Characteristic diffusion tensor tractography. in multiple system atrophy with predominant cerebellar ataxia. and cortical cerebellar atrophy

Transcription

1 Characteristic diffusion tensor tractography in multiple system atrophy with predominant cerebellar ataxia and cortical cerebellar atrophy Yusuke Fukui, Nozomi Hishikawa, Kota Sato, Yumiko Nakano, Ryuta Morihara, Yasuyuki Ohta, Toru Yamashita, and Koji Abe* Department of Neurology, Graduate School of Medicine, Dentistry and Pharmaceutical Sciences, Okayama University, Shikatacho, Kitaku, Okayama , Japan *Corresponding author: Prof. Koji Abe Department of Neurology, Okayama University Graduate School of Medicine, Dentistry and Pharmacy, Shikata-cho, Okayama , Japan. Tel.: ; Fax: ; Word count: 3616 Short running title: Differentiation of MSA-C from CCA by tractography. Keywords: cerebellar ataxia, cortical cerebellar atrophy, diffusion tensor imaging, multiple system atrophy, olivopontocerebellar atrophy, tractography. Conflict of interest: The authors declare no financial or other conflict of interest.

2 Abstract Objective To determine whether diffusion tensor imaging (DTI) tractography analysis is a potential method for differentiating cerebellar ataxia patients with multiple system atrophy with predominant cerebellar ataxia (MSA-C) and cortical cerebellar atrophy (CCA). Methods Forty-one MSA-C patients (62.7 ± 8.1 years old, mean ± SD) and age- and gender-matched 15 CCA patients (63.0 ± 8.6 years old) were examined. Tractography was performed using the DTI track module provided in the MedINRIA version 1.9.4, and regions of interest were drawn manually to reconstruct an efferent fiber tract and two afferent fiber tracts via the cerebellum. Results Compared with CCA, MSA-C patients showed significant declines of fractional anisotropy (FA) values of afferent 1 and 2 (p<0.01 respectively) and a significant increase of the radial diffusivity (RD) value in afferent 1 (p<0.05). Receiver-operator characteristic curve analysis showed 85.7% sensitivity and 75.0% specificity of FA values in afferent 1(cutoff value: 0.476). Linear regressions showed strong correlations between FA value and disease duration in CCA patients (efferent 1, r=-0.466; afferent 2, r=-0.543; both p<0.05), and between the FA value and the ratio of the standardized scale for the assessment and rating of ataxia (SARA)/disease duration in MSA-C patients (afferent 1, r=-0.407; p<0.01).

3 Conclusion The present DTI tractography newly showed that the FA values of two afferent fiber tracts showed significant declines in MSA-C patients, and afferent 1 showed good diagnostic sensitivity and specificity. When combining the FA values of efferent 1 with disease duration, the present DTI tractography analysis could be useful for differentiating MSA-C and CCA patients. Introduction Multiple system atrophy (MSA) is a sporadic and progressive neurodegenerative disease that clinically manifests cerebellar ataxia, parkinsonism, autonomic failure, and pyramidal signs in some combinations (1). Multiple system atrophy with predominant cerebellar ataxia (MSA-C) is the most common subtype of sporadic cerebellar ataxia in Japan (2). The pathological features of MSA-C are characterized by the accumulation of α-synuclein in glial cells (3), and degeneration of the afferent pathway (i.e. olivo-cerebellar tract and ponto-cerebellar tract) is remarkable (4). On the other hand, cortical cerebellar atrophy (CCA) does not belong to MSA, but is another dominant type of sporadic cerebellar ataxia (5, 6), pathologically characterized by extensive Purkinje cell loss of the vermis, neuronal loss in the flocculonodular lobe, and

4 degeneration of the efferent pathway (2, 7, 8). Diffusion tensor imaging (DTI) is a neuroimaging technique that quantifies the anisotropy of water diffusion in normal and abnormal brain tissues, and provides evidence of a damaged white matter tract (9, 10). When DTI is combined with tractography, which estimates fiber tracts, individual fiber tracts in the living human brain can be visualized in three dimensions (11). Thus DTI and tractography can detect microstructural abnormalities through alterations in fractional anisotropy (FA), mean diffusivity (MD) and radial diffusivity (RD) (9, 12). However, there are only a few reports on DTI and tractography imaging for cerebellar ataxia (9, 13, 14). In the present study, therefore, we examined DTI and tractography, focusing on the efferent and afferent pathways in patients with cerebellar ataxia as a potential method for differentiating MSA-C and CCA. Patients and Methods Participants For the present study, we investigated sporadic spinocerebellar ataxia (SCA) in the

