Acellular Normothermic Machine Perfusion Does not Exacerbate Acute Rejection in Rodent Vascularized Composite Allografts

Abstract

Background.

Vascularized composite allotransplantation faces significant challenges, including limited preservation time and high rates of acute rejection because of the ischemia susceptibility of muscle and immunogenicity of skin. However, no studies have investigated whether normothermic machine perfusion (NMP) has an effect on acute rejection compared with static cold storage (SCS).

Methods.

Heterotopic hindlimb transplants were performed in a full-mismatch model after 6 h of SCS or NMP. Postoperative assessments included clinical scoring, histological examination, cytokine profiling, and flow cytometry of splenocytes and lymphocytes at the end of the study.

Results.

Clinical assessment revealed significantly lower rejection scores in the NMP group on postoperative day (POD) 2 (P = 0.03). Notably, histology indicated delayed rejection, with Banff scores showing severe rejection from POD2 in the SCS group and from POD6 in the NMP group. Cytokine profiling demonstrated early upregulation of anti-inflammatory interleukin-10 in the NMP group at POD1, and flow cytometry showed a reduction in B cell populations.

Conclusions.

While NMP is not a cure-all, it is associated with a delayed onset of acute rejection compared with SCS in a full-mismatch rodent vascularized composite allotransplantation model. Future studies should focus on analysis earlier in the rejection process and on long-term graft survival to further elucidate the mechanism of the effect of NMP on the immune response.

INTRODUCTION

In the last 2 decades, over 200 vascularized composite allotransplantations (VCA) have been reported worldwide.¹ As a collective group, VCA refer to organs composed of multiple tissue types, such as skin, muscle, bone, and vasculature. These transplants are performed on patients who suffer from severe facial defects,² bilateral upper extremity amputations,³ and major composite tissue defects.⁴ Major challenges in VCA include limited storage time and rejection post-transplantation.⁵ Damage to the muscle tissue in VCA, but also in autologous tissue needed for replantation, occurs as early as 6 h of static cold storage (SCS), the current clinical gold standard.^6-8 After allotransplantation, about 89% of VCA transplants suffer acute rejection despite immunosuppression in the first year because of the high immunogenicity of skin tissue.^9-13

SCS, characterized by cold ischemia time,¹⁴ has been implicated in exacerbating immunogenicity and rejection because of increased ischemia-reperfusion injury (IRI), oxidative stress, and inflammatory response in solid organs. In rodent livers, a direct correlation between the molecular mechanisms of IRI and duration of ischemic time has been previously established.¹⁵ Contrary to SCS, use of machine perfusion systems allows for continuous perfusion and assessment of organs at desired temperatures, mitigating IRI, and increasing organ preservation time.^16,17 Benefits of machine perfusion have been shown to attenuate both ischemic damage and inflammatory response post-transplantation, leading to improved graft functionality and better outcomes in lung and human liver transplantation.^18-21 In human kidney transplantation, however, no differences in the prevalence of acute rejection were observed when comparing SCS and machine perfusion.²² In VCA, the effect of machine perfusion has shown a beneficial effect in SCS^16,23,24 and warm ischemic time²⁵ transplantation models, thereby focusing on IRI; however, no studies have been performed assessing its effect on rejection.

The immunological impact of machine perfusion on VCA, particularly cytokines at the tissue level and modulation, remains unclear. Therefore, this study aims to elucidate the impact of normothermic machine perfusion (NMP) versus SCS on acute rejection in a rodent VCA model.

MATERIALS AND METHODS

Animals

Rats (250 ± 50 g) were used for all experiments, of which recipients were inbred Lewis rats and donors were Brown Norway rats (Charles River Laboratories, Wilmington, MA). The animals received humane care in accordance with the National Research Council guidelines and the experimental protocols were approved by the Institutional Animal Care and Use Committee (IACUC) of Massachusetts General Hospital (Boston, MA) under protocol number 2021N000005.

Study Design

Partial hindlimb transplants were performed in 2 surgical groups as shown in Figure 1: (1) NMP (n = 8); (2) SCS group (n = 9). Donors were Brown Norway and recipients Lewis rats, a full-mismatch model.²⁶

FIGURE 1.

