An implantable seal-less centrifugal pump with integrated double-disk motor

Development of a Pivot Bearing Supported Sealless Centrifugal Pump for Ventricular Assist

Since 1991, in our laboratory, a pivot bearing-supported, sealless, centrifugal pump has been developed as an implantable ventricular assist device (VAD). For this application, the configuration of the total pump system should be relatively small. The C1E3 pump developed for this purpose was anatomically compatible with the small-sized patient population. To evaluate an-tithrombogenicity, ex vivo 2-week screening studies were conducted instead of studies involving an intracorpore-ally implanted VADs using calves. Five paracorporeal LVAD studies were performed using calves for longer than 2 weeks. The activated clotting time (ACT) was maintained at approximately 250 s using heparin. All of the devices demonstrated trouble-free performances over 2 weeks. Among these 5 studies, 3 implantations were subjected to 1-month system validation studies. There were no device-induced thrombus formations inside the pump housing, and plasma-free hemoglobin levels in calves were within the normal range throughout the experiment (35, 34, and 31 days). There were no incidents of system malfunction. Subsequently, the mass production model was fabricated and yielded a normalized index of hemolysis of 0.0014, which was comparable to that of clinically available pumps. The wear life of the impeller bearings was estimated at longer than 8 years. In the next series of in vivo studies, an implantable model of the C1E3 pump will be fabricated for longer term implantation. The pump-actuator will be implanted inside the body; thus the design calls for substituting plastic for metallic parts.

Current status of the gyro centrifugal blood pump--development of the permanently implantable centrifugal blood pump as a biventricular assist device (NEDO project)

The New Energy and Industrial Technology Development Organization (NEDO) project was started in 1995. The goal is the development of a multipurpose, totally implantable biventricular assist device (BVAD) that can be used for any patient who suffers from severe heart failure. Our C1E3 (two-week pump) centrifugal pump, called the Gyro pump, has three design characteristics: a magnetic coupling and double pivot bearing system, an eccentric inlet port, and secondary vanes on the bottom of the impeller. The pump was miniaturized. The C1E3 evolved into the NEDO PI-601, a totally implantable centrifugal pump for BVAD. The current NEDO PI-710 pump (five-year pump) system includes a centrifugal pump with pivot bearings, a hydraulically-levitated impeller, an rpm-controlled miniaturized actuator (all-in-one actuator plus controller), an emergency clamp on the left outflow, and a Frank-Starling-type flow control. The final mass production model is now finalized, and the final animal study and two-year endurance studies are ongoing.

Mechanical circulatory support for infants and children with cardiac disease

Mechanical circulatory support is assuming an expanding role in the practice of congenital cardiac surgery. Extracorporeal membrane oxygenation and centrifugal ventricular assist devices are still the mainstay of mechanical circulatory support for children; however, newly developed pulsatile, paracorporeal ventricular assist devices designed for pediatric applications are achieving increased utilization. In addition, several new, continuous flow devices that are under development as fully implantable systems for adults, ultimately may be useful for pediatric patients.

Development of the Baylor Gyro permanently implantable centrifugal blood pump as a biventricular assist device

