STandardized Ultra-conservative Or Physician-directed ICD Programming for Continuous Flow LVAD Support

Overview

This study will evaluate the utilization of an ultra-conservative programming strategy to reduce shocks for ventricular arrhythmias (VA) among patients with heart failure, an implantable cardioverter-defibrillator (ICD) and continuous flow (CF) left ventricular assist device (LVAD). Patients on LVAD support demonstrate unique hemodynamic tolerability of VA, and the role for ICDs among patients with newer generation CF LVADs remains less clear than the older generation devices. Prior studies have explored extended detection ICD programming to reduce unecessary or potentially avoidable shocks to patients. However, no prospective randomized study to-date has evaluated such programming strategies in the CF LVAD patient population. The study hypothesis is that ultra-conservative (UC) ICD programming will result in a reduction of shocks and an improvement in quality of life without increasing hospitalizations, syncope or death among patients on CF LVAD support, and the primary design is a 1:1 randomization between UC programming over standard, physician discretion programming.

The current International Society for Heart and Lung Transplantation (ISHLT) guidelines provide a class I recommendation for tachy-therapy re-activation in patients with ICDs undergoing LVAD implantation. However, these guidelines reflect a survival benefit observed in cohorts comprised primarily of older-generation pulsatile flow LVADs. While the overall VA prevalence in the LVAD population exceeds 50% with VA usually occurring \>200 beats per minute, VA are generally well-tolerated both hemodynamically and symptomatically. Observed LVAD flows do decrease during sustained VA but there are multiple case reports of LVAD patients remaining awake and asymptomatic despite hours of ventricular fibrillation and a large case series reported no deaths and rare syncope due to VA in CF LVAD patients. Contrary to guidelines, contemporary evaluations restricted to CF LVADs show no clear ICD survival benefit in single center studies nor a large, pooled meta-analysis. In fact, propensity-matched analyses of CF LVAD patients in the INTERMACS and UNOS registries have demonstrated increased mortality and more hospitalizations among patients with an active ICD and treated VA. It is unclear in the CF LVAD population whether ICD shocks themselves are harmful or rather identify a higher-acuity patient subgroup. In the general HF population, however, shocks have consistently been associated with worse outcomes. Shock reduction programming resulted in reduced mortality in the MADIT-RIT trial, suggesting shocks may be directly harmful and not only a marker of substrate. Therefore, the present study will evaluate whether ultra-conservative (UC) programming for ICDs can reduce shocks over standard (physician-discretion) programming among patients with heart failure on CF LVAD support. The model is 1:1 randomization between the two programming strategies for patients with newly implanted CF LVAD devices and a pre-existing ICD. They will be eligible for enrollment during their index hospitalization, and after discharge up to their first outpatient follow-up visit no later than 3 months post discharge. The primary endpoints will be any ICD shock, all-cause mortality, and syncope. The secondary endpoints will be time to first ICD shock, time to first ICD shock for VT/VF, quality of life questionnaire and hospitalizations, including for heart failure and ventricular arrhythmias. The principal exclusion criteria will be pediatric patients, or those without a functional ICD system. Standard of care follow-up will continue thru 24 months.