A MERICAN A SSOCIATION FOR

T HE STUDY OF LIVER D I S E ASES

HEPATOLOGY, VOL. 63, NO. 4, 2016

Primary Prophylaxis of Variceal Bleeding

in Children and the Role of MesoRex
Bypass: Summary of the Baveno VI
Pediatric Satellite Symposium
Benjamin L. Shneider,1 Jean de Ville de Goyet,2 Daniel H. Leung,1 Anshu Srivastava,3 Simon C. Ling,4 Mathieu Duche,5
Patrick McKiernan,6 Riccardo Superina,7 Robert H. Squires,8 Jaime Bosch,9 Roberto Groszmann,10 Shiv K. Sarin,11
Roberto de Franchis,12 and George V. Mazariegos8

Approaches to the management of portal hypertension and variceal hemorrhage in pediatrics remain controversial, in large
part because they are not well informed by rigorous clinical studies. Fundamental biological and clinical differences pre-
clude automatic application of approaches used for adults to children. On April 11-12, 2015, experts in the field convened
at the first Baveno Pediatric Satellite Meeting to discuss and explore current available evidence regarding indications for
MesoRex bypass (MRB) in extrahepatic portal vein obstruction and the role of primary prophylaxis of variceal hemorrhage
in children. Consensus was reached regarding MRB. The vast majority of children with extrahepatic portal vein obstruc-
tion will experience complications that can be prevented by successful MRB surgery. Therefore, children with extrahepatic
portal vein obstruction should be offered MRB for primary and secondary prophylaxis of variceal bleeding and other com-
plications, if appropriate surgical expertise is available, if preoperative and intraoperative evaluation demonstrates favorable
anatomy, and if appropriate multidisciplinary care is available for postoperative evaluation and management of shunt
thrombosis or stenosis. In contrast, consensus was not achieved regarding primary prophylaxis of varices. Although variceal
hemorrhage is a concerning complication of portal hypertension in children, the first bleed appears to be only rarely fatal
and the associated morbidity has not been well characterized. Conclusion: There are few pediatric data to indicate the effi-
cacy and safety of pharmacologic or endoscopic therapies as primary prophylaxis or that prevention of a sentinel variceal
bleed will ultimately improve survival; therefore, no recommendation for primary prophylaxis with endoscopic variceal liga-
tion, sclerotherapy, or nonspecific beta-blockade in children was proposed. (HEPATOLOGY 2016;63:1368-1380)

A
pproaches to the management of portal tinct from that in adults. Therefore, caution should
hypertension (PHT) and variceal hemor- be exercised in the direct extrapolation of the screen-
rhage in children remain controversial, in ing and treatment practices for cirrhosis in adults to
large part because they are not well informed by rigor- children. Primary prophylaxis of variceal hemorrhage
ous clinical studies.(1) Underpinning the practice of in adults with cirrhosis is widely accepted,(2) but it is
endoscopic screening, variceal ligation/banding, and not clear if this should be extended to children. Avail-
surgical interventions is an accurate understanding of able data in pediatrics are quite limited and mostly
the natural history of PHT in children, which is dis- retrospective.

Abbreviations: BA, biliary atresia; EHPVO, extrahepatic portal vein obstruction; EVL, endoscopic variceal ligation; EVS, endoscopic variceal sclero-
therapy; HVPG, hepatic venous pressure gradient; MRB, MesoRex bypass; NSBB, nonspecific beta-blocker; PHT, portal hypertension; RR, recessus of
Rex.
Received June 1, 2015; accepted September 5, 2015.
The Baveno VI Pediatric Satellite Symposium was sponsored by Children’s Hospital of Pittsburgh of UPMC.
Copyright VC 2015 by the American Association for the Study of Liver Diseases.

View this article online at wileyonlinelibrary.com.

DOI 10.1002/hep.28153
Potential conflict of interest: Dr. Shneider consults and owns stock in Bristol-Myers Squibb. He consults for Vertex. Dr. Bosch consults and received
grants from Exalenz. He consults for Gilead, Conatus, Intercept, Actelion, and Falk.

1368
HEPATOLOGY, Vol. 63, No. 4, 2016 SHNEIDER ET AL.

As such, there has been a multinational effort among care, prompt recognition of features of bleeding by
internationally recognized pediatric and adult hepatolo- caregivers, and anticipatory guidance to seek immedi-
gists and surgeons to advance our understanding of the ate medical attention for potential bleeding. This low
pathophysiology and natural history of PHT in chil- mortality, per se, may not drive clinical decision making
dren. These activities seek to develop strategies for the vis-a-vis primary intervention. The clinical consequen-
management of variceal hemorrhage and have taken ces of first variceal hemorrhage are not well docu-
place at the last three Baveno workshops, which have mented in the pediatric literature. For instance, it is
been held every 5 years since 1990.(2,3) On April 11-12, not known if an episode of bleeding predisposes to
2015, experts in the field convened the first Baveno acute or chronic liver failure, development of multior-
Pediatric Satellite Meeting, entitled “Controversies in gan failure, or a more rapid overall decompensation of
the Management of Varices in Children—An Interna- the underlying liver disease. These issues could influ-
tional Forum,” in Baveno, Italy, to participate in high- ence decision making regarding primary prophylaxis
level discussions and explore currently available evidence and need to be prospectively studied.
regarding two major unresolved issues in pediatrics: (1) A key factor in considering primary prophylaxis of
the role of primary prophylaxis of variceal hemorrhage variceal hemorrhage is an understanding of the timing
in children and (2) the indications for MesoRex bypass of that event. The timing of first variceal hemorrhage
(MRB) in extrahepatic portal vein obstruction in BA is profoundly impacted by liver transplantation
(EHPVO). as a therapeutic modality, making transplantation a
The disease processes that most commonly lead to competing risk factor in the “altered” natural history of
PHT in children are fundamentally different from this disease. A recent report from Bic^etre found that
those in adults. Biliary atresia (BA) and EHPVO are 75% of first variceal bleeding occurred by 2.5 years of
two common causes of PHT in children. The patho- age,(9) while studies from Helsinki found that early
physiologies of these diseases are clearly distinct from variceal hemorrhage correlated with a failed Kasai
the primarily hepatocellular diseases (e.g., hepatitis C, (defined as total bilirubin >40 mM, 2.3 mg/dL).(10) In
alcohol-related liver disease, and fatty liver disease) contrast, data from Denver found that 50% of the first
that lead to cirrhosis in adults. In BA and EHPVO, episodes of bleeding in longer-term follow-up children
PHT develops early in the clinical course of disease, with BA and their native liver occurred by 5-6 years of
leading to variceal hemorrhage in children who have age.(11) Most of the historical literature on EHPVO
well-compensated disease and less comorbidity and preceded use of MRB as a preemptive therapy.
can, therefore, “tolerate” the bleeding episodes more Increased understanding and use of preemptive MRB
readily than an adult with decompensated cirrhosis. may markedly change the natural history of EHPVO.
Careful analysis of large published series suggests that In Chandigarh, 50% of first variceal hemorrhage in
the mortality risk after first variceal hemorrhage in EHPVO occurred by 6 years of age.(12) The course of
children may be less than 1%.(4-8) This low mortality is disease after the first episode of variceal hemorrhage
absolutely dependent upon ready access to medical can be quite variable and dependent on the disease

