causes and management of portal hypertension in the pediatric population

PORTAL HYPERTENSION 1089-3261/01 $15.00 + .OO

CAUSES AND MANAGEMENT OF PORTAL HYPERTENSION IN THE

PEDIATRIC POPULATION

Frederick C. Ryckman, MD, and Maria H. Alonso, MD

The survival of children with portal hypertension has improved during the past decade. This improvement has resulted from (1) progress in the pharmacologic control of acute portal hypertensive hemorrhage; (2) improved efficacy and safety of endoscopic methods to treat acute esophageal variceal hemorrhage that reduce the risk of rebleeding; (3) recognition of the role for advanced surgical therapy (portocaval shunts); and (4) improved success with pediatric liver transplantation as a definitive treatment for children with end-stage liver disease. This article examines the role of each of these treatment modalities in the management of portal hypertension in children.

DEFINITION AND ETIOLOGY

Portal hypertension is defined as an elevation of the portal pressure above 10 to 12 mm Hg. In healthy children, portal pressure rarely exceeds 7 mm Hg. Elevation of the portal pressure is most commonly caused by obstruction of portal venous flow by a presinusoidal, sinusoidal, or postsinusoidal blockage, although increased splanchmic blood flow may contribute in some cases. In response to increased pressure within the portal circulation, collateral circulatory pathways develop connecting the high-pressure portal vasculature to the low-pressure systemic venous system. These collaterals may form at any site; however,

From the Department of Surgery, Division of Pediatric Surgery, University of Cincinnati Children’s Hospital Medical Center, Cincinnati, Ohio

CLINICS IN LIVER DISEASE

VOLUME 5 * NUMBER 3 * AUGUST 2001 789

790 RYCKMAN & ALONSO

the communications most commonly occur within the esophageal wall, connecting the coronary and short gastric veins to the esophageal venous plexus, which communicates with the intercostal, azygos, and hemiazy- gous veins. As portal pressure increases, esophageal varices developing within this plexus become the site with the highest risk for massive hemorrhage. Less threatening collateral communications can develop between the recanalized umbilical vein and abdominal wall systemic veins (caput medusa), the inferior rectal veins as hemorrhoids, and around the retroperitoneal pancreas and duodenum. In addition, any surgical union between the portal and systemic venous circulation, such as occurs with intestinal stomas, is a possible, and often problematic, site of variceal development. Favorable collaterals developing within the tissues surrounding the pancreas, duodenum, and left kidney, form spontaneous splenorenal shunts.

The progressive development of collaterals connecting the portal and systemic circulation has the beneficial effect of decreasing portal pressure. This effect is vitiated, however, by the concurrent development of a hyperdynamic circulatory Portal hypertension has been associated with the presence of autonomic nervous system dysfunction and an excess of circulating cytokines leading to tachycardia, decreased systemic and splanchnic vascular resistance resulting from vasodilatation, plasma volume expansion, increased cardiac output, and subse- quently increased portal inflow.

The combination of increased portal inflow, venous outflow obstruction, and the remarkable collateral circulation that develops accounts for many of the complications associated with portal hypertension. Superfi- cial submucosal collaterals, especially those in the esophagus and stomach and, to a lesser extent, those in the duodenum, colon, or rectum, are prone to rupture and bleeding. In addition, prominent submucosal arteriovenous communications between the muscularis mucosa and dilated precapillaries and veins within the stomach result in vascular ectasia, or congestive hypertensive gastropathy, significantly contribut- ing to the risk of hemorrhage from the stomach.

Each of the causes of elevated portal pressure shares the common mechanism of increased resistance to blood flow from the visceral or splanchnic portal circulation to the right atrium. In children, the location of this increased resistance can be (1) prehepatic (or presinusoida1)- usually within the portal vein and its primary feeding branches; (2) intrahepatic-caused by presinusoidal obstructions (congenital hepatic fibrosis, congenital or acquired arterial-portal fistula, or schistosomiasis), postsinusoidal cirrhosis, or veno-occlusive disease; or (3) posthepatic- caused by to hepatic vein obstruction. Although this anatomic descrip- tion is helpful for structurally organizing a differential diagnosis (as shown in the box), the primary factor influencing the prognosis and treatment algorithm is the intrinsic status of the liver. Presinusoidal obstruction does not result in impairment of hepatic synthetic function, and coagulopathy is absent. Treatment should be directed toward the prevention of hemorrhage through palliative interventional procedures

PORTAL HYPERTENSION IN THE PEDIATRIC POPULATION 791

while spontaneous collateral venous channels develop. In contrast, postsinusoidal obstruction is characterized by hepatic synthetic compro- mise, coagulopathy, and progressive hepatic failure. Although intervention to prevent or treat potentially fatal complications may be necessary, definitive correction with liver transplantation is often required.

Pediatric Diseases Associated with Portal Hypertension Presinusoidal Hypertension

Venous obstructions Portal vein thrombosis/cavernous transformation Splenic vein thrombosis Portal vein malformation (congenital)

Congenital hepatic fibrosis Arteriovenous fistula Schistosomiasis Hepatoportal sclerosis

Sinusoidal Hypertension

Hepatocellular disease Autoimmune hepatitis Hepatitis B, C Wilson’s disease a,-antitrypsin deficiency Glycogen storage disease type IV Toxins and drugs Histiocytosis X Gaucher’s disease Peliosis

Biliary tract disease Biliary atresia Cystic fibrosis Choledochal cyst lntrahepatic cholestasis syndromes Sclerosing cholangitis

Sinusoidal veno-occlusive disease

Postsinusoidal Hypertension

Budd-Chiari syndrome Inferior vena cava obstructions Chronic congestive heart failure Veno-occlusive disease (particularly following bone marrow transplantation) Prothrombotic disease

PRESINUSOIDAL OBSTRUCTION

The most common type of presinusoidal obstruction is extrahepatic portal vein obstruction (EPVO) at any level of the portal vein. Umbilical


vein infection in infancy, with or without umbilical vein cannulation, has been associated with the development of occult thrombosis of the portal vein. Infection can spread from the umbilical vein to the left branch of the portal vein and eventually to the main portal venous channels, leading to phlebitis and subsequent thrombosis. Similar infections in older children, such as perforated appendicitis, primary peritoni- tis, and inflammatory bowel disease have also been identified as predisposing factors, as have primary biliary tract infections or cholangitis. Inherited abnormalities predisposing to hypercoagulability, such as factor VLeiden mutation and protein C, protein S, and antithrombin I11 defi- ciencies, as well as hyperviscosity or polycythemia in infancy can all lead to secondary thrombosis, especially when accompanied by neonatal dehydration or systemic infection and phlebitis.20, 67 Ando et a1 suggested that embryologic malformations resulting in tortuous, poorly developed portal veins could be a primary cause for EPVO or predispose to an increased risk of thromb~sis.~ Anatomic abnormalities can also include webs or diaphragms within the portal vein leading to obstruction. The possible role of congenital abnormalities in EPVO is supported by the concurrent presence of other congenital anomalies in 40% of children with no identified postnatal cause for EPVO, compared with an incidence of only 12% in children with a defined cause such as umbilical vein ~atheterization.5~, 67 Presinusoidal obstruction can also result from congenital hepatic fibrosis, schistosomiasis, and hepatoportal sclerosis. In rare cases, increased portal blood flow is attributed to congenital or acquired arteriovenous fistula within the portal system.l4 Despite thor- ough evaluation, more than one half of reported EPVO cases have no identifiable cause.

POSTSINUSOIDAL OBSTRUCTION

Postsinusoidal obstruction of the portal venous system results from intrinsic liver disease, cirrhosis, or obstruction to the hepatic vein outflow from the liver. Cirrhosis resulting from primary liver diseases is the most common cause of postsinusoidal portal hypertension, with venous obstruction arising as the result of intrahepatic scarring. The numerous causes of cirrhosis include recognized disorders such as extrahepatic biliary atresia, metabolic liver disease such as or,-antitrypsin deficiency, Wilson’s disease, glycogen storage disease type IV, hereditary fructose intolerance, and cystic fibrosis. Because many of these conditions are associated with progressive liver failure, the primary treatment in most cases is liver transplantation. Successful transplantation corrects the portal hypertension and its complications such as hypersplenism, ascites, and synthetic liver failure. In the clinical situation when hepatic synthetic failure is not present or is only slowly progressive, direct treatment of portal hypertension or its complications is indicated.

