Author + information
- Received July 2, 2013
- Revision received September 12, 2013
- Accepted September 19, 2013
- Published online April 8, 2014.
- Anita Driessen-Mol, PhD∗,
- Maximilian Y. Emmert, MD, PhD†,‡,§,
- Petra E. Dijkman, PhD†,
- Laura Frese, PhD†,
- Bart Sanders, MSc∗,
- Benedikt Weber, PhD†,‡,
- Nikola Cesarovic, PhD‡,
- Michele Sidler, MD||,
- Jori Leenders, MSc∗,
- Rolf Jenni, MD, MSEE§,
- Jürg Grünenfelder, MD§,
- Volkmar Falk, MD§,
- Frank P.T. Baaijens, PhD∗ and
- Simon P. Hoerstrup, MD, PhD∗,†,‡,§,||∗ ()
- ∗Department of Biomedical Engineering, Eindhoven University of Technology, Eindhoven, the Netherlands
- †Swiss Center of Regenerative Medicine, University and University Hospital Zürich, Zürich, Switzerland
- ‡Center Surgery, Division of Surgical Research, University Hospital Zürich, Zürich, Switzerland
- §Heart Center Zürich, University Hospital Zürich, Zürich, Switzerland
- ||Center of Applied Biotechnology and Molecular Medicine (CABMM), University of Zürich, Zürich, Switzerland
- ↵∗Reprint requests and correspondence:
Dr. Simon P. Hoerstrup, Heart Center Zürich, University and University Hospital Zürich, Raemistrasse 100, 8091 CH-Zürich, Switzerland.
Objectives This study sought to evaluate long-term in vivo functionality, host cell repopulation, and remodeling of “off-the-shelf” tissue engineered transcatheter homologous heart valves.
Background Transcatheter valve implantation has emerged as a valid alternative to conventional surgery, in particular for elderly high-risk patients. However, currently used bioprosthetic transcatheter valves are prone to progressive dysfunctional degeneration, limiting their use in younger patients. To overcome these limitations, the concept of tissue engineered heart valves with self-repair capacity has been introduced as next-generation technology.
Methods In vivo functionality, host cell repopulation, and matrix remodeling of homologous transcatheter tissue-engineered heart valves (TEHVs) was evaluated up to 24 weeks as pulmonary valve replacements (transapical access) in sheep (n = 12). As a control, tissue composition and structure were analyzed in identical not implanted TEHVs (n = 5).
Results Transcatheter implantation was successful in all animals. Valve functionality was excellent displaying sufficient leaflet motion and coaptation with only minor paravalvular leakage in some animals. Mild central regurgitation was detected after 8 weeks, increasing to moderate after 24 weeks, correlating to a compromised leaflet coaptation. Mean and peak transvalvular pressure gradients were 4.4 ± 1.6 mm Hg and 9.7 ± 3.0 mm Hg, respectively. Significant matrix remodeling was observed in the entire valve and corresponded with the rate of host cell repopulation.
Conclusions For the first time, the feasibility and long-term functionality of transcatheter-based homologous off-the-shelf tissue engineered heart valves are demonstrated in a relevant pre-clinical model. Such engineered heart valves may represent an interesting alternative to current prostheses because of their rapid cellular repopulation, tissue remodeling, and therewith self-repair capacity. The concept of homologous off-the-shelf tissue engineered heart valves may therefore substantially simplify previous tissue engineering concepts toward clinical translation.
- off-the-shelf tissue engineering
- tissue-engineered heart valves
- transcatheter valve implantation
Valvular heart disease is an important cause of morbidity and mortality (1). The number of patients requiring heart valve replacements is approximately 280,000 annually worldwide (2) and is estimated to triple in the upcoming decades (3). In recent years, transcatheter valve implantations have emerged as a minimally invasive alternative to conventional surgery, in particular for elderly high-risk patients. However, being bioprosthetic in nature (consisting mainly of fixed, ovine, or porcine tissues) the currently used transcatheter prostheses are inherently associated with progressive dysfunctional degeneration, preventing their broader use in younger patient populations.
While preliminary attempts with decellularized xenogeneic and allogeneic heart valves have only shown limited host cell repopulation in pre-clinical and clinical trials (4–8), the concept of tissue engineered, living, and autologous heart valves with self-repair and remodeling capacity has been proposed as a promising alternative to overcome such limitations (9–15). In 2010, we reported successfully merging transcatheter-based technologies with heart valve tissue engineering (HVTE) on the basis of stem cell methodology (12). However, this “classical” heart valve tissue engineering concept comprising complex multistep procedures such as cell harvest, cell expansion, seeding on scaffolds, bioreactor in vitro culture, and time-critical implantation coordination of the delicate, living engineered autologous heart valves require high logistical and financial efforts. In this regard, the concept of off-the-shelf (decellularized) homologous tissue-engineered heart valves has recently been introduced as a promising alternative to substantially simplify current HVTE approaches toward routine clinical translation (16–18).
