Linking Computational Fluid Dynamics Modeling to Device-Induced Platelet Defects in Mechanically Assisted Circulation

Jiafeng Zhang; Dong Han; Zengsheng Chen; Shigang Wang; Wenji Sun; Bartley P Griffith; Zhongjun J Wu

doi:10.1097/MAT.0000000000002242

Linking Computational Fluid Dynamics Modeling to Device-Induced Platelet Defects in Mechanically Assisted Circulation

ASAIO J. 2024 May 20. doi: 10.1097/MAT.0000000000002242. Online ahead of print.

Authors

Jiafeng Zhang¹, Dong Han¹, Zengsheng Chen¹, Shigang Wang¹, Wenji Sun¹, Bartley P Griffith¹, Zhongjun J Wu^{1

2}

Affiliations

¹ From the Department of Surgery, University of Maryland School of Medicine, Baltimore, Maryland.
² Fischell Department of Bioengineering, A. James Clark School of Engineering, University of Maryland, College Park, Maryland.

PMID: 38768482
DOI: 10.1097/MAT.0000000000002242

Abstract

Thrombotic and bleeding events are the most common hematologic complications in patients with mechanically assisted circulation and are closely related to device-induced platelet dysfunction. In this study, we sought to link computational fluid dynamics (CFD) modeling of blood pumps with device-induced platelet defects. Fresh human blood was circulated in circulatory loops with four pumps (CentriMag, HVAD, HeartMate II, and CH-VAD) operated under a total of six clinically representative conditions. Blood samples were collected and analyzed for glycoprotein (GP) IIb/IIIa activation and receptor shedding of GPIbα and GPVI. In parallel, CFD modeling was performed to characterize the blood flow in these pumps. Numerical indices of platelet defects were derived from CFD modeling incorporating previously derived power-law models under constant shear conditions. Numerical results were correlated with experimental results by regression analysis. The results suggested that a scalar shear stress of less than 75 Pa may have limited contribution to platelet damage. The platelet defect indices predicted by the CFD power-law models after excluding shear stress <75 Pa correlated excellently with experimentally measured indices. Although numerical prediction based on the power-law model cannot directly reproduce the experimental data. The power-law model has proven its effectiveness, especially for quantitative comparisons.