Traumatic injuries (ballistic, high-energy motor vehicle accidents, burns, etc.), congenital diseases, and neoplasia are responsible for severe disfigurements and amputation of functional aesthetic units in patients of all ages. For cases where autologous reconstructive options have been exhausted, vascularized composite allograft (VCA) transplantation offers significant quality of life improvements [1]. Although immense progress has been made in VCA procurement techniques, graft adjustments to the recipient, and post-operative care to achieve optimal results [2, 3], a more effective and longer graft preservation method is needed to make VCA transplantation a routine reconstructive procedure.

tThe method used for ex vivo preservation is a key factor in post-transplantation graft function and rejection risks [4–6]. For about half a century now, the combination of University of Wisconsin (UW) solution (or equivalent), an organ bag, and an icebox have been used to store all organs from procurement to transplantation. This method was chosen as a gold standard for its simplicity and cost-effectiveness [7]. However, limitations for this approach are obvious considering that at +4°C tissue metabolism continues at ∼10% of nominal levels, resulting in a maximum preservation time of 4–6 h as the tissue effectively slowly suffocates. Further, cold ischemia causes ongoing injury to the graft, adversely affecting post-operative graft viability and functional outcomes [5, 8, 9].

Hence, the field has been pursuing better technologies to overcome ex vivo preservation injury. Machine perfusion (MP) is a promising alternative since it allows dynamic and monitored ex vivo organ preservation by supplying allografts with nutrients and oxygen at concentrations tailored to meet specific metabolic needs of the tissue [10, 11]. Oxygenated MP was first optimized in solid organ transplantation and is now clinically used in kidney [12], liver [13], heart [14], and lung [15] transplantations. VCA ex vivo perfusion studies are rather recent in the literature and are still mostly confined to animal models. Swine limb models (forelimbs or hindlimbs) are frequently used for VCA preservation studies since they are easy to procure and contain common tissue types required for reconstructive procedures (skin, fat, muscle, nerve, tendon, and bone). Previous studies have suggested the superiority of different machine perfusion protocols in swine limb or myocutaneous flap preservation, when used either ex vivo [16–18] or in situ after replantation [19–21]. However, to our knowledge, no study reports the outcomes of ex vivo perfused VCAs after allotransplantation to assess the impact of immunological incompatibility between donor tissues and the recipient animals, which is a central component of clinical VCA procedures [22].

In this study we used a previously described heterotopic limb allotransplantation model in swine [23] to evaluate the possibility of enhanced ex vivo preservation of VCA using MP. Our work relied on 24-h ex vivo preservation of VCA using either SCS or subnormothermic machine perfusion (SNMP, +21°C), demonstrating that SNMP preserves VCAs for up to 24 h before allotransplantation, increasing the total viable duration for porcine partial hindlimbs 4–6 times beyond what can be achieved on ice.

After several breakthrough clinical trials in solid organ preservation, machine perfusion is permeating the field of VCA, aiming to enhance preservation duration and allograft viability [7]. In this study, we demonstrated the superiority of preserving VCAs for 24 h with a SNMP protocol compared to conventional time-matched SCS in a porcine heterotopic limb allotransplantation model.

In previous reports, in vivo viability of VCA after machine perfusion preservation was assessed using replantation models with follow-up durations ranging from 12 h to 7 days. In 2017, Kueckelhaus et al. were to the first to replant porcine forelimbs after 12 h of hypothermic perfusion [16]. They observed reduced ischemic injuries in perfused limbs compared to standard 4-h SCS, thus extending VCA preservation up to 3 times longer than the gold standard. Two years later, the same team extended the perfusion time to 24 h and validated in vivo viability with a similar replantation model [19]. More recently, Kruit et al. replanted forelimbs after 18-h SNMP preservation and followed the recipient pig for 12 h to assess muscle function and histology compared to 4-h SCS preservation [20]. Our findings were in line with these results, as we were able to preserve VCAs for 24 h in subnormothermic conditions with stable arterial resistances, moderate edema, and outflow potassium and lactate concentrations remaining in physiological values. Additionally, pigs that received perfused VCAs showed good clinical tolerance and did not suffer any systemic reaction, infection, or acute allograft rejection. In contrast, 2 out of 3 SCS control pigs developed poor general condition most likely due to graft failure (inflammation, rhabdomyolysis as suggested by high potassium levels). The lactate spike found in one of the SCS animals was considered resulting from hepatic and multi-organ failure resulting from graft failure. Pain management was considered optimal by experienced veterinarians. Complete necropsies were performed on all animals euthanized before POD14, and no alternative conditions such as sepsis or disseminated intravascular coagulation were found. Despite the lack statistical significance, qualitative assessment of muscle histology in the SCS group showed major muscle degeneration, supporting the hypothesis of graft failure leading to the poor general condition of the recipient animals. However, graft failure was not clinically obvious given that there was no direct sign of graft ischemia and graft degeneration was concluded upon euthanasia dissection. Hence, delayed VCA failure can insidiously occur as late as POD9 with no or minimal skin lesions. This phenomenon was previously observed in rat hind limb transplantations after not 24 but 48 h of SCS. Interestingly, histological analyses of muscle biopsies showed no difference between SNMP and SCS immediately after ex vivo preservation but revealed a statistical difference at the end of study in favor of SNMP, underlining the importance of in vivo analysis of the allograft in a longer follow up period than previously described. Our hypothesis is that reperfusion injuries slowly deteriorate muscular cells leading to extended myocyte apoptosis and microvascular damage, viciously killing the muscle tissue and releasing necrosis marker in the systemic circulation although those markers have not been investigated in this study.

