Binswanger's disease

Magnetic resonance imaging (MRI) of diffusion and magnetization transfer was combined with ^sup 1^H-spectmscopic imaging (CSI) to evaluate the clinical potential of in-vivo profiles of various brain pathologies. Ten patients (multiple sclerosis, cerebrovascular disease, leukodystrophy, Alzheimer dementia) and five healthy volunteers were investigated with diffusion-weighted MRI, magnetization transfer imaging, and CSI. Proton spectra were analyzed as ratios of NAA/Cr and Cho/Cr calculated from the peak areas of N-acetylaspartate (NAA), (phospho)-creatine (Cr) and choline (Cho). The apparent diffusion coefficient (ADC) and the magnetization transfer ratio (MTR) were determined in identical voxels to ensure identical partial volume effects compared to CSI. Compared to MTR and ADC assessments, the lower spatial resolution of CSI clearly indicates a hindrance at 7.5 T. In most deinyelinating lesions, NAA/Cr reduction paralleled attenuated MTRs and elevated ADCs. By contrast, in acute stroke and some acute MS lesions the ADC was reduced, while MTR and NAA/Cr were also decreased. In Alzheimer's dementia, ADC was increased, MTR unchanged and Cho/Cr increased. In a case of leukodystrophy, ADC was pronouncedly increased, MTR and NAA/Cr both reduced, and Cho/Cr normal. Combined measurements of ADC, MTR and CSI are feasible and provide differential in-vivo information on various brain pathologies. [Neurol Res 2003; 25: 292-300]

Keywords: Proton MR spectroscopy; diffusion imaging; magnetization transfer; brain pathology

INTRODUCTION

Three new MR methods, diffusion-weighted magnetic resonance imaging (DWI), imaging of the magnetization transfer ratio (MTR), and ^sup 1^H MR spectroscopic imaging (also named chemical shift imaging (CSI)), have all improved the understanding of the underlying pathophysiology in a number of neurological diseases1-7. Measurements of the apparent diffusion coefficient (ADC), the MTR and proton spectra provide in-vivo information on local tissue diffusion, its macromolecular integrity, and a metabolic profile3,8-10. By using those MR technologies, a wide variety of brain lesions such as cytotoxic cell swelling, vasogenic edema formation, gliosis, inflammation, demyelination, ischemia, axonal loss etc., all appearing hyperintense in T^sub 2^-weighted MR images, may be further differentiated. Ideally, these three methods were to be combined and integrated into a single MR examination.

In the present feasibility study, MRI of diffusion, magnetization transfer, and CSI were combined in order to develop a clinically useful study protocol to analyze brain lesions of different origin. The hypothesis was tested that the combination of the MR methods mentioned and the comparison of regional results obtained would be feasible and, due to the multi-parametric approach, provide a more solid basis for the interpretation of common MRI and CSI findings.

PATIENTS AND METHODS

Patients and controls

Five healthy controls (mean age 30 + or -4 years) and 10 patients (mean age 45 + or - 22 years) were investigated with an identical MR protocol combining conventional MRI (proton density-, T^sub 2^- and T^sub 1^-weighted images), ADC maps, MTR maps and CSI. In two MS patients additional short echo time volume selective 1H-MRS data was also evaluated. All patients were in-patients at the Department of Neurology. The brain pathologies under investigation comprised multiple sclerosis (four patients), cerebrovascular disease (three patients), leukodystrophy (Alexander's disease), Alzheimer-type dementia and cerebral hemangioma (one patient each). The clinical data are given in Table 1. Informed consent was obtained in written form from all patients or the responsible relative prior to the MR examination (Table 7).

MR data acquisition

MRI was performed on a 1.5 Tesla clinical scanner (Vision, Siemens, Erlangen, Germany) with echo planar hardware (gradient power 25 mT/m, rise time 83 mT/m/ msec). The MRI protocol consisted of:

1. Transverse, coronal, and sagittal localizing sequences followed by transverse oblique contiguous images (slice thickness 5 mm) aligned with the inferior borders of the corpus callosum (sequences 2 to 5).

2. Proton density (PD)- and T^sub 2^-weighted images (TSE, 2620 msec/14 msec/85 msec, FOV 180240mm^sup 2^, matrix size 384 x 512).

3. T^sub 1^-weighted images (SE, TR 530 msec/TE 12 msec).

4. Diffusion-weighted echo planar spin echo images (TR 4000 msec/TE 144 msec, b = 0/160/360/640/ 1000 sec/mm^sup 2^) with sequential application of three separate diffusion sensitising gradients in perpendicular directions.

