Effect of Portacaval Shunt and Arterialization of the Liver on Bile Acid Metabolism

GASTHOENTEHOLOGY

Copyri~ht

69: 338-341 , 1975

Vol. 69, No.2 Printed in U. S .A .

© 1975 by the Willia ms & Wilkins Co.

EFFECT OF PORTACAVAL SHUNT AND ARTERIALIZATION OF THE LIVER ON BILE ACID METABOLISM WoLFGA NG HoRAK, GRABNER ,

M.D .,

ALFRED GANGL ,

M.D .,

Jos EF F uNOVIC::> .

M.D.,

ANn Gw~tr.

M .D.

Department of Gastroenterology and Department of Surgery (!), Univ ersity of Vienna, Austria

The effect on bile acid clearance of portacaval end-to-side shunt, with and without arterialization of the hepatic portal venous bed, was studied in dogs. Portacaval shunt reduced liver blood flow by 75%, and peripheral serum bile acid concentration rose from 3 to 239 ,uM. Arterialization maintained total liver blood f1ow within physiological range and serum bile acid concentration was only 19 J.LM. Since hepatic bile acid extraction was similar in both groups, bile acid clearance was determined largely by total liver blood f1ow. The results of this study indicate that hepatic clearance of substances originating from portal blood is impaired in portacaval shunt and can be restored by increasing the arterial blood supply to the liver. Portacaval end-to-side shunt combined with arterialization of the intrahepatic portal venous bed is a new method in the management of portal hypertension in patients with cirrhosis of the liver. 2-4 The effect of the substantial changes in hepatic hemodynamics cauaed by this procedure on liver function has not been studied in detail. To obtain such information we have determined the clearance of endogenous bile acids which normally reach the liver via the portal vein. Serum bile acid concentrations were measured in peripheral, portal, and hepatic venous blood of dogs with a portacaval shunt alone, and with portacaval shunt combined with arterialization of the liver. Received February 27 , 1974. Accepted March 28, 1975. A prelimina ry report was presented at the Seventh Meeting of the European Association for the Study of the Liver, Arnhem, September 1972, and published in abstract form . 1 Address requests for reprints to: Dr. W. Horak, Ordinariat fi.ir Gastroenterologie, Garnisongasse 13, A-1090 Wien, Austria. The authors are grateful for the technical assistance of Mr. K. Winterstein . 338

Materials and Methods Animals and procedures. Nine mongrel dogs (14 to 25 kg), maintained on a standard laboratory diet, were anesthetized with intravenous thiopentone (Pentothal), 30 mg per kg, followed by nitrous oxide through an endotracheal tube. The abdomen was entered, calibrated flow probes were placed around the hepatic artery and portal vein, and flows were measured by a square wave electromagnetic flow meter (Nycotron, model 372). Blood was taken from the portal and peripheral veins for estimation of serum bile acids, and a liver biopsy was obtained . In all dogs an end-to-side portacaval shunt was constructed . Additionally , in 4 dogs arterialization of the portal vein was performed by grafting a segment of the external jugular vein between the proximal cut end of the portal vein and the aorta. ' By plication of the graft and utilizing a flow probe the blood flow was adjusted to equal the preshunt portal flow. Blood was drawn preoperatively and at 2-week intervals thereafter for determination of bile acids and routine laboratory tests. Ten weeks postoperatively a second laparotomy was performed , and blood flows of the hepatic artery and the vein graft were again measured. Hepatic venous blood was obtained from a polyethylene catheter inserted directly through the wall of the left common hepatic vein. A liver

