Extracorporeal membrane oxygenation versus counterpulsatile, pulsatile, and continuous left ventricular unloading for pediatric mechanical circulatory support.

OBJECTIVES: Despite progress with adult ventricular assist devices, limited options exist to support pediatric patients with life-threatening heart disease. Extracorporeal membrane oxygenation remains the clinical standard. To characterize (patho)physiologic responses to different modes of mechanical unloading of the failing pediatric heart, extracorporeal membrane oxygenation was compared to intra-aortic balloon pump, pulsatile-flow ventricular assist device, or continuous-flow ventricular assist device support in a pediatric heart failure model. DESIGN: Experimental. SETTING: Large animal laboratory operating room. SUBJECTS: Yorkshire piglets (n = 47; 11.7 ± 2.6 kg). INTERVENTIONS: In piglets with coronary ligation-induced cardiac dysfunction, mechanical circulatory support devices were implanted and studied during maximum support. MEASUREMENTS AND MAIN RESULTS: Left ventricular, right ventricular, coronary, carotid, systemic arterial, and pulmonary arterial hemodynamics were measured with pressure and flow transducers. Myocardial oxygen consumption and total-body oxygen consumption were calculated from arterial, venous, and coronary sinus blood sampling. Blood flow was measured in 17 organs with microspheres. Paired Student t tests compared baseline and heart failure conditions. One-way repeated-measures analysis of variance compared heart failure, device support mode(s), and extracorporeal membrane oxygenation. Statistically significant (p < 0.05) findings included 1) an improved left ventricular blood supply/demand ratio during pulsatile-flow ventricular assist device, continuous-flow ventricular assist device, and extracorporeal membrane oxygenation but not intra-aortic balloon pump support, 2) an improved global myocardial blood supply/demand ratio during pulsatile-flow ventricular assist device and continuous-flow ventricular assist device but not intra-aortic balloon pump or extracorporeal membrane oxygenation support, and 3) diminished pulsatility during extracorporeal membrane oxygenation and continuous-flow ventricular assist device but not intra-aortic balloon pump and pulsatile-flow ventricular assist device support. A profile of systems-based responses was established for each type of support. CONCLUSIONS: Each type of pediatric ventricular assist device provided hemodynamic support by unloading the heart with a different mechanism that created a unique profile of physiological changes. These data contribute novel, clinically relevant insight into pediatric mechanical circulatory support and establish an important resource for pediatric device development and patient selection.

Intraaortic balloon pump timing discrepancies in adult patients.

The objective of this clinical study was to quantify the incidence and magnitude of intraaortic balloon pump (IABP) inflation and deflation landmark discrepancies associated with the IABP catheter arterial pressure waveform. Cardiac surgery patients with an IABP inserted prior to surgery were recruited. Following cardiac exposure, a high-fidelity pressure catheter was inserted into the aortic root for digital recording. The radial artery pressure signal was simultaneously recorded from the patient monitor along with the arterial pressure and electrocardiogram waveforms from the IABP console while operating at 1:1 and 1:2 synchronization. In selected patients, recordings were obtained with the IABP timed to the high-fidelity aortic root waveform. In all 11 patients, inflation and deflation landmark delays were observed when comparing the aortic root waveforms to the IABP arterial pressure waveforms (inflation delay = 74 ± 29 [23-117] ms; deflation delay = 71 ± 37 [24-141] ms, mean ± standard deviation [min-max]). Delays were greater when compared to the radial artery waveform (inflation delay = 175 ± 50 [100-233] ms; deflation delay = 168 ± 52 [100-274] ms). In all cases, the landmark delays were statistically different from zero (P < 0.001). Diastolic augmentation and afterload reduction varied with waveform source. Conflicting indications of afterload reduction occurred in four patients. Timing to the aortic root waveform resulted in greater diastolic pressure augmentation and afterload reduction but mixed changes in stroke volume. Delay and distortion of the arterial waveform was consistently found when measured through the IABP catheter lumen. These delays can alter IABP efficacy and may be eliminated by using high-fidelity sensing of aortic pressure.

The effect of gravitational acceleration on cardiac diastolic function: a biofluid mechanical perspective with initial results.

