Biomechanical Properties of Decellularized Porcine Pulmonary Valve Conduits

Transcription

1 Artificial Organs ( ):, Blackwell Publishing, Inc. 2007, Copyright the Authors Journal compilation 2007, International Center for Artificial Organs and Transplantation Biomechanical Properties of Decellularized Porcine Pulmonary Valve Conduits *Gernot Seebacher, Christian Grasl, Martin Stoiber, Erwin Rieder, *Marie-Theres Kasimir, Daniela Dunkler, *Paul Simon, *Günter Weigel, and Heinrich Schima *Department of Cardiothoracic Surgery, Medical University of Vienna; Institute of Biomedical Engineering and Physics, Medical University of Vienna; Department of General Surgery, Medical University of Vienna; Core Unit for Medical Statistics and Informatics, Medical University of Vienna; and Ludwig Boltzmann Cluster for Cardiovascular Research, Medical University of Vienna, Vienna, Austria Abstract: Tissue-engineered heart valves constructed from a xenogeneic or allogeneic decellularized matrix might overcome the disadvantages of current heart valve substitutes. One major necessity besides effective decellularization is to preserve the biomechanical properties of the valve. Native and decellularized porcine pulmonary heart valve conduits (PPVCs) (with [n = 10] or without [n = 10] cryopreservation) were compared to cryopreserved human pulmonary valve conduits (n = 7). Samples of the conduit were measured for wall thickness and underwent tensile tests. Elongation measurement was performed with a video extensometer. Decellularized PPVC showed a higher failure force both in longitudinal (+73%; P < 0.01) and transverse (+66%; P < 0.001) direction compared to human homografts. Failure force of the tissue after cryopreservation was still higher in the porcine group (longitudinal: +106%, P < 0.01; transverse: +58%, P < 0.001). In comparison to human homografts, both decellularized and decellularized cryopreserved porcine conduits showed a higher extensibility in longitudinal (decellularized: +61%, P < 0.001; decellularized + cryopreserved: +51%, P < 0.01) and transverse (decellularized: +126%, P < 0.001; decellularized + cryopreserved: +118%, P < 0.001) direction. Again, cryopreservation did not influence the biomechanical properties of the decellularized porcine matrix. Key Words: Tissue engineering Heart valve Biomechanical testing. The limited durability of conventionally used biological heart valve substitutes, and the consecutive high risk of thromboembolic events and bleeding after implantation of mechanical valve prostheses, are still major problems in cardiovascular surgery that remain unsolved (1,2). Due to limited durability and the lack of growth potential of current artificial heart valves especially in young patients who would benefit from a new durable biological heart valve, tissue engineering as a multidisciplinary approach doi: /j x Received September 2006; revised February Address correspondence and reprint requests to Dr. Günter Weigel, Department of Surgery, Division of Cardiothoracic Surgery, Surgical Research Laboratories, Medical University of Vienna, Vienna, Austria. guenter.weigel@meduniwien. ac.at could help to develop alternative devices. But the durable, nonimmunogenic, and nonthrombogenic heart valve prosthesis with the ability to grow is still not available (3 5). The decellularization approach of xenogeneic heart valves might be one option besides the use of biodegradable polymer as scaffolds for tissue engineering purposes. For clinical implementation, not only a thorough decellularization and eventual reseeding are needed, but also biomechanical stability has to be preserved. In former studies, we could demonstrate the decellularization of heart valve material and the successful seeding with human cells (6). The present study was carried out to describe the effects of decellularization, freezing, and thawing on the biomechanical properties of porcine heart valve tissue, and for the first time these data were compared to the stability of conventionally used homograft heart valve tissue. 1

