- Email: [email protected]
In vivo repopulation of xenogeneic and allogeneic acellular valve matrix conduits in the pulmonary circulation
Rainer G. Leyh, MD, Mathias Wilhelmi, MD, Philip Rebe, MD, Stefan Fischer, MD, Theo Kofidis, MD, Axel Haverich, MD, PhD, and Heike Mertsching, PhD Division of Thoracic and Cardiovascular Surgery, Hannover Medical School , and Leibniz Research Laboratories for Biotechnology and Artificial Organs (LEBAO), Hannover, Germany
Background. Approaches to in vivo repopulation of acellularized valve matrix constructs have been described recently. However, early calcification of acellularized matrices repopulated in vivo remains a major obstacle. We hypothesised that the matrix composition has a significant influence on the onset of early calcification. Therefore, we evaluated the calcification of acellularized allogenic ovine (AVMC) and xenogenic porcine (XVMC) valve matrix conduits in the pulmonary circulation in a sheep model. Methods. Porcine (n ⴝ 3) and sheep (n ⴝ 3) pulmonary valve conduits were acellularized by trypsin/EDTA digestion and then implanted into healthy sheep in pulmonary valve position using extracorporeal bypass support. Transthoracic echocardiography (TTE) was performed at 12 and 24 weeks after the implantation. The animals were sacrificed at week 24 or earlier when severe calcification of the valve conduit became evident by TTE. The valves were examined histologically and biochemically.
Results. All AVMC revealed severe calcification after 12 weeks with focal endothelial cell clustering and no interstitial valve tissue reconstitution. In contrast, after 24 weeks XVMC indicated mild calcification on histologic examination (von Kossa staining) with histologic reconstitution of valve tissue and confluent endothelial surface coverage. Furthermore, immunohistologic analysis revealed reconstitution of surface endothelial cell monolayer (von Willebrand factor), and interstitial myofibroblasts (Vimentin/Desmin). Conclusions. Porcine acellularized XVMC are resistant to early calcification during in vivo reseeding. Furthermore, XVMC are repopulated in vivo with valve-specific cell types within 24 weeks resembling native valve tissue.
M
matrix conduits after in vitro repopulation with autologous myofibroblast and endothelial cells [4, 5]. In vitro repopulation with autologous myofibroblast and endothelial cells resulted in complete reconstitution of functioning viable heart valve tissue. In contrast, in vivo repopulation was incomplete with subsequent early calcification. Whether this is true for biological matrices from other species has yet to be determined. We hypothesized that the matrix composition has a significant impact on the onset of early calcification of these tissue engineered heart valves. Therefore, we evaluated the calcification of acellularized allogenic ovine (AVMC) and xenogenic porcine (XVMC) valve matrix conduits after in vivo repopulation in the pulmonary circulation in a sheep model.
echanical and biological heart valve prostheses are nonviable constructs with the inherent drawbacks, that these structures have no growing, repairing, and remodelling ability. Tissue engineered heart valves may have the potential to eliminate the disadvantages of the commonly used mechanical and biological heart valve prostheses. The concept of tissue engineering is to create an isogenic tissue, which is based on an anatomically shaped matrix, repopulated with viable cellular tissue components. Biodegradable polymer scaffolds have been used to reconstruct valve conduits, which resulted in a partial cellular and functional reconstitution of valve tissue [1, 2]. However, limitations in cellular adhesion and tissue regeneration are an unsolved problem when biodegradable polymer scaffolds are used [2]. The concept of a biological scaffold material for heart valve tissue engineering might be an alternative to overcome these obstacles [3]. Recently, we proved the concept of tissue engineering of valve conduits on allogenic ovine Presented at the Thirty-eighth Annual Meeting of The Society of Thoracic Surgeons, Fort Lauderdale, FL, Jan 28 –30, 2002. Address reprint requests to Dr Leyh, Division of Thoracic and Cardiovascular Surgery, Hannover Medical School, Carl Neuberg St. 1, 30623 Hannover, Germany; e-mail: [email protected].
© 2003 by The Society of Thoracic Surgeons Published by Elsevier Science Inc
(Ann Thorac Surg 2003;75:1457– 63) © 2003 by The Society of Thoracic Surgeons
Material and Methods Acellularization of Pulmonary Valve Conduits Porcine pulmonary valve conduits (n ⫽ 3) were obtained from pigs (German landrace), ranging from 25 to 30 kg, and sheep (n ⫽ 3), ranging from 25 to 30 kg, from the Tierzuchanstalt, Mariensee, Germany. Hearts were obtained under sterile conditions, and pulmonary valve conduits were harvested with a thin ridge of subvalvular 0003-4975/03/$30.00 PII S0003-4975(02)04845-2
CARDIOVASCULAR
In Vivo Repopulation of Xenogeneic and Allogeneic Acellular Valve Matrix Conduits in the Pulmonary Circulation
CARDIOVASCULAR
1458
LEYH ET AL XENOGENIC MATRIX CONDUITS IN PULMONARY CIRCULATION
muscle tissue proximally and a short segment of the truncus pulmonalis distally. The macroscopic anatomy of ovine and porcine pulmonary valve conduits did not reveal any gross differences. The valve conduits were stored in Hanks buffered saline solution (HBSS; Biochrom, Cambridge, UK) at 4°C, and incubated with 0.05% trypsin (Biochrom) and 0.02% EDTA (Biochrom) for 48 hours, followed by PBS flushing for 48 hours to remove all cellular debris. Samples of the conduits were taken before and after this treatment to document acellularization by hematoxylin-eosin and movat-pentachrom staining using light microscopy.
Animal Model We used an immature sheep model to test our biological heart valve scaffolds because bioprosthetic tissue calcifies and degenerates progressively and rapidly in immature animals, analogous to the accelerated rate of calcification seen in young patients, with morphology similar to that seen in clinical specimens.
Surgical Techniques Acellularized xenogenic pulmonary valve conduits were implanted into 6 lambs (age 10 –12 weeks, weight 25–30 kg), in all valve conduits the storage solution was evaluated for bacterial colonalization. Anesthesia was induced with 30 mg/kg of ketamine, and maintained with an intravenous bolus injection of 2 mg/kg propofol. The heart was exposed by a left anterolateral thoracotomy entering the chest through the fourth intercostal space. Systemic anticoagulation was induced with heparin (400 IU/kg). By means of femoral arterial and right atrial venous cannulation, normothermic cardiopulmonary bypass was established. On bypass, 0.01 mg/kg fentanyl and 0.02 mg/kg pancuronium were administered. With the heart beating, the pulmonary artery was transected, and a segment of the main pulmonary artery and all three native leaflets were removed. The valve conduits were implanted in the orthotopic position by using running 5-0 monofilament sutures (Prolene; Ethicon, Inc, Somerville, NJ), thereby the subvalvular muscle of the conduits were in contact with recipient RV outflow tract muscle. For implantation of the conduits only a small rim of subvalvular muscle tissue was left in place, care was taken to keep the amount of subvalvular muscle identical from conduit to conduit. Heparin was reversed with 300 IU of protamine per kilogram after weaning from bypass. The thoracic wall was closed in layers using resorbable sutures, and an intercostal nerve block with 0.25% bupivacaine was administered. No further anticoagulation was given. All animals received 1000 mg of cefazolin for the first postoperative week on a daily basis. For pain control, intramuscular buprenorphin injections were performed for the first 3 days and thereafter as necessary. All animals received humane care in compliance with the “Guide for the Care and Use of Laboratory Animals” published by the National Institutes of Health (National Institutes of Health publication No. 85-23, revised 1985, Bethesda, MD). After 10 days in the Medizinische Hochschule Hannover research facilities, the animals were moved to an off-site indoor housing facility.
Ann Thorac Surg 2003;75:1457– 63
Postoperative Evaluation The animals were evaluated by echocardiography immediately after the operation, and 12 and 24 weeks thereafter. The animals were sacrificed at week 24 or earlier when severe calcification of the valve conduit became evident by transthoracic echocardiography. After termination specimens of the valve leaflets and conduit wall were prepared for histology and immunhistochemistry.
