Delivery of gelfoam-enabled cells and vectors into the pericardial space using a percutaneous approach in a porcine model.

Intrapericardial drug delivery is a promising procedure, with the ability to localize therapeutics with the heart. Gelfoam particles are nontoxic, inexpensive, nonimmunogenic and biodegradable compounds that can be used to deliver therapeutic agents. We developed a new percutaneous approach method for intrapericardial injection, puncturing the pericardial sac safely under fluoroscopy and intravascular ultrasound (IVUS) guidance. In a porcine model of myocardial infarction (MI), we deployed gelfoam particles carrying either (a) autologous mesenchymal stem cells (MSCs) or (b) an adenovirus encoding enhanced green fluorescent protein (eGFP) 48 h post-MI. The presence of MSCs and viral infection at the infarct zone was confirmed by immunoflourescence and PCR. Puncture was performed successfully in 16 animals. Using IVUS, we successfully determined the size of the pericardial space before the puncture, and safely accessed that space in setting of pericardial effusion and also adhesions induced by the MI. Intrapericardial injection of gelfoam was safe and reliable. Presence of the MSCs and eGFP expression from adenovirus in the myocardium were confirmed after delivery. Our novel percutaneous approach to deliver (stem-) cells or adenovirus was safe and efficient in this pre-clinical model. IVUS-guided delivery is a minimally invasive procedure that seems to be a promising new strategy to deliver therapeutic agents locally to the heart.

Prostaglandin synthesis inhibitor improves hypotension during normothermic cardiopulmonary bypass.

Hypotension is a major systemic side effect during cardiopu monary bypass (CPB), especially at normothermia. We previously reported that prostaglandin (PG) might play a substantial role in hypotension. The purpose of this study was to clarify whether a PG synthesis inhibitor (PGSI) could improve hypotension during CPB. Thirty-six patients undergoing cardiac surgery with normothermic CPB (35-37 degrees C) were divided into two groups: a PGSI group (n = 18), whose members wer given a PGSI before and during CPB, and a control group (n = 18). In both groups, perfusion flow was sufficient and pressure was maintained at above 45 mm Hg by infusion of metaraminol, a vasoconstrictor. The mean arterial pressure throughout CPB was significantly higher in the PGSI group than in the control group (57 +/- 4 vs. 48 +/- 3 mm Hg, p < 0.01), whereas the dose of infused metaraminol was significantly lower in the PGSI group (13 +/- 7 vs. 21 +/- 6 mg, p < 0.01). The blood base excess was not significantly different (1.0 +/- 1.6 vs. 1.7 +/- 1.9 mmol/L, p = 0.28), and urine output was significantly higher in the PGSI group (503 +/- 179 vs. 354 +/- 112 ml/hr, p < 0.01). In conclusion, PGSI can improve hypotension during CPB and increase urine output without impairing peripheral circulation.

Regulation of perfusion pressure during cardiopulmonary bypass using sevoflurane.

In hypothermic cardiopulmonary bypass (CPB), various vasodilators are used to control the perfusion pressure. These agents, however, often decrease the pressure excessively, and the low perfusion pressure may persist until the end of CPB. In this study we evaluate the safety and characteristics of the regulation of perfusion pressure during CPB using a volatile anesthetic, sevoflurane which has an extremely low partition coefficient. Twenty adult patients who underwent cardiac surgery were studied. Sevoflurane was applied by a vaporizer inserted into the oxygenator gas supply line. Pump flows were fixed at 2.4 L/min/m2 during the hypothermic period. Sevoflurane concentration was adjusted to keep mean arterial pressure (MAP) between 40 and 70 mmHg during CPB. Hemodynamic and metabolic parameters were measured and compared to the group we previously treated with chlorpromazine. In all cases, MAP could be maintained adequately. In the sevoflurane group, systemic vascular resistance indices (SVRI) during the rewarming period and at the end of CPB were higher, and doses of norepinephrine needed at the end of CPB were significantly lower than in the chlorpromazine group. The regulation of perfusion pressure during CPB using sevoflurane was safe and could easily maintain adequate SVRI.

Oxygen metabolism under various bypass flow conditions during cardiopulmonary support in awake goats.

