You are on page 1of 13


Non-invasive evaluation of portal hypertension using

ultrasound elastography

Annalisa Berzigotti⇑

Keywords: Advanced chronic liver
disease; Cirrhosis; Non-invasive;
Portal hypertension (PH) leads to serious complications, such as bleeding from gastroe- Varices; Decompensation.
sophageal varices, ascites and portosystemic encephalopathy in patients with chronic liver Received 20 December 2016;
disease (CLD). Gold standard methods for assessing PH and its complications include the mea- received in revised form 4 February
surement of hepatic venous pressure gradient and endoscopy; however, these are invasive, 2017; accepted 6 February 2017
expensive and not available at all centres. Therefore, non-invasive alternatives have been
the subject of extensive investigation over the last 20 years. The present review focuses on
the role of ultrasound elastography - a novel group of non-invasive techniques used to mea-
sure stiffness in target organs. In the context of CLD these methods are used to identify the
presence of PH, its severity, and the risk of PH-related complications. The rationale,
accumulated evidence, advantages and limitations of liver and spleen stiffness measurements
evaluated by different ultrasound elastography techniques in patients with advanced CLD is
discussed. Recent data regarding the use of ultrasound elastography techniques in patients
with non-cirrhotic forms of PH are also described.
Ó 2017 European Association for the Study of the Liver. Published by Elsevier B.V. All rights


Portal hypertension (PH) is a common clinical syn- obstruction, thrombosis of the hepatic veins Swiss Liver Center, Hepatology,
drome. It is haemodynamically defined by an (Budd-Chiari syndrome, BCS) and idiopathic PH.2 University Clinic for Visceral
Surgery and Medicine, Inselspital,
increase in the venous pressure gradient across Once PH passes a critical threshold of 10 mmHg,
University of Bern, Switzerland
the liver, calculated from its inflow through the extrahepatic vascular changes, driven by angio-
portal vein versus its outflow through the hepatic genesis, lead to portosystemic collateral develop-
veins.1 ment and splanchnic vasodilation, which in
Like any other vascular system, portal pressure turn contribute to further increases in portal
is the product of two independent factors: resis- pressure.3
tance to blood flow and the amount of flow, as sta- PH can remain asymptomatic for many years but
ted by Ohm’s law: imaging and laboratory testing may suggest its pres-
ence. Splenomegaly is a very common consequence
Pressure ¼ Resistance  Flow of PH; it usually leads to thrombocytopenia due to

hypersplenism,4 and is often the first manifestation
An increase in resistance to portal blood flow is to infer the presence of PH.
the initial factor that leads to a rise in portal pres- From the clinical point of view, PH is relevant
sure. This resistance can be located at any point because of its severe complications, which include,
in the liver circulation, i.e. at the prehepatic, intra- upper digestive bleeding caused by gastroe-
⇑ Corresponding author. Address:
hepatic or post-hepatic level. In the Western world, sophageal varices, ascites, spontaneous bacterial
Swiss Liver Center, Hepatology,
ca. 90% of cases of PH are due to advanced chronic peritonitis and hepatorenal syndrome, and hepatic University Clinic for Visceral
liver disease (ACLD) or cirrhosis, which cause struc- encephalopathy.5 The risk of developing clinical Surgery and Medicine, Inselspital,
tural damage through fibrogenesis, parenchymal complications can be effectively reduced using University of Bern, MEM F807,
extinction and regeneration. PH then develops at appropriate non-pharmacological measures and Murtenstrasse 35, CH - 3010
Berne, Switzerland. Tel.: +41 31
the intrahepatic sinusoidal site.1 medical therapy to lower PH,6,7 providing strong
632 87 27; fax: +41 31 632 49 97.
Other less common causes include vascular liver rationale for the early identification of at-risk E-mail address: annalisa.berzigotti@
diseases, such as extrahepatic portal vein patients.

