Local image variance of 7 Tesla SWI is a new technique for preoperative characterization of diffusely infiltrating gliomas: correlation with tumour grade and IDH1 mutational status

Eur Radiol DOI 10.1007/s00330-016-4451-y MAGNETIC RESONANCE Local image variance of 7 Tesla SWI is a new technique for preoperative characterization of diffusely infiltrating gliomas: correlation with tumour grade and IDH1 mutational status Günther Grabner 1,2,3 & Barbara Kiesel 2,4 & Adelheid Wöhrer 2,5 & Matthias Millesi 2,4 & Aygül Wurzer 2,4 & Sabine Göd 1 & Ammar Mallouhi 2,6 & Engelbert Knosp 2,4 & Christine Marosi 2,7 & Siegfried Trattnig 1,2 & Stefan Wolfsberger 2,4 & Matthias Preusser 2,7 & Georg Widhalm 2,4 Received: 16 February 2016 / Revised: 29 April 2016 / Accepted: 25 May 2016 # The Author(s) 2016. This article is published with open access at Springerlink.com Abstract Objectives To investigate the value of local image variance (LIV) as a new technique for quantification of hypointense microvascular susceptibility-weighted imaging (SWI) structures at 7 Tesla for preoperative glioma characterization. Electronic supplementary material The online version of this article (doi:10.1007/s00330-016-4451-y) contains supplementary material, which is available to authorized users. * Georg Widhalm 1 High Field Magnetic Resonance Centre, Department of Biomedical Imaging and Image-Guided Therapy, Medical University of Vienna, Waehringer Guertel 18-20, 1097 Vienna, Austria 2 Comprehensive Cancer Center, Central Nervous System Tumours Unit (CCC-CNS), Medical University of Vienna, Waehringer Guertel 18-20, 1097 Vienna, Austria 3 Department of Health Sciences and Social Work, Carinthia University of Applied Sciences, St. Veiterstraße 47, 9020 Klagenfurt am Wörthersee, Austria 4 Department of Neurosurgery, Medical University of Vienna, Waehringer Guertel 18-20, 1097 Vienna, Austria 5 Institute of Neurology, Medical University of Vienna, Waehringer Guertel 18-20, 1097 Vienna, Austria 6 Department of Radiology, Medical University of Vienna, Waehringer Guertel 18-20, 1097 Vienna, Austria 7 Department of Internal Medicine I, Medical University of Vienna, Waehringer Guertel 18-20, 1097 Vienna, Austria Methods Adult patients with neuroradiologically suspected diffusely infiltrating gliomas were prospectively recruited and 7 Tesla SWI was performed in addition to standard imaging. After tumour segmentation, quantification of intratumoural SWI hypointensities was conducted by the SWI-LIV technique. Following surgery, the histopathological tumour grade and isocitrate dehydrogenase 1 (IDH1)-R132H mutational status was determined and SWI-LIV values were compared between low-grade gliomas (LGG) and high-grade gliomas (HGG), IDH1-R132H negative and positive tumours, as well as gliomas with significant and non-significant contrast-enhancement (CE) on MRI. Results In 30 patients, 9 LGG and 21 HGG were diagnosed. The calculation of SWI-LIV values was feasible in all tumours. Significantly higher mean SWI-LIV values were found in HGG compared to LGG (92.7 versus 30.8; p < 0.0001), IDH1R132H negative compared to IDH1-R132H positive gliomas (109.9 versus 38.3; p < 0.0001) and tumours with significant CE compared to non-significant CE (120.1 versus 39.0; p < 0.0001). Conclusions Our data indicate that 7 Tesla SWI-LIV might improve preoperative characterization of diffusely infiltrating gliomas and thus optimize patient management by quantification of hypointense microvascular structures. Key Points • 7 Tesla local image variance helps to quantify hypointense susceptibility-weighted imaging structures. • SWI-LIV is significantly increased in high-grade and IDH1R132H negative gliomas. • SWI-LIV is a promising technique for improved preoperative glioma characterization. • Preoperative management of diffusely infiltrating gliomas will be optimized. Eur Radiol Keywords 7 Tesla MRI . Diffusely infiltrating gliomas . Susceptibility-weighted imaging . Local image variance . Glioma characterization Abbreviations 5-ALA 5-aminolevulinic acid CE Contrast-enhancement CM Contrast medium CSI Chemical shift imaging DWI Diffusion weighted imaging FWHM Full width at half maximum GBM Glioblastoma multiforme HGG High-grade gliomas IDH1 Isocitrate dehydrogenase 1 LGG Low-grade gliomas LIV Local image variance MINC Medical imaging network common data MPRAGE Magnetization prepared rapid gradient echo NAWM Normal appearing white matter nCBV Normalized cerebral blood volume PET Positron emission tomography RANO Response Assessment in Neuro-Oncology rCBV Relative cerebral blood volume ROI Region of interest SD Standard deviation SWI Susceptibility-weighted imaging WHO World Health Organization Introduction Diffusely infiltrating gliomas are the most frequent primary brain tumours in adults [1]. According to the current World Health Organization (WHO) criteria, the histopathological spectrum of diffusely infiltrating gliomas ranges from slowly growing tumours (low-grade gliomas = LGG; WHO grade II) to highly malignant neoplasms (high-grade gliomas = HGG; WHO grades III and IV) [2]. Following neurosurgical resection or biopsy of HGG, immediate postoperative treatment with radio- and/or chemotherapy is crucial, while in most LGG maximal safe tumour resection without initial postoperative therapy is performed [3–5]. LGG typically show malignant progression to HGG within several years, where the formation of pathological microvessels by neo-angiogenesis represents one of the key steps [2, 6]. Thus, the detection of these pathological microvascular structures is crucial for histopathological differentiation of LGG from HGG: while in LGG (WHO grade II) angiogenic features are typically absent, glioblastoma multiforme (GBM; WHO grade IV), the most common and malignant form of glioma, is characterized by the presence of pathognomonic microvascular proliferates [2, 6]. Nowadays, new molecular markers have been introduced that are capable of further refining the classification of gliomas into distinct subtypes [7, 8]. Most notably, presence of the isocitrate dehydrogenase 1 (IDH1) mutation was shown to be associated with WHO grade II/III gliomas and secondary GBM as well as a significantly longer progression-free and overall survival [7–10]. By far the most common IDH1 mutation involves the amino acid 132 at exon 4 (IDH1-R132H) [11]. More and more, molecular markers such as the IDH1 mutational status are increasingly incorporated in clinical decision making in addition to the tumour grade. Furthermore, it has been recently demonstrated that IDH1 mutant gliomas particularly profit from aggressive tumour resections [12, 13]. Similarly to different microvascular patterns in gliomas of various grades of malignancy, neo-angiogenesis, and thus formation of pathological microvessels was found to be associated with IDH1/2 mutational status with increased neoangiogenesis in IDH1/2 wild-type gliomas and inhibition of neo-angiogenesis in IDH1/2 mutant tumours [14]. Thus, reliable identification of these pathological microvascular structures is essential for preoperative glioma characterization to plan the appropr (...truncated)