5. Images of the magnetization transfer ratio MTR (3D FLASH sequence, TR 40 msec/TE 5 msec, 1.2 kHz off-resonance pulse, bandwidth 500 Hz, hsat = 500); T^sub 1^-weighted images were acquired both with and without application of a saturation pre-pulse to saturate the broad resonance of immobile macromolecular protons. The energy deposited by this pulse provided measurable differences between saturated and unsaturated images and ensured a good signal-to-noise ratio (SNR) in the calculated MTR image.

The T^sub 2^-weighted images in transverse, coronal, and sagittal planes were used for the positioning of the CSI slab. After choosing the volume-of-interest, the volume selective shim and the water-suppression were performed. A water-suppressed SE-pulse-sequence (TE = 135 msec, TR = 1500 msec; NA = 2, TA= 12 min 55 sec) and furthermore the same sequence without water-suppression (NA = 1, TA = 6 min 31 sec) were used. The VOI was usually 8x8x2 cm^sup 3^, resulting in a lattice containing 64 voxels according to 64 ^sup 1^H-MR-spectra (Figure 7).

Data processing and analysis

ADC maps were obtained by a linear least-squares fit on a pixel-by-pixel basis after averaging of the direction-dependent diffusion-weighted images. The directionally independent trace of the diffusion tensor (ADC/3) was determined. Quantitative MTR images were derived from the two images without (Mo) and with (Ms) saturation pulse on a pixel-by-pixel basis according to the equation: MTR = (Mo-Ms)/Mo. Signal intensities in the calculated image represented the amount of magnetization transfer between the free and bound water pool.

For the evaluation of CSI a post-processing program LUISE (Siemens Medical Systems, Erlengen, Germany) enclosing a water-reference processing, manually phase and baseline-correction for each spectrum were used. The resonance-lines of N-acetyl-aspartate, creatine and phospho-creatine, choline-containing compounds and lactate were manually fitted with gaussian line-shapes. Quality assurance was given by subtraction of the fitted lines from the original spectra. The results of the evaluation were expressed as ratios of NAA/Cr and Cho/Cr.

In order to compare measurements of MRI and CSI, those voxels of spectroscopic imaging were selected that were located in brain lesions identified on T^sub 2^-weighted images (e.g., MS plaques, ischemic lesions, affected white matter). Second, corresponding voxels were selected in the (unaffected) contralateral hemisphere and/or in normal appearing white and gray matter. After selection of the voxel-of-interest, the images of ADC and MTR were re-aligned to match the plane of the CSI slab (voxel size 1x1x2 cm^sup 3^; slab thickness 2 cm). The slice thickness of re-aligned MR images amounted to 4.0 mm (ADC maps) and 2.5 mm (MTR maps), respectively. Manually defined regions-of-interest (ROIs, size 1 x 1 cm^sup 2^) were positioned precisely at the location of the pre-selected CSI voxel. This procedure was performed in those four slices (ADC maps) or seven slices (MTR maps), respectively, that corresponded to the CSI slab (Figure 2). Per selected CSI voxel, four ADC values (derived from ROIs in four re-aligned slices) and seven MTR values (derived from ROIs in seven re-aligned slices) were averaged for direct comparison. Normal values for NAA/Cr, Cho/Cr, MTR and ADC were established in five healthy volunteers by analysis of 19 (cortical) and 21 (white matter) voxels-of-interest (VOI) or regions-of-interest (ROI), respectively. Significant alterations were defined by a confidence interval of two standard deviations above or below the respective normal values.

RESULTS

Normal values were established of the variables under investigation in a separate manner for cerebral cortex and subcortical white matter in five healthy controls (Table 2).

In patients with MS, chronic lesions showed an increased ADC, reduced MTR and attenuated NAA/Cr ratio. The Cho/Cr was normal or showed increased values in MS patients (Tables 3 and 4). In one MS patient with acute lesions, a reduced ADC was observed. Another acute MS lesion with marked edema showed a very marked decrease in MTR and reduction of all metabolites indicating a dilutional effect. Additional single voxel 1 ^sup 1^H-spectra showed signal at ~1.3 up to 1.4 ppm, reflecting pathological free lipids and/or lactate (not shown).

Our case with suspected M. Alexander, a rare form of leukodystrophy, presented with very high values of the ADC observed in the affected white matter and both, MTR and NAA/Cr ratios were attenuated, while the Cho/ Cr ratio was normal (Table 3, Figures 3 and 4).

