| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Ann Thorac Surg 2007;84:729-736
© 2007 The Society of Thoracic Surgeons
a Department of Cardiovascular Surgery, Charité Hospital, Medical University Berlin, Berlin, Germany
b Department of Radiology, Charité Hospital, Medical University Berlin, Berlin, Germany
c Department of Anesthesiology, Charité Hospital, Medical University Berlin, Berlin, Germany
d Tissue Bank, Charité Hospital, Medical University Berlin, Berlin, Germany
Accepted for publication April 16, 2007.
* Address correspondence to Dr Dohmen, Department of Cardiovascular Surgery, Charité, Humboldt University Berlin, Luisenstraße 13, D-10117, Berlin (Email: pascal.dohmen{at}charite.de).
Presented at the Poster Session of the Forty-second Annual Meeting of The Society of Thoracic Surgeons, Chicago, IL, Jan 30–Feb 1, 2006.
| Drs Dohmen and Konertz disclose a financial relationship with AutoTissue Ltd.
|
 |
Abstract |
|---|
 Top Abstract IntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
Methods: Between May 2000 and February 2003, 23 patients received tissue-engineered heart valves. Two to four weeks before the Ross operation, a piece of forearm or saphenous vein was harvested to isolate, characterize, and expand endothelial cells. A pulmonary allograft (n = 11) or xenograft (n = 12) was decellularized, coated with fibronectin, and seeded with autologous vascular endothelial cells, using a specially developed bioreactor. Follow-up was performed by clinical evaluation, transthoracic echocardiography, magnetic resonance imaging, and multislice computed tomography.
Results: The patient mean age was 44.0 ± 13.7 years. Cell seeding density was 1.1 x 105 ± 0.5 x 105 cells/cm2, with a viability of 90.2% ± 8.9%. All patients survived the operation. One patient died during follow-up, and 1 patient required reoperation. All surviving patients are currently in New York Heart Association functional class I. Transthoracic echocardiographic evaluation of the tissue-engineered heart valve showed a mean flow velocity of 0.9 ± 0.4 m/s at 5 years. Multislice computed tomography showed no calcification up to 5 years postoperatively.
Conclusions: Tissue-engineered heart valves showed excellent hemodynamic performance during mid-term follow-up. Decellularization of heart valves and seeding with autologous vascular endothelial cells may prevent tissue degeneration and improve valve durability.
Autografts are the only viable heart valves available during aortic valve operations. Commercially available glutaraldehyde bioprostheses are nonviable structures and therefore no remodeling or regeneration potential is given [1]. Cryopreserved allografts are at the most partially viable, but also carry a high bioburden [2]. Because of their allogenic origin, immunologic reactions are possible, resulting in degeneration and calcification of these bioprostheses [3]. Tissue engineering technology made it possible to create a viable heart valve with potential for growth, regeneration, and repair [4].
This study was designed to collect prospective safety and effectiveness data about a tissue-engineered (TE) heart valve to reduce tissue deterioration and assure excellent hemodynamic behavior. Modifications have been initially made to the regular cryopreserved pulmonary allograft, performing decellularization treatment and in vitro seeding with autologous vascular endothelial cells. Allografts have limited availability, and therefore, other sources needed to be evaluated to overcome this problem. Thus, we performed in a second step: a decellularized pulmonary xenograft was used as a scaffold to be seeded in vitro with autologous vascular endothelial cells to overcome allograft shortages.
Extensive evaluation was performed to exclude any risk for the patient [5] and after first results were available of the seeded decellularized allografts [6]. This study presents mid-term clinical results of the TE heart valve created by in vitro seeding of a decellularized pulmonary allograft or xenograft during the Ross procedure.
|
Material and Methods |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
Valve Prosthesis
The scaffold used for tissue engineering in 11 patients was a commercially available cryopreserved pulmonary allograft (Cryolife Inc, Kennesaw, GA), and in 12 patients, a porcine heart valve (AutoTissue Ltd, Berlin, Germany) was used.
Preparation of the TE heart valve has been previously extensively described [5, 7]. In brief, a piece of autologous vein was prepared, and vascular endothelial cells were harvested, characterized, and expanded in the tissue laboratory. If a sufficient number of endothelial cells were available, a cryopreserved allograft or a fresh xenograft was decellularized [5]. The decellularized scaffolds were coated with Pronectine F (PAA, Laboratories GmbH, Coelbe, Germany) and seeded with vascular endothelial cells by using a special developed bioreactor. After seeding, quality control was performed, sterility was proven, and the TE heart valves were available for implantation.
