Journal of Neurosurgery 2026;145(5):1326–1333
Two glioblastomas with similar enhancing volumes can present different surgical problems when one extends into deep periventricular structures. Arcidiacono and colleagues examined whether an anatomical grid could describe that difference on preoperative MRI. Its clearest practical contribution was to show how many regions the enhancing tumor occupied and which of them could constrain resection.
Study and key findings
———————-
The investigators retrospectively analyzed 116 adults with newly diagnosed IDH-wildtype glioblastoma. They registered contrast-enhanced MRI to a common anatomical space and divided the supratentorial brain into 48 regions. Each region was scored as involved or uninvolved. They compared the distribution of enhancing tumor with preoperative volume, residual volume, extent of resection and survival.
The median enhancing volume was 38.74 cm³, and tumors occupied a median of five grid regions. Central regions encompassing the basal ganglia and subventricular zone were frequently involved. Greater spatial involvement correlated with reduced resectability. Larger residual volumes and a lower extent of resection were also associated with shorter survival.
Interpretation and limitations
——————————
The prognostic cutoff deserves caution. The authors report a survival difference with six or more involved regions in one analysis, but their Kaplan–Meier comparison at that cutoff did not reach significance. No individual region emerged as an independent prognostic marker. Only enhancing tumor was segmented, so the grid describes visible disease distribution and misses nonenhancing infiltration. Functional mapping results and postoperative neurological deficits were unavailable in a standardized form across centers.
Clinical takeaway
—————–
At the tumor board, use the grid to make the proposed resection anatomically explicit. For example, two lesions of similar volume may leave very different expected residuals depending on their relationship to the basal ganglia and deep white matter. The map can structure that discussion. A specific grid location or a six-region count alone provides insufficient evidence to extend a surgical margin.
