PURPOSE

To evaluate the clinical value of 1H MR spectroscopy (1H MRSI) for follow-up of irradiated glioma compared to positron emission tomography (PET) with [18F]-2-fluoro-deoxy-D-glucose (FDG-PET) and single photon emission tomography with [123I]-a-methyl-L-tyrosine (IMT-SPECT).

MATERIALS AND METHODS

Twenty-four patients with irradiated gliomas were examined using 1H MRSI (2D spectroscopic imaging; PRESS; TE = 135 msec; 1.5T Magnetom Vision, Siemens; Voxel size 9 x 9 x 15 mm (3)). MR spectra (n = 233) were evaluated in areas suspicious of tumor (n = 86) as well as in healthy appearing brain tissue (n = 147). Relative signal intensity ratios of choline (Cho), creatine (Cr) and N-acetyl-aspartate (NAA) were calculated. PET scans (n = 19) were performed with 200 - 250 MBq FDG, IMT-SPECT examinations (n = 14) with 200 - 250 mBq IMT. Based on clinical and MRI/CT, follow-up lesions were classified as either neoplastic [PT] or non-neoplastic [nPT].

RESULTS

True positive results for the diagnosis of PT/nPT were 88/89 % (1H MRSI), 73/100 % (PET) and 100/75 % (SPECT). Cho/Cr showed highly significant changes for PT. Determinating a correlation between Cho, Cr, NAA and IMT-SPECT as well as FDG-PET was not possible because of different location of maximum tracer uptake and acquired 2D 1H MRSI.

CONCLUSION

IMT-SPECT seems to be superior to detect tumor progression in irradiated gliomas. 1H MRSI was more suitable than FDG-PET to differentiate between recurrence and radiation-induced changes. FDG-PET plays a role as sensitive method for detecting high-grade tumors. PET and SPECT allowed the examination of the entire tumor including surrounding brain tissue with higher spatial resolution than the acquired 2D 1H MRSI. A main limitation of our study was that only 2D 1H MRSI was used, with only parts of the tumor evaluated. The use of 3D MR spectroscopic imaging may further increase the diagnostic accuracy.