EASL Clinical Practice Guidelines for the management of patients with [616783]

EASL Clinical Practice Guidelines for the management of patients with
decompensated cirrhosisq
European Association for the Study of the Liver⇑
Summary
The natural history of cirrhosis is characterised by an asymp-
tomatic compensated phase followed by a decompensated
phase, marked by the development of overt clinical signs, themost frequent of which are ascites, bleeding, encephalopathy,
and jaundice. The following Clinical Practice Guidelines (CPGs)
represent the first CPGs on the management of decompensatedcirrhosis. In this context, the panel of experts, having empha-
sised the importance of initiating aetiologic treatment for any
degree of hepatic disease at the earliest possible stage, extendedits work to all the complications of cirrhosis, which had notbeen covered by the European Association for the Study of the
Liver guidelines, namely: ascites, refractory ascites, hypona-
tremia, gastrointestinal bleeding, bacterial infections, acute kid-ney injury, hepatorenal syndrome, acute-on-chronic liver
failure, relative adrenal failure, cirrhotic cardiomyopathy, hep-
atopulmonary syndrome, and porto-pulmonary hypertension.The panel of experts, produced these GPGs using evidence from
PubMed and Cochrane database searches providing up to date
guidance on the management of decompensated cirrhosis withthe only purpose of improving clinical practice.
/C2112018 European Association for the Study of the Liver. Published by
Elsevier B.V. All rights reserved.
Introduction
When the panel of experts nominated by the European Associ-
ation for the Study of the Liver (EASL) governing board began
work to update the Clinical Practice Guidelines (CPGs) onascites, spontaneous bacterial peritonitis (SBP), and hepatorenal
syndrome (HRS),
1it became obvious that all other complica-
tions of decompensated cirrhosis had to be covered. Within thisframework, a formal definition of decompensated cirrhosis wassought. The natural history of cirrhosis is characterised by a
silent, asymptomatic course until increasing portal pressure
and worsening liver function produce a clinical phenotype. Inthe asymptomatic phase of the disease, usually referred to as
compensated cirrhosis, patients may have a good quality of life,
and the disease may progress undetected for several years.Decompensation is marked by the development of overt clinicalsigns, the most frequent of which are ascites, bleeding,
encephalopathy, and jaundice. Following the first appearance
of any of these, the disease usually progresses more rapidly
towards death or liver transplantation (LT). This phase of thedisease has been designated ‘‘decompensated cirrhosis”.
2
Progression of the decompensated disease may be further accel-erated by the development of other complications such asrebleeding, acute kidney injury (AKI), with or without thefeatures of HRS, hepato-pulmonary syndrome (HPS), portopul-
monary hypertension (PPHT), cirrhotic cardiomyopathy (CCM),
and bacterial infections. Indeed, the development of bacterialinfections as well as hepatocellular carcinoma may accelerate
the course of the disease at any stage, but especially in decom-
pensated cirrhosis.
3Having defined the potential field of action,
and having emphasised the importance of initiating aetiologic
treatment for any degree of hepatic disease at the earliest pos-
sible stage, the panel decided to extend the work to all thosecomplications of cirrhosis which have not yet been covered by
EASL guidelines, namely: gastrointestinal (GI) bleeding, bacte-
rial infections other than SBP, acute-on-chronic liver failure(ACLF), adrenal failure, HPS, PPHT and CCM. In doing so, we havehad to deal with the recommendations regularly proposed by
very well recognised international expert groups who have
worked in the field of GI bleeding or ascites and ascites-relatedcomplications for many years. Given their extreme importance
in clinical practice, only specific aspects of their recommenda-
tions were further developed in an attempt to give a more inte-grated view of the pathophysiology and management of patients
with decompensated cirrhosis. Thus, this document can no longer
be considered an update of earlier guidelines, but rather the firstCPG on the management of decompensated cirrhosis with thesole purpose of improving clinical practice.
Guidelines development process
A panel of hepatologists with a great interest in decompen-
sated cirrhosis, approved by the EASL Governing Board, wroteand discussed this CPG between March 2017 and February
2018. The guidelines were independently peer reviewed, and
all contributors to the CPG disclosed their conflicts of interestby means of a disclosure form provided by the EASL Officeprior to work commencing. The EASL Ethics Committee
reviewed the composition of the panel to eliminate the
potential for real or perceived bias. The CPG panel conflict ofinterests are declared in this submission. These guidelines have
been produced using evidence from PubMed and Cochrane
database searches before 27 March 2018. Tables describing
Journal of Hepatology 2018 vol. xxx jxxx–xxxqClinical Practice Guideline Panel: Paolo Angeli (Chair), Mauro Bernardi
(Governing Board representative), CÁndid Villanueva, Claire Francoz, Rajeshwar
P. Mookerjee, Jonel Trebicka, Aleksander Krag, Wim Laleman, Pere Gines
⇑Corresponding author. Address: European Association for the Study of the Liver
(EASL), The EASL Building – Home of Hepatology, 7 rue Daubin, CH 1203 Geneva,
Switzerland. Tel.: +41 (0) 22 807 03 60; fax: +41 (0) 22 328 07 24.
E-mail address: easloffice@easloffice.eu .JOURNAL
OF HEPATOLOGYClinical Practice Guidelines
Please cite this article in press as: The European Association for the Study of the Liver. EASL Clinical Practice Guidelines for the management of pati ents with decompensated cirrhosis. J
Hepatol (2018), https://doi.org/10.1016/j.jhep.2018.03.024

the rationale behind the levels of evidence and of recommen-
dations are provided ( Table 1 ).
Pathophysiology of decompensated cirrhosis
The transition from compensated asymptomatic cirrhosis to
decompensated cirrhosis occurs at a rate of about 5% to 7% per
year.4Once decompensation has occurred, cirrhosis becomes a
systemic disease, with multi-organ/system dysfunction.5At this
stage, patients become highly susceptible to bacterial infections
because of complex cirrhosis-associated immune dysfunction,
which involves both innate and acquired immunity.6In turn,
patients with bacterial infections are burdened by severe mor-bidity, up to ACLF, and high mortality.
6,7Because of these events,
decompensation represents a prognostic watershed, as the med-
ian survival drops from more than 12 years for compensated cir-rhosis to about two years for decompensated cirrhosis.
4For
decades the clinical manifestations of decompensated cirrhosis
have been seen as the consequence of a haemodynamic distur-bance, the hyperdynamic circulatory syndrome, ascribable to
peripheral arterial vasodilation that mainly occurs in the
splanchnic circulatory area. The extent of such vasodilation isto endanger effective volaemia, ultimately leading to peripheral
organ hypoperfusion, the kidney being most affected.
8Indeed,
reduced effective volaemia brings about the activation of vaso-constrictor and water and sodium retaining mechanisms, suchas the renin-angiotensin-aldosterone (RAAS), sympathetic ner-
vous system and arginine-vasopressin secretion. This explains
some of the cardinal features of decompensated cirrhosis, suchas renal retention of sodium and water leading to ascites forma-
tion and HRS. Other manifestations attributable to haemody-
namic abnormalities include HPS, increased susceptibility toshock, and a reduced cardiovascular responsiveness to physio-
logical and pharmacological vasoconstrictor stimuli. Subsequent
studies have highlighted that a cardiac dysfunction, due to CCM,
9
is also involved in the pathogenesis of effective hypovolaemia.10
This occurs particularly in the most advanced stages of decom-
pensation, when such an abnormality prevents cardiac output
from increasing enough to comply with the needs of systemiccirculation. Although the molecular mechanisms responsible
for arterial vasodilation, consisting of an enhanced endothelial
production of vasodilating substances, such as nitric oxide, car-bon monoxide, prostacyclin and endocannabinoids have been
convincingly demonstrated,
11the primary causes of such abnor-
malities remained somewhat obscure until it became clear thatpatients with advanced cirrhosis present a state of chronic
inflammation, as witnessed by increased circulating levels of
pro-inflammatory cytokines and chemokines.
12This is likely
caused by the systemic spread of bacteria and bacterial products,called pathogen associated molecular patterns (PAMPs), as aresult of an abnormal bacterial translocation (BT). Changes in
the microbiome and increased intestinal permeability accountfor this phenomenon. A similar role is likely played by othermolecules, called danger associated molecular patterns
(DAMPs), released by the diseased liver because of local inflam-
mation and cell apoptosis and necrosis. Both PAMPs and DAMPsbind with innate recognition receptors of immune cells that,
once activated, produce and release pro-inflammatory mole-
cules, along with reactive oxygen and nitrogen species. This cas-cade of events contributes to the development of circulatory
dysfunction and, along with it, directly favours the development
of multi-organ dysfunction and failure ( Fig. 1 ).
5Current strate-
gies for prophylaxis and treatment of decompensation and organ
failure in cirrhosis rely on measures aimed to prevent or improve
the outcome of each complication, that is renal sodium retentionleading to ascites formation, ammonia production in hepaticencephalopathy, effective hypovolaemia after large-volume
paracentesis (LVP) or during HRS, renal dysfunction induced by
SBP, and intestinal dysbiosis or bacterial overgrowth in patientspredisposed to develop infections. All these strategies will be
discussed in these CPGs. However, the improved knowledge of
the pathophysiological background of decompensated cirrhosisnow offers the opportunity for more comprehensive therapeutic
and prophylactic approaches to disease management. Indeed,
besides treating the underlying aetiologic factor(s), wheneverpossible, mechanistic approaches to counteract key pathophysi-
ologic mechanisms may prevent or delay disease progression
and the incidence of complications and multi-organ dysfunction,thus improving patient survival and quality of life, as well asreducing the economic burden of the disease.
Management of decompensated cirrhosis
Ideally, the strategy of management of patients with decompen-
sated cirrhosis should be based on preventing cirrhosis progres-sion ( i.e. further decompensation) rather than treating
complications as they occur. The ultimate treatment for decom-
pensated cirrhosis would be one that targets primarily the
pathological alterations within the liver with the aim of restor-ing the integrity of liver architecture by suppressing inflamma-
tion, causing fibrosis regression, regularising the portal and
arterial circulation, and normalising cell number and function.Unfortunately, such a treatment does not exist at present. Sev-
eral antifibrotic or anti-inflammatory drugs have shown pro-
mise in experimental models of chronic liver diseases, but notreatment has yet been translated into clinical practice.
13Mean-
while, the overall management of decompensated cirrhosis can
be addressed using two approaches. The first approach is thesuppression of the aetiological factor(s) that has caused liverinflammation and cirrhosis development, whereas the secondTable 1 .Level of Evidence and Grade of Recommendations.
Level of evidence
I Randomised, controlled trials
II-1 Controlled trials without randomisation
II-2 Cohort and case-control analytical studies
II-3 Multiple time series, dramatic uncontrolled experiments
III Opinions of respected authorities, descriptive epidemiology
Grade of recommendations
1 Strong recommendation: Factors influencing the strength of the recommendation included the quality of the evidence, presumed
patient-important outcomes, and cost
2 Weaker recommendation: Variability in preferences and values, or more uncertainty: more likely a weak recommendation is
warranted. Recommendation is made with less certainty: higher cost or resource consumptionClinical Practice Guidelines
2 Journal of Hepatology 2018 vol. xxx jxxx–xxx
Please cite this article in press as: The European Association for the Study of the Liver. EASL Clinical Practice Guidelines for the management of pati ents with decompensated cirrhosis. J
Hepatol (2018), https://doi.org/10.1016/j.jhep.2018.03.024

approach is based on targeting key factors of pathogenesis of
cirrhosis decompensation and progression.
Effects of suppression of aetiological factor on
outcome of decompensated cirrhosis
Removal of the aetiological factor(s) causing liver injury is an
important cornerstone in the management of cirrhosis. This
approach is clearly effective in preventing decompensation
and improving outcome in patients with compensated cirrhosis.However, results in patients with decompensated cirrhosis are
less efficacious and probably depend, among other factors, on
the actual status of liver disease at the time of removing theaetiological factor of liver injury. For example, although in some
patients with decompensated alcoholic cirrhosis suppression of
alcohol consumption is associated with progressive ‘‘re-com-pensation” of cirrhosis and excellent long-term outcome, inother patients alcoholic cirrhosis progresses despite stopping
alcohol intake.
14,15Likewise, in patients with cirrhosis due to
hepatitis B virus (HBV) infection, treatment with antiviralagents is associated with improved outcome in some, but not
all patients.
16Moreover, treatment of patients with decompen-
sated cirrhosis due to hepatitis C virus infection with directantiviral agents is associated with beneficial effects in liver
function and portal hypertension and likely improves outcome,
but these effects are unfortunately not generalisable to allpatients treated.
17,18The beneficial effects of removing respon-
sible factors in other aetiologies of decompensated cirrhosis are
less clear, perhaps with the exception of autoimmune hepatitis.
Effects of targeting key pathogenic events in
prevention of cirrhosis progression
Several strategies have been evaluated to prevent disease pro-
gression in patients with decompensated cirrhosis, including i)
targeting microbiome abnormalities and BT, to improve gut-liver
axis; ii) improving the disturbed circulatory function; iii) treatingthe inflammatory state; and iv) targeting portal hypertension.
Administration of rifaximin has been shown to reduce the
risk of development of several complications of cirrhosis besideshepatic encephalopathy in retrospective studies and small caseseries.
19Nonetheless, data from prospective randomised dou-
ble-blind studies are lacking. In patients with decompensated
cirrhosis, treatment with norfloxacin reduces the risk of SBPand HRS,
20,21but its use is hampered by the possibility of
increased risk of infection by resistant bacteria. The potential
effectiveness of improving circulatory and kidney function bylong-term administration of albumin to patients with decom-
pensated cirrhosis has been explored in two recent randomised
controlled trials (RCTs), both published in abstract form, withcontradictory findings.
22,23The discrepant findings may be
related to different doses of albumin used and/or heterogeneity
in the study population. Further studies are needed to find outwhether long-term albumin administration is efficacious indecompensated cirrhosis. Interestingly, treatment with statins,
through their pleotropic effects, has been shown to reduce portal
hypertension and improve survival in patients with advancedcirrhosis.
24,25These remarkable effects require validation in
future studies. Another potential terapeutical strategy in the
prevention of decompensation may be anticoagulation. Indeed,in a small RCT, a 12-month course of enoxaparin was safe and
effective in preventing portal vein thrombosis (PVT) in patientswith cirrhosis and a Child-Pugh scores of 7–10. In addition,
enoxaparin appeared to delay the occurrence of hepatic decom-pensation and to improve survival suggesting that both PVT anddecompensation may be related to a worsening of portal hyper-
tension and the consequent progressive damage of the intestinal
mucosal barrier.
26From the same perspective, two other strate-
gies should be considered. In 2010, it was shown that pentoxi-
fylline treatment significantly reduced the risk of liver-related
complications compared to placebo in an RCT of patients withadvanced cirrhosis. The prevention of these complications,
which included bacterial infections, renal failure, and hepatic
encephalopathy was probably related to the fact that pentoxi-fylline prevents intestinal BT and the consequent development
of systemic inflammation.
27Finally, some investigations have
shown that treatment with propranolol is not only effective inreducing portal hypertension and the consequent the risk of var-iceal bleeding but also in decreasing the risk of other complica-
tions of cirrhosis related to portal hypertension, such as ascites,
HRS, SBP, and hepatic encephalopathy.
28These effects occur
specifically in patients who respond to propranolol treatment
by markedly decreasing portal pressure, emphasising the strong
relationship between pressure and complications of cirrhosis.Nevertheless, in these studies most of patients had compensated
cirrhosis. Therefore, studies should be performed in the group of
patients with decompensated cirrhosis with the objective ofassessing these beneficial effects in cirrhosis progression.
Recommendations
/C15In patients with decompensated cirrhosis, the aetiologi-
cal factor, should be removed, particularly alcohol con-
sumption and hepatitis B or C virus infection as thisstrategy is associated with decreased risk of decompen-sation and increased survival (II-2,1) .
/C15Strategies based on targeting abnormalities in gut-liver
axis by antibiotic administration ( i.e.rifaximin), improv-
ing the disturbed systemic circulatory function ( i.e.long-
term albumin administration), decreasing the inflamma-tory state ( i.e.statins), and reducing portal hypertension
(i.e.beta-blockers) have shown potential benefit to
decrease cirrhosis progression in patients with decom-pensated cirrhosis. However, further clinical research is
needed with these strategies to confirm their safety
and potential benefits as therapeutic approaches withthe aim of preventing cirrhosis progression in decom-pensated patients.
Management of specific complications of
decompensated cirrhosis
Ascites
Ascites is the most common cause of decompensation in cirrhosis,
as 5% to 10% of patients with compensated cirrhosis per year
develop this complication.29The mainstay of ascites formation
is renal sodium retention due to the activation of sodium retain-ing systems, such as the renin-angiotensin-aldosterone system
(RAAS) and sympathetic nervous system. The resulting positive
fluid balance ultimately leads to extracellular fluid volumeexpansion. Reduced effective volaemia secondary to splanchnicarterial vasodilation is a main determinant of these alterations,
8
but renalfunction abnormalitiesinduced by systemic inflammationJOURNAL
OF HEPATOLOGY
Journal of Hepatology 2018 vol. xxx jxxx–xxx 3
Please cite this article in press as: The European Association for the Study of the Liver. EASL Clinical Practice Guidelines for the management of pati ents with decompensated cirrhosis. J
Hepatol (2018), https://doi.org/10.1016/j.jhep.2018.03.024

also play a role, especially in the most advanced stages of cirrho-
sis.5Portal hypertension also contributes30by acting as a com-
partmentalising factor of the expanded extracellular fluid
volume.
The occurrence of ascites impairs patient working and social
life, often leads to hospitalisation, requires chronic treatment
and is a direct cause of further complications, such as SBP,
restrictive ventilatory dysfunction, or abdominal hernias. Theappearance of ascites heralds a poor prognosis, as the five-year
survival drops from about 80% in compensated patients to about
30% in patients with decompensated cirrhosis and ascites.
4
Uncomplicated ascites
Evaluation of patients with ascites
Cirrhosis is the main cause of ascites in the Western world, being
responsible for about 80% of cases. Malignancy, heart failure,
tuberculosis, pancreatic disease, or other rarer diseases account
for the remaining cases. Initial patient evaluation should includehistory, physical examination, abdominal ultrasound, and labo-
ratory assessment of liver and renal functions, serum and urine
electrolytes, as well as an analysis of the ascitic fluid.
Diagnosis of ascites
Ascites can be graded from 1 to 3 according to the amount of
fluid in the abdominal cavity
31(Table 2 ). The ascites that recurs
at least on three occasions within a 12-month period despite
dietary sodium restriction and adequate diuretic dosage is
defined as recidivant.32
Diagnostic paracentesis is indicated in all patients with new
onset of grade 2 or 3 ascites and in those admitted to the hospi-
tal for any complication of cirrhosis.31,32Manual or automated
neutrophil count, total protein and albumin concentration,and culture should be always assessed. A neutrophil count
above 250 cells/
ll denotes SBP.33A total protein concentration
<1.5 g/dl is generally considered a risk factor for SBP, althoughthere are conflicting data.
33,34Ascitic fluid culture requires thebedside inoculation of at least 10 ml into blood culture bottles
to enhance its sensitivity.35The calculation of serum-ascites
albumin gradient (SAAG) may be useful when the cause of
ascites is not immediately evident, as SAAG ≥1.1 g/dl indicates
that portal hypertension is involved in ascites formation withan accuracy of about 97%.
36Other tests, such as amylase, cytol-
ogy, or culture for mycobacteria should be guided by clinical
presentation. Ascitic cholesterol determination followed bycytology and carcinoembryonic antigen (CEA) determination
in samples where cholesterol concentration exceeds 45 mg/dl
appears to be a cost-effective method for the differential diag-nosis between malignancy-related and non-malignant ascites.
37
Recommendations
/C15A diagnostic paracentesis is recommended in all patients
with new onset grade 2 or 3 ascites, or in those hospi-
talised for worsening of ascites or any complication ofcirrhosis (II-2;1) .
/C15Neutrophil count and culture of ascitic fluid culture
(bedside inoculation blood culture bottles with 10 mlfluid each) should be performed to exclude bacterial
peritonitis. A neutrophil count above 250 cells/
lli s
required to diagnose SBP (II-2;1) .
/C15Ascitic total protein concentration should be performed
to identify patients at higher risk of developing SBP(II-2;1) .
/C15The SAAG should be calculated when the cause of ascites
is not immediately evident, and/or when conditionsother than cirrhosis are suspected (II-2;1) .
/C15Cytology should be performed to differentiate malig-
nancy-related from non-malignant ascites (II-2;1) .Adrenal
dysfunctionHEKidney
dysfunctionHPS++Bacterial translocation
PAMPs
Activation if innate pattern
recognition receptors
Release of pro-inflamamtory molecules
(ROS/RNS)
Splanchnic arteriolar vaso dilation
and cardiovascular dysfunctionPortal hypertensionCirrhosis
Liver injury
Damaged cells
DAMPs
Other potential
mechanisms
Fig. 1. The new theory on the development of complications and organ failure/s in patients with cirrhosis (adapted from Ref. 5).DAMP, damage-
associated molecular pattern; HE, hepatic encephalopathy; HPS, hepatopulmonary syndrome; PAMP, pathogen-associated molecular pattern; RNS, re active
nitrogen species; ROS, reactive oxygen species.Clinical Practice Guidelines
4 Journal of Hepatology 2018 vol. xxx jxxx–xxx
Please cite this article in press as: The European Association for the Study of the Liver. EASL Clinical Practice Guidelines for the management of pati ents with decompensated cirrhosis. J
Hepatol (2018), https://doi.org/10.1016/j.jhep.2018.03.024

Prognosis of patients with ascites
The development of ascites in patients with cirrhosis is associ-
ated with a poor prognosis, as their one and two-year mortalityis about 40 and 50%, respectively.
1Thus, patients with ascites
should generally be considered for referral for LT. Hypona-
traemia, low arterial pressure, glomerular filtration rate (GFR)and low renal sodium excretion are independent predictors of
mortality in cirrhosis with ascites.
38As these parameters are
not included in the Child-Pugh score, and only serum creatinine(SCr), which overestimates GFR in cirrhosis,
39is included in the
model for end-stage liver disease (MELD) score, the most com-
monly used prognostic scores can underestimate the mortality
risk in patients with ascites. Modifications of the MELD score,such as the MELD-Na and MELD-Ascites scores have only par-
tially overcome this limitation.
40Thus, patients with ascites
may not receive adequate priority in transplant lists, andimproved methods to assess prognosis in these patients are
needed. A prognostic score able to identify patients with low
MELD score (<18) at high risk of 12-month adverse outcomehas recently been proposed, but it still has limited application.
41
Recommendation
/C15Since the development of grade 2 or 3 ascites in patients
with cirrhosis is associated with reduced survival, LTshould be considered as a potential treatment option
(II-2;1) .
Management of uncomplicated ascites
Ascites is uncomplicated when it is not infected, refractory or
associated with HRS.
31,32
Grade 1 or mild ascites .No data on the evolution of grade 1
ascites are available, nor it is known whether its treatmentmodifies its natural history.
Grade 2 or moderate ascites
.Patients who develop grade 2 ascites
do not require hospitalisation, unless other complications are
present. They have a positive sodium balance, which can be cor-rected by reducing the dietary sodium intake and increasing
renal sodium excretion with diuretics. Although upright posture
favours renal sodium reabsorption
42and attenuates the
response to diuretics,43there is no evidence that a prolonged
maintenance of the supine position eases the treatment of
ascites.
Sodium restriction. The prophylactic use of salt restriction in
patients who never had ascites is not supported by evidence.
Dietary sodium restriction can lead to the resolution of ascites
in about 10% of patients,44especially in those with the first epi-
sode of ascites. A clear advantage from the use of low-sodium
diets associated with diuretics has not emerged from clinical tri-
als comparing different dietary regimens.44,45Extreme sodium
restriction favours the development of diuretic-induced
hyponatraemia and renal failure.46Moreover, even moderate
sodium restriction, when not prescribed with an adequate edu-cational programme, is often associated with reduced calorie
intake,
47and may impair nutritional status. The current opinion
is that dietary sodium should only be moderately restricted(80–120 mmol/day), mainly to avoid excess salt intake.Recommendations
/C15A moderate restriction of sodium intake (80–120 mmol/
day, corresponding to 4.6–6.9 g of salt) is recommended
in patients with moderate, uncomplicated ascites (I;1).
This is generally equivalent to a no added salt diet withavoidance of pre-prepared meals. Adequate nutritional
education of patients on how to manage dietary sodium
is also recommended (II-2;1) .
/C15Diets with a very low sodium content (<40 mmol/day)
should be avoided, as they favour diuretic-induced com-plications and can endanger a patient’s nutritional status(II-2;1) .
/C15Prolonged bed rest cannot be recommended because
there is insufficient evidence that it is beneficial in the
treatment of ascites (III;1) .
Diuretics. Neither diuretics nor LVP are associated with a survival
benefit because they act downstream of the pathophysiological
cascade, being symptomatic therapies. The negative fluid balanceinduced by diuretics should not lead to a body weight loss exceed-
ing 0.5 kg/day in patients without peripheral oedema and
1 kg/day in the presence of peripheral oedema to avoid plasmavolume contraction, ultimately leading to renal failure and
hyponatraemia.
48Since secondary hyperaldosteronism plays a
pivotal role in the renal sodium retention in patients with cirrho-sis,
49,50anti-mineralocorticoid drugs (spironolactone, canrenone
or K-canrenoate) represent a mainstay in the medical treatment
of ascites.50Four hundred mg/day represents the maximal dosage
usually recommended.31,32The mechanism of action of anti-min-
eralocorticoids explains their slow effect. In fact, the activated
aldosterone pathway, which involves interaction with a cytosolic
receptor and, then, a nuclear receptor, needs to be exhaustedbefore their natriuretic effect arises. Therefore, the dosage of
these drugs should not be increased earlier than 72 h. Amiloride,
a diuretic acting in the collecting duct, is less effective than anti-mineralocorticoids, and should only be used in patients who
develop severe side effects with aldosterone antagonists.
51
Proximal tubular sodium reabsorption promotes renal sodium
retention through various mechanisms, such as increased angio-
tensin II production, sympatho-adrenergic hyperactivity and
reduced renal perfusion.49As proximal tubular sodium reabsorp-
tion can become relatively prevalent in patients with long-stand-ing ascites,
52,53loop diuretics are indicated in this setting.
However, they should be combined with but not substituted for
anti-mineralocorticoids. Indeed, despite their potent activity,the natriuretic effect of loop diuretics can be completely blunted
by unopposed hyperaldosteronism.
54Whether diuretic treat-
ment should be initiated with anti-mineralocorticoids alone orshould also include a loop diuretic has long been debated. Two
studies have addressed this matter providing apparently conflict-
ing results because of differences in patient populations.
55,56In
both studies, the effects of a diuretic regimen initially consistingof spironolactone or K-canrenoate alone at stepwise increasing
Table 2 .Grading of ascites.
Grade 1. Mild ascites: it is only detectable by ultrasound examination
Grade 2. Moderate ascites: it is manifest by moderate symmetrical
distension of abdomen
Grade 3. Large or gross ascites: it provokes marked abdominal distensionJOURNAL
OF HEPATOLOGY
Journal of Hepatology 2018 vol. xxx jxxx–xxx 5
Please cite this article in press as: The European Association for the Study of the Liver. EASL Clinical Practice Guidelines for the management of pati ents with decompensated cirrhosis. J
Hepatol (2018), https://doi.org/10.1016/j.jhep.2018.03.024

dosages (from 100/200 to 400 mg/day), with furosemide added in
non-responder patients, were compared with those of the combi-nation of anti-mineralocorticoids with furosemide (from 40 to160 mg/day) from the beginning of treatment. In one study,
56
the response rate, the rapidity of ascites mobilisation and theincidence of diuretic-induced complications were similar inbothregimens.However,asthesequentialtreatmentrequiredless
dose adjustments, it appeared to be more suitable for treating
ascites on an outpatient basis. In the other study,
55the combined
regimen achieved the resolution of ascites in a shorter time, with
a lower incidence of side effects, mainly hyperkalemia. Such
divergent results likely arose from differences in the patientpopulations. In one study,
56patients with ascites at the first
appearance and well preserved renal function prevailed, while, in
the other,55most patients had recurrent ascites and many showed
a substantial reduction of GFR. Thus, patients with ascitesat the first appearance can confidently be treated with anti-
mineralocorticoids alone, as they will likely develop a satisfactory
response with few side effects. Patients with long-standing,recurrent ascites should receive the combination therapy, which
likely shortens the time to achieve natriuresis and lowers the
incidence of hyperkalemia.
1In a randomised double-blind cross-
over trial torasemide induced greater cumulative 24 h diuresis than
furosemide, suggesting that torasemide might be more advanta-
geous in patients exhibiting a weak response to furosemide.57
Following mobilisation of ascites, diuretics should be tapered
to the lowest dosages able to maintain patients with minimal or
no ascites, to minimise side effects. Whenever possible, an aeti-ologic treatment of the underlying cirrhosis should be insti-tuted, as this eases the control of ascites in many cases.
Complications of diuretic therapy. The haemodynamic status of
patients with cirrhosis and ascites
8makes them highly suscep-
tible to rapid reductions in extracellular fluid volume, which
mostly occur with loop diuretics. Thus, renal failure is frequent
in this setting,48as is hepatic encephalopathy, also favoured by
increased renal ammonia production. Loop diuretics can also
lead to potassium and magnesium depletion. Hyponatraemia
is another common diuretic-induced side effect in cirrhosis. Itmostly, but not exclusively, occurs with loop diuretics, as theyinhibit Na-K-Cl transporter and, therefore, solute-free water
generation. Plasma volume contraction can also enhance argi-
nine-vasopressin release. Thus, hyponatraemia can also ensuewith anti-mineralocorticoid administration, albeit infrequently.
Most experts agree on at least temporarily withdrawing diuret-
ics when serum sodium concentration decreases below 120–125 mmol/L. Hyperkalemia, especially in patients with reduced
renal perfusion, and painful gynecomastia are the most com-
mon side effects induced by anti-mineralocorticoids.
Muscle cramps can impair quality of life in patients receiving
diuretics. Albumin infusion can relieve cramps,
58as well as
baclofen (10 mg/day, with a weekly increase of 10 mg/day upto 30 mg/day), which was safely used in a recent RCT.
59One
RCT investigated the use of quinidine at the dose of 400 mg/day
for four weeks in patients with cirrhosis with painful muscle
cramps. Although more effective than placebo, quinidine wasassociated with diarrhoea in about one-third of cases requiring
treatment withdrawal.
60Because of the frequency of diuretic-
induced side effects, especially during the first month of treat-ment,
55serial measurements of SCr, sodium, and potassium
are warranted. The assessment of urine sodium excretion can
be limited to non-responders, to unveil excessive sodium intake.Recommendations
/C15Patients with the first episode of grade 2 (moderate)
ascites should receive an anti-mineralocorticoid drug
alone, starting at 100 mg/day with stepwise increases
every 72 h (in 100 mg steps) to a maximum of 400mg/day if there is no response to lower doses (I;1).
/C15In patients who do not respond to anti-mineralocorti-
coids, as defined by a body weight reduction of less than
2 kg/week, or in patients who develop hyperkalemia,
furosemide should be added at an increasing stepwisedose from 40 mg/day to a maximum of 160 mg/day (in40 mg steps) (I;1).
/C15Patients with long-standing or recurrent ascites should
be treated with a combination of an anti-mineralocorti-
coid drug and furosemide, the dose of which should be
increased sequentially according to the response, asexplained (I;1).
/C15Torasemide can be given in patients exhibiting a weak
response to furosemide (I;2).
/C15During diuretic therapy a maximum weight loss of 0.5
kg/day in patients without oedema and 1 kg/day inpatients with oedema is recommended (II-2;1) .
/C15Once ascites has largely resolved, the dose of diuretics
should be reduced to the lowest effective dose (III;1) .
/C15During the first weeks of treatment patients should
undergo frequent clinical and biochemical monitoring
particularly on first presentation (I;1).
/C15In patients presenting with GI haemorrhage, renal
impairment, hepatic encephalopathy, hyponatraemia,
or alterations in serum potassium concentration, these
abnormalities should be corrected before starting diure-tic therapy (III;1) . In these patients, cautious initiation of
diuretic therapy and frequent clinical and biochemical
assessments should be performed (III;1) . Diuretic ther-
apy is generally not recommended in patients with per-
sistent overt hepatic encephalopathy (III;1) .
/C15Diuretics should be discontinued if severe hypona-
traemia (serum sodium concentration <125 mmol/L),
AKI, worsening hepatic encephalopathy, or incapacitat-
ing muscle cramps develop (III;1) .
/C15Furosemide should be stopped if severe hypokalemia
occurs (<3 mmol/L). Anti-mineralocorticoids should bestopped if severe hyperkalemia occurs (>6 mmol/L) (III;1) .
/C15Albumin infusion or baclofen administration (10
mg/day, with a weekly increase of 10 mg/day up to 30
mg/day) are recommended in patients with muscle
cramps (I;1).
Grade 3 or large ascites
.The treatment of choice for the manage-
ment of patients with grade 3 ascites is represented by LVP.
Paracentesis should be performed under strict sterile conditionsusing disposable sterile materials. The procedure is associatedwith a very low risk of local complications, particularly bleed-
ing
61,62even in patients with international normalized ratioClinical Practice Guidelines
6 Journal of Hepatology 2018 vol. xxx jxxx–xxx
Please cite this article in press as: The European Association for the Study of the Liver. EASL Clinical Practice Guidelines for the management of pati ents with decompensated cirrhosis. J
Hepatol (2018), https://doi.org/10.1016/j.jhep.2018.03.024

(INR)>1.5 and platelet count <50,000/ ll, minor bleeding from
puncture site occurred in two out of 142 paracentesis.61Thus,
there are no data supporting the prophylactic use of fresh frozenplasma of pooled platelets, even though these are employed in
many centres when prothrombin activity is below 40% and pla-
telet count <40,000/
ll. LVP should be avoided in the presence of
disseminated intravascular coagulation. Other contraindications
to LVP are reported ( Table 3 ).
The removal of large volumes of ascitic fluid is potentially
associated with a further reduction of effective blood volume, a
condition known as post-paracentesis circulatory dysfunction
(PPCD).63The clinical manifestations of PPCD are renal failure,
dilutional hyponatraemia, hepatic encephalopathy and
decreased survival.63Plasma volume expansion should be per-
formed at the completion of LVP to prevent this complication.Artificial plasma expanders, such as dextran-70 (8 g/L of ascitesremoved)
64or polygeline (150 ml/L),64saline solution (170 ml/
L),65only show a similar efficacy to 20% albumin (8 g/L)64when
less than 5 L of ascites are removed. However, polygeline is nolonger used in many countries because of the potential risk of
transmission of prions and dextran carries the risk of severe aller-
gic reaction and renal failure. A meta-analysis of randomised tri-als showed that albumin is superior to any other plasma expander
or vasoconstrictor not only in preventing PPCD, but also its clini-
cal consequences, such as hyponatraemia and mortality.
66More-
over, albumin infusion after LVP appears to be more cost-effective
than a cheaper plasma volume expander, such as polygeline,
because of the lower number of liver-related complications andhospital costs for a 30-day period.
67LVP combined with infusion
of albumin in patients with grade 3 ascites is more effective and
safer than diuretics.68,69However, LVP does not modify the
underlying pathophysiological abnormalities leading to ascitesformation. Thus, patients treated with LVP require diuretic ther-
apy to prevent the re-accumulation of ascites.
70
Recommendations
/C15LVP is the first-line therapy in patients with large ascites
(grade 3 ascites), which should be completely removedin a single session (I;1).
/C15LVP should be followed with plasma volume expansion
to prevent PPCD (I;1).
/C15In patients undergoing LVP of greater than 5 L of ascites,
plasma volume expansion should be performed by infus-ing albumin (8 g/L of ascites removed), as it is more
effective than other plasma expanders, which are not
recommended for this setting (I;1).
/C15In patients undergoing LVP of less than 5 L of ascites, the
risk of developing PPCD is low. However, it is generallyagreed that these patients should still be treated with
albumin because of concerns about use of alternative
plasma expanders (III;1) .
/C15After LVP, patients should receive the minimum dose of
diuretics necessary to prevent re-accumulation of ascites(I;1).
/C15When needed, LVP should also be performed in patients
with AKI or SBP (III;1) .Drugs contraindicated in patients with ascites
As non-steroidal anti-inflammatory drugs inhibit renal prosta-
glandin synthesis, they should not be used in patients with cir-rhosis and ascites, where an increased vasodilatingprostaglandin synthesis counteracts the renal vasoconstrictor
effects of angiotensin II. Indeed, their administration can lead
to acute renal failure, hyponatraemia, and diuretic resistance.
71
It would appear that selective inhibitors of cyclooxygenase-2 do
not impair renal function and response to diuretics in patients
with ascites.72However, it is not known whether these drugs
can be safely used in clinical practice when analgesia is needed.
Patients with ascites are also particularly sensitive to the renal
vasoconstrictor effect of endogenous adenosine, and dipyri-damole can induce a marked reduction in renal perfusion.
73
The maintenance of an adequate arterial pressure in cirrhosis
with ascites is assured by the activation of endogenous vasocon-strictor systems. Thus, angiotensin-converting enzymeinhibitors,
74angiotensin II receptor antagonists, and a1-adren-
ergic blockers75should be avoided, as they can induce arterial
hypotension and renal function impairment. Aminoglycosidesshould be avoided in the treatment of bacterial infections,
except in specific cases (discussed later), because they are asso-
ciated with high incidence of nephrotoxicity.
76Although cirrho-
sis with ascites and preserved renal function does not appear to
be a risk factor for renal failure induced by contrast media,77
this cannot be excluded in patients with impaired renal func-
tion. In these cases, preventive measures such as plasma volumeexpansion with saline may be employed.
78
Recommendations
/C15Non-steroidal anti-inflammatory drugs should not be
used in patients with ascites because of the high risk of
developing further sodium retention, hyponatraemia,
and AKI (II-2;1) .
/C15Angiotensin-converting-enyzme inhibitors, angiotensin
II antagonists, or a1-adrenergic receptor blockers should
not generally be used in patients with ascites because of
increased risk of renal impairment (II-2;1) .
/C15The use of aminoglycosides is discouraged, as they are
associated with an increased risk of AKI. Their use should
be reserved for patients with severe bacterial infectionsthat cannot be treated with other antibiotics (II-2;1) .
/C15In patients with ascites and preserved renal function, the
use of contrast media does not appear to be associatedwith an increased risk of renal impairment (II-2) . There
are insufficient data in patients with renal failure. Never-
theless, a cautious use of contrast media and the use ofpreventive measures for renal impairment are recom-
mended (III;1) .Table 3 .Contraindications to paracentesis.
/C15Uncooperative patient
/C15Abdominal skin infection at the proposed puncture sites
/C15Pregnancy
/C15Severe coagulopathy (accelerated fibrinolysis or disseminated
intravascular coagulation)
/C15Severe bowel distensionJOURNAL
OF HEPATOLOGY
Journal of Hepatology 2018 vol. xxx jxxx–xxx 7
Please cite this article in press as: The European Association for the Study of the Liver. EASL Clinical Practice Guidelines for the management of pati ents with decompensated cirrhosis. J
Hepatol (2018), https://doi.org/10.1016/j.jhep.2018.03.024

Refractory ascites
Evaluation of patients with refractory ascites
According to the criteria of the International Ascites Club, refrac-
tory ascites is defined as ‘‘ascites that cannot be mobilised or the
early recurrence of which ( i.e., after LVP) cannot be satisfactorily
prevented by medical therapy”.31,32The diagnostic criteria of
refractory ascites are shown in Table 4 . Refractoriness of ascites
is associated with a poor prognosis, with a median survival of
about six months.79Therefore, if a patient with refractory ascites
has not yet been considered for LT, he/she should be immediately
referred to a liver transplant center. The potential underestimation
of the mortality risk by commonly used prognostic scores, as dis-cussed earlier also applies to patients with refractory ascites.
80
Recommendations
/C15The diagnosis of refractory ascites relies on the assess-
ment of the response of ascites to diuretic therapy and
salt restriction. Such an evaluation should be done in
stable patients without associated complications, suchas bleeding or infection, after ascertaining patient com-pliance to treatment (III;1) .
/C15Patients with refractory ascites should be evaluated for
LT(III;1) .
Management of refractory ascites
Large-volume paracentesis
.There is general agreement that LVP
is an effective and safe treatment of refractory ascites,31,35
which should be associated with albumin administration to pre-
vent PPCD.
Diuretics in patients with refractory ascites .Once refractoriness of
ascites has been ascertained, diuretics should be discontinued.Only when renal sodium excretion on diuretics exceeds 30
mmol/day, maintenance of diuretic therapy can be considered,
when tolerated.
31
Non-selective beta-blockers in patients with refractory ascites .The
controversial issue on the use of non-selective beta-blockers
(NSBBs) in patients with ascites and, particularly, in those withrefractory ascites will be developed in the section dedicated to
GI bleeding.
Transjugular intrahepatic portosystemic shunts .Transjugular
intrahepatic portosystemic shunts (TIPS) decompresses the portal
system by shunting an intrahepatic portal branch into a hepatic
vein. Its insertion accentuates perpheral arterial vasodilation inthe short term. However, within 4–6 weeks its result is an
improvement in effective volaemia and renal function, ultimately
leading to an increase in renal sodium excretion.
81–85TIPS-
induced natriuresis can be delayed by advanced age and reducedpre-TIPS GFR,
84and prevented by intrinsic kidney disease.86TIPS
may also exert beneficial effects on nitrogen balance and nutri-
tion87and quality of life.88A major complication after TIPS inser-
tion using bare stent grafts is the development of hepatic
encephalopathy, which can occur in up to 50% of patients.89,90
The incidence of this complication can be significantly reduced
to about 18% with the use of polytetrafluoroethylene (PTFE)-cov-
ered stent grafts of 8 mm,91a result confirmed by a recent ran-
domised trial comparing 8 mm and 10 mm stent grafts.92
Notably, this favourable effect is better than with larger stent
grafts underdilated to 8 mm. Indeed, it has been shown that
underdilated 10 mm stent grafts passively expand to almost thefull diameter within 1–6 weeks.
93It must be underlined that the
indication for TIPS insertion in these studies was the prevention
or treatment of recurrent bleeding, which may restrict the rele-
vance of these results in patients with refractory ascites. Dysfunc-tion of TIPS with bare stent grafts because of stent thrombosis and
stenosis can develop in up to 80% of cases.
89This complication has
been significantly reduced with the use of PTFE-covered stents.94
Controlled studies and meta-analysis. The clinical effects of TIPS
with bare stents in patients with refractory or recurrent asciteshave been assessed in six prospective RCTs,
95–100whose main
features are reported ( Table 5 ). Based on these RCTs, seven
meta-analyses were performed.101–107The final messages can
be summarised as follow: i) TIPS controlled ascites better thanLVP, and ii) TIPS is followed by a greater incidence of hepatic
encephalopathy. However, discrepant results were obtained
with respect to survival. A better survival with LVP, mainlybecause of a detrimental effect of TIPS in Child-Pugh class C
patients, was reported by one study,
96while no difference was
reported by two.95,100A better survival with TIPS was reported
Table 4 .Definition and diagnostic criteria for refractory ascites in cirrhosis.
Definition
Diuretic-resistant ascites Ascites that cannot be mobilized or the early recurrence of which cannot be prevented because of a lack of response to sodi um
restriction and diuretic treatment
Diuretic-intractable
ascitesAscites that cannot be mobilized or the early recurrence of which cannot be prevented because of the development of diuretic-
induced complications that preclude the use of an effective diuretic dosage
Diagnostic criteria
Treatment duration Patients must be on intensive diuretic therapy (spironolactone 400 mg/day and furosemide 160 mg/day) for at least one week and
on a salt-restricted diet of less than 90 mmol/day
Lack of response Mean weight loss of <0.8 kg over four days and urinary sodium output less than the sodium intake
Early ascites recurrence Reappearance of grade 2 or 3 ascites within four weeks of initial mobilisation
Diuretic-induced
complicationsDiuretic-induced hepatic encephalopathy is the development of encephalopathy in the absence of any other precipitating factor
Diuretic-induced renal impairment is an increase of serum creatinine by >100% to a value >2 mg/dl (177 lmol/L) in patients with
ascites responding to treatment
Diuretic-induced hyponatremia is defined as a decrease of serum sodium by >10 mmol/L to a serum sodium of <125 mmol/L
Diuretic-induced hypo- or hyperkalemia is defined as a change in serum potassium to <3 mmol/L or >6 mmol/L despite
appropriate measures
Invalidating muscle crampsClinical Practice Guidelines
8 Journal of Hepatology 2018 vol. xxx jxxx–xxx
Please cite this article in press as: The European Association for the Study of the Liver. EASL Clinical Practice Guidelines for the management of pati ents with decompensated cirrhosis. J
Hepatol (2018), https://doi.org/10.1016/j.jhep.2018.03.024

in another two studies,97,99while, in the remaining one,98
although a survival advantage was not found, TIPS was indepen-
dently associated with transplant-free survival at multivariateanalysis. In four meta-analyses including the five studies avail-
able at that time no survival advantantage with TIPS emerged.
However, a trend towards reduced mortality with TIPS wasseen
104after the exclusion of an outlier trial.96The latter was
also excluded in the only meta-analysis on individual patient
data, and an increased transplant-free survival was found.107
Finally, the two meta-analyses that included all six trials102,103
provided contrasting results, as an improved transplant-free sur-vival was found in one,
107while a survival advantage with TIPS
was limited to patients with recurrent ascites in the other.102
Fewer studies assessing the effects of TIPS with PTFE-covered
stent grafts are available. Two retrospective studies108,109
reported better control of ascites and one-year108or two-
year109survival with covered stent grafts than bare stent grafts
in patients with refractory ascites. A survival benefit of TIPS vs.
serial paracentesis in patients with refractory ascites has beenreported in a single-centre case-control propensity score analy-
sis.
110In a recent RCT comparing covered TIPS vs.LVP in
patients with recurrent ascites, a better one-year transplant-free survival was seen in patients treated with covered stents,
without any significant increase in occurrence of hepatic
encephalopathy.
111Thus, currently available data suggest that
TIPS improves survival compared to LVP in patients with recur-
rent ascites, but it does not in those with refractory ascites.
A careful selection of patients is also crucial to maximise the
beneficial effects of TIPS, as TIPS can even be detrimental inpatients with the most advanced stages of cirrhosis, such asthose belonging to Child-Pugh class C.
96The main exclusion cri-
teria for TIPS insertion in the seven RCTs are reported in Table 5 .
A score system based on SCr, INR, serum bilirubin and aetiologyof cirrhosis has been proposed to predict survival after TIPS
insertion for refractory ascites.
112Another simple predictor of
survival suggested for patients receiving TIPS for refractoryascites consists of the combination of serum bilirubin
concentration and platelet count.
113Another factor that seems
to influence mortality is the number of TIPS procedures per-formed in a centre, as the risk of inpatient mortality is lower
in hospitals performing ≥20 TIPS per year.
114
Other treatments .Based on the exclusion criteria reported
(Table 5 ), a substantial portion of patients with refractory
ascites are not candidates for TIPS insertion. Thus, the search
for alternative treatments is warranted.
Medical treatments. Therapies aimed at improving circulatory
and renal function have been proposed. The a1-adrenergic
agonist midodrine has been shown to improve systemic andrenal haemodynamics in patients with cirrhosis and uncom-
plicated ascites.
115In a small RCT comparing the addition of
midodrine (7.5 mg t.i.d) to diuretic treatment with diuretictreatment alone in patients with refractory or recurrentascites for six months, only a transient beneficial effect on
the control of ascites was seen at the third month.
116The
use of terlipressin, an analogue of vasopressin with a predom-inant vasoconstrictor effect in the splanchnic circulatory area
in patients with refractory ascites has only been assessed in
acute studies. In one,
117terlipressin administration (1 toTable 5 .Characteristics and results of six randomised controlled trials comparing bared TIPS and LVP in patients with cirrhosis and refractory or
recidivant ascites.
Refs. Refractory/
recidivant
ascites (%)Exclusion criteria Enrolled
patients
(N)Ascites
improved
(%)Encephalopathy
(%)Survival
(%)
TIPS LVP TIPS LVP TIPS LVP TIPS LVP
Lebrec et al.
1995100/0 Age >70 yr; severe extra-hepatic diseases; HCC; pulmonary
hypertension; HE, bacterial infection; severe alcoholic hepatitis; portal
or hepatic vein obstruction or thrombosis; obstruction of biliary tract;
obstruction of hepatic artery; serum creatinine >1.7 mg/dl13 12 38 0*15 6 29 60
Rössle et al.
200055/45 HE ≥grade 2; serum bilirubin >5 mg/dl, serum creatinine >3 mg/dl;
portal-vein thrombosis, hepatic hydrothorax; advanced cancer; failure
of LVP (ascites persisting after LVP or need for LVP >once per week)29 31 84 43*23 13 58 32
Ginés et al.
200290100/0 Age >18 or >75 yr; serum bilirubin >10 mg/dl; prothrombin time <40%
(INR 2.5); platelet count <than 40,000/mm3; serum creatinine >3 mg/
dl, HCC, complete portal vein thrombosis; cardiac or respiratory
failure; organic renal failure; bacterial infection; chronic HE35 35 51 17*60 34 26 30
Sanyal et al.
2003100100/0 Causes of ascites other than cirrhosis; advanced liver failure (serum
bilirubin bilirubin >5 mg/dl, PT INR >2); incurable cancers or
nonhepatic diseases that were likely to limit life expectancy to 1 yr;congestive heart failure; acute renal failure; parenchymal renal
disease; portal vein thrombosis; bacterial infections; HE ≥grade II;
florid alcoholic hepatitis, HCC; gastrointestinal hemorrhage within 6 w
of randomisation.52 57 58 16
*38 21 35 33
Salerno
et al. 20049968/32 Age > 72 yr; recurrent HE ≥grade 2; serum bilirubin >6 mg/dl; serum
creatinine >3 mg/dl; Child-Pugh score >11; complete portal vein
thrombosis; HCC; gastrointestinal bleeding within 15 d of
randomisation; serious cardiac or pulmonary dysfunctions;bacterial
infection; SAAG gradient <11 g/L.33 33 79 42*61 39 59 29*
Narahara
et al. 201197100/0 Age >70 yr, chronic HE, HCC and other malignancies, complete portal
vein thrombosis with cavernomatous transformation, bacterialinfection, severe cardiac or pulmonary disease, organic renal disease.30 30 87 30
*20 5 20 5*
HCC, hepatocellular carcinoma; HE, hepatic encephalopathy; INR, international normalized ratio; LVP, large volume paracentesis; PT, prothrombi n time; SAAG, serum-
ascites albumin gradient; TIPS, transjugular intrahepatic portosystemic shunt.
*Significantly lower than TIPS.JOURNAL
OF HEPATOLOGY
Journal of Hepatology 2018 vol. xxx jxxx–xxx 9
Please cite this article in press as: The European Association for the Study of the Liver. EASL Clinical Practice Guidelines for the management of pati ents with decompensated cirrhosis. J
Hepatol (2018), https://doi.org/10.1016/j.jhep.2018.03.024

