Volume changes of thalamus, hippocampus and cerebellum are associated with specific CSF profile in MS

Abstract

Background: The underlying pathogenesis of surface-in grey matter abnormalities in MS, demonstrated by both neuropathology and advanced MRI analyses, is under investigation and it might be related to CSF-mediated mechanism of inflammation and/or damage. Objective: To examine the link of CSF inflammatory profile with the damage of three regions early-involved in MS and bordering with CSF: thalamus, hippocampus and cerebellum. Methods: In this longitudinal, prospective study, we evaluated, in 109 relapsing–remitting MS patients, at diagnosis and after 2-year follow-up, the association between the baseline CSF level of 19 inflammatory mediators and the volume changes of thalamus, hippocampus, cerebellar cortex and control regions (globus pallidus, putamen). Results: The multivariable analysis showed that the CXCL13 and sCD163 CSF levels at baseline were independent predictors of thalamus ([Formula: see text]; p < 0.001) and hippocampus ([Formula: see text]; p < 0.001) volume change after 2-year follow-up. These molecules, plus CCL25, IFN-γ and fibrinogen, were independent predictors of the cerebellar cortex volume loss ([Formula: see text]; p < 0.001). No independent predictors of volume changes of the control regions were found. Conclusion: Our results indicate an association between the CSF inflammatory profile and grey matter volume loss of regions anatomically close to CSF boundaries, thus supporting the hypothesis of a surface-in GM damage in MS.

“Ependymal‐in” Gradient of Thalamic Damage in Progressive Multiple Sclerosis

“…The observed significant association observed in MS TLs between gradient of microglia activation and gradient of neuro‐axonal decreased density strongly support this hypothesis. Moreover, this idea is supported by a recent longitudinal, prospective study of patients with relapsing–remitting MS at the time of diagnosis showing that CXCL13 and sCD163 CSF protein levels were independent predictors of thalamic ( R 2 model = 0.80; p < 0.001) and hippocampal ( R 2 model = 0.47; p < 0.001) volume change after 2‐year follow‐up 46 . In addition, chronically elevated TNF and IFN‐γ cytokine concentrations in the CSF in vivo in a rat model have been demonstrated able to induce microglial activation, subpial demyelination, and neuronal loss in the cortical layers underlying the subarachnoid space 12,15 .…”

“…Inflammatory/ cytotoxic factors produced by meningeal infiltrates, possibly related to the elevated B cell activity and released into the CSF, could cause damage not only to the regions adjacent to the meningeal TLS, but also diffusely across all the brain 20 and spinal cord 44,45 surfaces. It was previously suggested that B cells from patients with MS but not controls, in addition to secretion of pro-inflammatory molecules, such as TNF, lymphotoxin-α, IL-6, and GM-CSF, may also secrete one or more factors toxic to oligodendrocytes and neurons that can possibly directly contribute to demyelination in patients with MS. [8][9][10][11]16,[44][45][46] Here, we indeed found that, among a large panel of inflammatory mediators examined in the CSF from the same MS cases used for thalamic cell counts, only a specific subgroup of molecules related to both innate immune activity and lymphoid neogenesis (CCL19, CXCL10, and CXCL13) and major inflammatory factors (including sTNFR1, fibrinogen, IFN-γ, IL-2, and IL-10), strongly associated with the gradient of increased microglial density in the thalamus of progressive MS cases. As previously suggested, 16 CSF levels of TNF and its pro-inflammatory soluble receptor, sTNFR1, appear to represent good surrogate correlates of the "surface-in" GM pathology.…”

“…Moreover, this idea is supported by a recent longitudinal, prospective study of patients with relapsing-remitting MS at the time of diagnosis showing that CXCL13 and sCD163 CSF protein levels were independent predictors of thalamic (R 2 model = 0.80; p < 0.001) and hippocampal (R 2 model = 0.47; p < 0.001) volume change after 2-year follow-up. 46 In addition, chronically elevated TNF and IFN-γ cytokine concentrations in the CSF in vivo in a rat model have been demonstrated able to induce microglial activation, subpial demyelination, and neuronal loss in the cortical layers underlying the subarachnoid space. 12,15 These data strongly suggest a key role of intrathecal inflammation in the slow build-up of diffuse GM pathology, in particular, in neuronal loss and aberrant rounding morphology, that as previously shown may be mediated by TNF/TNFR1-mediated necroptosis (rather than apoptosis), possibly related to chronic meningeal inflammation.…”

Analysis of CSF D-Dimer to Identify Intrathecal Fibrin-Driven Autoimmunity in Patients With Multiple Sclerosis

“…The measurement of fibrinogen in CSF allows insight into the extent of BBB disruption in patients with RRMS and correlates well with the presence of MS pathology. 4 , 5 , 7 , 28 Acute BBB disruption (CSF/serum albumin quotient) in patients with MS is known to correlate with CNS inflammation injury and was recently linked to increased intrathecal neurofilament light chain, chitinase 3-like 1, metalloproteinase 2, and interleukin-6 levels. 29 , 30 However, although levels of fibrinogen in CSF correlate with levels of acute (or past) BBB dysfunction, they do not necessarily reflect the extent of fibrinogenic pathway activation, e.g., processing of fibrinogen into fibrin, which effectively exerts proinflammatory effects.…”

“…Future studies shall assess additional markers of coagulation cascade activation in the CSF such as thrombin and prothrombin and relate the parameters of fibrin-derived autoimmunity to other prognostic markers of MS disease activity, such as cytokines, 28,32 chemokines related to B-cell activity, 9 and markers of neuroaxonal degeneration (e.g., neurofilament light chain 33 ) to deepen the understanding of the interconnection of coagulation system and neuroimmunologic mechanisms.…”