Hepatic vein obstruction (Budd-Chiari syndrome) can occur as the result of obstruction to the hepatic veins at any point from the sinusoids


to the entry of the hepatic veins into the right atrium/inferior vena cava. Although a specific cause is often not found, thrombosis can occur as a complication of neoplasms, collagen vascular disease, infection, trauma, or hypercoagulability states. Veno-occlusive disease has been recognized as one of the most frequent causes of hepatic vein obstruction in children. In this disorder, the centrilobular venules or sublobular hepatic veins are occluded. Most cases occur after total-body irradiation with or without cytotoxic drug therapy associated with bone marrow transplantation? This condition has also occurred after the ingestion of herbal remedies containing the pyrrolizidine alkaloids, which are sometimes taken as medicinal teas.n

DIAGNOSIS AND EVALUATION

Clinical history and examination should concentrate on identifying factors which predispose to the development of cirrhosis, including a family history of inherited metabolic disease and possible exposure to viral or toxic pathogens. Clinical examination findings suggesting underlying liver disease (ascites, liver size and contour, nutritional status), hypersplenism (spleen size, bruising), or hepatopulmonary syndrome (spider angiomas, clubbing, cyanosis) contribute to diagnostic evaluation and therapeutic planning. Historical events preceding portal vein thrombosis should be sought. Hypercoagulability and its complications should be evaluated in both the patient and family members because of the inherited basis for these protein abnormalities.

Imaging tests are essential to confirm the presence of portal hyper- 'tension, define the portal venous anatomy, and formulate options for future therapy. Initial screening with ultrasound can suggest the presence of chronic liver disease and should determine portal venous patency. Doppler examination can depict both the direction of portal flow and the degree of hepatopetal flow, which correlates with the risk of variceal hemorrhage. The branches of the portal venous system are examined to exclude splenic vein thrombosis or widespread portal system thrombosis. Magnetic resonance angiography or contrast-enhanced computed topography has replaced mesenteric angiography when further definition of portal anatomy is necessary (e.g., when liver transplantation or portosystemic shunt procedures are planned).42

Upper gastrointestinal endoscopy is the most accurate and reliable method for detecting esophageal varices and for detecting the source of acute gastrointestinal hemorrhage. This procedure is especially valuable during acute hemorrhage, because up to one third of patients with known varices may have bleeding from other sources such as portal hypertensive gastropathy or gastric or duodenal ~lcerat ions.~~ Endos- copy can also identify features associated with an increased risk for future hemorrhage, such as large varices, red spots apparent over varices representing fragile telangiectasis within the shallow submucosa, and


portal hypertensive gastropathy. Endoscopy is also used to initiate treatment when acute bleeding varices are identified.

Liver biopsy may be helpful in determining the cause of intrinsic liver disease and in defining further therapy or need for transplantation.

TREATMENT OF PORTAL HYPERTENSIVE COMPLICATIONS

The decision to undertake pharmacologic, endoscopic, or surgical treatment for portal hypertension must be based on the natural history of the disease and the possibility of life-threatening complications. The prognosis is related to the primary cause of the portal hypertension. It has been generally accepted that in patients with portal hypertension caused by EPVO the risk of acute variceal bleeding decreases with age, concurrent with the development of spontaneous portosystemic collaterals. This postulated natural history has been the primary argu- ment supporting conservative management of hemorrhage in these patients, using endoscopic therapy to obliterate esophageal varices while awaiting the development of favorable retroperitoneal and peripancreatic collaterals. In studying the natural history of patients with EPVO, however, Lykavieris et a1 found little to support this theory.50 Their review of 44 patients, followed for a mean of 8 years after their twelveth birthday, showed that the actuarial risk of hemorrhage increased with age, from 49% at 16 years to 76% at 24 years. Children who had experienced bleeding complications before age 12 years had a significantly greater chance of bleeding again by age 23 years than did those who had not bled by 12 years of age (93% versus 56%; P = 0.007). In those with grade I1 or I11 varices, the actuarial risk of hemorrhage was 60% at age 18 years and 85% at age 23 years, compared with a risk of zero in patients with no esophageal varices or grade I varices at 12 years of age who experienced no episodes of hemorrhage. This high rate of bleeding in adolescence and early adulthood challenges the assumption that these complications inevitably decrease with time and suggests that a high-risk population for rebleeding can be identified and selected for effective preemptive treatment.

In patients with intrinsic liver disease, therapeutic choices are influ- enced by the probability of disease progression and the potential need for liver transplantation in the future. A significant number of these patients will require temporizing endoscopic treatment or surgical portosystemic shunt therapy to treat complications or maintain stability before needing liver replacement.

The most common portal hypertensive complication is gastrointestinal bleeding. Regardless of the site and mechanism, initial therapy is directed toward fluid resuscitation and, when necessary, blood replacement. A nasogastric tube should be placed to confirm the upper gastrointestinal tract as the source of bleeding and for evacuation of blood from the stomach. An H, receptor blocker or proton pump inhibitor should


be administered to decrease the risk of further bleeding from gastric erosions. In patients with hepatic synthetic dysfunction and coagulopathy, administration of vitamin K, fresh-frozen plasma or cryoprecipitate, and platelets when thrombocytopenia is present may also be necessary. Adequate volume resuscitation is essential; however, volume overload from excessive transfusion or crystaloid administration is counterpro- ductive, leading to a further increase in portal pressure and continued hemorrhage.

PHARMACOLOGIC TREATMENT

Pharmacologic intervention to decrease portal pressure may be considered in patients with continued bleeding. A variety of options are now available when intervention is required.

Vasopressin

Vasopressin increases splanchnic vascular tone, decreasing splanchnic arterial inflow and thus decreasing portal venous pressure. It is administered as an initial bolus of 0.3 units/kg over 20 minutes followed by a continuous infusion of 0.002 to 0.005 U/kg/minute. Intra-arterial infusion into the superior mesenteric artery has no advantage over intravenous routes. Several randomized, controlled trials in adults have verified the significant beneficial effect of vasopressin in controlling variceal hemorrhage.16 Although vasopressin infusion has been associated with control of variceal hemorrhage in 53% to 85% of cases in children, its use has been limited by the associated systemic vasoconstriction to the heart, bowel, and kidneys that impairs cardiac function and exacerbates fluid retention.39 Nitroglycerin has been used to augment the decrease in portal pressure and ameliorate the untoward systemic effects but is inappropriate when systemic blood pressure is unstable.

Somatostatin and Octreotide

Somatostatin, a 14-amino acid peptide reduces splanchnic blood flow by selective mesenteric vascular smooth muscle constriction and therefore does not precipitate the systemic vasoconstriction seen with vasopressin infusions. Its short half-life complicated treatment, leading to the development of octreotide, an 8-amino acid synthetic somatostatin analogue. Octreotide can be administered subcutaneously but is best used as a continuous intravenous drip (25-50 pg/m2/hour or 1.0 pg/ kg/hour). In adult studies, both somatostatin and octreotide have achieved excellent results in controlling acute variceal hemorrhage compared with vasopressin or mechanical tamponade. Studies in children to confirm this success have not been undertaken, and its use and efficacy


in children can only be inferred. Adult trials have also shown somatostatin to be equivalent to endoscopic sclerotherapy for initial control of acute bleeding.16

p-Blockers

P-Blockers have no role in the treatment of acute variceal hemorrhage. They have, however, been used in an effort to prevent variceal hemorrhage in high-risk patients. Therapy is directed toward decreasing the heart rate by 25%, thereby decreasing cardiac output, portal inflow, and perhaps blocking P-receptor-moderated vasodilatation, allowing unopposed a stimulation within the mesenteric arterioles. All studies in both adults and children emphasize the necessity of adequate dosing to achieve this 25% reduction in heart rate if the therapeutic success is to be achieved.

Most studies analyzing the effectiveness of P-blockers have been conducted in adults with intrinsic liver disease. Efficacy has been evaluated in two groups: (1) patients with documented varices who under- went P-blocker treatment in an attempt to prevent the first episode of bleeding (primary prophylaxis); and (2) patients treated following the initial hemorrhage in an attempt to prevent recurrent bleeding (secondary prevention). The efficacy of therapy differs with the indication.

In patients treated for primary prophylaxis, propranolol was effective in one study in preventing the first variceal hemorrhage in Child- Pugh class A, class B, and class C adult patients. The overall bleeding rate was 3.9% in treated patents and 21.6% in the control group.15 This improvement was more effective in Child-Pugh class B and class C patients, and less so in class A patients. A similar result was reported in the Italian multicenter trial using vitamin K treatment as the control medi~ation.~~ In addition, two meta-analyses have shown that proprano- lo1 is effective in preventing first bleeding in adults with cirrhosis.62 In these studies patients treated with P-blockers had significantly lower rates of bleeding and death from bleeding.", 16, 64 Similar success in primary prophylaxis of first variceal hemorrhage has also been reported using nadolol with or without isosorbide mononitrate; a significant decrease in the risk of first variceal hemorrhage was Combined treatment with nadolol and isosorbide mononitrate further decreased the risk of first variceal hemorrhage by more than one half in comparison with nadolol alone. Morbidity and mortality were similar in the two treatment groups.54 These and many other studies have established the use of P-blockade as an effective method of primary prophylaxis in adults with established cirrhosis.

