Measurement of Volumetric Lesion Load in Multiple Sclerosis: Moving from Normal- to Dirty-Appearing White Matter

Sandy Cheng-Yu Chen; Hsiao-Wen Chung; Michelle Liou

It is generally accepted that conventional T2-weighted MR imaging (T2WI) is sensitive in revealing macroscopic multiple sclerosis (MS) lesions. Nonetheless, the definitive diagnosis of MS remains clinical (1). Although MR imaging remains the best diagnostic test for the workup of MS, several studies have shown only modest correlations between the clinical neurologic deficits and lesion count measured by T2WI (2). Therefore, conventional T2WI techniques appear inadequate to characterize MS lesions fully, because that sequence is not specific for tissue abnormalities such as acute edema, demyelination, gliosis, and axonal loss, which share similar hyperintensite appearances at T2WI. On the other hand, other MR images may show better tissue characterization. For example, MS lesions with T1 hypointensity usually indicate axonal loss or transient acute edema (3). Gadolinium-enhanced T1-weighted imaging (T1WI) allows the separation of active MS lesions from inactive ones by showing enhancement due to the increased blood-brain-barrier permeability of acute inflammatory lesions. Nonetheless, these techniques have not successfully balanced sensitivity to MS lesions and accuracy in characterizing subtypes of tissue damage.

More recently, investigators have examined MS lesion burden that exists at microscopic levels by using innovative MR techniques. Specifically, normal-appearing white matter (NAWM) in MS patients was found to have significantly lower magnetization transfer ratio (MTR) than that of healthy control subjects. These findings coincided with a pathology report (4) in which up to 72% of white matter lesions that appeared macroscopically normal were abnormal at the microscopic level. One study showed that microscopic MS lesions in NAWM may portend development of new macroscopic lesions, suggesting NAWM as prelesional white matter changes (5). Similarly, other studies suggested that these early microscopic MS lesions may further evolve to become subtly visible at T2WI before they develop into full-blown acute demyelinating plaques, which are readily seen at conventional MR imaging. These ill-defined MS lesions, which occur mainly in the deep and periventricular white matter, have been described as dirty-appearing white matter (DAWM) at T2WI (6).

Although one may argue that some of these DAWM findings are actually convalescent demyelinating MS lesions, the concept that DAWM should be measured as MS lesion burden should not be overlooked. Nevertheless, this important concept can be challenged by many issues. One of the issues is that the definition of DAWM is not as clear as that of NAWM in terms of the T2WI signal intensity changes, because DAWM may be difficult to differentiate from the normal range of variability in white matter myelination that may also appear “dirty” at fast spin-echo T2WI. Similarly, in acute MS lesions, the often-associated perilesional edema can be misclassified as DAWM bordering the acute lesions. Others may ask whether these DAWM MS lesions are useful in predicting patients’ clinical disability or outcome. To answer these questions confidently, a longitudinal correlational study between patients’ clinical scores and a robust quantitative analysis of MS lesion burden based solely on DAWM measurement is required.

In this issue of the AJNR, Ge et al have taken a further step toward analyzing the MTR behavior in DAWM in patients with relapsing-remitting MS (RRMS). Analysis of MTR histograms from different tissue categories showed clearly distinguishable patterns reflected in several statistical measures (eg, mean MTR and peak height), possibly allowing further understanding of MS lesions in RRMS. Because even in limited-size autopsy specimens MS lesions are generally highly heterogeneous and contain essentially all different stages of disease progress, imaging evaluation of MS by using a single parameter is certainly difficult. Thus, the reported data for MTR in DAWM may be particularly interesting to the neuroscientists, because it may add to our understanding of the complexity and course of MS and may potentially help monitor the response to new therapeutic regimens.