Experimental design. Hindlimbs were procured from Brown Norway animals. Grafts were subjected to either NMP or SCS for a period of 6 h, after which they were transplanted in a heterotopic partial hindlimb model. At the end of the study, splenocytes and lymphocytes were isolated, plasma was obtained, and skin and muscle samples underwent histological analysis.

Machine Perfusion System

Modified Steen (Steen+) was used as perfusate and prepared as described earlier¹⁶ with minor adjustments of the sodium chloride, potassium chloride, and sodium bicarbonate (decreased from 86 to 71.2 mmol/L, 4.6 to 2.68 mmol/L, and 15 to 5.95 mmol/L, respectively). Furthermore, 100 mg/L vancomycin and ceftriaxone were added. The perfusate composition used is displayed in Table S1 (SDC, https://links.lww.com/TXD/A859). Perfusate was circulated using a roller pump system (Masterflex L/S, Vernon Hills, IL) with tubing (Masterflex platinum-cured silicone tubing, L/S 13, Cole-Parmer, Vernon Hills, IL) delivering perfusate into and out of the perfusion reservoir (Figure 1). Temperature was regulated by a water bath (Polystat Cooling/Heating Circulating Bath, Cole-Parmer), set at 37 °C, through double-jacketed perfusion system components (Radnoti, Covina, CA, USA). Perfusate oxygen concentration was maintained within a close range of 450 mm Hg using a 95% O₂/5% CO₂ gas cylinder (Airgas, Radnor, PA, USA). Pressure transducer (PT-F, Living Systems Instrumentation, St Albans City, VT) was connected close to the angio catheter (BD Angiocath 24G) in the femoral artery during perfusion. Before the start of the perfusion, base pressures were obtained to account for the system pressure of each flow rate. A maximum flow of 1.5 mL/min was observed, and flow was regulated in a pressure-based manner, aiming at a vascular resistance of 30–35 mm Hg.

An open circuit was used during the stabilization phase of the perfusion. After 1 h of machine perfusion, a closed circuit was maintained with a total volume of about 220 mL of perfusate. At 5 h 45 m, the circuit was opened and fresh perfusate was used to washout the accumulated metabolites before transplantation.

Inflow and outflow perfusate samples were analyzed using a Siemens RapidPoint 500 (Siemens, Munich, Germany). VCA performance metrics were analyzed to determine viability during perfusion (pH, O₂ consumption, glucose consumption, lactate, potassium). NaHCO₃^- titration was performed to correct for acidosis if needed. Vascular resistance was calculated using pressure readings taken at every time point and defined as: pressure/flow × initial weight. Oxygen consumption was calculated using a modified Fick equation using circuit flow, limb weight, and pre- and post-limb oxygen contents (0.00314 × flow[Po₂ in − Po₂ out]/weight). Weight change was defined as: final weight − initial weight/initial weight, presented as a percentage.²⁷

VCA Transplantation

Whole hindlimbs were procured as described earlier by our group and others.^28,29 Briefly, skin was excised circularly at the level of the hip. The fat pad was dissected, and the femoral vessels were exposed and skeletonized. The femoral artery was cannulated with a 24G cannula, and the femoral vessels were ligated at the level of the inguinal ligament. Muscle tissue was separated with sharp scissors, and at the mid-femur level, the femur was transected with a bone cutter. Immediately after procurement, a pressure-controlled manual flush with 3 mL (200 IU) of heparin saline at room temperature was performed, during which venous return was observed. Next, the VCA was subjected to either 6 h of NMP or SCS. For SCS, VCA were flushed with 5 mL ice-cold histidine-tryptophan-ketoglutarate (HTK) during which venous return was observed, and stored in a fridge at 4 °C. After 6 h of NMP or SCS, the VCA were flushed with 3 mL of heparinized saline to washout the preservation solution, and transplanted. Recipient preparation was performed as described earlier by our group.²⁹ Briefly, a 1 cm skin incision was made over the inguinal ligament and a 3 cm skin incision was made in the flank. The fat pad was dissected, and the femoral vessels were exposed, skeletonized, and ligated distally, just proximal to the epigastric branches. The foot from the donor limb was removed at the mid-tibial level using bone cutters, and skin was excised to form a skin island. The donor limb was tunneled, presenting the skin island in the flank and the donor femoral vessels in the groin. Recipients did not receive any immunosuppression drugs before or after transplantation. On postoperative day (POD) 7, animals were sacrificed. To ensure no surgical failure had occurred, vascular patency of the transplanted VCA was confirmed in all animals by performing a milking test of the femoral vessels. Next, a cardiac puncture was performed to obtain whole blood, and the spleen and lymph nodes were procured for further analysis.

