Alpha-Synuclein Oligomers and Neurofilament Light Chain Predict Phenoconversion of Pure Autonomic Failure

Abstract

OBJECTIVE:

To explore the role of alpha-synuclein (αSyn) oligomers and neurofilament light chain (NfL) in cerebrospinal fluid (CSF) of patients with pure autonomic failure (PAF) as markers of future phenoconversion to multiple system atrophy (MSA).

METHODS:

Well-characterized patients with PAF (n=32) were enrolled between June 2016 and February 2019 at Mayo Clinic Rochester and followed prospectively with annual visits to determine future phenoconversion to MSA, Parkinson’s disease (PD), or dementia with Lewy bodies (DLB). ELISA was utilized to measure NfL and protein misfolding cyclic amplification (PMCA) to detect αSyn oligomers in CSF collected at baseline.

RESULTS:

Patients were followed for a median of 3.9 years. 5 patients converted to MSA, 2 to PD, and 2 to DLB. NfL at baseline was elevated only in patients who later developed MSA, perfectly separating those from future PD and DLB converters as well as non-converters. ASyn-PMCA was positive in all but 2 cases (94%). The PMCA reaction was markedly different in 5 samples with maximum fluorescence and reaction kinetics previously described in MSA patients; all of these patients later developed MSA.

INTERPRETATION:

αSyn-PMCA is almost invariably positive in the CSF of patients with PAF establishing this condition as α-synucleinopathy. Both NfL and the magnitude and reaction kinetics of αSyn PMCA faithfully predict which PAF patients will eventually phenoconvert to MSA. This finding has important implications not only for prognostication, but also for future trials of disease modifying therapies, allowing for differentiation of MSA from Lewy body synucleinopathies before motor symptoms develop.

INTRODUCTION

Parkinson’s disease (PD) and dementia with Lewy bodies (DLB) are the most common and well-known neurodegenerative disorders associated with abnormal aggregation and neuronal accumulation of alpha-synuclein (αSyn), with the classic pathologic substrate of Lewy bodies.^1-3 PD is characterized by tremor, rigidity, and bradykinesia, often accompanied by autonomic dysfunction, and sleep disorders.^{1, 2} DLB is the second most common form of dementia and is characterized by cognitive decline with prominent fluctuations, hallucinations, sleep disorders, and parkinsonism.³ In contrast to PD and DLB, multiple system atrophy (MSA) is a rare neurodegenerative disorder with an estimated point prevalence of 3.4 to 4.9 cases per 100,000 population, translating to an estimated 11,100 to 16,000 patients currently affected with MSA in the United States.^4-6 MSA is a rapidly progressive and universally fatal disorder, characterized by motor dysfunction that may predominantly involve parkinsonism (MSA-P) or cerebellar impairment (MSA-C), along with prominent autonomic failure.⁷ It is also associated with αSyn aggregation, but in contrast to PD and DLB, the pathologic hallmark of MSA is glial cytoplasmic αSyn inclusions.^{8, 9}

More recently, pure autonomic failure (PAF) has been added to the spectrum of synucleinopathies, and is clinically characterized by orthostatic hypotension (OH), anhidrosis, bladder dysfunction, and constipation.^10-12 Some patients with PAF survive for decades without clinical central nervous system involvement.¹³ However, the frequent presence of REM sleep behavior disorder (RBD) and anosmia in PAF, along with reduction of catecholamines in the spinal fluid, is indicative of subclinical involvement of central structures including noradrenergic and dopaminergic neurons.^14-16 Moreover, many patients with PAF are known to phenoconvert to one of the synucleinopathies with motor or cognitive involvement (PD, DLB, and MSA).^17-19 We have been able to define clinical markers of future phenoconversion, including severe bladder dysfunction, the presence of subtle motor signs, and the presence of RBD, but clinical prediction of phenoconversion remains imperfect, leaving clinicians and patients with considerable uncertainty about the prognosis and the possibility of harboring a rapidly progressive, terminal disease like MSA.^17-19 This situation highlights the need for reliable biomarkers that aid the prediction of future phenoconversion and the type of phenoconversion in patients with PAF.