Patients
Between May 2000 and February 2003, 23 consecutive patients received this TE heart valve. Selecting criteria for the Ross procedure were elective surgery, age younger that 65 years, acceptable life expectancy, and good quality of the native pulmonary valve. Additional selecting criteria to receive a TE heart valve were a patient-signed agreement, age older than 18 years, elective surgery with no risk of an additional 4 to 6 weeks of waiting time, and agreement for the additional harvesting a vein segment. The patient’s preoperative characteristics are listed in Table 1. Preoperatively, 6 patients (26.1%) were in New York Heart Association (NYHA) functional class III, and 17 (73.9%) were in class II.
|
A subcoronary implantation technique was performed in 22 patients. One patient received aortic root replacement due to previous multiple interventional and surgical treatments. A running suture line was performed for all anastomoses, using a double-armed 4-0 polypropylene suture for the proximal and a 5-0 polypropylene suture for the distal anastomosis. Mean aortic cross clamp-time was 108.2 ± 13.5 minutes, and mean extracorporeal circulation time was 120.0 ± 25.1 minutes.
Three patients also underwent reconstruction of the ascending aorta.
Follow-Up
Patients were schedule for follow-up visits at 3, 6, and 12 months postoperatively, and annually thereafter. During follow-up, patients were clinically evaluated and transthoracic echocardiography was performed. If echocardiography was difficult or suggestive of hemodynamic changes, magnetic resonance imaging was added. After 1-year of follow-up, multislice computed tomography (CT) was added to evaluate valve morphology.
Doppler Echocardiography
Transthoracic echocardiographic examination was performed preoperatively, at discharge, and at 3, 6, and 12 months, and annually thereafter. A Hewlett Packard Sonos 5500 (Agilent Technologies Ltd., Boeblingen, Germany) with a 2.5-MHz transducer was used. During each echocardiographic examination, aortic and pulmonary regurgitation were measured by the length and the width of the jet into the right or left ventricular outflow tract. Gradation was 0 (none), 1+ (trivial), 2+ (mild), 3+ (moderate), and 4+ (severe). The mean aortic and pulmonary flow velocities were measured with continuous-wave Doppler. Pulsed-wave Doppler was used for localization of the pressure drop at the right ventricular outflow tract and the TE heart valve to determine the origin of any increased flow velocity. All measurements were performed during normal quiet respiration. Spectral recordings were stored on magneto optical disk. Doppler echocardiography measurements of all heartbeats in three complete consecutive respiratory cycles were analyzed.
Magnetic Resonance Imaging
Examinations were performed at 1.5T (Magnetom Vision, Siemens, Erlangen, Germany), using a standardized protocol, which has been previously reported [10]. The maximum flow velocity encoding was set to 150 cm/s and increased to 250 cm/s if velocity aliasing was encountered. The standard electrocardiogram system provided by the magnetic resonance manufacturer was used.
Multislice Computed Tomography
Data were acquired on a multislice CT scanner (Aquilion16, Toshiba Medical Systems, Otawara, Japan) using a standardized protocol, which has been previously reported [11].
Histologic Evaluation
After each seeding procedure was completed, a piece of the valve wall was fixed and histologically evaluated by performing a Giemsa staining as well as routine histologic staining (hematoxylin and eosin, von Gieson, Goldner staining). Immunohistochemical evaluation was performed by von Willebrand factor, CD31, and CD34.
Histologic evaluation of explanted heart valves was performed by routine histologic staining (hematoxylin and eosin, von Gieson, Goldner, von Kossa staining) and immunohistochemical staining (von Willebrand Factor, antifibroblast factor, CD31, CD34, and CD68 staining).
Data Analysis
Quantitative data are presented as mean ± standard deviation. For comparing the endothelial cell seeding density and cell viability between decellularized allografts and xenografts, unpaired Student t test with SPSS 13.0 software (SPSS Inc. Chicago, IL) was used. The level for statistical significance was set at a p < 0.05.
|
Results |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
Giemsa staining showed a monolayer of endothelial cells at the inner surface of the seeded heart valves (Fig 1). None of the samples taken from the TE heart valves, cell cultures, or from the final prepared heart valve showed a positive result for bacteriologic or fungal contamination.
|
After the second passage, a mean number of 6.5 x 106 ± 2.9 x 106 endothelial cells were available, which was considered to be sufficient to cover a decellularized valve scaffold. The mean number of endothelial cells remaining in the solution after seeding was 1.5 x 106 ± 1.4 x 106, which means the mean number of endothelial cells covering the valves was 5.1 x 106 ± 2.1 x 106. The seeding percentage of endothelial cells was 80.1% ± 14.7%, with a density of 1.1 x 105 ± 0.5 x 105 cells/cm2 after a 4-hour duration.