Despite its wide clinical application, patient recovery from cardiopulmonary support (CPS) is not necessarily satisfactory. To clarify what influence CPS has on organ perfusion, we investigated the oxygen metabolism under various bypass flow (BF) conditions in a series of chronic animal CPS experiments. The CPS system, which consists of a pulsatile ventricular assist device and a compact artificial lung was installed without anesthesia in 6 adult goats weighing 49-51 kg. BF was adjusted stepwise from 0% to 50%, 75%, 90%, and 100% of total systemic blood flow (TSF) by balancing the pulmonary arterial flow. The animals' TSF and oxygen delivery (DO2) were sufficiently maintained throughout the experiments. The oxygen consumption (VO2) and the oxygen extraction rate (ExO2) increased from 178+/-14 to 342+/-19 ml/min, and from 28+/-2% to 64+/-1%, respectively, in proportion to the increase of CPBF dependency from 0% to 100%. The blood lactate level did not change appreciably even at 90% BF from 5.7+/-0.3 to 11.2+/-1.2 mg/dl, but drastically elevated to 23.5+/-4.6 mg/dl at the total bypass. This indicates that CPS leads to a relative lack of oxygen and can induce organ dysfunction due to increasing VO2 and ExO2 in proportion to the increase of BF dependence even if TSF and DO2 are sufficiently maintained.

Hemodynamic effects of prostaglandins and catecholamines in graded reduction of pulmonary flow during venoarterial bypass in awake goats.

The roles of prostaglandins and catecholamines in the hypotensive hemodynamic change during cardiopulmonary support with a venoarterial bypass (VAB) were investigated in a series of chronic animal experiments of gradually reduced pulmonary arterial blood flow (PAF). The VAB system consisted of a pulsatile ventricular assist device, an artificial lung, and the right atrium uptake and descending aorta return cannulae in four adult goats weighing 49-51 kg. The PAF was adjusted to 50, 10, and 0% of the total systemic blood flow. Indomethacin, an inhibitor of prostaglandin production; phentolamine, an alpha-antagonist of catecholamine; and noradrenaline, an agonist of catecholamine were administered at each PAF condition. The mean aortic pressure (mAoP) and the systemic vascular resistance decreased in proportion to the decrease in PAF. Indomethacin increased the mAoP at all PAF conditions, indicating a relationship between prostaglandins and hypotension. Phentolamine decreased the mAoP at all PAF conditions, indicating a normal response of catecholamine receptors. However, noradrenaline increased the mAoP at 50 and 10% PAF, but did not appreciably increase the value at 0% PAF, indicating complete response of catecholamine receptors to endogenous catecholamines at 0% PAF only. In conclusion, prostaglandins play a substantial role in hypotension during VAB, and catecholamines may subsequently increase in compensation for extreme hypotension.

Prolonged nonpulsatile left heart bypass with reduced systemic pulse pressure causes morphological changes in the aortic wall.

We investigated the morphological changes in the aorta due to reduced systemic pulse pressure in prolonged nonpulsatile left heart bypass (LHB). Nineteen adult goats were divided into 3 groups, the nonpulsatile group in which nonpulsatile LHB was conducted, the pulsatile group in which pulsatile LHB was conducted, and the control group used as the normal control. The average aortic pulse pressures were 12, 47, and 37 mm Hg, respectively. The descending aorta was subjected to morphological examination. In the nonpulsatile group, the wall was significantly thinner, and the volume ratio of smooth muscle cells (SMCs) was much lower. In terms of the SMC type classification, the proportion of SMCs with low activity and low contractility was higher, and the cell density of the SMCs was increased compared to those in the other groups. These results indicate that prolonged nonpulsatile LHB causes morphological atrophic changes in the aorta.

Hemodynamic and humoral conditions in stepwise reduction of pulmonary blood flow during venoarterial bypass in awake goats.

The effects of reduced pulmonary arterial blood flow (PAF) during venoarterial bypass (VAB) on hemodynamic and humoral conditions were investigated in a series of experiments in a chronic animal model. A biventricular bypass system was installed in five adult goats weighing 49.8 +/- 1.1 kg. Two weeks later, the extracorporeal circuitry was changed to VAB without anesthesia. The PAF was reduced stepwise from 100% to 50, 25, 10, and 0% of total systemic flow. The mean aortic pressure and systemic vascular resistance decreased from 110 +/- 14 to 66 +/- 3 mmHg and from 1,288 +/- 77 to 740 +/- 73 dyne.sec/cm5, respectively, in proportion to the decrease in PAF from 100 to 0%. The prostaglandin E2 concentration increased from 1.5 +/- 0.6 to 8.8 +/- 0.6 pg/ml following the decrease in PAF from 100 to 0%. The renin-angiotensin system increased in proportion to the decrease in PAF. In contrast, the epinephrine and norepinephrine concentrations (60 +/- 10 and 227 +/- 80 pg/ml, respectively, at 100% PAF) did not change appreciably even at 10% PAF, but were markedly elevated to 335 +/- 117 and 2,088 +/- 1,503 pg/ml at 0% PAF. The antidiuretic hormone level similarly changed. In conclusion, decrease in PAF during VAB exerts significant effects on hemodynamics in a proportional manner and on vasoactive humoral factors in a diverse manner.