ARTICLE INFORMATION:
From the 1Texas Children’s Hospital and the Department of Pediatrics, Baylor College of Medicine, Houston, TX; 2Bambino Gesu Child-
ren’s Hospital and Tor Vergata Roma University, Rome, Italy; 3Sanjay Gandhi Postgraduate Institute of Medical Sciences, Lucknow, India;
4
Hospital for Sick Children and the Department of Paediatrics, University of Toronto, Toronto, Canada; 5Hepatologie Pediatrique and
Centre de Reference National de l’Atresie des Voies Biliaires, Radiologie Pediatrique, Universite Paris-Sud 11, H^
opital Bic^etre, AP-HP, Le
Kremlin-Bic^etre, France; 6Birmingham Children’s Hospital, Birmingham, United Kingdom; 7Lurie Children’s Hospital, Chicago, IL;
8
Children’s Hospital of Pittsburgh of UPMC, Pittsburgh, PA; 9Hospital Clinic-IDIBAPS and CIBEREHD, Barcelona, Spain; 10Yale Uni-
versity School of Medicine, New Haven, CT; 11Institute of Liver and Biliary Sciences, New Delhi, India; 12University of Milan, Milan,
Italy.

ADDRESS CORRESPONDENCE AND REPRINT REQUESTS TO:

Benjamin L. Shneider, M.D. CCC 1010.00, Houston, TX 77030.
Pediatric Gastroenterology, Hepatology, and Nutrition E-mail Benjamin.Shneider@bcm.edu
Texas Children’s Hospital Tel: 11-832 822 3608
6701 Fannin Street Fax: 11-832 825 3633

1369
SHNEIDER ET AL. HEPATOLOGY, April 2016

process, interventions, and the development of sponta- scores other problems associated with EHPVO,
neous portosystemic shunts. One of the inherent com- including growth failure, mild disseminated intravas-
plexities of pediatrics is the remarkable variability of cular coagulation (attributed to chronic subclinical
disease processes leading to PHT and potential influen- endotoxemia(27)), and portal hypertensive biliopa-
ces of systemic diseases on overall prognosis, such as in thy.(28-30) Children with EHPVO commonly demon-
cystic fibrosis or congenital hepatic fibrosis/autosomal strate impaired quality of life affecting all domains
recessive polycystic kidney disease. when measured using a standard, validated tool that
assesses physical, social, scholastic functioning, and
emotional domains.(21) Growth failure and massive
Natural History of EHPVO splenomegaly are independent determinants of
EHPVO causes 60% of cases of PHT among chil- adverse quality of life, and scholastic functioning is
dren in the developing world compared to 20% in the the most affected domain. Successful variceal eradica-
tion has not been shown to improve quality of life for
developed world.(6,13) Up to 70% of children with
these children.(21)
EHPVO initially present to medical services with
Children with EHPVO are therefore commonly
upper gastrointestinal hemorrhage, which mostly
affected by its complications, which include not only
occurs before 10 years of age.(14,15) However, the
variceal hemorrhage but also the adverse effects of por-
cumulative incidence of variceal bleeding continues to
tosystemic shunting and metabolic disturbance
increase through adolescence; and after the index
(Table 1). The potential for MRB to provide restora-
bleed, episodes characteristically recur about once per
tion of normal physiologic blood flow (Fig. 1) and to
year in children left without secondary prophylactic
ameliorate the complications of EHPVO highlights
treatment.(14-17) Even among children treated with
the need for careful consideration of the optimal case
endoscopic sclerotherapy following their index bleed,
selection and timing for this surgical intervention.
up to 25% will have interval bleeding prior to comple-
tion of esophageal variceal eradication. The cumulative
incidence of bleeding after eradication of varices ranges MRB: Technical Issues in
up to 30% over a 15-year follow-up.(12,16,18) Endo-
scopic variceal ligation (EVL) performs better than Approach
sclerotherapy, with a 4% rate of interval bleeding The recessus of Rex (RR) is the entry point for sys-
before variceal eradication is complete and a low risk of temic blood flow to the body during intrauterine life and
rebleeding thereafter.(19) Portal hypertensive gastropa- is the only segment of the umbilical venous pathway that
thy is commonly present in children with EHPVO remains patent throughout life (Fig. 2A,B).(31-33) It is
undergoing treatment for variceal bleeding but, in con- beside the liver hilum in an advantageous position that
trast to adult data, is rarely a source of significant typically precludes its involvement when a cavernoma-
bleeding in children.(20) tous transformation occurs in the porta hepatis second-
Splenomegaly is uniformly present in most series of ary to portal vein trunk thrombosis. This advantage is
pediatric EHPVO, and up to 30% of affected children lost when there is direct involvement of the RR through
will show hypersplenism (defined by leukopenia, umbilical catheter placement or omphalitis. Most chil-
thrombocytopenia, and palpable splenomegaly).(21) dren with EHPVO have normal liver parenchyma with
Most children with splenomegaly are asymptomatic, low-pressure intrahepatic portal venous systems, making
but some develop acute abdominal pain due to splenic it possible to bypass the splanchnic blood to a patent RR
infarction or perisplenitis.(21) with a large vascular conduit to decompress and cure
Other complications of EHPVO arise secondarily PHT. In <5% of cases, parenchymal fibrosis or venopa-
to portosystemic shunting. Minimal hepatic ence- thy may limit the flow, preventing cure.
phalopathy occurs in about one-third of children with Not all RR and intrahepatic portal venous systems are
EHPVO, especially those who develop spontaneous patent, and if MRB is being considered earlier in the
portosystemic shunts or who have surgical shunts course (preemptive strategy), it is important to select
created.(22-24) Hepatopulmonary syndrome is a well- patients with favorable anatomy (type A) (Fig. 3), where
recognized complication of EHPVO, and portopul- 100% of children had a successful preemptive MRB.(31)
monary hypertension is also rarely reported.(25,26) Although computed tomographic angiography and
Complex multifactorial pathophysiology likely under- Doppler sonography are sometimes sufficient to assess

