| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
J Thorac Cardiovasc Surg 1998;116:609-616
© 1998 Mosby, Inc.
SURGERY FOR ADULT CARDIOVASCULAR DISEASE |
Kiel and Hannover, Germany, and Linz, Austria
From the Department of Cardiovascular Surgery,a University Hospital, "Christian Albrechts Universität zu Kiel," Germany; Department of Cardiothoracic and Vascular Surgery,b Medizinische Hochschule Hannover, Germany; and the Department of Surgery,c General Hospital, Linz, Austria.
Received for publication Aug 19, 1997. Revisions requested Oct 1, 1997; revisions received June 5, 1998. Accepted for publication June 8, 1998. Address for reprints: André R. Simon, MD, Klinik für Herz-, Thorax- und Gefässchirurgie, Medizinische Hochschule Hannover, Carl Neuberg Str, 30623 Hannover, Germany.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and methodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
1 However, the use of these prostheses is limited, because biologic heart valve prostheses undergo degeneration after implantation, which leads to their destruction. The factors causing this process are not understood, and growing evidence suggests that, besides mechanical stress, immunogenic properties of the implant (eg, viable endothelial cells) contribute to the long-term deterioration. We 2 have previously demonstrated the immunogenic potential of human cardiac valve endothelium (HCVE) in vitro.
In the current study, we have further examined the immunogenicity of HCVE from explanted valves and valve prostheses using antibodies against endothelial adhesion molecules. We found that HCVE can express molecules such as intercellular adhesion molecule–1 (ICAM-1, CD54) or E-selectin (CD62E), which play critical roles in peripheral blood mononuclear cell attachment to and arrest on the vascular wall, 3,4 an important step in the initiation of immune responses. For instance, using a mouse cardiac transplantation model, Isobe and associates 5 demonstrated that inhibition of CD54 receptor interaction with its ligand, leukocyte function–associated antigen (LFA-1), completely prevents rejection of the implant and leads to long-term graft acceptance.
Stimulation of endothelium can lead to its increased immunogenicity (eg, major histocompatibility complex [MHC] class II up-regulation) and stimulation of immune responses in vitro. On cardiac valve prostheses, this might lead to the removal of the endothelial lining, subsequently rendering the implant more susceptible to mechanical stress. To further address the role of HCVE in the degeneration of biologic heart valve implants, we analyzed the surface and the internal structure of such valvular substitutes using immunohistochemistry on cryostat sections of valves and valve prostheses obtained during replacement operation, rereplacement or cardiac transplantation, and retransplantation. We showed that HCVE expresses adhesion molecules such as platelet endothelial cell adhesion molecule (CD31) and human cellular adhesion molecule (CD44), known to be involved in homing, adhesion, and migration of immunocompetent cells to tissues as well as costimulation of T cells via the T-cell reactivity/costimulatory molecule pathway. 6,7 To differentiate between donor and recipient origin of cells, we used monoclonal antibodies specific for donor and recipient MHC class I molecules. We could show a partial re-endothelialization process on xenoprosthetic and alloprosthetic valves of recipient origin after long-term clinical implantation. In contrast, on surfaces of valves obtained from previously transplanted hearts, a confluent layer containing HCVE could be detected showing no signs of re-endothelialization by the recipient. This observation is striking, because these valves appear to be resistant to degeneration, being implanted when fully viable, without previous cryopreservation, into a recipient receiving systemic immunosuppression. These findings suggest an important role of the immune system in the long-term function of allograft valve prostheses and lead us to the hypothesis that implantation of a fully viable allograft prosthesis, matched to the HLA phenotype of the recipient, might be a way to further improve long-term results after allograft valve implantation.
|
Materials and methods |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
2 and cultured in chamber slide culture chambers, 6- and 96-well microtiter plates (Nunc A/S, Roskilde, Denmark) at 37°C and 5% carbon dioxide until confluence. Cell cultures were fed by exchange of medium when necessary. For all experiments, primary cultures were used as soon as a confluent monolayer had developed.