Echocardiography Epicardial echocardiography examinations were performed with a Hewlett-Packard Sonos 5500 Cardiac Imager (Palo Alto, CA) equipped with a 7.5-MHz phasedarrayed transducer to evaluate valve function. Twodimensional echocardiography color Doppler examination of the right ventricular outflow tract, conduit, and distal main pulmonary artery was performed. Qualitative evaluation of pulmonary valve competence was made by using color flow Doppler mapping: 0 ⫽ none, I° ⫽ mild, II° ⫽ moderate, and III° ⫽ severe pulmonary valve regurgitation.
Tissue Analysis Five specimens of each valve at each region (muscle, cups, wall) were used for histology. Thus, 15 specimens for each valve were examined. Native ovine and porcine pulmonary valves were evaluated to compare their cellular and extracellular matrix composition. After in vivo repopulation the explanted pulmonary valve conduits were evaluated macroscopically and histologically. For determination of cellular and extracellular components, as well as calcium content of the pulmonary valve conduits, histologic staining and histochemical assays were performed. Histologic characterization and semiquantification of valve conduit matrix components were performed by means of standard hematoxylin-eosin, movat-pentachrom, and van Giesson staining. For semiquantitative analysis the xenogenic valve conduits were compared with native ovine pulmonary valve conduits. Microscopic calcification within the conduit wall and valve leaflet was semiquantitatively analyzed (van Kossa staining).
HISTOLOGY.
IMMUNOHISTOLOGY. Snap frozen conduit wall and leaflet specimen were used for immunohistologic analysis. Immunhistochemical staining for endothelial cell characterization was performed using the avidin-biotinperoxidase technique. Endothelial cells were characterized by the presence of factor VIII related antigens (primary antibody, von Willebrandt factor; clone 8/86, DAKO, Hamburg, Germany). Leukocytes were characterized by the presence of CD11b-related antigen (primary antibody, CD11b monoclonal antibody, clone MM12A; VMRD, Inc, Pullman, WA). A goat anti-mouse antibody (DAKO) served as a secondary antibody. Streptavidinperoxidase conjugate was applied, and final staining was performed with diaminobenzidine (DAKO). For semiquantitative analysis of myocytes, fibroblasts xenogenic valve conduits were compared with native ovine pulmonary valve conduits. For characterization and semiquantitative analysis of myocytes and fibro-
LEYH ET AL XENOGENIC MATRIX CONDUITS IN PULMONARY CIRCULATION
1459
blasts a double immunofluoresecence technique with monoclonal antibodies against myocytes (monoclonal desmin antibody, clone C-18; Santa Cruz Biotechnology, Santa Cruz, CA) and fibroblasts (polyclonal vimentin antibody clone Vim H-84; Santa Cruz Biotechnology) were applied. The following protocol was used for immunofluorescence labeling: the tissue slides were thawed, rinsed with PBS, and incubated with 0.1% Triton X-100 (Sigma Chemical Inc, St. Lous, MD) in PBS. After rinsing in PBS ⫹ 0.5% bovine serum albumin (BSA, Sigma) three times the tissue slides were incubated with the primary antibody for 1 hour at room temperature. The secondary fluorophore-labeled rabbit anti-goat antibody against monoclonal desmin antibody or polyclonal vimentin antibody (Alexa Fluor 488 or Alexa Fluor 594; MoBiTec, Go¨ ttingen, Germany) was applied for 45minutes after three rinsing steps with BPS ⫹ 0.5% BSA. The cells were visualised by use of a conofocal scanning microscope (LSM 510; Carl Zeiss, Ober Kochen, Germany). For immunohistologic analysis positive controls consisted of cultured ovine endothelial cell and myofibroblast cytospot controls. Negative controls were simultaneously performed. Normal mouse or rabbit serum served as a primary antibody and a specific mouse or rabbit antibody served as the secondary antibody.
Results All animals survived the surgical procedure and recovered uneventfully. Evaluation of the storage solution for bacterial colonalization was negative for all pulmonary valve conduits. All animals receiving an AVMC were terminated at 12 weeks, whereas the animals receiving a XVMC were terminated at 24 weeks after the operation.
Fig 1. (Top) Valve morphology. Photograph of an explanted xenogenic valve conduit 24 weeks after implantation illustrating normal valve morphology. (Bottom) Photograph of an explanted allogenic valve conduit 12 weeks after implantation illustrating severe calcification.
Macroscopic Valve Conduit Morphology
and valve leaflets, resembling endothelial cells after 12 weeks of implantation. Staining with the von Willebrandt factor confirmed this observation. In contrast, the XVMC displayed a confluent endothelial monolayer of the valve leaflets and conduit wall after 24 weeks, as indicated by hematoxylin-eosin and positive von Willebrandt factor staining (Fig 3). The interstitium of AVMC was repopulated by only very few fibroblast and myocytes as indicated by hematoxylin-eosin and vimentin-desmin staining. However, the interstitium of XVMC indicated a fibroblast and myocyte population, which was similar to native ovine valvular tissue. The Movat-pentachrom stain, which stains collagen yellow, and glycosaminoglycans and proteoglycans blue-green, as well as the van Gieson stain, which stains elastin red, revealed reduced amounts of stainable collagen, and elastin, and an increased amount of glycosaminoglycans and proteoglycans in AVMC. In contrast in XVMC amounts of stainable collagen, glycosaminoglycans and proteoglycans, and elastin were similar to native ovine valvular tissue, indicating active matrix synthesis (Fig 4). All AVMC characterized macroscopic calcification after 12 weeks. Microscopically severe calcification was detected within the matrix adjacent to areas with reduced cellular population. In contrast, XVMC revealed only trace micro-
Gross macroscopic evaluation of the explanted valve conduits revealed valvular and subvalvular calcification in all AVMC after 12 weeks of implantation, the leaflets were degenerated, shrunken, and severely calcified resulting in combined valve pathology. In contrast, all XMCV appeared macroscopically to be within normal ranges with no signs of gross calcification after 24 weeks of implantation (Fig 1).
Echocardiography Mild to moderate pulmonary valve regurgitation was observed in all AVMC after 12 weeks. At this time severe calcification of the conduits became evident. Pulmonary valve regurgitation was not observed in the XVMC after 12 and 24 weeks, furthermore no calcification was detected on echocardiography.
Histology and Immunohistology The acellularization process resulted in a complete cell loss of AVMCs and XVMCs in all sections (muscle, wall, cusps) (Fig 2). The histologic and immunohistologic results of explanted valve conduits are depicted in Table 1 and 2. The hematoxylin-eosin staining of AVMC revealed a patchy focal flattened cell clustering of the conduit wall
CARDIOVASCULAR
Ann Thorac Surg 2003;75:1457– 63
CARDIOVASCULAR
1460
LEYH ET AL XENOGENIC MATRIX CONDUITS IN PULMONARY CIRCULATION
Ann Thorac Surg 2003;75:1457– 63
fibers of the endothelial cell monolayer. Calcification was localized within the matrix in areas with reduced cellular population. Inflammatory infiltration of leukocytes was not detected in any explanted conduit.
