LABORATORY INVESTIGATION VENTRICULAR PERFORMANCE Right heart function during left heart assist and the effects of volume loading in a canine preparation

A significant fraction of patients in whom mechanical left ventricular assist devices are implanted for refractory cardiac failure after open heart surgery have had the complication of right heart failure. To evaluate the effects of left ventricular assistance and pressure unloading on right ventricular function, we performed experiments in the normal hearts of open-chest, anesthetized, large mongrel dogs. We compared right ventricular function before and after left ventricular-to-aortic bypass with a roller pump at right atrial pressure levels of 1, 3, 5, and 7 mm Hg produced by volume loading. No significant changes were found in cardiac output or stroke volume over this range of right atrial pressures when comparing that before to that during left ventricular bypass, which at a right atrial pressure of 1 mm Hg reduced peak left ventricular pressure from 96 + 6 to 15 + 9 mm Hg and at a right atrial pressure of 5 mm Hg reduced it from 113 + 3 to 29 + 12 mm Hg, while maintaining aortic pressure. There was no evidence of right ventricular failure under these conditions: (from before to during bypass) at a right atrial pressure of 1 mm Hg cardiac output was 3.4 + 0.4 to 3.7 + 0.6 liter/min and stroke volume was 28 + 5 to 33 + 6 ml; during volume loading at a right atrial pressure of 7 mm Hg cardiac output was 5.6 + 0.6 to 5.7 + 0.7 liter/min and stroke volume was 47 + 5 to 52 + 5 ml. However, there were significant (25% to 40%) reductions in the maximum and minimum rates of change in right ventricular pressure and thc maximum rate of change in pulmonary arterial flow., suggesting reductions in right ventricular contractility perhaps via reduced forces in the interventricular septum during left ventricular pressure unloading. There were also significant 10% to 30% reductions in right ventricular peak systolic pressure and in mean pulmonary arterial pressure indicative of reduced afterload apparently produced by direct reductions in left ventricular filling pressures during left ventricular bypass. In separate experiments in which right ventricular peak systolic pressure was held at constant levels, maximum dP/dt was reduced from before to during bypass from 322 + 33 to 227 + 28 mm Hg/sec at a right ventricular peak systolic pressure of 21 mm Hg and from 427 22 to 318 + 18 at a right ventricular peak systolic pressure of 32 mm Hg. Reduced contractility is also suggested because there was no evidence that end-diastolic volume, stroke volume, or ejection fraction changed in response to the reduced afterload. The results illustrate that the beneficial effects of right ventricular afterload reduction with left ventricular assistance can balance the detrimental effects of impairment of contractility, leaving stroke volume unchanged. Thus, in the normal dog heart no change in the ability of the right ventricle to eject a required stroke volume was detected during volume loading or during left ventricular pressure unloading with left ventricular bypass and assistance. Circulation 70, No. 4, 708-716, 1984. THE SUCCESSFUL CLINICAL use of left ventricular assist devices (LVADs) depends not only on safe blood contacting biomaterials and effective engineering performance and control features, but also on an understanding of the effects of the device on host organ From the Department of Cardiovascular Surgery. Presbyterian Hospital of Pacific Medical Center. and the Medical Research Institute of San Francisco. Supported by NHLBI grant HL27275. Address for correspondence: Dr. David J. Farrar, Department of Cardiovascular Surgery. Presbyterian Hospital of Pacific Medical Center, San Francisco, CA 94120. Received Oct. 10, 1983; revision accepted July 19. 1984. 708 systems. One major problem area demonstrated in clinical experience with LVAD for refractory cardiac failure after open heart operations is right ventricular failure. Of 103 patients with LVADs reported by nine groups, including our own, 27 patients exhibited right ventricular failure that was refractory to pharmacologic support,' 1-3 although 41 additional LVAD patients were reported in articles that do not adequately discuss right ventricular failure. 14-16 The exact cause of right ventricular failure during left ventricular assistance is unknown, and there is a paucity of experimental data in the few reports and CIRCULATION by gest on A ril 0, 2017 http://ciajournals.org/ D ow nladed from LABORATORY INVESTIGATION-VENTRICULAR PERFORMANCE abstracts that have been published concerning the effects of an LVAD on the right ventricle.'1 2-2 One leading possibility is that right ventricular failure in patients who have undergone open heart surgery may be due to the same myocardial ischemia and depression of cardiac function during cardiopulmonary bypass that caused the left heart to fail.21 This preexisting right ventricular disease and poor right ventricular function is unmasked with the increased venous return to the right heart generated by the assist device. Another possibility is that volume and pressure unloading the left ventricle with a LVAD reduces the coupling between the ventricles via the interventricular septum and reduces the contribution of left ventricular contraction to an already marginal right ventricle. Thus, abnormal septal contraction or reduced septal forces may contribute to right ventricular failure. These are important phenomena to understand from the standpoint of proper control of the assist pump, for helping to interpret clinical findings from patients with various cardiac pathology, and for proper clinical decisions on whether to provide right or left ventricular assistance or a total artificial heart. The present study was designed to evaluate