Protocols based on Subnormothermic perfusion appear to strike a balance in temperature. Hypothermic machine perfusion has generally led to satisfactory outcomes in VCA preservation [16, 20, 28] without additional oxygen carriers [28]. However, cooling VCAs down to 8°C–10°C appears to result in vascular injuries and subsequent higher interstitial edema (considered clinical revelation of vascular leakage induced by endothelial damage during ex vivo storage), especially after 24 h of ex vivo perfusion. Krezdorn et al. reported an average weight gain of +41% of after 24 h of hypothermic perfusion [19]. Similarly, Kruit et al. reported +18.6% of weight variation after 18 h of hypothermic perfusion [20], while we were able to limit edema formation to an average of +15% weight for all limbs after 24 h of SNMP. These results are also in line with previous studies on SNMP organ preservation [25, 29], where SNMP was used as an organ reconditioning phase after an ischemic period and prior to transplantation. However, we did not directly compare SNMP with different temperatures in this study. A direct comparison utilizing the same protocols and models are needed to draw further conclusions and develop a mechanistic understanding of how perfusion temperature effects this very complex tissue. Low-temperature preservation has the potential to mitigate cell metabolism, therefore suggesting longer preservation durations [30]. Subzero preservation techniques seem to be promising alternatives for multi-day preservation of solid organs [31–35], but solid evidence still needs to be demonstrated in VCA. In the realm of Normothermic Machine Perfusion (NMP), Fahradyan et al. [36] demonstrated 12-h preservation of porcine VCAs with preserved weight and muscle contractility. They suggested that extending normothermic red blood cell perfusion to 24 h could be feasible, albeit with limited significant evidence versus SCS controls. A pioneering work from Werner et al. [37] in fresh human arms demonstrated encouraging results of 24-h near-normothermic (30°C–33°C) perfusion. This sub-physiologic protocol however imposed using a blood-derived perfusate for adequate oxygen transportation. In contrast, the use of subnormothermic “room” temperatures (20°C–25°C) in machine perfusion offers several advantages. First, it simplifies the perfusion device setup, eliminating the need for a heating or cooling system [29, 38]. Second, it allows clinicians to reach a balance between oxygen needs and subdued tissue metabolism resulting in perfusate simplifications [28]. Perfusates that require red blood cells or oxygen carriers that are essential to normothermic “physiological” conditions are no longer needed to ensure adequate oxygenation of the composite tissue at +20°C, thus eliminating the potential immunogenicity of such additional components [39].

In our experiments, all limbs were perfused with a modified acellular Steen solution [25] to minimize edema. Fine-tuned control to maintain a low-pressure circuit (around 30 mmHg) was also needed to mitigate edema—likely due to excessive endothelial shear stress at high pressures - using a continuous flow perfusion pump. Whether non-linear or continuous flow provides better outcomes in VCA machine perfusion still remains unclear [40, 41]. Our previous work demonstrated that non-linear flow has some advantages over continuous flow with improved endothelial function and decreased ischemic injury, although the differences were limited [41]. In the present study we utilized continuous flow for practical reasons.

Allotransplantation is only rarely performed in machine perfusion studies, likely due to cost-related issues. To our knowledge, this study is the first to report allotransplantation of perfused VCA in swine with a 2-week follow up. Our laboratory has been extensively working on VCA in miniature swine to implement new tolerance induction protocols in VCA, rising our expertise in immunosuppressants in swine [42, 43]. Consequently, our experience led us to use a bitherapy made of tacrolimus and steroids. As a matter of fact, on short term studies, the immunogenic risk is mostly confined to acute rejection episodes, which are usually mitigated with high-dose steroids [44].

A major limitation of this work was the small sample size (n = 3 per group). Performing allotransplantation instead of replantation to approximate clinical settings subsequently doubled the number of swine needed, thereby limiting the group sizes. While this limits the power of the statistical analysis, leading to increased risk of type-II errors, the SNMP group notably showed clear improved results including a higher recipient survival rate after transplantation, which in our opinion provided additional evidence to a growing VCA perfusion literature. We were also unable to assess limb function after transplantation with this surgical model. However, based on the positive results of this study, our SNMP protocol can be used in other swine VCA models to achieve this goal. Thus, one could argue that 24-h SCS is far more than what is usually the accepted limit for cold ischemia time in human clinical setting for VCA. Although, no studies have proven that swine VCA respond exactly as human tissues. Our previous experience in rat VCA preservation studies showed that rat VCA graft survival was similar after 6 and 24 h of SCS [25]. If graft injuries were evidently expected after 24 h SCS, it served as a negative control to our 24-h SNMP protocol. Since this work mainly aimed at testing the compatibility of 24-h SNMP with successful VCA allotransplantation, another limitation is the lack of focus on nerve histology and nerve conduction assessment, which should be specifically addressed in further studies.

Overall, we showed that a simplified SNMP protocol allowed 24-h VCA preservation in a swine allotransplantation model with confirmed tissue viability until 14 days after allotransplantation. Further studies should focus on assessing neuromuscular function in a long-term follow-up to match the clinical setting.