Andrew H. Goldstein

End organ function with prolonged nonpulsatile circulatory support.

The hypothesis tested in these studies was that long-term circulatory support with a nonpulsatile device is safe and causes no end organ dysfunction. An inexpensive, small centrifugal pump with a 7 L/min capacity was implanted in 6 sheep (15 acute implants have previously been reported). The inlet cannula was placed in the left atrium and the outlet graft anastomosed to the descending aorta. A percutaneous cable supplied DC power and heparinized saline (10 ml/hr) for lubrication. Outputs of pump flow, stator, animal core temperature, pump power consumption, and RPM were monitored throughout the course of each experiment. The sheep moved freely within a large pen using an overhead swivel/tether system that carried all input and output lines. Four sheep survived longer than 4 weeks, and the indices of end organ function were analyzed at 28 days. No animal revealed any neurologic dysfunction. Hemoglobin was 9.075 +/- 0.78 g/dl at 28 days, as opposed to 7.475 +/- 0.68 (p = 0.002) before surgery. The blood urea nitrogen was 9.250 +/- 4.57 versus 14 +/- 5.72 mg/dl (p = 0.041), creatinine was 0.775 +/- .10 versus 0.775 +/- 0.05 mg/dl (p > or = 0.999), total bilirubin was 0.425 +/- 0.2 versus 0.225 +/- 0.05 mg/dl (p = 0.092), serum glutamic oxaloacetic transaminase was 74.75 +/- 24 versus 106.25 +/- 15.84 IU/L (p = 0.015), serum glutamic pyruvic transaminase was 36 +/- 28.7 versus 28.3 +/- 5.7 IU/L (p = 0.25), and total protein was 6.675 +/- 0.49 versus 5.47 +/- 0.15 g/dl (p = 0.025). It is concluded that these animals adapted very well to pulseless circulatory support. The results of these studies support the concept of an inexpensive, implantable, centrifugal pump as a ventricular assist device.

Predictable reduction in left ventricular stroke work and oxygen utilization with an implantable centrifugal pump

Previous investigations with roller pumps and pneumatic pulsatile assist devices have demonstrated that nearly complete capture of normal left ventricular end-diastolic volume was necessary for appreciable reductions in oxygen consumption and stroke work. We tested the hypothesis that a centrifugal pump would decrease left ventricular stroke work and oxygen consumption as a function of pump flow. Ten sheep (35 to 50 kg) were instrumented and placed on left atrium-to-descending aorta bypass with a small, lightweight (112 g), implantable centrifugal pump. The relations between pump flow as a percent of cardiac output (% bypass), left ventricular stroke work, and oxygen consumption were studied. Left ventricular stroke work was calculated from the pressure-volume loops obtained with micromanometer and conductance catheters and was indexed per 100 g of left ventricular wet weight. Oxygen consumption was calculated from left main coronary artery blood flow and the arterial-coronary sinus oxygen content difference, normalized to 100 g of left ventricular wet weight and a heart rate of 100 beats/min. Measurements were made in stepwise increments of pump flow from zero to the maximum obtainable and then reversed in similar decrements. Analyses were made for 27 complete runs. Our data demonstrate that reductions in left ventricular stroke work and oxygen consumption were achieved from zero to maximal bypass. There was an approximate 66% and 50% reduction in left ventricular stroke work and oxygen consumption, respectively, at 60% bypass.(ABSTRACT TRUNCATED AT 250 WORDS)

Small, low-cost implantable centrifugal pump for short-term circulatory assistance

