JOURNALOF SURGICALRESEARCH50,595-601(1991)

PGE, Reduces Injury in Hepatic Allografts

following

Preservation’

KIM M. OLTHOFF, M.D., EVETTE WASEF, B.S., PHILIP SEU, M.D., DAVID K. IMAGAWA, M.D., PH.D., JULIE A. FREISCHLAG, M.D., JOHNHART, M.D.,* ANDRONALD W. BUSUTTIL, M.D., PH.D. Department

Presented

of Surgery

at the Annual

and *Department

Meeting

of Pathology,

of the Association

for Academic Surgery, Houston,

Prostaglandins of the E series have been shown to decrease renal and hepatic ischemic injury as well as improve hepatic function in patients with primary nonfunction following transplantation. We wished to determine the effect of prostaglandin E, (PGE,) on hepatic allograft reperfusion injury in the isolated perfused rat liver (IPRL) model. Livers were harvested from adult male Sprague-Dawley rats and the bile duct, portal vein, and suprahepatic vena cava were cannulated. Control livers were placed immediately on the IPRL apparatus and perfused for 2 hr with a bloodKreb’s solution. Group A and B allografts were stored for 8 hr in heparinized lactated Ringer’s solution at 4°C. Group A livers were then perfused with a PGE, infusion at 0.1 wg/kg/min while B livers received a placebo infusion of NS at the same rate. Temperature, pH, and inflow pressures were kept constant. Oxygen consumption, portal flow, and resistance were calculated for each group and found not to be statistically different. LDH, SGOT, superoxide anion (SOA), and bile flow were measured at 30-min intervals. At the end of the 2-hr perfusion, the placebo Group B (N = 5) had LDH, SGOT, and SOA higher than those of either Group A (N = 5) or control (N = 4) livers. The difference between Group A and Group B was significant for SGOT and SOA (P < 0.05). Bile flow was highest in the control group (24.2 k 1.8 ccl/g/30 min). Group A livers produced bile at a similar rate (19.1 + 2.5 pi/g/30 min), while Group B bile rate was markedly reduced (11.2 ? 2.5 pi/g/30 min). The difference between Groups A and B was not significant (P = 0.06). Histologically, allografts in Group B demonstrated cellular edema and pale staining while those in Group A showed normal hepatic architecture with no evidence of injury, comparable to those of control livers. In conclusion, PGE, infusion reduces hepatic injury following extended cold storage and reperfusion as determined by decreased

1 This work was supported Liver Transplant Foundation.

UCLA School of Medicine,

in part by the Joanne Barr Memorial

Los Angeles, California

Texas, November

SGOT, SOA, and preservation 0 1991AcademicPress,Inc. ture.

90024

14-17, 1990

of hepatic

architec-

INTRODUCTION

Since the recent introduction of new solutions for the preservation and storage of hepatic allografts, successful liver transplantation is no longer limited by short periods of preservation of 8 hr or less [l]. Unfortunately, some allografts still fail to function well initially after transplantation. The cause of this primary nonfunction is felt to be due to an injury which occurs during preservation and significantly affects the success of orthotopic liver transplantation (OLT) [2]. Reperfusion injury has been described in the heart [3], the cold-preserved ischemic kidney [4], the ischemic intestine [5], and more recently in the ischemic, reperfused liver [6] and the cold-stored liver [7]. This injury appears to occur to endothelial cells upon reintroduction of oxygen following ischemia [8]. Free radicals have been implicated as the mediators of this endothelial cell damage, leading to a no-reflow phenomenon associated with capillary plugging and cellular edema [9]. Prostaglandins of the E series act directly on the vascular smooth muscle, inhibit platelet aggregation, and have been shown to have a cytoprotective action on gastric and intestinal mucosa [lo, 111, kidneys [12], and lungs [13] following ischemic and chemical insults. Recently, prostaglandin E, (PGE,) has been shown to be beneficial in the treatment of fulminant hepatic failure [ 141 as well as in clinical primary nonfunction following OLT [15]. This study was designed to determine if the administration of PGE, following extended hypothermic preservation of the liver would reduce or reverse any hepatic injury seen with reperfusion. The isolated perfused rat liver model was used to mimic the three components of clinical preservation and transplantation: ischemia, hypothermia, and reperfusion. 595

0022-4804/91$1.50

Copyright 0 1991 by Academic Press, Inc. All rights of reproduction

in any form reserved.

596

JOURNAL

OF SURGICAL

RESEARCH:

TABLE Data

Control Group A Group B

Representing

VOL.

50, NO. 6, JUNE

1991

1

Values (Mean k SE) of Liver Weight Prior to Perfusion and Portal Portal Flow, and Portal Resistance at Midperfusion (60 min)

Pressure,

N

Liver weight k)

Portal pressure (cm perfusate)

Portal flow (ml/mm/g liver)

Portal resistance (10’ - dyn - set * cm-i)

