|

Received: 12 October 2016 Accepted: 19 October 2016

DOI: 10.1111/liv.13277

REVIEW ARTICLE

Prevention and treatment of variceal haemorrhage in 2017

Felix Brunner1|Annalisa Berzigotti1|Jaime Bosch1,2

1
Swiss Liver Center,Hepatology,University
Clinic for Visceral Surgery and
Medicine,Inselspital, Bern University
Hospital,University of Bern, Bern,
Switzerland
2
Hepatic Hemodynamic Laboratory,Hospital
Clinic-IDIBAPS and CIBEREHD,University
of Barcelona, Barcelona, Spain
Funding Information
None.
Correspondence
Prof. Jaime Bosch, MD, PhD, FRCP, Hepatic
Hemodynamic Laboratory, Hospital Clinic,
IDIBAPS and CIBEREHD, University of
Barcelona, Barcelona, Spain.
Email: jbosch@clinic.ub.es
Handling Editor: Francesco Negro
Abstract
Variceal haemorrhage is a major complication of portal hypertension that still causes
high mortality in patients with cirrhosis. Improved knowledge of the pathophysiology
of portal hypertension has recently led to a more comprehensive approach to prevent
all the complications of this condition. Thus, optimal treatment of portal hypertension
requires a strategy that takes into account the clinical stage of the disease and all the
major variables that affect the risk of progression to the next stage and death. In pa-
tients with compensated liver disease, the correction of factors influencing the pro-
gression of fibrosis, in particular aetiologic factors, is now feasible in many cases and
should be achieved to prevent the development or progression of gastroesophageal
varices and hepatic decompensation. Once gastroesophageal varices have developed,
non-selective beta-blockers remain the cornerstone of therapy. Carvedilol provides a
greater decrease in portal pressure and is currently indicated as a first-choice therapy
for primary prophylaxis. The treatment of acute variceal haemorrhage includes a com-
bination of vasoactive drugs, antibiotics and endoscopic variceal band ligation. In high-
risk patients, the early use of transjugular intrahepatic portosystemic shunt (TIPS)
lowers the risk of re-bleeding and improves survival. Transjugular intrahepatic porto-
systemic shunt is the choice for uncontrolled variceal bleeding; a self-expandable
metal stent or balloon tamponade can be used as a bridging measure. The combination
of non-selective beta-blockers and endoscopic variceal band ligation reduces the risk
of recurrent variceal bleeding and improves survival. In these cases, statins seem to
further improve survival. Transjugular intrahepatic portosystemic shunt is indicated in
patients who rebleed during secondary prophylaxis.

KEYWORDS
cirrhosis, portal hypertension, variceal bleeding, varices

1|INTRODUCTION bleeding from gastroesophageal varices (GEVs) and portal hyper-

tensive gastropathy,a scites, spontaneous bacterial peritonitis
Portal hypertension is the most severe and frequent compli- (SBP), hepatorenal syndrome (HRS), hepatic encephalopathy (HE),
cation of chronic liver disease. This complication is the cause hepatopulmonary andporto-pulmonary syndrome, bacteremia and
of most of the severe clinical symptoms of cirrhosis, such as hypersplenism.

Abbreviations: (c)ACLD, (compensated) advanced chronic liver disease; ACE, angiotensin-converting enzyme; BRTO, balloon-occluded retrograde transvenous obliteration; CSPH, clinically
significant portal hypertension; CT, computer tomography; EIS, endoscopic injection sclerotherapy; e-NOS, endothelial NO synthase; EV, oesophageal varices; EVL, endoscopic variceal band
ligation; GEV, gastroesophageal varices; GOV1+2, gastroesophageal varices type 1+2; HE, hepatic encephalopathy; HRS, hepatorenal syndrome; HSC, hepatic stellate cells; HVPG, hepatic
venous pressure gradient; ICU, intensive care unit; IGV1+2, isolated gastric varices type 1 +2; ISMN, isosorbide-5-mononitrate; LSM, liver stiffness measurements; MRI, magnetic resonance
imaging; NASH, non-alcoholic steatohepatitis; NO, nitric oxide; NSBB, non-selective beta-blockers; OCA, obeticholic acid; PPG, portal pressure gradient; PTFE, polytetrafluoroethylene; RCT,
randomized controlled trial; SBP, spontaneous bacterial peritonitis; SEC, sinusoidal endothelial cells; TE, transient elastography; TIPS, transjugular intrahepatic portosystemic shunt; VEGF,
vascular endothelial growth factor.

104 | 2017 John Wiley & Sons A/S.