Echocardiographic measurements of astronaut cardiac function have documented an initial increase, followed by a progressive reduction in both left ventricular end-diastolic volume index and stroke volume with entry into microgravity (micro-G). The investigators hypothesize that the observed reduction in cardiac filling may, in part, be due to the absence of a gravitational acceleration dependent, intraventricular hydrostatic pressure difference in micro-G that exists in the ventricle in normal gravity (1-G) due to its size and anatomic orientation. This acceleration-dependent pressure difference, DeltaP(LV), between the base and the apex of the heart for the upright posture can be estimated to be 6660 dynes/cm(2) ( approximately 5 mm Hg) on Earth. DeltaP(LV) promotes cardiac diastolic filling on Earth, but is absent in micro-G. If the proposed hypothesis is correct, cardiac pumping performance would be diminished in micro-G. To test this hypothesis, ventricular function experiments were conducted in the 1-G environment using an artificial ventricle pumping on a mock circulation system with the longitudinal axis anatomically oriented for the upright posture at 45 degrees to the horizon. Additional measurements were made with the ventricle horizontally oriented to null DeltaP(LV)along the apex-base axis of the heart as would be the case for the supine posture, but resulting in a lesser hydrostatic pressure difference along the minor (anterior-posterior) axis. Comparative experiments were also conducted in the micro-G environment of orbital space flight on board the Space Shuttle. This paper reviews the use of an automated cardiovascular simulator flown on STS-85 and STS-95 as a Get Away Special payload to test this hypothesis. The simulator consisted of a pneumatically actuated, artificial ventricle connected to a closed-loop, fluid circuit with adjustable compliance and resistance elements to create physiologic pressure and flow conditions. Ventricular instrumentation included pressure transducers in the apex and base as well as immediately upstream of the inflow valve and downstream of the outflow valve, and a flow probe downstream of the outflow valve. By varying the circulating fluid volume, ventricular function could be determined for varying preload pressures at a regulated, mean afterload pressure of 95 mm Hg. This variation in preload condition permitted the construction of a ventricular function curve for the micro-G environment for comparison to the same curve for the 1-G environment. Data were collected from both missions at the upper end of the ventricular function curve. Experiment operation in the 1-G, supine orientation or in the micro-G environment eliminated the DeltaP(LV) observed in the 1-G, upright orientation. Consistent with the hypothesis, additional atrial pressure was required in micro-G to obtain stroke volumes and flow rates similar to those measured in 1-G for the upright posture. The necessary increase in atrial pressure was approximately 5 mm Hg in these experiments. In the same range of flow rates and stroke volumes, similar flows were observed in the 1-G supine posture for atrial pressures intermediate to the 1-G upright and micro-G values, also consistent with the hypothesis. Additional experiments on board the Space Shuttle are in preparation to gather data across the rest of the normal physiologic range of the ventricular function curve.

Physiologic control of rotary blood pumps: an in vitro study.

Rotary blood pumps (RBPs) are currently being used as a bridge to transplantation as well as for myocardial recovery and destination therapy for patients with heart failure. Physiologic control systems for RBPs that can automatically and autonomously adjust the pump flow to match the physiologic requirement of the patient are needed to reduce human intervention and error, while improving the quality of life. Physiologic control systems for RBPs should ensure adequate perfusion while avoiding inflow occlusion via left ventricular (LV) suction for varying clinical and physical activity conditions. For RBPs used as left ventricular assist devices (LVADs), we hypothesize that maintaining a constant average pressure difference between the pulmonary vein and the aorta (deltaPa) would give rise to a physiologically adequate perfusion while avoiding LV suction. Using a mock circulatory system, we tested the performance of the control strategy of maintaining a constant average deltaPa and compared it with the results obtained when a constant average pump pressure head (deltaP) and constant rpm are maintained. The comparison was made for normal, failing, and asystolic left heart during rest and at light exercise. The deltaPa was maintained at 95 +/- 1 mm Hg for all the scenarios. The results indicate that the deltaPa control strategy maintained or restored the total flow rate to that of the physiologically normal heart during rest (3.8 L/m) and light exercise (5.4 L/m) conditions. The deltaPa approach adapted to changing exercise and clinical conditions better than the constant rpm and constant deltaP control strategies. The deltaPa control strategy requires the implantation of two pressure sensors, which may not be clinically feasible. Sensorless RBP control using the deltaPa algorithm, which can eliminate the failure prone pressure sensors, is being currently investigated.

Characterization of an adult mock circulation for testing cardiac support devices.

