Note: Descriptions are shown in the official language in which they were submitted.
CA 022~3313 1998-10-28
WO 97/40849 PCT/US97/06809
METHOD FOR TREATING OR PREVENTING
ISCHEMIA-REPERFUSION INJURY
BACKGROUND OF THE INVENTION
Ischemia-reperfusion injury frequently occurs when the flow of blood to a
region of the body is temporarily halted (ischemia) and then re-established
(reperfusion). Ischemia-reperfusion injury can occur during certain surgical
procedures, such as repair of aortic aneurysms and organ transplantalion.
Clinically ischemia-reperfusion injury may be manifested by such complications
20 as pulmonary dysfunction, including adult respiratory distress syndrome, renal
dysfunction, consumptive coagulopathies including thrombocytopenia, fibrin
deposition into the microvasculature and disseminated intravascular
coagulopathy, transient and permanent spinal cord injury, cardiac arrhythmias
and acute ischemic events, hepatic dysfunction including acute hepatocellular
25 damage and necrosis, gastrointestinal dysfunction including hemorrhage and/or infarction and multisystem organ dysfunction (MSOD) or acute systemic
inflammatory response syndromes (SIRS). The injury may occur in the parts of thebody to which the blood supply was interrupted, or it can occur in parts fully
supplied with blood during the period of ischemia.
International Patent Publication No. WO 96/01318 relates to polypeptides
other than interleukin -10 (IL-10) allegedly having one or more properties similar
to those of IL-10. Among the very long list of diseAces allegedly treatable withthese non-lL-10 proteins are tissue damage as a result of "hypoxia/ischemia
35 (infarction: reperfusion)", "ischemia", "reperfusion injury", and "reperfusion
syndrome". I lowevcr, there is no evidence in this publication that the non-lL-10
proteins would actually work for treating all of the diseases in the long list.
CA 022~3313 1998-10-28
W O 97/40849 rcTrusg7/o6809
-2-
SUMMARY OF THE INVENTION
The present invention comprises a method for treating ischemia-
reperfusion injury comprising administering to a patient in need of such treatment
5 an effective amount of IL-10. Another aspect of this invention comprises a method
for preventing ischemia-reperfusion injury in a patient about to undergo a
procedure capable of causing ischemia-reperfusion injury or to a patient who hasalready undergone such procedure in which ischemia-reperfusion injury has not
yet occurred comprising administering to the patient an effective amount of IL-10.
Preferred applications of this invention are preventing ischemia reperfusion
injury by administering the IL-10 in conjunction with surgical repair of the thoracic
or suprarenal aorta due to aneurysmal disease, but also in conjunction with those
surgical procedures that induce or require transient occlusion or bypass of the
15 visceral blood supply via the hepatic, renal and/or enteric arteries secondary to
major organ transplant, including liver, kidney, small intestine, and pancreas as
well as surgical procedures that result in the transient reduction or prevention of
blood flow to the viscera including hepatic and biliary surgical resections, total or
partial pancreatectomy (Whipple procedure), total and partial gastrectomy,
20 esophagectomy, colorectal surgery, vascular surgery for mesenteric vascular
cliseAce, or abdominal insufflation during laparoscopic surgical procedures.
Additional Applic~tions include blunt or penetrating trauma that results in
interrruption of blood flow to the visceral organs including those arising from
25 penetrating wounds to the abdomen resulting from gun shot wounds, stab wounds or from penetrating wounds or blunt abdominal trauma secondary to
deacceleration injury and/or motor vehicle accidents. Other preferred applications
include diseases or procedures that result in systemic hypotension that eithe
disrupts or decreases the flow of blood to the visceral organs, including
30 hemorrhagic shock due to blood loss, cardiogenic shock due to myocardial
infarction or cardiac failure, neurogenic shock or anaphylaxis.
Further A~F'~ ions of this invention include preventing or treating lower
torso or extremity ischemia reperfusion injury by administering IL-10 in
35 conjunction with surgical procedures that induce or require transient occlusion or
bypass of the blood supply to the torso or the upper or lower extremities. This
application is particularly relevant to the practice of vascular surgery that
encompasses controlled periods of visceral, torso, and extremity ischemia
CA 022~3313 1998-10-28
WO 97/40849 PCT/US97/06809
-3-
f~llDwEd by reperfusion. Procedures which involve such ischemia-reperfusion
include but are not limited to repair of abdominal aortic aneurysms, aortic femoral,
popliteal or tibial bypass for cl~llcliG~tion or limb threatening ischemia, repair of
popliteal or femoral aneurysms, bypass, thrombectomy or embolectomy for acute
5 limb ischemia, or vascular trauma. Administration of IL-10 may improve limb
salvage and survival after significant torso or extremity ischemia.
The amount of IL-10 to be administered is preferably between 0.1 to 500
~lg/kg of body weight, more preferably 1 to 50 ~g/kg. The IL-10 may be of human
10 or viral origin produced biologically from mammalian cellular sources or by
reco",binant DNA technology. Adn~i,)isl~lion preferably takes place by
intravenous, intramuscular or subcutaneous injection. The IL-10 is preferably
administered from one to zero hours before the blood flow is reestablished.
In those surgical procedures in which temporary or sustained disruption of
blood flow is anticipated to occur, as before surgical repair of thoracoabdominal or
supraceliac aneursymal disease, or surgical procedures to the abdomen that will
necessarily include the transient reduction in visceral blood flow, or for organtransplantation, the IL-tO is preferably given either as a single bolus injection one
to zero hours before the ischemic event or as a continuous intravenous injectionbeginning one to zero hours before the ischemic event and extending during the
perioperative period and continuing for at least eight hours after restoration of
visceral blood flow.
For individuals in whom disrupted visceral blood flow has already occurred,
as in those individuals with trauma or injury to the visceral organs or their blood
supply, or in patients with systemic hypotension due to shock, the IL-10 would be
preferably given either as a single bolus injection prior to or simultaneously with
restoration of normal visceral blood flow or as a continuous intravenous injection
prior to or simultaneously with restoration of normal visceral blood flow and
extending for at least eight hours after restoration of visceral blood flow.