wileyonlinelibrary.com/journal/liv Liver International 2017; 37(Suppl. 1): 104115
Published by John Wiley & Sons Ltd
BRUNNER etal.
Portal hypertension is defined as an increased portosystemic pres-
sure gradient (difference between pressures in the portal vein and the
inferior vena cava), which can be evaluated clinically by measuring the
hepatic venous pressure gradient (HVPG) through hepatic vein cath-
1
eterization. This review will only focus on portal hypertension due
to chronic liver disease (sinusoidal portal hypertension), which is the
leading aetiology, followed by schistosomiasis (in developing coun-
tries), portal and splenic vein thrombosis, Budd-Chiari syndrome and
other less frequent conditions due to presinusoidal or post-sinusoidal
block.
Portal hypertension is defined by a HVPG>5mm Hg, but the risk
of developing GEV and the clinical complications of decompensated
chronic liver disease (e.g. ascites, variceal bleeding and overt HE) are
2
only present when it reaches 10mm Hg. Thus, HVPG 10mm Hg is
called clinically significant portal hypertension (CSPH).
At diagnosis, up to half of the patients with compensated cirrhosis
have developed GEV. This is even more frequent in patients with se-
3
verely impaired liver function (Child-Pugh class B/C). GEVs develop
with an annual incidence of 5%-9% in patients without varices at diag-
nosis.4,5 The rate of progression from small to large varices is similar.
Prospective studies in the 1980s showed that the risk of first variceal
bleed increases with impairment of liver function, size of varices and
presence of red colour signs over the varices (red wales, red spots, dif-
fuse redness) with an annual risk of approximately 12% (5% for small
varices and 15% for large varices).6,7 Bleeding from ruptured GEV is a
major medical emergency, with a 6-week mortality of approximately
15% especially patients with poor liver function (Child-Pugh C). In
many cases, death does not occur due to bleeding, but to infections,
HRS and liver failure because of the bleeding episode. In the last three
decades, improvement in evidence-based-guided treatment of portal
hypertension and management of acute bleeding episodes has cut the
mortality due to GEV bleeding in half.8
2| STAGES OF CIRRHOSIS AND
PORTAL HYPERTENSION: FROM
FIBROSIS TO CLINICALLY SIGNIFICANT
PORTAL HYPERTENSION
In chronic liver disease, fibrosis progresses leading to cirrhosis and
continues to worsen as long as the aetiologic factor is present. In daily
practice, liver fibrosis is usually measured non-invasively by determin-
ing liver stiffness to assess the degree of fibrosis without a biopsy.9
Since cirrhosis is an anatomical diagnosis requiring a liver biopsy, the
term compensated advanced chronic liver disease (cACLD) has been
10
suggested rather than cirrhosis. In asymptomatic compensated pa-
tients, values below 10kPa on liver stiffness measurements (LSMs) by
transient elastography (TE) exclude cACLD, while values between 10
and 15kPa suggest sACLD but are not diagnostic, and correspond to
a grey zone, and cACLD is very likely in patients with values above
15kPa.10 Liver stiffness measurements have also been found to corre-
late with the presence of CSPH (values >13 suggest cirrhosis, and >20
are almost pathognomonic). The diagnostic accuracy of elastography
|
105
Key points
Treatment of portal hypertension must take into account
the clinical stage of chronic advanced liver disease.
Non-selective beta-blockers remain the cornerstone of
the prevention of variceal bleeding, re-bleeding and other
complications of portal hypertension, and they improve
survival in patients with optimal hemodynamic response.
Major advances include the use of carvedilol for primary
prophylaxis of variceal bleeding and the use of statins
combined with NSBB and EVL to reduce mortality in pa-
tients who have bled from variceal rupture.
TIPS is now used early after a first episode of bleeding in
high-risk patients, as a life-saving procedure in massive
bleeding, and to prevent rebleeding in patients not re-
sponding to standard therapy (NSBB+EVL).
increases if it is combined with other data suggesting portal hyperten-
sion, such as a reduced platelet count and increased spleen size in
the so-called LSPS index.11 The non-invasive diagnosis of CSPH in a
patient without varices can also be made by detecting portal-systemic
collaterals with imaging techniques (ultrasound, CT or MRI). The lat-
est Baveno consensus conference recommends that patients with an
LSM value 20kPa and/or a decreased platelet count (<150G/L) un-
dergo endoscopic screening for the presence of GEV, while patients
with both LSM<20kPa and platelet count >150G/L can safely skip
endoscopy due to the low risk of having varices requiring treatment.10
Based on the natural history of the disease, ACLD/cirrhosis can
be classified into five stages according to the presence or absence of
complications (Table1).12,13
3|PATHOPHYSIOLOGY OF PORTAL
HYPERTENSION AND ITS CHANGE WITH
DISEASE PROGRESSION
The pathophysiology of portal hypertension is essentially
based on applying Ohms law to hemodynamics: Pressure
Gradient=ResistanceFlow. Thus, portal hypertension may be
due to increased resistance to portal blood flow, increased blood
flow through the portal-collateral system or a combination of both
(Figure1). In ACLD, increased resistance to portal blood flow occurs
not only because of the obvious architectural changes in the liver vas-
culature (fibrosis, nodules, angiogenesis and sinusoidal remodelling)
but also due to dynamic changes in the sinusoids. Once they have
been activated through inflammatory signalling or other stimuli, the
adjacent hepatic stellate cells (HSCs), which reside in the space of
Disse (virtual space between hepatocytes and sinusoidal endothelial
cells [SECs]), begin to deposit extracellular matrix proteins and col-
lagen. Microthrombi in small hepatic veins aggravate the structural
component by causing hepatocyte drop-out and areas of parenchymal
 |
106
T A B L E 1 Stages of advanced chronic liver disease
Stage 1a Stage 1b Stage 2
Characteristics Compensated Compensated Compensated
without without with varices
varices varices
No CSPH CSPH CSPH
1-y mortality 1.5%
Complications (annual 7% GEV 6.6% Ascites
incidence) 4% Ascites 4% GEV
bleeding
CSPH, clinically significant portal hypertension (HVPG>10mm Hg); GEV, gastroesophageal varices; HE, hepatic encephalopathy.
extinction. This structural component is thought to account for about
70% of the increased hepatic resistance to portal blood flow.
The remaining 30% is a dynamic component including active con-
traction of activated HSC wrapped around sinusoids, myofibroblasts
in the portal tracts, and vascular smooth muscle cells in the hepatic
vasculature. In response to chronic liver damage, the SECs lose their
fenestrae, which normally allow passive transport of macromolecules
from the sinusoidal lumen to hepatocytes. A basement membrane de-
velops along the sinusoidal surface of liver SEC, leading to capillar-
ization of sinusoids that exhibit endothelial dysfunction. Endothelial
dysfunction is characterized by impaired release of vasodilatory agents,
mainly nitric oxide (NO) by the endothelial NO synthase (e-NOS). In a
setting of increased vasoconstrictor drive due to increased endothelin
production, activated renin-angiotensin and adrenergic systems and
enhanced release of thromboxane A2, the lack of sufficient NO results
in hepatic vasoconstriction further increasing hepatic resistance and
worsening portal hypertension (Figure1).14
More advanced portal hypertension with the presence of portal-
systemic collaterals (stage 2) is accompanied by splanchnic vasodila-
tion, increased inflow of blood into the splanchnic organs draining into
the portal venous system and increased portal-collateral blood flow.
Vasodilation is mediated by increased release of vasodilators, includ-
ing NO, carbon monoxide, endogenous cannabinoids and vasodilatory
peptides of splanchnic origin such as glucagon. This is the opposite of
the intrahepatic circulation (which exhibits vasoconstriction) and rep-