Neurofilament light chain (NfL) in the spinal fluid reflects central axonal degeneration and has been reported as a valuable biomarker in a number of neurologic disorders characterized by rapid disease progression or considerable neuronal injury.^20-22 Given the rapid disease progression and neuronal loss in MSA compared to other synucleinopathies, it comes as no surprise that NfL has been reported to be increased in both MSA-C and MSA-P.^{21, 23, 24} We could recently show that NfL can differentiate even cases of early MSA from early PD and DLB with remarkable sensitivity and specificity.²⁵

Another marker of significant recent interest in synucleinopathies is the misfolded, aggregated form of αSyn in CSF. Using a protein misfolding cyclic amplification (PMCA) assay, we recently showed that the product of the αSyn-PMCA assay is biochemically and biophysically distinct between MSA and PD.²⁶ Furthermore, we could demonstrate that this assay in CSF allows for faithful differentiation of MSA from Lewy body synucleinopathies.^{26, 27}

It was therefore logical to measure these powerful markers in prospectively collected CSF samples of clinically well-characterized and longitudinally followed patients with PAF, in order to explore their value in predicting future phenoconversion of PAF into MSA or one of the Lewy body synucleinopathies. The presence of misfolded αSyn would also further establish PAF as α-synucleinopathy.

METHODS

Study Design

In a prospective study design, we followed patients with PAF longitudinally in order to determine future phenoconversion to MSA, PD, and DLB. CSF from well-characterized patients with PAF was collected at baseline for biomarker studies.

Participants

Patients were enrolled as a part of a prospective, longitudinal study of synucleinopathies (Mayo Longitudinal Synucleinopathy Biomarker Study, NS092625). Only spinal fluid samples collected at baseline were utilized for the data presented in this manuscript.

All patients underwent standardized autonomic function testing and were diagnosed with PAF by a Mayo Clinic autonomic disorder specialist. Inclusion criteria included age between 30 and 80 years upon enrollment, presence of neurogenic OH (defined as OH and prolonged blood pressure recovery time following Valsalva maneuver), absence of clinical evidence of central neurodegeneration (parkinsonism, cerebellar ataxia, dementia), and absence of peripheral neuropathy.

Patients were excluded if they had potential alternative causes of neurogenic OH, including autoimmune, paraneoplastic, endocrine, and rheumatologic etiologies. Subjects were also excluded if they were pregnant or breastfeeding, scored 24 points or less on the Mini-Mental Status Examination, had a clinically significant or unstable medical or surgical condition that might preclude safe completion of the study or might affect the results of the study, or had taken any investigational products within 60 days prior to baseline.

Patients underwent annual clinical neurologic evaluation for evidence of central neurodegeneration; if the latter was suspected, patients were referred for additional testing and subspecialty evaluation as appropriate for confirmation. MSA, PD, and DLB were diagnosed based on published criteria.^28-30

Standard Protocol Approvals, Registrations, and Patient Consents

The study was approved by the Mayo Clinic Institutional Review Board, and written informed consent was obtained from all participants.

Annual Clinical and Laboratory Study Assessments

A medical and neurologic history was obtained from all subjects, and all underwent a comprehensive general and neurologic examination. Medications that could potentially bias evaluations were held for 5 half-lives prior to neurologic assessments, autonomic testing, and spinal fluid collection. All subjects underwent MRI of the head, blood draw for supine and standing plasma catecholamines, and standardized autonomic function testing including autonomic reflex screen and thermoregulatory sweat test. In order to quantify autonomic deficits, the Composite Autonomic Severity Score (CASS), a validated instrument to quantify the overall severity and distribution of autonomic failure based on standardized autonomic testing, was derived.³¹ Autonomic symptoms were assessed using the Composite Autonomic Severity Scale, COMPASS-select.³²

CSF Collection and Analysis

After completion of clinical, laboratory, and imaging assessments, CSF was collected via routine spinal tap. A lumbar spinal needle was placed in the subarachnoid space via a standard posterior, intervertebral approach between lumbar level 2 and 5; the specific level was determined individually for each patient based on anatomical considerations.

After collection of 5cc of CSF for safety assessments (cell count, protein), 10cc of CSF were collected into separate tubes for biomarker studies. CSF collection for baseline assessments took place between 6/29/2016 and 2/21/2019. CSF was placed on ice immediately after collection and hand-delivered to the laboratory for further processing. After centrifuging at 10,000rpm for 10 minutes at 4°C to remove any potential blood contamination, CSF was aliquoted into cryotubes and transferred to a −80° C freezer until the day needed for the biomarker assays. Assays were performed with the technologist blinded to clinical information.

