Abstracts

Rationale: Epilepsy surgery is increasingly offered to patients with extra-temporal non-lesional epilepsy. Inracranial EEG monitoring plays an important role in determining the epileptogenic zone and adjacent eloquent cortex. Intracranial EEG monitoring may be associated with significant morbidity. Following a study of the risk factors associated with complications during intracranial EEG monitoring at the Westmead campus (Acta Neurochirurgica 2009; 151:37-50) we created a revised protocol for implantation. Including: one to one nursing ratio, hourly observations for the initial 48 hour postoperatively, continuous heart rate and blood pressure, oxygen saturation and intracranial pressure monitoring, An external ventricular drain provides continuous ICP monitoring. PICC lines inserted in all patients. Post-operative MRI preformed 2-4h following implantation of the electrodes before transfer to the monitoring ward. An epilepsy trained registered nurse replaced the patient's relative as the sitter. Subdural electrode arrays not larger than 4x8 grid arrays were used. Methods: All patients who underwent invasive monitoring at Westmead Hospital and the Children's Hospital at Westmead between 1988 and 2004 (Gp A) were compared with those done after the protocol changes; between 2005 and 2012 (Gp B). Complications were categorised according to the relationship to the electrodes, outcome and the need for intervention. In those time periods a total of 296 patients (Gp A period) and 254 patients (Gp B period) underwent epilepsy surgery. All patients had intracranial monitoring because non-invasive investigations did not produce adequate localising information or the focus was adjacent to eloquent cortex. Results: 71 (median age: 24 years) patients in Gp A and 52 patients in Gp B (median age: 24 years) underwent intracranial electrode implantation. Transient complications requiring no treatment; Gp A 13% vs Gp B 2%. Transient complications requiring treatment; Gp A 10% vs Gp B 8%. Death occurred in Gp A (n=2) but not in Gp B. Indirect systemic complications in Gp A were 11% with none in Gp B. Complications directly related to intracranial EEG monitoring were significantly reduced using the revised protocol (p= 0.001). Infection rate was increased in Gp B (5.7%) compared with Gp A (2.8%), not statistically significant (p=0.90). Monitoring of continuous ICP prevented catastrophe in one case. Raised ICP prompted an urgent CT scan showing a 11.7 mm subdural collection and a new area of left frontal lobe hypodensity. Urgent removal of the subdural grids and resection of epileptogenic focus was followed by recovery. Histology confirmed venous infarction. Conclusions: Complication rates following intracranial implantation have been reduced after the introduction of our new protocol. Intracranial infection was increased in Gp B in comparison to Gp A. This maybe related to routine use of external ventricular drainage and further review is warranted.