In 1975, O’Brien et al. (25) successfully performed the first transplantation of cryopreserved human heart valves. Research on cryopreserved homografts has progressed internationally, with an emphasis on prolonging the durability of cryopreserved allogeneic heart valves. Cryopreserved valves demonstrate a 93% rate of freedom from reoperation over 10 years and 81% over 15 years, demonstrating marked advancements in long-term performance (22,34). Bioprosthetic valves, made from pig or cow heart tissue, have undergone continual refinement. However, compared with cryopreserved valves, bioprosthetics exhibit a faster rate of structural valvular degeneration (SVD), exhibiting roughly 31% freedom from reoperation at 15 years and 91% after 10 years (22,34). This trend is particularly pronounced in younger patients with active immune functions, where cryopreserved valves show 92% freedom from surgery after 5 years, compared with 53% for bioprosthetics, highlighting the advantages of homografts (4).

Generally, homografts effectively prevent thromboembolic events, eliminating the need for long-term anticoagulant use (9,35), and are more resistant to infection, making them suitable for treating infectious endocarditis. However, no significant differences were observed in the reoperation frequency or endocarditis incidence compared with that of homografts, leading to a recommendation for prosthetic valve use in cases of supply issues (15).

Bioprosthetic valves, once covered with autogenous endothelial cells, typically cease to cause coagulation abnormalities in humans, thereby reducing the need for long-term anticoagulation therapy (8). However, in dogs, the response to anticoagulants can differ significantly from that in humans, and the risk of thrombotic complications remains (5). While bioprosthetic valves are generally preferred over mechanical valves in dogs due to the reduced need for anticoagulation, the possibility of thrombosis still exists, particularly because the coagulation management in dogs may be less predictable (35). Additionally, complications such as severe bleeding and thrombosis related to anticoagulant therapy, as well as inflammatory conditions like pannus, can result in poor outcomes in canine valve replacement (2,5). Consequently, the choice of replacement in canine cardiac surgery is highly constrained, with suture repairs being preferred over prosthetics for valve insufficiency, offering improved durability and better long-term results.

Transplantation of fresh goat or porcine heterograft valves in dogs was associated with poor long-term outcomes (12,27). Thrombus formation around the valve was the primary cause of mortality in porcine valve transplantation, whereas necrosis and phanocytosis were the primary causes of valve degeneration in goats. In contrast, canine homograft valve transplantation demonstrated good results for up to 1 year, maintaining valve function (1,28). Clinical studies on cryopreserved canine homograft valve transplantation have not been conducted yet. This underscores the need for research on cryopreservation for canine valve transplantation and ultimately supports the establishment of a tissue bank.

Therefore, this study aimed to establish appropriate protocols for procurement and cryopreservation that are crucial for extending the lifespan of cryopreserved homografted heart valves.

Heart extirpation

This study was approved by the Institutional Animal Care and Use Committee of the National University of Chungnam (approval no. 202309A-CNU-154). In this study, six mature (>15 months old) healthy female dogs were weighed. A comprehensive physical examination, blood tests, and chest radiography were performed for all the dogs to confirm the absence of pre-existing diseases. The results of the physical examination and blood tests confirmed the suitability of the donors, and these parameters were included in the study analysis.

The initial step in cardiothoracic surgery involves precisely incising along the sternum during general anesthesia administration. This midline incision on the pericardium offers clear visibility of crucial anatomical structures, including the ascending aorta, aortic arch, and initial 2-3 cm of the arch vessels. Key anatomical structures were meticulously dissected and separated to ensure precision during the procedure. Subsequently, the cranial and caudal vena cava were gently transected. The heart was then turned outward, the pulmonary veins were divided, and the right and left pulmonary arteries were transected to ensure a comprehensive dissection. The final step in gaining access to the chest cavity involves cutting the aorta distally to the greatest extent possible. Following extirpation, the heart was thoroughly rinsed with cold saline (4°C) to remove residual blood. After harvesting, the appearance of the body from which the tissue was collected was restored.

