Identification of novel electrophysiologic biomarkers of cognition in glioma-infiltrated cortex.

Abstract

2077 Background: Diffuse gliomas, the most common primary brain cancers, often invade speech-critical areas. Maximal resection improves survival, but damage to functional cortex may cause permanent impairments. Direct cortical stimulation (DCS) differentiates functional (DCS+) from nonfunctional (DCS-) cortex by temporarily disrupting neuronal activity, yet it remains unknown how DCS+ sites elicit transient impairments. DCS is technically challenging and resource-intensive. As a result, fewer than 50% of glioma patients receive optimal surgical care. This translational study aims to identify electrophysiologic biomarkers of DCS+ cortex to 1) aid in safe resection by avoiding functional cortex and 2) to elucidate causal relations behind these transient impairments. Methods: Local field potentials of subdural array data from glioma infiltrated cortex was annotated as DCS+ or DCS- prospectively. We compared spectral electrophysiologic variations (mean Theta [4-8 Hz], Alpha [8-13 Hz], Beta [13-30 Hz], and Full Gamma [30-150 Hz] ranges) at resting state between DCS+ and DCS- sites using linear mixed-effects models (to account for patient-level differences). Results: 1421 cortical sites of language were studied in 91 patients including 21 Oligodendroglioma WHO grade 2-3, 19 Astrocytoma WHO 2-3, 3 Astrocytoma WHO 4, 48 IDHwt glioblastoma [GBM] WHO 4). 115 (8.0%) were DCS+. After alignment to ECoG electrode arrays, 512 cortical sites (49 DCS+) were assigned to electrodes. In oligodendrogliomas, DCS+ (N=16) vs DCS- (N=132) sites had higher alpha (77.7 ± 111.9 vs 38.6 ± 39.8, p=0.018), beta (23.1 ± 18.1 vs 11.9 ± 17.9, p=0.033), and full gamma (0.6 ± 0.6 vs 0.3 ± 0.3, p<0.001) power. Similarly, in astrocytoma, DCS+ (N=13) vs DCS- (N=147) sites had significantly higher theta (98.5 ± 94.9 vs 57.1 ± 67.2, p=0.020), alpha (83.5 ± 76.3 vs 39.9 ± 42.4, p=0.003), beta (55.9 ± 58.4 vs 16.3 ± 17.8, p<0.001), and gamma (0.9 ± 0.9 vs 0.4 ± 0.4, p=0.006) power. Interestingly, when comparing DCS+ (N=20) and DCS- (N=237) sites in GBM patients, no significant differences were found in any studied ranges (all p>0.05). Conclusions: This study is the first of its kind to identify unique electrophysiological biomarker differences (at resting state) for oligodendroglioma and astrocytoma speech cortex. It has two key implications. First, clinically, the identification of electrophysiologic biomarkers may improve direct cortical stimulation (DCS) mapping: It can make surgeries faster by identifying cortex critical for cognition (speech) based on these biomarkers, safer by helping neurosurgeons avoid resecting critical regions, and more accessible. Second, this research suggests that different tumor types (low-grade gliomas vs. GBM) remodel speech areas differently, prompting further investigation into tumor-specific effects on neural circuits.