In a study of multiparametric MRI and digital histopathologic images (DHI) in glioma patients obtained from the Cancer Imaging Archive, a team of researchers determined that combining MRI and DHI data resulted in a more accurate prediction of overall survival (OS) for both high- and low-grade glioma patients than either MRI or DHI alone.

Key Points

• A set of 14 MRI and 12 histopathologic imaging features were identified as discriminative with respect to overall survival (OS) on the basis of a log-rank P value less than .01.
• Cox proportional hazard models that accounted for censored data were trained, and the discriminative performance was quantified by the concordance index (C-index) and Kaplan-Meier analysis
• OS time was predicted with a C-index of 0.793 with combined MRI and histopathologic imaging features, which was higher than 0.697 for MRI features alone and 0.669 for histopathologic features alone.

“Gliomas are tumors that originate in glial cells of neuroepithelial tissue and are a major cause of death. Gliomas are classified on the basis of their histopathologic characteristics into astrocytoma, oligodendroglioma, oligoastrocytoma, and ependymoma subtypes. Furthermore, the severity of the glioma is categorized from low (grade I) to high (grade IV) according to the World Health Organization grading system).”

The study was published online July 23, 2021 in Radiology: Imaging Cancer.

Source:

Saima Rathore, Ahmad Chaddad, Muhammad A. Iftikhar, Michel Bilello, Ahmed Abdulkadir,
“Combining MRI and Histologic Imaging Features for Predicting Overall Survival in Patients with Glioma” Radiology: Imaging Cancer 2021 3:4
https://pubs.rsna.org/doi/abs/10.1148/rycan.2021200108
Accessed August 2, 2021