2 mg intravenous [i.v.], according to body weight) only
increased renal sodium excretion when associated withexogenous atrial natriuretic factor. In another,
1182 mg of ter-
lipressin led to an increase in GFR, renal plasma flow and
renal sodium excretion. However, in this study only eight
patients with refractory ascites were included. Whether aprolonged treatment with terlipressin may lead to a clinically
relevant improvement of renal function and sodium excretion
in refractory ascites is not known.
The
a2-adrenoceptor agonist clonidine, a sympatholytic
drug, which suppresses RAAS activity and improves the
response to diuretics in patients with cirrhosis and asciteswas tested in a large prospective RCT. It was shown that cloni-
dine administration on top of diuretics for three months led to
an overall response to diuretics in 60% of cases, while noresponse was seen with diuretics alone. This effect was asso-ciated with significant reductions of RAAS and sympathetic
nervous system activity. Interestingly, the favourable effects
of clonidine were predicted by the variant genotype of G pro-tein (GNB3 C825T) and adrenergic receptor (ADRA2C Del 322–
325) polymorphisms, and the baseline norepinephrine level.
119
Small scale or pilot studies evaluated the effects of various
combinations of midodrine with either clonidine,120the antag-
onist of vasopressin V2-receptors tolvaptan,121or octreotide
and albumin122in patients with refractory and recurrent
ascites. Some promising results were obtained, but they need
to be confirmed by sufficiently powered RCTs. A recent RCT123
compared the effects of the combined administration of mido-
drine (5 mg t.i.d) and rifaximin (550 mg b.i.d) on top of diuret-ics for 12 weeks with diuretics alone. After 12 weeks, 80% of
patients in the active arm were complete responders with a
significant improvement in survival in the midodrine/rifaximinarm. Due to weakness in the study design, these results are
not definitive, but they certainly warrant further investigation.
Alfapump
/C210.The automated low-flow ascites pump (Alfapump/C210)
system consists of a subcutaneously implanted battery-powered
programmable pump. It is connected to catheters that transferascites from the peritoneal cavity to the bladder, from which itis eliminated with urine. The device has internal sensors that
monitor pump function. In two multicentre safety and efficacy
studies,
124,125Alfapump/C210ensured a significant reduction of the
number and volume of paracentesis in patients with advanced
cirrhosis and refractory ascites. However, adverse effects
directly related to the device occurred in about one-third124to
half125of cases. In a multicentre RCT in patients with refractory
ascites, Alfapump/C210reduced the median number of paracentesis
per month by 85% with respect to LVP, and significantlyimproved quality of life and nutritional parameters, as assessed
by hand-grip strength and body mass index. Alfapump
/C210had no
effect on survival and was associated with a significantly higherincidence of serious adverse events (85.2 vs.45.2%), mainly rep-
resented by AKI.
126Thus, even though Alfapump/C210is effective in
reducing the need for paracentesis in patients with refractory
ascites, its frequent side effects require close monitoring ofpatients. Indeed, in addition to device-related adverse event, it
should be noted that the evaluation of kidney and circulatory
function in 10 patients with cirrhosis and refractory ascitescarrying Alfapump
/C210has shown a significant GFR decline within
six months, which was associated with a marked increase inplasma renin activity and norepinephrine concentration.
127This
likely represented the pathophysiological background of 18
episodes of AKI experienced by seven patients.
Recommendations
/C15Repeated LVP plus albumin (8 g/L of ascites removed)
are recommended as first line treatment for refractory
ascites (I;1).
/C15Diuretics should be discontinued in patients with refrac-
tory ascites who do not excrete >30 mmol/day of sodiumunder diuretic treatment (III;1) .
/C15Although controversial data exist on the use of NSBBs in
refractory ascites, caution should be exercised in cases of
severe or refractory ascites. High doses of NSBB should
be avoided ( i.e.propranolol >80 mg/day) (II-2;1) . The
use of carvedilol can not be recommended at present(I;2).
/C15Patients with refractory or recurrent ascites (I;1),o r
those for whom paracentesis is ineffective ( e.g.due to
the presence of loculated ascites) should be evaluated
for TIPS insertion (III;1) .
/C15TIPS insertion is recommended in patients with recur-
rent ascites (I;1) as it improves survival (I;1) and in
patients with refractory ascites as it improve the controlof ascites (I;1).
/C15The use of small-diameter PTFE-covered stents in
patients is recommended to reduce the risk of TIPS dys-
function and hepatic encephalopathy with a high risk of
hepatic encephalopathy is recommended (I;1).
/C15Diuretics and salt restriction should be continued after
TIPS insertion up to the resolution of ascites (II-2;1) ,a s
well as close clinical follow-up (III,1) .
/C15Careful selection of patients for elective TIPS insertion is
crucial, as is the experience of the centre performing thisprocedure. TIPS is not recommended in patients with
serum bilirubin > 3 mg/dl and a platelet count lower
than 75 x 10
9/L, current hepatic encephalopathy grade
≥2 or chronic hepatic encephalopathy, concomitant
active infection, progressive renal failure, severe systolic
or diastolic dysfunction, or pulmonary hypertension(III;1) .
/C15At present the addition of clonidine or midodrine to
diuretic treatment cannot be recommended (III;1) .
/C15Alfapump
/C210implantation in patients with refractory
ascites not amenable to TIPS insertion is suggested inexperienced centres. However, close patient monitoring
is warranted because of the high risk of adverse events
including renal dysfunction and technical difficulties (I;2).Clinical Practice Guidelines
10 Journal of Hepatology 2018 vol. xxx jxxx–xxx
Please cite this article in press as: The European Association for the Study of the Liver. EASL Clinical Practice Guidelines for the management of pati ents with decompensated cirrhosis. J
Hepatol (2018), https://doi.org/10.1016/j.jhep.2018.03.024

Hepatic hydrotorax
Hepatic hydrotorax describes the accumulation of transudate in
the pleural space of patients with decompensated cirrhosis inthe absence of cardiac, pulmonary or pleural disease. Its forma-
tion is secondary to small diaphragmatic defects, more often
located in the right side, through which ascites moves into thepleural space because of the negative intrathoracic pressure
induced by inspiration. Hepatic hydrothorax can lead to respira-
tory failure and can be complicated by spontaneous bacterialinfections (empyema). Its appearance is associated with poorprognosis, as the median survival of patients with hepatic
hydrothorax ranges from 8–12 months.
128,129Notably, the most
common prognostic scores, such as Child-Pugh and MELD, seemto underestimate such an adverse outcome.
128
Diagnosis of hepatic hydrothorax
Once pleural effusion has been ascertained, cardiopulmonary
and primary pleural diseases should be excluded by standard
clinical approaches. Diagnostic thoracentesis is required to rule
out bacterial infection, whose diagnosis relies on the same crite-ria described for ascites. The protein content of pleural effusion
in uncomplicated hepatic hydrothorax is low and the serum to
pleural fluid albumin gradient is greater than 1.1 g/dl.
128
The presence and extent of diaphragmatic defects can be
assessed indirectly, by radioisotope techniques, or directly by
magnetic resonance imaging or colour-Doppler
ultrasonography.130,131
Treatment of hepatic hydrothorax
The first-line management relies on the treatment of ascites
with diuretics and/or LVP as discussed earlier. However, it isnot rare for pleural effusion to persist despite successful treat-
ment of ascites (refractory hydrothorax). Therapeutic thoracen-
tesis is required to relieve dyspnoea. Its efficacy in refractoryhepatic hydrothorax is transient and repeated thoracentesis
are required, which increase the risk of complications such as
pneumothorax, pleural or soft tissue infection, and bleeding.
132
The frequent occurrence of these complications discourages the
use of chronic pleural drainage, which can also be followed by
renal dysfunction from fluid loss.133
Whenever indicated and possible, LT represents the best
option for patients with refractory hepatic hydrothorax, which
does not seem to adversely affect the outcome of transplanta-
tion.134,135TIPS has been effectively employed as definitive
treatment or bridge to transplantation in patients with refrac-
tory hepatic hydrotorax, whose general outcome seems to be
related to the severity of the underlying cirrhosis.136,137These
results have been confirmed by a more recent meta-analysis.138
Pleurodesis induced by various agents, such as talc, tetracy-
cline, doxycycline, bleomycin and povidone-iodine, can beoffered to patients who are not candidates for TIPS or LT. Arecent meta-analysis showed that the pooled rate of complete
response after pleurodesis was 72%. However, the pooled rate
of complications related to this procedure was as high as82%.
139Finally, thoracoscopic repair with mersilene mesh
appears to be effective in patients with well-defined diaphrag-
matic defects.140Advanced liver disease, as assessed by MELD
score, and preoperative renal dysfunction appear to adversely
affect three-month survival. Unfortunately, clear cut-off values
cannot be retrieved from that study.Recommendations
/C15Patients with hydrothorax should be evaluated for LT
(III;1) .
/C15Cardiopulmonary and primary pleural disease should be
ruled out before diagnosing hepatic hydrothorax (III;1) .
Diagnostic thoracentesis should be performed especially
when infection of the pleural effusion is suspected(III;1) .
/C15Diuretics and thoracentesis are recommended as the
first-line management of hepatic hydrothorax (III;1) .
/C15Therapeutic thoracentesis is indicated in patients with
dyspnoea (III;1) . Chronic pleural should not be per-
formed because of the frequent occurrence of complica-
tions (II-2;1) .
/C15In selected patients, TIPS insertion for recurrent symp-
tomatic hepatic hydrothorax is recommended (II-2;1) .
/C15Pleurodesis can be suggested to patients with refractory
hepatic hydrothorax not amenable to LT or TIPS inser-
tion. However, the frequent occurrence of side effects
related to this technique restricts its use to selectedpatients (I;2).
/C15Mesh repair of diaphragmatic defects is suggested for
the management of hepatic hydrothorax in very selected
patients. The best results can be achieved in patients
with non-advanced cirrhosis without renal dysfunction(II-2;2) .
Hyponatremia
Definition and pathophysiology
Hyponatremia is common in patients with advanced cirrhosis,
and has been arbitrarily defined as serum sodium concentra-tion lower than 130 mmol/L.
141,142However, according to
guidelines on hyponatremia in the general patient popula-
tion,143reductions below 135 mmol/L should also be consid-
ered. Patients with hyponatremia have a poor prognosis, as itis associated with increased mortality
144,145and morbidity,
particularly neurological complications,146,147and reduced sur-
vival after LT.148Incorporating serum sodium concentration
into the MELD score, a new score (MELD-Na) was generated
that provides more accurate survival predictions than MELD
alone,149especially in patients with ascites and hyponatremia
with intermediate MELD score values.150Both hypovolaemic
and hypervolaemic hyponatremia can occur in patients with
cirrhosis. The second, most common, is characterised by anexpansion of the extracellular fluid volume, with ascites and
oedema. It may occur spontaneously, or because of excessive
hypotonic fluids ( i.e., 5% dextrose), or secondary to complica-
tions of cirrhosis leading to an abrupt worsening of effectivevolaemia. The main drivers are non-osmotic hypersecretion
of vasopressin and enhanced proximal nephron sodium reab-
sorption, which impair free water generation and are bothcaused by effective hypovolaemia. As opposed to hyper-
volaemic hyponatremia, hypovolaemic hyponatremia is
characterised by the frequent absence of ascites and oedema.JOURNAL
OF HEPATOLOGY
Journal of Hepatology 2018 vol. xxx jxxx–xxx 11
Please cite this article in press as: The European Association for the Study of the Liver. EASL Clinical Practice Guidelines for the management of pati ents with decompensated cirrhosis. J
Hepatol (2018), https://doi.org/10.1016/j.jhep.2018.03.024

It is caused by a prolonged negative sodium balance with
marked loss of extracellular fluid often due to excessive diure-tic therapy.
Management of hyponatremia
It is generally considered that hyponatremia should be treated
when serum sodium is lower than 130 mmol/L, although thereis no good evidence regarding the level of serum sodium at
which treatment should be initiated. Hypovolaemic hypona-
tremia requires plasma volume expansion with saline solutionand the correction of the causative factor. The management of
hypervolemic hyponatremia requires attainment of a negative
water balance. Non-osmotic fluid restriction is helpful in pre-venting a further decrease in serum sodium levels, but it is sel-
dom effective in improving natremia. Hypertonic sodium
chloride administration to patients with decompensated cirrho-sis may improve natremia but enhances volume overload and
worsens the amount of ascites and oedema. Therefore, it should
be limited to severely symptomatic hyponatremia, as defined bylife-threatening manifestations, cardio-respiratory distress,abnormal and deep somnolence, seizures and coma, which do
not frequently occur in patients with cirrhosis. Furthermore,
hypertonic sodium chloride administration can be consideredin patients with severe hyponatremia who are expected to get
a liver transplant within a few days. In these cases, hypona-
tremia must not be corrected completely and rapidly to avoidthe risk of central pontine myelinolysis that is increased in
advanced cirrhosis.
143In practice, after an initial rapid correc-
tion aimed at attenuating clinical symptoms (5 mmol/L in thefirst hour), serum sodium concentration should not increase
more than 8 mmol/L per day.
143Albumin infusion appears to
improve serum sodium concentration, but more information isneeded.
151
Vaptans
Vaptans are selective antagonists of the V2-receptors of argi-
nine-vasopressin in the principal cells of the collecting ductsthat enhance solute-free water excretion.
152Indeed, these
drugs are effective in improving serum sodium concentration
in conditions associated with high vasopressin levels, such asthe syndrome of inappropriate antidiuretic hormone secretion
(SIADH) and heart failure.
152The effects of the administration
of vaptans to hyponatremic patients with cirrhosis and asciteshave been assessed in several studies. Namely, tolvaptan,
satavaptan and lixivaptan lead to an increased urine volume,
a solute-free water excretion, and an improvement of hypona-tremia in 45–82% of cases.
153–155In another study, a short-
term intravenous infusion of conivaptan for one to four days
in patients with end stage liver disease awaiting OLT was also
effective in increasing serum sodium concentration.156How-
ever, the safety of vaptans has only been established for
short-term treatments lasting from one week to one month.
When satavaptan was used long term, in addition to diuretics,despite improving both serum sodium concentration and con-
trol of ascites, a higher all-cause mortality rate, mostly associ-
ated with known complications of cirrhosis, was reportedcompared to standard medical treatment.
157,158Moreover, a
recent study cast doubt on the efficacy of tolvaptan in patients
with cirrhosis and severe hypervolemic hyponatremia (serumsodium ≤125 mEq/L) in a real-life setting.
159At present, both
conivaptan and tolvaptan have been approved in the US by
the FDA, while only tolvaptan in Europe has been approvedby the EMA for management of severe hypervolemic hypona-
tremia (<125 mmol/L). The unique indication given for tolvap-tan by the EMA is SIADH, while the FDA also included heartfailure and liver cirrhosis. However, the occurrence of serious
hepatic injury in three patients with autosomal dominant poly-
cystic kidney disease treated with tolvaptan in a double-blindplacebo-controlled trial
160led the FDA to conclude that this
drug should not be used in patients with underlying liver
disease.
Recommendations
/C15The development of hyponatremia (serum sodium con-
centration <130 mmol/L) in patients with cirrhosis car-ries an ominous prognosis, as it is associated with
increased mortality and morbidity. These patients
should be evaluated for LT (II-2,1) .
/C15The removal of the cause and administration of normal
saline are recommended in the management of hypov-olemic hyponatremia (III;1) .
/C15Fluid restriction to 1,000 ml/day is recommended in the
management of hypervolemic hyponatremia since itmay prevent a further reduction in serum sodium levels
(III;1) .
/C15The use of hypertonic saline in the management of
hypervolemic hyponatremia should be limited to the
rare cases presenting with life threatening complica-tions. It could also be considered in patients with severe
hyponatremia who are expected to get LT within a few
days. The correction of serum sodium concentration,once an attenuation of symptoms has been obtained,should be slow (≤8 mmol/L per day) to avoid irreversible
neurological sequelae, such as osmotic demyelination
(II-3;1) .
/C15Albumin administration can be suggested in hyperv-
olemic hyponatremia, but data are very limited to sup-port its use (II-3;2) .
/C15At present, the use of vaptans should be limited to con-
trolled clinical studies (III;1) .
Gastrointestinal bleeding
Pathophysiology
Variceal haemorrhage (VH) occurs because of the rupture of
the variceal wall due to excessive wall tension. Variceal walltension is an intrinsic property of the vessel wall that opposes
the expansive force determined by variceal transmural pres-
sure, which depends on portal pressure and vessel size. Tissuesupport surrounding the varix may counteract the increase invariceal pressure and size, protecting the wall from rupture.
161
Once variceal wall rupture occurs, the amount of bleeding isrelated to transmural pressure (which mainly depends on por-tal pressure), to the area of rupture in the vessel wall and to
blood viscosity and/or alterations of haemostasis.
161All these
factors can be influenced by therapy. Drug therapy and por-tal-systemic derivative procedures, reduce portal (and variceal)
pressure. Endoscopic therapies and other physical methods,
such as balloon tamponade or expandable prosthesis, actmerely by both interrupting the blood flow into the varixClinical Practice Guidelines
12 Journal of Hepatology 2018 vol. xxx jxxx–xxx
Please cite this article in press as: The European Association for the Study of the Liver. EASL Clinical Practice Guidelines for the management of pati ents with decompensated cirrhosis. J
Hepatol (2018), https://doi.org/10.1016/j.jhep.2018.03.024

and sealing the vascular wall. Portal pressure is a key factor
determining both variceal rupture and the severity of thebleeding episode.
161During acute bleeding, portal pressure
may increase because of different factors such as over-transfu-
sion or absorption of blood from the gut, which may have a
role in failure to control bleeding and/or precipitating rebleed-ing. Portal pressure is usually assessed by the hepatic venous
pressure gradient (HVPG).
Natural history of gastro-oesophageal varices and variceal
haemorrhage
Variceal haemorrhage, causative of 70% of all upper GI bleed-
ing events in patients with portal hypertension, remains one
of the most severe and immediate life-threatening complica-
tions in patients with cirrhosis and constitutes the secondmost frequent decompensating event after ascites.
162,163
Decompensated patients have ‘clinically significant portal
hypertension’ (CSPH) per definition and, consequently, a high
risk of having gastro-oesophageal varices. In fact, while only42% of Child A patients have gastro-oesophageal varices,
72% of Child B/C patients do.
164When decompensation devel-
ops, patients without varices on a previous endoscopy shouldhave a repeat endoscopy performed given the risk of develop-
ing varices due to worsening of portal hypertension and liver
dysfunction. In those without varices at screening, ‘‘ de novo ”
varices develop at a rate of 7–8%/year,
165,166which could be
higher in patients decompensated due to worsening of portal
hypertension and liver dysfunction. The progression rate from
small to large varices runs up to 22% at one year and 51% atthree years in patients with Child B/C cirrhosis, especially
when alcoholic in origin and/or when red wale marks are pre-
sent at first endoscopy, compared to 2% and 16%, respectively,in compensated patients without those risk factors.
165,166
Prospective studies have consistently demonstrated that therisk of VH, estimated overall at 5–15% per year, is relatedto variceal size.
166–170This risk is further amplified by the
severity of liver dysfunction (Child B/C) and/or the presence
of red wale marks on varices. Thus, not only medium/largevarices ( i.e.varices that do not collapse with insufflation at
endoscopy), but also small varices with red signs or in Child
C should be considered ‘ high-risk ’ varices. Despite improve-
ment in therapy, overall mortality with each episode of VHremains around 15% to 25% at six weeks. Such risk is much
higher in patients who develop VH in addition to other
decompensation (over 80% at five-years) than in those pre-senting with VH as an isolated decompensating event (20%
at five-years).
170,171Mortality risk is particularly high when
VH is associated with AKI and/or concomitant bacterial infec-tions.
172Without secondary prophylaxis, rebleeding occurs in
approximately 60% to 70% of patients, usually within one to
two years of the index haemorrhagic event.173Although
increasing efforts are performed to test non-invasively forthe presence of gastro-oesophageal varices, these efforts lar-
gely remain restricted to compensated cirrhosis.
167Given
the high prevalence of ‘high-risk varices’ in decompensatedcirrhosis, oesophago-gastroduodenoscopy (EGD) should be
performed to detect the presence, size of varices and presence
of red wale marks.
168,169Recommendations
/C15Because they are deemed high risk, patients in whom
decompensation develops should have EGD performed
to screen for gastro-oesophageal varices, unless previ-
ously diagnosed and treated (II-2;1) .
/C15If EGD is performed, the presence, size and presence of
red wale marks should be reported (II-2;1) .
/C15In patients without varices on screening EGD in whom
the aetiological factor persists and/or the state of decom-pensation continues, screening EGD should be repeatedevery year. In the remaining patients the screening could
be prolonged, but the exact interval is unclear and more
data is required (III;2) .
Prevention and treatment of variceal haemorrhage
Considering the high-risk of death when VH occurs in patients
with decompensated cirrhosis, implementation of strategies to
adequately treat VH and to prevent (re)bleeding and deathshould be actively pursued in patients with decompensated cir-
rhosis. It should be noted that the current recommendations
will concentrate on decompensated patients given the focus ofthese CPGs.
Primary and secondary prophylaxis of VH in decompensated
patients
The Baveno VI
168and American Association for the Study of
Liver Diseases (AASLD)169guidelines primarily recommend
NSBBs for primary prophylaxis of VH in patients with cirrhosiswho have high-risk varices and also, combined with endoscopic
band ligation (EBL), for secondary prophylaxis of VH. Both
NSBBs and EBL have shown to be equally effective in preventingfirst bleeding in patients with high-risk varices. The choice
between options depends on factors such as patient preference,
contraindications or adverse events. Although numerically EBLinduces less side effects, it has been associated with more sev-ere and potentially life-threatening complications, resulting
from bleeding EBL-ulcers. Moreover, EBL does not impact on
portal hypertension. Thus, it does not reduce/prevent othercomplications and surveillance endoscopies are required to
detect variceal recurrence, supporting overall primary prefer-
ence for NSBBs.
174,175For prevention of rebleeding (secondary
prophylaxis), combined therapy with NSBBs plus EBL is recom-
mended because combination therapy significantly decreases
the probability of rebleeding compared to monotherapy usingeither EBL or drug therapy. NSBBs are the cornerstone of com-
bined therapy because a meta-analysis shows an improvement
in survival with the addition of NSBBs (± nitrates) to EBL, whilethe addition of EBL to NSBBs (± nitrates) has no effect on mor-tality.
176Recent RCTs indicate that guiding therapy according
to the HVPG response to NSBBs can be valuable in this high-risk
setting.91,177HVPG-guided therapy may improve the outcomes
achieved with current first-line therapy combining NSBBs and
EBL,91and may achieve a similar survival as covered TIPS, which
is the most effective therapy in terms of preventing bleeding.177JOURNAL
OF HEPATOLOGY
Journal of Hepatology 2018 vol. xxx jxxx–xxx 13
Please cite this article in press as: The European Association for the Study of the Liver. EASL Clinical Practice Guidelines for the management of pati ents with decompensated cirrhosis. J
Hepatol (2018), https://doi.org/10.1016/j.jhep.2018.03.024

Accordingly, HVPG-guided therapy can be used when available.
However, this approach has relevant drawbacks such as inva-siveness and limited availability and, therefore, cannot bewidely recommended. NSBBs, such as propranolol and nadolol,
act on portal hypertension because non-selective beta-blockade
reduces cardiac output and splanchnic blood flow while theunopposed effect of alpha-1 adrenergic receptors leads to
splanchnic vasoconstriction, thus reducing portal pressure and
its consequential complications. Nonetheless, haemodynamicresponse rates to NSBBs are modest: approximately 46% of cases
according to meta-analyses,
178,179endorsing the overall search
for novel therapeutic options. Carvedilol, an NSBB with intrinsicanti-alpha-1 receptor activity, has been associated with a
greater reduction in portal pressure than the traditional NSBBs
and has therefore become a valuable alternative.
180Its benefi-
cial action on alpha-1 receptors reduces both porto-collateraland intrahepatic resistance, however, this is at the cost of more
profound effects on systemic arterial pressure, particularly in
decompensated patients. The problem with all the recommen-dations mentioned so far is that they are based on high quality
RCTs that usually excluded patients with more advanced cirrho-
sis, while major controversy has arisen in recent years regardingthe use and safety of NSBBs in patients with advanced disease,
particularly in those with refractory ascites and/or SBP. The dis-
cussion was initiated by the Clichy group,
181who reported poor
survival and increased risk of PPCD among patients with refrac-
tory ascites on NSBB therapy. The mechanism underlying these
findings was thought to relate to further induction of systemicarterial hypotension and exhaustion of cardiac reserve, in lightof the progressive hyperdynamic circulation typically associated
with end-stage disease. As a result, end-organ perfusion
becomes critical and sets off a multitude of complications, likeHRS. Therefore, ‘‘the window hypothesis” was proposed which
suggested refractory ascites as a critical juncture where the pro-
tective effects of NSBBs may cease and a detrimental impactmay begin.
182However, this hypothesis was challenged by
opposing reports suggesting protective effects with NSBBs even
in decompensated patients.183–186Illustratively, a recent post
hocanalysis of three RCTs where vaptans and NSBBs were co-
administered to patients with ascites showed that NSBBs did
not increase mortality.183On the contrary, during follow-up,
29% of initial NSBB users stopped taking NSBBs, inducing amarked rise in mortality and coinciding with variceal bleeding,
bacterial infection and/or development of HRS.
183Non-haemo-
dynamic effects of NSBBs, like reduction of intestinal permeabil-ity, inflammation and BT, are considered to contribute to the
beneficial effect, particularly in this advanced stage.
187–189
Whether NSBBs are detrimental in some patients with advanced
cirrhosis should be clarified by future studies (ideally RCTs), as
well as the optimal drug-schedule in such stages. Meanwhile
some considerations could be made regarding dosing, type ofNSBB and titration.
168,184,185,190Firstly, dosing of NSBBs was
suggested as a potential determinant according to a Danish
study in which low propranolol doses (<160 mg/day) were asso-
ciated with reduced mortality after experiencing an SBPcompared to higher doses.
184,190Secondly, not all NSBBs proved
equal. Carvedilol, which exhibits additional vasodilatory
anti-alfa-1-adrenergic activity, might be deleterious in decom-pensated patients as it is more likely to cause a systemichaemodynamic depressive effect and may be best avoided or
very closely monitored.
185Thirdly, the concept of titration of
NSBBs to a target heart rate of 50–55 bpm might be challengedin decompensated patients given that, in parallel to the progres-
sion of liver disease, the hyperdynamic state evolves similarly,
which may lead to treatment of the most vulnerable patientsparadoxically with higher, and potentially hazardous, doses.
Therefore, the use of NSBBs should be based on a critical risk/ben-
efit evaluation in patients with refractory ascites and signs of sys-temic circulatory dysfunction.
168,191Parameters such as severe
hyponatraemia,191low mean arterial pressure38or cardiac out-
put,192and increasing SCr193identify more vulnerable patients
among those with decompensated cirrhosis, in whom a dose
reduction or temporal discontinuation of NSBB treatment should
be considered. The recent BAVENO VI consensus168proposed that
in patients with refractory ascites and (i) systolic blood pressure<90 mmHg, or (ii) SCr >1.5 mg/dl, or (iii) hyponatraemia <130
mmol/L, the NSBB dose should be reduced or even temporarily
discontinued. Abrupt interruption of beta-blockers for a meanof three to six days was recently found to be associated with nei-
ther an apparent increase in the risk of variceal bleeding nor with
a haemodynamic rebound.
194If upon rechallenge, NSBB intoler-
ance occurs, EBL should be considered as an alternative in primary
prophylaxis.168In the setting of refractory ascites and secondary
prophylaxis, covered TIPS placement may be considered if thepatient is an appropriate candidate.
111,168
Recommendations
/C15Primary prophylaxis must be initiated upon detection of
‘‘high-risk varices” ( i.e.small varices with red signs,
medium or large varices irrespective of Child-Pugh clas-
sification or small varices in Child-Pugh C patients)because of increased risk of VH (I;1).
/C15Patients with small varices with red wale marks or
Child-Pugh C should be treated with NSBBs (III;1) .
/C15Patients with medium-large varices should be treated
with either NSBBs or EBL (I;1). The choice of treatment
can be based on local resources and expertise, patient
preference, contraindications and adverse events (III;2) .
NSBBs could be preferred because in addition to lower-ing portal pressure, they also exert other potential bene-ficial effects (II-2;2) .
/C15Although ascites is not a contraindication for NSBBs, cau-
tion should be exercised in cases of severe or refractory
ascites (I;1). High doses of NSBB should be avoided
(II-2;1) . The use of carvedilol can not be recommended
at present (I;2).
/C15In patients with progressive hypotension (systolic BP
<90 mmHg), or in patients who develop an acute inter-
current conditions such as bleeding, sepsis, SBP or AKI,
NSBBs should be discontinued (III,1) . After recovery,
reinstatement of NSBBs can be attempted (III,2) . When
NSBB intolerance or contraindications persist, patients
bleeding risk should be managed by expeditious EBL
(III,1) .Clinical Practice Guidelines
14 Journal of Hepatology 2018 vol. xxx jxxx–xxx
Please cite this article in press as: The European Association for the Study of the Liver. EASL Clinical Practice Guidelines for the management of pati ents with decompensated cirrhosis. J
Hepatol (2018), https://doi.org/10.1016/j.jhep.2018.03.024

Recommendations
/C15Combination therapy of NSBBs + EBL is recommended
since it reduces the risk of rebleeding compared withmonotherapy (I,1).
/C15Similar recommendations as for primary prophylaxis can
be made with respect to NSBB usage in patients with
ascites or developing an acute inter-current condition
(III,2) .
/C15If the patient continues to be intolerant to NSBB, covered
TIPS placement is recommended provided that there areno absolute contraindication (cf. criteria in ascites sec-
tion) (III,1) .
Variceal haemorrhage
Acute GI bleeding in cirrhosis, either because of gastro-oesopha-
geal varices or non-variceal lesions, is a medical emergencywith a high incidence of complications and high mortality and
therefore requires intensive care ( Fig. 2 ). Acute variceal haemor-
rhage (AVH) must be suspected in any cirrhotic patient present-ing with upper acute GI bleeding and treatment should be
started as soon as bleeding is clinically confirmed, regardless
the lack of confirmation by upper endoscopy.
195Initial therapy
should be directed at restoring volaemia.196Vasoactive drug
therapy197,198and antibiotic prophylaxis195,196should be initi-
ated as soon as AVH is suspected. Goals of therapy in AVHinclude the control of bleeding, as well as the prevention of
early recurrence and the prevention of six-week mortality,which is considered the main treatment outcome by consen-
sus.
168,199Blood volume restitution should be initiated
promptly to restore and maintain haemodynamic stability toensure tissue perfusion and oxygen delivery. To facilitate resus-
citation at least two catheters should be placed, large enough to
allow rapid volume expansion, which can usually be done withcrystalloids.
196No benefit has been demonstrated with the use
of colloids compared to crystalloids.200Red blood cells are used
to improve oxygen delivery to tissues in case of severe anaemia.A restrictive transfusion strategy is adequate in most patients
with acute GI bleeding, with a haemoglobin threshold for
transfusion of 7 g/dl and a target range after transfusion of7 to 9 g/dl.
201The threshold for transfusion may be higher in
patients with massive haemorrhage or in those with underlying
conditions that preclude an adequate physiological response toacute anaemia. Recommendations regarding management ofcoagulopathy and thrombocytopenia cannot be made based on
currently available data.
168,169,199
As mentioned above vasoactive drug therapy should be initi-
ated as soon as AVH is suspected. Starting vasoactive drugs
before endoscopy decreases the incidence of active bleeding dur-
ing endoscopy and facilitates endoscopic therapy, improving thecontrol of bleeding, and potentially survival.
197,198Either terli-
pressin, somatostatin or octreotide are accepted drugs with pro-
ven efficacy.202All these drugs require i.v administration. The
recommended dose of terlipressin is 2 mg/4 h during the first
48 h, followed by 1 mg/4 h thereafter. The recommended dose
of somatostatin is a continuous infusion of 250 lg/h (that can
be increased up to 500 lg/h) with an initial bolus of 250 lg.
The recommended dose of octreotide is a continuous infusion
of 50 lg/h with an initial bolus of 50 lg. A bolus of somatostatin
Acute gastrointestinal bleeding + portal hypertension
Immediate start of drug therapy
(somatostatin /terlipressin )
Antibiotic prophylaxis
(ceftriaxone or norfloxacine )
Control
(~85% of cases)Further bleeding
(~15% of cases)
Rescue with TIPS Consider early TIPS
in high-risk+ maintain vasoactive drug therapy 3-5 days
and antibiotic prophylaxis ( ceftriaxone or norfloxacine )Balloon tamponade or esophageal stenting
if masive bleedingEarly diagnostic endoscopy (<12 h)
Confirm variceal bleeding
Endoscopic therapy (band ligation )ENDOSCOPYENDOSCOPYInitial assesment (history, physical & blood exam, cultures)
& resuscitationAirway
Breathing
Circulation
– Volume replacement with crystalloids
(or colloids)
– Restrictive transfusion
Hb threshold of 7 g/dl & target of 7- 9 g/dl
+
Fig. 2. Algorithm for the management of acute gastrointestinal bleeding in patients with cirrhosis (adapted from Ref. 168).TIPS, transjugular
portosystemic shunts.JOURNAL
OF HEPATOLOGY
Journal of Hepatology 2018 vol. xxx jxxx–xxx 15
Please cite this article in press as: The European Association for the Study of the Liver. EASL Clinical Practice Guidelines for the management of pati ents with decompensated cirrhosis. J
Hepatol (2018), https://doi.org/10.1016/j.jhep.2018.03.024

or octreotide can be given again if bleeding is ongoing. Once AVH
is confirmed, vasoactive drug therapy should be administeredfor five days to avoid early rebleeding.
168,169Shorter administra-
tion of vasoactive drugs (48–72 h) can be considered in less sev-
ere episodes although more data are required.203Once blood
volume restitution has been initiated and haemodynamic stabil-ity has been achieved, upper endoscopy should be performed, as
soon as possible within the first 12 h after admission, to ascer-
tain the cause of haemorrhage (up to 30% of cirrhotic patientsbleed from non-variceal causes) and to provide endoscopic ther-
apy if indicated.
168,169Erythromycin should be considered
before emergency endoscopy (250 mg i.v., 30–120 min before)to facilitate the procedure by improving visibility, in the absence
of contraindications (QT prolongation).
204When AVH is con-
firmed by endoscopy, EBL should be performed within the sameprocedure. EBL is more effective than sclerotherapy to controlbleeding, with fewer adverse effects, and may even improve sur-
vival.
196Sclerotherapy can be used when ligation is not feasible.
The combination of endoscopic therapy and vasoactive drugs ismore effective than the isolated use of either of these options
alone,
205,206because it combines the local haemostatic effect
on the varices induced by endoscopic treatment and the portalhypotensive effect achieved with drugs. This combination is cur-
rently considered the standard of care in AVH.
168,169Cyanoacry-
late injection and EBL are accepted options for endoscopictherapy in patients bleeding from gastric (cardiofundal) varices
as both therapies are equally effective.
207However, EBL should
only be performed on small gastric varices in which the com-plete vessel can be suctioned into the ligation device. Otherendoscopic therapies, such as the endoscopic ultrasound–guided
insertion of coils and/or cyanoacrylate, are available for fundal
varices. Prevention of complications should run simultaneouslyto haemostatic therapies from admission of patients with cirrho-
sis and acute GI bleeding. The main complications, whatever the
cause of bleeding, include bacterial infections (such as aspirationpneumonia or SBP), hepatic encephalopathy and deterioration of
renal function. Bacterial infections are observed in more than
50% of patients and may already be present at the time of bleed-ing (20%) acting as a precipitating event.
196Moreover, the pres-
ence of bacterial infection is an independent predictor of failure
to control bleeding and death.208Antibiotic prophylaxis is rec-
ommended because it reduces the incidence of infections andimproves control of bleeding and survival.
199,208Ceftriaxone (1
g/24 h) for up to seven days, is the first choice in patients with
advanced cirrhosis, in those on quinolone prophylaxis and inhospital settings with high prevalence of quinolone-resistant
bacterial infections.
209,210Oral quinolones (norfloxacin 400 mg
b.i.d) can be used in the remaining patients. These recommenda-tions are however best evaluated and cross-checked from the
perspective of local resistance patterns. Renal function should
be preserved by the adequate replacement of fluids and elec-trolytes.
211Nephrotoxic drugs (such as aminoglycosides and
non-steroidal anti-inflammatory drugs [NSAIDs]) as well as
LVP, beta-blockers, vasodilators and other hypotensive drugs
should be avoided during the course of AVH. Oral non-absorb-able disaccharides may be used to prevent the development of
hepatic encephalopathy,
169although more studies are needed.
When encephalopathy develops, lactulose or lactitol should beused.
168,169Proton pump inhibitors (PPIs) have not shown effi-
cacy for the management of AVH. However, a short course ther-
apy with PPI after EBL may reduce the size of post-bandingulcers.
212Despite therapy with vasoactive drugs plus EBL andprophylactic antibiotics, up to 10–15% of patients with AVH have
persistent bleeding or early rebleeding.195,199In such cases, TIPS
should be considered as the rescue therapy of choice.168,169
When TIPS is not feasible or in case of modest rebleeding, a sec-
ond endoscopic therapy may be attempted while vasoactive
drugs can also be optimised, by doubling the dose of somato-statin and/or changing to terlipressin if not used previously. Bal-
loon tamponade should be used in case of massive bleeding, as a
temporary ‘‘bridge” until definitive treatment can be institutedand for a maximum of 24 h, preferably under intensive care facil-
ities.
168,169Because of the high risk of aspiration pneumonia,
tamponade should be preceded by prophylactic orotrachealintubation in comatose or encephalopathic patients. Removable,
covered and self-expanding oesophageal stents are an alterna-
tive to balloon tamponade, and may have lower rates of seriousadverse events.
213RCTs suggest that in high-risk patients, early
(preemptive) PTFE-coated TIPS placed within 72 h (ideally in less
than 24 h) may result in better permanent control of bleeding
and may improve survival.214,215However, these studies had rel-
evant drawbacks such as the inclusion of a highly selected pop-
ulation because of strict exclusion criteria, while observational
studies have not confirmed the effect on survival.216,217The
use of Child-Pugh class B plus active bleeding at endoscopy as
a criterion to select high-risk patients has also been criticised.218
It has also been suggested that a recalibrated MELD score may
better identify patients at high risk than other scores.219At pre-
sent, early TIPS should be considered in patients with Child-Pugh
class C, with a score <14. However, future studies should clarifywhich criteria may be preferred to select high-risk patientsbefore a wide implementation of early TIPS. Future studies
should also clarify whether an adequate stratification of risk in
patients with AVH can optimise therapy.
Recommendations
/C15Acute GI bleeding, both due to gastro-oesophageal
varices or to non-variceal lesions, carries a high inci-dence of complications and mortality in decompensated
cirrhosis and therefore requires close monitoring
(II-2;1) .
/C15Volume replacement should be initiated promptly to
restore and maintain haemodynamic stability (III;1) .
Either colloids and/or crystalloids should be used(III;1) . Starch should not be used for volume replace-
ment (I;1).
/C15A restrictive transfusion strategy is recommended in
most patients with a haemoglobin threshold for transfu-
sion of 7 g/dl and a target range of 7–9 g/dl (I;1).
/C15Antibiotic prophylaxis is recommended in cirrhotic
patients with acute GI bleeding because it reduces theincidence of infections and improves control of bleedingand survival. Treatment should be initiated on presenta-
tion of bleeding and continued for up to seven days (I;1).
Ceftriaxone (1 g/24 h) is the first choice in patients withdecompensated cirrhosis, those already on quinolone
prophylaxis, and in hospital settings with high preva-
lence of quinolone-resistant bacterial infections. Oralquinolones (norfloxacin 400 mg b.i.d) should be used in
the remaining patients (I;1).Clinical Practice Guidelines
16 Journal of Hepatology 2018 vol. xxx jxxx–xxx
Please cite this article in press as: The European Association for the Study of the Liver. EASL Clinical Practice Guidelines for the management of pati ents with decompensated cirrhosis. J
Hepatol (2018), https://doi.org/10.1016/j.jhep.2018.03.024

