Journal Logo

Original Basic Science—General

Hypothermic Ex Situ Perfusion of Human Limbs With Acellular Solution for 24 Hours

Haug, Valentin MD1,2; Kollar, Branislav MD1; Tasigiorgos, Sotirios MD1; Endo, Yori MD1; Kauke, Martin MD1; Safi, Ali-Farid MD, DDS1; Veeramani, Anamika BS1; Abdulrazzak, Obada1; Bausk, Bruce BS3; Walt, David PhD3; Pomahac, Bohdan MD1

Author Information
doi: 10.1097/TP.0000000000003221



As soon as a limb is detached from blood supply, the dogma “time is tissue” holds true. Preventing ischemia and its devastating consequences to the tissue, such as ischemia–reperfusion injury (IRI), remains a challenge. For decades, static cold storage (SCS) has been the gold-standard method for delaying ischemic tissue damage and reducing the metabolic demand following traumatic limb amputation, or surgical procurement of the limb for subsequent transplantation.1-3 For every 10°C reduction in temperature of the tissue, metabolic demand decreases by half.4 For vascularized composite allografts (VCAs), the allowable ischemia time is just 4–6 h, even under SCS.5,6 As skeletal muscle is the predominant tissue in the limb and the tissue type most sensitive to ischemic damage,7 its susceptibility to IRI in particular presents a significant obstacle to the functional success of limb transplantation.8 In solid organ transplantation, the concept of machine perfusion (MP) has emerged as a favorable alternative to SCS for ex situ preservation.9-11 Small and large animal studies confirmed the potential of MP in the ex situ preservation of VCA.12-14 The outcomes after replantation of ex situ preserved limbs seem in favor of MP versus SCS in those experimental studies.15,16 Presently, only 1 preclinical study has shown an increased extracorporeal preservation time in human arms.7 Recently, 2 traumatically amputated lower limbs were replanted after 12 and 15 h of hypothermic ex situ perfusion17; however, blood-based perfusates were used in both cases.7,12,13 Blood-based perfusates are not ideal for MP, as blood products can serve as vectors for diseases, may lead to transfusion reactions, and promote HLA sensitization.18-21 They are associated with logistical issues tied to blood bank and require refrigeration for transport and storage. Additionally, they have a relatively short shelf-life of 21–49 d.20,21 In contrast, acellular perfusates are readily available, significantly more stable and can be used in austere environments.22 The aim of this study was to evaluate the effects of MP with a colloid-enriched acellular perfusate in prolonging the viability of amputated human limbs and compare the effects of MP with that of SCS.


Study Approval

In collaboration with the New England Donor Services, this preclinical study was approved by the Institutional Review Board (Protocol number: 2016P0013320). After obtaining explicit consent, 6 upper extremities were procured from 3 adult brain-dead donors. One extremity of a donor was either assigned to hypothermic ex vivo MP (n = 3), or control group SCS (n = 3) for 24 h.

Procurements of the Limbs

The procurement surgeries were performed under sterile conditions by the senior author and 2 assistants. Thirty thousand units of heparin were administered and circulated for 3 min before placement of the cannulas. After aortic cross-clamping, 3–4 L of the UW-Belzer solution are run in aorta and 2 L in the portal vein. After procurement of the heart, liver, and kidneys (marking the start of warm ischemia time), procurement of the arms was initiated. Control blood samples were obtained before aortic cross-clamping. Control muscle biopsies were obtained before amputation. Transhumeral amputation was completed after identifying and labeling the brachial artery. Both limbs from each patient were immediately placed in a sterile bag with the UW-Belzer solution, then placed into a second bag filled with ice slurry (the start of cold ischemia time) and were transferred to the laboratory at Brigham and Women’s Hospital. One of the limbs remained sterile bagged in ice slurry for 24 h to serve as a SCS control. Direct contact with ice was avoided. The contralateral limb was connected to the perfusion apparatus (end of ischemia time). Both limbs were weighed before and after the experiment.

