PURPOSE: To determine the relative signal intensity ratios of choline (Cho), phosphocreatine (CR) and N-acetyl-aspartate (NAA) in MR spectroscopic imaging (proton-MRSI) for differentiating progressive tumors (PT) from non-progressive tumors (nPT) in follow-up and treatment planning of gliomas. Threshold values to indicate the probability of a progressive tumor were also calculated. MATERIAL AND METHODS: Thirty-four patients with histologically proven gliomas showing a suspicious brain lesion in MRI after stereotactic radiotherapy were evaluated on a 1.5 Tesla unit (Magnetom Vision, Siemens, Erlangen, Germany) using 2D proton MRSI (repetition time/echo time = 1500/135 msec, PRESS; voxel size 9 x 9 x 15 mm (3)). A total of 274 spectra were analyzed (92 voxel were localized within the suspicious brain lesion). Relative signal intensities Cho, Cr and NAA were measured and their ability to discern between PT and nPT was assessed using the linear discrimination method, logistic regression, and the cross-validation method. PT and nPT were differentiated between on the basis of clinical course and follow-up by MRI, CT and positron emission tomography. RESULTS: The Cho parameter and the relative signal intensity ratios of Cr and NAA were most effective in differentiating between PT and nPT. The logistic regression method using the parameter ln(Cho/Cr) and ln(Cho/NAA) had the best predictive results in cross-validation. A sensitivity of 93.8 % and specificity of 85.7 % were achieved in the differentiation of PT from nPT by proton-MRSI. CONCLUSION: (1)H-MRSI has a high sensitivity and specificity for differentiating between therapy-related effects and the relapse of irradiated gliomas. This method allows for assessment of the probability of radiotherapy response or failure.
PURPOSE: To determine the relative signal intensity ratios of choline (Cho), phosphocreatine (CR) and N-acetyl-aspartate (NAA) in MR spectroscopic imaging (proton-MRSI) for differentiating progressive tumors (PT) from non-progressive tumors (nPT) in follow-up and treatment planning of gliomas. Threshold values to indicate the probability of a progressive tumor were also calculated. MATERIAL AND METHODS: Thirty-four patients with histologically proven gliomas showing a suspicious brain lesion in MRI after stereotactic radiotherapy were evaluated on a 1.5 Tesla unit (Magnetom Vision, Siemens, Erlangen, Germany) using 2D proton MRSI (repetition time/echo time = 1500/135 msec, PRESS; voxel size 9 x 9 x 15 mm (3)). A total of 274 spectra were analyzed (92 voxel were localized within the suspicious brain lesion). Relative signal intensities Cho, Cr and NAA were measured and their ability to discern between PT and nPT was assessed using the linear discrimination method, logistic regression, and the cross-validation method. PT and nPT were differentiated between on the basis of clinical course and follow-up by MRI, CT and positron emission tomography. RESULTS: The Cho parameter and the relative signal intensity ratios of Cr and NAA were most effective in differentiating between PT and nPT. The logistic regression method using the parameter ln(Cho/Cr) and ln(Cho/NAA) had the best predictive results in cross-validation. A sensitivity of 93.8 % and specificity of 85.7 % were achieved in the differentiation of PT from nPT by proton-MRSI. CONCLUSION: (1)H-MRSI has a high sensitivity and specificity for differentiating between therapy-related effects and the relapse of irradiated gliomas. This method allows for assessment of the probability of radiotherapy response or failure.
Authors: T Scarabino; T Popolizio; F Trojsi; G Giannatempo; S Pollice; N Maggialetti; A Carriero; A Di Costanzo; G Tedeschi; U Salvolini Journal: Radiol Med Date: 2008-12-11 Impact factor: 3.469
Authors: Alexander M Spence; Mark Muzi; Jeanne M Link; Finbarr O'Sullivan; Janet F Eary; John M Hoffman; Lalitha K Shankar; Kenneth A Krohn Journal: Mol Imaging Biol Date: 2009-03-27 Impact factor: 3.488
Authors: Sarah J Nelson; Yan Li; Janine M Lupo; Marram Olson; Jason C Crane; Annette Molinaro; Ritu Roy; Jennifer Clarke; Nicholas Butowski; Michael Prados; Soonmee Cha; Susan M Chang Journal: J Neurooncol Date: 2016-08-17 Impact factor: 4.130