Background: CSF-specific oligoclonal bands (CSF-OCBs) can be used for dissemination in time (DIT) in the 2017 multiple sclerosis (MS) diagnostic criteria. A cut-off of ≥2 CSF-OCBs was recommended but studies have suggested ≥3 CSF-OCBs may be superior.

Objective: To assess utility of ≥2 and ≥3 CSF-OCBs as a cut-off for MS diagnosis.

Methods: Paired serum and CSF-OCBs sent to the Walton Centre, UK between July 2018 and June 2020 were included. CSF-OCBs were assessed using isoelectric focussing and reviewed by two blinded raters. Case records were reviewed.

Results: Of 1334 paired serum and CSF-OCB requests, 945 cases had sufficient clinical information. More than 1 CSF-OCB was detected in 268/945(28%) cases. Of these, 252 had ≥2 and 230 had ≥3 CSF-OCBs. The sensitivity and specificity for MS with ≥2 and ≥3 CSF-OCBs were 91.7%, 91.2%, 90.2% and 93.8% respectively. Only 3/22 patients with 2 CSF-OCBs had MS. In 25% of patients, CSF-OCBs reduced time to MS diagnosis (median 437.5 days (28-1332)).

Conclusions: Although cut-offs of ≥2 or ≥3 CSF-OCBs performed similarly well, 2 CSF-OCBs were frequently seen with non-inflammatory pathology. Use of ≥3 CSF-OCBs for MS diagnosis should be considered. CSF analysis reduced time to MS diagnosis by approximately 14 months.