Comment
Fig 2. Hematoxylin-eosin–stained section of fresh ovine (top) and decellularized ovine (bottom) pulmonary artery wall. (Top) Native cell density of pulmonary valve leaflets. (Bottom) Native cells have been removed from the pulmonary artery wall after trypsin/EDTA acellularization.
scopic valvular calcification in one animal and trace conduit wall calcification in two animals after 24 weeks, verified by van Kossa staining (Fig 5), in all conduits (AVMC, XVMC) calcification were detected between the
The basic concept of heart valve tissue engineering is the use of preformed scaffold material with in vitro or in vivo seeding of cellular tissue components. Previous studies focused on biodegradable polymer scaffolds [1, 2, 6, 7] or allogenic acellular matrix scaffolds [4, 5] for heart valve tissue engineering. Major disadvantages of biodegradable polymer scaffolds are that natural three-dimensional (3D) structures have to be reconstructed and that important extracellular matrix proteins are not present in synthetic polymers, eg, natural ligands of cellular attachment for integrin receptor binding [8, 9]. Thus, acellularized xenogenic and allogenic tissues, maintaining natural 3D structures and extracellular matrix composition are a promising alternative to biodegradable polymer scaffolds. Bader and coworkers [10] demonstrated physiologic cellular adhesive properties of endothelial cells on an acellularized matrix.[10] Furthermore, acellularized biological matrices have been successfully used for tissue engineering of heart valves, blood vessels, and urinary bladder [4, 11, 12]. In previous experiments we used in vitro repopulated AVMC for successful heart valve tissue engineering [4, 5]. However, the concept of repopulation is still a matter of intense debate. In general two different concepts to generate bioartificial prosthesis exist: in vitro repopulation and in vivo repopulation of acellularized biological scaffold on the other side [13]. Our result demonstrated the successful in vivo repopulation of acellularized heart valve conduits. However, here we provide evidence that the source of acellularized valve matrix conduits (allogeneic/xenogeneic) influence in vivo repopulation and early calcification. We demonstrated that porcine acellularized XVMC are repopulated in vivo with valve-specific cell types within 24 weeks resembling native valve tissue and that these conduits are resistant to early calcification
Table 1. Histologic Examination of Conduit Wall and Valve Leaflets Conduit Wall Valve Leaflet Follow-Up Animals (weeks) SC IC Collagen Elastin Proteoglycan Calcification SC IC Collagen Elastin Proteoglycan Calcification XA-6 XA-7 XA-8 AA-1 AA-2 AA-3
24 24 24 12 12 12
⫹ ⫹⫹ 0 – 0 0 – — – — – —
0 – – – – –
0 0 0 – – –
0 ⫹ 0 ⫹ ⫹ ⫹
0 ⫹ ⫹ ⫹⫹ ⫹⫹⫹ ⫹⫹
0 0 0 – – –
0 0 0 — — —
0 0 — – – –
⫹ ⫹ 0 – – –
0 0 0 ⫹ ⫹ ⫹
0 ⫹ 0 ⫹⫹⫹ ⫹⫹⫹ ⫹⫹⫹
Histological examination refers to presence of SC (multilayer, ⫹; monolayer, 0; focal, –; absent, —), IC (⬎ 30% more cells compared to native, ⫹⫹; 0%–30% more cells compared to native, ⫹; comparable to native, 0; 0%–30% less cells compared to native, –; ⬎ 30% less cells compared to native, —), collagen content (⬎ 30% more compared to native, ⫹⫹; 0%–30% more compared to native, ⫹, comparable to native, 0; 0%–30% less compared to native, –; ⬎ 30% less compared to native, —), elastin content (⬎ 30% more compared to native, ⫹⫹; 0%–30% more compared to native, ⫹, comparable to native, 0; 0%–30% less compared to native, –; ⬎ 30% less compared to native, —), proteoglycan (⬎ 30% more compared to native, ⫹⫹; 0%–30% more compared to native, ⫹, comparable to native, 0; 0%–30% less compared to native, –; ⬎ 30% less compared to native, —), and calcification (0, no calcification; ⫹, few calcified spots; ⫹⫹, small areas of calcification; ⫹⫹⫹, large confluent areas of calcification). AA ⫽ allogenic acellularized;
IC ⫽ intestitial cells;
SC ⫽ superficial cells;
XA ⫽ xenogenic acellularized.
LEYH ET AL XENOGENIC MATRIX CONDUITS IN PULMONARY CIRCULATION
1461
Table 2. Immunohistochemical Examination of Conduit Wall and Valve Leaflets Conduit Wall Animals XA-6 XA-7 XA-8 AA-1 AA-2 AA-3
Follow-Up (weeks) 24 24 24 12 12 12
Valve Leaflet
vWF
Fibroblasts (%)
Myocytes (%)
CD11b
⫹ ⫹ ⫹ ⫹ ⫹ ⫹
95 95 95 25 25 25
90 90 90 5 5 5
⫺ ⫺ ⫺ ⫺ ⫺ ⫺
vWF
Fibroblasts (%)
Myocytes (%)
CD11b
⫹ ⫹ ⫹ ⫹ ⫹ ⫹
95 95 95 25 25 25
90 90 90 5 5 5
⫺ ⫺ ⫺ ⫺ ⫺ ⫺
Endothelial staining by vWF (positive for endothelial cells, ⫹; negative for endothelial cells, ⫺), fibroblast cell content compared to native ovine pulmonary valve tissue, myocytes cell content compared to native ovine pulmonary valve tissue, inflammatory infiltrates by CD11 staining (positive for leukocytes, ⫹; negative for leukocytes, ⫺). AA ⫽ allogenic acellularized;
vWF ⫽ von Willebrandt factor;
XA ⫽ xenogenic acellularized.
during in vivo reseeding. In contrast, AVMC were only partially repopulated in vivo with occurrence of early calcification within 12 weeks after implantation. The xenogenic (porcine) matrix was repopulated with all three vascular cell types (endothelial, myocytes, and
Fig 3. Hematoxylin-eosin and von Willebrandt factor staining of explanted xenogenic pulmonary valve conduits after 24 weeks (top) illustrating endothelial cell monolayer and normal extracellular matrix cellularity. (Bottom) Explanted allogenic valve conduit 12 weeks after implantation revealing focal endothelial cell lining and a reduced cellular repopulation of the extracellular matrix.
fibroblasts) in an approximate natural proportion. The surface of the allogeneic (ovine) matrix on the other hand was only partially repopulated with endothelial cells. Within the matrix the number of cells was reduced, however, fibroblasts predominated and only very few myocytes were detectable. We speculate that this phenomenon is the consequence of different extracellular matrix microenvironments of different biological matrices. The acellularization process used in this study resulted in a complete cell loss of the biological scaffold (allogeneic/xenogeneic) with preservation of the specific architecture of the leaflet extracellular matrix, ie, the fibrosa, spongiosa, and ventricularis. The acellularization process has been previously validated in our laboratory demonstrating complete absence of cellular components as well as maintenance of matrix integrity by means of scanning electron microscopy [14]. Recently, we have demonstrated that acellular implanted scaffold contains up to 2% of DNA per weight unit [15]. Furthermore, cell
Fig 4. Movat-pentachrome staining of an explanted xenogenic pulmonary valve conduit after 24 weeks (left) revealing normal amounts of stainable collagen (yellow), glycosaminoglycans and proteoglycans (blue-green), and elastin similar to native ovine valvular tissue. (Right) Explanted allogenic valve conduit 12 weeks after implantation, demonstrating a reduced amount of stainable collagen and elastin, and an increased amount of glycosaminoglycans and proteoglycans.
CARDIOVASCULAR
Ann Thorac Surg 2003;75:1457– 63
CARDIOVASCULAR
1462
LEYH ET AL XENOGENIC MATRIX CONDUITS IN PULMONARY CIRCULATION
Ann Thorac Surg 2003;75:1457– 63
allogeneic valve conduits implies that immigrated vascular specific cells might have died off due to changes in scaffold microenvironment. This is supported by the fact that fibroblasts that are more resistant to microenvironmental changes predominated the allogeneic extracellular matrix after explantation. All these factors might be involved in the early calcification process and incomplete repopulation of the AVMC. However, both biological scaffolds (AVMC and XVMC) were acellularized with the same technique, thus species individual extracellular matrix composition of the native pulmonary valve could be a factor. However, semiquantitative analysis of the extracellular matrix composition revealed more glycosaminoglycans and proteoglycane in the ovine matrix. Whether this explains our results remains speculative. Furthermore, the acellularization process might have damaged the ovine extracellular matrix components and left the porcine extracellular matrix components intact, implicating a species-specific extracellular matrix component damaging effect of trypsin.
