Journal Logo

Original Basic Science—Liver

Normothermic Ex Vivo Liver Perfusion Prevents Intrahepatic Platelet Sequestration After Liver Transplantation

Kollmann, Dagmar MD, PhD1,2; Linares-Cervantes, Ivan MD, PhD1; Ganesh, Sujani MSc1; Rosales, Roizar MD1; Hamar, Matyas MD1; Goto, Toru MD1; Urbanellis, Peter MD1; Tessandier, Nicolas PhD3; Boilard, Eric PhD3; Bruguera, Claudia MD1; Wiebe, Aryn BSc1; Bartczak, Agata PhD1; Yip, Paul MD4; Adeyi, Oyedele MD4; Selzner, Markus MD1; Selzner, Nazia MD, PhD1

Author Information
doi: 10.1097/TP.0000000000003194



Various cell types and mediators have been shown to be involved in hepatic ischemia/reperfusion (I/R) injury, among which platelets play a critical role.1,2 They accumulate in the liver, adhere to sinusoidal endothelial cells (SECs), and release a variety of biologically active components. Platelets and their secreted factors can initiate angiogenesis, elicit immune responses, and mediate inflammation or regeneration of damaged tissue.2,3 During liver transplantation (LT), platelets are needed to seal blood vessels that have been surgically injured. However, it is known that hepatic I/R injury disrupts the integrity of SECs which then leads to extravasation of platelets, and consequently platelet-mediated damage.4-6 The accumulation of activated and degranulated platelets in liver sinusoids (platelet sequestration) in addition to hemodilution and immunologic reactions are the cause for a significant fall in platelet count (up to 60%) after reperfusion.7-9 This drop in platelet count after LT has been reported as an independent predictor of survival and can have a major impact on postoperative outcome.10-13 Additionally, platelet transfusions during LT have been independently associated with impaired 1-year patient survival.14 The severity of I/R injury during graft reperfusion and subsequent platelet sequestration is to a large extent determined by the degree of damage the graft acquired before, during, and after graft procurement. Cold ischemia during static cold storage (SCS) at 4°C, the standard of care for graft preservation, leads to anaerobic metabolism and therefore accumulation of lactic acid and depletion of ATP.15 Recently, normothermic ex vivo liver perfusion (NEVLP) has been implemented in the field of LT as an alternative to SCS.16-18 During NEVLP, the liver is continuously perfused with perfusion solutions under physiological conditions (37°C, oxygen-carrier, nutrition, etc.) and thus an active metabolism is maintained. A major benefit of NEVLP is the ability to assess graft function or to treat the graft before transplantation.19-21 Furthermore, improved outcome after transplantation of extended criteria grafts (eg, donation after circulatory death [DCD] grafts) subjected to NEVLP has been reported in large animal studies as well as clinical studies.22-26 This study aimed to investigate the impact of NEVLP and SCS on platelet aggregation and platelet-mediated SEC injury in heart-beating donor (HBD) and DCD LT.


Study Design

Pig LT was performed with livers from HBDs or DCD donors (30 min warm ischemia time [WIT]). Livers were subjected to 8 hours SCS (SCS-group) or NEVLP (NEVLP-group) before transplantation (n = 5 in each group; Figure 1). In both NEVLP-groups (HBD-NEVLP-group and 30’DCD-NEVLP-group), livers were kept at 4°C in histidine-tryptophan-ketolutarate (HTK) solution during the back-table preparation for ex vivo perfusion (including placing cannulas, as well as preparation of washed erythrocytes for the perfusion) and were subsequently subjected to 5 hours NEVLP at 37°C. In addition, another group was added, where both donor and recipient pigs received clopidogrel (drug identification number: 02293161, TEVA pharmaceutical industries, Petah Tikva, Israel) for 4 days before organ retrieval and organ transplantation (d 1: 300 mg, d 2–4: 37.5 mg/d; HBD-SCS-Clopidogrel group, n = 3). After a total preservation time of 8 hours for both SCS- and NEVLP-groups, the livers were transplanted into recipient pigs. Pigs were followed for a survival period of 4 days, during which blood samples were drawn daily.

Overview of the study protocol and the 4 groups investigated in the study. DCD, donation after circulatory death; HBD, heart-beating donor; LT, liver transplantation; NEVLP, normothermic ex vivo liver perfusion (37°C); SCS, static cold storage (4°C); WIT, warm ischemia time.


Male Yorkshire pigs (30–35 kg) were used for the study. All included animals received humane care, which is in compliance with the “Principles of Laboratory Animal Care” formulated by the National Society for Medical Research and the “Guide for the Care of Laboratory Animals” published by the National Institutes of Health. The Animal Care Committee of the Toronto General Hospital has approved this study.

