Bile and Pancreatic Juice Replacement Ameliorates Early Ligation-Induced Acute

PancreatiSs in Rats Isaac Samuel, MD, PRCS, Yasuo Toriumi, MD, PhD, Daniel P. Wilcockson, MS, Chicago, Illinois,

Charles M. Turkelson, PhD, Travis E. Solomon, MD, PhD, Kansas City, Missouri,

Raymond J. Joehl, MD, Chicago, Illinois

BACKGROUND: In healthy rats, combined bile and pancreatic juice diversion from gut has a synergistic rather than additive effect on stimula- tion of exocriue pancreatic protein secretion. We hypothesized that exclusion of combined bile and pancreatic juice from gut exacerbates bile and pancreatic-duet ligation-induced acute pan- creatitis iu rats to a greater extent than exclu- sion of either bile or pancreatic juice alone.

METHODS: Bile and pancreatic juice (obtained fresh from donor rats) were replaced, separately or together, via a duodenal fistula beginning im- mediately before 6 hours of duct ligation. Pan- creatic morphologic changes were evaluated with an acute pancreatitis histology score and morpho- metric quantitation of aciuat-cell necrosis. Plasma amylase and cholecystokiniu concentrations and pancreatic subcellular distribution of cathepsin B activity were determined. Characteristics of bile and pancreatic juice obtained from donor rats were also studied.

RESULTS: Combined bile and pancreatic juice re- placement limited the increase in acute pancre- atitis histology score by 77%, acinar cell necro- sis by 95%, hyperamylasemia by 77%, and hypercholecystokininemia by 99%, while pre- venting subcellular redistribution of cathepsin B. Amelioration of pancreatic morphologic changes was significantly greater with combined bile and pancreatic juice replacement than with replace- ment of either bile or pancreatic juice alone.

CONCLUSION: In this experimental corollary of early gallstone-induced acute pancreatitis, com- bined bile and pancreatic juice exclusion from gut contributes to disease pathogenesis to a greater extent than exclusion of either bile or pancreatic juice alone.

From the Department of Surgery (IS, YT. DPW, RJJ), Northwestern University Medical School, and Surgical and Research Services, Veterans Affairs Lakeside Medical Center, Chicago, Illinois; and Research Service (CMT, ‘ES), Veterans Affairs Medical Center, Kansas City, Missouri.

This study was supported by a Merit Review grant from Medical Research Service, Department of Veterans Affairs, Washington, DC.

Isaac Samuel, MD, was the recipient of a Research Fellowship from the Department of Surgery, Northwestern University Medical School, Chicago, Jllinois.

Requests for reprints should be addressed to Raymond J. Joehl, MD, Department of Surgery, Northwestern University Medical School, 250 East Superior Street, Suite 201, Chicago, Jllinois 60611.

Presented in part as a poster at the Annual Meeting of the Society for Surgery of the Alimentary Tract, Boston, Massachusetts, May 17-19, 1993.

Manuscript submitted May 2, 1994 and accepted in revised form July 29, 1994.

L igation of the common bile-pancreatic duct excludes bile and pancreatic juice from gut and induces acute

pancreatitis in rats. is4 Exclusion of bile from gut has pre- viously been implicated in the pathogenesis of ligation-in- duced acute pancreatitis in rats4,5 and in opossums.6 Additionally, in healthy rats, the increase in exocrine pan- creatic protein secretion after external diversion of pan- creatic juice alone is of a greater magnitude than that af- ter external diversion of bile alone.7 Moreover, combined external diversion of both bile and pancreatic juice has a synergistic rather than an additive effect on the stimula- tion of exocrine pancreatic protein secretion.7

Based on these observations, we hypothesized that early in the development of ligation-induced acute pancreatitis in rats: (1) combined bile and pancreatic juice exclusion from gut contributes to the pathogenesis of acute pancreatitis to a greater extent than exclusion of either bile or pancreatic juice alone; (2) exclusion of pancreatic juice alone from gut contributes to the pathogenesis of acute pancreatitis to a greater extent than exclusion of bile alone; and (3) com- bined bile and pancreatic juice replacement is useful in both prevention and treatment of acute pancreatitis.

To test these hypotheses, we studied the ameliorating ef- fects of duodenal replacement of bile and pancreatic juice, separately or together, on rats with early acute pancreatitis induced by ligation of the common bile-pancreatic duct for 6 hours. Bile and pancreatic juice were obtained fresh from donor rats for replacement in experimental rats (Figure 1). Additional experiments were performed to study character- istics of bile and pancreatic juice obtained from donor rats.

METHODS Male Sprague-Dawley rats were purchased from Charles

River Laboratories (Wilmington, Massachusetts). Rats weighing 350 to 450 g were used as experimental rats, while larger rats weighing 450 to 550 g were used as donor rats.

The 46 rats used in experimental groups were allowed free access to water and standard rat chow until induction of anesthesia. Midline laparotomy was performed under general anesthesia induced with subcutaneous ketamine hydrochloride 87 mg/kg and xylazine hydrochloride 13 mg/kg. A duodenal fistula was created using a polyethy- lene tube (PE-50, I.D. 0.58 mm, O.D. 0.965 mm; Clay Adams, Parsipanny, New Jersey). The rats were studied in the following experimental groups.

