Considering the extent of resection in diffuse glioma

The review by Young et al.¹ represents the latest contribution from the UCSF group to summarize a comprehensive approach to the surgical management of gliomas. In the detailed discussion of the preoperative, intraoperative, and postoperative phases that comprise a patient’s journey through a craniotomy for glioma, there is technical information that is relevant for all neurosurgeons who take care of patients with brain tumors—those in academic settings as well as those in various community settings. Importantly, the authors review the extensive literature on aggressive surgery and wisely emphasize the importance of surgical judgment in avoiding neurologic deficits. The message is clear: being aggressive and being safe are not mutually exclusive. To this end, the authors review the tools available to the surgeon to accomplish these goals. Of course, there is great variability between institutions and also individual surgeons as to how the goals of glioma surgery are achieved; ultimately, the quality of the clinical outcome justifies the approaches taken. The concepts reviewed by Young et al. are sufficiently important that this review is also highly suitable for non-surgeons as well as trainees from students to residents and fellows.

It is understood that gliomas, particularly high grade, require multi-disciplinary treatment because surgery never removes all of the invasive tumor cells and is therefore not a sufficient singular treatment modality. Nevertheless, substantial data show the survival benefit associated with surgery in these settings, thus ushering us beyond the days of nihilism in glioma surgery. In low-grade glioma, the data overwhelmingly demonstrate that the extent of resection improves overall survival.² Although the survival benefit of resection for glioblastoma is significantly less than in the low-grade setting, for patients those months—especially the quality months—matter. It is worth mentioning here the key studies that support approaches that maximize the extent of resection.^3–6

What exactly the goal of resection should be remains a salient point of discussion. Indeed, the authors emphasize that the goal of surgery is “supramaximal safe resection.” What does this mean? If “maximal safe resection” means removal of all contrast-enhancing (CE) disease, then supramaximal resection is the additional surgical removal of non-contrast-enhancing (nCE) areas, typically considered to be the T2/fluid-attenutated inversion recovery (FLAIR) signal abnormality on MRI. In fact, the term “FLAIRectomy” has been used to describe this approach.⁷ Accumulating evidence demonstrates the survival benefit from resecting nCE FLAIR disease in high-grade glioma.^6,8 Importantly, it is paramount to exercise judicious decision-making when deciding which patients can undergo these types of supramaximal resections; in specific anatomic locations, the high eloquence of adjacent cortical areas or white matter tracts rigidly constrains the extent of resection possible. It is clear that the induction of neurologic deficits that reduce a patient’s performance status almost completely negates the benefits of significant resection.

The surgical goal of removing significant nCE disease raises a number of key issues. We will limit this discussion here to glioblastoma, as the FLAIR signal in IDH mutant low-grade disease is more likely to represent expansile tumor. First, implicit in the supramaximal philosophy in glioblastoma is that FLAIR signal represents areas adjacent to CE masses that harbor infiltrative, rather than densely packed, disease. However, we do not have a good understanding of the FLAIR microenvironment, although some studies have explored this question.⁷ Specifically, it is currently impossible to determine the tumor infiltration gradient within FLAIR signal and to know whether the entirety of FLAIR signal is comprised of microscopic disease or whether this infiltration is heterogeneous. Thus, quantitative technologies for assessing this infiltrative tumor are needed. Further complicating radiographic assessment, FLAIR signal is nonspecific in that it also represents edema and other reactive components to the growing tumor mass. Second, surgical resection of nCE disease is technically distinct from the surgery that addresses CE areas. Unlike the resection of CE disease—in which 5-ALA-driven fluorescence, ultrasound, and/or the visually distinct tumor features observed under white light indicate where disease remains—there are no clear feedback cues to guide nCE resection. Thus, in many areas, this portion of nCE glioblastoma resection is akin to “flying with instruments only.” If nCE disease is limited to some areas such as the right frontal lobe, then anatomic boundaries delimiting a lobar resection can guide the surgery. In most other settings, however, there are no adjuncts, which can continuously guide the limits of resection in FLAIR-hyperintense tissue. Although compelling technologies with the potential of real-time cellular feedback include Raman spectroscopy and rapid intraoperative molecular testing, it is unclear whether even these approaches operate at the sensitivity level and kinetics needed to detect sporadic cells within tissue fast enough to be pragmatic at the time of surgery. For several centers, the use of intraoperative MRI is an additional approach to inform nCE resection.

Third, it is important to understand more granularly whether there are specific anatomic or molecular subgroups of glioblastoma patients in whom supramaximal surgery is particularly beneficial. Further work to define the genomic landscape of IDH wild-type nCE disease and how this correlates with outcome will be helpful. It may also be the case that we will move beyond the characterization of nCE disease purely by MRI signal intensity and, instead, complement imaging descriptions with additional data such as molecular features akin to the concept of “minimal residual disease.”

Ultimately, although we currently consider “supramaximal” to be the additional resection of nCE FLAIR disease, one can envision the extrapolation of this term to include even “non-FLAIR” infiltrative disease. Indeed, this approach is exemplified in studies describing lobectomies for appropriately located glioblastomas.⁹ While these approaches may be germane in select settings, we are likely approaching an asymptote in which the benefit of ever-increasing extent of resection will have diminishing returns when trying to avoid neurologic deficits. Nevertheless, knowing that the search for better medical, biological, and/or immune-based therapies remains an urgent priority, the excellent review by Young et al. underscores that aggressive but safe surgery remains a crucially important component of treatment for our patients and serves as an important reference for our field.

References