Rationale:
Electrical stimulation brain mapping (ESM) is the gold standard method for identification and preservation of eloquent language areas to reduce the risk of postoperative language decline. Subdural grid electrodes (SDE) have historically been the standard modality for mapping language. However, stereo-EEG (SEEG) has increasingly replaced SDE due to its less invasive nature, lower complication rates, and ability to sample deep brain structures. Despite these advantages, SEEG may provide limited information for language localization due to lack of broad cortical surface coverage provided by SDE. Although seizure recording with both SDE and SEEG have shown comparable seizure outcomes, it remains unclear whether ESM using SEEG electrodes can identify language cortex with the same success as SDE, as current research is scarce and confounded by interindividual variability in patient samples. We present a unique case series of 3 patients who had both SDE and SEEG language mapping, allowing direct comparison of results obtained with the two techniques.
Methods:
Patients were three male, right-handed, native English speakers with drug-resistant, left (language dominant) hemisphere epilepsy. Patients underwent pre-surgical evaluation including extraoperative ESM with both SDE and SEEG targeting the same regions (left frontal and/or temporal) at 2 different timepoints (timepoint 1: ages 14, 23, 28; timepoint 2: ages 15, 29, 40, respectively). Two patients had SEEG followed by SDE implant within 4 months, with no therapeutic intervention between procedures. One patient underwent SDE implant prior to limited left frontal resection of cortical dysplasia sparing positive language sites, followed by SEEG. ESM tasks included visual picture naming (VN), auditory description naming (AN), and reading. Sites were considered positive for language if errors occurred on at least 75% of stimulation trials and tasks were performed without error on control (i.e., non-stimulation) trials.
Results:
In all 3 patients, different sites were identified as positive with SEEG versus SDE language mapping with regard to location, number of sites found, and type of language disruption. Positive sites found/number of sites tested: Patient 1: SEEG=0/17 vs SDE=5/23 (VN: 1, VN+AN: 2, AN+VN+Reading: 2); Patient 2: SEEG=1/18 (VN: 1) vs SDE=2/26 (AN: 1, VN/AN/comprehension: 1); Patient 3: SEEG=2/11 (AN: 1, AN+VN+Reading: 1) vs SDE=7/46 (AN: 5, VN: 1, AN+VN: 1). Across patients, of the 46 SEEG sites tested, 3 (6.5%) were positive, whereas, of the 95 SDE sites tested, 14 (14.7%) were positive.
Conclusions:
In this unique case series, we observed discrepancies in both the number and type of critical language sites identified with SEEG vs SDE within the same individual. Not surprisingly, a higher number of positive sites were identified with SDE mapping, likely due to the broader area of coverage. Although SEEG has become the most commonly used extraoperative mapping technique, depending on the location and type of surgical intervention planned, these findings suggest following SEEG language mapping with SDE or intraoperative cortical language mapping to ensure identification and preservation of critical language areas.
Funding: None.