| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Neuroradiology |
1 From the Department of Radiology, University of Pennsylvania Medical Center, 3400 Spruce St, Philadelphia, PA 19104 (I.C., R.I.G., D.L.K., J.K.U., L.G.N., L.W., X.Z., L.J.M., J.C.M.), and MCP–Hahnemann University, School of Public Health (M.P.). From the 1998 RSNA scientific assembly. Received August 2, 1999; revision requested September 24; revision received December 8; accepted December 16. Supported in part by National Institutes of Health grants NS29029-03/04, 5M01-RR00040-34, NS37172, NS 34353; I.C. supported in part by a grant from the French Society of Radiology. Address correspondence to J.C.M. (e-mail: jmcgowan@seas.upenn.edu).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
MATERIALS AND METHODS: Conventional and MT magnetic resonance (MR) images were obtained in 23 patients and 25 healthy volunteers. Clinical tests for comparison with the MT histogram parameters included the Extended Disability Status Scale and the ambulation index. Lesion load calculated with T2-weighted MR images and whole-brain and white matter volumes were measured.
RESULTS: The location of the MT histogram peak and the mean MT ratio for NAWM were significantly lower in patients with MS than in control subjects. In longitudinal studies, the histogram peak location and mean MT ratio shifted in the direction of normal values as the duration of disease increased. A mean of 26.5% of the volume of new lesions identified on the later studies were demonstrated to have originated in NAWM corresponding to "lost" pixels on the histogram.
CONCLUSION: MT histogram analysis of NAWM, including longitudinal analysis, may provide new prognostic information regarding lesion formation and increase understanding of the course of the disease.
Index terms: Brain, MR, 10.121411, 10.121412, 10.121417 • Brain, white matter • Magnetic resonance (MR), magnetization transfer, 10.121412, 10.121417 • Sclerosis, multiple, 10.871
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Recently, there has been interest in the characterization of MS with newer MR techniques such as magnetization transfer (MT), with the overarching goal of increased specificity for diagnosis and prognosis. In earlier work (3), up to 72% of macroscopically normal white matter (WM) was shown to be diseased at the microscopic level (3). More recently, normal-appearing WM (NAWM) in patients with MS has been found (4–6) to be abnormal by means of analysis of both MR imaging and MR spectroscopic findings.
A typical approach for studies involving MT contrast is to compute the mean MT ratio (MTR) values from two-dimensional regions of interest, largely limiting the results to macroscopic disease that is apparent on MR images. A more recent approach (7) was designed to investigate the multifocal and diffuse extent of MS through the generation of histograms of MTR values over the total brain area. Results of this work indicated that the overall distribution of MT values is shifted toward lower values in patients with MS and that the peak height of the MTR histogram, which is thought to reflect the residual amount of normal brain tissue, is significantly lower in patients with MS than in healthy control subjects. The MTR histogram peak height exhibits a negative correlation with disease duration, and a negative correlation between whole-brain MTR histogram peak height and Extended Disability Status Scale (EDSS) score has also been shown in a population of patients whose disease was classified as relapsing-remitting MS (7–10). A negative correlation between lesion load as calculated from T2-weighted MR images and MTR histogram peak height also was found (10).
In the present study, we focused our investigation on WM that was judged to be normal on conventional MR images obtained with T2-weighted and intermediate-weighted sequences (the latter of which minimizes both T1 and T2 contrast) and obtained MTR histograms restricted to this tissue alone. Our hypothesis was that the distribution of MTR values in NAWM, as demonstrated with the NAWM MTR histogram, would reflect the presence and progression of the disease. We also tested the hypothesis that an MTR abnormality in NAWM would portend development of new lesions.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The magnitude of the MT effect was determined by calculating the MTR (4): MTR = [1 - (Ms/M0)] · 100%, where Ms refers to the intensity of a pixel on the MT image and M0 represents the intensity of the corresponding pixel on the control image. This ratio corresponds to the percentage of signal intensity loss due to MT. MTR maps were then produced wherein each pixel intensity value corresponded to the MTR at that location. MT imaging was conducted during two study periods: before and after installation of an imager software update that altered the MT pulse and resulted, during the second period, in reduced MT saturation due to a lower MT pulse amplitude. The pulse offset frequency was 2,000 Hz during period 1 and 1,200 Hz during period 2, when pulse amplitude was described in terms of an effective flip angle of 900°. Ten patients and 18 control subjects were examined in period 1, and 13 patients and seven control subjects were examined in period 2. Care was taken to compare MTR values and derived values only between patients and subjects imaged with identical protocols. We found no substantive difference in the qualitative MT effect between the two periods of the study.
