PAPER OF THE 21ST ANNUAL ESA MEETING

Hypothermic Oxygenated Perfusion (HOPE) Downregulates the

Immune Response in a Rat Model of Liver Transplantation
Andrea Schlegel, MD,∗ Philipp Kron, MD,∗ Rolf Graf, PhD,∗ Pierre-Alain Clavien, MD, PhD,∗ †
and Philipp Dutkowski, MD∗

Recently, we have presented the worldwide first results of a

Objective: To evaluate the impact of a novel oxygenated perfusion approach
newly developed machine liver perfusion approach for human liver
on rejection after orthotopic liver transplantation (OLT).
grafts, donated after cardiac death.5 The perfusion technique con-
Background: Hypothermic oxygenated perfusion (HOPE) was designed to
sisted of a short-term hypothermic oxygenated perfusion (HOPE) af-
prevent graft failure after OLT. One of the mechanisms is downregulation
ter conventional organ procurement and cold storage and was shown
of Kupffer cells (in situ macrophages). We, therefore, designed experiments
to protect effectively from ischemia-reperfusion (I/R) injury despite
to test the effects of HOPE on the immune response in an allogeneic rodent
long donor warm ischemia times.6–8 We have also reported, in several
model of nonarterialized OLT.
experimental studies, that such effects of HOPE depend on oxygena-
Methods: Livers from Lewis rats were transplanted into Brown Norway rats
tion of the perfusate6 with decreased downstream activation of numer-
to induce liver rejection in untreated recipients within 4 weeks. Next, Brown
ous inflammatory pathways.6,7 On the basis of these findings, the aim
Norway recipients were treated with tacrolimus (1 mg/kg), whereas in a third
of the current study was to analyze effects of HOPE treatment on the
group, liver grafts from Lewis rats underwent HOPE or deoxygenated ma-
immune response in a model of acute rejection, using incompatible
chine perfusion for 1 hour before implantation, but recipients received no
rat strains for OLT.9 We compared allogeneic liver transplants with
immunosuppression. In a last step, low-dose tacrolimus treatment (0.3 mg/kg)
and without immunosuppressive or HOPE treatment. In addition, to
was assessed with and without HOPE.
provide mechanistic insights, we added perfusion experiments with
Results: Allogeneic OLT without immunosuppression led to death within
a deoxygenated perfusate. We focus on T-cell response and survival
3 weeks after nonarterialized OLT due to severe acute rejection. Full-dose
within 30 days after liver transplantation.
tacrolimus prevented rejection, whereas low-dose tacrolimus led to graft fi-
brosis within 4 weeks. HOPE treatment without immunosuppression also
protected from lethal rejection. The combination of low-dose tacrolimus and METHODS
1-hour HOPE resulted in 100% survival within 4 weeks without any signs of Animals
rejection. Male Lewis and Brown Norway rats (250–320 g) were used for
Conclusions: We demonstrate that allograft treatment by HOPE not only all experiments. Animals received standard laboratory diet and water
protects against preservation injury but also impressively downregulates the according to the Swiss Animal Health Care law, and experiments were
immune system, blunting the alloimmune response. Therefore, HOPE may approved by the animal ethics committee. Anesthesia during liver
offer many beneficial effects, not only to rescue marginal grafts but also by procurement and transplantation was maintained with isoflurane.
preventing rejection and the need for immunosuppression.
Keywords: HOPE, immune response, liver transplantation, oxygen, rejection Study Design
(Ann Surg 2014;260:931–938) Transplantation of livers from Lewis rats into Brown Norway
recipients has previously been shown to induce rejection.9–12 Using
this strain combination, we compared in an allogeneic liver transplant
model the impact of HOPE against standard immunosuppression. We
O rthotopic liver transplantation (OLT) offers a unique chance
to treat patients with end-stage liver disease. The persistent
need for lifelong immunosuppressive treatment, however, is a major
selected the following experimental groups (see Supplemental Digital
Content Fig. 1, available at http://links.lww.com/SLA/A649):
shortcoming in terms of several serious side effects including infec- I. Syngeneic control: Livers from Brown Norway rats were pro-
tious complications,1 long-term nephrotoxicity,2 and primary tumor cured and implanted into Brown Norway recipients (Syngeneic
recurrence3 or secondary tumor growth.4 For these reasons, novel control).
strategies decreasing the need of immunosuppressive treatment after II. Untreated group: Livers from Lewis rats were procured and
transplantation seem highly attractive. transplanted into Brown Norway recipients without any addi-
tional treatment (Untreated).
III. Full immunosuppression: Livers from Lewis rats were procured
and transplanted into Brown Norway recipients, which received
From the ∗ Laboratory of the Swiss HPB and Transplant Center, Department of
Surgery, University Hospital Zurich, Switzerland; and †Hospital Paul Brousse,
full immunosuppressive treatment with tacrolimus intramuscu-
Université Paris Sud, Paris, France larly (1 mg/kg of bodyweight/d), starting before recipient hepa-
Disclosure: Supported by Swiss National Science Foundation grant 32003B- tectomy (TAC).
140776/1 to P.D. The authors declare no conflicts of interest. IV. Low-dose immunosuppression: Livers from Lewis rats were pro-
Supplemental digital content is available for this article. Direct URL citation appears
in the printed text and is provided in the HTML and PDF versions of this article
cured and transplanted into Brown Norway recipients, which re-
on the journal’s Web site (www.annalsofsurgery.com). ceived a reduced immunosuppressive treatment with tacrolimus
Reprints: Philipp Dutkowski, MD, Department of Surgery & Transplantation, intramuscularly (0.3 mg/kg of bodyweight/d), starting before
University Hospital Zurich, Raemistrasse 100, CH-8091 Zurich, Switzerland. recipient hepatectomy (low TAC).
E-mail: philipp.dutkowski@usz.ch.
Copyright C 2014 by Lippincott Williams & Wilkins
V. HOPE group: Livers from Lewis rats were procured, machine
ISSN: 0003-4932/14/26005-0931 perfused with an oxygenated perfusate (HOPE) for 1 hour, and
DOI: 10.1097/SLA.0000000000000941 transplanted without any additional recipient treatment (HOPE).

Annals of Surgery r Volume 260, Number 5, November 2014 www.annalsofsurgery.com | 931

Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Schlegel et al Annals of Surgery r Volume 260, Number 5, November 2014

VI. HNPE group: Livers from Lewis rats were procured, machine Liver Procurement and OLT
perfused with a nitrogenated perfusate (HNPE) for 1 hour, Donor livers were freed from ligaments and flushed with
and transplanted without any additional recipient treatment 6 mL of heparinized (1 U/mL) saline at room temperature via the
(HNPE). portal vein. Livers were excised (weight 9.7 ± 1.5 g) and placed
VII. HOPE + low-dose immunosuppression: Livers from Lewis rats in precooled UW solution (4◦ C). Afterward, non–machine-perfused
were procured, machine perfused with an oxygenated perfusate livers received cuffs for the portal vein and the infrahepatic vena
(HOPE) for 1 hour, and transplanted into Brown Norway recip- cava whereas livers allocated to the perfusion groups received first a
ients, which received a reduced immunosuppressive treatment stent for the portal vein and later cuffs after machine perfusion. Cold
with tacrolimus intramuscularly (0.3 mg/kg of bodyweight/d), storage was approximately 30 minutes in all experimental groups.
starting before recipient hepatectomy (HOPE + low TAC). Nonarterialized liver transplantation was performed after procure-
ment or machine perfusion according to the technique by Kamada and
Endpoints Calne.13
We analyzed hepatocyte necrosis, Kupffer cell activation, en-
dothelial cell activation, and T-cell activation by specific staining
procedures 24 hours after OLT in each group. In additional ex- HOPE and HNPE
periments, we documented plasma alanine aminotransferase (AST), Livers from Lewis rats to be cold perfused were connected to
bilirubin, high mobility group box-1 protein (HMGB-1), 8-hydroxy- the precooled perfusion device and perfused for 1 hour through the
2-deoxy guanosine (8-OHdG), interleukin (IL)-2, IL-10, and inter- portal vein with a constant perfusion pressure of 3 mm Hg or less.14
feron gamma (IFN-γ ) during 14 days after OLT. The T-cell response We used 50 mL of recirculating modified starch-free UW solution
after 2 weeks was investigated by fluorescence-activated cell-sorting as perfusate.7,8 Perfusion box and perfusate were maintained at 4◦ C
(FACS) analysis in blood samples. Additional staining procedures by an open bath thermostat (Huber, Germany). In all HOPE exper-
were performed 2 and 4 weeks after transplantation to confirm tissue iments, the cold perfusate was actively oxygenated (pO2 >60 kPa)
remodeling into liver fibrosis and rejection. Follow-up for survival (HOPE group). In the HNPE group, oxygen was replaced by nitrogen
after OLT was 4 weeks. (pO2 <2 kPa) (HNPE group).

Syngeneic control Untreated TAC HOPE HNPE

A

H&E
(Hepatocytes)

CD 68 positive
cells
(Kupffer cells)

vWF
(Sinusoidal
endothelial cells)

CD3-positive
cells
(T cells)

FIGURE 1. Liver injury 24 hours after OLT: allogeneic liver transplantation without immunosuppression induced hepatocyte injury
and Kupffer and endothelial cell activation during the first day after reperfusion (A–C). HOPE treatment significantly protected
hepatocytes from reperfusion injury (A). Macrophages and endothelial cells after HOPE also appeared less activated (B, C).
Machine perfusion with a deoxygenated perfusate (HNPE) induced the same degree of injury after OLT as untreated allogeneic
controls (A–C). Treatment with tacrolimus showed less effect on Kupffer cell activation compared with HOPE (B). One day after
OLT, CD3-positive T cells in livers were rarely detectable in all experimental groups (D). H&E indicates hematoxylin-eosin; TAC,
tacrolimus; vWF, von Willebrand factor.

932 | www.annalsofsurgery.com 
C 2014 Lippincott Williams & Wilkins

Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Annals of Surgery r Volume 260, Number 5, November 2014 Impact of HOPE on Immune Response After Liver Transplantation

Assays for fibrosis, CD68 staining for Kupffer cell activation, CD80/86
Hepatocyte injury after transplantation was measured by AST staining for antigen presenting cells, von Willebrand factor staining
release and total bilirubin (serum multiple biochemical analyzer DRI- for endothelial activation, and CD3 staining for T lymphocytes.
CHEM4000i; FUJIFILM, Japan). Oxidative damage of DNA by Quantification of CD68-positive cells, CD80-positive cells, von
oxygen free radicals was detected using an 8-OHdG enzyme-linked- Willebrand factor–positive endothelial cells, and CD3-positive
immunosorbent assay (Abnova, KA0444). Nuclear subcellular injury lymphocytes in liver tissue was performed in 8 random fields of 2
was measured by release of HMGB-1 using a specific enzyme-linked- slides per animal, resulting in 160 quantified fields per experimental
immunosorbent assay (IBL International GmbH, ST51011). In addi- group. Acute rejection was assessed on hematoxylin-eosin staining
tion, several markers indicating graft rejection were analyzed in the and based on portal inflammation, bile duct inflammation, and
rat plasma after transplantation, that is, IL-2 (IBL, 87728012), IL-10 endothelial inflammation,15 and signs of chronic rejection based on
(R&D, R1000), and IFN-γ (R&D, RIF00). portal inflammation, bile duct damage, obstructive arteriopathy, and
graft fibrosis.16
FACS Analysis
After blood retrieval, 1.5 mL of heparinized full rat blood Statistics
was used for this analysis. After white blood cell isolation us- Data are presented as means ± standard deviation. Statisti-
ing Ficoll and DMEM+P/S buffer, cells were incubated with var- cal analysis was performed using the nonparametric Mann-Whitney-
ious combinations of mAbs (anti-Rat CD3, PerCP-eFluor710 sin- Wilcoxon U test (GraphPadPrism, version 4.0; San Diego, CA).
gle staining, anti-rat CD4-PE combined with anti-rat CD25-FITC,
anti-rat CD8-FITC combined with anti-rat CD28-PE), washed twice
with FACS buffer, and fixed with 1% paraformaldehyde. Three- RESULTS
color immunofluorescence staining was analyzed using a FACS
Calibur instrument (FACS diva, version 6.1.2). The lymphocytes Acute Rejection Without and With
were gated using forward and side scatter to exclude debris and Immunosuppressive Therapy in an Allogeneic Rat
dead cells. Afterward, 50,000 events were acquired in each assay for Liver Transplant Model
analysis. In a first step, we documented the degree of liver graft injury in
an allogeneic rat liver transplant model (donor livers from Lewis rats
Histological Examinations into Brown Norway recipients)9,10 in contrast to syngeneic transplant
The following staining procedures were performed after controls. Recipients without immunosuppressive treatment developed
transplantation: hematoxylin-eosin staining for necrosis, Sirius red severe graft injury within 24 hours after OLT, confirmed by nuclear

A AST B Bilirubin C 8-OHdG D HMGB-1

untreated
HNPE
HOPE 25
Plasma Bilirubin (μmol/L)

Plasma 8-OHdG (ng/mL)

3000 TAC (full) 12000

Plasma HMGB I (ng/mL)

200
syngeneic control 20
Plasma AST (U/L)

150 8000
2000 15
100 10
1000 4000
50 5

0 0 0 0

01 3 5 7 14 01 3 5 7 14 01 3 5 7 14 01 3 5 7 14
Days after transplantation Days after transplantation Days after transplantation Days after transplantation

E IL-2 F IL-10 G IFN-y

6000 1000
600
Plasma IL-10 (pg/mL)

Plasma IFN- y (pg/mL)

800
Plasma IL-2 (pg/mL)

4000
600 400

400
2000 200
200

0 0
0

01 3 5 7 14 01 3 5 7 14
01 3 5 7 14
Days after transplantation Days after transplantation
Days after transplantation

FIGURE 2. Liver injury during 2 weeks after allogeneic OLT: Recipients without immunosuppression developed elevated levels
of liver enzymes (A), bilirubin (B), IL-2, IL-10 (E, F), and IFN-γ (G). Nuclear injury and oxidative stress markers (ie, HMGB-1,
and 8-OHdG) increased in untreated allogeneic grafts and also in livers perfused with deoxygenated perfusate (HNPE group)
(C, D). Immunosuppression (TAC) had no effect on initial DAMP molecule and reactive oxygen species release but protected from
the release of rejection markers (E–G). HOPE treatment led, first, to significantly decreased reperfusion injury, as shown by less
amounts of oxidized DNA (C) and HMGB-1 (D), and second, induced less IL-2, IL-10, and IFN-γ release (E–G) during 2 weeks
after OLT. Machine perfusion without oxygen abrogated the effect of HOPE (A–G). TAC indicates tacrolimus.


C 2014 Lippincott Williams & Wilkins www.annalsofsurgery.com | 933

Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Schlegel et al Annals of Surgery r Volume 260, Number 5, November 2014