Recommendations
/C15Vasoactive drug therapy should be initiated as soon as
acute variceal bleeding is suspected, and before endo-scopy. Terlipressin, somatostatin or octreotide are
accepted options. In patients with acute variceal bleed-
ing drug therapy should be administered for three to fivedays (I;1).
/C15Gastroscopy should be performed within the first 12 h
after admission once haemodynamic stability has been
achieved, to ascertain the cause of haemorrhage and to
provide endoscopic therapy (II-2;1) .
/C15When acute variceal bleeding is confirmed by endo-
scopy, variceal ligation should be performed within thesame procedure (I;1).
/C15In the absence of contraindications (QT prolongation)
pre-endoscopy erythromycin (250 mg i.v., 30–120 minbefore) can be used to facilitate the procedure (I;2).
/C15The combination of vasoactive drugs and ligation is rec-
ommended as the first therapeutic option in acute vari-
ceal bleeding (I;1).
/C15Early pre-emptive covered TIPS (placed within 24–72 h)
can be suggested in selected high-risk patients, such as
those with Child class C with score <14 (I;2). However,
the criteria for high-risk patients, particularly Child Bwith active bleeding, remains debatable and needs fur-
ther study.
Recommendations
/C15Up to 10–15% of patients have persistent bleeding or
early rebleeding despite treatment with vasoactive
drugs plus variceal ligation, and prophylactic antibiotics.TIPS should be used as the rescue therapy of choice in
such cases (I;1).
/C15Balloon tamponade should be used in case of uncon-
trolled bleeding, but with pre-requisite of expertise
and as a temporary ‘‘bridge” until definitive treatment
can be instituted and for a maximum of 24 h (III;1) .
Removable, covered and self-expanding oesophageal
stents can be used as alternative to balloon tamponade
(I;2).
/C15In the context of bleeding, where encephalopathy is
commonly encountered, prophylactic lactulose may beused to prevent encephalopathy, but further studies
are needed (I;2).
/C15Beta-blockers and vasodilators should be avoided during
the acute bleeding episode (III,1) .
Portal hypertension gastropathy and intestinopathy
Portal hypertension gastropathy (PHG) often presents in decom-
pensated patients given that its natural history is significantlyinfluenced by the severity of liver disease and portal hyperten-
sion. The presence of oesophageal varices and a Child-Pugh classB or C at enrollment were found to predict the incidence of PHG,
which might range between 30 and 45%.
220,221The incidence
and severity of PHG may increase following endoscopic treat-ment for oesophageal varices.
222Portal hypertension should
be distinguished from gastric antral vascular ectasia (GAVE or
watermelon stomach), which have different underlyingpathophysiologies and different therapeutic implications. The
diagnosis of PHG is made by endoscopy and typically shows a
snake-skin mosaic pattern (mild subtype), which may havesuperimposed red signs (severe PHG) and is most commonly
located in the proximal stomach (fundus and body) whereas
GAVE is characterised by the presence of red spots without abackground mosaic pattern, typically located in the gastric
antrum.
170Similar endoscopical lesions, as documented in
PHG, may be observed in other areas of the GI tract where theyhave been termed portal hypertensive duodenopathy, portalhypertensive enteropathy or portal hypertensive colopathy
depending on the location of the lesions.
223PHG and every form
of enteropathy might be clinically important because they aresometimes responsible for insidious blood loss (chronic iron
deficient anaemia) and in exceptional cases even overt acute
bleeding. When PHG is found as an incidental asymptomaticfinding without concomitant oesophageal or gastric varices, its
relevance is unclear and endoscopic follow-up or prophylactic
treatment is not recommended.
168First-line therapy for chronic
haemorrhage from PHG is an NSBB.168,224,225The same consid-
erations regarding the use of NSBBs in decompensated patients
should be made as for gastro-oesophageal varices, except thatthere is no alternative, endoscopic, standard intervention avail-able for PHG. In addition, iron supplementation should be pro-
vided.
168,226In patients with medically refractory PHG and
compensated cirrhosis, TIPS has shown to improve the endo-scopic appearance and decrease the transfusion requirement.
227
In case of acute PHG bleeding, albeit rare, small and uncon-trolled studies have suggested pharmacological interventionwith somatostatin-analogues or terlipressin because of their
portal hypotensive effects and reduction in gastric blood
flow.
226,228In addition, similar measures are to be taken as for
AVH (antibiotic prophylaxis, restrictive transfusion policy). Forportal hypertension intestinopathy, there is no established stan-
dard of treatment and an approach analogous to that for PHG is
suggested. As for any given complication, LT should be consid-ered as part of the management of decompensated patients.
Recommendations
/C15NSBB and iron supplementation and/or blood transfu-
sion, when indicated, are recommended as first-line
therapy for chronic haemorrhage from PHG is an (I;1).
/C15In patients with transfusion-dependent PHG in whom
NSBBs fail or are not tolerated, covered TIPS placement
may be used provided the patient has no contraindica-tion for TIPS (II-3;2) .
/C15Acute PHG bleeding may be treated with somatostatin-
analogues or terlipressin but substantiating data are lim-ited(I;2).
Gastric varices
The Sarin classification is most commonly used for risk stratifi-
cation and management of gastric varices ( Table 6 ).
229GastricJOURNAL
OF HEPATOLOGY
Journal of Hepatology 2018 vol. xxx jxxx–xxx 17
Please cite this article in press as: The European Association for the Study of the Liver. EASL Clinical Practice Guidelines for the management of pati ents with decompensated cirrhosis. J
Hepatol (2018), https://doi.org/10.1016/j.jhep.2018.03.024

varices are present in about 20% of patients with cirrhosis. Gas-
tro-oesophageal varices type 1, which are the most common(75% of gastric varices), are oesophageal varices extendingbelow the cardia into the lesser curvature and, in the absence
of specific studies, are commonly managed following guidelines
for oesophageal varices.
168Cardiofundal varices (gastro-oeso-
phageal varices type 2 & isolated gastric varices type 1) bleed
less frequently. However, haemorrhage from cardiofundal
varices is often more severe, more difficult to control and showsa higher risk of recurrent bleeding and mortality (up to 45%)
compared to oesophageal varices.
229Cardiofundal varices are
more frequent in patients with splanchnic venous thrombosis,which should be investigated by imaging. The evidence to sup-
port recommendations for management of gastric VH is much
less robust than that for oesophageal varices. Regarding primaryprophylaxis of bleeding from gastric varices, a single ran-domised trial suggested that cyanoacrylate injection may be
more effective than NSBBs in preventing first bleeding in
patients with large cardiofundal varices, although survival wassimilar.
230Therefore, the last BAVENO consensus concluded
that further studies are needed to evaluate the risk/benefit ratio
of using cyanoacrylate in this setting before a formal recom-mendation can be made and meanwhile propose NSBBs as the
primary approach.
168Acute gastric VH is medically treated like
bleeding oesophageal varices. However, injection therapy withcyanoacrylate (‘glue’) may be the preferable option for endo-
scopic haemostasis.
231Although equally effective as EBL in ini-
tial haemostasis, the rebleeding rate is significantly lower.232
TIPS, with or without additional embolisation of collaterals, isequally effective in gastric and oesophageal VH for control of
acute bleeding events and prevention of rebleeding.
233In case
of massive bleeding, balloon tamponade with the Linton-Nach-las tube may serve as a bridge to other treatments. Regarding
secondary prophylaxis, in one RCT repeated cyanoacrylate
injection was superior to NSBBs to prevent rebleeding from car-diofundal varices,
232while the addition of NSBBs to cyanoacry-
late did not improve the outcomes achieved with glue alone in
another RCT.234Another trial comparing TIPS to glue injection
showed that TIPS proved more effective in preventing rebleed-ing from gastric varices, with similar survival and frequency of
complications.
235The option of early TIPS should be strongly
considered, particularly in cardiofundal varices given the highrebleeding rate, provided that patient is an appropriate
candidate for such a procedure. Alternatively, balloon-occluded
retrograde transvenous obliteration (BRTO) can be considered.This interventional radiological procedure enables treatment
of fundal varices associated with a large gastro/splenorenal
collaterals, which has the theoretical advantage over TIPS ofnot diverting portal blood inflow from the liver. However, norandomised trials are available comparing BRTO with other
therapies. Several variations of this technique are available,such as balloon-occluded antegrade transvenous obliteration(BATO).
236
Recommendations
/C15NSBBs are suggested for primary prevention of VH from
gastro-oesophageal varices type 2 or isolated gastric
varices type 1 (III;2) .
/C15Primary prevention for gastro-oesophageal varices type
1 follow the recommendations of oesophageal varices
(III;2) .
/C15Acute gastric VH should be treated medically, like oeso-
phageal VH (I;1). Cyanoacrylate is the recommended
endoscopic haemostatic treatment for cardiofundal
varices (gastro-oesophageal varices type 2 or isolated
gastric varices type 1) (I;2).
/C15TIPS with potential embolisation efficiently controls
bleeding and prevents rebleeding in fundal VH (gastro-
oesophageal varices type 2 or isolated gastric varicestype 1) and should be considered in appropriate candi-
dates (II-2;1) .
/C15Selective embolisation (BRTO/BATO) may also be used to
treat bleeding from fundal varices associated with large
gastro/splenorenal collaterals, although more data isrequired (III;2) .
Bacterial infections
The risk of bacterial infection in cirrhosis is caused by multiple
factors that include liver dysfunction, portosystemic shunting,
gut dysbiosis, increased BT, cirrhosis-associated immune dys-
function,237,238and genetic factors This immune defect facili-
tates BT, induced by increased intestinal permeability and gut
bacterial overgrowth observed in cirrhosis.239Genetic immune
defects can contribute to the high risk of bacterial infectionsin cirrhosis, particularly SBP. Cirrhotic patients carrying NOD2
variants associated with impaired recognition of the bacterial
product muramyl dipeptide have a higher risk of SBP and areduced survival time.
240
Spontaneous bacterial peritonitis
Definition
Spontaneous bacterial peritonitis has been defined as a bacterial
infection of ascitic fluid without any intra-abdominal surgically
treatable source of infection. SBP is very common in patientswith cirrhosis and ascites.
241,242When first described, its mor-
tality exceeded 90% but it has been reduced to approximately
20% with early diagnosis and treatment.243
Diagnosis
The diagnosis of SBP is based on diagnostic paracentesis.33,244
All patients with cirrhosis and ascites are at risk of SBP and
the prevalence of SBP in outpatients is 1.5–3.5% and /C2410% in
hospitalised patients.245Half the episodes of SBP are present
at the time of hospital admission while the rest are acquiredduring hospitalisation.
33Patients with SBP may have one of
the following:33i) local symptoms and/or signs of peritonitis:Table 6 .Classification, prevalence and risk of bleeding of gastric varices.
Type Definition Relative
frequencyOverall bleeding
risk without
treatment
GOV 1 OV extending below cardia
into lesser curvature70% 28%
GOV 2 OV extending below cardia
into fundus21% 55%
IGV 1 Isolated varices in the
fundus7% 78%
IGV 2 Isolated varices else in the
stomach2% 9%
GOV, gastro-oesophageal varices; IGV, isolated gastric varices; OV, oesophageal
varices.Clinical Practice Guidelines
18 Journal of Hepatology 2018 vol. xxx jxxx–xxx
Please cite this article in press as: The European Association for the Study of the Liver. EASL Clinical Practice Guidelines for the management of pati ents with decompensated cirrhosis. J
Hepatol (2018), https://doi.org/10.1016/j.jhep.2018.03.024

abdominal pain, abdominal tenderness, vomiting, diarrhoea,
ileus; ii) signs of systemic inflammation: hyper or hypothermia,chills, altered white blood cell count, tachycardia, and/ortachypnoea; iii) worsening of liver function; iv) hepatic
encephalopathy; v) shock; vi) renal failure; and, vii) GI bleeding.
However, it is important to point out that SBP may be asymp-tomatic, particularly in outpatients.
245In an observational study
in 239 patients with SBP, delayed diagnostic paracenteseis (>12
h after admission) was associated with a 2.7-fold increase inmortality.
246Peritoneal infection causes an inflammatory reac-
tion resulting in an increased number of neutrophils in ascitic
fluid. Despite the use of sensitive methods, ascites culture isnegative in as many as 60% of patients with clinical manifesta-
tions suggestive of SBP and increased ascites neutrophil count.
33
The gold standard for ascitic neutrophil count is manual micro-
scopy, but it is labour intensive and associated with interob-server variability, time and costs. In most places this has been
substituted with automated counts based on flow cytometry
for counting and differentiating cells. This technique has beendocumented to have high linearity with manual microscopy
and thus sensitivity and specificity close to 100%.
247,248The
greatest sensitivity for the diagnosis of SBP is reached with acut-off neutrophil count of 250/mm
3, although the greatest
specificity is reached with a cut-off of 500 neutrophils/mm3.33
The use of reagent strips cannot be recommended for the rapid
diagnosis of SBP.249Although the presence of bacterial DNA in
plasma and/or ascites is associated with an impairment of circu-
latory function,250there are not enough data to support its use
in clinical practice.251Ascites culture is essential to guide antibi-
otic therapy. Patients with an ascitic fluid neutrophil count ≥250
cells/mm3and negative culture have culture-negative SBP.252
Their clinical presentation is like that of patients with culture-
positive SBP and should be treated in a similar manner. Some
patients have ‘bacterascites’ in which cultures are positive but
there is normal ascitic neutrophil count (<250/mm3).33In some
patients bacterascites is the result of secondary bacterial coloni-
sation of ascites from an extraperitoneal infection. These
patients usually have general symptoms and signs of infection.In other patients, bacterascites is due to the spontaneous coloni-sation of ascites, and can either be clinically asymptomatic or
lead to abdominal pain or fever. While in some patients, partic-
ularly in those who are asymptomatic, bacterascites representsa transient and spontaneously reversible colonisation of ascites,
in other patients, mainly those who are symptomatic, bac-
terascites may represent the first step in the development ofSBP.
33Spontaneous fungal peritonitis is a rare, less recognised
and studied complication, occurring in <5% of cases, but obser-
vational data suggest a worse prognosis.253
Spontaneous bacterial pleural empyema
Infection of a pre-existing hydrothorax, known as spontaneous
bacterial pleural empyema, is uncommon. One study followed3,390 patients with cirrhosis for four years and observed it in
2.4% of the overall population and 16% of patients with pre-
existing hydrothorax, with associated mortality of 38%.
254The
diagnosis is based on pleural fluid analysis obtained by diagnos-tic thoracocentesis. In the largest observational study reported
so far, the diagnosis of spontaneous bacterial empyema was
established when the pleural fluid analysis showed a positiveculture and more than 250 neutrophils/mm
3or a negative cul-
ture and more than 500 neutrophils/mm3, in the absence of lung
infection.255Pleural fluid culture in blood culture bottles waspositive in 75% of cases.255Spontaneous bacterial pleural
empyema was associated with SBP in /C2450% of cases.255
Secondary bacterial peritonitis
A small proportion ( /C245%) of patients with cirrhosis may develop
peritonitis due to perforation or inflammation of an intra-abdominal organ, a condition known as secondary bacterialperitonitis.
256The differentiation of this condition from SBP is
important. Secondary bacterial peritonitis should be suspected
in patients who have localised abdominal symptoms or signs,presence of multiple organisms on ascitic culture, very high
ascitic neutrophil count and/or high ascitic protein concentra-
tion, or in those patients with an inadequate response to ther-apy.
256Patients with suspected secondary bacterial peritonitis
should undergo prompt computed tomography (CT) scanning
and early consideration for surgery.
Recommendations
/C15A diagnostic paracentesis should be carried out in all
patients with cirrhosis and ascites without delay at hos-pital admission to rule out SBP. A diagnostic paracentesis
should also be performed in patients with GI bleeding,
shock, fever or other signs of systemic inflammation, GIsymptoms, as well as in patients with worsening liverand/or renal function, and hepatic encephalopathy
(II-2;1) .
/C15The diagnosis of SBP is based on neutrophil count in asci-
tic fluid of >250/mm
3(II-2;1) . Neutrophil count is deter-
mined by microscopy, but can be substituted with a flowcytometry based automated count. The use of reagent
strips has no clear evidence to support it in routine prac-
tice(II-2;1) .
/C15Although ascitic fluid culture positivity is not a pre-
requisite for the diagnosis of SBP, culture should beperformed in order to guide antibiotic therapy (II-2;1) .
/C15Blood cultures should be performed in all patients with
suspected SBP before starting antibiotic treatment (II-2;1) .
/C15Patients with bacterascites (neutrophil count less than
250/mm
3but positive bacterial culture) exhibiting signs
of systemic inflammation or infection should be treated
with antibiotics (II-2;1) . Otherwise, the patient should
undergo a second paracentesis. If the culture resultscome back positive again, regardless of the neutrophilcount, the patient should be treated (III;1) .
/C15The diagnosis of spontaneous bacterial pleural empyema
should be based on positive pleural fluid culture and
increased neutrophil count of >250/mm
3or negative
pleural fluid culture and a neutrophil count of >500/mm
3in the absence of pneumonia (II-2;1) .
/C15Secondary bacterial peritonitis should be suspected in
case of multiple organisms on ascitic culture, very high
ascitic neutrophil count and/or high ascitic protein con-
centration, or in those patients with an inadequateresponse to therapy. Patients with suspected secondarybacterial peritonitis should undergo prompt CT scanning
and early considerations for surgery (III,1) .JOURNAL
OF HEPATOLOGY
Journal of Hepatology 2018 vol. xxx jxxx–xxx 19
Please cite this article in press as: The European Association for the Study of the Liver. EASL Clinical Practice Guidelines for the management of pati ents with decompensated cirrhosis. J
Hepatol (2018), https://doi.org/10.1016/j.jhep.2018.03.024

Management of spontaneous bacterial peritonitis
Empirical antibiotic therapy .Empirical antibiotic therapy must
be initiated immediately after the diagnosis of SBP.33Potentially
nephrotoxic antibiotics ( i.e., aminoglycosides) should not be
used as empirical therapy.76In the 1990 s, cefotaxime, a third-
generation cephalosporin, was extensively investigated inpatients with SBP because at that time it covered most causative
organisms and because of its high ascitic fluid concentrations
during therapy.
1,33Infection resolution was obtained in 77 to
98% of patients. A dose of 4 g/day is as effective as a dose of 8g/day.
257A five-day therapy is as effective as a 10-day treat-
ment.258Alternatively, amoxicillin/clavulanic acid, first given i.
v. then orally, has similar results with respect to SBP resolutionand mortality as cefotaxime
259and at a much lower cost. How-
ever, there is only one comparative study with a small sample
size and results should be confirmed in larger trials. In addition,some concern exists regarding amoxicillin/clavulanic acid as its
use is associated with a high rate of drug induced liver injury
(DILI).
260Administration of i.v. ciprofloxacin for seven days
results in a similar SBP resolution rate and hospital survival ascefotaxime, but at a significantly higher cost.
261However,
switch therapy ( i.e., use of i.v. antibiotic initially, followed by
oral step-down administration) with ciprofloxacin is morecost-effective than i.v. ceftazidime.
262Oral ofloxacin has shown
similar results as i.v. cefotaxime in uncomplicated SBP, without
renal failure, hepatic encephalopathy, GI bleeding, ileus, orshock.
263However, the spread of resistant bacteria in the
healthcare environment during the last two decades has led to
an alarming increase in the number of infections caused bymulti-drug resistant organisms (MDROs)
264that are defined
by an acquired non-susceptibility to at least one agent in three
or more antimicrobial categories.265Patients with advanced cir-
rhosis are highly susceptible to the development of infectionscaused by MDROs, because they require repeated hospitalisa-
tions, are often submitted toinvasive procedures and are fre-
quently exposed to antibiotics, either as prophylaxis or astreatment. All these factors are well kown risk factors for the
development of infections sustained by MDROs.
266Bacterial
resistance increases four fold the risk of mortality of SBP.267In
particular, nosocomial SBP has been associated with multi-drug
resistance and poor outcomes.266The landscape of bacterial
resistance is continuously changing and challenging recommen-dations for antibiotics. Thus, it is crucial to separate commu-nity-acquired SBP from health care-associated and nosocomial
SBP
6,266–268and to consider both the severity of infection and
the local resistance profile in order to decide the empiricalantibiotic treatment of SBP. Piperacillin/tazobactam has been
recommended as the primary approach for health care and
nosocomial SBP in areas with low prevalence of infections sus-tained by MDROs. On the contrary meropenem alone or/and
combined with glycopeptides or with daptomycin has been sug-
gested as the primary approach for health care-associated SBPwhen severe, or in areas with high prevalence of MDROs, and
for nosocomial SBP in general.
6,266,268,269Regarding the severity
of infection, it should be highlighted that, recently, the new cri-teria for the definition of sepsis, namely qSOFA and Sepsis-3
270
have been validated in patients with cirrhosis and bacterialinfections, proving that they are more accurate than thoserelated to the systemic inflammatory response syndrome inpredicting hospital mortality.
271Accordingly, a new algorithm
has been proposed for the application of qSOFA and Sepsis-3
in the management of cirrhotic patients ( Fig. 3 ). Some more
detailed recommendations on the empirical antibiotic treat-
ment of SBP based on the severity and the environment of the
infection as well as on local resistance profiles are provided(Fig. 4 ). A randomised trial with 32 nosocomial episodes of
SBP, found meropenem plus daptomycin more effective
(86.7%) than ceftazidime (25%) to manage SBP, defined as>25% decrease of neutrophil count at 48 h and to <250/mm
3
at day seven.243If ascitic fluid neutrophil count fails to decrease
to less than 25% of the pretreatment value after two days of
If baseline SOFA score available?
Apply sepsis-3 criteria
Poor outcome
Patient con need transfer to ICUYes
Positive NegativeNo
Apply sepsis-3 criteria and qSOFA
Sepsis-3 and
qSOFA positiveSepsis-3 and
qSOFA negative
Good outcomeSepsis-3 positive
and qSOFA negativeGood outcome
Grey zone
Monitoring SOFA
score is required
Fig. 3. Algorithm for the application of qSOFA and Sepsis-3 criteria in patients with cirrhosis and bacterial infections (adapted form Ref. 271).ICU,
intensive care unit.Community-acquired
SBP or SBENosocomial
SBP or SBE
§SBP or SBE
Healthcare-associated
SBP or SBE
§AREA DEPENDENT:
Like nosocomial
infections if high
prevalence of MDRO
or sepsis Carbapenem alone or
+ daptomycin,
vancomycin or
linezold# if high
prevalence of MDR
Gram+ bacteria or
sepsis 3rd generation
cephalosporin or
piperacillin-tazobactam
Fig. 4. Recommended empirical antibiotic treatment of SBP or SBE
(adapted from Ref. 6).SBE, spontaneous bacterial empyema; SBP, sponta-
neous bacterial peritonitis; MDRO, multidrug resistant organism.Clinical Practice Guidelines
20 Journal of Hepatology 2018 vol. xxx jxxx–xxx
Please cite this article in press as: The European Association for the Study of the Liver. EASL Clinical Practice Guidelines for the management of pati ents with decompensated cirrhosis. J
Hepatol (2018), https://doi.org/10.1016/j.jhep.2018.03.024

antibiotic treatment, there is a high likelihood of failure to
respond to therapy.33This should raise the suspicion of an infec-
tion caused by bacteria resistant to antibiotic therapy, indicat-ing the need for modification of antibiotic treatment according
toin vitro sensitivity or on an empirical basis, or the presence
of ‘secondary peritonitis’. In this context, it should be high-lighted that the progressive increase of the use of carbapenems
because of the worldwide pandemic of extended spectrum beta-
lactamases (ESBLs) producing Enterobacteriaceae has promoted
the emergence of carbapenem-resistant Enterobacteriaceae . This
implies a potential shift from MDR bacteria to extensively drug
resistance (XDR) bacteria defined by a non-susceptibility to atleast one agent in all but two or fewer antimicrobial categories
or to pandrug resistance (PDR) bacteria defined by a non-sus-
ceptibility to all agents in all antimicrobial categories.
265The
shift requires an active surveillance in patients at risk, in orderto identify patients who are colonised or infected by these
clones and prevent their dissemination. The shift may also seri-
ously affect the effectiveness of the broadest spectrum empiri-cal antibiotic treatment among those previously
recommended for SBP and infections other than SBP. Carbapen-
emase-producing and carbapenem-resistant non-carbapene-mase-producing Enterobacteriaceae can be treated with
tigecycline or with the combination of tigecycline at high doses
and a carbapenem in continuous infusion. Addition of i.v. col-istin could be necessary in severe infections. Severe infections
caused by Pseudomonas aeuruginosa resistant to carbapenems
and quinolones usually require the combination of i.v.amikacin/tobramycin or colistin plus a carbapenem/ceftazidime(needed as synergic antibiotics despite antibiotic resistance).
Vancomycin resistant Enterococci should be treated with line-
zolid, daptomycin or tigecycline. All this means reintroducinginto clinical practice antibiotics known to be highly nephrotoxic
in patients with cirrhosis. It follows that serum levels of amino-
glycosides and vancomycin must be monitored closely in thesepatients, to decrease the risk of renal failure. The shift from MDR
to XDR bacteria re-emphasises the interest of the pharmaceuti-
cal industry for the development of new antibiotics. Several newglycopeptides such as oritavancin, new oxazolidinones such astedizolid phosphate, new cephalosporins, such as ceftaroline and
ceftobiprole and razupenem, a new carbapemen, display extended
activity against gram-positive bacteria including vancomycin-resistant Enterococci. In contrast, few newly developed
antibiotics are active against gram-negative MDROs. Temocillin,
a derivative of ticarcillin, is effective against organismsproducing ESBLs. Among cephalosporin-betalactamase inhibitor
combinations, ceftazidime/avibactam and ceftolozane/tazobac-
tam represent further new alternatives to carbapenems for thetreatment of patients with infections sustained by ESBL
producing, carbapenem-resistant Enterobacteriaceae and Pseu-
domonas aeruginosa . However, there are currently no data
regarding the clinical use of these drugs in cirrhosis.
266
Recommendations
/C15Empirical i.v. antibiotics should be started immediately
following the diagnosis of SBP (II-2;1) .
/C15Environment (nosocomial vs.community acquired), local
bacterial resistance profiles and severity of infectionshould guide empirical antibiotic treatment (I;1)./C15Third-generation cephalosporins are recommended as
first-line antibiotic treatment for community-acquired
SBP in countries with low rates of bacterial resistance(I;1). In countries with high rates of bacterial resistance
piperacillin/tazobactam or carbapenem should be con-
sidered (II-2;1) .
/C15Healthcare associated and nosocomial SBP is more likely
to harbour resistance to antibiotics. Piperacillin/tazobac-
tam should be given in areas with low prevalence ofmulti-drug resistance while carbapenem should be used
in areas with high prevalence of ESBL producing Enter-
obacteriaceae . Caarbapenem should be combined with
glycopeptides or daptomycin or linezolid in areas with
high prevalence of gram positive MDR bacteria (I;1).
/C15Severe infections sustained by XDR bacteria may require
the use of antibiotics known to be highly nephrotoxic in
patients with cirrhosis, such as vancomycin or aminogly-
cosides. In these cases, patients’ plasma level should bemonitored in accordance with local policy thresholds
(III;1) .
/C15De-escalation according to bacterial susceptibility based
on positive cultures is recommended to minimise resis-
tance selection pressure (II-2;1) .
/C15The efficacy of antibiotic therapy should be checked with
a second paracentesis at 48 h from starting treatment.Failure of first-line antibiotic therapy should be sus-pected if there is worsening of clinical signs and symp-
toms and/or increase or no marked reduction in
leucocyte count (at least 25%) in 48 h (II-2;1) .
/C15The duration of treatment should be at least 5–7 days
(III;1) .
/C15Spontaneous bacterial empyema should be managed
similarly to SBP (II-2;2) .
Intravenous albumin in patients with spontaneous bacterial peri-
tonitis .SBP without septic shock may precipitate deterioration
of circulatory function with severe liver failure, hepatic
encephalopathy, and type 1 HRS and has approximately 20%
hospital mortality despite infection resolution.272A ran-
domised, controlled study in patients with SBP treated withcefotaxime showed that albumin (1.5 g/kg body weight at diag-
nosis, followed by 1 g/kg on day three) significantly decreased
the incidence of type 1 HRS (from 30% to 10%) and reduced mor-tality from 29% to 10% compared with cefotaxime alone. Treat-
ment with albumin was particularly effective in patients with
baseline serum bilirubin ≥68
lmol/L (4 mg/dl) or SCr ≥88
lmol/L (1 mg/dl). It is unclear whether i.v. albumin is useful
in patients with baseline bilirubin <68 lmol/L and creatinine
<88lmol/L, as the incidence of type 1 HRS in patients meeting
these criteria was very low in the two treatment groups (7%without albumin and 0% with albumin).
272The application of
the schedule of this therapeutic option should be implemented
in clinical practice.273Non-randomised studies in patients with
SBP also show that the incidence of renal failure and death are
very low in patients with moderate liver failure and without
renal dysfunction at diagnosis of SBP, so albumin is probablynot necessary.
274JOURNAL
OF HEPATOLOGY
Journal of Hepatology 2018 vol. xxx jxxx–xxx 21
Please cite this article in press as: The European Association for the Study of the Liver. EASL Clinical Practice Guidelines for the management of pati ents with decompensated cirrhosis. J
Hepatol (2018), https://doi.org/10.1016/j.jhep.2018.03.024

Recommendation
/C15The administration of albumin (1.5 g/kg at diagnosis and
1 g/kg on day 3) is recommended in patients with SBP (I;1).
Prophylaxis of SBP
Since most episodes of SBP are thought to result from the translo-
cation of enteric gram-negative bacteria, the ideal prophylacticagent should be safe, affordable and effective at decreasing the
amounts of these organisms fro m the gut while preserving the
protective anaerobic flora (selective intestinal decontamina-
tion).
267Given the high cost and inevitable risk of developing
resistant organisms, the use of p rophylactic antibiotics must be
strictly restricted to patients at high risk of SBP.267Three high-risk
patient populations have been iden tified: i) patients with acute GI
haemorrhage; ii) patients with low total protein content in ascitic
fluid and no prior history of SBP (primary prophylaxis), and iii)
patients with a previous history of SBP (secondary prophylaxis).275
Primary prophylaxis in patients with low total protein content in
ascitic fluid without prior history of SBP .Cirrhotic patients with
low ascitic fluid protein concentration (<10 g/l) and/or highserum bilirubin levels are at high risk of developing a first epi-sode of SBP.
267Several studies have evaluated prophylaxis with
norfloxacin in patients without prior history of SBP.267Fernandez
et al. randomised 68 patients with cirrhosis and low ascites pro-
tein levels (<15 g/l) with advanced liver failure (Child-Pugh score
≥9 points with serum bilirubin level ≥3 mg/dl) or impaired renal
function (SCr level ≥1.2 mg/dl, blood urea nitrogen level ≥25 mg/dl,or serum sodium level ≤130 mEq/L) to receive norfloxacin
(400 mg/day for 12 months) or placebo.
276Norfloxacin signifi-
cantly improved the three-month probability of survival (94%vs.62%; p= 0.03) but at one year the difference in survival
was not significant (60% vs.48%; p= 0.05). Norfloxacin adminis-
tration significantly reduced the one-year probability of developingSBP (7% vs.61%) and HRS (28% vs.41%). In a double-blind pla-
cebo-controlled trial, 100 patients with ascitic fluid total protein
level <15 g/l were randomised to ciprofloxacin (500 mg/day for
12 months) or placebo.
277The probability of survival at one year
was higher in patients receiving ciprofloxacin (86% vs.66%;
p<0.04). Meta-analyses of all the trials together or including only
pure primary prophylaxis support a significant preventive effectagainst SBP (RR 0.2; 95% CI 0.07–0.52; p= 0.001).
278,279The
survival benefit is most pronounced at three months (94% vs.
62%, p= 0.003) and seems to decrease over time and may be lost
after 12 months follow-up (RR 0.65; 95% CI; 0.41–1.02).280
Recommendations
/C15Primary prophylaxsis with norfloxacin (400 mg/day) in
patients with Child-Pugh score ≥9 and serum bilirubin
level ≥3 mg/dl, with either impaired renal function or
hyponatraemia, and ascitic fluid protein lower than 15g/L is recommended (I;1).
/C15Norfloxacin prophylaxis should be stopped in patients
with long-lasting improvement of their clinical condi-tion and disappearance of ascites (III;1) .Patients with prior SBP
In patients who survive an episode of SBP, the cumulative recur-
rence rate at one year is approximately 70%.
33Probability of
survival at one year after an episode of SBP is 30–50% and fallsto 25–30% at two years. Therefore, patients recovering from an
episode of SBP should be considered for LT. There is only one
randomised, double-blind, placebo-controlled trial of nor-floxacin (400 mg/day orally) in patients who had a previous epi-
sode of SBP.
281Treatment with norfloxacin reduced the
probability of recurrence of SBP from 68% to 20%. In an open-label, randomised study comparing norfloxacin 400 mg/day to
rufloxacin 400 mg/week in the prevention of SBP recurrence,
the one-year probability of SBP recurrence was 26% and 36%,respectively ( p= 0.16).
282Norfloxacin was more effective in
the prevention of SBP recurrence due to Enterobacteriaceae (0%
vs.22%, p= 0.01). The use of intermittent ciprofloxacin has been
associated with a higher rate of quinolone-resistant organismsand should be avoided.
282,283It is uncertain whether prophy-
laxis should be continued without interruption until LT or death
in all patients with prior SBP, or if treatment could be discontin-ued in patients showing an improvement of liver disease. Many
patients receive rifaximin to prevent recurrent episodes of
HE.
284However, rifaximin may also be effective against
recurrent SBP.285There are no data to guide new indications
for primary or secondary prophylaxis of SBP among patients
already on rifaximin. More in detail, it is not known whethernorfloxacin prophylaxis should be started in patients being trea-ted with rifaximin for prevention of recurrent HE. Likewise, it is
not known whether norfloxacin prophylaxis should be stopped
in patients who would require rifaximin to prevent HE. Prospec-tive studies are required to investigate the potential benefits
and side effects of combined therapy with norfloxacin and
rifaximin.
Recommendations
/C15The administration of prophylactic Norfloxacin (400
mg/day, orally) is recommended in patients who recoverfrom an episode of SBP (I;1).
/C15Despite some promising evidence, at present, rifaximin
cannot be recommended as an alternative to norfloxacinfor secondary prophylaxis of SBP (I;2). Thus, at present,
no recommendation can be given to guide primary or
secondary prophylaxis of SBP among patients alreadyon rifaximin for the prevention of recurrent HE.
/C15Patients who recover from SBP have a poor long-term
survival and should be considered for LT (II-2,1) .
/C15Since it has been suggested that PPI may increase the
risk for the development of SBP, its use should berestricted to those with a clear indication (II-2,1) .
Concomitant medications
Very frequently PPIs are used in patients with cirrhosis, which
may increase the risk of SBP. Indications for long-term useshould be carefully assessed and PPIs discontinued when possi-
ble.
286,287NSBBs may be detrimental in end-stage liver disease
with haemodynamic derangement, patients should be moni-tored closely and doses adjusted or drug discontinued if con-
traindications occur.
168,190,288Probiotics have been assessed asClinical Practice Guidelines
22 Journal of Hepatology 2018 vol. xxx jxxx–xxx
Please cite this article in press as: The European Association for the Study of the Liver. EASL Clinical Practice Guidelines for the management of pati ents with decompensated cirrhosis. J
Hepatol (2018), https://doi.org/10.1016/j.jhep.2018.03.024

combination therapy with norfloxacin in one randomised trial
in a mixed group of patients on primary and secondary preven-tion of SBP. No additional benefits were demonstrated.
289
Infections other than SBP
Prevalence, diagnosis and impact on prognosis
Non-SBP infections are frequent in patients with cirrhosis and
present or develop during hospitalisation in 25–30% of patients.
The most frequent infections other than SBP are: urinary tract,pneumonia, skin and soft tissue infections, and bacter-
aemia.
242,290They constitute a heterogeneous group regarding
clinical course and prognosis. Non-SBP infections increase theodds ratio for death by 3.75 and are associated with a 30%
one-month and 63% 12-month mortality.
291Endocarditis, sec-
ondary peritonitis, pneumonia and bacteraemia have worseprognoses. The combination of data on liver and renal dysfunc-tion and the type of infection enables the identification of
patients with poor prognosis.
290In particular, non-SBP infec-
tions, as well as SBP, are known as common precipitating factorsfor ACLF.
3An early diagnosis of all these infections and of SBP is
a crucial step in the management of patients with cirrhosis.
Since the presentation and the initial course of any bacterialinfection may be subtle and not very specific, clinical suspicion
is important. Indeed, all inpatients with cirrhosis should be con-
sidered as potentially infected until proven otherwise. There-fore, a complete work-up should be carried out at admissionand at any time during the hospital stay when clinical deterio-
ration occurs.
6In addition, close microbiological surveillance is
needed in patients who are at risk of developing infectionscaused by methicillin resistant organisms. C reactive protein
and procalcitonin can be used for detecting infection and to
define the severity of the infection,
6while their use in the
stewardship of antibiotic treatment deserves further
investigation.292To optimise the empirical antibiotic treatment,
it is quite important to distinguish among community acquired,health care associated and nosocomial infections. Mortality for
nosocomial infections is higher (25–48%) than for community-
acquired infection (7–21%) since they are more commonlysustained by MDR bacteria.
6,266,268Like in SBP, there is an
increasing challenge of resistant bacteria among non-SBP infec-
tions. Among 312 patients with cirrhosis and blood stream
infections gram-negative bacteria, gram-positive bacteria andCandida were the cause of blood stream infections episodes in
53%, 47% and 7% of the cases, and 31% of the infections were
caused by MDR bacteria.
293
Management of infections other than SBP
In a randomised trial 94 patients with cirrhosis and infections
(most prevalent were urinary tract infections [46%], SBP [22%],and pneumonia [19%]) were randomised to a broad-spectrum
antibiotic regimen or a standard regimen. In–hospital mortality
was significantly higher in the standard than in the broad-spec-trum group (25% vs.6%;p= 0.01).
294Some more specific sugges-
tions on the empirical antibiotic treatment of infections other
than SBP based on the type, the severity and the environmentof the infection as well as on local resistance profiles are given(Figs. 5–7 ). In patients who fail to respond to a broad-spectrum
antibiotic treatment a fungal infection, including fungal SBP
295
should be suspected and investigated.296
Finally, in two randomised trials, concomitant albumin may
protect against deterioration in renal and circulatory func-
tion.297,298However, albumin did not improve survival and thusit cannot be recommended. The issue of the management of infec-
tions sustained by XDR bacteria has been previously developed.
Recommendations
/C15Infections other than SBP are frequent and associated
with increased mortality. Hospitalised patients with cir-rhosis should be assessed and monitored closely for thepresence of infections to enable early diagnosis and
appropriate treatment (II-1;1) .
/C15Empirical antibiotic therapy should be commenced
promptly at suspicion of infection (II-1;1) .
/C15The choice of empirical antibiotic therapy should be
based on several factors including: environment (noso-
comial vs. health care associated or community
acquired), local resistance profiles, severity and type of
infection (I;1).Cellulitis
Community-acquired
cellulitisHealthcare-
associated cellulitisNosocomial
cellulitis
Piperacillin-
tazobactam
or 3rdgeneration
cephalosporin +
oxacillin AREA DEPENDENT:
Like nosocomial
infections if high
prevalence of MDROs
or if sepsis3rdgeneration
cephalosporin or
meropemen + oxacillin
or glycopeptides or
daptomycin or
linezolid*
Fig. 5. Recommended empirical antibiotic treatment of soft tissue infec-
tions (adapted from Ref. 6).MDRO, multidrug resistant organism.
Pneumonia
Community-acquired
pneumoniaHealthcare-associated
pneumoniaNosocomial
pneumonia
Piperacillin-tazobactam
or ceftriaxone +
macrolide or
levofloxacin or
moxifloxacinAREA DEPENDENT:
Like nosocomial
infections if high
prevalence of MDROs§
or if sepsisCeftazidime or
meropemen§+
levofloxacin ±
glycopeptides or
linezolid#
Fig. 6. Recommended empirical antibiotic treatment of pneumonia
(adapted from Ref. 6).MDRO, multidrug resistant organism.
UTI
Community-acquired
UTIHealthcare-
associated UTINosocomial
UTI
UNCOMPLICATED:
ciprofloxacin or
cotrimoxazole
IF SEPSIS: 3rd
generation cephalospo-
rin or piperacillin-
tazobactamAREA DEPENDENT:
Like nosocomial
infections if high
prevalence of MDR§
or if sepsisUNCOMPLICATED:
fosfomycin or
nitrofurantoin
IF SEPSIS:
meropemen +
teicoplanin or
vancomycin§ #
Fig. 7. Recommended empirical antibiotic treatment of UTI (adapted from
Ref. 6).MDR, multidrug resistant; UTI, urinary tract infection.JOURNAL
OF HEPATOLOGY
Journal of Hepatology 2018 vol. xxx jxxx–xxx 23
Please cite this article in press as: The European Association for the Study of the Liver. EASL Clinical Practice Guidelines for the management of pati ents with decompensated cirrhosis. J
Hepatol (2018), https://doi.org/10.1016/j.jhep.2018.03.024

/C15In the context of high bacterial resistance to antibiotics,
carbapenem alone or in combination with other antibi-
otics proved to be superior to third-generation cephalos-porins in healthcare associated infections other than SBP,
and therefore, should be preferred (I;1).
/C15Severe infections sustained by XDR bacteria can require
the use of antibiotics known to be highly nephrotoxic
in patients with cirrhosis such as vancomycin or amino-
glycosides. In these cases, patients’ plasma level shouldbe monitored in accordance with local policy thresholds
(III;1) .
/C15Routine use of albumin is not recommended in infec-
tions other than SBP (I;1).
Prophylaxis of infections other than SBP
There is preliminary evidence that in patients with Child-Pugh
class C, norfloxacin administration can reduce the risk of infec-tions and can decrease six-month mortality. However, moredata are needed before a recommendation can be made.
21
Renal impairment
Definition and diagnosis
Renal impairment in patients with cirrhosis was defined more
than 30 years ago by an SCr value ≥1.5 mg/dl because this value
was considered an index of GFR ≤40 ml/min.32The use of SCr in
the evaluation of renal function in patients with cirrhosis, hasseveral well-known limitations. However, the diagnosis of renal
dysfunction in liver disease is still based on it.
32,299The diagno-
sis should be based on different diagnostic categories includingchronic kidney disease (CKD) and acute renal failure (ARF).
When only based on reduction of GFR, the diagnosis of CKD in
patients with cirrhosis is still challenging, because all the SCr-based equations that have been proposed overestimate GFR inpatients with cirrhosis.
300–304It can be reasonably assumed that
patients with decompensated cirrhosis frequently have CKD
caused by certain comorbidities ( i.e.diabetes, arterial hyperten-
sion) and/or specific causes ( i.e.IgA nephropathy, virus-inducedglomerulopathy),305however the prevalence of CKD in this pop-
ulation is still unknown. ARF is a common complication inpatients with decompensated cirrhosis.
306Historically, the diag-
nosis was based on an increase in SCr of 50% from baseline to a
final value >1.5 mg/dl (133 lmol/L).1,32,307Recently, the term
ARF was replaced by AKI,308–310irrespectively of its different
types. AKI is now defined, as proposed by the Kidney Disease
Improving Global Outcomes (KDIGO) group,310as either an
absolute increase in SCr of more than or equal to 0.3 mg/dl(≥26.4
lmol/L) in less than 48 h, or by a percentage increase
in SCr of more or equal to 50% (1.5-fold from baseline) in less
than seven days. A new staging system was also introduced,mainly based on the percentage increase of SCr from baseline
(Table 7 ), either at the time of the first fulfillment of the KDIGO
criteria (initial stage) or at the peak value of SCr during hospital-isation in case of progressive AKI (peak stage).
310Based on the
staging system and according to the results of several prospec-
tive studies,311–317a new algorithm for the management of AKI
in patients with cirrhosis has been proposed318(Fig. 8 ). Recent
studies have suggested that in patients with cirrhosis, in AKI
stage 1, SCr <1.5 mg/dl is associated with a worse outcome than
an SCr ≥1.5 mg/dl.313,314,317Thus, in contrast with the KDIGO
staging system, it has been proposed to distinguish between a
stage 1A (SCr <1.5 mg/dl) and a stage 1B (SCr ≥1.5 mg/dl) within
AKI stage 1.313,314,317It should be highlighted that the KDIGO
criteria also include criteria based on urinary output in the diag-
nosis of AKI ( Fig. 9 ).310These criteria were not considered by the
recent International Club of Ascites (ICA) consensus because (a)these patients are frequently oliguric with avid sodium reten-tion, despite a relatively normal GFR, (b) they may have an
increased urine output because of diuretics, and (c) on a regular
ward, urine collection is often inaccurate and alwaysuntimely.
318However, these criteria may also be applied when-
ever a patient with cirrhosis requires a bladder catheter. The
definition of baseline SCr used in the KDIGO criteria is crucialsince it has been observed that about 25–30% of episodes of
AKI occur before hospitalisation, representing the so-called
‘‘community-acquired AKI”. Ideally, ‘‘community-acquired AKI”should be diagnosed at the time of hospital admission, requir-ing, according to the KDIGO criteria, an SCr value dated within
Table 7 .International Club of Ascites (ICA-AKI) new definitions for the diagnosis and management of acute kidney injury in patients with cirrhosis.
Subject Definition
Baseline sCr A value of sCr obtained in the previous three months, when available, can be used as baseline sCr. In patients with more than one value withi n
the previous three months, the value closest to the admission time to the hospital should be used
In patients without a previous sCr value, the sCr on admission should be used as baseline.
Definition of
AKI- Increase in sCr ≥0.3 mg/dl (≥26.5 lmol/L) within 48 h; or,
– A percentage increase sCr ≥50% which is known, or presumed, to have occurred within the prior seven days
Staging of
AKI- Stage 1: increase in sCr ≥0.3 mg/dl (≥26.5 lmol/L) or an increase in sCr ≥1.5-fold to 2-fold from baseline;
– Stage 2: increase in sCr >2-fold to 3-fold from baseline;
– Stage 3: increase of sCr >3-fold from baseline or sCr ≥4.0 mg/dl (353.6 lmol/L) with an acute increase ≥0.3 mg/dl (≥26.5 lmol/L) or initiation
of renal replacement therapy
Progression
of AKIProgression Regression
Progression of AKI to a higher
stage and/or need for RRTRegression of AKI to a lower stage
Response to
treatmentNo response Partial response Full response
No regression of AKI Regression of AKI stage with a reduction of sCr to ≥0.3
mg/dl (≥26.5 lmol/L) above the baseline valueReturn of sCr to a value within 0.3 mg/dl
(≥26.5 lmol/L) of the baseline value
AKI, acute kidney injury; sCr, serum creatinine; RRT, renal replacement therapy.Clinical Practice Guidelines
24 Journal of Hepatology 2018 vol. xxx jxxx–xxx
Please cite this article in press as: The European Association for the Study of the Liver. EASL Clinical Practice Guidelines for the management of pati ents with decompensated cirrhosis. J
Hepatol (2018), https://doi.org/10.1016/j.jhep.2018.03.024

the last week before admission. This point is so crucial for the
application of the KDIGO criteria that it has been suggested thatan SCr value be calculated when not available the seven daysprior to admission. The baseline SCr can be calculated by inver-
sely applying the formulas that are used to calculate the esti-
mated GFR, considering normal values of GFR of 75 ml/min.
319
Whilst an imputed SCr is accepted in the general population,
it can not be used in patients with cirrhosis.320Indeed, all
SCr-based formulas overestimate the true GFR in these patientsleading to an overestimation of the baseline SCr and thus under-
estimating the prevalence of AKI on admission.
320Therefore, it
has been proposed that not only the value obtained in the lastseven days, but also that within the last three months be con-
sidered as a baseline value of SCr in patients with cirrhosis
(Table 7 ). In addition, an SCr value obtained within the last three
months is the reference to define acute kidney disease (AKD), athird category of renal impairment, along with AKI and CKD,
which has been recently proposed in KDIGO recommendations.
AKD is clearly a distinct category with different outcome,whether or not it is associated with AKI. AKD is defined by a
GFR <60 ml/min/1.73 m
2for less than three months, or a
decrease in GFR ≥35% for less than three months, or an increasein SCr <50% within the last three months ( Table 8 ). However, no
data exist about the prognostic impact of AKD, with or without
AKI, in patients with cirrhosis. Thus, waiting for these data, itseems even more justified to make the diagnosis of AKI in
patients with cirrhosis on an increase in SCr ≥50% during the
last three months. This assumption may also facilitate the diag-nosis of AKI overlapping CKD.Recommendations
/C15In patients with liver diseases, even a mild increase in
SCr should be considered since it may underlie a marked
decrease of GFR (II-2;1) .
/C15The first step to be addressed in the diagnostic process is
to establish if the patient has a CKD, AKD or AKI as well
as an overlap between these diagnostic categories(II-2;1) .
/C15The diagnosis of CKD should be based on a GFR <60 ml/
min/1.73 m
2estimated by SCr-based formulas, with or
without, signs of renal parenchymal damage (protein-
uria/haeamturia/ultrasongraphy abnormalities) for at
least three months (II-2;1) .
/C15The diagnostic process should be completed by staging
CKD, which relies on GFR levels, and by investigatingits cause. It should be highlighted that any SCr based
formula overestimates GFR in patients with cirrhosis
(II-2, 1) .Resolution
Close follow upResponse ?
NO
NO YESPersistance Progression YESClose monitoring
Remove risk factors (withdrawal of nephrotoxic drugs, vasodilators and
NSAIDs, taper/withdraw diuretics and β-blockers, expand plasma
volume, treat infections* when diagnosed)Initial AKI# stage 1a°
Further treatment of
AKI decided on a
case-by-case basisInitial AKI# stage >1a°
Withdrawal of diuretics (if not yet applied) and volume
expansion with albumin (1 g/kg) for 2 days
Does AKI meet criteria of HRS ?
Specific
treatment for
other AKI
phenotypesVasoconstrictors
and albumin#AKI at the first fulfilling of KDIGO criteria
Fig. 8. Algorithm for the management of AKI in patients with cirrhosis (adapted from Ref. 318).AKI, acute kidney injury; HRS, hepatorenal syndrome;
NSAID, non-steroidal anti-inflammatory.
KDIGO urine output criteria = an urinary output <0.5 ml/kg B.W./h/ x 6-12 h
Stage Serum creatinine criteria
1° An urinary output <0.5 ml/kg B. W./h x 6-12 h
2° An urinary output <0.5 ml/kg B.W./h x 12 h
3° An urinary output <0.5 ml/kg B.W./h x 24 h or anuria per 12 h
Fig. 9. Criteria based on urinary output for the diagnosis of AKI (adapted
from Ref. 310).AKI, acute kidney injury; B.W., body weight.Table 8 .Definitions of kidney disease.
Definition Functional criteria Structural
criteria
AKI Increase in sCr ≥50% within seven days,
or
increase in sCr ≥0.3 mg/dl within two daysNo criteria
AKD GFR <60 ml/min per 1.73 m2for <3 months,
ordecrease in GFR ≥35% for < 3 months,
or
increase in sCr ≥50 % for < 3 monthsKidney damage
for <3 months
CKD GFR <60 ml/min per 1.73 m
2for ≥3 months Kidney damage
for ≥3 months
AKD, acute kidney disease; AKI, acute kidney injury; CKD, chronic kidney disease;
GFR, glomerular filtration rate; sCr, serum creatinine.JOURNAL
OF HEPATOLOGY
Journal of Hepatology 2018 vol. xxx jxxx–xxx 25
Please cite this article in press as: The European Association for the Study of the Liver. EASL Clinical Practice Guidelines for the management of pati ents with decompensated cirrhosis. J
Hepatol (2018), https://doi.org/10.1016/j.jhep.2018.03.024