In this study and for the first time, we investigate the long-term in vivo functionality, host repopulation capacity, and matrix remodeling of “off-the-shelf” homologous transcatheter tissue-engineered heart valves (TEHVs) as pulmonary valve replacement in sheep.
In vitro production of homologous cell-based transcatheter TEHV
Living stented TEHVs were engineered as previously described (16,19) on the basis of rapid degrading synthetic scaffolds and vascular-derived cells and decellularized using a series of enzymatic treatments (16). Please see the expanded Methods section in the Online Appendix for further details. Sterilization was obtained by immersion in 70% EtOH and antibiotic treatment. Decellularized TEHVs were stored in fresh M-199 medium at 4°C.
Transapical implantation and in vivo performance of TEHVs
To enable transapical delivery, the stented TEHVs (length = 27 mm, outer diameter = 30 mm) were crimped and loaded onto a custom-made inducing system (outer diameter = 12 mm). To evaluate in vivo functionality and host repopulation capacity, TEHVs (n = 12) were transapically implanted as pulmonary valve replacement in adult sheep (pulmonary annulus 24 to 26 mm, age 2.8 ± 0.1 years, weight range 69 ± 2 kg). The remaining valves (n = 5) served as reference (control) valves. All animals received human care. The ethics committee (Veterinäramt, Gesundheitsdirektion, Kanton Zürich [197/2010], Switzerland) approved the study in compliance with the Guide for the Care and Use of Laboratory Animals, published by the National Institutes of Health (NIH publication No. 85-23) (20). Please see the expanded Methods section in the Online Appendix for further details. The appropriate position and functionality of the implanted valve was visualized by angiography. In vivo functionality (heart rate, mean and maximum transvalvular pressure gradient, and grade of insufficiency) was monitored using transesophageal echocardiography during the procedure, immediately after implantation and after 1, 4, 8, 16, and 24 weeks post-operatively. Insufficiency (central regurgitation) was graded as none to: 1) trivial; 2) mild; 3) moderate; and 4) severe. Anticoagulation therapy was maintained for 7 days after implantation. The TEHVs were explanted within 1 day (n = 2), and after 8 weeks (n = 2), 16 weeks (n = 4), and 24 weeks (n = 4). Photographs were taken from all explanted valves to macroscopically analyze tissue appearance.
Qualitative tissue analyses: (immuno-) histology and scanning electron microscopy
Control (n = 5) and explanted TEHVs (n = 12) were analyzed by (immuno-) histology and scanning electron microscopy. Please see the expanded Methods section in the Online Appendix for further details.
Quantitative tissue analyses: tissue composition
The total amount of DNA, sulfated glycosaminoglycans (sGAGs), and collagen in the leaflets of control (n = 5) and explanted TEHVs after 8 (n = 2), 16 (n = 4), and 24 weeks (n = 4) as well as of native ovine valve leaflets (n = 3) were analyzed using biochemical assays. For each valve 2 samples per leaflet were analyzed, resulting in 6 measurements per valve. These measurements were averaged to represent an average value for each valve. The values for DNA, sGAGs, and collagen are expressed per milligram dry weight. Please see the expanded Methods section in the Online Appendix for further details on the assays.
Data are represented as mean ± SD. To identify differences in tissue composition (DNA, sGAG, and collagen) between the control valve leaflets and the explanted leaflets at 16 and 24 weeks and that of native ovine valve leaflets, a 1-way analysis of variance was performed with Tukey’s post-hoc testing. The number of values for the explanted valves after 8 weeks was too low (n = 2) to include in the statistical analyses. Differences were considered significant when p < 0.05. Statistics were performed using GraphPad Prism software (version 5.0d, San Diego, California).