When the two different valve scaffolds were compared, no differences were found in the seeding density between the decellularized allograft and the decellularized xenograft, respectively, 1.05 x 105 ± 0.5 x 105 cells/cm2 and 1.2 x 105 ± 0.5 x 105 cells/cm2, which was not significant (p = 0.23). The seeding was considered complete after a 10-day maturation period in a humidified incubator (37°C, 5% carbon dioxide and 98% air saturation).
Follow-Up
Follow-up was 100% complete. Mean follow-up was 46.0 ± 12.1 months (range, 33 to 68 months). There were no operative deaths. One patient suddenly died 3 months after surgery because of arrhythmia. Two days before the event, however, transthoracic echocardiography showed a maximum flow velocity at the neoaortic valve of 1.3 m/s and a maximum flow velocity at the TE heart valve of 0.9 m/s. Regurgitation at the pulmonary and neoaortic heart valve was absent.
Increasing maximum flow velocity developed in a second patient, rising from 0.8 m/s at discharge to 3.5 m/s at 4 months of follow-up. Magnetic resonance imaging demonstrated an external compression at the level of the distal anastomosis; however, no dilatation of the TE valve wall was seen (Fig 2). An inflammatory reaction was found intraoperatively, with fibrotic tissue and narrowing at the distal anastomosis. Although the valve leaflets looked smooth and pliable without degeneration, the valve was replaced.
|
Histologic Examination
Histologic evaluation of the explanted TE heart valve with hematoxylin and eosin staining showed no structural deterioration but surrounding fibrotic tissue (Fig 3A). Goldner staining showed an excellent preservation of the extracellular matrix. Furthermore, the high pressure due to the stenosis at the distal anastomosis did not influence the extracellular matrix (Fig 3B). Von Kossa staining showed absence of calcification at the complete matrix, including the wall (Fig 3C) and the leaflets. Von Willebrand factor visualized a monolayer of endothelial cells at the leaflets. Similar results are found at the valve wall showing a monolayer of cells positive for CD31 (Fig 3D) and CD 34 (Fig 3E). Many cells positive for CD34 are present in the deeper layers of the media (Fig 3F). The cells show to be building channels, and erythrocytes can be seen with the hematoxylin and eosin staining. A minimal number of CD68-positive cells were found (Fig 3G). A part of the host cells, which grow into the decellularized matrix, are shown to be positive for antifibroblast staining (Fig 3H).
|
|
|
|
Comment |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
Wang and colleagues [15] showed in a recently published study that pulmonary autografts are superior compared with aortic allografts at rest (respectively, 5 ± 2 mm Hg and 11 ± 4 mm Hg; p = 0.027) and at maximum exercise (respectively, 10 ± 5 mm Hg and 15 ± 5 mm Hg; p = 0.003). The pulmonary allograft used to reconstruct the right ventricular outflow tract, however, showed limiting results at rest and maximum exercise in these patients. The pressure gradient at rest was 14 ± 10 mm Hg compared with 3 ± 1 mm Hg in the patients who received the aortic allograft without pulmonary valve replacement. At maximum exercise, the difference became even more pronounced at, respectively, 25 ± 22 mm Hg and 5 ± 4 mm Hg.
In a similar study, Da Costa and colleagues [16] showed excellent results of the autologous pulmonary valve, with an average mean gradient of 1.8 ± 0.6 mm Hg at rest and 4.3 ± 2.5 mm Hg at maximum exercise. However, the pressure gradient at the pulmonary allograft increased during time at exercise at discharge and mean follow-up of 12.8 months, respectively, 10.4 ± 6.1 mm Hg and 26.0 ± 13.2 mm Hg. This study was able to show an increase of pressure gradient during follow=up, especially at exercise. Chambers and colleagues [17] showed again the excellent results of the pulmonary autograft at 20 years; however, there was a 20% rate of reoperation due to allograft failure of the right ventricular outflow tract reconstruction.
