Introduction

The surgical repair, reconstruction or closure of different cardiac vessels and structures following cardiac surgery often requires the use of a replacement material of either biological or synthetic origin. A variety of different materials have been assessed as substitution materials during reconstructive surgical procedures, with variable success.

Differences in criteria for an ideal material have contributed to the long list of potential substitute materials, although today, most users agree that the ideal material must be inert, non-toxic, non-carcinogenic, impermeable to liquids, able to hold sutures, prevent meningeal adhesions or infections, be handled and sterilized easily and also be inexpensive. Bovine pericardium has been identified as a material which satisfies most of the criteria and seems to have suitable properties for use as a substitute material (Baharuddin et al. 2002).

In the resection and patch or prosthetic reconstruction of the pulmonary artery and superior vena cava, for example, the use of biological materials such as autologous or bovine pericardium, azygos vein and saphenous vein, has achieved greater acceptance than synthetic materials. This was mainly due to improved biocompatibility, a lower risk of infection and thrombolysis, and lower cost than synthetic materials (D’Andrilli et al. 2005).

The clinical use of bovine pericardium for the construction of an artificial heart valve was first reported in 1971 by Ionescu, and since then it has been used worldwide to treat various congenital cardiac defects (Neuhauser and Oldenburg 2003). However, severe calcification in vivo of pericardium fixed and stored in glutaraldehyde (GA) remained a big concern (Neethling et al. 1996). The aim of this study was to provide sufficient evidence that an alternative biochemical treatment of the tissue before storage in GA will yield substantial improvement regarding in vivo calcification, without sacrificing any of the required and proven properties, and that it would be safe to use as substitute material during cardiothoracic procedures.

Materials and methods

Fresh bovine pericardial sacs (n = 5) were obtained from the local abattoir, transported on ice to the laboratory and manually cleaned of fat and adventitial tissue while being washed in copious amounts of cold (4°C) Plasmalyte B solution (Adcock Ingram, Johannesburg, South Africa). Within 4–6 h after collection, six samples (3 × 4 cm) and another six samples (9 × 3 cm) were cut from each pericardial sac and divided into six groups for different biochemical treatments (GA and 5 GAG concentrations), subsequent protein denaturation temperature determination and tensile strength testing. Group VII consisted of similar samples cut from five donated Glycar patches.

Chemical treatments

In Group I (control group), pericardial tissue samples were cross-linked in 0.625% phosphate-buffered GA (Merck Chemicals, Johannesburg, South Africa), pH 7.4, at room temperature for 24 h and stored in the same solution.

Pericardial tissue samples in Groups II–VI were fixed in different concentrations (0.0025, 0.005, 0.01, 0.1 and 0.2 M) of sodium metaperiodate + 0.5% chondroitin sulphate (Merck Chemicals, Johannesburg, South Africa) in distilled water under constant shaking at 4°C for 24 h, followed by thorough washing for two cycles of 30 min with 4°C phosphate-buffered saline (PBS) (Highveld Biological, Johannesburg, South Africa), and stored in 0.625% GA until implantation (Arenaz et al. 2004).

Group VII consisted of commercially available Glycar pericardial tissue samples (n = 5) (donated by Glycar Inc.). Tissue treatment involved immediate cleaning after harvesting, followed by initial cross-linking in 0.625% GA for at least 72 h, starting in the abattoir to minimize the ischemic time, and changing the GA twice during this period. Selected pericardium were cut into the designated sizes and sterilized in formaldehyde for 48 h, washed in saline to remove excess GA and treated with a high concentration of propylene glycol for 7–14 days at room temperature. Samples were finally stored in 2% propylene oxide in sterile water, and sterilization and packaging were performed under environmental control in class 100 Clean Room conditions (Frater et al. 1997).

Protein denaturation temperatures

Five samples from each of the differently treated pericardiums (GA, 5 GAG concentrations, Glycar) (n = 35) were used to determine the protein denaturation temperatures. A small tissue sample was placed in a hermetically sealed pan of a differential scanning calorimeter (Mettler Toledo, DSC 822e, Microsep, Johannesburg, South Africa) and subjected to thermal analysis. The temperature was raised at a rate of 10°C/min from 25 to 95°C, and the temperature of thermal denaturation for each sample was electronically recorded as a peak maximum (Lovekamp and Vyavahare 2001).

