Hemolysis evaluation of centrifugal pumps using microcapsule suspension.

OBJECTIVES: Bovine and human blood has been widely used for in vitro hemolysis testing to evaluate centrifugal cardiac assist pumps. However, results from such tests are complicated by variations in the susceptibility of individual red blood cells to shear. The objective of this study was to evaluate the use of microcapsule suspension as an alternative to bovine or human blood for hemolysis testing. METHODS: Microcapsule suspensions of 100 micro m maximal diameter (average 79.1 micro m) with a polyurethane membrane were used. Each microcapsule contained a leuco dye, which was used to measure "hemolysis" in the suspension after exposure to mechanical stress. Six centrifugal pumps were used to measure and compare the hemolysis values of microcapsule suspensions, bovine blood and human blood. RESULTS: Correlations were significant between the hemolysis values measured using microcapsule suspensions and those using bovine blood (R = 0.965, p = 0.002) and human blood (R = 0.940, p = 0.005). CONCLUSIONS: Microcapsule suspension can be successfully used instead of blood to compare the relative hemolytic performance of centrifugal blood pumps.

In vivo hemodynamic performance of the Cleveland Clinic CorAide blood pump in calves.

BACKGROUND: The Cleveland Clinic CorAide left ventricular assist system is based on a small implantable continuous-flow centrifugal blood pump with a completely suspended rotating assembly designed for long-term circulatory support (5 to 10 years). METHODS: Between June 1999 and August 2000, the CorAide blood pump was implanted in 10 calves for 1 month and in 3 calves for 3 months. RESULTS: The mean pump flow and arterial pressure were 6.1 +/- 1.1 L/min and 97 +/- 5 mm Hg, respectively. The mean plasma free-hemoglobin level after postoperative day 3 was 2.0 +/- 1.8 mg/dL. Renal and hepatic function remained normal in all cases. There was no incidence of mechanical failure, hemolysis, bleeding, or systemic organ dysfunction in any of the cases. Significant findings at autopsy were limited to two cases of renal infarction, one of which was associated with an outflow graft infection. CONCLUSIONS: The CorAide blood pump is easily implanted, reliable, nonhemolytic, and nonthrombogenic, positioning it as a leading third-generation, continuous-flow left ventricular assist system with a completely suspended rotor.

Numeric flow simulation for an innovative ventricular assist system secondary impeller.

The secondary impeller of the Cleveland Clinic Foundation Innovative Ventricular Assist System heart pump is designed to pump the blood from the journal bearing through the annulus back to the discharge to rejoin the main flow. It is crucial, however, that the amount of hemolysis and potential for thrombus formation is minimized. Based on our earlier computational fluid analysis of the initial design, an improved secondary impeller has been proposed. The objective of this work was to analyze the flow with the new impeller configuration by computational fluid dynamics and to find out where in the geometry hemolysis becomes high, as well as where the potential for thrombosis exists. The present work includes: 1) visualizing the 3-dimensional flow pattern in the secondary impeller cavity, 2) investigating the tip clearance effect, and 3) estimating the shear stress distribution as a measure for blood damage caused in the secondary impeller cavity.

Numeric modeling of the cardiovascular system with a left ventricular assist device.

A numeric model consisting of a lump-parameter cardiovascular system (CVS) model and a model for the Cleveland Clinic Implantable Ventricular Assist System (IVAS), a nonpulsatile rotary pump designed to augment the failing left ventricle, are described in this paper. The purposes of this study were to 1) observe the hemodynamic interactions between CVS and IVAS under various physiologic and pathophysiologic conditions running at different speeds; and 2) allow testing and optimization of various IVAS control algorithms. An existing numeric model of CVS (24 coupled differential equations, representing all cardiac chambers and systemic and pulmonary vasculature) was modified to add the IVAS pump as an auxiliary chamber between the left ventricle and aorta with pressure-flow-speed characteristics derived from in vitro testing. Simulations were conducted for ventricles with normal and abnormal systolic and diastolic dysfunction at different exercise levels with the pump running at various speeds. Computer simulations show that 1) numeric modeling is useful for predicting hemodynamic response of CVS to IVAS in various circumstances; 2) IVAS results in normalization of cardiac output, especially in failing hearts, although with reduced pulse pressure; and 3) various control algorithms allowing adaptation of IVAS to physiologic demands of CVS could be developed based on the simulation study.