5 outpatient clinic of Okayama University Hospital from November 2009 to March Among all outpatients, 41 multiple system atrophy with predominant cerebellar ataxia (MSA-C) patients (21 male and 20 female) and 15 cortical cerebellar atrophy (CCA) patients (7 male and 8 female) were examined for the present analysis. There was no significant difference between the two groups in terms of disease onset age (MSA-C, 58.6 ± 8.7 years old, mean ± SD; CCA, 53.3 ± 15.1 years old), disease duration (MSA-C, 4.0 ± 2.7 years; CCA, 9.7 ± 9.4 years), and examined ages (MSA-C, 62.7 ± 8.1 years old; CCA, 63.0 ± 8.6 years old) (Table 1). All patients were assessed by the standardized scale for the assessment and rating of ataxia (SARA). SARA was higher in MSA-C patients (17.3 ± 8.5, * p<0.05) than in CCA patients (11.3 ± 2.9). In MSA-C patients, a hot cross bun sign was evident in the pons in 71.8% of patients and T2-hyperintensity of the middle cerebellar peduncle in 69.2% of patients, both of which were not positive in CCA patients (Table 1). The clinical diagnoses of MSA-C were based on the second consensus statement on the diagnosis of MSA (15). Then, in this study, we enrolled the MSA-C patients who stared to recognize the disease by cerebellar symptoms, and fulfilled criteria for probable MSA-C. The CCA diagnoses were based on diagnostic criteria according to Klockgether (6) are progressive ataxia, age at onset > 20years, no /(sub)acute disease onset, no established symptomatic cause, and no possible or probable MSA according to the Gilman criteria (15). In the present study, all CCA patients showed adult onset slowly-progressive cerebellar ataxia without a

6 family history. All patients were Japanese. The clinical information of patients with MSA-C and CCA is summarized in Table 1. Ethical permission for this study was provided by the Ethics Committee on Epidemiological Studies of the Okayama University Graduate School of Medicine, Dentistry and Pharmaceutical Sciences (approval #304), and written informed consent was obtained from all participants prior to enrollment in this study. MR imaging Magnetic resonance imaging (MRI) was performed with a 1.5 T MR scanner (Archieva, Philips, The Netherlands) using an 8-channel receive head coil. According to a previous report (16), an axial single-shot spin-echo echoplanar imaging sequence was used for DTI. The imaging parameters were as follows: repetition time/echo time, 8960/71 ms; matrix size, ; field of view, 224 mm; section thickness, 2.5 mm; intersection gap, 0 mm; section number, 60; b value, 1000 s/mm 2 ; motion-probing gradients, 15; flip angle, 90º; number of repetitions for averaging, 2; total acquisition time, 304 s. In addition to DTI, T1-and T2-weighted images were also obtained to exclude other neurological disorders. DTI processing and fiber tracking

7 Tractography was performed on a stand-alone MacOSX workstation using the DTI track module provided in the software MedINRIA version (17). Diffusion tensor was first estimated (background removal threshold, 600; high smoothing), and regions of interest (ROI) were then drawn manually to reconstruct an efferent fiber tract and two afferent fiber tracts via the cerebellum. Because the afferent fiber tracts of MSA-C patients and the efferent fiber tract of CCA patients are mainly impaired, the ROIs of cerebello-rubro tract (efferent 1) were placed over a red nucleus and superior cerebellar peduncle bilaterally (Fig. 1A, red). The ROIs of olivo-cerebellar tract (afferent 1) were placed on the inferior olivary nucleus and cerebellar cortex bilaterally (Fig. 1A, blue), and ponto-cerebellar tract (afferent 2) were placed on the pontine nucleus and middle cerebellar peduncle bilaterally (Fig. 1A, green). The fiber tracking started with defining ROIs on the high resolution T2-weighted images. The fiber tracking parameters were as follows: FA threshold, 300; smoothness, 20; minimum length, 10; sampling, 1. All placements of ROIs were performed by another rater who was blinded to clinical information. We obtained several quantitative parameters through these fiber tracking and calculations: mean fractional anisotropy (FA) values, mean diffusivity (MD) values, and mean radial diffusivity (RD) values. Statistical analysis