The Perfusion Setup

The perfusion device circuit (Figure 1) starts with a 4-L reservoir containing the perfusate. A peristaltic machine pump (Master Flex Pump L/S, Cole-Parmer, IL) generates a continuous flow through membrane filters (Critical Process Inc., NH). The perfusate then moves toward a heat exchanger with a cooling circuit (Hyper 212 EVO, Cooling Master, CA), where it is cooled to 8°C. A membrane oxygenator (Dideco Kids-D100 Neonatal Oxygenator, Sorin Group, Mirandola, Italy), connected to an oxygen tank, saturates the perfusate to a target Po2 of >500 mm Hg. The oxygenated perfusate is finally passed through the limb via an arterial catheter (Fem-Flex II Femoral Arterial Cannula, 4.7-mm diameter, Edwards Lifesciences LLC, CA). A bubble trap prevents free gas bubbles from entering the limb and obstructing the microvasculature. After passing through the limb, the solution drains via the open venous outflow into a collection tray, from which it is fed back into the perfusate reservoir by gravity thus completing the circuit. The circuit is equipped with various instrumentation to enable real-time monitoring of pressure (Pendotech Press-S-0000, Cole-Parmer, IL), Po2 (Dissolved Oxygen Probe, Atlas Scientific, NY), and perfusate temperature (Pendotech Tempc-N-999, Cole-Parmer, IL). Flow rates are regulated by a custom-built dynamic feedback algorithm (MIT, Edelman Lab) to maintain the desired arterial pressure of 30 mm Hg. Pressure, temperature, and oxygen partial pressure are measured constantly and recorded every minute using LabView 2016 (National Instruments, Austin, TX).

Bench-top perfusion device. Illustration by Medical Illustrator Kai-Ou Tang. Printed with permission and copyrights retained by Brian D. Polizzotti, PhD.

Perfusion Solution

For the perfusion experiments, we used acellular Steen solution (Xvivo Perfusion AB, Göteborg, Sweden). Steen solution was supplemented with 0.1% of 50% Dextrose (Hospira Inc., IL) to provide an energy supply, 0.0075% Insulin R (Lilly USA, LLC, Indianapolis, IN) to facilitate the intracellular uptake of the dextrose, and 125 mg/L methylprednisolone (Fresenius Kabi, IL) to reduce capillary leakage and edema.23 Additionally, 2500 units/L of heparin were added (Fresenius Kabi, IL) to the perfusion solution for the first hour only. The perfusate, which serves as a sink for the markers of muscle injury, was renewed after 1, 6, 12, and 18 h of perfusion to maintain a sufficient energy supply (Table 1). After mixing the Steen solution with the additives, osmotic pressure was assessed with a vapor pressure osmometer (Wescor Vapro, UT). Osmotic pressure averaged around ±307 mOsmol/kg.

Sample collection and perfusate exchange

Sample Collection

At the start of the experiment, afferent and efferent samples were obtained from the perfusate and analyzed at 0-, 1-, 2-, and 4-h timepoints, then every 2 h until the end of the experiment after 24 h of perfusion (Table 1). The perfusate samples were kept on ice until analysis for a maximum of 1 h. Samples for blood gas analysis were collected in a heparin-coated syringe. Samples for lactate were collected in a sodium fluoride/potassium oxalate Vacutainer (Becton, Dickinson and Co., NJ), samples for myoglobin, CK and lactate dehydrogenase (LDH) were collected in lithium heparin vacutainers (Becton, Dickinson and Co., NJ).

Muscle biopsies were obtained before the amputation then at 0-, 4-, 12-, and 24-h timepoints (Table 1). The biopsies were procured from the superficial flexor muscle group of the forearm. The biopsied samples were then snap-frozen in liquid nitrogen and stored at −80°C or put in RNAlater solution (Thermo Fisher Scientific, MA) and stored at −20°C. Samples for the histopathological workup were immersed in formalin.

Perfusate Analysis

All the perfusate samples were measured for Po2 and Pco2, pH, potassium (K+), CK, LDH, myoglobin, and lactate at the aforementioned timepoints, no later than 1 h after collection. Blood gas analysis was performed with 837 Flex Radiometer (Radiometer Inc., CA), myoglobin analysis with Cobas e602, lactate analysis with Cobas c702, and LDH and CK with Cobas c502 (Roche Diagnostics, IL).


The muscle biopsy samples were stored in formalin for 48 h, then in 70% ethanol and subsequently embedded in paraffin. The samples were then sectioned to 6-µm slides and stained with hematoxylin and eosin to evaluate the tissue architecture. To visualize the glycogen content of the skeletal muscle cells, the sections were stained with periodic acid-Schiff staining (PAS).

Histologic Analysis

Sections were visualized using a light microscope (Nikon Eclipse E400, Tokyo, Japan) and images were captured using a camera (Nikon DS-Fil, Tokyo, Japan). From each slide, 5 digital pictures from randomly selected fields were obtained at a 10-fold magnification using the identical camera settings. Photographs were deidentified and evaluated for intercellular distance, tissue architecture and glycogen by 2 reviewers independently (V.H. and A.V.). The optical density of glycogen-positive cells was calculated with ImageJ software (v.1.51).24,25

Quantification of Ischemia

Skeletal muscle biopsies were snap-frozen in liquid nitrogen and stored at −80°C until analysis. Western blot analysis was performed using 20 μg of whole lysate extracts. Hypoxia-inducible factor 1-alpha (HIF-1α) was probed using anti-HIF-1α antibody (Ab51608, Abcam, CA) diluted 1:2000 in 4% bovine serum albumin, followed by horseradish peroxidase-conjugated antirabbit secondary antibody (7074, Cell Signaling, MA). The antigen-antibody complexes were then visualized with enhanced chemiluminescence (Thermo Fisher Scientific, MA). The intensities of HIF-1α signals were quantified and plotted for each patient.