NfL was measured using the Uman Diagnostics (Umea, Sweden) enzyme-linked immunosorbent assay (ELISA) kit according to the manufacturer’s instructions. Standards and test samples were measured in duplicate and the average of the two measurements used for quantitative analyses. Samples were diluted 1:2 with kit diluent. Absorbance (450 nm) was measured using an Omega Fluorescence Base microplate reader (BMG, Ortenberg, Germany). NfL levels were quantified using a third order polynomial fit and measured in pg/ml.

The presence of αSyn oligomers was determined using PMCA taking advantage of the seeding-nucleation process of αSyn aggregation where misfolded oligomers seed the polymerization of monomeric protein.³³ In principle, using a cyclical process, monomeric protein in excess is combined with the study sample containing minute amounts of misfolded oligomers and incubated to induce growth of polymers. The sample is then subjected to a mechanical force to break down the polymers, multiplying the number of seeds, resulting in an exponential increase in the number of seeds.

Monomeric protein was purified and characterized as previously described.^{26, 27} ASyn monomers at a concentration of 1 mg/mL in 100mM piperazine-N,N’-bis(ethanesulfonic acid) at pH 6.5 in 500mM NaCl were placed in black 96-well plates in the presence of 5μM concentration of thioflavinT (ThT) at a final volume of 200μL. 40μL of CSF sample was added to each well. Samples were then subjected to cyclic agitation for 1 minute at 500 rpm followed by 29 minutes without shaking at 37°C. The increase in ThT fluorescence over time was monitored at excitation of 435nm and emission of 485nm daily for 15 days using a microplate spectrofluorometer (Gemini EM; Molecular Devices). Each sample was run in duplicate and the average of the maximum ThT fluorescence (arbitrary units, AU) measured during the 15 days of the assay for each sample used for quantitative analyses.

Statistical Analysis

Descriptive statistics were used to describe demographic, clinical, and autonomic characteristics by disease group, including mean, standard deviation (SD), median, interquartile range (IQR), frequencies, and percentages as appropriate.

Given the relatively small numbers in the converter groups, group differences in the spinal fluid markers between MSA-converters, PD/DLB converters, and non-converters were compared using non-parametric tests. Kruskal-Wallis test was used for comparing multiple groups with pairwise post-hoc comparisons using Dunn test. Mann-Whitney U-test was used for a priori comparisons of only two groups. All statistical tests were 2-sided, and p values <0.05 were considered statistically significant. Analysis was performed using SPSS (IBM SPSS Statistics 25).

RESULTS

Participants

A total of 32 patients with PAF were included in this study. Demographic, clinical, and autonomic characteristics of these patients are summarized in Table 1. Patients were predominantly male. Patients had experienced symptoms consistent with PAF for a median of 6.2 years. Bladder symptoms were common, male erectile dysfunction was invariably present. The majority of patients had a history consistent with RBD and admitted to hypo- or anosmia. There was an invariably high autonomic symptom burden and objective evidence of severe autonomic failure. Supine norepinephrine (NE) was low and rose to only modestly higher values in the upright position.

Table 1.

Demographic, clinical, and autonomic patient characteristics

Category	PAF (all)	MSA converters	PD/DLB converters	Non-converters
Number	32	5	4	23
Age, years	65.9±6.9	62.4±7.5	69.3±6.1	66.1±6.9
Sex (%) male	25 (78)	4 (80)	2 (50)	19 (83)
Disease duration, years	6.2 (IQR 4.3-9.5)	3.9 (IQR 3.5-5.3)	5.7 (IQR 4.5-6.9)	7.4 (IQR 4.8-10.2)
Bladder symptoms (%)	26 (81)	5 (100)	3 (75)	18 (78)
Male erectile dysfunction (%)	25 (100)	4 (100)	2 (100)	19 (100)
Constipation	20 (63)	4 (80)	4 (100)	12 (52)
REM-sleep behavior	20 (63)	4 (80)	4 (100)	12 (52)
Impaired sense of smell	20 (63)	1 (20)	3 (75)	16 (70)
Subtle motor signs	12 (38)	4 (80)	3 (75)	5 (22)
Mild cognitive symptoms	7 (22)	0 (0)	0 (0)	7 (30)
COMPASS-select	43.1±12.3	41.4±8.9	42.0±7.6	43.7±13.8
CASS	8.0±1.4	7.0±0.7	7.8±1.0	8.3±1.5
TST (% anhidrosis)	94.5 (IQR 58.4-99.1)	95.5 (IQR 61.1-99.4)	93.7 (IQR 82.9-95.0)	95.3 (IQR 57.7-99.1)
Norepinephrine (supine), pg/ml Normal range: 70-750	93 (IQR 37.5-157)	154 (IQR 141-228)	101 (IQR 79-150)	48 (IQR 30-152)
Norepinephrine (upright), pg/ml Normal range: 200-1700	154 (IQR 44-270)	304 (IQR 217-430)	201 (IQR 155-275)	87 (IQR 40-206)