Journal of Hepatology 2017 vol. 67 j 399–411

Portal hypertension in patients with chronic liver 6–9 mmHg; and an HVPG P10 mmHg represents
Key point disease the ‘clinically significant’ PH threshold.4 When
Chronic liver disease
HVPG reaches 10 mmHg or above, PH can become
accounts for 90% of cases of In chronic liver disease (CLD), the liver progres- symptomatic as patients can develop gastroe-
portal hypertension in sively accumulates extracellular matrix (fibrogene- sophageal varices and hyperdynamic circulation,
Western countries. sis); in ACLD,8,9 fibrosis is accompanied by a more increasing their risk of clinical decompensation.
complex derangement of intrahepatic architecture. Therefore, patients with compensated ACLD should
This includes vascular remodelling (capillarisation be screened for the presence of CSPH,7 and patients
of the liver sinusoids and neo-angiogenesis), forma- with CSPH constitute the population that should be
tion of intrahepatic shunts and an increase in hep- screened for the presence of varices at risk of bleed-
atic vascular tone (active contraction of activated ing. Severe PH (HVPG P12 mmHg), and very severe
hepatic stellate cells, myofibroblasts and smooth PH (HVPG P16 mmHg), are associated with an
muscle cells). There are two distinct phases of increased risk of variceal bleeding and mortality,
ACLD, characterised by different prognoses (‘com- respectively (Table 1).1
pensated’ and ‘decompensated’ phase). The ‘com- Clinical decompensation is defined by the onset
pensated’, asymptomatic stage, precedes the onset of ascites, variceal bleeding, jaundice or hepatic
of clinical events by years, and patients’ prognosis encephalopathy. It marks the symptomatic phase
is only slightly worse than the general population.7 of cirrhosis, which is characterised by a much higher
In the past, diagnosis of compensated ACLD was mortality rate. At this stage, 100% of patients have
difficult because it was exclusively based on liver PH and, thus, in patients who already developed
biopsy. Now, the availability of non-invasive meth- decompensation, PH does not need to be diag-
ods (ultrasound elastography in particular) has nosed.7 However, patients with decompensated
improved the clinicians’ identification of asymp- ACLD are far more likely to have varices that require
tomatic ACLD patients, who could further progress therapy, so endoscopy is always needed in this
to clinically significant PH (CSPH) and its subse- population.7
quent complications (decompensated phase).7,10 Upper gastrointestinal endoscopy is the best
The development of PH drives the progression of method to determine the presence of oesophageal
ACLD to the onset of complications.7 and gastric varices, and allows the identification of
additional signs used to stratify bleeding risk (size
Portal hypertension in patients with cirrhosis: of varices; presence of red colour signs and wale
Invasive diagnostic methods marks).11 Cross-sectional imaging techniques, such
as contrast-enhanced computed tomography (CT)
Sinusoidal PH in patients with cirrhosis can be reli- scanning and contrast-enhanced magnetic reso-
ably and safely evaluated by measuring the hepatic nance imaging (MRI) can be used to visualise large
venous pressure gradient (HVPG) via hepatic vein gastroesophageal collaterals but cannot replace
catheterization. However, HVPG measurement endoscopy because of their limited sensitivity.
could underestimate the severity of PH in patients Whenever large varices or varices with red colour
with primary biliary cholangitis as they may have signs are diagnosed, primary prophylaxis of bleed-
a pre-sinusoidal component (especially in early ing should be initiated using non-selective beta-
stages).4 blockers or endoscopic therapy (band ligation).7,12
An HVPG of up to 5 mmHg is considered In patients with decompensated Child-Pugh grade
normal; subclinical PH is defined by an HVPG of C cirrhosis, even small varices can bleed, and pri-
mary prophylaxis should be initiated promptly.7,12

Table 1. Hepatic venous pressure gradient thresholds associated with clinical risks in chronic liver A need for non-invasive methods to assess portal
disease. hypertension in patients with chronic liver disease
Hemodynamic Increase in the risk of the following events:
threshold HVPG measurement and endoscopy are the back-
Compensated 10 mmHg Presence and development of gastroesophageal bone for the assessment of PH in CLD.7,12 However,
ACLD varices;74,75 first clinical decompensation in patients they are invasive and may (in rare cases) lead to
with no varices;76 postoperative decompensation in complications; in addition, a specialised clinical set-
patients undergoing resection for HCC;77,78 development
ting and specific expertise are required to carry out
of HCC79
these tests, limiting their availability and increasing
12 mmHg Bleeding from varices74,80
the cost to health care systems. The development of
16 mmHg First clinical decompensation in patients with varices;81
mortality82 simple, non-invasive methods enabling accurate and
Decompensated 16 mmHg Rebleeding and mortality83 rapid diagnosis of patients with a low risk of both
cirrhosis 20 mmHg Failure to control variceal bleeding in bleeders;84 CSPH and varices requiring treatment (who could
mortality85 avoid invasive tests), and patients with CSPH (at
22 mmHg Mortality in patients with alcoholic cirrhosis and AAH86 high risk of complications and varices, requiring
30 mmHg Spontaneous bacterial peritonitis87 further testing), would further the advancement of
AAH, acute alcoholic hepatitis; ACLD, advanced chronic liver disease; HCC, hepatocellular carcinoma. personalised medicine in this field. Patients could