Cytokine Analysis via Digital ELISA

Perfusate was analyzed for cytokine levels with digital ELISA (Simoa) assays using the Simoa HD-1 Analyzer (Quanterix, Lexington, MA). Samples were analyzed with 4 multiplex (3-plex) panels, yielding a total of 12 unique cytokines measurements for each sample (Table S1, SDC, In depth, descriptions of the Simoa analysis have been previously reported.26,27 Briefly, immune complexes were formed on beads in a 3-steps format by capturing the target, binding with biotinylated detection antibody, and labeling with streptavidin-β-galactosidase. Beads were then suspended in enzyme substrate, loaded into microarrays of femtoliter wells, and imaged. One bead is loaded per well, and only those beads containing enzyme-labeled immune complexes (“on” beads) produce a fluorescent signal. Signal is reported as Average Enzymes per Bead.

Statistical Analysis

For the comparison between 2 groups, unpaired t-test was performed to detect statistically significant differences. For the comparison between 2 groups at different timepoints, either 2-way ANOVA or mixed-effect analysis was performed. In the case of multiple comparisons, P values were adjusted via Tukey correction. A P value of <0.05 was considered statistically significant. Results are shown as the median with error (range). All the statistical analyses and visualization of the results were performed using GraphPad Prism version 8.00 for MacOS (GraphPad Software, La Jolla, CA).



In total, 3 pairs of upper extremities were procured from brain-dead donors between July 2017 and July 2019 in the Boston area. Two donors were male, and 1 donor was female. The median age was 28 y (range: 24–51 y) and the median BMI was 28.9 kg/m2 (range: 22.3–29.1 kg/m2).

The median total ischemia time was 213 min (range: 127–222 min). The median warm ischemia time was 90 min (range: 65–155 min), and that of cold ischemia was 67 min (range: 37–148 min). After the initial incision, the procurement took 8 min per limb, at maximum. The limbs were immediately bagged and immersed in ice slurry. The limbs in the SCS group (n = 3) remained in ice slurry for the duration of the experiment (24 h). After transfer to the laboratory, the limbs in the MP group (n = 3) were connected to the already-prepared bench-top perfusion device via cannulation of the brachial artery and perfused for 24 h. Figure S1, SDC,, shows the appearance of the human limbs after 24 h of MP and SCS.

Machine Perfusion

The machine data of the perfusion are shown in Figure 2A. Perfusion pressure was maintained at a median of 30.4 mm Hg (range, 27.3–35.5 mm Hg) (Figure 2A). The temperature of the perfusate was maintained at a median of 9.43°C (range, 4.8–14.3°C) (Figure 2A). The circuit flow was automatically adjusted to maintain a constant target pressure of 30 mm Hg, with a median of 30.4 mL/min (range, 10.0–62.5 mL/min) (Figure 2A). Within the first 5 h of the perfusion, a gradual increase in the circuit flow was observed, then a decrease over the final 6 h of the experiment. Po2 of the perfusate ranged from 385.4 to 609.7 mm Hg with a median of 555.8 mm Hg (Figure 2A). The limb weight increased by 1.4% in the SCS group and 4.3% in the perfusion group over 24 h (Figure 2A), neither of which were found statistically significant.

Machine data (A), results of the clinical chemistry (B), and blood gas analysis (C). Median is depicted as plotted data points; the range is depicted as error bars.

Perfusate Analysis

A clinical chemistry analysis of the perfusate showed an increase in biomarkers of muscle injury such as myoglobin, CK, and LDH (Figure 2B). Myoglobin, CK, and LDH showed increased levels after ischemia before starting the perfusion, peaking after 6 h, and dropping after exchanging the perfusion solution. The LDH level then plateaued, whereas the myoglobin and CK levels (Figure 2B) further increased toward the end of the perfusion. The lactate levels were elevated up to a median of 6.9 mmol/L (range, 5.19–7.71 mmol/L), at the beginning of the MP but decreased to a median of 2.8 mmol/L (range, 1.66–3.9 mmol/L) after 24 h (Figure 2B).

Blood gas analysis of the “venous” postlimb efflux showed a gradual increase in pH from a median of 6.90 (range, 6.76–7.26) to 7.24 (range, 7.16–7.4), however, remaining in acidosis throughout (Figure 2C). A similar trend was observed in the potassium levels as that of the lactate levels, which rose to a median of 9.6 mmol/L (range, 9.48–106 mmol/L) with ischemia time, then descended to a median of 5.77 mmol/L (range, 5.6–5.8 mmol/L) after starting the perfusion over time (Figure 2C). Venous partial carbon dioxide pressure (Pco2) after ischemia was elevated to a median of 77.7 mm Hg (range, 25.8–103.1 mm Hg), leveling out to a median of 24.9 mm Hg (range, 17.4–26.8 mm Hg), until the end of the experiment (Figure 2C). The median partial oxygen pressure (Po2) indicating the degree of oxygenation of the acellular perfusate was 800 mm Hg (range, 283–800 mm Hg) at inflow, whereas the “venous” Po2 was 478.3 mm Hg (range, 202–784 mm Hg) (Figure 2C). The venous glucose levels remained stable at around 231.5 mg/dL (range, 190–264 mg/dL) throughout the duration of perfusion, differing from the median “arterial” glucose levels with a median of 232.0 mg/dL (range, 213–270 mg/dL) for 1.5 mg/dL (Figure 2C).

Histological Analysis

The mean intercellular distance between myocytes in the hematoxylin and eosin slides showed a significant increase in both the SCS and MP groups after 24 h compared to the baseline control samples. A notable difference between the 2 groups was observed at the 12-h timepoint, using mixed-effect analysis with Tukey correction (Figure 3).

Intercellular distance and HE staining of skeletal muscle biopsies. CTRL, control; HE, hematoxylin and eosin; MP, machine perfusion; SCS, static cold storage.

A histological analysis of the PAS-stained slides showed significantly greater number of PAS-positive cells in the perfusion group after 4-h timepoint compared with the SCS group (P = 0.0022), similar to the control samples taken before the ischemic events (Figure 4). At 12- and 24-h timepoints, no substantial differences between the 2 groups were observed, using 2-way ANOVA with the Tukey post hoc correction.

Results of intracellular glycogen and PAS staining of skeletal muscle samples after 4 h of MP or SCS. CTRL, control; MP, machine perfusion; PAS, periodic acid-Schiff staining; SCS, static cold storage.

Protein Expression Analysis

Protein expression analysis was conducted for the second and third pair of the donated limbs. Snap-frozen skeletal muscle biopsies from the first donor showed signs of protein degradation, rendering them unsuitable for a meaningful protein analysis.

Human limb (HL) 2 underwent warm ischemia for 65 min and cold ischemia for 148 min (total ischemia time: 213 min). The expression of HIF-1α in the SCS-limb increased after amputation and peaked at the 8-h timepoint (Figure 5). Subsequently, a gradual decrease in the HIF-1α expression to a minimum expression toward the end of 24 h was noted. The expression of HIF-1α in the perfused limb decreased over the first 8 h of the perfusion. The biopsy at 12-h timepoint showed an initial increase in the HIF-1α expression, however, to a lesser degree than that of the SCS-preserved limb at the equivalent timepoint. A further increase in the expression was observed in the 16- and 20-h biopsies, before its fall at 24-h timepoint.

Results of the protein expression of HL2 and HL3 including densitometry. Ctrl, control; HIF, hypoxia-inducible factor; HL, human limb; MP, machine perfusion; SCS, static cold storage.

HL3 underwent warm ischemia for 90 min followed by cold ischemia lasting 37 min (total ischemia time: 127 min). In the SCS-preserved limb, the expression of HIF-1α peaked after 127 min of ischemia, which corresponded to the 0-h timepoint of biopsy. The HIF-1α expression then decreased sequentially in the 4-, 12-, and 24-h biopsies. Expectedly, the perfused limb also showed a peak in the expression of HIF-1α at the 0-h timepoint. The subsequent HIF-1α expression in the 4-, 12-, and 24-h biopsies was lower than that of the SCS-preserved limb.

Cytokine Analysis

Cytokine concentration was measured in perfusate samples using digital ELISA (Simoa) assays. Hourly samples were taken at hour 1, hours 2–6, and hours 18–24. Selected data from HL2 and HL3 are shown in Figure 6. Ultrasensitive digital ELISAs allowed for the detection and quantification of cytokines in the highly dilute perfusate. There was a general trend of increasing cytokine levels from 2 to 6 h and from 19 to 24 h, each after the change of the perfusate solution. However, a drastic increase was not seen in either the absolute concentration of these cytokines, or the rate of concentration increase in the last circulating perfusate as compared to the first circulating perfusate. Additional results of the cytokine analysis are shown in Figure S2, SDC,

Selected date of the cytokine analysis. The dotted line parallel to the x-axis depicts the limit of detection (LOD). HL, human limb.


Using hypothermic ex situ perfusion, we preserved 3 human upper extremities with an acellular, colloid-enriched, oxygenated perfusate for 24 h.