All patients underwent a neurologic examination at the time of MR imaging. The Kurtzke EDSS, ambulation index, and duration of disease (defined according to the date of initial diagnosis) were determined for the patients.
As noted, a total of 25 healthy volunteers whose age covered the same range as that of the patients and who had no history of neurologic problems underwent an identical imaging protocol. Nine patients and six healthy subjects underwent a protocol of two conventional and MT MR imaging studies for the purpose of a longitudinal analysis. The time between the two studies varied from 12 to 32 months (mean, 24 months) for the patients and from 3 to 7 months (mean, 4 months) for the control subjects. All studies were approved by the institutional review board of the University of Pennsylvania, and informed consent was obtained from all patients and subjects.
The T2- and intermediate-weighted MR images were postprocessed for whole-brain segmentation of lesion, WM, and gray matter by using algorithms based on fuzzy connectedness principles reported previously (14,15) and an internal version of the 3DVIEWNIX software system (16). The results of software segmentation of the brain were manually corrected by a trained neuroradiologist (I.C.) to exclude parts of the skull and regions of cerebrospinal fluid that were occasionally included by the algorithms. Lesions (if present), WM, and gray matter were also identified with the software and were used to produce masks that corresponded to the specific tissue types. This technique exhibits 95.5% intrareader reproducibility for the segmentation of WM (and gray matter) on repeated images (Nyul LG, unpublished results, 1998). Identified lesions were confirmed by the neuroradiologist and were excluded from the WM mask to obtain NAWM masks.
Because the T2- and intermediate-weighted images were generally not in perfect registration with the MT images, the former were registered with the latter in three dimensions by using an intensity correlation method (17,18). Because the intensity patterns on the T2-weighted and MT images were relatively similar, we registered each whole set of T2-weighted images with the corresponding set of MT images. The resultant transformation (including translation and rotation in three dimensions) was then applied to the NAWM mask, which was subsequently also scaled to match the bigger voxels of the MT images. This process resulted in a transformed NAWM mask that matched, as accurately as possible, the NAWM region on the MTR maps. Application of the NAWM mask to the MTR maps allowed calculation of the MTR histogram of whole-brain NAWM. The histograms were normalized to the NAWM volume, and the histogram parameters of peak height, peak location, and mean MTR were calculated for each patient and subject. Data from patients and control subjects that corresponded to the two MT imaging protocols (periods 1 and 2) were grouped and processed as outlined subsequently.
We measured the differences between MTR histograms for NAWM in a subset of control subjects and patients who underwent follow-up conventional and MT MR imaging. In each case, a subtraction histogram was generated by subtracting the more recent histogram from the earlier one (Fig 1). Because we used normalized histograms that had equal total areas, the subtraction histogram, as demonstrated in Figures 1 and 2, can be divided into positive and negative sections that are equal in area. All of the subtraction histograms for patients had a general bimodal form (Fig 1), whereas subtraction histograms for healthy subjects did not have a particular form. In patients, the change in the distribution over time was reflected in the area under either the positive or the negative curve sections. We computed the integral of the positive section of the histogram to determine the relative change between the two MT studies. We also measured the maximum and minimum points of the curve.
