Ivan Vesely

MICROMECHANICS OF THE FIBROSA AND THE VENTRICULARIS IN AORTIC VALVE LEAFLETS

The elastic response of aortic valve cusps is a summation of its fibrous components. To investigate the micromechanical function of valve leaflet constituents, we separated the fibrosa and the ventricularis from fresh and glutaraldehyde-fixed leaflets and tested them individually. The ventricularis was stiffer circumferentially than radially (7.41 kPa vs 3.68 kPa, p less than 0.00001) and was more extensible radially (62.7% vs 21.8% strain to high modulus phase, p less than 0.00001). The fibrosa was also stiffer circumferentially than radially (13.02 kPa vs 4.65 kPa, p less than 0.0008), but had uniform extensibility. Glutaraldehyde fixation did not affect the circumferential elastic modulus of the fibrosa, but reduced its radial modulus from 4.65 kPa to 2.32 kPa (p less than 0.0078). The elastic modulus of the ventricularis remained unchanged. Fixation also reduced the extensibility of the ventricularis circumferentially (from 21.8% to 15.2% strain, p less than 0.018), but not radially, and increased the radial extensibility of the fibrosa from 27.7% to 46.1% (p less than 0.0048). These data show that while the ventricularis contains a large amount of elastin, the amount of radially oriented collagen is similar to that of the fibrosa. The fibrosa, by itself, has the same extensibility in both directions (about 23% strain), but can extend much more radially when connected to the rest of the leaflet because it is attached to the ventricularis in a highly folded configuration. The two layers therefore complement each other during aortic valve function, and become detrimentally altered by fixation in glutaraldehyde.

Tissue Buckling as a Mechanism of Bioprosthetic Valve Failure

Current reports indicate that collagen fiber disruption resulting from cyclic leaflet bending is a factor determining long-term durability of bioprosthetic heart valves. Examination of the opening characteristics of porcine xenografts has shown two areas of high bending curvature that correlate well with sites of leaflet tearing. These are at the free edge and near the attachment of the leaflets to the aortic root. To determine the potential effects of sharp bends in leaflet material, we examined 15 strips each of fresh and glutaraldehyde-treated porcine aortic valve tissue. Leaflet strips were bent to curvatures of 0.18 mm-1 to 6.67 mm-1, histologically processed, sectioned, and examined under a light microscope. We observed severe compressive buckling in the samples taken from bioprosthetic valves but little in the fresh-tissue samples. At physiological curvatures (less than 0.28 mm-1), no buckling occurred in the fresh tissue; at high bending curvatures (2.0 mm-1), the depth of buckling observed in the treated tissue was 100% greater than that in the fresh. We believe that porcine xenograft failure is related to compressive buckling of the aldehyde-treated tissue and is mediated by the systematic breaking of collagen fibers at the site of buckling. We suggest that alternative valve designs and preservation techniques be employed to prevent such abnormal leaflet deformations.

Numerical simulation of leaflet flexure in bioprosthetic valves mounted on rigid and expansile stents

Recent studies suggest that flexural stresses induced during the opening phase may be responsible for much of the mechanical failures of bioprosthetic heart valves. Sharp leaflet bending is promoted by the mounting of valves on rigid stents that do not mimic the systolic expansion of the natural aortic root. We, therefore, hypothesized that flexural stresses could be significantly reduced by incorporating a flexible or expansile supporting stent into the valve design. Using our own non-linear finite element code (INDAP) and the pre- and post-processor modules of a commercial finite element package (PATRAN), we simulated the opening and closing behaviour a trileaflet bovine pericardial valve. The leaflets of this valve were assumed to be of uniform thickness, with a non-linear elastic behaviour adapted from experimentally obtained bending stiffness data. Our simulations have shown that during maximal systolic valve opening, sharp curvatures are induced in the leaflets near their commissural attachment to the supporting stent. These areas of sharp flexure experience compressive stresses of similar magnitude to the tensile stresses induced in the leaflets during valve closure. By incorporating a stent with posts that pivot about their base, such that a 10% expansion at the commissures is realized, we were able to reduce the compressive commissural stressing from 250 to 150 kPa. This was a reduction of 40%. Conversely, a simple pliable stent with stent posts that deflect inward and outward under load did not achieve a significant reduction of compressive stresses. This numerical analysis, therefore, supports the theory that (i) high flexural and compressive stresses exist at sites of sharp leaflet bending and may promote bioprosthetic valve failure, and (ii) that proper design of the supporting stent can significantly reduce such flexural stresses.

Aortic valve cusp microstructure: The role of elastin**

The aortic valve cusp is a three-layered structure, composed of differing amounts of collagen, elastin, and glycosaminoglycans. Little quantitative information is presently available on the amount, location, orientation, and overall structure of these constituents, particularly of elastin. We developed a technique to isolate aortic valve elastin in a morphologically intact state. Whole leaflets were digested in 0.1 N sodium hydroxide solution at a temperature of 75 degrees C. Both scanning electron microscopy and computerized three-dimensional reconstructions of serial sections showed a well-defined honeycomb or spongelike structure, suggesting that elastin forms a matrix that surrounds and links the collagen fiber bundles. This relationship between collagen and elastin is further supported by the naturally wavy configuration of the valve cusps, permitting elongations of 40%, even though collagen fibrils typically strain to 1% to 2% before fracture. Elastin likely acts to return collagen fibers back to their undeformed state, maintaining rest geometry.