In vitro and in vivo heat dissipation of an electrohydraulic totally implantable artificial heart.

The authors evaluated the heat transfer characteristics of an electrohydraulic totally implantable artificial heart (EH-TAH) developed at our institute. In three in vitro experiments, the heat dissipation of the EH-TAH was investigated. First, the EH-TAH was connected to a closed mock circuit filled with 1 L of saline, and driven at an input power of 20 W. The estimated heat conducted to the blood was approximately 10.3 W, which was almost half of the input power. Second, we simulated heat transfer with the circulation of a calf by using a heat exchanger. The amount of heat dissipating directly from the EH-TAH surface was calculated to be 10 W. Third, the temperature of the actuator examined with thermography was found to be almost uniform, and no prominent high temperature area was observed. In an in vivo study, the EH-TAH was implanted for 10 days in a calf weighing 62 kg. The input power was 18 +/- 2 W, the temperature of the actuator-tissue contacting surface was 39.4 +/- 0.8 degrees C, and that of the pump blood chamber was 39.8 +/- 0.4 degrees C. This slight temperature elevation was thought to be attributable to heat dissipation to the blood. On histologic study of the chest wall and the lung in contact with the actuator, vascularized connective tissue envelopes were observed, but unfavorable side effects, such as tissue necrosis, were not observed. These results suggest that the thermal effect of this system is acceptable at the input power used.

In vitro and in vivo evaluation of a left-right balancing capacity of an interatrial shunt in an electrohydraulic total artificial heart system.

The authors evaluated the basic performance of an interatrial shunt (IAS) made by punching a hole in the atrial septum, in accommodating the left-right imbalance in our electrohydraulic total artificial heart (EHTAH) system. In an in vitro study conducted in a closed mock circuit connected with the EHTAH, the interatrial pressure gradient changed in compliance with the amount of bronchial flow and the size of the IAS. The IAS of 4.4 mm diameter or larger maintained the interatrial pressure gradient within physiologically permissible limits when the amount of bronchial flow was 5% of cardiac output or less. A left-to-right one-way valve made of a piece of pericardium, a possible option in this IAS method, successfully prevented right-to-left reverse shunt flow through the IAS. In a chronic in vivo study using a calf implanted with the EHTAH for 10 days, a 4.5 mm IAS without the one-way valve demonstrated satisfactory dynamic left-right balancing capacity with a stable interatrial pressure gradient of 4 +/- 1 mmHg over a wide range of atrial pressures. No thrombus was found in or around the IAS at autopsy. The authors conclude that the IAS is a simple and promising means of left-right balancing in the EHTAH system.

Early changes in circulating blood volume and volume regulating humoral factors after implantation of an electrohydraulic total artificial heart.

The early changes in circulating blood volume (CBV) and volume regulating humoral factors after implantation of an electrohydraulic total artificial heart (EH-TAH) were investigated in a calf and compared with results in a sham operated control calf. CBV was measured by the dye dilution method using indocyanine green. CBV and humoral factors were periodically investigated. In the EH-TAH implanted calf, the cardiac output was estimated at 6-7 L/ min (94-109 ml/kg/min), and the aortic pressure and aerobic metabolic condition were favorable. Nevertheless, the CBV was increased to 132 and 168% of the pre-operative value (range in the control calf, 83-103%) on post operative days 4 and 8, respectively. The atrial natriuretic peptide level on days 2, 5, and 7 was 23, 170, and 240 (in the control calf, 19-61) pg/ml, respectively, and the antidiuretic hormone level was 7.3, 2.0, and 1.3 (0.5-1.3) pg/ ml, respectively. The plasma renin activity was 3.2, 3.7, and 3.1 (0.5-0.3) ng/ml/hr, respectively. The angiotensin-I and angiotensin-II levels were also increased in the EH-TAH implanted calf. It is concluded that significant water retention occurs even at sufficient cardiac output early after EH-TAH implantation. The changes in humoral factors are suggested to arise secondary to the increased CBV or other unknown factors.