Limitation
Fig 5. Illustration of von Kossa staining of an explanted xenogenic pulmonary valve conduit after 24 weeks (top) revealing a few spots of calcification only within the valve leaflet. (Bottom) Severe calcification within the valvular leaflet matrix adjacent to areas with reduced cellular population.
debris might be trapped between the elastic laminae [16]. It has been reported that cell debris as well as organic phosphorus induces early calcification [17–19]. The source of organic phosphorus is numerous and might derive from cell debris and/or fragmented DNA. However, this cannot explain the varying occurrence of calcification in allogeneic and xenogeneic valve conduits. Assuming that both scaffolds (allogeneic/xenogeneic) are repopulated initially after implantation the combination of less repopulation and severe early calcification seen in
The authors wish to address several limitations of this study: (1) although the gross macroscopic appearance revealed no differences between the ovine and porcine pulmonary valve root, minor differences could have an influence on our results; (2) differences between ovine and porcine collagen type and collagen fiber arrangements in the pulmonary root were not determined in this study; however, studies comparing ovine and porcine tissue concerning collagen type and collagen fiber arrangements are sparse. According to the literature species-specific differences are present, these differences were only described for medial tibial plateau and pericardium, whether this is true for cardiac valve tissue remains unanswered. Further studies addressing this issue are mandatory because the knowledge about differences in collagen type and fiber arrangement are important for heart valve tissue engineering using biological scaffolds [20, 21]. In conclusion, in vivo repopulation of acellularized biological heart valve scaffold with valve-specific cell types resembling native valve tissue occurs within 24 weeks after implantation. However, the source of the biological matrix as well as the acellularization technique itself may impact on in vivo repopulation. Further studies comparing different acellularization techniques, e.g., static vs. dynamic acellularization, with different biological scaffolds are needed to elucidate the impact of the acellularization process on the onset of early calcification and quality of repopulation.
References 1. Shinoka T, Shum-Tim D, MA PX, et al. Creation of a viable pulmonary artery autograft through tissue engineering. J Thorac Cardiovasc Surg 1998;115:536 –46. 2. Stock U, Nagaashima M, Khalil PN, et al. Tissue-engineered valve conduits in the pulmonary circulation. J Thorac Cardiovasc Surg 2000;119:732–40.
LEYH ET AL XENOGENIC MATRIX CONDUITS IN PULMONARY CIRCULATION
3. Wilson GJ, Courtman DW, Klement P, Lee JM, Yeger H. Acellular matrix: a biomaterials approach for coronary artery bypass and heart valve replacement. Ann Thorac Surg 1995;60:S353–8. 4. Steinhoff G, Stock U, Karim N, et al. Tissue engineering of pulmonary heart valves on allogenic acellular matrix conduits: in vivo restoration of valve tissue. Circulation 2000;102: III 50 –5. 5. Mertsching H, Leyh R, Rebe P, et al. Tissue engineering of autologous heart valves—results of 3, 6 and 9 months implantation in a growing sheep model. EACTS/ESTS Joint Meeting. Lisbon, 2001. 6. Shinoka T, Breuer C, Anel R, et al. Tissue-engineering heart valves: valve leaflet replacement in a lamb model. Ann Thorac Surg 1995;60:513–6. 7. Hoerstrup SP, Sodian R, Daebritz S, et al. Functional living trileaflet heart valves grown in vitro. Circulation 2000;102:III 44 –9. 8. Cheresh DA, Berliner SA, Vicente V, Ruggeri ZM. Recognition of distinct adhesive sites on fibrinogen by related integrins on platelets and endothelial cells. Cell 1989;58:945– 53. 9. Joshi P, Chung CY, Aukhil I, Erickson HP. Endothelial cells adhere to the RGD domain and the fibrinogen-like terminal knob of tenascin. J Cell Sci 1993;106:389 –400. 10. Bader A, Schilling T, Teebken OE, et al. Tissue engineering of heart valves— human endothelial cell seeding of detergent acellularized porcine valves. Eur J Cardiothorac Surg 1998;14:279 –84. 11. Nerem RM, Selikatar D. Vascular tissue engineering. Annu Rev Biomed Eng 2001;3:225–43. 12. Atala A. Tissue engineering for bladder substitution. World J Urol 2000;18:364 –70.
1463
13. O’Brien MF, Goldstein S, Walsh S, Black KS, Elkins R, Clarke D. The SynerGraft valve: a new acellular (nonglutaraldehyde-fixed) tissue heart valve for autologous recellularization first experimental studies before clinical implantation. Semin Thorac Cardiovasc Surg 1999;11(Suppl):194 –20. 14. Teebken OE, Bader A, Steinhoff G, Haverich A. Tissue engineering of vascular grafts: human cell seeding of decellularised porcine matrix. Eur J Vasc Endovasc Surg 2000;19: 381–6. 15. Walles T, Ciubotaru S, Sorrentino S, et al. Cardiovascular tissue engineering: importance of scaffold matrix composition and scaffold thickness. J Am Coll Cardiol (in press). 16. Zeltinger J, Landeen LK, Alexander HG, Kidd ID, Sibanda B. Development and characterization of tissue-engineered aortic valves. Tissue Eng 2001;7:9 –22. 17. Schoen FJ, Golomb G, Levy RJ. Calcification of bioprosthetic heart valves: a perspective on models. J Heart Valve Dis 1992;1:110 –4. 18. Levy RJ, Schoen FJ, Levy JT, Nelson AC, Howard SL, Oshry LJ. Biologic determinants of dystrophic calcification and osteocalcin deposition in glutaraldehyde-preserved porcine aortic valve leaflets implanted subcutaneously in rats. Am J Pathol 1983;113:143–55. 19. Schoen FJ, Hirsch D, Bianco RW, Levy RJ. Onset and progression of calcification in porcine aortic bioprosthetic valves implanted as orthotopic mitral valve replacements in juvenile sheep. J Thorac Cardiovasc Surg 1994;108:880 –7. 20. Kaab MJ, Gwynn IA, Notzli HP. Collagen fibre arrangement in the tibial plateau articular cartilage of man and other mammalian species. J Anat 1998;193:23–34. 21. Naimark WA, Lee JM, Limeback H, Cheung DT. Correlation of structure and viscoelastic properties in the pericardia of four mammalian species. Am J Physiol 1992;263:H1095–106.
DISCUSSION DR DAVID CLARKE (Denver, CO): I would like to congratulate you on the performance of these experiments but I must question the results you observed. Experiences that I am familiar with have used the model you describe. Although the results with the xenografts are almost identical to yours in that repopulation occurs, there is measurable calcification, the acellular allogenic tissue that was implanted has shown virtually no calcification and at least as good, if not better, repopulation than the xenogenic tissue. How can you account for this difference that you observed in terms of the dramatically worse results with the allogenic tissue? Not only do your results seem intuitively unreasonable, but also, evidence exists which indicate that they may not be correct. DR LEYH: Well, actually I expected this question. It is a very simple question to ask but a very difficult question to answer. We assume that both scaffold materials were initially repopulated during or after implantation. Looking at the fact that we had severe calcification and reduced cell population in the allogenic mat rix only, we believe that these migrated cells just died due to changes in the scaffold microenvironment. The reason for this can be the acellularization process itself, because in this study we used a static acellularization technique. We believe that a dynamic acellularization process will result in more physiologic acellularization with less damage to the extracellular matrix. DR CLARKE: Were your valves cryopreserved during the process prior to implantation?
DR LEYH: No. DR GINO GEROSA (Padova, Italy): Did you observe any major differences in terms of matrix composition among the pig valves and the sheep valves? DR LEYH: No, we did not in terms of the examination we did. There were no differences in matrix composition seen on histologic examination, but we did not perform an electromicroscopic examination, and I think that we should find differences if we do this kind of examination. DR GEROSA: Do you think this can play a major role in terms of onset of calcification? DR LEYH: I do, because I think if we do use a sort of acellularization process, which might damage the scaffold material, this will result in early calcific ation. So we have to keep this in mind. DR MARK O’BRIEN (Brisbane, Australia): Could you make a comment on any differences that you may have seen in the conduit wall? DR LEYH: We saw calcification in the conduit wall in the xenogenic group as well as in the allogenic group, but, again, in the allogenic group calcification was pronounced compared to the xenogenic group, and we believe that the thickness of the conduit wall may be a factor influencing repopulation and, thus, calcification.