2 2 G. SEEBACHER ET AL. C min 30 min 45 min 60 min 75 min 90 min FIG. 1. Schematic presentation of the temperature gradient used for cryopreservation of heart valve conduits. MATERIALS AND METHODS Porcine pulmonary heart valve conduits (PPVCs) (n = 10) of pigs ( kg) were obtained from a local slaughterhouse. Within 2 h after explantation, the conduits were incubated in sterile water containing an antibiotic solution with vancomycin (12 mg/l), metronidazole (12 mg/l), amikacin (12 mg/l), and ciprofloxacine (3 mg/l) for 48 h at 4 C. Six porcine pulmonary conduits were treated with a detergentbased decellularization procedure as previously described (6). Briefly, the conduits were placed into 100-mL sterile water containing 0.05% tert-octylphenyl-polyoxyethylene (Triton X-100, Bio-Rad, Hercules, CA, USA), 0.05% sodiumdeoxycholate (Merck, Darmstadt, Germany), and 0.05% octylphenyl-polyethylene glycol (IGEPAL- CA630, ICN,Aurora, OH, USA) for 48 h at 4 C. Conduits were subsequently treated with ribonuclease (100 mg/ml) (RNase, Roche Diagnostics GmbH, Mannheim, Germany) and deoxyribonuclease (150 IU/mL) (DNase, Sigma, St. Louis, MO, USA) with 50-mmol/L MgCl 2 in phosphate buffered saline (Ca ++ and Mg ++ free; PBS /, Gibco, Paisley, Scotland) for 24 h at 37 C. To remove residual detergents, the heart valve conduits were subsequently washed for 12 days in PBS / changed every second day, with an additional antibiotic sterilization carried out for at least 3 days at the end of the washing procedure. All steps were conducted under continuous shaking. For detection of tissue integrity and cell removal, histological examination was performed. Two longitudinal slices excised from each decellularized conduit were embedded in paraffin. Three sections (10 mm) of each slice were stained with hematoxylineosin (HE) and examined by two independent assistants. To exclude the presence of nucleic remnants, immunofluorescent staining with the DNAspecific dye TOPRO-3 (Molecular Probes, Leiden, The Netherlands) was performed. Human pulmonary heart valves were obtained in the course of heart transplantations. Valve conduits (n = 7) that could not be used for implantation as homograft substitutes for several reasons (e.g., valvular incompetence) were cryopreserved after antibiotic sterilization (approved by the Ethics Committee of the Medical University of Vienna). Cryopreservation Porcine valve conduits (native or decellularized) and cadaveric homografts were treated according to European tissue bank standards (7) and stored at -180 C until use. In brief, valve conduits were cryopreserved in sterile tubes containing RPMI 1640 medium and 10% (vol/vol) dimethyl sulfoxide. Cryopreservation was carried out in an automatic freezing device (Ice Cube 1610 Computer Freezer, SY-LAB, Neupurkersdorf, Austria) using the gradient depicted in Fig. 1. Mechanical testing Tests were performed with a universal tensile tester (Meßphysik Beta 10-2,5, Messphysik Labongeräte GesmbH, Fürstenfeld, Austria) and a 50-N force transducer (Hottinger Baldwin Type 70-2, Hottinger Baldwin Messtechnik GmbH, Darmstadt, Germany) according to a modified procedure presented by Korossis et al. (8). Elongation was measured with a video extensometer (Meßphysik ME46, Messphysik Laborgeräte GesmbH) between the two elongation measurement marks on the specimen (Fig. 2). This optical strain measurement ensures to measure the real elongation at the specimen and therefore avoids measuring errors caused by clamping. The specimens were punched out of the tissue with a specially designed punching tool. The specimens are dumbbell-shaped, mm overall in size, 6 7 mm in the gauge area (Fig. 2). Due to variation of wall thickness within one conduit and even within one specimen, measurement FIG. 2. Test sample with applied elongation measurement marks. Dumbbell-shaped test sample, 25 mm long and 10 mm broad; arrows show the position of the elongation measurement marks applied in the narrow part of the sample.

