Impact of T-cell-mediated immune response on xenogeneic heart valve transplantation: short-term success and mid-term failure

Abstract

OBJECTIVES

Allogeneic frozen cryopreserved heart valves (allografts or homografts) are commonly used in clinical practice. A major obstacle for their application is the limited availability in particular for paediatrics. Allogeneic large animal studies revealed that alternative ice-free cryopreservation (IFC) results in better matrix preservation and reduced immunogenicity. The objective of this study was to evaluate xenogeneic (porcine) compared with allogeneic (ovine) IFC heart valves in a large animal study.

METHODS

IFC xenografts and allografts were transplanted in 12 juvenile merino sheep for 1–12 weeks. Immunohistochemistry, ex vivo computed tomography scans and transforming growth factor-β release profiles were analysed to evaluate postimplantation immunopathology. In addition, near-infrared multiphoton imaging and Raman spectroscopy were employed to evaluate matrix integrity of the leaflets.

RESULTS

Acellular leaflets were observed in both groups 1 week after implantation. Allogeneic leaflets remained acellular throughout the entire study. In contrast, xenogeneic valves were infiltrated with abundant T-cells and severely thickened over time. No collagen or elastin changes could be detected in either group using multiphoton imaging. Raman spectroscopy with principal component analysis focusing on matrix-specific peaks confirmed no significant differences for explanted allografts. However, xenografts demonstrated clear matrix changes, enabling detection of distinct inflammatory-driven changes but without variations in the level of transforming growth factor-β.

CONCLUSIONS

Despite short-term success, mid-term failure of xenogeneic IFC grafts due to a T-cell-mediated extracellular matrix-triggered immune response was shown.

INTRODUCTION

Human allogeneic heart valves are almost perfect heart valve replacements, offering optimal haemodynamic properties, outstanding resistance to infections without anticoagulation [1]. In the early 1960s, allografts were directly transplanted, as the so-called homovitals [2]. Since 1968, valves were cryopreserved with controlled-rate freezing and stored in vapour-phase liquid nitrogen due to logistic issues with fresh tissues [3], enabling long-term storage times up to 5 years. Frozen cryopreservation (FC) remains the preservation method of choice for human heart valves worldwide.

Shortcomings of FC valves include intensive and logistically unfavourable storage in the vapour phase of liquid nitrogen, as well as their limited long-term function due to inflammation, immune response, and subsequent structural deterioration [4]. In particular, early reinterventions in paediatrics occur due to allograft failure within the first 5 years [5], demonstrating the need for new optimized preservation methods. To overcome these limitations, an alternative ice-free cryopreservation (IFC) method was developed. This method prevents interstitial ice formation [6] and enables optimal preservation of the extracellular matrix (ECM) structures. A mid-term large animal study [7] and an unpublished long-term study confirmed preserved ECM integrity and excellent haemodynamics compared with standard cryopreserved valves.

Despite these promising results, the worldwide organ shortage remains unchanged. Therefore, the objective of this study was to explore the potential of IFC for xenogeneic pulmonary valves, by comparing xenogeneic (porcine) with allogeneic (ovine) valves in a sheep model.

METHODS

Tissue preparation

Seven porcine (domestic pig, ∼3-month-old, Laboratory animal facility in Frankfurt, Germany) and 7 control ovine hearts (Dorset Cross, ∼1-year-old, slaughterhouse in Minnesota, USA) were obtained using aseptic conditions, rinsed with ice-cold lactated Ringer’s solution. Pulmonary heart valves were excised aseptically and placed individually in 100 ml volumes of Dulbecco’s modified Eagle’s medium (Invitrogen, Carlsbad, CA, USA) containing 4.5 g/l glucose with 126 mg/l lincomycin, 52 mg/l vancomycin, 157 mg/l cefoxitin and 117 mg/l polymixin for overnight storage at 4 °C.

Ice-free cryopreservation

All valves were preserved using IFC. They were individually placed and cooled in sterile polyester bags containing 100 ml VS83 solution (4.65 mol/l dimethyl sulfoxide, 4.65 mol/l formamide and 3.31 mol/l propylene glycol in 1× EuroCollins solution) as previously described [8]. Valves were stored at −80 °C. After a total storage time of 10.6 ± 0.8 months (ovine) and 12 ± 1.8 months (porcine) at −80 °C, valves were implanted.