Diseased Controls: Bile and Pancreatic-Duct Ligation (BPDL) Group

In 8 rats, the hepatic bile duct was fist ligated proximal to its junction with the pancreas to prevent bile entry into the common bile-pancreatic duct. After an interval of

THE AMERICAN JOURNAL OF SURGERY” VOLUME 169 APRIL 1995 391

LIGATION-INDUCED ACUTE PANCREATITISlSAMUEL ET AL

Peristaltic Pump

C Donor

Donor Rat

Recipient Rat A Rat

Recipient Rat B Recipient Rat C

Figure 1. Donor rats were used to obtain bile and pancreatic juice for donation to experimental rats. A. Bile and pancreatic juice were di- verted together from a donor rat via a bile-pancreatic fistula and recirculated via a duodenal fistula during recovery periods. Enteral feeding was instituted via a gastric fistula. A liquid-level photodetector and peristaltic pump were used to facilitate collection and immediate recircula tion of bile and pancreatic juice. B. Bile and pancreatic juice were donated together to an experimental rat with bile and pancreatic duct liga- tion (Recipient Rat A). C. When experimental rats with bile and pancreatic duct ligation received either pancreatic juice alone (Recipient Rat B) or bile alone (Recipient Rat C), the donor rat was prepared with separate pancreatic and bile fistulas.

about 1 minute to allow residual bile in the common bile- pancreatic duct to pass into the duodenum, the common bile-pancreatic duct was ligated near its junction with the duodenum. Using a syringe pump (Sage Instruments, Cambridge, Massachusetts), 50 mmol/L NaHCO, was in- fused via the duodenal fistula at 2 mL/h.

Diseased/Treated Groups In 30 rats, bile and pancreatic-duct ligation was performed

as above, and bile and pancreatic juice (obtained fresh from a donor rat) were replaced separately or together via the duodenal fistula, as shown in Figure 1: (1) Replacement of combined bile and pancreatic juice (BPDL + BPJ group, n = 9); (2) replacement of pancreatic juice alone (BPDL + PJ group, n = 8); (3) replacement of bile alone (BPDL + B group, n = 7); and (4) late replacement of combined bile and pancreatic juice, begun 30 minutes after bile and pan- creatic-duct ligation (BPDL + BPJ-L group, n = 6). In the first three diseased/treated groups above, replacement was begun immediately before duct ligation.

Sham Controls In 8 rats, the hepatic bile duct and common bile-pan-

creatic duct were dissected but not ligated. Normal saline

was infused via the duodenal fistula at 2 mL/h using a sy- ringe pump.

Investigations Rats in experimental groups were killed 6 hours after op-

eration by inferior vena caval exsanguination under gen- eral anesthesia (induced as above). The pancreas was quickly excised. Plasma amylase concentrations were de- termined on a spectrophotometer (CX5, Beckman, Brea, California). Plasma cholecystokinin concentrations were measured using a sensitive and specific radioimmunoas- say.8 Tracer for the assay was prepared as described pre- viously.9 A portion of each pancreas was immediately fixed in 10% phosphate-buffered formaldehyde. Hematoxylin- and eosin-stained sections were examined in blinded fash- ion and assigned an acute pancreatitis histology score based on degree of white blood-cell infiltration, acinar-cell vac- uolation, and focal acinar-cell necrosis (Table).

After a review of the results obtained by the above his- tology scoring system, we performed morphometric quan- titation of focal acinar-cell necrosis. Each histology slide was examined again in blinded fashion and each focus of acinar-cell necrosis was photographed (Olympus BH2, Tokyo, Japan) at 200X magnification. Using the Bioquant

392 THE AMERICAN JOURNAL OF SURGERY” VOLUME 169 APRIL 1995

IV system of digitizing morphometry (R & M Biometrics, Nashville, Tennessee), the area of each focus of acinar- cell necrosis was measured on photomicrographs, and the area of pancreatic tissue examined was measured from his- tologic slides. A total of 257 photomicrographs was com- piled for morphometric analysis. Numerical density of fo- cal acinar-cell necrosis was expressed as the number of foci per square centimeter of pancreatic tissue. Areal den- sity of acinar-cell necrosis, which represents the total area of acinar-cell necrosis per square centimeter of pancreatic tissue, was expressed as X IO3 pm2/cm2 after correction for magnification. The correction factor for magnification was X40,000, as linear magnification of X200 results in areal magnification of X200*. Linear magnification was calibrated with a micrometer.

To study subcellular distribution of lysosomal enzyme cathepsin B, a portion of each pancreas stored at -70°C was homogenized in 0.25 M buffered sucrose.3 Differ- ential centrifugation was performed at gravitational speeds of 15Og for 15 minutes, 1,OOOg for 20 minutes, 12,000g for 20 minutes, and 105,000g for 60 minutes to obtain a postnuclear total homogenate, a zymogen-en- riched fraction, a lysosome-enriched fraction, and a post- microsomal soluble fraction, respectively.3J0 Cathepsin B activity, assayed with N-alpha-benzyloxycarbonyl-l-ly- sine p-nitrophenyl ester as substrate (Beckman DU65, Fullerton, California),” was expressed in units of enzyme activity per gram of total protein; total protein was deter- mined by measuring absorbance at 280 nm in the post- nuclear total homogenate. All chemicals used for cathep- sin B assay and homogenization buffer were purchased from Sigma Chemical Co. (St. Louis, Missouri).

