Diagnostic and Interventional Radiology
Neuroradiology - Original Article

The reliability and interobserver reproducibility of T2/FLAIR mismatch in the diagnosis of IDH-mutant astrocytomas

1.

Department of Radiology, Muğla Sıtkı Koçman University School of Medicine, Muğla, Turkey

2.

Department of Radiology Acıbadem University School of Medicine, İstanbul, Turkey

3.

Department of Neurosurgery, Acıbadem Mehmet Ali Aydınlar University School of Medicine, İstanbul, Turkey

4.

Department of Pathology, Acıbadem Mehmet Ali Aydınlar University School of Medicine, Acıbadem Altunizade Hospital, İstanbul, Turkey

5.

Departments of Neurosurgery Acıbadem University School of Medicine, İstanbul, Turkey

6.

Department of Radiology, Acıbadem Mehmet Ali Aydınlar University School of Medicine, İstanbul, Turkey

Diagn Interv Radiol 2021; 27: 796-801
DOI: 10.5152/dir.2021.20624
Read: 166 Downloads: 48 Published: 03 November 2021

PURPOSE: The reliability and reproducibility of T2-weighted imaging/ fluid-attenuated inversion recovery (T2/FLAIR) mismatch were investigated in the diagnosis of isocitrate dehydrogenase (IDH) mutant astrocytoma between WHO grade II and III diffuse hemispheric gliomas. 

METHODS: WHO grade II and grade III diffuse hemispheric gliomas (n=133) treated in our institute were included in the study. Pathological findings and molecular markers of the cases were reviewed with the criteria of WHO 2016. The finding of mismatch between T2-weighted and FLAIR images in preoperative magnetic resonance imaging (MRI) of the cases was evaluated by two different radiologists. The readers reviewed MRIs independently, blinded to the histopathologic diagnosis or molecular subset of tumors. The cases were classified as IDH-mutant astrocytoma, oligodendroglioma and IDH-wildtype (IDH-wt) astrocytoma according to molecular and genetic features.

RESULTS: T2/FLAIR mismatch positivity was observed in 46 patients (34.6%). T2/FLAIR mismatch positivity was observed in 42 of 75 IDH-mutant astrocytomas (56%) and 4 of 43 oligodendrogliomas (9.30%), while it was not seen among IDH-wt astrocytomas (0/15, 0%). The T2/FLAIR mismatch ratio was significantly different between IDH-mutant astrocytomas (WHO grade II and grade III) and oligodendrogliomas (chi-square, p <0.05). The sensitivity, specificity, positive predictive value, negative predictive value, and accuracy of T2/FLAIR mismatch in predicting IDH-mutant astrocytomas were 58.7%, 90.7%, 91.7%, 61.4%, and 70.3% respectively. Radiologist 1 diagnosed T2/FLAIR mismatch in 48 of 133 cases (36.1%) and Radiologist 2 in 66 of 133 cases (49.6%). The interrater agreement for the T2/FLAIR mismatch sign was 0.61 (p <0.05), 95% CI (0.55, 0.67).

CONCLUSION: T2/FLAIR mismatch appears to be an important MRI finding in distinguishing IDH-mutant astrocytomas from other diffuse hemispheric gliomas. However, it should be kept in mind that T2/FLAIR mismatch sign can be seen in a minority of oligodendrogliomas besides IDH-mutant astrocytomas.

You may cite this article as: Yeniçeri İÖ, Yılmaz ME, Erşen-Danyeli A, Özduman K, Pamir MN, Dinçer A. The reliability and interobserver reproducibility of T2/FLAIR mismatch in the diagnosis of IDH-mutant astrocytomas. Diagn Interv Radiol 2021; 27:796–801.

Files
EISSN 1305-3612