Limitations of Right Ventricular Outflow Tract Reconstruction
New heart valves are needed to reconstruct the right ventricular outflow tract. Several studies suggested that allograft degeneration is due to immunologic reactions [18, 19]. We performed a decellularization technique to eliminate all cells in regular available cryopreserved allografts to overcome immunogenicity. In vitro autologous vascular endothelial cell seeding was preferred for these decellularized matrices to protect the structure, because the vascular endothelium is characterized by anticoagulatory and profibrinolytic activities [20]. The Andrews and colleagues [21] study showed that glycoprotein Ib-IX-V and VI, two adhesion receptors that bind von Willebrand factor and collagen, are primarily responsible for regulating initial platelet adhesion and activation within the blood flow. The platelet adhesion that follows could lead to rapid signal transduction, with activation of integrins that support adhesion and aggregation. On the other hand, there is a natural cover of the collagen with autologous vascular endothelial cells to eventually overcome reactions against allogenic or xenogenic collagen.
In this study, autologous endothelial cells were used from the cephalic and the saphenous vein. An excellent study by Pascual and colleagues [22] showed that endothelial cells of vein origin can restore the function of arterial endothelial cells and create a functional platelet/endothelial cell adhesion molecule (PECAM 1)–type junction between both cell types. Simon and colleagues [23] showed that human umbilical vein endothelial cells are also able to prevent platelet adhesions and activation and eliminate thrombogenicity. In another clinical study, Deutsch and colleagues [24] showed that vein endothelial cells can significantly increase patency rates of polytetrafluoroethylene grafts due to reduction of thrombogenicity. We therefore believe using vein endothelial cells of two different areas does not create a problem.
We used a porcine decellularized valve matrix in 12 patients. The reason for this change was the difficulty in getting a suitable allograft available at the time of the seeding. At the time enough endothelial cells are present, the seeding should be performed as soon as possible, because the seeded valve needs to maturate after seeding and sterility needs to be proven.
The second problem was that although cryopreserved allografts are the gold standard, Brockbank and colleagues [25] suggested a possible cause of tissue failure and calcification due to interstitial ice formation in cryopreserved allografts. Furthermore, Goffin and colleagues [26] have mentioned the importance of applying dimethylsulfoxide (DMSO) as cryoprotective solution to the allograft and waiting at least 40 to 60 minutes to allow DMSO to penetrate completely into the tissue. The leaflets in this study showed ice crystals, so it would have been interesting to know what would have happened with the cryopreserved wall of these allografts.
A further important issue of this study would be to analyze the freezing curve. Goffin and colleagues [27] have always taken great care with this issue and used a grading scale for the quality of the allografts. Cryopreserved pulmonary allografts received a second grade quality due to the unsatisfactory freezing curve in 15%. If the current chamber temperature decreases too fast, the program should be corrected by hand. This, of course, does not always guarantee a perfect final freezing process. It seems that the authors of this recent study have had no cryopreservation failures.
We therefore preferred the use of a fresh valve matrix that has not been cryopreserved before decellularization and is in abundant in supply. Porcine matrices were used because several studies showed allergic reactions to bovine collagen, especially when the tissue was not cross-linked. Hanke and colleagues [28] showed abscesses persisting for days or weeks as a manifestation of hypersensitivity to bovine collagen in 4 of 10,000 cases. Periods of remission have been described up to 24 months afterwards. Local tissue necrosis occurred in 9 of 10,000 patients. Mehlisch [29] showed similar results with bovine collagen allergy in 6.5% of treated patients. No adverse reactions have been reported for porcine collagen, however.
Glutaraldehyde treatment has been believed for many years to reduce antigenicity of xenogenic collagen [30]. Those who argued for its use claimed that cross-link formation shields or modifies major antigenic situs, thus reducing the capacity to interact with antibodies. Courtman and colleagues [31] found in a more recent study that xenogenic grafts treated with glutaraldehyde actually showed increased antigenicity. Because previous studies showed complete decellularization of these valve matrices, they were used for TE heart valves [5].
We did not find any differences between the allogenic or xenogenic matrix when they were used. Hemodynamic behavior was not significantly different for either group, nor was seeding capacity; however, long-term follow-up will be needed to show if there is any difference between both matrices.