injury (8-OHdG and HMGB-1 release), AST release, and Kupffer survived (6/8) but developed liver fibrosis within 4 weeks after OLT
and endothelial cell activation (Figs. 1, 2). Subsequently, in these (Fig. 5).
animals, we observed acute rejection during the first 2 weeks after In summary, allogeneic nonarterialized rat liver transplantation
transplantation, as documented by massive T-cell infiltration (Fig. 3), without immunosuppression led to severe acute rejection, liver fibro-
along with persistent high blood levels of AST, bilirubin, IL-2, IL-10, sis, and animal death within 3 weeks. Lethal rejection and fibrosis
and IFN-γ (Fig. 2). In addition, Sirius red staining in liver samples within 4 weeks were prevented with adequate tacrolimus treatment
after 2 weeks revealed severe graft fibrosis (Fig. 4). Such degree of recipients, whereas low-dose tacrolimus induced graft fibrosis and
of injury was lethal in all untreated recipients during 18 days after chronic graft injury.
allogeneic OLT (Fig. 3) despite choosing experimental conditions
with minimal exposure of liver grafts to cold and warm ischemia (see
Supplementary Fig. 1, available at http://links.lww.com/SLA/A649). Liver Graft Protection by Short-term Application
In a next step, we treated recipients before graft implantation of HOPE Without Immunosuppressive Treatment
with tacrolimus (1 mg/kg of body weight/d),9 resulting in a trough To investigate the effects of HOPE on the immune response in
level of 8 to 10 ng/l, according to earlier studies.9 Such degree of the same allogeneic transplant model, we perfused donor livers from
immunosuppression protected significantly from endothelial cell ac- Lewis rats for 1 hour and implanted perfused liver grafts in Brown
tivation 24 hours after OLT (Fig. 1). In addition, within the following Norway recipients, without any immunosuppressive treatment. As
2 and 4 weeks, infiltration of T cells in liver grafts remained low expected from previous studies,6,8 HOPE reduced reperfusion injury
(Fig. 3). These findings were paralleled by the decreased amount of within the first 24 hours after OLT, confirmed by minimal 8-OHdG,
circulating activated T cells (Fig. 3) and low cytokine levels in blood HMGB-1, and AST release and decreased activation of Kupffer and
(Fig. 2). Recipient survival increased to 80% within 4 weeks (6/8) endothelial cells (Figs 1, 2). Furthermore, HOPE treatment prevented
(Fig. 3). infiltration of CD3-positive T cells in liver grafts (Fig. 3) consistent
Finally, we reduced immunosuppressive therapy to one third with low levels of IFN-γ , IL-2, and IL-10 (Fig. 2). Two weeks af-
of the daily dose (0.3 mg/kg of body weight/d) resulting in a trough ter HOPE treatment and OLT, quantification of circulating activated
level of 3 to 4 ng/l. With this low-dose tacrolimus therapy, recipients CD3-positive T cells remained low (Fig. 3) and graft fibrosis and

A Quantification of circulating total and activated T cells in blood C Survival

(FACS-Fluorescence-activated-cell-sorting )

CD3 + CD3 +/ CD28+ CD3 +/ CD25+

P=0.01 P=0.17 100 Syngeneic control

30 P=0.005 30 P=0.005 10 P=0.002

CD3+/CD28-positive T cells (%)

CD3-/CD25-positive T cells (%)

80 TAC
P=0.005 P=0.005 P=0.008 P=0.002 P=0.002 P=0.7 P=0.03 HOPE
8
CD3-positive cells (%)

Percent survival

P=0.005 P=0.81 P=0.005

20 60
20
6
HNPE
40
4 untreated
10 10
2 20

0 0 0 0
1 2 3 4
l

PE

PE
l

PE

ro

ed
ro

ed

PE

PE

PE

TA
ro

ed
TA

nt
TA

N
at
nt

Weeks after transplantation

O
at

nt

HO

HN
at

H
H
co
H

re
co

re

co

tre

nt
nt

c
c

ei

U
Un
ic
ei

en
en

ne

ng
ng

e
ng

sy
sy

sy

P<0.0001
B CD-3 positive cells in liver tissue P<0.0001
P<0.0001P<0.0001 P<0.0001
200
No. of CD3-positive cells/HPF

Syngeneic control Untreated TAC HOPE HNPE

150

100

50

0
l

PE

PE
ro

ed

TA
nt

HO

HN
at
co

tre
Un
c
ei
en
ng
sy

FIGURE 3. T-cell activation and survival 2 to 4 weeks after OLT: The number of circulating T cells in blood was quantified by FACS
analysis (A). CD28- and CD25-positive T cells were less observed in recipients of HOPE-treated livers and also in animals with full
immunosuppression (FACS analysis) (A). The number of infiltrating T cells in liver grafts was significantly reduced in HOPE- and
TAC-treated animals (B). HOPE and TAC treatment increased recipient survival, whereas untreated controls died within 2 weeks
after OLT (C). TAC indicates tacrolimus.

934 | www.annalsofsurgery.com 
C 2014 Lippincott Williams & Wilkins

Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Annals of Surgery r Volume 260, Number 5, November 2014 Impact of HOPE on Immune Response After Liver Transplantation

P=0.05

P<0.0001
A CD 80/86-positive cells (Antigen presenting cells, APC) P<0.0001 P<0.0001 P<0.0001
150

No. of CD80-positive cells/HPF

Syngeneic control Untreated TAC HOPE HNPE
100

50

PE

PE
ro

ed

TA
nt

N
at

H
H
co

re
nt
c
B H&E

ei

U
en
ng
sy
Syngeneic control Untreated TAC HOPE HNPE

C Sirius Red staining (Fibrosis)

P=0.2
100
P=0.0015
Syngeneic control Untreated TAC HOPE HNPE

% of fibrotic fibres/HPF
P<0.0001 P=0.0002 P=0.02
80

60

40

20

PE

PE
ro

ed

TA
nt

N
at

H
H
co

re
nt
c
ei

U
en
ng
sy
FIGURE 4. HOPE treatment led to significant reduction of CD80/86-positive cells and protected from high-grade fibrosis compared
with untreated recipients and machine perfusion without oxygen (HNPE) (A–C). However, HOPE without any immunosuppression
induced mild fibrosis (C) not seen under full immunosuppression (TAC) (A–C). H&E indicates hematoxylin-eosin; HPF, high power
field; TAC, tacrolimus.

survival appeared significantly improved by HOPE compared with absence of oxygen abrogated downstream the protective effect of
untreated liver grafts within 4 weeks after OLT (Figs. 3, 4). However, HOPE on activation of T cells (Fig. 3) and led to acute liver graft
graft histology after 4 weeks disclosed increased signs of chronic in- rejection, fibrosis (Fig. 4), and death of all recipients within 4 weeks
jury comparable with those seen with low-dose tacrolimus treatment after OLT (Fig. 3), similarly as in untreated recipients. On the basis
(Fig. 5). of this, the effect on the direct immune response by HOPE seems to
On the basis of these experiments, we conclude that HOPE be mediated by perfusate oxygen and points therefore to an initial
treatment of grafts protects from activation of the early immune re- mitochondrial protection during machine perfusion, with subsequent
sponse in recipients and rescued from lethal injury in an allogeneic decreased release of reactive oxygen species and DAMP molecules.
nonarterialized liver transplant model. Yet, HOPE alone, without any
further immunosuppression, failed to prevent chronic graft injury. Liver Graft Protection With Low-Dose Tacrolimus
in Combination With HOPE
Immune Response After Machine Perfusion In a last experimental step, we combined graft treatment by
Without Oxygen HOPE and low-dose immunosuppressive therapy. With this strategy,
Previous results have shown that protection against I/R injury all recipients survived for 4 weeks without signs of rejection or fibrosis
depended on the presence of oxygen in the machine perfusate during in contrast to recipients under low-dose tacrolimus without HOPE
HOPE.6 Here, we analyzed whether early immune response of the (Fig. 5). We conclude that the addition of HOPE before OLT allows to
recipient was also mediated through oxygenation effects. For this reduce immunosuppressive therapy without increased risk of rejection
purpose, we perfused liver grafts from Lewis rats with a nitrogenated in an allogeneic rodent transplant model.
perfusate (HNPE) at 4◦ C (pO2 level <2 kPa). Livers perfused with
this technique demonstrated significant higher degrees of reperfusion DISCUSSION
injury after OLT in Brown Norway recipients, confirmed by release of This study showed that liver allograft treatment by an easy
damage associated molecular pattern (DAMP) molecules (HMBG-1), applicable machine perfusion approach before OLT not only is
AST, and cytokines (IL-10, IL-2, IFN-γ ) and Kupffer and endothelial effective against reperfusion injury but also prevents activation of
cell activation (Figs. 1, 2). Furthermore, machine perfusion in the the immune response pathways. This was evident, first, by decreased


C 2014 Lippincott Williams & Wilkins www.annalsofsurgery.com | 935

Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Schlegel et al Annals of Surgery r Volume 260, Number 5, November 2014

A AST B Bilirubin C Tacrolimus Plasma Level

P=0.66
P=0.014 P=0.55
14

Tacrolimus Plasma Level (ng/l)

1200 P=0.008
P=0.002 P=0.18 P=0.14 140
12

Plasma Bilirubin (µmol/l)

120 P=0.008 P=0.046 P=0.07
TAC
Plasma AST (U/l)

10
800 100
80 8
60 6
400 low TAC
40 4 E Fibrotic Fibres in liver tissue
20 HOPE + low TAC
2 P=0.05
0 0 100

% of fibrotic fibres/HPF
0 P<0.0001
C C PE C
C

PE

C
TA TA TA
TA

TA

TA
80 P<0.0001 P=0.0006 P=0.02

O
O
H
H
w

w
w w
Lo lo
Lo

lo
+ 1357 14 28

+
PE 60
PE Days after transplantation
O
O
H

H 40

D Histology 20
TAC Low TAC HOPE HOPE + low TAC
0

PE

C
TA

TA

TA
O
H
w

w
Lo

lo
H&E

+
PE
O
H
F CD3-positive cells in liver tissue

No. of CD3-positive cells/HPF

P=0.06
200
P<0.0001
Sirius Red 150 P<0.0001 P=0.0004 P<0.0001
(Fibrosis)
100

50

PE

C
TA

TA

TA
O
H
w

w
CD-3

Lo

lo
+
PE
(T cells)

O
H
FIGURE 5. Immune response 4 weeks after OLT: Reduction of tacrolimus to levels between 4 and 5 ng/l (C) induced liver fibrosis
and T-cell infiltration within 4 weeks (D). Adding HOPE before OLT prevented significantly liver injury, fibrosis, and T-cell infiltration
within 4 weeks despite low-dose TAC treatment (A–F). TAC indicates tacrolimus.