|
|
We compared MTR histogram parameters between patients and control subjects for periods 1 and 2 by using the Wilcoxon rank sum test for independent samples. When two studies were available, only the earlier was used for this comparison. We compared changes between successive studies in patients and control subjects by using the Wilcoxon signed rank test for paired data. We performed Spearman correlations between MTR histogram parameters and EDSS score, ambulation index, age, and disease duration, as well as lesion load on T2-weighted images and brain and WM volumes; again, only the first of multiple studies was used.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Our present—largely negative—results with regard to correlation between NAWM MTR histogram parameters and clinical measures underscore that both microscopic and macroscopic assessments of cerebral disease burden must be included in a global index of disease status. In addition, there is recent evidence (20) supporting the argument for inclusion of gray matter abnormalities for recognition of neuronal loss associated with MS and the potential for "disconnection" of pathways due to WM disease. We note that the previous histogram analyses (7,8) took into account gray matter involvement as part of the whole-brain analysis. Atrophy in gray matter may yet be found to play an important role, and spinal cord involvement certainly has a large influence on clinical assessment (21). Thus, the present results and the techniques discussed in this report may offer the most value in terms of increased understanding of the natural history of the disease and of methods for further probing the prognostic implications.
Results of both phases of this study clearly indicate that the NAWM histogram differed between normal brains and those of patients with MS. Our observations in NAWM may reflect the pathologic processes occurring microscopically therein, including demyelination, edema, and gliosis. Previous results (4,22–24) have suggested that MTR can serve as an index of demyelination, and the results of a study of feline wallerian degeneration in visual pathways (25) established a relationship between a biphasic response of MTR and histopathologic findings at electron microscopy.
Our results are in contrast, to some degree, with those reported for whole-brain histograms. In particular, the normalized peak height of the NAWM histogram was not different between patients and control subjects, whereas in the whole-brain histogram (7,8), this parameter was shown to be reliably different. An explanation for this finding may be the inclusion of lesion pixels in the whole-brain histograms, which causes a profound depression of the peak.
In the comparative analysis of paired studies, we noted that, over time, the histogram distributions shifted to the right toward more normal values. This can be seen in the shape of the subtraction histogram, which in patients always was characterized by a negative-valued curve on the left and a corresponding curve to the right on the positive side of the abscissa (Fig 1). Subtraction histograms in healthy subjects did not follow a regular pattern and were smaller in magnitude, consistent with the expected absence of change from study to study (Fig 2).
The areas of the positive and negative sections of the subtraction histogram were equalized by means of the normalization process, which enabled us to study the distribution of values as opposed to raw pixel counts. In patients, the magnitude of the area under the negative portion of the curve corresponded to fewer pixels with lower MTR values in the second study, which represented an evolution toward normal values. The height of the subtraction histogram contained information about the distribution of the "shifted" pixels and might be useful in the discrimination of disease processes, although the preliminary nature of the present data precludes establishment of this contention. We speculated that the "lost" low-MTR pixels would represent a substantial volume of WM that became lesions, but we found that the fraction was variable and, in some cases, very small. An alternative explanation for the finding could be that the regions of NAWM associated with lost pixels were transiently edematous, particularly around lesions that existed at the time of the first study. Edema is known to be characterized by lower MTR values (4,26) and, thus, could have contributed to lower MTRs in NAWM in the first of any pair of studies. Resolution of this edema in the second study could then help explain the higher and more normal values of MTR in the second study. One might expect, however, that the process could occur in either direction and would not have an overall influence on the results. The question remains unresolved and is suggested for future study.
The results of our analysis with regard to the origin of new lesions are suggestive of a link between MTR abnormality in NAWM and subsequent development of lesions. Our results are consistent with some previous results (27,28) but not with others (29). This seeming discrepancy is, perhaps, reflected in the wide variation among the data. The issue remains unresolved and is deserving of further studies, which may provide more insight into the natural course of the disease.
In conclusion, our results add to the information on characterization of NAWM abnormality associated with MS and demonstrate that such abnormality may be explored in terms of the distribution of MTR values. Our results in a small sample indicate that the NAWM distribution of MTR values became closer to the control (disease-free) distribution over time, an unexpected finding that suggests the need for future study and leads to speculation on the nature of evolution from NAWM to lesion. Some regions of NAWM may be "at risk" for the development of new lesions, while other NAWM regions may be stable with regard to MS, although the histogram technique does not appear to help distinguish between those tissue types in an obvious manner. Rather, the future value of the technique may lie in providing an estimate of the degree to which the disease has progressed.