Tensile strength testing

Mechanical properties of tissues were examined by fixing both ends of the tissue sample in the clamps of a tensile testing machine (Lloyds LS100 Plus, IMP, Johannesburg, South Africa), gradually stretching it by applying constant tension on the two ends and recording the data on a personal computer (Thubrikar et al. 1983).

Six pericardial samples (9 × 3 cm) cut from each of the five pericardial sacs and treated with GA (control, n = 6) and the five different GAG concentrations (n = 30), and the Glycar patches (n = 6), were inserted between the clamps of the tensile strength tester and stretched at a controlled rate of 10 mm/min until the breaking point was reached.

Enzymatic degradation

Pericardial tissue samples from all seven treatment groups were dried in a temperature controlled incubator at 70°C for 24 h and weighed. Collagenase enzyme (Sigma-Aldrich, Johannesburg, South Africa) was suspended at a concentration of 440 U/ml in a solution of 50 mM Tris–HCl buffer (Sigma-Aldrich, Johannesburg, South Africa) and 0.36 mM CaCl2 at pH 7.4. Approximately 1.2 ml of this solution was added to each of the dried tissue samples and allowed to react for 24 h at 37°C under constant shaking, centrifuged for 5 min at 4,000 rpm and most of the liquid was discarded. Insoluble residues of tissue were again dried completely and weighed. Dry weights of the undigested samples were compared with those obtained prior to enzymatic digestion, and the tissue loss was calculated and expressed as a percentage of the dry weight.

In vivo studies: subcutaneous implantation in rats

Juvenile (6 weeks old) male Albino Wistar rats (100–150 g), were used as a standard screening model for in vivo calcification. All animals received humane care in compliance with the “Principles of Laboratory Care” (National Society for Medical Research) and the “Guide for the Care and Use of Laboratory Animals” published by the National Institutes of Health. The study protocol was approved by the Animal Ethics committee of the University of the Free State.

All animals were anaesthetised with Ketamine (45 mg/kg s.c.) (Centaur Labs, Isando, South Africa) and Medetomidine (0.3 mg/kg s.c.) (Pfizer Laboratories, Johannesburg, South Africa) for 45–60 min, shaved dorsally and a midline incision of ±3 cm made through the skin. Pericardial tissue samples were rinsed for 15 min in sterile 0.9% saline (Adcock Ingram, Johannesburg, South Africa), implanted subcutaneously into separate dorsal pockets (2 on each side) of the animal and secured with two 6/0 Prolene sutures (Johnson & Johnson, Johannesburg, South Africa). The incision was closed with a continuous 5/0 PDS absorbable suture (Johnson & Johnson, Johannesburg, South Africa) and the anaesthetic reversed with Antisedan (0.2–0.4 mg/kg s.c.) (Novartis, Kempton Park, South Africa). Analgesic (Buprenorphine, 0.01–0.05 mg/kg s.c. 8–12 hourly) (Schering-Plough, Isando, South Africa) was administered for 4 days post-operatively. Animals were sacrificed 8 weeks post-op by means of an overdose CO2-inhalation, and all samples were retrieved for determination of calcium and water content.

During phase I, rats (n = 11) received one pre-cut sample (0.8 × 1.5 cm) from each of the GA,0.01 M GAG-treated and Glycar patches.

During phase II, a total of nine rats received subcutaneous pericardial implants. Samples of pericardial patches treated with the highest (0.2 M), lowest (0.0025 M) and optimal (0.01 M) concentrations of GAGs together with a control patch (GA) were implanted subcutaneously into four rats (n = 4), and patches treated with the highest (0.2 M), lowest (0.0025 M) and optimal (0.01 M) concentrations of GAGs together with a sample from the commercial Glycar patch were implanted into another five rats. Samples were retrieved after 8 weeks for determination of extractable calcium content, water content and histological evaluation.

Quantitative calcium analysis

The quantitative calcium analysis was performed by the Eco-Analytica Laboratory, School for Environmental Sciences & Development, Northwest University, Potchefstroom, South Africa.