resents an adaptive homeostatic response to maintain portal perfusion
of the liver, which is reduced by the development of prehepatic and
intra-hepatic portal-systemic shunts. However, the increased portal-
collateral blood flow further worsens portal hypertension, accelerating
the development of collaterals causing further splanchnic vasodilation,
creating a vicious cycle. Severe splanchnic vasodilation may result in
peripheral resistance and hypotension and effective hypovolaemia,
causing activation of endogenous vasoactive systems, which trigger
plasma volume expansion (Figure1). This expanded plasma volume
and the additional blood volume returning from the portal-systemic
collaterals to the right atrium promotes increased cardiac output and
a full-blown hyperkinetic syndrome. In stages 3 and 4, vasodilation is
so intense that clinical sodium retention is triggered and located in the
peritoneal cavity in the form of ascites (Figure1).
BRUNNER etal.
Stage 3 Stage 4 Stage 5
Variceal bleeding First non-bleeding Any second
without any decompensation decompensating
other (e.g. ascites, HE, event
complication jaundice)
CSPH CSPH CSPH
2% 10% 21% 87% 5-y-mortality
21% Progress to 10% Progress to
stage 4 stage 5
Collateral formation is not only caused by increased portal
pressure, but requires active angiogenesis, driven by vascular en-
dothelial growth factor (VEGF). VEGF not only activatesamong
othersthe eNOS signalling pathway, but also promotes the forma-
tion of portosystemic collaterals and GEV. Physiologically, porto-
systemic collaterals are shunts between the portal and the systemic
circulation that contribute to the hyperkinetic systemic circulation
and hepatic encephalopathy as well as impaired metabolism of
drugs and toxins.
4|Current treatments for portal hypertension
(Table2)
4.1|Non-selective beta-blockers (NSBBs)
remain the cornerstone of prevention of bleeding
Traditional NSBB have been used for more than 30 years
Standard NSBBs (propranolol, nadolol and timolol) decrease portal
pressure by reducing portal-collateral blood flow through reduction
of the cardiac index (via beta1-receptor blockade) and splanchnic
vasoconstriction (via beta2-adrenoceptor blockade).5,15 Therefore,
hyperkinetic circulation is necessary for them to act. For more than
30years, NSBBs have been shown to effectively prevent a first GEV
bleed and recurrence of GEV bleeding, as well as to lower bleeding-
associated mortality.16,17 Moreover, NSBBs increase intestinal transit
time, reduce bacterial translocation and lower the risk of spontaneous
bacterial peritonitis.18 To effectively prevent variceal bleeding, NSBB
should reduce the HVPG to 12mm Hg or below (optimal response) or
at least by 20% of its baseline value (good hemodynamic response).
However, a long-term satisfactory hemodynamic response is only ob-
tained in 33%-50% of patients with NSBB. In non-responders, addi-
tion of low doses of an NO-donor such as isosorbide-5-mononitrate
(ISMN) causes an additional decrease in portal pressure, while in-
creasing side effects.19,20 However, the fixed combination of NSBB
and ISMN is not clearly better than NSBB alone in preventing a first
episode of GEV bleeding, but in an la carte HVPG-guided approach,
the addition of ISMN has been shown to rescue about one third of
non-responders to NSBB.20,21 The recommended dosages of NSBB
and most frequent side effects are listed in Table2.
BRUNNER etal.
F I G U R E 1 Pathophysiology of portal hypertension and potential
targets for drug therapy
Carvedilol is more effective than propranolol/nadolol
In addition to its capacity to block beta1- and beta2-receptors, the
NSBB carvedilol has an intrinsic anti--adrenergic activity and the po-
tential to release NO. This is why it is more effective in lowering portal
hypertension than propranolol/nadolol after acute or chronic adminis-
tration (in a face-to face randomized comparison, the mean HVPG de-
22
crease was 22.2% vs 15.6% respectively). Moreover, more than half
(56%) of non-responders to standard NSBB therapy achieve a good
hemodynamic response with carvedilol. In the same study, the use of
propranolol as a first-line primary prevention followed by carvedilol in
those who did not hemodynamically respond to propranolol resulted
in a good hemodynamic response in 72%.23 Low doses of carvedilol
(<25mg/day) are as effective as relatively high doses (25-50mg/day)
in decreasing HVPG. Low doses (12.5mg/day) of carvedilol are, there-
fore, recommended, as high doses can cause arterial hypotension,
which may enhance sodium retention and worsen ascites. In a primary
prophylaxis randomized controlled trial, carvedilol prevented a first
GEV bleeding episode more effectively than repeat EVL, although this
requires confirmation in further studies.24
4.2|Terlipressin is used for acute bleeding
Terlipressina long-acting synthetic vasopressin analogueis a po-
tent splanchnic vasoconstrictor that also has systemic circulatory ef-
fects, increasing arterial pressure and systemic vascular resistance,
and decreasing cardiac output. A single intravenous administration
leads to a marked decrease in the HVPG of more than 25%, which
|
107
is sustained for several hours.25 Due to its long biological activity,
terlipressin can be administered every 4hours as repeat intravenous
injections, although if it is given as a continuous infusion the total
dose may be decrease. Due to the intravenous administration of this
agent, it is only used for short periods, which in practice is limited to
the treatment of acute variceal bleeding (2-5days) or type-I hepato-
renal syndrome (1-2weeks). Because of the potential ischaemic and
arrhythmic complications, terlipressin should not be used in patients
with a history of coronary, cerebrovascular, peripheral or visceral ar-
terial diseases. Also, it should be used with caution in elderly and/or
hypertensive patients.
4.3|Somatostatin and long-acting somatostatin
analogues are also potent splanchnic vasoconstrictors
Somatostatin is also an effective splanchnic vasoconstrictor that
effectively decreases portal pressure through the inhibition of
glucagon and other vasoactive peptides and the facilitation of adr-
energic vasoconstriction.25,26 Because of its very short half-life
(1-3minutes), it should be administered in a continuous intravenous
infusion. Dose-response studies have shown that to significantly
decrease both HVPG and portal-collateral (azygos) blood flow, a
dose of 500g/hour is required.27 Long-acting analogues (octreo-
tide, vapreotide, and lanreotide) have been developed to overcome
the short half-life of somatostatin. After a bolus of octreotide the,
HVPG markedly decreases in a manner similar to somatostatin
(around 50%), but this effect is short and decreases with repeated
administration. The decrease in the HVPG is not maintained by a
continuous intravenous infusion, but limits the postprandial increase
in portal pressure in patients with portal hypertension. However,
this only partially explains why octreotide successfully controls GEV
bleeding.28
4.4|New drugs
4.4.1|Statins may improve survival
Statins improve endothelial dysfunction in many vascular diseases due
to its beneficial pleiotropic circulatory effects. Most effects are due to
enhanced expression of the KLF-2 transcription factor and reported
genes (eNOS, thrombomodulin and angiopoietin). In experimental cir-
rhosis, statins (simvastatin and atorvastatin) decrease hepatic vascular
resistance, improve endothelial dysfunction and decrease fibrosis. In
patients with cirrhosis, statins moderately lower HVPG (even in pa-
tients being treated by NSBB) and improve quantitative liver func-
tion tests but do not influence the systemic circulation.29 In a recent
RCT in patients who survived an episode of variceal bleeding adding
simvastatin to the standard of care (NSBB and EVL) did not improve
the prevention of variceal rebleeding, but improved survival, mainly
by reducing the number of deaths due to bleeding and infections.30
Large epidemiological surveys have shown that the progression of
liver disease is reduced and mortality decreased in patients with cALD
receiving statins.31
 |
108