/C15In patients with cirrhosis the diagnosis of AKI should be
based on adapted KDIGO criteria, thus, either on an
increase in SCr of >0.3 mg/dl from baseline within 48 h,or an increase of ≥50% from baseline within three
months (II-2,1) .
/C15The staging of AKI should be based on an adapted KDIGO
staging system, thus distinguishing within AKI stage 1,
between AKI stage 1A and AKI stage 1B according to a
value of SCr <1.5 or ≥1.5 mg/dl, respectively (II-2,1) .
Precipitating factors
Infections, diuretic-induced excessive diuresis, GI bleeding,
therapeutic paracentesis without adequate volume expansion,nephrotoxic drugs, and NSAIDs are the other common precipi-tating factors of AKI in patients with cirrhosis.
20,242,306The
nephrotoxicity of contrast agents is still debated in patients
with cirrhosis321but contrast imaging should be performed cau-
tiously, particularly in decompensated cirrhosis or in patients
with known CKD. Finally, the increase in intra-abdominal pres-
sure associated with tense ascites may lead to AKI, by increasingrenal venous pressure.
322–324
Management
The cause of AKI should be investigated as soon as possible, to
prevent AKI progression. However, even in the absence of a defini-
tive recognised cause of AKI, the management should be immedi-
ately started according to the initial stage ( Fig. 2 ). Irrespective of
the stage, diuretics should be discontinued. Similarly, even ifthere are controversial data, beta-blockers should be stopped.
168
Other precipitating factors of AKI should be identified and treated,including screening and treatment of infection, volume expan-sion when appropriate, and discontinuation of all nephrotoxic
drugs, such as vasodilators or NSAIDs.
318Volume replacement
should be used in accordance with the cause and the severity offluid loss. Patients with diarrhoea or excessive diuresis should
be treated with crystalloids, whilst patients with acute GI bleed-
ing should be given packed red blood cells to maintain haemoglo-bin level between 7–9 g/dl.
325In patients with AKI and tense
ascites, therapeutic paracentesis should be associated with albu-
min infusion since it improves renal function.326In case of no
obvious cause and AKI stage >1A, 20% albumin solution at thedose of 1 g of albumin/kg of body weight (with a maximum of
100 g of albumin) for two consecutive days should be given.
307
All other therapeutic options, especially renal replacement ther-
apy (RRT) and kidney transplantation will be discussed in the sec-
tion dedicated to the management of HRS-AKI.
Recommendations
/C15When a diagnosis of AKI is made, its cause should be
investigated as soon as possible to prevent AKI progres-
sion. Even in absence of an obvious cause, the manage-ment should be immediately started. Maximalattention in the screening and treatment of infections
should be carried out (II-2,1) .
/C15Diuretics and/or beta-blockers as well as other drugs
that could be associated with the occurrence of AKI such
as vasodilators, NSAIDs and nephrotoxic drugs should beimmediately stopped (II-2,1) ./C15Volume replacement should be used in accordance with
the cause and severity of fluid losses (II-2,1) .
/C15In case of no obvious cause of AKI, AKI stage >1A or infec-
tion-induced AKI, 20% albumin solution should be used at
the dose of 1 g of albumin/kg of body weight (with a max-imum of 100 g of albumin) for two consecutive days (III,1) .
/C15In patients with AKI and tense ascites, therapeutic para-
centesis should be associated with albumin infusion evenwhen a low volume of ascetic fluid is removed (III,1) .
Types of AKI
All types of AKI can occur in patients with cirrhosis, namely pre-
renal AKI, HRS-AKI, intrarenal or intrinsic AKI, and post-renal
AKI. The most common cause of AKI in hospitalised patientswith decompensated cirrhosis is pre-renal, accounting for
approximately 68% of the cases.
306,327,328Intrarenal-AKI is
mainly represented by acute tubular necrosis (ATN).306Finally,
post-renal AKI is uncommon in decompensated cirrhosis.328
Considering that most cases of pre-renal AKI are resolved by
volume expansion and that post-renal AKI is uncommon, the
key point is to differentiate HRS-AKI from ATN. As describedin the section ‘‘Hepatorenal syndrome”, the concept that HRS
is only a functional injury has been challenged during the last
decade and, thus, the definition of HRS probably has to berevised. In addition, as kidney biopsy is rarely performed in
the setting of AKI in clinical practice, the distinction between
HRS-AKI and ATN is difficult. Recently, novel biomarkers haveemerged in this setting and urinary neutrophil gelatinase-
associated lipocalin (NGAL) is the most promising. Indeed, sev-
eral studies have shown that urinary NGAL, a marker of tubulardamage, could help to determine the type of AKI.
329–335
However, cut-off values differ greatly according to series, there
are overlaps between the different types of AKI and it should be
highlighted that no study has confirmed the diagnosis byreference kidney biopsy. Diagnosis based on a combination of
multiple biomarkers may be interesting but needs further
evaluation.
329,330,332–334
Recommendations
/C15All types of AKI can occur in patients with cirrhosis,
namely pre-renal, HRS, intrinsic, particularly ATN, and
post-renal. Therefore, it is important to differentiateamong them (II-2,1) .
/C15The diagnosis of HRS-AKI is based on revised ICA criteria.
As kidney biopsy is rarely performed in the setting ofAKI, biomarkers should be implemented In clinical prac-
tice among the different biomarkers to date, urinary
NGAL can be used to distinguish between ATN and HRS(II-2;2) .
Prognosis
In patients with decompensated cirrhosis, AKI has a negative
impact on hospital survival according to either the initial stage,
314
or the peak stage.313,317Even transient episodes of AKI are
associated with a negative impact on mid-term survival.315
Nevertheless, a more comprehensive prognostic classificationClinical Practice Guidelines
26 Journal of Hepatology 2018 vol. xxx jxxx–xxx
Please cite this article in press as: The European Association for the Study of the Liver. EASL Clinical Practice Guidelines for the management of pati ents with decompensated cirrhosis. J
Hepatol (2018), https://doi.org/10.1016/j.jhep.2018.03.024

also considering extra-renal organ failures is much more accurate
than the KDIGO criteria for the prognosis in these patients.Finally, looking to the data in the general population, it shouldbe highlighted that the risk for developing CKD is higher in
patients with severe or repeated episodes of AKI.
336Since patients
with decompensated cirrhosis are prone to develop frequentepisodes of AKI, it can be speculated that they are at higher risk
of developing CKD.
Hepatorenal syndrome
Definition, diagnosis and classification
For a long time, HRS has been defined as ‘‘a functional renal fail-
ure caused by intrarenal vasoconstriction which occurs in
patients with end-stage liver disease as well as in patients with
acute liver failure or alcoholic hepatitis”.
32,307Several data chal-
lenge this definition of HRS as well as the classification in type 1and type 2. Firstly, as described below, pathogenesis of HRS
includes both haemodynamic and inflammatory changes. Sec-
ondly, the absence of renal parenchymal damage, defining thefunctional nature, has never been proven by renal biop-
sies.
337,338The absence of significant proteinuria and/or haema-
turia do not rule out renal lesions, particularly tubular andinterstitial lesions.
307In addition, studies assessing novel kidney
biomarkers have shown that tubular damage can occur in
patients with HRS-AKI when HRS is diagnosed according tothe traditional criteria.
328–330,332Finally, it should be noted that
HRS-AKI can occur in patients with underlying CKD. Type 1 and
type 2 were historically defined based on time frame SCr
increase.32,307In the recent revised classification, type 1 HRS
now corresponds to HRS-AKI.318Consequently, type 2 HRS
should now include renal impairment which fulfills the criteria
of HRS but not of AKI, namely non-AKI-HRS (NAKI), and onlyHRS-CKD as previously proposed.
339
Pathophysiology
According to the new theory that has been developed on the
pathophysiology of decompensated cirrhosis,5the view on
HRS has been changed in recent years, moving from the idea
that it was only related to renal hypoperfusion due to macrocir-culatory dysfunction ( i.e.splanchnic arterial vasodilation and
reduction of cardiac output).
192,338The new theory is that the
increased circulating levels of pro-inflammatory cytokines andchemokines
340,341may exercise a direct relevant role in the
development of HRS. Such cytokines have been associated with
renal impairment in patients and in animal models of cirrhosis
with infection.342–345Moving from the concept that AKI and
HRS-AKI are often precipitated by bacterial infection, the new
hypothesis on the pathogenesis of sepsis-induced AKI should
also be considered.346–348This theory proposes that a synergic
interplay of inflammation and microvascular dysfunction is
responsible for the amplification of the signal that PAMPs and
DAMPs exert on proximal epithelial tubular cells. The recogni-tion of this signal and its subsequent spread to all the other
proximal tubular epithelial cells cause a mitochondria-medi-
ated metabolic downregulation and reprioritisation of cell func-tions to favour survival processes above all else.
349The
sacrificed functions include the absorption on the lumen side
of sodium and chloride. The consequent increases of sodium
chloride delivery to the macula densa triggers further intrarenalactivation of the RAAS and thus lowers GFR. Finally, severe
cholestasis may further impair renal function by worsening
inflammation and/or macrocirculatory dysfunction, or bypromoting bile salt-related direct tubular damage.
350,351All
these findings suggest that the pathophysiology of AKI, and par-ticular of HRS-AKI, in patients with decompensated cirrhosisseems more complex than previously hypothesised, supporting
the concept that AKI-HRS is not purely functional in nature.
Management
The non-specific management of AKI as been previously
described. Thus, in this section, drug therapy, TIPS, RRT, LT
and simultaneous liver and kidney transplantation (SLK) willbe considered.
Drug therapy
.Once the diagnosis of HRS-AKI has been made,
patients should promptly receive vasoconstrictive drugs, inassociation with albumin. The rational for using vasoconstric-tors is to counteract the splanchnic arterial vasodilation,
improving renal perfusion.
352Terlipressin, a vasopressin ana-
logue, is the most commonly used. The efficacy of terlipressinplus albumin in the treatment of HRS has been proven in many
studies.
353–360In the most recent studies, rates of response
(complete or partial response) to this treatment range from 64to 76%, with a complete response, from 46 to 56%.
358–360These
response rates must now be evaluated according to the new def-
initions of responses in HRS-AKI recently proposed by the ICA(Table 7 ). In two meta-analyses terlipressin plus albumin was
proven to improve not only renal function but also short-term
survival in patients with HRS.
361,362Terlipressin was initially
proposed to be administered by i.v. boluses at a starting doseof 0.5–1 mg every 4–6 h, progressively increased to a maximum
of 2 mg every 4–6 h in case of a reduction of baseline SCr <
25%.
353–358Adding albumin to terlipressin is more effective
than terlipressin alone.354One possible explanation is that albu-
min, by increasing volaemia, may counteract the decrease in
cardiac output associated with HRS192but also by terli-
pressin.363In addition, antioxidant and anti-inflammatory prop-
erties of albumin may have a beneficial effect.364The dose of
albumin in HRS treatment has not been well established. Stud-ies have suggested adapting the dose according to the level of
central venous pressure (CVP), but there is evidence that CVP
is inaccurate to manage volume expansion and to assess cardiacoutput in patients with cirrhosis. In contrast, CVP may be help-ful to prevent circulatory overload. Albumin has been used
intravenously at the mean dose of 20–40 g/day. Treatment
should be maintained until a complete response (SCr below1.5 mg/dl) or for a maximum of 14 days either in case of partial
response (decrease of SCr ≥50 with a final value still higher than
1.5 mg/dl) or in case of non-response. More recently, continuousi.v. infusion of terlipressin at an initial dose of 2 mg/day was
proposed,
359,365demonstrating a similar rate of response but
lower adverse effects than the administration of the drug byi.v. boluses.
360Indeed, terlipressin, when administered by con-
tinuous i.v. infusion, has a more stable lowering effect on portal
pressure, even when used at lower doses than those provided by
i.v. boluses.360The most common side effects of terlipressin are
diarrhoea, abdominal pain, circulatory overload and cardiovas-
cular ischaemic complications which have been reported in up
to 45–46% of patients when the drug was delivered by i.v.boluses.
360The rate of discontinuation because of side effects,
mainly cardiovascular, is around 20%.360Accordingly, a careful
clinical screening including electrocardiogram is recommendedin all patients before starting treatment. Patients can be treated
on a regular ward but the decision to transfer to a higher level ofJOURNAL
OF HEPATOLOGY
Journal of Hepatology 2018 vol. xxx jxxx–xxx 27
Please cite this article in press as: The European Association for the Study of the Liver. EASL Clinical Practice Guidelines for the management of pati ents with decompensated cirrhosis. J
Hepatol (2018), https://doi.org/10.1016/j.jhep.2018.03.024

care should be case based. Recurrent HRS in responders, after
the end of the treatment, has been reported in up to 20% ofcases. Re-treatment is usually effective, however, in some cases,continuous recurrence occurs, thus a long-term treatment with
terlipressin plus albumin and a long-term hospitalisation are
required.
366The possibility of treating some of these patients
outside the hospital has recently been proposed367but even if
promising, further studies are needed. Other vasoconstrictive
drugs include i.v. noradrenaline and oral midodrine plus subcu-taneous or i.v. octretide, both in combination with albumin.
Noradrenaline, given by continuous i.v. infusion at the dose of
0.5–3 mg/h, has been proven to be as effective as terlipressinregarding the increase in mean arterial pressure, the reversal
of renal impairment and one-month survival.
368–371However,
the number of patients treated with noradrenaline remainstoo small to definitively confirm its efficacy. In addition, in con-trast to terlipressin, the use of noradrenaline always requires a
central venous line and, in most countries, the transfer of the
patient to an intensive care unit (ICU). The combination mido-drine plus octreotide, used in countries where terlipressin is
not yet available,
372has been shown to be much less effective
than terlipressin in the treatment of type 1 HRS in a recentRCT.
359Vasoconstrictors, in particular terlipressin, in associa-
tion with albumin, have also been proposed in the treatment
of type 2 HRS. The treatment has been proven to be effectivein most cases but, unfortunately, recurrence after the with-
drawal of treatment is the norm. In addition, there are contro-
versial data about the impact of this treatment on outcomes,especially in candidates for LT.
373,374This may be, at least in
part, due to the suboptimal definition of type 2 HRS, as previ-
ously discussed. The most relevant factors that may impair
the response to vasoconstrictors are: a) the baseline value ofSCr, b) the degree of inflammation and c) the degree of cholesta-
sis.
375–377The finding that the higher baseline values of SCr, the
lower rate of response to terlipressin plus albumin,375probably
reflects the presence of renal parenchymal damage337and rep-
resents one of the main reasons behind the adoption of the
KDIGO criteria for the definition of AKI in patients with cirrho-sis, and the introduction of the new algorithm for its manage-ment. Regarding inflammation, it has been recently shown
that, for the same value of baseline SCr, the rate of response is
related to the number of extra-renal organ failures.
376
Transjugular intrahepatic portosystemic shunts .The use of TIPS
may improve renal function in patients with type 1 HRS.378,379
However, the applicability of TIPS in this clinical setting is usu-ally very limited because, in most patients, TIPS is contraindi-
cated because of severe degree of liver failure. TIPS has been
studied in patients with type 2 HRS
380and in the management
of refractory ascites, frequently associated with type 2 HRS. In
these patients, TIPS has been shown to improve renal function.95,379
Renal replacement therapy .Renal replacement therapy should be
considered in the management of AKI, whatever the type. As faras HRS-AKI, it should be considered in non-responders to vaso-
constrictors. RRT should also be considered in patients with
end-stage kidney disease. The indications for RRT are the samein patients with cirrhosis as in the general population including:
severe and/or refractory electrolyte or acid-base imbalance,Recommendations
/C15Vasoconstrictors and albumin are recommended in all
patients meeting the current definition of AKI-HRS stage>1A, should be expeditiously treated with vasoconstric-
tors and albumin (III;1) .
/C15Terlipressin plus albumin should be considered as the
first-line therapeutic option for the treatment of HRS-
AKI. Telipressin can be used by i.v. boluses at the initial
dose of 1 mg every 4–6 h. However, giving terlipressinby continuous i.v. infusion at initial dose of 2 mg/day
makes it possible to reduce the global daily dose of the
drug and, thus, the rate of its adverse effects. In case ofnon-response (decrease in SCr <25% from the peak value),
after two days, the dose of terlipressin should be increased
in a stepwise manner to a maximum of 12 mg/day (I;1)./C15Albumin solution (20%) should be used at the dose
20–40 g/day. Ideally, apart from routinely monitoring
patients with HRS-AKI, the serial measurement of CVPor other measures of assessing central blood volume,
can help to prevent circulatory overload by optimising
the fluid balance and helping to titrate the dose of albu-min (II-2;1) .
/C15Noradrenaline can be an alternative to terlipressin. How-
ever, limited information is available (I;2).
/C15In contrast to terlipressin, the use of noradrenaline
always requires a central venous line and, in severalcountries, the transfer of the patient to an ICU. Mido-
drine plus octreotide can be an option only when terli-
pressin or noradrenaline are unavailable, but itsefficacy is much lower than that of terlipressin (I;1).
/C15According to the new definition of HRS-AKI, complete
response to the treatment should be defined by a finalSCr within 0.3 mg/dl (26.5
lmol/L) from the baseline
value, while partial response should be defined by the
regression of AKI stage to a final SCr ≥0.3 mg/dl (26.5
lmol/L) from the baseline value (III;1) .
/C15Adverse events related to terlipressin or noradrenaline
include ischaemic and cardiovascular events. Thus, a care-
ful clinical screening including electrocardiogram is rec-
ommended before starting the treatment. Patients can betreated on a regular ward, but the decision to transfer tohigher dependency care should be case based. For the dura-
tion of treatment, it is important to closely monitor the
patients. According to the type and severity of side effects,treatment should be modified or discontinued (I;1).
/C15In cases of recurrence of HRS-AKI upon treatment cessa-
tion, a repeat course of therapy should be given (I;1).
/C15Terlipressin plus albumin is also effective in the treat-
ment of HRS outside the criteria of AKI (HRS-NAKI), for-merly known as HRS type II. Unfortunately, recurrence
after the withdrawal of the treatment is the norm, and
controversial data exists on the impact of the treatmenton long-term clinical outcome, particularly from the per-
spective of LT. As such, vasoconstrictors and albumin are
not recommended in this clinical scenario (I;1).Clinical Practice Guidelines
28 Journal of Hepatology 2018 vol. xxx jxxx–xxx
Please cite this article in press as: The European Association for the Study of the Liver. EASL Clinical Practice Guidelines for the management of pati ents with decompensated cirrhosis. J
Hepatol (2018), https://doi.org/10.1016/j.jhep.2018.03.024

severe or refractory volume overload, and/or symptomatic
azotaemia. However, published data on RRT in patients withcirrhosis are scant, with controversial effects on survival.
381,382
It has been stated that indications for RRT depend on the per-
spective of LT. It has been stated that RRT may be considered
in patients who are candidates for LT, while, in contrast, thedecision to initiate RRT in non-candidates should be individu-
alised to avoid futility.
20However, it has recently been observed
that critically ill liver cirrhotic patients requiring RRT have veryhigh mortality independent of LT options. Thus, RRT and treat-
ment at the ICU should not be limited to LT candidates but
should be based on the individual severity of illness.
383There-
fore, repeated risk stratification is necessary during the course
of treatment, assisted by prognostic scores in addition to clinical
judgment and patients0wishes.383The ideal timing for RRT ini-
tiation has not been defined in patients with cirrhosis. However,data on AKI in patients with acute liver failure as well as in crit-
ically ill patients without liver disease suggest that early RRT
improves survival.
384–386Both haemodialysis or continuous
renal replacement therapy (CRRT), have been used in patients
with cirrhosis. Despite the available evidence,387CRRT is prob-
ably better tolerated, providing greater cardiovascular stabilityand allowing a slower correction of severe or refractory hypona-
traemia than haemodialysis.
Liver support systems
.In two controlled studies, both the so-
called artificial liver support systems, either the molecularadsorbents recirculating system (MARS
/C210) or Prometheus/C210,
showed promising beneficial effects in patients with type 1
HRS, but should be further investigated.388,389
Liver transplantation and simultaneous liver-kidney transplanta-
tion.The best therapeutic option in patients with HRS is LT.390
However, several studies have shown that SCr after LT is higher
in patients transplanted with HRS, compared to those withoutHRS at the time of LT. In addition, the presence of HRS at the
time of LT has a negative impact on survival after LT.
391The
treatment of type 2 HRS before LT has given conflicting resultson the clinical outcome after LT
373,374and thus, requires further
investigation. SLK can be indicated in patients with cirrhosis
and CKD in the following conditions: a) estimated GFR (usingMDRD6 equation) ≤40 ml/min or measured GFR using iothala-
mate clearance ≤30 ml/min, b) proteinuria ≥2 g a day, c) kidney
biopsy showing >30% global glomerulosclerosis or >30% intersti-tial fibrosis, or d) inherited metabolic disease. SLK is also indi-cated in patients with cirrhosis and sustained AKI irrespective
of its type, including HRS-AKI when refractory to drug therapy,
in the following conditions: a) AKI on RRT for ≥4 weeks or b)estimated GFR ≤ 35 ml/min or measured GFR ≤25 ml/min ≥4
weeks.
392Beyond these two conditions, in a candidate with high
priority for LT due to a high MELD score, the option of SLK maybe considered in the presence of risk factors for underlying
undiagnosed CKD (diabetes, hypertension, abnormal renal
imaging and proteinuria >2 g/day).
392The development of
new biomarkers of kidney fibrosis, a common and irreversiblefeature of CKD, is also promising in this context.
393
Regarding the priority allocation of patients with HRS-AKI to
the waiting list, some rules should be applied in case ofresponse to drug therapy. In fact, by lowering SCr and increasing
serum sodium concentration, the treatment can significantly
lower the MELD and MELD-Na score, potentially delaying LT.Considering that the survival rate in responders at three monthsis almost 50%, a specific policy of priority allocation is needed
for these patients. This can be made either by continuing to con-sider the baseline MELD and/or MELD-Na score
394rather than
those during or after the end of the treatment, or by providing
an exception to the MELD score.395
Recommendations
/C15There is insufficient data to advocate TIPS in HRS-AKI but
it could be suggested in selected patients with HRS-NAKI
(II-2;2) .
/C15LT is the best therapeutic option for patients with HRS
regardless of the response to drug therapy (I;1).
/C15The decision to initiate RRT should be based on the indi-
vidual severity of illness (I;2).
/C15The indication for liver-kidney transplantation remains
controversial. This procedure should be considered inpatients with significant CKD or with sustained AKIincluding HRS-AKI with no response to drug therapy
(II-2;1) .
Prevention of hepatorenal symdrome
The prevention of HRS-AKI, as for other causes of AKI, is based
on the use of albumin in patients who develop SBP
272and the
prevention of SBP using norfloxacin,276as discussed before. In
addition, the use of pentoxyfilline may decrease the incidenceof renal failure in patients with cirrhosis
27and of type 1 HRS
as well as mortality in patients with severe alcoholic hepati-
tis.396However, recent papers do not confirm these
results397,398and further studies are needed.
Recommendations
/C15Albumin (1.5 g/kg at diagnosis and 1 g/kg on day three)
should be given in patients with SBP to prevent AKI (I;1).
/C15Norfloxacin (400mg/day) should be given as prophylaxis
of SBP to prevent HRS-AKI (I;1).
Acute-on-chronic liver failure
Definitions and pathophysiology
Since the CANONIC study, the first major international observa-
tional study characterising the syndrome of ACLF,3a large num-
ber of publications have described the association of this
syndrome with different clinical, diagnostic and therapeutic
approaches. ACLF occurs in 30% of admitted patients3,399and
in 25% of outpatients,400and is a major cause of death in cirrho-
sis, with an approximately 50% mortality rate.400Even though
there is an ongoing debate regarding the definition of
ACLF,401–405the concept of the development of ACLF is similar
across different continents and health systems. There is agree-
ment that ACLF is not just decompensation of liver cirrhosis,
but a distinct syndrome.406The reason is that ACLF is defined
by a multi-organ failure and has a higher short-term mortality
than a ‘‘simple decompensation” of cirrhosis.3,401,406The risk
of developing ACLF is higher in outpatients with advanced liverdisease according to the presence of ascites, low mean arterialJOURNAL
OF HEPATOLOGY
Journal of Hepatology 2018 vol. xxx jxxx–xxx 29
Please cite this article in press as: The European Association for the Study of the Liver. EASL Clinical Practice Guidelines for the management of pati ents with decompensated cirrhosis. J
Hepatol (2018), https://doi.org/10.1016/j.jhep.2018.03.024

pressure or anaemia and with a high MELD score.400ACLF devel-
ops on the background of acute decompensation (AD) of cirrho-sis, but a remarkable number of patients ( /C2440%) admitted to
hospital developed ACLF on the first episode of AD of their liver
disease.
3Thus, the presence of AD is an important clinical fea-
ture for the diagnosis of ACLF.3,401,406The EASL-CLIF Consortium
has proposed and validated a prognostic score (CLIF-C AD score)
for patients with AD who do not develop ACLF407The CLIF-C AD
score ( Box 1 ) was proved to be more accurate for predicting out-
come in these patients than the MELD or MELD-Na score.407
Once developed, ACLF is characterised by hepatic and extrahep-atic organ dysfunction and/or failure, highly activated systemicinflammation, and a high 28-day mortality.
3,12The overwhelm-
ing and devasting inflammatory response is a key pathogenic
mechanism in the development of ACLF, probably explainingwhy ACLF frequently happens in younger patients.
3,401,406,408
The trigger of ACLF and this inflammatory response could not
be identified in 40–50% of the patients in CANONIC study,3
which might be associated with genetic predisposition, severe
portal hypertension or other factors predisposing the patients
to development of AD and ACLF.409However, identification of
the precipitating events of AD are of great importance to pre-vent and manage ACLF.
410,411
Precipitating events
The precipitating events vary between different populations,
geographic areas and aetiologies. While in Western countries
(Europe, North and Latin America) bacterial infection, followed
by active alcohol intake or binge are major precipitatingevents,
3,412,413in Eastern countries (Asia, Pacific region) the
exacerbation of hepatitis B, followed by alcohol or bacterial
infections are the major causes of AD and ACLF develop-ment.
414–416But there are a number of other insults, which
might induce ACLF, such as superimposed infection with hepa-
totropic viruses (especially HAV, HEV), DILI, GI bleeding, circula-
tory dysfunction upon different situation ( e.g. surgery, LVP
without albumin). Therefore, in general the precipitating factors
might be differentiated into three major categories, hepatotoxic
injury (active alcohol intake or binge, DILI), immunologicalinsults (flairs of viral or autoimmune hepatitis, bacterial, fungal
and viral infections, common cold, subclinical infections, etc.)
and haemodynamic derangement following procedures (haem-orrhage, surgery, LVP).
Bacterial infections
Overall, the major precipitating factor for ACLF is bacterial
infections accounting for 30–57% of cases.
409,410The importance
of bacterial infections for the development of organ failures and
ACLF was also underlined by the studies of North American Con-
sortium for End-stage Liver Disease (NACSELD), who havedefined ACLF by the development of two organ failures in pres-
ence of bacterial infections.
412By contrast, bacterial infections
were not considered to be precipitating events for ACLF accord-ing to the definition of Asian Pacific Association for the Study of
the Liver (APASL).402However, nowadays there is evidence that
bacterial infections are mainly involved in the development oforgan failures and thereby of ACLF in Asia as well.
415,416In Wes-
tern countries bacterial infections are the precipitating events in
one-third of patients admitted with ACLF and in two-thirds ofpatients developing ACLF during follow-up.
3,409,412,413Based
on these data, preventive and early therapeutic interventions
for the treatment of infections are of major importance to pre-vent the development of ACLF. The role of bacterial infections
as triggers of AD and development of organ failures has already
been discussed.
Active alcohol intake or binge
Alcoholic liver disease was the most prevalent in patients with
AD and ACLF in the CANONIC study, as well as in recent reportsfrom India.
415–417Interestingly, active alcoholism and alcohol
binge were not only a major trigger in these patients, but led
to a more severe syndrome than other triggers in alcoholic cir-rhosis patients without heavy active alcoholism.
3The role and
mechanisms of active alcoholism need further investigation,
especially regarding prevention and treatment.
Reactivation and superimposed viral hepatitis
Reactivation of HBV in patients with cirrhosis is the main pre-
cipitating event in the non-Caucasian Asian population,413,415
occurring mostly in genotypes B and D, and hepatitis B e antigen
positive patients. Interestingly, superimposed HAV and HEV can
also trigger ACLF in 14–18%.414,416According to the Western
experience, these are unusual causes.3,413However, the role of
HEV might have been overlooked, and might gain more impor-
tance now because of advances in diagnostics and increases in
awareness.3,418A timely recognition and treatment of the pre-
cipitating event might prevent ACLF and improve outcome in
these patients.
Clinical and diagnostic features of ACLF
As discussed previously, organ failures in the presence of AD of
cirrhosis are the basis for the diagnosis of ACLF. However, in the
CANONIC study the presence and grading of ACLF was based onmortality and the independent association of organ dysfunc-tion/failure with mortality, which was chosen to be ≥15% at
28 days.
3Organ failures were defined based on a sequential
organ failure assessment (SOFA) score, which was adapted topatients with cirrhosis, the CLIF SOFA score ( Table 9 ). However,
two organs received special attention, the kidney and the brain.
3
In fact, it has been observed that even mild renal or brain dys-
function in the presence of another organ failure, is associated
with a significant short-term mortality and therefore defines
the presence of ACLF. Thus, patients with renal failure, definedas creatinine ≥2 mg/dl, were classified as ACLF grade Ia whilepatients with a non-renal and non-cerebral organ failure
combined either with mild renal dysfunction (creatinine
between 1.5 and 1.9 mg/dl) and/or grade I and II hepaticencephalopathy, as well as those with cerebral failure combined
with mild renal dysfunction were classified as ACLF grade Ib
(Table 10 ).
3Thereafter, patients with two organ failures are
classified as grade II ACLF, and have a 28-day mortality rate of
32%. Patients with three or more organ failures are classified
as grade III ACLF and have an average 28-day mortality of 78%(Table 9 ). According to this EASL-CLIF definition of ACLF,
approximately one-quarter of patients admitted to the hospitalBox 1 .CLIF-C Acute Decompensation Score (Ref. 407).
10 x [0.03 x Age + 0.66 x Ln(Creatinine) + 1.71 x Ln(INR) +
0.88 x Ln(WBC) 0.05 x Sodium + 8]CLIF-C Acute Decompensation score
Age in years; creatinine in mg/dl; WBC (white blood count) in 109 cells/L; sodium in mmol/LClinical Practice Guidelines
30 Journal of Hepatology 2018 vol. xxx jxxx–xxx
Please cite this article in press as: The European Association for the Study of the Liver. EASL Clinical Practice Guidelines for the management of pati ents with decompensated cirrhosis. J
Hepatol (2018), https://doi.org/10.1016/j.jhep.2018.03.024

for AD of cirrhosis had ACLF at admission or develop it during
the hospitalisation. After having simplified the CLIF SOFA score
into the CLIF Organ Failure score ( Table 11 ), the EASL-CLIF Con-
sortium formulated a new score, the CLIF-C ACLF score, whichenables the prediction of mortality in patients with ACLF.
419
The CLIF-C ACLF score ( Box 2 ) has been validated by different
independent series of patients.417,420,421Other scores were
recently proposed by the APASL ACLF Research Consortium
and by the NACSELD, but they were not compared specifically
with the CLIF-C-ACLF score.422,423Recommendations
/C15The diagnosis of ACLF should be made in a patient with
cirrhosis and AD (defined as the acute development or
worsening of ascites, overt encephalopathy, GI-haemor-
rhage, non-obstructive jaundice and/or bacterial infec-tions), when organ failure(s) involving high short-term
mortality develop (II-2;1) .
/C15The diagnosis and the grading of ACLF should be based
on the assessment of organ function as defined by the
CLIF-C Organ Failure score (II-2,1) .
/C15Potential precipitating factor(s), either hepatic ( i.e.heavy
alcohol intake, viral hepatitis, DILI, autoimmune hepati-
tis) and/or extrahepatic ( i.e.infections haemodynamic
derangements following haemorrhage, surgery) should
be investigated. However, in a significant proportion of
patients, a precipitant factor may not be identified(II-2,1) .
Management of ACLF
General management
Unfortunately, there is no specific effective treatment for
ACLF.
424Therefore, treatment is currently based on organ sup-
port and management of associated complications. The cause
of liver injury can be specifically treated only in certain situa-tions such as in ACLF secondary to HBV infection, as describedTable 9 .CLIF-Sequential Organ Failure Assessment (SOFA) score (adapted from Ref. n /C1763).
The CLIF-Sequential Organ Failure Assessment (SOFA) score
Organ/system 0 1 2 3 4
Liver (bilirubin mg/dl) <1.2 ≥1.2–<2.0 ≥2.0–<6.0 ≥6.0–<12.0 ≥ 12.0
Kidney (creatinine, mg/dl) <1.2 ≥1.2–<2.0 ≥ 2.0–<3.5 ≥3.5–<5.0 ≥5.0
Cerebral (HE grade) No HE Grade I Grade II Grade III Grade IV
Coagulation (INR and PLT count) <1.1 ≥1.1–<1.25 ≥1.25–<1.5 ≥1.5–<2.5 ≥ 2.5 or PLT ≤20.000/mm3
Circulation (MAP, mmHg and vasopressors) ≥70 <70 Dopamine ≤5*or
dobutamine or
terlipressinDopamine >5*or
E≤0.1*or NE
≤0.1*Dopamine >15*
or E >0.1*or NE
>0.1*
Lungs
PaO 2/FiO 2,o r
SpO 2/FiO 2>400
>512>300–≤400
>357–≤512>200–≤300
>214–≤357>100–≤200
>89- ≤214≤100
≤89
E, epinephrine; FiO 2, fraction of inspired oxygen; HE, hepatic encephalopathy; NE, norepinephrine; PaO 2, partial pressure of arterial oxygen; SpO 2, pulse oximetric
saturation. The bold text indicates the diagnostic criteria for organ failures.
*lg/kg/min.
Table 10 .Classification and grades of ACLF (adapted from Ref. 3).
Grades of
ACLFClinical characteristics
No ACLF No organ failure, or single non-kidney organ failure, creatinine
<1.5 mg/dl, no HE
ACLF Ia Single renal failure
ACLF Ib Single non-kidney organ failure, creatinine 1.5–1.9 mg/dl and/
or HE grade 1–2
ACLF II Two organ failures
ACLF III Three or more organ failures
ACLF, acute-on-chronic liver failure; HE, hepatic encephalopathy.
Table 11 .Chronic Liver Failure – Organ Failure score system (adapted from Ref. 419).
Organ/system 1 point 2 points 3 points
Liver Bilirubin <6 mg/dl 6 ≤Bilirubin <12 mg/dl Bilirubin ≥12 mg/dl
Kidney Creatinine <2 mg/dl 2 Creatinine <3.5 mg/dl Creatinine ≥3.5 mg/dl or renal replacement
Brain/HE (West Haven criteria) Grade 0 Grades 1–2 Grades 3-4a
Coagulation INR <2.0 2.0 ≤INR <2.5 INR ≥2.5
Circulation MAP ≥70 mmHg MAP <70 mmHg Use of vasopressors
Lungs PaO 2/FiO 2>300,
or
SpO 2/FiO 2>357PaO 2/FiO 2≤300–>200,
or
SpO 2/FiO 2>214–≤357PaO 2/FiO 2≤200b
or
SpO 2/FiO 2≤214b
Note: The bold text denotes criteria for diagnosing organ failures.
FIO 2, fraction of inspired oxygen; HE, hepatic encephalopathy; INR, international normalized ratio; MAP, mean arterial pressure; PaO 2, partial pressure of arterial oxygen;
SpO 2, pulse oximetric saturation.
aPatients submitted to mechanical ventilation due to HE and not to a respiratory failure were considered as presenting a cerebral failure (cerebral sc ore = 3).
bOther patients enrolled in the study with mechanical ventilation were considered as presenting a respiratory failure (respiratory score = 3).Box 2 .CLIF-C Acute Liver Failure (ACLF) score (Ref. 419).
10 x [0.033 x Clif OFs + 0.04 x Age + 0.63 x Ln(WBC) 2]
Age in years; CLIF OF score as in Table 10; sodium in mmol/LCLIF-C ACLF scoreJOURNAL
OF HEPATOLOGY
Journal of Hepatology 2018 vol. xxx jxxx–xxx 31
Please cite this article in press as: The European Association for the Study of the Liver. EASL Clinical Practice Guidelines for the management of pati ents with decompensated cirrhosis. J
Hepatol (2018), https://doi.org/10.1016/j.jhep.2018.03.024

later. Patients with ACLF should ideally be admitted to intensive
care or intermediate care units, yet this decision should be indi-vidualised based on certain factors, particularly patients’ ageand associated comorbidities. Moreover, patients suitable for
LT should be referred to a transplant centre early in the course
of ACLF. Late referral may make transplantation impossible dueto the rapid evolution of ACLF in most patients.
425In patients in
whom ACLF is associated with precipitating factors, such as bac-
terial infections, GI bleeding, or drug toxicity, early identifica-tion and management of these conditions is crucial to patient
survival. Nonetheless, it should be emphasised that this early
treatment of triggering factors may not prevent the progressionof ACLF in all patients. Meanwhile, as already stated, in approx-
imately half of patients with ACLF a precipitating factor cannot
be identified.
3Organ support is very important in the manage-
ment of patients with ACLF.424Haemodynamic function should
be monitored and vasopressor therapy administered in case of
marked arterial hypotension. Hepatic encephalopathy should
be treated early with standard therapy. Special care should betaken to preserve airway patency to prevent aspiration pneu-
monia. In patients with coagulation failure, either because of
impairment of coagulation factors or low platelet count, substi-tutive therapy should be given only if there is clinically signifi-
cant bleeding. If there is respiratory failure, patients should be
given oxygen therapy and ventilation, if required. Finally, ifthere is kidney failure its cause should be identified and man-
aged accordingly. Volume expansion should be given to patients
with fluid loss or in the setting of SBP. Excessive volume expan-sion should be avoided. Patients meeting the criteria of AKI-HRSshould be treated with terlipressin and albumin or nore-
pinephrine, if terlipressin is not available. Patients with sus-
pected ATN should be treated with RRT if they meet criteriafor this treatment.
392
Specific therapies
Liver support systems .Extracorporeal liver support systems, par-
ticularly albumin dialysis (MARS system) and fractionated
plasma separation and adsorption (Prometheus system) have
been evaluated as therapies for ACLF. These systems removealbumin-bound substances and other substances that accumu-
late in the context of ACLF and may have deleterious effects
on the function of different organs. Both methods have beenevaluated in large RCTs in patients with ACLF and no significanteffects on survival could be demonstrated.
388,389It should be
emphasised however, that the definition of ACLF in both trials
was different than the current definition of ACLF based on theCANONIC study.
3Moreover, a sub-analysis of the Prometheus
study showed a beneficial effect on survival in patients with
MELD score higher than 30.389This finding deserves further
investigation. Nonetheless, based on the results of available
RCTs, extracorporeal liver support systems do not improve sur-
vival of patients with ACLF and should not be recommended inthis indication.
Antiviral therapy for chronic hepatitis B
.Reactivation of hepatitis
B is a very common cause of ACLF in certain areas of the world,
particularly in southwest Asia.414A number of non-randomised
studies and an RCT have shown that treatment with lamivudine,
tenofovir or entecavir is associated with inhibition of HBV repli-
cation, improvement of liver function, and higher survival inpatients with ACLF secondary to hepatitis B infection.
426–429
The only RCT included 24 patients, 14 treated with tenofovirand 13 treated with placebo, and showed significant differences
in three-month survival (57% vs.15%, respectively).426There-
fore, it seems evident that the presence of HBV infection shouldbe investigated in all patients with ACLF and antiviral therapy
should be started as soon as possible.
Other therapies
.A number of therapies have been assessed in
patients with ACLF, including dexamethasone, plasma
exchange, chinese herbs, caspase inhibitors, mesenchymal stem
cells transplantation, and administration of granulocyte-colonystimulating factor (G-CSF).
430–432In most cases, the information
is still very preliminary and no recommendations can be made
regarding their potential use in clinical practice. However, a
note on G-CSF seems pertinent because this approach has beenassessed in an RCT.
432The rationale behind this treatment
seems to be the mobilisation of stem cells from the bone mar-
row and their engraftment within the liver, although other ben-eficial effects may also occur. The only RCT evaluating this
therapy included 47 patients with ACLF, as defined by the APASL
criteria, 23 treated with G-CSF (12 doses of 5
lg/kg subcuta-
neously) and 24 treated with placebo in a double-blind manner.
The main findings were an improvement in 60-day survival in
the G-CSF group vs.the placebo group (66% vs.26%, respec-
tively; p= 0.001) along with a reduction in Child-Pugh, SOFA,
and MELD scores and a decrease in the occurrence of HRS, hep-
atic encephalopathy and sepsis in G-CSF treated patients.
Although these results are promising, additional studies in a lar-ger number of patients are needed.
Liver transplantation
.Liver transplantation is theoretically the
definitive treatment for ACLF because it allows the cure of ACLFsyndrome as well as the underlying liver disease.
425However,
some important issues regarding LT for ACLF deserve a com-
ment, particularly the accessibility of patients to LT, evaluation
of candidate subjects, the outcomes of LT on survival, and futility.The accessibility of patients with ALCF to LT is probably
decreased compared to that of patients with other indications
for LT, because patients with ACLF have a high mortality rateafter diagnosis of the condition. Early referral to transplant cen-
tres is therefore crucial. Then, because ACLF is a rapidly evolving
syndrome, candidate patients need to be submitted to a ‘‘fast-track” clinical evaluation of organ function and potential comor-
bidities that could contraindicate LT. Data on outcome of
patients with ACLF treated with liver LT are scarce but nonethe-less, patient survival at three-months after LT is about 80%,much higher than what would be anticipated if patients were
not transplanted.
425,433,434Almost all patients with ACLF-3
developed complications after LT, especially pulmonary, renaland infectious, compared to patients with no ACLF, or ACLF-1
and -2. This emphasises the need for special management when
transplanting patients with ACLF-3, with repeated systematicscreening for infection and careful monitoring of renal and res-
piratory parameters.
434Another point is that some patients
with ACLF are potentially too sick for LT. In the context of scar-city of donor livers, the potential benefit of LT for patients withACLF must be balanced with the rationing. Thus, more data is
needed to determine medical futility in patients with
ALCF.
425,434However, if LT is contraindicated or not available
for patients with organ failures ≥4 or CLIF-C ACLFs >64 at days
3–7 after diagnosis of ACLF-3, the intensive organ support
should be discontinued owing to futility.425Clinical Practice Guidelines
32 Journal of Hepatology 2018 vol. xxx jxxx–xxx
Please cite this article in press as: The European Association for the Study of the Liver. EASL Clinical Practice Guidelines for the management of pati ents with decompensated cirrhosis. J
Hepatol (2018), https://doi.org/10.1016/j.jhep.2018.03.024

Recommendations
/C15At present, there is no specific therapy for ACLF aside
from antiviral therapy in patients with ACLF due to reac-
tivation of HBV infection. Treatment of ACLF should be
based on organ support and management of precipitants(see point below) and associated complications. Patients
should be treated in intermediate care or intensive care
settings. Organ function, particularly, liver, kidney, brain,lung, coagulation, and circulation should be monitored
frequently and carefully throughout hospitalisation, as
ACLF is a dynamic condition. However, monitoring andmanagement should be individualised according tospecific circumstances, mainly patients’ age and comor-
bidities (III, 1) .
/C15Early identification and treatment of precipitating fac-
tors of ACLF, particularly bacterial infections, are recom-
mended. However, in some patients ACLF progressesdespite treatment of precipitating factors (III;1) .
/C15Administration of nucleoside analogues (tenofovir, ente-
cavir) should be instituted as early as possible in patientswith ACLF due to HBV infection (I;1).
/C15Early referral of patients with ACLF to liver transplant
centres for immediate evaluation is recommended (II-3;1) .
/C15Withdrawal of ongoing intensive care support can be
suggested in patients, who are not candidates for LT,
with four or more organ failures after one week of ade-
quate intensive treatment (II-2, 2) .
/C15Despite promising results, the administration of G-CSF
can not be recommended at present (1;2) .
Relative adrenal insufficiency
Definition and pathophysiology
Relative adrenal insufficiency (RAI) is a condition of inadequate
cortisol response to stress in the setting of critical illness,435also
named as ‘‘Critical Illness Related Corticosteroid Insufficiency”(CIRCI).
436RAI has also been described in patients with cirrhosis
and, although it is mainly present in critically ill patients with
sepsis or septic shock (68.9%), it also affects non-critically ill cir-rhotic patients (41.8%), including those with compensated cir-
rhosis.
437–442The pathophysiology of RAI in cirrhosis is not
well defined. Suppression of the hypothalamic-pituitaryadrenalaxis activity, reduced effective volemia, which may impair adre-nal perfusion, and both impaired cholesterol synthesis and
enhanced pro-inflammatory cytokine production likely con-
tribute to impair adrenal steroidogenesis.
443,444Adrenal dys-
function blunts the vascular effect of angiotensin II,
norepinephrine and vasopressin, leading to further sympathetic
nervous system activity.445These effects would worsen the car-
dio-circulatory dysfunction of advanced cirrhosis, and favour
gut bacterial overgrowth, and hence BT, by impairing intestinal
motility.445This explains why RAI in decompensated cirrhosis is
associated with a higher probability of severe sepsis and type-1
HRS, and higher short-term mortality.437,446
Diagnosis
The diagnosis of RAI is influenced by the method employed, as
the measurement of serum total cortisol, either at baseline orafter stimulation by the standard dose- or low dose-short Syn-
acthen tests can be utilised.447The consensus statements from
the American College of Critical Care Medicine recommendreferring to a delta total serum cortisol <250 nmol/L (9
lg/dl)
after adrenocorticotrophic hormone administration or a random
total cortisol <276 nmol/L (10 lg/dl) in critically ill patients.436
There is no reason for not employing these indications in
patients with cirrhosis. However, the diagnosis of RAI based
on serum total cortisol concentration, which is measured bystandard assays, may be flawed by the reduced serum levels
of cortisol binding globulin (CBG) and albumin frequently seen
in patients with cirrhosis. This may lead to an overestimationof RAI, as more than 90% of circulating cortisol is bound to these
proteins.
448The assessment of serum-free cortisol concentra-
tion would overcome this limitation. Serum-free cortisol levels<50 nmol/L at baseline, or <86 nmol/L (9
lg/dl) after adrenocor-
ticotrophic hormone suggest the presence of RAI in critically ill
patients.449By comparing RAI diagnosis in clinically stable
patients with cirrhosis based on either total or free plasma cor-tisol, a clear discrepancy emerged, as the prevalence of RAI was
58% using total cortisol criteria and 12% using free cortisol with
a peak plasma level <33 nmol/ after stimulation.
450Unfortu-
nately, the methods for determining free cortisol are complex
and expensive, so they are not used in routine clinical practice.
The surrogate methods that have been proposed for the calcula-tion of plasma free cortisol
451,452do not seem to be fully reliable
in patients with cirrhosis.450For these reasons, salivary cortisol
has received attention, as it correlates with free cortisol levelsirrespective of the concentration of binding proteins.
453,454
Baseline salivary cortisol <1.8 ng/ml (<0.18 lg/dl) or an incre-
ment <3 ng/ml (0.3 lg/dl)453following a standard-dose short
Synacthen test are suggestive of RAI. However, even the evalu-ation of salivary cortisol is not without shortcomings.
454
Recommendation
/C15Diagnosis of RAI should be based on a delta serum total
cortisol after 250 lg corticotropin injection of <248
nmol/L (9 lg/dl) or a random total cortisol of <276
nmol/L (<10 lg/dl) (II-2,1) . As serum free cortisol con-
centration can be influenced by the reduced serum levels
of CBG and albumin frequently seen in patients with cir-
rhosis, salivary cortisol determination can be preferred(II-2;2) .
Treatment of relative adrenal insufficiency
It is not known whether cortisol supplementation in clinically
stable cirrhosis with RAI is of any value. Two studies have eval-uated the effects of treating RAI in critically ill patients with cir-rhosis. In one study 17 patients with cirrhosis and sepsis, in
whom RAI was diagnosed, received i.v. hydrocortisone (50 mg/
6h), and were compared with 50 consecutive patients with cir-rhosis and septic shock who had previously been admitted to
the same ICU but did not receive steroids. A higher rate of shock
resolution, survival in the ICU and hospital survival were seen inthe patients treated with hydrocortisone.
455In the second
study, 57 patients with cirrhosis, septic shock and RAI were ran-
domised to receive either i.v. 50 mg of hydrocortisone or normalsaline every 6 h until haemodynamic stability was achieved, fol-lowed by steroid tapering over eight days. Lower vasopressorJOURNAL
OF HEPATOLOGY
Journal of Hepatology 2018 vol. xxx jxxx–xxx 33
Please cite this article in press as: The European Association for the Study of the Liver. EASL Clinical Practice Guidelines for the management of pati ents with decompensated cirrhosis. J
Hepatol (2018), https://doi.org/10.1016/j.jhep.2018.03.024

doses and higher rates of shock reversal were seen in patients
who received hydrocortisone. However, 28-day mortality didnot differ between the two groups. Moreover, shock relapseand GI bleeding occurred more often in the hydrocortisone
group.
456
Recommendation
/C15At present, hydrocortisone treatment (at a dose of 50
mg/6h) of RAI cannot be recommended (I-2) .
Cardiopulmonary complications
Cirrhotic cardiomyopathy
Definition and pathophysiology
Cirrhotic cardiomyopathy (CCM) refers to chronic cardiac dys-
function in a patient with established cirrhosis, characterised
by a blunted contractile response to stress (pharmacologi-
cal/surgical or inflammatory) and an altered diastolic relaxation,often associated with electrophysiological abnormalities such as
prolongation of the QTc interval. These phenomena occur in the
absence of any other cardiac disease.
457Systemic inflammation
is thought to be key in inducing myocardial dysfunction associ-
ated with impaired diastolic relaxation and decreased left ven-
tricular ejection fraction, however, there are few controlledstudies.
193,458,459Shear stress generated by portal hypertension
exhibiting mechanical forces on myocardial fibres, may also
play a part.460CCM is largely subclinical but its presence does
influence prognosis in advanced disease,461and it certainly
impacts on the course of interventions such as TIPS and LT.462
Diagnosis
Characterisation of systolic dysfunction in cirrhotic
cardiomyopathy
Systolic dysfunction refers to impaired left ventricle contractile
responses to stress on echo, translating to a resting left ventric-
ular ejection fraction (LVEF) <55%. For most patients with cirrho-
sis, the resting systolic function is normal or even increased, dueto the hyperdynamic circulation and reduced afterload to main-
tain cardiac output. To investigate systolic dysfunction in cirrho-
sis, it is necessary to induce circulatory stress eitherpharmacologically or through exercise. Systolic dysfunction thenmanifests as a lack of an appropriate left ventricular contractile
response to the applied stress. As disease advances, the progres-
sive reduction in peripheral vascular resistance unmasks systolicdysfunction. Early studies used exercise stress testing to demon-
strate a lack of increment in cardiac output or LVEF
463,464and
this was even shown when noradrenaline levels were increased,suggesting loss of sympathetic responsiveness.
465More recent
studies used pharmacological stress echo to show a blunted
response.466However, other studies using cardiac MRI, have
shown normal chronotropic and inotropic responses suggesting
the techniques used may give rise to variability.467
Myocardial strain imaging for assessing systolic dysfunction
Myocardial strain imaging is a more recent echocardiographic
technique evaluating the degree of shortening of myocardial
muscle fibres (‘strain’) influencing cardiac wall motion. Themeasurement of left ventricular global longitudinal systolicstrain (GLS) is believed to be a sensitive marker of left ventric-ular systolic function and facilitates the assessment of systolic
dysfunction at rest,
468,469as well as having prognostic impor-
tance in heart failure.470Studies of strain imaging in cirrhosis
have demonstrated variable results; some showing impaired
systolic strain in patients compared with healthy controls, albeit
with no correlation to Child-Pugh score.471,472Others demon-
strate systolic strain within normal range and not influenced
by the presence of ascites.473,474However, interestingly, when
patients undergo LT, the systolic strain improves.471
Characterisation of diastolic dysfunction in cirrhotic
cardiomyopathy
Numerous echocardiographic criteria along with transmitral
Doppler evaluation have been used to characterise diastolic dys-
function including, early diastolic/atrial filling ratio (E/A), earlydiastolic filling/mitral annular velocity (E/e’) and tricuspid sys-tolic jet velocity. Such measurements are influenced by the
pre- and afterload changes of portal hypertension. The latest
American Society of Echocardiography (ASE) and EuropeanAssociation of Cardiovascular Imaging guidelines for the evalu-
ation of diastolic dysfunction recommend the following criteria
based on a normal LVEF (often the case in cirrhosis):
475
i. Average E/e’>14
ii. Septal e’ velocity <7 cm/s OR Lateral e’ velocity <10 cm/s
iii. Tricuspid velocity >2.8 m/s
iv. Left atrial volume index (LAVI) >34 ml/m2
Diastolic dysfunction translates to impaired relaxation of the
left ventricle, abnormal filling of the left atrium, and a higherleft atrial volume. Indeed, increased LAVI has been associated
with greater risk of heart failure in ischaemic cardiac disease.
476
Based on these guidelines, diastolic dysfunction is classified as:
grade I if one of the three principle criteria (1,3 and 4 above) are
met; and grade II if two or more of the criteria are met.
However, there is heterogeneity in descriptions of the preva-
lence of diastolic dysfunction in cirrhosis, in part reflecting thedifferent echo techniques and/or diagnostic criteria applied,
and the influence of vasoactive agents such as beta-blockers
and terlipressin.
Several studies using an E/A ratio of ≤1 criteria have demon-
strated left atrial enlargement in patients with ascites and
advanced disease.
464,477Therapeutic paracentesis improves E/A
ratio and importantly, in all studies, there is no relation to aetiol-
ogy.478In patients treated with TIPS there was no relation to aeti-
ology, but diastolic dysfunction does show a positive correlationwith higher MELD scores.
462,479A further study using E/e’ ratio to
define diastolic dysfunction in patients with ascites and elevated
plasma renin, demonstrated that an increased E/e’ was an inde-
pendent predictor of development of HRS type 1 and one-yearmortality.
480By contrast, other studies fail to show a clear rela-
tionship with disease severity or survival,460,472,481albeit in two
studies echo criteria are not specified.460,472Studies using the
LAVI criteria suggest a closer association between left atrial
enlargement and Child-Pugh C disease.482
Newer techniques such as cardiac MRI with ‘T1 mapping’ and
Late Gadolinium enhancement are being deployed to assess
whether fibrosis or oedema modulates myocardial function in
conditions such as amyloid and Fabry disease.483Literature on
using such techniques in liver disease are very limited. A recentstudy in patients with chronic hepatitis C showed no significant
differences in echo parameters to controls, but demonstratedClinical Practice Guidelines
34 Journal of Hepatology 2018 vol. xxx jxxx–xxx
Please cite this article in press as: The European Association for the Study of the Liver. EASL Clinical Practice Guidelines for the management of pati ents with decompensated cirrhosis. J
Hepatol (2018), https://doi.org/10.1016/j.jhep.2018.03.024