Explanted samples were dried in a temperature controlled incubator (Scientific, Series 100, Lasec, Johannesburg, South Africa) at 45°C for 48 h, weighed, hydrolyzed in 1 ml 50% nitric acid (Protea Laboratories, Johannesburg, South Africa) + 50 μl hydrogen peroxide (H2O2) (Diagnostic Media Products, Sandringham, South Africa)/dry sample at 90°C for 30–40 min, and the extractable calcium content was determined by atomic absorption spectrophotometry (Agilent ICP-MS 7500c, Chemetrix, Midrand, South Africa) and expressed as μg calcium per mg tissue (dry weight).

Tissue water content

The water content of the different treated pericardial samples (5 × [GAG] + GA + Glycar) implanted into the nine rats during phase II (n = 36) was also determined and compared following explantation after 8 weeks. Explanted tissue samples were cleaned of excess host tissue, and weighed on a digital balance (Mettler AE 100, Protea Laboratories, Sandton, South Africa). Samples were then dried in a temperature controlled incubator at 45°C for 48 h, again weighed and the water content calculated as a percentage of the wet weight.

Histological analysis

Chemicals used were obtained from Merck, Johannesburg, South Africa unless otherwise indicated. All samples were routinely embedded in paraffin wax, sectioned and stained as required. All histological samples were routinely stained with hematoxylin and eosin, and also with Alcian blue 8GX and counterstained with nuclear fast red, to visualize the presence of GAGs.

Explanted samples were stained with (1) von Kossa stain to quantify calcium deposits (briefly, this involved staining with silver nitrate solution under strong light for 30 min, reducing the stain with 0.5% hydroquinone, treatment with 3% sodium thiosulphate and counterstained with 1% nuclear fast red); (2) Gomori Trichrome for assessment of possible infiltration of host immune cells. Sections were stained with Mayer’s hematoxylin, nuclei differentiated with 1% acid alcohol for 10 s, allowed to turn blue in Scott’s tap water for 30 s and stained in the Trichrome solution at pH 3.4.

Statistical analysis

Histological findings were categorised and summarised according to frequencies and percentages. Numeric data was expressed as means and standard deviations or percentiles, depending on the distribution of the data. Comparisons were done by paired t tests or signed rank tests (numerical variables), or McNemar tests (categorical variables) in the case of paired data. For unpaired data t tests, Mann–Whitney tests or χ2 tests were used. Confidence intervals were calculated for differences in means, medians or percentages.

Results

Thermal denaturation temperature

The mean peak temperature at which denaturation of the triple helix of the collagen molecules occured, was recorded in order to determine the optimal GAG concentration to be used for cross-linking of bovine pericardial tissue (Fig. 1). There was no statistically significant (p = 0.1245) difference in denaturation temperature between tissues treated with 0.0025 and 0.005 M GAGs. Tissues treated with 0.01, 0.1 and 0.2 M GAGs showed a significant (p ≤ 0.0443) decrease in denaturation temperature compared to the 0.0025 and 0.005 M GAG-treated tissue. GA- and Glycar-treated tissue showed a statistically significant (p ≤ 0.0005) increase in mean denaturation temperature compared to all the GAG-treated tissues.

Fig. 1
figure 1

Thermal denaturation temperatures (°C) for differently treated pericardial samples. Data are expressed as mean values ± SD

Full size image

Figure 2 demonstrates the marked increase in mean denaturation temperature and tissue strength following different chemical treatments, compared to fresh untreated bovine pericardial tissue.

Fig. 2
figure 2

Representative diagram of the cyclic warming data for fresh, GA, 0.01 M GAG and Glycar-treated patches as recorded by the differential scanning calorimeter, with the point of protein denaturation clearly indicated

Full size image

Tensile strength

Figure 3 shows the mean tensile strength at the breaking point of different pericardial tissues treated with (a) GA (control), (b) Five different concentrations of GAG and (c) Glycar method.

Fig. 3
figure 3

Tensile strength (MPa) measurements of differently treated pericardial strips. Data are expressed as mean values ± SD

Full size image

The mean tensile strength of control (GA) tissue was significantly (p ≤ 0.0313) lower compared to 0.005 M GAG, 0.01 M GAG-treated and Glycar tissue, but no significant (p ≥ 0.0938) difference existed between GA and the remaining GAG treatments. Tissue treated with the lowest (0.0025 M) and two highest (0.1 and 0.2 M) GAG concentrations appear to have a lower tensile strength compared to 0.005 and 0.01 M GAG-treated tissues. Only 0.1 M GAG-treated tissue showed a significant (p = 0.0313) decrease in tensile strength when compared to 0.005 M GAG-treated tissue.

