Prediction of Tumor Consistency of Meningiomas using Diffusion Tensor Imaging

Abstract

Introduction: Skull base meningiomas tend to be intimately attached to cranial nerves, major feeding vessel arteries as well as eloquent brain, thus making their resection quite challenging. In addition, endoscopic skull base approaches limit the natural dexterity available to traditional open surgeons. When deciding between a fully endoscopic approach versus an open approach, surgeons may benefit from knowledge of meningioma consistency (hard vs. soft) to facilitate decision making. Diffusion tensor imaging (DTI) relies on the principles of water diffusion, and we hypothesize that “soft” tumors have greater water diffusion in random directions whereas “hard” tumors have less water diffusion in random directions. In addition, we hypothesize that high content of intercellular collagen is responsible for the hard consistency. The goal of this study is to determine whether DTI can predict tumor consistency as determined by percent collagen contents. Materials and Methods: A total of 42 meningiomas with histopathologic diagnosis of atypical (n = 7; grade II, 7M; age, 47–80 years), anaplastic (n = 2; grade III, 2F; age, 44–56 years), and typical (n = 33;grade I; 8M/25F; age, 27–86 years) meningiomas were included in this study. Subtypes of typical meningiomas include 10 fibroblastic, 7 transitional, 15 meningothelial, and 1 angiomatous. All patients underwent MR studies before surgery on a 3T Siemens scanner. DTI data were acquired using a single shot spin echo EPI sequence with parallel imaging using GRAPPA (acceleration factor = 2). Contrast-enhanced T1-weighted images, FLAIR, and DTI metrics including fractional anisotropy (FA), mean diffusity (MD), linear anisotropy (CL), planar anisotropy (CP), and spherical anisotropy (CS) maps were coregistered and the median values were measured from the enhancing region. After the surgery, the histologic slides from the patients were evaluated for collagen content using sirius red stain. The percentage of collagen content was calculated using the following equation: Collagen content (%) = (collagen area/total tumor area) × 100. DTI metrics and collagen contents were compared between different subtypes. Results: Fibroblastic meningiomas demonstrated significantly higher collagen contents compared with other subtypes. The median FA and CP values were significantly higher in fibroblastic meningiomas compared with other subtypes and atypical meningiomas. The correlation coefficient between FA and collagen contents was 0.43 and between CP and collagen contents was 0.35. Conclusion: DTI may be helpful to predict tumor consistency as determined by collagen contents. This information may influence decision making in skull base surgery for meningiomas.