• Marcello Moccia, Steven van de Pavert, Arman Eshaghi, Lukas Haider, Jonas Pichat, Marios Yiannakas, Sebastien Ourselin, Yi Wang, Claudia Wheeler-Kingshott, Alan Thompson, Frederik Barkhof, and Olga Ciccarelli.
• From the Department of Neuroinflammation, Queen Square MS Centre, NMR Research Unit, UCL Queen Square Institute of Neurology, Faculty of Brain Sciences (M.M., S.v.d.P., A.E., L.H., M.Y., Y.W., C.W.-K., A.T., F.B., O.C.), Centre for Medical Image Computing, Department of Medical Physics and Bioengineering (J.P., S.O.), and Translational Imaging Group, UCL Institute of Healthcare Engineering (F.B.), University College London, UK; Multiple Sclerosis Clinical Care and Research Centre, Department of Neurosciences (M.M.), Federico II University, Naples, Italy; Department of Radiology and Nuclear Medicine (F.B.), VU University Medical Center, Amsterdam, the Netherlands; and National Institute for Health Research University College London Hospitals Biomedical Research Centre (A.T., F.B., O.C.), UK.
• Neurology. 2020 Dec 1; 95 (22): e2965-e2976.

ObjectiveTo identify pathologic correlates of magnetization transfer ratio (MTR) in multiple sclerosis (MS) in an MRI-pathology study.MethodsWe acquired MTR maps at 3T from 16 fixed MS brains and 4 controls, and immunostained 100 tissue blocks for neuronal neurofilaments, myelin (SMI94), tissue macrophages (CD68), microglia (IBA1), B-lymphocytes, T-lymphocytes, cytotoxic T-lymphocytes, astrocytes (glial fibrillary acidic protein), and mitochondrial damage (COX4, VDAC). We defined regions of interest in lesions, normal-appearing white matter (NAWM), and cortical normal-appearing gray matter (NAGM). Associations between MTR and immunostaining intensities were explored using linear mixed-effects models (with cassettes nested within patients) and interaction terms (for differences between regions of interest and between cases and controls); a multivariate linear mixed-effects model identified the best pathologic correlates of MTR.ResultsMTR was the lowest in white matter (WM) lesions (23.4 ± 9.4%) and the highest in NAWM (38.1 ± 8.7%). In MS brains, lower MTR was associated with lower immunostaining intensity for myelin (coefficient 0.31; 95% confidence interval [CI] 0.07-0.55), macrophages (coefficient 0.03; 95% CI 0.01-0.07), and astrocytes (coefficient 0.51; 95% CI 0.02-1.00), and with greater mitochondrial damage (coefficient 0.31; 95% CI 0.07-0.55). Based on interaction terms, MTR was more strongly associated with myelin in WM (coefficient 1.58; 95% CI 1.09-2.08) and gray matter (GM) lesions (coefficient 0.66; 95% CI 0.13-1.20), and with macrophages (coefficient 1.40; 95% CI 0.56-2.25), astrocytes (coefficient 2.66; 95% CI 1.31-4.01), and mitochondrial damage (coefficient -12.59; 95% CI -23.16 to -2.02) in MS brains than controls. In the multivariate model, myelin immunostaining intensity was the best correlate of MTR (coefficient 0.31; 95% CI 0.09-0.52; p = 0.004).ConclusionsMyelin was the strongest correlate of MTR, especially in WM and cortical GM lesions, but additional correlates should be kept in mind when designing and interpreting MTR observational and experimental studies in MS.© 2020 American Academy of Neurology.

26 keywords...

hide…