We speculate that although inclusion of gray matter changes and spinal cord involvement may be essential for full characterization of the disease, MT histogram analysis of NAWM, including longitudinal analysis, may provide new prognostic information regarding lesion formation, as well as increased understanding of the disease course. An improved understanding may further the evaluation of novel treatment strategies and pharmaceutical agents for which the preservation of NAWM could be essential.
|
|
FOOTNOTES |
|---|
Author contributions: Guarantor of integrity of entire study, J.C.M.; study concepts, R.I.G., J.C.M., J.K.U.; study design, R.I.G., J.C.M., J.K.U., D.L.K.; definition of intellectual content, J.C.M.; literature research, I.C., J.K.U.; clinical studies, I.C., D.L.K.; data acquisition, I.C., L.J.M.; data analysis, I.C., L.W., L.G.N., X.Z.; statistical analysis, M.P., J.C.M.; manuscript preparation, I.C., J.C.M.; manuscript editing, J.C.M., R.I.G., J.K.U.; manuscript review, R.I.G., J.K.U.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  H. Vrenken, P.J.W. Pouwels, S. Ropele, D.L. Knol, J.J.G. Geurts, C.H. Polman, F. Barkhof, and J.A. Castelijns Magnetization transfer ratio measurement in multiple sclerosis normal-appearing brain tissue: limited differences with controls but relationships with clinical and MR measures of disease Multiple Sclerosis, July 1, 2007; 13(6): 708 - 716. [Abstract] [PDF]
|
|
|
|
 |
|
 H. Vrenken, J.J.G. Geurts, D.L. Knol, C.H. Polman, J.A. Castelijns, P.J.W. Pouwels, and F. Barkhof Normal-appearing white matter changes vary with distance to lesions in multiple sclerosis. AJNR Am. J. Neuroradiol., October 1, 2006; 27(9): 2005 - 2011. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. Ge Multiple Sclerosis: The Role of MR Imaging AJNR Am. J. Neuroradiol., June 1, 2006; 27(6): 1165 - 1176. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. C.-Y. Chen, H.-W. Chung, and M. Liou Measurement of Volumetric Lesion Load in Multiple Sclerosis: Moving from Normal- to Dirty-Appearing White Matter AJNR Am. J. Neuroradiol., November 1, 2003; 24(10): 1929 - 1930. [Full Text] [PDF]
|
|
|
|
|
|
Y. Ge, R. I. Grossman, J. S. Babb, J. He, and L. J. Mannon Dirty-Appearing White Matter in Multiple Sclerosis: Volumetric MR Imaging and Magnetization Transfer Ratio Histogram Analysis AJNR Am. J. Neuroradiol., November 1, 2003; 24(10): 1935 - 1940. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. J. Lowe, M. D. Phillips, J. T. Lurito, D. Mattson, M. Dzemidzic, and V. P. Mathews Multiple Sclerosis: Low-Frequency Temporal Blood Oxygen Level-Dependent Fluctuations Indicate Reduced Functional Connectivity—Initial Results Radiology, July 1, 2002; 224(1): 184 - 192. [Abstract] [Full Text]
|
|
|
|
|
|
A. C. Guo, V. L. Jewells, and J. M. Provenzale Analysis of Normal-Appearing White Matter in Multiple Sclerosis: Comparison of Diffusion Tensor MR Imaging and Magnetization Transfer Imaging AJNR Am. J. Neuroradiol., November 1, 2001; 22(10): 1893 - 1900. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. L. P. Lindberg, C. J. A. De Groot, L. Montagne, P. Freitag, P. van der Valk, L. Kappos, and D. Leppert The expression profile of matrix metalloproteinases (MMPs) and their inhibitors (TIMPs) in lesions and normal appearing white matter of multiple sclerosis Brain, September 1, 2001; 124(9): 1743 - 1753. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