Valvular tissue harvest

Dissection procedures were conducted under sterile conditions in a laminar flow hood or operating room to ensure optimal sterility and precision. The heart was first carefully extracted from the transport solution, followed by its placement in a sterile basin containing 1 L of cold 0.9% normal saline solution, maintained at a constant temperature of 4°C.

The dissection was performed in a series of carefully executed steps. Initially, the aorta was delicately separated from the pulmonary artery. Subsequently, anterior entry into the right ventricular cavity enabled circumferential dissection, facilitating the separation of the pulmonary valve and artery from the right ventricle. Circumferential division of the right ventricular outflow tract and interconnected tissues ensured that a precise distance of 1 cm from the pulmonary valve was maintained. Following the removal of the homograft block from the donor heart, any surplus tissue, including muscle and adipose tissue, was removed thoroughly. After these critical steps, the pulmonary homografts were labeled following precise sizing procedures, including the total length and diameter of the annulus (Fig. 1).

Figure 1. Harvested heart valve. (A) Extirpated heart. Pericardium not removed. (B) Separated pulmonary valve. Approximately 2 cm of myocardium was included, and the excision was performed to encompass from the main pulmonary artery to the branching points of the left and right pulmonary arteries. (C, D) Diameter of the annulus measured: approximately 1 cm.

Procurement and packaging

Tissue-handling personnel followed meticulous procedures to ensure the safety and sterility of the collected tissue samples. Using sterile swabs, they carefully sampled different donor tissue surfaces, collecting five swabs on each surface. The swabs were then placed in transport media and sent for microbial testing to ensure that the samples were free of any potential contaminants. Subsequently, an antibiotic solution was prepared by adding specific antibiotics, such as cefotaxime (500 mg), lincomycin (1 million units), polymyxin B (1 million units), vancomycin (500 mg), and amphotericin B (50 mg), to a 1 L solution of RPMI 1640 solution. Each homograft was immersed in this antibiotic solution at 4°C for 24 h, allowing for a thorough antibiotic treatment. Subsequently, the homografts were removed from the solution, ensuring sterility. They were then rinsed for 12 min in fresh RPMI 1640 medium. These homografts were preserved in a cryoprotectant medium comprising RPMI 1640 medium, 10% fetal bovine serum, and 10% dimethyl sulfoxide (DMSO). Microbial culture tests were performed at each stage to confirm the absence of contaminants. To prepare the homografts for cryopreservation, 50 mL of the cryoprotectant medium was placed in a double plastic bag and all air was diligently removed to prevent bag rupture during thawing (32).

Cryopreservation

The samples were aseptically stored in a double bag and transferred to the freezing chamber of the programmable freezer. Cooling was conducted according to a preprogrammed protocol (Fig. 2A). Proper configuration of the cooling program is essential to prevent the release of latent heat. Once the tissue reached a freezing point of −70°C, it was transferred to the vapor phase of liquid nitrogen (<−135°C) for storage (Fig. 2B, C).

Figure 2. (A) The freezing rate graph over time and the cooling curve graph of the sample. The temperature of the actual sample (blue line) decreases in response to the dropping temperature of chamber (red line) according to the programmed temperature settings (green line). (B, C) The graft after freezing.

Thawing

The thawing protocol was a fast rewarming process consisting of rapid warming by transferring the bag from the nitrogen gas phase to a 42°C water bath to prevent ice recrystallization. This rewarming process had to be performed within 15 min to minimize the cytotoxic effect of DMSO (18,35). The cryoprotectant liquid was removed gradually in four 3-min steps via immersion in tapering concentrations of DMSO (10%, 5%, 0%, and 0% in RPMI 1640 medium with 10% fetal calf serum) at 4°C. Thereafter, the vascular homograft was stored in pre-cooled custodiol solution until implantation (32).

