Premium
Musculoskeletal tumors: Use of proton MR spectroscopic imaging for characterization
Author(s) -
Fayad Laura M.,
Bluemke David A.,
McCarthy Edward F.,
Weber Kristin L.,
Barker Peter B.,
Jacobs Michael A.
Publication year - 2006
Publication title -
journal of magnetic resonance imaging
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.563
H-Index - 160
eISSN - 1522-2586
pISSN - 1053-1807
DOI - 10.1002/jmri.20448
Subject(s) - malignancy , medicine , nuclear medicine , coronal plane , magnetic resonance imaging , histopathology , sarcoma , pathology , magnetic resonance spectroscopic imaging , radiology
Purpose To determine the value of multivoxel proton magnetic resonance spectroscopic imaging (MRSI) in distinguishing malignant skeletal tumors from benign tumors and normal bone marrow using the metabolite choline (Cho) as a marker for malignancy. Materials and Methods Pathologic specimens obtained from 13 patients who had undergone wide resection for skeletal tumors underwent evaluation by MRSI at 1.5 T. Coronal T1‐weighted gradient‐echo sequence obtained for localization purposes (TR/TE = 250/1.8 msec, field of view [FOV] = 18 × 18), and single‐slice MRSI (TR/TE = 2000/272 msec, FOV = 18 × 18, 10‐mm slice‐thickness) were performed. Water, lipid, and Cho images were reconstructed from MRSI data. Cho signal was measured in each specimen and expressed relative to background noise level (signal‐to‐noise ratio [SNR]) where noise was measured between 7.0 and 9.0 ppm. Cho SNRs were compared between areas containing malignant tumor and nonmalignant tissue (benign lesion or normal bone marrow) as determined by histopathology. Results Specimens included 13 skeletal sarcomas (seven osteosarcomas, three chondrosarcomas, one malignant fibrous histiocytoma, one fibrosarcoma, and one leiomyosarcoma). All specimens included a sample of normal bone marrow and two specimens also contained benign lesions. All sarcomas demonstrated a signal at 3.2 ppm assigned to Cho‐containing metabolites in areas of malignancy. Peak Cho SNR was significantly different for areas containing histologically‐proven malignancy compared to nonmalignant tissue (9.8 ± 5.1 vs. 2.7 ± 1.4, respectively, P < 0.002). Conclusion These preliminary results indicate that MRSI at 1.5 T is a promising noninvasive method of differentiating malignant skeletal tumors from nonmalignant tissue. Using MRSI, Cho can be detected in skeletal tumors and may serve as a marker for malignancy. J. Magn. Reson. Imaging 2006. © 2005 Wiley‐Liss, Inc.