K. Mizuguchi et al., DEVELOPMENT OF THE BAYLOR NASA AXIAL-FLOW VENTRICULAR ASSIST DEVICE -IN-VITRO PERFORMANCE AND SYSTEMATIC HEMOLYSIS TEST-RESULTS, Artificial organs, 18(1), 1994, pp. 32-43
Our newly developed axial flow pump consists of a flow tube, an intern
al rotating impeller, and a fixed flow stator (we call the stator) beh
ind the impeller. This pump produces a flow of 3 to 8 L/min against 50
to 150 mm Hg pressure difference, respectively, in the range of 10,00
0 to 16,000 rpm. An axial flow pump that will be used as a ventricular
assist device (VAD) has to have low hemolytic and good antithrombogen
ic characteristics. This paper will show how to decrease the hemolytic
properties of this axial flow pump systematically using a test matrix
. The test variables evaluated were impeller blade tip geometry, impel
ler flow tube clearance (radial clearance), impeller stator clearance
(axial clearance), impeller blade number, stator blade number, and imp
eller length. All in vitro hemolysis tests were performed at 5.0 L/min
against 100 mm Hg pressure difference using a total of 83 bags of fre
sh bovine blood. The results were as follows: the impeller blade tip g
eometry did not significantly effect hemolysis, a 0.005-inch and a 0.0
09-inch radial clearance were significantly (p < 0.01 or 0.001) less h
emolytic than the other clearances, a 0.075-inch axial clearance was s
ignificantly (p < 0.05) more hemolytic than the other clearances, two-
and six-bladed impellers were significantly (p < 0.01 and 0.02, respe
ctively) less hemolytic than a four-bladed impeller, a five-bladed sta
tor was significantly (p < 0.05 or 0.01) less hemolytic than the other
stators, and the impeller length did not make a significant differenc
e. Currently, the best index of hemolysis is 0.031 +/- 0.018 g/100 L,
and using parameters from these results, implantable devices are being
fabricated.