¹H-MRS metabolic patterns for distinguishing between meningiomas and other brain tumors

Abstract

Yet meningiomas have characteristic neuroimaging features, some other lesions are still confusing with meningiomas. The aim of this study was trying to find the typical ¹H-MRS metabolic factors of histologic subtyped meningiomas, schwannomas, metastases, and other brain tumors for differential diagnosis among them. ¹H-MRS using STEAM (TE/30 ms, TR/2 sec) and PRESS (TE/288 ms, TR/2 sec) sequences were performed on 44 untreated brain tumors. Obtained metabolic patterns from the typical spectra of meningioma, schwannoma, metastasis were compared with each other or other brain tumors to evaluate the usefulness for diagnosis between them. Alanine(Ala) was observed in 15 cases of the 19 meningiomas with a little variation to three histologic subtypes, while minimal lipids were observed in every 19 meningiomas. Elevated glutamate/glutamine(Glx) was detected in 12 cases of the meniningiomas. Increased myo-inositol(mI) was detected in 11 cases of the 13 schwannomas. Dominant lipids signals as well as long-T2 lipids were detected in every metastasis in conjunction with elevated choline (Cho). Enhanced Glx was observed in 4 cases of the 8 metastases without correlation of primary tumor site or types. Hemangiopericytoma showed different spectral patterns from typical meningiomas: only dominant Cho, minimal lipids and absence of Ala or Glx signals. These metabolic patterns in typical tumors may provide a basis for differential diagnosis (average value of χ² = 23.33, p < 0.01) between meningiomas and schwannomas as well as metastases. However proton spectral distinction among the different histologic subtypes of meningiomas was not definite.

Introduction

With advent of MRI, the quality of preoperative diagnostic imaging for meningiomas have been improved dramatically and most of cases it is not difficult to diagnose meningiomas with conventional MRI with gadolinium-DTPA enhancement. With current MR imaging technique, dural attachment, sinus involvement and even information about a histologic subtype, vascularity and tumor’s consistency are possible. [1] Yet meningiomas have characteristic neuroimaging features, some other lesions are still confusing with meningiomas. Especially it is not easy to distinguish menigiomas from hemagiopericytomas, schwannomas occurring at the cerebello-pontine angle, dural metastatic tumors mimicking meningiomas. Surgical procedures or therapeutic plan of each tumorous condition may be quite different depending on the preoperative impression. [2] When diagnosing atypical tumor or tumor-like transformation, and preoperative diagnosis is not confident just by using conventional MRI and even using contrast-enhanced MRI methods, it is frequently requested to rule out meningiomas unambiguously. In addition the more detailed identification of the subtype of meningiomas may provide valuable information prior to surgical planning.

Studies of extracts of surgical specimens have generally indicated an increase in Alanine in meningiomas and an increase in myo-inositol in schwannomas. [3], [4] Elevation of Glx were observed in the extracts of surgically excised samples of meningiomas [5] and in an in vivo study. [6] However these findings have been still controversial without extended in vivo studies with multiple cases of meningiomas or schwannomas. In vivo MRS studies of human metastasis show a similar correlation between metabolic features and histopathological grade. Lactate was more likely to be present in late stage than in early or intermediate stage metastasis. [7], [8] Lipid was recognized to occur in tumors of higher histopathological grade, [8] and was correlated with the amount of microscopic cellular necrosis. [7], [9] However at present MRS cannot reliably characterize histologic types or subtypes of tumors in the clinical routine as the metabolic patterns seen in in vivo MRS are not sufficiently correlating to histologic diagnosis.

In this study localized proton spectroscopy was investigated on the untreated tumors diagnosed as meningiomas, schwannomas, metastases and other brain lesions by only the MRI basis, which were substantially confirmed with biopsy. We evaluated the potential significance of MRS findings of new metabolic markers in the pre-operative differential diagnosis of these brain tumors. Spectral characteristics of meningiomas were investigated according to the histologic subtypes retrospectively.