Characteristics of mixed venous oxygen saturation and physical activity as parameters for artificial heart control.

Mixed venous oxygen saturation (SvO2) and physical activity (PhyAc) are practical candidates as parameters of total artificial heart (TAH) control because SvO2 can be measured through a transparent blood pump housing with infrared rays and PhyAc can be calculated from signals of an accelerometer used for rate response pacemakers. Although the methods for measurement of the parameters have already been developed, characteristics of these parameters for TAH control, such as during exercise, are still unclear and were examined in this study. SvO2, cardiac output (CO), and PhyAc were measured as parameters. Multi-stage treadmill exercise tests were performed. Difference values (DVs) from the value at the start of exercise showed better correlation than did absolute values. Correlation coefficients between DV in CO and DV in SvO2 and between DV in CO and PhyAc were high at -0.82 and 0.72, while the time constants for the change of SvO2 and CO to the PhyAc change were 26 and 32 sec. Although the correlation coefficient between the CO and SvO2 was higher than that between CO and PhyAc, PhyAc responded more quickly to the speed change compared with the response of SvO2 and CO. It was concluded that SvO2 and PhyAc were useful parameters with different characteristics for TAH control during exercise.

Morphologic changes of the aortic wall due to reduced systemic pulse pressure in prolonged non pulsatile left heart bypass.

The morphologic changes of the aortic wall due to reduced systemic pulse pressure in prolonged non pulsatile left heart bypass (LHB) were investigated. Sixteen adult goats were divided into three groups: the non pulsatile group in which non pulsatile LHB was conducted for 137 days on average, the pulsatile group in which pulsatile LHB was conducted for 79 days on average, and the control group used as the normal control. The average aortic pulse pressures were 12, 48, and 37 mmHg, respectively. At the end of the experiments, the descending aorta was excised and subjected to morphologic examination. The wall thickness of the aorta in the non pulsatile group (1.4 mm) was significantly thinner than that in the pulsatile group (2.2 mm) and the control group (2.0 mm), and the volume ratio of smooth muscle cells (SMC) in the non pulsatile group (37%) was lower than that in the pulsatile group (48%) and the control group (49%). In SMC classification, the proportion of SMC with low activity and low contractility in the non pulsatile group (57%) was high as compared with that in the pulsatile (2%) and control (5%) groups. These results strongly indicate that prolonged non pulsatile LHB causes substantial morphologic changes in the aorta.

Noninvasive pump flow estimation of a centrifugal blood pump.

A flow rate estimating method was investigated for a centrifugal blood pump developed in our institute. The estimated flow rate was determined by the power consumption, the rotating speed of the motor, and the hematocrit value. The power consumption and the rotating speed of the motor were measured with a wattmeter. The examinations were performed in a closed mock loop filled with goat blood with hematocrit values of 21.5%, 28%, 34%, and 42%. Measured values of blood viscosity were 2.47, 3.09, 3.71, and 5.07 mPa.s at a share rate of 37.5/s, respectively. A linear correlation between the power consumption and the pump flow rate was observed in all hematocrit values. But variations in hematocrit caused a difference in the flow rate up to 1.1 L/min at the same power consumption and rotating speed. Effects of blood viscosity on the flow estimation were corrected by the hematocrit value. The value of the coefficient of determination, R2, between the estimated flow rate and the measured flow rate was 0.988. These results may indicate that the flow estimating method calculated by the power consumption of the motor, the rotating speed, and the hematocrit value is useful in the clinical situation.

Long-term evaluation of a nonpulsatile mechanical circulatory support system.

Antithrombogenicity of a centrifugal pump (CP) developed in our institute is provided by a central balancing hole (BH) in the impeller. A current CP, the National Cardiovascular Center (NCVC)-2, was ameliorated to improve antithrombogenicity, whereby the BH diameter was widened to improve self washout flow velocity, and an edge of the thrust bearing was rounded off to minimize flow separation. Effects of these modifications were assessed in a long-term in vivo experiment. The antithrombogenicity, hemolytic property, and mechanical durability of the NCVC-2 were investigated in 3 goats. The NCVC-2 was installed paracorporeally between the left atrium and the aorta and driven as long as possible at rotating speeds of about 2,800 rpm. The NCVC-2 ran for 50, 200, and 367+ days. The mean bypass flow rates were 6.8, 5.0, and 5.3 L/min, respectively. Creatinine, blood urea nitrogen (BUN), glutamic-oxaloacetic transaminase (GOT), and glutamic-pyruvic transaminase (GPT) did not increase until one week before termination. Plasma free hemoglobin was kept to a level less than 15 mg/dl, except for the last week of the second case. These results indicate that the NCVC-2 has excellent antithrombogenicity, an acceptable hemolytic property and the necessary durability for prolonged use.