lower post-contrast myocardial T1 time and higher partition
coefficients, indicative of diffuse myocardial fibrosis.484
Natural history
Impact of disease related physiological stress: sepsis,
decompensation and/or GI bleeding
There are limited studies of cardiac assessment during acute cir-
rhosis decompensation and associated haemodynamic instabil-ity. In a seminal study of acute cirrhosis decompensation with
SBP, a subgroup with HRS were shown to have lower cardiac
outputs at diagnosis and this correlated inversely with highinflammatory indices.
193A follow-on study by the same group
identified that patients with HRS had cardiac outputs that were
further reduced at follow-up, compared to patients who do not
develop HRS after SBP, and these patients had higher plasmanoradrenaline and renin.
485Other studies recapitulate this with
a demonstration of reduced kidney blood flow and importantly,
suggest those with low cardiac index, also have increasedmortality.
192,485
In relation to systemic inflammation and sepsis, one study
showed lipopolysaccharide binding protein (LBP) levels (a sur-rogate for BT and lipopolysaccharide) in patients with ascitesto be associated with significant diastolic dysfunction and left
atrial enlargement. The E/e’ ratio in these patients correlated
with LBP levels. This data supports findings from experimentalstudies, which have shown a role for inflammation, signalling
through the inflammasome and macrophage activation, as key
pathological processes related to myocardial dysfunction.
486–488
Acute GI bleeding in cirrhosis is understandably associated
with significant haemodynamic disturbances and has not been
studied systematically in relation to cardiac function. Dataassessing chronotropic function suggest the QTc interval is
increased in cirrhotic patients during an acute bleeding episode
compared to non-cirrhotic patients, and that this is associatedwith higher MELD scores and independently predicts sur-vival.
489This contrasts with a more recent study that fails to
demonstrate a clear link between QTc prolongation and mortal-
ity.490Possible reasons for this heterogeneity in outcomes are
the variable nature of vasoactive agents and their respective
doses required for the control of bleeding in these studies. For
example, one study showed terlirpessin decreased cardiac out-put by 17% and the reduction in wall motion after terlipressin
correlated with the Child-Pugh score.
491
Impact of interventions on cirrhotic cardiomyopathy
TIPS.Cardiac reserve is a major clinical consideration for elective
TIPS placement and a 2D echo to assess LVEF is standard prac-
tice. Despite this, some patients do have cardiac decompensa-tion post TIPS insertion. Several studies show an associationbetween presence of diastolic dysfunction at the time of TIPS
and poor survival.
462,479By contrast, others have not shown
any difference in survival between patients with and withoutdiastolic dysfunction at the time of TIPS.
492However, uniformly,
studies suggest an increase in left ventricular and atrial volume
over time, implying that such patients may be at greater risk offuture heart failure, based on literature for ischaemic heart dis-
ease and dilated cardiomyopathy.
476,493
Liver transplantation .Just as data on the effects of cirrhosis com-
plications on cardiac function are variable, data on the impact ofthe physiological stress of LT on patients with pre-existing car-
diac dysfunction is heterogenous, largely because of the differ-ent echo criteria and thresholds applied.
One study in 173 transplant recipients assessed systolic
(resting ejection fraction <55%) and diastolic (E/A ratio <1 or a
deceleration time >200 ms) dysfunction and reported it occur-ring in 2% and 43% of patients, respectively. Whilst patients with
diastolic dysfunction were older, interestingly, outcomes were
not influenced by the presence of diastolic dysfunction.
494By
contrast, another study used echo and brain natriuretic peptide
(BNP) levels to grade severity of cardiac dysfunction. Those
patients with higher BNP levels (>391) on day one tended tohave higher mortality and longer dialysis requirements post
transplantation. Of these, a subset with BNP levels >567 had
ejection fractions <50%, and some of these died of cardiogenicshock within 72 h post-transplant. Autopsy in these patientsshowed diffuse myocardial fibrosis. In the main, BNP levels
tended to decrease towards normal values over a week.
495
A further study performed a detailed echo assessment,
including myocardial strain assessment with speckle tracking,
in patients undergoing LT compared to non-transplanted
patients over a median follow-up of 18 months. Whilst patientspre-transplant had increased left ventricular mass and diastolic
dysfunction, following transplantation, there was a significant
improvement in systolic strain and reduced left ventricularmass. Conversely, cirrhosis patients who were not transplanted
had an increase in left ventricular mass, albeit systolic strain did
not change significantly.
471This implies that some of the patho-
physiological changes in CCM, such as increased left ventricularmass and size, are reversible with resolution of the disease.
However, studies with comprehensive characterisation of car-
diac function post transplantation are limited.
Prognosis for cirrhotic cardiomyopathy
Data relating cardiac dysfunction (especially diastolic dysfunc-
tion) with survival is variable. Some prospective studies, despitedetailed evaluation of patients, including those with ascites and
using speckle tracking, show no relation between cardiac dys-
function and survival, even among more decompensatedpatients.
473,496Many of the patients in these studies have evi-
dence of diastolic dysfunction and some with even advanced
grade II diastolic dysfunction albeit the GLS values in thesestudies are within the normal range.
Conversely, other studies suggest an association between
presence of diastolic dysfunction and higher two-year mortality,
with diastolic dysfunction ranging from 38–67%, especially inpatients with severe ascites.
461,497Indeed, in one such study, a
multivariate analysis showed left ventricular diastolic dysfunc-
tion was an independent predictor of mortality.461Another
study followed 80 patients to assess one-year mortality, finding
46% had diastolic dysfunction on echo criteria and about half of
these had grade II dysfunction, in whom mean arterial bloodpressure was lower and MELD score higher than grade I
patients. The presence of diastolic dysfunction was associated
with a higher degree of ascites and plasma renin levels and38% of these patients developed criteria for HRS type I. Survivalwas 95% in those without diastolic dysfunction, compared to
79% in those with grade I dysfunction and 39% with grade II
diastolic dysfunction. E/e’ ratio was an independent predictorof survival.
480JOURNAL
OF HEPATOLOGY
Journal of Hepatology 2018 vol. xxx jxxx–xxx 35
Please cite this article in press as: The European Association for the Study of the Liver. EASL Clinical Practice Guidelines for the management of pati ents with decompensated cirrhosis. J
Hepatol (2018), https://doi.org/10.1016/j.jhep.2018.03.024

Recommendations
/C15Evaluation of cirrhosis patients with echocardiography
should be performed with dynamic stress testing either
pharmacologically, or through exercise, given that sys-
tolic dysfunction may be masked by the hyperdynamiccirculation and reduced afterload. Failure to increment
cardiac output after physiological/pharmacological
stress (and in the absence of influence of beta-blockade)indicates systolic dysfunction (II-1;1) .
/C15Myocardial strain imaging and assessment of GLS may
serve as a sensitive marker of left ventricular systolicfunction and facilitate its assessment at rest and in
decompensated patients (II-2;2) . Cardiac MRI may also
identify structural changes. However, with all thesetechniques, there is the need for more controlled studies
and correlation with clinical endpoints (III;2) .
/C15Diastolic dysfunction may occur as an early sign of car-
diomyopathy in the setting of normal systolic function,
and should be diagnosed using the recent ASE guideli-nes, namely: Average E/e’>14; Tricuspid velocity >2.8m/s and LAVI >34 ml/m
2(II-1;1) .
/C15In patients with AD of cirrhosis, reduced cardiac output
(as a manifestation of CCM) is of prognostic significance
as it is associated with the development of AKI (specifi-
cally hepatorenal dysfunction) after infections such asSBP(II-1;1) .
/C15Prolongation of the QTc interval is common in cirrhosis
and can be evaluated since it may indicate a poor out-
come. Agents that can prolong the QT interval should
be used cautiously (II-2;2) .
/C15Detailed functional cardiac characterisation should be
part of the assessment for TIPS insertion (II-2;2) or LT
(II-1;1) .
/C15Standardized criteria and protocols for the assessment of
systolic and diastolic function in cirrhosis are needed(II-2;2) .
Hepato-pulmonary syndrome
Definitions and clinical manifestations
The association of chronic liver disease with respiratory symp-
toms and hypoxia is well recognised. Four main pulmonarycomplications may occur in patients with chronic liver disease:
pneumonia, hepatic hydrotorax, HPS and PPHT. HPS is definedas a disorder in pulmonary oxygenation, caused by intrapul-monary vasodilatation and, less commonly, by pleural and pul-
monary arteriovenous communications occurring in the clinical
setting of portal hypertension.
498,499It is most commonly
diagnosed in patients with cirrhosis498,499and portal hyperten-
sion500but, it has also been described in patients with pre-hepatic
portal hypertension,501with venous obstruction but without
cirrhosis, and even in patients with acute or chronic hepatitis500
(Table 12 ). A severe impairment of liver function and a specific
aetiology of liver disease are not needed for the development ofHPS,
498based on the profiles of the patients studied. In terms of
prevalence, HPS has been reported in 10% of patients with
chronic viral hepatitis in 15–23% of those with cirrhosis andin 28% of those with Budd-Chiari syndrome.
502–504However,
the prevalence of HPS reported in patients with cirrhosis
undergoing LT evaluation ranges from 5–32%,504–508while
intrapulmonary vascular dilatation (IPVD) can be detected byechocardiography in 50–60% of cirrhosis patients undergoing
LT evaluation. No relationship seems to exist between HPS
and CCM.
504The clinical manifestations of HPS in patients
with chronic liver disease primarily involve dyspnoea and
platypnoea.498,502,506Dyspnoea is the most common respiratory
complaint in patients with HPS, but it is non-specific. Its onset isinsidious, usually occurring on exertion. Platypnoea, which is a
shortness of breath exacerbated by sitting up and improved by
lying supine, is a less sensitive but a more specific finding inthese patients. Hypoxemia with exertion or at rest is commonand it is exacerbated in the upright position (orthodeoxia).
There are no signs or hallmarks of HPS on physical examination.
However, tachypnoea and polypnoea, digital clubbing and/orcyanosis in patients with the hallmarks of chronic liver disease
suggest the presence of HPS.
498,502,506
Pathophysiology
The pathophysiology of HPS is characterised by an IPVD occur-
ring within the pulmonary arterial circulation. This vascular
abnormality consists of diffuse or localised abnormal dilatedpulmonary capillaries and, less commonly, pleural and pul-
monary arteriovenous communications,
509which result in
impaired oxygenation of venous blood as it passes throughthe pulmonary circulation. IPVD impairs ventilatory/perfusion(V/Q) ratio and may result in anatomic and functional shunt
leading to hypoxaemia ( Fig. 10 ). In patients with advanced liver
cirrhosis this leads to a subtle increase in intrapulmonary bloodshunting, which is more pronounced in patients with HPS. The
consequent increase of shunting and V/Q mismatch in the
Table 12 .Diagnostic criteria of hepatopulmonary syndrome.
Hypoxia with partial pressure of oxygen <80 mmHg or alveolar–arterial oxygen gradient ≥15 mmHg in ambient air ( ≥20 mmHg in patients older than
65 years).
Pulmonary vascular defect with positive findings on contrast-enhanced echocardiography or abnormal uptake in the brain (>6%) with radioactive
lung-perfusion scanning
Commonly in presence of portal hypertension, and in particular:
– hepatic portal hypertension with underlying cirrhosis
– pre-hepatic or hepatic portal hypertension in patients without underlying cirrhosis
Less commonly in presence of:
– acute liver failure, chronic hepatitis
All criteria were determined by means of positive contrast-enhanced echocardiography ( i.e., microbubble opacification of the left heart chambers three to six cycles after
right atrial passage). The abbreviated formula for the alveolar–arterial gradient is as follows: PaO 2/C0PaO 2= (FIO 2[Patm–PH 2O]/C0[PaCO 2/0.8]) /C0PaO 2, where PaO 2denotes
partial pressure of alveolar oxygen, PaO 2partial pressure of arterial oxygen, FIO 2fraction of inspired oxygen, Patm atmospheric pressure, PH 2O partial pressure of water
vapor at body temperature, and PaCO 2partial pressure of arterial carbon dioxide (0.8 corresponds to the standard gas-exchange respiratory ratio at rest); the normal range
is 4 to 8 mmHg.Clinical Practice Guidelines
36 Journal of Hepatology 2018 vol. xxx jxxx–xxx
Please cite this article in press as: The European Association for the Study of the Liver. EASL Clinical Practice Guidelines for the management of pati ents with decompensated cirrhosis. J
Hepatol (2018), https://doi.org/10.1016/j.jhep.2018.03.024

upright position is the cause of the orthodeoxia.510The patho-
genesis of IPVD is probably multifactorial ( Fig. 11 ). The release
of nitric oxide, which is a potent vasodilator, plays a critical rolein the development of HPS. The increased release of nitric oxide
in the pulmonary circulation is related to an increased expres-
sion and activity of two isoforms of nitric oxide synthase(NOS), the endothelial NOS (eNOS) and the inducible NOS
(iNOS).
511–516Meanwhile, BT and the BT-related endotoxaemia
and pro-inflammatory response also contribute to the accumu-lation of macrophages in the pulmonary microvasculature.
517
Endothelial activation of fractalkine (CX3CL1), a chemokine, inthe lung may favour the adherence of monocytes in the pul-monary microcirculation.
518Monocytes express iNOS and pro-
duce heme oxygenase-1, leading to increased carbon
monoxide production, further enhancing vasodilatation.519CX3CL1 and vascular endothelial growth factor (VEGF) A, pro-duced by circulating monocytes, also contribute to angiogene-sis, recently recognised as a further pathogenetic factor ofpulmonary IPVD in experimental HPS.
520–522A downregulation
of miRNA-199 a-5p has recently been described as a contribu-
tory mechanism of pulmonary microvascular endothelial cellproliferation and thus pathogenesis of HPS.
523Polymorphisms
in genes involved in the regulation of angiogenesis have also
been associated with the risk of HPS in patients with cirrho-sis
524(Fig. 11 ). Finally, it has recently been observed that rosu-
vastatin, by down-regulating protein expression of nuclear
factor kappa B and VEGF-1,2 and Rho-associated A kinase,may improve the intrapulmonary angiogenesis and the
alveolar-arterial oxygen pressure gradient in common bile duct
ligation rats.
525
Diagnosis
In patients with portal hypertension and the clinical suspicion
of HPS partial pressure of oxygen (PaO 2) in arterial blood gas
(ABG) should be assessed. A PaO 2lower than 80 mmHg and or
an alveolar-arterial oxygen gradient (P[A-a]O 2) ≥15 mmHg
while breathing ambient air at sea level should lead to further
investigations ( Table 12 ). For adults ≥65 years a P[A-a]O 2≥20
mmHg cut-off is used.526However, it should be highlighted that
although these criteria are well established, enabling one to
unify the diagnostic methods and thus to better understandthe disease, they are based on a consensus of experts. Pulse
oximetry indirectly measures oxygen saturation (SpO
2), it is
non-invasive and may be useful in the diagnosis of HPS in adultssince a SpO
2<96% was found to be highly sensitive (100%) and
specific (88%) for detecting HPS in patients with a PaO 2<70Normal HPS
Alveol
Capillary
8µ
500 µO2 O2
Fig. 10. The pathophysiology of hepatopulmonary syndrome (adapted
from Ref. 498).HPS, hepatopulmonary syndrome.
• Hepatic injury/failure
• Portal hypertension• Portosystemic shunt
• Hyperdynamic circulation
• Bacterial translocation
Increased release of
ET-1Systemic
inflammation
HEPATOPULMONARY
SYNDROMEETB
receptor
Increased expression and
activity of eNOS
Increased release of
NOEndothelial cellRecruitment of
macrophages
in the lungs
Increased expression and
activity of iNOS and HO
Increased release of
NO and COEndothelial cell
MacrophageEndothelial activation of
CX3CL1
Increased adherence of
macrophages/monocytes to
endothelial cells
VFG-A release
Endothelial cell
Endothelial cell
proliferation
VASODILATION VASODILATIONANGIOGENESIS
Pulmonary capillary
CX3CL1
Genetic factors
Fig. 11. The pathogenesis of hepatopulmonary syndrome. ET, endothelin; eNOS, endothelial nitric oxide synthase; iNOS, inducible nitric oxide synthase; NO,
nitric oxide; HO, Heme oxygenase-1; CO, carbon monoxide; CX3CL1, fractalkine; VGF-A, vascular endothelial growth factor A.JOURNAL
OF HEPATOLOGY
Journal of Hepatology 2018 vol. xxx jxxx–xxx 37
Please cite this article in press as: The European Association for the Study of the Liver. EASL Clinical Practice Guidelines for the management of pati ents with decompensated cirrhosis. J
Hepatol (2018), https://doi.org/10.1016/j.jhep.2018.03.024

mmHg, limiting ABG testing to only 14% of patients.527The
validity of this non-invasive approach was not confirmed,recently, in paediatric patients with HPS.
528Serial SpO 2mea-
surements may be useful to monitor impaired oxygenation over
time in patients with HPS. The ABG is essential for the staging of
the severity of HPS. HPS can be categorised as mild (PaO 2≥80
mmHg), moderate (PaO 260–79 mmHg), severe (PaO 250–59
mmHg), and very severe (PaO 2<50 mmHg).498,500,501,503
Recently, it has been observed that HPS is associated with ele-
vated von Willenbrand factor antigen (vWF-Ag) levels. Thus,
vWF-Ag has been proposed as a potentially useful screening tool
for early detection of HPS, but further studies are needed to val-idate it.
529The chest X ray is usually non-specific, nevertheless,
it can be used to effectively rule out other concomitant pul-
monary diseases since only a mild interstitial pattern in thelower part of the lungs may be found, because of pulmonaryvasodilatation.
498,500,501,503A decrease in the single-breath dif-
fusing capacity for carbon monoxide is the only alteration of
the routine pulmonary function test that is frequently and con-sistently abnormal in patients with HPS. However, it is not
specific and it may not normalise after LT.
498,500,501,503All the
other respiratory function tests are non-specific, showing nor-mal or reduced forced vital capacity or maximum forced expira-
tory volume during the first second (FEV1). Thus, they can only
be used to rule out other concomitant pulmonary diseases. Tho-racic CT scans have also been proposed as a complementary
technique to rule out another underlying pulmonary pathol-
ogy,
498,499although there is little information regarding their
specific role in the diagnosis of HPS. It has been suggested thatthoracic CT scans can be useful to measure the calibre of the
peripheral arteries and the bronchial/arterial relationship.
530,531
Furthermore, CT scanning makes it possible to define the vascu-
lar pattern of HPS in a similar manner to arteriography by
detecting pleural and pulmonary arteriovenous communica-
tions. Contrast-enhanced transthoracic echocardiography withsaline (shaken to produce microbubbles >10
lm in diameter)
is the most useful method to detect pulmonary vascular dilata-
tion. After the administration of agitated saline in a peripheralvein, microbubble opacification of the left atrium within threeto six cardiac cycles after right-atrial opacification indicates
microbubble passage through an abnormally dilated vascular
bed, since microbubbles do not pass through normal capillar-ies.
532The injection of technetium-99 m–labeled macro-aggre-
gated albumin (MAA) in the peripheral vein for lung scanning
(MAA scan) is a potential alternative diagnostic procedurealthough it is more invasive and less sensitive. Particles, with
a 20–50
lm size, escape through the abnormal pulmonary cap-
illaries and stay in downstream capillary beds supplied by sys-temic arteries, such as the brain, kidneys, and spleen.
Quantitative imaging of the MAA scan in the brain and lung
enables calculation of the degree of shunting.
533,534The mea-
surement of shunting with MAA scans may be useful as a com-plementary diagnostic tool in patients with HPS in two clinical
situations. Firstly, in patients with a severe hypoxaemia and a
coexistent HPS and intrinsic lung disease since a shunting >6%at MAA scan proves the major contribution of HPS to hypox-
aemia. Secondly, in patients with HPS and very severe hypox-
aemia (PaO
2<50 mmHg), since the presence of shunting >20%
is associated with a poor outcome after LT.535Despite the
potential role of lung perfusion scintigraphy for prognostic use
in patients with cirrhosis and IPVD, its diagnostic accuracy forHPS remains to be established.
536Finally, neither constrast eco-cardiography nor MAA scan can differentiate discrete arteriove-
nous communications from diffuse precapillary and capillarydilatations or intracardiac shunt. The former distinction can bemade by means of pulmonary angiography. The latter distinc-
tion can be made by means of transoesophageal contrast-
enhanced echocardiography that directly reveals the intra-atrialseptum. Pulmonary angiography should not be performed in all
patients with suspected HPS, but only in: a) patients with the
severe hypoxaemia (PaO
2<60 mmHg) poorly responsive to
administration of 100% oxygen, and b) patients strongly sus-
pected (by means of a CT chest scan) of having arteriovenous
communications that would be amenable to embolisation.
Recommendations
/C15In presence of tachypnoea and polypnoea, digital club-
bing and/or cyanosis in a patient with the hallmarks ofchronic liver disease, HPS should be suspected and
investigated (II-2,1) .
/C15Pulse oximetry is the screening tool for HPS in adult
patients, but not in paediatric patients. For patients with
SpO
2<96%, ABG analysis should be performed. A PaO 2
lower than 80 mmHg and or an alveolar-arterial oxygengradient (P[A-a]O
2) ≥15 mmHg while breathing ambient
air, should lead to further investigations. For adults ≥65
years a P[A-a]O 2≥20 mmHg cut-off should be used (II-2,1) .
/C15The use of contrast (microbubble) echocardiography to
characterise HPS is recommended (II-2;1) .
/C15Trans-oesophageal contrast-enhanced echocardiography
can be performed to exclude definitively intra-cardiacshunts, albeit this technique is not devoid of risks
(II-2;2) .
/C15An MAA scan should be performed as a complementary
tool to quantify the degree of shunting in patients with
severe hypoxaemia and coexistent intrinsic lung disease,
or to assess the prognosis in patients with HPS and verysevere hypoxaemia (PaO
2<50 mmHg) (II-2;1) .
/C15Neither contrast echocardiography nor MAA scan can
definitively differentiate discrete arteriovenous commu-
nications from diffuse precapillary and capillary dilata-
tions or cardiac shunts. Pulmonary angiography shouldbe performed only in patients with the severe hypox-
aemia (PaO
2<60 mmHg), poorly responsive to adminis-
tration of 100% oxygen, and in whom there is a strongsuspicion of arteriovenous communications that areamenable to embolisation (II-2;1) .
Natural history
The natural history of IPVD as well as of HPS is still unclear.
Most patients with IPVD maintain a normal gas exchange over
time, and it is not clear the reason why a subset of patients withIPVD develops HPS.
537A diagnosis of HPS is associated with a
poor outcome in terms of both survival and quality of
life.505,507,508Regarding survival, it should be highlighted that
in patients undergoing evaluation for LT, the mortality ratewas almost double in patients with HPS compared to patients
with cirrhosis without HPS, independent of other potential pre-
dictors of mortality such as age, MELD score and comorbidi-ties.
505In patients with cirrhosis and HPS, who were notClinical Practice Guidelines
38 Journal of Hepatology 2018 vol. xxx jxxx–xxx
Please cite this article in press as: The European Association for the Study of the Liver. EASL Clinical Practice Guidelines for the management of pati ents with decompensated cirrhosis. J
Hepatol (2018), https://doi.org/10.1016/j.jhep.2018.03.024

evaluated for LT, the five-year survival rate was 23% while it was
63% in patients with cirrhosis without HPS who were matchedfor the aetiology and severity of cirrhosis according to theChild-Pugh classification, age, and MELD score.
503Survival was
significantly worse among patients with HPS and a PaO 2of less
than 50 mmHg at the time of diagnosis.506,507
Management
Medical treatment
Spontaneous resolution of HPS is uncommon. There is no estab-
lished medical therapy currently available for HPS. Several drugs
have been applied forthe treatment of HPS with conflicting results.However, no large randomised trial has been conducted, probably
because of the low number of patients. Data from several uncon-
trolled clinical studies and anecdotal evidence indicate that treat-ment with beta-blockers, cyclooxygenase inhibitors, systemicglucocorticoids and cyclophosphamide, almitrine bismesylate,
inhaled nitric oxide, nitric oxide inhibitors, and antimicrobial
agents has been uniformly unsuccessful.
505Pentoxifylline has also
been tried in the treatment of HPS in adults and children in two
small pilot studies with contradictory results in terms of improve-
ments in oxygenation and frequent GI side effects.538,539Adminis-
tration of garlic was found to be associated with an improvement
in the PaO 2in a small randomised study.540However, a case of
moderate hepatotoxicity associated with short-term, high-dosegarlicin therapy in an LT recipient with persistent HPS wasrecently reported.
541The use of TIPS has been proposed to reduce
portal pressure in patients with HPS. However, data are insuffi-
cient even when a systemic analysis review is considered.542In
addition, there is some concern that TIPS can enhance pulmonary
vasodilation by exacerbating the hyperkinetic circulation. Thus, no
recommendation for the use TIPS to treat HPS can be given.498,505
Finally, coil embolisation (embolotherapy) has been shown to
improve arterial oxygenation temporarily in the context of angio-
graphic arteriovenous communications.531,543Endothelin-1
receptor antagonists or angiogenesis inhibitors have not been
tested up to now in patients with HPS. Thus, long-term oxygen
therapy remains the most frequently recommended therapy forsymptoms in patients with severe hypoxaemia. However, someaspects of this treatment such as efficacy, costs, and compliance,
remain to be evaluated.
Recommendations
/C15Long-term oxygen therapy is recommended in patients
with HPS and severe hypoxaemia. Nevertheless, there
is no available data concerning effectiveness, tolerance,cost-effectiveness, compliance and effects on survival
rates of this therapy (II-2;1) .
/C15No recommendation can be proposed regarding the use
of drugs or the placement of TIPS for the treatment of
HPS (I;1).
Liver transplantation
The most common and the only successful treatment for HPS is
LT. LT results in a complete reversal or in a significant improve-
ment of HPS in more than 85% of patients with severe hypox-aemia.
544In a prospective clinical study performed in the pre-
MELD era, a pre-LT severe hypoxaemia, in particular when itwas associated with a large shunting at MAA scan, was found
to be a very strong predictor of mortality after LT.535In 2007,
five years after the introduction of MELD in US, the United Net-work for Organ Sharing (UNOS) recommended assigning an
MELD score of 22 for the initial application of patients with sev-
ere HPS (PaO
2<60 mmHg), with further increases every three
months, to balance pre- and post-LT outcomes between HPS
and non-HPS candidates.545In the largest retrospective study
comparing the results of LT between the pre-MELD era andthe MELD era in patients with HPS, the five-year survival rate
after LT was found to improve from 67% during the pre-MELD
era to 88% in the MELD era.
546Other data showed that in
post-MELD era, there was no association between pre-LT oxy-
genation and waitlist survival in patients with HPS. These find-
ings reflect not only the results of the introduction of HPS as aMELD exception, but also of an improved perioperative manage-ment in patients with HPS. The regular assessment of the sever-
ity of hypoxaemia may facilitate LT prior to the occurrence of
very severe hypoxaemia. In fact, hypoxaemia can worsen inpatients with HPS who are on the active transplantation list,
with a median decrease in pO
2of 5.2 mmHg per year,508and
it has been recently confirmed that a pre-LT room-air PaO 2
≤44.0 mmHg is still associated with increased post-LT mortal-
ity.547Thus, it has been suggested that an ABG analysis should
be carried out every six months, but no study has clarifiedwhich is the best method for conducting this (ABG analysis vs.
pulse oximetry) nor how frequently it should be performed.
Despite the increased survival rate in patients with HPS afterLT in the MELD era, it has recently been observed that HPSMELD exception patients had lower overall mortality compared
to others awaiting LT, suggesting that the appropriateness of the
HPS MELD exception policy should be reassessed.
548There are
very few and small studies on the impact of HPS on anaesthetic
procedures, as well as in the post-LT management in the ICU.
Nevertheless, it seems that inhaled nitric oxide, methylene blue,extracorporeal membrane oxygenation and non-invasive venti-
lation may improve oxygenation immediately post-LT.
549–551
Recommendations
/C15Patients with HPS and PaO 2<60 mmHg should be evalu-
ated for LT since it is the only treatment for HPS that has
been proven to be effective to date (II-2;1) .
/C15Since a severe hypoxaemia (PaO 2<45–50 mmHg) is
associated with increased post-LT mortality, an ABG
analysis should be carried out every six months in orderto facilitate prioritisation to LT (II-2;1) .
Portopulmonary hypertension
Definition and diagnosis
A diagnosis of PPHT should be considered in a patient with
established portal hypertension in the absence of other causes
of pulmonary artery or venous hypertension. namely: chronicthromboembolism, chronic lung disease/hypoxia; chronic leftheart disease.
Patients may be asymptomatic but often present with exer-
tional dyspnoea and they may have clinical signs of right heartfailure when moderate to severe disease develops.
552Classifica-
tion of severity is based on mean pulmonary arterial pressureJOURNAL
OF HEPATOLOGY
Journal of Hepatology 2018 vol. xxx jxxx–xxx 39
Please cite this article in press as: The European Association for the Study of the Liver. EASL Clinical Practice Guidelines for the management of pati ents with decompensated cirrhosis. J
Hepatol (2018), https://doi.org/10.1016/j.jhep.2018.03.024

(mPAP) and assumes there is high pulmonary vascular resis-
tance (PVR). PPHT is graded as mild (mPAP ≥25 and <35mmHg); moderate (mPAP ≥35 and <45 mmHg), and severe(mPAP ≥45 mmHg).
498The diagnosis also requires there to be
normal pulmonary occlusion pressures, to exclude elevation of
pulmonary pressure resulting from elevated left ventricular fill-ing pressure. Transthoracic Doppler echocardiography (TDE) is
the main screening tool for evaluating the presence of PPHT
when screening high-risk patients, such as those being consid-ered for TIPS or LT.
553–555As a screening test, some studies sug-
gest a pulmonary artery systolic pressure of >30 mmHg on TDE
has a negative predictive value of 100%, but a positive predictivevalue of only 59%.
554However, when assessing patients for LT,
the threshold for right heart catheterisation is less clear, with
a right ventricular systolic pressure >50 mmHg and/or signifi-cant right ventricular hypertrophy seen as the trigger for thisinvestigation to rule out significant PPHT.
555
Pathophysiology
In patients with portal hypertension, PPHT is thought to arise
from limited blood flow in the pulmonary arterial circulation
because of vasoconstriction. Numerous factors are thought to
be responsible for this including: Changes in endogenous vaso-regulators; increased endothelin 1 and reduced prostacyclin syn-
thase from pulmonary endothelial cells; proliferation of smooth
muscle cells/endothelial activation and platelet aggregation.
Natural history and prognosis
From studies in patients evaluated for LT, the incidence is
thought to be between 3–10% based on haemodynamic criteria.Furthermore, female sex and pre-existing autoimmune liver dis-
ease are thought to be independent risk factors.
556Genetic vari-
ation in oestradiol levels may increase the predisposition topulmonary artery vasoconstriction. Indeed, women are at three
times greater risk than men.
557There is also an association
between patients who have moderate to severe PPHT and thepresence of large portosystemic shunts.
558However, there is
no clear association between the severity of liver disease or por-
tal hypertension and the development of severe PPHT.556,559
Studies quote survival rates at one year of between 35–46%
without specific treatment.560,561Mortality is often associated
with other complications of liver disease such as hepatocellular
cancer, sepsis and GI bleeding and right ventricular failure.Increased rates of mortality are related to higher right atrial
pressure and lower cardiac index.
559,562In a multicentre reg-
istry study, patients with PPHT were shown to have worse out-comes than patients with idiopathic pulmonary hypertension,with a five-year survival of 40% vs.64%.
563However, a retro-
spective French study challenges this, whilst reporting
increased mortality in those with a lower cardiac index, likelyreflecting failed compensation to increased right ventricular
dysfunction, and patients with more advanced liver disease.
564
Medical treatment
The evidence base for pharmacological therapies in PPHT is lim-
ited with most data extrapolated from studies in pulmonary
arterial hypertension not related to liver disease.565,566Drugs
to promote acute vasodilatation during right heart catheterisa-
tion assessment, theoretically may be deleterious as they run
the risk of further reducing cardiac index. There is a lack of datato clarify this.
567Conversely, whilst patients with advancedportal hypertension may be on treatment with beta-blockers,
withdrawing beta-blocker therapy may help to increase cardiacoutput and thereby help exertional dyspnoea, in patients withadvanced PPHT.
568
Endothelin receptor antagonists .Bosentan has been shown to
improve pulmonary artery haemodynamics and exercise toler-ance in patients with PPHT, independently of liver disease
severity.
569–572One retrospective study reports survival rates
of up to 89% at three years.573Others have shown improve-
ments in cardiac index up to 39%, albeit in a small number ofpatients, but an increase in aminotransferases, which responded
to dose reduction or discontinuation.
571The FDA places a cau-
tion on this class of drug in patients with advanced liver dys-function. There is limited data on the use of other members in
this family of agent, including ambrisentan and macitentan,
for PPHT.
574,575
Phosphodiesterase subtype-5 inhibitors .Blockade of phosphodi-
esterase-5 inhibitors facilitate the vasodilatory effects of nitric
oxide, through reduced metabolism of cGMP. Small case seriessuggest that sildenafil improves functional capacity andincreases cardiac output.
576–578It should be noted that silde-
nafil can precipitate variceal bleeding and as such, caution
should be exercised.579
Prostacyclin analogues .Prostacyclin analogues have many poten-
tial benefits including vasodilatory, reduced vascular smooth
muscle proliferation and anti-thrombotic. Case series suggestimproved pulmonary haemodynamics with i.v. epoprostenoland the potential for improved five-year survival compared to
registry data in pulmonary artery hypertension (70 vs.
40%).
580–583However, lower doses than those used in idiopathic
pulmonary hypertension are suggested to reduce the develop-
ment of thrombocytopenia and splenomegaly. Other studies
have also looked at use of inhaled iloprost and reported short-term haemodynamic benefit.
584
Impact of the management of other complications of cirrhosis
Caution should be exercised when considering TIPS placement
for the treatment of other complications of cirrhosis in patients
with proven PPHT. The anticipated increase in right ventricular
filling pressures and cardiac output may precipitate markedincreases in PVR and right-sided pressure overload.
585,586
Moderate PPHT (mPAP >35 and <45 mmHg) is a relative con-
traindication for TIPS placement, and severe PPHT is an absolute
contraindication.586
Recommendations
/C15Screening for PPHT should be via TDE in patients deemed
potential recipients for TIPS or LT; in those with a posi-tive screening test, right heart catheterisation should
be performed (II-1;1) .
/C15In patients with PPHT who are listed for transplantation,
echocardiography should be repeated on the waitlist,
albeit, the specific interval is unclear (III;1) .
/C15Beta-blockers should be stopped and varices managed
by endoscopic therapy in cases of proven PPHT (II-3;1) .Clinical Practice Guidelines
40 Journal of Hepatology 2018 vol. xxx jxxx–xxx
Please cite this article in press as: The European Association for the Study of the Liver. EASL Clinical Practice Guidelines for the management of pati ents with decompensated cirrhosis. J
Hepatol (2018), https://doi.org/10.1016/j.jhep.2018.03.024

/C15Therapies that have been approved for primary pul-
monary arterial hypertension may have benefit in PPHT
to improve exercise tolerance and haemodynamics. How-ever, endothelin antagonists should be used with caution
because of concerns over hepatic impairment (II-2;1) .
/C15TIPS should not be used in patients with PPHT (II-3;1) .
Liver transplantation
Historically, severe PPHT has been a relative contraindication
for LT because of very poor outcomes. However, with the adventof improved haemodynamic control with agents such as i.v.
prostacyclin, there are case series showing normal pulmonary
haemodynamics almost two years post LT.
587,588
Stratifying risk for LT .In patients with an mPAP ≥45–50 mmHg,
most centres would deem this an absolute contraindication to
transplantation irrespective of therapy applied.562,587,589
Patients with an mPAP >35 have increased risk post LT, associ-
ated with increased hospital stay and longer ventilator require-
ments.562,590,591If LT is considered in such patients, it is
suggested that their PPHT is treated aggressively to lower mPAPand improve right ventricular function.
588,592,593
To facilitate access to LT before there is further progression
of PPHT to a point where transplantation risks are deemed toohigh, MELD exception (MELD 22 points) has been granted to
patients with PPHT (mPAP >25 mmHg and PVR >240 dynes/s
per cm
/C05) with at least moderate disease severity (baseline
mPAP >35 mmHg).594Patients are considered surgical candi-
dates if, after targeted therapy to lower PAP, they have
improved mPAP (<35 mmHg) and PVR (<400 dyne/s per cm/C05)
and/or normalise their PVR. Applying this exception has beennoted to reduce waitlist mortality.
595
Per-operative considerations .All patients should be monitored
with a pulmonary artery catheter. Therapy to lower mPAPshould be continued throughout the operative period, given that
there is often a rise in cardiac output post re-perfusion and this
may add more stress on any pre-existing impaired right ventri-cle function.
595–597Indeed, in some cases, a severe acute rise in
PAP may lead to graft failure because of hepatic congestion
through a failing right ventricle. The management of such
adverse haemodynamics, in addition to i.v. prostacyclin orinhaled nitric oxide includes the use of extracorporeal mem-
brane oxygen therapy (ECMO).
598,599
Postoperative considerations .Monitoring PAP response to ther-
apy is via serial transthoracic echo with tissue Doppler at 4–6-month intervals and consideration of tapering pulmonary
artery targeted therapy, though no controlled data exists to
provide guidance on this.
581,600,601Case reports and series
suggest that 29–64% of patients with moderate to severe
PPHT under long-term follow-up post-transplant have been
able to discontinue therapy over time.599–602Indeed, some
suggest a return to normal right ventricle function following
therapy for PPHT in the pre-transplant period and then after
transplant surgery.581,601PPHT MELD exception patients have
worse one-year mortality or graft failure than patients with-
out PPHT.603Recommendations
/C15If mPAP <35 mmHg and right ventricular function is pre-
served, LT should be considered (II-2,1) . A mPAP of ≥45
mmHg should be considered an absolute contraindica-
tion to LT irrespective of therapy applied (III,1) .
/C15Therapy to lower mPAP and improve right ventricular
function should be commenced in patients with mPAP≥35 mmHg. Right ventricular function should be period-
ically evaluated (II-2,1) .
/C15MELD exception can be considered in patients with proven
PPHT in whom targeted therapy fails to decrease mPAP
<35 mmHg but does facilitate normalisation of PVR to
<240 dynes/s cm
/C05and right ventricular function (II-3;2) .
/C15MELD exception should be advocated in patients with
proven PPHT of moderate severity (assessment mPAP≥35 mmHg) in whom targeted treatment lowers mPAP
<35 mmHg and PVR <400 dynes/s cm
/C05(II-2;1) .
Conclusions
These guidelines on the management of patients with decom-
pensated cirrhosis were developed based on a new pathophys-iological background that offers the opportunity for more
comprehensive therapeutic or prophylactic approaches to man-
age the disease. The knowledge of the key pathophysiologicmechanisms makes it possible nowadays to counteract the pro-gression of cirrhosis and so to prevent its complications. This
represents a step forward, shifting our approach from treating
the complications of decompensated cirrhosis to preventingtheir occurrence. However, to make this possible it is crucial
to think about new models of specialist care for patients with
cirrhosis. A care coordination programme, has been proven toimprove survival and to reduce emergent readmission to the
hospital in these patients.
604Care coordinators can facilitate
the development of educational programmes for patients andcaregivers optimising their adherence to guideline recommen-dations. In addition, they can plan invasive procedures in a
day hospital, allowing transfer of real-time information to pri-
mary care physicians to improve quality and coordination ofcare. In so doing, it is possible to prevent unnecessary visits to
the emergency department and/or emergent readmission to
the hospital. These measures will progressively reduce the bur-den of cirrhosis.
Conflict of interest
Paolo Angeli: Consultancy fee from Sequana Medical AG, Gilead
Italy and Biovie; Patent inventor from Biovie; Research grant
from Gilead; Speaker’s fee from Bhering, Kedrion 2016. MauroBernardi: Consultancy fee from CLS Behring GmbH, Baxter
Healthcare SA, Grifols SA; Speaker’s fee from CLS Behring GmbH,
Baxter Healthcare SA, PPTA Europe, Octapharma AG, GileadSciences, AbbVie Italia. Wim Laleman: Speaker’s fee for Gore,
Norgine, 4C, Abbvie, Sirtex; Consultancy fee for AbbVie, Gilead,
MSD, Intercept; Research grant from Gilead. Jonel Trebicka:Speaker’s fee or Consultancy fee from Gore & associates (TIPS),
Sequana medical (alpha-pump), Alexion (PNH), Versantis (lipo-
somes). Aleksander Krag: None. Claire Francoz: None. PereGines: Advisory/Consultancy fee for Sequana Grifols, Mallinck-JOURNAL
OF HEPATOLOGY
Journal of Hepatology 2018 vol. xxx jxxx–xxx 41
Please cite this article in press as: The European Association for the Study of the Liver. EASL Clinical Practice Guidelines for the management of pati ents with decompensated cirrhosis. J
Hepatol (2018), https://doi.org/10.1016/j.jhep.2018.03.024

rodt, Ferring Pharmaceuticals; Research Funding from Sequana,
Grifols, Ferring Pharmaceuticals.
Please refer to the accompanying ICMJE disclosure forms for
further details.
Acknowledgments
We would like to thank Alessandra Brocca Dr, Marta Tonon MD,
Husain-Syed Faeq MD for the great editorial work. We acknowledge
ICREA for the ACADEMIA AWARD given to Pere Ginès. We would like
to thank the reviewers of this Clinical Practice Guideline for their
time and critical reviewing: EASL Governing Board, Alexander
Gerbes, Thierry Gustot, Guadalupe Garcia-Tsao.
References
[1]EASL clinical practice guidelines on the management of ascites,
spontaneous bacterial peritonitis, and hepatorenal syndrome in cirrho-
sis. J Hepatol 2010;53:397–417 .
[2]D’Amico G. The clinical course of cirrhosis. Population based studies andthe need of personalized medicine. J Hepatol 2014;60:241–242
.
[3]Moreau R, Jalan R, Gines P, Pavesi M, Angeli P, Cordoba J, et al. Acute-on-chronic liver failure is a distinct syndrome that develops in patients
with AD of cirrhosis. Gastroenterology 2013;144:1426–1437
.
[4]D’Amico G, Garcia-Tsao G, Pagliaro L. Natural history and prognosticindicators of survival in cirrhosis: a systematic review of 118 studies. JHepatol 2006;44:217–231
.
[5]Bernardi M, Moreau R, Angeli P, Schnabl B, Arroyo V. Mechanisms ofdecompensation and organ failure in cirrhosis: From peripheral arterialvasodilation to systemic inflammation hypothesis. J Hepatol
2015;63:1272–1284
.
[6]Jalan R, Fernandez J, Wiest R, Schnabl B, Moreau R, Angeli P, et al.Bacterial infections in cirrhosis: a position statement based on the EASLSpecial Conference. J Hepatol 2013;60:1310–1324
.
[7]Arvaniti V, D’Amico G, Fede G, Manousou P, Tsochatzis E, PleguezueloM, et al. Infections in patients with cirrhosis increase mortality four-
fold and should be used in determining prognosis. Gastroenterology
2010;139:1246–1256
.
[8]Schrier RW, Arroyo V, Bernardi M, Epstein M, Henriksen JH, Rodes J.Peripheral arterial vasodilation hypothesis: a proposal for the initiationof renal sodium and water retention in cirrhosis. Hepatology1988;8:1151–1157
.
[9]Wiese S, Hove JD, Bendtsen F, Moller S. Cirrhotic cardiomyopathy:
pathogenesis and clinical relevance. Nat Rev Gastroenterol Hepatol
2014;11:177–186 .
[10] Arroyo V, Terra C, Gines P. Advances in the pathogenesis and treatmentof type-1 and type-2 hepatorenal syndrome. J Hepatol 2007;46:935–946
.
[11] Iwakiri Y, Groszmann RJ. The hyperdynamic circulation of chronic liver
diseases: from the patient to the molecule. Hepatology 2006;43:
S121–S131 .
[12] Claria J, Stauber RE, Coenraad MJ, Moreau R, Jalan R, Pavesi M, et al.Systemic inflammation in decompensated cirrhosis: Characterizationand role in acute-on-chronic liver failure. Hepatology 2016;64:1249–1264
.
[13] Trautwein C, Friedman SL, Schuppan D, Pinzani M. Hepatic fibrosis:
Concept to treatment. J Hepatol 2015;62:S15–S24 .
[14] Alvarez MA, Cirera I, Sola R, Bargallo A, Morillas RM, Planas R.Long-term clinical course of decompensated alcoholic cirrhosis: aprospective study of 165 patients. J Clin Gastroenterol 2011;45:906–911
.
[15] Powell Jr WJ, Klatskin G. Duration of survival in patients with Laennec’s
cirrhosis. Influence of alcohol withdrawal, and possible effects of recent
changes in general management of the disease. Am J Med1968;44:406–420
.
[16] Shim JH, Lee HC, Kim KM, Lim YS, Chung YH, Lee YS, et al. Efficacy ofentecavir in treatment-naive patients with hepatitis B virus-relateddecompensated cirrhosis. J Hepatol 2010;52:176–182
.
[17] Cheung MC, Walker AJ, Hudson BE, Verma S, McLauchlan J, Mutimer DJ,
et al. Outcomes after successful direct-acting antiviral therapy forpatients with chronic hepatitis C and decompensated cirrhosis. JHepatol 2016;65:741–747
.
[18] Lens S, Alvarado E, Mariño Z, et al. Effects of all-oral antiviraltherapy on HVPG and systemic hemodynamics in patients withhepatitis C virus-associated cirrhosis. Gastroenterology 2017;153:1273–1283
.
[19] Kang SH, Lee YB, Lee JH, Nam JY, Chang Y, Cho H, et al. Rifaximintreatment is associated with reduced risk of cirrhotic complications andprolonged overall survival in patients experiencing hepaticencephalopathy. Aliment Pharmacol Ther 2017;46:845–855
.
[20] Ginès P, Schrier RW. Renal failure in cirrhosis. N Engl J Med
2009;361:1279–1290 .
[21] Moreau R, Elkrief L, Bureau C, Pararnau JM, Thavenot T, Saliba F, et al. Arandomized trial of 6-month norfloxacin therapy in patients withChild-Pugh class C cirrhosis. J Hepatol 2017;66:S1
.
[22] Caraceni P, Riggio O, Angeli P, Alessandria C, Neri S, Foschi FG, et al.
Long-term albumin administration in decompensated cirrhosis: an
open label randomized trial. Lancet 2018, [In press] .
[23] Sola E, Sola C, Simon-Talero M, Martin-Llahi M, Castellote J, Garcia-Martinez R, et al. Midodrine and albumin for prevention of complica-tions of cirrhosis in patients in the waiting list for liver transplantation.A randomized, multicenter, double-blind, placebo-controlled trial. JHepatol 2017;66:S11
.
[24] Abraldes JG, Albillos A, Banares R, Turnes J, Gonzalez R, Garcia-Pagan JC,
et al. Simvastatin lowers portal pressure in patients with cirrhosis andportal hypertension: a randomized controlled trial. Gastroenterology2009;136:1651–1658
.
[25] Abraldes JG, Villanueva C, Aracil C, Turnes J, Hernandez-Guerra M,Genesca J, et al. Addition of simvastatin to standard therapy for the
prevention of variceal rebleeding does not reduce rebleeding but
increases survival in patients with cirrhosis. Gastroenterology2016;150:1160–1170
.
[26] Villa E, Cammà C, Marietta M, Luongo M, Critelli R, Colopi S, et al.Enoxaparin prevents portal vein thrombosis and liver decompensationin patients with advanced cirrhosis. Gastroenterology 2012;143:1253–12609
.
[27] Lebrec D, Thabut D, Oberti F, Perarnau JM, Condat B, Barraud H, et al.
Pentoxifylline does not decrease short-term mortality but does reducecomplications in patients with advanced cirrhosis. Gastroenterology2010;138:1755–1762
.
[28] Abraldes JG, Tarantino I, Turnes J, Garcia-Pagan JC, Rodes J, Bosch J.Hemodynamic response to pharmacological treatment of portal hyper-
tension and long-term prognosis of cirrhosis. Hepatology 2003;37:
902–908
.
[29] Gines P, Quintero E, Arroyo V, Teres J, Bruguera M, Rimola A, et al.Compensated cirrhosis: natural history and prognostic factors. Hepa-tology 1987;7:122–128
.
[30] Ripoll C, Groszmann R, Garcia-Tsao G, Grace N, Burroughs A, Planas R,et al. Hepatic venous pressure gradient predicts clinical decompensa-
tion in patients with compensated cirrhosis. Gastroenterology
2007;133:481–488
.
[31] Moore KP, Wong F, Gines P, Bernardi M, Ochs A, Salerno F, et al. Themanagement of ascites in cirrhosis: report on the consensus conferenceof the International Ascites Club. Hepatology 2003;38:258–266
.
[32] Arroyo V, Gines P, Gerbes AL, Dudley FJ, Gentilini P, Laffi G, et al.
Definition and diagnostic criteria of refractory ascites and hepatorenal
syndrome in cirrhosis. International Ascites Club. Hepatology1996;23:164–176
.
[33] Rimola A, Garcia-Tsao G, Navasa M, Piddock LJ, Planas R, Bernard B,et al. Diagnosis, treatment and prophylaxis of spontaneous bacterialperitonitis: a consensus document. International Ascites Club. J Hepatol2000;32:142–153
.
[34] Bruns T, Lutz P, Stallmach A. Nischalke HD Low ascitic fluid protein does
not indicate an increased risk for spontaneous bacterial peritonitis incurrent cohorts. J Hepatol 2015;63:527–528
.
[35] Runyon BA, Canawati HN, Akriviadis EA. Optimization of ascitic fluidculture technique. Gastroenterology 1988;95:1351–1355
.
[36] Runyon BA, Montano AA, Akriviadis EA, Antillon MR, Irving MA,
McHutchison JG. The serum-ascites albumin gradient is superior to
the exudate-transudate concept in the differential diagnosis of ascites.Ann Intern Med 1992;117:215–220
.
[37] Gerbes AL, Jüngst D, Xie YN, Permanetter W, Paumgartner G. Asciticfluid analysis for the differentiation of malignancy-related and nonma-lignant ascites. Proposal of a diagnostic sequence. Cancer 1991;68:1808–1814
.
[38] Llach J, Gines P, Arroyo V, Rimola A, Tito L, Badalamenti S, et al.
Prognostic value of arterial pressure, endogenous vasoactive systems,and renal function in cirrhotic patients admitted to the hospital for thetreatment of ascites. Gastroenterology 1988;94:482–487
.Clinical Practice Guidelines
42 Journal of Hepatology 2018 vol. xxx jxxx–xxx
Please cite this article in press as: The European Association for the Study of the Liver. EASL Clinical Practice Guidelines for the management of pati ents with decompensated cirrhosis. J
Hepatol (2018), https://doi.org/10.1016/j.jhep.2018.03.024