Surgical Preparation of Donor Rats Ten rats (450 to 550 g) were fasted overnight and sub-

mitted to midline laparotomy under general anesthesia in- duced with ketamine and xylazine, administered as de- scribed above. Donor rats used for combined bile and pancreatic juice donation (n = 4) were prepared with a common bile-pancreatic fistula, while those used for sep- arate donation of bile and pancreatic juice (n = 2) were prepared with separate bile and pancreatic fistulas (Figure 1).5,7.12 A duodenal fistula was created in each donor rat for recirculation of externally diverted bile and pancreatic juice. In addition, a gastric fistula was instituted for en- teral feeding. A liquid-level photodetector (Skan-A-Matic, SENISYS, Plano, Texas) and peristaltic pump (Rainin, Boston, Massachusetts) were used for collection and im- mediate recirculation of bile and pancreatic juice.5*7.‘2 A 3-day postoperative recovery period with recirculation of bile and pancreatic juice was allowed prior to bile and pan- creatic juice donation, as a shorter recovery period is as- sociated with decreased exocrine pancreatic function.5J2

From the first postoperative day, each donor rat received enteral feeding (Ensure Plus 1.5 Cal/mL, Ross Laboratories, Columbus, Ohio) diluted 1: 1 with distilled water and infused with a syringe pump at 1 mIJh. In addition, water and stan- dard rat chow were freely available. When experiments were begun, the entire volume of bile and (or) pancreatic juice di- verted from a donor rat during the 6-hour period was do- nated to a recipient rat in the diseased/treated experimental

LIGATION-INDUCED ACUTE PANCREATITISlSAMUEL ET AL

TABLE Acute Pancreatitis Histology Score

White blood cell infiltration (range O-3) Periductal infiltration only 1.0 Periductal and occasional parenchymal infiltration 1.5 Periductal and mild parenchymal infiltration Periductal and moderate parenchymal infiltration G:i Periductal and severe parenchymal infiltration 3.0

Acinar-cell vacuolation (range O-1) Occasional vacuolation 0.5 Prominent vacuolation 1.0

Acinarcell necrosis (range O-1) Occasional acinar necrosis 0.5 Prominent acinar necrosis 1.0

An acute oancreatitis histoloav score (ranae O-5). assinned bv examina- tion of hematoxylin and eosi&ained sections of pancrek in blinded fasb ion, was used for qualitative assessment of pancreatic morphologic changes.

groups. During donation, the rate of enteral feeding was in- creased to 3 mIJh to maintain fluid balance in the donor rat. Each donor rat donated its bile and (or) pancreatic juice to a maximum total of 4 recipient rats on 4 consecutive days from the third to the sixth postoperative day. After each-6-hour period of donation, each donor rat was allowed an l&hour overnight period of recovery during which its bile and pan- creatic juice were recirculated (Figure lA), and enteral feed- ing was continued at 1 mI./h. Polyethylene catheters (Clay Adams) were used to create bile-duct, pancreatic, duodenal (PE-50, I.D. 0.58 mm, O.D. 0.965 mm) and gastric fistulas (PE-90, I.D. 0.86 mm, O.D. 1.27 mm). The polyethylene catheters were tunneled subcutaneously to the nape of the neck where they exited and were passed through the hollow cable coil of a tethering system (Harvard Apparatus Inc., South Natick, Massachusetts) secured to a stand.

Each donor rat was housed in a cage in the laboratory with controlled ambient room temperature of 23 + 2°C. The tethering system allowed ample mobility, grooming, and free access to water and chow. Donor rats were mon- itored daily with regard to grooming activity, ear color, patency of catheters, eating and drinking routines, urina- tion, and bowel action. Donor rats were killed after use with an injection of 0.5 mL saturated potassium chloride via the inferior vena cava under general anesthesia in- duced as above.

Characterization of the Donor Rat To study the effects of repeated donation on characteris-

tics of bile and pancreatic juice, 4 donor rats were prepared with separate bile and pancreatic fistulas. These donor rats used for the characterization experiment were not used for donation of bile and pancreatic juice to experimental rats, but underwent sham donation only. All other procedural details were identical to that described above, under sur- gical preparation of donor rats. This “characterization of the donor rat” experiment involved separate collection of bile and pancreatic juice every hour for 6 hours on 4 con- secutive days, from the third to the sixth postoperative day (ie, days of donation 1 to 4).

Hourly volumes of bile and pancreatic juice were re- corded; basal volumes of bile and pancreatic juice were

THE AMERICAN JOURNAL OF SURGERYa VOLUME 169 APRIL 1995 393

LIGATION-INDUCED ACUTE PANCREATITIS/SAMUEL ET AL

Figure 2. Representative photomicrographs of hematoxylin- and eosinstained sections of rat pancreas after 6 hours of bile and pancreatic- duct ligation in diseased controls (BPDL group; original magnification x 1000). A. Focal acinarcell necrosis involves two adjacent acini. Interstitial white blood cell infiltration is also seen (black arrows). B. Characteristic appearance of acinarcell vacuolation: several vacuoles are seen, some of which are immediately adjacent to the nucleus and contain cell debris.

not recorded. Pancreatic juice protein and bicarbonate RESULTS concentrations and bile salt and bicarbonate concentrations Pancreatic Morphologic Changes were determined in the basal state (“zero hours”) and af- Bile and pancreatic-duct ligation (BPDL group) in- ter each hour of donation. Protein concentration was de- duced acute pancreatitis (Figure 2), as assessed by our termined by measuring absorbance at 280 nm on a spectro- blinded acute pancreatitis histology scoring system and photometer (Beckman DU65, Beckman). Bicarbonate morphometric quantitation of focal acinar-cell necrosis concentration was determined on a pi-I/blood gas analyzer (Figure 3). Following the sham operation, the increase (Instrumentation Laboratories, Lexington, Massachusetts) in acute pancreatitis histology score was entirely due to using a 300~pL sample of bile or pancreatic juice. At each mild white blood cell infiltration (Figure 3). Following time point, a lOO-pL sample of bile was mixed with 900- bile and pancreatic-duct ligation, there was a greater than pL of methanol and stored at -70°C for measurement of threefold increase in acute pancreatitis histology score, bile salt concentration using the 3-alpha-hydroxysteroid compared with sham controls; this increase was due to dehydrogenase reaction. I3 All remaining bile and pancre- increases in all three components of the scoring system. atic juice collected during the 6-hour period of study were Numerical density of focal acinar-cell necrosis follow- discarded and not recirculated (sham donation). However, ing bile and pancreatic-duct ligation was 52 f 23/cm*, recirculation of bile and pancreatic juice was begun soon which amounted to a mean area1 density of 144 * 66 X after the 6-hour period of sham donation and continued lo3 pm*/cm*, while sham operation was not associated overnight until the next 6-hour period of study. with acinar-cell necrosis (Figure 3).

In diseased/treated rats that received combined bile Statistical Analysis and pancreatic juice replacement (BPDL + BPJ group),

In experimental groups, statistical computations em- there was substantial amelioration of ligation-induced ployed one-way analysis of variance or the Kruskal-Wallis acute pancreatitis. This amelioration was reflected in analysis of variance, as appropriate. In characterization significant reductions in all three individual components studies on the donor rat, the paired Student’s t-test was of the acute pancreatitis histology score, which limited used. A P value below 0.05 was considered significant. the increase in the total score by 77%. In addition, mean All values are reported as mean + SEM. numerical and area1 densities of acinar-cell necrosis

394 THE AMERICAN JOURNAL OF SURGERY” VOLUME 169 APRIL 1995

LIGATION-INDUCED ACUTE PANCREATITWSAMUEL ET AL

0 SHAM BPDL BPDL BPDL BPDL BPDL

* L L .

BGJ dJ Ii BtiJ-L

Figure 3. Qualitative assessment of pancreatic morphologic changes and quantitative assessment of acinarcell necrosis were performed on hematoxylin- and eosin-stained sections of pancreas from experi- mental groups. A. Depicts the total acute pancreatitis histology score and individual components of the total score (see Table). B. Depicts numerical density of focal acinarcell necrosis expressed as the num ber of foci per square centimeter of pancreatic tissue; C. Depicts areal density of acinarcell necrosis, which represents total area of acinarcell necrosis per square centimeter of pancreatic tissue and was quantitated from photomicrographs using digitizing morphome try. Values are mean f SEM. * = significantly greater than SHAM group; # = significantly different from SHAM and BPDL groups; and @ = significantly different from SHAM, BPDL, and BPDL + BPJ groups. Kruskal-Wallis analysis of variance, P ~0.05. SHAM = sham control group; BPDL = bile and pancreaticduct ligation group; BPDL + BPJ = group with BPDL and combined bile and pancreatic juice replace ment; BPDL + PJ = group with BPDL and replacement of pancreatic juice alone; BPDL + B = group with BPDL and replacement of bile alone; and BPDL + BPJ-L = group with BPDL and combined bile and pancreatic juice replacement begun 30 minutes after duct ligation.

1

‘2 were reduced to a mere 2.3 k 1.5km (96% ameliora- tion) and 7.6 + 5.0 X lo3 um2/cm2 (95% amelioration), respectively.

Although diseased/treated rats that received either bile or pancreatic juice replacement alone (BPDL + B and BPDL + PJ groups, respectively) showed significant ame- lioration of the acute pancreatitis histology score, the to- tal score was still significantly higher than that follow- ing combined bile and pancreatic juice replacement. Moreover, the acinar-cell necrosis component of the to- tal acute pancreatitis histology score and the numerical density of acinar-cell necrosis were not significantly be- low that seen in diseased controls (BPDL group) fol- lowing replacement of either bile or pancreatic juice alone (BPDL + B and BPDL + PJ groups, respectively). Pancreatic juice replacement alone significantly amelio- rated areal density of acinar-cell necrosis, but not to the extent seen with combined bile and pancreatic juice re- placement. Amelioration of area1 density of acinar-cell necrosis following bile replacement alone was not sig- nificant. Combined bile and pancreatic juice replacement, begun even 30 ‘minutes after bile and pancreatic-duct lig- ation (BPDL + BPJ-L group), achieved significant re- duction of the acute pancreatitis histology score and its individual components, and of the numerical and area1 densities of acinar-cell necrosis.

Plasma Amylase and Cholecystokinin Concentration Diseased controls developed hyperamylasemia, while all

four diseased/treated groups showed significant ameliora- tion of hyperamylasemia (77% reduction, Figure 4).

Diseased controls developed hypercholecystokininemia. All diseased/treated groups, except the one with bile re- placement alone, showed significant amelioration of hy- percholecystokininemia (99% reduction).

