Left Ventricular Assist Device Explantation Evaluation Protocol Using Comprehensive Cardiopulmonary Exercise Testing
Left ventricular assist devices (LVADs) provide decongestion of the left ventricle (LV) with associated reversal of cardiomyocyte hypertrophy, restoration of adrenergic receptor density, and improvement in calcium handling. Unfortunately, translation of these changes into definitive functional recovery at the organ level is infrequent. Reports of LVAD explantation rates because of cardiac recovery are highly variable depending on heart failure (HF) pathogenesis and weaning criteria used, with reported rates ranging from 4.5% to 45%.1 Institution-specific LVAD explantation evaluation protocols exist that use various hemodynamic, imaging, and gas exchange measurements in any combination. However, protocols that provide comprehensive assessment of myocardial performance under dynamic loading conditions and in response to the highly relevant physiological stress of exercise are lacking. The high reported rates of HF recurrence after LVAD explantation provide further motivation to carefully assess cardiac reserve capacity before explantation. This case series describes the use of a novel protocol in 2 patients, which integrates assessments of hemodynamic, imaging, and gas exchange measures during the state of rest, LVAD speed reduction, and exercise to uniquely characterize cardiac reserve capacity and guide LVAD explantation decision making.
A 44-year-old man underwent coronary bypass graft surgery in the setting of severe LV systolic dysfunction related to anabolic steroid use and coronary artery disease. Despite prolonged, postoperative mechanical support, there was no evidence of LV function recovery, and the patient underwent insertion of a Heartware LVAD (HVAD). Six months later, a transthoracic echo on full LVAD support revealed a low normal LV ejection fraction (LVEF), normal LV end-diastolic dimension, and aortic valve opening with each beat. These data prompted a LVAD explantation evaluation (algorithm outlined in Table and Figure 1).
Based on this algorithm, the integrated cardiopulmonary exercise test was performed on an upright cycle ergometer with HVAD speed turned down to 2000 rpm. The decrement in speed had minimal impact on measured Fick cardiac output (CO). His LVEF by first-pass radionuclide ventriculography remained normal at rest and at peak exercise. His peak VO2 (14.5 mL/kg/min) was reduced. However, the observed increment in pulmonary artery wedge pressure (PAWP; 5 to 16 mm Hg) was only modest relative to the increment in CO (7.1 to 14.6 L/min; PAWP–CO slope, 1.2 mm Hg/L/min; Figure 2; Table I in the Data Supplement). Heart rate augmented appropriately to 91% of predicted maximum. Peripheral oxygen extraction (Ca-vO2) was impaired despite normal hemoglobin, with a CvO2 remaining >10 mL/dL throughout exercise. Taken together, these findings indicated significant cardiac reserve capacity despite his abnormal peak VO2. The patient subsequently underwent successful LVAD explantation. One year after explantation, his LV systolic function remains normal (LVEF 50%), he is working full time, and he is able to perform 60 minutes of continuous treadmill exercise at a speed of 4.5 mph.
In contrast, we evaluated a 61-year-old patient in whom LVEF >45%, normal LV end-diastolic dimension, and New York Heart Association class I status prompted a comprehensive LVAD explantation evaluation. This patient had experienced an acute myocardial infarction complicated by cardiogenic shock 7 months prior requiring bridge-to-transplant HVAD placement. During his HVAD explantation evaluation, patient 2 had a peak VO2 of 14.1 mL/kg/min similar to patient 1. With LVAD speed reduction, his transthoracic echo indicated normal LV end-diastolic dimension, mildly depressed LVEF, and consistent aortic valve opening. During exercise, his heart rate increased to only 69% of predicted maximum despite exceeding a respiratory exchange ratio of 1.1. In contrast to patient 1, Ca-vO2 was normal, and he was able to augment his CO to only 8.3 L/min at peak exercise. His LV function was abnormal during exercise as evidenced by a significant increment in upright PAWP (7 to 20 mm Hg) coupled with impaired CO response to exercise (PAWP–CO slope, 3.2 mm Hg/L/min; Figure 3; Table I in the Data Supplement). There was also evident exercise oscillatory ventilation characteristic of advanced HF. These exercise-based findings suggested that the patient would not to be able to tolerate LVAD explantation. Thirty days later, the patient underwent cardiac transplantation.