For individuals in whom disrupted skeletal blood flow has already occurred,
as in those individuals with acute lower extremity ischemia due to embolic or
thrombotic occlusion of peripheral blood vessels or acute ischemia due to
vascular trauma, the IL-10 would be preferably given either as a single bolus
injection prior to or simultaneously with restoration of nommal blood flow or as a
continuous intravenous injection prior to or simultaneously with restoration of
~ . . . .
CA 022~3313 1998-10-28
W 097/40849 PCT~USg7/06809 -4-
normal blood flow and extending for at least eight hours after restoration of blood
flow.
Altematively, the IL-10 may be administered by gene therapy or transfer
5 using either liposomes and mammalian expression plasmids, mechanical delivery
systems (gene gun) of viral transfection schemes, including but not limited to
adenovirus, adeno-associated virus, retrovirus or herpes simplex virus constructs.
BRIFF DESCRIPTION OF THE DRAWINGS
Figures 1 (a), 1 (b) and 1 (c) illustrate the plasma TNF-a, IL-1~ and IL-8
concentrations"especti~/ely, following thoracoabdominal and infrarenal aortic
aneurysm repair.
Figures 2(a~, 2(b) and 2(c) illustrate the plasma IL-6, changes in plasma
p55 concenl~dlions, and changes in plasma p75 concelllldlions~ respectively,
following thoracoabdominal and infrarenal aortic aneurysm repair.
Figure 3 illustrates the changes in lung myeloperoxidase levels (neutrophil
ir,~ill,~lion) in mice following supraceliac aortic cross clamp and treatment with
inhibitors of TNF and IL-1.
Figure 4 illustrates the changes in lung perrneability (1251-albumin
leakage) in mice following supraceliac aortic cross clamp and treatment with
inhibitors of TNF and IL-1.
Figure 5, which illusllates the appearance of IL-10 in the circulation of mice
following supraceliac aortic cross clamp used, shows plasma IL-10 concer,l,alions
in mice following supraceliac aortic cross-clamping and treatment with
recol"binant human IL-10.
Figure 6 illustrates the changes in lung myeloperoxidase levels (neutrophil
infiltration) in mice following supraceliac aortic cross clamp and treatment with
recombinant human IL-10.
RE( 1 l~ ;1 (RULE 91)
ISAIEP
CA 022~3313 1998-10-28
W O 97/40849 PCT~US97/06809
-5-
DETAILED DESCRIPTION OF THE INVENTION
All references cited herein are hereby incorporated in their entirety by
reference.
As used herein, "interleukin-10" or "IL-10" is defined as a protein which (a)
has an amino acid sequence of mature IL-10 (e.g., lacking a secretory leader
sequence) as disclosed in U.S. Patent No. 5,231,012 and (b) has biological
activity that is common to native IL-10. Also included are muteins and other
10 analogs, including the Epstein-Barr Virus protein BCRF1 (viral IL-10), which retain
the biological activity of IL-10.
IL-10 suitAhlQ for use in the invention can be obtained from culture medium
conditioned by activated cells secreting the protein, and purified by standard
methods. Additionally, the IL-10, or active fragments thereof, can be chemically15 synthesized using standard techniques known in the art. See Merrifield, Science
233:341 (1986) and Atherton et al., Solid Phase Peptide Synthesis: A Practical
Approach, 1989, I.R.L. Press, Oxford. See also U.S. Patent No. 5,231,012.
Preferably, the protein or polypeptide is obtained by reco",binant
techniques using isolated nucleic acid encoding the IL-10 polypeptide. General
20 methods of molecular biology are described, e.g., by Sambrook et aL, Molecvlar
Cloning, A Laboratory Man~al, Cold Spring Harbor, New York, 2d ed., 1989, and
by Ausubel et al., (eds.) Current Protocols in Molecular Biology, Green/Woley,
New York (1987 and periodic supplements). The appropriate sequences can be
obtained using standard techniques from either genomic or cDNA libraries.
25 Polymerase chain reaction (PCR) techniques can be used. See, e.g., PCF~
Protocols: A Guide to Methods and Applications, 1990, Innis et al., (Ed.), Academic
Press, New York, New York.
Libraries are constructed from nucleic acid extracted from appropriate cells.
See, e.g., U.S. Patent No. 5,231,012, which ~iscloses reco",binant methods for
30 making IL-10. Useful gene sequences can be found, e.g., in various sequence
dAtAh:~ses, e.g., GenBank and BMPL or nucleic acid and PIR and Swiss-Prot for
protein, c/o Intelligenetics, Mountain View, California, or the Genetics Computer
Group, University of Wisconsin Biotechnology Center, Madison, Wisconsin.
Clones comprising sequences that encode human IL-10 have been
35 deposited with the American Type Culture Collection (ATCC), Rockville, Maryland,
under Accession Nos. 68191 and 68192. Identification of other clones harboring
CA 022~3313 1998-10-28
W 097/40849 PCT~US97/06809 -6-
the sequences encoding IL-10 is performed by either nucleic acid hyl,ridi~alion or
immunological detection of the encoded protein, if an expression vector is used.Oligonucleotide probes based on the deposited sequences disclosed in U.S.
Patent No. 5,231,012 are particularly useful. Oligonucleotide probes sequences
5 can also be prepared from conserved regions of related genes in other species.Alternatively, degenerate probes based on the amino acid sequences of IL-10 can
be used.
Standard methods can be used to produce transformed prokaryotic,
mammalian, yeast or insect cell lines which express large quantities of the
10 polypeptide. Exemplary E. coli strains suitable for both expression and cloning
include W3110 (ATCC Bi, 27325), X1776 (ATCC No. 31244). X2282, and RR1
(ATCC Mp/ 31343). Exemplary mammalian cell lines include COS-7 cells, mouse
L cells and CHP cells. See Sambrook (1989), supra and Ausubel etal., 1987
supplements, supra.
Various expression vectors can be used to express DNA encoding IL-10.
Conventional vectors used for expression of recombinant proteins in prokaryotic
or eukaryotic cells may be used. Preferred vectors include the pcD vectors
described by Okayama et al., Mol. Cell. Biol. 3:280 (1983); and Takebe et al., Mol.
Cell. Biol. 8:466 (1988). Other SV40-based mammalian ex~r~ssion vectors
20 include those disclosed in Kaufman et al., Mol. Cell. Biol. ~.1304 (1982) and U.S.
