Implantable cardioverter-defibrillator lead revision following left ventricular assist device implantation.

INTRODUCTION

Lead dysfunction can lead to serious consequences including failure to treat ventricular tachycardia or fibrillation (VT/VF). The incidence and mechanisms of lead dysfunction following left ventricular assist device (LVAD) implantation are not well-described. We sought to determine the incidence, mechanisms, timing, and complications of right ventricular lead dysfunction requiring revision following LVAD implantation.

METHODS

Retrospective observational chart review of all LVAD recipients with pre-existing implantable cardioverter-defibrillator (ICD) from 2009 to 2018 was performed including device interrogation reports, laboratory and imaging data, procedural reports, and clinical outcomes.

RESULTS

Among 583 patients with an ICD in situ undergoing LVAD implant, the median (interquartile range) age was 62.5 (15.7) years, 21% were female, and the types of LVADs included HeartWare HVAD (26%), HeartMate II (52%), and HeartMate III (22%). Right ventricular lead revision was performed in 38 patients (6.5%) at a median (25th, 75th) of 16.4 (3.6, 29.2) months following LVAD. Mechanisms of lead dysfunction included macrodislodgement (n = 4), surgical lead injury (n = 4), recall (n = 3), insulation failure (n = 8) or conductor fracture (n = 7), and alterations in the lead-myocardial interface (n = 12). Undersensing requiring revision occurred in 22 (58%) cases. Clinical sequelae of undersensing included failure to detect VT/VF (n = 4) and pacing-induced torsade de pointes (n = 1). Oversensing occurred in 12 (32%) and sequelae included inappropriate antitachycardia pacing ([ATP], n = 8), inappropriate ICD shock (n = 6), and ATP-induced VT (n = 1).

CONCLUSION

The incidence of right ventricular lead dysfunction following LVAD implantation is significant and has important clinical sequelae. Physicians should remain vigilant for lead dysfunction after LVAD surgery and test lead function before discharge.