Predicting TERT mutation in glioma patients using MRI-derived radiomics; A systematic review and meta-analysis.
Zahra Moradi, Saeed Mohammadzadeh, Ehsan Mehrtabar et al.
✦ AI-curated · Sources linked
In 30 seconds
This systematic review and meta-analysis assessed MRI-derived radiomics models for predicting TERT promoter mutation status in patients with glioma. Across 17 included studies, pooled sensitivity was 83 % (95 % CI: 76 %-87 %) and specificity was 79 % (95 % CI: 71 %-85 %); the area under the curve was 0.88. Subgroup analysis found that diagnostic performance differed significantly according to segmentation approach. The authors concluded that radiomics may provide a non-invasive adjunct for preoperative molecular characterization and risk stratification.
Key findings
- Across 17 studies, pooled sensitivity was 83 % (95 % CI: 76 %-87 %) and specificity was 79 % (95 % CI: 71 %-85 %).
- The pooled area under the curve was 0.88.
- Diagnostic performance differed significantly by segmentation approach.
Why it matters
TERT promoter status is associated with malignant progression and poor survival in glioma. MRI-derived radiomics could offer a non-invasive estimate of this molecular feature, potentially adding information to preoperative assessment. The pooled results indicate useful but imperfect discrimination, so predictions should be interpreted as adjunctive rather than definitive molecular characterization.
What to keep in mind
Performance varied significantly by segmentation approach, indicating heterogeneity related to how imaging regions were defined.