Neurodegeneration and energy depletion in MS: Links between tissue integrity loss and microvascular changes in white matter disease.

Disturbances in cerebral oxygen delivery and utilization are increasingly recognized as key features, and potential contributors to, neuronal damage in multiple sclerosis (MS). We recently discovered microvascular changes, which are thought to limit oxygen extraction, in MS-related white matter (WM) lesions compared to unspecific WM lesions. It is unclear whether such microvascular changes antedate demyelinating changes in MS, or whether they are secondary to subsequent disease changes, such as inflammation, tissue edema, and blood-brain barrier break-down. Diffusion kurtosis imaging (DKI) is sensitive to early MS-related disease changes, including altered myelin integrity, cellularity, and edema detecting deviations from Gaussian diffusion that serve as indirect markers of tissue integrity. The purpose of this study was to examine whether regions with altered DKI metrics overlap with regions with microvascular changes in MS patients, and to compare microvascular changes with parallel DKI changes in both MS-related lesions and unrelated WM lesions to learn more about their microstructural correlates. In this cross-sectional study, we assessed microstructural damage in 54 MS patients and 26 non-diseased symptomatic controls (SC) using diffusion kurtosis imaging (DKI) and explored the relationship between these findings and microvascular flow patterns and oxygen delivery measured by dynamic susceptibility contrast-enhanced MRI (DSC-MRI). MRI at 3T included three-dimensional (3D) T2-weighted fluid-attenuated inversion recovery (T2-FLAIR), 3D magnetization-prepared 2 rapid acquisition gradient-echo (MP2RAGE), post-contrast 3D T1-weighted images, DSC-MRI, and DKI. White matter lesions (WMLs) were manually outlined as MS-characteristic T2-FLAIR lesions, MS contrast-enhancing lesions and nonspecific lesions T2-FLAIR lesions. DKI-derived structural parameters, mean kurtosis (MK) and mean diffusivity (MD), were extracted from lesion masks and normal-appearing white matter (NAWM) and correlated with DSC-derived vascular parameters mean transit time (MTT), and the distribution of capillary transit times (CTH). Finally, an extended flow-diffusion model of oxygen transport was employed to evaluate tissue oxygen availability based on local blood flow and microvascular flow patterns. After adjusting for age and sex, NAWM in MS showed higher MD (+2.4 %, p = 0.01) and lower MK (-2.8 %, p = 0.01) compared with SC without concurrent changes in perfusion or oxygenation. Unspecific T2-FLAIR lesions demonstrated higher MD relative to NAWM (+13 %, p < 0.001) and reduced MK (-6.7 %, p < 0.001), but no microvascular impairment. By contrast, MS T2-FLAIR lesions showed more pronounced structural alterations, with higher MD than unspecific lesions (+13 %, p = 0.01) and markedly reduced MK (-16 %, p = 0.02), accompanied by increased CTH (+31 %, p = 0.02) and prolonged MTT (+32 %, p = 0.02), consistent with impaired oxygen extraction despite preserved CBF. Our findings indicate that, microstructural alterations, as assessed by DKI, are detectable in normal-appearing tissue before microvascular disturbances become evident on GE-DSC MRI. In MS lesions, however, microvascular flow heterogeneity coexists with tissue degeneration, suggesting inefficient oxygen extraction as a likely contributor to lesion pathology. These results emphasize the need for longitudinal studies to determine the temporal relationship between impaired oxygen extraction and disease progression.