Advanced MRI measures discern disability levels better than standard techniques
Brain scans routinely help doctors diagnose MS, but the discrepancy between the amount of visibly damaged tissue and the severity of symptoms is so common, it has a name: the clinicoradiological paradox. Now, by applying advanced imaging techniques to a short stretch of spinal cord, researchers have improved their ability to match imaging results with physical disability levels in MS patients, according to a January 16 online report in Neurology (Oh et al., 2013).
"It's a very important paper, because it shows that by looking at the spinal cord in these innovative ways, we can derive measures that correlate well with patients' clinical status," says Paul O'Connor, a neurologist at the University of Toronto in Canada. If the results are validated by others, he says, advanced MRI imaging of the spinal cord could provide a more precise and less error-prone measure of disease severity than a standard examination. Furthermore, the approach holds potential to detect treatment responses and predict future disability.
The idea for the study sprang from a typical day in the MS clinic for Jiwon Oh, a neurology fellow at Johns Hopkins University in Baltimore, Maryland, and first author of the paper. Two patients with back-to-back appointments were indistinguishable by conventional MRI measures, but one experienced no symptoms other than mild numbness, whereas the other required a wheelchair. "Both patients had exactly the same lesion load but completely different levels of disability," Oh says. "There has to be something else going on in the spinal cord that better explains the level of disability" in the two patients, she remembers thinking.
Spinal cord imaging was on Oh’s mind. Most patients don’t undergo the procedure until they develop a constellation of symptoms such as numbness below the neck, difficulty moving arms or legs, or bladder problems. Oh was just finishing an extensive study that compared how well advanced quantitative imaging techniques stack up against conventional MRIs of the spinal cord in their capacity to mirror neurological dysfunction—and the new methods seemed to outperform the typical imaging methods (Oh et al., 2012).
In the brain, about half of MS lesions cause no apparent symptoms. But tissue damage in the spinal cord is assumed to be more tightly linked to disability. "There is so much redundancy in the brain … a lot of scars there may not cause symptoms," says Anthony Traboulsee, a neurologist at the University of British Columbia in Vancouver, Canada. "The spinal cord is like a fuse box. All the wires from the house are tightly packed into that area. The same size of scar in the spinal cord is more likely to cause symptoms."
Relatively few imaging studies in MS focus on the spinal cord, mostly because it is hard to image. "There's a lot more real estate" in the brain, Oh says, and it doesn't move as much during scans. The spinal cord is barely thicker than the thumb, and a breath, heartbeat, or pulsing artery can move it enough to blur an MRI image. Other interference can come from competing signals thrown off by the protective packaging of bone and cerebrospinal fluid. Extracting a precise signal of spinal cord tissue damage "is like trying to pick out a diamond in the midst of rubies and sapphires," she says.
Despite the potential obstacles, Oh was in a good position to scrutinize spinal cords for clinically relevant indicators of MS because of her recent experience testing advanced techniques. Standard MRI methods pick up areas of tissue injury, but they cannot gauge the degree or type of damage. "We use the basic MRI sequences all day, every day in clinical practice," Oh says, but "we all know that there are several significant limitations to conventional MRIs, even though they are extraordinarily useful" (see "More Than Meets the Eye").
For her study, Oh selected diffusion tensor imaging (DTI), an MRI technique that detects distinctive changes in the direction and amount of water movement in three dimensions. For instance, water tends to travel along intact axons but across damaged fibers. Another technique generated a measure called the magnetization transfer ratio (MTR), which quantifies how much magnetization is exchanged between free water and complex molecules, such as those in tissue, and is believed to be particularly sensitive to myelin damage and axon density (Bosma and Stroman, 2012). By comparison, standard MRI protocols detect static differences in water density, which may reveal the presence or absence of an MS lesion but without such detail. The techniques also spit out different kinds of data. Each pixel of a conventional MRI image typically classifies tissue as damaged or normal, but advanced MRI imaging assigns a wider range of quantitative values to the way water moves through tissue. DTI and MTR can convey more detail about microstructural changes in tissue that looks normal in conventional images (Fox et al., 2011).
Building upon her earlier research, which found a population-level effect, Oh tested the real-world clinical relevance of advanced spinal cord imaging by sorting patients using lesion counts and disability levels. She divided 124 people with MS into two groups: those with two or fewer lesions as shown in standard spinal cord MRIs and those with more than two. More than half of the patients had relapsing-remitting MS, just under a third had secondary progressive MS, and about 10% had primary progressive MS. Cervical spinal cord MRIs had been taken in each patient at the start of the original study. (Oh didn’t know the clinical status of each patient when she counted the lesions.)
She further subdivided the two lesion groups into high or low disability, using an Expanded Disability Status Scale cutoff score of 6, which marks the transition from walking independently the length of a short city block to needing a cane or other assistance. Both the low- and high-lesion groups included patients of low and high disability. To assess how well the quantitative spinal cord MRI measurements denote disability levels, Oh evaluated a small section of spinal cord just above the Adam's apple. "There's always some sort of motion artifact," she says, but images from this region tend to be high quality. The strategy also takes the potential clinical use into account: The cervical segment can be conveniently imaged with the brain in about 10 to 20 minutes more of scanning time, and other studies have shown that damage there reflects overall MS spinal cord tissue injury.
In both low- and high- lesion groups, the advanced MRI measures correlated with disability levels better than the lesion counts, she and her co-authors report in the new paper.
The cross-sectional study results argue for a prospective study to determine whether the advanced techniques will be useful in the clinic, says Adrienne Dula, a biomedical engineer at Vanderbilt University in Nashville, Tennessee. "They chose very practical techniques that can be tagged on to the end of a standard-of-care imaging session," Dula says.
"Having new MRI techniques might—a big might—give us a clue about how to monitor and intervene in the progression of disease," Traboulsee says. He estimates that the quantitative techniques will need to prove their merit in clinical trials with outcomes that extend 5 years or more. Along the way, the techniques might provide information about the evolution of MS, including potentially new information about the disease processes in the spinal cord compared to the brain.
"For a long time, people have struggled to understand why the number of lesions in the brain and the spinal cord and the amount of space they take up doesn't fully explain a patient's clinical disability," says co-author Daniel Reich, a neurologist and radiologist at the National Institute of Neurological Disorders and Stroke in Bethesda, Maryland. "People have thought if we could look at the spinal cord really well, we could understand why this person is as disabled as they are. In this paper, the new observation is that we can glean information relating to the disease by looking at measures beyond whether or not there is a focal lesion."
Debate rages in the literature about the exact nature of the microscopic damage detected by each tiny measurement of quantitative MRI techniques, such as myelin loss here or axon damage there. "Imaging with MRI doesn't give you easy answers," Reich says. "We don't have the kind of molecular specificity that you have when you do the pathology. All we can do is look at how water molecules move and change properties in the field. But that doesn't mean you can't understand the biology." That may be a clinicoradiological paradox clinicians, researchers, and patients can live with.
Key open questions
- Will the findings be replicated with other MS populations?
- What microstructural changes are the quantitative MRI measures detecting in the spinal cord? Local inflammation? Long-range axonal degeneration from damaged neurons in the brain? Some other physiological perturbation?
- Can quantitative MRI be used to predict the course of MS?
Thumbnail image on landing page. "Spinal Cord" traced by Time3000, 2006. Original version created by ExplicitImplicity. Released under Creative Commons Attribution-ShareAlike 2.5 Generic (CC-BY-SA 2.5) license.