INTRODUCTION: The purpose was to evaluate the capacity of immunotherapies to normalize several putative inflammatory biomarkers of disease activity in OMS.

METHODS: BAFF and APRIL and a variety of chemokine ligands for CXCR3, CXCR5, CCR4, CCR7 were measured in cerebrospinal fluid (CSF) and serum by ELISA in 433 children: 296 OMS/109 controls/28 other inflammatory neurological disorders (OIND). A blinded scorer rated motor severity on the validated OMS Evaluation Scale.

RESULTS: In untreated OMS, elevations of CSF BAFF (+57%), CXCL10 (+2.7-fold), CXCL13 (+16.5-fold) and serum CCL22 (+55%) and CCL17 (+121%) were significant (P < 0.0001). In cross-sectional comparisons of treated groups, three different pharmacodynamic patterns emerged. Corticotropin (ACTH) was associated with lower production of CSF BAFF (-61%) and CXCL13 (-91%) and serum CCL22 (-51%) and CCL21 (-32%); response to corticosteroids was similar. In contrast, the IVIg group showed no such effects, but serum APRIL was higher (+6.7-fold), remaining normal in CSF. Rituximab-treated, not cyclophosphamide-treated, groups had higher serum BAFF (+2.6-fold), normalizing after B cell repopulation. Treatment with front-loaded ACTH, IVIg, and rituximab dropped CSF BAFF (-41%) and CXCL13 (-84%), and serum CCL22 (-34%), by 6-8 months, and diminished severity score (71%). By one week, serum CCL22 dropped 81% and CCL21 59%, with slow decline in CXCL13. Concentrations normalized over 12 weeks of ACTH tapering. Serum APRIL increased 2.9-fold after 1-2 g/kg IVIg monotherapy (P = 0.0003).

CONCLUSIONS: Striking distinctions in immunotherapy effects on BAFF/APRIL, CXCL13/CXCR5, CCL22/CCR4, and CCL21/CCR7 signaling were found. The contrasting pharmacodynamic patterns may offer utility as treatment biomarkers.