A need exists for a mock circulation that behaves in a physiologic manner for testing cardiac devices in normal and pathologic states. To address this need, an integrated mock cardiovascular system consisting of an atrium, ventricle, and systemic and coronary vasculature was developed specifically for testing ventricular assist devices (VADs). This test configuration enables atrial or ventricular apex inflow and aortic outflow cannulation connections. The objective of this study was to assess the ability of the mock ventricle to mimic the Frank-Starling response of normal, heart failure, and cardiac recovery conditions. The pressure-volume relationship of the mock ventricle was evaluated by varying ventricular volume over a wide range via atrial (preload) and aortic (afterload) occlusions. The input impedance of the mock vasculature was calculated using aortic pressure and flow measurements and also was used to estimate resistance, compliance, and inertial mechanical properties of the circulatory system. Results demonstrated that the mock ventricle pressure-volume loops and the end diastolic and end systolic pressure-volume relationships are representative of the Starling characteristics of the natural heart for each of the test conditions. The mock vasculature can be configured to mimic the input impedance and mechanical properties of native vasculature in the normal state. Although mock circulation testing systems cannot replace in vivo models, this configuration should be well suited for developing experimental protocols, testing device feedback control algorithms, investigating flow profiles, and training surgical staff on the operational procedures of cardiovascular devices.

Left ventricular and myocardial perfusion responses to volume unloading and afterload reduction in a computer simulation.

Ventricular assist devices (VADs) have been used successfully as a bridge to transplant in heart failure patients by unloading ventricular volume and restoring the circulation. In a few cases, patients have been successfully weaned from these devices after myocardial recovery. To promote myocardial recovery and alleviate the demand for donor organs, we are developing an artificial vasculature device (AVD) that is designed to allow the heart to fill to its normal volume but eject against a lower afterload. Using this approach, the heart ejects its stroke volume (SV) into an AVD anastomosed to the aortic arch, which has been programmed to produce any desired afterload condition defined by an input impedance profile. During diastole, the AVD returns this SV to the aorta, providing counterpulsation. Dynamic computer models of each of the assist devices (AVD, continuous, and pulsatile flow pumps) were developed and coupled to a model of the cardiovascular system. Computer simulations of these assist techniques were conducted to predict physiologic responses. Hemodynamic parameters, ventricular pressure-volume loops, and vascular impedance characteristics were calculated with AVD, continuous VAD, and asynchronous pulsatile VAD support for a range of clinical cardiac conditions (normal, failing, and recovering left ventricle). These simulation results indicate that the AVD may provide better coronary perfusion, as well as lower vascular resistance and elastance seen by the native heart during ejection compared with continuous and pulsatile VAD. Our working hypothesis is that by controlling afterload using the AVD approach, ventricular cannulation can be eliminated, myocardial perfusion improved, myocardial compliance and resistance restored, and effective weaning protocols developed that promote myocardial recovery.

Hemodynamic and pressure-volume responses to continuous and pulsatile ventricular assist in an adult mock circulation.

This study investigated the hemodynamic and left ventricular (LV) pressure-volume loop responses to continuous versus pulsatile assist techniques at 50% and 100% bypass flow rates during simulated ventricular pathophysiologic states (normal, failing, recovery) with Starling response behavior in an adult mock circulation. The rationale for this approach was the desire to conduct a preliminary investigation in a well controlled environment that cannot be as easily produced in an animal model or clinical setting. Continuous and pulsatile flow ventricular assist devices (VADs) were connected to ventricular apical and aortic root return cannulae. The mock circulation was instrumented with a pressure-volume conductance catheter for simultaneous measurement of aortic root pressure and LV pressure and volume; a left atrial pressure catheter; a distal aortic pressure catheter; and aortic root, aortic distal, VAD output, and coronary flow probes. Filling pressures (mean left atrial and LV end diastolic) were reduced with each assist technique; continuous assist reduced filling pressures by 50% more than pulsatile. This reduction, however, was at the expense of a higher mean distal aortic pressure and lower diastolic to systolic coronary artery flow ratio. At full bypass flow (100%) for both assist devices, there was a pronounced effect on hemodynamic parameters, whereas the lesser bypass flow (50%) had only a slight influence. Hemodynamic responses to continuous and pulsatile assist during simulated heart failure differed from normal and recovery states. These findings suggest the potential for differences in endocardial perfusion between assist techniques that may warrant further investigation in an in vivo model, the need for controlling the amount of bypass flow, and the importance in considering the choice of in vivo model.