The Baylor Gyro permanently implantable centrifugal blood pump (Gyro PI pump) has been under development since 1995 at Baylor College of Medicine. Excellent results were achieved as a left ventricular assist device (LVAD) with survival up to 284 days. Based on these results, we are now focusing on the development of a biventricular assist device (BVAD) system, which requires 2 pumps to be implanted simultaneously in the preperitoneal space. Our hypothesis was that the Gyro PI pump would be an appropriate device for an implantable BVAD system. The Gyro PI 700 pump is fabricated from titanium alloy and has a 25 ml priming volume, pump weight of 204 g, height of 45 mm, and pump diameter of 65 mm. This pump can provide 5 L/min against 100 mm Hg at 2,000 rpm. In this study, 6 half-Dexter healthy calves have been used as the experimental model. The right pump was applied between the infundibular of the right ventricle and the main pulmonary artery. The left pump was applied between the apex of the left ventricle and the thoracic descending aorta. As for anticoagulation, heparin was administered at the first postoperative week and then converted to warfarin sodium from the second week after surgery. Both pump flow rates were controlled maintaining a pulmonary arterial flow of less than 160 ml/kg/min for the sake of avoidance of pulmonary congestion. Blood sampling was done to assess visceral organ function, and the data regarding pump performance were collected. After encountering the endpoint, which the study could not keep for any reasons, necropsy and histopathological examinations were performed. The first 2 cases were terminated within 1 week. Deterioration of the pump flow due to suction phenomenon was recognized in both cases. To avoid the suction phenomenon, a flexible conduit attached on the inlet conduit was designed and implanted. After using the flexible inflow conduit, the required power and the rotational speed were reduced. Furthermore, the suction phenomenon was not observed except for 1 case. There was no deterioration regarding visceral organ function, and pulmonary function was maintained within normal range except for 1 case. Even though the experimental animal survived up to 45 days with the flexible inflow conduit, an increase in power consumption due to thrombus formation behind the impeller became a problem. Lower rotational speed, which was probably produced by the effectiveness of the flexible inflow conduit, was speculated to be one of the reasons. And the minimum range of rotational speed was 1,950 rpm in these 6 BVAD cases and the previous 3 cases of LVAD. In conclusion, 6 cases of BVAD implantation were performed as in vivo animal studies and were observed up to 45 days. The flexible inflow conduit was applied in 4 of 6 cases, and it was effective in avoiding a suction phenomenon. The proper rotational speed of the Gyro PI 700 pump was detected from the viewpoint of antithrombogenicity, which is more than 1,950 rpm.

Future devices and directions

This article summarizes the status of left ventricular assist devices currently in the stages of bench testing, animal experiments, and pilot clinical trials. The major design features and estimate of costs for 17 devices are described under 3 major categories of indications for use: destination therapy, bridge to transplant, and bridge to recovery. A sleeved piston pump located in the aorta and a unique, magnetically suspended centrifugal pump are described in the destination therapy section. Eight centrifugal and 4 axial flow devices are listed in the bridge to transplant category, and an external cup and a very low-cost centrifugal pump with a left atrium-to-aorta circuit are described in the bridge to recovery section. The key design features of the future, which will be required for success in both the clinical and marketplace arenas, will be simplicity, safety, low-power requirements, and low cost.

In vivo evaluation of the miniaturized Gyro centrifugal pump as an implantable ventricular assist device

A miniaturized Gyro centrifugal pump has been developed to be incorporated into a totally implantable artificial heart. The Gyro PI (permanently implantable) model is a pivot bearing supported centrifugal pump with a priming volume of 20 ml. With the miniaturized actuator, the pump-actuator package has a height of 53 mm, a diameter of 65 mm, and a displacement volume of 145 ml. To evaluate the hemocompatibility and efficiency of the Gyro PI pump system, a plastic prototype (Gyro PI-601) was implanted into a bovine model as a left or right ventricular assist device (LVAD or RVAD), bypassing from the left ventricular apex to the descending aorta or from the right ventricular infundibulum to the main pulmonary artery. The calves were anticoagulated with heparin to maintain activated clotting times from 150 to 200 s. Four calves were supported for 23, 24, and 50 days in the LVAD studies, and 40 days in the RVAD study. The first calf died due to intrathoracic bleeding associated with sepsis. The second calf was euthanized for a low flow rate less than 2 L/min due to an obstructed inflow with growing pannus. The third and fourth calves were euthanized as scheduled. Renal and hepatic functions remained normal, and plasma free hemoglobin values were less than 8 mg/dL throughout the experiments. The fourth case showed flow rates of 4.83 +/- 0.57 L/min, input power of 6.16 +/- 0.49 W, and the inside temperature of the actuator of 43.5 +/- 0.52 degrees C. The pumps implanted in the fourth calf demonstrated no thrombus formation at the autopsy. These in vivo experiments revealed that the Gyro PI pump can provide adequate flow as an easily implantable, efficient, antithrombogenic, and nonhemolytic centrifugal LVAD or RVAD with miniaturized actuators.