Biaxial strain analysis of the porcine aortic valve

The function of a bioprosthetic heart valve is determined largely by the material properties of the valve cusps. The mechanics of natural and bioprosthetic valve cusps have been studied extensively using uniaxial tensile testing. This type of testing, however, does not duplicate the natural biaxial loading condition. Whole-valve biaxial testing therefore is preferred. The objective of the present study was to investigate the heterogeneity of the valve cusps by mapping out the regional variability of the biaxial strain versus pressure relationship. Whole porcine aortic valves were mounted horizontally, submerged in physiologic saline solution at 37 degrees C, and pressurized in the range of 0 to 130 mm Hg of pressure. The ventricular side of the cusps were marked with black dots and the three-dimensional position of these dots was recorded together with the aortic pressure. By calculating the distance between the dots in the radial and circumferential directions in different regions, the local strain versus pressure relationship was determined. The results showed that the valve cusp material strained by 23% +/- 0.8% in the radial direction and 10.0% +/- 0.5% in the circumferential direction before lock-up. It was also found that while the valve cusp was highly anisotropic in the central region, the basal region was relatively isotropic, and the cusp as a whole was asymmetrical in its distensibility.

The pulmonary valve. Is it mechanically suitable for use as an aortic valve replacement

Pulmonary autografts have shown a low incidence of early failure and late structural deterioration when placed in the aortic position yet the potential value of the pulmonary valve as a replacement device has not been widely considered. Since the mechanical suitability of pulmonary valves for the high stress aortic position is unclear, we set out to define and compare the tensile mechanical properties of these two valves. We removed all 72 cusps from 12 fresh porcine aortic and pulmonary valves. Eighteen cusps from each of the two groups were fixed flat in 0.25% glutaraldehyde for greater than 24 hr. Circumferential or radial strips were cut from each cusp creating eight test groups, each with nine specimens. Stress-strain curves for each specimen were obtained using an Instron tensile testing machine. Stress-strain curves were obtained at 2, 10, 50, and 200 mm/min strain rates, each specimen was then strained to fracture. We found pulmonary leaflets were thinner than aortic leaflets (0.49 mm versus 0.67 mm) and glutaraldehyde fixation did not affect this relationship. The elastic moduli were comparable within the fresh and glutaraldehyde fixed treatment groups and within the circumferential and radial strip groups. Radial strips were more extensible than circumferential strips. Tissue viscous properties were similar and glutaraldehyde fixation produced minimal changes in stress relaxation rates. Tissue fracture tests emphasized tissue anisotropy with the failure point depending upon strip orientation. Glutaraldehyde had no effect on fracture stress but strain at fracture doubled due to increased collagen fiber crimping during fixation. Overall, porcine pulmonary valve cusps, whether fresh or glutaraldehyde fixed, exhibited similar tensile properties to aortic valve cusps.

Longitudinal and radial distensibility of the porcine aortic root

The aortic root has been shown to be a highly distensible structure. The function of the aortic valve is intimately related to the expansion of the aortic root, and current nonexpansible stent designs may affect its performance. We therefore measured the radial and longitudinal expansion of the porcine aortic root as a function of pressure in both a static pressurization model and in an isolated working heart model. The radial and longitudinal expansion of the aortic root was measured using a custom-built digital sonomicrometer. Multiple ultrasonic crystals were sutured exterior to the commissures and along the length of the aortic root, and their separation was tracked at varying aortic pressures. In static testing, we found that commissural separation at zero pressure was 26% +/- 7% (mean +/- standard deviation) less than at 120 mm Hg, whereas the longitudinal distance between the base of the valve and the commissures decreased by 11% +/- 9%. Approximately one quarter of the total dimensional change occurred over the physiologic range of 80 to 120 mm Hg. In the isolated porcine heart model, we measured a greater distensibility than in the static tests. For example, at aortic pressures of 120/80 mm Hg (systolic/diastolic), the diameter of the aortic root would be 22% +/- 6% less at 80 mm Hg than at 120 mm Hg. The longitudinal dimensions would be 15% +/- 8% less at 80 mm Hg than at 120 mm Hg. We conclude that the aortic root contracts significantly when depressurized, as during valve replacement surgery, and that the in vivo distensibility of the aortic root is much greater that what is generally measured in vitro. These results suggest that dimensional changes in the implanted prosthetic valve and the recipient aortic root must be considered to achieve both optimal valve orifice and, in the case of distensible valves such as allografts, a proper valve cusp geometry.

Dynamic glutaraldehyde fixation of a porcine aortic valve xenograft. I. Effect of fixation conditions on the final tissue viscoelastic properties.

