Effect of cyclic deformation on xenogeneic heart valve biomaterials

Abstract

Glutaraldehyde-fixed bovine pericardium is currently the most popular biomaterial utilized in the creation of bioprosthetic heart valves. However, recent studies indicate that glutaraldehyde fixation results in calcification and structural valve deterioration, limiting the longevity of bioprosthetic heart valves. Additionally, glutaraldehyde fixation renders the tissue incompatible with constructive recipient cellular repopulation, remodeling and growth. Use of unfixed xenogeneic biomaterials devoid of antigenic burden has potential to overcome the limitations of current glutaraldehyde-fixed biomaterials. Heart valves undergo billion cycles of opening and closing throughout the patient's lifetime. Therefore, understanding the response of unfixed tissues to cyclic loading is crucial to these in a heart valve leaflet configuration. In this manuscript we quantify the effect of cyclic deformation on cycle dependent strain, structural, compositional and mechanical properties of fixed and unfixed tissues. Glutaraldehyde-fixed bovine pericardium underwent marked cyclic dependent strain, resulting from significant changes in structure, composition and mechanical function of the material. Conversely, unfixed bovine pericardium underwent minimal strain and maintained its structure, composition and mechanical integrity. This manuscript demonstrates that unfixed bovine pericardium can withstand cyclic deformations equivalent to 6 months of in vivo heart valve leaflet performance.

Fig 1. Flow diagram depicting how strips of pericardia were loaded to undergo consistent cyclic deformation.

Tissues were harvested and cut into strips, marked and imaged. Strips were then loaded to maintain a standard radius of curvature in cyclic deformation and ran in Dynatek’s M6 Heart Valve Tester.

Fig 2. Finite element model of the testing procedure to investigate the stresses applied to the tissue and whether they are physiological.

Continuum elements are in blue and rigid bodies are in grey. The dashed line indicates a plane of symmetry and the red markers indicate the region where displacement boundary conditions were applied. The simulation consisted of two steps; the first used the larger spherical rigid body to deform the sample to the radius of curvature at the start of the in vitro testing. The elements on the left edge highlighted were fixed in the y direction but could move in the x direction to allow the tissue to be pulled over the larger contact body. The second step removed this contact body then applied the maximum pressure in the in vitro testing (260 mmHg) to the base of the in silico sample. Here the highlighted elements were then held in both directions to mimic the sample being pinned.

Fig 3

Finite element model of the testing procedure at the end of the first step (A) and the end of the second step (B). For A, the tissue deformed around the two contact bodies with the profile of the tissue resembling that in the in vitro test. While for B, the tissue deformation exceeds that originally applied by the larger contact body and the maximum principal stress is also higher than at the end of the first step. As expected, the maximum principal stress is far lower in A than B. In the second step (B), the maximum principal stress in the central region of the model where the largest deflection occurred was 1.2 MPa and was located on the upper side of the tissue. The lower maximum principal stress on the underside of the tissue was approximately 0.42 MPa. The stress distribution is relatively consistent along the tissue likely because of the uniform loading applied.

Fig 4. Cyclic dependent strain is dependent on tissue source and fixation.

Longitudinal cycle dependent strain did not occur in unfixed bovine pericardium (A). Conversely, glutaraldehyde-fixed bovine pericardium exhibited negative longitudinal strain with increased cycle number (A). Porcine pericardium exhibited positive cycle dependent longitudinal strain which plateaued by 10 million cycles (A). Both bovine and porcine pericardium underwent negative cycle dependent perpendicular strain with increased cycle number (B). Glutaraldehyde-fixed bovine pericardium underwent positive cycle dependent perpendicular strain at 20 million cycles (B). Samples were assessed using a two-way analysis of variance, comparing means between tissue types and cycle number. n = 6 per group, per cycle number. Groups not connected by the same lower case letter (porcine pericardium), lower case double apostrophe (bovine pericardium), or lower case single apostrophe (glutaraldehyde-fixed bovine pericardium) are statistically significantly different. Data represent the mean ± s.d.