In microangiopathic white matter lesions due to Binswanger's disease, the ADC was increased, MTR unchanged or slightly decreased and ^sup 1^H-MR-spectra without consistent changes (Tables 3 and 4). In subacute ischemie stroke, the values of ADC and MTR were slightly reduced, NAA/Cr however markedly decreased (Table 3). Proton spectra showed a lactate peak in infarcted areas (Table 4, Figures 4 and 5). In a case of Alzheimer's type of dementia, the ADC was elevated and MTR unchanged. Proton MR spectra revealed a slight decrease in NAA/Cr and a slight increase of Cho/ Cr, respectively (Table 3). In a case of thalamic hemangioma, increased levels of ADC and a reduced NAA/Cr were observed. The individual values of ADC, MTR and cerebral metabolite ratios are given in Table 3. For synopsis of imaging and spectroscopic results observed in various pathologic states, see Table 4 and Figure 4.

To demonstrate the influence of the individual lesion geometry on partial volume effects in the assessments of ADC and MTR, the MS lesion shown on Figure 2 (voxel marked by black square) was analyzed. In four consecutive re-aligned ADC maps and seven corresponding MTR maps, respectively, the individual mean values of ADC/MTR were measured in a ROI that was identical to the voxel position in the CSI slab. Individual ADC values (from top to bottom): 1 540, 1540, 1310, and 1970 [mu]m^sup 2^/sec (mean 1590 [mu]m^sup 2^/sec); individual MTR values: 22, 22, 23, 24, 27, 31, and 33% (mean 26.2%).

DISCUSSION

In the present feasibility study, measurements of tissue diffusion and magnetization transfer were combined with CSI information in a variety of pathological states of the brain. Beyond the lesion size and pattern on conventional (PD-, T^sub 2^-, T^sub 1^-weighted) MR images, a more detailed description offers various potential advantages. Our hypothesis was that different brain pathologies would be reflected by a differential profile of the quantitative MRI and CSI variables. Not only the direction of change, but also the extent of potential alterations could be taken into account and might be particularly helpful in follow-up assessements to identify even small changes. Indeed, the synopsis of alterations observed confirms typical patterns for the different underlying pathologies as they have been suggested in previous studies that were performed on those pathologies separately6,9-18.

Most MS lesions presented with high ADC, low MTR, consistently low NAA/Cr and normal or high Cho/Cr ratios. These findings correspond very well to the literature where reduced NAA levels (by part transiently) have been reported in conjunction with increased choline-containing compounds probably reflecting increased membrane turnover7. Increased ADC and decreased MTR have been observed in MS plaques by several investigators4,9,11-14, whereas the finding of a reduced ADC is unusual and can be only found in very acute MS lesions as also shown recently by Tievsky et al.19.

Patients with subcortical vascular lesions due to microangiopathy showed an increased ADC, normal or reduced MTR and no consistent CSI changes in the affected white matter. This is in line with a study by Constans ef al.15 in which no consistent metabolite changes in ^sup 1^H-MR-spectra were reported in non-demented patients with white matter signal hyperintensities, and confirms the nonspecific nature of T^sub 2^weighted white matter hyperintensities in patients with microangiopathy and the usefulness of quantitative evaluations using MTR or ADC in order to estimate the severity of tissue damage.

A similar, but not identical, pattern was seen in the case of Alzheimer's dementia with increased ADC, normal MTR, slightly reduced NAA/Cr and slightly increased Cho/Cr. There is good evidence for reduced NAA levels in Alzheimer's disease as shown by several investigators15'20'21. Most of those measurements have been performed in the frontal or parietal cortex of patients, i.e., in regions that can be expected to be affected by disease. However, there are conflicting results with regard to choline-containing compounds: Tedeschi et al. 1 reported a reduced Cho/Cr in some patients with Alzheimer's disease, whereas Constans ef a/.15 found increased Cho/Cr as we observed it in our patient. It has been proposed that increased levels of choline or Cho/Cr reflect changes in membrane lipids22, but the exact pathogenetic mechanism in Alzheimer's disease remains unclear.

In cerebral ischemia, a disturbance of tissue diffusion occurs very early after vessel obstruction and has been extensively documented both in experimental and human stroke8'16. Much less is known about changes in MTR in these conditions. There are recent reports on reduced MTR that occur following cerebral artery occlusion in animals23 and in human stroke patients . We observed a trend toward lower MTR in the ischemic cortex compared to contralateral nonischemic regions. In good agreement with previous reports, proton MR spectra showed a lactate peak indicating anerobic metabolism in the ischemic region as well as an impressive decrease of NAA/Cr that points out a remarkable neuronal loss in the area of investigation6'18'24.

