Rofo 2005; 177 - 9
DOI: 10.1055/s-2005-865258

Quantification of White Matter myo-Inositol by 1H-MRS at 3.0 T in Normal Controls and in Patients with Multiple Sclerosis

F Träber 1, MP Wattjes 1, M Harzheim 2, O Gür 1, W Block 1, HH Schild 1
  • 1Department of Radiology
  • 2Department of Neurology, University of Bonn, Germany

Purpose: To optimize the 1H-MRS acquisition protocol for measuring absolute cerebral concentrations of myo-inositol (mIns) at 3.0 T, in order to investigate metabolic alterations in the normal appearing white matter (NAWM) of patients with a clinically isolated syndrome (CIS) suggestive of multiple sclerosis (MS).

Methods: On a clinical 3.0 T whole-body MR system (Gyroscan Intera 3.0 T, Philips Medical Systems), single-voxel 1H-MRS (PRESS-localized, water suppression by dual inversion prepulses) in the parietal white matter of the centrum semiovale was performed in 10 healthy controls. Echo time was varied in the range TE 30–140 ms in phantom experiments and in-vivo acquisitions to find the optimum TE value for mIns quantification, with respect to the intensity of the spin-coupled mIns multiplets at 3.54 ppm (C1/C3) and 3.63 ppm (C4/C6) and to the separation from adjacent glutamate/glutamine (Glx) signals. Metabolite signal ratios were then obtained from water-suppressed measurements with TR/TE 2000/38 ms. Absolute concentrations of mIns, of choline compounds (Cho), of total creatine (tCr), and of N-acetyl aspartate (NAA) in the parietal white matter of the controls were determined by relating NAA to the internal water signal from the MRS voxel in unsuppressed acquisitions with TR/TE 3500/140 ms and correcting for partial CSF volume by water T2 relaxometry. Additionally, relaxation times T2 of Cho, tCr, and NAA at 3.0 T were measured by mono-exponential fits to a series of 5 water-suppressed spectra with TR/TE 2500/50–400 ms. With voxel localization and acquisition parameters held identical to the control investigations, T2 relaxation times and absolute metabolite concentrations were also determined in NAWM of 22 patients with CIS (5 of them classified as early-stage relapsing-remitting MS according to McDonald's criteria). In 15 of the CIS cases, metabolite quantification was also done at 1.5 T (Philips Gyroscan Intera 1.5 T) for intraindividual comparison. Postprocessing of all data was performed by time-domain analysis using the MRUI software package.

Results: Intra-individual comparison of relative linewidths δν/γ obtained in the short-TE measurements at both field strengths revealed a reduction by about 10–20% at 3.0 T for the singlet resonances of Cho and tCr and almost unchanged values for the C1/C3 and C4/C6 pseudo-doublet of mIns. T2 relaxation times of Cho and NAA at 3.0 T were slightly prolonged in the NAWM of the CIS patients (257±34 ms and 349±44 ms, respectively, vs. 232±20 ms and 321±16 ms in the controls). Preliminary results for the metabolite ratios at 3.0 T and for the absolute concentrations showed a reduction of NAA/Cho, a 10% increase for mIns, and a 10% decrease for NAA in NAWM of the patient group. Absolute metabolite concentrations determined from the measurements at 3.0 T were consistent with the values obtained at 1.5 T. In the NAWM of the patient subgroup with early-stage MS, an even more pronounced elevation in myo-inositol concentration was observed in comparison to healthy controls.

Conclusion: The increased chemical shift dispersion at 3.0 T strongly alters the 1H-MRS pattern of mIns multiplets compared to lower field strengths, resulting in a well resolved doublet peak from C1/C3 and C4/C6 protons with almost equal component intensity at TE 30–40 ms and improved separation from Glx. First results of our clinical study indicate higher mIns concentration already in lesion-free NAWM during the very early stages of MS. With an enlarged number of cases, the significance of these findings and the sensitivity at 3.0 T compared to 1.5 T may be assessed.