Patent No. 4,675,285. These SV40-based vectors are particularly useful in COS-
7 monkey cells (ATCC No. CRL 1651), as well as in other mammalian cells such
as mouse L cells. See also, Pouwels etal., (1989 and supplements) Cloning
Vectors: A Laboratory Manual, Elsevier, New York.
The IL-10 may be produced in soluble form, such as a secreted product of
transfommed or transfected yeast, insect om,lar,l",alian cells. The peptides canthen be purified by standard procedures that are known in the art. For example,
pu,i~ic~liol1 steps could include ammonium sulfate precipil~lion, ion exchange
chromatography, gel filtration, electrophoresis, affinity chromatography, and the
like. See Methods in Enzymology Purification Principles and Practices (Springer-Verlag, New York, 1982).
Alternatively, IL-10 may be produced in insoluble form, such as aggregates
or inclusion bodies. The IL-10 in such a form is purified by standard proceduresthat are well known in the art. Examples of purification steps include separating
the inclusion bodies from disrupted host cells by centrifugation, and then
CA 022~3313 1998-10-28
W O 97/40849 PCTrUS97/06809
-7-
solubilizing the inclusion bodies with chaotropic agent and reducing agent so that
the peptide assumes a biologically active conformation. For specifics of these
procedures, see, e.g. Winkler et al., Biochemistry 25:4041 (1986), Winkler et al.,
Bio/Technology 3:9923 (1985); Koths etal., and U.S. Patent No. 4,569,790.
The nucleotide sequences used to transfect the host cells can be modified
using standard techniques to make IL-10 or fragments thereof with a variety of
desired properties. Such modified IL-10 can vary from the naturally-occurring
sequences at the primary structure level, e.g., by amino acid, insertions,
suhstihltions, deletions and fusions. These modifications can be used in a
1 0 number of combi. ~alions to produce the final modified protein chain.
The amino acid sequence variants can be prepared with various objectives
in mind, including increasing serum half-life, facilitating purification or preparation,
improving therapeutic efficacy, and lessening the severity or occurrence of sideeffects during therapeutic use. The amino acid sequence variants are usually
predetermined variants not found in nature, although others may be post-
translational variants. Such variants can be used in this invention as long as they
retain the biological activity of IL-10.
Modifications of the sequences encoding the polypeptides may be readily
accomplished by a variety of techniques, such as site-directed mutagenesis
(Gillman etal., Gene 8:81 (1987)). Most modificalions are evaluated by routine
screening in a suitable assay for the desired characteristics. For instance, U.S.
Patent No. 5,231,012 describes a number of in vitro assays suitable for measuring
IL-10 activity.
Preferably, human IL-10 is used for the treatment of humans, although viral
IL-10 could possibly be used. Most preferably, the IL-10 used is recombinant
human IL-10. The preparation of human IL-10 has been described in U.S. Patent
No. 5,231,012. The cloning and expression of viral IL-10 (BCRF1 protein) from
Epstein-Barr virus has been disclosed by Moore et al., Science 248:1230 (1990).
For examples of procedures and assays to determine IL-10 activity, see
United States Patent No. 5,231,012. This patent also provides proteins having IL-
10 activity and production of such proteins including recombinant and synthetic
techniques.
To prepare pharm~ceutic~l compositions of IL-10 for practice of this
invention, the IL-10 is admixed with a pharmaceutically acceptable carrier or
CA 022~3313 1998-10-28
WO 97/40849 PCT/US97/06809
-8-
excipient which is preferably inert. A phammaceutical carrier can be any
compatible non-toxic substance sllitAhle for delivery of the polypeptide to a
patient. Preparation of such pharmaceutical compositions is known in the art; see,
e.g., Remington's Pharmaceutical Sciences, and U.S. Pharmacopeia: National
5 Formulary, Mack Publishing Company, Easton, PA (1984).
Compositions may be ingested orally or injected into the body.
Formulations for oral use include compounds to protect the polypeptides from
proteases which occur in the ga~ lestinal tract. Injections are usually
intramuscular, subcutaneous, intradermal or intravenous. Altematively, intra-
10 articular injection or other routes could be used in appropriate circumstances.
When administered parenterally, the compositions can be formulated in aunit dos~ge injectable form (solution, suspension, emulsion) in association with a
pharmaceutical carrier. For instance, the IL-10 may be administered in aqueous
vehicles such as water, saline or buffered vehicles with or without various
15 additives and/or diluting agents. Examples of suit~hle carriers are normal saline,
Ringer's solution, dextrose solution, and Hank's solution. Non-aqueous carriers
such as fixed oils and ethyl oleate may also be used. A preferred carrier is 5%
dextrose/saline. The carrier may contain minor amounts of additives such as
substances that enhance isotonicily and chemical stability, e.g., buffers and
20 preservatives. However, the IL-10 in the composition is preferably formulated in
purified fomm subslanlially free of aggregates and other proteins. In addition, it
should be noted that a suspension, such as a zinc suspension, can be prepared toinclude the polypeptide. Such a suspension can be useful for subcutaneous (SQ)
or intramuscular (IM) injection.
It is believed that ischemia-reperfusion injury is caused, at least in part, by
the release of excess amounts of proinflammatory cytokines, such as TNF-a, IL-1,IL-6, and IL-8. Examples 1 and 2 were performed to test this theory and the effect
IL-10 has on visceral ischemia-reperfusion injury. Example 3 sets forth the
30 ~pplic~tion of the invention to a human patient undergoing aortic aneurysm repair.
Example 4 was performed to demonstrate that IL-10 will attenuate the skeletal
muscle and pulmonary injury after hindlimb ischemia-reperfusion in a rat model.
EXAMPLE 1
Initial studies investigated prospectively the associative relationship
between the proinflammatory cytokine response and morbidity and mortality
CA 022~3313 1998-10-28
W097/40849 rcTrusg7/o6809
_9_
following visceral ischemia and reperfusion in humans by measuring
proinflammatory cytokine levels in patients undergoing thoracoabdominal or
infrarenal aortic aneurysm repair, and comparing these results to the incidence of
postoperative organ dysfunction.
Sixteen human patients undergoing elective repair of a thoracoabdominal
aortic aneurysm and 9 patients undergoing elective infrarenal aortic aneurysm
repair agreed to arterial blood sampling for proinflammatory cytokine
msasurements. Each thoracoabdominal aortic aneurysm was repaired through a
10 left flank incision using a retroperitoneal approach. The diaphragm was divided
circumferentially, allowing exposure of the descending thoracic aorta. Prior to
cross-clamping, each patient was given mannitol (0.5 gm/kg) and solumedrol (15
mg/kg). Depending upon the location of the aneurysm, the visceral arteries were
sewn onto the graft as a Carrel patch or as part of the proximal anastomosis with
15 an extensive posterior taper to the graft. Once the repair was completed,
coagulation products (platelets and fresh frozen plasma) were infused as needed.Preoperatively, a catheter was placed in the lumbar spinal column and
cerebrospinal fluid drained to maintain intrathecal pressure at 5-10 cm water.
Infrarenal abdominal aortic aneurysms were repaired transperitoneally using
20 standard surgical tecl",i~ues and the aorta was reconstructed using either a
straight tube graft to the aortic bifurcation or a bifurcated graft to the
internal/external iliac artery bifurcation.
In both groups of patients, arterial blood samples (7 ml) were obtained
25 fo"~ ing induction of anesthesia, just prior to aortic cross-clamp place",enl, just
prior to clamp release, and at timed intervals (1, 2, 4, 6 to 8, 24 hrs and daily for 7
days) after reperfusion. Clinical and laboratory data were colleGted prospectively
from all patients to determine preoperdlive risk factors and postoperative organdysfunction pattems. Data collected included operative parameters (total
30 operative time, aortic cross-clamp time, estimated blood loss, intraoperativecomplications), postoperative course (complications, organ dysfunction) and
causes of death. Laboratory values were analyzed during the initial 7
postoperative days to focus on the injury associated with tissue ischemia-
reperfusion after thoracoabdominal and infrarenal aortic aneurysm repair.
Postoperative pulmonary dysfunction was defined as the need for positive-
pressure mechanical ventilatory assistance for greater than 7 days while
postoperative hepatic dysfunction was defined as peak lactate dehydrogenase
CA 022~3313 1998-10-28
W 097/40849 PCT~US97/06809
-10-
(LDH) levels greater than 500 U/L and either serum transaminase levels
(AST/ALT) greater than 200 U/L or an increase in total bilirubin levels greater than
3 mg/dl. Renal dysfunction was defined as an increase in serum creatinine of 2
mg/dl or more over preoperative baseline, while a platelet count less than
50,000/mm3 or a drop in white blood cell count below 4,500 /mm3 indicated the
presence of hematopoietic dysfunction. Patients with 2 or more organ systems
meeting these criteria were designated as having multiple system organ
dysfunction [MSOD].
1 0 Freshly thawed plasma samples were assayed for TNF-a, IL-1, IL-6, IL-8
and TNF-oc shed receptors (p55 and p75) by ELISA. The sensitivity of the TNF-a,
IL-1, IL-6, IL-8, p55 and p75 assays are 14,10, 27, 313, 14 and 17 pg/ml,
respectively.
1 5 The mortality and morbidity data from the 16 patients undergoing
thoracoabdominal aortic aneurysm repair and the 9 patients undergoing infrarenal aortic aneurysm repair are reported in Table 1.
TABLE 1
Incidence of organ dysfunction following thoracoabdominal or infrarenal
aortic aneurysm repair.
Data presented shows that the frequency of pulmonary dysfunction and
MSOD following thoracoabdominal aortic aneurysm repair was significantly
higher than following abdominal aortic aneurysm repair.
Thoracoabdominal Infrarenal
Aortic Aneurysm Aortic
(n=16) Aneurysm
(n=9)
Mortality 19% 0%
Pulmonary Dysfunction 56%* 11 %
Tracheostomy 25% 0%
Renal Dysfunction 38% ** 0%
Dialysis 13% 0%
Hepatic Dysfunction 31% 0%
Hematopoietic Dysfunction 38% ** 0%
Leukopenia 13% 0%
MSOD 44%* o%
* p<0.05 by Fisher's exact test
** p=0.057 by Fisher's exact test
Three patients died after thoracoabdominal aortic aneurysm repair, 2 from
40 MSOD and 1 from cardiac arrest. Pulmonary dysfunction occurred in 9 patients
CA 022~3313 1998-10-28
W O 97/40849 PCTAUS97/06809
-1 1-
and placement of a temporary tracheostomy was eventually required in 4 patients.Renal dysfunction developed in 6 patients and hemodialysis was necessary in 2
~ of them. Hepatic dysfunction, thrombocytopenia, and leukopenia developed after
thoracoabdomii-al aortic aneurysm repair in 5, 6, and 2 patients, respectively, and
5 lower exl~elllily dysfunction due to spinal cord injury occurred in 2 patients. In
contrast, there were no operative deaths after infrarenal aortic aneurysm repair(Table 1). Pulmonary dysfunction occurred in only 1 patient and there was no
cvidence of renal, hepatic, hematopoietic or lower ekl~ ily dysfunction in any
patient.
The peak plasma cytokine responses in both groups of patients are
reported in Table 2.
TABLE 2
Peak proinflammatory cytokine concentrations following thoracoabdominal
or infrarenal aortic aneurysm repair.
Plasma samples were obtained 0, 1, 2, 4, 6-8, 24, 48, 72 hours and daily for
up to seven days fcllo- ;ng thoracoabdominal or infrarenal aortic aneurysm repair.
Peak conce~ alions are reported here. Levels of all proinflammatory cytokine
20 were significantly higher in patients following thoracoabdominal than infrarenal
aortic aneurysm repair (p<0.05).
Thoracoabdominal Infrarenal
Aortic Aortic
Aneurysm Aneurysm
(n=16) (n=9)
TNF-a pgs/ml 161i58 10_10
IL-1 b pgs/ml 133+59 24+10
IL-6, pgs/ml 1,280i664 181+108
IL-8, pgslml 410+139 137+77
pS5, change from baseline in pgs/ml751i668 204+21 8
p75, change from baseline in pgs/ml5,201+1,983 383+171
C3a, llg/ml 111+~1 30~7
all values are significantly different between the two groups, by two-way ANOVA,p<0.05
Plasma TNF-a IL-1, IL-6 and IL-8 concer~l~dlions were undetectable prior to
35 surgery. Following surgical repair of thoracoabdominal aortic aneurysms, a
monophasic TNF-a response was detected in 11 of 16 pdtientS (69%) (Figures
1 (a), 1 (b) and 1 (c)). TNF-a levels peaked 4 hours after reperfusion and then
gradually decreased toward baseline over the next 24 hours. lL-6 and IL-8 levelsalso increased in a monophasic pattem with peak levels again occurring 4 hours
RE~'ll~l~ SEIEET ~RULE 91)
ISAIEP
CA 02253313 1998-10-28
WO 97/40849 PCT/US97/06809
-12
after reperfusion in 16 (100%) and 11 (70%) patients, respectively; however,
unlike the pattem seen with TNF-~, circulating IL-6 and IL-8 levels decreased tobaseline within 8 hours. IL-1 was also detected in a monophasic pattem in 50% ofthe thoracoabdominal aortic aneurysm patients, but its peak levels occurred at 1hour after reper~usion and IL-1 levels retumed to baseline levels 4-6 hours a~ter
reperfusion. The plasma co(,centlations of the soluble TNF-a receptors, p55 and
p75, were increased after thor~co~hdominal aortic aneurysm repair in 12(75%)
and 16 (100%) of the patients assayed, respectively (Figure 2). p55leceptor
concenlI~Iions reached a zenith at 24 hours and remained elevated for several
10 days while p75 receptor conce~ lions continued to increase throughout the
initial 48 hours after reperfusion. In contrast to thoracoabdominal aortic aneu~sm
repair patients, peak serum levels of TNF-a, IL-1, IL-6, IL-8, p55 and p75 were 3
to 15-fold less in patients undergoing infrarenal abdominal aortic aneurysm repair
(Table 2 and Figures 1 (a), 1 (b), 1 (c) and 2(a), 2(b) and 2(c)).
A retrospective analysis was performed in an effort to establish an
associati~/e relatiGnship between patient clinical outcome and the conce"lraliol,s
of various proinflammatory cytokines. Patients undergoing thoracoabdominal
aortic aneurysm repair in whom peak TNF-a levels were less than 150 pg/ml did
20 not experience single or multiple organ dysfunction, while single organ
dysfunction and MSOD were common in patients whose peak TNF-a levels were
greater than 150 pg/ml (Table 3).
TABLE 3
Relationship between post-operative organ dysfunction and peak
circulating TNF-a levels.
TNF-~c c150 pgs/ml TNF-a ~150
pgs/ml
MortaJity 1 death cardiac 2 deaths - MSOD
Pvlmonary Dysfunction 0% 57%~*
Renal Dysfunction 0% 71% *
Dialysis 0% 29%
Hepatic Dysfunction 0% 71% *
Hematopoietic Dysfunction 0% 71% *
Leukopenia 0% 28%
MSOD 0% 86%*
* p<0.05 by Fisher's exact test
**p=0.07 by Fisher's exact test
RE~ l l~'l~;l~ SHEET (RULE 91)
ISA/EP
CA 022~33l3 l998-l0-28
W O 97/40849 PCT~US97/06809
-13-
ln addition, patients who developed MSOD after thoracoabdominal aortic
~ aneurysm repair had higher circulating levels of all assayed cytokine and soluble
TNF-a receptors (p55 and p75) as compared to patients without MSOD (Table 4);
5 however, only TNF-a and p55 receptor levels were sl~lislicAlly different (pcO.05)
while there was a trend toward higher levels of IL-1, IL-6, IL-8 and p75 receptors in
patients who developed MSOD as compared to patients without MSOD (Table 4).
TABLE 4
Plasma proinflammatory cytokine concenl,~tions in patients with and
without evidence of multisystem organ dysfunction (MSOD).
Peak plasma co"cenl~dlions of TNF, IL-6, p55 and p75 were significantly
higher in patients following thoracoabdominal aortic aneurysm repair with MSOD
than in patients either following thoracoabdominal aortic aneurysm repair without
15 MSOD or in patients following infrarenal aortic aneurysm repair.
Thoracoabdominal Thoracoabdominal Infrarenal Aortic
Aortic Repair with Aortic Repair w/o Repair
MSOD MSOD
cross- 56+5 mins 33+4 mins nr
clamp time
TNFa 414+5g~ 86i55 1 Oi10
L- ~ 173i112 02+ 2 '4= 0
- 4,907+1887~ '4~+ 6 8 -108
- 601+259 ~ 7 + 07 3, =77
p5, + ,~ 5i711 * +4 2=415 +204+218
p7, + ,4 9+1940~ +4,13 +1,884 +382+171
values for p55 and p75 are changes from baseline. All values are in pgs/ml.
~ p~O.05 versus no MSOD by 2-way ANOVA
nr = not reported
The results pr~sented here der"onsl,ate that surgical repair of
thoracoabdominal aortic aneurysms which causes visceral ischemia-reperfusion
injury results in a systemic proinflammatory cytokine response characterized by
25 the appearance of TNF-a,IL-1,IL-6 and IL-8 in the blood as early as 1 to 4 hours
after release of the cross-clamp. Additionally, the presence and magnitude of this
proinflammatory cytokine response is associated with the incidence of
postoperative organ dysfunction after thoracoabdominal aortic aneurysm repair.
Ischemia and subsequent reperfusion injury of the viscera appear to be critical for
30 the induction of this systemic proinflammatory cytokine response, because themagnitude of the proinflammatory cytokine response is 3 to 15-fold less in patients
undergoing repair of the infrarenal aorta where visceral ichemia/reperfusion does
CA 022~33l3 l998-l0-28
W 0 97/40849 PCT~US97/06809
-14-
not occur than following thoracoabdominal aortic repair. In addition, patients
having infrarenal aortic aneurysm repair, in whom visceral ischemia is avoided,
have a significantly lower incidence of postoperative organ dysfunction.