8 Continuous demographic and clinical data are presented as mean ± S.D. in the text and table. Statistical analyses were performed using statistical software (SPSS ; IBM, Armonk, New York, USA). After having checked for normality, we performed Mann-Whitney tests to compare FA, MD, and RD values between MSA-C and CCA patients. P-values < 0.05 were considered to be significantly different. Receiver-operator characteristic (ROC) curves were computed for both FA and RD values. Cutoff levels were selected from the ROC curves as the values with optimum sensitivity and specificity for this study. In order to examine the influence of disease-duration, SARA, SARA/disease-duration, FA, MD, and RD values on each fiber, we performed Spearman s rank correlation coefficient tests among MSA-C patients and CCA patients. The FA values of each fiber were assessed as a function of disease-duration (M) and SARA/disease-duration using linear regression analysis was calculated. Results All patients were examined with MRI. The demographic and clinical features of MSA-C and CCA patients are shown in Table 1. Fig. 1 demonstrates an efferent fiber tract (red)

9 and two afferent fiber tracts (blue and green) via the cerebellum in sagittal (Fig. 1B), coronal (Fig. 1C), and axial (Fig. 1D) sections. Imaging findings Fig. 2 shows a comparison of mean fractional anisotropy (FA), mean diffusivity (MD), and mean radial diffusivity (RD) values of an efferent fiber tract and two afferent fiber tracts in MSA-C and CCA patients. The FA values of efferent 1 showed no significant difference between MSA-C (0.50 ± 0.07) and CCA patients (0.47 ± 0.04). In contrast, the FA values of afferent 1 and afferent 2 were significantly reduced in MSA-C than CCA patients (afferent 1: MSA-C vs CCA, 0.45 ± 0.03 vs 0.50 ± 0.03; afferent 2: 0.40 ± 0.03 vs 0.43 ± 0.04, ** p<0.01 respectively) (Fig. 2A). On the other hand, the MD values (10-3 mm/s) of these three tracts were not significantly different between MSA-C (efferent 1: 2.19 ± 0.32, afferent 1: 1.94 ± 0.28, afferent 2: 2.08 ± 0.45) and CCA (2.27 ± 0.36, 1.82 ± 0.18, 2.22 ± 0.43) patients (Fig. 2B). The RD values (10-3 mm/s) of efferent 1 and afferent 2 were not significantly different between MSA-C (efferent 1: 0.51 ± 0.13, afferent 2: 0.53 ± 0.15) and CCA (efferent 1: 0.56 ± 0.10, afferent 2: 0.57 ± 0.13) patients. However, the RD value of afferent 1 was significantly higher in MSA-C patients (0.47 ± 0.11, * p<0.05) than in CCA patients (0.41 ± 0.12) (Fig. 2C).

10 ROC curve analysis To evaluate the diagnostic sensitivity and specificity of FA and RD values that were significantly different (Fig. 2A and C), ROC curves were computed for MSA-C vs CCA patients (Fig. 3). Evaluation of the FA values indicated that the sensitivity and specificity to discriminate MSA-C patients from CCA patients was 85.7% and 75.0%, respectively with a cutoff value of for afferent 1 (area under the curve (AUC): 0.834). In the same manner (cutoff: 0.396), the sensitivity and specificity of afferent 2 was 85.7% and 60.0%, respectively (AUC: 0.757). The cutoff value of efferent 1 was not set (Fig. 3A) due to a lack of difference in FA between MSA-C and CCA patients (Fig. 2A). Evaluation of the RD values revealed that the sensitivity and specificity of afferent 1 was 82.9 and 46.7 %, respectively (cutoff: 0.421, AUC: 0.671). The cutoff values of efferent 1 and afferent 2 were not set (Fig. 3B) due to a lack of difference in RD between MSA-C and CCA patients (Fig. 2C). Correlation and Regression Analyses To evaluate the disease-duration and SARA/disease-duration effects on DTI characteristics, the data of MSA-C and CCA patients were plotted to create a scatter diagram, and regression lines were calculated. The linear regressions of efferent 1 between the FA values and disease-duration revealed a strong correlation in CCA patients (diamonds and dotted line: r=-0.466,