One seeded valve needed to be explanted because of distal anastomosis narrowing. Reconstruction of this valve was not possible because the anastomosis was too distal and it was technically impossible to perform a patch reconstruction. Histologic examination of this explanted valve, however, showed similar results as in experimental animals [6, 32]. The decellularized matrices withstand the high pressure due to the narrowing of the distal anastomosis, without dilatation and fragmentation of the collagen structures. In the animal model, we saw the matrix repopulating with interstitial cells as early as 3 months after implantation. We were able to demonstrate the persistence of endothelial cells to the blood flow on a seeded TE heart valve was explanted after 7 days. Similar results were found in this clinical case. The seeded endothelial cells persisted, and as a result of the blood flow, the deeper layers started to be recellularized with interstitial cells. Furthermore, we were able to see a revascularization similar to what we saw in valves we implanted in a juvenile sheep model [4].
This study also showed the impact of the applied decellularization technique. Simon and colleagues [33] reported early failure of a decellularized porcine heart valve (SynerGraft, CryoLife), due to severe foreign body reaction dominated by neutrophil granulocytes and macrophages in the early implants. At 1 year, a pronounced lymphocytic reaction could be seen. Histologic evaluation showed very early calcific deposits within the scaffold; furthermore, a pseudointimal layer at the inner surface was present. The changes seen in the SynerGraft xenogenic heart valve seemed to be due to incomplete decellularization, with a lot of cellular debris remaining in the scaffold and also damage to the collagen elastin scaffold caused by an inadequate decellularization method.
We have seen a pseudointimal layer in our experimental animal model only in cryopreserved allografts. Our decellularized porcine valve matrix never showed a pseudointima. At the inner surface, a smooth endothelial cell layer was seen, and interstitial cells had grown into the scaffold [4]. Bechtel and colleagues [34] showed in another study that decellularization of pulmonary allografts (SynerGrafts) showed absence of positive anti-human leukocyte antigen antibodies. Hemodynamic behavior, however, was not positively influenced because transvalvular pressure gradients rose significantly during follow-up. It is evident that different decellularization techniques do not necessarily achieve similar results. Next to this, no transvalvular pressure increase was seen at up to 5 years of follow-up in this clinical study.
Study Limitations
Ideal would be a randomized study, using regular cryopreserved allografts and seeded decellularized allografts; however, experimental data and the safety and feasibility data of the initial three implantations were retrospectively analyzed and compared with previously used cryopreserved allografts implanted by the same surgeon. A different clinical course could be shown retrospectively, so the ongoing use of conventional cryopreserved allografts was considered obsolete. This was performed in compliance with the Declaration of Helsinki. A patient selection was, unfortunately, made by the selection criteria because all patients were elective due to the waiting time until the TE valve was ready to be implanted.
Conclusion
Tissue-engineered heart valves showed excellent hemodynamic performance during mid-term follow-up, without increase of transvalvular flow velocity. Decellularization of heart valves and seeding with autologous vascular endothelial cells may prevent tissue degeneration and improve valve durability.
|
Acknowledgments |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
|
References |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  F. Godart, I. Bouzguenda, F. Juthier, F. Wautot, A. Prat, C. Rey, D. Corseaux, A. Ung, B. Jude, and A. Vincentelli Experimental off-pump transventricular pulmonary valve replacement using a self-expandable valved stent: A new approach for pulmonary incompetence after repaired tetralogy of Fallot? J. Thorac. Cardiovasc. Surg., May 1, 2009; 137(5): 1141 - 1145. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Gomez-Caro, E. Martinez, A. Rodriguez, D. Sanchez, J. Martorell, J. M. Gimferrer, A. Haverich, W. Harringer, J. L. Pomar, and P. Macchiarini Cryopreserved Arterial Allograft Reconstruction After Excision of Thoracic Malignancies Ann. Thorac. Surg., December 1, 2008; 86(6): 1753 - 1761. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Schreiber, A. Eicken, S. Seidl, and R. Lange Unexpected Early Failure of a Decellularized Right Ventricle to Pulmonary Artery Graft Ann. Thorac. Surg., December 1, 2008; 86(6): 2026 - 2027. [Full Text] [PDF]
|
|
|
|
 |
|
 F. J. Schoen Evolving Concepts of Cardiac Valve Dynamics: The Continuum of Development, Functional Structure, Pathobiology, and Tissue Engineering Circulation, October 28, 2008; 118(18): 1864 - 1880. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. M. Dohmen and W. Konertz Decellularization of Xenogenic Biologic Tissue Ann. Thorac. Surg., June 1, 2008; 85(6): 2163 - 2163. [Full Text] [PDF]
|
|
|
|
|
|
A. Vincentelli, F. Juthier, and B. Jude Reply to the editor. J. Thorac. Cardiovasc. Surg., May 1, 2008; 135(5): 1190 - 1190. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |
