CSF Myeloid Microvesicles Demonstrate Diagnostic Value in Predicting MS Disease Course

Roberto Furlan, MD, PhD

Recently published findings suggest that cerebrospinal fluid (CSF) myeloid microvesicles (MVs) have diagnostic and prognostic value in predicting neuroinflammation and microglial/macrophage activity in vivo. The results also support possible use in clinical practice to compliment MRI and clinical evaluation at the moment of multiple sclerosis (MS) diagnosis.

Senior author Roberto Furlan, MD, PhD, deputy director and head, Clinical Neuroimmunology Unit, San Raffaele Scientific Institute, and colleagues evaluated CSF myeloid MVs in 601 individuals with a diagnosis of neuroinflammatory, neurodegenerative, or no neurological disease at discharge or follow-up. Myeloid MVs were measured with flow cytometry as isolectin B4-positive events in fresh CSF.

Among the 601 individuals included in the analysis, 298 had a neuroinflammatory disease; 263 were within the MS spectrum; 176 had available follow-up data. Investigators observed a difference in CSF myeloid MV concentration in neuroinflammatory patients, patients affected by neurodegenerative diseases, and the control group (Kruskahl-Wallis; P <.0001). A post-hoc analysis showed statistically significant differences in median MV concentration between the neuroinflammatory group (median, 1615/mL) and the neurodegenerative/dementia group (325/mL; P <.0001), and between the neuroinflammatory group and the control group (CG; 270/mL; P <.0001). Notably, no differences were found between the neurodegenerative/dementia group and the CG.

READ MORE: CNM-Au8 Meets Primary End Point in Phase 2 Parkinson Disease, Multiple Sclerosis Studies

Among those with a diagnosis on the MS spectrum, patients with an active presence of gadolinium enhancement had higher CSF myeloid MV concentration (median, 3185/mL; n = 116) compared to the nonactive group (1340/mL; n = 79; Mann­–Whitney, P <.0001). Additionally, these results remained true when only considering patients with relapsing-remitting MS (RRMS; n = 66; 4085/mL vs 1450/mL; n = 65; Mann–Whitney; P <.0001), whereas in the other subgroups (radiologically isolated syndrome [RIS], clinically isolated syndrome [CIS], progressive MS [PMS], idiopathic central nervous system neuroinflammatory disease [INID], recurrent myelitis [RM]), differences did not reach statistical significance.

The performance of CSF myeloid MVs in diagnosis and prognosis supports their potential use in clinical practice, to complement MRI and clinical evaluation at the moment of diagnosis,” Furlan et al wrote. "CSF, despite requiring an invasive procedure to be collected, is routinely analyzed at the time of MS diagnosis. Therefore, CSF myeloid MVs analysis would not increase the burden of invasiveness in MS management."

Using a Jonckheere­–Terpstra test, investigators found an increase in median MV concentration with the increase of enhancing lesion number (P <.0001; Kendall = 0.365). Among just patients with RRMS without MRI disease activity, those who had a relapse in the previous 30 days had higher CSF myeloid MV levels (n = 11; 3740/mL) compared to patients without recent clinical disease activity (n = 54; median, 1370/mL; Mann–Whitney, P = .0137).

CSF MVs distinguished RRMS from the CG with an outstanding level of discrimination (area under the curve [AUC] = 0.939 [95% CI, 0.94-0.974]; P <.0001). In differentiating patients with RRMS from patients of the neurodegenerative/dementia group, the AUC was 0.91 (95% CI, 0.877-0.944; P <.0001).

Furlan et al also evaluated the diagnostic performance of CSF myeloid MVs in distinguishing asymptomatic lesions of inflammatory (RIS) or noninflammatory (UBL) origin. They found the concentration to be significantly higher among patients with RIS compared to UBL (median, 450 [n = 11] vs 1820 [n = 21]; Mann–Whitney U = 13.5; P <.0001). Furthermore, ROC curve analysis showed an outstanding level of discrimination, with an AUC of 0.942 (95% CI, 0.865-1.0; P <.0001).

Due to the small sample size for patients with PMS who had follow-up (n = 24), investigators divided patients by the median CSF myeloid MV concentration (low MVs group, <1030/ml; high MVs group, ≥1030/mL). At 36 months, PMS with high CSF myeloid MV levels displayed a shorter mean time to confirmed disability progression as compared to the low group (24.96 vs 32.31 months). Additionally, a log-rank test demonstrated a significant different in the survival distribution between groups (X2 = 3.826; P = .05).

REFERENCE

Gelibter S, Pisa M, Croese T, et al. Spinal fluid myeloid microvesicles predict disease course in multiple sclerosis. Ann Neurol. Published online July 3, 2021. doi:10.1002/ana.26154