Subcellular Distribution of Cathepsin B Confirming observations of other investigators,3 liga-

tion of bile and pancreatic ducts resulted in subcellular redistribution of cathepsin B, as evidenced by a signifi- cant shift of cathepsin B activity from the lysosome-en- riched subcellular fraction to the zymogen-enriched sub- cellular fraction (Figure 5). All diseased/treated groups, except that with bile replacement alone, showed signif- icant amelioration of the subcellular shift of cathepsin B following duct ligation.

Chacterization of the Donor Rat For assessment of changes in volume and composition

of bile and pancreatic juice during 6 consecutive hours of

$ 250

< 225

-b Ii 200

-t 175

ti p 150

5 125

B E loo

Y B 75 h $ 5 50

0 g 25

1

B+pJ ;J

C I /

THE AMERICAN JOURNAL OF SURGERY’ VOLUME 169 APRLL 1995 395

z f 300 a

ii 5 250

P

z ” 2 200

v 12-- .- .2

6 150 3 IO-- * 2

m

6 6-w 3 100

E 8

5 3 6-- J 50

a 4-- 0

2 --

O-

9000 + t

6000 i- 3 7000

3 6000

$ 5000

2 4000

m" i 3000

Figure 5. Graph depicts cathepsin B activity in subcellular fractions of differentially centrifuged pancreatic homogenates in the various ex- perimental groups. l = significantly greater than SHAM group; # = significantly different from SHAM and BPDL groups. One-way analysis of variance, P ~0.05. SHAM = sham control group; BPDL = bile and pancreatic-duct ligation group; BPDL + BPJ = group with BPDL and combined bile and pancreatic juice replacement; BPDL + PJ = group with BPDL and replacement of pancreatic juice alone; BPDL + B = group with BPDL and replacement of bile alone; and BPDL + BPJ-L = group with BPDL and combined bile and pancreatic juice replacement begun 30 minutes after duct ligation.

mencement of diversion. Bicarbonate concentrations of pancreatic juice and bile were essentially maintained dur- ing the entire period of study. However, hourly bile vol- ume and bile salt concentration were not maintained dur- ing the periods of sham donation described above.

Figure 4. Top graph depicts plasma cholecystokinin concentration measured by radioimmunoassay. Bottom graph depicts plasma amy- lase concentration in the various experimental groups. * = significantly greater than SHAM group; # = significantly different from SHAM and BPDL groups. One-way analysis of variance, P ~0.05. SHAM = sham control group; BPDL = bile and pancreaticduct ligation group; BPDL + BPJ = group with BPDL and combined bile and pancreatic juice re- placement; BPDL + PJ = group with BPDL and replacement of pan- creatic juice alone; BPDL + B = group with BPDL and replacement of bile alone; and BPDL + BPK = group with BPDL and combined bile and pancreatic juice replacement begun 30 minutes after duct ligation.

sham donation (ie, total diversion of bile and pancreatic juice from gut) on any 1 of the 4 days of sham donation, values were compared with the earliest value of the same day of sham donation (Figure 6). To assess effect of re- peated 6-hour periods of sham donation for 4 days on vol- ume and composition of bile and pancreatic juice, the value at each time point during the second, third, and fourth days of sham donation was compared with the value at the cor- responding time point on the first day of sham donation.

The total exocrine pancreatic protein output during the entire 6-hour period of diversion each day was increased rather than compromised, owing to increases in both pan- creatic juice volume and protein concentration after com-

0 Postnuclear total homopnate m Zymogec-enriched fradion I Lysosame-enriched fraction e Poslmiaosomal solubls fractiin

SHAM BPDL BPDL BPDL BCDL + +

LIGATION-INDUCED ACUTE PANCREATITKYSAMUEL ET AL

I B PDL +

BPJ PJ B BPJ-L 1

COMMENTS The results of this study indicate that early in the devel-

opment of ligation-induced acute pancreatitis in rats: (1) combined bile and pancreatic juice exclusion from gut ex- acerbates pancreatic morphologic changes to a greater ex- tent than exclusion of either bile or pancreatic juice alone; (2) exclusion of pancreatic juice alone from gut exacer- bates pancreatic morphologic changes to a greater extent than exclusion of bile alone; (3) combined bile and pan- creatic juice replacement is useful in both prevention and treatment of the disease; (4) amelioration of pancreatic morphologic changes after combined bile and pancreatic juice replacement, whether begun before duct ligation (pre- vention) or 30 minutes after duct ligation (treatment), is associated with amelioration of hypercholecystokininemia; and (5) pancreatic juice exclusion from gut, with or with- out bile exclusion from gut, influences the shift of catbep- sin B activity from the lysosome-enriched to the zymogen- enriched subcellular fraction.

The results of this study also indicate that the protein con- centration and the volume of pancreatic juice obtained from donor rats remain satisfactory during the course of donation, whereas bile volume and bile salt concentration are not maintained during each 6-hour period of donation.