CARDIOVASCULAR
Ann Thorac Surg 2003;75:1457– 63
Background. Approaches to in vivo repopulation of acellularized valve matrix constructs have been described recently. However, early calcification of acellularized matrices repopulated in vivo remains a major obstacle. We hypothesised that the matrix composition has a significant influence on the onset of early calcification. Therefore, we evaluated the calcification of acellularized allogenic ovine (AVMC) and xenogenic porcine (XVMC) valve matrix conduits in the pulmonary circulation in a sheep model. Methods. Porcine (n ⴝ 3) and sheep (n ⴝ 3) pulmonary valve conduits were acellularized by trypsin/EDTA digestion and then implanted into healthy sheep in pulmonary valve position using extracorporeal bypass support. Transthoracic echocardiography (TTE) was performed at 12 and 24 weeks after the implantation. The animals were sacrificed at week 24 or earlier when severe calcification of the valve conduit became evident by TTE. The valves were examined histologically and biochemically.
Results. All AVMC revealed severe calcification after 12 weeks with focal endothelial cell clustering and no interstitial valve tissue reconstitution. In contrast, after 24 weeks XVMC indicated mild calcification on histologic examination (von Kossa staining) with histologic reconstitution of valve tissue and confluent endothelial surface coverage. Furthermore, immunohistologic analysis revealed reconstitution of surface endothelial cell monolayer (von Willebrand factor), and interstitial myofibroblasts (Vimentin/Desmin). Conclusions. Porcine acellularized XVMC are resistant to early calcification during in vivo reseeding. Furthermore, XVMC are repopulated in vivo with valve-specific cell types within 24 weeks resembling native valve tissue.
M
matrix conduits after in vitro repopulation with autologous myofibroblast and endothelial cells [4, 5]. In vitro repopulation with autologous myofibroblast and endothelial cells resulted in complete reconstitution of functioning viable heart valve tissue. In contrast, in vivo repopulation was incomplete with subsequent early calcification. Whether this is true for biological matrices from other species has yet to be determined. We hypothesized that the matrix composition has a significant impact on the onset of early calcification of these tissue engineered heart valves. Therefore, we evaluated the calcification of acellularized allogenic ovine (AVMC) and xenogenic porcine (XVMC) valve matrix conduits after in vivo repopulation in the pulmonary circulation in a sheep model.
echanical and biological heart valve prostheses are nonviable constructs with the inherent drawbacks, that these structures have no growing, repairing, and remodelling ability. Tissue engineered heart valves may have the potential to eliminate the disadvantages of the commonly used mechanical and biological heart valve prostheses. The concept of tissue engineering is to create an isogenic tissue, which is based on an anatomically shaped matrix, repopulated with viable cellular tissue components. Biodegradable polymer scaffolds have been used to reconstruct valve conduits, which resulted in a partial cellular and functional reconstitution of valve tissue [1, 2]. However, limitations in cellular adhesion and tissue regeneration are an unsolved problem when biodegradable polymer scaffolds are used [2]. The concept of a biological scaffold material for heart valve tissue engineering might be an alternative to overcome these obstacles [3]. Recently, we proved the concept of tissue engineering of valve conduits on allogenic ovine Presented at the Thirty-eighth Annual Meeting of The Society of Thoracic Surgeons, Fort Lauderdale, FL, Jan 28 –30, 2002. Address reprint requests to Dr Leyh, Division of Thoracic and Cardiovascular Surgery, Hannover Medical School, Carl Neuberg St. 1, 30623 Hannover, Germany; e-mail: [email protected].
© 2003 by The Society of Thoracic Surgeons Published by Elsevier Science Inc
(Ann Thorac Surg 2003;75:1457– 63) © 2003 by The Society of Thoracic Surgeons
Material and Methods Acellularization of Pulmonary Valve Conduits Porcine pulmonary valve conduits (n ⫽ 3) were obtained from pigs (German landrace), ranging from 25 to 30 kg, and sheep (n ⫽ 3), ranging from 25 to 30 kg, from the Tierzuchanstalt, Mariensee, Germany. Hearts were obtained under sterile conditions, and pulmonary valve conduits were harvested with a thin ridge of subvalvular 0003-4975/03/$30.00 PII S0003-4975(02)04845-2
CARDIOVASCULAR
In Vivo Repopulation of Xenogeneic and Allogeneic Acellular Valve Matrix Conduits in the Pulmonary Circulation
CARDIOVASCULAR
1458
LEYH ET AL XENOGENIC MATRIX CONDUITS IN PULMONARY CIRCULATION
muscle tissue proximally and a short segment of the truncus pulmonalis distally. The macroscopic anatomy of ovine and porcine pulmonary valve conduits did not reveal any gross differences. The valve conduits were stored in Hanks buffered saline solution (HBSS; Biochrom, Cambridge, UK) at 4°C, and incubated with 0.05% trypsin (Biochrom) and 0.02% EDTA (Biochrom) for 48 hours, followed by PBS flushing for 48 hours to remove all cellular debris. Samples of the conduits were taken before and after this treatment to document acellularization by hematoxylin-eosin and movat-pentachrom staining using light microscopy.
Animal Model We used an immature sheep model to test our biological heart valve scaffolds because bioprosthetic tissue calcifies and degenerates progressively and rapidly in immature animals, analogous to the accelerated rate of calcification seen in young patients, with morphology similar to that seen in clinical specimens.
Surgical Techniques Acellularized xenogenic pulmonary valve conduits were implanted into 6 lambs (age 10 –12 weeks, weight 25–30 kg), in all valve conduits the storage solution was evaluated for bacterial colonalization. Anesthesia was induced with 30 mg/kg of ketamine, and maintained with an intravenous bolus injection of 2 mg/kg propofol. The heart was exposed by a left anterolateral thoracotomy entering the chest through the fourth intercostal space. Systemic anticoagulation was induced with heparin (400 IU/kg). By means of femoral arterial and right atrial venous cannulation, normothermic cardiopulmonary bypass was established. On bypass, 0.01 mg/kg fentanyl and 0.02 mg/kg pancuronium were administered. With the heart beating, the pulmonary artery was transected, and a segment of the main pulmonary artery and all three native leaflets were removed. The valve conduits were implanted in the orthotopic position by using running 5-0 monofilament sutures (Prolene; Ethicon, Inc, Somerville, NJ), thereby the subvalvular muscle of the conduits were in contact with recipient RV outflow tract muscle. For implantation of the conduits only a small rim of subvalvular muscle tissue was left in place, care was taken to keep the amount of subvalvular muscle identical from conduit to conduit. Heparin was reversed with 300 IU of protamine per kilogram after weaning from bypass. The thoracic wall was closed in layers using resorbable sutures, and an intercostal nerve block with 0.25% bupivacaine was administered. No further anticoagulation was given. All animals received 1000 mg of cefazolin for the first postoperative week on a daily basis. For pain control, intramuscular buprenorphin injections were performed for the first 3 days and thereafter as necessary. All animals received humane care in compliance with the “Guide for the Care and Use of Laboratory Animals” published by the National Institutes of Health (National Institutes of Health publication No. 85-23, revised 1985, Bethesda, MD). After 10 days in the Medizinische Hochschule Hannover research facilities, the animals were moved to an off-site indoor housing facility.
Ann Thorac Surg 2003;75:1457– 63
Postoperative Evaluation The animals were evaluated by echocardiography immediately after the operation, and 12 and 24 weeks thereafter. The animals were sacrificed at week 24 or earlier when severe calcification of the valve conduit became evident by transthoracic echocardiography. After termination specimens of the valve leaflets and conduit wall were prepared for histology and immunhistochemistry.
