frequently in Japanese patients with multiple system atrophy (151). Transcranial sonography could also distinguish between patients with Parkinson disease and atypical parkinsonism in a German population (242). Of note, 10% of the population has been found to have abnormal echogenicity of the substantia nigra (258). Differences in cardiac and whole-body metaiodobenzylguanidine (MIBG) imaging are seen between multiple system atrophy and Parkinson disease (130).

Voxel-based morphometry as well as tractography are not appropriate for the diagnostic workup of individual patients, but they offer important research insights (23). Morphometric measurement of midbrain and pons can distinguish MSA-P and progressive supranuclear palsy (36). Additionally, putamen/caudate volume ratios on MRI scans can also differentiate between multiple system atrophy and Parkinson disease (201). Voxel-based morphometry and relaxometry comparison of MSA-P and MSA-C shows infratentorial differences, namely, stronger reduction of gray matter in the cerebellum and white matter in the brainstem, as well as reduction of the relaxation rate R2 in the cerebellum and brainstem (135). MSA-P does not show any changes in the superior cerebellar peduncle (98). On tractography-based analysis, MSA-C cases display decreased fractional anisotropy and volume as well as increased apparent diffusion coefficient values in the middle cerebellar peduncle, compared to controls (226). Diffusion tensor imaging indexes correlate with severity of orthostatic hypotension and disease duration in MSA-C (228). Location of degenerated pontocerebellar tracts on diffusion tensor imaging corresponds to the transverse portion of the hot cross bun sign (62). MSA-C can be differentiated from Parkinson disease via computer-assisted statistical analysis of brain perfusion SPECT images (244). Song and colleagues found a difference in occipital Tc-99m hexamethylpropyleneamine oxime (HMPAO) perfusion SPECT between MSA-P and mild Parkinson disease (210). Three-dimensional fractal analysis (259) and voxel-based morphometry (143) can also show cerebellar volume loss in MSA-C. Three-dimensional “gyrification index” of the cerebellum based on T1-weighted MRI images can be used to differentiate MSA-C from controls (260). Mean diffusivity values of brainstem and cerebellar hemispheres derived from diffusion-weighted MRI imaging were higher in both subtypes of multiple system atrophy as compared to the Richardson syndrome variant of progressive supranuclear palsy and Parkinson disease (146).

11C-raclopride PET scanning, using both D2 receptor binding potential and cerebral influx ratio, can distinguish patients with MSA-P and Parkinson disease better than current striatal D2 receptor-only analysis (238). There is a higher rate of signal reduction in caudate and anterior putamen in MSA-P as compared to Parkinson disease on [(123)I]β-CIT SPECT (148).

There is a 3-year premotor phase with autonomic dysfunction, including orthostatic hypotension and erectile dysfunction prior to the onset of parkinsonism or ataxia in multiple system atrophy (159). Premotor Parkinson disease may have similar features, but cardiac sympathetic denervation and hyposmia are not seen in premotor multiple system atrophy. There is a single case report of open bladder neck during storage phase of the bladder as visualized on urodynamics, which preceded the development of generalized motor symptoms of multiple system atrophy (269).

Nerve conduction and EMG studies may show subclinical polyneuropathy in multiple system atrophy. External urethral sphincter EMG shows denervation in almost all patients with multiple system atrophy (164), but not in Parkinson disease or in other cerebellar degenerations (106; 250). The ability of this procedure to distinguish multiple system atrophy from Parkinson disease has, however, been disputed (67), and false positives arise from many nonneurologic causes of sphincter denervation, such as multiparity or pelvic surgery.

Management

Drug treatment: Parkinsonian symptoms. Available treatments are primarily symptomatic (51). Although estimates vary, about 30% to 40% of patients with multiple system atrophy may be responsive to levodopa (32). Facial dystonia occurs more frequently than limb dyskinesias with levodopa therapy. Levodopa therapy may worsen orthostatic hypotension (249; 32). Treatment may be initiated with carbidopa/levodopa 25/100 half to one tablet twice a day and increased every few days to efficacy or side effects. Patients with multiple system atrophy may require and tolerate far larger dosages than patients with Parkinson disease. High-dose levodopa therapy has been found to be nontoxic in a transgenic mouse multiple system atrophy model (216). Ancillary antiparkinsonian agents have not added significantly to the efficacy of levodopa, but adequate studies are lacking. In a double-blind 48-week study, treatment of MSA-P with rasagiline did not show a significant benefit (170).