Influence of non pulsatile systemic circulation on tissue blood flow and oxygen metabolism.

Influence of non pulsatile systemic circulation on oxygen metabolism was examined regarding tissue perfusion in 12 adult goats weighing from 46 to 55 kg. Under general anesthesia, a flow character, changeable total left heart bypass circuit consisting of pulsatile and non pulsatile pumps was installed through a left thoracotomy. Systemic flow was converted from pulsatile to non pulsatile in 7 of 12 animals, and in the reverse order in the other 5, by changing the driving pump. Esophageal mucosal blood flow was determined by a colored microsphere method that estimated tissue blood flow at the pre capillary level. Esophageal intramucosal pH was evaluated with a silicone balloon tonometer catheter surgically placed in the submucosal space. Hemodynamic and arterial blood gas parameters were unchanged at flow mode conversion. Although oxygen delivery was comparable between pulsatile and non pulsatile circulation, oxygen extraction ratio was lower and venous oxygen saturation was higher in non pulsatile than pulsatile circulation. Although statistically not significant, serum lactate level tended to be higher with non pulsatile circulation. No difference was observed in esophageal mucosal blood flow between pulsatile and non pulsatile circulation, whereas intramucosal pH, which strongly correlated with arterial pH regardless of the flow mode, was significantly lower under non pulsatile than pulsatile conditions. In conclusion, systemic oxygen uptake is less efficient in non pulsatile than pulsatile circulation in the setting of an acute experiment using animals, which may be accounted for by the disparity between the pre capillary blood flow and actual tissue oxygen metabolism.

Importance of metabolic function of the natural lung evaluated by prolonged exclusion of the pulmonary circulation.

It is generally considered that the natural lung metabolizes various vasoactive substances through the pulmonary circulation. However, the influences of bypassing or eliminating the pulmonary circulation have not been fully elucidated, especially for prolonged periods. In this study, we performed total cardiopulmonary bypass and exclusion of the pulmonary circulation for up to 336 hr in awake goats to clarify the importance of the metabolic function of the lung. In seven adult goats, biventricular bypass with a pulsatile ventricular assist system was first established. After 2 weeks, the biventricular bypass was converted to total cardiopulmonary bypass without anesthesia. Adequate gas exchange and perfusion support were achieved in all animals. However, the institution of total cardiopulmonary bypass led to marked decreases in the mean aortic pressure and systemic vascular resistance, and they remained low thereafter. The arterial levels of prostaglandin E2 and norepinephrine, which are inactivated naturally through the pulmonary circulation, increased remarkably. These results indicate that the natural lung plays an important role in controlling systemic circulation by metabolizing various vasoactive substances. Understanding the non respiratory function of the natural lung is of prime importance for advancement of lung and heart-lung support.

Development of an integrated artificial heart-lung device for long-term cardiopulmonary support.

An integrated artificial heart-lung device has been developed as a long-term cardiopulmonary support system. The device is composed of gas exchange and pumping units. The gas exchange unit consists of a special hollow fiber membrane that can prevent serum leakage. The entire blood contacting surface of the gas exchange unit is treated with covalent heparin bonding. The pumping unit consists of two pusher-plate artificial hearts joined to each end of the artificial lung unit. The core size and priming volume of the device are 11 x 14 x 17 cm and 400 ml, respectively. In in vitro evaluation, the device exhibited a maximum output of 7.0 L/ min, with a pressure gradient of 10 mmHg per 1 L/min flow rate. In acute in vivo evaluation with adult goats, the device satisfactorily replaced the animals' circulation and respiration for 6 hr. Pumping output ranged from 5.5 to 6.0 L/min, whereas PaO2 and PaCO2 were kept above 500 mmHg and below 30 mmHg, respectively. The device demonstrated prominent thrombus resistant properties in nonheparin animal use trials. These results indicate that the integrated artificial heart-lung device has a potential to be a long-term cardiopulmonary support system that can be used with minimal anticoagulant therapy.