August 1975

BILE ACIDS IN SHUNT ANIJ ARTMUAUZATION

biopsy was taken, the dogs were killed, and the liver was removed and weighed . Analytical methods. Bile acids were extra cted from serum as described by Mayer et 6 al. using tetraheptylammonium chloride 7 (Serva, Heidelberg) , and were determined fluorimetrically and enzymatically according to Murphy et a !. 8 with purified hydroxyste roid dehydrogenase (Worthington Biochemica ls Corp., Freehold, N.J.). In duplicate de te rminations the coefficient of varia tion was :l.5%. Standard methods were used for determination of serum bilirubin, SGOT, ammonia, and cholesterol.• Calculations. Hepatic bile acid extract ion factor (E) 10 weeks postoperat ively was estimated from bile acid concentrations in systemic (BA s) and in hepatic venous (BAh .,) blood. E = BA s - BAh u RAs

Bile acid clearance ( C l) was calculated from t otal hepat ic blood t1ow (HBF) and the extraction factor (E). Cll ml / mln l kf.! bcJCiy w(•ight

1

=

HBF

X

E

Res ults are presented as mean ± 1 SEM . St u dent's t-test was performed for statistical analysis .

Results Before construction of the portacaval shunt, portal venous blood flow was 31.9 ± 3.4 ml per min per kg of body weight and hepatic artery f1ow was 10.4 ± 1.2 ml per min per kg. Portacaval shunting caused a TABLE

decrease in total hepatic blood flow by 7fi %, which could he prev ented by arterialization of the shunted liver (table I) . Body weight remained consta nt in all a nimals during 10 weeks . Liver weight was 2.8:1 :!: 0.17'Y,. of body weight in fi normal dogs. After porta cava l shunt it was i .9G I · 0.15 %, which indicates a significant reduction of liver mass (P < O.!lOOG) , whereas in arterialized dogs liver weight was ~.44 1 0.14% of body weight, not significantly different from controls nor from shunted dogs. Light microscopy of livers 10 weeks a fter portacava l shunt revealed marked fatty meta morphos is, which was virtual ly a bsent in a rterialized livers. Serum bilirubin levels, 0 .~ 1 0.1 mg p e r 100 ml preopera tively, remained un changed throughout the entire experiment. in both shunted a nd arteria li zed anima ls . SGOT showed an increase (!> .- O.Ofi) in s hunted anima ls from 11 :1. I mlJ to 77 1: 52 mU 10 weeks postoperatively, whereas in a rterial ized animals SGOT remained at preoperative levels. Serum a mmonia con centration (48 ± 5 J.li!: per 100 ml preoperative ly) increased markedly after operation in both groups, but was significantly lower in arterial izecl (82 .:Jc 1:3 11g per 100 ml) than in shunted dogs (Ififi : t Hi 11g per 100 ml; f> < O.OOfi). Tota l serum cholesterol (204 1. fi rng per 100 ml preoperat ively) , d eclined significantly (P < O.OG) and reached a new level by the 4th week, which was maintained

l. Effect of portacaval shunt and arterialization of the liver on hepat ic blood flow and hile acid metabolism"

No. ani·

mals

T ota l hepat ic blood now

Serum hilc nc id concentration

Pcriphcrnl vein

m l/min/k~

9 5

P o rtacaval shunt + arterialization•

4

42.3 ± 4.3 10.5 ± 0.8 (P < O.OOl) 40.3 ± 5.1 (NS)

fa ctor

2.5 238.6 (P < 18 .7

± 0.5 ± 41.5 0.001) ± 9.2 (P < O.OM

body wt

47.5 ± 21.3 2.2

ClcnranC'c ml/min/h~

!-'moles/liter

body wt

Pre-shu nt P o r tacaval shunt•

Hepatic vei n

Hcpal ic hilc acid E xtrac tion

± 1.~

0.92' 0.81 ± 0.06 0 .90 ± 0.04

39.0 ± :l.9 8.6 ± 1.0 (P < 0.001) 36. 1 ± 4.1) (NSl

a All values are means ± SEM; significance of differen ce to pre-shunt values was obtain ed by Student 's t test. NS, not s ignifican t. • Ten weeks postoperatively. ' Value taken from O ' M a ill e et al.'"

