Multiple System Atrophy (MSA) is a rare and rapidly progressive neurodegenerative disorder that is characterized by a combination of parkinsonism, cerebellar impairment and autonomic dysfunction.1 Patients are divided into MSA-Parkinsonism (MSA-P) or MSA-Cerebellar (MSA-C) depending on the predominance of parkinsonism or cerebellar symptoms.2 The current diagnostic criteria also distinguish between levels of certainty, differentiating definite MSA for subjects with autopsy demonstration, probable MSA for patients with autonomic failure plus parkinsonism or cerebellar ataxia, and possible MSA for people with sporadic, progressive adult-onset disease including parkinsonism or cerebellar ataxia and one autonomic feature plus other clinical or neuroimaging abnormality.1

In 1969, Graham and Oppenheimer first introduced the term MSA to encompass the pathological entities of olivopontocerebellar atrophy, striatonigral degeneration, and Shy-Drager syndrome on the basis of the neuropathological findings in these disorders.3 The neuropathological hallmark of the disease is the presence of alpha-synuclein (α-synuclein) aggregates in oligodendrocytes forming glial cytoplasmic inclusions4,5 (Papp-Lantos bodies),4 observed in all types of oligodendroglia that contain aggregates of misfolded α-Synuclein, however GCIs are present not only in known atrophic areas but also in areas without atrophy such as the cortex. Hence, MSA is a oligodendroglial α-synucleinopathy. Macroscopically, MSA is characterized by striatal atrophy and depigmentation of the substantia nigra. In some cases, striatal degeneration is histological without macroscopic atrophy. These findings are associated with olivopontocerebellar atrophy, which may be macroscopically evident. The neuropathology of sporadic MSA also includes varying degrees of demyelination and atrophy, especially in the cerebellum, pons and medulla, the loss of pigmented cells in the substantia nigra pars compacta and the classic MSA microscopic oligondendroglial inclusion pathology.6 The neuronal loss is mainly present in the putamen, substantia nigra, pons, inferior olive, and cerebellum,7 which involves striatonigral or olivopontocerebellar structures. Loss of Purkinje cells affects mainly the vermis and gliosis and neurodegeneration are observed in the pons and cerebellum.

The etiology is still unknown and the origin of α-synuclein deposits and its role in the pathophysiology of the disease is still unclear. Other relevant aspects regarding the pathogenesis of MSA are neuroinflammation, mitochondrial dysfunction, oxidative stress and impaired autophagy and insulin resistance.8 The diagnosis of MSA is based on a medical history and neurological examination. Diagnostic criteria have been described1,9 and recently some of the critical issues have been reviewed in order to achieve a better sensitivity.10 Ancillary investigations help to support the diagnosis and complete the differential diagnosis, but specificity and sensitivity of current diagnostic criteria are not high enough. In this regard, biomarkers could help as to increase sensitivity and specificity.11 Although Magnetic Resonance Imaging (MRI) can be normal in MSA, especially in the early stages of the disease,12 volume measures of different regions has been performed to discriminate MSA from PD with 100% accuracy13 and some signs such as the “hot cross bun” can be suggestive.14 Due to the heterogeneity of clinical phenotype and lack of specific biomarkers, it is challenging to make a correct diagnosis of MSA.15

We here review the main features of a large retrospective series of cases with MSA, with the aim of describing clinical and neuropathological findings of MSA, to try to improve sensitivity and specificity of the current diagnostic criteria from early stages.

We retrospectively studied 51 MSA patients, who were referred to the Movement Disorders Unit, Hospital Universitario de Cruces (Barakaldo, Vizcaya), from 2000 to 2019, with parkinsonism, ataxia, autonomic failure or multiple system atrophy diagnosis and met MSA criteria. Established MSA criteria1 were applied by senior investigators to include patients in the study. Therefore, patients with probable or possible MSA-P or MSA-C were included. We also included three definite MSA who had post-mortem confirmed diagnosis. Neuropathological study was performed following the standard protocol from the Biobank. Other cases with adult onset ataxia or parkinsonism, not meeting criteria at the end of follow-up, despite of suspicion, were excluded from the analysis. The local Ethics Committee approved the research protocol (CEIC E09/25).

Data collection from the medical records included demographic and clinical data, and assessments such as neurological examination, UMSARS (Unified Multiple System Atrophy Rating Scale), and response to Levodopa (l-Dopa). Clinical features were assessed by means of UMSARS, including functional status (UMSARS I), neurological deficits (UMSARS II), autonomic examination (UMSAR III) and disability status (UMSARS-IV). We performed UMSAR scale at baseline in all patients and at 2 years follow-up in 28 patients. The age at onset was defined as the age of the first symptom, according to medical records. In case of death, date and cause of death was obtained. At the autonomic disorders unit, we systematically assess patients with MSA diagnosis by means of the UMSARS part I (a functional subscale to score symptoms and ability to undertake activities of daily living), UMSARS II (neurological motor evaluation), Autonomic examination and UMSARS IV (global disability scale).

