Clinical features and outcome of patients with autoimmune cerebellar ataxia evaluated with the Scale for the Assessment and Rating of Ataxia

Background and purpose This study was undertaken to assess the long-term outcome of patients with paraneoplastic and non paraneoplastic autoimmune cerebellar ataxia (ACA) using the Scale for the Assessment and Rating of Ataxia (SARA). Methods Patients with subacute cerebellar ataxia admitted to our institution between September 2012 and April 2020 were prospectively recruited. Serum and/or cerebrospinal fluid was tested for neural autoantibodies by indirect immunofluorescence on mouse brain, cell-based assays, and radioimmunoassay. SARA and modified Rankin Scale (mRS) score were employed to assess patients' outcome. Results Fifty-five patients were recruited, of whom 23 (42%) met the criteria for cerebellar ataxia of autoimmune etiology. Neural autoantibodies were detected in 22 of 23 patients (Yo-immunoglobulin G [IgG], n = 6; glutamic acid decarboxylase 65-IgG, n = 3; metabotropic glutamate receptor 1-IgG, n = 2; voltage-gated calcium channel P/Q type-IgG, n = 2; Hu-IgG, n = 1; glial fibrillary acidic protein-IgG, n = 1; IgG-binding unclassified antigens, n = 7). Thirteen patients were diagnosed with paraneoplastic cerebellar syndrome (PCS) and 10 with idiopathic ACA. All patients received immunotherapy. Median SARA score was higher in the PCS group at all time points (p = 0.0002), while it decreased significantly within the ACA group (p = 0.049) after immunotherapy. Patients with good outcome (mRS <= 2) had less neurological disability (SARA < 15) at disease nadir (p = 0.039) and presented less frequently with paraneoplastic neurological syndrome (p = 0.0028). The univariate linear regression model revealed a good correlation between mRS and SARA score both at disease onset (p < 0.0001) and at last follow-up (p < 0.0001). SARA score < 11 identified patients with good outcome. Conclusions Patients with idiopathic ACA significantly improved after immunotherapy. SARA score accurately reflects patients' clinical status and may be a suitable outcome measure for patients with ACA.