The models of ligation-induced acute pancreatitis in rats’-4s’4 and in opossum&r5 are useful experimental corollaries of gallstone-induced acute pancreatitis in hu- mans to investigate disease pathogenesis. Exclusion of

396 THE AMERICAN JOURNAL OF SURGERY” VOLUME 169 APRIL 1995

LIGATION-INDUCED ACUTE PANCREATITIS/SAMUEL ET AL

7 2000

c = 1

1600

=I600

$w

p 1200

.fj 1000

5 600

fj 600

5 s 400

n 200

0 1

2400 + T

g 2200 32ooo t 1600

E" 1600

2 14Gu

p 1200

g 1000

m600

600

400

200

0 5 6 6 2

r f f 60

!Y

$ 50

g40 .o co

8 30 .- 4 .s 20

ii t c 10

d

40

s

5 a 30 I

ii s

5 2o

m al z * 10

0 = 5 6 6

-

: Days of sham donation:

T 0 One W Two m Three I Four

1 2 3 4 5 6 4 5 6 Hours Hours

Figure 6. Graphs depict hourly values for volume and composition of pancreatic juice and bile from donor rats (n = 4). After a 3-day recov- ery, each rat underwent sham donation for 6 hours on 4 consecutive days (days of sham donation 1 to 41, which corresponded to postop erative days 3 to 6. The key to the bar graphs is provided in the graph of pancreatic juice protein. All values are mean 2 SEM. * = signifi- cantly different from first value on the same day of sham donation; # = significantly different from value at the same time on the first day of sham donation. Paired Students t-test, P ~0.05. Note that basal (zero hours) pancreatic juice protein, bicarbonate, bile salt, and bile bicar- bonate were measured while basal volumes were not.

bile from gut has previously been implicated in the patho- of early ligation-induced acute pancreatitis in rats to a genesis of ligation-induced acute pancreatitis in rats4,5 greater extent than exclusion of either bile or pancreatic and in opossums.6 In the present study, we show that juice alone. This finding is consistent with the observa- combined exclusion of both bile and pancreatic juice tion of Nakamura et al’ that exocrine pancreatic hyper- from gut exacerbates the pancreatic morphologic changes stimulation in healthy rats is of a greater magnitude af-

THE AMERICAN JOURNAL OF SURGERY” VOLUME 169 APRIL 1995 397

ter external diversion of both bile and pancreatic juice than after external diversion of either bile or pancreatic juice alone. However, whether exocrine pancreatic hy- perstimulation mediates the exacerbating effects of bile- pancreatic juice exclusion from gut in early ligation-in- duced acute pancreatitis in rats was not investigated in the present study.

Here, measurement of plasma cholecystokinin concen- tration was used as an indirect indicator of the adequacy of bile and pancreatic juice replacement. In addition, char- acterization studies on the donor rat were performed to fur- ther assess the adequacy of replacement. Although bile salt replacement was probably less than adequate, it is clear that bile obtained from donor rats certainly contributed to the amelioration of acute pancreatitis in experimental rats, as combined bile and pancreatic juice replacement achieved significantly greater amelioration of morphologic changes and acinar-cell necrosis than replacement of pan- creatic juice alone (Figure 3). Additionally, the marginal yet significant amelioration of the acute pancreatitis his- tology score seen in this study with bile replacement alone was similar to that achieved in a previous study where bile flow to the duodenum was maintained using a polyethy- lene shunt between the hepatic bile duct and duodenum.4

A pattern is identified, where the amelioration of ligation- induced acute pancreatitis in rats after combined bile and pancreatic juice replacement is greater than that achieved with replacement of pancreatic juice alone, and the ame- lioration achieved with replacement of pancreatic juice alone is greater than that achieved with replacement of bile alone. This pattern bears a striking similarity with the pat- tern observed in the experiments of Nakamura et al,7 in which stimulation of exocrine pancreatic protein secretion in healthy rats after external diversion of bile and (or) pan- creatic juice was greater after combined bile and pancre- atic juice diversion than after pancreatic juice diversion alone, and was greater after pancreatic juice diversion alone than after bile diversion alone. In view of these similari- ties, it is tempting to speculate that hyperstimulation of the exocrine pancreas against an obstructed duct exacerbates early ligation-induced acute pancreatitis in rats.

Cholecystokinin is implicated in the pathogenesis of var- ious forms of experimentally induced acute pancreati- ti~.~,~,‘~-‘~ If cholecystokinin is claimed to be the primary factor responsible for exacerbation of ligation-induced acute pancreatitis in rats,2%4 then one cannot explain the presence of considerable residual acute pancreatitis after replacement of pancreatic juice alone in the present study, where the plasma cholecystokinin concentrations were not significantly different from those of sham controls. Furthermore, previous studies have shown that L-364,7 18, a potent, highly specific, and long-acting competitive cholecystokinin-receptor antagonist, does not ameliorate morphologic changes associated with early ligation-in- duced acute pancreatitis in rats.3J4

There are two possible explanations for these incon- gruities: fist, that other factors in addition to cholecys- tokinin, such as the cholinergic system and other gut hor- mones (eg, secretin), are also involved in modulating basal and bile and pancreatic juice exclusion-stimulated exocrine pancreatic secretion20-25; and second, that the exacerbating

effects of bile and pancreatic juice exclusion from gut may be mediated by mechanisms other than hyperstimulation of the exocrine pancreas hitherto not evaluated in this model. Therefore, the multiple effects of bile and pancre- atic juice exclusion from gut should not be equated merely with hypercholecystokininemia alone, or even with ex- ocrine pancreatic hyperstimulation alone, and cholecys- tokinin may play only a contributory rather than a primary role in disease pathogenesis.

