Neuropsychological outcomes after epilepsy surgery: A comparison of stereo electroencephalography and subdural electrodes

Abstract

Background and purpose: We analyzed the association of neuropsychological outcomes after epilepsy surgery with the intracranial electrode type (stereo electroencephalography [SEEG] and subdural electrodes [SDE]), and electrical stimulation mapping (ESM) of speech/language.

Methods: Drug-resistant epilepsy patients who underwent comprehensive neuropsychological evaluation before and 1 year after epilepsy surgery were included. SEEG and SDE subgroups were matched by age, handedness, operated hemisphere, and seizure freedom. Postsurgical neuropsychological outcomes (adjusted for presurgical scores) and reliable change indices were analyzed as functions of electrode type and ESM.

Results: Ninety-nine patients aged 6-29 years were included with similar surgical resection/ablation volumes in the SEEG and SDE subgroups. Most of the neuropsychological outcomes were comparable between SEEG and SDE subgroups; however, Working Memory and Processing Speed were significantly improved in the SEEG subgroup. Undergoing language ESM was associated with significant improvements in Spelling, Letter-Word Identification, Vocabulary, Verbal Comprehension, Verbal Learning, and Story Memory scores, but a decline in Calculation scores.

Conclusions: Intracranial evaluations with SEEG and SDE are comparable in terms of long-term postsurgical neuropsychological outcomes. Our data suggest that SEEG may be associated with improvements in working memory and processing speed, representing cognitive domains served by spatially distributed networks. Our study also supports wider use of language ESM before epilepsy surgery, preferably using other language tasks in addition to visual naming. Rather than the type of electrode, postsurgical neuropsychological outcomes are driven by whether language ESM was performed or not, with beneficial effects of language mapping.

Keywords: drug-resistant epilepsy; functional brain mapping; intracranial electroencephalography; language mapping.

Figure 1.

Differences in the proportions of parcels lesioned during epilepsy surgery after monitoring with SEEG and SDE. Standardized mean difference between SEEG and SDE are shown along with 95% confidence intervals. All standardized mean differences lie between ±2 with no significant p-values for independent sample t-tests (not shown). Abbreviations: Inferior Frontal Gyrus (IFG), Subdural Electrodes (SDE), Stereotactic Electroencephalography (SEEG), Standardized Mean Difference (SMD)

Figure 2.

Effect of electrode type (SEEG/SDE) and performing language ESM (yes/no) on reliable change indices of neuropsychological scores. Slopes of linear regression models along with 95% confidence intervals are shown. *indicates Bonferroni corrected p-value ≤0.05. Abbreviations: Confidence Interval (CI), Electrical Stimulation Mapping (ESM), Subdural Electrodes (SDE), Stereotactic Electroencephalography (SEEG). Verbal Comprehension (VC), Working Memory (WM), Processing Speed (PS), and Visual Spatial (VS) subtests of Wechsler scales; Letter-Word Identification (LWI), Calculation (CALC), and Spelling (SPEL) subtests of the Woodcock-Johnson Tests of Achievement; Story Memory Immediate (SMI), Story Memory Delayed (SMD), Verbal Learning Immediate (VLI), and Verbal Learning Delayed (VLD) subtests of Wide Range Assessment of Memory and Learning (WRAML); Peabody Picture Vocabulary Test (PPVT); Boston Naming Test (BNT).

Figure 3.

Effect of electrode type (SEEG/SDE) (A) and performing language ESM (yes/no) (B) on mean neuropsychological scores pre and post epilepsy surgery. The direction of change is different for the two electrode types as well as for performing ESM in case of SMI, SMD, VLI, VLD (all subtests of WRAML), PPVT, and BNT. In all these instances, patients monitored with SEEG and patients who underwent ESM showed an increase whereas those monitored with SDE and those who did not undergo ESM (respectively) showed a decrease in mean scores after epilepsy surgery. Also note that presurgical scores are lower for the ESM group than for no-ESM group across the neuropsychological domains Abbreviations: Standard Error (SE), Verbal Comprehension (VC), Working Memory (WM), Processing Speed (PS), and Visual Spatial (VS) subtests of Wechsler scales; Letter-Word Identification (LWI), Calculation (CALC), and Spelling (SPEL) subtests of the Woodcock-Johnson Tests of Achievement; Story Memory Immediate (SMI), Story Memory Delayed (SMD), Verbal Learning Immediate (VLI), and Verbal Learning Delayed (VLD) subtests of Wide Range Assessment of Memory and Learning (WRAML); Peabody Picture Vocabulary Test (PPVT); Boston Naming Test (BNT)