Assessment of tissue structural preservation

Tissue samples were collected immediately after antibiotic treatment from each harvested heart valve leaflet. A portion of the graft was sampled for subsequent comprehensive histological examination for control groups. The remaining tissues were then cryopreserved for 6 months and thawed for histological analysis. Consistent processing and sectioning techniques were employed for all the specimens. Serial sections were generated for each leaflet and were specifically intended for histological analyses. To evaluate the overall cellular and tissue structures, representative sections from paraffin-embedded samples were stained using hematoxylin and eosin (HE).

Statistical analysis

The results of the organizational assessment were quantified by measuring the extent of preservation within each area, number of cells per unit area, and area size. Statistical analysis was performed using Statistical Package for Social Sciences (version 26.0; IBM Corporation, Chicago, USA); independent variables for normality were assessed using the Shapiro–Wilk test, and significance was determined using the Mann–Whitney test, as appropriate. Data are reported as mean ± standard deviation, with a significance level set at p < 0.05 for all the tests.

Gross evaluation

Histological evaluation of both naive and cryopreserved pulmonary valve tissues revealed that the trilaminar structure—composed of the fibrosa, spongiosa, and ventricularis—was maintained in both groups, confirming that the overall architecture of the valve was preserved following the cryopreservation and thawing processes (Fig. 3). The general morphology of the tissues was comparable between groups, although minor variations were observed. In particular, the ventricularis layer appeared slightly thinner in cryopreserved samples compared to naive tissues.

Figure 3. Histology of native (A, B) and cryopreserved (C, D) pulmonary valve after staining with H&E. Despite cryopreservation, the trilaminated architecture is maintained, but overall cellular preservation has decreased (original magnification: A and C,
×100; B and D, ×400). v, ventricularis; s, spongiosa; f, fibrosa.

Structural preservation

When assessing the entire structure, structural preservation remained largely consistent between naive and cryopreserved tissues, with no statistically significant differences in total preservation observed across the layers (Fig. 4). The fibrosa layer, primarily composed of collagen, exhibited a significant decrease in structural preservation in cryopreserved samples, as highlighted in the bar graph (Fig. 5B, fibrosa). This disruption in collagen organization was evident in the histological sections (Fig. 3B, D), confirming the partial deterioration of the matrix. The spongiosa layer, characterized by its proteoglycan-rich matrix, showed a relatively consistent preservation of structure, with only minor disorganization detected in cryopreserved samples. These changes were not statistically significant (Fig. 5B, spongiosa), and the layer maintained its loose and hydrated structure. For the ventricularis, which is abundant in elastic fibers, no significant differences in structural preservation were observed between naive and cryopreserved tissues (Fig. 5B, ventricularis). Considering that no significant differences were observed in the overall structure, these findings suggest that, despite layer-specific differences, the valve maintained overall structural integrity post-cryopreservation.

Figure 4. Histology of native (A, C) and cryopreserved (B, D) pulmonary valve after staining with H&E (original magnification; A and B, ×100; C and D, ×400). (E) In the total structure, no significant differences were observed between naive tissue and cryopreserved tissue for the three metrics.

Figure 5. (A) Histology of each layer of native and cryopreserved pulmonary valve after staining with H&E (original magnification ×400). (B) A bar graph representing the mean and standard deviation for each metric in each layer. Closer examination of individual layer structures, a significant decrease in both the level of structural preservation and the number of cells per unit area was observed in the fibrosa layer.

