| Purpose:To explore the clinical practical value of1H-MRSI in the diagnosis and differential diagnosis of intracranial lesions, in the differentiation between recurrent glioma and radiation necrosis, in the grading of gliomas, in the surgical planning of gliomas and in the judgement to whether glioma was totally resected during the operation.Methods:Part one:The data o1'H-MRSI was analyzed prospectively in 275 patients(154 males and 121 females; mean age:43.2 years; range:1-76 years). The value of Choline (Cho) and N-acetyle aspartate(NAA) was studied to make clear tumors or nontumorous lesions, and gliomas or non-gliomas. Glioma was graded, recurrent glioma and radiation necrosis was differentiated, according to the value of Cho, the ratio of Cho/NAA and Creatine(Cr)/NAA, especially the ratio of Cr/NAA in the peritumoural regions. Magnetic resonance spectroscopy (MRS) diagnosis was compared with the conventionial histopathology of tumor.Part two:In correspondence with1H-MRSI, the boundary of each supratentorial gliomas was outlined, surgical planning was prepared with neuro-navigation so as to guide the resection of tumor. Part three:'H-MRSI was performed with intra-operation coil in 19 glioma cases. It was utilized to point out whether the tumor was totally resected or residual in 17 cases. MRS diagnosis was supposed to identify the type of the tumor in 2 cases.Results:Part one:Among all of the 275 cases, there were 100 cases of non-gliomas, and 175 cases of gliomas and recurrent gliomas. There were 30 incorrect MRS diagnosis in these 175 cases. In the differentiation between nontumorous lesions and tumors, there were 246 correct MRS diagnosis and 29 incorrect, the Kappa index was 0.662, the Youden index was 0.816, and the coincidence ratio was 89.5%. In the differentiation between gliomas and non-gliomas, there were 238 correct MRS diagnosis and 37 incorrect, the Kappa index was 0.723, the Youden index was 0.759, and the coincidence ratio was 86.5%. Of 33 patients who had previous tumor resection, there were 31 correct MRS diagnosis and 2 incorrect, the Kappa index was 0.798, the Youden index was 0.929, and the coincidence ratio was 93.9% in the differentiation between recurrent glioma and radiation necrosis. In the grading of 145 gliomas, the Kappa index, the Youden index, and the coincidence ratio was 0.865,0.897,93.8% vs 0.764,0.771,88.3% in the groups of low grade glioma (LGG) and high grade glioma (HGG) respectively. Part Two:The data of 1H-MRSI was analyzed in 134 gliomas, of which MRS grading was correct in 119 cases. In correspondence with 1H-MRSI, the boundary of these tumors was outlined exactly.43 cases underwent ordinary neuronavigation surgery, and 76 underwent iMRI neuronavigation surgery. Among all of the 76 cases with iMRI neuronavigation surgery, there were 22 LGG,6 anaplastic gliomas, and 48 HGG. Total tumor resection was achieved in 62 cases, subtotal resection 13 cases, and partial resection 1 case. Part three: Among the 19 cases which were examined with intraoperative coil, it was unclear to point out whether the tumor was totally resected or still residual in 17 cases. There were 2 correct MRS diagnosis to dentify the type of the tumor.Conclusion:There is higher clinical practical value of 1H-MRSI in the differentiation between non-tumorous lesions and tumors, between gliomas and non-gliomas, between recurrent glioma and radiation necrosis, and in the grading of gliomas, especially in the last two.'H-MRSI is helpful to outline the edge of glioma exactly, to prepare for the surgical planning with neuro-navigation perfectly and to guide tumor resection effectively. For several factors, it still need more work to do to confirm the clinical practical value of intra-operative 'H-MRSI in pointing out whether the tumor was totally resected or still residual. |
PDF Full Text Request