Analysis of a new PM motor design for a rotary dynamic blood Pump.

The permanent magnet (PM) motor for a rotary dynamic blood pump requires high power density to coordinate the motor size with the limited pump space and high efficiency to reduce the size and weight of the associated batteries. The motor also serves as a passive axial magnetic thrust bearing, a reacting hydraulic force, and provides a stabilizing force for the radial journal bearing. This article presents analysis of a new PM motor for the blood pump application. High power density is achieved by using the Halbach magnetic array, and high efficiency is accomplished by optimizing the rotor magnet assembly and the stator slots/windings. While both radial and axial forces are greatly enhanced, pulsating components of the torque and force are also significantly reduced.

Design of a DSP controller for an innovative ventricular assist system.

The design and development of the digital signal processor controller for an innovative ventricular assist system is presented. A DSP56005 is used as the central processor, with other peripheral components. System hardware and software were developed through the advanced development system, and stand alone operation of the system was also accomplished. Two different control modes--current control mode and speed control mode--were developed and investigated. Performance of efficiency and dynamic response were examined through experimental testing.

Development of a 6 x 18 inch rheology tunnel for experimental fluid dynamics investigation.

The Ohio State University (OSU) and the Cleveland Clinic Foundation (CCF) developed a 6 x 18 inch low velocity Rheologic Research Tunnel to do flow visualization and other experimental fluid studies, particularly on scaled-up models of cardiovascular devices, such as the CCF's Innovative Ventricular Assist System. The large test section (TS) permits detailed data to be obtained that would be inaccessible with a smaller test prototype. A particular feature of the OSU-CCF program is the use of a non-Newtonian blood analog (NNBA), so the effect of the shear-thinning behavior of blood on the local development of separation, stagnation, and flow patterns can be studied. The TS can simulate a pressure driven slit flow of 6 x 18 in., or the external flow around a vane or blade having an aspect ratio of 1. Maximum pressure is 8.5 psig, while the maximum velocity is 21.7 in/sec. The fluid supply tank has a capacity of 500 gal of NNBA and, with its associated filtration and circulation systems, can be adapted to studies of large transparent models better studied outside the TS. Using 2 pumps, flow rates of 98-610 gal/min can be provided. Instrumentation includes thermistors, a 48 port pressure scanner with pressure transducers, a data acquisition system, and a digital video camera. Dye and hydrogen bubble systems have been developed. Development of such a facility presents problems not encountered in more typical water tables or wind tunnels. These include fundamental issues such as providing a uniform flowfield; practical issues with respect to priming, operating, and obtaining data from the system; and safety considerations. For the very large volume of NNBA, a xanthan gum solution is used, whose shear-thinning behavior depends not only on concentration, but also on age and prior shear history. The lessons learned are presented, permitting others to efficiently develop systems suitable to their testing needs.

Blood flow analysis for the secondary impeller of an IVAS heart pump.

The rotodynamic heart pump (IVAS), designed by the Cleveland Clinic Foundation, includes a secondary flow path along the journal bearing, through a secondary impeller, and over the rotor outer surface. The flow behaviors of the blood through the journal bearing and the secondary impeller are investigated by a computational fluid dynamics method that solves the 3-dimensional Navier-Stokes equations using a new solution algorithm. Results of the analyses include: 1) the blood flow patterns within the journal bearing, 2) the effect of the non-uniform bearing clearance on the flow patterns of the impeller cavity, 3) the flow patterns around a secondary impeller blade that include effects of tip clearance and the gap between the blade and the inner or outer side wall, 4) effects of the blade angles on the secondary impeller performance, and 5) the shear stress distribution.

Flow analysis of the Cleveland Clinic centrifugal pump.

