Prognostic Factors and Treatment Efficacy in Spinal Cord Sarcoidosis: An Observational Cohort With Long-term Follow-up.

ABSTRACT

BACKGROUND AND OBJECTIVES: Spinal cord sarcoidosis is a rare manifestation of sarcoidosis with a consequent risk of neurologic sequelae for the patient. We investigated prognostic factors and efficacy of immunosuppressive treatments in a longitudinal cohort. METHODS: We retrospectively studied patients with spinal cord sarcoidosis followed between 1995 and 2021 in 7 centers in France. Patients with definite, probable, or possible spinal cord sarcoidosis according to the Neurosarcoidosis Consortium Consensus Group criteria and with spinal cord involvement confirmed by MRI were included. We analyzed relapse or progression rate with a Poisson model, initial Rankin score with a linear model, and change in the Rankin score during follow-up with a logistic model. RESULTS: A total of 97 patients were followed for a median of 7.8 years. Overall mean relapse or progression rate was 0.17 per person-year and decreased over time. At last visit, 46 (47.4%) patients had a loss of autonomy (Rankin score ≥2). The main prognostic factors significantly associated with relapse or progression rate were gadolinium enhancement (relative rate [95% CI] 0.61 [0.4, 0.95]) or meningeal involvement (relative rate [95% CI] 2.05 [1.31, 3.19]) on spinal cord MRI and cell count (relative rate [95% CI] per 1 log increase 1.16 [1.01, 1.33]) on CSF analysis. Relapse or progression rate was not significantly associated with initial Rankin score or Expanded Disability Status Scale. Tumor necrosis factor-α (TNF-α) antagonists significantly decreased relapse or progression rate compared with corticosteroids alone (relative rate [95% CI] 0.33 [0.11, 0.98]). Azathioprine was significantly less effective than methotrexate on relapse or progression rate (relative rate [95% CI] 2.83 [1.04, 7.75]) and change in Rankin score (mean difference [95% CI] 0.65 [0.23, 1.08]). DISCUSSION: Regarding the relapse or progression rate, meningeal localization of sarcoidosis was associated with a worse prognosis, TNF-α antagonists resulted in a significant decrease compared to corticosteroids alone, and methotrexate was more effective than azathioprine. CLASSIFICATION OF EVIDENCE This study provides Class IV evidence that in individuals with spinal cord neurosarcoidosis, TNF-α antagonists were associated with decreased relapse or progression rate compared to corticosteroids alone, but other therapies showed no significant benefit.