OBJECTIVE: To observe metabolic differences between spinal tumor and other diseases in human spinal mass lesions, in vivo 1H magnetic resonance spectroscopy (MRS) was attempted to obtain metabolic signals in patients with various spinal mass lesions. METHODS: 1H nuclear magnetic resonance (NMR) spectra were obtained from 14 patients before surgery using a receive-only surface coil on a 1.5 T clinical magnetic resonance imaging (MRI) unit. MRS findings were compared with the histopathologic results from biopsy. In addition, tumor spectra were compared with the spectra of other benign diseases including disc herniation, which can mimic spinal cord tumor. In vitro 1H-NMR spectra were also collected from perchloric acid extracts of some spinal tumors. RESULTS: Typical water resonance line widths were in the 6- to 10-Hz range, but the metabolic signals observed were sufficiently resolved to be assigned from comparison with the 1H spectra of brain tissue. Choline was detected in all tumor spectra (n = 6) except ependymoma, whereas it was absent in other benign diseases including disc herniation (mimicking spinal cord tumors), dermoid cyst, tuberculosis, and non-multiple sclerosis myelitis. Spectral patterns of meningiomas, schwannomas, metastasis from renal cell carcinoma, and ependymomas in the spinal cord were similar to those of central nervous system (CNS) tumors. It was not possible to observe distinctive metabolic differences between benign diseases owing to relatively larger line broadening of some signals compared with that in CNS tissue. CONCLUSIONS: It appeared that acquisition of in vivo 1H-NMR signals was possible in human spinal mass lesions on a 1.5 T clinical MRI unit. Detection of choline only in the spinal tumors may indicate that there is some potential in using in vivo 1H-MRS to distinguish spinal tumors from disc herniation mimicking spinal cord tumors, non-multiple sclerosis myelitis, and dermoid cysts. On the basis of our NMR findings, however, it was not possible to distinguish between benign diseases.
OBJECTIVE: To observe metabolic differences between spinal tumor and other diseases in human spinal mass lesions, in vivo 1H magnetic resonance spectroscopy (MRS) was attempted to obtain metabolic signals in patients with various spinal mass lesions. METHODS:1H nuclear magnetic resonance (NMR) spectra were obtained from 14 patients before surgery using a receive-only surface coil on a 1.5 T clinical magnetic resonance imaging (MRI) unit. MRS findings were compared with the histopathologic results from biopsy. In addition, tumor spectra were compared with the spectra of other benign diseases including disc herniation, which can mimic spinal cord tumor. In vitro 1H-NMR spectra were also collected from perchloric acid extracts of some spinal tumors. RESULTS: Typical water resonance line widths were in the 6- to 10-Hz range, but the metabolic signals observed were sufficiently resolved to be assigned from comparison with the 1H spectra of brain tissue. Choline was detected in all tumor spectra (n = 6) except ependymoma, whereas it was absent in other benign diseases including disc herniation (mimicking spinal cord tumors), dermoid cyst, tuberculosis, and non-multiple sclerosis myelitis. Spectral patterns of meningiomas, schwannomas, metastasis from renal cell carcinoma, and ependymomas in the spinal cord were similar to those of central nervous system (CNS) tumors. It was not possible to observe distinctive metabolic differences between benign diseases owing to relatively larger line broadening of some signals compared with that in CNS tissue. CONCLUSIONS: It appeared that acquisition of in vivo 1H-NMR signals was possible in human spinal mass lesions on a 1.5 T clinical MRI unit. Detection of choline only in the spinal tumors may indicate that there is some potential in using in vivo 1H-MRS to distinguish spinal tumors from disc herniation mimicking spinal cord tumors, non-multiple sclerosis myelitis, and dermoid cysts. On the basis of our NMR findings, however, it was not possible to distinguish between benign diseases.