An implantable ventricular assist blood pump is being developed by the Cleveland Clinic Foundation in cooperation with the NASA Lewis Research Center. At the nominal design condition, the pump provides blood flow at the rate of 5 L/min at a pressure rise of 100 mmHg and a rotation speed of 3000 RPM. Bench testing of the centrifugal pump in a water/glycerin mixture has provided flow and pressure data at several rotative speeds. A one-dimensional empirically based pump flow analysis computer code developed at NASA Lewis Research Center has been used in the design process to simulate the flow in the primary radial pump stage. The computer model was used to size key impeller and volute geometric parameters that influence pressure rise and flow. Input requirements to the computer model include a simple representation of the pump geometry. The model estimates the flow conditions under design and off-design operating conditions at the impeller leading and trailing edges, and the volute inlet and exit. Output from the computer model is compared to flow and pressure data obtained from bench testing.

The Cleveland Clinic rotodynamic pump program.

The Cleveland Clinic Foundation has developed a unique rotodynamic blood pump for future use as a permanent implant. This pump is small (2.5 x 2.5 inches) and requires an electric input power of 7 watts to produce 5 L/min of blood flow against 100 mm Hg at 3,000 rpm. Initial in vivo testing has confirmed in vitro function and shown low hemolysis. Endurance bench testing has exceeded 12 months of continuous function. This pump is the basis of an innovative ventricular assist system in which power is supplied by a tranocutaneous electrical transmission system, and pacer technology is used for both control logic and telemetry functions. The resulting system will be completed and tested under an NHLBI contract during the next 5 years.

Chronic nonpulsatile blood flow. II. Hemodynamic responses to progressive exercise in calves with chronic nonpulsatile biventricular bypass.

We investigated the effects of stepwise treadmill exercise on animal (calf) hemodynamic variables during chronic nonpulsatile biventricular bypass with ventricular fibrillation. Seven days was allowed for recovery from the effects of anesthesia and surgery; each animal's natural heart was then fibrillated. The pump flows were maintained at nominal rates of 90, 100, and 120 ml.kg-1.min-1 for 1 week each, with the order varying from experiment to experiment. A total of 30 incremental exercise tests were performed on five animals. No significant changes in mean aortic pressure were observed during nonpulsatile perfusion at the three nominal flow rates of nonpulsatile flow either before or during exercise. The systemic vascular resistance decreased significantly during exercise (from 705 +/- 22 to 547 +/- 81 dyne.sec.cm-5, p < 0.01, and from 604 +/- 25 to 510 +/- 15 dyne.sec.cm-5, p < 0.05, at nominal flow rates of 100 and 120 ml.kg-1.min-1, respectively). There were also significant (analysis of variance, Scheffe test, p < 0.05) differences in systemic vascular resistance among three nominal flow rates both before and during exercise. These results suggest that the autonomic nerve reflex control of the cardiovascular system in physical exercise was functioning normally in animals with chronic nonpulsatile blood flow.

Chronic nonpulsatile blood flow. III. Effects of pump flow rate on oxygen transport and utilization in chronic nonpulsatile biventricular bypass.

The relationship between blood flow and oxygen transport was studied in five calves with chronic nonpulsatile biventricular bypass. Seven days was allowed for recovery from the effects of anesthesia and operation; the natural heart was then fibrillated. Pump flows were maintained at nominal rates of 90, 100, or 120 ml.kg-1.min for 1 week each, with the sequence varied from experiment to experiment. Venous and arterial blood samples were taken at rest for blood gas analysis. Serum lactate analysis was done twice a week, on the third and seventh days after each pump flow change. Serum catecholamine levels were assayed on the seventh day of each flow rate. Progressive exercise tests were also conducted during each test segment. Basal oxygen consumption of a 4-month-old calf was 6.3 +/- 0.3 ml.kg-1.min-1. The mixed venous oxygen tension decreased when pump flow rate was reduced (29.6 +/- 1.0, 28.3 +/- 1.2, and 23.8 +/- 0.9 mm Hg at 120, 100, and 90 ml.kg-1.min-1 of pump flow, respectively), and oxygen extraction increased linearly when pump flow rate was reduced. Hemoglobin concentration significantly affected oxygen extraction rate. Serum lactate concentration increased significantly at a 90 ml.kg-1.min-1 perfusion compared with concentrations at other pump flow rates (7.81 +/- 2.42 mEq/L at 90 ml.kg-1.min-1 vs 0.71 +/- 0.19 and 0.73 +/- 0.81 mEq/L at 100 and 120 ml.kg-1.min-1, respectively; p < 0.01, analysis of variance, Scheffe F test). Maximum oxygen extraction during exercise was 78%. These results suggest that a critical flow level between 90 and 100 ml.kg-1.min-1 maintains oxidative metabolism in the calf with chronic nonpulsatile flow. The resulting oxygen delivery was slightly higher than that indicated in the literature. Maximal oxygen extraction was normal.

