The temporal dynamics and clinical relevance of choroid plexus measures in multiple sclerosis

Choroid plexus enlargement is a promising biomarker of disease activity in multiple sclerosis. However, longitudinal changes in choroid plexus volume and microstructural integrity remain unclear. This study investigated temporal changes in choroid plexus measures and their correlations with clinical disability and brain volume changes over 18 months and the entire disease duration. We recruited consecutive relapsing-remitting multiple sclerosis patients at treatment initiation who were then invited to come back for clinical, neuropsychological and brain MRI assessments at 6 and 18 months. Choroid plexus volume was measured using FreeSurfer and Gaussian Mixture Models on 3D-T1-weighted scans, and choroid plexus T1/T2 ratio was calculated from conventional 3D-T1- and T2-weighted images. Brain lesion, whole brain, grey matter, and white matter volumes were measured. Alternating conditional expectation algorithm was used to estimate trajectories of changes in choroid plexus measures over the entire disease course. Multiple linear regression and mixed effects models were used to investigate associations of choroid plexus measures with clinical and MRI measures. False discovery rate correction was applied. 422 RRMS patients were recruited [mean age: 40.8 years (SD 10.9), mean disease duration: 9.5 years (SD 17.4), median expanded disability status scale: 2.0 (IQR: 1.5-3.5); mean symbol digit modalities test score: 50.6 (SD 14.7), mean brief visuospatial memory test-revised score: 25 (SD7.6)]; 276 participants were studied at 6-months follow-up and 80 at 18-months. During the entire disease course, an initial increase in normalized choroid plexus volume was observed, followed by a plateau; T1/T2 ratio decreased initially, but then increased once the volume had stabilized. When examining changes in choroid plexus volumes over a median follow-up of 8.6 months, significant increases in both choroid plexus volumes [β = 0.45, standard error = 0.11, False discovery rate-corrected P < 0.001)] and T1/T2 ratios (β = 0.29, standard error = 0.14, False discovery rate-corrected P = 0.05) were observed. A higher baseline choroid plexus T1/T2 ratio was linked to a faster rate of decrease in normalized brain volume (β = -0.21, standard error = 0.08, False discovery rate-corrected P = 0.01) and deep grey matter volume (β = -0.25, standard error = 0.10, False discovery rate-corrected P = 0.03) over time. Higher baseline choroid plexus T1/T2 values were associated with worsening performance on brief visuospatial memory test-revised over time (β = -0.23, standard error = 0.10, False discovery rate-corrected P = 0.04). Changes in choroid plexus measures over time appear non-linear, with volumes increasing earlier in the disease course and T1/T2 ratios rising later. After a mean disease duration of 9.5 years, higher choroid plexus T1/T2 ratios, but not volume, predicted faster memory decline and whole brain and deep grey matter volume loss, underscoring the value of assessing choroid plexus microstructure, alongside volumes, in predicting clinical and MRI outcomes.