1370
HEPATOLOGY, Vol. 63, No. 4, 2016 SHNEIDER ET AL.

TABLE 1. Complications of EHPVO Shown to Improve or Resolve With MRB With Comparison to
Distal Splenorenal Shunt
Comparison to Distal
Complication Outcome Mechanism Splenorenal Shunt
Variceal hemorrhage(4,14,15) 100% efficacy with patent MRB(35,58) Reduction/normalization of portal pressures Similar
Biliopathy(29,59) Potential reversal(30) (Fig. 1) Decompression of porto-portal collateral veins(30) Similar
Hypersplenism(4) Resolution(58) (Fig. 1) Reduction/normalization of portal pressures Superior(34) (Fig. 1)
Minimal hepatic Improved(37) Restoration of normal portal blood flow(37) Superior(24,37)
encephalopathy(22,23,37) 1. Short-term memory
2. Motor skill dexterity
Hepatic dysfunction Resolution of Restoration of normal portal blood flow Superior(60)
1. Coagulopathy
2. Protein C and S deficiencies(60)
Liver atrophy Normalization of liver volume (Fig. 1) Restoration of normal portal blood flow Superior
Somatic growth failure(28) Catch-up growth: weight and height z IGF-1 and HGH production amelioration of Superior(34,36)
scores(36) portal enteropathy; metabolic rescue(36)
Portopulmonary syndrome 1: Complete resolution within 6 Restoration of normal portal flow Superior
hepatopulmonary syndrome months(61)
Portopulmonary syndrome 2: Uncertain effect: two children died Restoration of normal portal flow Potentially superior
arterial pulmonary hypertension from pulmonary hypertensive crises
after successful MRB; one child is
stable and continues to be followed
for improvement (R. Superina, per-
sonal communication)
Regenerative nodules and adenoma Complete resolution (Fig. 1) Restoration of normal portal flow Superior
related to portosystemic shunting

Abbreviations: HGH, human growth hormone; IGF-1, insulin-like growth factor 1.

for favorable anatomy, portography (retrograde internal or thrombosed intrahepatic portal venous branches
jugular venography) achieves the highest sensitivity and were contraindications. Two-thirds of patients were
specificity by allowing a detailed anatomical study of the referred secondary to bleeding that required ongoing
intrahepatic venous system (Fig. 4).(31) The creation of a interventions. In this cohort, all patients with patent
bypass between the superior mesenteric vein and the RR MRB were cured of their PHT bleeding with no
with a large autologous venous conduit is currently the recurrences. No subsequent esophageal endoscopy,
procedure of choice (Fig. 2C).(33) The bypass is ideally a banding, or nonspecific beta-blocker (NSBB) therapy
large-diameter conduit, with the most direct routing was reported. After MRB, mean spleen size decreased
possible; this allows low resistance, high flow, and low from 7 6 5 cm below the left costal margin to non-
risk of thrombosis. The material for the bypass varies palpable and platelet counts increased by a mean of
according to surgeon preference, although the best 78 6 64 3 103/mL to levels higher than with portosys-
results have been achieved using an internal jugular vein temic shunting and in the normal range in most
autograft. Homografts, cryopreserved or not; arterial cases.(34)
grafts; and prosthetic material have poorer outcomes. Age at time of surgery does not appear to impact
The pre-pancreatic retropyloric route of MRB is the the success of MRB; however, the volume of blood
most direct and optimal approach.(33) In order to provide flow in the vein graft is inversely proportional to age at
sufficient space for the conduit and avoid compression surgery. Infants and young children are more likely to
by the liver edge, resecting a portion of segments 4 and 3 have blood flow in the vein graft that approximates a
of the liver is necessary in all cases (Fig. 5).(33) Additional normal portal vein for a child of similar body surface
resection of segment 3 is recommended when a bypass is area.(35) Importantly, up to 15% of children with
MRB may require postoperative percutaneous dilata-
created closer to the gastric or splenic veins.
tion of the anastomosis or an intrahepatic venous
stricture to maintain satisfactory patency of the
Outcomes of MRB bypass.(36) Ultimately, MRB not only leads to success-
ful restoration of portal flow but has the potential to
In one of the largest single-center experiences of reverse a variety of known complications of EHPVO
161 children presenting with idiopathic EHPVO, (Table 1). This is remarkably distinct from decom-
84% were successfully treated with MRB. Inadequate pressive surgeries, like the distal splenorenal shunt,

1371
SHNEIDER ET AL. HEPATOLOGY, April 2016



FIG. 1. Results of MRB. Before (A, C, E) and after (B, D, F) imaging is shown for MRB performed for extrahepatic portal vein
obstruction. Case 1: A and B show resolution of biliopathy (black arrows, A) after reconstitution of portal blood flow (white arrows,
B) by MRB (1, superior mesenteric vein; 2, bypass; 3, RR). Case 2: C and D demonstrate increase in liver mass (L) and decrease in
spleen size. E and F show resolution of hepatic nodules (white arrows, E). (A and B reproduced with permission from the publisher,
Millar(62).)


which reduce portal pressure but increase portosyste- significant complications like minimal hepatic ence-
mic shunting (Table 1). Persistent shunting after these phalopathy, hepatic adenoma, and portopulmonary
procedures has the potential to lead to predictable and hypertension.(37-40)

1372
HEPATOLOGY, Vol. 63, No. 4, 2016 SHNEIDER ET AL.



FIG. 2. Embryologic origins of the RR and diagram of the MRB. (A) Prenatal vascular anatomy. (B) Postnatal vascular anatomy indicating
involution of umbilical vein and the ductus venosus but persistence of the RR. (C) The approach to the MRB is depicted. Abbreviations: A,
ductus venosus; B, bypass; M 5 superior mesenteric vein; P, portal vein (in A,B) or extrahepatic portal vein thrombosis (in C); R, Rex recessus;
S, splenic vein; U, umbilical vein.