Cellularity

Quantitative analysis of cell density (cells/mm₂) showed variable results across the layers. In the fibrosa layer, a statistically significant reduction in cellular density was detected in cryopreserved tissues compared to naive tissues (Fig. 5B, fibrosa). The cell-to-area ratio of fibrosa layer in cryopreserved samples was also significantly lower, which corresponds to the partial disruption of the collagen matrix. The spongiosa layer, despite minor reductions in cellularity, did not exhibit statistically significant differences between naive and cryopreserved tissues (Fig. 5B, spongiosa). The ventricularis layer displayed similar cellular densities between naive and cryopreserved tissues, with no significant reduction in cell count observed (Fig. 5B, ventricularis). This is consistent with the overall preservation of the elastic fiber network noted in the structural preservation analysis (Fig. 4). For the entire structure, when all three layers were considered together, no significant differences in cellular density or cell-to-area ratio were observed between naive and cryopreserved tissues (Fig. 4E). While layer-specific changes were noted, the overall cellularity of the valve tissue was relatively well-preserved, supporting the viability of the cryopreservation process across the entire valve.

Homograft valves, introduced in the 1960s, have been a significant advancement in cardiac surgery in human medicine owing to their inherent advantages (9). Homograft valves were initially recognized for their ability to prevent thromboembolic events without the need for long-term anticoagulation, distinguishing them from mechanical valves. Compared with bioprosthetic valves, homograft valves also exhibit a lower incidence of SVD (9,35). Additionally, they are resistant to infection, making them a preferred choice for patients with infective endocarditis or abscess (15). While no significant differences were observed in preventing infective endocarditis between bioprosthetic and homograft valves (17,38), the consensus leans toward the effectiveness of homografts in preventing this condition (21).

When examining freedom from SVD as a conclusive indicator of valve selection, comparing homografts and bioprosthetic valves have yielded varying results (20). Medtronic Freestyle bioprosthetic valves showed 86% freedom from valve degeneration 8 years after implantation, whereas homografts showed 37% freedom. Concerning Perimount-stented bovine pericardial bioprosthetics and homografts, freedom from explanation for SVD after 10 years was 93% and 91%, respectively (33). While no significant difference was observed in the first 5 and 10 years after pulmonary valve transplantation, a higher risk of structural valular degeneration was observed with bioprosthetics at 15 years (81% vs. 31%). In younger patients, homografts showed a 92% freedom from structural valve degeneration 5 years after pulmonary valve transplantation, whereas bioprosthetics had a 53% freedom from SVD (4).

SVD is generally associated with several pathophysiological factors, including calcification, immune responses, and ischemia (20,21). Among these, calcification is the primary contributor to SVD, and it differs between bioprosthetic and homograft valves. Homograft valve calcification is presumed to result from homograft processing, possibly because of antibiotic treatment and cryopreservation (21). Cryopreserved homografts stored in liquid nitrogen vapor exhibited macroscopic or microscopic cracks after rapid thawing (36). Such damage can be attributed to the impairment of ECM, including elastic and collagenous fibers, which can serve as deposition sites for calcium phosphate minerals. Generally, unlike crosslinked bioprosthetic valves with glutaraldehyde, extensive calcification does not occur, and cuspal calcification is observed less than that in the aortic wall. Glutaraldehyde-treated bioprosthetic valves undergo two main calcification mechanisms (20,21). First, glutaraldehyde-induced cell death increases the calcium influx into cells due to the inactivity of ion pumps in the cell membrane. Second, in living tissues, proteoglycans prevent the calcification of collagen fibers, but glutaraldehyde cannot adequately cross-link proteoglycans, leading to gradual proteoglycan degradation and loss and promotion of calcification in bioprosthetic valves (20). Owing to these mechanistic differences, calcification in homograft valves does not lead to significant primary tissue failure, unlike that in bioprosthetic valves (21).

Immune reactions are identified as a mechanism of homograft destruction and have been supported by clinical studies. The immune responses include pannus formation, inflammation, and thrombosis. These immune reactions are more pronounced when valve replacement is performed in younger patients. Homograft valves are characterized by very low immunogenicity and show superior outcomes than bioprosthetic valves in young patients. This contrasts with the common knowledge that glutaraldehyde, which is used in bioprosthetic valve treatment, removes immunogenicity from xenografts. However, recent research has suggested that glutaraldehyde is insufficient to eliminate immunogenicity (24,32). In addition to calcification and immune reactions, cryopreservation processes involving controlled-rate freezing, storage and thawing can lead to ice crystal formation and freezing artifacts, potentially contributing to homograft valve deterioration. These processes have been observed to cause more destruction of ECM structures than in fresh or vitrified tissues, eventually contributing to graft dysfunction (4,14,16,21).

