Permissive ICD Programming for LVAD Patients

Abstract

Left ventricular assist devices (LVAD) improve quality of life and survival of advanced heart failure patients. Ventricular arrhythmias (VA) are common in these patients and are often hemodynamically tolerated for prolonged periods. It is uncertain whether implantable cardioverter-defibrillators (ICD) should be programmed permissively to avoid ICD shocks or non-permissively to prevent complications of VA. We retrospectively analyzed data from patients discharged after LVAD implant at Barnes-Jewish Hospital from 2005-2013. The follow-up duration was up to one year from LVAD implant. Based on ICD programming at discharge interrogations, patients were assigned to a programming group: permissive or non-permissive. The permissive strategy would allow for faster heart rates and greater use of anti-tachycardia pacing (ATP) prior to delivering a shock. Primary and secondary outcomes were proportion of patients with ICD shocks, time to death and time to first rehospitalization. There were 222 patients in the final analysis. The rate of shocks was 18% in the permissive group and 33% in the non-permissive group (p=0.08). The rate of inappropriate ICD shocks was 3% in the permissive group and 7% in the non-permissive group (p=0.47). Compared to the permissive group, the non-permissive group had a significantly higher rate of antiarrhythmic drug therapy (76% vs 54%, p=0.01), particularly with mexiletine (20% vs 0%, p=0.03). There was no significant difference in time to death (logrank p=0.52) or time to first rehospitalization (logrank p=0.64) between the two groups. In LVAD patients permissive ICD programming had a nonsignificant trend toward fewer ICD shocks and did not have a significant difference in mortality or time to first rehospitalization. The non-permissive group had a significantly higher rate of antiarrhythmic drug therapy, specifically with mexiletine. This study suggests permissive programming is safe in this population and might reduce unnecessary device therapy. Further, a safe reduction in therapy could prolong ICD battery life and lead to fewer generator changes, thereby minimizing complications in a population on chronic anticoagulation with high morbidity. Larger studies are needed to further define optimal ICD programming settings for LVAD patients, in an effort to reduce ICD shocks.Table 1Primary and Secondary OutcomesOverall (222)Permissive (39)Non- Permissive (183)P-valueTotal ICD Shocks66 (30%)7 (18%)59 (33%)0.08Inappropriate ICD shocks14 (6%)1 (3%)13 (7%)0.47Mortality62 (28%)11 (28%)51 (28%)0.52Time to First Rehospitalization180 (81%)31 (79%)149 (82%)0.64Antiarrhythmic Drugs160 (72%)21 (54%)139 (76%)0.01 Open table in a new tab