Cleveland clinic rotodynamic pump.

BACKGROUND: It is now accepted that 70% to 80% of patients with end-stage heart failure would benefit from a permanent implanted left ventricular assist device. Previously there was little consideration of the use of nonpulsatile pumps for this function. METHODS: An extensive 5-year engineering research and development program to develop a permanent implanted nonpulsatile blood pump has been undertaken. RESULTS: We have developed a continuous-flow blood pump of small size (207 g) and low power requirement (6.5 watts) producing 5 L/min flow with low hemolysis. CONCLUSIONS: This pump has the potential to be the basis of an innovative ventricular assist system.

Chronic nonpulsatile blood flow. I. Cerebral autoregulation in chronic nonpulsatile biventricular bypass: carotid blood flow response to hypercapnia.

To investigate the response of the carotid blood flow and general circulation to hypercapnia in chronic nonpulsatile blood flow, we performed 18 carbon dioxide gas inhalation studies on three calves undergoing a centrifugal biventricular bypass with ventricular fibrillation. An ultrasonic flow probe was put on the carotid artery during biventricular bypass pump implantation, and pump flows were maintained at 90, 100, and 120 ml/kg per minute for 1 week each. The carbon dioxide inhalation studies were performed twice a week. Hypercapnia was induced by administering pure carbon dioxide gas through a nasal tube at flow rates of 0, 5, 7.5, 10, 12.5, and 15 L/min for 5 minutes each at three different nominal pump flow rates, and the resultant arterial blood gas and hemodynamic changes were recorded. No significant correlation existed between the carotid blood flow and mean aortic pressure, which varied from 70 to 140 mm Hg, but the carotid blood flow correlated significantly (p < 0.01) with the systemic pump flow rate. A significant (p < 0.01) linear relationship was found between the carotid blood flow and arterial carbon dioxide tension. For each 1 mm Hg change in arterial carbon dioxide tension, there was a 2.8 % change in the carotid blood flow. The percent changes in the carotid blood flow in response to arterial carbon dioxide tension were calculated as 2.9%, 3.7%, and 2.5% for each 1 mm Hg change in arterial carbon dioxide tension at pump flows of 90, 100 and 120 ml/kg per minute. No significant differences in the carotid blood flow response to hypercapnia were detected among the three systemic pump flow rates. These results thus suggested that chronic nonpulsatile blood flow had no detrimental effects on cerebral autoregulation.

Effect of respiration on the arterial pressure wave in calves with nonpulsatile biventricular bypass.

The possibility of idioperipheral pulsation in calves with chronic nonpulsatile biventricular bypass has been previously reported. To test the hypothesis that both spontaneous respiration and mechanical ventilation are the cause of pulsation in the arterial pressure recording in calves with nonpulsatile biventricular bypass, two protocols were used to generate data. Both right and left ventricles were bypassed using two centrifugal pumps followed by electrical fibrillation in seven calves. In protocol 1, the spontaneous respiratory rate was increased in awake calves by carbon dioxide gas inhalation. In protocol 2, the animals were anesthetized and intubated to control ventilation, and data were collected as the ventilation rate or tidal volume was changed in a stepwise fashion, as well as when the ventilation was stopped for several seconds. The results of protocol 1 showed good correlation between the arterial pulse rate and spontaneous respiratory rate (Y = 1.03 X, r = 0.822, p < 0.001). The results of protocol 2 showed that the relationship between the ventilation rate and the arterial pulse cycle was identical (Y = X, r = 1.000, p < 0.001), and no pulsation was seen when the ventilator was stopped. The authors conclude that the arterial pulsation observed in calves with nonpulsatile biventricular bypass is caused by respiration.