Beyond the differences in lesion distribution, that frequently identifies leukodystrophies, a combination of a gross increase in ADC and a marked decrease in MTR was noted in the white matter of patient No. 9. In line with these findings a decrease in NAA levels was observed24'25 whereas interestingly the Cho/Cr ratio was unaltered. The normal Cho may indicate long-standing disease without recent activity in the presence of severe chronic tissue changes.

Although applied to a limited number of patients this study demonstrates that a multiparametric approach including CSI is feasible and provides complementary data to routine MR examinations. Currently, it cannot be applied routinely mainly due to the extensive postprocessing procedures of spectroscopic and imaging data. Technical advances, however, may enable a broader clinical application in the near future. In this study special emphasis was put on the consideration of the partial volume effects that are derived from the marked difference in spatial resolution between the spectroscopic and imaging data. We made sure that all ADC and MTR values were calculated as a mean of the same VOI, i.e., in a voxel of 1 x 1x2 cm^sup 3^ size. For future studies in the field, an improved CSI resolution appears most necessary and promising. Given the dilutional effect of the large spectroscopic voxels applied to MTR and ADC assessments, one may speculate that the higher SNR of high field systems should improve the detection and quantitation of CSI abnormalities to a large degree and, thereby, make proton MR spectroscopy much more attractive to the investigation of small brain lesions.

CONCLUSION

This study demonstrates that beyond the lesion size and pattern on conventional (PD-, T^sub 2^-, T^sub 1^-weighted) images, a more detailed description can be derived from an integrated MR protocol. It is now feasible to combine MRI of diffusion and magnetization transfer with 1H-MR spectroscopy, thereby providing a view from different angles at the pathology. This combination may lead to confirming results, provide a grading or staging of lesions and help to differentiate tissue changes that are more difficult to distinguish with a more simplistic approach. In the future, it may provide a promising tool to enhance the insight into brain pathophysiology.

ACKNOWLEDGEMENTS

Dr Mockel was supported by a grant of the Baden-Wurttemberg Research Funding Agency. Dr Back was supported by a grant of the European Neurological Society which is gratefully acknowledged.

REFERENCES

1 Warach S, Chien D, Li W, Ronthal M, Edelman RR. Fast magnetic resonance diffusion-weighted imaging of acute human stroke. Neurology 1992; 42: 171 7-1 723

2 Moseley ME, Butts K, Yenari MA, de Crespigny A. Clinical aspects of DWI. NMR Biomed 1995; 8: 387-396

3 Wolff SD, Balaban RS. Magnetization transfer contrast (MTC) and tissue water relaxation in vivo. Magn Res Med 1 989; 10: 1 35-1 44

4 Gass A, Barker GJ, Kidd D, Thorpe JW, MacManus D, Brennan A, Tofts PS, Thompson AJ, McDonald Wl, Miller DH. Correlation of magnetization transfer ratio with clinical disability in multiple sclerosis. Ann Neurol 1994; 36: 62-67

5 Hugg JW, Laxer KD, Maison GB, Maudsley AA, Weiner MW. Neuron loss localizes human temporal lobe epilepsy by in vivo proton magnetic resonance spectroscopic imaging. Ann Neuro! 1993; 34: 788-794

6 Gillard JH, Barker PB, van ZiJl PCM, Bryan RN, Oppenheimer SM. Proton MR spectroscopy in acute middle cerebral artery stroke. Am J Neurorad 1996; 17: 873-886

7 Narayana PA, Doyle TJ, Lai D, Wolinsky JS. Serial proton magnetic resonance spectroscopic imaging, contrast-enhanced magnetic resonance imaging, and quantitative lesion volumetry in multiple sclerosis. Ann Neurol 1998; 43: 56-71

8 Hoehn-Berlage M. Diffusion-weighted NMR imaging: Application to experimental focal cerebral ischemia. NMR Biomed 1995; 8: 345-358

9 Loevner LA, Grossman Rl, McGowan JC, Ramer KN, Cohen JA. Characterization of multiple sclerosis plaques with T1-weighted MR and quantitative magnetization transfer. Am j Neuroradiol 1995; 16: 1473-1479

10 Ross B, Michaelis T. Clinical applications of magnetic resonance spectroscopy. Magn Reson Q 1994; 10: 191-247

11 Pike GB, de Stefano N, Narayanan S, Francis GS, Antel JP, Arnold DL. Combined magnetization transfer and proton spectroscopic imaging in the assessment of pathologic brain lesions in multiple sclerosis. AmJ Neuroradiol 1999; 20: 829-837