The study by Ge et al represents a classic example of applying advanced imaging techniques (in this case, MTR combined with elegant image segmentation for analysis) to a particular biomedical target (in this case, the DAWM in RRMS), hoping to answer specific questions that are essential in clinical neurology. The study performed by Ge et al should bring to the attention of neuroradiologists the increasing capability of advanced MR techniques to help visualize both the invisible and the uncertain on routine images. Changes in MTR likely indicate, for example, alterations in chemical exchange of bulk water associated with that bound to the macromolecular environment (7). Therefore, from an analysis of MTR in tissues that are supposedly in different stages of the disease progress (ie, NAWM, DAWM, and lesion plaques), it might be possible to unravel the ongoing pathologic processes related to macromolecular changes in those T2-uncertain tissues (ie, DAWM at T2WI). In fact, techniques other than MTR may be used for such a purpose. Specifically, just as diffusion tensor imaging (DTI) can be used to study traumatic axonal injury (8) in which there is trauma-induced disorientation of neuronal fibers, DTI (9) has also been shown to reveal significant changes of microstructural anisotropy in NAWM (9) and hence may likely be another effective technique to assess MS disease burden. Without doubt, technical developments in modern imaging modalities, together with clear understanding of the pathophysiology behind the diseases, are the essential elements pushing further improvements in the diagnostic efficacy of MR.

As with almost all research reports, however, ample room exists for further improvement. For example, the analysis employed by Ge et al compares histographic parameters (mean MTR, peak height) across multiple lesion types. Because of factors such as technical difficulties in precise control of the radio-frequency power for the magnetization transfer preparation pulses, the reported mean MTR values may not be directly comparable with those of other studies that used similar methods (6). On the other hand, the MTR peak height, which represents the percentage of tissue having MTR values corresponding to the value at the highest occurrence rate in a single tissue category (eg, DAWM), is at least subject to the choice of the number of groups when classifying MTR into a histogram. In such a case, the measure of kurtosis, in which a larger positive value indicates that the distributions are narrower and more “peaked” around its center, can be used as a histogram-free counterpart of the MTR peak height. More important, one needs to bear in mind that a histogram analysis only compares the statistical “trends” among different tissues rather than individually identifying the regional disease progress unambiguously. Even with the potential of cross-validated MTR analysis with T2WI findings and clinical features such as disease duration and the expanded disability status scores, definition of different lesion types continues to be determined manually. Unsupervised statistical or neural-network segmentation techniques, particularly sensitive to “peaks in a multivariate attenuation,” might be helpful in this regard, if multitechnique MR imaging combining T2WI, contrast-enhanced T1WI, MTR, and even DTI were used in the future for better tissue classification. Ideally, different lesion types should be visually discernable, which is perhaps the most difficult goal in radiologic diagnosis of MS because of the heterogeneity and different types of the disease.

References

↵
McDonald WI, Compston A, Edan G, et al. Recommended diagnostic criteria for multiple sclerosis: Guidelines from the International Panel on the Diagnosis of Multiple Sclerosis. Ann Neurol 2001;50:121–127
CrossRef PubMed
↵
Kalkers NF, Bergers E, Castelijns JA, et al. Optimizing the association between disability and biological markers in MS. Neurology 2001;57:1253–1258
Abstract/FREE Full Text
↵
von Walderveen MA, Kamphorst W, Scheltens P, et al. Histopathological correlate of hypointense lesions on T1-weighted spin-echo MRI in multiple sclerosis. Neurology 1998;50:1282–1288
Abstract/FREE Full Text
↵
Allen I, McKeown S. A histological histochemical and biochemical study of the macroscopically normal white matter in multiple sclerosis. J Neurol Sci 1979;41:81–91
CrossRef PubMed
↵
Catalaa I, Grossman RI, Kolson DL, et al. Multiple sclerosis: magnetization transfer histogram analysis of segmented normal-appearing white matter. Radiology 2000;216:351–355
PubMed
↵
Ropele S, Strasser-Fuchs S, Augustin M, et al. A comparison of magnetization transfer ratio, magnetization transfer rate, and the native relaxation time of water protons related to relapsing-remitting multiple sclerosis. AJNR Am J Neuroradiol 2000;21:1885–1891
Abstract/FREE Full Text
↵
Wolff SD, Balaban RS. Magnetization transfer contrast (MTC) and tissue water proton relaxation in vivo. Magn Reson Med 1989;10:135–244
CrossRef PubMed
↵
Arfanakis K, Haughton VM, Carew JD, et al. Diffusion tensor MR imaging in diffuse axonal injury. AJNR Am J Neuroradiol 2002;23:794–802
Abstract/FREE Full Text
↵
Guo AC, MacFall JR, Provenzale JM. Multiple sclerosis: diffusion tensor MR imaging for evaluation of normal-appearing white matter. Radiology 2002;222:729–736
PubMed