To further explore the direct role of acute visceral ischemia in mediating this
proinflammatory cytokine response and associated organ dysfunction, an
additional 8 patients were studied following elective thoracoabdominal aortic
aneurysm repair. I lowever, in this case, duration of visceral ischemia was
reduced by left atrial-femoral artery bypass (LAFBP) and retrograde perfusion ofthe visceral arteries. LAFBP provides distal blood flow during repair of
thoracoabdominal aneurysms and reduces visceral ischemia time. We
prospectively examined the effect of LAFBP on patients undergoing
thoracoabdominal aortic repair (n=8) and compared the cytokine response,
morbidity, and mortality to patients undergoing standard thoracoabdominal aorticaneurysm repair (n=16) without the benefit of LAFBP.
Timed measurement of cytokine levels was done during the 48 hour
perioperative period and cytokine levels were measured by ELISA. Clinical data
concerning postoperative pulmonary, hepatic, renal, and hematopoietic
dysfunction were also prospectively collected. Patients undergoing repair of
thoracoabdominal aortic aneurysms with LAFBP had shorter visceral ischemia
times (18_5 min. vs 45+12 min.) and slalislically significant reductions in
circulating TNF-a, IL-10, and p75 levels (p<0.05 by two-way ANOVA) when
compared to the control group (Table 5).
TABLE 5
Plasma proinflammatory cytokine concenl~dlions in patients undergoing
thoracoabdominal aortic aneurysm with left atrial femoral bypass (LAFB) or
without LAFB
Peak plasma concenl,dlions of TNF-a, IL-10 and p75 were siy"ilica"lly
higher in patients following thoracoabdominal aortic aneurysm repair without
LAFB than in patients following thoracoabdominal aortic aneurysm repair with
LAFB.
Repair with Repair W/O LAFBP
LAFBP(n+8) (n=16)
TNF-a, pgs/ml 10_10 161 +58~
IL-6, pgs/ml 2320+1644 1280+664
CA 022~3313 1998-10-28
W O 97/40849 PCT~US97/06809
-15-
IL-8, pgs/ml 410+139 458+402
IL-10, pgs/ml 60+29 806+260~
p55, baseline pgs/ml +741+298 +751+668
p75, baseline pgs/ml +641+406 +5201+1983
p<0.05
Additionally, the incidence of pulmonary dysfunction, renal dysfunction,
5 thrombocytopenia, multisystem organ dysfunction, and mortality were reduced inpatients undergoing LAFBP, although the numbers were too small to show any
s~ lic~l difference.
These findings suggest that acute visceral ischemia-reperfusion injury
10 secondary to thoracoabdominal aortic aneurysm repair is associated with a high
rate of morbidity and multisystem organ dysfunction that is not seen with similar
surgical procedures that do not cause visceral ischemia. Furthermore, techniquesaimed at reducing the duration of ischemia during aortic cross-clamping (left atrial-
femoral bypass) appear to reduce the magnitude of the TNF-a and IL-1
15 responses.
EXAMPLE 2
Ex~,eri~"e~ in mice have been conducted that demon~l,ate that
pret,e~l",ent with recombinant human IL-10 can reduce distant organ injury in a
clinically relevant model of acute visceral ischemia-reperfusion injury. The initial
goal of these studies was to develop a clinically relevant model of acute ischemia-
reperfusion injury that demonstrated evidence of organ injury that was dependentupon an endogenous proinflammatory cytokine response that could be inhibited
by either a TNF-a receptor construct or a monoclonal antibody against the IL-1
type I (p80) receptor (35F5, Hoffmann-LaRoche, Nutley, NJ).
Thirty mice (C57BU6, approx. 20 gm) were anesthetized with pentobarbital.
In 16 of these animals, the supraceliac aorta was cross-clamped for 30 minutes.
Six animals had their infrarenal aorta cross-clamped for 30 minutes, while another
8 animals received only anesthesia, incision and bowel mobilization without aortic
cross-clamping. Two hours prior to supraceliac aortic cross-clamping, 8 of the 16
animals were pretreated with the intraperitoneal injection of 10 mg/kg BW of TNF-
CA 022~3313 1998-10-28
W 097/40849 PCTAUS97/06809
-16-
bp (a TNF-a binding protein that is comprised of the extracellular domains of two
p55 TNF-a leceptors covalently linked to polyethylene glycol). Two hours after
aortic clamp removal and abdominal wound closure, the animals were sacrificed
and lung neutrophil infiltration was evaluated by MPO content. Results are shown5 in Figure 3. Supraceliac aortic cross-clamping resulted in a significant increase in
pulmonary neutrophil infiltration at 2 hours, which was not seen in animals thathad the infrarenal aorta cross-clamped. Pretreatment of the animals undergoing
supraceliac aortic cross-clamping with TNF-bp significantly attenuated this
Increase.
To determine the effect of visceral ischemia-reperfusion on lung capillary
function, 50 mice were anesthetized with pentobarbital, and in 34 animals the
supraceliac aorta was cross-clamped for 30 minutes. Eleven of these animals
were pretreated with TNF-bp (10 mg/kg) while 9 were pretreated with 150 119 of a15 monoclonal antibody directed against the murine IL-1 receptor type I (35F5). It has
been previously reported that this antibody blocks IL-1 binding to the functional IL-
1 type I receptor and attenuates IL-1-mediated inflammation. Control groups
consisted of a sham operation group (n=10) and an infrarenal cross-clamp group
(n=6). After the removal of the aortic cross-clamps and the onset of reperfusion,
20 the animals were injected with 1 IlCi of 1125 labeled albumin i.v. via the inferior
vena cava. At the end of 4 hours of reperfusion the animals were sacrificed and
the lungs were treated with bronchoalveolar lavage (BAL) with 1.75 ml of normal
saline. The pulmonary mean permeability index was c~lcul~ted as the ratio of
CPM/gm BAL over CPM/gm blood. The results are shown in Figure 4. Both
25 pretreatment with TNF-bp and 35F5 decreased pulmonary capillary injury (p
0.05), with 35F5 having a more pronounced effect.
Thus, these findings demo"~lld~e that the lung injury secondary to
supraceliac cross-clamping in the mouse is a result of endogenous production of
30 TNF-a or IL-1. Inhibiting either of these cytokines with novel inhibitors of either
TNF-a or the IL-1 type I receptor can minimize the lung injury secondary to
visceral ischemia-reperfusion injury.
To demonstrate that similar effects can be obtained by immediate
35 pretreatment with recombinant human IL-10, an additional study was conducted in
mice subjected to supraceliac aortic cross-clamping. Visceral ischemia was
induced in 90 female C57BU6 mice (20-22gm) by supraceliac aortic cross-
clamping for 25-30 minutes. An additional 38 mice received sham procedures.
