Preoperative Nomogram-Based Assessment to Identify GBM Patients Who Do not Derive Survival Benefit From GTR Compared to STR.

Abstract

Rationale and objectives: Glioblastoma, IDH-wildtype (GBM), the most common primary malignant brain tumor in adults, has a median overall survival of 11-15 months. While gross total resection (GTR) generally improves survival compared to subtotal resection (STR), certain patient subgroups may not benefit from more extensive resection.

Methods: This study developed a nomogram-based predictive model using preoperative clinical and imaging data to identify GBM patients who may or may not benefit from GTR compared to STR. Data from the UCSF-PDGM dataset (N=371) were used to construct the model, with external validation performed using the UPENN-GBM dataset (N=457).

Results: Multivariate Cox regression identified age, extent of resection (EOR), and volume all (necrotic, enhancing, and peritumoral regions of tumor) as independent prognostic factors. The nomogram stratified patients into low-, medium-, and high-score groups based on age and tumor volume. Results showed that GTR significantly improved survival in patients with scores between 55 and 95, but not in those with scores below 55 or above 95. Younger patients with smaller tumors (usually with score <55) and older patients with larger tumors (usually with score >95) derived limited additional survival benefit from GTR compared to STR. The nomogram-based classification outperformed methods relying solely on age or tumor volume.

Conclusion: These findings suggest that preoperative assessment using the nomogram can guide individualized surgical strategies, optimizing the extent of resection for GBM patients. However, prospective studies are warranted to further validate the reliability of the findings in this research.