Anatomical consideration for an implantable centrifugal biventricular assist system

A miniaturized pivot bearing-supported centrifugal blood pump (Gyro PI) has been developed as a long-term biventricular assist system (BiVAS). In this study we determined the anatomical configuration of this system using a bovine model. Under general anesthesia, a left lateral thoracotomy was performed to open the chest. Two Gyro PI-601 pumps for left and right assists were placed in the preperitoneal pocket by a subcostal abdominal incision. The left pump could be placed along the dome of the diaphragm just beneath the apex of the left ventricle. The right pump could be placed next to the left pump. The inlet and outlet ports of both pumps penetrated the diaphragm. The inlet port of the left pump, with a length of 55 mm, was inserted directly into the apex of the left ventricle. A woven Dacron graft (150 mm long, 11 mm inner diameter) was placed between the outlet port of the left pump and the descending aorta. As for the right pump, a 100 mm long and 120 degree angled inflow conduit was placed between the inlet port and the right ventricular infundibulum. The outlet port of the right pump was connected to the main trunk of the pulmonary artery using a 90 mm long, 11 mm inner diameter Dacron graft. We could perform biventricular assistance to confirm the anatomical feasibility of the Gyro implantable centrifugal BiVAS.

Flow measurement at the pump head of centrifugal pumps: comparison of ultrasonic transit time and ultrasonic Doppler systems

Determination of blood flow is essential for monitoring rotary blood pumps. However, accurate measurement directly adjacent to the pump housing is difficult because of the highly irregular flow profiles near the fast spinning rotor. Therefore, a specially adapted flow probe based on the ultrasound transit time (USTT) principle was designed to evaluate the flow in centrifugal blood pumps. The probe can be directly mounted at the housing and creates 2 crossed measuring ultrasound beams. The mean value, Qm, of the 2 output signals corresponds to the blood flow and the difference, Qd, correlates to the vorticity of the flow profile in the pump outflow tract. In vitro measurements obtained an accuracy for mean flow values of better than +/-0.6 L/min in extreme working points and for vorticity values even as high as Qd = 3.5 L/min. Because of vorticity, however, the output signal contained considerable noise, and that required the application of a 10 Hz filter. Positioning of the ultrasound (US) beams parallel to the axial direction of the pump was superior to radial positioning. Additional measurement of the flow profile demonstrated that a large vorticity occurred (up to Qd equal to 3.5 L/min), and this vorticity was highly dependent upon the afterload of the pump. In vivo experiments demonstrated the reliability of the method. We concluded that USTT flow measurement can determine blood flow immediately adjacent to the pump housing with sufficient accuracy, and these measurements are superior to those from US-Doppler systems (which cannot handle the vorticity accurately enough) and electromagnetic devices (which lack zero stability).

Development and initial testing of a permanently implantable centrifugal pump

To be able to salvage heart failure patients, the need for an economical permanent ventricular assist device is increasing. To meet this increasing demand, a miniaturized centrifugal blood pump has been developed as a permanently implantable device. The Gyro permanently implantable model (PI-601) incorporates a sealless design with a blood stagnation free structure. The pump impeller is magnetically coupled to the driver magnet in a sealless manner. This pump is atraumatic and antithrombogenic and incorporates a double pivot bearing system. A miniaturized actuator was utilized in this system in collaboration with the University of Vienna. The priming volume of this pump is 20 ml. The overall size of the pump actuator package is 53 mm in height and 65 mm in diameter, 145 ml of displacement volume, and 305 g in weight. Testing to date has included in vitro hydraulic performance and hemolysis. This pump can provide 5 L/min against a 110 mm Hg total pressure head at 2,000 rpm and 8 L/min against 150 mm Hg at 2,500 rpm. The normalized index of hemolysis (NIH) value of this pump was 0.0028 g/100 L at 5 L/min against 100 mm Hg. A preliminary anatomical study revealed the possibility of the implantability of 2 such systems in biventricular bypass at a preperitoneal location. This system is feasible for use as a permanently implantable biventricular assist device.