Normothermic Ex Vivo Liver Perfusion Setup

The setup for the ex vivo perfusion was built up as described previously,27 similar to the protocol of the OrganOx Metra recently published in human clinical trials.24,27,28 Briefly, the circuit setup included a centrifugal pump (Rotaflow centrifugal pump), 2 hard shell reservoirs (Maquet, Hirrlingen, Germany), a hollow-fiber dialyzer (NR16, Fresenius, Bad Homburg, Germany), and a leukocyte filter. The liver was perfused via the portal vein and the hepatic artery, and perfusate was drained through the cannulated upper and lower cava to the main reservoir. The hepatic artery pressure was set to 50–60 mm Hg resulting in a flow up to 400 mL/h. A second reservoir was used to regulate the portal vein pressure, which was aimed to reach 2–4 mm Hg and a flow of 900–1400 mL/h. Porcine blood was obtained from the donor animal shortly before liver retrieval and erythrocytes were passed through leukocyte filters. For the perfusate, 1.5 L of Steen Solution (XVIVO Perfusion, Goteborg, Sweden) was mixed with the washed porcine erythrocytes (around 400 g) and a final hematocrit of 15% with a hemoglobin level of 45 mg/dL was achieved. Steen is a buffered extracellular-type solution based on dextran and albumin, used to stabilize the perfusate with a near physiologic environment. During the priming of the circuit, Heparin (10 000 international units (IU), Sandoz Canada, Quebec, QC, Canada), amino acid concentrate (Travasol 4, 25%, 50 mL Bolus, Baxter, Hamilton, ON, Canada), sodium bicarbonate (20 mmol), calcium chloride (9.2 mmol/L), and antibiotics (Cefazolin—1 g, Pharmaceutical Partners of Canada, Richmond Hill, ON, Canada and Metronidazole—500 mg, Baxter, Mississauga, ON, Canada) were added. Additionally, amino acid concentrate (Travasol 4, 25%, 8 mL/h, Baxter, Hamilton, ON, Canada), Insulin (125 IU/h, NovoRapid, Novo Nordisk, Mississauga, ON, Canada), 2% Taurocholic acid (7 mL/h, Sigma-Aldrich, St Louis, MO, infused as a precursor for bile production), and Prostaglandin E1 (500 μg/3 h, Pfizer, Kirkland, QC, Canada) were continuously administered during the perfusion.

Pig Liver Transplantation

For the DCD organ retrieval, the donor livers were retrieved 30 minutes after induction of cardiac arrest by injecting K+Cl (20 mEq) intracardially. All donor pigs received 500 IU/kg of heparin 5 minutes before circulatory arrest (aortic cross-clamping or K+Cl injection). All livers were flushed with a total volume of 3 L cold (4°C) Custodiol-HTK (Essential Pharmaceuticals, LCC, Ewing, NJ) through the aorta and portal vein. In the SCS-groups, the livers were packed in bags filled with Custodiol-HTK and then stored in an icebox (4°C) for 8 hours. In the NEVLP-groups, the livers were cannulated and prepared for the perfusion on ice (4°C). To keep preservation time comparable for all experiments, livers from the NEVLP-groups were stored on ice for a total of 2 hours before they were perfused for 5 hours at 37°C. After 5 hours of NEVLP, the livers were flushed with cold (4°C) Custodiol-HTK and shortly stored on ice before implantation was performed. In liver recipients, a venovenous portojugular bypass (Rotaflow centrifugal pump, Maquet, Hirrlingen, Germany) was implemented during the anhepatic phase as described previously.22,29 The bypass was removed following portal vein reperfusion. Fine needle biopsies were obtained 3 hours after graft reperfusion. The animals were weaned from the respirator 3 hours after completion of the surgery. Signs of recipient distress or suffering were evaluated regularly during the 4-day survival period based on the institutional animal use protocol. Animals were euthanized under deep anesthesia in case of severe suffering (excessive bleeding, uncoordinated movements, lethargy, severe acid-base disturbances) or on postoperative day (POD) 4. Patency of vascular anastomoses were assessed at the time of euthanasia and samples from the liver parenchyma were obtained and stored in 10% formalin for further pathological analyses.

Preparation of Platelet Free Plasma and Plasma

For the preparation of platelet free plasma (PFP), blood was collected in tubes containing buffered sodium citrate, theophylline, adenosine, and dipyridamole (CTAD-tubes, BD Vacutainer, Belliver Industrial Estate, Plymouth, United Kingdom), immediately placed on ice and processed within 30 minutes. The tubes were then centrifuged at 1.000 g at 4°C for 10 minutes, followed by a second centrifugation step of the plasma supernatant at 10.000 g at 4°C for 10 minutes to remove the remaining platelets according to a previously described protocol.30 For the collection of normal plasma, blood was collected in citrated tubes (BD Vacutainer, Belliver Industrial Estate, Plymouth, United Kingdom) and centrifuged at 10.000 g for 10 minutes.

Assessment of Liver Injury, Liver Function, and Complete Blood Cell Count

In order to evaluate liver injury during ex vivo machine perfusion, liver enzymes were measured using Piccolo Xpress (Abbot, Princeton, NJ). Levels of aspartate transaminase (AST) and alanine transaminase (ALT) were determined at baseline, hour 2, and hour 4 of perfusion. Additionally, posttransplant serum AST, ALT, total bilirubin levels, and alkaline phosphatase levels were analyzed 3 hours and 5 hours after reperfusion and then daily until euthanasia. Furthermore, international normalized ratio, prothrombin time, platelet count, hemoglobin levels, and white blood cell count were measured on a daily basis after transplantation.

Tumor growth factor-β (TGF-β), platelet factor 4 (PF4), hyaluronic acid (HA), and von Willebrand factor (vWF) levels were measured by ELISA. Plasma levels of the platelet activation markers TGF-β (RND systems, Minneapolis, MN) and PF4 (Cloud Clone Corporation, Houston, TX) were determined in PFP at baseline, 3 hours after reperfusion, on POD1 and POD3. HA levels were determined for the evaluation of endothelial cell function (quantitative sandwich enzyme immunoassay technique, R&D Systems, Minneapolis, MN).31,32

Flow Cytometry Analysis of Platelet-derived Extracellular Vesicles

Platelet-derived extracellular vesicles (EVs) were determined in PFP at baseline, 3 hours after reperfusion, and on POD1 and POD3. Samples (10 μL) were incubated with 0.5 μg of fluorescein isothiocyanate isomer 1-conjugated mouse antipig CD61 (anti-CD61-fluorescein isothiocyanate) (Bio-Rad Laboratories, CA, USA) in combination with 1 μM of CellTracker Deep Red (Thermo Fisher Scientific, MA, USA). The samples were labeled during 30 minutes at room temperature in a final volume of 100 μL of PBS, and diluted to 500 μL with PBS before flow cytometry analysis. Before acquisition, a known quantity of fluorescent microspheres (2 μm diameter, Cy5-labeled, Nanocs Inc, NY, USA) were added to each tube. Samples were acquired with a BD Canto II Special Order Research Product, mounted with a forward scatter coupled to a photomultiplier tube, adapted to EV quantification (BD Biosciences, CA, USA).

Histological Evaluation

Liver biopsies obtained 3 hours after reperfusion and at the day of euthanasia were stored in 10% formalin for 48 hours followed by 70% ethanol. Sections of 5 µm were cut and processed after the biopsies were embedded in paraffin. Sections were stained by hematoxylin and eosin following standard procedure. The platelet-specific marker CD61 (JM2E5, mouse anti pig CD61, Bio-Rad Laboratories Inc, Hercules, CA) was applied in immunohistochemistry to detect platelets in the liver sinusoids. Platelet aggregates were identified as clumps of 3 or more cells within the liver and were quantified as number of clumps/5 20× high power fields. Additionally, SECs were stained using CD31 (PECAM, Santa Cruz Biotechnology, Dallas, TX). The CD31 stain was evaluated by assigning a score ranging between 1 and 4 (grade 1 = none or only scattered CD31 staining with no zonal distribution, grade 2 = reduced panzonal CD31 staining, grade 3 = reduced CD31 staining in zone 3, grade 4 = intact panzonal CD31 expression in endothelial cell lining).22 Three investigators blinded for the group allocations, including an experienced liver pathologist, evaluated all immunohistochemistry slides.

Statistical Analysis

For statistical analysis, SPSS 24.0 (IBM, Chicago, IL) and GraphPad Prism (Version 7.00 for Mac OS X, GraphPad Software, La Jolla, CA) were used. Flow cytometer analyses were done with FlowJo software (FlowJo, LLC, Ashland, OR). Results are expressed as median and interquartile range or mean ± SD. To compare continuous parameters between the groups, nonparametric tests (Kruskal-Wallis test, Wilcoxon signed-rank test, and Mann-Whitney U test) or ANOVA were applied as appropriate. For the comparison of categorical outcomes, the chi-square test was used. All statistical tests were 2-sided and P < 0.05 was considered as statistically significant.


Improved Posttransplant Outcome by NEVLP Compared to SCS

The study groups are summarized in Figure 1. All pigs from the HBD-NEVLP-group and HBD-SCS-group (n = 10) survived for 4 days post-LT. While in the DCD-NEVLP-group, all pigs (n = 5) survived for 4 days, 1 pig had to be euthanized on POD1 in the DCD-SCS-group due to primary nonfunction, resulting in a survival rate of 80% (n = 4/5) in this group.

Levels of AST were significantly higher after transplantation in the SCS-groups compared with the NEVLP-groups (median AST levels [U/L]; POD2: HBD-SCS: 1297 [626–2351], DCD-SCS: 1421 [649–2374], HBD-NEVLP: 355 [217–520], DCD-NEVLP: 249 [156–342]; P = 0.005; Figure 2A). Similarly, ALT levels were higher in the SCS-groups compared with the NEVLP-groups, with a significant difference 3 hours after reperfusion (median ALT levels [U/L]; HBD-SCS: 55 [48–57], DCD-SCS: 58 [52–71], HBD-NEVLP: 36 [26–42], DCD-NEVLP: 36 [25–58]; P = 0.028; Figure 2B).

Perioperative and postoperative time course of liver function and liver injury parameters stratified by the 4 groups. Results for baseline (BL), 3 h, and 12 h after reperfusion as well as for d 1, d 2, d 3, and d 4 are demonstrated as median and interquartile range (IQR) for (A) aspartate transaminase (AST), (B) alanine transaminase (ALT), (C) international normalized ratio (INR), (D) prothrombin time, (E) total bilirubin, and (F) hemoglobin. *P < 0.05; **P < 0.01; ***P < 0.001. DCD, donation after circulatory death; HBD, heart-beating donor; NEVLP, normothermic ex vivo liver perfusion; SCS, static cold storage; U/L, units per liter; WIT, warm ischemia time.

International normalized ratio (INR) and prothrombin time (markers for posttransplant liver function) were significantly higher in the SCS-groups on POD2 (P = 0.007 and P = 0.008, respectively) and on POD3 (P = 0.047 and P = 0.044; Figure 2C and D). Total bilirubin levels were significantly higher in the DCD-SCS group compared with the other groups at 12 hours (P = 0.016) and 2 days (P = 0.012) after transplantation (Figure 2E). Hemoglobin levels were similar between the 4 groups during the postoperative follow-up (Figure 2F), indicating only a minor and comparable blood loss in all 4 study groups.

NEVLP Promotes Posttransplant Platelet Recovery and Prevents Platelet Sequestration in Graft Sinusoids

All pigs had a comparable total platelet count at baseline (HBD-SCS: 375 × 109/L [217–401 × 109/L], HBD-NEVLP: 285 × 109/L [199–348 × 109/L], DCD-SCS: 238 × 109/L [234–378 × 109/L], DCD-NEVLP: 297 × 109/L [204–448 × 109/L]; P = 0.86). However, platelet count recovery after transplantation was significantly faster in the NEVLP-groups compared with the SCS-groups at 12 hours (median % of baseline, HBD-SCS: 22% [19%–47%] versus HBD-NEVLP: 57% [48%–92%] versus DCD-SCS: 19% [10%–33%] versus DCD-NEVLP: 58% [26%–78%], respectively; P = 0.023) and on day 1 (HBD-SCS: 14% [10%–36%] versus HBD-NEVLP: 64% [50%–98%] versus DCD-SCS: 12% [6%–33%] versus DCD-NEVLP: 69% [22%–80%], respectively; P = 0.038; [Figure 3A]). Furthermore, intrahepatic sequestration of platelets was higher in the SCS-groups, with significantly more aggregation of platelets in liver sinusoids (median number of clumps; HBD-SCS: 50 [33–115], HBD-NEVLP: 3 [1–16], DCD-SCS: 60 [40–90], DCD-NEVLP: 2 [2–13], respectively; P = 0.004; Figure 3B). Representative examples of CD61 staining for the HBD-SCS (Figure 3C), the HBD-NEVLP (Figure 3D), DCD-SCS (Figure 3E), and DCD-NEVLP (Figure 3F) groups are shown in Figure 3. Platelet count increased in the perfusate during NEVLP of the HBD and the DCD graft, indicating a wash out of aggregated platelets (Figure S1, SDC,

Platelet count and platelet sequestration in the liver sinusoids during the perioperative and postoperative course. A, Platelet count relative to baseline (BL), (B) median number of platelet aggregates counted per 5 high power fields (HPFs) in histological analyses obtained 3 h after reperfusion stained with CD61, and examples for CD61 stain of (C) HBD-SCS group, (D) HBD-NEVLP group, (E) DCD-SCS group, and (F) DCD-NEVLP group are shown. *P < 0.05; **P < 0.01; ***P < 0.001. DCD, donation after circulatory death; HBD, heart-beating donor; NEVLP, normothermic ex vivo liver perfusion; SCS, static cold storage; WIT, warm ischemia time.

NEVLP Reduces Levels of TGF-ß, PF4, and Circulating Platelet-derived Extracellular Vesicles

Levels of TGF-ß and PF4, 2 antiproliferative substances released from activated platelets, were measured in plasma at baseline, 3 hours after reperfusion, and on POD1 and POD3. There was a significant difference in TGF-ß levels between the groups 3 hours posttransplantation, on POD1 and on POD3 (P = 0.025, P = 0.004, and P = 0.016, respectively; Figure 4A). TGF-ß levels correlated with the grade of liver injury, with the DCD-SCS group showing the highest TGF-ß levels posttransplant (median plasma TGF-ß levels [pg/mL] 3 h after transplantation: HBD-SCS: 767 [532–1132], HBD-NEVLP: 555 [362–688], DCD-SCS: 1239 [998–1458], DCD-NEVLP: 772 [630–1197]; P = 0.025). PF4 levels calculated relative to platelet count increased after transplantation in all 4 groups reaching the highest levels on POD1, however, with a higher but not significant increase comparing the HBD-SCS versus HBD-NEVLP and the DCD-SCS versus DCD-NEVLP groups (median PF4 levels [ng/mL]/platelet count on POD1: HBD-SCS: 4.5 [2.3–8.1], HBD-NEVLP: 3.1 [2.4–4.2], DCD-SCS: 2.7 [1–3.8], DCD-NEVLP: 1 [0.7–2.5]; P = 0.001; Figure 4B).

Platelet activation markers and endothelial cell function after transplantation. A–C, Platelet activation markers in platelet free plasma (PFP): (A) tumor growth factor-β (TGF-ß), (B) platelet factor 4 (PF4) levels relative to platelet count, and (C) platelet-derived extracellular vesicles relative to platelet count. Postoperative endothelial cell function in the different groups is represented by hyaluronic acid levels (D) and endothelial cell integrity was determined by staining of liver biopsies obtained 3 h after reperfusion with CD31: (E) example from the HBD-NEVLP group and (F) example from the HBD-SCS group. *P < 0.05; **P < 0.01; ***P < 0.001. BL, baseline; DCD, donation after circulatory death; HBD, heart-beating donor; NEVLP, normothermic ex vivo liver perfusion; SCS, static cold storage; WIT, warm ischemia time.

Next, platelet-derived EVs were analyzed in PFP. These small EVs consist of various molecules and organelles (eg, functional enzymes, cytokines, transcription factors) and are generated via shedding from the membrane of activated platelets.33,34 EVs are involved in thrombosis, promote the recruitment of leukocytes and monocyte adhesion to activated vascular endothelium.34-36 There was a significant increase of platelet-derived EVs/platelet count on POD1 in all groups; however, the SCS-groups had a more pronounced increase compared with the NEVLP-groups (median EVs/platelet count on POD1: HBD-SCS: 3.3 × 105 [1.5–7.2 × 105], HBD-NEVLP: 2 × 105 [0.9–2.3 × 105], DCD-SCS: 2.3 × 105 [1.5–6.8 × 105], DCD-NEVLP: 0.8 × 105 [0.4–4 × 105]; P = 0.339; Figure 4C).

