Biopsy targeting with dynamic contrast-enhanced versus standard neuronavigation MRI in glioma: a prospective double-blinded evaluation of selection benefits.

Current biopsy planning based on contrast-enhanced T1W (CET1W) or FLAIR sequences frequently delivers biopsy samples that are not in concordance with the gross tumor diagnosis. This study investigates whether the quantitative information of transfer constant K maps derived from T1W dynamic contrast-enhanced MRI (DCE-MRI) can help enhance the quality of biopsy target selection in glioma. 28 patients with suspected glioma received MRI including DCE-MRI and a standard neuronavigation protocol of 3D FLAIR- and CET1W data sets (0.1 mmol/kg gadobutrol) at 3.0 T. After exclusion of five cases with no K-elevation, 2-6 biopsy targets were independently selected by a neurosurgeon (samples based on standard imaging) and a neuroradiologist (samples based on kinetic parameter K) per case and tissue samples corresponding to these targets were collected by a separate independent neurosurgeon. Standard technique and K-based samples were rated for diagnostic concordance with the gross tumor resection reference diagnosis (67 WHO IV; 24 WHO III and II) by a neuropathologist blinded for selection mode. K-based sample targets differed from standard technique sample targets in 90/91 cases. More K-based than standard imaging-based samples could be extracted. Diagnoses from K-based samples were more frequently concordant with the reference gross tumor diagnoses than those from standard imaging-based samples (WHO IV: 30/39 vs. 11/20; p = 0.08; WHO III/II: 12/13 vs. 6/11; p = 0.06). In 4/5 non-contrast-enhancing gliomas, K-based selection revealed significantly more accurate samples than standard technique sample-selection (10/12 vs. 2/8 samples; p = 0.02). If K elevation is present, K-based biopsy targeting provides significantly more diagnostic tissue samples in non-contrast-enhancing glioma than selection based on CET1W and FLAIR-weighted images alone.