Diagnostic Performance of MRI,MRS and Perfusion weighted in Characterization of Brain Tumors

Abstract

The objectives of this study were to evaluate the effectiveness of perfusion and magnetic resonance spectroscopy in the characterization of brain tumors. And to evaluate the usefulness of MR spectroscopy (MRS) in grading of brain tumors as well to evaluate which metabolite values/ratios could provide better classification/grading of brain tumors when using, short or long TE. The study was conducted during the period extended from 2014 up to 2017, using 1.5 Tesla superconducting Syngo MRI system at Neuro-Surgery and Oncology department at Jazan- Saudi Arabia. The study was designed by obtaining a spectrum, that was analyzed, and was influenced by many parameters, including chemical characteristics of each metabolite and the compounds in which they are located; N-acetyl aspartate (NAA) , choline (Cho), creatine (Cr) Cho/ NAA, Cho/ Cr, NAA/ Cr , Lactate and Lipid at the selection of the lesion area to be studied. The perfusion MRI (rCBV) was characterized as normal, hyper, and hypo-perfused. Choice of the technique type to be employed was (PRESS and STEAM) and the choice of the echo time was (short and long). The diagnostic strategy was evaluated based on results of imaging of 128 patients who had completed data including age, tumor tertiary in the brain. Conventional MRI with standard diagnostic criteria, findings and application of scan with T1:TR˂500ms and TE ˂50ms and T2 :TR˃1500 and TE˃80ms were obtained.MRI with contrast enhancement in T1 weighted image and nullify the signal from CSF were also been achieved . Adequate scan was attained with Short echo times (TEs) less than 30ms for STEAM and TE as long as 144ms and 288ms milliseconds for PRESS technique. For MRS long TE (1500/144ms) was used; for identification (Cho), (Cr), (NAA), and lactate. IV

Conversely, short TE (2000/35ms) was applied for identification of Lipids, lactate, ALA, myoinositol, glutamate, and glutamine. The current study showed that the most common affected age was age between 41-50 years old constituting 34(26.6%). Distribution of tumor territory in the brain site has significant relation with age at p-value= 0.006.The entire lesions were diagnosed according to the standard criteria of diagnosis done by radiologist. Astrocytomas, were found in 58(45.3%) of the cases, Gliomatosis Cerebri, Glioblastic Multiform ( GBM) and Oligodendroglioma were 44 (34.4%) ,where the Lymphoma and Meningioma were found in 2(1.6%) and 8(6.3%) of the cases in respectively , where the metastases constituting 3(2.3%).Ependymal tumors were found to be 13(10.1%) 5 cases were Ependymoma and 8 were Subependymoma. Significant results between the perfusion findings and the MRS values regarding Cho/ NAA at p= (0.000) and Cho/ Cr at p = (.000).Astrocytomas showed a relative reduction in NAA and Cr, and Cho comparing with the lymphoma .The spectra of metastases are similar to those of meningioma and gliomatos cerebri and ependemal tumors, with low NAA, low Cr, and high Cho levels. The difference was found to be significant in the NAA values at different brain lesions with no significant reduction or increasing in Cho and Cr. Cho/Cr has significant impact in differentiation of lymphoma from other lesions at p value = 0.004 where the other parameters including Cho/ NAA, NAA/ Cr, Lactate and Lipid showed no significant relations. By combining both MRS and perfusion MRI, the diagnosis of lesions was achieved with value of 0.94±.89 for NAA and Cho/ NAA 1.83±1.22.

V

A significant relation was found between the diagnosis and tumor character in Perfusion-MRI (rCBV) as normal, hypo-perfused and hyper- perfused at p-value=0.001 as well; the STEAM and PRESS results at P-value = 0.042 and 0.042 in respectively. Perfusion-MRI (rCBV) and MRS are useful for establishing the differential diagnosis between brain metastases and brain tumors. Both has significant relation in differentiation and diagnosis of brain tumors regarding its perfusion as well in chemical characteristics of each metabolite and the compounds in which they are located; N-acetyl aspartate (NAA) , Cho/ NAA, Cho/ Cr at the selection of the area to study. So the combination of MRS and perfusion-weighted imaging may improve characterization of brain lesions. NAA, Cho, Cr, NAA/ Cr, Lactate and Lipids metabolic values showed no significant relation between the values to be increased or decreased according to grading therefore the dependency of grading should be based upon the Cho/ NAA and Cho/ Cr .From this account and metabolic ratios; the study suggested to diagnose, differentiate and grade gliomas after evaluating the Cho/ NAA and Cho/ Cr. Conclusion :MRS metabolic ratios (Cho/Cr and Cho/NAA) can be used to grade and differentiate gliomas. Ratios less than 1.5 were suggested to be considered as normal values, ratios from 1.5 to 2 were suggested to subsist as low grade glioma and ratios higher than 2 were suggested to be high grade glioma or metastasis. Meningioma can be diagnosed by conventional MRI images. MRS should be added to routine MR imaging studies as it provides greater information concerning tissue characterization than what is possible with MR imaging studies alone.