Implantation and in vivo performance of TEHVs
The transapical implantation procedures (n = 12) were successful. No perioperative morbidity or mortality occurred and all valves could be deployed successfully at the target site (Fig. 1, Online Video 1). In the early post-operative phase, 2 animals presented with valve migration into the right ventricular outflow tract and died within 24 h post-operatively. The remaining animals (n = 10) made a swift recovery and completed their respective follow-up without any complications. Serial echocardiography confirmed sufficient valve function with mobile leaflets and excellent coaptation at the early follow-up time points (Fig. 2, Online Videos 2 and 3). No central regurgitation was observed at 4 weeks (n = 10) and mild central regurgitation at 8 weeks (n = 10) (Table 1). Only minor paravalvular leakage was initially observed in some animals up to 1 week after implantation.
While leaflet mobility was maintained on long-term follow-up, the coaptation slowly decreased over time (Fig. 2, Online Video 4), which was most likely due to merging of the leaflets with the valvular wall at the level of the hinge area as well as the occurrence of a single leaflet prolapse in some animals. Consequently, mild to moderate central regurgitation could be observed at 16 weeks (n = 8), that further increased to moderate central regurgitation at 24 weeks (n = 4), with 1 animal presenting with severe insufficiency. Functional measurements (Table 1) demonstrated stable mean and peak transvalvular pressure gradients over time (4.4 ± 1.6 mm Hg and 9.7 ± 3.0 mm Hg, respectively).
Macroscopic TEHV appearance
The implanted in vitro grown homologous TEHV revealed thin and shiny tissue formation in both valvular wall and leaflets (Figs. 3A to 3C). All explanted TEHVs demonstrated shiny and pliable leaflets and dense whitish valvular wall tissue, irrespective of the implantation period (Figs. 3D to 3L). In all explants, the valvular wall tissue was integrated into the surrounding native valvular wall. Excellent coaptation of the explanted valves was evident in the implant (Fig. 3A) and was maintained up to 8 weeks of implantation (Fig. 3D). Thereafter, valve closure was incomplete in line with the observed central regurgitation (Figs. 3G and 3J). Apparently, the line of attachment of the leaflets to the valvular wall shifted upward in time, indicating tissue merging process at the level of the hinge area (Figs. 3E, 3F, 3H, 3I, 3K, and 3L), associated with a reduction in leaflet size with time.
TEHV repopulation and remodeling
Prior to implantation the TEHVs revealed no cellular remnants (Figs. 4A, 4F, and 4K) and a well-developed extracellular matrix, mainly consisting of collagen demonstrating an efficient decellularization procedure (Figs. 4P, 4U, and 4Z). After TEHV deployment in the pulmonary position endogenous cellular repopulation occurred rapidly with first signs of cell infiltration as early as 5 h after implantation (Figs. 4B, 4G, and 4L). Both leaflets and wall tissues were homogeneously repopulated over time (Figs. 4C, 4D, 4H, 4I, 4M, and 4N), with fastest repopulation at highest densities in the wall. After 24 weeks, cell repopulation density in the leaflets (Fig. 4E) and hinge area (Fig. 4J) approached that in the valvular wall (Fig. 4O). Scaffold remnants remained longest visible in the leaflets (Figs. 4B to 4E) with local increased cell densities. Minimal depositions, most likely blood platelets and fibrin, were present at the surface of the whole valve after 5 h (Figs. 4B, 4G, and 4L), but disappeared with time. The valvular tissues demonstrated abundant amounts of collagen that increased in density over time, in particular in the hinge area (Figs. 4V, 4W to 4Y) and wall (Figs. 4AA to 4DD), and to a lesser extent in the leaflets (Figs. 4Q,R,S,T). Elastic matrix formation was evident in the wall at 8 weeks and later time points (Figs. 4BB to 4DD). In the hinge area the formation of elastic fibers was visible after 16 weeks (Figs. 4X and 4Y) and in the leaflets at 24 weeks (Fig. 4T). Calcification was not detected in any of the valves (data not shown).
TEHV cell phenotypes and distributions
Cells infiltrating the valve within 5 h after implantation were all vimentin positive (Figs. 5A, 5E, and 5I) and alpha-smooth muscle actin (α-SMA) negative (Figs. 5M, 5Q, and 5U). After 8 weeks, vimentin positive cells were identified mainly in close vicinity of the polymeric scaffold remnants in the leaflet (Fig. 5B) and hinge area (Fig. 5F), while more homogenously distributed and in higher amounts in the valvular wall (Fig. 5J). The cells in the leaflet and hinge area were all α-SMA negative (Figs. 5N and 5R) and α-SMA positive in the valvular wall (Fig. 5V). After 16 weeks, homogenously distributed vimentin positive cells were observed in all regions of the valve (Figs. 5C, 5G, and 5K). These cells were α-SMA positive in the hinge area (Fig. 5S) and wall (Fig. 5W) and sparsely α-SMA positive in the leaflet (Fig. 5Q). After 24 weeks, vimentin positive cells were homogenously distributed over the valve (Figs. 5D, 5H, and 5L). The level of α-SMA seemed lower in the hinge area (Fig. 5T) and wall (Fig. 5X) as compared with the 16 weeks explants. In the leaflets, more α-SMA positive cells were identified (Fig. 5P) as compared with earlier time points.