CA 022~33l3 l998-l0-28
W O 97/40849 PCTrUS97/06809
-17-
Plasma IL-10 levels were measured by ELISA at 1, 2, 4 and 8 hrs after
reperfusion, and lung neutrophil infiltration was determined by MPO assay at 2
hrs, as previous studies had revealed that maximal neutrophil infiltration occurred
in the lung at 2 hrs. Thirty-six of the mice undergoing visceral ischemia-
reperfusion were pretreated with 0.2~Lg (n=7), 2~1g (n=13), 5,ug (n=6), or 20~Lg(n=10) of recombinant human IL-10.
Mean plasma IL-10 conce~ lions peaked at 9,120 pg/ml 2 hours following
25-30 minutes of supraceliac aortic cross-cla",ping (Figure 5). Visceral ischemia-
reperfusion injury also resulted in an 6-fold increase in lung neutrophil infiltration
(p~0.05) (Figure 6). When mice were pretreated with exogenous IL-10, neutrophil
infiltration was significantly reduced (p~0.05 for all doses). Maximal
improvements in pulmonary neutrophil i"fillration were attained with 5 ~Lg/mouse(250 ~lg/kg BW) of IL-10.
Visceral ischemia-reperfusion injury ~soci~ted with supraceliac aortic
cross-clamping promotes the release of IL-10, while exogenous IL-10
administration prior to aortic cross-clamping limits pulmonary injury in this model
of acute visceral ischemia-reperfusion injury. Thus, exogenous IL-10 may offer anovel therapeutic approach to decrease complications associated with
thoracoabdominal aortic aneurysm repair and other ischemia-reperfusion injuries.
Hypothetical Example 3 illustrates a preferred application of the invention
conte""~lated for treating humans.
EXAMPLE 3
A 58 year-old white male presents to the emergency room of a local
University hospital complaining of several months of inte""itlent sharp epigastric
and periumbilical abdominal pain, with no other significant sy",ptoms. The patient
has no history of any significant medical pr~bl~ms other than a history of
atherosclerotic ~ise~ce. On physical exam, the patient is found to have a
nontender, pulsatile mid-abdominal mass, with an audible bruit. Laboratory
examination including hematology, biochemistries, liver function tests, urinalysis
and amylase are all within normal limits. Flat and upright abdominal x-rays, as
well as chest x-rays, are unremarkable. An abdominal CT scan with cuts through
the lower chest reveals an aortic aneurysm extending from the level of the
diaphragmatic hiatus to the aortic bifurcation, 6.5 cm in largest diameter.
.
CA 022~3313 1998-10-28
W 097/40849 PCTrUS97tO6809
-18-
After informed consent is obtained, the patient is prepared for surgery. One
hour prior to skin incision, the patient is given a single bolus administration of
recombinant human IL-10 at a dose of 1011g/kg body weight through an indwelling
catheter in the median cubital vein. In addition, a lumbar catheter is placed to5 drain cerebrospinal fluid to maintain intrathecal pressure at 5-10 cm water
pressure. Under a general inhalation anesthetic, a left flank incision is made,
gaining access to the aorta via a retroperitoneal approach. The diaphragm is
divided circumferentially to allow exposure of the thoracic aorta. After the patient
is given intravenous doses of mannitol (0.5 gm/kg) and solumedrol (15 mg/kg), the
10 aorta is cross-clamped proximal to the cephalad aspect of the aneurysm and distal
to the aortic bifurcation at the level of the proximal external iliac arteries. The aorta
is then reconstructed utilizing a bifurcated graft from the level of the caudal
thoracic aorta to the external iliac arteries bilaterally. The celiac and superior
mesenteric arteries are then sewn to the graft as a Carrel patch. Cross-clamp time
15 and period of warm visceral ischemia is 42 minutes. The aortic cross-clamps are
thereafter removed, restoring perfusion of the viscera, pelvis, and lower
extremities. Three units of packed red blood cells and two units of fresh frozenplasma are infused. Incisions are then closed, and the patient is transported to the
surgical intensive care unit intubated and receiving ventilatory ~ssist~nce, but20 hemodynamically stable. After an unremarkable night, the patient is extllh~ted on
post-operative day 1. On post-operative day 2, the patient is transferred out of the
intensive care unit to the surgical ward. The patient has return of bowel function
on post-operative day 5, and is discharged home, ambulating without difficulty,
tolerating a regular diet, with his incision healing nicely, with no evidence of25 infection on post-operative day 7.
Another preferred applicalion of this invention is administration of IL-10 to a
patient one to zero hours before the patient receives a major organ transplant.
This invention is especially applicable to treatment of ischemia-reperfusion
30 occurring in the visceral section of the body. Regardless of which procedure
causes or is expected to cause the ischemia- reperfusion injuly, the inventive
method of treatment will be deemed successful if one or more of the signs or
s~""ptol"s of ischemia-reperfusion injury are alleviated or fail to appear at all.
CA 022~3313 1998-10-28
W O 97/40849 PCT~US97/06809
_19_
EXAMPLE 4
The following experiments in rats demonstrate that pretreatment with
exogenous human IL-10 may decrease lung and soleus muscle injury in a
5 clinically relevant model of hindlimb ischemia-reperfusion injury.
Twenty eight male Sprague-Dawley rats (Charles River Laboratories,
Wilmington, MA., approx. 350 gm) were anesthetized with pentobarbital
intraperitoneally (40 mg/kg, Abbott Laboratories, Chicago, IL.). In twenty of the
10 rats, bilateral hindlimb ischemia was produced by placement of a rubber band
tourniquet across the upper thigh of both lower extremities. The cessation of
arterial blood flow was confirmed by the absence of a Doppler signal in the
superficial femoral artery. The remaining eight rats received anesthesia alone.
Half of the animals in each group (10 in the ischemic group and 4 of the
non-ischemic conl~ols) were pretreated with 10 llg of recombinant IL-10. After the
induction of anesthesia, a catheter was placed into the right atrium through theexternal jugular vein for blood sampling and infusion of normal saline (1 cc/hr).