400 Journal of Hepatology 2017 vol. 67 j 399–411

then be ascribed to different categories of risk and quality criteria for the correct interpretation of
ideally to different tests algorithms. Such a results are not yet completely defined.
‘‘point-of-care” non-invasive sphygmomanometer
in patients with CLD is one step closer, thanks to Liver stiffness measurement for portal hypertension:
the development of ultrasound elastography. Rationale and accuracy Key point

Liver stiffness measurement

Ultrasound elastography techniques in a nutshell Liver stiffness measurement (LSM) accurately (LSM) improves the non-
reflects liver fibrosis in CLD.17 In patients with invasive risk stratification
Elastography techniques are all based on a com- CLD, fibrosis is the major component of increased of patients with compen-
sated advanced chronic liver
mon principle: all tissues have intrinsic mechani- intrahepatic vascular resistance leading to PH (as
disease; more than 90% of
cal/elastic properties that can be measured by discussed above). Therefore, LSM has been studied patients with an LSM [20-
creating a distortion in the tissue and evaluating as a possible surrogate for PH (Fig. 1). The first data 25 kPa (evaluated by tran-
its response. There are several ways of creating reporting the close relationship between LSM and sient elastography) will have
the distortion, detecting and then analysing the HVPG in patients with CLD were obtained using TE clinically significant portal
response of the tissue. The term ‘‘ultrasound elas- in patients with post-orthotopic liver transplanta- hypertension.
tography” groups the techniques using ultrasound tion recurrence of hepatitis C infection, and was
to detect the velocity of the microdisplacements published about 10 years ago;18 since then, over
(shear waves) induced in the tissue. Comprehen- 20 studies comparing LSM (using different elastog-
sive technical details can be found elsewhere.13,14 raphy techniques) and HVPG in patients with cirrho-
The use of elastography to identify liver disease, sis have been published. These are summarised in
depends on the detection of changes in the Table 2.
mechanical properties of the liver, which occur In a study targeting patients with hepatitis C
when the structure of its tissue is modified, e.g. virus (HCV)-related ACLD, Vizzutti et al. observed a
when fibrosis accumulates. close correlation between LSM and an HVPG of up
Briefly, the first method that has been tested to 10–12 mmHg. Above this haemodynamic thresh-
and validated for the assessment of liver fibrosis old, however, the strength of the correlation Key point
is vibration-controlled transient elastography (TE, between the two parameters decreased markedly.19 Transient elastography is the
FibroScanÒ), which is a stand-alone machine. A This is in line with the concept that once CSPH most validated ultrasound
specific probe applied to the skin at a right inter- develops, the severity of PH partially depends on elastography technique;
costal space produces a vibration that is transmit- the increase in porto-systemic flow (splanchnic most data and proposed
cut-offs refer and apply to
ted to the liver. The measurement of the shear vasodilatation, hyperdynamic circulation, and por-
it. Other methods utilising
wave displacement is provided by an ultrasound tosystemic collaterals) that cannot be measured by high-end ultrasound devices
beam on the tip of the probe and is reported in kilo LSM. Interestingly in patients with severe PH, reduc- are now available and are
Pascals (kPa). The operator has very limited control ing the blood flow-dependent component of PH using undergoing validation.
over the area of interest (monodimensional chronic non-selective beta-blocker (NSBB) therapy,
view).15 Newer methods that are built-into high strengthens the correlation between LSM and HVPG.20
end ultrasound devices, focus high-intensity Even though LSM cannot be used to estimate
short-duration acoustic pulses to generate the tis- reliably an exact HVPG value (correlation (r) range:
sue displacement, either at one point (point shear 0.59–0.70 in published series), it enables accurate
wave elastography, pSWE), or in larger, distinct discrimination between patients with and without
portions of the insonated area (two-dimensional CSPH;19,21 in the published data, the area under
shear wave elastography, 2D-SWE).13,14,16 All these the receiver operating characteristic curve (AUROC)
techniques allow the real-time visualisation of the ranged between 0.82–0.94.22 A recent meta-analysis
region of interest, where elasticity can be mea- verified these data.23 In the 11 studies included, the

