Preoperative proton MR spectroscopic imaging of brain tumors: Correlation with histopathologic analysis of resection specimens

Citation
C. Dowling et al., Preoperative proton MR spectroscopic imaging of brain tumors: Correlation with histopathologic analysis of resection specimens, AM J NEUROR, 22(4), 2001, pp. 604-612
Citations number
67
Categorie Soggetti
Radiology ,Nuclear Medicine & Imaging","Neurosciences & Behavoir
Journal title
AMERICAN JOURNAL OF NEURORADIOLOGY
ISSN journal
01956108 → ACNP
Volume
22
Issue
4
Year of publication
2001
Pages
604 - 612
Database
ISI
SICI code
0195-6108(200104)22:4<604:PPMSIO>2.0.ZU;2-9
Abstract
BACKGROUND AND PURPOSE: Tumor progression is often difficult to distinguish from nonneoplastic treatment response on the basis of MR images alone. Thi s study correlates metabolite levels measured by preoperative MR spectrosco pic (MRS) imaging with histologic findings of biopsies, obtained during ima ge-guided resections of brain mass lesions, to clarify the potential role o f MRS in making this distinction. METHODS: Twenty-nine patients with brain tumors underwent high-resolution ( 0.2-1 cc) 3D proton MRS imaging and MR imaging before undergoing surgery; 1 1 had a newly diagnosed neoplasm, and 18 had recurrent disease. Surgical bi opsies were obtained from locations referenced on MR images by guidance wit h a surgical navigation system, MR spectral voxels were retrospectively cen tered on each of 79 biopsy locations, and metabolite levels were correlated with histologic examination of each specimen. RESULTS: All mass lesions studied, whether attributable to tumor or noncanc erous effects of previous therapy, showed abnormal MR spectra compared with normal parenchyma, When the pattern of MRS metabolites consisted of abnorm ally increased choline and decreased N-acetyl aspartate (NAA) resonances, h istologic findings of the biopsy specimen invariably was positive for tumor . When choline and NAA resonances were below the normal range, histologic f indings were variable, ranging from radiation necrosis, astrogliosis, and m acrophage infiltration to mixed tissues that contained some low-, intermedi ate-, and high-grade tumor. CONCLUSION: This study demonstrated that 3D MRS imaging can identify region s of viable cancer, which may be valuable for guiding surgical biopsies and focal therapy. Regions manifesting abnormal MR spectra had a mixture of hi stologic findings, including astrogliosis, necrosis, and neoplasm.