Sixty porcine aortic valves were fixed under dynamic conditions at specific durations, pressures and vibration rates in a 0.5% glutaraldehyde phosphate buffer (pH 7.4, 0.2 M). Tensile relaxation tests were performed at low through high extension rates (0.3, 3 and 30 mm s-1) and tissue denaturation temperatures were determined by the hydrothermal isometric tension method. Conventional statically fixed valves and fresh valves were used as controls. No differences between dynamic and static treatment were observed at pulsation rates above those expected in the physiological range (i.e. above 1.2 Hz) or at higher pressures such as 30 mmHg. However, differences in both stress relaxation rates and denaturation temperatures were delineated in milder fixation conditions, i.e. at low pressures (< 4 mmHg) and low vibration rates similar to that of the normal heart beat (approximately 1.2 Hz). In these conditions the relaxation rate of the dynamically fixed tissue (-7.4 +/- 0.7% of stress remaining per log(s)) was similar to that of the fresh tissue (-6.7 +/- 1.2% log(s-1)) and significantly higher than the statically treated tissue (-3.9 +/- 1.7% log(s-1)). The rates of stress relaxation appeared to be strain rate dependent in both radial and circumferential directions when the tissues were strained at physiological rates during testing (> approximately 15000% min-1). Dynamically treated valves showed higher denaturation temperatures (mean +/- SD) (89.4 +/- 0.5 degree C) compared with the statically fixed (82.7 +/- 1.4 degrees C) or untreated (fresh) valves (65.5 +/- 0.8 degree C). The results suggest a higher degree of internal cross-linking owing possibly to enhanced penetration of the glutaraldehyde reagent and a greater accessability of reactive cross-linking sites on the collagen molecules. Better stress relaxation rates are likely associated with an increase in potential shearing between adjacent collagen fibres thus preserving the natural stress-reducing mechanism of the fresh, untreated valves. The dynamically treated valves therefore possess characteristics that may enable them to better resist long-term mechanical fatigue and in vivo degradation.

A comparison of macroscopic lipid content within porcine pulmonary and aortic valves: Implications for bioprosthetic valves

Lipid droplets have been demonstrated within both explanted porcine bioprostheses and normal porcine aortic valves. Because of the increasing interest in pulmonary valves as an allograft or xenograft aortic valve substitute, we examined the incidence and distribution of such lipid deposits in 50 porcine aortic valves and 50 matched porcine pulmonary valves. All 300 cusps were removed with surgical scissors and, under a dissecting microscope, the ventricularis layer was removed to expose the spongiosal layer. Macroscopic extracellular lipid droplets analyzed by means of a dissecting microscope with an eyepiece grid and stereology point-counting techniques to provide an area-density average spatial probability map for each cusp. Only 8% of porcine aortic valves were free of lipid, with the distribution of the lipids being 52% +/- 14% right coronary cusp, 90% +/- 8% left coronary cusp, and 68% +/- 13% noncoronary cusp. Of the pulmonary valves, 60% were free of lipid, with the incidence of lipids being 26% +/- 12% left cusp, 6% +/- 7% right cusp, and 12% +/- 9% anterior cusp. Subsequently, lipid cluster samples underwent thin-layer chromatography, which showed them to be phospholipids, oleic acid (fatty acid), triglycerides, and unesterified cholesterol. One primary mode of bioprosthetic valve failure is leaflet calcification. The similarity of distribution within the spongiosal layer between leaflet calcification and intrinsic cusp lipids suggests that the observed lipids might act as a nucleation site for calcification. The substantially lower incidence of lipid in pulmonary valves therefore may represent a potential benefit when these valves are considered for use as aortic valve replacements.

The hybrid xenograft/autograft bioprosthetic heart valve: In vivo evaluation of tissue extraction

The major functional problem with bioprostheses is poor long-term durability. Bioprosthetic valves fail because of calcification and mechanical fatigue, both of which result from the glutaraldehyde fixation process. In an effort to develop a biologically active, non-cross-linked bioprosthetic valve, we devised a cellular extraction process. We tested the mechanical integrity of the processed valves and cultured both human and porcine cells on this material. To test the potential for calcification, we implanted strips of fresh, extracted, and glutaraldehyde-treated porcine heart valve tissue subcutaneously into 3-week-old Sprague Dawley rats for 21 days. We used atomic absorption spectroscopy to measure the extent of calcium accumulation and histopathologic assessment to evaluate the antigenic response. We found that the cell extraction process significantly reduced the propensity of the material to calcify in vivo (mean +/- standard deviation, 4.12 +/- 1.02 mg/g calcium extracted versus 10.75 +/- 3.9 mg/g calcium fresh versus 79.6 +/- 18.3 mg/g calcium glutaraldehyde fixed) but increased the antigenicity, as evidenced by increased cellular activity and resorption. Although they may reduce calcification, conventional detergent-based cell extraction techniques do not completely remove porcine aortic valve antigens and may in fact increase the antigenicity of the valve cusp material.