Centrifugal pump failures.

Centrifugal pumps will not pass gross quantities of gaseous emboli due to the nonocclusive nature of the pump. However, retrograde flow can occur under circumstances that include: product malfunctions, low flows, and human errors. Negative pressure created by falling arterial perfusate can draw air into the cannula. Food and Drug Administration (FDA) records about centrifugal pump malfunctions were obtained. Out of 350,000 cases completed with centrifugal pumps over a 23 month period, the FDA received reports of 68 malfunctions, 22 electrical burning smells, and three speed surges, yielding a failure rate of 1 in 3,763 cases. FDA records revealed five death reports and three serious injury reports. A survey was sent to 2,424 Society of Thoracic Surgeons' members to obtain more information; 285 who use centrifugal pumps responded. Sixty surgeons (21%) reported 108 malfunctions, including 46 complete pump failures. Fifty-one of 243 surgeons (21%) who use centrifugal pumps for bypass reported that perfusionists have forgotten to clamp the pump line, resulting in backflow. We conclude centrifugal pumps are generally safe, but malfunctions, low flows, and human errors can lead to retrograde flow and occasionally air embolization. There are valves that can be added to the bypass circuitry to prevent this risk.

Results of the unidirectional Centri-Safe arterial valve for prevention of retrograde flow during cardiopulmonary bypass.

Retrograde blood from the aortic cannula into the cardiopulmonary circuit may lead to aortic air emboli when nonocclusive centrifugal pumps are used. The authors tested a nonregurgitant, unidirectional valve containing a Teflon ball occluder to prevent backflow. In vitro measurements of leakage rate, forward flow pressure drop, burst strength, and hemolysis levels along with animal (n = 12) and human (n = 12) in vivo hematologic and hemolysis levels were measured. Data were analyzed by paired and unpaired Student's t-test. Pressure drop differences at flows of 5 l/min were 7.3 +/- 0.3 mmHg before and 7.6 +/- 0.1 mmHg after 10,800 cycles of pulsatile pumping. (P = NS). Leakage rate during this period at pressures of 100 mmHg was not significant. Volume required to close the Teflon ball was less than 1 ml. Hemolysis analysis done in vitro and in vivo in control (no valve) and experimental (valve) groups used hemoglobin, hematocrit, platelets, plasma free hemoglobin, and lactic dehydrogenase as hemolysis indices. There were no statistical differences. The authors conclude that the CentriSafe valve (Cardiac Systems, Inc., Conshohocken, PA) is safe and prevents fatal backflow and air emboli. The valve is nonthrombotic in anticoagulated blood, can be opened and closed thousands of times, and has a burst strength equal to or greater than other components in the perfusion setup.

In vitro hemodynamic analysis of flexible artificial ventricles.

An in vitro fluid dynamic study was performed to compare the hemodynamic characteristics of a rigid and a flexible total artificial heart. The artificial ventricles were incorporated into a mock circulatory system, and pressure signals within the ventricular chamber, proximal to the inflow valve and distal to the outflow valve, were obtained. The instantaneous flow rate through the inflow and outflow valves was measured with electromagnetic flow probes. Flow visualization studies performed on the flexible ventricle suggested a vortical motion within the chamber with a smooth washout of fluid in the next pumping phase, but flow disturbances were observed near the wall of the ventricle as well as near the outflow valve. The rate of pressure increase (dP/dt) was smaller in the flexible ventricle as compared with the rigid ventricle for comparable flows and heart rates. The results of the present study indicated that the flexible ventricle with polyurethane valves, having the advantage of ease of implantation and cost savings, can be a viable alternative as a bridge to transplant.

In vitro continuous wave Doppler gradients of mechanical valves in less than optimal orientations.

Continuous wave Doppler ultrasound is a noninvasive method of determining transvalvular gradients that may overestimate the gradients of prosthetic valves. It is theorized that a valve rotated or tilted in its annulus will create abnormally high velocities that calculate into derived gradients that are greater than those actually present. An in vitro apparatus was constructed to analyze the velocities and the actual and derived gradients of prosthetic valves in less than optimal orientations. Continuous flow studies of a 27 mm Omniscience prosthesis indicated good agreement between the actual and derived gradients with the valve's major orifice oriented toward the aorta's larger radius. These values increase as the valve is tilted to decrease its net opening angle. When the valve is rotated 180 degrees, the Doppler gradients tend to overestimate the actual pressure drops. These data indicate that certain valve misorientations create velocities that overestimate the transvalvular gradients.

Hemodynamic comparisons of polyurethane trileaflet and bioprosthetic heart valves.