3 DECELLULARIZATION AND BIOMECHANICAL PROPERTIES 3 TABLE 1. Allocation of the samples that was matched in the porcine tissue groups whereas a slight hangover compared to the sample size in the homograft group must be reported Group Conduits n Longitudinal samples n Transverse samples n Porcine, native Porcine, native, cryo Porcine, decell Porcine, decell, cryo Human, cryo of the sample cross section necessary for calculation of tensile strength seems impractical. Therefore, we chose to present the mechanics of the tissue in terms of tension or force per unit width of specimen. This method ignores thickness and avoids the effect of varying tissue thickness within the specimen. Nevertheless, the thickness of all tissue specimens was measured with a vernier caliper and an averaged thickness value for each matrix group was calculated. Tissue specimens were gripped in the tensile tester by brass clamps with attached sand paper to prevent slipping. The tissue was loaded at 10 mm/min until rupture of the sample; the data from the load cell and the strain between the elongation marks were recorded online by the built-in personal computer. From the recorded load data, the tension was calculated according to the formula: T = F/I, where T is the tension in N/mm, F is the acquired force in Newtons (N), and I is the width of the specimen in millimeters (mm). Strain was calculated from the extension data according to the formula: e=dl/l 0 100, where e is the strain expressed in % of specimen elongation, Dl the extension of the specimen in mm, and l 0 the original length of the specimen in mm, at zero stress. Study groups Studies on five different groups of biological heart valve conduits were performed in order to obtain information about the natural strength of the material and to evaluate the influence of decellularization and preservation on the biomechanical stability of the matrix: Group 1: native, PPVC. Group 2: native, but cryopreserved PPVC. Group 3: decellularized PPVC. Group 4: decellularized and cryopreserved PPVC. Group 5: native, human cryopreserved pulmonary homografts. Each treatment group provided samples for biomechanical testing that were punched out of the pulmonary trunk in the size of mm (Fig. 1) in longitudinal and transverse direction and marked with ballpoint ink for elongation measurement. Details about the specimens that underwent biomechanical testing are given in Table 1. Statistics Load at failure normalized to the width of the specimen and strain at failure were analyzed and compared between the different groups of valve conduits. All values are shown as mean SD. Maximum strain and maximum tension in longitudinal and transverse directions are graphically summarized in boxplots. Wilcoxon s signed-rank test was used to assess the differences between the groups. Calculation was done by SPSS software package 14.0 (SSPS, Inc., Chicago, IL, USA). A P value less than 0.05 was considered significant. RESULTS Decellularization After the decellularization process, no microscopy cells or cell remnants could be detected within the examined porcine tissue sections, neither by light (HE staining) nor by immunofluorescence (TOPRO- 3-staining), which is in accordance with recent findings (6). A representative photomicrograph is shown in Fig. 3. Biomechanical testing No statistically significant differences of tension were observed between native and decellularized porcine specimens, thus indicating the preservation of biomechanical stability during decellularization of PPVC. Compared to human homografts, tension of decellularized PPVC was statistically significantly higher in longitudinal (P < 0.05, Fig. 4a) as well as in transverse direction (P < 0.001, Fig. 4b). Rupture of the porcine tissue occurred at much higher levels of strain. In longitudinal direction, the strain level of decellularized porcine samples was % while

4 4 G. SEEBACHER ET AL. FIG. 3. HE (left, 200 ) and TOPRO-3- staining (right, laser scanning microscopy 400 ) of tissue sections. After decellularization, no cell remnants could be detected, neither with HE staining nor with TOPRO-3-staining. human tissue showed a level of 71 30% (P < 0.001, Fig. 5a). In specimens punched out in transverse direction, the difference was even higher (P < 0.001, Fig. 5b). Porcine tissue reached a maximum strain of % while human tissue ruptured at a mean strain level of 47 18%. Cryopreservation did not alter the extensibility of the decellularized porcine tissue. Decellularized cryopreserved PPVC revealed a significantly higher tension than did human homografts in longitudinal (P < 0.01, Fig. 4a) as well as in transverse direction (P < 0.001, Fig. 4b). Results similar to that obtained in force experiment were found when measuring the extensibility in longitudinal (P < 0.01, Fig. 5a) and transverse direction (P < , Fig. 5b). The porcine conduits had a mean thickness of 2.2 mm and the human homograft of 1.4 mm (Fig. 6). Moreover, a comparison of decellularized porcine cryopreserved tissue with porcine native (with