Lifestyle management: orthostatic hypotension. Treatment of symptomatic orthostatic hypotension starts with sodium and volume repletion unless the patient is at risk of congestive heart failure or renal insufficiency. Patients with multiple system atrophy and dysautonomia should be informed that the frequent recommendation to minimize dietary sodium does not necessarily apply to them. Drinking 350 mL of tap water once daily in the early morning demonstrated a benefit in treating orthostatic hypotension in a group of patients with multiple system atrophy without any adverse effects (39). Four hundred and fifty milliliters of clear water ingestion was confirmed to have a pressor effect in another study of orthostatic insufficiency, but eating clear soup seemed to have a postprandial hypotensive effect (268).

Ancillary measures such as pressure stockings to increase central venous volume and elevating the head of the bed 6 inches to increase renin secretion may also be attempted, but often prove uncomfortable. It is also useful to avoid extreme heat with its reflex peripheral vasodilation and to avoid overeating and straining at stool, which increase vagal activity.

Droxidopa has been shown to be effective for neurogenic orthostatic hypotension in a double-blind trial (99). Typical starting dosage is 100 mg TID titrated in 100 mg increments every 24 to 48 hours up to a maximum dosage of 600 mg TID.

The mineralocorticoid fludrocortisone may be started at 0.1 mg daily and increased to a maximum of four tablets per day in two divided doses, given with fluid repletion. Midodrine, an alpha-adrenergic agonist, is a good alternative (92). Treatment starts with 2.5 mg three times a day increasing to 10 mg three times a day. Supine hypertension is an uncommon but important adverse effect that may be lessened by raising the head of the bed. If these drugs are unsuccessful, inhibition of vasodilator prostaglandin synthesis may be initiated with indomethacin 25 mg three times a day with meals increasing to 50 mg three times a day. Alternatives include the alpha-adrenergic agonist clonidine 0.1 mg twice a day increasing to 0.3 mg twice a day, and the peripheral vasoconstrictors ephedrine starting at 25 mg three times a day and propranolol starting at 20 mg twice a day.

Pyridostigmine has been found to ameliorate the hypotension of multiple system atrophy without causing supine hypertension, at least after a single 60 mg dose (205).

Drug treatment: urological complaints. Urinary frequency or incontinence may respond to a peripherally acting anticholinergic agent such as oxybutynin 5 to 10 mg at bedtime, tolterodine 2 mg at bedtime, or propantheline 15 to 30 mg at bedtime, if the pathophysiology is detrusor hyperreflexia. However, anticholinergic treatment may worsen the constipation of multiple system atrophy. The impotence of multiple system atrophy may respond to yohimbine 5.6 mg one to three times daily (179). Sildenafil, although efficacious, often exacerbates the orthostatic hypotension of multiple system atrophy (85).

Drug treatment: hallucinations. The hallucinations caused by dopaminergic therapeutic agents usually respond to clozapine 6.25 to 50 mg or quetiapine 25 to 75 mg at bedtime. These must be used cautiously in patients with hypotension. All patients receiving clozapine must be monitored for agranulocytosis with weekly white blood cell counts.

Ataxia. There is no known treatment for the ataxia of multiple system atrophy, olivopontocerebellar atrophy type.

Associated sleep disorders and respiratory complaints. Rapid eye movement behavior disorder usually responds well to clonazepam 0.5 mg at bedtime. Patients with multiple system atrophy have a higher risk of obstructive sleep apnea (66). Central apnea due to degeneration of pontomedullary respiratory centers may be masked by untreated obstructive sleep apnea (219). Nocturnal stridor (223) due to vocal cord paralysis and possible excessive adductor activation (240) is treated with continuous positive airway pressure (CPAP) (199).