Warming procedure

Individual ice-free cryopreserved valves were directly warmed in a 37 °C water bath without external manipulation until the solution moved freely. Bags were cleansed externally with 70% ethanol, and cryoprotectant solution was removed in 5 steps by placing the valve for 5 min each in ice-cold EuroCollins solution. Subsequently, valves were stored in ice-cold lactated Ringer’s solution (Ringerlösung, B.Braun, Melsungen, Germany) until implantation.

Implantation

Twelve juvenile merino sheep (average age 16 ± 2 weeks; weight 35 ± 5 kg) underwent pulmonary heart valve replacement as previously described [9]. In brief, the right ventricular outflow tract was exposed by a left anterolateral thoracotomy through the fourth intercostal space. Normothermic cardiopulmonary bypass was established using femoral arterial and right atrial venous cannulation. With the heart beating, the pulmonary artery was transected and a segment of the artery and all 3 leaflets were removed. The trimmed heart valves were subsequently implanted using running 5-0 monofilament sutures (Fig. 1E). The cardiovascular sheep model was applied due to the similar heart size, frequency, stoke volume and pressure ratios as humans. Six xenogeneic and 6 allogeneic IFC valves were implanted. All animals were kept in an indoor housing facility and received care in compliance with the ‘Guide for the Care and Use of Laboratory Animals’ published by the National Institute of Health (NIH publication no. 85-23 revised 2011).

Figure 1:

Gross morphology of allogeneic and xenogeneic pulmonary heart valves. (A) Allogeneic leaflets were thin and translucent as the representative explant after 8 weeks displays. (B) Early xenogeneic explants, as well as the representative explant after 8 weeks, had thin and translucent leaflets. Strong inflammation was detected for (C) xenogeneic explant at 4 weeks (D) and at 12 weeks. The inflamed inside of leaflets as well as thickened leaflets vegetation are indicated with black arrows. (E) Lateral thoracotomy with implanted xenogeneic porcine heart valve with running 5-0 monofilament sutures before the excision of the valve at 2 weeks after implantation.

Explantation

After 1, 2, 3, 4, 8 and 12 weeks, 2 animals, 1 from each group, were euthanized. Valves were explanted and the following gross inspection further processed.

Ex vivo computed tomography

Ex vivo computed tomography (CT) phantom scans were performed for paraffin-embedded heart valve blocks before tissue sectioning. Examinations were performed with a third-generation dual-source CT system (SOMATOM Force, Siemens Healthcare, Forchheim, Germany), equipped with a fully integrated circuit detector system (Stellar Infinity; Siemens) and 2 X-ray tubes (Vectron; Siemens). All studies were reconstructed with advanced modelled iterative reconstruction (ADMIRE; Siemens) at a standard strength level of 3 using a medium-sharp convolution kernel (Bv36), 0.6 mm section thickness and 0.4 mm increment.

Histology

One leaflet of each heart valve was formalin fixed for 24 h and paraffin embedded (Merck, Darmstadt, Germany). For general morphology, representative 3-μm-thick sections were stained with standard Hematoxylin-Eosin, Elastica van Gieson and Movat Pentachrome nach Verhoeff stains. Calcium deposits were detected via von Kossa stain. CD3 (DCS, Hamburg, Germany) stain was performed to identify T-cell infiltration.

Transforming growth factor-β1 cytokine release

Tissue punches of the second leaflet with 5 mm thickness of each explanted valve were obtained for the cytokine release assessment (n = 2–3). Excised native leaflets and preserved but not implanted leaflets served as additional controls. Punches were washed twice in ice-cold phosphate-buffered saline and incubated individually in 24 well plates with 500 μl Dulbecco’s modified Eagle’s medium (Biochrom, Berlin, Germany) for 4 days at 37 °C and 5% carbon dioxide. Coculture supernatants were centrifuged (1 min, 1000g) and stored at −80 °C until further analysis. Active and inactive transforming growth factor (TGF)-β1 in supernatants were analysed using LEGEND MAX™ Total TGF-β1 ELISA Kit (Biolegend, Fell, Germany) according to the manufacturer’s protocol. Mean ± standard deviation was determined and a non-parametric 1-way analysis of variance was conducted with Prism 6 (GraphPad, La Jolla, CA, USA). Differences with P-value <0.05 were considered statistically significant.