In comparison, the current “gold standard,” SCS only allows for an ex situ preservation of VCA for a maximum of 4–6 h before replantation. We have shown that hypothermic MP enables for a 4-fold increase in the viable preservation time, not necessitating blood products.28 VCA recipients would greatly benefit from this extended ex situ preservation time, as it would contribute to an enlarged donor pool.29 When donor and recipient are not located in the same hospital, a prolonged ischemia time is inevitable. Subsequently, functional impairments of the transplanted limb, as with the intrinsic muscles of the hand—for example, are vulnerable to the ischemic damage. Experimental studies in rodents showed a ischemia time-dependent, however not significant, decrease of muscle function after conservation in terms of contractility.30-32 Similarly, it was found in a clinical study of hand transplantation that the allografts that did not show good function after transplantation had undergone longer periods of ischemia.33 Furthermore, studies in pigs showed that the rates of acute rejections of VCA were correlated with ischemia time, but could be reduced by subnormothermic and normothermic perfusion without immunosuppressive therapy.34,35 These results indicate that ex situ perfusion could not only improve functional preservation of the VCAs by a sufficient oxygenation of the VCA during procurement and transplantation but might also delay the onset or decrease the frequency of acute rejections, which occurs in up to 80% of the cases within the first year after transplantation.36

Another potential benefiter of ex citu perfusion might be victims of traumatic major limb amputation (TLA). This technology enables for stabilization of the patient and viability maintenance of the amputated limb until the patient is ready for surgery. Taeger et al17 demonstrated successful replantation of 2 lower extremities after ex citu perfusion. Extending the time window before replantation can make a substantial change in the therapeutic approach of patients with TLA. The paradigm “life before limb” could change to “life first, limb second,” salvaging the limbs as well as life. If replantation is not possible, at least, spare-part surgery could be facilitated by this technique.37

The average ischemic times of the upper extremities were substantially longer in our donors (187.3 ± 52.4 min), because of the geographical spread of donors, when comparing it to the limbs Werner et al7 used for their study (mean of 76 min) and the porcine limbs Krezdorn et al28 used in their replantation experiments. Furthermore, the donors of our study were younger with lower BMIs (mean 34.3 y; 26.8 kg/m2).

To maintain a higher degree of comparability and based on the preliminary data from previous limb perfusion studies in large animals by Krezdorn et al,28 we maintained the mean arterial pressure at an average of 30.88 mm Hg also with a continuous flow to minimize the sheer stress-induced endothelial damage and edema caused by nonphysiological pressure and high flow rate of MP.14,15 Pressure and temperature were comparable to the results of Krezdorn et al28 (pressure 29.4 ± 0.6 mm Hg; temperature 8.2 ± 0.7°C). The perfusion solution was cooled and maintained at a mean of 9.8°C, thus reducing the limbs’ metabolic demand, whereas normothermic perfusion, on the other hand, allows the assessment of muscle function.7,13,38 Therefore, in extracorporeal liver perfusion, organ function in marginal donors can be evaluated under normothermic conditions.9,10,39,40 However, normothermic perfusion necessitates either blood-based perfusion solutions, as used in the study of Werner et al,7 or artificial oxygen carriers, as the tissue’s metabolic demand is not decreased. Because of the disadvantages of blood products, such as short shelf-life, the need for refrigeration, HLA sensitization, and risk of disease transmission, we chose a readily available commercial perfusate.18-21 Interestingly, a study in large animals has reported better outcomes and accumulation of less toxic metabolites when perfused with acellular solutions compared to blood.41 However, the most obvious advantage that blood as a perfusion solution offers over acellular perfusates is the oxygen carrying capacity of hemoglobin.

Analyzing the perfusate, a fall in the lactate level was observed in the MP group from levels of 6.6 ± 1.29 mmol/L at the start to levels of 2.79 ± 1.1 mmol/L at the end of perfusion, indicating that anaerobic metabolism of the tissue was alleviated over time. Lactate levels in the study by Werner et al7 were lower, when starting the experiment, most likely due to the shorter ischemic time (mean 76 min). Interestingly, other studies of normothermic or subnormothermic perfusion of human or porcine limbs showed the opposite trend over time with an average of 12 mmol/L after 11 or 12 h or 15 mmol/L after 24 h.7,13,42 Potential reasons for that could be a smaller circuit volume (Werner et al7: 300 mL) and different timepoints for the perfusate exchange (Werner et al7: every 3–5 h). Biomarkers of muscle injury, such as myoglobin, showed similar trends and levels compared to the human limbs perfused with a normothermic, blood-based perfusate by Werner et al.7

Histological analysis of the skeletal muscle biopsies revealed no significant tissue damage, necrosis, or pathological signs of edema in either groups. Changes in the tissue architecture would be expected after replantation of the limb with an infiltration of leukocytes to happen. This is in agreement with other ex situ perfusion studies.7,13

