Hidden Brain Damage After Glioma Surgery Is Far More Common Than Thought, Study Finds

Brain tumor surgery has become remarkably precise over the past two decades, with surgeons using fluorescent dyes, intraoperative imaging and electrical mapping to strip away as much glioma as possible while sparing healthy tissue. Yet a new prospective study suggests that even in skilled hands, the operation itself frequently leaves behind invisible scars. Researchers at Sun Yat-sen University Cancer Center in Guangzhou, China, tracked 64 patients with newly diagnosed supratentorial gliomas through a battery of serial magnetic resonance imaging scans in the first month after surgery, and what they found challenges the assumption that a clean resection means an uninjured brain. Nearly one in four patients had suffered a postoperative stroke, more than half showed microscopic bleeding, and every single patient had swelling around the surgical cavity. The findings, published in the Journal of Neuro-Oncology, offer one of the most detailed time-lapse portraits yet of what neuroscientists call surgical brain injury, the collateral damage inflicted on surrounding tissue when a tumor is removed.

The study’s design is what sets it apart from much of the existing literature. Most previous estimates of postoperative infarction and edema have come from retrospective chart reviews, in which radiologists examined scans ordered for clinical reasons rather than for research. That approach risks missing injuries that never produce obvious symptoms. In this prospective observational study, the Chinese team enrolled patients before their operations and mandated a strict imaging schedule: an early MRI within 72 hours of surgery, followed by repeat scans on postoperative days 7, 14 and 28. This serial protocol allowed the investigators to watch injuries evolve in real time, distinguishing fresh infarcts from older changes, tracking the waxing and waning of brain swelling, and capturing microhemorrhages that a single routine scan could easily overlook. Neurological function was then formally assessed at day 28 using the modified Rankin Scale, a standard clinical measure in which scores of 2 or higher indicate disability that interferes with independent daily life.

The headline result concerns stroke. Postoperative infarction, dead brain tissue caused by interrupted blood supply, appeared in 14 of the 64 patients, an incidence of 21.9 percent. Strikingly, every one of these infarcts was localized to the same trio of deep brain regions: the basal ganglia, the thalamus and the insula. These are not random locations. They are the territories supplied by small, deeply penetrating arteries, the lenticulostriate vessels and their counterparts, which branch directly off major arteries such as the middle cerebral artery and run upward through the surgical corridor to reach the tumor. When gliomas sit in or near these deep gray-matter structures, the dissection plane can pass within millimeters of these fragile perforating vessels. The study’s anatomical pattern therefore points to a mechanical mechanism: perforating arteries are being stretched, cauterized, or sacrificed during resection, and the deep nuclei they supply pay the price.

When the researchers searched for risk factors, one variable stood out with statistical force. Intraoperative blood loss of 300 milliliters or more was independently associated with postoperative infarction, with an odds ratio of 4.62 and a p-value of 0.035, meaning the association was unlikely to be due to chance. The connection makes physiological sense. Heavy bleeding during surgery can drop blood pressure, reduce perfusion through small vessels, force more aggressive coagulation of bleeding points, and obscure the surgical field so that fine perforators are harder to identify and preserve. Blood loss, in this framing, is not merely a marker of a bloody operation; it is a proxy for hemodynamic stress and hemostatic manipulation throughout the case. The authors’ conclusion is direct: meticulous intraoperative hemostatic management, careful control of bleeding from the first incision to the final closure, is essential to prevent postoperative infarction and the disability that can follow.

The second major finding concerned microhemorrhage, tiny punctate bleeds visible on susceptibility-weighted MRI sequences. These were remarkably common, detected in 53.8 percent of patients, more than half of the cohort. Microhemorrhages are often dismissed as clinically silent, and in many contexts they are. But their prevalence here suggests that blood-brain barrier disruption and capillary trauma extend far beyond the visible margins of the resection cavity. Whether these microbleeds accumulate into lasting cognitive consequences, as they appear to do after traumatic brain injury and in cerebral amyloid angiopathy, remains an open question that this study was not designed to answer. Still, their frequency reframes glioma surgery: the zone of injury is not a sharp line at the tumor border but a gradient of vascular damage that radiates outward into apparently normal brain.