or without cryopreservation) and porcine decellularized specimens without cryopreservation, did not show any significant differences concerning tension and failure strain both in longitudinal and transverse directions. DISCUSSION As currently used cardiovascular prostheses are far from being the ideal substitute due to major postimplant complications such as thromboembolic events or early degradation and calcification, efforts have been undertaken in the evaluation and development of bioartificial heart valve substitutes. Tissue engineering, defined as the regeneration of tissue by cell transplantation with the aid of supporting structures, is thought to overcome the obstacles of artificial tissue prostheses in cardiovascular surgery. Two major principles are followed in matrix research: (i) development of biodegradable polymer constructs emulating the initial anatomical structure; and (ii) decellularization of preformed tissue, such as heart valves or blood vessels is under investigation. Decellularization generally refers to the removal of cells and cellular remnants while leaving a biological material composed essentially of extracellular matrix (ECM) components. The concept of decellularization or tissue engineering of heart valves was developed and modified by several groups (6,7,9 11). It is believed that the process of decellularization can reduce or eliminate the immune response or calcification associated with nonautologous cells and cell membranes. However, due to extraction of cells, it can be hypothesized that some components of the ECM are simultaneously removed, thus influencing the biomechanical properties of decellularized tissue. Hopkins (12) suggested to carefully define and measure the degradation occurring during decellularization and to relate them to appropriate standards derived from similar measurements of fresh human functioning heart valves. If a decellularization process weakens relevant physical properties of the putative transplant valve below a level where safety can be ensured, then additional steps will be needed to strengthen the ECM platform. This study has been designed to evaluate whether a decellularization procedure will alter the biomechanical properties of a PPVC in comparison with a conventional pulmonary homograft root. As for logistic purposes (e.g., shipping), freezing and thawing cycles might be included. Therefore, cryopreserved tissue samples were tested for their biomechanical performance. Although biomechanical testing was used in the evaluation of heart valves, most studies were focusing on aortic material and due to very heterogeneous

5 % strain Newton/millimeter DECELLULARIZATION AND BIOMECHANICAL PROPERTIES 5 a a 2, ,00 2,000 % strain 150,00 *** ** 1,500 ** * 100,00 1,000 50,00 0,500 porcine native porcine decell human homograft porcine cryo porcine cryo decell Groups porcine native porcine decell human homograft porcine cryo porcine cryo decell Groups b b 2,500 *** *** 200 Newton/millimeter 2, *** *** 1, , ,500 porcine native porcine decell human homograft porcine cryo porcine cryo decell Groups FIG. 4. (a) Maximum tension in longitudinal direction. Maximum force that the different materials could withstand until rupture (N/mm). Force was applied in longitudinal direction of the bloodstream. Porcine decellularized samples showed statistically significant higher tension than did human homografts (*P < 0.05). Porcine decellularized and cryopreserved tissue was even more resistant compared to human homografts (**P < 0.01). Wilcoxon s signed-rank test. (b) Maximum tension in transverse direction. Maximum force that the different materials could withstand until rupture (N/mm). Force was applied in transverse direction of the bloodstream. Porcine decellularized samples as well as porcine decellularized and cryopreserved tissue showed statistically significant higher tension than did human homografts (***P < 0.001). Wilcoxon s signed-rank test. porcine native porcine decell human homograft porcine cryo porcine cryo decell Groups FIG. 5. (a) Maximum strain in longitudinal direction. Maximum strain that the different materials could withstand until rupture (percentage of elongation compared to the original length of the specimen). Force was applied in longitudinal direction of the bloodstream. Porcine cryopreserved decellularized samples showed statistically significant higher strain than did human homografts (**P < 0.01). Porcine decellularized tissue was even more extensible compared to human homografts (***P < 0.001). Wilcoxon s signed-rank test. (b) Maximum strain in transverse direction. Maximum strain that the different materials could withstand until rupture (percentage of elongation compared to the original length of the specimen). Force was applied in transverse direction of the bloodstream. Porcine decellularized samples as well as porcine decellularized and cryopreserved tissue were statistically equal but showed both higher extensibility than did human homografts (***P < 0.001). Wilcoxon s signed-rank test.