Near-infrared multiphoton imaging

The third leaflet was placed individually in 100 ml ice-cold Dulbecco’s modified Eagle’s medium with 4.5 g/l glucose and antibiotics (1.2 g/l amikacin, 3 g/l flucytosin, 1.2 g/l vancomycin, 0.3 g/l ciprofloxacin and 1.2 g/l metronidazol). Autofluorescence and second harmonic generation imaging were performed within 72 h. Leaflets were placed in a glass bottom dish (ibidi GmbH, Martinsried, Germany) and assessed on the microscope stage (40× oil immersion objective; N.A. 1.3; Carl Zeiss, Jena, Germany) of a multiphoton laser system (JenLab GmbH, Jena, Germany), as previously described [10]. Samples were excited with a laser power of 20 mW and a wavelength of 760 nm. Detector settings were 1000 V contrast and 52.7% brightness.

Raman spectroscopy

Explanted native ovine, porcine pulmonary and non-transplanted (NT) IFC leaflets were analysed within 48 h, as described previously [11]. Samples were measured with a custom-built confocal Raman microspectroscope [12]. In each tissue, 30 randomly selected spots, focusing on collagen structures, were measured through a 60× water immersion objective (N.A. 1.2; Olympus, Japan). Spectra were acquired at 85 mW laser power during 100 s exposures. Data sets were pretreated by baseline correction, background subtraction and pursued by the spectroscopy software OPUS 4.2 (Bruker Optik GmbH, Ettlingen, Germany). Multivariate data analysis software, The Unscrambler^® (Camo, Oslo, Norway), was used for principal component analysis, including spectral differences and peak shifts. Mean ± standard deviation score values were determined and an analysis of variance with a Tukey’s test was performed (n = 30) with OriginPro 9.1G software (OriginLab, Northampton, MA, USA). Differences with P-values <0.05 were considered statistically significant.

RESULTS

All animals survived heart valve replacement. One allogeneic heart valve required explantation due to an acute severe infection (pneumonia) at 18 days instead of 4 weeks.

Gross morphology

Gross morphology revealed thin and translucent leaflets for all allogeneic valves (Fig. 1A). Xenogeneic valves generally had thicker leaflets (Fig. 1B). After 4 weeks, the arterial side of xenogeneic leaflet became severely inflamed (Fig. 1C), and at 12 weeks, additional severe vegetations on the ventricularis (Fig. 1D) were observed.

Ex vivo computed tomography

Ex vivo CT scans of formalin-fixed and paraffin-embedded explanted valves indicated increasing calcium deposition restricted to the conduit wall for both groups starting at 3 weeks after implantation. Allogeneic tissue calcifications were limited to the hinge region near the former annulus muscle and the suture lines. In contrast, the entire conduit wall in xenogeneic tissue was severely calcified after 12 weeks (Fig. 2B).

Figure 2:

Calcium depositions in explanted allogeneic and xenogeneic heart valves. Calcium depositions were demonstrated with (A) van Kossa stain for 1 week implanted allograft and xenograft in the former annulus muscle and (B) computed tomography phantom scans of formalin-fixed and paraffin-embedded heart valves. Bright white areas indicate calcium depositions in the tissue.

Histology

No thickening of allogeneic leaflets was observed. In contrast, severe leaflet thickening of all xenogeneic leaflets was observed by 4 weeks of implantation (Fig. 3). Hinge regions including the annulus and lumen of allogeneic valves were almost completely acellular after 1 week and remained virtually cell free throughout the 12-week study (Fig. 4V, VI, IX, X). Sporadic CD3-positive cells were observed at 12 weeks in the hinge region (Fig. 4XI). A few T-cells could be seen after 1 week near the annulus (Fig. 4VIII). Small numbers of CD3-positive T-cells surrounded but never infiltrated the former muscular annulus (Fig. 4VIII, XII). Increased cellularity, with abundant CD3-positive cells, was detected at the suture lines. In addition, changes in the proteoglycan and collagen make-up of ECM were detected over the 12-week period. Increasing proteoglycan deposition and decreasing collagen content in the valves annulus were detected over time using Movat Pentachrome after Verhoeff staining (Fig. 4II, VI, X). Leaflet tissue appeared slightly more condensed, particularly elastin fibres in the lamina ventricularis (data not shown).

Figure 3:

Leaflet diameters of allogeneic and xenogeneic heart valves. The mean ± standard deviation (n = 5) of (A) leaflet tip and (B) belly diameters of transplanted ice-free cryopreserved allogeneic and xenogeneic pulmonic heart valves are displayed. Native leaflet tip and belly diameter (open circle) is shown as a control. ALLO: allograft; XENO: xenograft.