Characteristics of all PAF patients combined, as well as separately for those who later converted to MSA, those who converted to PD or DLB, and those without conversion to date. Normally distributed numeric variables are shown as mean and standard deviation, not normally distributed variables as median and IQR. Frequencies are reported as number and category percentage.

Follow-up and Phenoconversion

Patients were followed for a median of 3.9 years (IQR 2.7-4.5). A total of 9 patients converted to either MSA (n=5), PD (n=2), or DLB (n=2) during follow up. 3 of the MSA converters were MSA-C, the other 2 developed MSA-P. Duration from baseline to confirmed phenoconversion ranged from 1 to 2 years in MSA converters, and from 1 to 3 years in PD and DLB converters. Those who converted to MSA had shorter disease duration at the time of enrollment, were less likely to report an impaired sense of smell, and had higher supine and upright plasma NE. Subtle motor signs were more common in all converters.

Two patients with a clinical diagnosis of MSA conversion came to autopsy, and both were pathologically confirmed as MSA. One patient who had not phenoconverted came to autopsy, and that patient had Lewy body pathology limited to medulla and pons.

CSF Markers - NfL

NfL in the CSF was markedly elevated in 5 patients, all of which later converted to MSA. All other samples showed NfL levels within the range we have previously established for normal controls and patients with Lewy body synucleinopathies (Fig 1).²⁵ None of those patients converted to MSA, but 4 of those later converted to either PD or DLB.

Figure 1. Neurofilament light chain in CSF.

PAF patients are shown on the right; those who later converted to MSA are indicated in red, those who later converted to PD or DLB are shown in green. In order to provide context to NfL levels in patients with established MSA, PD, DLB, and controls (CON), we have included findings in these patient groups (reported previously).²⁵ There was perfect separation of samples from patients who later developed MSA from all other samples. NfL was not different in samples from those who later developed PD or DLB compared to patients with stable PAF.

As expected, the finding of higher NfL levels in future MSA converters compared to other samples was statistically highly significant (p<0.001, Mann-Whitney). When separating patients into 3 groups (MSA converters, PD converters, non-converters), MSA converters remained significantly different compared to the other subgroups, but there was no difference between PD/DLB converters and non-converters (p=0.817, Dunn).

The previously suggested best NfL cut-off value between MSA and PD/DLB patients of 1,400pg/ml was found to also hold true for predicting future phenoconversion of PAF to MSA with both sensitivity and specificity of 100%.²⁵

CSF Markers – ASyn oligomers

The PMCA assay was reactive in all but 2 cases. Of the reactive cases, 5 strongly separated from the other samples with markedly lower maximum ThT fluorescence that fell perfectly into the previously established range specific for MSA (Fig 2).²⁵ All 5 of those patients later developed MSA. ThT fluorescence of all other reactive samples was markedly higher and fell into the range that was previously established for PD/DLB. 2 of those patients indeed developed PD and 2 DLB during follow-up, none developed MSA.

Figure 2. Alpha-synuclein oligomers in CSF by PMCA.

PAF patients are shown on the right; those who later converted to MSA are indicated in red, those who later converted to PD or DLB are shown in green. In order to provide context to NfL levels in patients with established MSA, PD, DLB, and controls (CON), we have included findings in these patient groups (reported previously).²⁵ The previously described “MSA range” between 150IU and 2,000IU is shaded in pink. Samples of PAF patients who later phenoconverted to MSA perfectly separated from samples of those who later developed PD or DLB, and from non-converters.

As expected, the finding of lower maximum ThT fluorescence in future MSA converters compared to other samples was highly significant (p=0.001, Mann-Whitney). As with NfL, when separating patients into 3 groups (MSA converters, PD converters, non-converters), MSA converters remained significantly different compared to the other subgroups, but there was no difference between PD/DLB converters and non-converters (p=0.436, Dunn).

