Abnormal blood-brain barrier permeability in normal appearing white matter in multiple sclerosis investigated by MRI

Abstract

Objectives: To investigate whether blood-brain barrier (BBB) permeability is disrupted in normal appearing white matter in MS patients, when compared to healthy controls and whether it is correlated with MS clinical characteristics.

Methods: Dynamic contrast-enhanced MRI was used to measure BBB permeability in 27 patients with MS and compared to 24 matched healthy controls.

Results: Permeability measured as K(trans) was significantly higher in periventricular normal appearing white matter (NAWM) and thalamic gray matter in MS patients when compared to healthy controls, with periventricular NAWM showing the most pronounced difference. Recent relapse coincided with significantly higher permeability in periventricular NAWM, thalamic gray matter, and MS lesions. Immunomodulatory treatment and recent relapse were significant predictors of permeability in MS lesions and periventricular NAWM. Our results suggest that after an MS relapse permeability gradually decreases, possibly an effect of immunomodulatory treatment.

Conclusions: Our results emphasize the importance of BBB pathology in MS, which we find to be most prominent in the periventricular NAWM, an area prone to development of MS lesions. Both the facts that recent relapse appears to cause widespread BBB disruption and that immunomodulatory treatment seems to attenuate this effect indicate that BBB permeability is intricately linked to the presence of MS relapse activity. This may reveal further insights into the pathophysiology of MS.

Keywords: Blood–brain barrier; DCE MRI; Disease modifying therapy; MRI; Multiple sclerosis; Perfusion MRI.

Fig. 1

A) On a T₂-weighted anatomical MRI sequence, with same geometry as the four acquired DCE perfusion slices, we placed regions of interest in periventricular normal appearing white matter (green), normal appearing gray matter (blue), thalamic gray matter (white), non-enhancing MS lesions (red) and contrast-enhancing MS lesions (white). B) Map of voxel-wise calculation of K^trans values for one MS patient. Marked with a red arrow is a visibly enhancing MS lesion. Note also that the scalp is showing relatively high values of permeability, due to the presence of extracranial vessels not expressing a blood–brain barrier.

Fig. 2

Tissue enhancement curves, Patlak plots and derived parameters for one MS patient (left column) and one healthy control (right column). “Time” on the x-axis of the Patlak plot, corresponds to ∫₀^tCa(τ)dτ/Ca(t). Abbreviations: NAWM = normal appearing white matter. NEL = non-enhancing lesion. V_b = blood volume.

Fig. 3

Permeability of the blood–brain barrier showing significantly higher values both in MS periventricular NAWM compared to cerebral white matter of healthy controls (p = 0.9 × 10^− 5; one-tailed T test), and in thalamic gray matter (p = 0.003; one-tailed T test). Black line = median. Boxes = 25% and 75% percentiles. Whiskers = sample range, outliers marked by circles. NB! All outliers were included in statistical analysis.

Fig. 4

In MS patients with clinical relapse within the last three months (n = 10), we found significantly higher permeability in periventricular NAWM (p = 0.004), thalamic gray matter (p = 0.004), and non-enhancing lesions (NELs) (p = 0.003; two-tailed T tests) compared to those with no relapse within three months. Immunomodulatory treatment coincided with significantly lower permeability in NEL (p = 0.01; two-tailed T test). Linear regression analysis showed that treatment (Beta = − 0.021 mL/100 g/min, p = 0.039) and relapse within the last three months (Beta = 0.034 mL/100 g/min, p = 0.001) were significant predictors of permeability in MS NEL. The overall model fit was R² = 0.37 and p = 0.001 with no significant interaction between the two parameters. Values for healthy controls are added for comparison. Black line = median. Boxes = 25% and 75% percentiles. Whiskers = sample range, outliers marked by a circle. Treatment = immunomodulatory treatment with IFN-beta or glatiramer acetate. Relapse = one or more relapses within the last three months.

Fig. 5

In MS patients receiving immunomodulatory treatment, we found the number of days since onset of last relapse to be a significant predictor of permeability in MS non-enhancing lesions (A), and periventricular normal appearing white matter (B). In the untreated group we found no such relation. Data points = mean value for each subject. Continuous line = logarithmic fit line; y = a − b∗log(x). R sq = R². IFN-beta = interferon beta. GA = glatiramer acetate.