[39] Caregaro L, Menon F, Angeli P, Amodio P, Merkel C, Bortoluzzi A, et al.
Limitations of serum creatinine level and creatinine clearance asfiltration markers in cirrhosis. Arch Intern Med 1994;154:201–205
.
[40] Bernardi M, Gitto S, Biselli M. The MELD score in patients awaiting livertransplant: strengths and weaknesses. J Hepatol 2011;54:1297–1306
.
[41] Biselli M, Dall’Agata M, Gramenzi A, Gitto S, Liberati C, Brodosi L, et al. A
new prognostic model to predict dropout from the waiting list in
cirrhotic candidates for liver transplantation with MELD score <18.Liver Int 2015;35:184–191
.
[42] Bernardi M, Santini C, Trevisani F, Baraldini M, Ligabue A, Gasbarrini G.Renal function impairment induced by change in posture in patientswith cirrhosis and ascites. Gut 1985;26:629–635
.
[43] Ring-Larsen H, Henriksen JH, Wilken C, Clausen J, Pals H, Christensen NJ.
Diuretic treatment in decompensated cirrhosis and congestive heartfailure: effect of posture. Br Med J 1986;292:1351–1353
.
[44] Bernardi M, Laffi G, Salvagnini M, Azzena G, Bonato S, Marra F, et al.Efficacy and safety of the stepped care medical treatment of ascites inliver cirrhosis: a randomized controlled clinical trial comparing twodiets with different sodium content. Liver 1993;13:156–162
.
[45] Gauthier A, Levy VG, Quinton A, Michel H, Rueff B, Descos L, et al. Salt or
no salt in the treatment of cirrhotic ascites: a randomised study. Gut1986;27:705–709
.
[46] Reynolds TB, Lieberman FL, Goodman AR. Advantages of treatment ofascites without sodium restriction and without complete removal ofexcess fluid. Gut 1978;19:549–553
.
[47] Morando F, Rosi S, Gola E, Nardi M, Piano S, Fasolato S, et al. Adherence
to a moderate sodium restriction diet in outpatients with cirrhosis andascites: a real-life cross-sectional study. Liver Int 2015;35:1508–1515
.
[48] Pockros PJ, Reynolds TB. Rapid diuresis in patients with ascites fromchronic liver disease: the importance of peripheral edema. Gastroen-terology 1986;90:1827–1833
.
[49] Bernardi M, Trevisani F, Gasbarrini A, Gasbarrini G. Hepatorenal
disorders: role of the renin-angiotensin-aldosterone system. Semin
Liver Dis 1994;14:23–34 .
[50] Bernardi M, Servadei D, Trevisani F, Rusticali AG, Gasbarrini G.Importance of plasma aldosterone concentration on the natriureticeffect of spironolactone in patients with liver cirrhosis and ascites.Digestion 1985;31:189–193
.
[51] Angeli P, Dalla Pria M, De Bei E, Albino G, Caregaro L, Merkel C, et al.
Randomized clinical study of the efficacy of amiloride and potassiumcanrenoate in nonazotemic cirrhotic patients with ascites. Hepatology1994;19:72–79
.
[52] Angeli P, Gatta A, Caregaro L, Menon F, Sacerdoti D, Merkel C, et al.Tubular site of renal sodium retention in ascitic liver cirrhosis evalu-ated by lithium clearance. Eur J Clin Invest 1990;20:111–117
.
[53] Gatta A, Angeli P, Caregaro L, Menon F, Sacerdoti D, Merkel C. A
pathophysiological interpretation of unresponsiveness to spironolac-tone in a stepped-care approach to the diuretic treatment of ascites innonazotemic cirrhotic patients. Hepatology 1991;14:231–236
.
[54] Perez-Ayuso RM, Arroyo V, Planas R, Gaya J, Bory F, Rimola A, et al.Randomized comparative study of efficacy of furosemide vs.spirono-
lactone in nonazotemic cirrhosis with ascites. Relationship between the
diuretic response and the activity of the renin-aldosterone system.Gastroenterology 1983;84:961–968
.
[55] Angeli P, Fasolato S, Mazza E, Okolicsanyi L, Maresio G, Velo E, et al.Combined vs.sequential diuretic treatment of ascites in non-azotaemic
patients with cirrhosis: results of an open randomised clinical trial. Gut2010;59:98–104
.
[56] Santos J, Planas R, Pardo A, Durandez R, Cabre E, Morillas RM, et al.
Spironolactone alone or in combination with furosemide in thetreatment of moderate ascites in nonazotemic cirrhosis. A randomizedcomparative study of efficacy and safety. J Hepatol 2003;39:187–192
.
[57] Gerbes AL, Bertheau-Reitha U, Falkner C, Jüngst D, Paumgartner G.Advantages of the new loop diuretic torasemide over furosemide in
patients with cirrhosis and ascites. A randomized, double blind cross-
over trial. J Hepatol 1993;17:353–358
.
[58] Angeli P, Albino G, Carraro P, Dalla Pria M, Merkel C, Caregaro L, et al.Cirrhosis and muscle cramps: evidence of a causal relationship.Hepatology 1996;23:264–273
.
[59] Elfert AA, Abo Ali L, Soliman S, Zakaria S, Shehab El-Din I, ElkhalawanyW, et al. Randomized placebo-controlled study of baclofen in the
treatment of muscle cramps in patients with liver cirrhosis. Eur J
Gastroenterol Hepatol 2016;28:1280–1284
.
[60] Lee FY, Lee SD, Tsai YT, et al. A randomized controlled trial of quinidinein the treatment of cirrhotic patients with muscle cramps. J Hepatol1991;12:236–240
.[61] Lin CH, Shih FY, Ma MH, Chiang WC, Yang CW, Ko PC. Should bleeding
tendency deter abdominal paracentesis? Dig Liver Dis 2005;37:946–951
.
[62] Pache I, Bilodeau M. Severe haemorrhage following abdominal para-centesis for ascites in patients with liver disease. Aliment PharmacolTher 2005;21:525–529
.
[63] Gines P, Tito L, Arroyo V, Planas R, Panes J, Viver J, et al. Randomized
comparative study of therapeutic paracentesis with and withoutintravenous albumin in cirrhosis. Gastroenterology 1988;94:1493–1502
.
[64] Gines A, Fernandez-Esparrach G, Monescillo A, Vila C, Domenech E,Abecasis R, et al. Randomized trial comparing albumin, dextran 70, and
polygeline in cirrhotic patients with ascites treated by paracentesis.
Gastroenterology 1996;111:1002–1010
.
[65] Sola-Vera J, Minana J, Ricart E, Planella M, Gonzalez B, Torras X, et al.Randomized trial comparing albumin and saline in the prevention ofparacentesis-induced circulatory dysfunction in cirrhotic patients withascites. Hepatology 2003;37:1147–1153
.
[66] Bernardi M, Caraceni P, Navickis RJ, Wilkes MM. Albumin infusion in
patients undergoing large-volume paracentesis: a meta-analysis of
randomized trials. Hepatology 2012;55:1172–1181 .
[67] Moreau R, Valla DC, Durand-Zaleski I, Bronowicki JP, Durand F, ChaputJC, et al. Comparison of outcome in patients with cirrhosis and ascitesfollowing treatment with albumin or a synthetic colloid: a randomisedcontrolled pilot trail. Liver Int 2006;26:46–54
.
[68] Gines P, Arroyo V, Quintero E, Planas R, Bory F, Cabrera J, et al.
Comparison of paracentesis and diuretics in the treatment of cirrhoticswith tense ascites. Results of a randomized study. Gastroenterology1987;93:234–241
.
[69] Salerno F, Badalamenti S, Incerti P, Tempini S, Restelli B, Bruno S, et al.Repeated paracentesis and i.v. albumin infusion to treat ’tense’ ascitesin cirrhotic patients. A safe alternative therapy. J Hepatol
1987;5:102–108
.
[70] Fernandez-Esparrach G, Guevara M, Sort P, Pardo A, Jimenez W, Gines P,et al. Diuretic requirements after therapeutic paracentesis in non-azotemic patients with cirrhosis. A randomized double-blind trial ofspironolactone vs.placebo. J Hepatol 1997;26:614–620 .
[71] Elia C, Graupera I, Barreto R, Solà E, Moreira R, Huelin P, et al. Severe
acute kidney injury associated with non-steroidal antiinflammatory
drugs in cirrhosis: a case-control study. J Hepatol 2015;63:593–600
.
[72] Claria J, Kent JD, Lopez-Parra M, Escolar G, Ruiz-Del-Arbol L, Gines P,et al. Effects of celecoxib and naproxen on renal function in nona-zotemic patients with cirrhosis and ascites. Hepatology2005;41:579–587
.
[73] Llach J, Gines P, Arroyo V, Salmeron JM, Gines A, Jimenez W, et al. Effect
of dipyridamole on kidney function in cirrhosis. Hepatology
1993;17:59–64 .
[74] Pariente EA, Bataille C, Bercoff E, Lebrec D. Acute effects of captopril onsystemic and renal hemodynamics and on renal function in cirrhoticpatients with ascites. Gastroenterology 1985;88:1255–1259
.
[75] Albillos A, Lledo JL, Rossi I, Perez-Paramo M, Tabuenca MJ, Banares R,
et al. Continuous prazosin administration in cirrhotic patients: effects
on portal hemodynamics and on liver and renal function. Gastroen-terology 1995;109:1257–1265
.
[76] Cabrera J, Arroyo V, Ballesta AM, Rimola A, Gual J, Elena M, et al.Aminoglycoside nephrotoxicity in cirrhosis. Value of urinary beta 2-microglobulin to discriminate functional renal failure from acutetubular damage. Gastroenterology 1982;82:97–105
.
[77] Guevara M, Fernández-Esparrach G, Alessandria C, Torre A, Terra C,
Montañà X, et al. Effects of contrast media on renal function in patientswith cirrhosis: a prospective study. Hepatology 2004;40:646–651
.
[78] Solomon R, Werner C, Mann D, D’Elia J, Silva P. Comparison of saline,mannitol, and furosemide to prevent acute decreases in renalfunction induced by radiocontrast agents. N Engl J Med 1994;331:
1416–1420
.
[79] Salerno F, Borroni G, Moser P, Badalamenti S, Cassara L, Maggi A, et al.Survival and prognostic factors of cirrhotic patients with ascites: astudy of 134 outpatients. Am J Gastroenterol 1993;88:514–519
.
[80] Guardiola J, Baliellas C, Xiol X, Fernandez Esparrach G, Gines P, VenturaP, et al. External validation of a prognostic model for predicting survivalof cirrhotic patients with refractory ascites. Am J Gastroenterol
2002;97:2374–2378
.
[81] Huonker M, Schumacher YO, Ochs A, Sorichter S, Keul J, Rossle M.Cardiac function and haemodynamics in alcoholic cirrhosis and effectsof the transjugular intrahepatic portosystemic stent shunt. Gut1999;44:743–748
.JOURNAL
OF HEPATOLOGY
Journal of Hepatology 2018 vol. xxx jxxx–xxx 43
Please cite this article in press as: The European Association for the Study of the Liver. EASL Clinical Practice Guidelines for the management of pati ents with decompensated cirrhosis. J
Hepatol (2018), https://doi.org/10.1016/j.jhep.2018.03.024

[82] Sanyal AJ, Freedman AM, Luketic VA, Purdum 3rd PP, Shiffman ML,
DeMeo J, et al. The natural history of portal hypertension aftertransjugular intrahepatic portosystemic shunts. Gastroenterology1997;112:889–898
.
[83] Wong F, Sniderman K, Liu P, Allidina Y, Sherman M, Blendis L.Transjugular intrahepatic portosystemic stent shunt: effects on hemo-
dynamics and sodium homeostasis in cirrhosis and refractory ascites.
Ann Intern Med 1995;122:816–822
.
[84] Wong F, Sniderman K, Liu P, Blendis L. The mechanism of the initialnatriuresis after transjugular intrahepatic portosystemic shunt. Gas-troenterology 1997;112:899–907
.
[85] Gerbes AL, Gülberg V, Waggershauser T, Holl J, Reiser M. Renal effects of
transjugular intrahepatic portosystemic shunt in cirrhosis: comparison
of patients with ascites, with refractory ascites, or without ascites.Hepatology 1998;28:683–688
.
[86] Ochs A, Rossle M, Haag K, Hauenstein KH, Deibert P, Siegerstetter V,et al. The transjugular intrahepatic portosystemic stent-shunt proce-dure for refractory ascites. N Engl J Med 1995;332:1192–1197
.
[87] Plauth M, Schutz T, Buckendahl DP, Kreymann G, Pirlich M, Grungreiff S,
et al. Weight gain after transjugular intrahepatic portosystemic shunt is
associated with improvement in body composition in malnourishedpatients with cirrhosis and hypermetabolism. J Hepatol 2004;40:228–233
.
[88] Gülberg V, Liss I, Bilzer M, Waggershauser T, Reiser M, Gerbes AL.Improved quality of life in patients with refractory or recidivant ascites
after insertion of transjugular intrahepatic portosystemic shunts.
Digestion 2002;66:127–130
.
[89] Casado M, Bosch J, Garcia-Pagan JC, Bru C, Banares R, Bandi JC, et al.Clinical events after transjugular intrahepatic portosystemic shunt:correlation with hemodynamic findings. Gastroenterology1998;114:1296–1303
.
[90] Riggio O, Angeloni S, Salvatori FM, De Santis A, Cerini F, Farcomeni A,
et al. Incidence, natural history, and risk factors of hepatic encephalopa-
thy after transjugular intrahepatic portosystemic shunt with polyte-trafluoroethylene-covered stent grafts. Am J Gastroenterol2008;103:2738–2746
.
[91] Sauerbruch T, Mengel M, Dollinger M, Zipprich A, Rossle M, Panther E,et al. Prevention of rebleeding from esophageal varices in patients with
cirrhosis receiving small-diameter stents vs.hemodynamically con-
trolled medical therapy. Gastroenterology 2015;149:660–668
.
[92] Wang Q, Lv Y, Bai M, Wang Z, Liu H, He C, et al. Eight millimetre coveredTIPS does not compromise shunt function but reduces hepaticencephalopathy in preventing variceal rebleeding. J Hepatol 2017;67:508–516
.
[93] Pieper CC, Jansen C, Meyer C, Nadal J, Lehmann J, Schild HH, et al.
Prospective evaluation of passive expansion of partially dilated tran-
sjugular intrahepatic portosystemic shunt stent grafts-a three-dimen-sional sonography study. J Vasc Interv Radiol 2017;28:117–125
.
[94] Bureau C, Garcia-Pagan JC, Otal P, Pomier-Layrargues G, Chabbert V,Cortez C, et al. Improved clinical outcome using polytetrafluo-roethylene-coated stents for TIPS: results of a randomized study.
Gastroenterology 2004;126:469–475
.
[95] Gines P, Uriz J, Calahorra B, Garcia-Tsao G, Kamath PS, Del Arbol LR,et al. Transjugular intrahepatic portosystemic shunting vs.paracentesis
plus albumin for refractory ascites in cirrhosis. Gastroenterology2002;123:1839–1847
.
[96] Lebrec D, Giuily N, Hadengue A, Vilgrain V, Moreau R, Poynard T, et al.Transjugular intrahepatic portosystemic shunts: comparison with
paracentesis in patients with cirrhosis and refractory ascites: a
randomized trial. French Group of Clinicians and a Group of Biologists.J Hepatol 1996;25:135–144
.
[97] Narahara Y, Kanazawa H, Fukuda T, Matsushita Y, Harimoto H, KidokoroH, et al. Transjugular intrahepatic portosystemic shunt vs.paracentesis
plus albumin in patients with refractory ascites who have good hepatic
and renal function: a prospective randomized trial. J Gastroenterol
2011;46:78–85
.
[98] Rossle M, Ochs A, Gulberg V, Siegerstetter V, Holl J, Deibert P, et al. Acomparison of paracentesis and transjugular intrahepatic portosys-temic shunting in patients with ascites. N Engl J Med 2000;342:1701–1707
.
[99] Salerno F, Merli M, Riggio O, Cazzaniga M, Valeriano V, Pozzi M, et al.
Randomized controlled study of TIPS vs.paracentesis plus albumin in
cirrhosis with severe ascites. Hepatology 2004;40:629–635 .
[100] Sanyal AJ, Genning C, Reddy KR, Wong F, Kowdley KV, Benner K, et al.The North American study for the treatment of refractory ascites.Gastroenterology 2003;124:634–641
.[101] Albillos A, Banares R, Gonzalez M, Catalina MV, Molinero LM. A meta-
analysis of transjugular intrahepatic portosystemic shunt vs.paracen-
tesis for refractory ascites. J Hepatol 2005;43:990–996 .
[102] Bai M, Qi XS, Yang ZP, Yang M, Fan DM, Han GH. TIPS improves livertransplantation-free survival in cirrhotic patients with refractoryascites: an updated meta-analysis. World J Gastroenterol
2014;20:2704–2714
.
[103] Chen RP, Zhu Ge XJ, Huang ZM, Ye XH, Hu CY, Lu GR, et al. Prophylacticuse of transjugular intrahepatic portosystemic shunt aids in thetreatment of refractory ascites: metaregression and trial sequentialmeta-analysis. J Clin Gastroenterol 2014;48:290–299
.
[104] D’Amico G, Luca A, Morabito A, Miraglia R, D’Amico M. Uncovered
transjugular intrahepatic portosystemic shunt for refractory ascites: a
meta-analysis. Gastroenterology 2005;129:1282–1293 .
[105] Deltenre P, Mathurin P, Dharancy S, Moreau R, Bulois P, Henrion J, et al.Transjugular intrahepatic portosystemic shunt in refractory ascites: ameta-analysis. Liver Int 2005;25:349–356
.
[106] Saab S, Nieto JM, Lewis SK, Runyon BA. TIPS vs.paracentesis for
cirrhotic patients with refractory ascites. Cochrane Database Syst Rev
2006:CD004889 .
[107] Salerno F, Camma C, Enea M, Rossle M, Wong F. Transjugularintrahepatic portosystemic shunt for refractory ascites: a meta-analysisof individual patient data. Gastroenterology 2007;133:825–834
.
[108] Maleux G, Perez-Gutierrez NA, Evrard S, Mroue A, Le Moine O, LalemanW, et al. Covered stents are better than uncovered stents for
transjugular intrahepatic portosystemic shunts in cirrhotic patients
with refractory ascites: a retrospective cohort study. Acta GastroenterolBelg 2010;73:336–341
.
[109] Tan HK, James PD, Sniderman KW, Wong F. Long-term clinical outcomeof patients with cirrhosis and refractory ascites treated with tran-sjugular intrahepatic portosystemic shunt insertion. J GastroenterolHepatol 2015;30:389–395
.
[110] Gaba RC, Parvinian A, Casadaban LC, Couture PM, Zivin SP, Lakhoo J,
et al. Survival benefit of TIPS vs.serial paracentesis in patients with
refractory ascites: a single institution case-control propensity scoreanalysis. Clin Radiol 2015;70:e51–e57
.
[111] Bureau C, Thabut D, Oberti F, Dharancy S, Carbonell N, Bouvier A, et al.Transjugular intrahepatic portosystemic shunts with covered stents
increase transplant-free survival of patients with cirrhosis and recur-
rent ascites. Gastroenterology 2017;152:157–163
.
[112] Malinchoc M, Kamath PS, Gordon FD, Peine CJ, Rank J, ter Borg PC. Amodel to predict poor survival in patients undergoing transjugularintrahepatic portosystemic shunts. Hepatology 2000;31:864–871
.
[113] Bureau C, Metivier S, D’Amico M, Peron JM, Otal P, Pagan JC, et al. Serumbilirubin and platelet count: a simple predictive model for survival in
patients with refractory ascites treated by TIPS. J Hepatol
2011;54:901–907
.
[114] Sarwar A, Zhou L, Novack V, Tapper EB, Curry M, Malik R, et al. Hospitalvolume and mortality after trans-jugular intrahepatic portosystemicshunt creation in the United States. Hepatology 2017
.
[115] Angeli P, Volpin R, Piovan D, Bortoluzzi A, Craighero R, Bottaro S, et al.
Acute effects of the oral administration of midodrine, an alpha-
adrenergic agonist, on renal hemodynamics and renal function incirrhotic patients with ascites. Hepatology 1998;28:937–943
.
[116] Singh V, Dhungana SP, Singh B, Vijayverghia R, Nain CK, Sharma N, et al.Midodrine in patients with cirrhosis and refractory or recurrent ascites:a randomized pilot study. J Hepatol 2012;56:348–354
.
[117] Gadano A, Moreau R, Vachiery F, Soupison T, Yang S, Cailmail S, et al.
Natriuretic response to the combination of atrial natriuretic peptide
and terlipressin in patients with cirrhosis and refractory ascites. JHepatol 1997;26:1229–1234
.
[118] Krag A, Moller S, Henriksen JH, Holstein-Rathlou NH, Larsen FS,Bendtsen F. Terlipressin improves renal function in patients withcirrhosis and ascites without hepatorenal syndrome. Hepatology
2007;46:1863–1871
.
[119] Lenaerts A, Codden T, Meunier JC, Henry JP, Ligny G. Effects of clonidineon diuretic response in ascitic patients with cirrhosis and activation ofsympathetic nervous system. Hepatology 2006;44:844–849
.
[120] Singh V, Singh A, Singh B, Vijayvergiya R, Sharma N, Ghai A, et al.Midodrine and clonidine in patients with cirrhosis and refractory orrecurrent ascites: a randomized pilot study. Am J Gastroenterol
2013;108:560–567
.
[121] Rai N, Singh B, Singh A, Vijayvergiya R, Sharma N, Bhalla A, et al.Midodrine and tolvaptan in patients with cirrhosis and refractoryor recurrent ascites: a randomised pilot study. Liver Int 2017;37:406–414
.Clinical Practice Guidelines
44 Journal of Hepatology 2018 vol. xxx jxxx–xxx
Please cite this article in press as: The European Association for the Study of the Liver. EASL Clinical Practice Guidelines for the management of pati ents with decompensated cirrhosis. J
Hepatol (2018), https://doi.org/10.1016/j.jhep.2018.03.024

[122] Tandon P, Tsuyuki RT, Mitchell L, Hoskinson M, Ma MM, Wong WW,
et al. The effect of 1 month of therapy with midodrine, octreotide-LARand albumin in refractory ascites: a pilot study. Liver Int 2009;29:169–174
.
[123] Hanafy AS, Hassaneen AM. Rifaximin and midodrine improve clinicaloutcome in refractory ascites including renal function, weight loss, and
short-term survival. Eur J Gastroenterol Hepatol 2016;28:1455–1461
.
[124] Bellot P, Welker MW, Soriano G, von Schaewen M, Appenrodt B, WiestR, et al. Automated low flow pump system for the treatment ofrefractory ascites: a multi-center safety and efficacy study. J Hepatol2013;58:922–927
.
[125] Stirnimann G, Berg T, Spahr L, Zeuzem S, McPherson S, Lammert F, et al.
Treatment of refractory ascites with an automated low-flow ascites
pump in patients with cirrhosis. Aliment Pharmacol Ther2017;46:981–991
.
[126] Bureau C, Adebayo D, Chalret de Rieu M, Elkrief L, Valla D, et al.Alfapump
/C210system vs.large volume paracentesis for refractory ascites:
A multicenter randomized controlled study. J Hepatol2017;67:940–949
.
[127] Sola E, Sanchez-Cabus S, Rodriguez E, Elia C, Cela R, Moreira R, et al.
Effects of alfapump system on kidney and circulatory function inpatients with cirrhosis and refractory ascites. Liver Transpl2017;23:583–593
.
[128] Badillo R, Rockey DC. Hepatic hydrothorax. Clinical features, manage-ment, and outcomes in 77 patients and review of the literature.
Medicine 2014;93:135–142
.
[129] Garbuzenko DV, Arefyev NO. Hepatic hydrothorax: An update andreview of the literature. World J Hepatol 2017:1197–1204
.
[130] Zenda T, Miyamoto S, Murata S, Mabuchi H. Detection of diaphragmaticdefect as the cause of severe hepatic hydrothorax with magneticresonance imaging. Am J Gastroenterol 1998;9:2288–2289
.
[131] Hewett LJ, Bradshaw ML, Gordon LL, Rockey DC. Diagnosis of isolated
hepatic hydrothorax using peritoneal scintigraphy. Hepatology
2016;64:1364–1366 .
[132] Alonso JC. Pleural effusion in liver disease. Semin Respir Crit Care Med2010;31:698–705
.
[133] Orman ES, Lok AS. Outcomes of patients with chest tube insertion forhepatic hydrothorax. Hepatol Int 2009;3:582–586
.
[134] Xiol X, Tremosa G, Castellote J, Gornals J, Lama C, et al. Liver
transplantation in patients with hepatic hydrothorax. Transpl Int2005;18:672–675
.
[135] Sersté T, Moreno C, Francoz C, Razek WA, Paugham C, et al. The impactof preoperative hepatic hydrothorax on the outcome of adult livertransplantation. Eur J Gastroenterol Hepatol 2010;22:207–212
.
[136] Gordon FD, Anastopoulos HT, Crenshaw W, Gilchrist B, McEniff N,
Falchuk KR, et al. The successful treatment of symptomatic, refractory
hepatic hydrothorax with transjugular intrahepatic portosystemicshunt. Hepatology 1997;25:1366–1369
.
[137] Siegerstetter V, Deibert P, Ochs A, Olschewski M, Blum HE, Rossle M.Treatment of refractory hepatic hydrothorax with transjugular intra-hepatic portosystemic shunt: long-term results in 40 patients. Eur J
Gastroenterol Hepatol 2001;13:529–534
.
[138] Ditah IC, Al Bawardy BF, Saberi B, Ditah C, Kamath PS. Transjugularintrahepatic portosystemic stent shunt for medically refractory hepatichydrothorax: A systematic review and cumulative meta-analysis.World J Hepatol 2015;7:1797–1806
.
[139] Hou F, Qi X, Guo X. Effectiveness and safety of pleurodesis for hepatichydrothorax: A systematic review and meta-analysis. Dig Dis Sci
2016;61:3321–3334
.
[140] Huang PM, Kuo SW, Chen JS, Lee JM. Thoracoscopic mesh repair ofdiaphragmatic defects in hepatic hydrothorax: A 10-year experience.Ann Thorac Surg 2016;101:1921–1927
.
[141] Angeli P, Wong F, Watson H. Gines P, and the participants to CAPPS.Hyponatremia in cirrhosis: results of a survey. Hepatology
2006;44:1535–1542
.
[142] Gines P, Berl T, Bernardi M, Bichet DG, Hamon G, Jimenez W, et al.Hyponatremia in cirrhosis: from pathogenesis to treatment. Hepatol-ogy 1998;28:851–864
.
[143] Spasovski G, Vanholder R, Allolio B, Annane D, Ball S, Bichet D, et al.Clinical practice guideline on diagnosis and treatment of hypona-traemia. Intensive Care Med 2014;40:320–331
.
[144] Biggins SW, Rodriguez HJ, Bacchetti P, Bass NM, Roberts JP, Terrault NA.
Serum sodium predicts mortality in patients listed for liver transplan-tation. Hepatology 2005;41:32–39
.
[145] Porcel A, Diaz F, Rendon P, Macias M, Martin-Herrera L, Giron-GonzalezJA. Dilutional hyponatremia in patients with cirrhosis and ascites. ArchIntern Med 2002;162:323–328 ;
Cordoba J, Garcia-Martinez R, Simon-Talero M. Hyponatremic andhepatic encephalopathies: similarities, differences and coexistence.Metab Brain Dis 2010;25:73–80
.
[146] Cordoba J, Garcia-Martinez R, Simon-Talero M. Hyponatremic andhepatic encephalopathies: similarities, differences and coexistence.
Metab Brain Dis 2010;25:73–80
.
[147] Amodio P, Del Piccolo F, Petteno E, Mapelli D, Angeli P, Iemmolo R, et al.Prevalence and prognostic value of quantified electroencephalogram(EEG) alterations in cirrhotic patients. J Hepatol 2001;35:37–45
.
[148] Londono MC, Guevara M, Rimola A, Navasa M, Taura P, Mas A, et al.Hyponatremia impairs early posttransplantation outcome in patients
with cirrhosis undergoing liver transplantation. Gastroenterology
2006;130:1135–1143
.
[149] Biggins SW, Kim WR, Terrault NA, Saab S, Balan V, Schiano T, et al.Evidence-based incorporation of serum sodium concentration intoMELD. Gastroenterology 2006;130:1652–1660
.
[150] Kim WR, Biggins SW, Kremers WK, Wiesner RH, Kamath PS, Benson JT,et al. Hyponatremia and mortality among patients on the liver-
transplant waiting list. N Engl J Med 2008;359:1018–1026
.
[151] McCormick PA, Mistry P, Kaye G, Burroughs AK, McIntyre N. Intra-venous albumin infusion is an effective therapy for hyponatraemia incirrhotic patients with ascites. Gut 1990;31:204–207
.
[152] Quittnat F, Gross P. Vaptans and the treatment of water-retainingdisorders. Semin Nephrol 2006;26:234–243
.
[153] Cardenas A, Gines P, Marotta P, Czerwiec F, Oyuang J, Guevara M, et al.
Tolvaptan, an oral vasopressin antagonist, in the treatment of hypona-tremia in cirrhosis. J Hepatol 2012;56:571–578
.
[154] Gerbes AL, Gulberg V, Gines P, Decaux G, Gross P, Gandjini H, et al.Therapy of hyponatremia in cirrhosis with a vasopressin receptorantagonist: a randomized double-blind multicenter trial. Gastroen-terology 2003;124:933–939
.
[155] Gines P, Wong F, Watson H, Milutinovic S, del Arbol LR, Olteanu D.
Effects of satavaptan, a selective vasopressin V(2) receptor antagonist,on ascites and serum sodium in cirrhosis with hyponatremia: arandomized trial. Hepatology 2008;48:204–213
.
[156] O’Leary JG, Davis GL. Conivaptan increases serum sodium in hypona-tremic patients with end-stage liver disease. Liver Transpl
2009;15:1325–1329
.
[157] Wong F, Gines P, Watson H, Horsmans Y, Angeli P, Gow P, et al. Effectsof a selective vasopressin V2 receptor antagonist, satavaptan, on ascitesrecurrence after paracentesis in patients with cirrhosis. J Hepatol2010;53:283–290
.
[158] Wong F, Watson H, Gerbes A, Vilstrup H, Badalamenti S, Bernardi M,et al. Satavaptan for the management of ascites in cirrhosis: efficacy
and safety across the spectrum of ascites severity. Gut
2012;61:108–116
.
[159] Pose E, Sola E, Piano S, Gola E, Graupera I, Guevara M, et al. Limitedefficacy of tolvaptan in patients with cirrhosis and severe hypona-tremia: real-life experience. Am J Med 2017;130:372–375
.
[160] Torres VE, Chapman AB, Devuyst O, Gansevoort RT, Grantham JJ,
Higashihara E, et al. Tolvaptan in patients with autosomal dominant
polycystic kidney disease. N Engl J Med 2012;367:2407–2418 .
[161] Bosch J, Groszmann RJ, Shah VH. Evolution in the understanding of thepathophysiological basis of portal hypertension: How changes inparadigm are leading to successful new treatments. J Hepatol2015;62:S121–S130
.
[162] D’Amico G, De Franchis R. Upper digestive bleeding in cirrhosis. Post-
therapeutic outcome and prognostic indicators. Hepatology
2003;38:599–612 .
[163] Jepsen P, Ott P, Andersen PK, Sorensen HT, Vilstrup H. Clinical course ofalcoholic liver cirrhosis: a Danish population-based cohort study.Hepatology 2010;51:1675–1682
.
[164] Kovalak M, Lake J, Mattek N, Eisen G, Lieberman D, Zaman A.
Endoscopic screening for varices in cirrhotic patients: data from a
national endoscopic database. Gastrointest Endosc 2007;65:82–88 .
[165] Merli M, Nicolini G, Angeloni S, Rinaldi V, De Santis A, Merkel C, et al.Incidence and natural history of small esophageal varices in cirrhoticpatients. J Hepatol 2003;38:266–272
.
[166] Groszmann RJ, Garcia-Tsao G, Bosch J, Grace ND, Burroughs AK, PlanasR, et al. Beta-blockers to prevent gastroesophageal varices in patients
with cirrhosis. N Engl J Med 2005;353:2254–2261
.
[167] Abraldes JG, Bureau C, Stefanescu H, Augustin S, Ney M, Blasco H, et al.Non invasive tools and risk of clinically significant portal hypertensionand varices in compensated cirrhosis : the ‘‘ANTICIPATE” study.Hepatology 2016;64:2173–2184
.JOURNAL
OF HEPATOLOGY
Journal of Hepatology 2018 vol. xxx jxxx–xxx 45
Please cite this article in press as: The European Association for the Study of the Liver. EASL Clinical Practice Guidelines for the management of pati ents with decompensated cirrhosis. J
Hepatol (2018), https://doi.org/10.1016/j.jhep.2018.03.024

[168] De Franchis RBaveno VI faculty. Expanding consensus in portal
hypertension: report of the BAVENO VI Consensus Workshop: Strati-fying risk and individualizing care for portal hypertension. J Hepatol2015;63:743–752
.
[169] Garcia-Tsao G, Sanyal AJ, Grace ND, Carey WPractice GuidelinesCommittee of the American Association for the Study of Liver
DiseasesPractice Parameters Committee of the American College of
Gastroenterology. Prevention and management of gastroesophagealvarices and variceal hemorrhage in cirrhosis. Hepatology2007;46:922–938
.
[170] North Italian Endoscopic Club for the Study and Treatment ofEsophageal Varices. Prediction of the first variceal hemorrhage in
patients with cirrhosis of the liver and esophageal varices. A prospec-
tive multicenter study. N Engl J Med 1988;319:983–989
.
[171] D’Amico G, Pasta L, Morabito A, et al. Competing risks and prognosticstages of cirrhosis: a 25-year inception cohort study of 494 patients.Aliment Pharmacol Ther 2014;39:1180–1193
.
[172] Augustin S, Muntaner L, Altamirano JT, Gonzalez A, Saperas E, Dot J,et al. Predicting early mortality after acute variceal hemorrhage based
on classification and regression tree analysis. Clin Gastroenterol
Hepatol 2009;7:1347–1354
.
[173] Bosch J, Garcia-Pagan JC. Prevention of variceal rebleeding. Lancet2003;361:952–954
.
[174] Garcia-Pagan JC, De Gottardi A, Bosch J. Review article: the modernmanagement of portal hypertension–primary and secondary prophy-
laxis of variceal bleeding in cirrhotic patients. Aliment Pharmacol Ther
2008;28:178–186
.
[175] Hernandez-Gea V, Aracil C, Colomo A, Garupera I, Poca M, Torras X,et al. Development of ascites in compensated cirrhosis with severeportal hypertension treated with beta-blockers. Am J Gastroenterol2012;107:418–427
.
[176] Puente A, Hernandez-Gea V, Graupera I, Roque M, Colomo A, Poca M,
et al. Drugs plus ligation to prevent rebleeding in cirrhosis: an updated
systematic review. Liver Int 2014;34:823–833 .
[177] Villanueva C, Graupera I, Aracil C, Alvarado E, Minana J, Puente A, et al.A randomized trial to assess whether portal pressure guided therapy toprevent variceal rebleeding improves survival in cirrhosis. Hepatology2017;65:1693–1707
.
[178] D’Amico G, Pagliaro L, Bosch J. Pharmacological treatment of portal
hypertension: an evidence-based approach. Semin Liver Dis1999;19:475–505
.
[179] Albillos A, Banares R, Gonzalez M, Ripoll C, Gonzalez R, Catalina MV,et al. Value of the hepatic venous pressure gradient to monitor drugtherapy for portal hypertension: a meta-analysis. Am J Gastroenterol2007;102:1116–1126
.
[180] Li T, Ke W, Sun P, Chen X, Belgaumkar A, Huang Y, et al. Carvedilol for
portal hypertension in cirrhosis: systematic review with meta-analysis.BMJ Open 2016;6:e010902
.
[181] Serste T, Melot C, Francoz C, Durand F, Rautou PE, Valla D, et al.Deleterious effects of beta-blockers on survival in patients withcirrhosis and refractory ascites. Hepatology 2010;52:1017–1022
.
[182] Wiest R, Albillos A, Gluud LL. The window hypothesis: haemodynamic
and non-haemodynamic effects of beta-blockers improve survival ofpatients with cirrhosis during a window in the disease. Gut2012;61:967–969
.
[183] Bossen L, Krag A, Vilstrup H, Watson H, Jepsen P. Non-selective beta-blockers do not affect mortality in cirrhosis patients with ascites: Posthoc analysis of three RCTs with 1198 patients. Hepatology
2016;63:1968–1976
.
[184] Bang UC, Benfield T, Hyldstrup L, Jensen JE, Bendtsen F. Effect ofpropranolol on survival in patients with decompensated cirrhosis: anationwide study based Danish patient registers. Liver Int2016;36:1304–1312
.
[185] Reiberger T, Mandorfer M. Beta adrenergic blockade and decompen-
sated cirrhosis. J Hepatol 2017;66:849–859 .
[186] Leithead JA, Rajoriya N, Tehami N, Hodson J, Gunson BK, Tripathi D,et al. Non-selective b-blockers are associated with improved survival in
patients with ascites listed for liver transplantation. Gut 2015;64:1111–1119
.
[187] Senzolo M, Cholongitas E, Burra P, Leandro G, Thalheimer U, Patch D,et al. Beta-blockers protect against spontaneous bacterial peritonitis in
cirrhotic patients: a meta-analysis. Liver Int 2009;29:1189–1193
.
[188] Reiberger T, Ferlitsch A, Payer BA, Mandorfer M, Heinisch BB, Hayden H,et al. Non-selective beta-blocker therapy decreases intestinal perme-ability and serum levels of LBP and IL-6 in patients with cirrhosis. JHepatol 2013;58:911–921
.[189] Mookerjee RP, Pavesi M, Thomsen KL, Mehta G, Macnaughtan J,
Bendtsen F, et al. Treatment with non-selective beta blockers isassociated with reduced severity of systemic inflammation andimproved survival of patients with acute-on-chronic liver failure. JHepatol 2016;64:574–582
.
[190] Madsen BS, Nielsen KF, Fialla AD, Krag A. Keep the sick from harm in
spontaneous bacterial peritonitis: Dose of beta blockers matters. J
Hepatol 2016;64:1455–1456 .
[191] Serste T, Gustot T, Rautou PE, Francoz C, Njimi H, Durand F, et al. Severehyponatremia is a better predictor of mortality than MELDNa inpatients with cirrhosis and refractory ascites. J Hepatol 2012;57:274–280
.
[192] Krag A, Bendtsen F, Henriksen JH, Moller S. Low cardiac output predicts
development of hepatorenal syndrome and survival in patients withcirrhosis and ascites. Gut 2010;59:105–110
.
[193] Ruiz-del-Arbol L, Urman J, Fernandez J, Gonzalez M, Navasa M,Monescillo A, et al. Systemic, renal, and hepatic hemodynamicderangement in cirrhotic patients with spontaneous bacterial peritoni-tis. Hepatology 2003;38:1210–1218
.
[194] Payance A, Bissonnette J, Roux O, Elkrief L, Gault N, Francoz C, et al. Lack
of clinical or haemodynamic rebound after abrupt interruption of beta-blockers in patients with cirrhosis. Aliment Pharmacol Ther2016;43:966–973
.
[195] Garcia-Tsao G, Bosch J. Management of varices and variceal hemor-rhage in cirrhosis. N Engl J Med 2010;362:823–832
.
[196] Villanueva C, Escorsell A. Optimizing general management of acute
variceal bleeding in cirrhosis. Curr Hepatol Rep 2014;13:198–207 .
[197] Avgerinos A, Nevens F, Raptis S, Fevery J. Early administration ofsomatostatin and efficacy of sclerotherapy in acute oesophagealvariceal bleeds: the European Acute Bleeding OesophagealVariceal Episodes (ABOVE) randomised trial. Lancet 1997;350:1495–1499
.
[198] Levacher S, Letoumelin P, Pateron D, Blaise M, Lapandry C, Pourriat JL.
Early administration of terlipressin plus glyceryl trinitrate to controlactive upper gastrointestinal bleeding in cirrhotic patients. Lancet1995;346:865–868
.
[199] Garcia-Tsao G, Abraldes JG, Berzigotti A, Bosch J. Portal hypertensivebleeding in cirrhosis: Risk stratification, diagnosis, and management:
2016 practice guidance by the American Association for the study of
liver diseases. Hepatology 2017;65:310–335
.
[200] Myburgh JA. Fluid resuscitation in acute illness–time to reappraise thebasics. N Engl J Med 2011;364:2543–2544
.
[201] Villanueva C, Colomo A, Bosch A, Concepcion M, Hernandez-Gea V,Aracil C, et al. Transfusion strategies for acute upper gastrointestinalbleeding. N Engl J Med 2013;368:11–21
.
[202] Seo YS, Park SY, Kim MY, Kim JH, Park JY, Yim HJ, et al. Lack of difference
among terlipressin, somatostatin, and octreotide in the control ofacute gastroesophageal variceal hemorrhage. Hepatology 2014;60:954–963
.
[203] Azam Z, Hamid S, Jafri W, Salih M, Abbas Z, Abid S, et al. Short courseadjuvant terlipressin in acute variceal bleeding: a randomized double
blind dummy controlled trial. J Hepatol 2012;56:819–824
.
[204] Altraif I, Handoo FA, Aljumah A, Alalwan A, Dafalla M, Saeed AM, et al.Effect of erythromycin before endoscopy in patients presenting withvariceal bleeding: a prospective, randomized, double-blind, placebo-controlled trial. Gastrointest Endosc 2011;73:245–250
.
[205] Banares R, Albillos A, Rincon D, Alonso S, Gonzalez M, Ruiz-del-Arbol L,et al. Endoscopic treatment vs.endoscopic plus pharmacologic treat-
ment for acute variceal bleeding: a meta-analysis. Hepatology
2002;35:609–615
.
[206] Villanueva C, Piqueras M, Aracil C, Gomez C, Lopez-Balaguer JM,Gonzalez B, et al. A randomized controlled trial comparing ligation andsclerotherapy as emergency endoscopic treatment added to somato-statin in acute variceal bleeding. J Hepatol 2006;45:560–567
.
[207] Rios Castellanos E, Seron P, Gisbert JP, Bonfill Cosp X. Endoscopic
injection of cyanoacrylate glue vs.other endoscopic procedures for
acute bleeding gastric varices in people with portal hypertension.Cochrane Database Syst Rev 2015;5:CD010180
.
[208] Bernard B, Grange JD, Khac EN, Amiot X, Opolon P, Poynard T. Antibioticprophylaxis for the prevention of bacterial infections in cirrhoticpatients with gastrointestinal bleeding: a meta-analysis. Hepatology
1999;29:1655–1661
.
[209] Fernandez J, Ruiz del Arbol L, Gomez C, Durandez R, Serradilla R,Guarner C, et al. Norfloxacin vs ceftriaxone in the prophylaxis ofinfections in patients with advanced cirrhosis and hemorrhage.Gastroenterology 2006;131:1049–1056
.Clinical Practice Guidelines
46 Journal of Hepatology 2018 vol. xxx jxxx–xxx
Please cite this article in press as: The European Association for the Study of the Liver. EASL Clinical Practice Guidelines for the management of pati ents with decompensated cirrhosis. J
Hepatol (2018), https://doi.org/10.1016/j.jhep.2018.03.024