Postoperatively, daily flap images were taken for blinded clinical assessment by an experienced microsurgeon using a clinical VCA rejection score. Briefly, grade 0 constitutes no difference between graft and native skin. Grade 1 shows mild erythema, grade 2 moderate erythema with the beginning of scaling and scabbing, grade 3 severe erythema and scabbing with areas of epidermolysis, and grade 4 constitutes full-thickness graft epidermolysis with areas of necrosis.³⁰ Blood draws were performed on PODs 1, 3, and 7, as described earlier by our group.³¹

Daily skin biopsies were performed in 1 replicate per group, as taking biopsies can also elicit a rejection response. In another replicate, additional POD 3 biopsies were obtained. In all replicates, POD 7 skin and muscle biopsies were obtained. Slides were stained with hematoxylin and eosin. Blinded evaluation by 2 clinicians was performed for all biopsy samples using the Banff criteria to assess acute cell-mediated rejection.^32,33 Muscle tissues were evaluated and scored using the histology injury scoring system for hypoxia-induced muscle injury.³⁴

Plasma Assessments

Cytokine arrays on rat plasma samples were performed by Eve Technologies (Calgary, Alberta, Canada), using their Rat Cytokine/Chemokine 27-Plex Discovery Assay Array (RD27). Using a multiplexed immunoassay platform based on Luminex xMAP technology,³⁵ the following cytokines, chemokines, and growth factors were measured in each sample: Eotaxin, epidermal growth factor, Fractalkine, IFNγ, interleukin (IL)-1α, IL-1β, IL-2, IL-4, IL-5, IL-6, IL-10, IL-12p70, IL-13, IL-17A, IL-18, IP-10, GRO/KC, TNFα, granulocyte-colony stimulating factor, GM-CSF, monocyte chemmoattractant protein-1, Leptin, LIX, macrophage inflammatory protein (MIP)-1α, MIP-2, RANTES, and vascular endothelial growth factor (VEGF)-A. Hydrocortisone concentrations were measured in perfusate and plasma samples using a rat cortisol ELISA kit (LS-F27363, LS Bio, Newark, CA) to assess potential carryover of steroid-containing perfusate following graft implantation.

Tissue Isolation and Flow Cytometry Analysis

Splenocytes and lymph node cells were isolated from each animal at the end of the study. Briefly, tissue was washed with cold phosphate buffered saline before being passed through a 70 μm cell strainer. The filtrate was then centrifuged, and red blood cells were lysed using Ammonium-Chloride-Potassium lysis buffer (ThermoFisher). Then, the cell suspension was treated with an antibody cocktail for the following markers: CD45, CD3, CD4, CD8a, CD43, HIS48, CD45R, and CD161.³⁶ Fixable Viability Dye eFluor 780 (ThermoFisher) was used as a viability marker. Flow cytometry was performed in a BD FACSAria instrument at The Harvard Stem Cell Institute-Center for Regenerative Medicine (HSCI-CRM) Flow Cytometry Core Facility at MGH. Flow cytometry results were analyzed using FlowJo v10.8 Software (BD Life Sciences).

Statistical Analysis

Clinical rejection score and histology score differences between groups were analyzed using a mixed-effect analysis with multiple comparisons. The time-series plots and bar charts are represented as mean with range. Statistical tests used are also noted in the figure legends. Outliers were identified using ROUT, Q = 1%. All statistical analyses were performed using Prism 10 for Mac OSX (GraphPad Software, La Jolla, CA). P values ≤0.05 were considered significant.