Kupffer and endothelial cells activation after HOPE treatment and Liver graft injury arises during the transplantation process from
OLT, followed by decreased T-cell infiltration in liver grafts, and also several sources, that is, as a result of donor brain or cardiac death,
by a decreased amount of circulating and activated T cells in blood. from cold storage, and from warm ischemia and reperfusion in the
Of note, the initial effect of HOPE treatment on the direct immune recipient. It is believed that hypoxia during procurement, preserva-
response pathway appeared comparable with the immunosuppression tion, and implantation triggers the release of reactive oxygen species
conferred by tacrolimus. Furthermore, whereas a significant reduc- in different compartments and the release of DAMP molecules, char-
tion of immunosuppressive treatment provoked graft fibrosis within acterized by extracellular matrix fragments, nucleic acids, histones,
4 weeks, the addition of HOPE before OLT protected from signs of and HMGB-1.19,20 DAMP molecules bind to Kupffer cells, dendritic
chronic graft injury. cells, leukocytes, and endothelial cells by numerous toll like recep-
Next, we demonstrated that the effect of HOPE against an tors (TRL-2, TLR-3, TLR-4. TLR-7, TLR-9) and receptor for ad-
immune response depends on oxygenation of the perfusate along vanced glycation end products (RAGE).20–24 Recognition of DAMP
with strong impact against reperfusion injury.7 Consistently, HOPE molecules through these receptors activate both donor- and recipient-
treatment prevented from lethal graft rejection in an allogeneic trans- derived dendritic cells, besides activation by other factors such as
plant model whereas hypothermic perfusion using a deoxygenated complement and lymphocytes.14,25–27 It has recently been suggested,
perfusate failed to protect from acute rejection. It seems therefore therefore, that prevention of an initial oxidative stress and DAMP
unlikely that washout effects of immunocompetent cells during ma- molecule release in donors and recipients could be a key option for
chine perfusion contribute to the decreased immune response ob- subsequent modulation of immune and inflammatory responses.28
served in HOPE-treated livers. This important finding is consistent However, available free radical scavengers or other pharmacological
with previous studies, which have underlined a key role of oxy- approaches have failed in clinical practice, due to their low activity
gen and mitochondrial function during HOPE.6,7,17 In parallel, end- at the time and site of graft injury.29 In contrast, hypothermic oxy-
ischemic oxygenated cold perfusion of porcine kidneys has recently genated machine perfusion before implantation, initially developed
been shown to reduce the innate immune response, as evaluated to rescue marginal liver grafts,5 offers a unique chance to impact on
by HMGB-1 release and gene expression of toll like receptor-4 the main source of intracellular oxidative stress due to changes in the
(TRL-4).18 mitochondrial redox state6 before implantation. On the basis of this,

936 | www.annalsofsurgery.com 
C 2014 Lippincott Williams & Wilkins

Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Annals of Surgery r Volume 260, Number 5, November 2014 Impact of HOPE on Immune Response After Liver Transplantation