After 8 weeks TEHV demonstrated partly confluent endothelial lining as observed by CD31 staining (Figs. 6A, 6D, and 6G). Similar features were observed at the surface of the explants at 16 (Figs. 6B, 6E, and 6H) and 24 weeks (Figs. 6C, 6F, and 6I). The cell lining showed the typical cobblestone morphology at all time points, representative for endothelial cells as visualized by scanning electron microscopy (Figs. 6J to 6O). The degree of endothelialization varied between locations and explants with decreasing endothelialization at the hinge area with implantation time (Figs. 6D to 6F) and increasing endothelialization of the leaflet (Figs. 6A to 6C) and valvular wall surface with implantation time (Figs. 6G to 6I).
Quantitative TEHV tissue analyses
DNA content (Fig. 7A) of the leaflets increased with implantation time as compared with the values before implantation (p < 0.01 at 16 weeks and p < 0.001 after 24 weeks). After 16 weeks, DNA content was still lower than that in native leaflets (p < 0.05), but after 24 weeks the DNA content in the explanted TEHVs was similar to that in native ovine valve leaflets. The amount of sGAGs (Fig. 7B) was lower in TEHVs before implantation as compared with that in native leaflets (p < 0.01) and was still lower in the 16-week explants (p < 0.05). After 24 weeks sGAG content approached native values. Collagen content (Fig. 7C) was higher after 16 and 24 weeks in vivo as compared with the values of the TEHVs before implantation (p < 0.05). Collagen content was equal to that in native ovine leaflets at all time points.
Since the emergence of transcatheter heart valve implantations, which today are mostly used in elderly inoperable patients, an expansion to younger patient populations is anticipated in the near future. A major limitation however—as in classical surgical bioprostheses—is the progressive dysfunctional deterioration inherent to the nonliving animal-derived tissue used in today’s bioprostheses. Furthermore, accumulating clinical findings suggest that such degenerative processes may be even aggravated in transcatheter valves due to the substantial mechanical stresses during the crimping and deployment procedure. In the search for a next-generation technology platform to overcome the limitations of today’s bioprosthetic transcatheter heart valve prostheses, the concept of HVTE using living autologous tissues with self-repair capacity has created substantial hope for future heart valve therapy concepts. In 2010, we have demonstrated the principal feasibility to merge transcatheter-based technologies and HVTE demonstrating transcatheter implantation of stent-based TEHVs as pulmonary valve replacements in a pre-clinical animal model (12). However, in this proof-of-concept study, the generation of the used TEHV followed the algorithm of the classical HVTE approach comprising complex logistical and financial efforts including cell and expansion, scaffold seeding, bioreactor in vitro pre-conditioning, and time-critical implantation coordination of the delicate, living, autologous tissue-engineered heart valves.
In this study, we simplified the previously used HVTE concept toward a substantially more translational and clinically relevant methodology demonstrating long-term in vivo functionality, host repopulation capacity, and matrix remodeling of off-the-shelf, tissue engineered decellularized homologous transcatheter TEHVs in sheep. Conceptually, off-the-shelf, homologous TEHVs carry significant advantages either when compared with decellularized xenogeneic/allogeneic natural heart valves (16,18) or to the technologically and logistically demanding “classical” HVTE approaches. By decellularization of homologous TEHV on the basis of biodegradable PGA/P4HB scaffolds, we are able to produce largely available off-the-shelf homologous starter matrices at any clinically relevant size without the risk for xenogeneic disease transmission (16). Importantly, this off-the-shelf concept greatly simplifies previous HVTE concepts with regard to financial and logistical efforts (16). Last but not least, these decellularized matrices are designed and adapted for transcatheter implantation, therewith further enhancing their clinical relevance.
Overall, the TEHV function was sufficient with mobile leaflets and excellent coaptation at the early follow-up time points. No central regurgitation was observed at 4 weeks and mainly mild central regurgitation at 8 weeks. While leaflet mobility as well as a low transvalvular mean gradient was maintained on all follow-up time points, valvular coaptation slowly decreased over time, which was most likely due to a principal design flaw of these prototype TEHVs resulting in merging of the leaflet base with the respective hinge areas. This may be primarily related to the prototype stent and valve design lacking important anatomical features such as sinuses and thus leading to nonphysiological loading and insufficient washout in the hinge area, in particular during diastole.