Human heart valve decontamination protocols vary slightly among studies. Most studies used broad-spectrum antibiotics, primarily penicillin, streptomycin, cefoxitin, vancomycin, amikacin, and gentamicin. Of the 22 reviewed studies, only 11 included antifungal agents (26). This antibiotic and antifungal cocktail was mixed in a RPMI 1640 or custodial solution. Some studies explored different incubation times and temperatures, but culturing at 37°C for 6-12 h did not significantly reduce contamination compared with culturing at 4°C for 24-48 h (8.1% vs. 5.9%, respectively) (26). Consequently, most studies opted for antibiotic incubation at 4°C, as it is hypothesized to maintain tissue integrity while allowing antibiotics to function effectively. Because amphotericin B exhibits cytotoxicity against fibroblasts in human heart valve leaflets, its use is not strongly recommended as an antifungal agent (6,13).

The properties of the cryoprotectant DMSO were discovered in 1959. DMSO enhances cell membrane permeation, making the membrane thinner and more hydrophilic (11). Increasing the concentration of intracellular solutes can prevent ice crystal formation in water, ultimately aiding in vitrification (23,37). Although 10% DMSO is highly effective as a cryoprotectant, it also induces cytotoxicity by causing excessive pore formation in cell membranes (3). Consequently, research has been conducted to identify substances that can maintain cryoprotective qualities while reducing DMSO concentration through various mixing methods. The addition of substances, such as sorbitol, disaccharides, and hydroxyethyl starch, can lower DMSO concentration (3,35). However, the most widely used technique is using 10% DMSO, followed by stepwise dilution and washing after thawing to reduce cytotoxicity (37).

Subsequently, controlled-rate freezers were used to gradually lower the temperature of the samples at a consistent rate. Excessively rapid or slow cooling can adversely affect cell viability. Rapid cooling proceeded without cell shrinkage, leading to the formation of microscopic ice crystals. In contrast, slow cooling induces significant cell shrinkage, leading to increased concentrations of extracellular and intracellular solutes owing to ice formation and intracellular dehydration, ultimately causing cell damage. Therefore, the optimal cooling rate typically falls within the range of 0.3°C to 10°C per min, with specific speeds varying depending on the cell type. Cooling protocols also vary among studies, but in general, cooling is conducted at a rate of −1°C per min down to −40°C, followed by cooling at a rate of −5°C per min down to −70°C (29). In the present study, to minimize cell damage caused by the heat released during the freezing process, known as ice crystallization (Fig. 6A), cooling was conducted at a rate of −0.5°C per min until just before heat release occurred, followed by rapid temperature reduction. A warming process was necessary for post-excessive cooling. If not appropriately managed, this could induce an artificial temperature elevation, which must be prevented (Fig. 6B).

Figure 6. Inadequate cooling graphs observed during the cooling process. (A) Cooling proceeded at a rate of –0.5°C/min down to –20°C, during which latent heat was generated (asterisk). (B) Following a rapid temperature decrease, there was an increase in the sample's temperature during the artificial warming process (arrow).

