Grey matter network measures predict disability and cognition

Data-driven, MRI network-based measures of co-varying grey matter volumes predict disability progression better than volumetric measures of grey and white matter lesion loads, a new study found [1]. Independent component analysis (ICA) of MRI can help select clinical MS study participants who are more likely to respond to treatment.

Baseline MRI and longitudinal clinical data were used from 988 participants of the randomised, double-blind, placebo-controlled ASCEND trial, which evaluated the effect of natalizumab on disease progression in secondary progressive MS. Spatial ICA was applied to baseline structural grey matter probability maps to identify co-varying grey matter regions. Correlations between ICA components and EDSS, 9-Hole Peg Test (9HPT), and Symbol Digit Modalities Test (SDMT) scores were computed.

A total of 15 clinically relevant networks of co-varying grey matter patterns were identified. Compared with conventional MRI measures, SDMT and 9HPT baseline scores correlated more strongly with ICA components, especially main basal ganglia components including the thalamus, caudate, putamen, and frontal and temporal lobes. EDSS correlated more closely with an ICA component involving cerebellum, brainstem, and temporal and parietal lobes (R=-0.11; P<0.001). EDSS progression was predicted by baseline caudate volume (HR 0.81; P<0.05). Descending SDMT scores were best predicted by 2 ICA components (HR 1.26; P<0.005; and HR 1.25; P<0.005). Two other ICA components predicted worsening of 9HPT scores (HR 1.30; P<0.01; and HR 1.21; P<0.05).

1. Colato E, et al. Predicting disability progression and cognitive worsening in multiple sclerosis with gray matter network measures. MSVirtual 2020, Abstract PS07.03.

 

© 2023 The Author(s). Published by Medicom Medical Publishers.
User license: Creative Commons Attribution – NonCommercial (CC BY-NC 4.0)

Posted on 25 November, 2020