Similarity of functional connectivity patterns in patients with multiple sclerosis who void spontaneously versus patients with voiding dysfunction

Abstract

Aim: To investigate if Multiple Sclerosis (MS) lesion characteristics affect functional brain connectivity (FC) during bladder voiding.

Methods: Twenty-seven ambulatory female patients with MS completed our functional magnetic resonance imaging (fMRI)/urodynamic testing (UDS) platform. Individual fMRI activation maps were generated at initiation of voiding. FC patterns of these regions were calculated and compared. Similarity of the FC pattern from one patient relative to all others was expressed by a parameter FC_sim. A statistical analysis was performed to reveal the relationship of the existence of an enhancing brain lesion, the size of the largest lesion and the ability to void spontaneously to this FC similarity measure.

Results: FC_sim values were significantly lower for patients with an enhancing MS lesion (11.7 ± 3.1 vs 5.3 ± 2.1 P < 0.001). Lesion size smaller than 20 mm inversely correlated significantly with FC_sim (R = -0.43, P = 0.05). Patients with the ability to void spontaneously had a higher FC_sim value (12.0 ± 2.8 vs 9.3 ± 4.4 s, P = 0.08). Patients that exhibited a decrease of compliance also showed a significantly lower FC_sim value (11.3 ± 3.5 vs 4.7 ± 0.7, P < 1e-5).

Conclusion: FC connectivity analysis derived from an fMRI task-based study including repetitive voiding cycles is able to quantify the heterogeneity of connectivity patterns in the brain of MS patients. FC similarity decreased with maximum lesion size or the presence of enhancing lesions affecting the ability to void spontaneously.

Keywords: fMRI; functional connectivity; urodynamics; voiding.

© 2018 Wiley Periodicals, Inc.

Figures

FIGURE 1

A, Representative FC networks in the anatomical space of the human brain. Subject 1 has low FC_sim, subject 4 high FC_sim. B, Adjacency matrices representing the FC networks shown in (A). Columns and rows represent all the voxels in the brain that exhibited on average a significant increase in BOLD effect during the initiation of voiding. Positive correlations between the BOLD signal time courses of each voxel are shown in red, negative correlations are shown in blue. Each entry represented a bivariate measure of connective strength between two voxels. Positive correlations are shown in A as connections between brain regions. C, The correlation matrix of the adjacency vector matrix discussed in the text. Magnitude of values are indicated by the size of the circles. Subjects 1–5 and 19 have predominantly low values. FC similarity (FC_sim) defined as the column sums of the correlation matrix is consequently low for these subjects (shown in a bar plot at the bottom of the figure). Positive correlations are displayed in blue and negative correlations in red color (not to be confused with the customary BOLD signal color-coding for fMRI). Color intensity and the size of the circle are proportional to the correlation coefficients

FIGURE 2

Upper row: Left: Correlation between FC_sim and lesion size (size_lesion) Right: Box plot of difference in FC_sim for patients that showed decreased compliance. Lower row: Left: Results of Student t-tests differentiating MS patients with enhancing and non-enhancing lesions and, right, patients that are able to void spontaneously