Radiographic characteristics fail to predict clinical course after subdural electrode placement

Subdural arrays of grid and strip electrodes are frequently used in epilepsy patients to localize the seizure focus and determine the location of critical brain functions via stimulation mapping. Postoperatively, the majority of patients with implanted subdural electrodes develop subacute extra-axial collections (EACs). Although conservative management is appropriate in most of these cases, occasionally patients manifest neurological symptoms that may necessitate reoperation for collection evacuation. Currently, there is little information available regarding the range of EAC size and the potential correlation between EAC size and symptom development. To facilitate treatment decision-making in postoperative subdural electrode patients, we reviewed and compared the computed tomographic (CT) features of postelectrode placement EACs in asymptomatic and symptomatic patients.

We retrospectively reviewed the medical records and CT scans of 22 consecutive patients who underwent craniotomy for placement of subdural grid and strip electrodes at Columbia University Medical Center. Medical records were reviewed for neurological complications from the time of grid placement until its removal. Each EAC was measured on CT for volume (% of total cranial volume), maximal thickness, and midline shift. One patient was excluded secondary to the development of an intracerebral hemorrhage.

Thirteen of 21 patients remained asymptomatic or minimally symptomatic during their hospitalization, with only mild to moderate, intermittent, postoperative headaches. The remaining eight developed symptoms such as persistent and severe headache, transient motor deficit, or speech impairment. Two of these patients underwent reoperation for hematoma evacuation. EACs in asymptomatic patients had a mean volume, maximal thickness, and midline shift of 5.7%, 1.25 cm, and 0.33 cm, respectively. EACs in symptomatic patients had a mean volume, maximal thickness, and midline shift of 7.7%, 1.46 cm, and 0.5 cm, respectively. Differences between maximal thickness and midline shift did not approach statistical significance. Despite this, the difference between the mean volume of symptomatic and asymptomatic EACs was statistically significant (P = 0.04).

The conventional methods of midline shift and maximal thickness for assessing EAC size did not adequately differentiate symptomatic and asymptomatic subdural electrode patients with EACs. Although total volume calculation using digital planimetric analysis demonstrated a statistically significant difference, we found no clear threshold volume that correlated with clinical course. Therefore, the appearance of EACs on CT scans is of limited use in predicting the development of symptoms and possible postoperative complications after subdural grid placement. Clinical judgment must guide management and determine the potential need for reoperation.