It is essential to assess the suitability of the entire process, from harvesting to thawing, to predict the functionality of homograft valves after cryopreservation and transplantation. Evaluation of tissue integrity and viability through histology is crucial for assessing the results. Valves typically consist of valvular endothelial cells (VEC), valvular interstitial cells (VIC), and ECM (19,31). While endothelial cells were previously believed to have only minor physiological roles, recent research indicates that the glycocalyx located on the surface of the endothelial cells can activate signaling pathways, leading to the secretion of substances such as nitric oxide, prostacyclin, and endothelin 1 (19,20). Following the harvest of heart valves, a significant loss of VEC occurs owing to ischemic damage, handling, and sterilization processes. Preserving VEC may increase durability after valve transplantation but can also induce homograft immunogenicity and immune rejection reactions. Therefore, the benefits of VEC preservation have not been clearly established yet. VICs play a crucial role in regulating protein synthesis and enzymatic degradation of the ECM, thereby maintaining the structural integrity of the valve. Changes in ECM stiffness can occur during homograft valve processing and cryopreservation, potentially altering the VIC phenotype. The subsequent loss of cellularity and changes in the remaining VICs’ phenotype after transplantation can lead to valve failure. Despite its thickness of 300-700 μm, the valve can maintain strength owing to the delicate arrangement and composition of proteins within the ECM (7,31).

Histologically, the heart valve consists of three distinct layers: fibrosa, spongiosa, and ventricularis, each playing a crucial role in the valve’s function. The fibrosa, which accounts for approximately 45% of the valve thickness, provides most of the structural strength due to its abundance of type I collagen fibers. The ventricularis layer, composed of aligned elastin fibers, contributes to the elasticity of the valve, while the spongiosa is rich in proteoglycans and glycosaminoglycans, which provide high hydrous content to dampen the effects of high-pressure blood flow. During cryopreservation, it is commonly reported that the ECM, particularly collagenous and elastic fibers, undergoes deterioration or alteration, leading to disruptions in the trilaminar structure of the valve (7). However, in this study, the trilaminar structure was generally well-preserved post-cryopreservation, with only specific structural deficits observed in each layer.

In the fibrosa layer, despite the abundance of collagen fibers, cryopreserved tissues exhibited a noticeable decrease in cellularity and partial matrix disruption. This observation is consistent with previous findings indicating that collagen fibers are particularly susceptible to damage from extracellular ice formation (31). In addition to ice formation, cold ischemic injury likely contributed to the reduced cellularity observed in this layer. Prolonged exposure to suboptimal conditions during the pre-freezing period could have led to cellular injury or death, which might explain the diminished cellularity in the fibrosa. The damage to the collagen matrix, compounded by ischemic stress, could affect the biomechanical properties of the fibrosa layer. Reduced cellularity and alterations in the collagen matrix may compromise valve durability, potentially accelerating structural degeneration of the valve. However, the degree of damage observed in this study was relatively moderate compared to more severe degradation reported in other studies (10,30). The spongiosa layer showed better preservation during cryopreservation. There were minor reductions in cellularity and ECM organization, but these changes were not statistically significant. In the ventricularis layer, thinning of the elastic fiber network and a reduction in layer thickness were evident in the cryopreserved samples. However, overall cellularity between naive and cryopreserved tissues did not differ significantly. The separation between the ventricularis and spongiosa layers became more distinct in the cryopreserved samples, which may be attributed to subtle changes in tissue composition or reorganization during the cryopreservation process. Despite the thinning of the elastic fibers, the structural integrity of the ventricularis layer was generally maintained, indicating that cryopreservation had only a moderate impact on this layer. The findings of this study suggest that the impact of cryopreservation on the ECM and cellular integrity of pulmonary valve tissues was relatively mild. While there were noticeable changes, particularly in the fibrosa layer, these were less severe than those reported in other studies (10,30). This suggests that cryopreserved heart valves may retain sufficient structural integrity and function, making cryopreservation a viable method for long-term valve storage.

In conclusion, cryopreservation and thawing did not cause significant structural or cellular damage to the pulmonary valve tissues. Additional studies are required to explore the functional viability of these valves after clinical adaptation and to better understand the mechanisms of minor tissue damage during cryopreservation.

Not applicable.

This work was supported by a research fund from the Chungnam National University (2024-1150).

The authors have no conflicting interests.