[210] Tandon P, Abraldes JG, Keough A, Bastiampillai R, Jayakumar S,
Carbonneau M, et al. Risk of bacterial infection in patients withcirrhosis and acute variceal hemorrhage, based on child-pugh class, andeffects of antibiotics. Clin Gastroenterol Hepatol 2015;13:1189–1196
.
[211] Cardenas A, Gines P, Uriz J, Bessa X, Salmeron JM, Mas A, et al. Renalfailure after upper gastrointestinal bleeding in cirrhosis: incidence,
clinical course, predictive factors, and short-term prognosis. Hepatol-
ogy 2001;34:671–676
.
[212] Shaheen NJ, Stuart E, Schmitz SM, Mitchell KL, Fried MW, Zacks S, et al.Pantoprazole reduces the size of postbanding ulcers after variceal bandligation: a randomized, controlled trial. Hepatology 2005;41:588–594
.
[213] Escorsell A, Pavel O, Cardenas A, Morillas R, Llop E, Villanueva C, et al.
Esophageal balloon tamponade vs.esophageal stent in controlling acute
refractory variceal bleeding: A multicenter randomized, controlledtrial. Hepatology 2016;63:1957–1996
.
[214] Monescillo A, Martínez-Lagares F, Ruiz-del-Arbol L, et al. Influence ofportal hypertension and its early decompression by TIPS placement onthe outcome of variceal bleeding. Hepatology 2004;40:793–801
.
[215] García-Pagán JC, Caca K, Bureau C, et al. Early use of TIPS in patients
with cirrhosis and variceal bleeding. N Engl J Med 2010;362:
2370–2379 .
[216] Garcia-Pagan JC, Di Pascoli M, Caca K, Laleman W, Bureau C, AppenrodtB, et al. Use of early-TIPS for high-risk variceal bleeding: results of apost-RCT surveillance study. J Hepatol 2013;58:45–50
.
[217] Rudler M, Cluzel P, Corvec TL, Benosman H, Rousseau G, Poynard T,
et al. Early-TIPSS placement prevents rebleeding inhigh-risk patients
with variceal bleeding, without improving survival. Aliment PharmacolTher 2014;40:1074–1080
.
[218] Augustin S, Altamirano J, Gonzalez A, Dot J, Abu-Suboh M, Armengol JR,et al. Effectiveness of combined pharmacologic and ligation therapy inhigh-risk patients with acute esophageal variceal bleeding. Am JGastroenterol 2011;106:1787–1795
.
[219] Reverter E, Tandon P, Augustin S, Turon F, Casu S, Bastiampillai R, et al.
A MELD-based model to determine risk of mortality among patientswith acute variceal bleeding. Gastroenterology 2014;146:412–419
.
[220] Merli M, Nicolini G, Angeloni S, Gentili F, Attili AF, Riggio O. The naturalhistory of portal hypertensive gastropathy in patients with livercirrhosis and mild portal hypertension. Am J Gastroenterol
2004;99:1959–1965
.
[221] Thuluvath PJ, Yoo HY. Portal Hypertensive gastropathy. Am J Gastroen-terol 2002;97:2973–2978
.
[222] Yoshikawa I, Murata I, Nakano S, Otsuki M. Effects of endoscopicvariceal ligation on portal hypertensive gastropathy and gastricmucosal blood flow. Am J Gastroenterol 1998;93:71–74
.
[223] Urrunaga NH, Rockey DC. Portal hypertensive gastropathy and colopa-
thy. Clin Liver Dis 2014;18:389–406 .
[224] Hosking SW, Kennedy HJ, Seddon I, Triger DR. The role of propranolol incongestive gastropathy of portal hypertension. Hepatology1987;7:437–441
.
[225] Perez-Ayuso RM, Pique JM, Bosch J, Panes J, Gonzalez A, Perez R, et al.Propranolol in prevention of recurrent bleeding from severe portal
hypertensive gastropathy in cirrhosis. Lancet 1991;337:1431–1434
.
[226] Ripoll C, Garcia-Tsao G. Treatment of gastropathy and gastric antralvascular ectasia in patients with portal hypertension. Curr TreatOptions Gastroenterol 2007;10:483–494
.
[227] Kamath P, Lacerda M, Ahlquist D, McKusick MA, Andrews JC, NagorneyDA. Gastric Mucosal Responses to intrahepatic portosystemic shuntingin patients with cirrhosis. Gastroenterology 2000;118:905–911
.
[228] Zhou Y, Qiao L, Wu J, Hu H, Xu C. Comparison of the efficacy of
octreotide, vasopressin, and omeprazole in the control of acutebleeding in patients with portal hypertensive gastropathy: a controlledstudy. J Gastroenterol Hepatol 2002;17:973–979
.
[229] Sarin SK, Lahoti D, Saxena SP, Murthy NS, Makwana UK. Prevalence,classification and natural history of gastric varices: a long-term follow-
up study in 568 portal hypertension patients. Hepatology
1992;16:1343–1349
.
[230] Mishra SR, Sharma BC, Kumar A, Sarin SK. Primary prophylaxis ofgastric variceal bleeding comparing cyanoacrylate injection and beta-blockers: a randomized controlled trial. J Hepatol 2011;54:1161–1167
.
[231] Rios CE, Seron P, Gisbert JP, Bonfill CX. Endoscopic injection ofcyanoacrylate glue vs.other endoscopic procedures for acute bleeding
gastric varices in people with portal hypertension. Cochrane Database
Syst Rev 2015;5:CD010180
.
[232] Mishra SR, Chander Sharma B, Kumar A, Sarin SK. Endoscopiccyanoacrylate injection vs.beta-blocker for secondary prophylaxis ofgastric variceal bleed: a randomised controlled trial. Gut
2010;59:729–735 .
[233] Chau TN, Patch D, Chan YW, Nagral A, Dick R, Burroughs AK. ‘‘Salvage”transjugular intrahepatic portosystemic shunts: gastric fundal com-pared with esophageal variceal bleeding. Gastroenterology1998;114:981–998
.
[234] Hung HH, Chang CJ, Hou MC, Liao WC, Chan CC, Huang HC, et al.
Efficacy of non-selective beta-blockers as adjunct to endoscopicprophylactic treatment for gastric variceal bleeding: a randomizedcontrolled trial. J Hepatol 2012;56:1025–1032
.
[235] Lo GH, Liang HL, Chen WC, Chen MH, Lai KH, Hsu PI, et al. A prospective,randomized controlled trial of transjugular intrahepatic portosystemic
shunt vs.cyanoacrylate injection in the prevention of gastric variceal
rebleeding. Endoscopy 2007;39:679–685
.
[236] Saad WE. Endovascular management of gastric varices. Clin Liver Dis2014;18:829–885
.
[237] Mookerjee RP, Stadlbauer V, Lidder S, Wright GA, Hodges SJ, Davies NA,et al. Neutrophil dysfunction in alcoholic hepatitis superimposed oncirrhosis is reversible and predicts the outcome. Hepatology
2007;46:831–840
.
[238] Wasmuth HE, Kunz D, Yagmur E, Timmer-Stranghoner A, Vidacek D,Siewert E, et al. Patients with acute on chronic liver failure display‘‘sepsis-like” immune paralysis. J Hepatol 2005;42:195–201
.
[239] Wiest R, Garcia-Tsao G. Bacterial translocation (BT) in cirrhosis.Hepatology 2005;41:422–433
.
[240] Appenrodt B, Grunhage F, Gentemann MG, Thyssen L, Sauerbruch T,
Lammert F. Nucleotide-binding oligomerization domain containing 2(NOD2) variants are genetic risk factors for death and spontaneousbacterial peritonitis in liver cirrhosis. Hepatology 2010;51:1327–1333
.
[241] Fernandez J, Navasa M, Gomez J, et al. Bacterial infections in cirrhosis:epidemiological changes with invasive procedures and norfloxacin
prophylaxis. Hepatology 2002;35:140–148
.
[242] Fasolato S, Angeli P, Dallagnese L, et al. Renal failure and bacterialinfections in patients with cirrhosis: epidemiology and clinical features.Hepatology 2007;45:223–229
.
[243] Piano S, Fasolato S, Salinas F, Romano A, Tonon M, Morando F, et al. Theempirical antibiotic treatment of nosocomial spontaneous bacterial
peritonitis: Results of a randomized, controlled clinical trial. Hepatol-
ogy 2016;63:1299–1309
.
[244] Runyon BA. The evolution of ascitic fluid analysis in the diagnosis ofspontaneous bacterial peritonitis. Runyon BA. Am J Gastroenterol2003;98:1675–1677
.
[245] Evans LT, Kim WR, Poterucha JJ, Kamath PS. Spontaneous bacterialperitonitis in asymptomatic outpatients with cirrhotic ascites. Hepa-
tology 2003;37:897–901
.
[246] Kim JJ, Tsukamoto MM, Mathur AK, Ghomri YM, Hou LA, Sheibani S,et al. Delayed paracentesis is associated with increased in-hospitalmortality in patients with spontaneous bacterial peritonitis. Am JGastroenterol 2014;109:1436–1442
.
[247] Van de Geijn GM, van Gent M, van Pul-Bom N, Beunis MH, van Tilburg
AJ, Njo TL. A new flow cytometric method for differential cell counting
in ascitic fluid. Cytometry B Clin Cytom 2016;90:506–511 .
[248] Fleming C, Brouwer R, van Alphen A, Lindemans J, de Jonge R. UF-1000i:validation of the body fluid mode for counting cells in body fluids. ClinChem Lab Med 2014;52:1781–1790
.
[249] Gülberg V, Gerbes AL, Sauerbruch T, Appenrodt B. Insufficient sensi-tivity of reagent strips for spontaneous bacterial peritonitis. Hepatology
2007;46:1669
.
[250] Bellot P, García-Pagán JC, Francés R, Abraldes JG, Navasa M, Pérez-Mateo M, et al. Bacterial DNA translocation is associated with systemiccirculatory abnormalities and intrahepatic endothelial dysfunction inpatients with cirrhosis. Hepatology 2010;52:2044–2052
.
[251] Bruns T, Reuken PA, Stengel S, Gerber L, Appenrodt B, Schade JH, et al.
The prognostic significance of bacterial DNA in patients with decom-
pensated cirrhosis and suspected infection. Liver Int 2016;36:1133–1142
.
[252] Runyon BA, Hoefs JC. Culture-negative neutrocytic ascites: a variant ofspontaneous bacterial peritonitis. Hepatology 1984;4:1209–1211
.
[253] Gravito-Soares M, Gravito-Soares E, Lopes S, Ribeiro G, Figueiredo P.Spontaneous fungal peritonitis: a rare but severe complication of liver
cirrhosis. Eur J Gastroenterol Hepatol 2017;29:1010–1016
.
[254] Chen CH, Shih CM, Chou JW, Liu YH, Hang LW, Hsia TC, et al. Outcomepredictors of cirrhotic patients with spontaneous bacterial empyema.Liver Int 2011;31:417–424
.JOURNAL
OF HEPATOLOGY
Journal of Hepatology 2018 vol. xxx jxxx–xxx 47
Please cite this article in press as: The European Association for the Study of the Liver. EASL Clinical Practice Guidelines for the management of pati ents with decompensated cirrhosis. J
Hepatol (2018), https://doi.org/10.1016/j.jhep.2018.03.024

[255] Xiol X, Castellvi JM, Guardiola J, Sese E, Castellote J, Perello A, et al.
Spontaneous bacterial empyema in cirrhotic patients: a prospectivestudy. Hepatology 1996;23:719–723
.
[256] Soriano G, Castellote J, Alvarez C, Girbau A, Gordillo J, Baliellas C, et al.Secondary bacterial peritonitis in cirrhosis: a retrospective study ofclinical and analytical characteristics, diagnosis and management. J
Hepatol 2010;52:39–44
.
[257] Rimola A, Salmeron JM, Clemente G, Rodrigo L, Obrador A, Miranda ML,et al. Two different dosages of cefotaxime in the treatment ofspontaneous bacterial peritonitis in cirrhosis: results of a prospective,randomized, multicenter study. Hepatology 1995;21:674–679
.
[258] Runyon Runyon BA, McHutchison JG, Antillon MR, Akriviadis EA,
Montano AA. Short-course vs.long-course antibiotic treatment of
spontaneous bacterial peritonitis. A randomized controlled study of100 patients. Gastroenterology 1991;100:1737–1742
.
[259] Ricart E, Soriano G, Novella MT, Ortiz J, Sabat M, Kolle L, et al.Amoxicillin-clavulanic acid vs.cefotaxime in the therapy of bacterial
infections in cirrhotic patients. J Hepatol 2000;32:596–602
.
[260] deLemos AS, Ghabril M, Rockey DC, Gu J, Barnhart HX, Fontana RJ, et al.
Drug-induced liver injury network (DILIN). Amoxicillin-clavulanate-
induced liver injury. Dig Dis Sci 2016;61:2406–2416 .
[261] Terg R, Cobas S, Fassio E, Landeira G, Rios B, Vasen W, et al. Oralciprofloxacin after a short course of intravenous ciprofloxacin in thetreatment of spontaneous bacterial peritonitis: results of a multicenter,randomized study. J Hepatol 2000;33:564–569
.
[262] Angeli P, Guarda S, Fasolato S, Miola E, Craighero R, Piccolo F, et al.
Switch therapy with ciprofloxacin vs.intravenous ceftazidime in the
treatment of spontaneous bacterial peritonitis in patients with cirrho-sis: similar efficacy at lower cost. Aliment Pharmacol Ther 2006;23:75–84
.
[263] Navasa M, Follo A, Llovet JM, Clemente G, Vargas V, Rimola A, et al.Randomized, comparative study of oral ofloxacin vs.intravenous
cefotaxime in spontaneous bacterial peritonitis. Gastroenterology
1996;111:1011–1017
.
[264] Fernandez J, Bert F, Nicolas-Chanoine MH. The challenges of multi-drug-resistance in hepatology. J Hepatol 2016;65:1043–1054
.
[265] Magiorakos AP, Srinivasan A, Carey RB, Carmeli Y, Falagas ME, Giske CG,et al. Multidrug-resistant, extensively drug-resistant and pandrug-resistant bacteria: an international expert proposal for interim stan-
dard definitions for acquired resistance. Clin Microbiol Infect
2012;18:268–281
.
[266] Fernandez J, Acevedo J, Castro M, Garcia O, de Lope CR, Roca D, et al.Prevalence and risk factors of infections by multiresistant bacteria incirrhosis: a prospective study. Hepatology 2012;55:1551–1561
.
[267] Wiest R, Krag A, Gerbes A. Spontaneous bacterial peritonitis: recentguidelines and beyond. Gut 2012;61:297–310
.
[268] Piano S, Brocca A, Mareso S, Angeli P. Infections complicating cirrhosis.
Liver Int 2018;38:126–133 .
[269] Lutz P, Nischalke HD, Kramer B, Goeser F, Kaczmarek DJ, Schlabe S, et al.Antibiotic resistance in healthcare-related and nosocomial sponta-neous bacterial peritonitis. Eur J Clin Invest 2017;47:44–52
.
[270] Singer M, Deutschman CS, Seymour CW, Shankar-Hari M, Annane D,
Bauer M, et al. The third international consensus definitions for sepsis
and septic shock (Sepsis-3). JAMA 2016;315:801–810 .
[271] Piano S, Bartoletti M, Tonon M, Baldassarre M, Chies G, Romano A, et al.Assessment of Sepsis-3 criteria and quick SOFA in patients withcirrhosis and bacterial infections. Gut 2017
.
[272] Sort P, Navasa M, Arroyo V, Aldeguer X, Planas R, Ruiz-del-Arbol L, et al.Effect of intravenous albumin on renal impairment and mortality in
patients with cirrhosis and spontaneous bacterial peritonitis. N EnglJ
Med 1999;341:403–409
.
[273] Garioud A, Cadranel JF, Pauwels A, Nousbaum JB, Thévenot T, Dao T,et al. Association Nationale des Hépato-gastroentérologues des Hôpi-taux Généraux de France, Association Française pour l’Etude du Foie,Club de Réflexion des Cabinets et Groupes d’Hépato-Gastroentérologie.
Albumin Use in Patients With Cirrhosis in France: Results of the ‘‘ALBU-
LIVE” Survey: A Case for Better EASL Guidelines Diffusion and/orRevision. J Clin Gastroenterol 2017;51:831–838
.
[274] Poca M, Concepcion M, Casas M, Alvarez-Urturi C, Gordillo J, Hernan-dez-Gea V, et al. Role of albumin treatment in patients with sponta-neous bacterial peritonitis. Clin Gastroenterol Hepatol 2012;10:309–315
.
[275] Fernandez J, Tandon P, Mensa J, Garcia-Tsao G. Antibiotic prophylaxis in
cirrhosis: Good and bad. Hepatology 2016;63:2019–2031 .
[276] Fernandez J, Navasa M, Planas R, Montoliu S, Monfort D, Soriano G, et al.Primary prophylaxis of spontaneous bacterial peritonitis delays hepa-torenal syndrome and improves survival in cirrhosis. Gastroenterology2007;133:818–824
.
[277] Terg R, Fassio E, Guevara M, Cartier M, Longo C, Lucero R, et al.Ciprofloxacin in primary prophylaxis of spontaneous bacterial peri-tonitis: A randomized, placebo-controlled study. J Hepatol2008;48:774–779
.
[278] Loomba R, Wesley R, Bain A, et al. Role of fluoroquinolones in the
primary prophylaxis of spontaneous bacterial peritonitis: meta-analy-sis. Clin Gastroenterol Hepatol 2009;7:487–493
.
[279] Saab S, Hernandez JC, Chi AC, et al. Oral antibiotic prophylaxis reducesspontaneous bacterial peritonitis occurrence and improves short-termsurvival in cirrhosis: a meta-analysis. Am J Gastroenterol 2009;104:
993–1001
.
[280] Krag A, Wiest R, Gluud LL. Fluoroquinolones in the primary prophylaxisof spontaneous bacterial peritonitis. Am J Gastroenterol2010;105:1444–1445
.
[281] Gines P, Rimola A, Planas R, Vargas V, Marco F, Almela M, et al.Norfloxacin prevents spontaneous bacterial peritonitis recurrence incirrhosis: results of a double-blind, placebo-controlled trial. Hepatology
1990;12:716–724
.
[282] Bauer TM, Follo A, Navasa M, Vila J, Planas R, Clemente G, et al. Dailynorfloxacin is more effective than weekly rufloxacin in prevention ofspontaneous bacterial peritonitis recurrence. Dig Dis Sci2002;47:1356–1361
.
[283] Terg R, Llano K, Cobas SM, Brotto C, Barrios A, Levi D, et al. Effects of oral
ciprofloxacin on aerobic gram-negative fecal flora in patients with
cirrhosis: results of short- and long-term administration, with daily andweekly dosages. J Hepatol 1998;29:437–442
.
[284] Bass NM, Mullen KD, Sanyal A, Poordad F, Neff G, Leevy CB, et al.Rifaximin treatment in hepatic encephalopathy. N Engl J Med2010;362:071–1081
.
[285] Elfert A, Abo Ali L, Soliman S, Ibrahim S, Abd-Elsalam S. Randomized-
controlled trial of rifaximin vs.norfloxacin for secondary prophylaxis of
spontaneous bacterial peritonitis. Eur J Gastroenterol Hepatol2016;28:1450–1454
.
[286] Min YW, Lim KS, Min BH, Gwak GY, Paik YH, Choi MS, et al. Protonpump inhibitor use significantly increases the risk of spontaneousbacterial peritonitis in 1965 patients with cirrhosis and ascites: a
propensity score matched cohort study. Aliment Pharmacol Ther
2014;40:695–704
.
[287] Dam G, Vilstrup H, Watson H, Jepsen P. Proton pump inhibitors as a riskfactor for hepatic encephalopathy and spontaneous bacterial peritonitisin patients with cirrhosis with ascites. Hepatology 2016;64:1265–1272
.
[288] Mandorfer M, Bota S, Schwabl P, Bucsics T, Pfisterer N, Kruzik M, et al.
Nonselective beta blockers increase risk for hepatorenal syndrome and
death in patients with cirrhosis and spontaneous bacterial peritonitis.Gastroenterology 2014;146:1680–1690
.
[289] Pande C, Kumar A, Sarin SK. Addition of probiotics to norfloxacin doesnot improve efficacy in the prevention of spontaneous bacterialperitonitis: a double-blind placebo-controlled randomized-controlled
trial. Eur J Gastroenterol Hepatol 2012;24:831–839
.
[290] Fernandez J, Acevedo J, Arroyo V. Response to the clinical course andshort-term mortality of cirrhotic patients with non-spontaneous bac-terial peritonitis infections. Liver Int 2017;37:623
.
[291] Arvaniti V, D’Amico G, Fede G, Manousou P, Tsochatzis E, PleguezueloM, et al. Infections in patients with cirrhosis increase mortality four-fold and should be used in determining prognosis. Gastroenterology
2010;139:1256
.
[292] Piano S, Morando F, Carretta G, Tonon M, Vettore E, Rosi S, et al.Predictors of early readmission in patients with cirrhosis after theresolution of bacterial infections. Am J Gastroenterol 2017;112:1575–1583
.
[293] Bartoletti M, Giannella M, Lewis R, Caraceni P, Tedeschi S, Paul M, et al.
A prospective multicentre study of the epidemiology and outcomes of
bloodstream infection in cirrhotic patients. Clin Microbiol Infect 2017 .
[294] Merli M, Lucidi C, Di Gregorio V, Lattanzi B, Giannelli V, Giusto M, et al.An empirical broad spectrum antibiotic therapy in health-care-associ-ated infections improves survival in patients with cirrhosis: A ran-domized trial. Hepatology 2016;63:1632–1639
.
[295] Fiore M, Chiodini P, Pota V, Sansone P, Passavanti MB, Leone S, et al.
Risk of spontaneous fungal peritonitis in hospitalized cirrhotic patients
with ascites: a systematic review of observational studies and meta-analysis. Minerva Anestesiol 2017;83:1309–1316
.
[296] Piano S, Angeli P. Reply letter. Hepatology 2016;64:998–999 .Clinical Practice Guidelines
48 Journal of Hepatology 2018 vol. xxx jxxx–xxx
Please cite this article in press as: The European Association for the Study of the Liver. EASL Clinical Practice Guidelines for the management of pati ents with decompensated cirrhosis. J
Hepatol (2018), https://doi.org/10.1016/j.jhep.2018.03.024

[297] Guevara M, Terra C, Nazar A, Sola E, Fernandez J, Pavesi M, et al.
Albumin for bacterial infections other than spontaneous bacterialperitonitis in cirrhosis. A randomized, controlled study. J Hepatol2012;57:759–765
.
[298] Thevenot T, Bureau C, Oberti F, Anty R, Louvet A, Plessier A, et al. Effectof albumin in cirrhotic patients with infection other than spontaneous
bacterial peritonitis. A randomized trial. J Hepatol 2015;62:822–830
.
[299] Piano S, Romano A, Di Pascoli M, Angeli P. Why and how to measurerenal function in patients with liver disease. Liver Int 2017;37:116–122
.
[300] Gonwa TA, Jennings L, Mai ML, Stark PC, Levey AS, Klintmalm GB.Estimation of glomerular filtration rates before and after orthotopic
liver transplantation: evaluation of current equations. Liver Transpl
2004;10:301–309
.
[301] Hoek FJ, Kemperman FA, Krediet RT. A comparison between cystatin C,plasma creatinine and the Cockcroft and Gault formula for theestimation of glomerular filtration rate. Nephrol Dial Transplant2003;18:2024–2031
.
[302] Inker LA, Schmid CH, Tighiouart H, Eckfeldt JH, Feldman HI, Greene T,
et al. Estimating glomerular filtration rate from serum creatinine and
cystatin C. N Engl J Med 2012;367:20–29 .
[303] Roy L, Legault L, Pomier-Layrargues G. Glomerular filtration ratemeasurement in cirrhotic patients with renal failure. Clin Nephrol1998;50:342–346
.
[304] Francoz C, Nadim MK, Baron A, Prie D, Antoine C, Belghiti J, et al.
Glomerular filtration rate equations for liver-kidney transplantation in
patients with cirrhosis: validation of current recommendations. Hepa-tology 2014;59:1514–1521
.
[305] Francoz C, Glotz D, Moreau R, Durand F. The evaluation of renalfunction and disease in patients with cirrhosis. J Hepatol2010;52:605–613
.
[306] Garcia-Tsao G, Parikh CR, Viola A. Acute kidney injury in cirrhosis.
Hepatology 2008;48:2064–2077 .
[307] Salerno F, Gerbes A, Gines P, Wong F, Arroyo V. Diagnosis, preventionand treatment of hepatorenal syndrome in cirrhosis. Gut 2007;56:310–1318
.
[308] Mehta RL, Kellum JA, Shah SV, Molitoris BA, Ronco C, Warnock DG, et al.Acute Kidney Injury Network: report of an initiative to improve
outcomes in acute kidney injury. Crit Care 2007;11:R31
.
[309] Nadim MK, Kellum JA, Davenport A, Wong F, Davis C, Pannu N, et al.Hepatorenal syndrome: the 8th International Consensus Conference ofthe Acute Dialysis Quality Initiative (ADQI) Group. Crit Care 2012;16:R23
.
[310] Khwaja A. KDIGO clinical practice guidelines for acute kidney injury.Nephron Clin Pract 2012;120:c179–c184
.
[311] Belcher JM, Garcia-Tsao G, Sanyal AJ, Bhogal H, Lim JK, Ansari N, et al.
Association of AKI with mortality and complications in hospitalizedpatients with cirrhosis. Hepatology 2013;57:753–762
.
[312] de Carvalho JR, Villela-Nogueira CA, Luiz RR, Guzzo PL, da Silva Rosa JM,Rocha E, et al. Acute kidney injury network criteria as a predictor ofhospital mortality in cirrhotic patients with ascites. J Clin Gastroenterol
2012;46:21–26
.
[313] Fagundes C, Barreto R, Guevara M, Garcia E, Solà E, Rodríguez E, et al. Amodified acute kidney injury classification for diagnosis and riskstratification of impairment of kidney function in cirrhosis. J Hepatol2013;59:474–481
.
[314] Piano S, Rosi S, Maresio G, Fasolato S, Cavallin M, Romano A, et al.Evaluation of the Acute Kidney Injury Network criteria in hospi-
talized patients with c irrhosis and ascites. J Hepatol 2013;59:482–489
.
[315] Tsien CD, Rabie R, Wong F. Acute kidney injury in decompensatedcirrhosis. Gut 2013;62:131–137
.
[316] Wong F, O’Leary JG, Reddy KR, Patton H, Kamath PS, Fallon MB, et al.New consensus definition of acute kidney injury accurately predicts 30-day mortality in patients with cirrhosis and infection. Gastroenterology
2013;145:1280–1288
.
[317] Huelin P, Piano S, Solà E, Stanco M, Solé C, Moreira R, et al. Validation ofa staging system for acute kidney injury in patients with cirrhosis andassociation with acute-on-chronic liver failure. Clin GastroenterolHepatol 2017;15:438–445
.
[318] Angeli P, Gines P, Wong F, Bernardi M, Boyer TD, Gerbes A, et al.Diagnosis and management of acute kidney injury in patients with
cirrhosis: revised consensus recommendations of the International
Club of Ascites. J Hepatol 2015;62:968–974
.
[319] Kidney Disease: Inproving Global Outcomes (KDIGO) CKD Work Group.KDIGO. Clinical practice guideline for the evaluation and managementof chronic kidney disease. Kidny Int 2012;2013:1–150
.[320] Rosi S, Piano S, Frigo AC, Morando F, Fasolato S, Cavallin M, et al. New
ICA criteria for the diagnosis of acute kidney injury in cirrhotic patients:can we use an imputed value of serum creatinine? Liver Int2015;35:2108–2114
.
[321] Guevara M, Fernandez-Esparrach G, Alessandria C, Torre A, Terra C,Montana X, et al. Effects of contrast media on renal function in patients
with cirrhosis: a prospective study. Hepatology 2004;40:646–651
.
[322] Umgelter A, Reindl W, Franzen M, Lenhardt C, Huber W, Schmid RM.Renal resistive index and renal function before and after paracentesis inpatients with hepatorenal syndrome and tense ascites. Intensive CareMed 2009;35:152–156
.
[323] Cabrera J, Falcón L, Gorriz E, Pardo MD, Granados R, Quinones A,
Maynar M. Abdominal decompression plays a major role in early
postparacentesis haemodynamic changes in cirrhotic patients withtense ascites. Gut 2001;48:384–389
.
[324] de Cleva R, Silva FP, Zilberstein B, Machado DJ. Acute renal failure dueto abdominal compartment syndrome: report on four cases andliterature review. Rev Hospital das Clin 2001;56:123–130
.
[325] Nadim MK, Durand F, Kellum JA, Levitsky J, O’Leary JG, Karvellas CJ,
et al. Management of the critically ill patient with cirrhosis: A
multidisciplinary perspective. J Hepatol 2016;64:717–735 .
[326] Umgelter A, Reindl W, Wagner KS, Franzen M, Stock K, Schmid RM,et al. Effects of plasma expansion with albumin and paracentesis onhaemodynamics and kidney function in critically ill cirrhotic patientswith tense ascites and hepatorenal syndrome: a prospective uncon-
trolled trial. Critical care 2008;12:R4
.
[327] Moreau R, Lebrec D. Diagnosis and treatment of acute renal failure inpatients with cirrhosis. Best Pract Res Clin Gastroenterol2007;21:111–123
.
[328] Moreau R, Lebrec D. Acute renal failure in patients with cirrhosis:perspectives in the age of MELD. Hepatology 2003;37:233–243
.
[329] Fagundes C, Pepin MN, Guevara M, Barreto R, Casals G, Sola E, et al.
Urinary neutrophil gelatinase-associated lipocalin as biomarker in the
differential diagnosis of impairment of kidney function in cirrhosis. JHepatol 2012;57:267–273
.
[330] Belcher JM, Sanyal AJ, Peixoto AJ, Perazella MA, Lim J, Thiessen-Philbrook H, et al. Kidney biomarkers and differential diagnosis ofpatients with cirrhosis and acute kidney injury. Hepatology
2014;60:622–632
.
[331] Qasem AA, Farag SE, Hamed E, Emara M, Bihery A, Pasha H. Urinarybiomarkers of acute kidney injury in patients with liver cirrhosis. ISRNNephrol 2014;2014:376795
.
[332] Barreto R, Elia C, Sola E, Moreira R, Ariza X, Rodriguez E, et al. Urinaryneutrophil gelatinase-associated lipocalin predicts kidney outcome anddeath in patients with cirrhosis and bacterial infections. J Hepatol
2014;61:35–42
.
[333] Ariza X, Sola E, Elia C, Barreto R, Moreira R, Morales-Ruiz M, et al.Analysis of a urinary biomarker panel for clinical outcomes assessmentin cirrhosis. PLoS One 2015;10:e0128145
.
[334] Puthumana J, Ariza X, Belcher JM, Graupera I, Ginès P, Parikh CR. Urineinterleukin 18 and lipocalin 2 are biomarkers of acute tubular necrosis
in patients with cirrhosis: a systematic review and meta-analysis. Clin
Gastroenterol Hepatol 2017;15:1003–1013
.
[335] Huelin P, Elia C, Solà E, Solé C, Moreira R, Carol M. New diagnosticalgorithm of acute kidney injury in cirrhosis that includes categoriza-tion of stage 1 and assessment of urine NGAL. Relevance for thedifferential diagnosis and clinical outcomes. J Hepatol 2017;66:S11
.
[336] Chawla LS, Eggers PW, Star RA, Kimmel PL. Acute kidney injury and
chronic kidney disease as interconnected syndromes. N Engl J Med
2014;371:58–66 .
[337] Trawale JM, Paradis V, Rautou PE, Francoz C, Escolano S, Sallee M, et al.The spectrum of renal lesions in patients with cirrhosis: a clinico-pathological study. Liver Int 2010;30:725–732
.
[338] Wadei HM, Geiger XJ, Cortese C, Mai ML, Kramer DJ, Rosser BG, et al.
Kidney allocation to liver transplant candidates with renal failure of
undetermined etiology: role of percutaneous renal biopsy. Am JTransplant 2008;8:2618–2626
.
[339] Wong F, Nadim MK, Kellum JA, Salerno F, Bellomo R, Gerbes A, et al.Working Party proposal for a revised classification system of renaldysfunction in patients with cirrhosis. Gut 2011;60:702–709
.
[340] Stadlbauer V, Wright GA, Banaji M, Mukhopadhya A, Mookerjee RP,
Moore K, et al. Relationship between activation of the sympathetic
nervous system and renal blood flow autoregulation in cirrhosis.Gastroenterology 2008;134:111–119
.
[341] Wiest R, Lawson M, Geuking M. Pathological bacterial translocation inliver cirrhosis. J Hepatol 2014;60:197–209
.JOURNAL
OF HEPATOLOGY
Journal of Hepatology 2018 vol. xxx jxxx–xxx 49
Please cite this article in press as: The European Association for the Study of the Liver. EASL Clinical Practice Guidelines for the management of pati ents with decompensated cirrhosis. J
Hepatol (2018), https://doi.org/10.1016/j.jhep.2018.03.024

[342] Albillos A, de la Hera A, Gonzalez M, Moya JL, Calleja JL, Monserrat J,
et al. Increased lipopolysaccharide binding protein in cirrhotic patientswith marked immune and hemodynamic derangement. Hepatology2003;37:208–217
.
[343] Navasa M, Follo A, Filella X, Jimenez W, Francitorra A, Planas R, et al.Tumor necrosis factor and interleukin-6 in spontaneous bacterial
peritonitis in cirrhosis: relationship with the development of renal
impairment and mortality. Hepatology 1998;27:1227–1232
.
[344] Shah N, Dhar D, El Zahraa Mohammed F, Habtesion A, Davies NA,Jover-Cobos M, et al. Prevention of acute kidney injury in a rodentmodel of cirrhosis following selective gut decontamination is asso-ciated with reduced renal TLR4 expression. J Hepatol 2012;56:
1047–1053
.
[345] Shah N, Mohamed FE, Jover-Cobos M, Macnaughtan J, Davies N, MoreauR, et al. Increased renal expression and urinary excretion of TLR4 inacute kidney injury associated with cirrhosis. Liver Int 2013;33:398–409
.
[346] Alobaidi R, Basu RK, Goldstein SL, Bagshaw SM. Sepsis-associated acutekidney injury. Semin Nephrol 2015;35:2–11
.
[347] Emlet DR, Shaw AD, Kellum JA. Sepsis-associated AKI: epithelial cell
dysfunction. Semin Nephrol 2015;35:85–95 .
[348] Prowle JR, Bellomo R. Sepsis-associated acute kidney injury: macrohe-modynamic and microhemodynamic alterations in the renal circula-tion. Semin Nephrol 2015;35:64–74
.
[349] de Seigneux S, Martin PY. Preventing the progression of AKI to CKD: the
role of mitochondria. J Am Soc Nephrol 2017;28:1327–1329 .
[350] Bairaktari E, Liamis G, Tsolas O, Elisaf M. Partially reversible renaltubular damage in patients with obstructive jaundice. Hepatology2001;33:1365–1369
.
[351] van Slambrouck CM, Salem F, Meehan SM, Chang A. Bile castnephropathy is a common pathologic finding fo kidney injury associ-ated with severe liver dysfunction. Kidney Int 2013;84:192–197
.
[352] Durand F, Graupera I, Gines P, Olson JC, Nadim MK. Pathogenesis of
hepatorenal syndrome: implications for therapy. Am J Kidney Dis2016;67:318–328
.
[353] Moreau R, Durand F, Poynard T, Duhamel C, Cervoni JP, Ichai P, et al.Terlipressin in patients with cirrhosis and type 1 hepatorenal syn-drome: a retrospective multicenter study. Gastroenterology 2002;122:
923–930
.
[354] Ortega R, Gines P, Uriz J, et al. Terlipressin therapy with and withoutalbumin for patients with hepatorenal syndrome: results of a prospec-tive, nonrandomized study. Hepatology 2002;36:941–948
.
[355] Sanyal AJ, Boyer T, Garcia-Tsao G, Regenstein F, Rossaro L, Appenrodt B,et al. A randomized prospective double blind, placebo controlled studyof terlipressin for type 1 hepatorenal syndrome. Gastroenterology
2008;134:1360–1368
.
[356] Martin-Llahi M, Pepin MN, Guevara M, Diaz F, Torre A, Monescillo A,et al. Terlipressin and albumin vs albumin in patients with cirrhosisand hepatorenal syndrome: a randomized study. Gastroenterology2008;134:1352–1359
.
[357] Boyer TD, Sanyal AJ, Wong F, Frederick RT, Lake JR, O’Leary JG, et al.
REVERSE study investigators terlipressin plus albumin is more effective
than albumin alone in improving renal function in patients withcirrhosis and hepatorenal syndrome type 1. Gastroenterology2016;150:1579–1589
.
[358] Rodriguez E, Elia C, Sola E, Barreto R, Graupera I, Andrealli A, et al.Terlipressin and albumin for type-1 hepatorenal syndrome associatedwith sepsis. J Hepatol 2014;60:955–961
.
[359] Cavallin M, Kamath PS, Merli M, Fasolato S, Toniutto P, Salerno F, et al.
Terlipressin plus albumin vs.midodrine and octreotide plus albumin in
the treatment of hepatorenal syndrome: A randomized trial. Hepatol-ogy 2015;62:567–574
.
[360] Cavallin M, Piano S, Romano A, Fasolato S, Frigo AC, Benetti G, et al.Terlipressin given by continuous intravenous infusion vs.intravenous
boluses in the treatment of hepatorenal syndrome: A randomized
controlled study. Hepatology 2016;63:983–992
.
[361] Gluud LL, Christensen K, Christensen E, Krag A. Terlipressin forhepatorenal syndrome. Cochrane Database Syst Rev 2012:CD005162
.
[362] Facciorusso A, Chandar AK, Murad MH, Prokop LJ, Muscatiello N,Kamath PS. Singh SComparative efficacy of pharmacological strategiesfor management of type 1 hepatorenal syndrome: a systematic review
and network meta-analysis. Lancet Gastroenterol Hepatol
2017;2:94–102
.
[363] Narahara Y, Kanazawa H, Taki Y, Kimura Y, Atsukawa M, Katakura T,et al. Effects of terlipressin on systemic, hepatic and renal hemody-namics in patients with cirrhosis. J Gastroenterol Hepatol 2009;24:1791–1797
.
[364] Garcia-Martinez R, Caraceni P, Bernardi M, Gines P, Arroyo V, Jalan R.Albumin: pathophysiologic basis of its role in the treatment of cirrhosisand its complications. Hepatology 2013;58:1836–1846
.
[365] Gerbes AL, Huber E, Gulberg V. Terlipressin for hepatorenal syndrome:
continuous infusion as an alternative to i.v. bolus administration.
Gastroenterology 2009;137:1179–1181 .
[366] Piano S, Morando F, Fasolato S, Cavallin M, Boscato N, Boccagni P, et al.Continuous recurrence of type 1 hepatorenal syndrome and long-termtreatment with terlipressin and albumin: a new exception to MELDscore in the allocation system to liver transplantation? J Hepatol
2011;55:491–496
.
[367] Gow PJ, Ardalan ZS, Vasudevan A, Testro AG, Ye B, Angus PW.Outpatient terlipressin infusion for the treatment of refractory ascites.Am J Gastroenterol 2016;111:1041–1042
.
[368] Alessandria C, Ottobrelli A, Debernardi-Venon W, Todros L, CerenziaMT, Martini S, et al. Noradrenalin vs terlipressin in patients withhepatorenal syndrome: a prospective, randomized, unblinded, pilot
study. J Hepatol 2007;47:499–505
.
[369] Duvoux C, Zanditenas D, Hezode C, Chauvat A, Monin JL, Roudot-Thoraval F, et al. Effects of noradrenalin and albumin in patients withtype I hepatorenal syndrome: a pilot study. Hepatology 2002;36:374–380
.
[370] Singh V, Ghosh S, Singh B, Kumar P, Sharma N, Bhalla A, et al.Noradrenaline vs.terlipressin in the treatment of hepatorenal syn-
drome: a randomized study. J Hepatol 2012;56:1293–1298
.
[371] Sharma P, Kumar A, Shrama BC, Sarin SK. An open label, pilot,randomized controlled trial of noradrenaline vs.terlipressin in the
treatment of type 1 hepatorenal syndrome and predictors of response.Am J Gastroenterol 2008;103:1689–1697
.
[372] Esrailian E, Pantangco ER, Kyulo NL, Hu KQ, Runyon BA. Octreotide/Midodrine therapy significantly improves renal function and 30-day
survival in patients with type 1 hepatorenal syndrome. Dig Dis Sci
2007;52:742–748
.
[373] Restuccia T, Ortega R, Guevara M, Gines P, Alessandria C, Ozdogan O,et al. Effects of treatment of hepatorenal syndrome before transplan-tation on posttransplantation outcome. A case-control study. J Hepatol2004;40:140–146
.
[374] Rodriguez E, Henrique Pereira G, Sola E, Elia C, Barreto R, Pose E, et al.
Treatment of type 2 hepatorenal syndrome in patients awaitingtransplantation: Effects on kidney function and transplantation out-comes. Liver Transpl 2015;21:1347–1354
.
[375] Boyer TD, Sanyal AJ, Garcia-Tsao G, Blei A, Carl D, Bexon AS, et al.Predictors of response to terliupressin plus albumin in hepatorenalsyndrome (HRS) type 1: relationship of serum creatinine to hemody-
namics. J Hepatol 2011;55:315–321
.
[376] Piano S, Schmidt HH, Ariza X, Amoros A, Romano A, Solà E, et al. Impactof Acute-on-Chronic Liver Failure on response to treatment withterlipressin and albumin in patients with type 1 hepatorenal syndrome.J Hepatol 2017;66:S572
.
[377] Nazar A, Pereira GH, Guevara M, Martín-Llahi M, Pepin MN, Marinelli
M, et al. Predictors of response to therapy with terlipressin and
albumin in patients with cirrhosis and type 1 hepatorenal syndrome.Hepatology 2010;51:219–226
.
[378] Brensing KA, Textor J, Perz J, Schiedermaier P, Raab P, Strunk H, et al.Long term outcome after transjugular intrahepatic portosystemic stent-shunt in non-transplant cirrhotics with hepatorenal syndrome: a phaseII study. Gut 2000;47:288–295
.
[379] Guevara M, Gines P, Bandi JC, Gilabert R, Sort P, Jimenez W, et al.
Transjugular intrahepatic portosystemic shunt in hepatorenal syn-drome: effects on renal function and vasoactive systems. Hepatology1998;28:416–422
.
[380] Testino G, Ferro C, Sumberaz A, Messa P, Morelli N, Guadagni B, et al.Type-2 hepatorenal syndrome and refractory ascites: role of transjugu-
lar intrahepatic portosystemic stent-shunt in eighteen patients with
advanced cirrhosis awaiting orthotopic liver transplantation. Hepato-gastroenterology 2003;50:1753–1755
.
[381] Keller F, Heinze H, Jochimsen F, Passfall J, Schuppan D, Buttner P. Riskfactors and outcome of 107 patients with decompensated liver diseaseand acute renal failure (including 26 patients with hepatorenalsyndrome): the role of hemodialysis. Ren Fail 1995;17:135–146
.
[382] Sourianarayanane A, Raina R, Garg G, McCullough AJ, O’Shea RS.
Management and outcome in hepatorenal syndrome: need for renalreplacement therapy in non-transplanted patients. Int Urol Nephrol2014;46:793–800
.Clinical Practice Guidelines
50 Journal of Hepatology 2018 vol. xxx jxxx–xxx
Please cite this article in press as: The European Association for the Study of the Liver. EASL Clinical Practice Guidelines for the management of pati ents with decompensated cirrhosis. J
Hepatol (2018), https://doi.org/10.1016/j.jhep.2018.03.024

[383] Staufer K, Roedl K, Kivaranovic D, Drolz A, Horvatits T, Rasoul-
Rockenschaub S, et al. Renal replacement therapy in critically ill livercirrhotic patients-outcome and clinical implications. Liver Int2017;37:843–850
.
[384] Wu VC, Ko WJ, Chang HW, Chen YS, Chen YW, Chen YM, et al. Earlyrenal replacement therapy in patients with postoperative acute liver
failure associated with acute renal failure: effect on postoperative
outcomes. J Am Coll Surg 2007;205:266–276
.
[385] Gaudry S, Hajage D, Schortgen F, Martin-Lefevre L, Pons B, Boulet E,et al. Initiation strategies for renal-replacement therapy in the inten-sive care unit. N Engl J Med 2016;375:122–133
.
[386] Zarbock A, Kellum JA, Schmidt C, Van Aken H, Wempe C, Pavenstadt
H, et al. Effect of early vs delayed initiation of renal replacement
therapy on mortality in critically ill patients with acute kidneyinjury: the elain randomized clinical trial. JAMA 2016;315:2190–2199
.
[387] Wong LP, Blackley MP, Andreoni KA, Chin H, Falk RJ, Klemmer PJ.Survival of liver transplant candidates with acute renal failure receivingrenal replacement therapy. Kidney Int 2005;68:362–370
.
[388] Banares R, Nevens F, Larsen FS, Jalan R, Albillos A, Dollinger M, et al.
Extracorporeal albumin dialysis with the molecular adsorbent recircu-lating system in acute-on-chronic liver failure: the RELIEF trial.Hepatology 2013;57:1153–1162
.
[389] Kribben A, Gerken G, Haag S, Herget-Rosenthal S, Treichel U, Betz C,et al. Effects of fractionated plasma separation and adsorption on
survival in patients with acute-on-chronic liver failure. Gastroenterol-
ogy 2012;142:782–789
.
[390] Boyer TD, Sanyal AJ, Garcia-Tsao G, Regenstein F, Rossaro L, AppenrodtB, et al. Impact of liver transplantation on the survival of patientstreated for hepatorenal syndrome type 1. Liver Transpl 2011;17:1328–1332
.
[391] Gonwa TA, Klintmalm GB, Levy M, Jennings LS, Goldstein RM, Husberg
BS. Impact of pretransplant renal function on survival after liver
transplantation. Transplantation 1995;59:361–365 .
[392] Nadim MK, Sung RS, Davis CL, Andreoni KA, Biggins SW, Danovitch GM,et al. Simultaneous liver-kidney transplantation summit: current stateand future directions. Am J Transplant 2012;12:2901–2908
.
[393] Francoz C, Nadim MK, Durand F. Kidney biomarkers in cirrhosis. J
Hepatol 2016;65:809–824 .
[394] Angeli P, Gines P. Hepatorenal syndrome, MELD score and livertransplantation: an evolving issue with relevant implications forclinical practice. J Hepatol 2012;57:1135–1140
.
[395] Cillo U, Burra P, Mazzaferro V, Belli L, Pinna AD, Spada M, et al. Amultistep, consensus-based approach to organ allocation in livertransplantation: toward a ‘‘blended principle model”. Am J Transplant
2015;1:2552–2561
.
[396] Akriviadis E, Botla R, Briggs W, Han S, Reynolds T, Shakil O. Pentox-ifylline improves short-term survival in severe acute alcoholic hepati-tis: a double-blind, placebo-controlled trial. Gastroenterology2000;119:1637–1648
.
[397] Mathurin P, Louvet A, Duhamel A, Nahon P, Carbonell N, Boursier J,
et al. Prednisolone with vs without pentoxifylline and survival of
patients with severe alcoholic hepatitis: a randomized clinical trial.JAMA 2013;310:1033–1041
.
[398] Parker R, Armstrong MJ, Corbett C, Rowe IA, Houlihan DD. Systematicreview: pentoxifylline for the treatment of severe alcoholic hepatitis.Aliment Pharmacol Ther 2013;37:845–854
.
[399] Allen AM, Kim WR. Epidemiology and healthcare burden of acute-on-
chronic liver failure. Semin Liver Dis 2016;36:123–126 .
[400] Piano S, Tonon M, Vettore E, Stanco M, Pilutti C, Romano A, et al.Incidence, predictors and outcomes of acute-on-chronic liver failure inoutpatients with cirrhosis. J Hepatol 2017;67:1177–1184
.
[401] Arroyo V, Jalan R. Acute-on-chronic liver failure: definition, diagnosis,and clinical characteristics. Semin Liver Dis 2016;36:109–116
.
[402] Sarin SK, Kedarisetty CK, Abbas Z, Amarapurkar D, Bihari C, Chan AC,
et al. Acute-on-chronic liver failure: consensus recommendations of theAsian Pacific Association for the Study of the Liver (APASL) 2014.Hepatol Int 2014;8:453–471
.
[403] Jalan R, Yurdaydin C, Bajaj JS, Acharya SK, Arroyo V, Lin HC, et al.Toward an improved definition of acute on chronic liver failure.Gastroenterology 2014;147:4–10
.
[404] Bajaj JS. Defining acute-on-chronic liver failure: will East and West ever
meet? Gastroenterology 2013;144:1337–1339 .
[405] Kim TY, Song DS, Kim HY, Sinn DH, Yoon EL, Kim CW, et al.Characteristics and discrepancies in acute-on-chronic liver failure:need for a unified definition. PLoS One 2016;11:e0146745
.[406] Arroyo V, Moreau R, Jalan R, Gines P. Acute-on-chronic liver failure: A
new syndrome that will re-classify cirrhosis. J Hepatol 2015;62:S131–S143
.
[407] Jalan R, Pavesi M, Saliba F, Amoros A, Fernandez J, Holland-Fischer P,et al. The CLIF Consortium Acute Decompensation score (CLIF-C ADs)for prognosis of hospitalised cirrhotic patients without acute-on-
chronic liver failure. J Hepatol 2015;62:831–840
.
[408] Arroyo V, Moreau R, Kamath PS, Jalan R, Gines P, Nevens F, et al. Acute-on-chronic liver failure in cirrhosis. Nat Rev Dis Primers 2016;2:16041
.
[409] Trebicka J. Predisposing factors in acute-on-chronic liver failure. SeminLiver Dis 2016;36:167–173
.
[410] Sarin SK, Choudhury A. Acute-on-chronic liver failure. Curr Gastroen-
terol Rep 2016;18:61 .
[411] Bajaj JS, O’Leary JG, Reddy KR, Wong F, Biggins SW, Patton H, et al.Survival in infection-related acute-on-chronic liver failure is defined byextrahepatic organ failures. Hepatology 2014;60:250–256
.
[412] Silva PE, Fayad L, Lazzarotto C, Ronsoni MF, Bazzo ML, Colombo BS,et al. Single-centre validation of the EASL-CLIF consortium definition ofacute-on-chronic liver failure and CLIF-SOFA for prediction of mortality
in cirrhosis. Liver Int 2015;35:1516–1523
.
[413] Li H, Chen LY, Zhang NN, Li ST, Zeng B, Pavesi M, et al. Characteristics,diagnosis and prognosis of acute-on-chronic liver failure in cirrhosisassociated to hepatitis B. Sci Rep 2016;6:25487
.
[414] Shalimar, Saraswat V, Singh SP, Duseja A, Shukla A, Eapen CE, et al.Acute-on-chronic liver failure in India: The Indian National Association
for Study of the Liver consortium experience. J Gastroenterol Hepatol
2016;31:1742–1749
.
[415] Shi Y, Yang Y, Hu Y, Wu W, Yang Q, Zheng M, et al. Acute-on-chronicliver failure precipitated by hepatic injury is distinct from thatprecipitated by extrahepatic insults. Hepatology 2015;62:232–242
.
[416] Shalimar, Kumar D, Vadiraja PK, Nayak B, Thakur B, Das P, et al. Acuteon chronic liver failure because of acute hepatic insults: Etiologies,
course, extrahepatic organ failure and predictors of mortality. J
Gastroenterol Hepatol 2016;31:856–864
.
[417] Dhiman RK, Agrawal S, Gupta T, Duseja A, Chawla Y. Chronic liverfailure-sequential organ failure assessment is better than the asia-pacific association for the study of Liver criteria for defining acute-on-chronic liver failure and predicting outcome. World J Gastroenterol
2014;20:14934–14941
.
[418] Pischke S, Suneetha PV, Baechlein C, Barg-Hock H, Heim A, Kamar N,et al. Hepatitis E virus infection as a cause of graft hepatitis in livertransplant recipients. Liver Transpl 2010;16:74–82
.
[419] Jalan R, Saliba F, Pavesi M, Amoros A, Moreau R, Gines P, et al.Development and validation of a prognostic score to predict mortalityin patients with acute-on-chronic liver failure. J Hepatol 2014;61:
1038–1047
.
[420] McPhail MJ, Shawcross DL, Abeles RD, Chang A, Patel V, Lee GH, et al.Increased survival for patients with cirrhosis and organ failure in liverintensive care and validation of the chronic liver failure-sequentialorgan failure scoring system. Clin Gastroenterol Hepatol 2015;13:1353–1360
.
[421] Lee M, Lee JH, Oh S, Jang Y, Lee W, Lee HJ, et al. CLIF-SOFA scoring
system accurately predicts short-term mortality in acutely decompen-sated patients with alcoholic cirrhosis: a retrospective analysis. LiverInt 2015;35:46–57
.
[422] O’Leary JG, Reddy KR, Garcia-Tsao G, Biggins SW, Wong F, Fallon MB,et al. NACSELD acute-on-chronic liver failure (NACSELD-ACLF) scorepredicts 30-day survival in hospitalized patients with cirrhosis. Hepa-
tology 2018
.
[423] Choudhury A, Jindal A, Maiwall R, Sharma MK, Sharma BC, Pamecha V,et al. APASL ACLF Working PartyLiver failure determines the outcomein patients of acute-on-chronic liver failure (ACLF): comparison ofAPASL ACLF research consortium (AARC) and CLIF-SOFA models.Hepatol Int 2017;11:461–471
.
[424] Hernaez R, Sola E, Moreau R, Gines P. Acute-on-chronic liver failure: an
update. Gut 2017;66:541–553 .
[425] Gustot T, Fernandez J, Garcia E, Morando F, Caraceni P, Alessandria C,et al. Clinical Course of acute-on-chronic liver failure syndrome andeffects on prognosis. Hepatology 2015;62:243–252
.
[426] Garg H, Sarin SK, Kumar M, Garg V, Sharma BC, Kumar A. Tenofovirimproves the outcome in patients with spontaneous reactivation of
hepatitis B presenting as acute-on-chronic liver failure. Hepatology
2011;53:774–780
.
[427] Lin B, Pan CQ, Xie D, Xie J, Xie S, Zhang X, et al. Entecavir improves theoutcome of acute-on-chronic liver failure due to the acute exacerbationof chronic hepatitis B. Hepatol Int 2013;7:460–467
.JOURNAL
OF HEPATOLOGY
Journal of Hepatology 2018 vol. xxx jxxx–xxx 51
Please cite this article in press as: The European Association for the Study of the Liver. EASL Clinical Practice Guidelines for the management of pati ents with decompensated cirrhosis. J
Hepatol (2018), https://doi.org/10.1016/j.jhep.2018.03.024