The Cleveland Clinic-Nimbus total artificial heart. In vivo hemodynamic performance in calves and preclinical studies.

In vitro function of the Cleveland Clinic-Nimbus electrohydraulic total artificial heart met National Heart, Lung, and Blood Institute hemodynamic guidelines for such devices. In a series of in vivo experiments, we implanted the total artificial heart in eight calves (mean weight 87 kg), one for a short-term experiment and seven for long-term experiments. The mean blood flow during support was 7.7 +/- 1.6 L/min with left atrial pressure 13 +/- 6 mm Hg, right atrial pressure 13 +/- 4 mm Hg, and aortic pressure 97 +/- 9 mm hg. Maximum pump flow (9.6 L/min) occurred after 4 days of support as a result of the high resting cardiac output of the animals. A 10% to 15% right pump stroke-volume limit effectively balanced atrial pressures, and afterload insensitivity was confirmed by the in vivo studies. Calves tolerated treadmill exercise studies well, with an average duration of 22 minutes and an average top speed of 2.1 mph. The experiments were terminated after 1 day to 120 days of support (mean 32 days). Most experiments were terminated as a result of correctable mechanical problems. In a separate study of six adult human patients undergoing orthotopic cardiac transplantation, five showed an excellent fit for the Cleveland Clinic-Nimbus total artificial heart. Further studies using chest roentgenograms, chest measurements, and transesophageal echocardiography should help predict fit of the total artificial heart in potential candidates. Initial candidates for a "vented-electric" version of the Cleveland Clinic-Nimbus total artificial heart are patients for whom univentricular (left ventricular assist device) support is not appropriate, but who require mechanical support as a bridge to cardiac transplantation.

Blood coagulability and hematological changes in calves with chronic centrifugal biventricular bypass pumps.

A Hemadyne centrifugal pump was used to determine the effects of long-term circulatory assist on platelet number and aggregability. Five calves were supported on biventricular bypass with a pair of Hemadyne centrifugal pumps. On the 7th postoperative day (POD), the heart was fibrillated after pump flow rates were increased to compensate for the total cardiac output. The animals were anticoagulated with heparin throughout the study and adjustments were made to maintain the activated clotting time at between 180 and 220 sec. To maintain the activated clotting time level in the therapeutic range, it was necessary to gradually increase the heparin dose with time due to the developing heparin tolerance. Fibrinogen levels increased significantly on the 3rd postoperative day when compared to the preoperative control value. Platelet aggregation was measured using a whole blood impedance method with adenosine diphosphate and collagen as agonists. Platelet function was found to be significantly depressed following the biventricular bypass procedure (P < 0.05 on the 7th, 14th, and 21st POD). Platelet numbers were significantly decreased compared to the preoperative control values, being the lowest on the 1st and 3rd POD, recovering to 76% of the control values by the 7th POD, and then gradually decreasing to the 55% level by the 28th POD. These results show that, when tested for up to 4 weeks following surgery, both platelet number and function in calves supported with centrifugal pumps remain depressed. Despite the progressive heparin tolerance and suppressed platelet number and function, none of the animals showed abnormal bleeding tendencies on biventricular bypass.

Catastrophic vasoconstriction after protamine reversal of heparin anticoagulation in the goat.

Catastrophic pulmonary vasoconstriction after protamine reversal of heparin anticoagulation was observed frequently in goats that underwent centrifugal biventricular bypass pump implant. Care should be taken when protamine is used to reverse heparin anticoagulation in goats.