T A B L E 2 Drugs for the treatment of portal hypertension

Propranolol Nadolol ISMN Carvedilol Terlipressin Somatostatin Octreotide Simvastatin

Beta-1 and beta-2 Beta-1 and beta-2 No donor Beta-1 and beta-2 Long-acting Peptide hormone, Long-acting Improves endothelial
adrenergic adrenergic adrenergic vasopressin inhibits glucagon SMT analogue dysfunction and reduces
receptor receptor receptor analogue and facilitates intra-hepatic vascular
antagonist (NSBB) antagonist (NSBB) antagonist Splanchnic adrenergic resistance
(NSBB) with vasoconstriction, vasoconstriction
intrinsic anti-1 increase in
activity systemic vascular
resistance
Route Oral Oral Oral Oral Intravenous Intravenous Intravenous Oral
Dosage Start with 10-20mg Start with 20mg Start with 10mg Start with 2mg injection every Bolus 250g Bolus 50g i.v., 20mg once a day.
twice daily daily daily at night 3.125mg twice 4h for 24-48h, followed by followed by After 2wk, increase to
Increase the dose Increase the dose Increase the daily then 1mg/4h for continuous i.v. continuous i.v. 40mg/d if CPK and ALT do
every 2-3d up to every 2-3d up to dose after Increase the up to 5d infusion 250- infusion 50g/h not increase over twice the
the maximal the maximal 2-3d up to dose every 2-3d 500g/h for up to for up to 5d baseline values
tolerated dose tolerated dose 20mg twice a up to 12.5mg/d 5d
Systolic blood Systolic blood day Systolic blood
pressure should pressure should pressure should
remain 100mm remain 100mm remain
Hg and HR Hg and HR 100mm Hg and
50bpm 50bpm heart
rate50bpm
Max dose 320mg/d 160mg/d 40 mg/d 25mg/d 8mg/d for 24-48h 500g/h 50g/h 40mg/d
4mg/d thereafter
Side Bradycardia, Bradycardia, Headache, Arterial Abdominal pain, Hyperglycaemia, Hyperglycaemia, Elevated
effects orthostatic orthostatic orthostatic hypotension, arterial hyperten- vomiting vomiting aminotransferases (not more
hypotension, hypotension, hypotension Sodium sion; less than 3% frequent than in the general
Bronchospasm, bronchospasm, retention, ischaemia population)
erectile erectile ascites (peripheral, Up to 3% rhabdomyolysis in
dysfunction dysfunction intestinal or advanced cirrhosis (bilirubin
myocardial) >5mg)
BRUNNER etal.
BRUNNER etal. |
109

4.4.2|Renin-angiotensin-aldosterone inhibitors: Cyanoacrylate

more studies needed injection

Although NSBB and ACE inhibitors and angiotensin receptor block-

EVL
ers lower the HVPG at a nearly similar rate and appear to be safe in
patients with good liver function (Child-Pugh A),32 the role of these
agents in preventing GEV bleeding and the safety profile in decom-
pensated ACLD require further studies.

