Early strong intrathecal inflammation in cerebellar type multiple system atrophy by cerebrospinal fluid cytokine/chemokine profiles: a case control study

All patients were examined at the Department of Neurology at Kyushu University Hospital, Japan, between 1 January 2005 and 30 June 2013. CSF samples from 32 patients with MSA-C or SCA and 15 control participants with other non-inflammatory neurological diseases (OND) were examined in the study. For diagnosis, we defined MSA-C according to the second consensus statement on the diagnosis of MSA [7]. Briefly, patients with a sporadic, progressive, adult-onset (>30 years) disease characterized by autonomic failure involving urinary incontinence, erectile dysfunction and orthostatic hypotension, poorly levodopa-responsive parkinsonism, and cerebellar syndromes including gait ataxia, dysarthria, limb ataxia, or cerebellar oculomotor dysfunction were enrolled. Patients predominantly with cerebellar ataxia were designated MSA-C. Clinical diagnosis of SCA was made on the basis of the diagnostic criteria for hereditary SCA [8]. Diagnostic guidelines for hereditary SCA comprise: (1) age at onset of symptoms >18 years; (2) predominantly progressive cerebellar ataxia with disease duration of >1 year; (3) cases with a family history of the presence of a similar disorder or with confirmed pathological genotypes, or a history of unexplained gait disturbance in first- and second-degree relatives; (4) exclusion of secondary ataxia caused by cerebrovascular disease, tumor, alcoholism, vitamin B1 or B12 deficiency, folate deficiency, drugs, neurosyphilis, multiple sclerosis, paraneoplastic cerebellar degeneration, immune-mediated cerebellitis, and hypothyroidism. We enrolled 12 hereditary SCA cases including two with SCA type 6, one with SCA type 8, two with SCA type 31, and seven with unknown mutations. The OND group included four patients with cervical spondylotic radiculomyelopathy, four with lumbar radiculopathy, four with normal pressure hydrocephalus, and one each with myopathy, cerebral vein malformation, and psychosomatic disorder. MSA patients’ clinical severity was assessed using the Unified Multiple System Atrophy Rating Scale (UMSARS) [9]. The demographic characteristics of the participants at CSF collection are summarized in Table 1. Disease duration was significantly longer in SCA patients compared with MSA-C patients, while age at disease onset did not differ significantly between the two diseases.

CSF samples were obtained from all patients by non-traumatic lumbar puncture, centrifuged within 30 min at 800 rpm (100×g) at 4 °C for 5 min, and the liquid phase of the CSF was stored at −80 °C until use. The levels of 27 cytokines/chemokines and growth factors in the liquid phase of the CSF, namely, interleukin (IL)-1β, IL-2, IL-4, IL-5, IL-6, IL-7, IL-9, IL-10, IL-12 (p70), IL-13, IL-15, IL-17A, interferon (IFN)-γ, TNF-α, C-X-C motif ligand (CXCL)8/IL-8, CXCL10/interferon-inducible protein-10 (IP-10), C-C motif ligand (CCL)2/macrophage chemoattractant protein-1 (MCP-1), CCL3/macrophage inflammatory protein (MIP)-1α, CCL4/MIP-1β, CCL5/RANTES, CCL11/eotaxin, granulocyte colony-stimulating factor (G-CSF), granulocyte-macrophage colony-stimulating factor (GM-CSF), platelet-derived growth factor (PDGF)-bb, basic fibroblast growth factor (FGF), vascular endothelial growth factor (VEGF), and IL-1 receptor antagonist (IL-1ra), were measured, as described previously [10, 11]. The Bio-Plex Cytokine Assay System (Bio-Rad Laboratories, Hercules, CA) was used according to the manufacturer’s instructions. The concentrations of cytokines/chemokines and growth factors were calculated by reference to a standard curve for each molecule derived from serially diluted concentrations of standards assayed in the same manner as the CSF samples. The detection limit for each molecule was determined by the recovery of the corresponding standard (calculated by: (observed concentration)/(expected concentration) × 100), and the lowest values with more than 70% recovery were set as the lower detection limits. The lower and upper detection limits and the detection rates are shown in Additional file 1: Table S1. No samples were beyond the upper detection limits. Some samples, especially in the OND samples, were below the lower detection limits.

All magnetic resonance images (MRI) were acquired with a 1.5 T or 3 T imager (Vision or Symphony; Siemens, Erlangen, Germany; Achieva, Philips, Best, the Netherlands). Sagittal T1-weighted images were used to measure the length of the pontine base. Imaging parameters for T1-weighted images were as follows: repetition time (TR)/echo time (TE) = 400–518/10–12 m/s; field-of-view (FOV) = 240 × 267–275 mm; and acquisition matrix = 256–512 × 192. Transverse T2-weighted images were used to evaluate the hot cross bun sign (HCBS). Imaging parameters for T2-weighted images were: TR/TE = 2650–4993/80–128 m/s; FOV = 230–240 × 256–267 mm; and acquisition matrix = 256–512 × 184–373. The length of each part of the brainstem was measured: (1) vertical width of vermis; (2) horizontal width of vermis on the line through the posterior medullary velum; (3) width of pontine base; (4) horizontal width of medulla oblongata on the line through the obex (Additional file 1: Figure S1). The degrees of HCBS, which is commonly used as a disease marker for MSA-C in the pons, was classified into four grades: Grade 0, no abnormal signals; Grade 1, faint hyper-intense anteroposterior line compared with horizontal line; Grade 2, definite HCBS on a single slice; and Grade 3, prominent HCBS on two or more sequential slices [12]. These measurements/classifications were performed by a trained neuro-radiologist without knowledge of the patients’ clinical information.

Data are presented as mean ± SEM. The Fisher’s exact probability test was used for comparisons of the detection rates of cytokines/chemokines and growth factors in each group, followed by Bonferroni’s correction for multiple comparisons. The statistical significance of the differences in CSF cytokine values among MSA-C, SCA, and OND patients was determined by one-way analysis of variance (ANOVA) with Bonferroni’s correction for multiple comparisons. For differences in cytokines with a p value <0.05 after correction, the Tukey test was used to determine significant differences between pairs of groups. Heatmap and clustering analyses for the correlations between each CSF cytokine were performed using JMP Pro software (ver. 12.2.0; SAS Institute, Cary, NC). Briefly, variables for each disease type were put into a multivariate platform to create correlation matrices comparing the combinations of all variables. A heatmap was created for the correlations and examined using cluster analyses. Heatmap analyses for correlations between CSF cytokine levels with disease duration, UMSARS, and MRI measurements were determined using Pearson’s correlation coefficient analysis and one-way ANOVA. Correlation analyses between CSF cytokine levels and CSF grouping were performed using Pearson’s correlation coefficient analysis and one-way ANOVA, followed by the Fisher protected least significant differentiation (PLSD) multiple comparison test. All statistical analyses were performed using JMP Pro software (ver. 11.0.0; SAS Institute Japan, Tokyo, Japan). The threshold for significance was set at p = 0.05.