To ensure viability, prevent deterioration, and allow for growth and remodeling it is crucial that the host rapidly repopulates these matrices after implantation. In our study, the TEHVs demonstrated a remarkable repopulation and remodeling capacity, pointing to the high potential of this approach with regard to self-repair capacity.
Repopulation occurred rapidly already after a few hours. It was fastest in the valvular wall, followed by the hinge area and the leaflets, reaching similar DNA content to that of native valves. In line with that, the collagen amount reached native levels with increasing density over time and the formation of elastic fibers was visible. Remodeling of the matrix occurred throughout the whole valve, but fastest in the wall, followed by the hinge area and the leaflets, and as such correlates with the observed rates of repopulation. Furthermore, endothelialization of the surfaces occurred at the valvular wall and the leaflets.
In comparison with native decellularized allogeneic or xenogeneic valve prostheses that have been repeatedly reported to show if at all only limited cell repopulation (4–8), repopulation of our TEHVs appears to be substantially more efficient. Importantly, thus far, repopulation of valve starter matrices was assumed to occur from the valve basis upward associated with remodeling of the valve matrix. This study suggests a potentially faster and more important route of repopulation, which is by blood-borne cells. The efficacy of repopulation and subsequent production and remodeling of extracellular matrix via this route may be very potent and its mechanisms are subject to subsequent studies.
Translation of this homologous off-the-shelf heart valve tissue engineering from the low-pressure pulmonary circulation to the systemic circulation appears feasible. Recent pilot experiments by our group have demonstrated structural integrity at systemic pressures of comparable valves (21–23). As soon as the stent and valve delivery system have been adapted to the aortic root, future studies will focus on long-term systemic valve replacements.
We demonstrate for the first time feasibility and long-term functionality of transcatheter delivered homologous off-the-shelf tissue-engineered heart valves with the potential for self-repair. The concept of off-the-shelf TEHV may represent a promising alternative to currently used valve prostheses as of their rapid host cell repopulation and remodeling capabilities toward native valve features within a short time-span. Moreover, the off-the-shelf TEHV concept may significantly simplify previous, classical HVTE concepts toward clinical translation.
Valvular coaptation decreased over time by merging of the leaflet base with the respective hinge areas. This shortcoming of the prototype TEHV requires stent and scaffold design modifications and optimizations (i.e., the implementation of anatomical sinuses ) to ensure and further improve long-term functionality of these next generation valves.
The authors thank Marina van Doeselaar (Department of Biomedical Engineering, TU/e) for help with TEHV culture, Pia Fuchs (Department of Surgical Research, USZ) for her support as to the histological examination, the Laboratory for Special Techniques (Institute for Clinical Pathology, USZ) as to the (immuno-) histochemical examination, and Mr. Klaus Marquardt (EMZ, University of Zürich) as to the scanning electron microscopy investigations.
The research leading to these results has received funding from the European Union's Seventh Framework Programme ([FP7/2007-2013]) under grant agreement n° 242008. All authors have reported that they have no relationships relevant to the contents of this paper to disclose. Drs. Driessen-Mol, Emmert, and Dijkman are joint first authors and contributed equally to this work. Drs. Baaijens and Hoerstrup are joint last authors and contributed equally to this work.
- Abbreviations and Acronyms
- heart valve tissue engineering
- sulfated glycosaminoglycan
- tissue-engineered heart valve
- Received July 2, 2013.
- Revision received September 12, 2013.
- Accepted September 19, 2013.
- American College of Cardiology Foundation
- Rosengart T.K.,
- Feldman T.,
- Borger M.A.,
- et al.
- Pibarot P.,
- Dumesnil J.G.
- Honge J.L.,
- Funder J.,
- Hansen E.,
- Dohmen P.M.,
- Konertz W.,
- Hasenkam J.M.
- Schmidt D.,
- Dijkman P.E.,
- Driessen-Mol A.,
- et al.
- Syedain Z.H.,
- Lahti M.T.,
- Johnson S.L.,
- et al.
- ↵(2011) Guide for the Care and Use of Laboratory Animals (National Academies Press (US), Washington, DC), 8th edition.
- Emmert M.Y.,
- Weber B.,
- Behr L.,
- et al.
- Emmert M.Y.,
- Weber B.,
- Behr L.,
- et al.