4.4.3|Aetiologic treatments reduce portal pressure

with time
GOV 1 GOV 2
Studies in several liver diseases (hepatitis B, hemochromatosis, hepatitis
C, alcoholic liver disease) have shown that successful specific treatment
of liver disease can markedly reduce portal pressure. The time neces- Cyanoacrylate
sary for this beneficial effect varies and is probably longer in clinically injection
decompensated patients. This is highly relevant since the introduction
of direct acting antivirals for hepatitis C virus infection, which is one of
Cyanoacrylate
the main causes of cirrhosis worldwide. It is still uncertain whether the
injection
risk of portal hypertension-related complications will completely disap-
pear, when this will occur and how it can be non-invasively detected,
but these elements are being evaluated in large prospective studies.

4.4.4|Antifibrotic drugs are promising IGV 1 IGV 2

Obeticholic acid (OCA) is a Farnesoid X receptor agonist, a nuclear re-
ceptor in the liver and intestine that suppresses bile acid synthesis
and increases the transport of bile acids from hepatocytes into the
canaliculi. OCA significantly improves liver biochemistry in patients
with primary biliary cholangitis who do not respond to treatment with
ursodeoxycholic acid. OCA also improves the histological features of
patients with non-alcoholic steatohepatitis (NASH). Further trials in
the field of primary biliary cholangitis, NASH, primary sclerosing chol-
33
angitis and biliary atresia are ongoing.
Simtuzumab, a monoclonal antibody directed against lysyl oxidase-
like 2 (LOXL2) enzyme, was developed to target changes in fibrosis. In
a mouse model, simtuzumab reduced bridging fibrosis, total collagen
expression and improved survival. In a recent open-label safety trial,
6 months of treatment with this agent was well tolerated in patients
with hepatitis C and/or HIV and advanced fibrosis, but did not affect
liver histology.34 A large RCT in patients with NASH-related cirrhosis
examining the long-term effects of simtuzumab on liver histology and
HVPG is in progress.
Emricasan, an anti-apoptotic agent with anti-inflammatory and antifi-
brotic properties, has been shown to decrease portal pressure in a recent
35
pilot study in patients with cirrhosis as well as in experimental models.
4.5|Endoscopic therapy: band ligation for
oesophageal varices and obliteration for gastric varices
Endoscopic variceal band ligation (EVL) is the endoscopic treatment
of choice for oesophageal varices (EV) because it is more effective
and safer than endoscopic injection sclerotherapy (EIS). 36
Once
F I G U R E 2 Gastroesophageal and isolated gastric varices:
classification according to Sarin55 and choice of endoscopic
treatment [Colour figure can be viewed at wileyonlinelibrary.com]
EVL is begun for acute bleeding or scheduled primary/secondary
prophylaxis, it should be repeated every 2-4weeks until complete
eradication of the varices is obtained. The term eradication is
not literally the absence of varices, but defines varices that are ab-
sent or small (remnants) and cannot or need not be further ligated.
This definition, therefore, involves a subjective component. After
eradication, endoscopic screening at 1, 6, 12months, and every
12months thereafter is recommended to detect recurrent high-risk
varices.
4.5.1|Endoscopic therapy for gastric varices
Isolated gastric varices (IGV1 and 2) and type 2 gastroesophageal
varices (GOV2), which reach the stomach from the oesophagus on the
side of the greater curvature (Figure2), should be treated by endo-
scopic variceal obliteration with intravariceal injection of a tissue ad-
hesive (e.g. n-butyl-2-cyanoacrylat). This has been shown to be more
effective than band ligation in acute bleeding from gastric varices. A
second glue injection into IGV 2-4weeks later can be considered to
further reduce the risk of re-bleeding. Although gastroesophageal
varices type 1 (GOV1), which extend into the stomach on the side
of the lesser curvature, are usually treated with EVL, centres in Asia
support using variceal obliteration with glue injection (Figure2).10
Thrombin has been used instead of glue.37
 |
110
4.6|Interventional radiology treatments:
intrahepatic shunts and retrograde
obliteration of varices
Transjugular intrahepatic portosystemic shunt (TIPS) is a minimally in-
vasive fluoroscopic-guided procedure that creates a shunt between
a hepatic vein and the intrahepatic portal vein sustained by a metal
stent. TIPS is a portosystemic shunt that is large enough to effectively
decrease portal pressure and small enough not to cause liver failure
and hepatic encephalopathy. Ideally, it should maintain a portal pres-
sure gradient (PPG, the difference between portal and inferior vena
cava pressure) between 10 and 12mm Hg. This requires progressive
dilatation of the stent with balloon angioplasty catheters (usually up
to 8-10mm in diameter) while measuring the PPG. The use of stents
covered with polytetrafluoroethylene (PTFE) is recommended to pre-
vent occlusion of the shunt and maintain rates of primary unassisted
patency similar to those of surgical shunts.38 TIPS has almost com-
pletely replaced surgical shunts because of the reduced surgical mor-
bidity and mortality and similar efficacy. The main indications for TIPS
are the treatment of acute variceal bleeding (as preventive therapy in
high-risk patients or as rescue therapy after failure of medical plus en-
doscopic treatment) and refractory ascites. TIPS is more effective than
alternative therapy in both these indications and may decrease mor-
tality. Evaluation of TIPS candidates should include a detailed history
of previous episodes of HE, assessment of portal vein patency and
anatomy on imaging, absence of biliary dilatation or tumours in the
intrahepatic tract, and a transthoracic echocardiography. Because of
the risk of cardiac decompensation, TIPS is contraindicated in patients
with congestive heart failure, severe pulmonary hypertension and tri-
cuspid regurgitation. Recurrent episodes of HE, portal vein thrombo-
sis and hepatocellular carcinoma are relative contraindications. HE is
more frequent in patients over 70 or with large bore shunts (12mm
or more in diameter).38. HE occurs in about one third of the cases.
Although in most cases it can be managed medically, in about 6%-
10% of patients HE is severe and may require closure or reduction
of TIPS.39
Balloon-occluded retrograde transvenous obliteration (BRTO) is
another minimally invasive procedure in which gastric or mesenteric
varices are retrogradely accessed from the femoral vein via a spon-
taneous splenorenal shunt and occluded with tissue adhesives or
thrombogenic intravascular coils. This method has been found to be
very efficient in controlling acute bleeding from gastric varices, but
no randomized controlled trials are available. It should be considered
when endoscopic and medical therapies fail and the patient is not a
good candidate for TIPS.40
5| CLINICAL SCENARIOS FOR THERAPY
The treatment of varices and variceal haemorrhage vary based on
the clinical scenario and the previously described stages of ACLD.
An overview of the different treatment strategies is provided in
Table3.
BRUNNER etal.
5.1|Compensated patients without
gastroesophageal varices: prefer aetiologic treatments
In early stages, the goal of treatment is to avoid the progression of
cirrhosis to hepatic decompensation by preventing a further increase
in portal hypertension. In this early stage, the aetiologic factors can
be targeted (e.g. antiviral treatment, alcohol withdrawal, weight loss,
phlebotomies and chelator therapy). Successful antiviral treatment
may lead to improvement of fibrosis, thus improving portal hyperten-
sion, although little is still known about how long this can take, if there
are no-return points, and how this resolution can be assessed non-
invasively. Exercise and weight loss have been reported to decrease
HVPG in obese patients independently of the underlying cause.41
Also, patients with ACLD should abstain from alcohol.10
The results of a large study in cACLD on the so-called preprimary
prophylaxis of the development of GEV by long-term administration
of timolol (a NSBB) were negative.5,15 However, this study importantly
showed that risk of developing varices, and especially ascites, bleeding
or hepatic encephalopathy, is more frequent in compensated patients
with CSPH (HVPG of at least 10mm Hg). Accordingly, the recommen-
dations from the most recent Baveno conference state that the goal
of therapy in these patients should be the prevention of decompen-
sation. The preliminary results of a multicentre randomized controlled
trial (ClinicalTrials number: NCT01059396) designed for this purpose
suggest that the long-term administration of propranolol (or carvedilol
in patients whose HVPG does not decrease by at least 10% from base-
line after intravenous propranolol) may prevent decompensation-free
survival and almost halve the incidence of ascites.