Protein expression analysis showed a delayed peak in the HIF-1α expression in the MP group, when compared with that of the SCS-group. These results suggest prolonged viability and improved nutritional and oxygen support for the limb over the duration of 24 h. When compared with the SCS group, ex vivo preservation time was extended by 4- to 6-folds in the MP group. This is consistent with the findings of the extracorporeal perfusion trial by Werner et al7, which reported viability after 24 h of perfusion with a blood-based perfusate. Similarly, Kueckelhaus et al14 found significantly higher levels of HIF-1α expression after replantation in the SCS-preserved limbs compared to the machine-perfused limbs in pigs, indicating less ischemic damage in the MP group. One could speculate that similar outcomes regarding damages in tissue architecture could be expected, when replanting the limbs.

The ultrasensitive Simoa assays could provide new insights into the expiration of human limbs after amputation, which could not be studied using traditional protein quantification methods. While we were largely able to measure these inflammatory markers, we did not observe a relative increase in cytokine concentration in the final perfusion (h 19–24) versus the initial perfusion (h 2–6). We therefore speculate that the observed increases in the perfusate cytokine concentration were due to cytokine accumulation over the course of perfusion, rather than due to an inflammatory response upon limb expiration. Extending the analysis of the perfusion solution after 24 h or after replantation could provide further insight into the mechanism of limb expiration associated with an inflammatory response.

Although this study brings new insights into the ex situ perfusion of human VCA, the findings should be interpreted in the context of the study’s limitations like the small sample size. This nonreplantation model does not allow the assessment of IRI and its consequences on short- and long-term muscle function. To circumvent this problem, future studies could expand the experiments by a subsequent perfusion of full blood to simulate IRI ex situ. Another potential improvement, when replantation is not an option, might be the use of a spectrometer, to assess mitochondrial function and myoglobin saturation of the skeletal muscle.43 Additionally, analysis of insulin-receptors could be beneficial to assess energy metabolism on a cellular level.44 For replantation studies, strict sterile conditions have to be maintained during procurement and ex situ perfusion, which challenges this technology. To draw conclusions regarding the optimal perfusion solution and settings, a larger sample size would allow for a direct comparison between blood-based and multiple acellular perfusates, such as UW solution, and SCS with a subsequent reperfusion with full blood. Future clinical studies may aim to compare the functional long-term results of each method within the same recipient in the case of bilateral arm transplantation. Last, as the major veins were not cannulated, the efflux of the perfusate was in contact with the atmosphere, which prevented a meaningful calculation of the oxygen consumption of the limb.

Hypothermic ex situ perfusion with an oxygenated, colloid-enriched acellular solution may extend the allowable extracorporeal preservation time at least 4-fold compared to SCS. This technology is a promising development for improved outcomes in VCA recipients and victims of TLA.


The authors want to express deep gratitude to the donors and their families without whom this research would not have been possible. In addition, the authors thank the New England Organ Bank for generously providing funding to support our experiments. They also thank Elazar Edelman and Fiona MacLeod for their assistance with designing the of our perfusion device. They also thank Petr Jarolim MD, PhD and his team of laboratory technicians for analyzing the perfusate samples. Last but not least, the authors thank the OR teams in the donor hospitals for their cooperation.


