MULTI-LABORATORY STUDY OF FLOW-INDUCED HEMOLYSIS USING THE FDA BENCHMARK NOZZLE MODEL

• Published in: Artificial organs 2014-09, Vol.39 (3), p.237-248
• Main Authors: Herbertson, Luke H., Olia, Salim E., Daly, Amanda, Noatch, Christopher P., Smith, William A., Kameneva, Marina V., Malinauskas, Richard A.
• Format: Article
• Language: English
• ISSN: 0160-564X; 1525-1594
• DOI: 10.1111/aor.12368

Multilaboratory in vitro blood damage testing was performed on a simple nozzle model to determine how different flow parameters and blood properties affect device-induced hemolysis and to generate data for comparison with computational fluid dynamics-based predictions of blood damage as part of an FDA initiative for assessing medical device safety. Three independent laboratories evaluated hemolysis as a function of nozzle entrance geometry, flow rate, and blood properties. Bovine blood anticoagulated with acid citrate dextrose solution (2–80 h post-draw) was recirculated through nozzle-containing and paired nozzle-free control loops for 2 h. Controlled parameters included hematocrit (36 ± 1.5%), temperature (25°C), blood volume, flow rate, and pressure. Three nozzle test conditions were evaluated ( n = 26–36 trials each): (i) sudden contraction at the entrance with a blood flow rate of 5 L/min, (ii) gradual cone at the entrance with a 6-L/min blood flow rate, and (iii) sudden-contraction inlet at 6 L/min. The blood damage caused only by the nozzle model was calculated by subtracting the hemolysis generated by the paired control loop test. Despite high intralaboratory variability, significant differences among the three test conditions were observed, with the sharp nozzle entrance causing the most hemolysis. Modified index of hemolysis (MIH nozzle ) values were 0.292 ± 0.249, 0.021 ± 0.128, and 1.239 ± 0.667 for conditions i–iii, respectively. Porcine blood generated hemolysis results similar to those obtained with bovine blood. Although the interlaboratory hemolysis results are only applicable for the specific blood parameters and nozzle model used here, these empirical data may help to advance computational fluid dynamics models for predicting blood damage.