sured, enabling both a semi-quantitative assess- hierarchical summary AUROC for CSPH discrimina-
ment of elasticity by a colour-coding, and a tion was 0.90, with sensitivity and specificity above
quantitative measurement expressed either in m/s 85% (sensitivity: 87.5%; 95% confidence interval [CI]:
or in kPa. Of all the pSWE techniques, Virtual Touch 75.8–93.9%; specificity: 85.3;%; 95% CI: 76.9–90.9%),
Quantification (VTQ) by acoustic radiation force and the summary HVPG-LSM correlation coefficient
impulse (ARFI) imaging (Siemens, Germany) is the was 0.783 (95% CI: 0.737–0.823). A cut-off threshold
most validated for liver fibrosis, and of 2D tech- of 13.6 kPa has a high sensitivity (over 90%),19,21
niqes SWE techniques, supersonic shear wave elas- whereas a threshold of 21 kPa has a high specificity
tography (SSI; Aixplorer, Supersonic Imagine, (over 90%) and can be used to confirm the presence
France) is close to full validation.13,14,16 Virtually of CSPH.21 This cut-off maintains a high specificity
all ultrasound device producers have now made value in patients with potentially resectable hepato-
either pSWE or 2D-SWE available, each with differ- cellular carcinoma, requiring risk stratification prior
ent software characteristics. Even though, similar to treatment allocation.24 Given the amount and
to TE, all ultrasound elastography techniques can quality of the existing data, the Baveno VI consensus
be considered point-of-care methods, the optimal on PH agreed that values of LSM by TE [20–25 kPa
use of newer techniques requires at least a basic can be used to identify CSPH12 (Fig. 2). However,
knowledge of ultrasound imaging principles, and most of the patients included in the studies leading

Journal of Hepatology 2017 vol. 67 j 399–411 401

LSM increases as fibrosis
accumulates in the liver and
indirectly reflects PH in patients
with ACLD

Hepatic vascular R
is mostly due SSM directly reflects PH
to liver fibrosis


Stiffness is due to congestion

and other changes related
directly to PH
independent of its cause

Fig. 1. Rationale for the use of liver and spleen stiffness measurements for the assessment of portal hypertension. Spleen
stiffness measurement could reflect PH more accurately, irrespective of its cause. ACLD, advanced chronic liver disease; LSM, liver
stiffness measurement; PH, portal hypertension; R, resistance. *P = R * Q: Pressure = Resistance * Flow.

to the identification of these cut-off values had a whereas in the other, no significant difference
viral aetiology, so more data regarding other dis- between the techniques was observed, once quality
ease causes are needed. The use of cut-off values criteria were applied to both methods.32
allows simple and straightforward risk stratifica-
tion in patients with compensated cirrhosis. Liver stiffness measurement prediction of
Key point
However, LSM provides a numerical, continuous gastroesophageal varices and combination with
In patients with compen- value, and therefore, the use of thresholds may lead unrelated methods
sated advanced chronic liver to the loss of relevant information (e.g. a value of
disease of non-cholestatic 50 kPa holds a much higher risk compared to a The diagnostic accuracy of LSM to predict the pres-
aetiology, endoscopy can be
value of 21 kPa). Fine-tuning of an individual’s risk ence and size of varices has been the subject of more
safely avoided by using liver
stiffness measurement of a specific endpoint (such as CSPH) without loss than 50 studies. LSM values are higher in patients
(LSM) and platelet count in of information, may be achieved by calibrating with esophageal varices (EV), and tend to be higher
combination. If such a the model probability, and using nomograms based in patients with large EV; overall, this parameter is
patient has an LSM of on this mathematical calculation.25 the most accurate single non-invasive predictor in
\20 kPa and a platelet count The precision of LSM for the diagnosis of CSPH this field. However, LSM is less accurate for the pre-
[150 g/L, it is very unlikely
can be improved further by combining other diction of EV than for CSPH.35 In a systematic review
(\5% risk) that on endo-
scopy he/she will have parameters associated with PH, such as platelet and meta-analysis including 18 studies with 3,644