1. Chaui-Berlinck JG, Monteiro LH, Navas CA, et al. Temperature effects on energy metabolism: a dynamic system analysis. Proc Biol Sci. 2002; 269:15–19. doi:10.1098/rspb.2001.1845
2. Harris K, Walker PM, Mickle DA, et al. Metabolic response of skeletal muscle to ischemia. Am J Physiol. 1986; 2502 Pt 2H213–H220. doi:10.1152/ajpheart.1986.250.2.H213
3. Krezdorn N, Tasigiorgos S, Wo L, et al. Tissue conservation for transplantation. Innov Surg Sci. 2017; 2:171–187. doi:10.1515/iss-2017-0010
4. Amin KR, Wong JKF, Fildes JE. Strategies to reduce ischemia reperfusion injury in vascularized composite allotransplantation of the limb. J Hand Surg Am. 2017; 42:1019–1024. doi:10.1016/j.jhsa.2017.09.013
5. Müller S, Constantinescu MA, Kiermeir DM, et al. Ischemia/reperfusion injury of porcine limbs after extracorporeal perfusion. J Surg Res. 2013; 181:170–182. doi:10.1016/j.jss.2012.05.088
6. Messner F, Grahammer J, Hautz T, et al. Ischemia/reperfusion injury in vascularized tissue allotransplantation: tissue damage and clinical relevance. Curr Opin Organ Transplant. 2016; 21:503–509. doi:10.1097/MOT.0000000000000343
7. Werner NL, Alghanem F, Rakestraw SL, et al. Ex situ perfusion of human limb allografts for 24 hours. Transplantation. 2017; 101:e68–e74. doi:10.1097/TP.0000000000001500
8. Blaisdell FW. The pathophysiology of skeletal muscle ischemia and the reperfusion syndrome: a review. Cardiovasc Surg. 2002; 10:620–630. doi:10.1016/s0967-2109(02)00070-4
9. Dutkowski P, Schlegel A, de Oliveira M, et al. HOPE for human liver grafts obtained from donors after cardiac death. J Hepatol. 2014; 60:765–772. doi:10.1016/j.jhep.2013.11.023
10. Brockmann J, Reddy S, Coussios C, et al. Normothermic perfusion: a new paradigm for organ preservation. Ann Surg. 2009; 250:1–6. doi:10.1097/SLA.0b013e3181a63c10
11. Imber CJ, St Peter SD, Lopez de Cenarruzabeitia I, et al. Advantages of normothermic perfusion over cold storage in liver preservation. Transplantation. 2002; 73:701–709. doi:10.1097/00007890-200203150-00008
12. Araki J, Sakai H, Takeuchi D, et al. Normothermic preservation of the rat hind limb with artificial oxygen-carrying hemoglobin vesicles. Transplantation. 2015; 99:687–692. doi:10.1097/TP.0000000000000528
13. Duraes EFR, Madajka M, Frautschi R, et al. Developing a protocol for normothermic ex-situ limb perfusion. Microsurgery. 2018; 38:185–194. doi:10.1002/micr.30252
14. Kueckelhaus M, Fischer S, Sisk G, et al. A mobile extracorporeal extremity salvage system for replantation and transplantation. Ann Plast Surg. 2016; 76:355–360. doi:10.1097/SAP.0000000000000681
15. Kueckelhaus M, Dermietzel A, Alhefzi M, et al. Acellular hypothermic extracorporeal perfusion extends allowable ischemia time in a porcine whole limb replantation model. Plast Reconstr Surg. 2017; 139:922e–932e. doi:10.1097/PRS.0000000000003208
16. Gok E, Alghanem F, Moon R, et al. Development of an ex-situ limb perfusion system for a rodent model. Asaio J. 2019; 65:167–172. doi:10.1097/MAT.0000000000000786
17. Taeger CD, Lamby P, Dolderer J, et al. Extracorporeal perfusion for salvage of major amputates. Ann Surg. 2019; 270:e5–e6. doi:10.1097/SLA.0000000000003226
18. Schreiber GB, Busch MP, Kleinman SH, et al. The risk of transfusion-transmitted viral infections. The Retrovirus Epidemiology Donor Study. N Engl J Med. 1996; 334:1685–1690. doi:10.1056/NEJM199606273342601
19. Rees L, Kim JJ. HLA sensitisation: can it be prevented? Pediatr Nephrol. 2015; 30:577–587. doi:10.1007/s00467-014-2868-6
20. Knowles S. Blood transfusion: challenges and limitations Transfus Altern Transfus Med. 2007; 9:2–9. doi:10.1111/j.1778-428X.2007.00062.x
21. Ng MSY, David M, Middelburg RA, et al. Transfusion of packed red blood cells at the end of shelf life is associated with increased risk of mortality– a pooled patient data analysis of 16 observational trials. Haematologica. 2018; 103:1542–1548. doi:10.3324/haematol.2018.191932
22. Becker S, Steinmeyer J, Avsar M, et al. Evaluating acellular versus cellular perfusate composition during prolonged ex vivo lung perfusion after initial cold ischaemia for 24 hours. Transpl Int. 2016; 29:88–97. doi:10.1111/tri.12649
23. Paulus P, Holfeld J, Urbschat A, et al. Prednisolone as preservation additive prevents from ischemia reperfusion injury in a rat model of orthotopic lung transplantation. PLoS One. 2013; 8:e73298. doi:10.1371/journal.pone.0073298
24. Schneider CA, Rasband WS, Eliceiri KW. NIH image to imagej: 25 years of image analysis. Nat Methods. 2012; 9:671–675. doi:10.1038/nmeth.2089
25. Inagi K, Connor NP, Ford CN, et al. Physiologic assessment of botulinum toxin effects in the rat larynx. Laryngoscope. 1998; 108:1048–1054. doi:10.1097/00005537-199807000-00018
26. Wilson DH, Rissin DM, Kan CW, et al. The Simoa HD-1 analyzer: a novel fully automated digital immunoassay analyzer with single-molecule sensitivity and multiplexing. J Lab Autom. 2016; 21:533–547. doi:10.1177/2211068215589580
27. Wu D, Dinh TL, Bausk BP, et al. Long-term measurements of human inflammatory cytokines reveal complex baseline variations between individuals. Am J Pathol. 2017; 187:2620–2626. doi:10.1016/j.ajpath.2017.08.007
28. Krezdorn N, Macleod F, Tasigiorgos S, et al. Twenty-four-hour ex vivo perfusion with acellular solution enables successful replantation of porcine forelimbs. Plast Reconstr Surg. 2019; 144:608e–618e. doi:10.1097/PRS.0000000000006084
29. Ben-Amotz O, Kruger EA, McAndrew C, et al. Logistics in coordinating the first adult transatlantic bilateral hand transplant: lessons learned. Plast Reconstr Surg. 2018; 142:730–735. doi:10.1097/PRS.0000000000004672
30. de With MC, van der Heijden EP, van Oosterhout MF, et al. Contractile and morphological properties of hamster retractor muscle following 16 h of cold preservation. Cryobiology. 2009; 59:308–316. doi:10.1016/j.cryobiol.2009.08.008
31. Hautz T, Hickethier T, Blumer MJ, et al. Histomorphometric evaluation of ischemia-reperfusion injury and the effect of preservation solutions histidine-tryptophan-ketoglutarate and University of Wisconsin in limb transplantation. Transplantation. 2014; 98:713–720. doi:10.1097/TP.0000000000000300
32. Lovic A, Landin L, Diez J, et al. Outcomes with respect to disabilities of the upper limb after hand allograft transplantation: a systematic review Transpl Int. 2012; 25:424–432. doi:10.1111/j.1432-2277.2012.01433.x
33. Landin L, Cavadas PC, Garcia-Cosmes P, et al. Perioperative ischemic injury and fibrotic degeneration of muscle in a forearm allograft: functional follow-up at 32 months post transplantation. Ann Plast Surg. 2011; 66:202–209. doi:10.1097/SAP.0b013e318206a365
34. Fries CA, Villamaria CY, Spencer JR, et al. A hyperbaric warm perfusion system preserves tissue composites ex vivo and delays the onset of acute rejection. J Reconstr Microsurg. 2019; 35:97–107. doi:10.1055/s-0038-1667298
35. Wang LC, Lawson SD, Fries CA, et al. Hyperbaric sub-normothermic ex-vivo perfusion delays the onset of acute rejection in a porcine VCA model Plast Reconstr Surg. 2015; 1364S34. doi:10.1097/01.prs.0000472317.36007.b1
36. Petruzzo P, Sardu C, Lanzetta M, et al. Report (2017) of the International Registry on Hand and Composite Tissue Allotransplantation (IRHCTT) Curr Transplant Reports. 2017; 4:294–303. doi:10.1007/s40472-017-0168-3
37. Peng YP, Lahiri A. Spare-part surgery. Semin Plast Surg. 2013; 27:190–197. doi:10.1055/s-0033-1360586
38. Ozer K, Rojas-Pena A, Mendias CL, et al. The effect of ex situ perfusion in a swine limb vascularized composite tissue allograft on survival up to 24 hours. J Hand Surg Am. 2016; 41:3–12. doi:10.1016/j.jhsa.2015.11.003
39. Nickkholgh A, Nikdad M, Shafie S, et al. Ex situ liver machine perfusion as an emerging graft protective strategy in clinical liver transplantation: the dawn of a new era. Transplantation. 2019; 103:2003–2011. doi:10.1097/TP.0000000000002772
40. Nasralla D, Coussios CC, Mergental H, et al.; Consortium for Organ Preservation in Europe. A randomized trial of normothermic preservation in liver transplantation. Nature. 2018; 557:50–56. doi:10.1038/s41586-018-0047-9
41. Slater NJ, Zegers HJH, Küsters B, et al. Ex-vivo oxygenated perfusion of free flaps during ischemia time: a feasibility study in a porcine model and preliminary results. J Surg Res. 2016; 205:292–295. doi:10.1016/j.jss.2016.06.096
42. Ozer K, Rojas-Pena A, Mendias CL, et al. Ex situ limb perfusion system to extend vascularized composite tissue allograft survival in swine. Transplantation. 2015; 99:2095–2101. doi:10.1097/TP.0000000000000756
43. Erjavec N, Pinato G, Ramser K. Raman spectroscopy as a tool for detecting mitochondrial fitness J Raman Spectrosc. 2016; 47:933–939. doi:10.1002/jrs.4930
44. Abdul-Ghani MA, DeFronzo RA. Pathogenesis of insulin resistance in skeletal muscle. J Biomed Biotechnol. 2010; 2010:476279. doi:10.1155/2010/476279

Supplemental Digital Content

Copyright © 2020 Wolters Kluwer Health, Inc. All rights reserved.