Consensus: MRB as Satellite meeting was that MRB, when highly feasi-
ble (see below), should be considered as the primary
Preprimary and Primary approach in children with EHPVO who have evi-
dence of a cavernoma, especially if there have been
Prophylaxis of any clinical complications of EHPVO. In order to
propose preemptive MRB for EHPVO, the follow-
Complications of PHT in ing criteria should be met to ensure a success rate of
EHPVO >90%.
1. Patient qualifications for MRB include favorable
In light of the morbidity associated with EHPVO anatomy and medical condition defined as
and the physiologic responses to MRB, the consen- a. “A” type anatomy confirmed by wedge two-
sus opinion of the panel at the Baveno VI Pediatric dimensional portography (retrograde internal

1373
SHNEIDER ET AL. HEPATOLOGY, April 2016


jugular venogram) by an experienced and mul-
tidisciplinary team (Fig. 3)
b. Normal hepatic venous pressure gradient
(HVPG) determined at the time of wedged
portography and/or confirmatory liver biopsy
revealing no fibrosis or underlying liver disease
that would impact portal flow
c. Patent superior mesenteric vein and splenic vein
without thrombosis confirmed using site-
specific expertise in either computed tomo-
graphic or magnetic resonance angiography
d. Bilateral internal jugular vein patency with
equivalent symmetry confirmed by imaging
such as ultrasound or computed tomographic
angiography
e. Hematological evaluation excluding an intrinsic
prothrombotic condition that would not be
reversed by MRB
f. Body weight >8 kg
FIG. 3. Classification of patency of the intrahepatic portal vein
in patients with cavernomatous transformation of the portal vein. g. Echocardiography excluding significant cardio-
(A) Complete patency of the intrahepatic portal venous system vascular abnormalities and significant pulmo-
including the RR and liver segmentation (SII-SVIII). (B-E) Par- nary hypertension
tial obstruction of the left liver (B), thrombosis of the right liver
(C), thrombosis of the left liver (D), and extensive thrombosis of 2. Surgical/institutional qualifications for performing
the intrahepatic portal veins (E) Pattern A is associated with preemptive MRB include
greatest likelihood of MRB success. Abbreviation: R, Rex reces- a. Pediatric liver transplant and vascular and hep-
sus. (Reproduced with permission of the publisher, Bertocchini
et al.(31))
atobiliary expertise with multidisciplinary care
including pediatric hepatology and interven-

tional radiology
b. Prior successful MRB experience
The panel also agreed by consensus that MRB is the
preferred approach for secondary prophylaxis per prior



FIG. 4. Imaging of the intrahepatic portal vasculature. Left: Direct (intraoperative) portal venography demonstrating complete pat-
ency. Right: Transjugular retrograde wedged portography also indicating complete patency.


1374
HEPATOLOGY, Vol. 63, No. 4, 2016 SHNEIDER ET AL.


and is rarely, if ever, indicated for the treatment of
EHPVO. Splenectomy is associated with lifelong risk
of infection with encapsulated bacteria and the poten-
tial risk for pancreatic injury. It may preclude future
use of distal splenorenal shunting. It provides no bene-
ficial impact on variceal hemorrhage or on the compli-
cations of portosystemic shunting and, as such,
typically has no role in the treatment of EHPVO.

Endoscopic Therapy as
Primary Prophylaxis of
Variceal Hemorrhage
Primary prophylaxis of variceal hemorrhage, while
FIG. 5. Intraoperative view of completed MRB procedure. Note generally accepted for adults, has not been endorsed,
the resection of a large portion of parenchyma of segments III by and large, for children with PHT for multiple
and IV yielding direct access of the internal jugular vein graft to
the RR, creating a route with no risk of extrinsic compression. reasons.(1-3,41) These include (1) limited prospective
natural history data including clinical consequences of

variceal hemorrhage, (2) limited data defining an
Pediatric Baveno V statement guidelines.(3) The degree endoscopic pattern of gastroesophageal varices predic-
of required certainty of the feasibility of MRB is less tive of high risk of bleeding, (3) minimal efficacy and
when used for secondary prophylaxis and would be safety data on the use of NSBBs in pediatric PHT,
decided by local clinical expertise. Alternatives to (4) limited reports of efficacy and safety of primary
MRB for secondary prophylaxis include endoscopic prophylaxis of varices in children, 5)
improvements in the emergency management of vari-
1. EVL ceal bleeding, and (6) use of liver transplantation as
2. Endoscopic variceal sclerotherapy (EVS) dictated definitive therapy for decompensated disease or for
by patient size restrictions disease with poor near-term prognosis (e.g., “failed”
3. Distal splenorenal shunt hepatoportoenterostomy).
The panel was in uniform agreement that splenec- Experiences with primary prophylaxis of variceal
tomy only addresses issues related to hypersplenism hemorrhage in children are limited and mostly focused

TABLE 2. Endoscopic Summary and Outcomes of Selected Case Series of Pediatric Primary Prophylaxis
Duche et al.(8) Lampela et al.(10) Celinska-Cedro et al.(43) Goncalves et al.(42)
Liver disease BA BA Cirrhosis, n 5 16 EHPVO, Cirrhosis, n 5 40
n 5 15 EHPVO, n 5 10
Form of primary prophylaxis EVS/EVL/combined EVS EVL* EVS
Number 16/15/5 27 31 50
Eradication of varices, n (%) 27 (75) 8 (29.6) 28 (90.3) 47 (94)
Mean no. of sessions to eradicate 4.2 2 2 6 0.9 3.5 6 2.6
Rebleeding, n (%) 2 (6) 12 (44) 0 (0) 12 (24)
Complications, n (%) 3 (9.6)† 4 (8.8)‡ 0 (0) 0 (0)§
Bleeding after eradication 0 – 0 (0) 3 (6)
Variceal recurrence % 10 (37) – 3 (9.6%) –
Outcome 58% underwent liver trans- 79% transplanted – 18% died, no
plantation within 14 months within 5 years effect on survival
3-year follow-up, 97%
survival

*Sixteen subjects (52%) had been pretreated with propranolol without success.
†
Complications included fever after sclerotherapy and early bleeding.
‡
Complications included superficial ulceration (n 5 3) and perforation (n 5 1), leading to death.
§
Fever and pain were not considered complications.