HORAK ETAL.

340

during the remaining observation period at 159 ± 13 mg per 100 ml in shunted and 161 ± 8 mg per 100 ml in arterialized dogs. Mean serum bile acid concentration was 3 ± 1 ,umoles per liter in systemic blood and 51 ± 9 ,umoles per liter in portal venous blood before construction of the portacaval shunt. Ten weeks postoperatively serum bile acids were 239 ± 42 ).£moles per liter in dogs with a portacaval shunt alone, which was significantly higher than the pre-shunt values in peripheral (P < 0.001) and in portal blood (p < 0.005). By contrast, in shunted dogs with arterialization of the liver, serum bile acids were 19 ± 9 ).£moles per liter, significantly (P < 0.005) less than in dogs with a shunt alone (fig. 1). Hepatic bile acid extraction factor was 0.81 in the dogs with a shunt alone and 0.90 in the arterialized animals, similar to 0.92 described in non-shunted dogs. 10 Therefore a linear correlation was found between hepatic bile acid clearance and total liver blood flow (fig. 2). Clearance was significantly (P < 0.001) reduced in dogs with a portacaval shunt alone and was maintained at normal values in the dogs with shunt and arterialization (table 1). Discussion Results of this study indicate that combination of portacaval shunt with arterialization of the liver is associated with less severe sequels both histologically and bio3oo Serum bile acids J.Jmoles/1

2oo

1oo

• •

8

10

Weeks

FIG. 1. Serum bile acid concentrations in 5 dogs with portacaval shunt (e) and 4 dogs with portacaval shunt and arteria lization of the portal vein (A) during 10 weeks after operation. The vertical lin es show the standard errors of the means. Statistically significant differences between the two groups are designated by an asterisk (P < 0.005).

Vol. 69, N o.2

0, 70 ~

......

c::

.E 60

/

......

E 'ID 50 g

..

(I)

ro 40

QJ

(3

~30 (I)

~ 20 (]) 10

..

..·······

_!········

00~~~~--~--~--~--~~

10

20

30

40

50

60

70

Hepatic blood flow(ml/min/kg) FIG. 2. Relationship between hepatic blood flow and hepatic bile acid clearance in 9 dogs pre-shunt (0) , 5 dogs with portacaval shunt (e) , and 4 dogs with shunt and arterialization of the portal vein (A). The slopes of the regression lines represent hepatic bile acid extraction. Bile acid clearance for pre-shunt animals was calculated by multiplying hepatic blood flow by the extract ion factor 0.92, described by O'Maille et al. ' 0

chemically than portacaval shunt alone. This is in agreement with findings described by others. 11 In animals with a portacaval shunt a striking increase of systemic serum bile acid concentration was observed which theoretically can be related to two main factors. First, bile acid-rich mesenteric venous blood is shunted into the systemic circulation . Second, the decrease of hepatic bile acid clearance causes accumulation of bile acids in peripheral blood to concentrations which exceed those in portal blood before shunting. Clearance depends on blood flow and extraction; the latter may be influenced by several factors. Although a decrease in liver blood flow could improve hepatic extraction fraction, fatty metamorphosis and atrophy of the liver after portacaval shunt may have the opposite effect. Interference of these factors can explain our finding that hepatic bile acid extraction after portacaval shunt with and without arterialization was similar to normal dogs. Therefore, bile acid clearance in animals with a portacaval shunt was determined largely by the reduc-