The shift of cathepsin B activity from the lysosome-en- riched subcellular fraction to the zymogen-enriched sub- cellular fraction observed in various forms of experimen- tal acute pancreatitis has been the focus of much attention.3J0,26 Based on in vitro observations that cathep- sin B activates trypsinogen to trypsin, it is suggested that colocalization of cathepsin B and zymogen enzymes is the triggering event in the pathogenesis of experimenta13s’0 and humanZ6 acute pancreatitis. However, immunocytochemi- cal studies show the presence of cathepsin B within zy- mogen granules of pancreatic acinar-cells in healthy rats,‘O and in vitro studies show that inhibitors of cathepsin B do not prevent zymogen enzyme activation when dispersed pancreatic acini are hyperstimulated with cholecystokinin.27 We studied the subcellular redistribution of cathepsin B af- ter bile and pancreatic-duct ligation merely as a parameter that reflects altered homeostasis at the subcellular level. Amelioration of the subcellular shift of cathepsin B with bile and pancreatic juice replacement in ligation-induced acute pancreatitis in rats is evidence at the subcellular level that bile and pancreatic juice exclusion from gut contributes to altered pancreatic acinar-cell homeostasis.

The primary aim of our study was to test the central hy- pothesis that combined bile and pancreatic juice exclusion from gut contributes to the early pathogenesis of ligation- induced acute pancreatitis in rats to a greater extent than exclusion of either bile or pancreatic juice alone. This cen- tral hypothesis was founded on physiologic principles eluci- dated in the elegant study by Nakamura et al,’ who showed that-in healthy rats-the magnitude of increase in ex- ocrine pancreatic protein secretion and plasma cholecys- tokinin concentration early after combined external diver- sion of both bile and pancreatic juice was much greater than that expected by the simple addition of the increase after external diversion of bile alone and the increase after ex- ternal diversion of pancreatic juice alone. In later studies, Miyasaka et a12* showed that luminal bile exerts feedback inhibition of exocrine pancreatic secretion in rats by direct and indirect mechanisms. According to their study, one mechanism is attributed to the stabilization of pancreatic juice proteases by calcium ions contained in bile (indirect mechanism), while a second mechanism is independent of protease activity and is attributed to a direct action of lu- minal bile salts probably mediated via neural pathways.28

Using the rat model, our present investigations have con- centrated entirely on the early phase of disease pathogen- esis, as it is advantageous to well understand the early events before later events can be explained. In a separate study, we investigated our hypothesis using the opossum model of early ligation-induced acute pancreatitis.29 The advantage of the opossum model was the ability to ligate the pancreatic duct without excluding bile from gut, as

LIGATION-INDUCED ACUTE PANCRElATITWSAMUEL ET AL

398 THE AMERICAN JOURNAL OF SURGERY” VOLUME 169 APRIL 1995

LIGATION-INDUCED ACUTE PANCREATITIS/SAMUEL ET AL

opossums possess a separate bile duct and a separate pan- creatic duct that unite to form a common channel close to the duodenum.6 By administering pancreatic enzyme ex- tracts via a duodenal fistula and leaving the bile duct in- tact, it was possible to avoid the development of a “donor opossum model” for the time being. There was also the added advantage of testing our hypothesis in ligation-in- duced acute pancreatitis in two species. Furthermore, the opossum model is regarded as a more severe model of lig- ation-induced acute pancreatitis than the rat model.6*15,30

The results of our initial experiments in the opossum model of ligation-induced acute pancreatitis were encour- aging. With qualitative assessment of pancreatic morpho- logic changes associated with 24 hours of bile-and pancre- atic-duct ligation in opossums, maintaining bile flow with pancreatic enzyme replacement ameliorated the acute pan- creatitis histology score by 51%; conversely, maintaining bile flow without pancreatic enzyme replacement did not significantly decrease the histology score.2g Nevertheless, with the determination of numerical and areal densities of acinar-cell necrosis (quantitated as described in the present- study), maintaining bile flow without pancreatic-enzyme replacement significantly ameliorated acinar-cell necrosis; and maintaining bile flow with pancreatic enzyme re- placement achieved even greater amelioration of acinar-cell necrosis, to values not significantly different from that of sham controls (unpublished observations).

In summary, we have developed and characterized a new experimental model to study early events in the patho- genesis of ligation-induced acute pancreatitis in rats. Using this model we have shown that, in this experimental corol- lary of early gallstone-induced acute pancreatitis, com- bined bile and pancreatic juice exclusion from gut con- tributes to disease pathogenesis to a greater extent than does exclusion of either bile or pancreatic juice alone.

9. Turkelson Ch$ Hamilton J. Distribution of cholecystokinin forms inintestinal secretory granule subtypes. Endocrinology. 1992; 13 1: 2533-2539. 10. Willemer S, Bialek R, Adler G. Localization of lysosomal and di- gestive enzymes in cytoplasmic vacuoles in caerulein-pancreatitis. Histochemistry. 1990;94:161-170. Il. Bajkowski AS, Frankfater A. Specific spectrophotometric assays for cathepsin Bl. Anal Biochem. 1975;68:119-127. 12. Miyasaka K. Kitani K, Green G. The sequential changes in pan- creatic exocrine function after abdominal surgery in the rat. Pancreas. 1986;1:347-353.

REFERENCES 1. Kueppers PM, Russell DH, Moody FG. Reversibility of pancre- atitis after temporary pancreaticobiliary duct obstruction in rats. Pancreas. 1993;8:632-637.