Echocardiography Epicardial echocardiography examinations were performed with a Hewlett-Packard Sonos 5500 Cardiac Imager (Palo Alto, CA) equipped with a 7.5-MHz phasedarrayed transducer to evaluate valve function. Twodimensional echocardiography color Doppler examination of the right ventricular outflow tract, conduit, and distal main pulmonary artery was performed. Qualitative evaluation of pulmonary valve competence was made by using color flow Doppler mapping: 0 ⫽ none, I° ⫽ mild, II° ⫽ moderate, and III° ⫽ severe pulmonary valve regurgitation.
Tissue Analysis Five specimens of each valve at each region (muscle, cups, wall) were used for histology. Thus, 15 specimens for each valve were examined. Native ovine and porcine pulmonary valves were evaluated to compare their cellular and extracellular matrix composition. After in vivo repopulation the explanted pulmonary valve conduits were evaluated macroscopically and histologically. For determination of cellular and extracellular components, as well as calcium content of the pulmonary valve conduits, histologic staining and histochemical assays were performed. Histologic characterization and semiquantification of valve conduit matrix components were performed by means of standard hematoxylin-eosin, movat-pentachrom, and van Giesson staining. For semiquantitative analysis the xenogenic valve conduits were compared with native ovine pulmonary valve conduits. Microscopic calcification within the conduit wall and valve leaflet was semiquantitatively analyzed (van Kossa staining).
HISTOLOGY.
IMMUNOHISTOLOGY. Snap frozen conduit wall and leaflet specimen were used for immunohistologic analysis. Immunhistochemical staining for endothelial cell characterization was performed using the avidin-biotinperoxidase technique. Endothelial cells were characterized by the presence of factor VIII related antigens (primary antibody, von Willebrandt factor; clone 8/86, DAKO, Hamburg, Germany). Leukocytes were characterized by the presence of CD11b-related antigen (primary antibody, CD11b monoclonal antibody, clone MM12A; VMRD, Inc, Pullman, WA). A goat anti-mouse antibody (DAKO) served as a secondary antibody. Streptavidinperoxidase conjugate was applied, and final staining was performed with diaminobenzidine (DAKO). For semiquantitative analysis of myocytes, fibroblasts xenogenic valve conduits were compared with native ovine pulmonary valve conduits. For characterization and semiquantitative analysis of myocytes and fibro-
LEYH ET AL XENOGENIC MATRIX CONDUITS IN PULMONARY CIRCULATION
1459
blasts a double immunofluoresecence technique with monoclonal antibodies against myocytes (monoclonal desmin antibody, clone C-18; Santa Cruz Biotechnology, Santa Cruz, CA) and fibroblasts (polyclonal vimentin antibody clone Vim H-84; Santa Cruz Biotechnology) were applied. The following protocol was used for immunofluorescence labeling: the tissue slides were thawed, rinsed with PBS, and incubated with 0.1% Triton X-100 (Sigma Chemical Inc, St. Lous, MD) in PBS. After rinsing in PBS ⫹ 0.5% bovine serum albumin (BSA, Sigma) three times the tissue slides were incubated with the primary antibody for 1 hour at room temperature. The secondary fluorophore-labeled rabbit anti-goat antibody against monoclonal desmin antibody or polyclonal vimentin antibody (Alexa Fluor 488 or Alexa Fluor 594; MoBiTec, Go¨ ttingen, Germany) was applied for 45minutes after three rinsing steps with BPS ⫹ 0.5% BSA. The cells were visualised by use of a conofocal scanning microscope (LSM 510; Carl Zeiss, Ober Kochen, Germany). For immunohistologic analysis positive controls consisted of cultured ovine endothelial cell and myofibroblast cytospot controls. Negative controls were simultaneously performed. Normal mouse or rabbit serum served as a primary antibody and a specific mouse or rabbit antibody served as the secondary antibody.
Results All animals survived the surgical procedure and recovered uneventfully. Evaluation of the storage solution for bacterial colonalization was negative for all pulmonary valve conduits. All animals receiving an AVMC were terminated at 12 weeks, whereas the animals receiving a XVMC were terminated at 24 weeks after the operation.
Fig 1. (Top) Valve morphology. Photograph of an explanted xenogenic valve conduit 24 weeks after implantation illustrating normal valve morphology. (Bottom) Photograph of an explanted allogenic valve conduit 12 weeks after implantation illustrating severe calcification.
Macroscopic Valve Conduit Morphology
and valve leaflets, resembling endothelial cells after 12 weeks of implantation. Staining with the von Willebrandt factor confirmed this observation. In contrast, the XVMC displayed a confluent endothelial monolayer of the valve leaflets and conduit wall after 24 weeks, as indicated by hematoxylin-eosin and positive von Willebrandt factor staining (Fig 3). The interstitium of AVMC was repopulated by only very few fibroblast and myocytes as indicated by hematoxylin-eosin and vimentin-desmin staining. However, the interstitium of XVMC indicated a fibroblast and myocyte population, which was similar to native ovine valvular tissue. The Movat-pentachrom stain, which stains collagen yellow, and glycosaminoglycans and proteoglycans blue-green, as well as the van Gieson stain, which stains elastin red, revealed reduced amounts of stainable collagen, and elastin, and an increased amount of glycosaminoglycans and proteoglycans in AVMC. In contrast in XVMC amounts of stainable collagen, glycosaminoglycans and proteoglycans, and elastin were similar to native ovine valvular tissue, indicating active matrix synthesis (Fig 4). All AVMC characterized macroscopic calcification after 12 weeks. Microscopically severe calcification was detected within the matrix adjacent to areas with reduced cellular population. In contrast, XVMC revealed only trace micro-
Gross macroscopic evaluation of the explanted valve conduits revealed valvular and subvalvular calcification in all AVMC after 12 weeks of implantation, the leaflets were degenerated, shrunken, and severely calcified resulting in combined valve pathology. In contrast, all XMCV appeared macroscopically to be within normal ranges with no signs of gross calcification after 24 weeks of implantation (Fig 1).
Echocardiography Mild to moderate pulmonary valve regurgitation was observed in all AVMC after 12 weeks. At this time severe calcification of the conduits became evident. Pulmonary valve regurgitation was not observed in the XVMC after 12 and 24 weeks, furthermore no calcification was detected on echocardiography.
Histology and Immunohistology The acellularization process resulted in a complete cell loss of AVMCs and XVMCs in all sections (muscle, wall, cusps) (Fig 2). The histologic and immunohistologic results of explanted valve conduits are depicted in Table 1 and 2. The hematoxylin-eosin staining of AVMC revealed a patchy focal flattened cell clustering of the conduit wall
CARDIOVASCULAR
Ann Thorac Surg 2003;75:1457– 63
CARDIOVASCULAR
1460
LEYH ET AL XENOGENIC MATRIX CONDUITS IN PULMONARY CIRCULATION
Ann Thorac Surg 2003;75:1457– 63
fibers of the endothelial cell monolayer. Calcification was localized within the matrix in areas with reduced cellular population. Inflammatory infiltration of leukocytes was not detected in any explanted conduit.