Sinusoidal Endothelial Cell Injury Correlates With Platelet Aggregation

SECs were evaluated by staining for the endothelial cell marker CD31. Whereas the endothelial cell lining was well maintained in the NEVLP-groups, the architecture was in large parts interrupted in both SCS-groups (mean score [1–4]: HBD-NEVLP versus HBD-SCS = 3.6 ± 1 versus 1.9 ± 0.2, P = 0.018; DCD-NEVLP versus DCD-SCS = 3.5 ± 0.4 versus 2 ± 0.9, P = 0.006). Figure 4 shows representative images from the HBD-NEVLP group (Figure 4E) and from the HBD-SCS group (Figure 4F). To evaluate the endothelial cell function, systemic HA levels were measured after transplantation. HA is metabolized by endothelial cells under normal conditions, and therefore high levels of HA are an indication of poor endothelial cell function.27,32,37 HA levels were significantly higher in the SCS-groups after transplantation when compared with the NEVLP-groups at 3 hours and on POD1 (P = 0.001 and P = 0.004, respectively; Figure 4D). On POD1, the following median HA levels (µg/mL) were measured (HBD-SCS = 1393 [776–1517] versus HBD-NEVLP = 193 [131–256]; P = 0.014 and DCD-SCS = 1178 [386–1291] versus DCD-NEVLP = 173 [143–260]; P = 0.016; [Figure 4D]). Additionally, vWF is a known factor mediating platelet aggregation. Increased consumption of total vWF caused by platelet adhesion to the graft endothelium during and after graft reperfusion, has been previously associated with platelet aggregation and increased hepatocellular and SEC injury.38 Plasma vWF levels decreased after transplantation in the DCD-SCS group (P = 0.044; Figure S2, SDC,, but remained rather stable in the other groups.

Pretreatment of Donors and Recipients With Clopidogrel Reduce Platelet-induced I/R Injury

To further provide evidence for the important role of platelets in I/R injury, another study group with clopidogrel pretreatment was added. Pig liver donors and recipients were pretreated with Clopidogrel to inhibit platelet function for 4 days before surgery. Serum levels of AST, a marker of reperfusion injury, were significantly lower in the HBD-SCS group treated with clopidogrel when compared with the HBD-SCS group without treatment at 3 hours and 5 hours posttransplantation (P = 0.028, P = 0.026, respectively; Figure 5A). The percentage of platelets dropped after transplantation in both the HBD-SCS and HBD-SCS-Clopidogrel groups; however, this drop was less prominent in the clopidogrel-treatment group (median % of platelets from baseline 12 h after transplantation: HBD-SCS: 22% [19%–47%] versus HBD-SCS-Clopidogrel: 40% [31%–72%] versus HBD-NEVLP: 57% [48%–92%], P = 0.051 and on POD1: HBD-SCS: 14% [10%–35%] versus HBD-SCS-Clopidogrel: 25% [20%–60%] versus HBD-NEVLP: 64% [50%–98%], P = 0.034). Importantly, there were almost no platelet aggregates in the liver sinusoids, as detected by CD61 staining, 3 hours after reperfusion (Figure 5B). Additionally, TGF-ß levels were lower in the HBD-SCS-Clopidogrel group compared with the HBD-SCS group (median plasma TGF-ß levels [pg/mL] on POD3: HBD-SCS-group: 854 [407–876] versus HBD-SCS-Clopidogrel: 350 [346–424] versus HBD-NEVLP-group: 295 [261–450]; P = 0.091; Figure 5C). Although the PF4 levels measured relative to the platelet count were higher in the HBD-SCS-Clopidogrel group compared with the HBD-NEVLP group, they were still lower than in the HBD-SCS group on POD1 (median PF4 levels relative to platelet count: HBD-SCS: 4.5 [2.3–8.1] versus HBD-SCS-Clopidogrel: 3.5 [1.3–6.7] versus HBD-NEVLP: 3.1 [2.4–4.2]; P = 0.470; Figure 5D). Similarly, HA levels were higher in the HBD-SCS-Clopidogrel group compared with the HBD-NEVLP group, but lower compared with the HBD-SCS group at POD 1 (median HA levels [µg/mL]: HBD-SCS: 1393 [776–1517] versus HBD-SCS-Clopidogrel: 783 [661–1091] versus HBD-NEVLP: 193 [131–256]; P = 0.016; Figure 5E). Notably, CD31 staining of the liver biopsies obtained 3 hours after reperfusion revealed a conservation of endothelial cell integrity, which was comparable to the HBD-NEVLP group (mean score [1–4]: HBD-SCS-Clopidogrel: 3.8 ± 0.2 versus HBD-SCS: 1.9 ± 0.2; P = 0.026).

Donors and recipients received clopidogrel before the standard HBD-SCS protocol (HBD-SCS-Clopidogrel) and postoperative outcome of this group was compared with the HBD-SCS group and the HBD-NEVLP group. A, aspartate transaminase (AST), (B) median number of platelet aggregates counted per 5 high power fields (HPFs) in histological analyses obtained 3 h after reperfusion stained with CD61, (C) tumor growth factor-β (TGF-ß), (D) platelet factor 4 (PF4) levels relative to platelet count, and (E) hyaluronic acid levels. BL, baseline; HBD, heart-beating donor; NEVLP, normothermic ex vivo liver perfusion; SCS, static cold storage; U/L, units per liter.


After LT, almost 90% of patients develop thrombocytopenia.10,39 Posttransplant thrombocytopenia is associated with postsurgical bleeding complications and cerebral hemorrhage but also heavily impacts graft function and patient survival.11-13 The pathophysiology of thrombocytopenia includes decreased platelet production (eg, drug-induced, infection, or liver disease), increased platelet destruction (eg, use of a bypass, sepsis, immune thrombocytopenia), and platelet sequestration or hemodilution.40 After LT, next to perioperative hemodilution, the main cause for the drop in platelet count is platelet sequestration in the liver.4,10 We could show in our experiments that SCS for 8 hours—the current gold standard—resulted in a significant drop in platelet count in liver recipients of both HBD or DCD grafts after transplantation. This drop in platelet count was almost completely absent in livers treated with NEVLP before LT. Importantly, perioperative blood loss was minimal and comparable between all 4 groups. Furthermore, histological analysis revealed a high number of platelet aggregates in the liver sinusoids after transplantation of cold stored livers, which was not found in transplanted liver grafts (HBD and DCD) pretreated with NEVLP.

Previous studies have shown that platelets entrapped in liver grafts are found alongside Kupffer cells in the space of Disse after transplantation.41 Liver SECs are most prone to I/R injury, and their damage leads to extravasation and aggregation of platelets in the sinusoids.4 Platelets have been shown to be fundamentally involved in hepatic I/R injury, further aggravating graft injury.42,43 Platelets secrete different procoagulant and proinflammatory molecules such as TGF-ß and PF4 and recently have been shown to contribute to various diseases via the release of platelet-derived EVs.34,44 Sindram et al5 have reported that after reperfusion of rat livers, which were previously kept on ice for 24 hours, platelets were rapidly sequestrated in the grafts. These aggregated platelets promoted activation of Kupffer cells, recruitment of leukocytes, and finally, SEC apoptosis.5 Our data indicate that proapoptotic and proinflammatory molecules are released from platelets after transplantation of HBD and DCD livers stored on ice, including TGF-ß, PF4, or platelet-derived EVs. This activation cascade could be significantly reduced when livers were subjected to NEVLP before LT. Consistent with these observations, the integrity of liver SECs was preserved in NEVLP-pretreated liver grafts, but not in SCS grafts. Furthermore, endothelial cell function, represented by systemic HA levels, was maintained in transplanted NEVLP-treated grafts, but not in SCS grafts. NEVLP therefore helps to protect liver injury and SEC injury from several perspectives. First, the liver injury is reduced by keeping the cold ischemia time short. SECs are protected and this leads to less platelet aggregation. Second, because of less SEC injury and platelet aggregation, there is less platelet-induced SEC injury. Additionally, platelets that have been trapped in the liver during the retrieval process are washed out during perfusion and the constant circulation in this system prevents further platelet aggregation. The combination of less cold ischemia time, reperfusion under controlled conditions and the wash out of platelets might be the crucial factors with which NEVLP is able to break through this vicious cycle.

To date, treatment options to mitigate platelet-mediated I/R injury are limited. The treatment of patients undergoing LT with antiplatelet therapy carries the risk of severe intraoperative and postoperative bleeding.45 Vivarelli et al46 examined the long-term administration of aspirin shortly after LT in 838 patients and reported reduced risk of late hepatic artery thrombosis without an increase in bleeding complications. However, to avoid platelet-mediated I/R injury, the therapy would ideally have to start before LT, which is currently not feasible. In our large animal experiments, we treated donor and recipient pigs with clopidogrel for 4 days before surgery and compared the outcomes to the SCS and NEVLP groups. We found that clopidogrel pretreatment reversed several platelet-associated detrimental effects on the grafts. This reduction of injury after preoperative treatment with clopidogrel provides further evidence that platelets play a fundamental role in I/R injury and SEC injury. Pretreating liver transplant recipients with clopidogrel is clinically difficult due to the high risk of bleeding. In our animal model, bleeding complications are rare because of well-preserved tissue layers and a lack of intra-abdominal adhesions. This is often not the case in clinical LT of cirrhotic patients. Nevertheless, we only treated the HBD-group with clopidogrel, as pretreatment of the DCD-group would have been too stressful for the animals with high risk of bleeding. Although platelet-induced SEC could be reduced by pretreatment with clopidogrel, clopidogrel only affects the platelet functions regulated by ADP, and it is possible that other platelet mediators, not regulated by ADP, may also be involved in liver injury. Furthermore, the biological effects of clopidogrel are different from using NEVLP before transplantation. There are several benefits provided during machine perfusion of the liver graft that are not modified with clopidogrel-treatment. First, grafts of the clopidogrel group were still exposed to the cold ischemia time, which compromises SEC-integrity and function. Contrarily, NEVLP reduces cold ischemia time significantly. Second, during NEVLP, the liver is provided with oxygen and nutrition during the preservation period, and thus the normal metabolic activity can be restored.47 Third, unlike in the clopidogrel group, the initial I/R injury occurs already in the machine under controlled conditions and not in the recipient.

In recent years, platelets have been attributed an important role in liver regeneration.48,49 Platelet-derived serotonin50 and a distinct alpha-granule release profile30 have been associated with improved liver regeneration after liver resection. To understand the ambivalent role of platelets, it is important to recognize the different processes involved in deceased donor LT, liver resection, or living donor LT. While ischemia-reperfusion injury is evident in all deceased donor LT, after liver resection or living donor LT the liver is mainly subjected to tissue edema and vascular engorgement followed by volume increase.51 During living donor LT, platelet transfusions were associated with improved graft regeneration, which is in total contrast to published evidence of LT using deceased donors.2 In deceased donor LT, platelet transfusions are uniformly described as an independent risk factor associated with reduced survival.14,52 This is in line with our data, which clearly underlines the negative effects of platelet aggregation in grafts from deceased donors.

In this study, both, HBD grafts and DCD grafts were tested, because both types of grafts are used in the clinical routine. The DCD-SCS-group showed increased post-LT AST levels, indicating a higher degree of injury in this model. Furthermore, 1 pig died in the DCD-SCS group due to primary nonfunction, whereas all other pigs survived. This indicates that the organs of the DCD-SCS-group had higher injury compared to the HBD-SCS-groups. Nevertheless, this is a large animal model and the experiments include an extensive surgical trauma, therefore a certain variability remains and some parameters were not higher in the DCD-SCS-group compared with the HBD-SCS-group. However, with a protocol of 8 hours cold ischemia time, 30 minutes WIT was the maximum we could apply to reach a survival of ≥80%. Despite this limitation, this study shows that both types of grafts, HBD and DCD, can benefit from NEVLP. Furthermore, due to the demanding experimental setup, this study is limited to a small number of animals per group. Additionally, we were limited by the availability of pig-specific reagents for platelet-specific factors. This study is not able to particularly distinguish between platelet-mediated SEC injury and SEC injury coming from cold ischemia and I/R injury. However, we believe that the protective effect of NEVLP includes both, less SEC injury because of reduced cold ischemia time and reduced platelet-induced SEC injury due to less platelet aggregation. The strengths of this study are a well-established pig LT model and the pig liver NEVLP system, which mimics the system currently used in the clinical setting.

