Long-Term Seizure Outcome With or Without Vagal Nerve Stimulation Therapy.

BACKGROUND AND OBJECTIVES

To compare long-term seizure control in patients with long-term VNS (vagal nerve stimulator) stimulation (VNS-on) with those who discontinued VNS after >3 years (VNS-off).

METHODS

Patients with refractory epilepsy with VNS therapy for >3 years (and follow-up for >2 years after VNS discontinuation for VNS-off patients) were included. Patients with brain surgery <3 years after VNS were excluded. We compared the percentage of patients with ≥50% seizure reduction (50% responder rate) and change in seizure frequency within and between groups in follow-up.

RESULTS

Thirty-three VNS-on and 16 VNS-off patients were evaluated. VNS-on patients underwent stimulation for 9.7 years (mean). VNS-off patients had VNS treatment for 6.5 years (mean), discontinued treatment, then had additional 8.0 years (mean) follow-up. 50% responder rates were similar between groups (VNS-on: 54.5% vs VNS-off at last-on: 37.5%, = 0.26; vs VNS-off at the last follow-up: 62.5%, = 0.60). VNS-on patients had a significant reduction in seizure frequency at the last follow-up compared with baseline (median [Mdn] = -4.5 seizures/month, interquartile range [IQR] = 14.0, 56% reduction, = 0.013). VNS-off patients also showed significant seizure reduction while still continuing VNS therapy (Mdn = -1.0 seizures/month, IQR = 13.0, 35% reduction, = 0.020) and, after discontinuing therapy, at the last follow-up compared with baseline (Mdn = -3.2, IQR = 11.0, 52% reduction, = 0.020). The 2 groups were comparable in seizure frequency change both at the last-on visit (absolute change, = 0.62; relative change, = 0.50) at the last follow-up (absolute change, = 0.67; relative change, = 0.76).

DISCUSSION

Patients who discontinued VNS therapy and those who continued therapy had similar response during active treatment and similar long-term outcomes, suggesting that factors such as the natural disease course and/or medication treatment strongly affect long-term outcomes.