Fig 5. Nominal stress versus strain curve of native tissue, and tissue that has undergone 10 million (10m) or 20 million (20m) cycles of loading in the heart valve tester.

Tissue types were porcine pericardium (PP), bovine pericardium (BP), or glutaraldehyde fixed bovine pericardium (GFBP). Overall despite the differences between tissue types and fatigue the curves show a similar trend of initial strain-stiffening followed by a linear region and then a decrease in gradient to breaking. This may be interpreted as collagen fibers unfurling then becoming stretched before gradually damaging and ultimate failure.

Fig 6. Cyclic deformation alters mechanical properties of glutaraldehyde-fixed bovine pericardium, but leaves porcine pericardium and bovine pericardium unchanged.

Tangent modulus (A) and ultimate tensile strength (B) of glutaraldehyde-fixed bovine pericardium decreased following 10 million cycles of deformation but returned to baseline levels at 20 million cycles. No alteration in tangent modulus or ultimate tensile strength of either porcine or bovine pericardium was found following cyclic deformation. Tangent modulus and ultimate tensile strength were assessed using a two-way analysis of variance, comparing means between tissue types and cycle number. n = 6 per group, per cycle number. Groups not connected by the same lower case letter (porcine pericardium), lower case double apostrophe (bovine pericardium), or lower case single apostrophe (glutaraldehyde-fixed bovine pericardium) are statistically significantly different. Data represent the mean ± s.d.

Fig 7. Biochemical composition of glutaraldehyde-fixed bovine pericardium is altered following cyclic deformation, but maintained in porcine pericardium and bovine pericardium.

Both collagen (A) and elastin (B) content of glutaraldehyde-fixed bovine pericardium decreased following 20 million cycles. Biochemical composition of both porcine and bovine pericardium was unaltered by cyclic deformation. There was no change in tissue hydration with increased cycle number for any of the tissues (C). Biochemical composition was assessed using a two-way analysis of variance and comparing means between tissue types and cycle number. n = 6 per group, per cycle number. Groups not connected by the same lower case letter (porcine pericardium), lower case double apostrophe (bovine pericardium), or lower case single apostrophe (glutaraldehyde-fixed bovine pericardium) are statistically significantly different. Data represent the mean ± s.d.

Fig 8. Qualitative histological assessment of fixed and unfixed tissue biomaterials following cyclic deformation.

Nuclei density was maintained across all groups regardless of cycle number (hematoxylin and eosin (H&E)) (A). Elastin content decreased with increasing number of cycles for glutaraldehyde-fixed bovine pericardium, whereas both porcine and bovine pericardium maintained elastin content (Verhoeff-Van Gieson (VVG)) (A). However, elastin in porcine pericardium appeared fragmented with increasing number of cycles. All three tissue types underwent some cycle dependent separation of collagen bundles, although this was more dramatic in porcine pericardium and glutaraldehyde-fixed bovine pericardium than in bovine pericardium. n = 6 per group, per cycle number. Scale bar represents 50 μm (A). Collagen birefringence (B) of glutaraldehyde-fixed bovine pericardium at baseline is less than that of either of the non-fixed tissues (C). Collagen birefringence was significantly reduced following 20 million cycles for bovine pericardium (C). Glutaraldehyde-fixed bovine pericardium exhibited a trend towards increased collagen birefringence with increased cycle number, although this finding failed to reach significance (C). Percentage of birefringence was assessed using a two-way analysis of variance, comparing means between tissue types and cycle number. Scale bar represents 10 μm. n = 6 per group, per cycle number. Groups not connected by the same lower-case letter (porcine pericardium), lower-case double apostrophe (bovine pericardium), or lower-case single apostrophe (glutaraldehyde-fixed bovine pericardium) are statistically significantly different. Data represent the mean ± s.d.