Figure 4:

Histology of representative explanted allogeneic ice-free cryopreservation pulmonic valves and native control sections. Acellular leaflets were observed using Haematoxylin-Eosin stain (I, V, IX), Movat Pentachrome nach Verhoeff stain (II, VI, X) showed extracellular matrix changes, detected in the valves annulus. Minimal CD3-positive T-cell infiltration over time of the hinge region (III, VII, XI) and around the former annulus muscle (IV, VIII, XII) was found. Scale bars equal 100 μm.

Acellular leaflets and hinge regions were also observed after 1 week for xenogeneic valves (Fig. 5AII, V and BI). In contrast to allogeneic tissues, subsequently, severe immune response characterized by cell infiltration and leaflet thickening was detected. After 4, 8 and 12 weeks, 33, 50 and 66% of the leaflets were infiltrated with abundant CD3-positive T-cells (Fig. 5AI) initiating from the hinge region. Explants at 4 and 12 weeks had the most thickened and inflamed leaflets, consistent with gross morphology (Fig. 5AV, IV). Importantly, the donor leaflet tissue was not generally infiltrated with cells. Neotissue formation of 2 thick cell-rich layers with abundant CD3-positive T-cells on arterial and ventricularis sides were detected (Fig. 5AIV and BIV). The lamina ventricularis with elastin fibres was overgrown and localized inside acellular leaflet tissue (Fig. 5AVIII).

Figure 5:

Histology of representative explanted xenogeneic ice-free cryopreservation pulmonic valves and native control sections. (A) Acellular leaflets were rapidly and severely infiltrated and overgrown with neotissue as indicated with the Haematoxylin-Eosin stain (I–IV), Elastica van Gieson stain (V–VIII). Black arrow indicates elastin fibres in the overgrown former lamina ventricularis. Extracellular matrix changes were detected in the annulus (IX–XII) using a Movat Pentachrome nach Verhoeff stain. (B) Acute and increasing T-cell infiltrations and inflammation are indicated in the leaflet (I, IV), conduit (II, V) and the former annulus muscle (III, VI) tissue especially around the borders of neotissue formation and former transplanted porcine tissue after 1 and 12 weeks.

In general, multiple foci of inflammatory infiltration with subsequent structural changes were found to a much greater extent compared with allogeneic tissue. The former annulus muscle was surrounded rapidly by abundant CD3-positive cells (Fig. 5BIII, VI). CD3-positive cells also infiltrated the cell-rich tunica externa of the conduit starting 1 week after implantation (Fig. 5BII, V). New microvessel formation was detected in the cell-rich tunica externa and slow degradation of the transplanted acellular conduit tissue was observed at 12 weeks originating from the tunica externa. Strong CD3-positive T-cell infiltration was observed at these degradation sites. Additionally, proteoglycan-rich neointima formation on the luminal surface of the allograft wall was found after 4 weeks of implantation. As observed in allogeneic tissue, similar structural changes of increasing proteoglycan depositions and decreasing collagen amounts were detected in the annulus over time. The xenogeneic annulus lost its original appearance and shrank dramatically (Fig. 5AIX–XII).

Calcifications

Von Kossa stains revealed no calcification in the leaflets of either groups. However, microcalcifications in the former muscular annulus of both groups were more abundant in xenogeneic valves (Fig. 2A). Stronger calcifications throughout the entire xenogeneic conduit wall were consistent with CT scans (Fig. 2B).

Transforming growth factor-β1 cytokine release

TGF-β1 is highly conserved across mammalian species and exists in a latent and active form. This cytokine is an important mediator and an indicator of tissue remodelling, inflammation and fibrosis. For native (ovine) leaflet tissue, a total (latent and active) TGF-β1 release of around 2500 pg/ml with 1% active TGF-β was detected (Fig. 6). The xenogeneic leaflets had slightly elevated total TGF-β levels compared with the allogeneic release profiles. No significant differences in TGF-β release profiles were observed for at different explant times.

Figure 6:

TGF-β1 release profiles of leaflet tissue punches. Total latent and active TGF-β1 (A, C) and the active TGF-β1 (B, C) release profiles for ovine allogeneic (A, B) and porcine xenogeneic (C, D) leaflets are presented. The mean ± standard deviation is displayed (n = 3). TGF-β: transforming growth factor beta.

Near-infrared multiphoton imaging

Autofluorescence and second harmonic generation images revealed no elastin or collagen changes in either allogeneic or xenogeneic tissues compared with the native controls (Fig. 7).

