Compare long-term seizure (sz) control in patients (pts) with continued vagal nerve stimulation (VNS) with those who discontinued VNS. 

Long term VNS follow-up (f/u) studies show progressive improvement but are limited by lack of comparison groups. This study compared long-term sz control in pts who continued VNS stimulation (“VNS-on”) to those who underwent VNS then discontinued treatment (“VNS-off”). Question: Do pts with long-term VNS have greater sz reduction than those who discontinue VNS?

Refractory epilepsy pts with VNS therapy for >3 yrs (and f/u >2 yrs after VNS discontinuation for VNS-off) with clear baseline (pre-VNS) and last f/u sz frequencies were included. Pts with brain surgery <3 yrs after VNS excluded; only last f/u prior to brain surgery, if performed, included. We assessed and compared 1) percent of pts with >50% sz reduction (50% responder rate), and 2) change in sz frequency at last f/u compared with baseline in VNS-on and VNS-off pts.

49 pts included: 33 VNS-on, 16 VNS-off. VNS-on pts had treatment for 116.6 ± 43 months. VNS-off pts had VNS treatment for 78 ± 37 months, discontinued treatment, then had 95.63 ± 38 months additional follow-up. 50% responder rates between groups (VNS-on: 54.5%; VNS-off: 50%, p=NS) were similar. However, VNS-on pts had significant reduction in sz frequency at last f/u compared with baseline (-12.7 sz/month± 34.1, p = 0.04) compared with VNS-off pts both right before VNS was discontinued (-2.8 sz/month ± 18.5, p = 0.55) and last f/u (-4.4 sz/month ± 14.7, p = 0.25), though effect size was small (0.28).

Pts with long-term VNS therapy and those who discontinued therapy had similar long-term 50% responder rates. However, pts with long-term VNS treatment may experience greater magnitude of sz reduction. Pts with poor VNS response may be more refractory. Larger confirmatory studies are needed.