Perhaps the most conceptually interesting result involves edema, the swelling that accompanies virtually every brain tumor and every brain operation. Peritumoral edema was present in all 64 patients, but the serial scans revealed that it came in two distinct flavors. In 42 patients, the swelling was residual tumor-associated edema, the vasogenic swelling driven by the tumor’s leaky vessels and inflammatory secretions, which shrank over time as the mass was removed. In 22 patients, however, the team identified de novo surgery-induced edema, swelling that appeared or expanded after the operation in areas where the tumor had driven the original fluid accumulation. The two patterns moved in opposite directions on serial imaging: the average volume of tumor-associated edema decreased over the four weeks of follow-up, while surgery-induced edema increased. This distinction matters clinically, because the two types of swelling can mimic one another on a single scan, and misreading postoperative edema as residual tumor, or vice versa, can lead to premature decisions about additional treatment.

The neurological stakes of all this radiological injury became clear in the outcome data. Of the 53 patients who completed the day-28 functional assessment, 13 had developed disability, scoring 2 or higher on the modified Rankin Scale. That is roughly one in four of those assessed, a sobering figure for operations performed with curative or life-extending intent. Two factors independently predicted poor outcome: higher intraoperative blood loss, with an odds ratio of 8.88 and a p-value of 0.013, and postoperative infarction, with an odds ratio of 6.33 and a p-value of 0.031. The blood loss association was the stronger of the two, reinforcing the study’s central message that what happens inside the operating room, particularly the management of bleeding, echoes directly into the patient’s functional recovery a month later.

These findings land in a field that has long been pulled in two directions. On one side stands the well-documented survival benefit of maximal resection: large studies have repeatedly shown that removing more contrast-enhancing tumor, and even supratotal resection extending beyond the MRI-visible margin, is associated with longer survival in glioblastoma and in lower-grade gliomas. On the other side is the growing recognition that the deep perforating arteries coursing through glioma territory are vulnerable, and that strokes after glioma surgery are neither rare nor benign. Earlier retrospective work, including studies from European groups, has reported postoperative ischemia rates ranging from a few percent to more than 20 percent depending on tumor location and detection methods, with insular and deep-seated tumors consistently flagged as high risk. The new prospective data confirm those signals with a rigor that retrospective series cannot match, and they add the temporal dimension, showing how infarcts, edema and microbleeds each follow their own trajectory in the first postoperative month.

The practical implications are already taking shape. For surgeons, the study argues for treating blood loss as a modifiable target rather than an unavoidable byproduct of tumor resection: aggressive early hemostasis, careful preservation of perforating vessels identified on preoperative tractography and angiography, and perhaps a lower threshold for staged or subtotal resection when a critical perforator is at risk. For oncologists and radiologists, the two-pattern model of edema provides a framework for interpreting postoperative scans, suggesting that serial imaging rather than a single early study may be needed to separate resolving tumor edema from progressive surgical swelling. For patients and families, the numbers offer a realistic picture of the risk-benefit calculus: in this cohort, roughly one in five patients sustained a radiological stroke, and one in four assessed patients had measurable disability at one month, but the study also identifies a concrete, actionable lever, intraoperative hemostatic management, that could shift those odds. Larger multicenter prospective studies will be needed to confirm the risk factors and to test whether protocol changes in the operating room actually reduce disability, but this work marks an important step toward making glioma surgery not only more complete, but also more gentle on the brain that surrounds the tumor.

Subject of Research: Radiological features of surgical brain injury and neurological outcomes after glioma resection

Article Title: Radiological features of peritumoral brain injury and neurological outcomes after glioma surgery: a prospective observational study

Article References: Radiological features of peritumoral brain injury and neurological outcomes after glioma surgery: a prospective observational study. (n.d.). https://doi.org/10.1007/s11060-026-05828-5

Image Credits: AI Generated

DOI: 10.1007/s11060-026-05828-5

Keywords: glioma surgery, surgical brain injury, brain infarction, brain edema, microhemorrhage, intraoperative blood loss, modified Rankin Scale, MRI, neurosurgery, perforating arteries, neurological outcomes, prospective observational study