[428] Zhang Y, Hu XY, Zhong S, Yang F, Zhou TY, Chen G, et al. Entecavir vs
lamivudine therapy for naive patients with spontaneous reactivation ofhepatitis B presenting as acute-on-chronic liver failure. World JGastroenterol. 2014;20:4745–4752
.
[429] Xiang-Hui Y, Lang X, Yan Z, Li Z, Xiao-Feng S, Hong R. Prediction ofprognosis to lamivudine in patients with spontaneous reactivation of
hepatitis B virus-related acute-on-chronic liver failure: using virologic
response at week 4. Eur J Intern Med 2014;25:860–864
.
[430] Chen JF, Wang KW, Zhang SQ, Lei ZY, Xie JQ, Zhu JY, et al. Dexametha-sone in outcome of patients with hepatitis B virus-related acute-on-chronic liver failure. J Gastroenterol Hepatol 2014;29:800–806
.
[431] Guo YM, Li FY, Gong M, Zhang L, Wang JB, Xiao XH, et al. Short-term
efficacy of treating hepatitis B virus-related acute-on-chronic liver
failure based on cold pattern differentiation with hot herbs: Arandomized controlled trial. Chin J Integr Med 2016;22:573–580
.
[432] Garg V, Garg H, Khan A, Trehanpati N, Kumar A, Sharma BC, et al.Granulocyte colony-stimulating factor mobilizes CD34(+) cells andimproves survival of patients with acute-on-chronic liver failure.Gastroenterology 2012;142:505–512
.
[433] Finkenstedt A, Nachbaur K, Zoller H, Joannidis M, Pratschke J, Graziadei
IW, et al. Acute-on-chronic liver failure: excellent outcomes after livertransplantation but high mortality on the wait list. Liver Transpl2013;19:879–886
.
[434] Artru F, Louvet A, Ruiz I, Levesque E, Labreuche J, Ursic-Bedoya J, et al.Liver transplantation in the most severely ill cirrhotic patients: A
multicenter study in acute-on-chronic liver failure grade 3. J Hepatol
2017;67:708–715
.
[435] Bouachour G, Tirot P, Gouello JP, Mathieu E, Vincent JF, Alquier P.Adrenocortical function during septic shock. Intensive Care Med1995;21:57–62
.
[436] Marik PE, Pastores SM, Annane D, Meduri GU, Sprung CL, Arlt W, et al.Recommendations for the diagnosis and management of corticosteroid
insufficiency in critically ill adult patients: consensus statements from
an international task force by the American College of Critical CareMedicine. Crit Care Med 2008;36:1937–1949
.
[437] Tsai MH, Peng YS, Chen YC, Liu NJ, Ho YP, Fang JT, et al. Adrenalinsufficiency in patients with cirrhosis, severe sepsis and septic shock.Hepatology 2006;43:673–681
.
[438] Acevedo J, Fernandez J, Prado V, Silva A, Castro M, Pavesi M, et al.
Relative adrenal insufficiency in decompensated cirrhosis: Relationshipto short-term risk of severe sepsis, hepatorenal syndrome, and death.Hepatology 2013;58:1757–1765
.
[439] Jang JY, Kim TY, Sohn JH, Lee TH, Jeong SW, Park EJ, et al. Relativeadrenal insufficiency in chronic liver disease: its prevalence and effectson long-term mortality. Aliment Pharmacol Ther 2014;40:819–826
.
[440] Tsai MH, Huang HC, Peng YS, Chen YC, Tian YC, Yang CW, et al. Critical
illness-related corticosteroid insufficiency in cirrhotic patients withacute gastroesophageal variceal bleeding: risk factors and associationwith outcome. Crit Care Med 2014;42:2546–2555
.
[441] Fede G, Spadaro L, Tomaselli T, Privitera G, Scicali R, Vasianopoulou P,et al. Comparison of total cortisol, free cortisol, and surrogate markers
of free cortisol in diagnosis of adrenal insufficiency in patients with
stable cirrhosis. Clin Gastroenterol Hepatol 2014;12:504–512
.
[442] Kim G, Huh JH, Lee KJ, Kim MY, Shim KY, Baik SK. Relative adrenalinsufficiency in patients with cirrhosis: a systematic review and meta-analysis. Dig Dis Sci 2017;62:1067–1079
.
[443] Fede G, Spadaro L, Tomaselli T, Privitera G, Germani G, Tsochatzis E,et al. Adrenocortical dysfunction in liver disease: a systematic review.
Hepatology 2012;55:1282–1291
.
[444] McNeilly AD, Macfarlane DP, O’Flaherty E, Livingstone DE, Mitic T,McConnell KM, et al. Bile acids modulate glucocorticoid metabolismand the hypothalamic–pituitary–adrenal axis in obstructive jaundice. JHepatol 2010;52:705–711
.
[445] Worlicek M, Knebel K, Linde HJ, Moleda L, Scholmerich J, Straub RH,
et al. Splanchnic sympathectomy prevents translocation and
spreading of E coli but not S aureus in liver cirrhosis. Gut 2010;59:1127–1134
.
[446] Cholongitas E, Goulis I, Pagkalidou E, Haidich AB, Karagiannis AKA,Nakouti T, et al. Relative adrenal insufficiency is associated with theclinical outcome in patients with stable decompensated cirrhosis. AnnHepatol 2017;16:584–590
.
[447] Trifan A, Chiriac S, Stanciu C. Update on adrenal insufficiency in
patients with liver cirrhosis. World J Gastroenterol 2013;19:445–456 .
[448] Galbois A, Rudler M, Massard J, Fulla Y, Bennani A, Bonnefont-RousselotD, et al. Assessment of adrenal function in cirrhotic patients: salivarycortisol should be preferred. J Hepatol 2010;52:839–845
.[449] Hamrahian AH, Oseni TS, Arafah BM. Measurements of serum-free
cortisol in critically ill patients. N Engl J Med 2004;350:1629–1638 .
[450] Tan T, Chang L, Woodward A, McWhinney B, Galligan J, Macdonald GA,et al. Characterising adrenal function using directly measured plasmafree cortisol in stable severe liver disease. J Hepatol 2010;53:841–848
.
[451] Coolens JL, Van Baelen H, Heyns W. Clinical use of unbound plasma
cortisol as calculated from total cortisol and corticosteroid-binding
globulin. J Steroid Biochem 1987;26:197–202 .
[452] le Roux CW, Sivakumaran S, Alaghband-Zadeh J, Dhillo W, Kong WM,Wheeler MJ. Free cortisol index as a surrogate marker for serum freecortisol. Ann Clin Biochem 2002;39:406–408
.
[453] Arafah BM, Nishiyama FJ, Tlaygeh H, Hejal R. Measurement of salivary
cortisol concentration in the assessment of adrenal function in critically
ill subjects: a surrogate marker of the circulating free cortisol. J ClinEndocrinol Metab 2007;92:2965–2971
.
[454] Thevenot T, Borot S, Remy-Martin A, Sapin R, Cervoni JP, Richou C, et al.Assessment of adrenal function in cirrhotic patients using concentra-tion of serum-free and salivary cortisol. Liver Int 2011;31:425–433
.
[455] Fernandez J, Escorsell A, Zabalza M, Felipe V, Navasa M, Mas A, et al.
Adrenal insufficiency in patients with cirrhosis and septic shock: Effect
of treatment with hydrocortisone on survival. Hepatology 2006;44:1288–1295
.
[456] Arabi YM, Aljumah A, Dabbagh O, Tamim HM, Rishu AH, Al-Abdulka-reem A, et al. Low-dose hydrocortisone in patients with cirrhosis andseptic shock: a randomized controlled trial. CMAJ 2010;182:
1971–1977
.
[457] Moller S, Henriksen JH. Cirrhotic cardiomyopathy: a pathophysiologicalreview of circulatory dysfunction in liver disease. Heart 2002;87:9–15
.
[458] Hunter JD, Doddi M. Sepsis and the heart. Br J Anaesth 2010;104:3–11 .
[459] Mehta G, Gustot T, Mookerjee RP, Garcia-Pagan JC, Fallon MB, Shah VH,et al. Inflammation and portal hypertension – the undiscoveredcountry. J Hepatol 2014;61:155–163
.
[460] De BK, Majumdar D, Das D, Biswas PK, Mandal SK, Ray S, et al. Cardiac
dysfunction in portal hypertension among patients with cirrhosis andnon-cirrhotic portal fibrosis. J Hepatol 2003;39:315–319
.
[461] Karagiannakis DS, Vlachogiannakos J, Anastasiadis G, Vafiadis-Zoubou-lis I, Ladas SD. Diastolic cardiac dysfunction is a predictor of dismalprognosis in patients with liver cirrhosis. Hepatol Int 2014;8:588–594
.
[462] Rabie RN, Cazzaniga M, Salerno F, Wong F. The use of E/A ratio as a
predictor of outcome in cirrhotic patients treated with transjugularintrahepatic portosystemic shunt. Am J Gastroenterol 2009;104:2458–2466
.
[463] Grose RD, Nolan J, Dillon JF, Errington M, Hannan WJ, Bouchier IA, et al.Exercise-induced left ventricular dysfunction in alcoholic and non-alcoholic cirrhosis. J Hepatol 1995;22:326–332
.
[464] Wong F, Girgrah N, Graba J, Allidina Y, Liu P, Blendis L. The cardiac
response to exercise in cirrhosis. Gut 2001;49:268–275 .
[465] Bernardi M, Rubboli A, Trevisani F, Cancellieri C, Ligabue A, Baraldini M,et al. Reduced cardiovascular responsiveness to exercise-inducedsympathoadrenergic stimulation in patients with cirrhosis. J Hepatol1991;12:207–216
.
[466] Kim MY, Baik SK, Won CS, Park HJ, Jeon HK, Hong HI, et al. Dobutamine
stress echocardiography for evaluating cirrhotic cardiomyopathy inliver cirrhosis. Korean J Hepatol 2010;16:376–382
.
[467] Krag A, Bendtsen F, Dahl EK, Kjaer A, Petersen CL, Moller S. Cardiacfunction in patients with early cirrhosis during maximal beta-adren-ergic drive: a dobutamine stress study. PLoS One 2014;9:e109179
.
[468] Stanton T, Leano R, Marwick TH. Prediction of all-cause mortality from
global longitudinal speckle strain: comparison with ejection fraction
and wall motion scoring. Circ Cardiovasc Imaging 2009;2:356–364 .
[469] Lang RM, Bierig M, Devereux RB, Flachskampf FA, Foster E, Pellikka PA,et al. Recommendations for chamber quantification: a report from theAmerican Society of Echocardiography’s Guidelines and StandardsCommittee and the Chamber Quantification Writing Group, developed
in conjunction with the European Association of Echocardiography, a
branch of the European Society of Cardiology. J Am Soc Echocardiogr2005;18:1440–1463
.
[470] Stampehl MR, Mann DL, Nguyen JS, Cota F, Colmenares C, Dokainish H.Speckle strain echocardiography predicts outcome in patients withheart failure with both depressed and preserved left ventricularejection fraction. Echocardiography 2015;32:71–78
.
[471] Chen Y, Chan AC, Chan SC, Chok SH, Sharr W, Fung J, et al. A detailed
evaluation of cardiac function in cirrhotic patients and its alterationwith or without liver transplantation. J Cardiol 2016;67:140–146
.
[472] Sampaio F, Pimenta J, Bettencourt N, Fontes-Carvalho R, Silva AP,Valente J, et al. Systolic and diastolic dysfunction in cirrhosis: a tissue-Clinical Practice Guidelines
52 Journal of Hepatology 2018 vol. xxx jxxx–xxx
Please cite this article in press as: The European Association for the Study of the Liver. EASL Clinical Practice Guidelines for the management of pati ents with decompensated cirrhosis. J
Hepatol (2018), https://doi.org/10.1016/j.jhep.2018.03.024

Doppler and speckle tracking echocardiography study. Liver Int
2013;33:1158–1165 .
[473] Nazar A, Guevara M, Sitges M, Terra C, Sola E, Guigou C, et al. LEFTventricular function assessed by echocardiography in cirrhosis: rela-tionship to systemic hemodynamics and renal dysfunction. J Hepatol2013;58:51–57
.
[474] Cesari M, Fasolato S, Rosi S, Angeli P. Cardiac dysfunction in patients
with cirrhosis: is the systolic component its main feature? Eur JGastroenterol Hepatol 2015;27:660–666
.
[475] Nagueh SF, Smiseth OA, Appleton CP, Byrd 3rd BF, Dokainish H,Edvardsen T, et al. Recommendations for the evaluation of leftventricular diastolic function by echocardiography: an update from
the American society of echocardiography and the european associa-
tion of cardiovascular imaging. Eur Heart J Cardiovasc Imaging2016;17:1321–1360
.
[476] Takemoto Y, Barnes ME, Seward JB, Lester SJ, Appleton CA, Gersh BJ,et al. Usefulness of left atrial volume in predicting first congestive heartfailure in patients > or = 65 years of age with well-preserved leftventricular systolic function. Am J Cardiol 2005;96:832–836
.
[477] Valeriano V, Funaro S, Lionetti R, Riggio O, Pulcinelli G, Fiore P, et al.
Modification of cardiac function in cirrhotic patients with and withoutascites. Am J Gastroenterol 2000;95:3200–3205
.
[478] Pozzi M, Carugo S, Boari G, Pecci V, de Ceglia S, Maggiolini S, et al.Evidence of functional and structural cardiac abnormalities incirrhotic patients with and without ascites. Hepatology 1997;26:
1131–1137
.
[479] Cazzaniga M, Salerno F, Pagnozzi G, Dionigi E, Visentin S, Cirello I, et al.Diastolic dysfunction is associated with poor survival in patients withcirrhosis with transjugular intrahepatic portosystemic shunt. Gut2007;56:869–875
.
[480] Ruiz-del-Arbol L, Achecar L, Serradilla R, Rodriguez-Gandia MA, RiveroM, Garrido E, et al. Diastolic dysfunction is a predictor of poor outcomes
in patients with cirrhosis, portal hypertension, and a normal creatinine.
Hepatology 2013;58:1732–1741
.
[481] Cesari M, Frigo AC, Tonon M, Angeli P. Cardiovascular predictors ofdeath in patients with cirrhosis. Hepatology 2017
.
[482] Merli M, Torromeo C, Giusto M, Iacovone G, Riggio O, Puddu PE.Survival at 2 years among liver cirrhotic patients is influenced by left
atrial volume and left ventricular mass. Liver Int 2017;37:700–706
.
[483] Kozor R, Nordin S, Treibel TA, Rosmini S, Castelletti S, Fontana M, et al.Insight into hypertrophied hearts: a cardiovascular magnetic resonancestudy of papillary muscle mass and T1 mapping. Eur Heart J CardiovascImaging 2016;18:1034–1040
.
[484] Ngu PJ, Butler M, Pham A, Roberts SK, Taylor AJ. Cardiac remodellingidentified by cardiovascular magnetic resonance in patients with
hepatitis C infection and liver disease. Int J Cardiovasc Imaging
2016;32:629–636
.
[485] Ruiz-del-Arbol L, Monescillo A, Arocena C, Valer P, Gines P, Moreira V,et al. Circulatory function and hepatorenal syndrome in cirrhosis.Hepatology 2005;42:439–447
.
[486] Zhang W, Xu X, Kao R, Mele T, Kvietys P, Martin CM, et al. Cardiac
fibroblasts contribute to myocardial dysfunction in mice with sepsis:
the role of NLRP3 inflammasome activation. PLoS One 2014;9:e107639 .
[487] Gaskari SA, Liu H, D’Mello C, Kunos G, Lee SS. Blunted cardiac responseto hemorrhage in cirrhotic rats is mediated by local macrophage-released endocannabinoids. J Hepatol 2015;62:1272–1277
.
[488] Liu H, Lee SS. Nuclear factor-kappaB inhibition improves myocardialcontractility in rats with cirrhotic cardiomyopathy. Liver Int 2008;28:
640–648
.
[489] Trevisani F, Di Micoli A, Zambruni A, Biselli M, Santi V, Erroi V, et al. QTinterval prolongation by acute gastrointestinal bleeding in patientswith cirrhosis. Liver Int 2012;32:1510–1515
.
[490] Zhao J, Qi X, Hou F, Ning Z, Zhang X, Deng H, et al. Prevalence, riskfactors and in-hospital outcomes of QTc interval prolongation in liver
cirrhosis. Am J Med Sci 2016;352:285–295
.
[491] Krag A, Bendtsen F, Mortensen C, Henriksen JH, Moller S. Effects of asingle terlipressin administration on cardiac function and perfusion incirrhosis. Eur J Gastroenterol Hepatol 2010;22:1085–1092
.
[492] Wannhoff A, Hippchen T, Weiss CS, Friedrich K, Rupp C, Neumann-Haefelin C, et al. Cardiac volume overload and pulmonary hypertensionin long-term follow-up of patients with a transjugular intrahepatic
portosystemic shunt. Aliment Pharmacol Ther 2016;43:955–965
.
[493] Busk TM, Bendtsen F, Henriksen JH, Fuglsang S, Clemmesen JO, LarsenFS, et al. Effects of transjugular intrahepatic portosystemic shunt (TIPS)on blood volume distribution in patients with cirrhosis. Dig Liver Dis2017;49:1353–1359
.[494] Raevens S, De Pauw M, Geerts A, Berrevoet F, Rogiers X, Troisi RI, et al.
Prevalence and outcome of diastolic dysfunction in liver transplanta-tion recipients. Acta Cardiol 2014;69:273–280
.
[495] Saner FH, Neumann T, Canbay A, Treckmann JW, Hartmann M,Goerlinger K, et al. High brain-natriuretic peptide level predictscirrhotic cardiomyopathy in liver transplant patients. Transpl Int
2011;24:425–432
.
[496] Sampaio F, Pimenta J, Bettencourt N, Fontes-Carvalho R, Silva AP,Valente J, et al. Systolic dysfunction and diastolic dysfunction do notinfluence medium-term prognosis in patients with cirrhosis. Eur JIntern Med 2014;25:241–246
.
[497] Alexopoulou A, Papatheodoridis G, Pouriki S, Chrysohoou C, Raftopou-
los L, Stefanadis C, et al. Diastolic myocardial dysfunction does not
affect survival in patients with cirrhosis. Transpl Int 2012;25:1174–1181
.
[498] Rodriguez-Roisin R, Krowka MJ, Herve P, Fallon MB. ERS task forcepulmonary-hepatic vascular disorders (PHD) scientific committee.Pulmonary-hepatic vascular disorders (PHD). Eur Respir J 2004;24:861–880
.
[499] Machicao VI, Balakrishnan M, Fallon MB. Pulmonary complications in
chronic liver disease. Hepatology 2014;59:1627–1637 .
[500] Kaymakoglu S, Kahraman T, Kudat H, Demir K, Cakaloglu Y, Adalet I,et al. Hepatopulmonary syndrome in noncirrhotic portal hypertensivepatients. Dig Dis Sci 2003;48:556–560
.
[501] Fuhrmann V, Madl C, Mueller C, Holzinger U, Kitzberger R, Funk GC,
et al. Hepatopulmonary syndrome in patients with hypoxic hepatitis.
Gastroenterology 2006;131:69–75 .
[502] Rodríguez-Roisin R, Agustí AG, Roca J. The hepatopulmonary syndrome:new name, old complexities. Thorax 1992;47:897–902
.
[503] Fallon MB, Abrams GA. Pulmonary dysfunction in chronic liver disease.Hepatology 2000;32:859–865
.
[504] Voiosu AM, Daha IC, Voiosu TA, Mateescu BR, Dan GA, Ba ˘icus/C223CR, et al.
Prevalence and impact on survival of hepatopulmonary syndrome and
cirrhotic cardiomyopathy in a cohort of cirrhotic patients. Liver Int2015;35:2547–2555
.
[505] Rodríguez-Roisin R, Krowka MJ. Hepatopulmonary syndrome; a liver-induced lung vascular disorder. N Engl J Med 2008;358:2378–2387
.
[506] Schenk P, Schoniger-Hekele M, Fuhrmann V, Madl C, Silberhumer G,
Muller C. Prognostic significance of the hepatopulmonary syndrome in
patients with cirrhosis. Gastroenterology 2003;125:1042–1052 .
[507] Swanson KL, Wiesner RH, Krowka MJ. Natural history of hepatopul-monary syndrome: impact of liver transplantation. Hepatology2005;41:1122–1129
.
[508] Fallon MB, Krowka MJ, Brown RS, Trotter JF, Zacks S, Roberts KE, et al.Impact of hepatopulmonary syndrome on quality of life and survival in
liver transplant candidates. Gastroenterology 2008;135:1168–1175
.
[509] Schraufnagel DE, Kay JM. Structural and pathologic changes in the lungvasculature in chronic liver disease. Clin Chest Med 2006;17:1–15
.
[510] Boryczka G, Hartleb M, Rudzki K, Janik MA. Influence of an upright bodyposition on the size of intrapulmonary blood shunts in patients withadvanced liver cirrhosis. J Physiol Pharmacol 2015;66:855–861
.
[511] Fallon MB, Abrams GA, Luo B, Hou Z, Dai J, Ku DD. The role of
endothelial nitric oxide synthase in the pathogenesis of a rat model ofhepatopulmonary syndrome. Gastroenterology 1997;113:606–614
.
[512] Luo B, Liu L, Tang L, Zhang J, Ling Y, Fallon MB. ET-1 and TNF- alpha inHPS: analysis in prehepatic portal hypertension and biliary andnonbiliary cirrhosis in rats. Am J Physiol Gastrointest Liver Physiol2004;286:G294–G303
.
[513] Wiest R, Groszmann RJ. The paradox of nitric oxide in cirrhosis and
portal hypertension: too much, not enough. Hepatology 2002;35:478–491
.
[514] Ling Y, Zhang J, Luo B, Song D, Liu L, Tang L, et al. The role of endothelin-1 and the endothelin B receptor in the pathogenesis of hepatopul-monary syndrome in the rat. Hepatology 2004;39:1593–1602
.
[515] Tang L, Luo B, Patel RP, Ling Y, Zhang J, Fallon MB. Modulation of
pulmonary endothelial endothelin B receptor expression and signaling:implications for experimental hepatopulmonary syndrome. Am JPhysiol Lung Cell Mol Physiol 2007;292:L1467–L1472
.
[516] Frossard JL, Schiffer E, Cikirikcioglu B, Bourquin BJ, Morel DR, PastorMC. Opposite regulation of endothelial NO synthase by HSP90 andcaveolin in liver and lungs of rats with hepatopulmonary syndrome.
Am J Physiol Gastrointest Liver Physiol 2007;293:G864–G870
.
[517] Rabiller A, Nunes H, Lebrec D, Tazi KA, Wartski M, Dulmet E, et al.Prevention of gram-negative translocation reduces the severity ofhepatopulmonary syndrome. Am J Respir Crit Care Med 2002;166:514–517
.JOURNAL
OF HEPATOLOGY
Journal of Hepatology 2018 vol. xxx jxxx–xxx 53
Please cite this article in press as: The European Association for the Study of the Liver. EASL Clinical Practice Guidelines for the management of pati ents with decompensated cirrhosis. J
Hepatol (2018), https://doi.org/10.1016/j.jhep.2018.03.024

[518] Thenappan T, Goel A, Marsboom G, Fang YH, Toth PT, Zhang HJ, et al. A
central role for CD68(1) macrophages in hepatopulmonary syndrome:reversal by macrophage depletion. Am J Respir Crit Care Med2011;183:1080–1091
.
[519] Zhang J, Yang W, Luo B, Hu B, Maheshwari A, Fallon MB. The role of CX(3)CL1/CX(3)CR1 in pulmonary angiogenesis and intravascular mono-
cyte accumulation in rat experimental hepatopulmonary syndrome. J
Hepatol 2012;57:752–758
.
[520] Carter EP, Hartsfield CL, Miyazono M, Jakkula M, Morris Jr KG,McMurtry IF. Regulation of heme oxygenase-1 by nitric oxide duringhepatopulmonary syndrome. Am J Physiol Lung Cell Mol Physiol2002;283:L346–L353
.
[521] Zhang J, Luo B, Tang L, Wang Y, Stockard CR, Kadish I, et al. Pulmonary
angiogenesis in a rat model of hepatopulmonary syndrome. Gastroen-terology 2009;136:1070–1080
.
[522] Zhang J, Yang W, Hu B, Wu W, Fallon MB. Endothelin-1 activation of theendothelin B receptor modulates pulmonary endothelial CX3CL1 andcontributes to pulmonary angiogenesis in experimental hepatopul-monary syndrome. Am J Pathol 2014;184:1706–1714
.
[523] Zeng J, Chen L, Chen B, Lu K, Belguise K, Wang X, et al. MicroRNA-199 a-
5p Regulates the Proliferation of Pulmonary Microvascular EndothelialCells in Hepatopulmonary Syndrome. Cell Physiol Biochem2015;37:1289–1300
.
[524] Roberts KE, Kawut SM, Krowka MJ, Brown Jr RS, Trotter JF, Shah V, et al.Genetic risk factors for hepatopulmonary syndrome in patients with
advanced liver disease. Gastroenterology 2010;39:130–139
.
[525] Chang CC, Wang SS, Hsieh HG, Lee WS, Chuang CL, Hc Lin, et al.Rosuvastatin improves hepatopulmonary syndrome through inhibitionof inflammatory angiogenesis of lung. Clin Sci 2015;129:449–460
.
[526] Rodriquez-Roisin R, Krowka MJ, Herve P, Fallon MB. ERS (EuropeanRespiratory Society) Task ForcePHD Scientific Committee. Highlights ofthe ERS task force on pulmonary-hepatic vascular disorders (PHD). J
Hepatol 2005;42:924–927
.
[527] Arguedas MR, Singh H, Faulk DK, Fallon MB. Utility of pulse oximetryscreening for hepatopulmonary syndrome. Clin Gastroenterol Hepatol2007;5:749–754
.
[528] Hoerning A, Raub S, Neudorf U, Muntjes C, Kathemann S, Lainka E, et al.Pulse oximetry is insufficient for timely diagnosis of hepatopulmonary
syndrome in children with liver cirrhosis. J Pediatr 2014;164:546–552
.
[529] Horvatits T, Drolz A, Roedl K, Herkner H, Ferlitsch A, Perkmann T, et al.Von Willebrand factor antigen for detection of hepatopulmonarysyndrome in patients with cirrhosis. J Hepatol 2014;61:544–549
.
[530] Köksal D, Kaçar S, Köksal AS, Tüfekçiog ˘lu O, Küçükay F, Okten S, et al.
Evaluation of intrapulmonary vascular dilatations with high-resolutioncomputed thorax tomography in patients with hepatopulmonary
syndrome. J Clin Gastroenterol 2006;40:77–83
.
[531] Lee KN, Lee HJ, Shin WW, Webb WR. Hypoxemia and liver cirrhosis(hepatopulmonary syndrome) in eight patients: comparison of thecentral and peripheral pulmonary vasculature. Radiology 1999;211:549–553
.
[532] Abrams GA, Jaffe CC, Hoffer PB, Binder HJ, Fallon MB. Diagnostic utilityof contrast echocardiography and lung perfusion scan in patients with
hepatopulmonary syndrome. Gastroenterology 1995;109:
1283–1288
.
[533] Wolfe JD, Tashkin DP, Holly FE, Brachman MB, Genovesi MG. Hypox-emia of cirrhosis: detection of abnormal small pulmonary vascularchannels by a quantitative radionuclide method. Am J Med1977;63:746–754
.
[534] Abrams GA, Nanda NC, Dubovsky EV, Krowka MJ, Fallon MB. Use of
macroaggregated albumin lung perfusion scan to diagnose hepatopul-
monary syndrome: a new approach. Gastroenterology 1998;114:305–310
.
[535] Arguedas MR, Abrams GA, Krowka MJ, Fallon MB. Prospective evalu-ation of outcomes and predictors of mortality in patients with hepato-pulmonary syndrome undergoing liver transplantation. Hepatology
2003;37:192–197
.
[536] Kalambokis G, Tsianos EV. Pitfalls in the assessment of intrapulmonaryshunt using lung perfusion scintigraphy in patients with cirrhosis. LiverInt 2010;31:138–139
.
[537] Kochar R, Tanikella R, Fallon MB. Serial pulse oximetry in hepatopul-monary syndrome. Dig Dis Sci 2011;56:1862–1868
.
[538] Tanikella R, Philips GM, Faulk DK, Kawut SM, Fallon MB. Pilot study of
pentoxifylline in hepatopulmonary syndrome. Liver Transpl 2008;14:
1199–1203 .
[539] Gupta LB, Kumar A, Jaiswal AK, Yusuf J, Metha V, Tyagi S, et al.Pentoxyfylline therapy for hepatopulmonary syndrome: a piloto study.Arch Intern Med 2008;168:1820–1823
.[540] De BK, Dutta D, Pal SK, Gangopadhyay S, Das Baksi S, Pani A. The role of
garlic in hepatopulmonary syndrome: a randomized controlled trial.Can J Gastroenterol 2010;24:183–188
.
[541] Shaikh SA, Tischer S, Choi EK, Fontana RJ. Good for the lung but bad forthe liver? Garlic-induced hepatotoxicity following liver transplanta-tion. J Clin Pharm Ther 2017;42:646–648
.
[542] Tsauo J, Weng N, Ma H, Jiang M, Zhao H, Li X. Role of transjugular
intrahepatic portosystemic shunts in the management of hepatopul-monary syndrome: a systemic literature review. J Vasc Interv Radiol2015;26:1266–1271
.
[543] Poterucha JJ, Krowka MJ, Dickson ER, Cortese DA, Stanson AW, KromRA. Failure of hepatopulmonary syndrome to resolve after liver
transplantation and successful treatment with embolotherapy. Hepa-
tology 1995;21:96–100
.
[544] Krowka MJ, Mandell MS, Ramsay MA, Kawut SM, Fallon MB, Man-zarbeitia C, et al. Hepatopulmonary syndrome and portopulmonaryhypertension: a report of the multicenter liver transplant database.Liver Transpl 2004;10:174–182
.
[545] Gupta S, Castel H, Rao RV, Picard M, Lilly L, Faughnan ME, et al.
Improved survival after liver transplantation in patients with hep-
atopulmonary syndrome. Am J Transplant 2010;10:354–363 .
[546] Fallon MB, Mulligan DC, Gish RG, Krowka MJ. Model for end-stage liverdisease (MELD) exception for hepatopulmonary syndrome. LiverTranspl 2006;12:S105–S107
.
[547] Iyer VN, Swanson KL, Cartin-Ceba R, Dierkhising RA, Rosen CB,
Heimbach JK, et al. Hepatopulmonary syndrome: favorable outcomes
in the MELD exception era. Hepatology 2013;57:427–2435 .
[548] Goldberg DS, Krok K, Batra S, Trotter JF, Kawut SM, Fallon MB. Impact ofthe hepatopulmonary syndrome MELD exception policy on outcomesof patients after liver transplantation: an analysis of the UNOSdatabase. Gastroenterology 2014;146:1256–1265
.
[549] Fleming GM, Cornell TT, Welling TH, Magee JC, Annich GM. Hepatopul-
monary syndrome use of extracorporeal life support for life-threatening
hypoxia following liver transplantation. Liver Transpl 2008;14:966–970 .
[550] Monsel A, Mal H, Brisson H, Luo R, Eyraud D, Vénizet C, et al.Extracorporeal membrane oxygenation as a brifge to liver trnsplanta-tion for acute respiratory distress syndrome-induced life-threateninghypoxemia aggravated by hepatopulmonary syndrome. Crit Care
2011;15:R234
.
[551] Chihara Y, Egawa H, Tsuboi T, Oga T, Handa T, Yamamoto K, et al.Immediate nonivasive ventilation may improve mortality in patientswith hepatopulmonary syndrome after liver transplantation. LiverTranspl 2011;17:44–148
.
[552] McLaughlin VV, Archer SL, Badesch DB, Barst RJ, Farber HW, Lindner JR,et al. ACCF/AHA 2009 expert consensus document on pulmonary
hypertension a report of the American College of Cardiology Founda-
tion Task Force on Expert Consensus Documents and the AmericanHeart Association developed in collaboration with the AmericanCollege of Chest Physicians; American Thoracic Society, Inc.; and thePulmonary Hypertension Association. J Am Coll Cardiol 2009;53:1573–1619
.
[553] Torregrosa M, Genesca J, Gonzalez A, Evangelista A, Mora A, Margarit C,
et al. Role of Doppler echocardiography in the assessment of portopul-monary hypertension in liver transplantation candidates. Transplanta-tion 2001;71:572–574
.
[554] Raevens S, Colle I, Reyntjens K, Geerts A, Berrevoet F, Rogiers X, et al.Echocardiography for the detection of portopulmonary hypertension inliver transplant candidates: an analysis of cutoff values. Liver Transpl
2013;19:602–610
.
[555] Cotton CL, Gandhi S, Vaitkus PT, Massad MG, Benedetti E, Mrtek RG,et al. Role of echocardiography in detecting portopulmonary hyper-tension in liver transplant candidates. Liver Transpl 2002;8:1051–1054
.
[556] Kawut SM, Krowka MJ, Trotter JF, Roberts KE, Benza RL, Badesch DB,et al. Clinical risk factors for portopulmonary hypertension. Hepatology
2008;48:196–203
.
[557] Paulus JK, Roberts KE. Oestrogen and sexual Dimorphism of pulmonaryarterial hypertension; a transitional challenge. Eur Respir J 2013;41:1014–1016
.
[558] Talwalkar JA, Swanson KL, Krowka MJ, Andrews JC, Kamath PS.Prevalence of spontaneous portosystemic shunts in patients withportopulmonary hypertension and effect on treatment. Gastroenterol-
ogy 2011;141:1673–1679
.
[559] Hadengue A, Benhayoun MK, Lebrec D, Benhamou JP. Pulmonaryhypertension complicating portal hypertension: prevalence and rela-tion to splanchnic hemodynamics. Gastroenterology 1991;100:520–528
.Clinical Practice Guidelines
54 Journal of Hepatology 2018 vol. xxx jxxx–xxx
Please cite this article in press as: The European Association for the Study of the Liver. EASL Clinical Practice Guidelines for the management of pati ents with decompensated cirrhosis. J
Hepatol (2018), https://doi.org/10.1016/j.jhep.2018.03.024

[560] Robalino BD, Moodie DS. Association between primary pulmonary
hypertension and portal hypertension: analysis of its pathophysiologyand clinical, laboratory and hemodynamic manifestations. J Am CollCardiol 1991;17:492–498
.
[561] Swanson KL, Wiesner RH, Nyberg SL, Rosen CB, Krowka MJ. Survival inportopulmonary hypertension: Mayo Clinic experience categorized by
treatment subgroups. Am J Transplant 2008;8:2445–2453
.
[562] Cartin-Ceba R, Krowka MJ. Portopulmonary hypertension. Clin Liver Dis2014;18:421–438
.
[563] Condino AA, Ivy DD, O’Connor JA, Narkewicz MR, Mengshol S,Whitworth JR, et al. Portopulmonary hypertension in pediatric patients.J Pediatr 2005;147:20–26
.
[564] Le Pavec J, Souza R, Herve P, Lebrec D, Savale L, Tcherakian C, et al.
Portopulmonary hypertension: survival and prognostic factors. Am JRespir Crit Care Med 2008;178:637–643
.
[565] Krowka MJ, Cartin-Ceba R. Portopulmonary hypertension: formidabledual threat vs.hopeful dual therapy. Liver Transpl 2014;20:635–636 .
[566] Kawut SM, Horn EM, Berekashvili KK, Garofano RP, Goldsmith RL,
Widlitz AC, et al. New predictors of outcome in idiopathic pulmonary
arterial hypertension. Am J Cardiol 2005;95:199–203
.
[567] Ricci GL, Melgosa MT, Burgos F, Valera JL, Pizarro S, Roca J, et al.Assessment of acute pulmonary vascular reactivity in portopulmonaryhypertension. Liver Transpl 2007;13:1506–1514
.
[568] Provencher S, Herve P, Jais X, Lebrec D, Humbert M, Simonneau G, et al.Deleterious effects of beta-blockers on exercise capacity and hemody-
namics in patients with portopulmonary hypertension. Gastroenterol-
ogy 2006;130:120–126
.
[569] Halank M, Miehlke S, Hoeffken G, Schmeisser A, Schulze M, Strasser RH.Use of oral endothelin-receptor antagonist bosentan in the treatment ofportopulmonary hypertension. Transplantation 2004;77:1775–1776
.
[570] Hoeper MM, Halank M, Marx C, Hoeffken G, Seyfarth HJ, Schauer J, et al.Bosentan therapy for portopulmonary hypertension. Eur Respir J
2005;25:502–508
.
[571] Savale L, Magnier R, Le Pavec J, Jais X, Montani D, O’Callaghan DS, et al.Efficacy, safety and pharmacokinetics of bosentan in portopulmonaryhypertension. Eur Respir J 2013;41:96–103
.
[572] Barth F, Gerber PJ, Reichen J, Dufour JF, Nicod LP. Efficiency andsafety of bosentan in child C cirrhosis with portopulmonary
hypertension and renal insufficiency. Eur J Gastroenterol Hepatol
2006;18:1117–1119
.
[573] Hoeper MM, Seyfarth HJ, Hoeffken G, Wirtz H, Spiekerkoetter E, PletzMW, et al. Experience with inhaled iloprost and bosentan in portopul-monary hypertension. Eur Respir J 2007;30:1096–1102
.
[574] Cartin-Ceba R, Swanson K, Iyer V, Wiesner RH, Krowka MJ. Safety andefficacy of ambrisentan for the treatment of portopulmonary hyper-
tension. Chest 2011;139:109–114
.
[575] DuBrock HM, Channick RN, Krowka MJ. What’s new in the treatment ofportopulmonary hypertension? Expert Rev Gastroenterol Hepatol2015;9:983–992
.
[576] Reichenberger F, Voswinckel R, Steveling E, Enke B, Kreckel A,Olschewski H, et al. Sildenafil treatment for portopulmonary hyper-
tension. Eur Respir J 2006;28:563–567
.
[577] Cadden IS, Greanya ED, Erb SR, Scudamore CH, Yoshida EM. The use ofsildenafil to treat portopulmonary hypertension prior to liver trans-plantation. Ann Hepatol 2009;8:158–161
.
[578] Fisher JH, Johnson SR, Chau C, Kron AT, Granton JT. Effectiveness ofphosphodiesterase-5 inhibitor therapy for portopulmonary hyperten-sion. Can Respir J 2015;22:42–46
.
[579] Tzathas C, Christidou A, Ladas SD. Sildenafil (viagra) is a risk factor for
acute variceal bleeding. Am J Gastroenterol 2002;97:1856 .
[580] Kuo PC, Johnson LB, Plotkin JS, Howell CD, Bartlett ST, Rubin LJ.Continuous intravenous infusion of epoprostenol for the treat-ment of portopulmonary hypertension. Transplantation 1997;63:604–606
.
[581] Sussman N, Kaza V, Barshes N, Stribling R, Goss J, O’Mahony C, et al.
Successful liver transplantation following medical management ofportopulmonary hypertension: a single-center series. Am J Transplant2006;6:2177–2182
.
[582] Fix OK, Bass NM, De Marco T, Merriman RB. Long-term follow-up ofportopulmonary hypertension: effect of treatment with epoprostenol.Liver Transpl 2007;13:875–885
.
[583] Awdish RL, Cajigas HR. Early initiation of prostacyclin in portopul-
monary hypertension: 10 years of a transplant center’s experience.Lung 2013;191:593–600
.[584] Melgosa MT, Ricci GL, Garcia-Pagan JC, Blanco I, Escribano P, Abraldes
JG, et al. Acute and long-term effects of inhaled iloprost in portopul-monary hypertension. Liver Transpl 2010;16:348–356
.
[585] Colombato LA, Spahr L, Martinet JP, Dufresne MP, Lafortune M, FenyvesD, et al. Haemodynamic adaptation two months after transjugularintrahepatic portosystemic shunt (TIPS) in cirrhotic patients. Gut
1996;39:600–604
.
[586] Boyer TD, Haskal ZJ. American Association for the Study of Liver D. TheRole of Transjugular Intrahepatic Portosystemic Shunt (TIPS) in theManagement of Portal Hypertension: update 2009. Hepatology2010;51:306
.
[587] Yoshida EM, Erb SR, Pflugfelder PW, Ostrow DN, Ricci DR, Ghent CN,
et al. Single-lung vs.liver transplantation for the treatment of
portopulmonary hypertension–a comparison of two patients. Trans-plantation 1993;55:688–690
.
[588] Krowka MJ, Wiesner RH, Heimbach JK. Pulmonary contraindications,indications and MELD exceptions for liver transplantation: a contem-porary view and look forward. J Hepatol 2013;59:367–374
.
[589] Kuo PC, Plotkin JS, Gaine S, Schroeder RA, Rustgi VK, Rubin LJ, et al.
Portopulmonary hypertension and the liver transplant candidate.
Transplantation 1999;67:1087–1093 .
[590] Krowka MJ, Plevak DJ, Findlay JY, Rosen CB, Wiesner RH, Krom RA.Pulmonary hemodynamics and perioperative cardiopulmonary-relatedmortality in patients with portopulmonary hypertension undergoingliver transplantation. Liver Transpl 2000;6:443–450
.
[591] Mangus RS, Kinsella SB, Marshall GR, Fridell JA, Wilkes KR, Tector AJ.
Mild to moderate pulmonary hypertension in liver transplantation. JSurg Res 2013;184:1150–1156
.
[592] Raevens S, De Pauw M, Reyntjens K, Geerts A, Verhelst X, Berrevoet F,et al. Oral vasodilator therapy in patients with moderate to severeportopulmonary hypertension as a bridge to liver transplantation. Eur JGastroenterol Hepatol 2013;25:495–502
.
[593] Ramsay MA, Simpson BR, Nguyen AT, Ramsay KJ, East C, Klintmalm GB.
Severe pulmonary hypertension in liver transplant candidates. LiverTranspl Surg 1997;3:494–500
.
[594] Krowka MJ, Fallon MB, Mulligan DC, Gish RG. Model for end-stage liverdisease (MELD) exception for portopulmonary hypertension. LiverTranspl 2006;12:S114–S116
.
[595] DuBrock HM, Goldberg DS, Sussman NL, Bartolome SD, Kadry Z, Salgia
RJ, et al. Predictors of waitlist mortality in portopulmonary hyperten-sion. Transplantation 2017;101:1609–1615
.
[596] Ramsay M. Portopulmonary hypertension and right heart failure inpatients with cirrhosis. Curr Opin Anaesthesiol 2010;23:145–150
.
[597] Paulsen AW, Whitten CW, Ramsay MA, Klintmalm GB. Considerationsfor anesthetic management during veno-venous bypass in adult
hepatic transplantation. Anesth Analg 1989;68:489–496
.
[598] Ramsay MA, Spikes C, East CA, Lynch K, Hein HA, Ramsay KJ, et al. Theperioperative management of portopulmonary hypertension withnitric oxide and epoprostenol. Anesthesiology 1999;90:299–301
.
[599] Stratta C, Lavezzo B, Ballaris MA, Panio A, Crucitti M, Andruetto P, et al.Extracorporeal membrane oxygenation rescue therapy in a case of
portopulmonary hypertension during liver transplantation: a case
report. Transplant Proc 2013;45:2774–2775
.
[600] Ashfaq M, Chinnakotla S, Rogers L, Ausloos K, Saadeh S, Klintmalm GB,et al. The impact of treatment of portopulmonary hypertension onsurvival following liver transplantation. Am J Transplant2007;7:1258–1264
.
[601] Hollatz TJ, Musat A, Westphal S, Decker C, D’Alessandro AM, Keevil J,
et al. Treatment with sildenafil and treprostinil allows successful liver
transplantation of patients with moderate to severe portopulmonaryhypertension. Liver Transpl 2012;18:686–695
.
[602] Khaderi S, Khan R, Safdar Z, Stribling R, Vierling JM, Goss JA, et al. Long-term follow-up of portopulmonary hypertension patients after livertransplantation. Liver Transpl 2014;20:724–727
.
[603] Goldberg DS, Batra S, Sahay S, Kawut SM, Fallon MB. MELD exceptions
for portopulmonary hypertension: current policy and future imple-mentation. Am J Transplant 2014;14:2081–2087
.
[604] Morando F, Maresio G, Piano S, Fasolato S, Cavallin M, Romano A, et al.How to improve care in outpatients with cirrhosis and ascites: a newmodel of care coordination by consultant hepatologists. J Hepatol2013;59:257–264
.JOURNAL
OF HEPATOLOGY
Journal of Hepatology 2018 vol. xxx jxxx–xxx 55
Please cite this article in press as: The European Association for the Study of the Liver. EASL Clinical Practice Guidelines for the management of pati ents with decompensated cirrhosis. J
Hepatol (2018), https://doi.org/10.1016/j.jhep.2018.03.024