Comment
Fig 2. Hematoxylin-eosin–stained section of fresh ovine (top) and decellularized ovine (bottom) pulmonary artery wall. (Top) Native cell density of pulmonary valve leaflets. (Bottom) Native cells have been removed from the pulmonary artery wall after trypsin/EDTA acellularization.
scopic valvular calcification in one animal and trace conduit wall calcification in two animals after 24 weeks, verified by van Kossa staining (Fig 5), in all conduits (AVMC, XVMC) calcification were detected between the
The basic concept of heart valve tissue engineering is the use of preformed scaffold material with in vitro or in vivo seeding of cellular tissue components. Previous studies focused on biodegradable polymer scaffolds [1, 2, 6, 7] or allogenic acellular matrix scaffolds [4, 5] for heart valve tissue engineering. Major disadvantages of biodegradable polymer scaffolds are that natural three-dimensional (3D) structures have to be reconstructed and that important extracellular matrix proteins are not present in synthetic polymers, eg, natural ligands of cellular attachment for integrin receptor binding [8, 9]. Thus, acellularized xenogenic and allogenic tissues, maintaining natural 3D structures and extracellular matrix composition are a promising alternative to biodegradable polymer scaffolds. Bader and coworkers [10] demonstrated physiologic cellular adhesive properties of endothelial cells on an acellularized matrix.[10] Furthermore, acellularized biological matrices have been successfully used for tissue engineering of heart valves, blood vessels, and urinary bladder [4, 11, 12]. In previous experiments we used in vitro repopulated AVMC for successful heart valve tissue engineering [4, 5]. However, the concept of repopulation is still a matter of intense debate. In general two different concepts to generate bioartificial prosthesis exist: in vitro repopulation and in vivo repopulation of acellularized biological scaffold on the other side [13]. Our result demonstrated the successful in vivo repopulation of acellularized heart valve conduits. However, here we provide evidence that the source of acellularized valve matrix conduits (allogeneic/xenogeneic) influence in vivo repopulation and early calcification. We demonstrated that porcine acellularized XVMC are repopulated in vivo with valve-specific cell types within 24 weeks resembling native valve tissue and that these conduits are resistant to early calcification
Table 1. Histologic Examination of Conduit Wall and Valve Leaflets Conduit Wall Valve Leaflet Follow-Up Animals (weeks) SC IC Collagen Elastin Proteoglycan Calcification SC IC Collagen Elastin Proteoglycan Calcification XA-6 XA-7 XA-8 AA-1 AA-2 AA-3
24 24 24 12 12 12
⫹ ⫹⫹ 0 – 0 0 – — – — – —
0 – – – – –
0 0 0 – – –
0 ⫹ 0 ⫹ ⫹ ⫹
0 ⫹ ⫹ ⫹⫹ ⫹⫹⫹ ⫹⫹
0 0 0 – – –
0 0 0 — — —
0 0 — – – –
⫹ ⫹ 0 – – –
0 0 0 ⫹ ⫹ ⫹
0 ⫹ 0 ⫹⫹⫹ ⫹⫹⫹ ⫹⫹⫹
Histological examination refers to presence of SC (multilayer, ⫹; monolayer, 0; focal, –; absent, —), IC (⬎ 30% more cells compared to native, ⫹⫹; 0%–30% more cells compared to native, ⫹; comparable to native, 0; 0%–30% less cells compared to native, –; ⬎ 30% less cells compared to native, —), collagen content (⬎ 30% more compared to native, ⫹⫹; 0%–30% more compared to native, ⫹, comparable to native, 0; 0%–30% less compared to native, –; ⬎ 30% less compared to native, —), elastin content (⬎ 30% more compared to native, ⫹⫹; 0%–30% more compared to native, ⫹, comparable to native, 0; 0%–30% less compared to native, –; ⬎ 30% less compared to native, —), proteoglycan (⬎ 30% more compared to native, ⫹⫹; 0%–30% more compared to native, ⫹, comparable to native, 0; 0%–30% less compared to native, –; ⬎ 30% less compared to native, —), and calcification (0, no calcification; ⫹, few calcified spots; ⫹⫹, small areas of calcification; ⫹⫹⫹, large confluent areas of calcification). AA ⫽ allogenic acellularized;
IC ⫽ intestitial cells;
SC ⫽ superficial cells;
XA ⫽ xenogenic acellularized.
LEYH ET AL XENOGENIC MATRIX CONDUITS IN PULMONARY CIRCULATION
1461
Table 2. Immunohistochemical Examination of Conduit Wall and Valve Leaflets Conduit Wall Animals XA-6 XA-7 XA-8 AA-1 AA-2 AA-3
Follow-Up (weeks) 24 24 24 12 12 12
Valve Leaflet
vWF
Fibroblasts (%)
Myocytes (%)
CD11b
⫹ ⫹ ⫹ ⫹ ⫹ ⫹
95 95 95 25 25 25
90 90 90 5 5 5
⫺ ⫺ ⫺ ⫺ ⫺ ⫺
vWF
Fibroblasts (%)
Myocytes (%)
CD11b
⫹ ⫹ ⫹ ⫹ ⫹ ⫹
95 95 95 25 25 25
90 90 90 5 5 5
⫺ ⫺ ⫺ ⫺ ⫺ ⫺
Endothelial staining by vWF (positive for endothelial cells, ⫹; negative for endothelial cells, ⫺), fibroblast cell content compared to native ovine pulmonary valve tissue, myocytes cell content compared to native ovine pulmonary valve tissue, inflammatory infiltrates by CD11 staining (positive for leukocytes, ⫹; negative for leukocytes, ⫺). AA ⫽ allogenic acellularized;
vWF ⫽ von Willebrandt factor;
XA ⫽ xenogenic acellularized.
during in vivo reseeding. In contrast, AVMC were only partially repopulated in vivo with occurrence of early calcification within 12 weeks after implantation. The xenogenic (porcine) matrix was repopulated with all three vascular cell types (endothelial, myocytes, and
Fig 3. Hematoxylin-eosin and von Willebrandt factor staining of explanted xenogenic pulmonary valve conduits after 24 weeks (top) illustrating endothelial cell monolayer and normal extracellular matrix cellularity. (Bottom) Explanted allogenic valve conduit 12 weeks after implantation revealing focal endothelial cell lining and a reduced cellular repopulation of the extracellular matrix.
fibroblasts) in an approximate natural proportion. The surface of the allogeneic (ovine) matrix on the other hand was only partially repopulated with endothelial cells. Within the matrix the number of cells was reduced, however, fibroblasts predominated and only very few myocytes were detectable. We speculate that this phenomenon is the consequence of different extracellular matrix microenvironments of different biological matrices. The acellularization process used in this study resulted in a complete cell loss of the biological scaffold (allogeneic/xenogeneic) with preservation of the specific architecture of the leaflet extracellular matrix, ie, the fibrosa, spongiosa, and ventricularis. The acellularization process has been previously validated in our laboratory demonstrating complete absence of cellular components as well as maintenance of matrix integrity by means of scanning electron microscopy [14]. Recently, we have demonstrated that acellular implanted scaffold contains up to 2% of DNA per weight unit [15]. Furthermore, cell
Fig 4. Movat-pentachrome staining of an explanted xenogenic pulmonary valve conduit after 24 weeks (left) revealing normal amounts of stainable collagen (yellow), glycosaminoglycans and proteoglycans (blue-green), and elastin similar to native ovine valvular tissue. (Right) Explanted allogenic valve conduit 12 weeks after implantation, demonstrating a reduced amount of stainable collagen and elastin, and an increased amount of glycosaminoglycans and proteoglycans.
CARDIOVASCULAR
Ann Thorac Surg 2003;75:1457– 63
CARDIOVASCULAR
1462
LEYH ET AL XENOGENIC MATRIX CONDUITS IN PULMONARY CIRCULATION
Ann Thorac Surg 2003;75:1457– 63
allogeneic valve conduits implies that immigrated vascular specific cells might have died off due to changes in scaffold microenvironment. This is supported by the fact that fibroblasts that are more resistant to microenvironmental changes predominated the allogeneic extracellular matrix after explantation. All these factors might be involved in the early calcification process and incomplete repopulation of the AVMC. However, both biological scaffolds (AVMC and XVMC) were acellularized with the same technique, thus species individual extracellular matrix composition of the native pulmonary valve could be a factor. However, semiquantitative analysis of the extracellular matrix composition revealed more glycosaminoglycans and proteoglycane in the ovine matrix. Whether this explains our results remains speculative. Furthermore, the acellularization process might have damaged the ovine extracellular matrix components and left the porcine extracellular matrix components intact, implicating a species-specific extracellular matrix component damaging effect of trypsin.
