A Einstien1, Rahul A Virani2. 1. Assistant Professor, Department of Radiology, Chettinad Hospital & Research Institute , Kelambakkam, Tamil Nadu, India . 2. Consultant, Department of Radiology, Horizon Imaging , Morbi, Rajkot, Gujarat, India .
Abstract
INTRODUTION: Spatially resolved metabolic data obtained from Proton Magnetic Resonance Spectroscopy ((1)H MRS) provides information which increases the diagnostic accuracy of imaging sequences in predicting the histology of suprasellar tumors. AIM: To evaluate the role of (1)H MRS in the diagnosis of various suprasellar tumors. MATERIALS AND METHODS: Sixty cases of various suprasellar, hypothalamic and third ventricular neoplasms were investigated with long-echo single voxel (1)H -MRS using 1.5 Tesla clinical imager. Single-voxel spectroscopic examinations were guided by T1-weighted or T2-weighted images. Statistical analysis was carried out using IBM SPSS software version 19. RESULTS: We observed that whenever brain tissue was damaged or replaced by any process, NAA was markedly reduced. Extra-axial lesions which do not infiltrate brain or contain neuroglial tissue, didn't demonstrate any NAA resonances. Cr was used as an internal standard for semi-quantitative evaluation of metabolic changes of other brain metabolites. Increased Cho was seen in processes with elevated cell-membrane turnover. CONCLUSION: Spectra obtained from different tumors exhibit reproducible differences while histologically similar tumors yield characteristic spectra with only minor differences. Pituitary tumors were typically characterized by significant reduction of NAA, Cr peak and moderate elevation of Cho peak. Gliomas were typically characterized by decrease of NAA and Cr peaks and increase of Cho peak. Craniopharyngiomas were typically characterized by significant decrease of all metabolites.
INTRODUTION: Spatially resolved metabolic data obtained from Proton Magnetic Resonance Spectroscopy ((1)H MRS) provides information which increases the diagnostic accuracy of imaging sequences in predicting the histology of suprasellar tumors. AIM: To evaluate the role of (1)H MRS in the diagnosis of various suprasellar tumors. MATERIALS AND METHODS: Sixty cases of various suprasellar, hypothalamic and third ventricular neoplasms were investigated with long-echo single voxel (1)H -MRS using 1.5 Tesla clinical imager. Single-voxel spectroscopic examinations were guided by T1-weighted or T2-weighted images. Statistical analysis was carried out using IBM SPSS software version 19. RESULTS: We observed that whenever brain tissue was damaged or replaced by any process, NAA was markedly reduced. Extra-axial lesions which do not infiltrate brain or contain neuroglial tissue, didn't demonstrate any NAA resonances. Cr was used as an internal standard for semi-quantitative evaluation of metabolic changes of other brain metabolites. Increased Cho was seen in processes with elevated cell-membrane turnover. CONCLUSION: Spectra obtained from different tumors exhibit reproducible differences while histologically similar tumors yield characteristic spectra with only minor differences. Pituitary tumors were typically characterized by significant reduction of NAA, Cr peak and moderate elevation of Cho peak. Gliomas were typically characterized by decrease of NAA and Cr peaks and increase of Cho peak. Craniopharyngiomas were typically characterized by significant decrease of all metabolites.
Authors: Lester Kwock; J Keith Smith; Mauricio Castillo; Matthew G Ewend; Sharon Cush; Thomas Hensing; Mahesh Varia; David Morris; Thomas W Bouldin Journal: Technol Cancer Res Treat Date: 2002-02