12 Horsfield MA, Larsson HB, Jones DK, Gass A. Diffusion magnetic resonance imaging in multiple sclerosis. ] Neurol Neurosurg Psychiatry 1998; 64 (Suppl. 1): S80-S84

13 Hiehle JF, Lenkinski RE, Grossman Rl, Dousset V, Ramer KN, Schnall MD, Cohen JA, Gonzalez-Scarano F. Correlation of spectroscopy and magentization transfer imaging in the evaluation of demyelinating lesions and normal appearing white matter in multiple sclerosis. Magn Res Med 1994; 32: 285-293

14 Droogan AC, Clark CA, Werring DJ, Barker GJ, McDonald Wl, Miller DH. Comparison of multiple sclerosis clinical subgroups using navigated spin echo diffusion-weighted imaging. Magn Reson Imaging 1999; 17: 653-661

15 Constans JM, Meyerhoff DJ, Gerson J, MacKay S, Norman D, Fein G, Weiner MW. H-1 MR spectroscopic imaging of white matter signal hyperintensities: Alzheimer disease and ischemie vascular dementia. Radiology 1995; 197: 517-523

16 Warach S, Gaa J, Siewert B, Wielopowski P, Edelman RR. Acute human stroke studied by whole brain echo planar diffusionweighted magnetic resonance imaging. Ann Neurol 1995; 37: 231-241

17 Hanyu H, lmon Y, Sakurai H, Iwamoto T, Takasaki M, Shindo H, Abe K. Diffusion-weighted magnetic resonance and magnetization transfer imaging in the assessment of ischemie human stroke. Intern Med 1998; 37: 360-365

18 Barker PB, Gillard JH, van ZiJl PC, Soher BJ, Hanley DF, Agildere AM, Oppenheimer SM, Bryan RN. Acute stroke: evaluation with serial proton MR spectroscopic imaging. Radiology 1994; 192: 723-732

19 Tievsky AL, Ptak T, Farkas J. Investigation of apparent diffusion coefficient and diffusion tensor anisotrophy in acute and chronic multiple sclerosis lesions. Am J Neuroradiol 1999; 20: 1491-1499

20 Schuff N, Amend DL, Meyerhoff DJ, TanabeJL, Norman D, Fein C, Weiner MW. Alzheimer disease: Quantitative H-1 MR spectroscopic imaging of frontoparietal brain. Radiology 1998; 207: 91-102

21 Tedeschi G, Bertolino A, Lundbom N, Bonavita S, Patronas NJ, Duyn JH, Metman LV, Chase TN, Di Chiro G. Cortical and subcortical chemical pathology in Alzheimer's disease as assessed by multislice proton magnetic resonance spectroscopic imaging. Neurology 1 996; 47: 696-704

22 Miller BL. A review of chemical issues in 1 H NMR spectroscopy: N-acetyl-L-aspartate, creatine and choline. NMR Biomed 1991; 4: 47-52

23 Ewing JR, Jiang Q, Boska M, Zhang ZG, Brown SL, Li GH, Divine GW, Chopp M. T1 and magnetization transfer at 7 Tesla in acute ischemie infarct in the rat. Magn Reson Med 1999; 41: 696-705

24 Hwang JH, Graham GD, Behar KL, Alger JR, Prichard JW, Rothman DL. Short echo time proton magnetic resonance spectroscopic imaging of macromolecule and metabolite signal intensities in the human brain. Magn Reson Med 1996; 35: 633-639

25 Schiffmann R, Moller JR, Trapp BD, Shih HH, Farrer RG, Katz DA, Alger JR, Parker CC, Hauer PE, Kaneski CR. Childhood ataxia with diffuse central nervous system hypomyelination. Ann Neurol 1994; 35: 331-340

Tobias Back*, Regina Mockel[dagger], Jochen G. Hirsch[dagger], Jochen Gaa[double dagger], Wolfgang H. Oertel*, Michael G. Hennerici[dagger] and Achim Gass[dagger]

* Department of Neurology, Philipps University Marburg

[dagger] Department of Neurology, [dagger] Department of Radiology, Klinikum Mannheim, Ruprecht-Karls University, Heidelberg, Germany

Correspondence and reprint requests to: Prof. Dr T. Back, Department of Neurology, Philipps University Marburg, R-Bultmann-Str 8, D-35039 Marburg, Germany, [back@mailer.uni-marburg.de] Accepted for publication December 2002.

Copyright Forefront Publishing Group Apr 2003
Provided by ProQuest Information and Learning Company. All rights Reserved