BACKGROUND In 1991, Allegheny General Hospital and Allegheny-Singer Research Institute purchased a centrifugal pump, then a 2-year-old technology, from Medtronic Bio-Medicus, as part of its research program for novel treatments of acute and chronic heart failure. During a 4-year development program, we then established and met goals of durability, performance, thromboresistance, and low cost. METHODS In vitro testing involved extensive hydraulic characterizations using Penn State mock loops. Calorimetry was used to determine efficiency. Durability studies used heated (37 degrees C) seawater for 28 to 45 days. In vivo studies used 46 sheep to test performance and engineering changes and to determine myocardial oxygen consumption, thromboresistance, and long-term durability. A left atrium-to-aorta circuit was used in all. RESULTS Hydraulic testing showed no preload sensitivity but moderate afterload sensitivity at all impeller speeds (2,000 to 6,000 rpm). The heat load was low, and overall efficiency was 13% to 15%. Bench durability studies showed no electrical malfunction of the stator or console without degradation of the biomaterials used. Acute in vitro studies showed a near-linear relationship of myocardial oxygen consumption and left ventricular stroke work, pump flow, and pump speed. At speeds of 2 to 3 L/min (50% bypass), left ventricular stroke work and myocardial oxygen consumption were decreased approximately 50%. Additionally, 5 animals have had implants for 28 to 154 days with no macroemboli or microemboli detected in any animal. Hematologic and biochemical studies became normal 3 to 7 days after implantation. Hemolysis was low at less than 10 mg/dL. Clinical costs of the device are estimated to be 80% less than those of currently available devices. CONCLUSIONS We conclude that an old technology has been made into new technology by application of sound engineering design principles, microchips, and new biomaterials. Qualifying trails for a Food and Drug Agency investigational device exemption application are in progress.

Aortic arch replacement for recurrent cerebral embolization

Replacement of the aortic arch for atheroma with cerebral embolization is in its infancy. The appropriateness of such intervention is controversial. Over a 10-month period, a 58-year-old woman suffered multiple debilitating cerebral vascular accidents manifested by motor, sensory, and memory deficits and documented by computed tomographic scanning and magnetic resonance imaging. Carotid and vertebral arteries were free of arteriosclerotic disease. Transesophageal echocardiography demonstrated two large atheromas with friable, pedunculated forms, one in the aortic arch and one in the very proximal descending thoracic aorta. Transcranial ultrasound revealed recurrent cerebral microembolic events. Cerebrovascular events continued, and the atheromas increased in size, despite treatment with Coumadin and aspirin. Under deep hypothermic arrest, the segment of the aortic arch harboring the atheroma was excised and replaced with a Dacron graft. Repeat transcranial ultrasound revealed cessation of embolic signals. All cerebrovascular events ceased. No further anticoagulation therapy was required. The patient has made substantial recovery from the preoperative deficits and continues to do well 1 year after aortic arch replacement. Resection of mobile aortic arch atheromas is likely to become increasingly important in the future as transesophageal echocardiography leads to their more common identification as a cause of cerebral ischemic events.

Modified fabrication techniques lead to improved centrifugal blood pump performance.

The authors are developing an implantable centrifugal blood pump for short- and medium-term (1-6 months) left ventricular assist. They hypothesized that the application of result dependent modifications to this pump would lead to overall improved performance in long-term implantation studies. Essential requirements for pump operation, such as durability and resistance to clot formation, have been achieved through specialized fabrication techniques. The antithrombogenic character of the pump has been improved through coating at the cannula-housing interfaces and the baffle seal, and through changing the impeller blade material from polysulfone to pyrolytic carbon. The electronic components of the pump have been sealed for implantable use through specialized processes of dipping and potting, and the surfaces of the internal pump components have been treated to increase durability. The device has demonstrated efficacy in five chronic sheep implantation studies of 14, 10, 28, 35, and 154 day duration. Post mortem findings from the 14 day experiment showed stable fibrin entangled around the impeller shaft and blades. After pump modification, autopsy findings of the 10 day study showed no evidence of clot. Additionally, the results of the 28 day experiment showed only a small (2.0 mm) ring of fibrin at the shaft-seal interface. In the 35 and 154 day experiments, redesign of the stators have resulted in improved motor corrosion resistance. The 35 day study showed a small, 0.5 mm wide fibrin deposit at the lip seal, but no motor failure. In the 154 day experiment, the motor failed because of stator fluid corrosion, while the explanted pump was devoid of thrombus. Based on these findings, the authors believe that these pump refinements have contributed significantly to improvements in durability and resistance to clot formation.