Predictions and cross-center validation of IDH status and 1p/19q status in adult diffuse glioma with non-invasive magnetic resonance spectroscopy.
Non-invasive metabolic profiling shows promise for predicting glioma subtypes, but further validation is needed for clinical use.
Where it sits
this study against the rest of the vip (vasoactive intestinal polypeptide) corpusSummary and findings
This study evaluated a non-invasive magnetic resonance spectroscopy framework for predicting IDH mutation and 1p/19q codeletion status in adult diffuse gliomas. A cross-center cohort of 268 patients was used to assess 48 features from 18 metabolites. The model showed high AUCs for IDH and 1p/19q predictions, suggesting potential for preoperative molecular subtyping.
Abstract
This study developed and validated a non-invasive multi-metabolite magnetic resonance spectroscopy framework for preoperative molecular subtyping of adult diffuse gliomas. Using a cross-center, cross-vendor cohort of 268 patients, 48 features derived from 18 metabolites were systematically evaluated to identify discriminative metabolic signatures for predicting IDH mutation and 1p/19q codeletion status. The integrated metabolic model showed robust intra-center and cross-center performance for IDH prediction, with AUCs of 0.906 and 0.857, respectively, and for 1p/19q prediction, with AUCs of 0.858 and 0.787, respectively. These results suggest that synergistic metabolic profiling support molecular assessment in patients who may not be suitable for invasive biopsy.
Background
The study addresses the need for non-invasive methods to determine molecular subtypes of adult diffuse gliomas, which are critical for prognosis and treatment planning. Current methods often require invasive biopsies, which may not be feasible for all patients. This research explores the potential of magnetic resonance spectroscopy to fill this gap.
Methods
A cross-center, cross-vendor cohort study was conducted with 268 patients diagnosed with adult diffuse gliomas. The study evaluated 48 features derived from 18 metabolites using magnetic resonance spectroscopy. The primary outcomes were the predictive accuracy of IDH mutation and 1p/19q codeletion status, measured by AUC values.
Results
The integrated metabolic model demonstrated robust performance, with AUCs of 0.906 and 0.857 for IDH mutation prediction intra-center and cross-center, respectively. For 1p/19q codeletion prediction, the AUCs were 0.858 intra-center and 0.787 cross-center. These findings suggest a strong potential for non-invasive molecular subtyping.
Interpretation
The study's findings align with the growing interest in non-invasive diagnostic tools for gliomas. While the AUC values indicate strong predictive power, the clinical significance of these findings remains to be fully validated in broader and more diverse patient populations. The reliance on metabolic signatures could offer a less invasive alternative to biopsies, but practical implementation challenges remain.
Key findings
- 268 patients in cross-center, cross-vendor cohort.
- 48 features derived from 18 metabolites evaluated.
- IDH prediction AUCs: 0.906 intra-center, 0.857 cross-center.
- 1p/19q prediction AUCs: 0.858 intra-center, 0.787 cross-center.
Limitations
- AUC metrics may not capture full clinical applicability.
- Requires further validation in diverse clinical settings.
- Potential implementation challenges in routine practice.