Recombinant human IL-10 (rhlL-10, 10 ,ug approx. 30 llg/kg BW IV) or a
20 comparable volume of normal saline was administered twenty minutes prior to the
onset of ischemia or at comparable times for the non-ischemic CGI Illols.
After 4 hours of ischemia, the tourniquets were removed and the exl,t:",ity
was reperfused. The restoration of arterial blood flow was confirmed by the
25 presence of a Doppler signal in the superficial femoral artery. Blood (0.5 cc) was
sampled at the time of central venous line placement, at reperfusion, 30 minutesafter reperfusion, 60 minutes after reperfusion, and hourly thereafter. Blood was
sampled at comparable time periods in the non-ischemic controls.
The animals were euthanized (pentobarbitol 100 mg/kg BW IV)
after 4 hours of reperfusion or at comparable times for the non-ischemic controls.
The soleus muscle from one hindlimb and one lung were analyzed for
assessment of neutrophil infiltration. Soleus muscle and pulmonary neutrophil
sequestration were quantified by the tissue myeloperoxidase (MPO) levels
(Warren etal., 1989, J.Clin.lnvest. 84:1873).
The remaining soleus muscle and lung tissue were analyzed to quantify the
capillary and/or cellular injury. Skeletal muscle and lung capillary endothelial cell
CA 022~3313 1998-10-28
W 097/40849 PCTrUS97/06809
-20-
injury were quantified by uptake of 1125 labeled albumin (Welbourn et al., 1991, J.
Appl. Physiol. 70:2645). Skeletal muscle cellular injury was quantified by the
uptake of Tc99 labeled pyrophosphate (Blebea etal., 1988, J. Vasc. Surg. 8:117).The mean capillary permeability index (CPI) and the skeletal muscle injury index5 (SMII) were c~lc~ ted using the following formulas:
CPI = (l125 muscle/muscle mass) / (1125 blood/blood mass).
SMII = (Tc99 muscle/muscle mass) / (Tc99 blood/blood mass).
10Circulating bioactive TNF was measured using the TNF-sensitive WEHI
murine fibrosarcoma cell line (Van Zeed et al., 1992, PNAS 89:4845).
Skeletal Muscle Injury:
15The results are shown in Table 6. The hindlimb l/R resulted in
significant skeletal muscle injury. Both the mean soleus muscle capillary
permeability index (MCPI) and the mean soleus skeletal muscle injury index
(SMII) after hindlimb l/R were significantly greater than the non-ischemic conlrols.
Pretreatment of the animals with recombinant human IL-10 prior to hindlimb
20 ischemia resulted in a significantly lower skeletal muscle capillary injury that was
not significantly different from the non-ischemic control. Pretreatment with human
IL-10 prior to ischemia also resulted in a decrease of the skeletal muscle cellular
injury, although the difference did not reach significance. However, again the
skeletal muscle cellular injury in the ischemic animals pretreated with human
25 recombinant IL-10 was not different than the non-ischemic controls. Neutrophil
infiltration in the skeletal muscle was not detected by the MPO assay for any of the
4 treatment groups.
CA 022~33l3 l998-l0-28
W 097/40849 PCT~US97/06809
-21-
Table 6
Skeletal Muscle Injury
Muscle Capillary Skeletal Muscle
Permeability Index Injury Index
l/R 1.51 + 1.46 0.94 + 0.83
I/R + IL-10 0.39 i 0.38* 0.36 + 0.51
SHAM 0.03 + 0.01* 0.05 + 0.02
SHAM+IL-10 0.19+0.23* 0.26+0.30
* significantly different from l/R (ANOVA, Duncan's multiple range test; p<.05)
Lung Injury
The results are shown in Table 7. The hindlimb ischemia-reperfusion also
resulted in significant pulmonary vascular injury as determined by the leakage of
112~ albumin into the lungs. Both the mean pulmonary capillary permeability index
and the mean pulmonary neutrophil infiltration in the animals subjected to
15 hindlimb ischemia-reperfusion were significantly greater than the non-ischemic
controls. Pretreatment with human recombinant IL-10 significantly reduced the
lung capillary injury after hindlimb ischemia-reperfusion and the PCPI values inthe pretreated animals were not different from the non-ischemic controls. In
contrast, pretreatment with human recombinant IL-10 resulted in a significant
20 increase in the lung myeloperoxidase content after hindlimb ischemia-reperfusion.
Although a ready explanation for this latter finding is not immediately forthcoming
and it is in no way essential to this invention, it may well have been that IL-10
prevented the activation and degranulation of neutrophils in the lung. In this
model, IL-10 may not have prevented the recruitment of neutrophils into the lung,
25 but prevented the degranulation of their toxic contents, thus explaining both the
higher MPO levels and reduced endothelial injury. Treatment of the non-ischemic
controls with human recombinant IL-10 also increased the pulmonary neutrophil
infiltration, although this difference was not significant.
~ _ , . .
CA 022~3313 1998-10-28
WO 97/40849 PCTIUS97/06809
-22-
Table 7
Lung Injury
Lung Capillary Lung Myeloperoxidase
Permeability Index Levels (U/g)
UR 2.32 + 1.25 19.54 + 11.07
I/R + IL-101.02 + 1.22* 37.06 + 4.35~
SHAM 0.65 + 0.22~ 7.82 + 3.03*#
SHAM + IL-100.44 + 0.43~ 11.02 + 2.01~#
5 *siy"i~icanlly different from l/R (ANOVA Duncan's; p<.05)
#significantly different from l/R + IL-10 (ANOVA Duncans multiple range test;
p~.05)
TNF Assay: Serum was assessed for circulating TNF in 6/10 rats undergoing
1 0 ischemia-reperfusion and TNF levels 2 50 pg/ml were detected in 67% (4/6). In
cGrlllasl significant circulating TNF levels were found in only 30% (3/10) of the
ischemic animals pretreated with human recombinant IL-10. Serum TNF levels of
2 50 pg/ml were not detected in any of the non-ischemic control animals.
These findings demonstrate that the anti-inflammatory cytokine IL-10
attenuates both local and distant organ injury resulting from hindlimb ischemia-reperfusion. The findings therefore provide indirect evidence that the associated
injuries are mediated in part by the proinflammatory cytokines and are amenable
to IL-10 based treal",e"ts.