Glycar-treated tissue demonstrated a significantly (p = 0.0039) higher tensile strength compared to GA-fixed control tissue (median values), but no significant difference (p ≥ 0.0547) in tensile strength existed between Glycar tissue and any of the GAG treatments (median values).

Enzymatic degradation

The extent of cross-linking of biological tissues is demonstrated by its resistance to enzymatic degradation. Figure 4 shows the resistance of different pericardial samples, expressed as a percentage of tissue dry weight after digestion with collagenase at 37°C for 24 h.

Fig. 4
figure 4

Enzymatic degradation of different pericardial tissues, with tissue loss expressed as a percentage of tissue dry weight before digestion by collagenase. Data are expressed as mean values ± SD

Full size image

No significant difference (p ≥ 0.0938) in mean tissue loss was found between the different treatment groups, except for the Glycar tissue, which had significantly (p < 0.05) less collagen digested when compared to all the other treated tissues.

Extractable calcium and water content

Figure 5 shows the calcium and water content for control samples (GA) compared to the lowest, highest and optimal concentration GAG-treated samples and commercial tissue after 8 weeks in the subcutaneous rat model. All treatments showed a significantly (p < 0.0001) higher mean water content compared to control (GA) tissue. There was no significant difference (p ≥ 0.2) between the water content of all the GAG-treated tissues.

Fig. 5
figure 5

The calcium and water content of the different pericardial samples after 8 weeks in the subcutaneous rat model. Data are expressed as mean values ± SD

Full size image

A statistically significant (p < 0.0001) decrease in the extractable calcium content of all the GAG-treated and the Glycar-treated patches compared to the control (GA) tissue was shown. When tissue treated with the different GAG concentrations were compared with each other as well as with the Glycar-treated tissue, no significant difference (p ≥ 0.0625) was found.

Histology

Histological appearance with von Kossa staining following 8 weeks subcutaneous implantation in rats showed significantly (p ≤ 0.0160) reduced calcification of the Glycar-treated tissue compared to all the other tissues, while 0.2 M GAG-treatment showed significantly (p = 0.0280) less calcification than the control tissue. Control (GA) tissue showed severe calcification (Fig. 6).

Fig. 6
figure 6

Histological comparison of light microscopy images of the degree of calcification of the explanted pericardial samples treated with (a) 0.625% GA, (b) 0.01 M GAG and (c) Glycar method (von Kossa, magnification (a, b, c) ×40)

Full size image

Histological appearance with Alcian blue staining of a pre-implantation pericardial sample treated with 0.01 M GAG showed that the GAG was only bound superficially to the outer surface of the pericardium, with limited penetration into the deeper layers (Fig. 7). Alcian blue staining of all the explanted pericardial samples showed no significant difference (p ≥ 0.1329) in the presence of GAG between any of the differently treated samples.

Fig. 7
figure 7

Light microscopy image of a 0.01 M GAG-treated implant, demonstrating the GAG (light blue), superficially bound to the outer surface of the pericardium (Alcian blue, magnification ×4). (Color figure online)

Full size image

Gomori trichrome staining to indicate host cell infiltration into explants showed a mild to moderate presence of host lymphocytes in the superficial layers of all explants, with mild infiltration into deeper layers only in some explants. One sample showed an acute inflammatory response due to infection after partial excavation of the patch by the recipient animal.

Discussion

Glutaraldehyde has been widely used as cross-linking agent for biological tissues for many years, because of some unique cross-linking and sterilizing properties. It produces materials with the highest degree of cross-linking, excellent haemodynamic performance, increased mechanical strength and a low antigenicity. Despite these advantages, GA has also been indicated as one of the major causes of clinical failure of bioprosthetic tissues. Calcification, lipid deposition and mechanical stresses further impact negatively on tissue durability and outcomes, while GA treatment also induces structural deterioration of collagen molecules and promotes GAGs extraction (Arenaz et al. 2004).

Many alternative cross-linking methods and chemical compounds have been investigated, each showing different degrees of success regarding reduction of calcification while maintaining tissue integrity and mechanical durability. In this study the optimal concentration to be used in pre-treating pericardial tissue with glycosaminoglycans was determined, and its effect on protein stabilization, mechanical strength and calcification potential was compared to GA-fixed and propylene glycol-treated tissue.