Explanted valves
Reference aortic valves (n = 6) were obtained at the time of replacement for valvular insufficiency (4/6) and from donor hearts rejected for transplantation (2/6). Explanted allograft prostheses were obtained at the time of reoperation for insufficiency and stenosis (n = 12, duration of implantation 3-12 years, 4 pulmonary frozen homografts [group A]; 6 aortic frozen homografts [group B], 2 undetermined [group C]). Explanted xenograft prostheses (n = 6, 5 glutaraldehyde-treated stented porcine xenografts, 1 glutaraldehyde-treated stented bovine xenograft) were obtained at the time of second replacement for insufficiency (duration of implantation 4-12 years). Valves from previously transplanted hearts were obtained at the time of retransplantation or valve replacement (n = 4, duration of implantation 1-4 years). Thus obtained specimens were snap-frozen in liquid nitrogen, cut into 5 µm sections, and used for immunohistochemical staining.
Aortic endothelium and endocardium
Aortic endothelium and endocardium were obtained at the time of transplantation and cardiectomy for transplantation, both from donors and recipients or from hearts used for heart valve banking. Additional endothelial cells were obtained at routine endomyocardial biopsies after cardiac transplantation. The material was immediately snap-frozen and stored in liquid nitrogen until further use. Tissue was cut in 5 µm sections and stained with the use of the panel of antibodies described herein.
HLA phenotypes.
Four allograft recipients (2 group A, 2 group B) and all patients undergoing retransplantation underwent HLA typing to differentiate between potential donor and recipient cells on the surface of the prostheses. Complete HLA data were available for 1 valve donor.
Valve donor:
VD1: HLA-A2 B7,40 BW3,6 Cw3,7 DR2,4(53) DQ3,7
Valve recipients:
VR1:HLA-A1,3 B8,7 Cw7 DR2(15),5(11) DQ1,3(7)
VR2: HLA-A2,9(23) B21(50),12(44) Cw5,6 DR1,7 DQ1(5),2
VR3: HLA-A2,3 B7 (no DR available)
VR4(VD1): HLA-A1,2 B13,7 Cw6 DR6,7
Patients undergoing retransplantation:
rTX1: HLA-A2,26 B60 CW3 DR1,12
rTX2: HLA-A2,30 B39,51 DR1,4
rTX3: HLA-A2,26 B51,60 DR1,6
rTX4: HLA-A2,11 B37,52 Cw2 DR2,1
Donors of original (first) hearts of patients undergoing retransplantation:
D1(rTX2): HLA-A2 B27,35 Cw2,4 DR1,11
D2(rTX4): HLA-A1,10(26) B8,12 DR,6
Antibodies
All primary antibodies were monoclonal (mAb). W6/32 (anti–class I heavy chain) was obtained from American Type Culture Collection (ATCC, Manassas, Va). Anti-CD29, CD31, CD34, CD44, CD49a, CD49c, CD49d, CD49e, CD49f, CD51, CD54, CD62P, CD106, anti-H/Y mAb and anti–factor VIII–related antigen monoclonal antibodies (anti–FVIII–related Ag mAb) were purchased from Immunotech (Hamburg, Germany). Additional anti–FVIII– related Ag mAb were purchased from DAKO (Dakopats, Hamburg, Germany). Anti-CD49b mAb were from Becton Dickinson (Heidelberg, Germany). Anti-CD62E mAb were from Dianova (Hamburg, Germany). Anti-CD102 mAb were from Bender & Co GmbH (Vienna, Austria). Alkaline phosphatase anti–alkaline phosphatase (APAAP) and immune peroxidase (POD) secondary mAb were from DAKO. Enzyme coupled anti–secondary APAAP mAb were from Dianova. Primary antibodies used and the primary function of their corresponding antigens are summarized in Table IA. Anti-HLA antibodies (see Table IB), which were used to differentiate between donor and recipient cells, were kindly provided by Dr Westphal (Institute of Immunology, Christian Albrechts Universität Kiel, Kiel, Germany).
|
|
2 |
Results |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
2 In brief, cells were shaped irregularly, showed contact inhibition, and grew in monolayers without detectable growth patterns. In all cell cultures, characteristic stigmata could be observed. Fibroblast contamination could be detected easily by their typical spindle-shaped appearance, their whorl-like growth, and lack of contact inhibition.
|
).
|
8,9 Basically, valves obtained at initial replacement or retransplantation appeared as normal valves with intact endothelial lining and interior fibroelastic structure in which stromal cells could be demonstrated (Fig 2, A
, and 4). Allogeneic or xenogeneic valves replaced at rereplacement showed breakup of internal collagenous structure (Fig 2, B, and 3). The interior of the prostheses showed areas of calcification and infiltrating cells. No stromal cells of donor origin could be shown.
|
|
). This layer stained negatively for FVIII-related Ag, as we previously reported, and negative for CD49b, CD49d, CD62E, CD62P, CD102, and CD106 (Table II).
|
, Fig 3).
Valves from transplanted hearts
On valves from transplanted hearts, the observed monolayer was not always continuous. The endothelial cells showed a similar phenotype (Fig 4), although some differences were observed (Table II). These layers partially stained positive for FVIII-related Ag also. In addition, cells could be detected on the surface of those valves that were negative for H/Y, surprisingly. In contrast to the cell population described earlier, these cells were positive for CD49b but negative for CD49f, CD51, and CD54 (Table II).
Explanted xenograft valves
On 2 xenograft valves, a discontinuous cell lining could be detected directly adherent to the valve, staining positive for the panel characteristic for HCVE (Table II) after stimulation with cytokines. These results confirm previously mentioned observations. 10 The xenograft valve taken from a heart at the time of retransplantation was negative for all antibodies on its surface. It showed a completely denuded surface. No cell lining could be detected.
Donor-recipient differentiation of cells on valve surfaces
Explanted allograft valves
With the use of antibodies specific for donor and recipient HLA types, cells of definite recipient origin could be detected on the surface of 1 of the allograft valves studied. On 2 additional valves, cells could be detected staining only positive for the recipient panel, while being negative for nonrecipient phenotypes. Therefore it is extremely likely that these cells are of recipient origin.
Valves from transplanted hearts
When donor- and recipient-specific antibodies were used to differentiate cells on the surface of valves from transplanted hearts, no cells could be detected on the valve surface that were of definite recipient origin. However, infiltration of cells clearly shown to be of recipient origin could be demonstrated in the interior of the valves.
Explanted xenograft valves
Because all cells found on xenograft surfaces must be of recipient origin, no differentiation was attempted. We could show viable cells resembling human umbilical vein endothelial cells on the surface of 2 explants.
Expression of cell surface molecules of aortic endothelium and endocardium.
Aortic endothelium and endocardium were stained by means of the protocol mentioned above. Aortic endothelial cells were positive for the MHC class I, anti-CD31, CD34, CD44, CD49a, CD49c, CD49e, CD49f, CD51, and CD54, which is the same panel of antibodies as HCVE. However, differences in intensity could be detected. In contrast to HCVE, aortic endothelium stained negative for H/Y (Table II). Endothelial cells stained similarly to aortic endothelial cells but were negative for CD49e, CD49f, and CD51 (Table II).
HCVE in vitro
Control stains of HCVE in vitro expressed a panel of molecules identical to that of reference aortic valve endothelium.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
. 2 Also, we reported on the immunogenic properties of HCVE tested in a mixed lymphocyte endothelial cell culture. 2
In the current study, we analyzed the expression of cell surface molecules known to be involved in rejection and inflammatory responses. Yacoub 11 could not show MHC class II molecule expression in situ: we demonstrate such expression not only on the surface of valves from transplanted hearts, but also on endothelialized portions of explanted prostheses and valves. The differing expression of class II molecules may be due to different sensitivities of the mAb used or different processing of the valves, but taken together with the previously described expression in vitro and the increase after cytokine stimulation, this suggests that viable HCVE can have an immunostimulatory and sensitizing effect in recipients after allogeneic valve replacement.
An initial step in leukocyte-endothelial attachment is the "rolling" of leukocytes via selectins. 4 On valvular surfaces from normal valves and nonstimulated cells in vitro, no expression of E-selectin was observed, although it has been shown to be inducible in vitro. 2 In rejection or inflammation, leading to CD62E induction on HCVE, lymphocytes could therefore attach to the valvular cusps via these fast-acting interactions. Once host immune cells attach to the endothelial surface, they become firmly adherent via other adhesive interactions (eg, integrins). 12,13 The specificity of this arrest is assured by expression of homing molecules such as CD44, CD31, and CD34,7,14,15 which are able to regulate the avidity of the leukocyte surface receptors by "outside-in" signaling. If arrest occurs, the cells can attack the endothelium, transmigrate through the endothelial lining, or detach themselves. We have demonstrated that HCVE expresses homing receptors such as CD31, CD34, and CD44 and molecules mediating firm adhesion such as CD106, CD102, and CD54 in situ, which enable host leukocytes to attach to the endothelial lining of a valve prosthesis. In addition, certain adhesion molecules have been implicated in providing costimulatory signals.1618 The observed expression of stimulatory molecules and adhesion/ costimulatory receptors on HCVE makes the involvement of immunologic processes in the destruction of valve implants even more probable.
One of the known functions of an endothelial layer is to sustain the collagenous structures beneath it. 19 The above findings corroborate the hypothesis that deterioration of internal valve structures may be a consequence of de-endothelialization, rendering the valve susceptible to mechanical stress. Johnson and Fass 20 have shown a relative deficiency of fibronectin synthesis by porcine cardiac valvular endothelium in vitro. If such a deficiency is also to be found in HCVE in vivo, it could have a significant impact on the reparative capacity of HCVE on valvular surfaces, leading to a vicious circle of endothelial and subsequent extracellular matrix destruction on a valve whose endothelial lining has been partially destroyed because of its immunogenic properties.
Although dendritic cells are thought to play critical roles in the stimulation of alloimmune responses, its has been well established that endothelial cells are central to the stimulation of an alloimmune response too. 21-24 It is not surprising that the immunologic reaction to a partially viable allograft valve is much weaker than the allogeneic response to a fully viable organ transplant, which is rejected rapidly and completely destroyed. However, valves in transplanted hearts showed almost complete endothelial linings, whereas cryopreserved allograft valves were almost always completely denuded. In addition, while the internal structure of an allograft valve prosthesis seemed largely disrupted, valves obtained from cardiac transplants had intact internal structures. An important factor in the lack of degeneration of valves in cardiac transplants may be the immunosuppression of the recipient, which may lead to better conservation of HCVE and reduced cell loss on the valve surface, thus preserving natural repair mechanisms in the valve.
Interestingly, we could detect cells of different phenotypes (as established by immunohistochemistry) on the surface of the valves from transplanted hearts. One appeared to be quite similar to stimulated HCVE on the basis of H/Y and FVIII-related Ag molecule expression, although the integrin expression differed. The other expressed a panel of cell surface molecules that was different from both the first cell type and normal HCVE. Whereas HCVE expressed an almost identical panel of surface molecules to aortic endothelium, the panel expressed by the cell populations found on the valves from transplanted hearts was closer to that of endocardial cells. Whether this means that the endocardium or any other cell group, such as the valve endothelium itself, is able to repopulate the valve surface of a transplanted heart, however, remains purely speculative. An alternative explanation is that the differences in surface molecule expression are simply the result of different stages of cell activation and stimulation.
The endothelial marker FVIII-related Ag (von Willebrand factor), a coagulation factor, seems not to be expressed by HCVE on normal valves or in a nonstimulated stage in vitro. Once cells are stimulated by interferon- in vitro or an inflammatory or rejection processes in vivo, they do express FVIII-related Ag on their surface. 2 Also, denuded acellular surface areas of valves stained positive for FVIII-related Ag. These findings suggest that inflammation, rejection, or loss of the endothelial lining may increase the thrombogenicity of the valve surface. Our results are supported by repeated reports of thrombotic material adherent to the surface of valves after endocarditis or on biologic valve prostheses after various durations of implantation in recipients.
Xenograft valves had internal structures that were comparable with those of explanted allograft valves. It is unclear how intense and significant the immunologic reaction of xenograft valve recipients to their prosthesis is over the years. However, it is very probable that mechanical stress plays a major role in the degeneration process of this type of biologic heart valve prosthesis. Of course, this also suggests that the long-term degeneration of quickly denuded allograft valves might be due to a large extent to an increase in sensitivity to mechanical stress after their immunologic de-endothelialization. In contrast to studies previously published by other groups, we were able to show a partial re-endothelialization of xenograft and allograft valve graft surfaces in vivo. In an earlier study we 10 reported on the presence of HCVE on surfaces of explanted xenograft valves. The results of the present study corroborate these findings. We could show small HCVE-bearing areas on the surface of 3 allograft valves. To our knowledge, we are the first group reporting such re-endothelialization. Of course, the functional value of these patches is questionable. On valves of transplanted hearts, no endothelial cells of recipient origin were detected. Also, the internal structures of the valves seemed to be largely intact. However, cells of recipient origin could be detected in the interior of these valves, as has been reported by us and others. 8,25
The results reported in this study suggest that HCVE may play an important role in the long-term function of allograft valve prostheses and that stimulation of a re-endothelialization process in vivo or in vitro might be possible. This may limit degeneration due to mechanical stress. Our findings suggest that factors leading to the destruction of valve allografts may include the endothelium, which triggers immune response mechanisms in the recipient. This might lead to the loss of the endothelial lining and a subsequent structural deterioration of the valve matrix resulting from loss of nutrients and increased susceptibility to mechanical stress. Finally, the stimulation of HCVE and de-endothelialization may increase the valve's thrombogenicity, which could also trigger inflammatory processes and further destroy the prosthetic implant. In this situation, the implant must finally succumb to the combination of ongoing mechanical stress and immunologic reaction.
It seems that the implantation of a biologic heart valve prosthesis, be it a xenogeneic or even a viable allogeneic one, in essence still results in an early loss of viability because of uncontrolled immunologic reactions. It is therefore not surprising that the long-term results after xenograft or allograft valve implantation are reported to reach comparable plateaus while the immunologically protected, fully viable aortic valve in a cardiac transplant seems to be almost completely resistant to this degeneration. 26 Therefore, since immunosuppression of valve recipients is not an alternative but storage of allograft valves in liquid nitrogen is possible for any amount of time, HLA-matching of prospective recipients to their valve could be a feasible approach toward a solution to this problem. This may lead to an increase in the long-term survival of allograft valves after implantation. This hypothesis is corroborated by the observed superior long-term results in HLA-matched over mismatched organs (eg, cardiac transplantations). 27-29 In xenografts, reseeding of autologous endothelial cells to prevent structural degeneration and decrease thrombogenicity may be used to achieve a similar effect and has already been attempted in vascular prostheses. 30
|
Acknowledgments |
|---|
|
References |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  V. J. Nina, P. M. Pomerantzeff, I. S. Casagrande, D. T Cheung, C. M. Brandao, and S. A Oliveira Comparative Study of the L-Hydro Process and Glutaraldehyde Preservation Asian Cardiovasc Thorac Ann, September 1, 2005; 13(3): 203 - 207. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. S. Farivar, L. H. Cohn, E. G. Soltesz, T. Mihaljevic, J. D. Rawn, and J. G. Byrne Transcriptional profiling and growth kinetics of endothelium reveals differences between cells derived from porcine aorta versus aortic valve Eur. J. Cardiothorac. Surg., October 1, 2003; 24(4): 527 - 534. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. S. Farivar, F. Filsoufi, and D. H. Adams Mechanisms of Gal{alpha}1-3Gal{beta}1-4GlcNAc-R ({alpha}Gal) expression on porcine valve endothelial cells J. Thorac. Cardiovasc. Surg., February 1, 2003; 125(2): 306 - 314. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. R. Clarke Presidential address: Value, viability, and valves J. Thorac. Cardiovasc. Surg., July 1, 2002; 124(1): 1 - 6. [Full Text] [PDF]
|
|
|
|
|
|
K. Jansson, L. Bengtsson, J. Swedenborg, and A. Haegerstrand In vitro endothelialization of bioprosthetic heart valves provides a cell monolayer with proliferative capacities and resistance to pulsatile flow J. Thorac. Cardiovasc. Surg., January 1, 2001; 121(1): 0108 - 115. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. Raanani, T. M. Yau, T. E. David, G. Dellgren, B. D. Sonnenberg, and A. Omran Risk factors for late pulmonary homograft stenosis after the Ross procedure Ann. Thorac. Surg., December 1, 2000; 70(6): 1953 - 1957. [Abstract] [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 |