Figure 7:

Multiphoton and second harmonic generation images of fibrosa and ventricularis of control and explanted heart valve leaflets. Collagen and elastic fibres of the fibrosa show parallel alignment. Scale bars represent 60 µm.

Raman spectroscopy

Raman spectroscopic analyses and multivariate data analysis of excised native, NT and explanted ice-free ovine and porcine leaflets analysis (Fig. 8A) revealed a clustering between all 4 groups. However, native and NT data of ovine tissues clustered more closely than porcine tissues, where a clear separation between native and NT tissues was shown.

Figure 8:

Multivariate data analysis. (A) Principal component analysis scores plot of native and NT porcine and ovine data sets: ovine native and NT data clusters overlap, while porcine native and NT differ more. (B) Significant collagen peaks of the loadings plots and their correlating molecular groups. (C) Score values of PC3 for allogeneic and (D) of PC1 for xenogeneic explants. *P-value ≤ 0.001 (n = 30). Loading plot (E) of PC3 of allogeneic explants and (F) of PC1 of xenogeneic explants. The mean ± standard deviation is displayed. NT: non-transplanted.

Principal component analysis of allogeneic leaflets (Fig. 8C) demonstrated significant differences between mean score values of PC3 compared with native tissue for the first 3 weeks. The 8- and 12-week implants were not statistically different. Corresponding loadings plot for allogeneic valves showed 5 typical influencing peaks on the spectral differences in the fingerprint region all correlated with certain molecular bonds in the collagen molecule (Fig. 8E) [13, 14]. In contrast, more pronounced statistically significant differences between the mean score valves of PC1 of xenogeneic and native tissues were detected, particularly at 12 weeks (Fig. 8D). Although the characteristic collagen peaks were still found in the loading plot, a clear shift of the peak intensities was observed, as indicated in Fig. 8F.

DISCUSSION

The objective of this study was to evaluate the potential of xenogeneic ice-free cryopreserved heart valves in comparison with ice-free allogeneic valves with natural unaltered ECM structures to overcome organ scarcity. Our results demonstrated an almost immediate and complete loss of cellular elements in the 1st and 2nd week after implantation. Acellular allogeneic heart valves revealed no immune rejection leaflet architecture and ECM preservation compared with native tissues. The loss of cellular components in IFC grafts was previously described after 7 months in vivo [7, 8]. The IFC procedure uses highly concentrated chemicals resulting in grafts without cell viability [8, 15, 16]. Based on the fast disappearance of cells, we hypothesized a passive washout of cell remnants. Sheer stress due to blood flow, as well as constant stretching and contraction of tissue, may enable washout of cellular debris. However, the acute cell removal process requires further in-depth investigation to evaluate possible early macrophage involvement within the 1st week. In contrast to allogeneic valves, an early and elevated T-cell infiltration was detected after 1 week in xenogeneic valves turning into a T-cell-mediated immune response with subsequent structural degeneration, which was observed to start after 4 weeks.

Compared with xenogeneic implanted valves, allogeneic valves show less adverse immune responses, consistent with previous mid-term results [7]. We postulate that the immediate cell loss and the lack of triggering soluble mediators prevented infiltration and activation of T-cells. In particular, cell-mediated T-cell activation by endothelial or dendritic cells [17] is unlikely in acellular grafts. The minimal T-cell infiltration restricted to the conduit wall is most likely caused by remaining cell remnants. Possibly, the additional washing processes prior to implantation using pulsatile bioreactors [18] might enable complete elimination of cell remnants and avoid subsequent T-cell infiltration.

As opposed to allografts, we observed an active rejection of xenogeneic valves. Multiple foci of inflammatory infiltrations, predominantly CD3-positive T lymphocytes, as well as pannus formation on the luminal side, indicate an ongoing rejection as previously described by others [19]. The inflammation that occurred was predominantly extravascular similar to glutaraldehyde-fixed bioprosthetic heart valves [20]. The immediate acute severe T-cell infiltration of acellular xenogeneic tissue confirms recent observations that the xenogeneic matrix itself is the culprit for the immunogenicity [19, 21]. Because of the fact that T-cell infiltration has been linked to ECM destruction, especially collagen fibres [22], this could partly explain the fast structural deterioration of T-cell-infiltrated xenogeneic tissues already after 4 weeks. Even though T-cell infiltration has always been seen as a major contributing factor for heart valve failure [21, 23], further in-depth studies into other immune cells of the innate immunity, such as granulocytes and macrophages, are warranted.