varices needing treatment. count and spleen size.26 subjects, summary AUROC was 0.84 for EV and
pSWE (ARFI technology, VTQ, Siemens, Ger- 0.78 for large EV, with summary sensitivity and
many) has been used in three studies to date, com- specificity of 0.87 (95% CI: 0.80–0.92) and 0.53
paring it to haemodynamic measurements,27–29 (95% CI: 0.36–0.69) for EV; 0.86 (95% CI:
and describing an applicability close to 100%. The 0.71–0.94) and 0.59 (95% CI: 0.45–0.72) for large
discriminative accuracy for CSPH was very good EV. Overall, the probability of correctly diagnosing
(AUROC 0.82–0.90). EV or large EV following a positive measurement
2D-SWE (specifically using the software pro- did not exceed 70%.35
vided in Aixplorer, Supersonic Imagine, France) Further studies were then carried out to evaluate
was the subject of four studies comparing LSM to whether the combination of LSM with unrelated
HVPG.30–34 The accuracy of discrimination between parameters associated with portal pressure, such
the presence and absence of CSPH of this method as platelet count and spleen size, could improve
(AUROC 0.82–0.90) was similar to the results the probability of correctly diagnosing EV. In the ini-
obtained using pSWE and TE. Two head-to-head tial study, which assessed patients with HBV-related
studies comparing 2D-SWE with TE30,32 for the CLD, LSM, platelet count and spleen size were
diagnosis of CSPH showed inconsistent results; in combined to form the LSM-spleen diameter to
one case 2D-SWE was more accurate than TE,30 platelet ratio score (LSPS),35 which had a higher

402 Journal of Hepatology 2017 vol. 67 j 399–411

Table 2. Accuracy of liver stiffness measurement using ultrasound elastography techniques for the diagnosis of clinically significant portal hypertension.

Study Year Design and Population Correlation AUROC for CSPH Cut-off for CSPH Sensitivity Specificity
method used coefficient (kPa)
between LSM and
Carrion 2006 Prospective, 129 OLT recipients with HCV- 0.840 n.a.0.930 n.a. 100% 60.8%
et al.18 TE recurrence for P6 mmHg 8.74 kPa for
HVPG [6 mmHg
Vizzutti 2007 Retrospective, 61 patients with HCV ACLD 0.781 0.990 13.6 kPa 97% 92%
et al.19 TE
Bureau 2008 Prospective, 144 patients with cirrhosis 0.858 0.945 21 kPa 89.9% 93.2%
et al.21 TE (alcohol and HCV) and varices
Lemoine 2008 Retrospective, 92 compensated patients with 0.728 0.840 all n.a. n.a. n.a.
et al.88 TE alcoholic or HCV cirrhosis 0.940 alcohol 34.9 kPa 90% 88%
Sanchez- 2011 Prospective, 38 patients with HCV 0.678 0.800 14 kPa 92.8% 50%
Conde TE cirrhosis and HIV coinfection 0.800
et al.89 for [12 mmHg
Colecchia 2012 Prospective, 100 HCV cirrhosis 0.836 0.920 24.2 kPa 52.3% 97.1%
et al.52 TE
Llop 2012 Prospective, 79 Child-Pugh A patients 0.552 0.840 Rule-out: 13.6 kPa 91% 57%
et al.24 TE (mostly viral) with potentially Rule-in: 21 kPa
resectable HCC 58% 91%
Reiberger 2012 Retrospective, 502 patients, mixed 0.794 0.871 18 kPa 82.2% 83.4%
et al.90 TE etiologies, some not
cirrhotics, some
Hong 2013 Retrospective, 59 patients with cirrhosis 0.704 0.851 21.95 kPa 82.5% 73.7%
et al.91 TE
Augustin 2014 Prospective, 40 asymptomatic ACLD n.a. n.a. 25 kPa 65% 93%
et al.39 TE patients
Schwabl 2015 Retrospective, 188 patients with CLD 0.846 0.957 16.1 kPa 94.8% 86.9%
et al.92 TE
Kitson 2015 Prospective, 95 patients with cirrhosis n.a. 0.900 29.0 kPa 71.9% 100%
et al.93 TE
Cho 2015 Retrospective, 219 consecutive patients with n.a. 0.850 n.a. n.a. n.a.
et al.94 TE alcoholic cirrhosis (some
Zykus 2015 Prospective, 107 patients with cirrhosis, 0.750 0.949 17.4 88.0% 87.5%
et al.95 TE mixed etiologies (67% Child-
Pugh A)
Salzl 2014 Prospective, 88 patients with cirrhosis TE: 0.765 0.870 16.8 kPa 89.7% 75%
et al.28 TE and pSWE (half decompensated)
(VTQ) pSWE: 0.646 0.855 2.58 m/s 71.4% 87.5%
Takuma 2016 Prospective, 60 patients, viral 0.609 0.833 n.a. n.a. n.a.
et al.29 pSWE (VTQ)
Attia 2015 pSWE (VTQ) 78 patients, mixed (some 0.444 0.929 2.17 m/s 97% 89%
et al.27 decomp.; 90% CSPH; 76% EV)
Procopet 2015 Prospective, 88 consecutive patients, all All cases: 2D-SWE: 0.858 2D-SWE: 17 kPa 80.8% 82.1%
et al.32 TE and 2D- compensated 2D-SWE: 0.611 (compensated)

