Abstract
Pilz et al. (Fluids Barriers CNS 17:7; 2020) investigated how CSF CXCL13 concentrations are influenced by CXCL13 serum concentrations and blood-CSF barrier (BCSFB) function, comparing the impact of serum CXCL13 levels and Q(albumin) (CSF albumin/serum albumin) on CSF CXCL13 among patients with CNS inflammation categorized as CXCL13 negative, low, medium, or high. Among all CXCL13 groups, their results showed no correlation between CSF CXCL13 concentrations and serum CXCL13 or Q(albumin). The authors argue that, in contrast to other proteins, CXCL13 passage across the BCSFB does not occur, regardless of BCSFB function, and is instead solely influenced by intrathecal production. In contrast to the authors' findings, in our studies including both non-inflammatory neurological disorders (NIND; n = 62) and multiple sclerosis (MS) patients we observed a significant correlation between serum CXCL13 concentrations and CSF CXCL13 concentrations. We review several observations which may underlie these contrasting results, including (1) the impact of serum CXCL13 concentrations on CSF CXCL13 in patients with lower intrathecal CXCL13 production and thus lower CXCL13 concentrations (i.e. NIND and MS), (2) the proposed diffusion dynamics of the small molecule CXCL13 across the BCSFB, and (3) differing definitions of negative versus elevated CSF CXCL13 concentrations determined by an assay's relative sensitivity. In conclusion, we argue that for patients with moderately elevated CSF CXCL13 concentrations, serum CXCL13 concentrations influence CSF CXCL13 levels, and thus the appropriate corrections including incorporation of CSF/serum ratios and Q(albumin) values should be utilized.