Conventional neuroimaging based on MRI was performed at baseline for all patients to rule out other diagnoses and complete differential diagnosis, using a 3T scan. MRI images were available in 40 patients. In 31 patients with parkinsonism 123I-metaiodobenzylguanidine (MIBG) scintigraphy was performed to establish the differential diagnosis with idiopathic Parkinson's Disease (PD). We analyzed the early (15min) and late (4h) heart-to-mediastinum uptake ratios (HMR). We selected healthy controls who had undergone MIBG scintigraphy as routine screening for suspicion of neurovascular dysautonomia, not confirmed.

Fifty-one patients with a clinical diagnosis of possible (n=7) or probable (n=44) MSA according to the current diagnostic criteria1 were evaluated. Regarding subtype, 32 patients (62.7%) were diagnosed as MSA-cerebellar and 19 patients (37.3%) were diagnosed as MSA-parkinsonian. The mean age for all subjects included was 62±8 years. Overall, there were 23 women, with no differences in sex distribution when analyzing all the cases, whereas by subtype there were significant differences in groups classification (p<0.05), as there were more women in the MSA-P group. Globally, disease duration was 6.9±3.9 years, from the first symptom to death. Both groups (MSA-P and MSA-C) had similar ages of disease onset (60.65±6.873 vs 62.63±8.469 years old) and similar disease duration (7.4±4.1 vs 6.6±3.8 years), with no statistically significant differences (p>0.05). As expected, all MSA-C patients had cerebellar signs and 100% of MSA-P patients had parkinsonism. Nevertheless, both groups of patients developed concomitant cerebellar and parkinsonian symptoms. In detail, 89.5% of MSA-P patients developed cerebellar signs, whilst 63% of patients from MSA-C group had at least one parkinsonian sign. l-Dopa therapy was given to all the subjects, 56% of MSA-C and 84% of MSA-P had poor response to l-Dopa treatment.

In this series of MSA patients, 54% of the subjects died during the follow-up, 10 from the MSA-P group and 17 MSA-C subjects. Mean time (months) from diagnosis to death was 68.65 (±30) (Fig. 1). Regarding causes of death, pneumonia was the most common cause of dead (66%), specifically aspiration pneumonia (33%), and other infectious diseases (22%) were the main causes of death among patients with MSA.

Figure 1.

Cumulative survival rate. Kaplan–Meier curves illustrating survival rate in MSA cerebellar and parkinsonian phenotype.

(0.07MB).

Multiple system atrophy is a rare, fatal, neurodegenerative disorder with symptoms of autonomic failure associated with parkinsonism, cerebellar ataxia, or both. According to the consensus criteria, in our cohort 63% of the patients were classified as cerebellar subtype. The frequencies of cerebellar and parkinsonism-predominant cases in this population are different from the European registry (Table 5), where the majority of patients have MSA-P,16 whereas the olivopontocerebellar-predominant pathology was more frequent in Japanese MSA.17 The reason for this discrepancy could be that patients with cerebellar subtype are more commonly referred to our autonomic disorders unit, whereas parkinsonian phenotype is less frequently referred. The majority of patients with MSA-P were females, whereas among MSA-C there were no differences in sex, as described in other series.

We here show that MSA-C and MSA-P patients have similar clinical characteristics, with no differences at baseline in clinical evaluation by means of UMSARS total score. At baseline, patients with MSA-P showed worse scores in parkinsonian items compared to MSA-C patients. It is interesting to highlight here that at least one parkinsonian sign was reported in 63% of MSA-C patients, being rigidity, rapid alternating movements and posture statistically significate different when comparing MSA-P with MSA-C (p<0.05). MSA-P cases showed cerebellar signs in a large number of cases, indicating that probably cerebellar features are less specific. In this line, cerebellar atrophy was present not only in MRIs from MSA-C cases but also in some MSA-P cases. The neuropathological study in MSA-P cases showed subthalamic atrophy and gliosis in the substantia nigra and locus coeruleus although volumetric MRI imaging did not show alterations.