Stability of the triple helix in collagenous tissues is increased by forming additional chemical bonds or cross-links between the molecules, thus increasing the denaturation temperature of the material. This is reflected in the significant elevation of denaturation temperatures compared to that of fresh pericardium by various chemical treatments, with GA and Glycar tissue giving the optimal results (Figs. 1, 2).

The significantly lower denaturation temperatures of tissues treated with all five GAG concentrations compared to GA and Glycar tissue (Fig. 1) could be due to the fact that GAG was only bound superficially without deep penetration into the tissue (Fig. 7). Treatment with lower GAG concentrations (0.0025 and 0.005 M) also resulted in a limited amount of GAG-collagen cross-links being formed, allowing GA to form the majority of cross-links during the final fixation treatment, and thus the higher denaturation temperatures. Tissues treated with high (0.1 and 0.2 M) concentrations of GAGs yielded denaturation temperatures below the accepted benchmark of 80°C, indicating limited cross-links formed by GA fixation following GAG fixation and confirming the superficial binding of GAGs. This corresponds to a previous study which suggested that high concentrations of GA promote rapid cross-linking of the tissue surface during fixation, generating a barrier that impedes or prevents further diffusion of GA into the tissue bulk (Cheung et al. 1985).

Treatment with 0.01 M GAG revealed a significantly lower denaturation temperature than tissues treated with 0.0025 and 0.005 M GAG, but it was still above the acceptable minimum of 80°C. This was indicative of an adequate degree of stable cross-linking, and therefore the 0.01 M GAG treatment was identified as the optimal concentration for further use.

The thickness of the layer of GAGs bound to the outer surface of the pericardium for different GAG concentrations was not investigated. A difference in tissue thickness and the thickness of the superficially bound layer of GAGs might affect the rate of GA penetration and cross-linking formation during the final fixation phase. This could result in differences in thermal denaturation temperatures, but this was however not investigated in this study. Fisher and colleagues did however demonstrate that the penetration of GA into bovine pericardium and the resultant cross-linking was uniform, even after only 2 h of fixation (Fisher et al. 1987).

Propylene glycol-treated tissue showed superior resistance to thermal denaturation (Fig. 1), indicating that the cross-links introduced by cross-linking with GA were adequately maintained. Limiting the ischemic time between harvesting of the tissue to initial fixation in 0.625% GA to introduce the collagen cross-links will be crucial in order to minimize the degradation of the tissue before fixation.

The presence of solvents in polyol solutions will introduce additional functional groups, which may react with the amino groups of the collagen. This will reduce the formation of cross-links by the aldehyde groups of GA. Using concentrated propylene glycol and other polyols instead of solvent-containing solutions to ‘cap’ the free aldehyde groups in the tissue will also minimize the possibility of impeding in the development of cross-links in the GA-crosslinked tissue, which might reduce the tissue strength.

Contrary to the high thermal denaturation temperatures achieved with GA fixation and Glycar-treated tissues compared to GAG treatments, the tensile strength of GA-treated pericardium and high concentrations of GAG was significantly lower compared to other treatments (Fig. 3). The substantially lower water content of GA-treated tissue had a negative effect on the elasticity and suppleness of the tissue, reducing its mechanical properties. The smaller effect of GA cross-linking on tissue treated with high concentrations of GAGs, also resulted in reduction of tensile strength. Cross-linking of fresh bovine pericardium with GA resulted in an increase in tensile strength due to the increase in the stiffness (reduction in stress relaxation) of the tissue as a result of added interfibrillar cross-links, as well as an increase in the elongation of the tissue at break (Lee et al. 1994).

The treatment of Glycar patches with concentrated propylene glycol following GA-cross-linking, ensured that the tissue maintained a high water content and remained supple. This, together with the adequate cross-linking of the pericardium by GA during processing as proven by DSC, resulted in the increased tensile strength of tissue treated with the lower GAG concentrations and the Glycar patches. This is supported by Maestro and colleagues, arguing that the higher water content may facilitate the reorganization of collagen fibers within the tissue when tensile forces are exerted, improving the visco-elastic properties of the pericardium (Maestro et al. 2006).