5.2|Compensated Patients with Varices
5.2.1|Patients with small varices and no other risk
factors for bleeding: beta-blockers
The use of NSBB could be beneficial in patients with small (<5mm)
EV without red colour signs and good/moderate liver function (Child-
Pugh A/B), but this must be confirmed in further trials.
5.2.2|Patients with high-risk varices
(primary prophylaxis of variceal bleeding): beta-
blockers or ligation
These include patients with small EV, but with red colour signs and/or
Child-Pugh C, and patients with medium or large (>5mm) EV. The goal
of therapy is to prevent first variceal bleeding and other complications
from portal hypertension thus improving survival.
In patients with high risk but small varices, primary prophylactic
treatment with carvedilol or a standard NSBB can be initiated.
In patients with medium or large varices (>5mm), primary pro-
phylaxis can be provided using either standard NSBB or carvedilol or
EVL. EVL is as effective and can be used as an alternative to NSBB
based on local resources and patient preferences. In general, EVL is
the preferred option in patients with contraindications or intolerance
BRUNNER etal.
T A B L E 3 Overview of treatment goals and strategies for portal hypertension according to the clinical scenario
Clinical scenario
Compensated patients in whom
CSPH has been excluded by
HVPG measurement
Compensated patients with
likely or confirmed CSPH
(LSM>20kPa and/or platelet
count<150G/L, or with
measured HVPG10mm Hg)
Compensated patients with
small varices without RCS in
Child-Pugh A/B
Compensated patients with
small varices with RCS AND
decompensated patients Child
C with small varices
Any patient with EV at least
medium/large in size or with
gastric varices
Acute variceal bleeding
Prevention of recurrent
bleeding after first
haemorrhage
CSPH, clinical significant portal hypertension; EV, oesophageal varices; EVL, endoscopic variceal band ligation; GOV, gastroesophageal varices; HVPG,
hepato-venous pressure gradient; IGV, isolated gastric varices; ISMN, isosorbide-5-mononitrate; NSBB, non-selective beta-blockers; RCS, red colour signs;
SEMS, self-expandable metal stent.
a
Insufficiently explored.
b
Role of carvedilol not yet defined.
Goal of treatment
Prevent
Progression to CSPH
Prevent development of
varices
Prevent decompensation
presence or absence of
Prevent decompensation
Stabilize/lower portal
pressure in order to reduce
the risk of variceal
progression
Prevent the first variceal
bleeding episode
Prevent decompensation if
compensated
Prevent further decompensa-
tion if decompensated
Prevent the first variceal
bleeding episode
Prevent decompensation if
compensated
Prevent further decompensa-
tion if decompensated
Achieve haemostasis
Prevent rebleeding within 5 d
Minimize 6-wk mortality
Prevent recurrent bleeding
Prevent other decompensa-
tion
Improve survival
Baveno recommendations
regarding screening and
surveillance endoscopy
No need for screening of
varices (no CSPH).
Screen for varices
If no varices are seen:
repeat endoscopy after 2
or 3y, depending on
ongoing liver damage
if varices are present: see
below
Repeat endoscopy after
1 or 2y, depending on
the presence or absence
of ongoing liver damage
Begin NSBB or carvedilol
No further endoscopies
needed
Begin NSBB or carvedilol
EVL if intolerant or
contraindications to
NSBB
No further diagnostic
endoscopies needed
For patients on EVL:
repeat ligation every
2-4wk until complete
eradication of EV,
afterwards 1, 6, 12mo
and every year
N/A
Repeated EVL every
2-4wk until complete
eradication, afterwards 1,
6, 12mo and every year
|
111
Accepted pharmacological therapies and
general measures
Aetiologic therapy (e.g. alcohol withdrawal,
antiviral therapy, weight reduction if over-
weight/obese)
Aetiologic therapy (e.g. alcohol withdrawal,
antiviral therapy, weight reduction if over-
weight/obese)
Aetiologic therapy (e.g. alcohol withdrawal,
antiviral therapy, weight reduction if over-
weight/obese)
NSBB can be considereda
Aetiologic therapy (e.g. alcohol withdrawal,
antiviral therapy, weight reduction if over-
weight/obese)
NSBB/carvedilol
Aetiologic therapy (e.g. alcohol withdrawal,
antiviral therapy, weight reduction if over-
weight/obese)
NSBB/carvedilol
Vasoactive drugs and prophylactic antibiotics
Restrictive transfusion policy
Perform endoscopy within 12h (achieve
stabilization first)
EVL for EV and cyanoacrylate injection for
GOV2/IGV
Consider SEMS for refractory/massive bleeding
Consider early TIPS in high-risk patients
(Child-Pugh C<13 points or Child B with active
haemorrhage) and in patients with early
rebleeding within 5d
First option (standard): NSBBb +/ ISMN plus
EVL
If repeated EVL not possible: NSBB+/ ISMN
if NSBB not tolerated: consider TIPS
Consider adding simvastatin
After bleeding from gastric varices: consider
TIPS
 |
112
to NSBB (after trying to switch to carvedilol, which is usually better
tolerated). However, contrary to NSBB, EVL does not prevent other
complications of portal hypertension, such as bleeding from portal hy-
pertensive gastropathy or colopathy, ascites and SBP.10,41,42 Gastric
varices without previous bleeding can also be treated with NSBB.10
5.3|Treatment of acute bleeding
Acute variceal haemorrhage is a medical emergency that requires
prompt and intensive medical care, ideally in an ICU. Initial manage-
ment should target control of the haemorrhage, prevent early (within
5days) re-bleeding, deterioration of liver function and other bleeding-
related complications, mainly infections, acute kidney injury and HE.
5.3.1| GENERAL MANAGEMENT:
RESUSCITATION, ANTIBIOTICS AND
PREVENTION OF ENCEPHALOPATHY
Resuscitation measures should follow the classical ABC (Airway,
Breathing, Circulation) principle to restore hemodynamic stability
and to maintain appropriate delivery of oxygen to the tissues. Airway
protection with orotracheal intubation is mandatory in unconscious
patients or those with severe haemorrhage (hematemesis) before en-
doscopy to prevent airway aspiration. Plasma volume expansion is pri-
marily infused for volume replacement. Packed red blood cells should
be transfused restrictively (target haemoglobin level7g/dL), because
a less restrictive policy (target haemoglobin 9g/dL) can increase mor-
tality.43 This restriction does not apply to Child C patients, in cases
of rapid ongoing bleeding or patients with a history of ischaemic car-
diovascular disease. Early prophylactic antibiotics, to prevent the occur-
rence of infections after a bleeding episode and improve survival should
be also started on admission.44 The Baveno consensus conference rec-
ommends the use of intravenous ceftriaxone 1g/24hours in patients
with advanced cirrhosis, in areas with high prevalence of quinolone-
resistant bacterial infections and in patients on previous quinolone
prophylaxis. The choice of the ideal antimicrobial prophylaxis must be
made in each centre depending on local antibiotic resistances. Also, the
patients individual risk factors (e.g. advanced cirrhosis and previous
infections) must be taken into account.10 Furthermore, patients with
signs of infection on admission should be extensively evaluated and
treated. Prevention of HE: Lactulose per os, via a nasogastric tube or
rectally, is recommended to prevent HE. Rifaximin may also be effec-
tive, but has not been specifically investigated yet in these cases.
5.3.2| SPECIFIC THERAPY: FROM
VASOACTIVE DRUGS TO TIPS
5.3.2.1|Vasoactive drugs
When variceal haemorrhage is suspected, intravenous vasoactive
therapy (terlipressin, somatostatin or octreotide) should be begun.
This is usually started on arrival in the emergency room, before
BRUNNER etal.
endoscopy. All drugs are thought to be equally effective in control-
ling bleeding and preventing re-bleeding within 5days, but the quality
of the evidence varies, because terlipressin is the only drug that has
been shown to be effective to control bleeding, decrease transfusion
requirements and reduce the 6-week mortality in placebo-controlled
clinical trials.45 A recent large face-to-face study that showed equal
efficacy should be considered with caution because subtherapeu-
tic doses of terlipressin and suboptimal doses of somatostatin were
used.46 Recommended doses are given in Table2. Vasoactive drug
infusion is maintained for 2-5days.
5.3.2.2|Endoscopic therapy
Emergency endoscopy should be scheduled as soon as possible, at
the latest 12hours after admission; 250mg erythromycin intrave-
nously prior to endoscopy accelerates the gastric emptying of clots
and improves visibility during the endoscopy.10 A positive diagnosis
of variceal bleeding requires endoscopic confirmation of active bleed-
ing from varices, the presence of a fibrin clot (white nipple sign)
signalling the point of variceal rupture, or blood in the stomach with
varices as the only potential source of the bleeding. EVL is indicated
in endoscopically demonstrated variceal bleeding and should be per-
formed immediately (at the diagnostic endoscopy) by an experienced
endoscopist. If the bleeding originates from GOV2 or IGV variceal
obliteration is preferred with injection of tissue adhesives.
Recommended treatment with combined vasoactive drugs and en-
doscopic therapy (plus prophylactic antibiotics) results in successful 5-
day control of bleeding in 85%-90% of cases. After this period, NSBB
treatment can be begun as secondary prophylaxis (see section 5.4).
5.3.2.3|Preemptive TIPS
The above treatment strategy does not stratify therapy for the known
risk factors. These are the presence of advanced liver failure (Child-
Pugh C), active bleeding at endoscopy despite vasoactive drug infu-
sion and very high portal pressure (HVPG20mm Hg). A multicentre
RCT recently has shown that making an early decision to perform a
PTFE-covered TIPS (during the first 24-48hours of admission), be-
fore declaring treatment failure, resulted in decreased failure to con-
trol bleeding, rebleeding and mortality in high-risk patients.47 Patients
were either Child-Pugh class C (<14 points) or Child-Pugh class B and
with active bleeding on endoscopy, with no contraindications to TIPS.
Further studies have confirmed the findings of this important report,
but showed that survival is especially improved in class C patients.48
This approach is now recommended for approximately 20% of pa-
tients who fulfil these criteria and are treated in centres with enough
experience with these procedures.
5.3.2.4|Treatment of patients that fail
standard therapy (Rescue Therapy)
The most demanding situation is the treatment of patients in whom
standard therapy with vasoactive drugs and endoscopic treatment
BRUNNER etal.
fails to control bleeding or the bleeding begins again within 5days.
The recommended therapy in these cases is to perform an emergency
TIPS, if there is no contraindication. In case of non-severe rebleed-
ing within 5 days, treatment with vasoactive drugs is continued and a
second attempt with endoscopic therapy can be made before a rescue
TIPS is considered. Rescue TIPS is very effective in controlling bleed-
ing, but it is still associated with a high mortality of about 40%, prob-
ably because liver function markedly deteriorates during uncontrolled
bleeding until the TIPS is finally performed. BRTO may be considered
in patients with refractory bleeding from gastric varices and a con-
traindication to TIPS.
5.3.2.5|Oesophageal Tamponade
Balloon tamponade with a Sengstaken-Blakemore or a Linton-Nachlas
tube may still be used in the very rare case of massive exsanguinating
bleeding or recurrent bleeding as a bridge to TIPS or another defini-
tive treatment. However, oesophageal self-expandable metallic stents
have been shown to be very effective49 and a recent RCT showed that
oesophageal stents are as effective but safer than balloon tampon-
ade.50 This is part of current recommendations.10
5.4| PATIENTS WHO SURVIVE AN
EPISODE OF VARICEAL BLEEDING: BETA-
BLOCKERS, LIGATION AND TIPS
This clinical setting includes the so-called secondary prophylaxis of
variceal haemorrhage, but as mentioned in the paragraph on primary
prophylaxis, the concept of improving outcomes in variceal bleeding is
evolving since the goal of therapy in these cases is to reduce the risk
of death, and thus prevent the onset of complications of cirrhosis that
can lead to death. They include re-bleeding, but should not be limited
to this, and all of them should be taken into account as endpoints in
future RCTs in this setting.
The goal of secondary prophylaxis of variceal bleeding is to pre-
vent recurrent variceal haemorrhage once the first episode of bleeding
has been under control for at least 5days, since the risk of rebleeding
is high (60% within first year) and associated with high mortality (up
to 33%).41 The first-line treatment includes standard NSBB lifelong
plus EVL until complete eradication of varices. As shown in a recent
meta-analysis, this combination is better than EVL alone to prevent
bleeding, but only slightly better than pharmacological therapy alone
and has no effect on survival.51 Thus, NSBBs (combined with nitrates if
tolerated) are the cornerstone of treatment. The most favourable out-
come is observed in patients with a good hemodynamic response to
NSBB, namely those in whom the HVPG is decreased by at least 20%
or below 12mm Hg on treatment.52 The HVPG response should be
measured regularly for better risk stratification and to guide therapy in
experienced centres.
Because EVL alone does not seem to provide sufficient protection
from bleeding, TIPS should be considered in patients who do not tol-
erate NSBB rather than EVL alone, in particular if other complications
|
113
such as ascites or portal vein thrombosis have already occurred. In
secondary prophylaxis, the use of carvedilol is not recommended, be-
cause data from RCT comparing carvedilol with NSBB in combination
with EVL are lacking. Statins provided additional survival benefit in
patients who survived an episode of variceal bleeding included in a
recent multicentre RCT, and should be considered in patients without
severe liver failure (benefit in Child C patients is unclear).30
PTFE-covered TIPS is better than standard secondary prophylaxis
(NSBB + EVL) in reducing the risk of rebleeding and is indicated in
patients with recurrent bleeding on standard medical and endoscopic
therapy. However, TIPS does not improve survival and is associated
with the development of HE in one third of patients.53,54
5.4.1|Secondary prophylaxis in special situations
5.4.1.1|Fundal and ectopic varices
Transjugular intrahepatic portosystemic shunt is indicated for the pre-
vention of rebleeding from fundal varices (GOV2 and IGV1). In expert
centres, cyanoacrylate glue injection can be used as an alternative.
In very selected cases (large gastro-or splenorenal collaterals), BRTO
could be an option. Ectopic varices (IGV2), which are more frequent in
non-cirrhotic portal hypertension, require a case-by-case assessment
preceded by cross-sectional imaging to map collateral vessels.
5.4.1.2|Patients with complicated ascites
Recent data have warned against the use of NSBB in decompensated
patients with refractory ascites, HRS and/or SBP. Although data are
not conclusive, these studies do not support the decision to stop
NSBB in all patients with complicated ascites. Nevertheless, the rec-
ommendations of the Baveno VI consensus conference suggest that
blood pressure, sodium levels and kidney function should be carefully
monitored and NSBB should be temporarily discontinued or reduced
in case of severe arterial hypotension, hyponatremia or acute kidney
injury.10
6|PERSONAL CONCLUSION
The treatment of portal hypertension has markedly improved in the
past years, and we now have successful therapies that have increased
patient survival in each clinical scenario. This has been possible mainly
by following a pathophysiological approach aimed at correcting the
basic disturbance, the increased portal pressure, by improving the
abnormalities causing it. There has been a continued systematic ef-
fort in converting each advance in knowledge into a new therapeutic
target, in what should be considered an example of highly successful
translational research. This has been done in parallel with careful clini-
cal observations and cooperative efforts that have allowed better risk
stratification and the definition and standardization of treatment end-
points, exemplified by the Baveno Consensus Conferences. We hope
that merging translational and clinical research efforts in the recently
 |
114
launched Baveno Cooperation will continue to improve patient out-
comes in the years to come.
CO NFLI CTS OF I NTE RE ST
None pertinent to this article.
REFERENCES
1. Bosch J, Abraldes JG, Berzigotti A, et al. The clinical use of HVPG
measurements in chronic liver disease. Nat Rev Gastroenterol Hepatol.
2009;6:573582.
2. Ripoll C, Groszmann R, Garcia-Tsao G, etal. Hepatic venous pressure
gradient predicts clinical decompensation in patients with compen-
sated cirrhosis. Gastroenterology. 2007;133:481488.
3. Kovalak M, Lake J, Mattek N, et al. Endoscopic screening for vari-
ces in cirrhotic patients: data from a national endoscopic database.
Gastrointest Endosc. 2007;65:8288.
4. Merli M, Nicolini G, Angeloni S, et al. Incidence and natural his-
tory of small esophageal varices in cirrhotic patients. J Hepatol.
2003;38:266272.
5. Groszmann RJ, Garcia-Tsao G, Bosch J, et al. Beta-blockers to pre-
vent gastroesophageal varices in patients with cirrhosis. N Engl J Med.
2005;353:22542261.
6. The North Italian Endoscopic Club for the Study, Treatment of
Esophageal Varices. Prediction of the first variceal hemorrhage in pa-
tients with cirrhosis of the liver and esophageal varices. A prospective
multicenter study. N Engl J Med. 1988;319:983989.
7. Garcia-Tsao G, Bosch J. Management of varices and variceal hemor-
rhage in cirrhosis. N Engl J Med. 2010;362:823832.
8. Carbonell N, Pauwels A, Serfaty L, etal. Improved survival after var-
