(75c) Exploring Hemolysis Due to Suture Dehiscence in Prosthetic Heart Valves | AIChE

(75c) Exploring Hemolysis Due to Suture Dehiscence in Prosthetic Heart Valves


Twitchell, K. - Presenter, University of Oklahoma
Papavassiliou, D. - Presenter, University of Oklahoma
O'Rear, E. A., University of Oklahoma
Zhao, J., University of Auckland
Maruyama, T., Kyushu University
Hemolysis is a serious medical condition in which red blood cells (RBCs) lose hemoglobin and are removed from the circulation. One known cause of this condition is exposure of RBCs to high shear stresses, including those created by intravascular prosthetics such as artificial heart valves. In 1975, as many as 15% of patients with artificial heart valves had severe hemolytic anemia[1], but more modern designs and improvements to surgical technique have brought that number below 1%. The anemia caused by modern devices is often attributable not to the device itself, but rather to paravalvular leakage. When this leakage occurs, a small opening appears between the normal heart tissue and the rim of the prosthetic, creating a channel through which blood can flow as a jet in the improper direction from the lower chamber to the upper chamber of the heart. This condition, called dehiscence, which typically occurs when a suture stitching pulls out of the tissue, subjects RBCs to shear stresses outside the normal physiological range, and can lead to an increased rate of hemolysis and hemolytic anemia.

The relationship between flow parameters and the hemolysis index (HI), a measurement of the degree of hemolysis occurring in a blood sample, has been the object of some study. Leverett et al.[2] made an observation in 1972 that has allowed for empirical correlations between the shear stress, the exposure time, and the HI for laminar flows. For the turbulent flow regime, prior work done by Ozturk et al.[3] in this research group matched the HI to the Kolmogorov Length Scale (KLS) size of turbulent eddies.

An anatomically representative model of a human left atrium was constructed in part using a software package developed by Xiong et al.[4] which transcribed MRI slices into epicardium outlines. These outlines were then stacked together using Free-D[5], a specialized software program designed to connect several cross sections into a 3D object. The resulting file was imported to ANSYS Fluent 19.1[6] for use in computational fluid dynamics (CFD) calculations. A 3 mm long, semi-circular channel of diameter 1 mm was added to the atrium to simulate the paravalvular leakage. A Casson constitutive equation was used to describe variable blood viscosity.

Ozturk[3] describes a process known as eddy analysis that is used to connect turbulence to the HI. The KLS, defined in this case as the fourth root of the ratio of the kinematic viscosity of blood to the turbulent kinetic energy dissipation rate, is calculated in the volume of the atrium including the dehiscence site. The total volume occupied by fluid structures that correspond to KLS eddies comparable to the size of RBCs is determined. Eddies up to a KLS of 10µm, roughly the same diameter as RBCs, are shown to have significant impact on hemolysis, since eddies of such dimensions apply greater shear to RBCs and have been shown to be a main cause of RBC trauma.

This eddy analysis was preformed on the atrium model to predict the hemolysis that occurs under paravalvular leakage conditions. The goal is to be able to evaluate how severe the problem is when it occurs in a patient, potentially preventing an unnecessary and life-threatening post-op surgery, or recommending a necessary and life-saving repair.


  1. Sethi, P.; Murtaza, G.; Rhaman, Z.; Zaidi, S.; Helton, T.; Paul, T. “Valvular Hemolysis Masquerading as Prosthetic Valve Stenosis,” Cureus. 9, e1143 (2017).
  2. Leverett, L. B., Hellums, J. D., Alfrey, C. P., and Lynch, E. C., “Red Blood Cell Damage by Shear Stress,” Biophys. J., 12(3), pp. 257–273 (1972).
  3. Ozturk, M., Papavassiliou, D. V., O’Rear, E. A., “An approach of Assessing Turbulent Flow Damage to Blood in Medical Devices,” J. Biomech Eng. 139, 011001-011008 (2017).
  4. Xiong, Z., Fedorov, V. V., Fu, X., Cheng, E., Macleod, R., & Zhao, J. “Fully Automatic Left Atrium Segmentation From Late Gadolinium Enhanced Magnetic Resonance Imaging Using a Dual Fully Convolutional Neural Network,” IEEE transactions on medical imaging, 38(2), 515-524. (2019).
  5. Andrey P & Maurin Y “Free-D: an integrated environment for three-dimensional reconstruction from serial sections”. Journal of Neuroscience Methods, 145, 233-244 (2005).
  6. Ansys® Fluent, Release 19.1.