We have found a similar natural history in both MSA-C and MSA-P, with a wide overlap in signs and symptoms between both subtypes, from baseline. As previously described and expected, parkinsonism and Cerebellar forms are characterized by a predominance of parkinsonism with striatonigral pathology and cerebellar ataxia with predominant olivopontocerebellar pathology.18 Nevertheless, due to the overlap and the clinical heterogeneity in phenotypes the neuropathological study is still essential to make an accurate diagnosis. One limitation of the study is the retrospective analyses that could have introduced some bias. To minimize bias, we have described clinical assessments based on diagnostic tests and specific neurological scales. Data from the UMSARS-III (autonomic examination) are in concordance with previous descriptions,19 highlighting the importance of autonomic tests when suspected diagnosis. The clinical assessment was performed systematically and follow-up time was similar to that in other prospective studies.20 After two years of follow up, we observed a mild, but significant worsening, indicating the fast progression of the disease in both subtypes (p<0.05). Interestingly, MSA-C patients respond poorly to l-Dopa but there was a 28% of patients with mild l-Dopa response. In the MSA-P group only one patient had a good response to l-Dopa.23 This could be explained by the fact that among MSA-C, parkinsonian signs were mild, it could be hypothesized that mild striatal atrophy among MSA-C patients would contribute to a better response to l-Dopa in this group of patients. MSA-P patients group, showed more severe parkinsonian features, accompanied by cerebellar signs in the majority of patients, without improvement after l-Dopa treatment was initiated. Another limitation of the study could be the absence of genetic studies in this cohort, in future studies it would be interesting to investigate which genes are related with the disease or predispose to the condition. In our cohort there is a subgroup of patients with initially Pure Autonomic Failure (PAF) diagnoses and progression to MSA, as previously reported approximately 8% of patients developed MSA after 4 years follow-up.21 This group could increase the heterogeneity of the cohort, but we should certainly include them in the natural history of MSA to better characterize the clinical spectrum of patients with MSA diagnosis.10

The mean time to death from diagnosis was 5.6 years (Median 5.7 years±2.5). Compared with other series,20 time to death is shorter in this study. A selection bias might be present because patients were referred when a probable diagnosis is reached. Rate of mortality was the same among both subtypes, with no differences between MSA-C and MSA-P (Fig. 1). Overall, the main cause of death was pneumonia, associated with bronco-aspiration in many cases, due to early dysphagia.2 When analyzing hazard ratios for shorter survival, we observed that patients with severe UMSAR-I and UMSAR-II score at baseline had a shorter median survival. MSA-P has been reported as more aggressive in an European cohort22 (Table 5). In our cohort median survival and clinical status were similar both in MSA-P and MSP-C, as described in other cohorts.20

Neuropsychological tests were performed, but different tests were used over the years, not being comparable. This is one important aspect that we should include in future descriptions, following a necessary general consensus. Neuropsychological assessment should be included since the beginning of the disease in future prospective studies, to better characterize this group of patients.24 Another important aspect is the lack of peripheral or fluid biomarkers, not available at the moment,25,26 that could help to differentiate MSA cases from PD cases. In the absence of other biomarkers, MIBG scintigraphy is useful in some cases, however, MRI findings seems to provide higher accuracy to distinguish between PD and MSA. In our series, 9 patients with MSA diagnosis had cardiac sympathetic denervation. In some cases MIBG did not match with clinical nor neuropathological phenotype, as we previously described.27 In fact, one of the MSA-C cases was neuropathologically confirmed and MIBG showed cardiac denervation. In this case, there was a massive loss of neurons in the substantia nigra, with reactive gliosis and the cerebellum (dentate nucleus) was massively depopulated.

In conclusion, in our cohort there is a substantial overlap of parkinsonian and cerebellar features, from early stages of the disease. In fact, 89.5% of MSA-P patients developed cerebellar signs, and 63% of patients with MSA-C had at least one parkinsonian sign. Existing diagnostic criteria have a low sensitivity at the early stages of the disease probably due to this overlap, which agrees with mixed neuropathological findings. Efforts should focus on early stages, to increase sensitivity and specificity.

Rigidity, alternating movements of hands and posture were significantly different between groups at baseline, while there were no differences in autonomic examination. After 2 years of follow-up there were differences in all the items between baseline and after 24 months evaluation, except for action tremor and rigidity. As we here report, UMSARS performed at baseline and follow-up could be helpful to rate prognosis from early diagnosis. Finally, UMSARS, including autonomic function testing, could be useful to differentiate between MSA-C and MSA-P.

MCA and JGE have contributed with conception and design of the research. BT, KB, MA, IG have contributed with acquisition of data. RP, AM and MCA have contributed with the analyses. JGE supervised the manuscript. All authors discussed the results and implications and commented on the manuscript at all stages.

Patients had been informed that their data may be used for scientific purposes and had given their informed consent.

The data that support the findings of this study are available from the corresponding author, upon reasonable request.

No funding has been received for this work.

The corresponding author confirms on behalf of all authors that there have been no involvements that might raise the question of bias in the work reported or in the conclusions, implications, or opinions stated.

The authors declare that they have not conflict of interest.