iceal bleeding in patients with cirrhosis over the past two decades.
Hepatology. 2004;40:652659.
9. European Association for the Study of the Liver, Asociacion
Latinoamericana para el Estudio del Higado. EASL-ALEH clinical prac-
tice guidelines: non-invasive tests for evaluation of liver disease se-
verity and prognosis. J Hepatol. 2015;63:237264.
10. de Franchis R. Expanding consensus in portal hypertension. J Hepatol.
2015;63:743752.
11. Berzigotti A, Seijo S, Arena U, et al. Elastography, spleen size, and
platelet count identify portal hypertension in patients with compen-
sated cirrhosis. Gastroenterology. 2013;144:102111.
12. DAmico G, Garcia-Tsao G, Pagliaro L. Natural history and prognostic
indicators of survival in cirrhosis: a systematic review of 118 studies.
J Hepatol. 2006;44:217231.
13. DAmico G, Pasta L, Morabito A, etal. Competing risks and prognostic
stages of cirrhosis: a 25-year inception cohort study of 494 patients.
Aliment Pharmacol Ther. 2014;39:11801193.
14. Bosch J, Groszmann RJ, Shah VH. Evolution in the understanding of
the pathophysiological basis of portal hypertension: how changes
in paradigm are leading to successful new treatments. J Hepatol.
2015;62(S1):S121S130.
15. Cals P, Oberti F, Payen JL, etal. Lack of effect of propranolol in the
prevention of large oesophageal varices in patients with cirrhosis. Eur
J Gastroenterol Hepatol. 1999;11:741745.
16. Lebrec D, Poynard T, Hillon P, et al. Propranolol for prevention of
recurrent gastrointestinal bleeding in patients with cirrhosis. N Engl
JMed. 1981;305:13711374.
17. Poynard T, Cals P, Pasta L, etal. Beta-adrenergic-antagonist drugs in
the prevention of gastrointestinal bleeding in patients with cirrhosis
and esophageal varices. N Engl J Med. 1991;324:15321538.
18. Senzolo M, Cholongitas E, Burra P, et al. Beta-Blockers protect
against spontaneous bacterial peritonitis in cirrhotic patients: a meta-
analysis. Liver Int. 2009;29:11891193.
BRUNNER etal.
19. Garcia-Pagan JC, Navasa M, Bosch J, et al. Enhancement of portal
pressure reduction by the association of isosorbide-5-mononitrate
to propranolol administration in patients with cirrhosis. Hepatology.
1990;11:230238.
20. Bureau C, Pron J-M, Alric JL, etal. A La Carte treatment of portal
hypertension: adapting medical therapy to hemodynamic response
for the prevention of bleeding. Hepatology. 2002;36:13611366.
21. Garca-Pagn JC. Propranolol plus placebo versus propranolol plus
isosorbide-5-mononitrate in the prevention of a first variceal bleed:
a double-blind RCT. Hepatology. 2003;37:12601266.
22. Sinagra E, Perricone G, DAmico M, etal. Systematic review with meta-
analysis: the haemodynamic effects of carvedilol compared with pro-
pranolol for portal hypertension in cirrhosis. Aliment Pharmacol Ther.
2014;39:557568.
23. Reiberger T, Ulbrich G, Ferlitsch A, etal. Carvedilol for primary pro-
phylaxis of variceal bleeding in cirrhotic patients with haemodynamic
non-response to propranolol. Gut. 2013;62:16341641.
24. Tripathi D, Ferguson JW, Kochar N, etal. Randomized controlled trial
of carvedilol versus variceal band ligation for the prevention of the
first variceal bleed. Hepatology. 2009;50:825833.
25. Berzigotti A, Bosch J. Pharmacologic management of portal hyperten-
sion. Clin Liver Dis. 2014;18:303317.
26. Wiest R, Tsai MH, Groszmann RJ. Octreotide potentiates PKC-
dependent vasoconstrictors in portal-hypertensive and control rats.
Gastroenterology. 2001;120:975983.
27. Cirera I, Feu F, Luca A, et al. Effects of bolus injections and contin-
uous infusions of somatostatin and placebo in patients with cir-
rhosis: a double-blind hemodynamic investigation. Hepatology.
1995;22:106111.
28. Abraldes JG, Bosch J. Somatostatin and analogues in portal hyperten-
sion. Hepatology. 2002;35:13051312.
29. Abraldes JG, Albillos A, Baares R, et al. Simvastatin lowers portal
pressure in patients with cirrhosis and portal hypertension: a random-
ized controlled trial. Gastroenterology. 2009;136:16511658.
30. Abraldes JG, Villanueva C, Aracil C, etal. Addition of simvastatin to
standard therapy for the prevention of variceal rebleeding does not
reduce rebleeding but increases survival in patients with cirrhosis.
Gastroenterology. 2016;150:11601170.e3.
31. Mohanty A, Tate JP, Garcia-Tsao G. Statins are associated with a de-
creased risk of decompensation and death in veterans with hepatitis C-
related compensated cirrhosis. Gastroenterology 2016;150:430440.e1.
32. Tandon P, Abraldes JG, Berzigotti A, et al. Renin-angiotensin-
aldosterone inhibitors in the reduction of portal pressure: a system-
atic review and meta-analysis. J Hepatol. 2010;53:273282.
33. Markham A, Keam SJ. Obeticholic acid: first global approval. Drugs.
2016;76:12211226.
34. Meissner EG, McLaughlin M, Matthews L, et al. Simtuzumab treat-
ment of advanced liver fibrosis in HIV and HCV-infected adults:
results of a 6-month open-label safety trial. Liver Int. 2016 May 27.
doi: 10.1111/liv.13177. [Epub ahead of print]
35. Garcia-Tsao G, Fuchs M, Shifman ML, etal. Emricasan (IDN-6556) ad-
ministered orally for 28days lowers portal pressure in patients with
compensated cirrhosis and severe portal hypertension. AASLD 2015.
Hepatology. 2015;62(6 Suppl):13791399.
36. Dai C, Liu WX, Jiang M, etal. Endoscopic variceal ligation compared
with endoscopic injection sclerotherapy for treatment of esopha-
geal variceal hemorrhage: a meta- analysis. World J Gastroenterol.
2015;21:25342541.
37. Smith MR, Tidswell R, Tripathi D. Outcomes of endoscopic human
thrombin injection in the management of gastric varices. Eur J
Gastroenterol Hepatol. 2014;26:846852.
38. Bureau C, Garcia-Pagan JC, Layrargues GP, et al. Patency of stents
covered with polytetrafluoroethylene in patients treated by trans-
jugular intrahepatic portosystemic shunts: long-term results of a ran-
domized multicentre study. Liver Int. 2007;27:742747.
BRUNNER etal.
39. Riggio O, Angeloni S, Salvatori FM, et al. Incidence, natural history,
and risk factors of hepatic encephalopathy after transjugular intra-
hepatic portosystemic shunt with polytetrafluoroethylene-covered
stent grafts. Am J Gastroenterol. 2008;103:27382746.
40. Park JK, Saab S, Kee ST, etal. Balloon-occluded retrograde transve-
nous obliteration (BRTO) for treatment of gastric varices: review and
meta-analysis. Dig Dis Sci. 2015;60:15431553.
41. Garcia-Tsao G, Bosch J. Varices and variceal hemorrhage in cir-
rhosis: a new view of an old problem. Clin Gastroenterol Hepatol.
2015;13:21092117.
42. Tsochatzis EA, Bosch J, Burroughs AK. Future treatments of cirrhosis.
Expert Rev Gastroenterol Hepatol. 2014;8:571581.
43. Villanueva C, Colomo A, Bosch A, et al. Transfusion strategies for
acute upper gastrointestinal bleeding. N Engl J Med. 2013;368:
1121.
44. Bernard B, Grang JD, Khac EN, et al. Antibiotic prophylaxis for
the prevention of bacterial infections in cirrhotic patients with
gastrointestinal bleeding: a meta-analysis. Hepatology. 1999;29:
16551661.
45. DAmico G, Pagliaro L, Bosch J. Pharmacological treatment of por-
tal hypertension: an evidence- based approach. Semin Liver Dis.
1999;19:475505.
46. Seo YS, Park SY, Kim MY, etal. Lack of difference among terlipressin,
somatostatin, and octreotide in the control of acute gastroesophageal
variceal hemorrhage. Hepatology. 2014;60:954963.
47. Garca-Pagn JC, Caca K, Bureau C, etal. Early use of tips in patients
with cirrhosis and variceal bleeding for the early TIPS (Transjugular
|
115
Intrahepatic Portosystemic Shunt) Cooperative Study Group. N Engl J
Med. 2010;362:23702379.
48. Garcia-Pagn JC, Di Pascoli M, Caca K, et al. Use of early-TIPS for
high-risk variceal bleeding: results of a post-RCT surveillance study.
JHepatol. 2013;58:4550.
49. McCarty TR, Njei B. Self-expanding metal stents for acute refractory
esophageal variceal bleeding: a systematic review and meta-analysis.
Dig Endosc 2016;28(5):53947.
50. Escorsell , Pavel O, Cardenas A, etal. Esophageal balloon tampon-
ade versus esophageal stent in controlling acute refractory variceal
bleeding: a multicenter randomized, controlled trial. Hepatology.
2016;63:19571967.
51. Puente A, Hernandez-Gea V, Graupera I, etal. Drugs plus ligation to
prevent rebleeding in cirrhosis: an updated systematic review. Liver
Int. 2014;34:823833.
52. DAmico G, Garcia-Pagan JC, Luca A, etal. Hepatic vein pressure gra-
dient reduction and prevention of variceal bleeding in cirrhosis: a sys-
tematic review. Gastroenterology. 2006;131:16111624.
53. Holster IL, Tjwa ETTL, Moelker A, etal. Covered transjugular intrahe-
patic portosystemic shunt versus endoscopic therapy+beta-blocker
for prevention of variceal rebleeding. Hepatology. 2016;63:581589.
54. Sauerbruch T, Mengel M, Dollinger M, etal. Prevention of rebleeding
from esophageal varices in patients with cirrhosis receiving small-
diameter stents versus hemodynamically controlled medical therapy.
Gastroenterology. 2015;149:6608.e1.
55. Sarin SK. Long-term follow-up of gastric variceal sclerotherapy: an
eleven-year experience. Gastrointest Endosc. 1997;46:814.