TY - JOUR
T1 - Reliability of noncontrast-enhancing tumor as a biomarker of IDH1 mutation status in glioblastoma
AU - Lasocki, Arian
AU - Tsui, Alpha
AU - Gaillard, Frank
AU - Tacey, Mark
AU - Drummond, Katharine
AU - Stuckey, Stephen
PY - 2017/5/1
Y1 - 2017/5/1
N2 - Isocitrate dehydrogenase 1 (IDH1) mutations in gliomas have been associated with a frontal lobe location and a greater proportion of noncontrast-enhancing tumour (nCET). The purpose of our study was to validate the utility of MRI imaging features in predicting IDH1 mutations in glioblastomas. Pre-operative MRIs of new glioblastoma patients, consisting of at least FLAIR and T1-weighted post-contrast sequences, were reviewed by a neuroradiologist based primarily on the VASARI feature set. IDH1 mutation testing was performed on all patients using immunohistochemistry. 153 patients met the inclusion criteria, of whom five had IDH1 mutations (3.3%). A frontal lobe location had equivalent frequency in both the IDH1-mutated and IDH1-wildtype cohorts (p = 1.000). Three (60%) of the IDH1-mutated tumours had >33% nCET, compared to 21% of IDH1-wildtype (p = 0.073). 12 tumours had a frontal lobe epicentre and >33% nCET, all being IDH1-wildtype. All five IDH1-mutated tumours had either a frontal lobe epicentre or >33% nCET, but none had both these features. Our results question the strength of the association between frontal lobe glioblastomas with substantial nCET and IDH1 mutations, as these features are also relatively frequent in IDH1-wildtype tumours, which are much more common. MRI is thus more useful for ruling out an IDH1 mutation rather than strongly suggesting its presence: if a particular glioblastoma does not have a frontal lobe epicentre and has less than 33% nCET, it can be predicted to be IDH1-wildtype with a high degree of confidence.
AB - Isocitrate dehydrogenase 1 (IDH1) mutations in gliomas have been associated with a frontal lobe location and a greater proportion of noncontrast-enhancing tumour (nCET). The purpose of our study was to validate the utility of MRI imaging features in predicting IDH1 mutations in glioblastomas. Pre-operative MRIs of new glioblastoma patients, consisting of at least FLAIR and T1-weighted post-contrast sequences, were reviewed by a neuroradiologist based primarily on the VASARI feature set. IDH1 mutation testing was performed on all patients using immunohistochemistry. 153 patients met the inclusion criteria, of whom five had IDH1 mutations (3.3%). A frontal lobe location had equivalent frequency in both the IDH1-mutated and IDH1-wildtype cohorts (p = 1.000). Three (60%) of the IDH1-mutated tumours had >33% nCET, compared to 21% of IDH1-wildtype (p = 0.073). 12 tumours had a frontal lobe epicentre and >33% nCET, all being IDH1-wildtype. All five IDH1-mutated tumours had either a frontal lobe epicentre or >33% nCET, but none had both these features. Our results question the strength of the association between frontal lobe glioblastomas with substantial nCET and IDH1 mutations, as these features are also relatively frequent in IDH1-wildtype tumours, which are much more common. MRI is thus more useful for ruling out an IDH1 mutation rather than strongly suggesting its presence: if a particular glioblastoma does not have a frontal lobe epicentre and has less than 33% nCET, it can be predicted to be IDH1-wildtype with a high degree of confidence.
KW - Glioblastoma
KW - Isocitrate dehydrogenase
KW - Magnetic Resonance Imaging
KW - Noncontrast-enhancing tumour
UR - http://www.scopus.com/inward/record.url?scp=85012873108&partnerID=8YFLogxK
U2 - 10.1016/j.jocn.2017.01.007
DO - 10.1016/j.jocn.2017.01.007
M3 - Article
C2 - 28214089
AN - SCOPUS:85012873108
VL - 39
SP - 170
EP - 175
JO - Journal of Clinical Neuroscience
JF - Journal of Clinical Neuroscience
SN - 0967-5868
ER -