Use of motor current in flow rate measurement for the magnetically suspended centrifugal blood pump

Indirect measurement of the flow rate of a centrifugal blood pump using the driving motor current was studied. The pump flow rate can be expressed as a function of the motor current under a given motor speed in the absence of energy loss resulting from uncertain mechanical contact friction. The magnetically suspended centrifugal blood pump (MSCP), developed by the collaboration of Kyoto University and NTN Inc., was suitable for the application of this measuring method because the impeller is suspended magnetically inside the pump housing without any mechanical contact. The effect of fluid viscosity on the pump performance was investigated in detail, and it was possible to estimate the pump flow rate and the pressure difference through the pump (from inlet port to outlet port) accurately by monitoring the motor current and speed when the kinematic viscosity of working fluids was known. The kinematic viscosity of working fluids can also be measured with the MSCP. The motor current and motor speed were monitored in a chronic animal experiment, and the estimated flow rate and pressure difference showed good correlation with directly measured data.

Vibration assessment for thrombus formation in the centrifugal pump

To clarify the correlation of vibration and thrombus formation inside a rotary blood pump, 40 preliminary vibration studies were performed on pivot bearing centrifugal pumps. No such studies were found in the literature. The primary data acquisition equipment included an accelerometer (Isotron PE accelerometer, ENDEVCO, San Juan Capistrano, CA, U.S.A.), digitizing oscilloscope (TDS 420, Tektronix Inc., Pittsfield, MA, U.S.A.), and pivot bearing centrifugal pumps. The pump impeller was coupled magnetically to the driver magnet. The accelerometer was mounted on the top of the pump casing to sense radial and axial accelerations. To simulate the 3 common areas of thrombus formation, a piece of silicone rubber was attached to each of the following 3 locations as described: a circular shape on the center bottom of the impeller (CI), an eccentric shape on the bottom of the impeller (EI), and a circular shape on the center bottom casing (CC). A fast Fourier transform (FFT) method at 5 L/min against 100 mm Hg, with a pump rotating speed of 1,600 rpm was used. The frequency response of the vibration sensors used spans of 40 Hz to 2 kHz. The frequency domain was already integrated into the oscilloscope, allowing for comparison of the vibration results. The area of frequency domain at a radial direction was 206 +/- 12.7 mVHz in CI, 239.5 +/- 12.1 mVHz in EI, 365 +/- 12.9 mVHz in CC, and 163 +/- 7.9 mVHz in the control (control vs. CI p = 0.07, control vs. EI p < 0.001, control vs. CC p < 0.001, EI vs. CC p < 0.001, CI vs. CC p < 0.001). Three types of imitation thrombus formations were roughly distinguishable. These results suggested the possibility of detecting thrombus formation using vibration signals, and these studies revealed the usefulness of vibration monitoring to detect thrombus formation in a centrifugal pump.

Flow characteristics and required control algorithm of an implantable centrifugal left ventricular assist device

As the clinical application of LVADs has increased, attempts have been made to develop smaller, less expensive, more durable and efficient implantable devices using rotary blood pumps. Since chronic circulatory support with implantable continuous-flow LVADs will be established in the near future, we need to determine the flow characteristics through an implantable continuous-flow LVAD. This study describes the flow characteristics through an implantable centrifugal blood pump as a left ventricular assist device (LVAD) to obtain a simple non-invasive algorithm to control its assist flow rate adequately. A prototype of the completely seal-less and pivot bearing-supported centrifugal blood pump was implanted into two calves, bypassing from the left ventricle to the descending aorta. Device motor speed, voltage, current, flow rate, and aortic blood pressure were monitored continuously. The flow patterns revealed forward flow in ventricular systole and backward flow in diastole. As the pump speed increased, an end-diastolic notch became evident in the flow profile. Although the flow rate (Q [l/min]) and rotational speed (R [rpm]) had a linear correlation (Q = 0.0042R - 5.159; r = 0.96), this linearity was altered after the end-diastolic notch was evident. The end-diastolic notch is considered to be a sign of the sucking phenomenon of the centrifugal pump. Also, although the consumed current (I [A]) and flow rate had a linear correlation (I = 0.212Q + 0.29; r = 0.97), this linearity also changed after the end-diastolic notch was evident. Based upon the above findings, we propose a simple algorithm to maintain submaximal flow without inducing sucking. To maintain the submaximal flow rate without measuring flow rate, the sucking point is determined by monitoring consumed current according to gradual increases in voltage.