Laryngoscopy under anesthesia may be useful for evaluating upper airway obstruction in multiple system atrophy (200). Tremulous arytenoid movements are a marker of severity of glottic stenosis, which worsens as the disease progresses (157).

Stridor presents controversies in prognosis and management, as the long-term survival value of interventions for stridor is unclear. According to a consensus statement by Cortelli and colleagues, the prognostic value of stridor is uncertain (Cortelli at al 2019). CPAP ventilation during sleep can be useful for symptomatic control of stridor and should be considered as a first-line treatment. If stridor occurs during wakefulness or CPAP is not tolerated, tracheostomy should be considered (89). The consensus statement cites tracheostomy as being effective in symptomatic control of stridor and may be required for severe stridor (35), but sudden death may occur in a subset of patients, even with tracheostomy (220).

Gastrostomy and tube feeding may be necessary in later stages of the disease (140).

Surgical treatment. Deep brain stimulation of the internal pallidum remains ineffective for multiple system atrophy and may even be harmful (114).

Experimental treatment. Intravenous immunoglobulin therapy has no effect on multiple system atrophy (144).

Nonsteroidal anti-inflammatory drugs (NSAIDs) have been found to inhibit the formation of alpha-synuclein fibrils, which are involved in the pathogenesis of multiple system atrophy and could serve as a potential drug development target for multiple system atrophy treatment (81). Fibril formation is also inhibited by nicotine (153). The Unified Multiple System Atrophy Rating Scale (UMSARS) has been evaluated as an outcome measure to be used in therapeutic trials (249; 132). The UMSARS Motor Examination score, which has been demonstrated to have satisfactory intrarater reliability, seems to be the best outcome measure for future therapeutic trials (111). Pontine and cerebellar atrophy rates on MRI can also serve as outcome measures (163). The PDQ-39 questionnaire, a disease-specific quality-of-life instrument for patients with Parkinson disease, has only limited validity in multiple system atrophy (193). A multiple system atrophy health-related quality-of-life scale (MSA-QoL) has been reported to have good potential for use in clinical trials (194). A prospective, 48-week, randomized, double-blind, multinational study failed to show a clinical effect of minocycline on symptom severity (45). More studies of the serotonergic system may also provide important insights due to reports of serotonergic depletion in ventrolateral medulla as well as a possible motor symptoms benefit from paroxetine (61; 220).

In a transgenic mouse model of multiple system atrophy based on 3-nitropropionic acid intoxication, rasagiline was found to be neuroprotective (217). A clinical trial for rasagiline in multiple system atrophy is currently ongoing. Therapies with rifampicin and autologous mesenchymal stem cells are also being investigated (251). Rifampicin failed to slow progression of autonomic deficits over a 12-month period (128; 204). A small double-blind trial of 33 patients with MSA-C treated with autologous stem cells showed clinical improvement, but there was a safety concern about cerebral ischemic lesions (56). A global MSA registry (GLOMSAR) was created to assist with research endeavors (56). Reduction of alpha-synuclein and suppression of microglial activation by minocycline or myeloperoxidase inhibition have shown benefit in animal models, but caution needs to be employed when extrapolating these results to potential for human disease modification (218; 109). In a mouse model, CLR01, which reduces alpha-synuclein oligomer formation and promotes destruction of formed oligomers, may reduce alpha-synuclein pathology (126). VX‐765, a caspase‐1 inhibitor precursor, and anle138b, an alpha-synuclein aggregation inhibitor that targets oligomeric α‐synuclein, are being tested in animal models (133). Current research also focuses on alpha-synuclein clearance via nilotinib or immunotherapy methods as well as alpha-synuclein suppression via antisense oligonucleotides, such as BIIB101 (118).

Outcomes

Please see specific outcome and side effect discussion in relation to particular treatment modalities above.

Special considerations

Pregnancy

Average age of onset at 53 years old is in the post-menopausal group. Information in effect of disease on pregnancy and vice versa is not available.

Anesthesia

Patients with any form of parkinsonism should avoid neuroleptic tranquilizers. Hypotension may be aggravated by opiates or other sedatives. The dysphagia of some patients with multiple system atrophy dictates careful postoperative airway management.