Search

In vitro testing of hemolysis is essential for the validation and development of ventricular assist devices. However, as many factors influence hemolysis, such tests' inter- and intralaboratory reproducibility is poor. In this work, CentriMag blood pumps were used to conduct a hemolysis study according to ASTM F1841 with blood from 23 bovine donors. Complementary blood analysis, including cell count, plasma composition, and viscosity, was performed to identify factors relevant to the variability of hemolysis testing results. Three strategies were tested to improve reproducibility: albumin supplementation, maintaining glucose concentration, and replacement of plasma with plasma-like buffer. Differences in red blood cell stability among donors were responsible for the largest portion of the total variance. Hematocrit varied widely among donors, and its adjustment to a standard value led to the artificial introduction of between-donor differences, especially in viscosity. It seems likely, that a more careful selection of donors with similar characteristics or repeated blood collection from the same donor could improve reproducibility. However, no direct correlations were found between the susceptibility to hemolysis and individual donor or blood characteristics in this study. The addition of albumin and glucose had a negligible effect while washing blood samples with artificial plasma significantly reduced mean hemolysis, although not its variation. The findings contribute to the understanding of variability in hemolysis experiments and give reason to question the common standard operating procedures, such as hemodilution or maintaining glucose concentration. To confirm the factors identified here, additional studies isolating the effects of individual factors are necessary.

Additive manufacturing (AM) is an effective tool for accelerating knowledge gain in development processes, as it enables the production of complex prototypes at low cost and with short lead times. In the development of mechanical circulatory support, the use of cheap polymer-based AM techniques for prototype manufacturing allows more design variations to be tested, promoting a better understanding of the respective system and its optimization parameters. Here, we compare four commonly used AM processes for polymers with respect to manufacturing accuracy, surface roughness, and shape fidelity in an aqueous environment. Impeller replicas of the CentriMag blood pump were manufactured with each process and integrated into original pump housings. The assemblies were tested for hydraulic properties and hemolysis in reference to the commercially available pump. Computational fluid dynamic simulations were carried out to support the transfer of the results to other applications. In hydraulic testing, the deviation in pressure head and motor current of all additively manufactured replicas from the reference pump remained below 2% over the entire operating range of the pump. In contrast, significant deviations of up to 620% were observed in hemolysis testing. Only the replicas produced by stereolithography showed a nonsignificant deviation from the reference pump, which we attribute to the low surface roughness of parts manufactured thereby. The results suggest that there is a flow-dependent threshold of roughness above which a surface strongly contributes to cell lysis by promoting a hydraulically rough boundary flow.