such machine perfusion technique potentially impacts on donor im- 13. Kamada N, Calne RY. Orthotopic liver transplantation in the rat. Technique
munogenic cells before any exposure to recipient cells. Accordingly, using cuff for portal vein anastomosis and biliary drainage. Transplantation.
1979;28:47–50.
we show here, for the first time, in a rodent liver transplant model that
14. Dutkowski P, Furrer K, Tian Y, et al. Novel short-term hypothermic oxygenated
1-hour perfusion with a cold oxygenated perfusate before OLT was perfusion (HOPE) system prevents injury in rat liver graft from non-heart
highly efficient in preventing lethal rejection, suggesting a direct link beating donor. Ann Surg. 2006;244:968–976; discussion 976–977.
between early reperfusion injury after organ transplantation and the 15. Banff schema for grading liver allograft rejection: an international consensus
initiation of the immune response.30 document. Hepatology. 1997;25:658–663.
Machine perfusion strategies may, therefore, not only improve 16. Demetris A, Adams D, Bellamy C, et al. Update of the International Banff
initial organ function by decreasing reperfusion injury but also reduce Schema for Liver Allograft Rejection: working recommendations for the
histopathologic staging and reporting of chronic rejection. An International
immune response by less activation of tissue-resident dendritic cells. Panel. Hepatology. 2000;31:792–799.
We opted in this study for a nonarterialized allogeneic rat OLT 17. Luer B, Koetting M, Efferz P, et al. Role of oxygen during hypothermic machine
model for 2 reasons. First, previous studies in rats have shown con- perfusion preservation of the liver. Transpl Int. 2010;23:944–950.
sistently that a nonarterialized rat OLT model was highly suitable for 18. Gallinat A, Paul A, Efferz P, et al. Hypothermic reconditioning of porcine kid-
rejection studies within 4 weeks after OLT.9,11,31,32 Second, additional ney grafts by short-term preimplantation machine perfusion. Transplantation.
hepatic artery reconstruction prolonged survival but finally failed to 2012;93:787–793.
prevent chronic rejection and graft dysfunction.31,32 19. Tang D, Kang R, Zeh HJ III, et al. High-mobility group box 1, oxidative stress,
and disease. Antioxid Redox Signal. 2011;14:1315–1335.
Translation of our results to the human situation may require
20. Land WG. Emerging role of innate immunity in organ transplantation, part II:
longer follow-up and experiments in large animals. Further investi- potential of damage-associated molecular patterns to generate immunostimu-
gations should also include nonstandard liver grafts, that is, extended latory dendritic cells. Transplant Rev (Orlando). 2012;26:73–87.
criteria grafts or liver grafts donated after cardiac death, as we choose 21. Land WG. Emerging role of innate immunity in organ transplantation, part I:
in this study an experimental model without previous exposure to evolution of innate immunity and oxidative allograft injury. Transplant Rev
warm or cold ischemia. (Orlando). 2012;26:60–72.
22. Land WG. Emerging role of innate immunity in organ transplantation, part III:
the quest for transplant tolerance via prevention of oxidative allograft injury
CONCLUSIONS and its consequences. Transplant Rev (Orlando). 2012;26:88–102.
HOPE-treated liver grafts may tolerate a significant lower im- 23. van Golen RF, van Gulik TM, Heger M. Mechanistic overview of reactive
munosuppressive treatment. If confirmed in the clinic, the use of species-induced degradation of the endothelial glycocalyx during hepatic is-
chemia/reperfusion injury. Free Radic Biol Med. 2012;52:1382–1402.
HOPE may open the door for regimens with significant low or de- 24. Lukacs-Kornek V, Schuppan D. Dendritic cells in liver injury and fibrosis:
layed calcineurin inhibitors without a risk of graft rejection, but less shortcomings and promises. J Hepatol. 2013;59:1124–1126.
toxic effects in high MELD recipients or donated after cardiac death 25. Sacks SH, Zhou W. The role of complement in the early immune response to
liver transplants.33,34 In addition, underlying recipient diseases that transplantation. Nat Rev Immunol. 2012;12:431–442.
require minimal immunosuppressive treatment, that is, recurrent hep- 26. Schlichting CL, Schareck WD, Weis M. Renal ischemia-reperfusion injury:
atitis C,35 or advanced hepatic tumors36 could benefit from this new new implications of dendritic cell-endothelial cell interactions. Transplant
strategy. Proc. 2006;38:670–673.
27. Zhang Q, Raoof M, Chen Y, et al. Circulating mitochondrial DAMPs cause
inflammatory responses to injury. Nature. 2010;464:104–107.
REFERENCES 28. van Golen RF, van Gulik TM, Heger M. The sterile immune response dur-
1. Leichtman AB. Balancing efficacy and toxicity in kidney-transplant immuno- ing hepatic ischemia/reperfusion. Cytokine Growth Factor Rev. 2012;23:69–
suppression. N Engl J Med. 2007;357:2625–2627. 84.
2. Morard I, Mentha G, Spahr L, et al. Long-term renal function after liver 29. Peralta C, Jimenez-Castro MB, Gracia-Sancho J. Hepatic ischemia and reper-
transplantation is related to calcineurin inhibitors blood levels. Clin Transplant. fusion injury: effects on the liver sinusoidal milieu. J Hepatol. 2013;59:1094–
2006;20:96–101. 1106.
3. Toso C, Cader S, Mentha-Dugerdil A, et al. Factors predicting survival after 30. Wang Y, Wu J, Jiang B, et al. Relationship between ischemia/reperfusion
post-transplant hepatocellular carcinoma recurrence. J Hepatobiliary Pancreat injury and acute rejection of allogeneic liver transplant in rats. Transplant
Sci. 2013;20:342–347. Proc. 2014;46:50–55.
4. Gutierrez-Dalmau A, Campistol JM. Immunosuppressive therapy and ma- 31. Ishii E, Shimizu A, Kuwahara N, et al. Hepatic artery reconstruction pre-
lignancy in organ transplant recipients: a systematic review. Drugs. vents ischemic graft injury, inhibits graft rejection, and mediates long-
2007;67:1167–1198. term graft acceptance in rat liver transplantation. Transplant Proc. 2013;45:
1748–1753.
5. 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. 32. Shimizu A, Ishii E, Kuwahara N, et al. Chronic antibody-mediated responses
may mediate chronic rejection in rat orthotopic liver transplantation. Transplant
6. Schlegel A, Rougemont O, Graf R, et al. Protective mechanisms of end-
Proc. 2013;45:1743–1747.
ischemic cold machine perfusion in DCD liver grafts. J Hepatol. 2013;58:278–
286. 33. Vincenti F. Immunosuppression minimization: current and future trends in
transplant immunosuppression. J Am Soc Nephrol. 2003;14:1940–1948.
7. Schlegel A, Graf R, Clavien PA, et al. Hypothermic oxygenated perfusion
(HOPE) protects from biliary injury in a rodent model of DCD liver transplan- 34. Barbier L, Garcia S, Cros J, et al. Assessment of chronic rejection in liver graft
tation. J Hepatol. 2013;59:984–991. recipients receiving immunosuppression with low-dose calcineurin inhibitors.
J Hepatol. 2013;59:1223–1230.
8. Dutkowski P, Graf R, Clavien PA. Rescue of the cold preserved rat liver by
hypothermic oxygenated machine perfusion. Am J Transplant. 2006;6:903– 35. Knolle PA, Thimme R. Hepatic immune regulation and its involvement in viral
912. hepatitis infection. Gastroenterology. 2014;46:1193–1207.
9. Hu J, Wang Z, Tan CJ, et al. Plasma microRNA, a potential biomarker for acute 36. Vivarelli M, Dazzi A, Cucchetti A, et al. Sirolimus in liver transplant recipients:
rejection after liver transplantation. Transplantation. 2013;95:991–999. a large single-center experience. Transplant Proc. 2010;42:2579–2584.
10. Qin L, Guan HG, Zhou XJ, et al. Blockade of 4-1BB/4-1BB ligand interac-
tions prevents acute rejection in rat liver transplantation. Chin Med J (Engl).
2010;123:212–215. DISCUSSANTS
11. Wu YB, Huang QM, Liu JF. Establishment of a chronic rejection model for J. Lerut (Brussels, Belgium):
orthotopic liver transplantation in the rat. Eur Surg Res. 2011;46:94–101.
12. Cheng J, Zhou L, Jiang JW, et al. Proteomic analysis of differentially ex-
Thank you very much for your elegant presentation and
pressed proteins in rat liver allografts developed acute rejection. Eur Surg Res. nice article on the potential immunologic benefit of hypothermic
2010;44:43–51. oxygenated machine perfusion in liver transplantation. This work is a