The results of P-blockade for secondary prevention of recurrent variceal hemorrhage are more controversial. In 11 randomized, controlled trials including 755 adult patients, rebleeding was reduced in all trials, and the death rate was reduced.16 In low-risk patients (Child-Pugh class A), the risk of rebleeding has been reported to be as low as 3%.16,62


This advantage seems to be lost in class B and class C patients, for whom the reported risk for recurrent bleeding is 46% to 72%.

Fewer studies have been done in children. Because of the presence of extrahepatic portal hypertension in up to one half of all children who are candidates for treatment, pediatric results may not correlate directly with those of adult patients who universally have intrinsic liver disease. Several investigators, however, have confirmed the safety of propranolol administration for pediatric patients. As in adults, dosage in children must be adjusted to achieve a reduction of 25% in resting heart rate to achieve a therapeutic result. Shashidhar et a1 reported that 7 their study population of 21 patients experienced bleeding on propranolol treatment, 2 were noncompliant, and 4 were inadequately dosed.” Ozsoylu et a1 showed that propranolol was efficient for preventing the first variceal hemorrhage in Child-Pugh class A, class B, and class C patients, but only class A patients benefited when propranolol was administered to prevent recurrent bleeding6* On the basis of these results, P-blocker treatment is recommended for primary prophylaxis in all patient groups but may have limited benefit for secondary prevention in poor-risk patients.

MECHANICAL TAMPONADE

Mechanical tamponade using balloon catheter tubes (Sengstaken- Blakemore, or Minnesota tubes) provide mechanical compression of esophageal and gastric fundal varices. These devices must be carefully placed by an individual skilled in their use, often with fluoroscopic assistance. Monitoring is necessary to keep the esophageal balloon pressure below mean arterial blood pressure to avoid mucosal ischemia during long-term placement. Suction of secretions from the upper esophagus and pharynx is necessary to prevent aspiration. Endotracheal intu- bation is inevitably required in children. Although balloon tamponade is usually successful in stopping refractory hemorrhage, the effect is often transient, and recurrence following removal is common. This high rate of complications has limited their use to emergency control until other measures or surgical intervention can be instituted.

ENDOSCOPIC INTERVENTION FOR ESOPHAGEAL VARICES

In most cases, variceal hemorrhage can be controlled in children by fluid resuscitation, correction of coagulation, and pharmacologic support. The risk of recurrent hemorrhage and the need for accurate diagnosis of the site of hemorrhage often mandate endoscopy during the early posthemorrhage period. When variceal hemorrhage is confirmed or strongly suspected, variceal sclerotherapy or variceal band ligation can be used to eradicate the present or future sites of bleeding.


ENDOSCOPICSCLEROTHERAPYANDENDOSCOPIC BAND LIGATION

The use of endoscopic methods to control acute variceal hemorrhage is well established. Endoscopic sclerotherapy (EST) has been widely used as the primary treatment for refractory variceal bleeding because of its high success rate (> 9oy0) and the ability to institute initial treatment at the time of diagnostic endoscopy. The procedure is successful using both intravariceal and perivariceal injection techniques (Fig. 1). In children, 5% ethanolamine, 1% to 1.5% tetradecyl sulfate, or 5% sodium morrhuate have been used with equal success. Most patients require 3 to 6 sessions at intervals of 2 to 4 weeks to eradicate esophageal varices." Minor early complications, which occur in almost all patients, include retrosternal pain, fever, and transient dysphagia. Esophageal ulceration at the site of injection, a direct consequence of the procedure, is seen in 70% to 80% of patients and can be the source of recurrent bleeding. Serious complications such as esophageal strictures, esophageal perforation, or mediastinitis occur in 10% to 20% of patients. Despite these risks, emergency sclerotherapy has been successful in terminating acute

Figure 1. Injection sclerotherapy techniques-intravascular (solid area) and perivascular (dashed line) techniques. (From Terblanche J, Krige J: Endoscopic therapy in the management of esophageal varices: Injection sclerotherapy and variceal ligation. In Nyhus LM, Baker RJ, Fischer JE [eds]: Mastery of Surgery. Boston, Little, Brown and Company, 1997, pp 1329-1 349; with permission.)


variceal hemorrhage. In a meta-analysis reviewing primarily adult patients, EST was more successful than vasopressin in controlling initial bleeding and in reducing the incidence of early rebleeding; however, no significant improvement was seem when EST was compared with somatostatin or octreotide administration for pharmacologic control.16 The long-term outcome following EST is related to the primary liver disease. In children with EPVO, recurrent variceal bleeding developed in 31% of patients followed for a mean period of 8.7 years79 however, in children with intrahepatic disease, the rate was 75% (Table l).,,

In an attempt to overcome these complications but preserve the treatment success of EST, esophageal band ligation (EBL) of varices has been developed, using techniques similar to those for banding internal hemorrhoids (Fig. 2). Band ligation has the advantage of ligating only the submucosal venous varices, without harming the submucosal lining. This procedure should minimize the technical complications associated with injection of the submucosa and esophageal wall during sclerotherapy. Several bands are placed during each procedure, requiring multiple passes of the endoscope which can bear only a single band at a time. In limited experience, rates of control for acute variceal hemorrhage (90% to 96%) are equivalent to those of ~clerotherapy.~~ In three recent large, randomized studies in adult patients, EBL had lower complication rates, less recurrent bleeding, and equivalent but more rapid obliteration of ~arices.,~, 48, 76 Application of this experience has made EBL the procedure of choice in many centers for bleeding varices in ad~l t s .4~

Experience with EBL in children is less extensive but reports similar success (Table 2).", 56, 63, 67 Variceal obliteration was achieved in 73% to 100% of cases, rebleeding before completion of obliteration was less common than with EST; however, recurrence was seen in 75% of patients with intrahepatic disease. In addition, the small size of the child's esophagus limits the number of O-rings that can be placed at one treatment session, requiring multiple treatments. The thinner esophageal wall of small children makes full-thickness ligation a risk in children younger than 1 year and precludes EBL below this age.22

Both EST and EBL have been used for the primary prophylaxis of variceal hemorrhage. Prophylactic EST decreased the incidence of bleeding in children from 42% to 6% in the controlled, randomized trial by Goncalves et a1 who used EST every 21 days until initial obliteration was complete.25 Following successful EST, however, 16% developed congestive hypertensive gastropathy, and 38% bled from this lesion. In the control group only 6% had congestive hypertensive gastropathy, and there were no gastropathy-associated bleeding episodes. Prophylactic EST did not alter patient survival compared with the control group. Esophageal band ligation has also been used in a limited number of children. When undertaken for intrahepatic disease, 72% of varices were eradicated or improved, but two thirds required EST in addition to EBL to achieve control. Varices were not controlled or recurred in 2770, and the emergence of congestive hypertensive gastropathy was also noted.69 Long-term control with EST seems to exceed that with EBL, and a

aY 0

0

Tabl

e 1.

PE

DIA

TRIC

EX

PE

RIE

NC

E W

ITH

EN

DO

SC

OP

IC S

CLE

RO

THE

RA

PY

Mea

n N

o.

Y G

astr

ic

No.

in

Obl

itera

ted

Reb

leed

ing

Rec

urre

nce

Eso

phag

eal

Eso

phag

eal

Str

ictu

re

of T

reat

men

ts

Var

icie

s (G

V)

Aut

hor

Stu

dy

YO

%

YO

Ulc

erat

ion

-

-

Dila

war

i 38

92

23

16

23

10

Y

achh

an7

50

88

26

10

20

6 8

70%

GV

H

orig

omeM

7

100

14

29

0 0

2.3

Ma

kso

~d

~~

62

45

53

-

-

-

-

-

-

45

96

18

-

-

3 -

18%

GV

-

-

-

-

-

prop

hyla

ctic

49

6

-

108

92 E

PVO

39

12

29

16

5

4

75 in

trah

epat

ic

dise

ase

-

3.8

-

-

prop

hyla

ctic

17

65

St

ring

ern

36

80

31

28

-

3 -

13%

GV

T

hapa

n3

30

90

10

13

20

20

5 13

% G

V

Go

n~

lav

es~

~

50

94

24

-

-

-

-

12%

GV

~i

i1

33

33

10

0 36

33

18

12

4-

5 16

% C

HG

-

EPV

O =

ext

rahe

patic

por

tal

vein

obs

truc

tion;

CH

G =

chr

onic

hyp

erte

nsiv

e ga

stro

path

y.


A

Figure 2. Endoscopic band ligation technique. A, Varix is drawn into the band ligator using suction to facilitate band placement. 6, Band applied to varix to occlude proximal and distal flow. (From Terblanche J, Krige J: Endoscopic therapy in the management of esophageal varices: Injection sclerotherapy and variceal ligation. In Nyhus LM, Baker RJ, Fischer JE (eds): Mastery of Surgery. Boston, Little, Brown and Company, 1997, pp 1329-1349; with permission.)

combination of both may be more effective than EBL alone. The risk of accelerated formation of gastric varices may overshadow any prophylactic benefit for either EST or EBL; however, sufficient experience to assess this risk is not available at the present time.