August 1975

BILE ACIDS IN SHUNT AND ARTERIAUZATION

tion of total hepatic blood f1ow . The results obtained in dogs with shunt and arterialization support this concept: arterialization maintained total hepatic blood flow without changes in hepatic bile acid extraction, thus restoring bile acid clearance and preventing accumulation of bile acids in systemic blood, despite the portacaval shunt. It appears from our data that the shunted animals removed more bile acids from blood than normal or arterialized dogs. This suggests that the redistribution of bile acids between the systemic and the enterohepatic circulation leads to further derangements of bile acid kinetics, possibly changes in synthesis or frequency of enterohepatic cycling of bile acids. In contrast to serum bile acids serum bilirubin levels did not change in a qualitatively similar way. This presumably is due to the fact that, unlike bile acids, conjugated bilirubin is not reabsorbed efficiently by intestinal mucosa, 12 and therefore does not reenter the systemic circulation via shunted mesenteric venous blood. A decrease in plasma cholesterol level after end-to-side portacaval shunt in familial hypercholesteremia has been reported.13 It is interesting that we observed a similar effect of portacaval shunting in healthy dogs, whether they were arterialized or not. The finding that the cholesterol-lowering effect was not inf1uenced by restoring total hepatic blood flow suggests that it depends on quality rather than on quantity of blood that perfuses the liver. Patients with portal-systemic anastomoses frequently suffer from encephalopathy. Although the factors responsible are not yet known, shunt encephalopathy may be due to inadequate hepatic catabolism of neurotoxic substances derived from intestinal absorption. If such hypothetical compounds behave as bile acids, they will bypass the liver and, because of impaired hepatic clearance, may accumulate to high concentrations in systemic blood. One might speculate that arterialization will improve the clearance of these substances.

341

Our results obtained in normal dogs do not necessarily apply to patients with cirrhosis of the liver; however, they could explain the less frequent incidence of encephalopathy reported in patients after portacaval shunt combined with arterialization. • REFERENCES 1. Horak W, Gangl A, Funovics ,J , et al: The effect of

2.

:3.

4.

5.

6.

7.

8.

9.

10 .

11 .

12. 13.

portacaval shunt and arterialization of the I iver on serum bile salts in the dog (abstr). Digestion 6:262, 1972 Adamsons IU, Kinkhabwala M. Moskowitz H, et al: Portacaval shunt with arterialization of the hepatic port ion of the portal vein. Surg Gynecol Obstet 1:l5:!i29-5:l!i. 1972 Maillard .JN, Henhamou .JP, HuetT B: Arterial izntion of the li ver with portacaval shunt in the treatment of portal hypertension due to intrahepatic block. Surgery H7 :88:l-R90, UJ70 Maillard ,JN, RuefT H, Prandi D. et al: Hepatic arterialization and portacnval shunt in hepatic cirrhosis. Arch Surg 108::ltG-:l20. 1974 Fisher B, Russ C, Updegraff H: A suitable technique for total arterialization of the dog liver. Surgery :l5:879-8il4, 195:l Mayer D, Haindl H , Koss F'W, ct. al : Enzymatische Bestimmung von Gallensauren in KorperfiUssigkeiten unci Gcwehen. Z Anal Chern 24:3:242-248 , 196il Hofmann AF': Efficient extraction of bile acid conjugates with tetraheptylammoniurnchloridc. a liquid ion exchanger. J Lipid Hes R:!ifi-!iR, 19(;7 Murphy GM, Billing BH, Baron DN: A lluorimctric and enzymatic method f(>r the estimation of serum total bile acids. ,J Clin Pathol 2:l:!i94 --fi98. 1970 Watson D: A simple method f<>r the determination of serum cholesterol. Clin Chim Acta 5:6:37-64:l. 1960 O'Maille ERL. Richards TG . Short AH: The in11uence of conjugation of cholic acid on its uptake and secretion: hepatic extraction of taurocholate and cholate in the dog. ,J Physiol 189:337-350. 1967 Adarnsons R: Arterializat ion of the liver in combination with a portacaval shunt (abstr) . Gast roenterology 66:287, 1974 Schmid R: Bilirubin metabolism in man. N Eng! J Med 287:703-709, 1972 Starzl TE, Putnam CW. Chase HP, et al: Portacaval shunt in hyperlipoproteinaemia. Lancet. 2:940-944, 197:1