C 2014 Lippincott Williams & Wilkins www.annalsofsurgery.com | 937

Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Schlegel et al Annals of Surgery r Volume 260, Number 5, November 2014

good example of a well-conducted experimental work in the field of DISCUSSANTS

small animal liver transplantation, going 7 steps further from a syn-
geneic control group to a HOPE-tacrolimus group. Scientific work M. Krawczyk (Warsaw, Poland):
on the minimization of immunosuppression is especially welcomed Professor Dutkowski, I appreciated your article very much.
in an era of extended criteria donors and a shift of interest within the Your excellent study is very important for the future of liver trans-
transplant community to long-term survival. plantation, as the shortage of donors is a problem for most transplant
In the “low-dose tacrolimus and HOPE” group, acute rejection centers.
and fibrosis were markers of chronic rejection and were eliminated in Here are my 2 questions:
a model on nonarterialized liver transplantation. First, you say that for these types of liver recipients, you would
In summary, all your findings come down to a well-known recommend the use of hypothermic oxygenated machine perfusion.
relationship between the degree of the I/R injury and the incidence of Do you mean that it would be possible to use it for urgent recipients?
immunologic events, such as rejection. This link was a substantial part If we have an urgent recipient, at the time of transplantation, this is
of the research, conducted by Prof Tilney, from Boston. As HOPE, really crucial. We know that the use of HOPE is time-consuming. Do
through the impact on the mitochondrial function, downregulates the you share my opinion that a benefit of HOPE has a higher effect than
I/R injury and the innate immune response, it was logical to also the time of the transplantation?
search for a possible downregulation of adaptive immunologic events Second, I would like to ask you, whether HOPE could be
in your experimental setting. used for blood group incompatibility in liver transplantation. From
You concluded that the combination of HOPE and low-dose personal experience, I know that the number of acute rejections, in
tacrolimus eliminated rejection and fibrosis. It must, however, be this group of patients, is higher than in blood group compatibilities
said that it is nowadays possible, in clinical practice, to obtain ex- in liver transplantation.
cellent (immunologic) outcome under minimal immunosuppression. Once again, congratulations on your excellent presentation. I
It could also be possible that the unfavorable immunologic events will be pleased to hear the answers to my 2 questions.
you documented in the non-tacrolimus groups were reinforced by the
nonarterialized model of liver transplantation. So, a question arises, Response From P. Dutkowski (Zurich, Switzerland):
pertaining to the real impact of HOPE on your observations. There- Thank you very much, Prof Krawczyk, for these good ques-
fore, I have 2 questions in relation to this. tions. In fact, applying HOPE is possible without losing time, as the
The first relates to the use of a nonarterialized rat model of liver oxygenated perfusion is done during recipient hepatectomy, which
transplantation. Indeed, the Tokyo Shimizu group showed that T-cell makes this method very simple.
infiltration and immunoglobulin deposition, in sync with antibody- The question, regarding blood group incompatibility, is dif-
mediated rejection, are much more pronounced in the nonarterialized ficult to address at the moment, because our study is the first to
liver transplantation model. So, in the arterialized model, the effect of suggest an effect on the immune response, by a simple ex vivo graft
HOPE could be less pronounced. Do you intend to add an arterialized treatment, before transplantation. Whether this technique is strong
model of liver transplantation to your already extensive number of enough to avoid rejection in incompatible livers requires further
models? investigation.
My second question, which is of great significance, relates to
a more in-depth explanation of the underlying immunologic mecha- DISCUSSANTS
nisms of your findings. HOPE alone postpones the rejection and trans-
forms this event into a chronic add-on; even adding small doses of R. Adam (Villejuif, France):
tacrolimus to HOPE may inhibit the B-cell response. It is, thus, neces- I congratulate you and your team for acquiring further evidence
sary to observe the production of anti-HLA antibodies to confirm the for the advantage of machine perfusion on the liver.
immunologic protective effect of HOPE. Indeed, accelerated fibrosis I have one, simple question: Do you think that the minimization
may be linked to antibody production, with the following activa- of ischemic injury to the liver directly influenced this type of effect
tion of the complement cascade and innate immunity (macrophages, on the minimization of allograph rejection, or is this a completely
monocytes, etc). Did you have the opportunity to observe the im- independent factor?
mune B response in your research, and if so, what was the result? Second, to demonstrate that oxygen is key in the minimiza-
This information is absolutely imperative, before one can claim that tion of the graph rejection, have you done or do you plan to do an
HOPE alone is able to downregulate the immune response of the liver experiment with HOPE, without oxygen?
recipient.
Response From P. Dutkowski (Zurich, Switzerland):
Response From P. Dutkowski (Zurich, Switzerland): Thank you very much, Prof Adam, for your kind comments.
Thank you very much Prof Lerut for these kind comments and Let me begin with your last question. We included experiments with
good questions. I would like to return to your first point, regarding an deoxygenated perfusion, showing that the effect on the immune re-
arterialized transplant model. We chose, on purpose, a nonarterialized sponse was dependent on oxygen in the perfusate, consistent with
rat liver transplantation, because, as you also mentioned, rejection is earlier studies on reperfusion injury.
known to occur to a higher degree without arterialization. To test the Indeed, as we minimized liver ischemia in this model of
maximum effect of HOPE, we, therefore, opted for a nonarterialized the allogeneic rat liver transplantation, we would expect a higher
transplantation in this study and showed a strong effect. reperfusion injury and also a higher immune response in grafts
The second question focuses on the effect of HOPE on the in- exposed to additional ischemia. Whether HOPE is also effective,
direct immune pathway. At this stage, we cannot say something about for example, in allogeneic donated after cardiac death livers or
the effect of HOPE on B-cell response and downstream mediators. in allogeneic liver grafts with long cold ischemia requires further
This issue should be addressed by further studies. investigation.

938 | www.annalsofsurgery.com 
C 2014 Lippincott Williams & Wilkins

Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.