PORTOSYSTEMIC SHUNTS

Numerous surgical procedures have been devised to divert portal blood into the low-pressure systemic venous circulation, thereby decreasing the portal venous pressure. Enthusiasm for the use of portosystemic shunting in children was limited by early reports suggesting that children less than 8 years of age and those with vessels for the shunt anastomosis less than 8 to 10 mm in diameter would be unsuitable candidates because of the high risk of shunt thrombo~is .~~ In addition, Voorhees et a1 suggested a high incidence of neuropsychiatric distur- bances following nonselective shunts in children.56, 84 Recent experience in centers skilled in pediatric vascular reconstruction has established that a high rate of success can be achieved with minimal complications even in small pediatric patients. The following principles should be followed:

1. Anastomosis should be constructed using fine (6.0-7.0) monofil- ament sutures with provision for growth.

00 0

N

Tabl

e 2.

PE

DIA

TRIC

EX

PE

RIE

NC

E W

ITH

VA

RIC

EA

L B

AN

D L

IGA

TIO

N

No.

in

Obl

itera

ted

Reb

leed

ing

Rec

urre

nce

Follo

w-U

p B

and

Mea

n N

o.

% G

astr

ic

%

%

Mea

n Fa

ilure

of

Tre

atm

ents

V

aric

ies

Aut

hor

Stu

dy

Ori

gin

%

-

22

36%

EPV

O

66

33

9 5.

3 ye

ars

0 4

64%

hep

atic

di

seas

e -

Nijh

awan

S6

15

100%

EPV

O

93

-

9 m

onth

s 0

1.8

7 Sa

sakP

9 11

10

0% b

iliar

y 18

-

-

1

-

-

-

atre

sia

54 i

mpr

oved

intr

ahep

atic

di

seas

e

FoxZ

Z 7

100%

10

0 33

75

13

.8 m

onth

s 0

4 33

% G

V

EPV

O =

ext

rahe

patic

vei

n ob

stru

ctio

n.


2. Sufficient mobilization of vessels is necessary to prevent kinking or twisting of the shunt after the viscera are returned to their normal location.

3. Venography should be performed after reconstruction to ensure division of all collaterals and adequate shunt flow.

4. Selective postoperative anticoagulation should be performed, and antiplatelet drugs should be administered.

5. The surgical team should be skilled in pediatric portal vascular reconstruction.

In general, portosystemic shunts can be classified into two groups- non-selective and selective shunts.

Nonselective shunts are constructed to communicate with the entire portal venous system, and therefore have the potential to divert blood from the normal antegrade perfusion to the liver. Historically, the most commonly used shunt in children was the mesocaval shunt (Clatworthy shunt), in which the distal inferior vena cava was ligated, divided, and its proximal portion then anastomosed to the side of the superior mesenteric vein. This procedure was often complicated by the development of transient lower-extremity edema but had the advantage of using a larger vein for the shunt anastomosis. A similar nonselective shunt, using a short segment of internal jugular vein to connect the superior mesenteric vein or splenic vein and the inferior vena cava, has now replaced the mesocaval shunt; this procedure retains the advantage of using a larger vessel for the anastomosis and avoids ligation of the inferior vena cava (Fig. 3)?3 An excellent patency rate (93%) and an absence of significant episodes of encephalopathy support its use in pediatric patients.23 The limited intra-abdominal dissection needed to complete this shunt contributes to its technical ease, and the shunt can be easily occluded if liver transplantation is needed. Other nonselective shunts (proximal splenorenal, end-to-side or side-to-side porto caval shunts) have significant disadvantages in children because they require splenectomy or dissection of the main portal vein which compromises liver transplantation.

Selective shunts are constructed to divert the gastvosplenic portion of the portal venous flow into a systemic vein, most frequently the left renal vein or the immediately adjacent inferior rena cava. Communica- tion between the central mesenteric portal circulation, which perfuses the liver, and the gastrosplenic portal circulation is severed by dividing the gastroepiploic veins, the coronary vein termination at the portal vein, and the retroperitoneal pancreatic collaterals. The most common selective shunt, the distal splenorenal shunt (DSRS, Warren shunt), pre- serves antegrade perfusion to the liver within the mesenteric portion of the portal circulation while decompressing the esophageal venous plexus through the short gastric veins and splenic vein (Fig. 4). This theoretical selective advantage can unfortunately be lost over time because the high- pressure mesenteric component progressively decompresses into the lower pressure gastrosplenic compartment. Complete isolation of the


m p . Splenic vein (

inferior vena cava.

Figure 3. Interposition “H” Graft Mesocaval Shunt A Nonselective shunt using either internal jugular vein harvested from the patient, or Polytetrafluoroethylene (PTFE) vascular graft material. This shunt allows the superior mesenteric vein to communicate with the infrarenal inferior vena cava. (From Zollinger RM: Atlas of Surgical Operations. New York, Macmillan Publishing, 1975; with permission.)

pancreas from the splenic vein will decrease the frequency and rapidity of this Despite this potential for progressive loss of selectivity, the authors use this shunt as the primary option in children. When the adrenal vein is appropriately located and dilated, it serves as an alternative anastomotic site to access the left renal vein.45, 52

When performed in centers experienced in complex vascular reconstruction of the portal system, as is necessary in pediatric liver transplantation, shunt patency has ranged from 83% to 100% (Table 3).52, 56, 59

When shunt patency is maintained, recurrence of variceal bleeding is uncommon, although decompressed varices may still be identifiable on endoscopy. Encephalopathy is uncommon in children following successful portosystemic shunting, even in patients with intrinsic liver disease. In these children this encephalopathy is controlled with lactulose and dietary inter~ention.~~, 66, 74

Methods of direct reconstruction of portal circulation in patients with EPVO into the left branch of the portal vein have been described and represent ideal solutions.8* l9 This mesenterico-portal shunt (Rex shunt) reestablishes normal portal inflow into the intrahepatic portal


Left gastroepiploic

mssenterlc vein

vena cava

Figure 4. Distal splenorenal shunt. A selective shunt allowing communication between the splenic vein and the left renal vein. The esophagogastric venous complex communicates by way of the short gastric veins, decompressing esophageal varices without decreasing perfusion through the mesenteric portal system to the liver. (From Zollinger RM: Atlas of Surgical Operations. New York, Macmillan Publishing, 1975; with permission.)

vein, using either an interposed jugular venous graft or the dilated coronary vein (Fig. 5). Candidates for this procedure must fulfill three conditions: the liver parenchyma must be normal, they must not have a hypercoagulable state, and the umbilical portion of the left hepatic vein must be accessible and patent. The procedure is unique in that it restores hepatopetal portal perfusion and the inflow of hepatatrophic substances to the liver. Patients with diffuse portal vein thrombosis are not candidates for this reconstruction. Doppler ultrasound studies and direct portography suggest, however, that approximately two thirds of children with EPVO have sufficient left portal vein patency to undergo this procedure. In the reported experience (16 cases), all shunts have maintained initial patency, although 3 required revision because of stenosis and 1 shunt occluded at 7 months.8, l9 Hypertensive gastropathy re- solved, and variceal bleeding did not recur. In these selected patients,

Tabl

e 3.

PE

DIA

TRIC

EX

PE

RIE

NC

E W

ITH

PO

RTO

SY

STE

MIC

SH

UN

T P

RO

CE

DU

RE

S

Type

of

Aut

hor

No.

in

Stu

dy

Shu

nt

% P

aten

cy

Yo E

ncep

halo

path

y S

urv

ival

YO

Cas

es 1

-7 =

71

-

100

peri

oper

ativ

e M

akso

ud51

42

D

SRS

Bem

ard'O

92

EPV

O

-~

Cas

es -2

=

92

MC-

ITV

93

0

100

17 C

HF

94

0 10

0 37

cir

rhos

is

84

29

81

12 B

udd-

Chi

ari s

yndr

ome

83

0 75

O

riof

f 59

162

EPV

O

SS-S

R-7

5 10

0 0

99%

-5 y

ears

PS

R-3

4 88

0

96Y

0-10

yea

rs

MC

-53

100

0 G

auth

ierZ

3 84

SR

-20

93

0 10

0% p

erio

pera

tive

(59

EPV

O, 2

3 in

trah

epat

ic

MC

-55

0 di

seas

e. 4 B

udd-

Chi

ari

PC

4

0 sy

ndro

me)

M

akes

hift

-7

0

(unp

ublis

hed

data

) (2

0 EP

VO

, 18

intr

ahep

atic

C

SR

S6

100

0 R

yckm

an a

nd A

lons

o 40

D

SRS2

5 96

0

100%

per

iope

rativ

e, 9

2% o

vera

ll

dise

ase,

2 B

uddi

-Chi

ari

MC

-9

89

11

100%

per

iope

rativ

e, 8

9% o

vera

ll sy

ndro

me)

83%

ove

rall

EPO

= e

xtra

hepa

tic v

ein

obst

ruct

ion;

DSR

S =

dis

tal s

plen

oren

al s

hunt

; MC

-IJV

= m

esoc

aval

-int

erna

l jug

ular

vei

n in

terp

ositi

on; S

S-SR

= s

ide-

to-s

ide

sple

nore

nal s

hunt

; C

HT

= c

hron

ic h

epat

ic f

ibro

sis;

PSR

= p

roxi

mal

spl

enor

enal

shu

nt; C

SRS

= c

entr

al s

plen

oren

al s

hunt

.