6 6 G. SEEBACHER ET AL. wall thickness [mm] 3 2,5 2 1,5 1 0,5 0 homograft porcine decellularized, cryopreserved porcine decellularized porcine cryopreserved porcine native FIG. 6. Mean values of the wall thickness from each study group. mechanical tests, comparison and learning from these data are difficult. Lu et al. (13) tested tissue samples derived from porcine aortic walls and compared completely decellularized collagen and elastin scaffolds in stress strain. Collagen scaffolds exhibited lower tensile strength and lower distensibility than did fresh aorta while elastin scaffolds maintained their distensibility but had reduced tensile strength. Other studies investigated native aortic leaflets and compared them with those that were decellularized, polymer coated, and glutaraldehyde fixed (14 16). Hybrid heart valve scaffolds were fabricated from decellularized porcine aortic heart valve matrices and enhanced with bioresorbable polymers, and properties of the hybrid structures were compared with native, decellularized, and glutaraldehyde-fixed specimens. It was demonstrated that polymer impregnation stiffened the leaflets and increased tensile strength unlike polymer coating which exhibited a converse behavior with diminishing stiffness and tensile strength. Comparing human aortic and pulmonary roots as well as cusps on their biomechanical behavior has shown that pulmonary roots are more distensible and less compliant when loaded to aortic pressures. The pulmonary valve cusp and root tissue also showed greater extensibility and greater stiffness (lower compliance) when subjected to the same loads. Mechanical differences therefore were minimal between aortic and pulmonary valve tissues. It was postulated that the pulmonary root should withstand the forces imposed on it when placed in the aortic position. However, the pulmonary root will distend about 30% more than the aortic root when subjected to aortic pressures. These geometric changes may affect valve function in the long term and should be appreciated when implanting a pulmonary valve graft (16). Elkins et al. (17) compared native and decellularized wall and leaflet tissue of human pulmonary arteries. Samples were evaluated under uniaxial testing and compared with cryopreserved human pulmonary valve tissue. Samples of conduit and myocardium were also evaluated for suture retention ability. Furthermore, the valves underwent hydrodynamic testing and interestingly no alterations of biomechanical properties could be detected. A comparison of porcine native and acellular pulmonary leaflets revealed a loss of integrity with increasing decellularization time (18). Another study demonstrated that decellularization of porcine pulmonary leaflets leads to a significant loss of strength in radial and circumferential direction, but after reseeding in a bioreactor, a significant increase in stability was regained and was comparable with the mechanical properties of native ovine and porcine tissues. Biochemical and mechanical analyses revealed a continuous increase of cell mass, collagen, and elastin contents as well as an increase of strength under pulsatile culture conditions compared to significantly lower values in the static controls (19). Also calf pericardium, native or cell extracted, was tested for its biomechanical properties (20). No significant difference could be observed for fracture tension or percentage of strain at fracture. Mechanical properties of the fresh tissue seemed to be preserved in the pericardial acellular matrix. In the present study, biomechanical testing of porcine pulmonary conduits revealed that the extensibility of native but also of decellularized porcine pulmonary conduit tissue is significantly higher when compared to human homograft tissue. In contrast, the tension of the homograft tissue was much lower than that of porcine conduits. This different biomechanical characteristic can be explained by the smaller wall thickness of the homograft. Moreover, a comparison of decellularized porcine cryopreserved tissue with porcine native (with or without cryopreservation) and porcine decellularized specimens without cryopreservation, did not show any significant differences concerning tension and failure strain both in longitudinal and transverse directions. This clearly indicates that the several steps during processing of porcine pulmonary tissue (decellularization, cryopreservation, and thawing) do not alter tissue integrity that would result in impaired biomechanical stability. Cryopreservation is mandatory for planning optimal time of valve replacement and having the possibility of storage of a certain pool of valves in case of immediate and acute surgery.