1375
SHNEIDER ET AL. HEPATOLOGY, April 2016

on the management of BA (Table 2).(8,10,42,43) The

approach seems feasible, although questions remain as
Complexity in the Use of
to the impact of the intervention on long-term out- NSBBs as Potential Primary
come. Patients with BA, particularly those with persis-
tent jaundice, are at high risk for developing Prophylaxis in Children with
worrisome endoscopic signs of PHT early in life.(9) In
BA, specific endoscopic variceal patterns have been
PHT
shown to be predictive of the risk of gastrointestinal One approach to reducing risk of variceal bleeding is
bleeding.(8,9,42) Grade 3 varices, grade 2 varices with the use of NSBBs as preemptive therapy, which is well
esophageal red markings, and, most significantly, the established in adults.(2) The optimal use of NSBBs
presence of gastric varices along the cardia have been may now be restricted to a “window” of optimal effec-
reported to be independent risk factors for bleeding.(8) tiveness with minimal harm.(45) The mechanism of
A 10-year follow-up of 225 children with BA with action is to induce splanchnic vasoconstriction through
low-risk signs at the first endoscopic examination ß2 receptor blockade while reducing cardiac output
revealed emergence of high-risk gastroesophageal vari- through ß1 blockade. Reports that include the use of
ces in nearly 34% at a median age of 4.25 years.(44) NSBBs in children are few.(3) Small numbers, a retro-
Progression to high-risk varices was related to bilirubi- spective approach with divergent etiologies of PHT,
nemia (>100 mM, 5.8 mg/dL) in children older than and absence of controls make these studies of limited
18 months at the first endoscopy. On the basis of these quality to serve as evidence to support the use of
findings, primary prophylaxis of varices was under- NSBBs in children with PHT. Adding propranolol to
taken in 36 children with BA using EVL in 15, EVS a regimen of esophageal variceal ligation did not yield
in 16, and combination therapy in five. On average 4.2 benefit.(46) Despite these findings in the available liter-
endoscopy sessions were required for eradication. ature, NSBBs are commonly used by pediatric gastro-
Rebleeding occurred in four children, while varices enterologists who responded to a survey on the
relapsed in 10. Twenty-one of these children under- management of PHT in children with BA.(1) The use
went liver transplantation at a mean age of 29 months. of NSBBs in severe infantile hemangiomas provides
Primary prophylaxis was well tolerated with one tran- useful observations into the cardiac effects of 1-3 mg/
sient fever and two early bleeding events from incom- kg/day of propranolol as well as the adverse effects
plete eradication of the varices.(8) associated with this dose in infants without liver dis-
In Helsinki, an endoscopic surveillance program ease or PHT.
offering prophylactic EVS reported outcomes in 47 Assessment of the effectiveness of NSBBs in the
children with BA beginning at 12 months of age.(10) prevention of initial and subsequent esophageal variceal
EVS was repeated at 2-week to 4-week intervals until hemorrhage remains complicated, even for adults.
varices were eradicated. By a median age of 11 months, Measurement of the HVPG with the goal of decreas-
60% demonstrated varices. Of these, 38% had either ing it to less than 12 mm Hg or by more than 20% of
grade 2 or grade 3 esophageal varices at baseline endos- baseline has become the best, albeit invasive, measure
copy. Overall, of the 28 patients with varices, 57% of the pharmacologic benefit from NSBBs.(47) A sim-
received EVS before the first bleeding episode, and plification for assessing the HVPG response to NSBBs
four of them (25%) bled. Recurrent bleeding after is to measure the acute change in HVPG 15 minutes
commencing sclerotherapy occurred in 44%. Esopha- after an intravenous propranolol infusion (0.15 mg/kg
geal varices occurred with similar frequency after failed in 10 minutes). A decrease >10% or below 12 mm Hg
and successful portoenterostomy in 28 (60%) patients. defines an adequate response after acute propranolol
All 16 patients with upper gastrointestinal bleeding testing, both in primary and in secondary prophy-
had undergone a failed portoenterostomy, whereas no laxis.(48) HVPG is feasible and can be safely measured
bleeding occurred after successful portoenterostomy. in children, but intrahepatic venovenous shunts may
Bile flow after portoenterostomy did not impact the complicate its interpretation.(49) There are no studies
development of varices, but lack of bile flow was asso- in children assessing the effect of NSBBs on HVPG,
ciated with an increased risk of bleeding from those although effects on splenic pulp pressure have been
varices. This issue needs to be considered in the devel- reported.(50)
opment of treatment plans and in the design of future Available data do not support the routine use of
clinical studies. NSBBs in infants, children, or adolescents with PHT.

1376
HEPATOLOGY, Vol. 63, No. 4, 2016 SHNEIDER ET AL.

The physiologic impact of NSBBs on the hyperdy- for varices to address parents’ desire to know the level
namic physiology of children with PHT with or with- of risk of variceal bleeding for their child.(1) Family cir-
out liver disease is not known, nor is the impact on cumstances, including understanding anticipatory
hemodynamics after a bleeding episode well described guidance with plans for ready access to medical care in
in children. NSBBs may have deleterious effects on the event of an episode of bleeding, have the potential
survival in adults with decompensated liver disease and to impact clinical decision making related to primary
refractory ascites, leading to the idea of a window of prophylaxis.
effectiveness for NSBBs and a potential change in the
approach to the use of NSBBs.(45,51,52) Well-designed
studies to assess the physiological impact of NSBBs in
Endoscopic Management for
children would be an important starting point to deter- Secondary Prophylaxis of
mine the efficacy and safety of NSBBs in children with
PHT. Variceal Hemorrhage
The rich experience in secondary prophylaxis poten-
Lack of Consensus on tially informs primary prophylaxis from a technical per-
spective and in the context of safety and efficacy. EVL
Primary Prophylaxis of is highly effective at controlling acute bleeding and
Variceal Hemorrhage not safer and more effective than sclerotherapy for second-
ary prophylaxis.(3,19,56) Where technically feasible
From EHPVO based upon size, EVL is the preferred endoscopic
method for acute esophageal variceal bleeding, primary
Because of the lack of sufficient evidence, no con- and secondary prophylaxis.(3,43) EVL is generally well
sensus could be achieved related to primary prophylaxis tolerated. Though 10%-15% of children complain of
of varices in children due to cirrhosis. The strength of transient retrosternal pain, esophageal stricture second-
the indication for primary prophylaxis in children is ary to EVL has not been reported in the pediatric liter-
unclear because mortality from the first bleed appears ature. The major limitation of this technique is the
to be very low and the associated morbidity needs to be lack of pediatric-specific equipment, and hence, it is
better defined, ideally in prospective longitudinal stud- not always possible to pass a loaded bander (four-
ies. To the many uncertainties of the risks and benefits shooter or six-shooter) through the pharynx in small
of beta-blockers and endoscopic management of vari- children.(56)
ces in children have now been added the possible EVS has been used in pediatric practice for more
adverse neurodevelopmental consequences of repeated than 50 years and likely will continue to be a part of
general anesthesia in young children, such as would be the therapeutic armamentarium in children given
required for primary endoscopic prophylaxis.(53,54) patient-size limitations. Most pediatric experience is
To reduce the need for invasive procedures and gen- with intravariceal injection of ethanolamine oleate,
eral anesthesia, there is interest in identifying noninva- sodium tetradecyl sulfate, or sodium morrhuate.(56)
sive tests that may replace endoscopy as a screening Cyanoacrylate and human thrombin injections have
test for bleeding risk or help to triage high-risk chil- been used in the treatment of gastric varices.(57) EVS is
dren to undergo endoscopy. Several such tests have quite effective, with control of acute bleeding in 90%
been studied in children, including platelet count, of cases. However, complications are more likely fol-
spleen size, platelet/spleen size ratio, a clinical predic- lowing emergency EVS and include ulceration, medi-
tion rule, and transient elastography of liver or astinitis, esophageal perforation, chylothorax, and
spleen.(55) Some of these tests show promise in pre- pneumothorax. Procedural mortality for EVS is
dicting varices in children and potential to limit the approximately 0.3% (unpublished review of 19 peer-
number of endoscopic and anesthesia sessions, reviewed reports). In the only randomized control
although their accuracy is inadequate for the diagnosis trial,(19) Zarger et al. demonstrated that banding
of varices that may specifically benefit from primary required fewer treatments, had reduced complications,
prophylaxis. Even in the absence of consensus over pri- and had lower rebleeding (4% versus 25%) versus scle-
mary prophylaxis of varices in children, some special- rotherapy. EVS for esophageal varices is only indicated
ists use endoscopy and/or noninvasive tests to screen for acute bleeding and for secondary prophylaxis in

1377
SHNEIDER ET AL. HEPATOLOGY, April 2016

infants and in children where EVL is technically

odology of Diagnosis and Therapy in Portal Hypertension. Pediatr
impossible or unavailable.(3) Secondary prophylaxis Transplant 2012;16:426-437.
often begins at day 6 after the index bleeding episode, 4) Alvarez F, Bernard O, Brunelle F, Hadchouel P, Odievre M,
followed by EVL every 2 weeks to 4 weeks for up to Alagille D. Portal obstruction in children. I. Clinical investiga-
tion and hemorrhage risk. J Pediatr 1983;103:696-702.
five sessions to eradicate varices.(3) Failure to eradicate
5) Howard ER, Stringer MD, Mowat AP. Assessment of injection
varices should lead to consideration of an alternative sclerotherapy in the management of 152 children with oesopha-
approach such as transjugular intrahepatic portosyste- geal varices. Br J Surg 1988;75:404-408.
mic shunting, portosystemic shunting, or liver 6) Poddar U, Thapa BR, Rao KL, Singh K. Etiological spectrum of
esophageal varices due to portal hypertension in Indian children:
transplantation. is it different from the West? J Gastroenterol Hepatol 2008;23:
1354-1357.

Conclusion 7) Miloh T, Kratimenos P, Shneider B, Arnon R, Suchy F, Kerkar

N. Acute variceal bleeding in children: single center experience.
HEPATOLOGY 2010;52:260A.
Although variceal hemorrhage is a concerning com- 8) Duche M, Ducot B, Ackermann O, Baujard C, Chevret L,
plication of PHT in children, the first bleed appears to Frank-Soltysiak M, et al. Experience with endoscopic manage-
ment of high-risk gastroesophageal varices, with and without
be only rarely fatal and the associated morbidity has
bleeding, in children with biliary atresia. Gastroenterology 2013;
not been well characterized. There are few pediatric 145:801-807.
data to indicate the efficacy and safety of pharmaco- 9) Duche M, Ducot B, Tournay E, Fabre M, Cohen J, Jacquemin
logic or endoscopic therapies as primary prophylaxis or E, et al. Prognostic value of endoscopy in children with biliary
atresia at risk for early development of varices and bleeding. Gas-
that prevention of a sentinel variceal bleed will ulti- troenterology 2010;139:1952-1960.
mately improve survival. Therefore, no recommenda- 10) Lampela H, Kosola S, Koivusalo A, Lauronen J, Jalanko H, Rintala R,
tion for primary prophylaxis with EVL, EVS, or et al. Endoscopic surveillance and primary prophylaxis sclerotherapy
NSBBs in children can be proposed at this time. of esophageal varices in biliary atresia. J Pediatr Gastroenterol Nutr
2012;55:574-579.
However, the vast majority of children with 11) Miga D, Sokol RJ, Mackenzie T, Narkewicz MR, Smith D,
EHPVO will experience complications that can be Karrer FM. Survival after first esophageal variceal hemorrhage in
prevented by successful MRB surgery. Therefore, chil- patients with biliary atresia. J Pediatr 2001;139:291-296.
12) Poddar U, Thapa BR, Singh K. Endoscopic sclerotherapy in
dren with EHPVO should be offered MRB for pri-
children: experience with 257 cases of extrahepatic portal venous
mary and secondary prophylaxis of variceal bleeding obstruction. Gastrointest Endosc 2003;57:683-686.
and other complications, if appropriate surgical exper- 13) Howard ER, Stamatakis JD, Mowat AP. Management of esopha-
tise is available, if preoperative and intraoperative eval- geal varices in children by injection sclerotherapy. J Pediatr Surg
1984;19:2-5.
uation demonstrates favorable anatomy, and if 14) Abd El-Hamid N, Taylor RM, Marinello D, Mufti GJ, Patel R,
appropriate multidisciplinary care is available for post- Mieli-Vergani G, et al. Aetiology and management of extrahe-
operative evaluation and management of shunt throm- patic portal vein obstruction in children: King’s College Hospital
bosis or stenosis. experience. J Pediatr Gastroenterol Nutr 2008;47:630-634.
15) Weiss B, Shteyer E, Vivante A, Berkowitz D, Reif S, Weizman
Z, et al. Etiology and long-term outcome of extrahepatic portal
Acknowledgment: We thank Anne McGinnis and vein obstruction in children. World J Gastroenterol 2010;16:
Maggie Bray for their outstanding organization of 4968-4972.
the meeting and the Baveno VI Scientific Committee 16) Zargar SA, Yattoo GN, Javid G, Khan BA, Shah AH, Shah
NA, et al. Fifteen-year follow up of endoscopic injection sclero-
for endorsing the satellite symposium. therapy in children with extrahepatic portal venous obstruction.
J Gastroenterol Hepatol 2004;19:139-145.
REFERENCES 17) Shah SR, Mathur SK. Presentation and natural history of variceal
bleeding in patients with portal hypertension due to extrahepatic
1) Gana JC, Valentino PL, Morinville V, O’Connor C, Ling SC. portal venous obstruction. Indian J Gastroenterol 2003;22:217-
Variation in care for children with esophageal varices: a study of 220.
physicians’, patients’, and families’ approaches and attitudes. 18) Itha S, Yachha SK. Endoscopic outcome beyond esophageal vari-
J Pediatr Gastroenterol Nutr 2011;52:751-755. ceal eradication in children with extrahepatic portal venous
2) de Franchis R; Baveno V Faculty. Revising consensus in portal hyper- obstruction. J Pediatr Gastroenterol Nutr 2006;42:196-200.
tension: report of the Baveno V Consensus Workshop on Methodol- 19) Zargar SA, Javid G, Khan BA, Yattoo GN, Shah AH, Gulzar
ogy of Diagnosis and Therapy in Portal Hypertension. J Hepatol GM, et al. Endoscopic ligation compared with sclerotherapy for
2010;53:762-768. bleeding esophageal varices in children with extrahepatic portal
3) Shneider BL, Bosch J, de Franchis R, Emre SH, Groszmann RJ, venous obstruction. HEPATOLOGY 2002;36:666-672.
Ling SC, et al. Portal hypertension in children: expert pediatric opin- 20) Poddar U, Thapa BR, Singh K. Frequency of gastropathy and
ion on the report of the Baveno V Consensus Workshop on Meth- gastric varices in children with extrahepatic portal venous