Figure 5. Computed Tomographic Scan showing the development of peripancreatic collateral venous channels connecting the mesenteric portal venous system to the decompressed esophagogastric system in a patient with primary extrahepatic portal venous obstruction 8 years post-operation following a distal splenorenal shunt without pancreatic isolation. (See also color plate 1, Fig. 7.)

this option should be considered despite the more complex technical challenge it presents.

The indications for portosystemic shunting have been altered by the growing success of endoscopic methods to control variceal bleeding and the improvements in pediatric liver transplantation. The authors now consider the following children to be candidates for portosystemic shunting

1. Children with documented variceal hemorrhage, who have progressive or continued esophageal variceal bleeding despite endoscopic intervention, and who have preserved hepatic synthetic function.

2. Children who fail endoscopic treatment and have intrinsic liver disease but have adequate liver synthetic function to predict that liver transplantation will not be needed for several years (selective shunt only).

3. Children with severe portal hypertension accompanying cystic fibrosis and variceal hemorrhage whose microbiologic flora com- promise liver transplant ~urvival.’~

4. Children with severe portal hypertension who reside a great distance from emergency medical care so that their survival would be endangered should significant hemorrhage occur.

.

5. Children with EPVO and uncontrolled hypersplenism.


TRANSJUGULAR INTRAHEPATIC PORTOSYSTEMIC SHUNT

The introduction of transjugular intrahepatic portosystemic shunt (TIPS) has added another therapeutic option for the physician confronted with complex portal hypertension. This procedure uses interventional radiographic techniques to place an intrahepatic expandable metallic shunt between a portal vein branch and the hepatic vein, forming a central nonselective portocaval shunt. The procedure is undertaken using access through the right internal jugular vein. The hepatic veins are identified and a puncture from the hepatic vein into an intrahepatic portion of the portal vein is undertaken with fluroscopic or ultrasono- graphic guidance. This tract is dilated, and an expandable mesh stent is placed, forming a communication between the intrahepatic portal vein and the hepatic vein branch (Fig. 6). Technical difficulties in establishing a safe but adequately large tract for sufficient shunt flow limit the usefulness of this procedure in infants. In children with biliary atresia, the close proximity of the biliary Roux-en-Y conduit to the portal vein and the often diminutive size of the portal vein increase the risk of stent perforation, malposition, or perforat i~n.~~ This procedure has great benefit in the control of refractory portal hypertensive bleeding unresponsive to the pharmacologic treatment and in patients needing tempo- rary portal decompression before liver transplantation.21, 31 The ability to embolize bleeding varices from the coronary vein at the time of TIPS placement assists in achieving primary control of bleeding

The two principal long-term complications of TIPS are encephalopathy and shunt occlusion. Because it is a central, nonselective shunt, this procedure can precipitate hepatic encephalopathy, especially when used in patients with severe intrinsic liver disease. The overall risk of encephalopathy ranges from 5% to 35% in adult patients, a rate similar to that seen with side-to-side surgical shunts. Most episodes of encephalopathy can be controlled with dietary protein restriction and lactulose administration. Selection of a shunt size that allows sufficient portal decompression without shunting excessive amounts of blood from the liver is also a theoretical solution. Stenosis of the shunt or shunt thrombosis remains a major complication following TIPS. Shunt stenosis occurs in 25% to 75% of cases, with shunt patency decreasing with the length of time that the shunt is in place. Intimal hyperplasia or incorrect shunt placement most commonly causes the stenosis. Regular monitoring for shunt patency and periodic shunt dilation or restenting is ne~essary.~~

Pediatric experience with TIPS is still limited, primarily because of a lack of appropriate candidates for the procedure. Most children with biliary atresia and an ineffective portoenterostomy procedure develop end-stage cirrhosis within their first 2 years of life. These patients are poor candidates because of their size and the frequency of portal vein abnormalities within this population. Most experience is in children over 5 years of age (Table 4). Success rates seem to approximate the adult experience, with 75% to 90% initial success in TIPS placement. The


I A I B

SMV

- . S.V.

SMV

Figure 6. Mesentericoportal shunt to restore portal blood flow to the left branch of the intrahepatic portal vein. This shunt returns portal blood flow directly to the hepatic portal circulation, and is useful only in patients with extrahepatic portal vein obstruction and a patent intrahepatic portal venous system. An interposition jugular venous graft (stippled area) is often required. (From de Ville DG, Alberti D, Clapuyt P, et al: Direct bypassing of extrahepatic portal venous obstruction in children: A new technique for combined hepatic portal revascularization and treatment of extrahepatic portal hypertension. J Pediatri Surg 33597 to 601, 1998; with permission.)

smaller size of the liver and its venous structures requires special skill and equipment. Shorter stent lengths and smaller-diameter stents have been constructed for pediatric applications; however, the risks of hepatic perforation and stent malposition are greater in small patients. Postpro- cedural encephalopathy seems to be less common in children, although limited clinical experience and the difficulties in diagnosing subtle encephalopathy in children makes this observation tentative. The complications of shunt stenosis are equally problematic, and patient growth over time may cause the initial shunt to be too short, requiring revision or restenting to maintain access to both the portal and hepatic venous circulation. These limitations and risks make TIPS a reasonable and suitable treatment for acute, unresponsive variceal hemorrhage in children with established intrinsic liver disease, often while awaiting liver transplantation. The procedure is particularly helpful when used as a

Tabl

e 4.

PE

DIA

TRIC

EX

PE

RIE

NC

E W

ITH

TR

AN

SJU

GU

LAR

IN

TRA

HE

PA

TIC

PO

RTO

SY

STE

MIC

SH

UN

T

FoIIo

w-U

P

Com

plic

atio

ns

Aut

hor

Age

of

Pat

ient

D

iagn

osis

Wa

ne

Cao

l3

Schw

eize

r70

Ast

falk

4 L

agie

P6

Ber

gerg

Wei

nber

gs6

Azo

ulay

5 Fl

eet2

]

Hac

kwor

th%

13 y

ears

21

mon

ths

4 ye

ars

2.5

year

s 10

mon

ths

7 pa

tient

s: 3

-13

year

s 4.

5 ye

ars

7 ye

ars

14 y

ears

10

, 11

year

s 6

year

s 7

year

s 11

yea

rs

2.5,

4 y

ears

10 p

atie

nts

14 y

ears

10

yea

rs

9 ye

ars

15 m

onth

s 4

patie

nts:

5,

5, 7

, 12

year

s

5 pa

tient

s: 7

, 9,

10,

15,

15

year

s

4 pa

tient

s: 2

, 3,

7,

9 ye

ars

8 pa

tient

s: 5

, 7,

9, 9

, 10,

15,

16

year

s

cyst

ic fi

bros

is

bilia

ry a

tres

ia

bilia

ry a

tres

ia

bilia

ry a

tres

ia

bilia

ry a

tres

ia

pseu

do-o

bstr

uctio

n cy

stic

fibr

osis

bi

liary

atr

esia

ch

oles

tasi

s co

ngen

ital h

epat

ic f

ibro

sis

bilia

ry a

tres

ia

shor

t-gu

t syn

drom

e

bone

mar

row

tra

nspl

ant

cyst

ic f

ibro

sis

cyst

ic f

ibro

sis

auto

imm

une

hepa

titis

cy

stic

fibr

osis

bi

liary

atr

esia

cirr

hosi

s

bilia

ry a

tres

ia

cirr

hosi

s (7

), as

cite

s (1

)

6 m

onth

s

-

4-24

m

onth

s 5

mon

ths

4 m

onth

s 11

mon

ths

4,9

mon

ths

1,lO

mon

ths

6 m

onth

s 30

mon

ths

2 ye

ars

22 d

ays

5, 6

, 66

, 0

days

0, 24

,40,

35,

800

days

9-12

7 da

ys to

live

r tr

ansp

lant

atio

n 30

1, 3

57 d

ays

(no

liver

tr

ansp

lant

atio

n)

ence

phal

opat

hy

liver

tran

spla

ntat

ion

1 li

ver

caps

ule

punc

ture

sh

unt o

cclu

sion

(an

giop

last

y)

1 o

cclu

ded

1 sm

all b

owel

tran

spla

ntat

ion

at

1 m

onth

1 su

rviv

or

shun

t ste

nosi

s-re

vise

d di

ed-p

ulm

onar

y di

seas

e

died

-sep

sis

3 liv

er tr

ansp

lant

atio

ns,

12 y

/o

3 oc

clud

ed, 2

revi

sed,

1 D

SRS,

1

occl

uded

-DSR

S

hem

orrh

age,

1 li

ver

tran

spla

ntat

ion

repl

aced

10

live

r tr

ansp

lant

atio

ns,

1

DSR

S =

dis

tal s

plen

oren

al s

hunt


bridge to achieve stability by controlling refractory hemorrhage in patients awaiting liver tran~plantation.~~ Long-term decompression is better achieved through surgical shunts at the present time, and TIPS is not indicated in patients with extrahepatic portal vein occlusion.