7 DECELLULARIZATION AND BIOMECHANICAL PROPERTIES 7 2,5 2 Force [N/mm] 1,5 1 decellularized, transversal homograft, transversal FIG. 7. Stress strain curves of transverse specimens of a cryopreserved homograft and a cryopreserved decellularized PPVC. Decellularized porcine tissue revealed a longer elastin phase than did a homograft, while the collagen phase of the curve and the transition phase run parallel. 0, Strain [%] Regarding the stress strain curve of the porcine tissue, a longer elastin phase for porcine tissue could be observed while the collagen phase showed a parallel course (Fig. 7). It could be estimated that the elastin concentration is higher in porcine pulmonary conduits than in human homografts which might be responsible for the higher extensibility observed in our experimental setting. To the best of our knowledge, this is the first study comparing decellularized vascular material with conventionally used homograft valve tissue regarding mechanical properties. Our results demonstrate that cell-extracted porcine cardiovascular tissue should perform similarly to native tissue in an in vivo setting. However, as it has been shown that even after decellularization, an immune response toward porcine tissue will occur, the mechanical properties of the acellular tissue might be impaired due to inflammatory reactions and could explain the aneurysmatic alterations of decellularized xenogeneic vascular substitutes described after in vivo implantation. Although many issues on decellularized xenogeneic cardiovascular tissue are still to be solved, our results give new important information on the safety of tissue-engineered cardiovascular prostheses. REFERENCES 1. Cannegieter SC, Rosendaal FR, Briet E. Thromboembolic and bleeding complications in patients with mechanical heart valve prostheses. Circulation 1994;89: Teoh KH, Ivanov J, Weisel RD, Darcel IC, Rakowski H. Survival and bioprosthetic valve failure. Ten year follow up. Circulation 1998;80: Breuer CK, Mettler BA, Anthony T, Sales VL, Schoen FJ, Mayer JE. Application of tissue-engineering principles toward the development of a semilunar heart valve substitute. Tissue Eng 2004;10: Flanagan TC, Pandit A. Living artificial heart valve alternatives. A review. Eur Cell Mater 2003;6: Neuenschwander S, Hoerstrup SP. Heart valve tissue engineering. Transpl Immunol 2004;12: Rieder E, Kasimir MT, Silberhumer G, et al. Decellularization protocols of porcine heart valves differ importantly in efficiency of cell removal and susceptibility of the matrix to recellularization with human vascular cells. J Thorac Cardiovasc Surg 2004;127: Parker P, Hunt C. European Association of Tissue Banks standards for cryopreserved cardiovascular tissue banking. Cell Tissue Bank 2000;1: Korossis SA, Booth C, Wilcox HE, et al. Tissue engineering of cardiac valve prostheses II: biomechanical characterization of decellularized porcine aortic heart valves. J Heart Valve Dis 2002;11: Lee JM, Haberer SA, Boughner DR. The bovine pericardial xenograft: I. Effect of fixation in aldehydes without constraint on the tensile viscoelastic properties of bovine pericardium. J Biomed Mater Res 1998;23: Schoen FJ. Aortic valve structure-function correlations: role of elastic fibers no longer a stretch of the imagination. J Heart Valve Dis 1997;6: Rieder E, Seebacher G, Kasimir MT, et al. Tissue engineering of heart valves: decellularized porcine and human valve scaffolds differ importantly in residual potential to attract monocytic cells. Circulation 2005;111: Hopkins RA. Tissue engineering of heart valves decellularized valve scaffolds. Circulation 2005;111: Lu Q, Ganesan K, Simionescu DT, Vyavahare NR. Novel porous aortic elastin and collagen scaffolds for tissue engineering. Biomaterials 2004;25: Grabow N, Schmohl K, Khosravi A, et al. Mechanical and structural properties of a novel hybrid heart valve scaffold for tissue engineering. Artif Organs 2004;28: Spina M, Ortolani F, Messlemani AE, et al. Isolation of intact aortic valve scaffolds for heart-valve bioprostheses: extracellular matrix structure, prevention from calcification, and cell repopulation features. J Biomed Mater Res 2003;67A: Vesely I, Casarotto D, Gerosa G. Mechanics of cryopreserved aortic and pulmonary allografts. J Heart Valve Dis 2000;9: Elkins RC, Dawson PE, Goldstein S, Walsh SP, Black KS. Decellularized human valve allografts. Ann Thorac Surg 2001;71:S Schenke-Layland K, Vasilevski O, Opitz F, et al. Impact of decellularization of xenogenic tissue on extracellular matrix

8 8 G. SEEBACHER ET AL. integrity for tissue engineering of heart valves. J Struct Biol 2003;143: Schenke-Layland K, Opitz F, Gross M, et al. Complete dynamic repopulation of decellularized heart valves by application of defined physical signals an in vitro study. Cardiovasc Res 2003;60: Courtman DW, Pereira CA, Kashef V, McComb D, Lee JM, Wilson GJ. Development of a pericardial acellular matrix biomaterial: biochemical and mechanical effects of cell extraction. J Biomed Mater Res 1994;28: