Decellularized mitral valve in a long-term sheep model

Decellularized mitral valve in a long-term sheep model Abstract OBJECTIVES The objective of this study was to evaluate surgical handling, in vivo hemodynamic performance and morphological characteristics of decellularized mitral valves (DMVs) in a long-term sheep model. METHODS Ovine mitral valves were decellularized using detergents and β-mercaptoethanol. Orthotopic implantations were performed in 6-month-old sheep (41.3 ± 1.2 kg, n = 11) without annulus reinforcement. Commercially available stented porcine aortic valves [biological mitral valve (BMV), n = 3] were implanted conventionally and used as controls. Valve function was evaluated by transoesophageal echocardiography and explants were investigated by a routine bright field microscopy and immunofluorescent histology. RESULTS During implantation, 2 DMVs required cleft closure of the anterior leaflet. All valves were competent on water test and early postoperative transoesophageal echocardiography. Six animals (DMV, n = 4; BMV, n = 2) survived 12 months. Six animals died within the first 4 months due to valve-related complications. At 12 months, transoesophageal echocardiography revealed severe degeneration in all BMVs. Macroscopically, BMV revealed calcification at the commissures and leaflet insertion area. Histological examination showed sporadic cells negative for endothelial nitric oxide synthase, von Willebrand factor and CD45 on their surface. In contrast, DMV showed no calcification or stenosis, and the regurgitation was trivial to moderate in all animals. Fibrotic hardening occurred only along the suture line of the valve annulus, immunostaining revealed collagen IV covering the entire leaflet surface and a repopulation with endothelial cells. CONCLUSIONS Surgical implantation of DMV is feasible and results in good early graft function. Additional in vivo investigations are required to minimize the procedure-related complications and to increase the reproducibility of surgical implantation. Degenerative profile of allogeneic DMV is superior to commercially available porcine aortic prosthesis. Tissue engineering, Mitral valve replacement, Decellularization INTRODUCTION Mitral valve replacement is a traditional surgical approach to treat severe mitral valve defects that are not suitable for repair. Thrombosis of mechanical prostheses and degeneration of xenogeneic biological valves are the main concerns after implantation [1]. Implantation of fresh and cryopreserved allogeneic valves was thought to overcome the drawbacks of mechanical valves and rapid degeneration of biological valve options, especially in young population. However, mitral valve allografts are rarely used for mitral valve replacement because of the complexity of surgical implantation and limited durability [2]. Degeneration of allogeneic tissue is mostly attributed to the remaining cells that are the target of immunological response in valve recipients [3]. As a consequence, the idea of tissue decellularization was already adopted for several other valve positions, as for example, the pulmonary and aortic valves. Moreover, the clinical use of such grafts resulted in enhanced freedom from valve degeneration and lower reoperation rate when implanted in the pulmonary position and compared with cryopreserved homografts or xenogeneic grafts [4–7]. For the mitral valve, some efforts were taken to provide only the most needed structures of mitral valve—chordae tendineae [8, 9]. Our group recently described the promising results for decellularization of the entire mitral valve [10, 11]. Decellularization of the mitral grafts using a modified protocol revealed acceptable mechanical strength and showed no toxicity. The aim of this study was to compare decellularized mitral valve (DMV) allografts with commercially available conventional stented porcine aortic valve prostheses in the orthotopic position in a long-term ovine animal model. MATERIALS AND METHODS Harvesting and decellularization Ovine mitral valves were harvested at the local slaughterhouse shortly after death. The valves were excised with a small rim of the left atrium of 3–4 mm including the chordae tendineae and papillary muscles. For surgical reason, just grafts with chordae attached only to 2 papillary muscles were further used for implantation. The valves were stored in phosphate-buffered saline (Sigma-Aldrich, St Louis, MO, USA) at 4 °C for 4–6 h until being decellularized using a previously described method [10]. Briefly, after being disinfected in povidone iodide (Braunol®; B. Braun, Melsungen, Germany) for 5 min followed by 20 min of washing in phosphate-buffered saline, the whole valves were then decellularized using sodium deoxycholate (SD; Sigma-Aldrich), sodium dodecyl sulphate (SDS ultra pure; Carl Roth, Karlsruhe, Germany) and β-mercaptoethanol (β-ME; AppliChem, Darmstadt, Germany). Subsequently, washing was done in ten 12-h steps; the overall duration of the decellularization process was 154 h. All decellularized valves were stored in sterile conditions at +4 °C. Orthotopic implantation The operations were performed under combined endotracheal and intravenous anaesthesia in 6-month-old sheep (weighing 41.3 ± 1.2 kg). A left-sided thoracotomy in the fourth intercostal space was followed by establishing extracorporeal circulation through cannulation of the descending aorta and right atrial appendage. After cross-clamping of the aorta, Buckberg cardioplegic solution was applied using a combined cardioplegic cannula. Opening of the left atrium was followed by excision of the cusps and chordae of the native mitral valve. We used slightly oversized DMV (approximately 4–5 mm) for the implantation to compensate the suture line along the annulus. First, we estimated the level of implantation at the DMV papillary muscles. For this purpose, the annulus of the DMV was positioned onto the native annulus and papillary muscles were introduced into the left ventricle. Afterwards, U-stitches with polytetrafluoroethylene pledgets (3 × 6 mm) were initially placed on the appropriate position of each native annulus and then on the DMV papillary muscles. In addition, 2 guiding polypropylene 4-0 sutures were placed at the trigons of the aortomitral transition to prevent dislocation of the valve during the implantation. These sutures were then continued into the simple running suture along the annulus of the mitral valve (Figs 1 and 2). After the implantation, a water test was performed to check the valve competence. Figure 1: View largeDownload slide Schematic of the operation. The pledgeted U-sutures (left picture) and running polypropylene suture along the annulus (right picture) are shown. Ao: aorta ascendens. Figure 1: View largeDownload slide Schematic of the operation. The pledgeted U-sutures (left picture) and running polypropylene suture along the annulus (right picture) are shown. Ao: aorta ascendens. Figure 2: View largeDownload slide Surgical technique. (A) Sizing of the DMV: the oversized (+4–5 mm) grafts were implanted. (B) Running polypropylene 4-0 suture along the annulus. (C) End view of the implanted decellularized mitral valves. Figure 2: View largeDownload slide Surgical technique. (A) Sizing of the DMV: the oversized (+4–5 mm) grafts were implanted. (B) Running polypropylene 4-0 suture along the annulus. (C) End view of the implanted decellularized mitral valves. In the control group, commercially available stented porcine aortic prostheses were implanted in a traditional way. In all animals, prostheses with a diameter of 23 mm were implanted. Thereafter, the left atrium was closed using the double running polypropylene 5-0 suture and the aortic clamp was opened. Transoesophageal echocardiography (TOE) was performed to assess the performance of the grafts after weaning from the heart–lung machine (Philips CX50 ultrasound system with Philips X7–2 T probe, Koninklijke Philips N.V., Amsterdam, Netherlands), within the first 2 weeks postoperatively and shortly before the animals were sacrificed. Histological staining A half of the A2 zone of each explanted valve (or one of the leaflets of xenograft) was embedded in paraffin, sliced into 8-μm-thick sections and stained by haematoxylin–eosin, Movat’s pentachromic, van Gieson and von Kossa stains. Slides were analysed using a routine bright field microscopy (Olympus BX40, Olympus, Tokyo, Japan); with colour camera (AxioCam MRc, Carl Zeiss, Jena, Germany). Immunofluorescent staining The other half of the A2 zone of each explanted valve (and another leaflet of xenograft) was snap-frozen in liquid nitrogen, mounted in Tissue-Tek (Sakura Finetek Europe, AV Alphen an den Rijn, Netherlands) and cut into 7-μm-thick sections. All samples were stained against the following antigens: DNA [4,6-diamidino-2-phenylindoledihydrochloride (DAPI) stain, Invitrogen, Carlsbad, CA, USA] Collagen IV (clone CI22, DAKO, Hamburg, Germany) Collagen I C2345 clone COL-I (Sigma-Aldrich Chemie GmbH, Munich, Germany) CD31 [monoclonal mouse immunoglobulin G (IgG)2a, Serotec, Raleigh, NC, USA] von Willebrand factor (polyclonal rabbit IgG, DAKO) Endothelial nitric oxide synthase (monoclonal mouse IgG1, eNOS/NOS Type III, BD Transduction Laboratories, San Jose, CA, USA) CD45 [mouse monoclonal (clone OX-1) IgG, BioRad, Oxford, UK] Procollagen (monoclonal mouse IgG1, M-38-c, Developmental Studies Hybridoma Bank, Iowa City, IA, USA) The images were obtained using the microscope AxioObserver A1, with the camera AxioCamMRm (Carl Zeiss, Jena, Germany). RESULTS Surgical performance The mean bypass time was 108.5 ± 32.3 min, and the mean cross-clamp time was 64.5 ± 24.4 min. The pleural drainage was removed on the 2nd postoperative day under suction. During the follow-up, no sheep showed fever or any other abnormalities. There was no valve-related early postoperative mortality. Fourteen animals (Animal nos. 1–14) survived the first 2 weeks (Table 1). Table 1: Distribution of the animals Animal number  Type of graft  Survival (weeks)  Cause of death  LD EF (%)  LD MR  1  DMV  56  Sacrificed  55  II  2  DMV  52  Sacrificed  60  II  3  DMV  30  Sacrificed  30  II  4  DMV  64  Sacrificed  70  II–III  5  BMV  61  Sacrificed  60  I  6  BMV  61  Sacrificed  50  0  7  DMV  18  Heart failure due to MR  40  0–I  8  DMV  8  Rupture of the donor papillary muscle  55  0–I  9  DMV  18  Heart failure due to MR  40  0–I  10  BMV  14  Heart failure  n/d  0–I  11  DMV  11  Rupture of chordae tendineae  60  0–I  12  DMV  2  Rupture of papillary muscle (loose suture)  n/d  0–I  13  DMV  3  Partial rupture of the annulus  n/d  0–I  14  DMV  40  Heart failure due to MR  60  I  Animal number  Type of graft  Survival (weeks)  Cause of death  LD EF (%)  LD MR  1  DMV  56  Sacrificed  55  II  2  DMV  52  Sacrificed  60  II  3  DMV  30  Sacrificed  30  II  4  DMV  64  Sacrificed  70  II–III  5  BMV  61  Sacrificed  60  I  6  BMV  61  Sacrificed  50  0  7  DMV  18  Heart failure due to MR  40  0–I  8  DMV  8  Rupture of the donor papillary muscle  55  0–I  9  DMV  18  Heart failure due to MR  40  0–I  10  BMV  14  Heart failure  n/d  0–I  11  DMV  11  Rupture of chordae tendineae  60  0–I  12  DMV  2  Rupture of papillary muscle (loose suture)  n/d  0–I  13  DMV  3  Partial rupture of the annulus  n/d  0–I  14  DMV  40  Heart failure due to MR  60  I  BMV: biological mitral valve; DMV: decellularized mitral valve; LD: last documented; EF: ejection fraction; MR: mitral regurgitation. Table 1: Distribution of the animals Animal number  Type of graft  Survival (weeks)  Cause of death  LD EF (%)  LD MR  1  DMV  56  Sacrificed  55  II  2  DMV  52  Sacrificed  60  II  3  DMV  30  Sacrificed  30  II  4  DMV  64  Sacrificed  70  II–III  5  BMV  61  Sacrificed  60  I  6  BMV  61  Sacrificed  50  0  7  DMV  18  Heart failure due to MR  40  0–I  8  DMV  8  Rupture of the donor papillary muscle  55  0–I  9  DMV  18  Heart failure due to MR  40  0–I  10  BMV  14  Heart failure  n/d  0–I  11  DMV  11  Rupture of chordae tendineae  60  0–I  12  DMV  2  Rupture of papillary muscle (loose suture)  n/d  0–I  13  DMV  3  Partial rupture of the annulus  n/d  0–I  14  DMV  40  Heart failure due to MR  60  I  Animal number  Type of graft  Survival (weeks)  Cause of death  LD EF (%)  LD MR  1  DMV  56  Sacrificed  55  II  2  DMV  52  Sacrificed  60  II  3  DMV  30  Sacrificed  30  II  4  DMV  64  Sacrificed  70  II–III  5  BMV  61  Sacrificed  60  I  6  BMV  61  Sacrificed  50  0  7  DMV  18  Heart failure due to MR  40  0–I  8  DMV  8  Rupture of the donor papillary muscle  55  0–I  9  DMV  18  Heart failure due to MR  40  0–I  10  BMV  14  Heart failure  n/d  0–I  11  DMV  11  Rupture of chordae tendineae  60  0–I  12  DMV  2  Rupture of papillary muscle (loose suture)  n/d  0–I  13  DMV  3  Partial rupture of the annulus  n/d  0–I  14  DMV  40  Heart failure due to MR  60  I  BMV: biological mitral valve; DMV: decellularized mitral valve; LD: last documented; EF: ejection fraction; MR: mitral regurgitation. There were 4 surgery-related deaths: 1 (Animal no. 13, after 3 weeks) due to partial rupture of the annulus over 5–7 mm in the P1/P2 zone resulting in a significant paravalvular leak, 2 deaths (Animal no. 12 after 2 weeks and Animal no. 8 after 8 weeks) due to rupture of the decellularized papillary muscle (one because of loose suture) resulting in severe mitral insufficiency, 1 death (Animal no. 11 after 11 weeks) due to rupture of the chordae tendineae. Three animals with documented moderate transvalvular regurgitation died after 4 months (Animal nos. 7 and 9) and 7 months (Animal no. 14). One animal from the control group died after 4 months (Animal no. 10) having clinical signs of heart failure. Six animals (DMV, n = 4 and BMV, n = 2) survived for 12 months. Because of the normal growth, the weight of the animals increased from initial 41.3 ± 1.2 kg to 63 ± 1.1 kg. The final TOE of the DMV animals (Animal nos. 1–4) revealed low transvalvular gradients with trivial-to-moderate insufficiency due to central jet (Fig. 3). Sonographic signs of fibrosis were only seen along the mitral valve annulus. In contrast, in BMV animals (Animal nos. 5 and 6), TOE showed increased transvalvular gradients (max/mean, Animal no. 5: 19/14 mmHg and Animal no. 6: 17/11 mmHg) and moderate enlargement of the left atrium as a result of moderate-to-severe mitral stenosis. Figure 3: View largeDownload slide Peak pressure gradient on the mitral valve substitutes during the follow-up. The maximum gradients on the biological mitral valves increased significantly, while those on the DMVs were stable or decreasing. DMV: decellularized mitral valves. Figure 3: View largeDownload slide Peak pressure gradient on the mitral valve substitutes during the follow-up. The maximum gradients on the biological mitral valves increased significantly, while those on the DMVs were stable or decreasing. DMV: decellularized mitral valves. Macroscopic findings DMVs showed soft leaflets comparable with the native ones with some film-like thrombotic depositions typical for early deceased animals. The atrial and the ventricular surfaces along the suture line were completely covered with neointima starting from 8 weeks postoperatively (Animal no. 8), so that the suture material was completely invisible (Fig. 4C). The 4 DMVs explanted at 12 months revealed almost circular moderate hardening along the mitral valve annulus, without macroscopic signs of calcification. Animal no. 4 had a 3-mm paravalvular leak due to local disruption of the stitch at the posterior portion of the annulus; all other samples did not show any defects. The chordae tendineae were also not calcified. Samples, explanted at 12 months, showed almost completely organized donor papillary muscles that formed a solid scar together with the tips of the recipient’s papillary muscle (Fig. 4D). Figure 4: View largeDownload slide Decellularized mitral valve (DMV) and biological mitral valve. (A) DMV after the end of decellularization. (B) Biological mitral valve at 61 weeks postoperatively, arrow shows the calcified regions. (C) Atrial view of the DMV at 56 weeks postoperatively. (D) Ventricular view of the DMV at 56 weeks postoperatively, the soft and transparent leaflets and a firm scar on the papillary muscle are seen. Figure 4: View largeDownload slide Decellularized mitral valve (DMV) and biological mitral valve. (A) DMV after the end of decellularization. (B) Biological mitral valve at 61 weeks postoperatively, arrow shows the calcified regions. (C) Atrial view of the DMV at 56 weeks postoperatively. (D) Ventricular view of the DMV at 56 weeks postoperatively, the soft and transparent leaflets and a firm scar on the papillary muscle are seen. Xenografts showed typical degeneration and calcification at the commissures and sinuses with subsequent severe reduction of leaflet motion (Fig. 4B). Histological findings The routine light microscopy of DMVs revealed preserved structure of all mitral valve components. The 4-layered leaflets showed typical organization of the collagen and elastin fibres, and there was no difference in the extracellular matrix in comparison with the native ovine mitral valve: the leaflets had typical 4-layered structure and showed no matrix disturbance (haematoxylin-eosin, van Gieson and pentachromic staining) (Fig. 5A–C). The cell amount in the proximal parts of the leaflets was comparable with that of the native ones, descending to the free margin—distally only the endothelial layer and several cells in pars spongiosa could be seen. The repopulation of the leaflet differed from layer to layer—the pars ventricularis showed the best performance, followed by pars spongiosa and pars atrialis. Pars fibrosa at 12 months had only several cells in the proximal areas but almost none were seen further distally. However, in comparison with the animals which died during the follow-up, the cell count was constantly increasing according to the time in vivo. Figure 5: View largeDownload slide Histology of the decellularized mitral valve (DMV) and the biological mitral valve (BMV). (A–C)—DMV and (D–F) BMV. (A, D) Movat´s pentachrome; (B, E) van Gieson staining; (C, F) von Kossa staining. The structure of the BMV is completely disturbed due to calcification, while the DMV shows almost repopulated typical 4-layered leaflet with no signs of calcification, but preserved elastin fibres. Figure 5: View largeDownload slide Histology of the decellularized mitral valve (DMV) and the biological mitral valve (BMV). (A–C)—DMV and (D–F) BMV. (A, D) Movat´s pentachrome; (B, E) van Gieson staining; (C, F) von Kossa staining. The structure of the BMV is completely disturbed due to calcification, while the DMV shows almost repopulated typical 4-layered leaflet with no signs of calcification, but preserved elastin fibres. Von Kossa staining revealed small single spots of calcification, which were around 4–12 µm in diameter and tended to be localized in the proximal parts of leaflets (Fig. 5). No calcification was seen in the chordae tendineae. The immunofluorescent staining of Animal nos. 1–4 showed typical structure of collagen I, collagen IV covering the entire surface of all valve components as well as several interstitial cells positive for procollagen. At the same time, repopulation of the decellularized tissues of DMVs was clearly seen. Thus, at 12 months, the endothelial layer was completed along the entire surface of the leaflets and chordae with cobblestone-like cells that were shown to express CD31, von Willebrandt factor and eNOS (Fig. 6). The small number of cells, positive for CD45, was seen forming groups mostly in the proximal parts of the leaflets, along the suture and in scars on papillary muscles (Fig. 7). Figure 6: View largeDownload slide Immunostaining of the decellularized mitral valve (DMV) and the biological mitral valve (BMV). (A and B) DMV and (C and D) BMV. (A, C) von Willenbrandt factor and DNA; (B, D) collagen IV and DNA. DMV, in contrast to BMV, shows complete collagen IV layer with von Willenbrandt factor-positive cells on it. Figure 6: View largeDownload slide Immunostaining of the decellularized mitral valve (DMV) and the biological mitral valve (BMV). (A and B) DMV and (C and D) BMV. (A, C) von Willenbrandt factor and DNA; (B, D) collagen IV and DNA. DMV, in contrast to BMV, shows complete collagen IV layer with von Willenbrandt factor-positive cells on it. Figure 7: View largeDownload slide Immunostaining of the decellularized mitral valve (DMV) and the biological mitral valve (BMV). (A and B) DMV and (C and D) BMV. (A, C) CD45 and DNA; (B, D) procollagen I and DNA. DMV shows collagen-producing cells within the leaflets. BMV has almost no cells, positive for procollagen I. The CD45-positive cells in the BMV tend to form nodules (picture of the region, close to the annulus). BMV has only several CD45-positive cells on the surface. Ao: Aorta ascendens. Figure 7: View largeDownload slide Immunostaining of the decellularized mitral valve (DMV) and the biological mitral valve (BMV). (A and B) DMV and (C and D) BMV. (A, C) CD45 and DNA; (B, D) procollagen I and DNA. DMV shows collagen-producing cells within the leaflets. BMV has almost no cells, positive for procollagen I. The CD45-positive cells in the BMV tend to form nodules (picture of the region, close to the annulus). BMV has only several CD45-positive cells on the surface. Ao: Aorta ascendens. Xenografts showed lower amount of elastin fibres compared with DMVs but with the presence of most of the glycosaminoglycans. Collagen fibres and the whole 4-layered structure were preserved in regions not destructed by rough calcification spots over 1 mm in diameter (Fig. 5). However, in non-degenerated regions, von Kossa staining revealed multiple small calcified spots throughout the leaflet. Immunofluorescent staining revealed several regions covered with collagen IV with single CD31, von Willebrandt factor and eNOS-positive cells (Fig. 6), but most part of the surface was free from any cellular elements or had some cells negative for all tested antigens. Rare CD45-positive cells were also found on the surface and within the leaflets (Fig. 7). DISCUSSION This study demonstrates the feasibility of complete DMV implantation in an allogeneic ovine growing model and represents a first step towards long-term testing of DMV under physiological conditions. Our implantation procedure was based on previously described techniques of fresh and cryopreserved human mitral valves [12–14]. Still, we faced several significant problems in our animal model: rupture of the annulus, chordae and papillary muscle leading to mitral regurgitation in a relevant number of implantations within 12 months of follow-up. Previously performed biomechanical testing showed no significant difference in mechanical strength between native and DMV leaflets, but due to methodological limitations, only the A2 zone was investigated [10, 15, 16]. The complexity of the mitral valve structure led to several problems not only for the establishment of a decellularization protocol but also for the in vitro testing of the functionality of the isolated mitral valve components or the whole valvular graft. Variability of chordae structure and papillary muscles did not allow performing adequate mechanical test in vitro. Optimizing the implantation technique In vivo, we observed problems with the posterior part of the annulus, which is relatively weak, both in the native valves and in the DMVs. Here, we have seen rupture of the tissue resulting in paravalvular leak in 2 animals, in both animals the polypropylene suture was torn off at the recipient’s P1/P2 zone of the annulus and subsequently the suture thread remained on the DMV. One death occurred due to rupture of the chordae tendineae at the P2 zone causing severe regurgitation. Potentially, a concomitant implantation of an additional artificial chordae would be helpful to prevent flail of the posterior leaflet if a decellularized chordae ruptures. However, in contrast to our concerns at the beginning of the study, we observed chordae rupture in only 1 case. Another problem was identified at the level of the papillary muscles. Two animals died due to the rupture on this site after side-by-side implantation using pledged U-suture. We believe that the muscle fixation within a transected recipient’s papillary muscle could be an alternative and may facilitate the ingrowth of the decellularized papillary muscle head. However, care should be taken while using this technique to avoid shortening of the chordae tendineae. Growth aspects We did not use artificial rings for stabilization of the mitral valve annulus to estimate the potential of the DMVs in growing individuals. The advantages of allogeneic decellularized aortic and pulmonary valve grafts over the traditional valves in children have been already reported by several groups [4–7, 17]. Moreover, these grafts revealed in vivo remodelling and adaptive growth with preservation of excellent function during physiological development of the young patients. Our growing model showed that the DMV surface was almost completely reseeded with the host cells within 12 months. However, TOE at 12 months showed a slight increase in valve regurgitation, probably because of the increase in the size of the annulus. Mitral regurgitation observed in those animals is the result of heart growth on the one hand and the result of too slow repopulation and remodelling of the DMV matrix because of its high tissue volume on the other hand. It has been already shown that the mitral valve leaflet is at least 3 times thicker as one of the aortic valves, and the area of the leaflet is also significantly larger [10, 15]. This hypothesis is supported by the fact that the cell count seen in the explanted allografts was not comparable with the native mitral valve, especially in the pars fibrosa. Because of these findings, the use in children appears questionable, as DMV was not reseeded and, probably, could not be remodelled rapidly enough during somatic growth. However, it should be taken into account that the weight of the animals increased about 50% (>20 kg) within 12 months, which by far exceeds the normal human growth. DMV implantation appears much easier in adult patients because it allows to perform annulus stabilization with a ring, which may reduce the risk of valve insufficiency development in the long-term follow-up and minimize annulus-related complications. In addition to leaflet quality and annulus fixation, the configuration of papillary muscles and the number and the distribution of the chordae tendineae are important and need to be carefully considered. Modern imaging techniques such as high-resolution computed tomography or magnetic resonance imaging will further facilitate planning of concomitant procedures such as prophylactic neochordae implantation. Histological aspects The porcine valves had typical calcification profile—on the commissures and at the leaflets insertion sites, so that their mobility was significantly reduced resulting in high transvalvular gradients. At the same time, the BMVs showed the impossibility of their repopulation and only degenerative changes of the structure. At 12 months, DMV showed full coverage with the basal membrane positive for the collagen IV as well as the complete repopulation of the entire leaflet surface, chordae and papillary muscles with the endothelial cells. However, within the leaflet, the recellularization was not complete, which differs significantly from the aortic allografts [18]. The least reseeded parts were the chordae tendineae and pars fibrosa of the leaflet. On the other hand, the repopulation was slowly ongoing: the DMVs obtained from the dead animals showed almost no cells within the leaflet, but at 12 months up to 20% of pars fibrosa and at least two-third of the pars spongiosa were reseeded. A clear migration trend could be seen—the cell amount was maximal on and close to the surface, decreasing distally and towards the centre of the pars fibrosa. Probably, some additional methods to enhance the repopulation could be helpful to increase the reseeding speed and to stimulate the remodelling [19]. CONCLUSION We demonstrated the feasibility of orthotopic implantation of DMVs in an ovine growing model. The novel substitute was shown to be superior to a commercially available porcine xenograft regarding resistance to calcification and degeneration and demonstrated good, but not complete, recellularization. Further in vivo tests are pending to increase the reproducibility of DMV implantation and to evaluate techniques for enhancing faster recellularization with respect to valve growth and long-term durability. Funding This work was supported by CORTISS Hannover, Herz- und Gewebeforschungs GmbH. Conflict of interest: none declared. REFERENCES 1 Head SJ, Çelik M, Kappetein AP. Mechanical versus bioprosthetic aortic valve replacement. Eur Heart J  2017; 38: 2183– 91. Google Scholar CrossRef Search ADS PubMed  2 Nappi F. Cryopreserved mitral homograft valve: 19 years experience. JACC Cardiovasc Interv  2014; 7: S58. Google Scholar CrossRef Search ADS   3 Schoen FJ, Levy RJ. Calcification of tissue heart valve substitutes: progress toward understanding and prevention. Ann Thorac Surg  2005; 79: 1072– 80. Google Scholar CrossRef Search ADS PubMed  4 da Costa FDA, Dohmen PM, Vieira E, Lopes SV, Colatusso C, Pereira EWL et al.   Ross Operation with decellularized pulmonary allografts: medium-term results. Rev Bras Cir Cardiovasc  2007; 22: 454– 62. Google Scholar CrossRef Search ADS PubMed  5 Da Costa FD., Costa AC, Prestes R, Domanski AC, Balbi EM, Ferreira AD et al.   The early and midterm function of decellularized aortic valve allografts. Ann Thorac Surg  2010; 90: 1854– 60. Google Scholar CrossRef Search ADS PubMed  6 Cebotari S, Tudorache I, Ciubotaru A, Boethig D, Sarikouch S, Goerler A et al.   Use of fresh decellularized allografts for pulmonary valve replacement may reduce the reoperation rate in children and young adults: early report. Circulation  2011; 124: S115. Google Scholar CrossRef Search ADS PubMed  7 Tudorache I, Horke A, Cebotari S, Sarikouch S, Boethig D, Breymann T et al.   Decellularized aortic homografts for aortic valve and aorta ascendens replacement. Eur J Cardiothorac Surg  2016; 50: 89– 97. Google Scholar CrossRef Search ADS PubMed  8 Shi Y, Vesely I. Characterization of statically loaded tissue-engineered mitral valve chordae tendineae. J Biomed Mater Res A  2004; 69: 26– 39. Google Scholar CrossRef Search ADS PubMed  9 Gong W, Li S, Lei D, Huang P, Yuan Z, You Z et al.   Tissue-engineered mitral valve chordae tendineae: biomechanical and biological characterization of decellularized porcine chordae. J Mech Behav Biomed Mater  2016; 56: 205– 17. Google Scholar CrossRef Search ADS PubMed  10 Iablonskii P, Cebotari S, Tudorache I, Granados M, Morticelli L, Goecke T et al.   Tissue-engineered mitral valve: morphology and biomechanics. Interact CardioVasc Thorac Surg  2015; 20: 712– 19. Google Scholar CrossRef Search ADS PubMed  11 Granados M, Morticelli L, Andriopoulou S, Kalozoumis P, Pflaum M, Iablonskii P et al.   Development and characterization of a porcine mitral valve scaffold for tissue engineering. J Cardiovasc Transl Res  2017; 10: 1– 17. Google Scholar CrossRef Search ADS PubMed  12 Hubka M, Siska K, Brozman M, Holec V. Replacement of mitral and tricuspid valves by homograft. J Thorac Cardiovasc Surg  1966; 51: 195– 204. Google Scholar PubMed  13 Baird RJ, Williams WG, Spratt EH, Cohoon WJ. Experimental homograft replacement of the mitral valve. Can J Surg  1969; 12: 144– 54. Google Scholar PubMed  14 Plunkett MD, Schneider DJ, Shah JJ, Bash SE, Bond LM, Geiss DM. Homograft replacement of mitral valve in children. Ann Thorac Surg  1998; 66: 849– 52. Google Scholar CrossRef Search ADS PubMed  15 Grande-Allen KJ, Liao J. The heterogeneous biomechanics and mechanobiology of the mitral valve: implications for tissue engineering. Curr Cardiol Rep  2011; 13: 113– 20. Google Scholar CrossRef Search ADS PubMed  16 Roberts N, Morticelli L, Jin Z, Ingham E, Korossis S. Regional biomechanical and histological characterization of the mitral valve apparatus: implications for mitral repair strategies. J Biomech  2016; 49: 2491– 501. Google Scholar CrossRef Search ADS PubMed  17 Sarikouch S, Horke A, Tudorache I, Beerbaum P, Westhoff-Bleck M, Boethig D et al.   Decellularized fresh homografts for pulmonary valve replacement: a decade of clinical experience. Eur J Cardiothorac Surg  2016; 50: 281– 90. Google Scholar CrossRef Search ADS PubMed  18 Baraki H, Tudorache I, Braun M, Höffler K, Görler A, Lichtenberg A et al.   Orthotopic replacement of the aortic valve with decellularized allograft in a sheep model. Biomaterials  2009; 30: 6240– 6. Google Scholar CrossRef Search ADS PubMed  19 Assmann A, Delfs C, Munakata H, Schiffer F, Horstkötter K, Huynh K et al.   Acceleration of autologous invivo recellularization of decellularized aortic conduits by fibronectin surface coating. Biomaterials  2013; 34: 6015– 26. Google Scholar CrossRef Search ADS PubMed  © The Author(s) 2018. Published by Oxford University Press on behalf of the European Association for Cardio-Thoracic Surgery. All rights reserved. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png European Journal of Cardio-Thoracic Surgery Oxford University Press

Loading next page...
 
/lp/ou_press/decellularized-mitral-valve-in-a-long-term-sheep-model-zfKCvNa0Ge
Publisher
Oxford University Press
Copyright
© The Author(s) 2018. Published by Oxford University Press on behalf of the European Association for Cardio-Thoracic Surgery. All rights reserved.
ISSN
1010-7940
eISSN
1873-734X
D.O.I.
10.1093/ejcts/ezx485
Publisher site
See Article on Publisher Site

Abstract

Abstract OBJECTIVES The objective of this study was to evaluate surgical handling, in vivo hemodynamic performance and morphological characteristics of decellularized mitral valves (DMVs) in a long-term sheep model. METHODS Ovine mitral valves were decellularized using detergents and β-mercaptoethanol. Orthotopic implantations were performed in 6-month-old sheep (41.3 ± 1.2 kg, n = 11) without annulus reinforcement. Commercially available stented porcine aortic valves [biological mitral valve (BMV), n = 3] were implanted conventionally and used as controls. Valve function was evaluated by transoesophageal echocardiography and explants were investigated by a routine bright field microscopy and immunofluorescent histology. RESULTS During implantation, 2 DMVs required cleft closure of the anterior leaflet. All valves were competent on water test and early postoperative transoesophageal echocardiography. Six animals (DMV, n = 4; BMV, n = 2) survived 12 months. Six animals died within the first 4 months due to valve-related complications. At 12 months, transoesophageal echocardiography revealed severe degeneration in all BMVs. Macroscopically, BMV revealed calcification at the commissures and leaflet insertion area. Histological examination showed sporadic cells negative for endothelial nitric oxide synthase, von Willebrand factor and CD45 on their surface. In contrast, DMV showed no calcification or stenosis, and the regurgitation was trivial to moderate in all animals. Fibrotic hardening occurred only along the suture line of the valve annulus, immunostaining revealed collagen IV covering the entire leaflet surface and a repopulation with endothelial cells. CONCLUSIONS Surgical implantation of DMV is feasible and results in good early graft function. Additional in vivo investigations are required to minimize the procedure-related complications and to increase the reproducibility of surgical implantation. Degenerative profile of allogeneic DMV is superior to commercially available porcine aortic prosthesis. Tissue engineering, Mitral valve replacement, Decellularization INTRODUCTION Mitral valve replacement is a traditional surgical approach to treat severe mitral valve defects that are not suitable for repair. Thrombosis of mechanical prostheses and degeneration of xenogeneic biological valves are the main concerns after implantation [1]. Implantation of fresh and cryopreserved allogeneic valves was thought to overcome the drawbacks of mechanical valves and rapid degeneration of biological valve options, especially in young population. However, mitral valve allografts are rarely used for mitral valve replacement because of the complexity of surgical implantation and limited durability [2]. Degeneration of allogeneic tissue is mostly attributed to the remaining cells that are the target of immunological response in valve recipients [3]. As a consequence, the idea of tissue decellularization was already adopted for several other valve positions, as for example, the pulmonary and aortic valves. Moreover, the clinical use of such grafts resulted in enhanced freedom from valve degeneration and lower reoperation rate when implanted in the pulmonary position and compared with cryopreserved homografts or xenogeneic grafts [4–7]. For the mitral valve, some efforts were taken to provide only the most needed structures of mitral valve—chordae tendineae [8, 9]. Our group recently described the promising results for decellularization of the entire mitral valve [10, 11]. Decellularization of the mitral grafts using a modified protocol revealed acceptable mechanical strength and showed no toxicity. The aim of this study was to compare decellularized mitral valve (DMV) allografts with commercially available conventional stented porcine aortic valve prostheses in the orthotopic position in a long-term ovine animal model. MATERIALS AND METHODS Harvesting and decellularization Ovine mitral valves were harvested at the local slaughterhouse shortly after death. The valves were excised with a small rim of the left atrium of 3–4 mm including the chordae tendineae and papillary muscles. For surgical reason, just grafts with chordae attached only to 2 papillary muscles were further used for implantation. The valves were stored in phosphate-buffered saline (Sigma-Aldrich, St Louis, MO, USA) at 4 °C for 4–6 h until being decellularized using a previously described method [10]. Briefly, after being disinfected in povidone iodide (Braunol®; B. Braun, Melsungen, Germany) for 5 min followed by 20 min of washing in phosphate-buffered saline, the whole valves were then decellularized using sodium deoxycholate (SD; Sigma-Aldrich), sodium dodecyl sulphate (SDS ultra pure; Carl Roth, Karlsruhe, Germany) and β-mercaptoethanol (β-ME; AppliChem, Darmstadt, Germany). Subsequently, washing was done in ten 12-h steps; the overall duration of the decellularization process was 154 h. All decellularized valves were stored in sterile conditions at +4 °C. Orthotopic implantation The operations were performed under combined endotracheal and intravenous anaesthesia in 6-month-old sheep (weighing 41.3 ± 1.2 kg). A left-sided thoracotomy in the fourth intercostal space was followed by establishing extracorporeal circulation through cannulation of the descending aorta and right atrial appendage. After cross-clamping of the aorta, Buckberg cardioplegic solution was applied using a combined cardioplegic cannula. Opening of the left atrium was followed by excision of the cusps and chordae of the native mitral valve. We used slightly oversized DMV (approximately 4–5 mm) for the implantation to compensate the suture line along the annulus. First, we estimated the level of implantation at the DMV papillary muscles. For this purpose, the annulus of the DMV was positioned onto the native annulus and papillary muscles were introduced into the left ventricle. Afterwards, U-stitches with polytetrafluoroethylene pledgets (3 × 6 mm) were initially placed on the appropriate position of each native annulus and then on the DMV papillary muscles. In addition, 2 guiding polypropylene 4-0 sutures were placed at the trigons of the aortomitral transition to prevent dislocation of the valve during the implantation. These sutures were then continued into the simple running suture along the annulus of the mitral valve (Figs 1 and 2). After the implantation, a water test was performed to check the valve competence. Figure 1: View largeDownload slide Schematic of the operation. The pledgeted U-sutures (left picture) and running polypropylene suture along the annulus (right picture) are shown. Ao: aorta ascendens. Figure 1: View largeDownload slide Schematic of the operation. The pledgeted U-sutures (left picture) and running polypropylene suture along the annulus (right picture) are shown. Ao: aorta ascendens. Figure 2: View largeDownload slide Surgical technique. (A) Sizing of the DMV: the oversized (+4–5 mm) grafts were implanted. (B) Running polypropylene 4-0 suture along the annulus. (C) End view of the implanted decellularized mitral valves. Figure 2: View largeDownload slide Surgical technique. (A) Sizing of the DMV: the oversized (+4–5 mm) grafts were implanted. (B) Running polypropylene 4-0 suture along the annulus. (C) End view of the implanted decellularized mitral valves. In the control group, commercially available stented porcine aortic prostheses were implanted in a traditional way. In all animals, prostheses with a diameter of 23 mm were implanted. Thereafter, the left atrium was closed using the double running polypropylene 5-0 suture and the aortic clamp was opened. Transoesophageal echocardiography (TOE) was performed to assess the performance of the grafts after weaning from the heart–lung machine (Philips CX50 ultrasound system with Philips X7–2 T probe, Koninklijke Philips N.V., Amsterdam, Netherlands), within the first 2 weeks postoperatively and shortly before the animals were sacrificed. Histological staining A half of the A2 zone of each explanted valve (or one of the leaflets of xenograft) was embedded in paraffin, sliced into 8-μm-thick sections and stained by haematoxylin–eosin, Movat’s pentachromic, van Gieson and von Kossa stains. Slides were analysed using a routine bright field microscopy (Olympus BX40, Olympus, Tokyo, Japan); with colour camera (AxioCam MRc, Carl Zeiss, Jena, Germany). Immunofluorescent staining The other half of the A2 zone of each explanted valve (and another leaflet of xenograft) was snap-frozen in liquid nitrogen, mounted in Tissue-Tek (Sakura Finetek Europe, AV Alphen an den Rijn, Netherlands) and cut into 7-μm-thick sections. All samples were stained against the following antigens: DNA [4,6-diamidino-2-phenylindoledihydrochloride (DAPI) stain, Invitrogen, Carlsbad, CA, USA] Collagen IV (clone CI22, DAKO, Hamburg, Germany) Collagen I C2345 clone COL-I (Sigma-Aldrich Chemie GmbH, Munich, Germany) CD31 [monoclonal mouse immunoglobulin G (IgG)2a, Serotec, Raleigh, NC, USA] von Willebrand factor (polyclonal rabbit IgG, DAKO) Endothelial nitric oxide synthase (monoclonal mouse IgG1, eNOS/NOS Type III, BD Transduction Laboratories, San Jose, CA, USA) CD45 [mouse monoclonal (clone OX-1) IgG, BioRad, Oxford, UK] Procollagen (monoclonal mouse IgG1, M-38-c, Developmental Studies Hybridoma Bank, Iowa City, IA, USA) The images were obtained using the microscope AxioObserver A1, with the camera AxioCamMRm (Carl Zeiss, Jena, Germany). RESULTS Surgical performance The mean bypass time was 108.5 ± 32.3 min, and the mean cross-clamp time was 64.5 ± 24.4 min. The pleural drainage was removed on the 2nd postoperative day under suction. During the follow-up, no sheep showed fever or any other abnormalities. There was no valve-related early postoperative mortality. Fourteen animals (Animal nos. 1–14) survived the first 2 weeks (Table 1). Table 1: Distribution of the animals Animal number  Type of graft  Survival (weeks)  Cause of death  LD EF (%)  LD MR  1  DMV  56  Sacrificed  55  II  2  DMV  52  Sacrificed  60  II  3  DMV  30  Sacrificed  30  II  4  DMV  64  Sacrificed  70  II–III  5  BMV  61  Sacrificed  60  I  6  BMV  61  Sacrificed  50  0  7  DMV  18  Heart failure due to MR  40  0–I  8  DMV  8  Rupture of the donor papillary muscle  55  0–I  9  DMV  18  Heart failure due to MR  40  0–I  10  BMV  14  Heart failure  n/d  0–I  11  DMV  11  Rupture of chordae tendineae  60  0–I  12  DMV  2  Rupture of papillary muscle (loose suture)  n/d  0–I  13  DMV  3  Partial rupture of the annulus  n/d  0–I  14  DMV  40  Heart failure due to MR  60  I  Animal number  Type of graft  Survival (weeks)  Cause of death  LD EF (%)  LD MR  1  DMV  56  Sacrificed  55  II  2  DMV  52  Sacrificed  60  II  3  DMV  30  Sacrificed  30  II  4  DMV  64  Sacrificed  70  II–III  5  BMV  61  Sacrificed  60  I  6  BMV  61  Sacrificed  50  0  7  DMV  18  Heart failure due to MR  40  0–I  8  DMV  8  Rupture of the donor papillary muscle  55  0–I  9  DMV  18  Heart failure due to MR  40  0–I  10  BMV  14  Heart failure  n/d  0–I  11  DMV  11  Rupture of chordae tendineae  60  0–I  12  DMV  2  Rupture of papillary muscle (loose suture)  n/d  0–I  13  DMV  3  Partial rupture of the annulus  n/d  0–I  14  DMV  40  Heart failure due to MR  60  I  BMV: biological mitral valve; DMV: decellularized mitral valve; LD: last documented; EF: ejection fraction; MR: mitral regurgitation. Table 1: Distribution of the animals Animal number  Type of graft  Survival (weeks)  Cause of death  LD EF (%)  LD MR  1  DMV  56  Sacrificed  55  II  2  DMV  52  Sacrificed  60  II  3  DMV  30  Sacrificed  30  II  4  DMV  64  Sacrificed  70  II–III  5  BMV  61  Sacrificed  60  I  6  BMV  61  Sacrificed  50  0  7  DMV  18  Heart failure due to MR  40  0–I  8  DMV  8  Rupture of the donor papillary muscle  55  0–I  9  DMV  18  Heart failure due to MR  40  0–I  10  BMV  14  Heart failure  n/d  0–I  11  DMV  11  Rupture of chordae tendineae  60  0–I  12  DMV  2  Rupture of papillary muscle (loose suture)  n/d  0–I  13  DMV  3  Partial rupture of the annulus  n/d  0–I  14  DMV  40  Heart failure due to MR  60  I  Animal number  Type of graft  Survival (weeks)  Cause of death  LD EF (%)  LD MR  1  DMV  56  Sacrificed  55  II  2  DMV  52  Sacrificed  60  II  3  DMV  30  Sacrificed  30  II  4  DMV  64  Sacrificed  70  II–III  5  BMV  61  Sacrificed  60  I  6  BMV  61  Sacrificed  50  0  7  DMV  18  Heart failure due to MR  40  0–I  8  DMV  8  Rupture of the donor papillary muscle  55  0–I  9  DMV  18  Heart failure due to MR  40  0–I  10  BMV  14  Heart failure  n/d  0–I  11  DMV  11  Rupture of chordae tendineae  60  0–I  12  DMV  2  Rupture of papillary muscle (loose suture)  n/d  0–I  13  DMV  3  Partial rupture of the annulus  n/d  0–I  14  DMV  40  Heart failure due to MR  60  I  BMV: biological mitral valve; DMV: decellularized mitral valve; LD: last documented; EF: ejection fraction; MR: mitral regurgitation. There were 4 surgery-related deaths: 1 (Animal no. 13, after 3 weeks) due to partial rupture of the annulus over 5–7 mm in the P1/P2 zone resulting in a significant paravalvular leak, 2 deaths (Animal no. 12 after 2 weeks and Animal no. 8 after 8 weeks) due to rupture of the decellularized papillary muscle (one because of loose suture) resulting in severe mitral insufficiency, 1 death (Animal no. 11 after 11 weeks) due to rupture of the chordae tendineae. Three animals with documented moderate transvalvular regurgitation died after 4 months (Animal nos. 7 and 9) and 7 months (Animal no. 14). One animal from the control group died after 4 months (Animal no. 10) having clinical signs of heart failure. Six animals (DMV, n = 4 and BMV, n = 2) survived for 12 months. Because of the normal growth, the weight of the animals increased from initial 41.3 ± 1.2 kg to 63 ± 1.1 kg. The final TOE of the DMV animals (Animal nos. 1–4) revealed low transvalvular gradients with trivial-to-moderate insufficiency due to central jet (Fig. 3). Sonographic signs of fibrosis were only seen along the mitral valve annulus. In contrast, in BMV animals (Animal nos. 5 and 6), TOE showed increased transvalvular gradients (max/mean, Animal no. 5: 19/14 mmHg and Animal no. 6: 17/11 mmHg) and moderate enlargement of the left atrium as a result of moderate-to-severe mitral stenosis. Figure 3: View largeDownload slide Peak pressure gradient on the mitral valve substitutes during the follow-up. The maximum gradients on the biological mitral valves increased significantly, while those on the DMVs were stable or decreasing. DMV: decellularized mitral valves. Figure 3: View largeDownload slide Peak pressure gradient on the mitral valve substitutes during the follow-up. The maximum gradients on the biological mitral valves increased significantly, while those on the DMVs were stable or decreasing. DMV: decellularized mitral valves. Macroscopic findings DMVs showed soft leaflets comparable with the native ones with some film-like thrombotic depositions typical for early deceased animals. The atrial and the ventricular surfaces along the suture line were completely covered with neointima starting from 8 weeks postoperatively (Animal no. 8), so that the suture material was completely invisible (Fig. 4C). The 4 DMVs explanted at 12 months revealed almost circular moderate hardening along the mitral valve annulus, without macroscopic signs of calcification. Animal no. 4 had a 3-mm paravalvular leak due to local disruption of the stitch at the posterior portion of the annulus; all other samples did not show any defects. The chordae tendineae were also not calcified. Samples, explanted at 12 months, showed almost completely organized donor papillary muscles that formed a solid scar together with the tips of the recipient’s papillary muscle (Fig. 4D). Figure 4: View largeDownload slide Decellularized mitral valve (DMV) and biological mitral valve. (A) DMV after the end of decellularization. (B) Biological mitral valve at 61 weeks postoperatively, arrow shows the calcified regions. (C) Atrial view of the DMV at 56 weeks postoperatively. (D) Ventricular view of the DMV at 56 weeks postoperatively, the soft and transparent leaflets and a firm scar on the papillary muscle are seen. Figure 4: View largeDownload slide Decellularized mitral valve (DMV) and biological mitral valve. (A) DMV after the end of decellularization. (B) Biological mitral valve at 61 weeks postoperatively, arrow shows the calcified regions. (C) Atrial view of the DMV at 56 weeks postoperatively. (D) Ventricular view of the DMV at 56 weeks postoperatively, the soft and transparent leaflets and a firm scar on the papillary muscle are seen. Xenografts showed typical degeneration and calcification at the commissures and sinuses with subsequent severe reduction of leaflet motion (Fig. 4B). Histological findings The routine light microscopy of DMVs revealed preserved structure of all mitral valve components. The 4-layered leaflets showed typical organization of the collagen and elastin fibres, and there was no difference in the extracellular matrix in comparison with the native ovine mitral valve: the leaflets had typical 4-layered structure and showed no matrix disturbance (haematoxylin-eosin, van Gieson and pentachromic staining) (Fig. 5A–C). The cell amount in the proximal parts of the leaflets was comparable with that of the native ones, descending to the free margin—distally only the endothelial layer and several cells in pars spongiosa could be seen. The repopulation of the leaflet differed from layer to layer—the pars ventricularis showed the best performance, followed by pars spongiosa and pars atrialis. Pars fibrosa at 12 months had only several cells in the proximal areas but almost none were seen further distally. However, in comparison with the animals which died during the follow-up, the cell count was constantly increasing according to the time in vivo. Figure 5: View largeDownload slide Histology of the decellularized mitral valve (DMV) and the biological mitral valve (BMV). (A–C)—DMV and (D–F) BMV. (A, D) Movat´s pentachrome; (B, E) van Gieson staining; (C, F) von Kossa staining. The structure of the BMV is completely disturbed due to calcification, while the DMV shows almost repopulated typical 4-layered leaflet with no signs of calcification, but preserved elastin fibres. Figure 5: View largeDownload slide Histology of the decellularized mitral valve (DMV) and the biological mitral valve (BMV). (A–C)—DMV and (D–F) BMV. (A, D) Movat´s pentachrome; (B, E) van Gieson staining; (C, F) von Kossa staining. The structure of the BMV is completely disturbed due to calcification, while the DMV shows almost repopulated typical 4-layered leaflet with no signs of calcification, but preserved elastin fibres. Von Kossa staining revealed small single spots of calcification, which were around 4–12 µm in diameter and tended to be localized in the proximal parts of leaflets (Fig. 5). No calcification was seen in the chordae tendineae. The immunofluorescent staining of Animal nos. 1–4 showed typical structure of collagen I, collagen IV covering the entire surface of all valve components as well as several interstitial cells positive for procollagen. At the same time, repopulation of the decellularized tissues of DMVs was clearly seen. Thus, at 12 months, the endothelial layer was completed along the entire surface of the leaflets and chordae with cobblestone-like cells that were shown to express CD31, von Willebrandt factor and eNOS (Fig. 6). The small number of cells, positive for CD45, was seen forming groups mostly in the proximal parts of the leaflets, along the suture and in scars on papillary muscles (Fig. 7). Figure 6: View largeDownload slide Immunostaining of the decellularized mitral valve (DMV) and the biological mitral valve (BMV). (A and B) DMV and (C and D) BMV. (A, C) von Willenbrandt factor and DNA; (B, D) collagen IV and DNA. DMV, in contrast to BMV, shows complete collagen IV layer with von Willenbrandt factor-positive cells on it. Figure 6: View largeDownload slide Immunostaining of the decellularized mitral valve (DMV) and the biological mitral valve (BMV). (A and B) DMV and (C and D) BMV. (A, C) von Willenbrandt factor and DNA; (B, D) collagen IV and DNA. DMV, in contrast to BMV, shows complete collagen IV layer with von Willenbrandt factor-positive cells on it. Figure 7: View largeDownload slide Immunostaining of the decellularized mitral valve (DMV) and the biological mitral valve (BMV). (A and B) DMV and (C and D) BMV. (A, C) CD45 and DNA; (B, D) procollagen I and DNA. DMV shows collagen-producing cells within the leaflets. BMV has almost no cells, positive for procollagen I. The CD45-positive cells in the BMV tend to form nodules (picture of the region, close to the annulus). BMV has only several CD45-positive cells on the surface. Ao: Aorta ascendens. Figure 7: View largeDownload slide Immunostaining of the decellularized mitral valve (DMV) and the biological mitral valve (BMV). (A and B) DMV and (C and D) BMV. (A, C) CD45 and DNA; (B, D) procollagen I and DNA. DMV shows collagen-producing cells within the leaflets. BMV has almost no cells, positive for procollagen I. The CD45-positive cells in the BMV tend to form nodules (picture of the region, close to the annulus). BMV has only several CD45-positive cells on the surface. Ao: Aorta ascendens. Xenografts showed lower amount of elastin fibres compared with DMVs but with the presence of most of the glycosaminoglycans. Collagen fibres and the whole 4-layered structure were preserved in regions not destructed by rough calcification spots over 1 mm in diameter (Fig. 5). However, in non-degenerated regions, von Kossa staining revealed multiple small calcified spots throughout the leaflet. Immunofluorescent staining revealed several regions covered with collagen IV with single CD31, von Willebrandt factor and eNOS-positive cells (Fig. 6), but most part of the surface was free from any cellular elements or had some cells negative for all tested antigens. Rare CD45-positive cells were also found on the surface and within the leaflets (Fig. 7). DISCUSSION This study demonstrates the feasibility of complete DMV implantation in an allogeneic ovine growing model and represents a first step towards long-term testing of DMV under physiological conditions. Our implantation procedure was based on previously described techniques of fresh and cryopreserved human mitral valves [12–14]. Still, we faced several significant problems in our animal model: rupture of the annulus, chordae and papillary muscle leading to mitral regurgitation in a relevant number of implantations within 12 months of follow-up. Previously performed biomechanical testing showed no significant difference in mechanical strength between native and DMV leaflets, but due to methodological limitations, only the A2 zone was investigated [10, 15, 16]. The complexity of the mitral valve structure led to several problems not only for the establishment of a decellularization protocol but also for the in vitro testing of the functionality of the isolated mitral valve components or the whole valvular graft. Variability of chordae structure and papillary muscles did not allow performing adequate mechanical test in vitro. Optimizing the implantation technique In vivo, we observed problems with the posterior part of the annulus, which is relatively weak, both in the native valves and in the DMVs. Here, we have seen rupture of the tissue resulting in paravalvular leak in 2 animals, in both animals the polypropylene suture was torn off at the recipient’s P1/P2 zone of the annulus and subsequently the suture thread remained on the DMV. One death occurred due to rupture of the chordae tendineae at the P2 zone causing severe regurgitation. Potentially, a concomitant implantation of an additional artificial chordae would be helpful to prevent flail of the posterior leaflet if a decellularized chordae ruptures. However, in contrast to our concerns at the beginning of the study, we observed chordae rupture in only 1 case. Another problem was identified at the level of the papillary muscles. Two animals died due to the rupture on this site after side-by-side implantation using pledged U-suture. We believe that the muscle fixation within a transected recipient’s papillary muscle could be an alternative and may facilitate the ingrowth of the decellularized papillary muscle head. However, care should be taken while using this technique to avoid shortening of the chordae tendineae. Growth aspects We did not use artificial rings for stabilization of the mitral valve annulus to estimate the potential of the DMVs in growing individuals. The advantages of allogeneic decellularized aortic and pulmonary valve grafts over the traditional valves in children have been already reported by several groups [4–7, 17]. Moreover, these grafts revealed in vivo remodelling and adaptive growth with preservation of excellent function during physiological development of the young patients. Our growing model showed that the DMV surface was almost completely reseeded with the host cells within 12 months. However, TOE at 12 months showed a slight increase in valve regurgitation, probably because of the increase in the size of the annulus. Mitral regurgitation observed in those animals is the result of heart growth on the one hand and the result of too slow repopulation and remodelling of the DMV matrix because of its high tissue volume on the other hand. It has been already shown that the mitral valve leaflet is at least 3 times thicker as one of the aortic valves, and the area of the leaflet is also significantly larger [10, 15]. This hypothesis is supported by the fact that the cell count seen in the explanted allografts was not comparable with the native mitral valve, especially in the pars fibrosa. Because of these findings, the use in children appears questionable, as DMV was not reseeded and, probably, could not be remodelled rapidly enough during somatic growth. However, it should be taken into account that the weight of the animals increased about 50% (>20 kg) within 12 months, which by far exceeds the normal human growth. DMV implantation appears much easier in adult patients because it allows to perform annulus stabilization with a ring, which may reduce the risk of valve insufficiency development in the long-term follow-up and minimize annulus-related complications. In addition to leaflet quality and annulus fixation, the configuration of papillary muscles and the number and the distribution of the chordae tendineae are important and need to be carefully considered. Modern imaging techniques such as high-resolution computed tomography or magnetic resonance imaging will further facilitate planning of concomitant procedures such as prophylactic neochordae implantation. Histological aspects The porcine valves had typical calcification profile—on the commissures and at the leaflets insertion sites, so that their mobility was significantly reduced resulting in high transvalvular gradients. At the same time, the BMVs showed the impossibility of their repopulation and only degenerative changes of the structure. At 12 months, DMV showed full coverage with the basal membrane positive for the collagen IV as well as the complete repopulation of the entire leaflet surface, chordae and papillary muscles with the endothelial cells. However, within the leaflet, the recellularization was not complete, which differs significantly from the aortic allografts [18]. The least reseeded parts were the chordae tendineae and pars fibrosa of the leaflet. On the other hand, the repopulation was slowly ongoing: the DMVs obtained from the dead animals showed almost no cells within the leaflet, but at 12 months up to 20% of pars fibrosa and at least two-third of the pars spongiosa were reseeded. A clear migration trend could be seen—the cell amount was maximal on and close to the surface, decreasing distally and towards the centre of the pars fibrosa. Probably, some additional methods to enhance the repopulation could be helpful to increase the reseeding speed and to stimulate the remodelling [19]. CONCLUSION We demonstrated the feasibility of orthotopic implantation of DMVs in an ovine growing model. The novel substitute was shown to be superior to a commercially available porcine xenograft regarding resistance to calcification and degeneration and demonstrated good, but not complete, recellularization. Further in vivo tests are pending to increase the reproducibility of DMV implantation and to evaluate techniques for enhancing faster recellularization with respect to valve growth and long-term durability. Funding This work was supported by CORTISS Hannover, Herz- und Gewebeforschungs GmbH. Conflict of interest: none declared. REFERENCES 1 Head SJ, Çelik M, Kappetein AP. Mechanical versus bioprosthetic aortic valve replacement. Eur Heart J  2017; 38: 2183– 91. Google Scholar CrossRef Search ADS PubMed  2 Nappi F. Cryopreserved mitral homograft valve: 19 years experience. JACC Cardiovasc Interv  2014; 7: S58. Google Scholar CrossRef Search ADS   3 Schoen FJ, Levy RJ. Calcification of tissue heart valve substitutes: progress toward understanding and prevention. Ann Thorac Surg  2005; 79: 1072– 80. Google Scholar CrossRef Search ADS PubMed  4 da Costa FDA, Dohmen PM, Vieira E, Lopes SV, Colatusso C, Pereira EWL et al.   Ross Operation with decellularized pulmonary allografts: medium-term results. Rev Bras Cir Cardiovasc  2007; 22: 454– 62. Google Scholar CrossRef Search ADS PubMed  5 Da Costa FD., Costa AC, Prestes R, Domanski AC, Balbi EM, Ferreira AD et al.   The early and midterm function of decellularized aortic valve allografts. Ann Thorac Surg  2010; 90: 1854– 60. Google Scholar CrossRef Search ADS PubMed  6 Cebotari S, Tudorache I, Ciubotaru A, Boethig D, Sarikouch S, Goerler A et al.   Use of fresh decellularized allografts for pulmonary valve replacement may reduce the reoperation rate in children and young adults: early report. Circulation  2011; 124: S115. Google Scholar CrossRef Search ADS PubMed  7 Tudorache I, Horke A, Cebotari S, Sarikouch S, Boethig D, Breymann T et al.   Decellularized aortic homografts for aortic valve and aorta ascendens replacement. Eur J Cardiothorac Surg  2016; 50: 89– 97. Google Scholar CrossRef Search ADS PubMed  8 Shi Y, Vesely I. Characterization of statically loaded tissue-engineered mitral valve chordae tendineae. J Biomed Mater Res A  2004; 69: 26– 39. Google Scholar CrossRef Search ADS PubMed  9 Gong W, Li S, Lei D, Huang P, Yuan Z, You Z et al.   Tissue-engineered mitral valve chordae tendineae: biomechanical and biological characterization of decellularized porcine chordae. J Mech Behav Biomed Mater  2016; 56: 205– 17. Google Scholar CrossRef Search ADS PubMed  10 Iablonskii P, Cebotari S, Tudorache I, Granados M, Morticelli L, Goecke T et al.   Tissue-engineered mitral valve: morphology and biomechanics. Interact CardioVasc Thorac Surg  2015; 20: 712– 19. Google Scholar CrossRef Search ADS PubMed  11 Granados M, Morticelli L, Andriopoulou S, Kalozoumis P, Pflaum M, Iablonskii P et al.   Development and characterization of a porcine mitral valve scaffold for tissue engineering. J Cardiovasc Transl Res  2017; 10: 1– 17. Google Scholar CrossRef Search ADS PubMed  12 Hubka M, Siska K, Brozman M, Holec V. Replacement of mitral and tricuspid valves by homograft. J Thorac Cardiovasc Surg  1966; 51: 195– 204. Google Scholar PubMed  13 Baird RJ, Williams WG, Spratt EH, Cohoon WJ. Experimental homograft replacement of the mitral valve. Can J Surg  1969; 12: 144– 54. Google Scholar PubMed  14 Plunkett MD, Schneider DJ, Shah JJ, Bash SE, Bond LM, Geiss DM. Homograft replacement of mitral valve in children. Ann Thorac Surg  1998; 66: 849– 52. Google Scholar CrossRef Search ADS PubMed  15 Grande-Allen KJ, Liao J. The heterogeneous biomechanics and mechanobiology of the mitral valve: implications for tissue engineering. Curr Cardiol Rep  2011; 13: 113– 20. Google Scholar CrossRef Search ADS PubMed  16 Roberts N, Morticelli L, Jin Z, Ingham E, Korossis S. Regional biomechanical and histological characterization of the mitral valve apparatus: implications for mitral repair strategies. J Biomech  2016; 49: 2491– 501. Google Scholar CrossRef Search ADS PubMed  17 Sarikouch S, Horke A, Tudorache I, Beerbaum P, Westhoff-Bleck M, Boethig D et al.   Decellularized fresh homografts for pulmonary valve replacement: a decade of clinical experience. Eur J Cardiothorac Surg  2016; 50: 281– 90. Google Scholar CrossRef Search ADS PubMed  18 Baraki H, Tudorache I, Braun M, Höffler K, Görler A, Lichtenberg A et al.   Orthotopic replacement of the aortic valve with decellularized allograft in a sheep model. Biomaterials  2009; 30: 6240– 6. Google Scholar CrossRef Search ADS PubMed  19 Assmann A, Delfs C, Munakata H, Schiffer F, Horstkötter K, Huynh K et al.   Acceleration of autologous invivo recellularization of decellularized aortic conduits by fibronectin surface coating. Biomaterials  2013; 34: 6015– 26. Google Scholar CrossRef Search ADS PubMed  © The Author(s) 2018. Published by Oxford University Press on behalf of the European Association for Cardio-Thoracic Surgery. All rights reserved. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices)

Journal

European Journal of Cardio-Thoracic SurgeryOxford University Press

Published: Jan 29, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off