Work

A Comparative Study of Advanced DCE Tracer Kinetic Models for IDH1 Genotyping and Progression-Free Survival Prediction in Gliomas

Year: 2026
Type: preprint
Study design: Observational Study
Abstract: Objective To compare quantitative dynamic contrast-enhanced (qDCE)-MRI tracer kinetic models for differentiating IDH1 subtypes and to assess their value in predicting progression-free survival (PFS) in high-grade gliomas (HGGs). Materials and Methods This retrospective study analyzed preoperative qDCE-MRI from 80 patients with pathologically confirmed glioma. Quantitative parameters from TOFTS, EXTOFTS, ATH, BRIX, DP, and QTM models were computed. ROC analysis was used to assess IDH1 prediction. Cox proportional hazards regression was used to evaluate PFS prediction in 34 HGGs. Results The distributed parameter (DP) model’s cerebral blood volume (CBV) was the best single predictor of IDH1 status (AUC = 0.846). A combined model incorporating DP-CBV, QTM-derived fluid flow velocity, and tumor location achieved the highest diagnostic performance, with an apparent AUC of 0.909, a bootstrap .632 + AUC of 0.898, and a stratified 5-fold cross-validation AUC of 0.881 ± 0.011. The DP model’s permeability-surface area product (PS) was an independent predictor of poor PFS in HGG (HR = 2.22; P < 0.001). The prognostic model combining DP parameters and clinical factors had the highest apparent C-index (0.870), significantly outperforming a clinical-only model. Correlations were observed between key parameters (e.g., DP-CBV/PS) and angiogenesis/proliferation markers (VEGF, CD31, Ki-67; P < 0.05). Conclusion Advanced qDCE-MRI models, particularly the DP model, effectively distinguish IDH1 molecular subtypes and predict PFS in HGG. DP-CBV is the optimal single imaging biomarker for IDH1 genotyping, further enhanced by combining with QTM-u and tumor location. In the exploratory PFS analysis, DP-PS is an independent prognostic imaging marker for poor PFS. Integration of these qDCE-MRI parameters with clinical assessment may enhance diagnostic and prognostic stratification for personalized glioma management.
Language: English
Subfield: Oncology
Field: Medicine
Open Access status: green
OpenAlex ID: W7236683270
Cites: 44
Cited by:
Related to: 10