RESULTS

Perfusion Parameters

VCA were procured from donors in 17 (±9) min on average and warm ischemic time was a maximum of 12 min in all hindlimbs. In the SCS group, VCA were flushed with heparin saline and HTK and stored at 4 °C. In the NMP group, VCA were also flushed with heparin saline and the 6-h perfusion was started thereafter. Perfusion parameters are displayed in Figure 2. After 1 h, a stable flow of 1.5 mL/min was reached in all replicates, and vascular resistance remained stable with a mean of 29.51 mm Hg/mL/min throughout the perfusion. Potassium remained within the physiological range, with a mean of 4.89 mmol/L, despite the closed circuit phase. Potassium and lactate accumulation were apparent between 1 and 5 h because of the closed circuit phase. After reopening the circuit with fresh perfusate, potassium was a mean of 4.3 mmol/L, and lactate was 1.8 mmol/L. Oxygen and glucose consumption were stable throughout perfusion with a mean of 0.10 mL/g/min and 1.48 mg/(h·g), respectively. Weight increase at the end of perfusion was a mean of 8.24%, which is well below the transplantable range and showed no significant difference compared with the SCS group.

FIGURE 2.

Perfusion parameters of the NMP group. A, Arterial resistance decreased from a mean of 83.8 mm Hg (min/mL) on connection to the system to 20.4 mm Hg (min/mL) at 1 h as a sign of stabilization. B, Potassium and (C) lactate accumulated during the closed circuit phase and returned to a mean of 4.3 and 1.8 mmol/L, respectively, after the circuit was reopened and fresh perfusate was circulated. D, Oxygen consumption and (E) glucose uptake remained stable throughout perfusion. F, Weight change is between −0.14% and 12.5% in all NMP replicates and between 4.04% and 9.68% in all SCS replicates, and was not found to be significant (mean with range, 1-way ANOVA). ANOVA, analysis of variance; NMP, normothermic machine perfusion; SCS, static cold storage.

Clinical Assessment

All transplants were successful from a technical standpoint until the end of the study as defined by visual assessment of the vascular patency test. Representations of daily images are shown in Figure 3. It should be noted that donor animals have a non-agouti dark brown coat and that none of the recipients received immunosuppression. Additional quantification of circulating hydrocortisone levels between PODs 0 and 7—performed to assess potential carryover of steroid-containing perfusate in the NMP group—is provided in Figure S1 (SDC, https://links.lww.com/TXD/A859), which demonstrates no significant differences between NMP and SCS recipients and levels compared with baseline by POD1. From POD 3 onward, erythema is observed in both groups. Furthermore, edema is visible, and induration of the tissue is palpable. From POD 5 onward, discoloration becomes patchy, suggesting areas of necrosis. On POD 6 and 7, severe epidermolysis with lymphatic fluid oozing is observed. Using clinical rejection scores, the NMP group had a significantly lower score on POD 2 compared with the SCS group (mean difference 0.57, P = 0.03) (Figure 3C). However, no differences in clinical assessment of rejection were found on the other PODs.

FIGURE 3.

Blinded clinical rejection score reveals delayed clinical signs in the NMP group. A, Representative daily clinical images of both rejection groups. B, Clinical rejection score used to assess the clinical images. C, Daily clinical scores show no significant differences beyond POD 2 (mean with range, mixed effects analysis). *P ≤ 0.05. NMP, normothermic machine perfusion; SCS, static cold storage.

Histological Assessment

Blinded histology assessment is shown of skin tissue in Figure 4A and of muscle tissue in Figure 4B. Microscopic histology at POD 1, 3, and 7 is shown in Figure 4C. Histological assessment on POD 7 did not reveal any differences between the groups, as all replicates showed frank necrosis, resulting in a Banff score of IV (Figure 4A). In the SCS group, a Banff III–IV was found as early as POD 2 based on epidermal necrosis and ulceration. This score persisted until the end of the study when frank necrosis was seen. Conversely, the NMP group did not show any signs of rejection until POD 4, severe edema and dermal inflammation were found, resulting in a Banff score of I. As late as POD 6, this group demonstrated a Banff score of III–IV based on early dermal necrosis and on POD 7 frank necrosis in both skin and muscle tissue. Assessment of the muscle injury score showed no clear differences, although there may be a trend of a lower score in the NMP group (Figure 4B).

FIGURE 4.