right heart function during left ventricular assistance in the normal dog heart to determine if right ventricular pump function is impaired with left ventricular pressure unloading and with increased venous return during volume loading.s that have been published concerning the effects of an LVAD on the right ventricle.'1 2-2 One leading possibility is that right ventricular failure in patients who have undergone open heart surgery may be due to the same myocardial ischemia and depression of cardiac function during cardiopulmonary bypass that caused the left heart to fail.21 This preexisting right ventricular disease and poor right ventricular function is unmasked with the increased venous return to the right heart generated by the assist device. Another possibility is that volume and pressure unloading the left ventricle with a LVAD reduces the coupling between the ventricles via the interventricular septum and reduces the contribution of left ventricular contraction to an already marginal right ventricle. Thus, abnormal septal contraction or reduced septal forces may contribute to right ventricular failure. These are important phenomena to understand from the standpoint of proper control of the assist pump, for helping to interpret clinical findings from patients with various cardiac pathology, and for proper clinical decisions on whether to provide right or left ventricular assistance or a total artificial heart. The present study was designed to evaluate right heart function during left ventricular assistance in the normal dog heart to determine if right ventricular pump function is impaired with left ventricular pressure unloading and with increased venous return during volume loading. Methods Eleven mongrel dogs of both sexes weighing between 27 and 35 kg were anesthetized with intravenous sodium pentobarbitol (35 mg/kg), intubated, and connected to a respirator adjusted to maintain blood gases within a normal range. Anesthesia was maintained by periodic intravenous doses of fentanyl (4 to 6 ,gl kg) as required. A left thoracotomy was performed on each dog and the heart was suspended in a pericardial cradle. Two No. 32F cannulas were inserted into the left atrium, one of which was placed across the mitral valve and into the left ventricle. These cannulas were attached to a Y connector and then to a roller pump that returned blood to the abdominal aorta at the bifurcation. In each of the first seven dogs, a reservoir containing dextran 40 premixed with the dog's blood and heated to body temperature was connected to the right atrium for volume loading and for producing increases in right atrial pressure (figure 1). The first group of dogs were instrumented for the following: High-fidelity right ventricular and pulmonary arterial pressures were measured with Konigsberg P9.5 and Millar PC350 pressure transducers, respectively. A specially modified SwanGanz catheter (Edwards Lab) was introduced with an injection port in the right ventricle and with a fast-response thermistor in the pulmonary artery. The average time constant of the thermistor-catheter systems was 173 msec, as measured in water baths at two different temperatures. The right ventricular and pulmonary arterial ports of the catheter were connected to Statham P23Db pressure transducers for periodic check in vivo of zero for the high-fidelity transducers. Left ventricular peak systolic Vol. 70, No. 4, October 1984 FIGURE 1. Experimental set up for determining right ventricular function in the dog preparation of left ventricular bypass. Pressures were measured from the right atrium (RA), right ventricle (RV), left ventricle (LV). pulmonary artery (PA). and aorta (A). and flow was measured from the PA. and end-diastolic pressures were measured with a Millar transducer connected to a short polyethelene catheter inserted through the left ventricular bypass cannula and into the left ventricle. Cardiac output was measured from the pulmonary artery with an electromagnetic blood flowmeter and flowprobes calibrated at the factory with whole blood (Carolina Medical Electronics). Right atrial and aortic pressures were measured through polyethelene catheters connected to Statham P23Db transducers. Zero pressure was taken at the level at the right atrium. All signals were connected to a Gould model 2800 eightchannel recorder. The first derivative of right ventricular pressure (dP/dt) and of pulmonary arterial blood flow (dQ/dt) were determined electronically with a Gould differentiator preamplifier. Measurements of all parameters were made with the left ventricular bypass pump off and on. Flow rates of the bypass pump were adjusted to obtain a maximum unloading effect by reducing maximum left ventricular pressure to under 30 mm Hg. After baseline measurements were made with the left ventricular bypass off and on, volume from the right atrial pressure reservoir was slowly added to raise right atrial pressure by approximately 2 mm Hg after a 10 min stabilization period. Measurements were then repeated with and without left ventricular bypass. This procedure of volume loading and comparing values obtained with left ventricular bypass off and on was repeated two more times to obtain data over a range of right atrial pressures from approximately 1 to 7 mm Hg. Estimates of right ventricular volume were determined by the thermodilution technique based on the work of Rapaport et al.23' 24 and Balcon and Oram.25 A rapid injection of 10 ml of iced 5% dextrose solution was made into the right ventricular port on the Swan-Ganz catheter and the runoff curve was measured on the chart recorder. Injections were repeated five to seven times at each experimental intervention. Right ventricular ejection fraction (RVEF) was estimated as RVEF = (T(k1) -Tk)/T(k 1) 709 by gest on A ril 0, 2017 http://ciajournals.org/ D ow nladed from