Myelin Water Measurement in the Spinal Cord Evan P. Minty, 1 Thorarin A. Bjarnason, 2 Cornelia Laule, 3 and Alex L. MacKay 3,4 * The desire to monitor the spatial–temporal characteristics of myelination in the spinal cord (SC), in the context of patholog- ical change in demyelinating diseases or proposed neuroregen- erative protocols, has led to an interest in noninvasive image- based myelin measurement methods. We present one strategy: a magnetic resonance-based measure that capitalizes on the characteristics of T 2 relaxation of water compartmentalized within tissue. In this study, 32-echo relaxation studies for mea- suring the myelin water fraction (MWF) were applied in healthy control SC in vivo using a sagittal inversion recovery multiecho sequence, and findings were supported with supplemental studies in bovine SC samples in vitro. Mean human MWF varied according the level of the SC examined: cervical, thoracic, and lumbar MWF was found to be 21.8 (SD  2.1)%, 24.3 (3.6)%, and 11.4 (6.4)%, respectively. Noteworthy reductions were observed in areas consistent with the expected locations of the cervical and lumbar enlargements. Average bovine MWF was 30.0 (2.7)% in white matter and 8.2 (0.4)% in gray matter. The poten- tial applications of T 2 measurement in SC, both in characteriz- ing disease processes like multiple sclerosis and in monitoring neuroregenerative therapies, should encourage future research in this area. Magn Reson Med 61:883– 892, 2009. © 2009 Wiley- Liss, Inc. Key words: spine; myelin water; T 2 ; human; bovine The human spinal cord (SC) is a 0.5-m long cylindrical structure in the central nervous system (CNS) that occu- pies the upper two thirds of the vertebral canal, and serves as a liaison for impulses between the brain and the periph- eral nerves. Transverse axial sections through the SC re- veal a characteristic central “butterfly” distribution of gray matter (GM) within the surrounding white matter (WM) as depicted in Figure 1; the distinction between the two depends on the presence of myelin, a lipoprotein mem- brane that ensheathes axons and gives rise to the glistening white appearance of WM. Myelin’s primary role of poten- tiating action potential propagation is well known; its im- portance in that respect is exemplified by the devastating pathologies of demyelinating diseases such as multiple sclerosis (MS). The motivation to explore a magnetic resonance marker of myelination in the SC is two-fold. The SC is often involved in the demyelinating pathology of MS and SC. Magnetic resonance imaging (MRI) has recently been rec- ommended as a paraclinical test for MS (1,2). Meanwhile, demyelination has been suggested as a potential strategy toward creating a permissive extracellular environment for axonal regeneration after SC injury (3). In both cases, mag- netic resonance markers of myelination in the SC are of interest in monitoring disease pathology and potential re- generative therapies. Myelin and T 2 Relaxation Indirect measurement of myelin is possible with MRI. T 2 relaxation aspires to resolve multiexponential T 2 decay data into its constituent exponential components, each of which is attributable to water protons within a distinct diffusion restricted relaxation environment in tissue. The T 2 component with the shortest relaxation time (20 ms) is believed to originate from water trapped within the bilayers of the myelin sheath (4,5); this “short” relaxation time is believed to arise due to the large interfacial contact area in the multilamellar myelin compartment. The frac- tional contribution of the signal from this short T 2 compo- nent relative to the total area under the T 2 distribution is called the myelin water fraction (MWF) (5). Magnetic res- onance studies have amassed substantial evidence that the short T 2 component is specific to myelin. Its presence was noted in the myelinated trigeminal and optic nerves of the garfish, but not in unmyelinated olfactory nerve of the same species (6). In vitro studies in a fixed human brain have shown a strong quantitative correlation between MWF and Luxol Fast Blue staining for myelin (7–9). Other authors have also noted that changes in the T 2 distribution are the most effective way to distinguish inflammation from myelin loss (10,11), suggesting that T 2 relaxation provides a specific correlate of myelination during patho- logical change. The majority of work studying MWF in human CNS tissue has been in the brain and, to date, few studies have been performed in human SC. Wu et al. (12) performed test–retest validation in human controls, Laule et al. (13) monitored MWF changes in MS patients, and Kozlowski et al. (9) correlated MWF with Luxol Fast Blue staining for myelin. Although these studies have shown utility in mea- suring MWF in SC, superior–inferior variation along the cord has not been investigated, but would be expected given varying contributions of white and GM structures to the image slice thickness along the length of the SC as depicted in Figure 1. 1 Department of Physics, University of Alberta, Edmonton, Canada. 2 Department of Electrical & Computer Engineering, University of Calgary, Calgary, Canada. 3 Department of Radiology, University of British Columbia, Vancouver, Can- ada. 4 Department of Physics & Astronomy, University of British Columbia, Van- couver, Canada. Grant sponsor: the Multiple Sclerosis Society of Canada. Grant sponsor: the Natural Sciences and Engineering Research Council (NSERC). *Correspondence to: Alex MacKay, Department of Physics and Astronomy, 6224 Agricultural Road, University of British Columbia, Vancouver, British Columbia, V6T 1Z1 Canada. E-mail: mackay@phas.ubc.ca Received September 14, 2007; revised October 12, 2008; accepted Novem- ber 24, 2008. DOI 10.1002/mrm.21936 Published online 3 February 2009 in Wiley InterScience (www.interscience. wiley.com). Magnetic Resonance in Medicine 61:883– 892 (2009) © 2009 Wiley-Liss, Inc. 883