Supramarginal Resection of Glioblastoma

This article discusses the evidence supporting the resection of glioblastoma beyond the borders of contrast-enhancing tumor. While several techniques for this have been described, including a so-called FLAIRectomy, lobectomy, or via the use of adjuncts such as fluorescence or intraoperative MRI, the optimal extent of additional resection has yet to be established. Many authors have noted a survival benefit with supramarginal resection without significant additional morbidity.

Key points

•

Glioblastoma cells are commonly found beyond the borders of contrast-enhancing disease.
•

Early evidence suggests that resection of invaded but non-enhancing brain (ie, a supramarginal resection) may safely confer additional survival benefit.
•

Supramarginal resection may be guided by FLAIR resection, anatomic resection, or intraoperative adjuncts such as MRI or fluorescence.

Glioblastoma (GBM) is the most common primary brain tumor. Treatment consists of maximal safe resection followed by chemoradiation and then adjuvant temozolomide. ^, The standard of care has not changed in nearly 2 decades, as clinical trials have failed to show a survival benefit of other treatments. While there is relatively little debate as to the timing and dosage of chemotherapy and radiation, the optimal extent of resection has not clearly been defined. However, recurrence after even complete resection is universal. The discovery that tumor cells are present outside the contrast-enhancing portion of the tumor, within the region of fluid attenuation inversion recovery (FLAIR) attenuation, raises the question of supramarginal resection. While it may seem reasonable that additional resection resulting in removal of more infiltrated brain would prolong survival, a balance must be struck between possible oncologic benefit and the risks of neurologic deficit and of prolonged recovery in patients with little time to do so. The following review explores the role of supramarginal resection in patients with GBM.

Extent of resection in glioblastoma

Surgery for GBM serves to provide tissue for diagnostic confirmation and evaluation of molecular markers as well as to reduce oncologic burden and tumor mass effect. ^, The value of incremental increases in extent of resection (EOR) was once the subject of some debate, as early studies were plagued by statistical weakness and the unfeasibility of conducting a randomized trial of resection. Initially, it seemed surgery only benefitted patients in cases of near-total resection. In a series of 420 patients with GBM, LaCroix and colleague demonstrated that at least 89% resection was required to confer a meaningful survival benefit. A 2011 retrospective study by Sanai and colleagues analyzed the survival of 500 patients undergoing resection for GBM and found that 78% resection could confer a survival benefit. Survival curves generated at 2% increments above 78% resection demonstrated successive increases in survival with greater EOR, with patients undergoing 78% resection having a median survival of 12.7 months and patients with 100% resection surviving a median of 16 months. In this series, multivariate analysis showed that EOR and residual tumor volume independently predicted survival. As the molecular understanding of astrocytic tumors becomes more advanced, modern series have shed additional light on the effect of EOR on survival. Beiko and colleagues retrospectively analyzed 335 patients with malignant astrocytoma and found that complete resection of isocitrate dehydrogenase (IDH)-wildtype glioma (now GBM according to the World Health Organization [WHO] 2021 classification) conferred survival benefit over incomplete resection (19.6 months vs 10.7 months). Thus, the utility of removal of at least as much contrast-enhancing tumor as safely possible has been well established. However, the infiltrative nature of GBM obviates a surgical cure and has prompted consideration more aggressive resection to remove cells that have migrated beyond the tumor core.

The extent of cell migration and rationale for supramarginal resection

Contrast-enhanced, T1-weighted MRIs effectively define areas of blood–brain barrier disruption caused by GBM, but multiple factors suggest that they do not adequately assess the extent of the disease. First, 96% of GBMs recur within 3 cm of the original resection cavity. Second, viable tumor cells can be found within the area of FLAIR attenuation 4 to 10 cm beyond the border of contrast enhancement. Accordingly, the pattern and the extent of FLAIR attenuation surrounding glioma have been shown to be associated with survival. Histopathologic analysis of GBM samples from 51 patients conducted by Barajas and colleagues demonstrated that 81% of samples from non-enhancing tissue contained viable tumor cells. The extent of this invasion is not trivial: an autopsy study of patients with GBM conducted by Yamahara and colleagues found that neoplastic cell density 60% to 100% of that of the tumor itself was found up to an average of 1 cm from the border of contrast-enhancing tumor.

The cells isolated from beyond the margin of GBM are notably different from their intratumoral counterparts in ways that warrant their aggressive removal when safe to do so. These cells have been shown to be more aggressive than their counterparts at the tumor core. They display upregulation of genes required for protein catabolism, hypothesized to fulfill the increased energy requirement for migration. They are more resistant to cell–cell adhesion forces. Migrating GBM cells display increased expression of survival-related genes including fibroblast growth factor receptor 3 (FGFR3), which may promote chemotherapy resistance, and LMO3, which inhibits TP-53-mediated apoptosis. A histopathologic study by Altieri and colleagues found equal distribution of glioma stem cells in core and peripheral samples taken from patients with GBM, a finding that underscores the malignant potential of cells that have migrated from the tumor and the possible utility of supramarginal resection.

As migrating GBM cells appear to have a genomic signature distinct from those at the core, the environment encountered by these cells also varies. Immune cells isolated from the periphery of GBM have been shown to display increased expression of immunosuppressive genes. NF-kB activation by cells in the GBM periphery leads to anti-inflammatory polarization of regional immune cells and a more pro-oncogenic environment. Increased neural activity and the hyperexcitable milieu of the peritumoral microenvironment have not only been shown to promote the proliferation and migration of glioma cells but are also not mechanisms targeted by conventional GBM treatment.

Thus, regions potentially targeted by supramarginal resection are characterized by genetically distinct populations of cells and a pro-tumoral microenvironment that likely explain the propensity of GBM to invade and recur. This underscores the importance of evaluating the safety and benefit of removal of at least part of this region of tumor.

Methods and benefits of supramarginal resection

Supramarginal resection is defined as removal of presumably infiltrated brain tissue beyond the borders of T1 contrast-enhanced tumor. Dandy was the first to attempt supramarginal resection in GBM by conducting hemispherectomies, though his series of 5 patients was fraught with complications. Indeed, 2 of these 5 patients still experienced recurrence. Since then, the role of supramarginal resection and the optimal ways in which to carry it out have been debated.

One method of achieving supramarginal resection is the use of FLAIR attenuation as a guide for resection beyond the tumor bulk ( Fig. 1 ). Grabowski and colleagues found that the absolute residual FLAIR volume was a univariate predictor of survival in patients with GBM, though residual contrast enhancement was a stronger predictor. However, because these areas can be quite large, authors using this framework have attempted to quantify the amount of resection needed to confer significant survival benefits. A large retrospective review by Li and colleagues analyzed the survival of 1229 patients with GBM following surgery. While this study again confirmed the contribution of total resection of the enhancing portion to survival (15.2 months vs 9.8 months for incomplete resection), this study was notable in that it showed that resection of greater than 53.2% of the peritumoral FLAIR attenuation resulted in significantly better survival than less extensive resections (20.7 vs 15.5 months; Fig. 2 ). A survival benefit may be seen with even less aggressive resection. In a series of 101 patients with newly diagnosed, completely resected GBM, Vivas-Buitrago and colleague found that between 20% and 50% supramarginal resection resulted in a survival benefit, while resection beyond 60% did not significantly impact survival ( Fig. 3 ). In this study, supramarginal resection was not associated with a decrement in postoperative Karnofsky Performance Status (KPS). Pessina and colleagues found a survival benefit with resection of at least 45% of the FLAIR attenuating region (2 year OS of 54% vs 12%); in this study, no statistical comparison of the rate of new or worsened neurologic deficit was performed.