1378
HEPATOLOGY, Vol. 63, No. 4, 2016 SHNEIDER ET AL.

obstruction treated with sclerotherapy. J Gastroenterol Hepatol children with primary extrahepatic portal vein thrombosis
2004;19:1253-1256. improves fluid neurocognitive ability. Pediatrics 2006;117:e405-
21) Krishna YR, Yachha SK, Srivastava A, Negi D, Lal R, Poddar U. e412.
Quality of life in children managed for extrahepatic portal venous 38) Guerin F, Porras J, Fabre M, Guettier C, Pariente D, Bernard
obstruction. J Pediatr Gastroenterol Nutr 2010;50:531-536. O, et al. Liver nodules after portal systemic shunt surgery for
22) D’Antiga L, Dacchille P, Boniver C, Poledri S, Schiff S, Zancan extrahepatic portal vein obstruction in children. J Pediatr Surg
L, et al. Clues for minimal hepatic encephalopathy in children 2009;44:1337-1343.
with noncirrhotic portal hypertension. J Pediatr Gastroenterol 39) Moon SB, Jung SE, Ha JW, Park KW, Seo JK, Kim WK. The
Nutr 2014;59:689-694. usefulness of distal splenorenal shunt in children with portal
23) Yadav SK, Srivastava A, Srivastava A, Thomas MA, Agarwal J, hypertension for the treatment of severe thrombocytopenia and
Pandey CM, et al. Encephalopathy assessment in children with leukopenia. World J Surg 2008;32:483-487.
extra-hepatic portal vein obstruction with MR, psychometry and 40) Ecochard-Dugelay E, Lambert V, Schleich JM, Duche M,
critical flicker frequency. J Hepatol 2010;52:348-354. Jacquemin E, Bernard O. Portopulmonary hypertension in liver
24) Srivastava A, Yadav SK, Lal R, Yachha SK, Thomas MA, disease presenting in childhood. J Pediatr Gastroenterol Nutr
Saraswat VA, et al. Effect of surgical portosystemic shunt on 2015;61:346-354.
prevalence of minimal hepatic encephalopathy in children with 41) Molleston JP, Shneider BL. Preventing variceal bleeding in
extrahepatic portal venous obstruction: assessment by magnetic infants and children: is less more? Gastroenterology 2013;145:
resonance imaging and psychometry. J Pediatr Gastroenterol 719-722.
Nutr 2010;51:766-772. 42) Goncalves ME, Cardoso SR, Maksoud JG. Prophylactic sclero-
25) Borkar VV, Poddar U, Kapoor A, Ns S, Srivastava A, Yachha SK. therapy in children with esophageal varices: long-term results of
Hepatopulmonary syndrome in children: a comparative study of a controlled prospective randomized trial. J Pediatr Surg 2000;35:
non-cirrhotic vs. cirrhotic portal hypertension. Liver Int 2015;35: 401-405.
1665-1672. 43) Celinska-Cedro D, Teisseyre M, Woynarowski M, Socha P,
26) Condino AA, Ivy DD, O’Connor JA, Narkewicz MR, Socha J, Ryzko J. Endoscopic ligation of esophageal varices for
Mengshol S, Whitworth JR, et al. Portopulmonary hypertension prophylaxis of first bleeding in children and adolescents with por-
in pediatric patients. J Pediatr 2005;147:20-26. tal hypertension: preliminary results of a prospective study.
27) Bajaj JS, Bhattacharjee J, Sarin SK. Coagulation profile and pla- J Pediatr Surg 2003;38:1008-1011.
telet function in patients with extrahepatic portal vein obstruction 44) Duche M, Ducot B, Ackermann O, Jacquemin E, Bernard O.
and non-cirrhotic portal fibrosis. J Gastroenterol Hepatol 2001; Progression to high-risk gastroesophageal varices in children with
16:641-646. biliary atresia with low-risk signs at first endoscopy. J Pediatr
28) Sarin SK, Bansal A, Sasan S, Nigam A. Portal-vein obstruction Gastroenterol Nutr 2015;60:664-668.
in children leads to growth retardation. HEPATOLOGY 1992;15: 45) Ge PS, Runyon BA. When should the beta-blocker window in
229-233. cirrhosis close? Gastroenterology 2014;146:1597-1599.
29) Dhiman RK, Behera A, Chawla YK, Dilawari JB, Suri S. Portal 46) dos Santos JM, Ferreira AR, Fagundes ED, Ferreira AP,
hypertensive biliopathy. Gut 2007;56:1001-1008. Ferreira LS, Magalhaes MC, et al. Endoscopic and pharmaco-
30) Dhiman RK, Saraswat VA, Valla DC, Chawla Y, Behera A, logical secondary prophylaxis in children and adolescents
Varma V, et al. Portal cavernoma cholangiopathy: consensus with esophageal varices. J Pediatr Gastroenterol Nutr 2013;56:
statement of a working party of the Indian National Association 93-98.
for Study of the Liver. J Clin Exp Hepatol 2014;4:S2-S14. 47) D’Amico G, Garcia-Pagan JC, Luca A, Bosch J. Hepatic vein
31) Bertocchini A, Falappa P, Grimaldi C, Bolla G, Monti L, de pressure gradient reduction and prevention of variceal bleeding in
Ville de Goyet J. Intrahepatic portal venous systems in children cirrhosis: a systematic review. Gastroenterology 2006;131:1611-
with noncirrhotic prehepatic portal hypertension: anatomy and 1624.
clinical relevance. J Pediatr Surg 2014;49:1268-1275. 48) La Mura V, Abraldes JG, Raffa S, Retto O, Berzigotti A,
32) de Ville de Goyet J, Gibbs P, Clapuyt P, Reding R, Sokal EM, Garcia-Pagan JC, et al. Prognostic value of acute hemodynamic
Otte JB. Original extrahilar approach for hepatic portal revascu- response to i.v. propranolol in patients with cirrhosis and portal
larization and relief of extrahepatic portal hypertension related to hypertension. J Hepatol 2009;51:279-287.
later portal vein thrombosis after pediatric liver transplantation. 49) Miraglia R, Luca A, Maruzzelli L, Spada M, Riva S, Caruso
Long term results. Transplantation 1996;62:71-75. S, et al. Measurement of hepatic vein pressure gradient in
33) di Francesco F, Grimaldi C, de Ville de Goyet J. Meso-Rex children with chronic liver diseases. J Hepatol 2010;53:624-
bypass—a procedure to cure prehepatic portal hypertension: the 629.
insight and the inside. J Am Coll Surg 2014;218:e23-e36. 50) Ozsoylu S, Kocak N, Yuce A. Propranalol therapy for portal
34) Lautz TB, Keys LA, Melvin JC, Ito J, Superina RA. Advantages hypertension in children. J Pediatr 1985;106:317-320.
of the Meso-Rex bypass compared with portosystemic shunts in 51) Serste T, Melot C, Francoz C, Durand F, Rautou PE, Valla D,
the management of extrahepatic portal vein obstruction in chil- et al. Deleterious effects of beta-blockers on survival in patients
dren. J Am Coll Surg 2013;216:83-89. with cirrhosis and refractory ascites. HEPATOLOGY 2010;52:1017-
35) Superina R, Bambini DA, Lokar J, Rigsby C, Whitington PF. 1022.
Correction of extrahepatic portal vein thrombosis by the mesen- 52) Krag A, Wiest R, Albillos A, Gluud LL. The window hypothe-
teric to left portal vein bypass. Ann Surg 2006;243:515-521. sis: haemodynamic and non-haemodynamic effects of beta-
36) Lautz TB, Sundaram SS, Whitington PF, Keys L, Superina RA. blockers improve survival of patients with cirrhosis during a win-
Growth impairment in children with extrahepatic portal vein dow in the disease. Gut 2012;61:967-969.
obstruction is improved by mesenterico-left portal vein bypass. 53) Wang X, Xu Z, Miao CH. Current clinical evidence on the
J Pediatr Surg 2009;44:2067-2070. effect of general anesthesia on neurodevelopment in children: an
37) Mack CL, Zelko FA, Lokar J, Superina R, Alonso EM, Blei updated systematic review with meta-regression. PLoS One
AT, et al. Surgically restoring portal blood flow to the liver in 2014;9:e85760.

1379
SHNEIDER ET AL. HEPATOLOGY, April 2016

54) Backeljauw B, Holland SK, Altaye M, Loepke AW. Cognition 59) Khuroo MS, Yattoo GN, Zargar SA, Javid G, Dar MY, Khan
and brain structure following early childhood surgery with anes- BA, et al. Biliary abnormalities associated with extrahepatic por-
thesia. Pediatrics 2015;136:e1-e12. tal venous obstruction. HEPATOLOGY 1993;17:807-813.
55) Adami MR, Ferreira CT, Kieling CO, Hirakata V, Vieira SM. 60) Mack CL, Superina RA, Whitington PF. Surgical restoration of
Noninvasive methods for prediction of esophageal varices in portal flow corrects procoagulant and anticoagulant deficiencies
pediatric patients with portal hypertension. World J Gastroen- associated with extrahepatic portal vein thrombosis. J Pediatr
terol 2013;19:2053-2059. 2003;142:197-199.
56) McKiernan P, Abdel-Hady M. Advances in the management of 61) Fuchs J, Warmann S, Kardorff R, Rosenthal H, Rodeck B,
childhood portal hypertension. Expert Rev Gastroenterol Hepatol Ure B, et al. Mesenterico-left portal vein bypass in children
2015;9:575-583. with congenital extrahepatic portal vein thrombosis: a unique
57) Oh SH, Kim SJ, Rhee KW, Kim KM. Endoscopic cyanoacrylate curative approach. J Pediatr Gastroenterol Nutr 2003;36:213-
injection for the treatment of gastric varices in children. World J 216.
Gastroenterol 2015;21:2719-2724. 62) Millar AJW. Surgical disorders of the liver and bile ducts and
58) Sharif K, McKiernan P, de Ville de Goyet J. Mesoportal bypass portal hypertension. In: Kelly D, ed. Diseases of the Liver and
for extrahepatic portal vein obstruction in children: close to a Biliary System in Children. 3rd ed. Oxford, UK: Wiley-Black-
cure for most! J Pediatr Surg 2010;45:272-276. well; 2008:433-474.

1380