The assay was repeated for the two non-reactive samples using separate aliquots and they remained non-reactive. Clinical characteristics of the patients with non-reactive PMCA assay did not differ noticeably from the other patients. Both had laboratory evidence of severe autonomic failure with neurogenic OH, and both had urinary symptoms. One had a history suggestive of RBD and anosmia, but normal supine NE, while the other one did not have RBD or anosmia, but markedly low supine NE. No alternative etiologies for these patients’ autonomic failure could be identified.

The previously established “MSA range” of maximum ThT fluorescence between 150 and 2,000AU was found to also apply to predicting future phenoconversion of PAF to MSA, allowing for prediction of conversion to MSA with 100% sensitivity and specificity.

Further analysis of the reaction kinetics of the PMCA assay demonstrated notable differences beyond the magnitude of the reaction. Identical to what we have previously shown for MSA and PD/DLB, the start of polymerization and formation of a reaction plateau occurred invariably earlier in samples from patients who later developed MSA, both occurring in less than half the time compared to samples from patients who later developed PD or DLB (Fig 3).²⁵

Figure 3. PMCA assay kinetics.

Mean and SEM at all measurement timepoints during the PMCA assay in PAF patients later converting to MSA (PAF_MSA), PAF patients later converting to PD or DLB, or not converting (PAF_LB), and of the two non-reactive samples (PAF_NEG). Note early start of polymerization and early, lower plateau in MSA converter samples, and late start of polymerization and formation of a late plateau at high polymerization levels in non-converters and PD/DLB converters, which perfectly matches the previously reported kinetics in established synucleinopathies.^{25, 26}

DISCUSSION

PAF was first described by Bradbury and Eggleston in 1925, and was subsequently referred to as Bradbury-Eggleston syndrome or idiopathic orthostatic hypotension.^{34, 35} However, it was not until the 1990s that alongside the degeneration of central and peripheral autonomic neurons αSyn deposits and Lewy body formation was described in autopsies of patients with PAF, allowing for recognition of PAF as a limited synucleinopathy.^{36, 37} More recently, PAF has been identified as an early presentation of the classic motor and cognitive synucleinopathies as some patients presenting as PAF will later develop motor and/or cognitive symptoms and phenoconvert to PD, DLB, or MSA.^{17, 18} We and others have therefore studied the clinical characteristics of PAF patients prior to phenoconversion with the goal to identify clinical predictors of future conversion, and to differentiate the type of future phenoconversion.^17-19 In fact, a number of clinical predictors were identified which include predictors of future phenoconversion in general (such as early motor signs or presence of RBD), but also predictors of the specific phenotype of conversion (such as severe bladder dysfunction for MSA).¹⁹ Although certain combinations of these clinical characteristics can allow for a high prediction probability of conversion and conversion type, those clinical prediction models are far from perfect.¹⁹ This situation leaves patients diagnosed with PAF with considerable uncertainty about the future, and with the knowledge that they carry a relatively high risk of having the early signs of an eventually rapidly progressive and fatal neurodegenerative disorder. A biomarker that reliably predicts future phenoconversion or at least allows for differentiating those who will eventually develop MSA from those who will not would therefore be of tremendous value. However, such markers would not only have implications for prognostication, but also for the development of disease-modifying therapies, where early stratification would allow the opportunity to test such therapies at the earliest disease stages when treatment is most likely to be effective, and may delay or even prevent further disease manifestations.

Recently, we described and validated two CSF markers, NfL and αSyn oligomers detected via PMCA, as diagnostic biomarkers that confidently distinguish early MSA from healthy controls on the one hand, and MSA from Lewy body synucleinopathies on the other hand.^{25, 26} In the present prospective study we applied these markers to carefully phenotyped patients with PAF and followed these patients longitudinally with careful surveillance for signs of phenoconversion to MSA, PD, or DLB with the goal to explore the value of these markers in predicting or differentiating future phenoconversion. In an analysis blinded to clinical data we found that both markers provided clean separation of those patients who later developed MSA from all other patients. The markers did not differ between those who later developed PD or DLB and those who did not phenoconvert during the observation period. Specifically, all patients who later developed MSA had CSF NfL levels greater than 1400pg/ml, and αSyn oligomers detected utilizing PMCA in these patients all fell into the previously established “MSA range” between 150 and 2000AU; all other PAF samples fell well below the level of 1400pg/ml for NfL and all but the two non-reactive samples into the PD/DLB range (>2000AU) for PMCA.²⁵ Utilizing these cut-off values, the prediction of future phenoconversion to MSA was therefore perfect for both markers.