In summary, we could show that NEVLP effectively prevents the sequestration of platelets in liver grafts following reperfusion. Furthermore, we found that NEVLP protects the liver from SEC injury. We therefore conclude that NEVLP is a useful strategy to reduce platelet-mediated I/R injury during and after LT of both, HBD organs and DCD organs.


1. Pereboom IT, Lisman T, Porte RJ. Platelets in liver transplantation: friend or foe? Liver Transpl. 2008; 14:923–931
2. Han S, Park HW, Song JH, et al. Association between intraoperative platelet transfusion and early graft regeneration in living donor liver transplantation. Ann Surg. 2016; 264:1065–1072
3. Gawaz M, Vogel S. Platelets in tissue repair: control of apoptosis and interactions with regenerative cells. Blood. 2013; 122:2550–2554
4. Miyashita T, Nakanuma S, Ahmed AK, et al. Ischemia reperfusion-facilitated sinusoidal endothelial cell injury in liver transplantation and the resulting impact of extravasated platelet aggregation. Eur Surg. 2016; 48:92–98
5. Sindram D, Porte RJ, Hoffman MR, et al. Platelets induce sinusoidal endothelial cell apoptosis upon reperfusion of the cold ischemic rat liver. Gastroenterology. 2000; 118:183–191
6. Sindram D, Porte RJ, Hoffman MR, et al. Synergism between platelets and leukocytes in inducing endothelial cell apoptosis in the cold ischemic rat liver: a Kupffer cell-mediated injury. Faseb J. 2001; 15:1230–1232
7. Porte RJ, Blauw E, Knot EA, et al. Role of the donor liver in the origin of platelet disorders and hyperfibrinolysis in liver transplantation. J Hepatol. 1994; 21:592–600
8. Himmelreich G, Hundt K, Isenberg C, et al. Thrombocytopenia and platelet dysfunction in orthotopic liver transplantation. Semin Thromb Hemost. 1993; 19:209–212
9. Takahashi K, Nagai S, Safwan M, et al. Thrombocytopenia after liver transplantation: should we care? World J Gastroenterol. 2018; 24:1386–1397
10. McCaughan GW, Herkes R, Powers B, et al. Thrombocytopenia post liver transplantation. Correlations with pre-operative platelet count, blood transfusion requirements, allograft function and outcome. J Hepatol. 1992; 16:16–22
11. Chatzipetrou MA, Tsaroucha AK, Weppler D, et al. Thrombocytopenia after liver transplantation. Transplantation. 1999; 67:702–706
12. Chang FY, Singh N, Gayowski T, et al. Thrombocytopenia in liver transplant recipients: predictors, impact on fungal infections, and role of endogenous thrombopoietin. Transplantation. 2000; 69:70–75
13. Lesurtel M, Raptis DA, Melloul E, et al. Low platelet counts after liver transplantation predict early posttransplant survival: the 60-5 criterion. Liver Transpl. 2014; 20:147–155
14. de Boer MT, Christensen MC, Asmussen M, et al. The impact of intraoperative transfusion of platelets and red blood cells on survival after liver transplantation. Anesth Analg. 2008; 106:32–44table of contents
15. Reddy S, Greenwood J, Maniakin N, et al. Non-heart-beating donor porcine livers: the adverse effect of cooling. Liver Transpl. 2005; 11:35–38
16. Kollmann D, Selzner M. Recent advances in the field of warm ex-vivo liver perfusion. Curr Opin Organ Transplant. 2017; 22:555–562
17. Barbas AS, Goldaracena N, Dib MJ, et al. Ex-vivo liver perfusion for organ preservation: recent advances in the field. Transplant Rev (Orlando). 2016; 30:154–160
18. Goldaracena N, Barbas AS, Selzner M. Normothermic and subnormothermic ex-vivo liver perfusion in liver transplantation. Curr Opin Organ Transplant. 2016; 21:315–321
19. Karangwa SA, Burlage LC, Adelmeijer J, et al. Activation of fibrinolysis, but not coagulation, during end-ischemic ex situ normothermic machine perfusion of human donor livers. Transplantation. 2017; 101:e42–e48
20. Gillooly AR, Perry J, Martins PN. First report of sirna uptake (for RNA interference) during ex vivo hypothermic and normothermic liver machine perfusion. Transplantation. 2019; 103:e56–e57
21. Rigo F, De Stefano N, Navarro-Tableros V, et al. Extracellular vesicles from human liver stem cells reduce injury in an ex vivo normothermic hypoxic rat liver perfusion model. Transplantation. 2018; 102:e205–e210
22. Spetzler VN, Goldaracena N, Echiverri J, et al. Subnormothermic ex vivo liver perfusion is a safe alternative to cold static storage for preserving standard criteria grafts. Liver Transpl. 2016; 22:111–119
23. Echeverri J, Goldaracena N, Kaths JM, et al. Comparison of BQ123, epoprostenol, and verapamil as vasodilators during normothermic ex vivo liver machine perfusion. Transplantation. 2018; 102:601–608
24. Selzner M, Goldaracena N, Echeverri J, et al. Normothermic ex vivo liver perfusion using steen solution as perfusate for human liver transplantation: first North American results. Liver Transpl. 2016; 22:1501–1508
25. Nasralla D, Coussios CC, Mergental H, et al.; Consortium for Organ Preservation in EuropeA randomized trial of normothermic preservation in liver transplantation. Nature. 2018; 557:50–56