The more pronounced influence of GA-induced cross-links formed in the 0.0025 M GAG-treated tissue compared to higher GAG concentrations, resulted in a tensile strength comparable to GA-fixed tissue. The presence of chondroitin sulphate was shown to decrease the tensile strength of cross-linked collagenous matrices, due to interfibrillar slippage of collagen (Pieper et al. 1999). This, in combination with the limited GA cross-links formed, might explain why the tensile strength of the 0.1 and 0.2 M GAG treatments was not increased significantly compared to the GA-fixed tissue. The hydrophilic nature of the GAG-treated tissue, because of their highly negatively charged units, was responsible for the significant increase in the water content compared to GA-fixed tissue. GAG-treated and Glycar-treated patches displayed similar water contents, which would be indicative of a hydrophilic nature of the Glycar treatment on pericardial tissue as well (Fig. 5).

The resistance of bovine pericardium to collagenase digestion indirectly reveals the degree of cross-linking of collagen (Jee et al. 2003). The comparable resistance of the GAG-treated and GA tissues to enzymatic digestion thus confirms the adequate degree of cross-linking obtained by pre-treatment of pericardium with GAG, without sacrificing thermal stability or tensile strength. Propylene glycol-treated tissue demonstrated significantly higher resistance to enzymatic digestion when compared to all the other treatments (Fig. 4). This, combined with its low calcification potential, high degree of cross-linking stability and tensile strength, makes propylene glycol-treated pericardium a superior substitute material for use in cardiac and other reconstruction procedures.

Treatment of biological tissue with GA is detrimental towards calcification, and results in severe calcification in rat subcutaneous implants (Fig. 6a). Incomplete cross-linking of the deeper collagen layers could occur, due to the potential barrier formed by GAG on the outer surface. The concentration and temperature during cross-linking can also make penetration of GA into dense pericardial tissue very slow (Khor 1997), and complete stabilization of collagen fibers might take up to 1 month (Chanda et al. 1997). This could present binding sites for calcium and serve as nucleation sites for calcification, as seen at histological evaluation in the fairly evenly dispersed calcification sites in the inner layers of the GAG explants (Fig. 6b).

Blocking of the free aldehyde groups in GA-treated bovine pericardium with a liquid polyol like propylene glycol (Glycar patch), showed a substantial decrease in the aldehyde-induced calcification of the tissue following subcutaneous implantation in rats (Fig. 6c). Treatment with di-substituted diols like the various propane diols and 2,3-butylene glycol rather than tri-substituted triols such as glycerol, proved to be more stable and effective in blocking the free aldehyde groups and thus minimizing the calcification of the tissue even further (Seifter and Frater 1995).

The presence of GAGs on the outer surface of only about 50% of GAG-treated explants and a moderate, evenly dispersed presence in surrounding host tissue indicated a leaching out of the GAGs into the surrounding host tissue, but no correlation with the calcification of the tissue could be made. The total absence of GAG was also seen in unimplanted GAG-treated tissue stored for several months in 0.625% GA, and correlates with finding by others that GAG continue to be lost from bioprosthetic heart valve leaflet tissue following storage in a 0.2% GA solution (Lovekamp et al. 2006). Conventional GA cross-linking of bioprosthetic heart valves does not stabilize the GAGs present, and the gradual loss of these GAGs might make these implants prone to calcification and tissue failure (Lovekamp and Vyavahare 2001).

Conclusion

Treatment of biological tissue with GA is highly detrimental towards calcification, as demonstrated in the rat subcutaneous implant model. Binding sulphated GAG to proteins in pericardial tissue before final fixation with GA resulted in significant reduction in calcification, while maintaining good structural integrity, mechanical properties and low antigenicity.

The leaching of GAGs from treated biological tissues presents a major challenge, and permanent cross-linking and fixation of the GAGs in order to produce stable and durable bioprostheses will be required before they can be used safely and with confidence, especially in a pulsatile haemodynamic system. GAG that bound only to the outer surface makes tissue prone to calcification and degradation, and penetration and fixation into the deeper layers, as well as more effective stabilization of the cross-links that are formed, will be required.

The chemical composition of the storage medium for biological tissue transplants will require agents free of potentially harmful aldehyde groups in solution while still rendering suitable sterilization, and propylene oxide might prove to be a good alternative.

Determinants of the calcification of bioprosthetic tissue implants has been proven to be multifactorial, and optimal management of all these factors will be required in order to enhance the longevity of bioprosthetic implants.