This separation of “premotor” or “prodromal” MSA cases from Lewy body synucleinopathies achieved by two independent CSF markers confirms these markers as valuable disease biomarkers for MSA that achieve better predictive certitude than any previously described clinical, autonomic, or biofluid marker, even before a clinical diagnosis of MSA is possible or even suspected. Combining these markers as a dual biofluid marker may not purport much additional diagnostic value at a first glance, except maybe in the scenario when one of the markers is borderline or the assay non-reactive; however, when it comes to making a diagnosis of a terminal disease, the added certainty of combining those markers may nevertheless add a reassuring level of certainty. In addition, NfL alone – while demonstrated to be a powerful marker in this clinical scenario – is a relatively nonspecific marker that is reflective of the tempo and distribution of neuronal degeneration in MSA, but has been reported to be increased in a number of neurodegenerative, neuroinflammatory, and neurotraumatic conditions.^{22, 38, 39}

In addition to the clear identification of future MSA converters, our PMCA data also provide compelling additional evidence that PAF represents an α-synucleinopathy in almost all cases considering that all but two PAF CSF samples triggered αSyn protein misfolding in our PMCA assay. The finding that all PAF but the MSA converters fall into the same range of maximum ThT-fluorescence and furthermore have identical PMCA reaction kinetics as PD and DLB cases, regardless if they later converted to PD or DLB, is similarly intriguing and suggests Lewy body pathology in all of these cases. But what about the two cases in whom the PMCA assay remained non-reactive? Both also had normal NfL levels and neither showed any evidence of conversion on follow-up. The assay was repeated and remained negative, making a technical cause unlikely. Those patients could harbor an alternative cause for autonomic failure, but both had additional indicators to suggest an underlying synucleinopathy (RBD, anosmia, low NE) and careful investigations for alternative causes, making that explanation also unlikely, but not impossible; a single case of “PAF” without synucleinopathy has previously been described.⁴⁰ Perhaps the most likely explanation is very limited, maybe predominantly peripheral αSyn pathology in these cases that escapes the sensitivity of the assay or is not reflected in CSF.

The question could be raised if eventually all patients with PAF would develop either PD or DLB if they were followed long enough, or if they lived long enough for the disease to spread to involve basal ganglia or cortex. This assumption would be supported by previous autopsy reports of PAF patients, since all but one report describe Lewy body pathology beyond the autonomic nervous system, with involvement of substantia nigra and the basal forebrain.^{11, 36, 41-43} However, even after long-term follow-up, only about a third of patients has been shown to phenoconvert.¹⁹ It is therefore plausible that Lewy body pathology can remain limited to selected areas, though this important topic requires further study.

The strengths of the study presented here include the novelty of applying recently developed synucleinopathy biomarkers to a limited, “premotor” synucleinopathy; the prospective collection of CSF samples; comprehensive longitudinal follow-up evaluations of all patients; meticulous clinical, autonomic, laboratory, and imaging phenotyping of all subjects; and the standardized, rigorously careful handling of samples in this single center study. Weaknesses include lack of a validation cohort, a thus far limited number of autopsy-confirmation of the clinical diagnosis of phenoconversion, and that follow-up of patients is still limited to 2-5 years, with additional phenoconversion therefore possible. That being said, our overall conversion rate of 28% is in line with conversion rates reported previously, ranging from 24 to 34%.^17-19 Furthermore, surveillance continues and an updated report will be provided after an additional 5 years of follow-up; based on our results we would anticipate that remaining phenoconversion will be exclusively conversion to PD or DLB. That would also be supported by the observation in previous studies that conversion to MSA occurred significantly earlier and within a narrower time window than conversion to PD/DLB.^{18, 19}

In conclusion, our findings demonstrate that two CSF markers, NfL and αSyn oligomers detected via PMCA, allow for faithful prediction of future phenoconversion to MSA in patients presenting with isolated autonomic failure. These markers, alone and in combination, possess higher accuracy for predicting MSA than any other currently available marker. This finding has important implications not only for prognostication, but also for future trials of disease modifying therapies in the synucleinopathies, allowing for differentiation of MSA from Lewy body synucleinopathies before motor symptoms develop.

Our data furthermore strongly solidify the concept of PAF as a synucleinopathy. Continued surveillance of the patients who have not converted to a motor or cognitive synucleinopathy and further autopsy studies should eventually answer the question if late development of PD or DLB is eventually the fate of all patients with PAF or if a truly “stable” form of PAF exists with αSyn misfolding remaining confined to autonomic structures.