Abstracts

Rationale: Surgical resection failure in patients with refractory epilepsy approximates  20-60%. For these patients, nonsurgical and nonpharmacologic therapeutic options are highly desirable. We previously reported on an open-label pilot clinical trial under an FDA IDE, where 17 pediatric and adult patients with refractory epilepsy were treated using transcranial direct current stimulation (tDCS),  a non-invasive neurostimulation technique that delivers low levels of electric currents (2mA) to the brain via scalp electrodes. Our patients had a mean decrease in seizure frequency to 56% of baseline during the treatment and to 42% eights weeks post-treatment. However, patients with previous surgical interventions were specifically excluded, due to concerns that irregularities in the topology of the brain and skull could have unintended and unanticipated effects on the applied electrical currents. Methods: After clearing this with the FDA, we expanded the inclusion criteria to include patients with prior surgical interventions. For each of these patients, a personalized stimulation montage was determined using an MRI of the patient's head to build a finite-element computational model, including any openings in the skull, gaps due to removed tissue, implants and any other surgical or other abnormalities in the brain topology, to simulate current flows in the brain. This personalized biophysical model of the brain and skull was then used to determine the placement of electrodes on the scalp (by a computer optimization method in part using genetic algorithms) so that the applied currents would, in theory, reach the seizure focus with the correct orientation. (See Figure 1 for an example of this.) After a baseline seizure monitoring for 8 weeks, patients followed the tDCS protocol of 10 sessions, 20 or 30 minutes per session, 2mA stimulation sessions over two weeks, with eight weeks of follow-up. In all instances, the simulation montage was arranged to deliver cathodal tDCS to the seizure focus. The primary outcome measure reduction in seizure frequency from baseline as measured using a daily seizure diary. No changes in medications were made. Side effects were recorded.  Results: Five patients - one with prior resective surgery, one with burr holes from an implanted EEG, one with a VR shunt, one with a corpus callosotomy, one with a left temporal craniectomy - were treated under our protocol. These patients reported seizure reduction from baseline in the eight weeks following treatment of 75%, 44%, 90%, 77%, and 7% respectively. No adverse events were reported for these patients. Conclusions: Targeted non-invasive neurostimulation using transcranial current stimulation in craniectomized patients may be an effective therapy for controlling seizures in surgical non-responders. The authors intend to continue this work in a blinded, placebo-controlled study under an FDA IDE for eventual device approval.  Funding: Massachusetts Life Science Center; Neuroelectrics Corporation