Limitation
Fig 5. Illustration of von Kossa staining of an explanted xenogenic pulmonary valve conduit after 24 weeks (top) revealing a few spots of calcification only within the valve leaflet. (Bottom) Severe calcification within the valvular leaflet matrix adjacent to areas with reduced cellular population.
debris might be trapped between the elastic laminae [16]. It has been reported that cell debris as well as organic phosphorus induces early calcification [17–19]. The source of organic phosphorus is numerous and might derive from cell debris and/or fragmented DNA. However, this cannot explain the varying occurrence of calcification in allogeneic and xenogeneic valve conduits. Assuming that both scaffolds (allogeneic/xenogeneic) are repopulated initially after implantation the combination of less repopulation and severe early calcification seen in
The authors wish to address several limitations of this study: (1) although the gross macroscopic appearance revealed no differences between the ovine and porcine pulmonary valve root, minor differences could have an influence on our results; (2) differences between ovine and porcine collagen type and collagen fiber arrangements in the pulmonary root were not determined in this study; however, studies comparing ovine and porcine tissue concerning collagen type and collagen fiber arrangements are sparse. According to the literature species-specific differences are present, these differences were only described for medial tibial plateau and pericardium, whether this is true for cardiac valve tissue remains unanswered. Further studies addressing this issue are mandatory because the knowledge about differences in collagen type and fiber arrangement are important for heart valve tissue engineering using biological scaffolds [20, 21]. In conclusion, in vivo repopulation of acellularized biological heart valve scaffold with valve-specific cell types resembling native valve tissue occurs within 24 weeks after implantation. However, the source of the biological matrix as well as the acellularization technique itself may impact on in vivo repopulation. Further studies comparing different acellularization techniques, e.g., static vs. dynamic acellularization, with different biological scaffolds are needed to elucidate the impact of the acellularization process on the onset of early calcification and quality of repopulation.
References 1. Shinoka T, Shum-Tim D, MA PX, et al. Creation of a viable pulmonary artery autograft through tissue engineering. J Thorac Cardiovasc Surg 1998;115:536 –46. 2. Stock U, Nagaashima M, Khalil PN, et al. Tissue-engineered valve conduits in the pulmonary circulation. J Thorac Cardiovasc Surg 2000;119:732–40.
LEYH ET AL XENOGENIC MATRIX CONDUITS IN PULMONARY CIRCULATION
3. Wilson GJ, Courtman DW, Klement P, Lee JM, Yeger H. Acellular matrix: a biomaterials approach for coronary artery bypass and heart valve replacement. Ann Thorac Surg 1995;60:S353–8. 4. Steinhoff G, Stock U, Karim N, et al. Tissue engineering of pulmonary heart valves on allogenic acellular matrix conduits: in vivo restoration of valve tissue. Circulation 2000;102: III 50 –5. 5. Mertsching H, Leyh R, Rebe P, et al. Tissue engineering of autologous heart valves—results of 3, 6 and 9 months implantation in a growing sheep model. EACTS/ESTS Joint Meeting. Lisbon, 2001. 6. Shinoka T, Breuer C, Anel R, et al. Tissue-engineering heart valves: valve leaflet replacement in a lamb model. Ann Thorac Surg 1995;60:513–6. 7. Hoerstrup SP, Sodian R, Daebritz S, et al. Functional living trileaflet heart valves grown in vitro. Circulation 2000;102:III 44 –9. 8. Cheresh DA, Berliner SA, Vicente V, Ruggeri ZM. Recognition of distinct adhesive sites on fibrinogen by related integrins on platelets and endothelial cells. Cell 1989;58:945– 53. 9. Joshi P, Chung CY, Aukhil I, Erickson HP. Endothelial cells adhere to the RGD domain and the fibrinogen-like terminal knob of tenascin. J Cell Sci 1993;106:389 –400. 10. Bader A, Schilling T, Teebken OE, et al. Tissue engineering of heart valves— human endothelial cell seeding of detergent acellularized porcine valves. Eur J Cardiothorac Surg 1998;14:279 –84. 11. Nerem RM, Selikatar D. Vascular tissue engineering. Annu Rev Biomed Eng 2001;3:225–43. 12. Atala A. Tissue engineering for bladder substitution. World J Urol 2000;18:364 –70.
1463
13. O’Brien MF, Goldstein S, Walsh S, Black KS, Elkins R, Clarke D. The SynerGraft valve: a new acellular (nonglutaraldehyde-fixed) tissue heart valve for autologous recellularization first experimental studies before clinical implantation. Semin Thorac Cardiovasc Surg 1999;11(Suppl):194 –20. 14. Teebken OE, Bader A, Steinhoff G, Haverich A. Tissue engineering of vascular grafts: human cell seeding of decellularised porcine matrix. Eur J Vasc Endovasc Surg 2000;19: 381–6. 15. Walles T, Ciubotaru S, Sorrentino S, et al. Cardiovascular tissue engineering: importance of scaffold matrix composition and scaffold thickness. J Am Coll Cardiol (in press). 16. Zeltinger J, Landeen LK, Alexander HG, Kidd ID, Sibanda B. Development and characterization of tissue-engineered aortic valves. Tissue Eng 2001;7:9 –22. 17. Schoen FJ, Golomb G, Levy RJ. Calcification of bioprosthetic heart valves: a perspective on models. J Heart Valve Dis 1992;1:110 –4. 18. Levy RJ, Schoen FJ, Levy JT, Nelson AC, Howard SL, Oshry LJ. Biologic determinants of dystrophic calcification and osteocalcin deposition in glutaraldehyde-preserved porcine aortic valve leaflets implanted subcutaneously in rats. Am J Pathol 1983;113:143–55. 19. Schoen FJ, Hirsch D, Bianco RW, Levy RJ. Onset and progression of calcification in porcine aortic bioprosthetic valves implanted as orthotopic mitral valve replacements in juvenile sheep. J Thorac Cardiovasc Surg 1994;108:880 –7. 20. Kaab MJ, Gwynn IA, Notzli HP. Collagen fibre arrangement in the tibial plateau articular cartilage of man and other mammalian species. J Anat 1998;193:23–34. 21. Naimark WA, Lee JM, Limeback H, Cheung DT. Correlation of structure and viscoelastic properties in the pericardia of four mammalian species. Am J Physiol 1992;263:H1095–106.
DISCUSSION DR DAVID CLARKE (Denver, CO): I would like to congratulate you on the performance of these experiments but I must question the results you observed. Experiences that I am familiar with have used the model you describe. Although the results with the xenografts are almost identical to yours in that repopulation occurs, there is measurable calcification, the acellular allogenic tissue that was implanted has shown virtually no calcification and at least as good, if not better, repopulation than the xenogenic tissue. How can you account for this difference that you observed in terms of the dramatically worse results with the allogenic tissue? Not only do your results seem intuitively unreasonable, but also, evidence exists which indicate that they may not be correct. DR LEYH: Well, actually I expected this question. It is a very simple question to ask but a very difficult question to answer. We assume that both scaffold materials were initially repopulated during or after implantation. Looking at the fact that we had severe calcification and reduced cell population in the allogenic mat rix only, we believe that these migrated cells just died due to changes in the scaffold microenvironment. The reason for this can be the acellularization process itself, because in this study we used a static acellularization technique. We believe that a dynamic acellularization process will result in more physiologic acellularization with less damage to the extracellular matrix. DR CLARKE: Were your valves cryopreserved during the process prior to implantation?
DR LEYH: No. DR GINO GEROSA (Padova, Italy): Did you observe any major differences in terms of matrix composition among the pig valves and the sheep valves? DR LEYH: No, we did not in terms of the examination we did. There were no differences in matrix composition seen on histologic examination, but we did not perform an electromicroscopic examination, and I think that we should find differences if we do this kind of examination. DR GEROSA: Do you think this can play a major role in terms of onset of calcification? DR LEYH: I do, because I think if we do use a sort of acellularization process, which might damage the scaffold material, this will result in early calcific ation. So we have to keep this in mind. DR MARK O’BRIEN (Brisbane, Australia): Could you make a comment on any differences that you may have seen in the conduit wall? DR LEYH: We saw calcification in the conduit wall in the xenogenic group as well as in the allogenic group, but, again, in the allogenic group calcification was pronounced compared to the xenogenic group, and we believe that the thickness of the conduit wall may be a factor influencing repopulation and, thus, calcification.
CARDIOVASCULAR
Ann Thorac Surg 2003;75:1457– 63