26. Nostedt JJ, Skubleny DT, Shapiro AMJ, et al. Normothermic ex vivo machine perfusion for liver grafts recovered from donors after circulatory death: a systematic review and meta-analysis. HPB Surg. 2018; 2018:6867986
27. Knaak JM, Spetzler VN, Goldaracena N, et al. Subnormothermic ex vivo liver perfusion reduces endothelial cell and bile duct injury after donation after cardiac death pig liver transplantation. Liver Transpl. 2014; 20:1296–1305
28. Linares-Cervantes I, Kollmann D, Goto T, et al. Impact of different clinical perfusates during normothermic ex situ liver perfusion on pig liver transplant outcomes in a DCD model. Transplant Direct. 2019; 5:e437
29. Spetzler VN, Goldaracena N, Knaak JM, et al. Technique of porcine liver procurement and orthotopic transplantation using an active porto-caval shunt. J Vis Exp. 2015; 99:e52055
30. Starlinger P, Haegele S, Offensperger F, et al. The profile of platelet α-granule released molecules affects postoperative liver regeneration. Hepatology. 2016; 63:1675–1688
31. Deaciuc IV, Bagby GJ, Lang CH, et al. Hyaluronic acid uptake by the isolated, perfused rat liver: an index of hepatic sinusoidal endothelial cell function. Hepatology. 1993; 17:266–272
32. Soejima Y, Yanaga K, Wakiyama S, et al. Serum hyaluronic acid as a reliable parameter of allograft viability in porcine liver transplantation. Hepatogastroenterology. 1996; 43:590–595
33. Wolf P. The nature and significance of platelet products in human plasma. Br J Haematol. 1967; 13:269–288
34. Melki I, Tessandier N, Zufferey A, et al. Platelet microvesicles in health and disease. Platelets. 2017; 28:214–221
35. Barry OP, Praticò D, Savani RC, et al. Modulation of monocyte-endothelial cell interactions by platelet microparticles. J Clin Invest. 1998; 102:136–144
36. Forlow SB, McEver RP, Nollert MU. Leukocyte-leukocyte interactions mediated by platelet microparticles under flow. Blood. 2000; 95:1317–1323
37. Goldaracena N, Echeverri J, Spetzler VN, et al. Anti-inflammatory signaling during ex vivo liver perfusion improves the preservation of pig liver grafts before transplantation. Liver Transpl. 2016; 22:1573–1583
38. am Esch JS 2nd, Tustas RY, Robson SC, et al. Recipient levels and function of von Willebrand factor prior to liver transplantation and its consumption in the course of grafting correlate with hepatocellular damage and outcome. Transpl Int. 2005; 18:1258–1265
39. Beltrame P, Rodriguez S, Brandão ABM. Low platelet count: predictor of death and graft loss after liver transplantation. World J Hepatol. 2019; 11:99–108
40. Nagrebetsky A, Al-Samkari H, Davis NM, et al. Perioperative thrombocytopenia: evidence, evaluation, and emerging therapies. Br J Anaesth. 2019; 122:19–31
41. Hutchison DE, Genton E, Porter KA, et al. Platelet changes following clinical and experimental hepatic homotransplantation. Arch Surg. 1968; 97:27–33
42. Yadav SS, Howell DN, Steeber DA, et al. P-selectin mediates reperfusion injury through neutrophil and platelet sequestration in the warm ischemic mouse liver. Hepatology. 1999; 29:1494–1502
43. Peralta C, Jiménez-Castro MB, Gracia-Sancho J. Hepatic ischemia and reperfusion injury: effects on the liver sinusoidal milieu. J Hepatol. 2013; 59:1094–1106
44. Li J, Liu B, Yan LN, et al. The roles and potential therapeutic implications of CXCL4 and its variant CXCL4L1 in the pathogenesis of chronic liver allograft dysfunction. Cytokine Growth Factor Rev. 2015; 26:67–74
45. Lisman T, Porte RJ. Antiplatelet medication after liver transplantation: does it affect outcome? Liver Transpl. 2007; 13:644–646
46. Vivarelli M, La Barba G, Cucchetti A, et al. Can antiplatelet prophylaxis reduce the incidence of hepatic artery thrombosis after liver transplantation? Liver Transpl. 2007; 13:651–654
47. Jayant K, Reccia I, Shapiro AMJ. Normothermic ex-vivo liver perfusion: where do we stand and where to reach? Expert Rev Gastroenterol Hepatol. 2018; 12:1045–1058
48. Starlinger P, Assinger A. Importance of platelet-derived growth factors in liver regeneration. Expert Rev Gastroenterol Hepatol. 2016; 10:557–559
49. Starlinger P, Assinger A, Haegele S, et al. Evidence for serotonin as a relevant inducer of liver regeneration after liver resection in humans. Hepatology. 2014; 60:257–266
50. Lesurtel M, Graf R, Aleil B, et al. Platelet-derived serotonin mediates liver regeneration. Science. 2006; 312:104–107
51. Taki-Eldin A, Zhou L, Xie HY, et al. Liver regeneration after liver transplantation. Eur Surg Res. 2012; 48:139–153
52. Pereboom IT, de Boer MT, Haagsma EB, et al. Platelet transfusion during liver transplantation is associated with increased postoperative mortality due to acute lung injury. Anesth Analg. 2009; 108:1083–1091

Supplemental Digital Content

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