13. Mashige F, Imai K, Osuga T. A simple and sensitive assay of to- tal serum bile acids. Clin Chim Acta. 1976;70:79-86. 14. Toriumi Y, Samuel I, Wilcockson DP, et al. Octreotide and chole- cystokinin antagonist reduce edema in obstruction-induced acute pan- creatitis. J Lab Clin Med. 1993;122:450-454. 15. Lerch MM, Saluja AK, Dawra R, et al. Acute necrotizing pan- creatitis in the opossum: earliest morphological changes involve aci- nar cells. Gastroenterology. 1992;103:205-213. 16. Gomez G, Townsend CM, Green D, et al. Involvement of chole- cystokinin receptors in the adverse effect of glucocorticoids on diet- induced necrotizing paucreatitis. Surgery. 1989;106:23&238. 17. Lampel M, Kern HF. Acute interstitial pancreatitis in the rat in- duced by excessive doses of a pancreatic secretagogue. Virchows Arch A Path01 Anat Histopathol. 1977;373:97-117. 18. Niederau C, Liddle RA, Ferrel LD, Grendell JH. Beneficial ef- fects of cholecystokinin receptor blockade and inhibition of prote- olytic enzyme activity in experimental acute hemorrhagic pancreati- tis in mice. Evidence for CCK as a major factor in the development of acute pancreatitis. .I Clin Invest. 1986;78: 105&1063. 19. Nordback lH, Clemens JA, Cameron JL. The role of cholecys- tokinin in the pathogenesis of acute pancreatitis in the isolated pan- creas preparation. Surgery. 1991;109:301-306. 20. Dressel TD, Goodale RL, Arneson MA, Bomer JW. Pancreatitis as a complication of anticholinesterase insecticide intoxication. Ann Surg. 1978;189:199-204. 21. Kim CK, Lee KY, Wang T, et al. Role of endogenous cholecys- tokinin on vagally stimulated pancreatic secretion in dogs. Am J Physiol. 1989;257 (suppl: Gastrointest Liver Physiol2CJ):G944-G949. 22. Louie DS, May D, Miller P, Owyang C. Cholecystokinin medi- ates feedback regulation of pancreatic enzyme secretion in rats. Am J Physiol. 1986;250 (suppl: Gastrointest Liver Physiol 13): G2524259.

2. Murayama KM, Samuel I, Toriumi Y, et al. Increased circulating cholecystokinin in obstruction-induced acute pancreatitis: I. Bile duct obstruction with and without pancreatic duct obstruction. J Surg Res. 1993;54:12&131. 3. Ohshio G, Saluja A, Steer ML. Effects of short-term pancreatic duct obstruction in rats. Gastroenterology. 1991; 100: 196202. 4. Toriumi Y, Samuel I, Wilcockson DP, et al. Increased circulating cholecystokinin in obstruction-induced acute pancreatitis: II. Pancreatic duct obstruction with and without bile duct obstruction. J Surg Rex 1993;54:132-135. 5. Murayama KM, Drew JB, Yokoo H, Jochl RJ. Bile exclusion from gut exacerbates acute pancreatitis caused by pancreatic duct obstruc- tion in rats. Pancreas. 1991;6:175-181. 6. Senninger N, Moody FG, Coelho JCU, Van Buren DH. The role of biliary obstruction in the pathogenesis of acute pancreatitis in the opossum. Surgery. 1986;99:688-693. 7. Nakamura R, Miyasaka K, Funakoshi A, Kitani K. Interactions be- tween bile and pancreatic juice diversions on cholecystokinin release and pancreas in conscious rats. Proc Sot Exp Biol Med. 1989;192: 182-186. 8. Turkelson CM, Dale WE, Reidelberger R, Solomon TE. Development of CCK radioimmunoassay using synthetic CCK 10 im- munogen. Regul Pept. 1986;15:206-217.

23. Moriyoshi Y, Shiratori K, Watanabe S, Takeuchi T. Potentiating effect of CCK and secretin on rat exocrine pancreas and its choQner- gic dependence. Pancreas. 1991;6:603-608. 24. Nakano I, Tawil T, Spannagel AW, et al. Atropine enhances food- stimulated CCK secretion in the rat. Pancreas. 1990;5:621-625. 25. Watanabe S, Takeuchi T, Chey WY. Mediation of trypsin in- hibitor-induced pancreatic hypersecretion by secretin and cholecys- tokinin in rats. Gastroenterology. 1992;102:621428. 26. Figarella C, Miszczuk-Jamska B, Barrett AJ. Possible lysosomal activation of pancreatic zymogens. Biol Chem Hoppe Seyler. 1988;369(suppl):293-298. 27. Leach SD, Modlin IM, Scheele GA, Gorelick FS. Intracellular ac- tivation of digestive zymogens in rat pancreatic acini. J C/in Invest. 1991;87:362-366. 28. Miyasaka K, Sazaki N, Funakoshi A, et al. Two mechanisms of inhibition by bile on luminal feedback regulation of rat pancreas. Gastroenterology. 1993;104: 1780-1785. 29. Samuel I, Toriumi Y, Wilcockson DP, Joehl RJ. Patbogenesis of pancreatic duct obstruction-induced acute paucreatitis in opossums is influenced by duodenal exclusion of pancreatic enzymes. Am J Surg. 1993;165:742. Abstract. 30. Samuel I, Toriumi Y, Yokoo H, et al. Ligation-induced acute pan- creatitis in rats and opossums-a comparative morphologic study of the early phase. J Surg Rex August 1994;57:299-311.

THE AMERICAN JOURNAL OF SURGERY” VOLUME 169 APRIL 1995 399