11 * p<0.05) but not in MSA-C patients (open circles and solid line: r=-0.194, p=0.231) (Fig. 4A). In contrast, a significant correlation of the linear regressions of afferent 1 was also observed between the FA values and SARA/disease-duration in MSA-C patients (open circles and solid line: r=-0.407, ** p<0.01), but not in CCA patients (diamonds and dotted line: r=0.140, p=0.634) (Fig. 4B). The FA values of afferent 2 were strongly correlated with disease-duration in CCA patients (diamonds and dotted line: r=-0.543, * p<0.05) but not in MSA-C patients (open circles and solid line: r=0.020, p=0.903) (Fig. 4C). Discussion In the present study, an efferent fiber tract and two afferent fiber tracts were successfully reconstructed as DTI track images via the cerebellum (Fig. 1). Compared to CCA patients, the FA values of afferent 1 and 2 showed a significant reduction in MSA-C patients (Fig. 2). ROC curve analysis showed that the FA values of afferent 1 especially discriminated MSA-C from CCA patients (Fig. 3A, solid line). Although there was no difference in the FA value of efferent 1 (Fig. 2), the FA values of efferent 1 and afferent 2 were strongly correlated with disease

12 duration in CCA patients (Fig. 4A and 4C, dotted lines). Moreover, those of afferent 1 showed a significant correlation with the ratio SARA/duration in MSA-C patients (Fig. 4B, solid line). In MSA-C patients, a hot cross bun sign is frequently found in the pons as a cruciform hyperintensity on transverse T2-weighted images. The sign is mainly due to selective losses of myelinated transverse pontocerebellar fibers and neurons in the pontine raphe with preservation of the pontine tegmentum and corticospinal tracts (4, 18). On the other hand, CCA may be characterized by extensive Purkinje cell loss in the vermis and neuronal loss in the flocculonodular lobe (2, 7, 8, 19). However, autopsy reports showed that the cerebellar hemisphere could also be involved in CCA (19, 20), suggesting the pathological lesion of CCA both in the efferent and afferent fiber tracts, but predominantly with the efferent fiber (Fig. 4A, C). In spite of distinct neuropathological features (21), cerebellar symptoms of MSA-C and CCA are quite similar, making differential diagnosis difficult, especially at an early stage. Many DTI and tractography studies have been performed to develop accurate diagnostic biomarkers for neurological disorders (10, 11, 22), where decreasing FA value was mainly reported at the legions of neuronal/ axonal loss and myelin degeneration with a concomitant increase of the MD value. An isolated decrease of the FA value was observed at the legions where α-synuclein accumulated (3, 9, 23, 24). In the present study, two afferent fiber tracts of MSA-C patients showed such an isolated decrease of the FA value (Fig. 2A and 2B). In

13 addition, the FA values of afferent 2 had already decreased at an early stage of MSA-C (Fig. 4C, solid line), suggesting that α-synuclein-positive glial cytoplasmic inclusions (GCI) were present in the pons before evident neuronal loss. The present tractography study (Fig. 1-4) may detect such insidious pathological processes. The present study has a limitation, because we could not get the information on their genetic testing to establish sporadic adult onset ataxia of unknown etiology. However we checked their family history thoroughly to carry out precise diagnosis (5, 25). Second, deterministic tractography cannot resolve the crossing of projections, such as efferent 1 where fibers originating in the dentate nucleus cross inside pons to reach the opposite red nucleus. However we thought that it is high certainty that automatically selecting the fiber via the 2 ROIs. In that case, bilateral tracking was performed automatically. Third, in this study, the member of our patients was too small to apply corrections for multiple comparisons. In conclusion, the present study demonstrated different features of efferent and afferent fiber tracts when comparing MSA-C and CCA patients. In particular, the FA values of two afferent fiber tracts showed a significant decline (Fig. 2A), and afferent 1 showed good diagnostic sensitivity, specificity, and area under the curve (AUC) (Fig. 3A). Furthermore, the FA values of each fiber tract suggest that this diffusion tensor tractography analysis could be useful for differentiating CCA from MSA-C patients (Fig. 4).

