Fractal radiomics as a virtual biopsy in glioma: A systematic review and meta-analysis of molecular status, radiogenomic associations, and survival outcomes.

BACKGROUND

Glioma diagnosis increasingly requires molecular evaluation; however, surgical biopsy has inherent limitations due to intratumoral heterogeneity and invasiveness. Fractal geometry provides a mathematical framework to quantify tumor architecture, contour irregularity, and subcomponent complexity as a non-invasive approach. Isocitrate dehydrogenase (IDH) and telomerase reverse transcriptase (TERT) alterations are important molecular markers in glioma classification and prognosis.

OBJECTIVES

This systematic review and meta-analysis aimed to evaluate the diagnostic performance of fractal-based imaging features for predicting molecular alterations, particularly IDH and TERT, and to assess their prognostic significance for survival and disease progression in glioma.

METHODS

A systematic search across PubMed, SCOPUS, and ScienceDirect identified was performed. This study followed PRISMA guidelines and JBI critical appraisal tools and the Radiomics Quality Score (RQS). Meta-analysis was performed using a random-effects inverse-variance model to pool diagnostic effect sizes and hazard ratios (HRs) for survival outcomes. Statistical heterogeneity was assessed via the I2 statistic.

RESULTS

Quantitative pooling for IDH mutation prediction showed a significant diagnostic effect size of 2.10 (95 % CI: 1.20-3.01). Fractal dimension (FD) of tumor contours and enhancing subcomponents had an area under the curve (AUC) values up to 0.97. Prediction of TERTp mutations showed a pooled effect size of 1.29, with peritumoral edema identified as the most predictive region of interest (ROI). A pooled HR for OS of 2.10 (95 % CI: 1.48-2.97), suggested that high geometric complexity in the enhancing rim or spatial chaos in the necrotic core (FD < 1.56) is associated with a two-fold increase in mortality risk.

CONCLUSION

Fractal radiogenomics and fractal analysis show promise as non-invasive virtual biopsy tools for molecular characterization and survival prediction in glioma. Quantification of structural complexity may complement conventional imaging and provide additional information regarding IDH and TERTp status and clinical outcomes.