Despite clear differences of T-cell infiltration for allografts and xenografts, no influenced immunological response profiles for the different explanation time points in neither of the groups were observed. Release of TGF-β1, an important mediator and indicator of tissue remodelling, as well as inflammation and fibrosis modulation, was analysed. Despite a high variability, similar active and inactive (latent) TGF-β1 levels were seen. A constant TGF-β release from transplanted and acellular leaflets indicate a prolonged leakage of matrix-bound TGF-β. Slightly increased absolute values for xenogeneic tissue might be related to an increased leaflet thickness and infiltrating TGF-β1 producing T-cells or macrophages. Other TGF-β isoforms, and especially their relative ratios, influence tissue regeneration [24]. Although TGF-β3 shares a similar structure to TGF-β1, it shows different effects on immune responses [25]. Furthermore, TGF-β3 is anti-fibrotic and might be involved in fibrosis modulation [26]. Whether other isoforms of TGF-β are also detectable and involved in immune responses to implanted IFC heart valves have to be elucidated in the future.

Additional in-depth structural analysis was performed because T-cell infiltration is known to be associated with ECM destruction, especially of collagen [22]. Near-infrared multiphoton imaging did not detect structural changes of collagen or elastin, unlike Raman spectroscopy analysis. Principal component analyses of the Raman showed native-like results for allograft. Although xenogeneic score values for longer implanted valves were more different compared with native fresh control tissues. Clear collagen changes, possibly caused by leucocytes, were found for xenogeneic tissue, particularly for the more severely infiltrated and inflamed leaflets after 4 and 12 weeks. A clear shift in the loading plot for the xenogeneic tissue away from typical collagen-associated loading plots seen in allogeneic tissues also indicated collagen changes. A limitation of our study is the lack of haemodynamic data, which could have underlined possible valvular dysfunction due to the structural ECM alterations.

Surprisingly, early microcalcifications were detected in both groups starting from 1 week after implantation in the muscular annulus. No calcifications in the conduit wall were found in mid-term or long-term implanted IFC valves [7]. Microcalcifications are formed at sites of cell death. These cytoskeletal remnants allow nodule formation and calcium deposition [27]. The phospholipids-rich membranous debris is linked to nucleate apatite [20, 28] and calcium phosphate crystal deposition has been identified as the source of calcification [28]. They are found especially close to smooth muscle cells, as these cells commonly form matrix vesicles to attract calcium, leading to microcalcifications [27]. Destruction of cardiomyocytes in the annulus may have caused an acute release of these plasma membrane-derived matrix vesicles causing microcalcifications [27–29]. Possible elevated calcium levels, which induce mineralization of cardiomyocytes, might accelerate rapid calcification [28]. Inflammatory reactions, connected with pathogenesis of acquired calcifications in valves [27, 30] as well as α-Gal antibodies interacting with the Gal epitope (galactose-α(1, 3)-galactose antigen) accelerate calcification in heart valves [22] and therefore favour a progression of the calcification processes, particularly in xenogeneic tissues. The potential elimination of α-Gal antibodies using knock-out models are the current focus of the ongoing xenogeneic IFC heart valve studies.

CONCLUSION

In summary, our data implicate that xenogeneic ECM in wild-type porcine tissue elicits an immune response rendering ice-free cryopreserved xenogeneic valves unsuitable for heart valve replacement. Allograft transplantation still remains the current preferable choice of replacement until new alternatives such as porcine α-Gal knock-out, human CD46 transgenic porcine xenograft valves or tissue-engineered valves can overcome these limitations and organ scarcity in the future. In contrast, IFC allografts perform well in vivo in sheep and are ready for evaluation in human patients.

Funding

This work was supported by the German Research Foundation [Sto 359/10-1 to U.S., SE 657/9-1 to M.S., SCHE 701/10-1 to K.S.-L. and INST 2388/30-1 to K.S.-L.] and the Ministry of Baden-Wuerttemberg for Sciences, Research and Arts (33-729.55-3/214 and SI-BW 01222-91). KGMB was supported by the National Institute of Allergy and Infectious Disease, National Institutes of Health grant R43 AI114486.

Conflict of interest: Julian L. Wichmann received speaker fees from GE Healthcare and Siemens Healthcare. Kelvin G.M. Brockbank is the owner and employee of Tissue Testing Technologies LLC.