14 Acknowledgments This work was partly supported by Grants-in-Aid for Scientific Research (B) , (C) 15K09316 and Challenging Research 15K15527 and Young Research 15K21181, and by Grants-in-Aid from the Research Committees (Mizusawa H, Nakashima K, Nishizawa M, Sasaki H, and Aoki M) from the Ministry of Health, Labour and Welfare of Japan. References 1. Gilman S, Low PA, Quinn N, Albanese A, Ben-Shlomo Y, Fowler CJ, et al. Consensus statement on the diagnosis of multiple system atrophy. J Neurol Sci. 1999;163(1): Tsuji S, Onodera O, Goto J, Nishizawa M, Study Group on Ataxic D. Sporadic ataxias in Japan--a population-based epidemiological study. Cerebellum. 2008;7(2): Wakabayashi K, Takahashi H. Cellular pathology in multiple system atrophy. Neuropathology. 2006;26(4): Schrag A, Kingsley D, Phatouros C, Mathias CJ, Lees AJ, Daniel SE, et al. Clinical usefulness of magnetic resonance imaging in multiple system atrophy. J Neurol Neurosurg Psychiatry. 1998;65(1): Abele M, Burk K, Schols L, Schwartz S, Besenthal I, Dichgans J, et al. The aetiology of sporadic adult-onset ataxia. Brain. 2002;125(Pt 5): Klockgether T. Sporadic adult-onset ataxia of unknown etiology. Handb Clin Neurol.

15 2012;103: Kanazawa I, Kwak S, Sasaki H, Mizusawa H, Muramoto O, Yoshizawa K, et al. Studies on neurotransmitter markers and neuronal cell density in the cerebellar system in olivopontocerebellar atrophy and cortical cerebellar atrophy. J Neurol Sci. 1985;71(2-3): Baloh RW, Yee RD, Honrubia V. Late cortical cerebellar atrophy. Clinical and oculographic features. Brain. 1986;109 ( Pt 1): Tha KK, Terae S, Yabe I, Miyamoto T, Soma H, Zaitsu Y, et al. Microstructural white matter abnormalities of multiple system atrophy: in vivo topographic illustration by using diffusion-tensor MR imaging. Radiology. 2010;255(2): Foerster BR, Carlos RC, Dwamena BA, Callaghan BC, Petrou M, Edden RA, et al. Multimodal MRI as a diagnostic biomarker for amyotrophic lateral sclerosis. Ann Clin Transl Neurol. 2014;1(2): Marek M, Paus S, Allert N, Madler B, Klockgether T, Urbach H, et al. Ataxia and tremor due to lesions involving cerebellar projection pathways: a DTI tractographic study in six patients. J Neurol. 2015;262(1): Nagesh V, Tsien CI, Chenevert TL, Ross BD, Lawrence TS, Junick L, et al. Radiation-induced changes in normal-appearing white matter in patients with cerebral tumors: a diffusion tensor imaging study. Int J Radiat Oncol Biol Phys. 2008;70(4): Ito M, Watanabe H, Kawai Y, Atsuta N, Tanaka F, Naganawa S, et al. Usefulness of combined fractional anisotropy and apparent diffusion coefficient values for detection of involvement in multiple system atrophy. J Neurol Neurosurg Psychiatry. 2007;78(7): Prakash N, Hageman N, Hua X, Toga AW, Perlman SL, Salamon N. Patterns of fractional anisotropy changes in white matter of cerebellar peduncles distinguish spinocerebellar ataxia-1 from multiple system atrophy and other ataxia syndromes. Neuroimage. 2009;47 Suppl 2:T Gilman S, Wenning GK, Low PA, Brooks DJ, Mathias CJ, Trojanowski JQ, et al. Second consensus statement on the diagnosis of multiple system atrophy. Neurology. 2008;71(9): Fujiwara S, Sasaki M, Kanbara Y, Inoue T, Hirooka R, Ogawa A. Feasibility of 1.6-mm isotropic voxel diffusion tensor tractography in depicting limbic fibers. Neuroradiology. 2008;50(2): N Toussaint, J C Souplet, P Fillard. MedINRIA: Medical Image Navigation and Research Tool by INRIA. MICCAI'07 Workshop on Interaction in medical image