Expanded pediatric cardiovascular simulator for research and training.

A mock circulation system has been developed to approximate key anatomic features and simulate the pressures and flows of an infant. Pulsatile flow is generated by 10 cc pulsatile ventricles (Utah infant ventricular assist device). Systemic vasculature is mimicked through the use of 3/8" ID bypass tubing with two flexible reservoirs to provide compliance. Vascular resistance, including pulmonary, aortic, and major branches, is controlled via a series of variable pinch clamps. The coronary branch has a dynamic resistor so that the majority of flow occurs during diastole. The system is instrumented to measure key pressures and flows. Right atrial pressure, left atrial pressure, pulmonary artery pressure, and mean aortic pressure are measured with high-fidelity pressure catheters (Millar Instruments, Houston, TX). Flows are measured by transit time ultrasonic flow probes (Transonic Systems, Ithaca, NY) in the pulmonary artery, aorta, coronary artery, and brachiocephalic artery along with assist device flow. The system can be tuned to create the hemodynamic values of a pediatric patient under normal or heart failure conditions. Once tuned to the desired hemodynamic conditions, the loop may be used to test the performance of various circulatory support systems including the intra-aortic balloon pump, left and right ventricular assist devices, or cardiopulmonary support systems such as extracorporeal membrane oxygenation.

The influence of mock circulation input impedance on valve acceleration during in vitro cardiac device testing.

For a mechanical heart valve, a strong spike in pressure during closing is associated with valve wear and erythrocyte damage; thus, for valid in vitro testing, the mock circulation system should replicate the conditions, including pressure spikes, expected in vivo. To address this issue, a study was performed to investigate how mock circulation input impedance affects valve closure dynamics. A left ventricular model with polyurethane trileaflet inflow valve and tilting disc outflow valve was connected to a Louisville mock circulation system, which incorporates 2 adjustable flow resistors and 2 compliances. In the study, 116 cases matched zero frequency modulus well (982-1147 dyn x s/cm), but higher harmonics were purposely varied. Acceleration measured at the outflow valve ring (42.4-89.4 milli-Gs) was uncorrelated with impedance error (74.1-237 dyn x s/cm relative to target impedance), but was correlated with end-systolic impedance (1082-1319 dyn x s/cm) for cases with high zero frequency modulus, which exhibited just less than full ejection. These differences demonstrate that mock circulation response affects the magnitude of the closing spike, indicating that control of this parameter is necessary for authentic testing of valves. Correlation of acceleration to end-systolic impedance was weak for low zero frequency modulus, which tended toward full or hyperejection, reinforcing common laboratory observations that valve closing also depends on ventricular operating conditions.

Ascending aorta outflow graft location and pulsatile ventricular assist provide optimal hemodynamic support in an adult mock circulation.

Although continuous flow (CF) and pulsatile flow (PF) ventricular assist devices (VADs) are being clinically used, their effects on aortic blood flow, as a measure of overall blood distribution, remain unclear. In acute VAD support animal experiments, our group has described a zone of turbulent mixing in the aortic arch. The objective of this study was to confirm this finding in the controlled setting of an adult mock circulation, simulating ventricular pathophysiologic states (normal and failing ventricle). CF and PF flow VADs were connected to ventricular apical inflow and ascending aorta (AA) or descending aorta (DA) outflow cannulae. Cardiovascular pressure and flow waveforms were recorded at varying levels of VAD bypass resulting in four test conditions: (i) CF-AA; (ii) CF-DA; (iii) PF-AA; and (iv) PF-DA. Confirming the animal data, no differences in mean aortic flow between CF and PF VADs were found, and significantly lower mean aortic arch flow with DA cannulation was noted. Mean aortic root flow decreased with increasing VAD bypass flow. As in the animal studies, despite similar mean flow rates, significant differences in waveform morphology were observed for AA and DA outflow graft locations and varying levels of VAD bypass. At 100% VAD support in the failing heart, PF restored waveform pulsatility to normal baseline while CF resulted in little pulsatility. These results confirm our earlier findings in the animal model, suggesting that outflow graft location may have a significant effect on aortic blood flow distribution. The long-term implications of these findings are being examined in ongoing studies.