LV Unloading with Continuous-Flow Left Ventricular Assist Devices: The Association Between Left Ventricular Size and Late Survival

Abstract

Purpose Left ventricular assist device (LVAD) therapy for the failing ventricle dramatically improves survival, but the nuances of optimal ventricular unloading remain unknown. Ideally, clinicians would titrate LVAD speed using an easily attainable metric that reflects LV unloading and correlates well with an important outcome such as death. We sought to determine if left ventricular end-diastolic dimension (LVEDD) meets these ideal criteria. Methods and Materials A retrospective review of all patients receiving a continuous-flow LVAD at a single center was undertaken. Patients who underwent an echocardiogram at our institution and had LVEDD recorded in the database were included. Simple descriptive statistics were employed, and survival was modeled using Kaplan-Meier analysis. A Cox proportional hazards model with LVEDD as a time-dependent covariate was used to produce hazard ratio estimates. Results Of 247 continuous-flow LVADs implanted from June 2006 to August 2012, echocardiographic data was available for 215 patients. This cohort included 37 females and 178 males, with an average age of 55+/-14 years. Indications for LVAD implantation was 161 patients bridge to transplant, 85 were destination therapy, and one patient was a bridge to recovery. Mean LVEDD at LVAD implant was 6.53+/-1.23 cm. LVEDD was not associated with survival (HR 2.01, 95% CI 0.45 – 9.0, P=0.36), however as depicted in Figure 1 , a trend towards improved survival is seen in those with LVEDD of 5 cm when compared with LVEDD of > 6 cm. Conclusions The current study does not find a correlation between LVEDD and survival, however further investigation with a larger study is required to rule out type II error.