Open Access

Oliver Söllner

• In this study the clinical value of the method of 31P und 1 H MRI spectroscopy is analyzed in the evaluation of tumors of the liver and the cerebrum. At first 39 patients (HCC n=30, metastases of colorectal carcinomas n=9) undergoing transarterial chemoembolization (TACE) were evaluated MR tomographically with 1.5 Tesla using 31P CSI spectroscopy. Moreover, 53 patients with cerebral tumors (17 meningiomas, 11 gliomas WHO grades I-II, 6 gliomas WHO grade III, 13 gliomas WHO grade IV and 6 metastases) were evaluated 1 H spectroscopically with the ISIS technique in different echo times. The results of both groups were correlated with the histopathological findings and compared with a study group. For evaluation the area under the curve of the measurable signal intensities were calculated, the ratios were determined and statistically evaluated. In patients with livertumors undergoing TACE, the 31P spectroscopy was performed before and after each course of TACE. Pretherapeutic evaluation revealed the tumor tissue with increased PME peak, PME/ß-ATP ratio, and PME/PDE ratio. In all cases the tumor spectres were to be differentiated from the spectra of the study group. If chemoembolization was technically successful, we found an increase in the Pi peak (+90.1%) and a decrease in the ß-ATP peak (-19.1%). After each course of therapy a number of patient groups could be differentiated depending on the changes in the different peaks and ratios. A response was characterized by a decrease of the PME/ß-ATP and PME/PDE ratios and an increase of the PDE/ß-ATP ratio. In non-responders, there was no decrease of the PME/ß-ATP and PME/PDE ratios, and these ratios increased 6 weeks later. The PDE/ß-ATP ratio decreased. Constant ratios were found if a steady state of the disease was achieved. Regrowth of tumor was accompanied by elevated PME and decreased PDE peaks. With regard to the 1 H spectroscopical findings the following statements can be made: The tumor spectra can be distinctly differentiated from the study group spectres. In this respect highly significant differences for the NAA/Cho and PCr/Cho ratios can be seen. The spectra of the meningiomas can be often characterized by the missing NAA. A small peak at 2.0 ppm can probably be due to a part of healthy brain tissue in the VOI at the rim of the tumor in some of the spectra. Moreover, some of the meningiomas show Alanin at 1.47 ppm, which, however, can also be overlain by fat signal in this area. On average, the PCr peak is reduced by half with regard to the referene; Inositol can hardly be detected even with short echo times. The metastases show a decreased NAA/Cho and PCr/Cho ratio. In few cases Ins/Cho can be measured, and then below the level of the study group. Additionally, two distinct peaks could be seen at 0.9 and 1.25 ppm according to strongly increased free fatty acids. All gliomas show a reduced NAA signal. In this respect, the reduction of the NAA/Cho ratio shows a nonsignificant dependence on malignity, which can be reflected in an almost completely reduced NAA signal in glioblastomas. PCr and Ins are also decreased. With increasing malignity of the lesion the Inositol signal increases and reaches the normal values of the study group. Using 1 H spectroscopy it is possible to support the differential diagnosis of the imaging modalities. Due to its sensitivity it is possible to use the 31P spectroscopy in therapy control. In order to establish these methods in the daily routine further improvements are necessary, particularly in regard to measurement sequences, automatisms and standardized evaluation protocols.