4 5 5

9.75 It 1.8 9.60 + 0.7 10.88 k 1.5

13.6 f 0.5 14.3 * 0.1 14.4 ? 0.4

1.44 f 0.18 1.32 k 0.12 1.21 * 0.14

6.30 + 0.30 6.83 2 0.34 7.02 f 0.47

MATERIALS

AND

METHODS

Male Sprague-Dawley rats (Bantin and Kingman, Fremont, CA) weighing between 250 and 350 g were used as donor animals. The animals were allowed free access to rat chow and water prior to surgery. Hepatic allografts were harvested under ether anesthesia as initially described by Miller [16]. The portal vein, bile duct, and suprahepatic vena cava were cannulated and the livers were flushed with heparinized lactated Ringer’s solution (LR), weighed, and stored in the same solution for 8 hr at 4°C. It has been previously shown that rat livers stored in a simple salt solution for 8 hr will not survive in the rat OLT model, representing severe preservation injury [ 171. Control livers were not stored and were placed immediately on the perfusion apparatus. The isolatedperfused rat liver (IPRL) apparatus. The IPRL apparatus was enclosed. in a temperature-controlled plexiglass cabinet maintained at 37°C. The perfusate was circulated through Tygon tubing by a Masterflex peristaltic pump (Cole Parmer Instruments, Chicago, IL) with inline blood filter, flowmeter (Gilmont Instruments, Chicago, IL), bubble trap, pressure flow monitor, pH and temperature probes, and inflow and outflow oxygen electrodes (Instech Laboratories, Her-

sham, PA). The perfusate was maintained at a pH of 7.4 and oxygenated with a Hamilton Silastic tube lung and the p0, was maintained over 400 mm Hg [18]. Perfusate. The perfusate consisted of rat whole blood diluted 2:l with Krebs-Ringer bicarbonate medium (Sigma Chemical Co., St. Louis, MO) to a hematocrit of 15%. A total of 200 cc of circulating perfusate was used. Liver perfusion. Following preservation, the livers were flushed with heparinized LR and placed in the perfusion circuit. Group A livers received a PGE, infusion. PGE, (Prostin-VR, Upjohn Co., Kalamazoo, MI) was diluted in normal saline and infused at a rate of 0.1 pg/kg body wt/min by using a Y-connector into the circuit near the inflow cannula. Group B allografts received a placebo infusion of normal saline at a similar rate. Control livers (N = 4) were placed immediately on the perfusion apparatus and received no infusion. Temperature, pH, and inflow pressures were kept constant. Outflow samples were collected at 30-min intervals and centrifuged. Portal flow and portal pressure were calculated at 30-min intervals. Arterial and venous blood samples were analyzed to determine oxygen consumption at 30 and 60 min. Bile was collected for 30-min

30

T

3.0

20

CONTROL GROUP A GROUP B

OXYGEN CONBUMPTION prides / gn. liver. min 1.0

BILE FLOW pllgnl30min

20

q CONTROL GROUP A GROUP B

10

* pd5 contrd vs. B

‘p=.c6 Avs.B

60

30 MINUIES

FIG. 1. Oxygen consumption (mean f SE) in perfused livers at 30 and 60 min. Values are significantly higher in controls than in Group B placebo allografts. There was no significant difference between Group A PGE, allografts and Group B allografts.

FIG. 2. Bile flow over 30.min intervals of perfusion. The flow in the control livers was consistently better than in either Group A or B, but significant only when compared to Group B (‘P < 0.05). Flow rates were higher in Group A when compared to Group B, but only approached significance (*P = 0.06).

OLTHOFF

ET AL.: PGE,

AND

HEPATIC

ALLOGRAFT

597

INJURY

0.6 1

0.4

i!im

---

LDH U/ml

CONTROL GROUP A

1

1 I

/............. T

/

*IA54i.of.a /..-*

* p&6 AVC3.B

.. . . . . . . GROUP B

loo0

500

CONTROL --GROUPA . . . . . . . . GROUP B

‘p=NS AW3.B

30

0

60 MINUTES

90

o

120

FIG. 3. LDH concentration in the perfusate (mean f SE). There is no significant difference between Groups A and B. Group B is significantly higher than control at 30 and 60 min (P < 0.05).

intervals and measured to determine bile flow rate. The supernatant was analyzed for LDH, SGOT, and superoxide anion (SOA) levels at 30-min intervals. Perfusions were performed for 120 min. Chemical techniques. LDH and SGOT were measured using commercially available kits (Sigma). Blood gas determinations were done using a blood gas analyzer (Radiometer, Copenhagen). The production of SOA was determined using a modification of the technique described by Cohen and Chovaniec [ 191. Thirty microliters of cytochrome c in a 1 mmole concentration was added to 920 ~1 of prewarmed Hanks’ solution (37°C) in a l-ml cuvette. Fifty microliters of supernatant was then added to the cuvette. The sample was then read in a spectrophotometer at an OD of 550 nm over a lo-min interval. After an initial activation period, the change in absorbance between 2 and 6 min was noted, reflecting the reduction of cytochrome c by the SOA release. This same reduction was run in the presence of superoxide dismutase (Sigma)

4 0

30

60 MINU-IES

90

120

FIG. 5. Superoxide anion production (mean + SE) by 50 pl of perfusate supernatant as measured by reduction of cytochrome c. Asterick indicates a significantly different data point between Group A and Group B. Results for Group A and control livers are nearly identical.

which blocked the reaction and served as a control. The rate of SOA generation was derived by dividing the linear absorbance change by the molar extinction coefficient for reduced cytochrome c minus oxidized cytochrome c and expressed as pmole OJmin/50 ~1 supernatant. Histology. Immediately following 2 hr of perfusion, the allografts were fixed in a 10% buffered formalin solution for over 24 hr and processed for routine histology with hematoxylin and eosin staining. All slides were read by the same pathologist who was blinded to the treatment groups and each allograft was graded on a severity of injury from 0 (normal) to +++ (severe injury). Statistics. Results are expressed as the mean f SE unless otherwise stated. Treatment groups were compared using an unpaired two-tailed t test. Data were analyzed on a Macintosh II personal computer with the Statview II statistics program. RESULTS

4m-

--

SGOT

--

CONTROL GROUPA

. .. . . . . . GROUP B

uldl

‘p