Development of a membrane oxygenator for long-term respiratory support and its experimental evaluation in prolonged ECMO.

The authors developed a new membrane oxygenator (MO) for long-term respiratory support and evaluated its performance in animal experiments for as long as 336 hr. The MO, with a membrane area of 1.2 m2 and priming volume of 140 ml, is compact and designed to be interposed in a ventricular assist system (VAS) conduit. It is made with a novel hollow fiber membrane, in which micropores are blind-ended so that serum leakage can be prevented during prolonged use. The blood contacting surface of the MO is heparinized with a newly developed covalent bonding technique that ensures good thrombus resistance. In vivo evaluation with five adult goats was performed by installing the MO into a venoarterial or venovenous bypass circuit. No systemic anticoagulant therapy was used, except for a heparin-added fluid infusion to keep the pressure monitoring lines open (2-3 U/kg/hr). Throughout the experiments, no plasma leakage was observed, and gas transfer rates were maintained in a satisfactory range. Platelet counts did not decrease to less than 60% of levels before bypass, and hemolysis was negligible. The levels of coagulation parameters including fibrinogen, fibrin degradation products (FDP), antithrombin III (AT III), antiplasmin, prothrombin time (PT) and activated partial thromboplastic time (APTT) remained within physiologic ranges and relatively constant. At the end of the evaluation, no thrombus formation was noted in three of five MOs. These results suggest that this MO is a promising device for long-term respiratory support.

Altered oxygen metabolic conditions associated with increased norepinephrine levels in a nonpulsatile systemic circulation.

Change in oxygen metabolic conditions accompanying the conversion of systemic flow from pulsatile to nonpulsatile (from P-mode to N-mode) was investigated in association with blood norepinephrine levels. Total left heart bypass was instituted through a left thoracotomy under general anesthesia in 10 adult goats. Pulsatile and nonpulsatile pumps were incorporated in the circuit in parallel, and the flow character was rapidly converted from the P-mode to the N-mode. Norepinephrine levels increased significantly after the conversion, from 222 +/- 54 pg/ml to 285 +/- 65 pg/ml. While oxygen delivery (DO2) was kept constant, the oxygen extraction ratio significantly decreased, from 21 +/- 3% to 16 +/- 3%, and venous oxygen saturation (SvO2) significantly increased, from 77 +/- 6% to 84 +/- 6% after depulsation. The serum lactate level was significantly higher in the N-mode than the P-mode (P-mode: 35 +/- 2 mg/dl, N-mode: 45 +/- 5 mg/dl). Strong positive and negative correlations of norepinephrine levels were observed with oxygen extraction ratio and SvO2, respectively, whereas norepinephrine levels did not correlated with DO2. Regression lines in these correlations unveiled higher oxygen uptake in the P-mode than the N-mode at the same norepinephrine level. These results indicate that, in the setting of an acute animal experiment, oxygen uptake is less efficient with the absence of pulsatility, and the higher norepinephrine concentration functioned to tune the oxygen metabolism in the initial stage of nonpulsatile systemic circulation.

Tandem operation of a turbo blood pump (BP-80--type centrifugal pump) to reduce hemolysis.

When operating turbo blood pumps in tandem, the strength of shear stress is reduced, but the exposure duration of the stress is increased. The purpose of this experiment was to compare the degree of contribution of these two factors on hemolysis as well as to evaluate the effectiveness of the tandem operation of turbo blood pumps. Tandem operation of two Bio-pumps (BP-80; Medtronics Bio-medicus, Inc., Eden Prairie, Minnesota, U.S.A.) were compared with single operation of a BP-80 in in vitro hemolysis tests in three different driving conditions, that is, pumping heads of 200, 350, and 500 mm Hg under a pump flow rate of 5 L/min. The Allen's hemolytic indexes of the tandem operation at pumping heads of 200, 350, and 500 mm Hg were 0.014, 0.020, and 0.080 mg/dl, respectively. The hemolytic indexes of the single operation at pumping heads of 200, 35, and 500 mm Hg were 0.014, 0.056, and 0.12 mg/dl, respectively. These results indicate that tandem operation is a useful method of reducing hemolysis with the BP-80 under high pumping heads and that the effect on hemolysis of exposure to higher shear stresses may be more serious than that of longer durations of exposure to shear stress in turbo blood pumps.