Predicting IDH and 1p/19q molecular status of gliomas with multi-b values DWI

Abstract

Background and purpose: In the 2021 WHO Classification, the importance of molecular pathology in glioma diagnosis has been emphasized, particularly the status of isocitrate dehydrogenase (IDH) mutation and 1p/19q co-deletion. Advanced magnetic resonance diffusion-weighted imaging (DWI) including mono-exponential (Mono), intravoxel incoherent motion (IVIM), stretched exponential model (SEM) techniques are beneficial for non-invasive prediction of these molecular markers. The continuous-time random walk (CTRW) model mitigates the empirical nature of the SEM and has shown promising results in grading gliomas. However, the application of CTRW model in prediction of IDH and 1p/19q molecular phenotypes in adult diffuse gliomas remains underreported. This study compares the clinical utility of mono-exponential, IVIM, SEM, and CTRW models for predicting IDH and 1p/19q molecular status in adult diffuse gliomas.

Materials and methods: Data of adult diffuse glioma patients from January 2021 to August 2023 were collected. The multi-b-value DWI was acquired using a spin-echo echo-planar imaging sequence with 13 b-values (0, 10, 20, 30, 50, 70, 100, 150, 200, 400, 800, 1500, 2000 s/mm²) in 30 diffusion-encoding directions. Multi-b-value DWI images were post-processed to generate parametric maps based on the mono-exponential (Mono), the intravoxel incoherent motion (IVIM), the stretched exponential model (SEM) and the continuous-time random walk (CTRW) models. The mean parameter values of solid tumor regions were calculated. An independent sample t-test or Mann-Whitney U test was used for comparisons between different subtypes of glioma. Receiver operating characteristic (ROC) analyses were used to assess diagnostic performance.

Results: A total of 95 glioma patients were included in the study. For predicting IDH status, CTRW_α exhibited the largest effect size and best diagnostic performance with an AUC of 0.761. At a threshold of 0.855, the sensitivity was 0.651, the specificity was 0.846, and the accuracy was 0.758. In predicting 1p/19q status in IDH-mutant gliomas, CTRW_α again showed the largest effect size and the best diagnostic performance with an AUC of 0.790. At a threshold of 0.886, sensitivity was 0.750, specificity was 0.903, and accuracy was 0.860.

Conclusions: The CTRW model could help predict IDH and 1p/19q status in adult diffuse gliomas.

Keywords: 1p/19q; IDH; continuous time random walk; diffusion magnetic resonance imaging; glioma.

Figure 1

The red areas represent the delineated VOIs. (A) A 52-year-old female with IDH wild-type glioblastoma in the left basal ganglia, CTRW_α = 0.714, below the IDH threshold (0.855), correctly classified by imaging analysis. (B) A 38-year-old female with IDH-mutant & 1p/19q co-deletion oligodendroglioma in the left temporal lobe, CTRW_α = 0.872, above the IDH threshold and below the 1p/19q threshold (0.886), correctly classified by imaging analysis. (C) A 38-year-old female with IDH-mutant astrocytoma without 1p/19q co-deletion in the right frontal lobe, CTRW_α = 0.916, above the IDH or 1p/19q thresholds, correctly classified by imaging analysis.

Figure 2

ROC curve of IDH status predicting by DWI parameters.

Figure 3

ROC curve of 1p/19q status predicting by DWI parameters.

Figure 4

Box plot of the distribution of each parameter in three subtypes of gliomas. n.s.: no significant difference, *P = 0.01~ 0.05, **P = 0.001~ 0.01, ***P < 0.001.