Use of multiple biomarkers to improve the prediction of multiple sclerosis in patients with clinically isolated syndromes

      Background: The early identification of patients at high risk of Clinically Definite Multiple Sclerosis (CDMS) represents the main purpose of diagnostic criteria and of clinicians in everyday clinical practice.
      Objective: To investigate whether the incorporation of different biomarkers in a model with established MRI criteria improves the prediction of MS.
      Methods: We evaluated baseline clinical data as well as MRI, multimodal evoked potentials and cerebrospinal fluid (CSF) data of patients with a first demyelinating episode. We used discrimination and calibration characteristics and reclassification of risk categories to assess incremental utility of different biomarkers for CDMS prediction.
      Results: During follow-up (median 7.2 years), 127 of the 243 participants in our study (mean age 31.6 years) developed a second clinical attack (CDMS). In Cox proportional-hazards models adjusted for established MRI criteria, age at onset, number of T1 lesions and presence of CSF oligoclonal bands significantly predicted the risk of developing MS within 2 and 5 years. The C-statistic increased significantly when the three biomarkers were incorporated into a model with established MRI criteria, both at 2 years (C-statistic with biomarkers vs. without biomarkers, 0.74 vs. 0.69) and at 5 years (0.66 vs. 0.70). The use of multiple biomarkers led to a 29% net-reclassification improvement at 2 years (p < 0.001) and 30% at 5 years (p < 0.001).
      Conclusions: The simultaneous addition of several biomarkers improves the risk stratification for MS in patients with clinically isolated syndromes beyond that of a model that is based only on MRI criteria.