Histology assessment shows earlier rejection in the SCS group. A, Light microscopy, H&E staining shows a significant difference between groups with a Banff III–IV score, severe rejection, from POD2 onward in the SCS group, whereas the NMP group does not reach Banff III–IV until POD6 (mean with range, mixed effects analysis). B, Blinded, microscopic muscle injury score²⁴ shows similar results between groups; however, a large variation is seen within groups. C, On POD 1 in both groups, normal skin architecture is seen, without any signs of rejection. At POD 3, the SCS group shows a Banff of III based on infiltration, apoptotic bodies, and microthrombi, while the NMP does not show signs of rejection. By POD 7, both groups show frank necrosis with complete loss of architecture, infiltration, and apoptotic bodies resulting in a Banff score IV. *P ≤ 0.05, **P ≤ 0.01, ****P ≤ 0.0001. H&E, hematoxylin and eosin; NMP, normothermic machine perfusion; SCS, static cold storage.

Immunological Assessment Post-transplantation

Figure 5A presents a comprehensive overview of cytokine and chemokine responses under allogeneic conditions, expressed as log₂(NMP/SCS) fold changes across PODs 0–7. Positive values (>0) indicate higher expression in the NMP group, whereas negative values (<0) represent higher expression in SCS-preserved grafts. Overall, the cytokine landscape reveals a dynamic and time-dependent immune modulation between the 2 preservation methods. At POD 1, a cluster of proinflammatory and regulatory cytokines (IL-1β, IL-6, IL-10, tumor necrosis factor alpha [TNF-α]) exhibited higher expression in the NMP group, suggesting early immune activation and regulatory signaling shortly after reperfusion. By contrast, later time points (PODs 3–7) were dominated by analytes upregulated in the SCS group, reflecting an inflammatory rebound after static preservation. The accompanying heatmap (Figure 5B) highlights coordinated temporal shifts: at POD 1, NMP shows concomitant upregulation of IL-10 (anti-inflammatory) alongside select chemokines (eg, MIP-1α) and growth-associated factors (epidermal growth factor, granulocyte-colony stimulating factor). At later time points (PODs 3–7), the overall profile shifts toward higher levels in the SCS group for several analytes. Notably, IL-10 exhibits the strongest positive log₂ fold change at POD 1 among all cytokines, consistent with an early immunoregulatory milieu under NMP, whereas LIX (CXCL5) shows its largest deviation at POD 7, consistent with later neutrophil-recruitment signals. Individual cytokine trajectories (Figure 5C–H) further illustrate these patterns. IL-10 and MIP-1α both peaked early in the NMP group, whereas IL-1β and TNF-α rose later in the SCS group, aligning with a sustained inflammatory surge. Collectively, these data indicate that NMP triggers a self-limited inflammatory and regulatory response, whereas SCS leads to a more sustained proinflammatory profile.

FIGURE 5.

Cytokine and chemokine profiling reveals distinct immune dynamics between NMP and SCS groups. A, Overview of cytokine and chemokine expression represented as log₂ (NMP/SCS) fold changes across postoperative days 0–7 under allogeneic conditions. Positive values indicate higher levels in NMP-preserved grafts, whereas negative values indicate higher levels in SCS-preserved grafts. B, Heatmap summarizing the temporal distribution of cytokines and chemokines in NMP and SCS groups. C–H, Individual trajectories of representative cytokines and chemokines over 7 d post-transplantation, illustrating early but regulated activation under NMP and a more sustained inflammatory response under SCS (2-way ANOVA, mean with range). **P ≤ 0.01. ANOVA, analysis of variance; IL, interleukin; LIX, lipopolysaccharide-induced CXC chemokine; MIP, macrophage inflammatory protein; NMP, normothermic machine perfusion; SCS, static cold storage; TNF, tumor necrosis factor.

Flow cytometry analyses of total T cell, B cell, and NK cell populations are depicted in Figure 6. Allogeneic transplants showed roughly twice the number of B cells compared with T cells and a 15-fold lower population of NK cells (Figure 6B). Importantly, B cell levels were significantly elevated in the SCS group compared with NMP, indicating a slight reduction in the immune response to rejection following perfusion. Further, in the rejection group, CD4⁺ effector T cell subpopulation was approximately 2.5-fold higher than CD8⁺ cytotoxic T cells in both the SCS and NMP groups (Figure 6B), suggesting that perfusion did not significantly impact T cell differentiation into the effector and cytotoxic subsets.

FIGURE 6.

Flow cytometry analysis reveals slight changes in immune cell populations post-NMP. A, Percentage of inflammatory cell populations in rejection (allogeneic) groups post-SCS and NMP at POD 7. B, Percentage of CD4⁺ and CD8⁺ T cell subsets in allogeneic groups post-SCS or NMP at POD 7 (2-way ANOVA, mean with range). *P ≤ 0.05; **P ≤ 0.01. ANOVA, analysis of variance; NMP, normothermic machine perfusion; SCS, static cold storage.

DISCUSSION

Clinically, VCA face ischemic and immunological challenges, with IRI and acute rejection leading to severe complications and even graft loss. Thus far, preservation has relied on SCS. Meanwhile, animal studies have shown that machine perfusion is an improved preservation method that mitigates IRI.^16,25 However, no previous studies have explored the immunological effects of machine perfusion on rejection. In contrast, this study assesses the impact of NMP, relative to SCS, on acute rejection in a rodent VCA model. Findings indicate that NMP, compared with SCS, does not worsen the acute rejection response and is associated with a delayed onset based on clinical, histological, and cytokine data.

The delayed onset of rejection observed in the NMP group suggests that preservation strategy may influence early post-transplant inflammatory dynamics. At POD 1, NMP grafts exhibited a mixed cytokine profile marked by transient increases in both pro- and anti-inflammatory mediators, indicative of a controlled early activation phase. In contrast, the SCS group showed a delayed escalation of proinflammatory cytokines beyond POD 1, suggesting a more prolonged and unregulated inflammatory response. As expected, this immunological advantage under NMP was temporary—by POD 7, both groups exhibited severe rejection, reflecting the eventual convergence of inflammatory processes over time. In muscle tissue, no major differences were found in the injury score, which may be because of the tissue’s susceptibility to ischemia, and the lack of nuance in the scoring system to differentiate IRI from rejection,³⁷ since the muscle injury score was developed in the context of IRI.³⁴ Skin, being more IRI-resistant and more immunogenic than muscle, better reflects rejection than IRI in this study. Therefore, the analysis in this study has focused more on the skin components.

The rationale behind the potential benefits of NMP touches multiple facets. First, NMP reduces IRI by maintaining a physiological environment,^38,39 as shown by stable perfusion parameters and reduced lactate and potassium. Furthermore, weight gain at the end of perfusion, an indicator of poor performance,^16,25,40-42 is limited. Second, a consequence of continuous flow and the use of acellular perfusate is that immune cells are washed out of the graft.⁴³ Third, perfusate contains hydrocortisone and dexamethasone, which are immunosuppressants and are routinely included in standard NMP protocols, as reported in several clinical and preclinical perfusion studies.^44-46 While it is the recipient who creates an immune response, not the donor graft, Cao et al⁴⁷ showed delayed rejection in murine VCA that were administered FK506-loaded nanoparticles. Therefore, it is possible that perfusing the graft with immunosuppressants provides temporary alleviation and delays the onset of an immunological response to the transplanted graft. While this study was not designed to isolate individual perfusate components to mechanistically dissect the contribution of corticosteroids, post-transplant plasma hydrocortisone levels showed no significant differences between NMP and SCS recipients and declined rapidly after transplantation, arguing against a sustained systemic steroid effect as a major contributor to the observed outcomes.

IRI increases proinflammatory cytokine production, infiltration of inflammatory cells, and oxidative stress,^48,49 causing tissue damage. The crosstalk between these signaling pathways is crucial to the regulation of oxidative stress responses, extracellular matrix remodeling, angiogenesis, inflammation, fibrosis, and apoptosis. As a result, several proinflammatory genes are upregulated, leading to overexpression of cytokines such as IL-6, IL-1α, IL-1β, TNF-α, and interferon-gamma.⁵⁰ TNF-α is a well-known trigger of regulated cell death pathways, activating both caspase-dependent apoptosis and caspase-independent necroptosis.⁵¹ Tissue-resident macrophages and other cells of nonhematopoietic origin present in donor transplanted organs, such as fibroblasts and endothelial cells from the vasculature, can elicit inflammation by priming adaptive immune responses, resulting in cytokine storm and T cell proliferation.⁵² In the context of VCA, skin is known to be a critical regulator of the connectivity and dynamics of inflammatory networks.⁵³ Studies have identified IL-1α, IL-18, and leptin as key central hubs in the skin’s inflammatory network in VCA.⁵⁴ Meanwhile, vascular endothelial growth factor, interferon-gamma, and monocyte chemmoattractant protein-1 are identified as key regulators of inflammatory cascades in muscle tissues.⁵⁴

Our findings revealed that key proinflammatory cytokines and chemokines were overall higher in the SCS group, particularly at later PODs, supporting the notion of a more intensified inflammatory response after static preservation. In contrast, the NMP group showed a transient and more balanced cytokine profile at early time points, characterized by concomitant increases in both pro- and anti-inflammatory mediators. This controlled activation may reflect the effects of ex vivo perfusion, during which donor-derived immune cells are largely removed,^43,55,56 thereby limiting the priming of adaptive immune responses after transplantation. The early upregulation of IL-10 in NMP-preserved grafts further supports an immunoregulatory shift, consistent with reduced antigen-presenting cell activation and an attenuated downstream inflammatory cascade.^57,58 NMP also exhibited a significantly lower B cell population, possibly implying a less robust antibody-mediated immune response within the NMP-preserved allografts. Elevated CD4⁺ effector T cells over CD8⁺ cytotoxic T cells in both groups suggest that NMP did not substantially alter T cell differentiation pathways. However, this does not rule out the possibility that NMP could still influence the overall dynamics and connectivity of the immune response in rejection settings. As there are no current studies comparing rejection responses in a full-mismatch VCA model, future research could provide deeper insights into the specific mechanisms involved. Understanding these mechanisms may help refine strategies for improving graft acceptance and minimizing rejection in such models.

The findings of this study suggest that NMP does not worsen early graft outcomes in VCA and is associated with a delayed onset of acute rejection relative to SCS. These results support the hypothesis that machine perfusion does not exacerbate the early stages of acute rejection in addition to its known positive effects on limiting IRI.^{16,23,25,41,59} However, the long-term benefits remain unclear, and further studies are needed to explore the sustained impact of NMP on VCA rejection, if any, and to optimize perfusion protocols. While data suggest there is a benefit to NMP in terms of IRI and rejection, it is not a cure-all; its effects seem significant but not transformative.

Limitations of this study include the number of replicates with daily histology samples, even if these samples were assessed by multiple blinded pathologists. This consideration was made because obtaining biopsies can elicit an immune response in itself. Further limitations include the performance of the flow cytometry analysis and histology assessment at the stage of severe rejection (POD 7), and the use of a small, and specifically rodent, animal model.⁶⁰

Future studies should examine flow cytometry around POD 3 or use a slower rejection model by transplanting between partial major histocompatibility complex mismatch or by using (subtherapeutic) immunosuppression. Additional immunological markers could be included, and conducting long-term follow-up would serve to better understand the chronic rejection processes. Exploring the genetic and molecular mechanisms underlying observed differences in cytokine profiles and immune cell populations will provide deeper insights into the benefits of machine perfusion in VCA, and upscaling to large animal models or trials on human VCA could be considered. Moreover, in the future, machine perfusion can serve as a platform for further function improvement, for example, by therapeutic treatment,⁶¹ immunomodulation, or by genetic modification^27,62-64 to attenuate rejection, independent of or in addition to systemic immune tolerance induction treatments.⁶⁵ To extend preservation time, machine perfusion has enabled cryopreservation strategies, significantly extending preservation times of VCA,^42,66,67 livers,^68,69 and kidneys^70,71 from hours to days and beyond in experimental setting. Even as a technology in itself, machine perfusion has been shown to extend preservation time.^72,73 This study contributes to the ongoing efforts to improve graft outcomes and patient survival in transplant surgery, with implications that extend beyond VCA to broader transplant methodologies.

ACKNOWLEDGMENTS

The authors thank MGH Histology Core, especially Dr Ivy Rosales, for their work in providing us with stained histology slides and blinded pathology assessment. The authors further thank the MGH Department of Photography for taking the pictures, displayed in Figure 1. The Shriners Children’s Genomics and Proteomics (84090-BOS-21), Morphology & Image Analysis (84050-BOS-21), and Translational Regenerative Medicine (84051-BOS-21) Cores are also gratefully acknowledged.