NONSHUNT PROCEDURES FOR PORTAL HYPERTENSION

The use of nonshunt surgical procedures for the management of portal hypertension are not as successful as shunt therapy. These operations have included direct variceal ligation through a transthoracic or abdominal approach, gastroesophageal devascularization procedures (Sugiura procedure), or, rarely, translocation of the spleen into the tho- ax.^^, 58, 65 In general, these procedures have been abandoned except where widespread thrombosis of the mesenteric venous vasculature makes shunt therapy or transplantation poor alternatives.

The Sugiura procedure has been used with the most success in children (Fig. 7). This procedure includes devascularization of the upper two thirds of the greater and lesser curvatures of the stomach and ligation of the left gastroepiploic, short gastric, and left gastric vessels. Ligation of all retrogastric collaterals, transhiatal devascularization of the lower esophagus, and esophageal transection with fundoplication and pyloroplasty if the vagus nerves are damaged complete the operative procedure.58 Splenectomy was advocated in the original descriptions of this procedure as well (Fig. 8). Splenectomy, however, has been associated with a greater risk of intraoperative bleeding, need for intra-

Hepaticvein I I

Figure 7. Transjugular intrahepatic portosystemic shunt (TIPS) procedure. (From Schneider BL, Groszmann RJ: Portal hypertension. In Suchy FJ [ed]: Liver Disease in Children. St. Louis, Mosby, 1994, pp 249-266; with permission.)


I I\Fhyghly Selective 19 (agotomy

ilication

Figure 8. Sugiura procedure for gastroesophageal devascularization, modified for pediatric patients to include division of the short gastric veins without splenectomy. (From Superina RA, Weber JL, Shandling B: A modified Sugiura operation for bleeding varices in children. J Pediatr Surg 18:794-799, 1983; with permission.)

operative blood transfusions, and postoperative portal vein thrombosis. Because of these risks and the known increased potential for postoperative infectious complications of splenectomy, the authors do not rou- tinely perform splenectomy in children. Specific indications for splenectomy during nonshunt operations are severe hypersplenism, massive splenomegaly, and splenic vein thrombo~is.~~, 6o Following extensive devascularization to achieve portal-azygos disconnection, rebleeding rates of 5% to 10% can be achieved during long-term follow-up.53, 6o The survival rates following these operations have been reported to be 88% at 5 years and 80% at 10 year^.^^,^* This procedure offers a safe alternative for variceal control in patients with anatomy unsuitable for shunting or when the expertise for emergency portocaval shunt or liver transplantation is not available.

Portopulmonary shunting, induced by splenopneumopexy, is in- tended to produce collateral circulation between the portal system and the pulmonary veins. This communication is created by amputating the superior pole of the spleen, transposing it through an opening in the left hemidiaphragm, and exposing the raw splenic surface to the left lower lung. Splenic artery occlusion by angiographic or direct ligation is also undertaken. A parenchymatous anastomosis is induced between the splenic pulp veins and the pulmonary venous structures. This uncom-


mon procedure has been successfully used in children with EPVO, hepatic fibrosis, intrinsic cirrhosis, and Budd-Chiari syndrome.', 65, A decrease in splenic pulp pressure of 25% has been reported in most cases in which the decrease was recorded. Indications for this procedure are limited to the treatment of children with widespread occlusion of the portal vein and its radicals.',

LIVER TRANSPLANTATION

The progressive improvement in operative techniques and in the immunosuppression management of children who have undergone liver transplantation have led to 1-year survival rates approaching 90% in many centers, with 5-year survival rates of 85?!0.~, 6, 7, 12, 26, 29, 36, 8o Regard- less of the primary cause of portal hypertension, liver transplantation successfully reverses the portal flow obstruction and allows the resolution of hypersplenism and hypertensive portal gastropathy. The introduction of innovative surgical procedures to allow transplantation of liver segments and reduced-size grafts has improved donor availability and donor access for children of all ages. The use of primary transplantation as a treatment modality for portal hypertension is limited, however, by the availability of suitable donor organs and the long-term risks of immunosuppression, opportunistic infections, and lymphoproliferative disease. When children have progressive intrinsic hepatic disease, the course of their progression and the amount of hepatic functional reserve should determine the use of primary transplantation or temporizing treatments such as sclerotherapy or surgical shunts. At present, primary transplantation is recommended for children who have significant portal hypertensive complications such as bleeding, hypersplenism, or hepatopulmonary syndrome, and for those who have progressive hepatic synthetic failure. Children with intrinsic liver disease but preserved hepatic synthetic function, who may not require transplantation for several years will achieve excellent palliation with selective DSRS. Transjugular intrahepatic portosystemic shunt is reserved for patients who have unresponsive variceal bleeding as a therapeutic bridge to transplantation, allowing them to achieve suitable stability while awaiting transplant donor organ availability.

PULMONARY SYNDROMES ASSOCIATED WITH PORTAL HYPERTENSION

Hepatopulmonary syndrome (HPS), a syndrome associated with hepatic dysfunction, hypoxemia, and pulmonary vascular dilatations, characteristically occurs in children with long-standing liver disease and portal hypertension. This constellation of symptoms is thought to be caused by the effects of vasoactive substances, normally inactivated in the liver, on the pulmonary vasculature. Although this syndrome is


usually seen in patients with portal hypertension and spontaneous portosystemic communications, it has also been reported following surgical portosystemic shuntingz9 Progressive hypoxemia, which is exacerbated by the horizontal position (orthodeoxia and platypnea), is commonly seen on evaluation. Hypoxemia does not correct with 100% oxygen ventilation, confirming the presence of intrapulmonary shunts contribut- ing to the alveolar perfusion abnormalities. Abnormal extrapulmonary (brain) uptake of Technetium-99m macroaggregated albumin (MAA) after lung perfusion scanning or the presence of early microbubble perfusion on echocardiography are diagnostic and predictive of morbidity with general anesthesia or liver transplantation." Present experience suggests that these abnormalities are reversible following successful liver transplantati~n.~~, 43

Pulmonary hypertension has been associated with severe liver disease and portal hypertension in 1% to 2% of patients. In the presence of spontaneous or surgically created portosystemic communication, pulmonary vasoconstrictive substances can pass directly into the systemic circulation, bypassing hepatic metabolism. Possible agents include hista- mine, serotonin, purine, pyrimidine, neuropeptide Y, and thrombox- a n e ~ . ~ ~ Symptoms of dyspnea, exercise intolerance, and palpitation occur, however, many patients remain asymptomatic until late in the course of their disease. Initial treatment with oxygen and vasodilators is occasion- ally helpful. The pathologic lesion, plexogenic pulmonary arteriopathy, results in increased smooth muscle, increased thickness of the media, concentric luminal fibrosis, and eventually fibrinoid necrosis. Case reports suggest that liver transplantation may assist in reversal, provided diagnosis and transplantation are undertaken at the early stage of pathologic ev0lution.4~ In later stages, resolution cannot be anticipated.

SUMMARY

Therapeutic options for children with portal hypertension now include a broad range of pharmacologic, endoscopic, and surgical procedures. Thoughtful application of all of these options can improve quality of life by decreasing the complications of portal hypertension and can decrease mortality by preventing the consequences of variceal hemorrhage. The development of portal hypertensive gastropathy following palliative procedures such as endoscopic sclerotherapy and band ligation may limit their long-term success in children. The excellent results now obtained with selective portosystemic shunts and liver transplantation assure that definitive surgical treatments will continue to be important components in the treatment of children with portal hypertensive complications or progressive liver disease. Evolving procedures, such as TIPS, represent excellent short-term life-preserving techniques to stabi- lize critically ill patients while awaiting liver transplantation. Their role in the future, long-term management of children is yet to be defined.


References

1.

2.

3.

4.

5.

6.

7.

8.

9.

10.

11.

12. 13.