LVAD explantation evaluation protocols are highly variable between centers. Proposed criteria have included LVEF >45%, LV end-diastolic dimension <60 mm, PAWP <12 mm Hg, cardiac index >2.8 L/min/m2, and solitary exercise measurement variables in the form of peak VO2 >16 mL/kg/min or an increase in minute ventilation relative to the production of carbon dioxide (VE/VCO2 slope) of <34 during low LVAD speed testing.2 In this report, we describe the incremental value of performing hemodynamic measurements during exercise to define cardiac reserve capacity.
In patients with cardiovascular disease and specifically with HF, there is increasing recognition of the incremental value of exercise-based measurements to assessments performed at rest. Vital sign assessment (heart rate and blood pressure), gas exchange parameters beyond peak VO2 (VE/VCO2 and exercise oscillatory ventilation), and hemodynamic measurements have all been shown to be prognostically significant in HF.3 The high rates of postexplant HF recurrence serves as a strong rationale to exquisitely characterize cardiac reserve in response to exercise.
Our explantation evaluation protocol is unique in that it consists of comprehensive assessments of biventricular performance using both invasive and noninvasive parameters simultaneously, combined with a maximal exercise test. Reliance on peak VO2 measurement as a sole exercise parameter for consideration of LVAD explantation is inadequate. Our first patient did not meet traditional cardiopulmonary exercise test explantation criteria based on a reduced peak VO2. In fact, in the 2 cases presented, the peak VO2 in isolation was nearly identical. However, relative contributions of CO to peak VO2 differed significantly, as did the excursion in PAWP relative to CO and the relative contribution of Ca-vO2 to peak VO2. Ca-vO2 is recognized to be an important determinant of exercise capacity in advanced HF and can be highly variable among individuals.4 In this case, patient 1’s peak VO2 of only 14.5 mL/kg/min could be explained largely by an impaired peak Ca-vO2. This finding may relate to lack of appropriate redistribution of blood flow to skeletal muscle during exercise and significant anemia.
Patient 1 was able to augment CO to >14 L/min in the setting of derived LVAD flows of only 2.4 to 3 L/min. In this patient’s case, a normal PAWP–CO slope and lack of exercise oscillatory ventilation, coupled with appropriate heart rate response, normal exercise LVEF, and a significant estimated intrinsic cardiac contribution to directly measure CO of >10 L/min at peak exercise, indicated robust cardiac reserve that prompted successful LVAD explantation.
LVAD explantation evaluation protocols remain heterogeneous across institutions. The use of a peak VO2 cut point in isolation may not adequately reflect cardiac reserve capacity. Consideration should be given to performing comprehensive exercise-based evaluations that permit ascertainment of filling pressures, CO, and variance in Ca-vO2 in evaluating LVAD explanation candidacy.
Sources of Funding
This article is supported by National Health Institutes and National Heart, Lung, and Blood Institute funding (1R01HL131029).
The Data Supplement is available at http://circheartfailure.ahajournals.org/lookup/suppl/doi:10.1161/CIRCHEARTFAILURE.116.003694/-/DC1.
- © 2016 American Heart Association, Inc.
- Ibrahim M,
- Yacoub MH.
- Malhotra R,
- Bakken K,
- D’Elia E,
- Lewis GD.
- Dhakal BP,
- Malhotra R,
- Murphy RM,
- Pappagianopoulos PP,
- Baggish AL,
- Weiner RB,
- Houstis NE,
- Eisman AS,
- Hough SS,
- Lewis GD.