16 analysis and visualization, Brisbane, Australia,. 18. Shrivastava A. The hot cross bun sign. Radiology. 2007;245(2): Ota S, Tsuchiya K, Anno M, Niizato K, Akiyama H. Distribution of cerebello-olivary degeneration in idiopathic late cortical cerebellar atrophy: clinicopathological study of four autopsy cases. Neuropathology. 2008;28(1): Tsuchiya K, Ozawa E, Saito F, Irie H, Mizutani T. Neuropathology of late cortical cerebellar atrophy in Japan: distribution of cerebellar change on an autopsy case and review of Japanese cases. Eur Neurol. 1994;34(5): Mascalchi M. Spinocerebellar ataxias. Neurol Sci. 2008;29 Suppl 3: Hesseltine SM, Law M, Babb J, Rad M, Lopez S, Ge Y, et al. Diffusion tensor imaging in multiple sclerosis: assessment of regional differences in the axial plane within normal-appearing cervical spinal cord. AJNR Am J Neuroradiol. 2006;27(6): Wenning GK, Tison F, Ben Shlomo Y, Daniel SE, Quinn NP. Multiple system atrophy: a review of 203 pathologically proven cases. Mov Disord. 1997;12(2): Nilsson C, Markenroth Bloch K, Brockstedt S, Latt J, Widner H, Larsson EM. Tracking the neurodegeneration of parkinsonian disorders--a pilot study. Neuroradiology. 2007;49(2): Schols L, Szymanski S, Peters S, Przuntek H, Epplen JT, Hardt C, et al. Genetic background of apparently idiopathic sporadic cerebellar ataxia. Hum Genet. 2000;107(2): Figure legends Fig. 1 ROIs of cerebello-rubro tract were placed over the red nucleus and superior cerebellar peduncle bilaterally (red: efferent 1), of olivo-cerbellar tract were placed on the inferior olivary

17 nucleus and cerebellar cortex bilaterally (blue: afferent 1), and of ponto-cerbellar tract were placed on the pontine nucleus and middle cerebellar peduncle bilaterally (green: afferent 2) (A). Tractography of the patient was reconstructed by an efferent fiber tract (red) and two afferent fiber tracts (blue and green) via the cerebellum in sagittal (B), coronal (C), and axial (D) sections. Fig. 2 Comparison of FA values (A), MD values (10-3 mm/s) (B), and RD values (10-3 mm/s) (C) of an efferent fiber tract and two afferent fibers tracts in MSA-C (dotted boxes) and CCA (diagonal striped boxes) patients. Note the significant declines of FA values of afferent 1 and 2 in MSA-C patients (** p<0.01 respectively), and the significant increase of the RD value of afferent 1 in MSA-C patients (* p<0.05). Fig. 3 ROC curve analysis of the FA values shows 85.7% sensitivity and 75.0% specificity to discriminate MSA-C from CCA patients with a cutoff value of of afferent 1 (area under the curves (AUC): 0.834) (A), in contrast to the lower sensitivity and specificity of afferent 1, 82.9 and 46.7%, respectively (cutoff: 0.421, AUC: 0.671) (B). Fig. 4 Scatter diagram and individual regression lines of MSA-C (open circles and solid lines) and CCA (diamonds and dotted lines) patients between the FA value and disease duration (month,

18 M) (A, C) or the ratio SARA/disease duration (B). Note the strong correlations between the FA value and disease duration in CCA patients in efferent 1 (A) and afferent 2 (C), and between the FA value and the ratio SARA/disease duration in MSA-C patients in afferent 1 (B).

19 Table 1 Demographic and clinical features of imaging study patients Subjects MSA-C CCA p-value n Gender (M/F) 21/20 7/8 p=0.763 a Disease onset (years) 58.6 ± ± 15.1 p=0.350 c Disease duration (years) 4.0 ± ± 9.4 p=0.055 c Examined age (years) 62.7 ± ± 8.6 p=0.901 b SARA 17.3 ± ± 2.9 *p=0.024 c Hot cross bun sign (%) **p<0.001 a T2-hyper intensity of middle cerebellar peduncle (%) **p<0.001 a a: χ 2 test b: t test c: Mann-Whitney test

20 Fig. 1 A B ROI efferent 1 ROI afferent 2 afferent 1 ROI C D afferent 2 efferent 1 afferent 2 afferent 1 afferent 1 efferent 1

21 Fig. 2 A ** ** FA B MD (10-3 mm/s) MD (10-3 mm/s) C * RD (10-3 mm/s) * 0 efferent 1 afferent 1 afferent 2

22 Fig A FA MSA-C vs CCA 0.8 Sensitivity B RD efferent 1 afferent 1 afferent Sensitivity Specificity

23 Fig. 4 A efferent 1 B afferent 1 C afferent 2 r=0.140 p=0.634 FA r= p=0.231 r= *p<0.05 r= **p<0.01 r=0.020 p=0.903 r= *p<0.05 duration (M) 0 SARA/duration 0 duration (M)