14.

15.

16.

17.

18.

Akita H, Sakoda K Portopulmonary shunt by splenopneumopexy as a surgical treatment of Budd-Chiari syndrome. Surgery 8785-94, 1980 Alonso MH, Ryckman FC: Current concepts in pediatric liver transplant. Semin Liver Dis 18:295-307, 1998 Ando H, Kaneko K, Ito F, et al: Anatomy and etiology of extrahepatic portal vein obstruction in children leading to bleeding esophageal varices. J Am Coll Surg 183:543-547,1996 Astfalk W, Huppert PE, Schweizer P, et al: Recurrent intestinal bleeding from jejunoje- junostomy caused by portal hypertension following hepatoportojejunostomy in extra hepatic biliary atresia (EHBA)-successful treatment by transjugular intrahepatic portosystemic shunt (TIPS). Eur J Pediatr Surg 7147-148,1997 Azoulay D, Castaing D, Lemoine A, et al: Transjugular intrahepatic portosystemic shunt (TIPS) for severe veno-occlusive disease of the liver following bone marrow transplantation. Bone Marrow Transplant 25:987-992, 2000 Balistreri W F Transplantation for childhood liver disease: An overview. Liver Transpl Surg 4:S18-S23, 1998 Balistreri WF, Grand R, Hoofnagle JH, et al: Biliary atresia: Current concepts and research directions. Summary of a symposium. Hepatology 23:1682-1692, 1996 Bambini DA, Superina R, Almond PS, et a1 Experience with the Rex shunt (mesenterico-left portal bypass) in children with extrahepatic portal hypertension. J Pediatr Surg 35:13-18,2000 Berger KJ, Schreiber RA, Tchervenkov J, et al: Decompression of portal hypertension in a child with cystic fibrosis after transjugular intrahepatic portosystemic shunt placement. J Pediatr Gastroenterol Nutr 19:322-325, 1994 Bernard 0, Alvarez F, Brunelle F, et a1 Portal hypertension in children. Clinical Gastroenterology 1433-55, 1985 Bernard B, Lebrec D, Mathurin P, et a1 Beta-adrenergic antagonists in the prevention of gastrointestinal rebleeding in patients with cirrhosis: A meta-analysis. Hepatology 25:63-70, 1997 Busuttil RW, Goss JA: Split liver transplantation. Ann Surg 229:313-321, 1999 Cao S, Monge H, Semba C, et a1 Emergency transjugular intrahepatic portosystemic shunt (TIPS) in an infant: A case report. J Pediatr Surg 32125-127, 1997 Carson JA, Tunell WP, Barnes P, et al: Hepatoportal sclerosis in childhood: A mimic of extrahepatic portal vein obstruction. J Pediatr Surg 16291-296,1981 COM HO, Grace ND, Bosch J, et a1 Propranolol in the prevention of the first hemorrhage from esophagogastric varices: A multicenter, randomized clinical trial. . - - The Bostgn-New Haven-Barcelona Portal Hypertension .Study Group. Hepatology 13~902-912.1991 DAmico G, Pagliaro L, Bosch J: The treatment of portal hypertension: A meta- analytic review. Hepatology 22:332-354, 1995 Debray D, Lykavieris P, Gauthier F, et al: Outcome of cystic fibrosis-associated liver cirrhosis: Management of portal hypertension. J Hepatol 31:77-83, 1999 de Ville dG, Alberti D, Clapuyt P, et al: Direct bypassing of extrahepatic portal venous obstruction in children: A new technique for combined hepatic portal revascularization and treatment of extrahepatic portal hwertension. T Pediatr Sure 33:597-601.1998

19. de Ville dG, Alberti D, Falchitti D,'et a1 Tr;?atment of extrahepatic pgrtal hypertension in children by mesenteric-to-left portal vein bypass: A new physiological procedure. Eur J Surg 165:777-781, 1999

19a. Dilawari JB, Chawla YK, Ramish GN, et al: Endoscopic sclerotherapy in children. Journal of Gastroenterology and Hepatology 4155-160, 1989

20. Dubuisson C, Boyer-Neumann C, Wolf M, et al: C, protein S and antithrombin I11 in children with portal vein obstruction. J Hepatol27132-135,1997

21. Fleet M, Stanley AJ, Forrest EH, et al: Transjugular intrahepatic portosystemic stent shunt placement in a patient with cystic fibrosis complicated by portal hypertension. Clin Radio1 55:236-237, 2000


22. Fox VL, Carr-Locke DL, Connors PJ, et al: Endoscopic ligation of esophageal varices in children. J Pediatr Gastroenterol Nutr 20:202-208, 1995

23. Gauthier F, de Dreuzy 0, Valayer J, et al: H-type shunt with an autologous venous graft for treatment of portal hypertension in children. J Pediatr Surg 241041-1043, 1989

24. Gimson AE, Ramage JK, Panos MZ, et al: Randomized trial of variceal banding ligation versus injection sclerotherapy for bleeding esophageal varices. Lancet

25. Goncalves ME, Cardoso SR, Maksoud JG: Prophylactic sclerotherapy in children with esophageal varices: Long-term results of a controlled prospective randomized trial. J Pediatr Surg 35:401405, 2000

26. Goretsky MJ, Greenhalgh DG, Warden GD, et al: The use of extracorporeal life support in pediatric burn patients with respiratory failure. J Pediatr Surg 30:620- 623, 1995

27. Goss JA, Shackleton CR, McDiarmid SV, et al: Long-term results of pediatric liver transplantation: An analysis of 569 transplants. Ann Surg 228411420, 1998

28. Hackworth CA, Leef JA, Rosenblum JD, et al: Transjugular intrahepatic portosystemic shunt creation in children: Initial clinical experience. Radiology 206109-114, 1998

29. Hannam PD, Sandokji AK, Machan LS, et a1 Post-surgical shunt hepatopulmonary syndrome in a case of non-cirrhotic portal hypertension: Lack of efficacy of shunt reversal. Eur J Gastroenterol Hepatol 11:1425-1427, 1999

30. Hasegawa T, Tamada H, Fukui Y, et al: Distal splenorenal shunt with splenopancreatic disconnection for portal hypertension in biliary atresia. Pediatr Surg Int 1592-96,1999

31. Heyman MB, Laberge JM: Role of transjugular intrahepatic portosystemic shunt in the treatment of portal hypertension in pediatric patients. J Pediatr Gastroenterol Nutr 29:240-249, 1999

32. Heyman MB, Laberge JM, Somberg KA, et al: Transjugular intrahepatic portosystemic shunts (TIPS) in children. J Pediatr 131:914-919, 1997

33. Hill ID, Bowie M D Endoscopic sclerotherapy for control of bleeding varices in children. Am J Gastroenterol 86472476, 1991

34. Horigome H, Nomura T, Saso K, et al: Endoscopic injection sclerotherapy for esophagogastric variceal bleeding in children with biliary atresia. Hepatogastroenterology 46:3060-3062, 1999

35. Howard ER, Stringer MD, Mowat AP: Assessment of injection sclerotherapy in the management of 152 children with oesophageal varices. Br J Surg 75:404-408, 1988

36. Inomata Y, Tanaka K, Uemoto S, et al: Living donor liver transplantation: An 8-year experience with 379 consecutive cases. Transplant Proc 31:381, 1999

37. The Italian Multicenter Project for Propranolol in Prevention of Bleeding: Propranolol for prophylaxis of bleeding in cirrhotic patients with large varices: A multicenter, randomized clinical trial. Hepatology 8:l-5, 1988

38. Johnson SP, Leyendecker JR, Joseph FB, et al: Transjugular portosystemic shunts in pediatric patients awaiting liver transplantation. Transplantation 62:1178-1181, 1996

39. Karrer FM, Narkewicz MR Esophageal varices: Current management in children. Semin Pediatr Surg 8:193-201, 1999

40. Kato T, Romero R, Koutouby R, et al: Portosystemic shunting in children during the era of endoscopic therapy: Improved postoperative growth parameters. J Pediatr Gastroenterol Nutr 30:419425, 2000

41. Kerns SR, Hawkins IF Jr: Transjugular intrahepatic portosystemic shunt in a child with cystic fibrosis. AJR Am J Roentgen01 159:1277-1278, 1992

42. Kreft B, Strunk H, Flacke S, et al: Detection of thrombosis in the portal venous system: Comparison of contrast-enhanced MR angiography with intraarterial digital subtraction angiography. Radiology 216536-92, 2000

43. Krowka MJ: Hepatopulmonary syndrome: Recent literature (1997 to 1999) and impli- cations for liver transplantation. Liver Transplantation 6:S31-S35, 2000

44. Krowka MJ, Wiseman GA, Burnett OL, et a1 Hepatopulmonary syndrome: A prospective study of relationships between severity of liver disease, PaO, response to 100% oxygen, and brain uptake after 99mTc MAA lung scanning. Chest 118:615-624, 2000

