Histogram analysis of diffusion  weighted magnetic resonance imaging for predicting isocitrate dehydrogenase  1 mutation status in brain gliomas

Fatemeh  Shahedi; Shahrokh  Naseri; Mahdi  Momennezhad; Kazem  Anvari; Babak  Ganjeifar; Marzieh  Maleki; Parvaneh  Layegh; Masoumeh  Gharib; Amin  Ghaemmaghami; Hoda  Zare

doi:10.18502/cjn.v24i1.20297

Fatemeh Shahedi Medical Physics Research Center, Mashhad University of Medical Sciences, Mashhad, Iran
Shahrokh Naseri Medical Physics Research Center, Mashhad University of Medical Sciences, Mashhad, Iran
Mahdi Momennezhad Medical Physics Research Center, Mashhad University of Medical Sciences, Mashhad, Iran
Kazem Anvari Cancer Research Center, Mashhad University of Medical Sciences, Mashhad, Iran
Babak Ganjeifar Department of Neurosurgery, Ghaem Hospital, Mashhad University of Medical Sciences, Mashhad, Iran
Marzieh Maleki Department of Neurosurgery, Ghaem Hospital, Mashhad University of Medical Sciences, Mashhad, Iran
Parvaneh Layegh Department of Radiology, School of Medicine, Mashhad University of Medical Sciences, Mashhad, Iran
Masoumeh Gharib Department of Pathology, School of Medicine, Mashhad University of Medical Sciences, Mashhad, Iran
Amin Ghaemmaghami Medical Physics Research Center, Mashhad University of Medical Sciences, Mashhad, Iran
Hoda Zare Medical Physics Research Center, Mashhad University of Medical Sciences, Mashhad, Iran

DOI: https://doi.org/10.18502/cjn.v24i1.20297

Keywords: Diffusion Magnetic Resonance Imaging; Glioma; Isocitrate Dehydrogenase; Mutation; Brain

Abstract

Background: Gliomas are a major type of central nervous system (CNS) tumor. Accurate diagnosis of glioma grade and molecular subtype such as isocitrate dehydrogenase 1 (IDH1) mutation status remains a challenge as required invasive biopsy, which is limited by sampling bias and procedural risks. Quantitative analysis of functional magnetic resonance imaging (MRI), particularly apparent diffusion coefficient (ADC) maps, can serve as a non-invasive diagnostic tool for gliomas. However, using ADC values from different tumor regions may not accurately reflect the tumors’ heterogeneous nature. This study aims to investigate the diagnostic accuracy of histogram features of ADC maps across the entire tumor volume in differentiating between low-grade gliomas (LGGs) and high-grade gliomas (HGGs), as well as IDH1-wildtype from IDH1-mutated tumors.

Methods: This cross-sectional study included 30 patients with glioma who were assessed prior to undergoing surgical resectionmean, minimum, median, maximum, 10^th, 25^th, 75^th, and 90^th percentiles, mode, standard deviation (SD), kurtosis, inhomogeneity, skewness, and entropy, were obtained from the ADC maps. Statistical analysis was conducted to clarify associations between ADC histogram parameters, grade, and IDH1 mutation status. The sensitivity was determined to evaluate the performance of each parameter.

Results: The analysis revealed that 10^th percentile ADC (ADC10^th) had the highest sensitivity (87.5%, P = 0.0423) for discriminating between glioma grades and IDH1 mutation status, respectively.

Conclusion: The whole-tumor ADC histogram-profiling indicates potential value for predicting glioma grades and IDH1 molecular subtypes. However, further validation is required before clinical adoption.