Late-onset cardiac arrhythmia associated with vagus nerve stimulation

We report on a 13-year-old boy who first presented at the age of 5 years with refractory complex partial seizures (CPS) with and without secondary generalization. Despite a thorough workup no cause could be identified. After unsuccessful medical treatment attempts with 13 different antiepileptic drugs (AED) finally a vagus nerve stimulator (VNS; vagus nerve stimulation) was implanted in 2002 (NCP Type 101; Cyberonics, Houston, TX, USA). The intraoperative findings revealed no anatomic abnormalities and the lead test passed uneventfully. Eighteen months prior to the reported incident, seizure frequency was stable with three to five CPS per month under medication with phenobarbital (PB), felbamate (FBM), and zonisamide (ZNS). Due to former probably drug-related rickets the patient received also calcium and vitamin D supplementation with parameters of calcium metabolism stable within normal limits. VNS stimulation parameters were unchanged for the last years prior to the reported adverse effect (30 s on, 5 min off, 2.25 mA, frequency 20 Hz, pulse width 250 ms). Six and a half years after implantation a significant increase of seizure frequency was observed without apparent cause. Interventions with lorazepam, chloralhydrate, clobazam, and acetazolamide and a change of basic medication (i.e., stopping FBM) were unsuccessful. Suddenly an intermittent and self-terminating complete heart block with bradycardia occurring every 15–25 min lasting for 20–40 s over a time span of 2 h was recorded (Fig. 1). Several minutes later an asystole for 6 s was witnessed on the electrocardiogram (ECG) monitor. Assuming an influence of the VNS device, the stimulator was turned off and the bradyarrhythmia dissolved. Breakage of leads or abnormal placement of electrodes was ruled out by X-ray. A cardiological examination as well as laboratory workup did not reveal any abnormalities. Seizures were finally terminated with thiopental. Outpatient followup including close ECG monitoring did not show any further episodes of cardiac arrhythmias. More than 50,000 VNS implantations have been performed worldwide. Large studies have reported both effectiveness and safety of the device. Despite the close proximity of VNS and centers controlling cardiovascular function initial studies in both animals and humans failed to demonstrate a relevant effect on cardiac function. Recent studies produced inconsistent results [7, 8, 11, 13, 15], mostly focusing on severe incidents associated with the intraoperative lead test during VNS implantation [1, 3, 5, 6, 14]. All patients made an uneventful recovery (Table 1). Only recently were the first two patients with late onset of cardiac arrhythmia, years after implantation, reported [2, 9]. In our patient an influence of the VNS is the most likely explanation, as cardiac arrhythmias subsided immediately following inactivation of VNS. However, since the incident took place during a phase of increased seizure frequency, and seizures (in particular status epilepticus) are well known to affect cardiac rhythm, it might be possible that these circumstances as well as changes in medication predisposed our patient to bradyarrhythmia. Most data for people with VNS compared with those with severe complex partial-onset epilepsy without the device failed to demonstrate an increased mortality rate, albeit encompassing less than 5 years after VNS implantation [4]. Considering the long time interval between the VNS implantation and the adverse event in our patient it might P. Borusiak (&) M. Zilbauer S. Cagnoli M. Heldmann A. Jenke Zentrum fur Kinderund Jugendmedizin, HELIOS Klinikum Wuppertal, Heusnerstr. 40, 42283 Wuppertal, Germany e-mail: peter.borusiak@helios-kliniken.de