342~391-394, 1993


45. Kulkarni VM, Nagral SS, Mathur SK Use of adrenal vein conduit for splenorenal

46. Lagier E, Rousseau H, Maquin P, et al: Treatment of bleeding stoma1 varices using shunts: A case report. Hepatogastroenterology 46:2033-2034, 1999

transjugular intrahepatic portosystemic shunt. J Pediatr Gastioenterol Nutr 18:501- 503,1994

47. Laine L, Cook D Endoscopic ligation compared with sclerotherapy for treatment of esophageal variceal bleeding. A meta-analysis. Ann Intern Med 123:280-287, 1995

48. Laine L, el Newihi HM, Migikovsky B, et al: Endoscopic ligation compared with sclerotherapy for the treatment of bleeding esophageal varices. Ann Intern Med

49. Losay J, Piot D, Bougaran J, et al: Early liver transplantation is crucial in children with liver disease and pulmonary artery hypertension. J Hepatol28:337-342, 1998

50. Lykavieris P, Gauthier F, Hadchouel P, et a1 Risk of gastrointestinal bleeding during adolescence and early adulthood in children with portal vein obstruction. J Pediatr 136805-808,2000

51. Maksoud JG, Goncalves ME: Treatment of portal hypertension in children. World J Surg 18:251-258, 1994

52. Mazariegos GV, Reyes J: A technique for distal splenoadrenal shunting in pediatric portal hypertension. J Am Coll Surg 187634-636, 1998

53. Mathur SK, Shah SR, Nagral SS, et al: Transabdominal extensive esophagogastric devascularization with gastroesophageal stapling for management of noncirrhotic portal hypertension: Long-term results. World J Surg 23:116&1174, 1999

54. Merkel C, Marin R, Sacerdoti D, et al: Long-term results of a clinical trial of nadolol with or without isosorbide mononitrate for primary prophylaxis of variceal bleeding in cirrhosis. Hepatology 31:324-329, 2000

55. Miura H, Kondo S, Shimada T, et a1 Long-term effects of distal splenorenal shunt with splenopancreatic and gastric disconnection on hypersplenism due to liver cirrhosis. Hepatogastroenterology 462995-2998, 1999

56. Nijhawan S, Patni T, Sharma U, et al: Endoscopic variceal ligation in children. J Pediatr Surg 30:145.51456,1995

57. Odievre M, Pige G, Alagille D Congenital abnormalities associated with extrahepatic portal hypertension. Arch Dis Child 52383-385, 1977

58. Ohashi K, Kojima K, Fukazawa M, et al: Long-term prognosis of non-shunt operation for idiopathic portal hypertension. J Gastroenterol 33:241-246, 1998

.59. Orloff MJ, Orloff MS, Rambotti M Treatment of bleeding esophagogastric varices due to extrahepatic portal hypertension: Results of portal-systemic shunts during 35 years. J Pediatr Surg 29:142-151, 1994

60. Orozco H, Mercado MA, Martinez R, et al: Is splenectomy necessary in devascularization procedures for treatment of bleeding portal hypertension? Arch Surg 133:36-38, 1998

61. Otte JB, de Ville D, Reding R, et al: Pediatric liver transplantation: From the full-size liver graft to reduced, split, and living related liver transplantation. Pediatr Surg Int 13308-318, 1998

62. Ozsoylu S, Kocak N, Demir H, et al: Propranolol for primary and secondary prophylaxis of variceal bleeding in children with cirrhosis. Turk J Pediatr 4231-33, 2000

63. Price MR, Sartorelli KH, Karrer FM, et al: Management of esophageal varices in children by endoscopic variceal ligation. J Pediatr Surg 31:1056-1059, 1996

64. Poynard T, Cales P, Pasta L, et al: Beta-adrenergic-antagonist drugs in the prevention of gastrointestinal bleeding in patients with cirrhosis and esophageal varices. An analysis of data and prognostic factors in 589 patients from four randomized clinical trials. Franco-Italian Multicenter Study Group. N Engl J Med 3241532-1538, 1991

65. Reese JC, Fairchild RB, Brems JJ, et al: Splenopneumopexy to treat portal hypertension produced by venous occlusive disease. Arch Surg 1271129-1132, 1992

66. Reyes J, Mazariegos GV, Bueno J, et al: The role of portosystemic shunting in children in the transplant era. J Pediatr Surg 34117-122,1999

67. Ryckman FC: New issues concerning the etiology of an old problem. J Am Coll Surg 183:637-639. 1996

119~1-7, 1993


68. Sakoda K, Ono J, Kawada T, et al: Portopulmonary shunt by splenopneumopexy for portal hypertension in children. J Pediatr Surg 23:323-327, 1988

69. Sasaki T, Hasegawa T, Nakajima K, et al: Endoscopic variceal ligation in the management of gastroesophageal varices in postoperative biliary atresia. J Pediatr Surg

70. Schweizer P, Brambs HJ, Schweizer M, et al: TIPS: A new therapy for esophageal variceal bleeding caused by EHBA. Eur J Pediatr Surg 5211-215,1995

71. Sergent G, Gottrand F, Delemazure 0, et al: Transjugular intrahepatic portosystemic shunt in an infant. Pediatr Radiol 27588-590, 1997

72. Shashidhar H, Langhans N, Grand RJ: Propranolol in prevention of portal hypertensive hemorrhage in children: A pilot study. J Pediatr Gastroenterol Nutr 29:12-17,1999

73. Shneider BL, Groszmann RJ: Portal hypertension. In Suchy FJ (ed): Liver Disease in Children. St. Louis, Mosby 1994, pp 249-266

74. Shun A, Delaney DP, Martin HC, et al: Portosystemic shunting for pediatric portal hypertension. J Pediatr Surg 32489-493, 1997

75. Soh H, Hasegawa T, Sasaki T, et al: Pulmonary hypertension associated with postoperative biliary atresia: Report of two cases. J Pediatr Surg 341779-1781, 1999

76. Stiegmann GV, Goff JS, Michaletz-Onody PA, et al: Endoscopic sclerotherapy as compared with endoscopic ligation for bleeding esophageal varices. N Engl J Med

77. Sperl W, Stuppner H, Gassner I, et al: Reversible hepatic veno-occlusive disease in an infant after consumption of pyrrolizidine-containing herbal tea. Eur J Pediatr

78. Steventon DM, Kelly DA, McKiernan P, et a1 Emergency transjugular intrahepatic portosystemic shunt prior to liver transplantation. Pediatr Radiol 2784-86, 1997

79. Stringer MD, Howard ER: Long-term outcome after injection sclerotherapy for esophageal varices in children with extrahepatic portal hypertension. Gut 35257-259, 1994

80. Sudan DL, Shaw BWJ, Langnas A N Causes of late mortality in pediatric liver transplant recipients. Ann Surg 227289-295, 1998

81. Superina RA, Weber JL, Shandling B: A modified Sugiura operation for bleeding varices in children. J Pediatr Surg 18:794-799, 1983

82. Terblanche J, Krige J: Endoscopic therapy in the management of esophageal varices: Injection sclerotherapy and variceal ligation. In Nyhus LM, Baker RJ, Fischer JE (eds): Mastery of Surgery. Boston, Little, Brown and Company, 1997, pp 1329-1349

83. Thapa BR, Mehta S Endoscopic sclerotherapy of esophageal varices in infants and children. J Pediatr Gastroenterol Nutr 10:430-434, 1990

84. Voorhees AB Jr, Chaitman E, Schneider S, et al: Portal-systemic encephalopathy in the noncirrhotic patient. Effect of portal-systemic shunting. Arch Surg 1076594563, 1973

85. Wang J, Cox KL, Dake M, et a1 Transjugular intrahepatic portosystemic shunt placement in a child complicated by perforated Roux-en-Y portoenterostomy. J Pediatr Gastroenterol Nutr 25:421-425, 1997

86. Weinberg GD, Matalon TA, Brunner MC, et al: Bleeding stoma1 varices: Treatment with a transjugular intrahepatic portosystemic shunt in two pediatric patients. J Vasc Interv Radiol 6:233-236, 1995

87. Yachha SK, Sharma BC, Kumar M, et al: Endoscopic sclerotherapy for esophageal varices in children with extrahepatic portal venous obstruction: A follow-up study. J Pediatr Gastroenterol Nutr 24:49-52, 1997

33:1628-1632, 1998

32611527-1532, 1992

154~112-116, 1995

88. Zollinger RM: Atlas of Surgical Operations. New York, Macmillan, 1975

Address reprint requests to F. C. Ryckman, MD

Children’s Hospital Medical Center 3333 Bumet Avenue

OSB - 3, Pediatric Surgery Office Cincinnati, OH 45229

e-mail: [email protected]

causes and management of portal hypertension in the pediatric population

Documents