Hemodynamic comparison of two polyurethane prosthetic heart valves with a bioprosthetic valve is presented. The valves were incorporated in a pulse duplicator simulating physiologic pulsatile flow, and comparisons between the valves were made on the transvalvular pressure drop, percent regurgitation, valve orifice area, rate of opening and closing as well as the performance index. The results showed that the functional characteristics of the polyurethane valve compared favorably with that of the bioprosthetic valve. The polyurethane valve can be a viable and inexpensive alternative, especially for short-term use in a total artificial heart as a bridge to transplant.

In vitro comparison of velocity profiles and turbulent shear distal to polyurethane trileaflet and pericardial prosthetic valves.

A comparative study of flow dynamics past biomer trileaflet valves and a pericardial bioprosthetic valve under steady and physiological pulsatile flow conditions in vitro is reported in this paper. The velocity profiles and the turbulent shear stresses distal to the valves were measured using laser Doppler anemometry. The authors' results showed that the velocity profiles distal to the trileaflet valves were similar to that measured distal to the pericardial valve. Higher magnitudes of absolute turbulent shear stresses were measured distal to the synthetic valves in comparison to the pericardial valves. However, when the stresses were nondimensionalized with respect to the orifice diameter at the inlet aspect, the stresses were comparable for all of the three valves. With design modifications to increase the orifice diameter at the inlet aspect of the polyurethane valves, the turbulent stresses distal to the valves can be minimized. Such in vitro studies on the flow dynamics past the polyurethane valves can provide information towards design changes to improve the performance characteristics of these valves. Polyurethane valves with flow characteristics comparable to the pericardial valves can be manufactured relatively inexpensively compared to mechanical or tissue valve prosthesis. Hence, the synthetic valves may be a viable alternative for short-term use in total artificial heart devices as a bridge to transplant.

Development of the Philadelphia Heart System.

A new pneumatic artificial heart system has been developed. The design criteria have been to produce an integrated series of blood pumps and drive systems that would reduce blood trauma and reactivity, while incorporating industrial, mass-production techniques. The system attempts to reproduce the natural heart's pressure and flow waveforms and allows the prosthetic valves to be installed in a manner consistent with their design. The system's ventricles are constructed entirely of polyurethane by a combination of vacuum-forming and solution-casting techniques. The atrial cuffs and arterial grafts are permanently attached to the pumps and do not incorporate a quick connect system. The prosthetic valves are sewn into the inflow and outflow tracts using their clinical sewing rings. Besides eliminating the crevices normally found in quick connect systems, this method mounts the valves in an extremely compliant housing to increase shock absorption. The drive system produces a systolic air flow with a variable pressure rise (dP/dt) to reduce mitral valve closing velocity. This system has been implanted into 25 calves to date, of which 17 were chronic experiments. In 14 animals, St. Jude bileaflet valves were used and these animals had a mean survival of 39 days. Six of these animals survived over 30 days, with the longest being 129 days. All of the animals showed the characteristic postoperative drop in red blood cell count and hematocrit that returned to near preoperative values in about 3 weeks. The plasma free hemoglobin values generally remained below 5 mg/dl. The necropsies performed on several of the earlier animals revealed renal infarcts. However, in two of the later experiments, no kidney damage was found. The blood contacting surfaces of the atrial cuffs from the animals surviving over 100 days were covered with a fibroproliferative pseudoneointimal growth that extended from the sewing rings to the natural atrial tissue. Grossly, this appears to be the same type of tissue response seen when only a valve is implanted in a natural calf heart.

Deposition of blood elements on vacuum-formed ventricles in calves. A gross and electron microscopic study.

The thrombogenicity of the Philadelphia Heart System was evaluated by implanting left and right ventricles in calves for periods ranging from 1.5 hours to 128 days. Explanted hearts were examined grossly and by scanning and transmission electron microscopy for accumulation of thrombus and ingrowth of pseudoneointima. Visible material was essentially limited to inflow and outflow valves, atrial cuff, and a circumferential band of white thrombus around the junction of the blood diaphragm with the housing. The band was barely visible at 33 days, but was 1 to 3 mm wide at 128 days. SEM examination revealed microscopic deposits of blood elements, mostly platelets, that were present between 10 days and 128 days. The deposits were rather uniformly distributed over the blood contacting surface and remained constant over time. At 128 days, the Dacron (Meadox Medical, Oakland, NJ) graft was partly covered by a thin layer of tissue composed of multiple layers of cells separated by collagen. New blood vessels were frequently found, but fibrin-rich thrombus was present in some areas. Thus, this system was capable of minimizing thrombus accumulation and remained stable for 4 months.