SWE (SSI) TE: 0.699 2D-SWE reliable 2D-SWE reliable
measures: 0.948 measures: 90.9%
15.4 kPa 91.3%
Kim 2015 Prospective, 115 cirrhotic patients, mixed 0.646 0.819 15.2 kPa for CSPH 85.7% 80.0%
et al.31 2D-SWE (SSI) etiologies (some (0.587 if ascites)
decompensated); 92 analysed 0.867 for 21.6 kPa for
HVPG P12 mmHg HVPG P12 mmHg 83.3% 80.8%
Elkrief 2015 Prospective, 79 patients, mixed (55 with 0.578 (measures 2D-SWE: 0.790 2D-SWE:24.5 kPa 81% 88%
et al.30 2D-SWE (SSI) ascites; 70 with CSPH) with
and TE variable \10%) TE: 0.780 TE: 65.3 kPa
TE: n.a. 52% 100%
Jansen 2016 Prospective, 158 patients, mixed (some 0.626 0.86 24.6 kPa 68.3% 80.4%
et al.59 2D-SWE (SSI) decompensated) \16 kPa rule-out
[29.5 kPa rule-in
ACLD, advanced chronic liver disease; AUROC, area under receiver operator curve; CSPH, clinically significant portal hypertension; HBV, hepatitis B virus; HCC, hepato-
cellular carcinoma; LSM, liver stiffness measurement; n.a., not applicable; pSWE, point shear wave elastography; SSI, supersonic imagine; SWE, shear wave elastography;
TE, transient elastography; VTQ, virtual touch quantification.

Journal of Hepatology 2017 vol. 67 j 399–411 403

CLD: ever decompensated? Patient with compensated CLD

Advanced Chronic Liver Disease?
CSPH in 100% LSM >10 kPA: No
High likelihood varices needing probable; >15 kPa Tailor need of therapy and follow-up
treatment: very probable Yes
perform endoscopy


LSM >21 kPA very Perform HVPG measurement

Yes LSM 13.6-21 kPA: uncertainty
likely (~90%) to assess PH

Perform endoscopic screening?

LSM >20 kPA and/or If LSM <20 kPa and platelets Safely avoid endoscopic screening
Yes >150 G/L varices needing
platelets <150 G/L Repeat LSM + platelets every year
treatment are very unlikely (<5%)*

Varices needing treatment?

Endoscopic surveillance according
No to the findings and presence or
Yes absence of ongoing exposure
to liver injury
Start NSBB or endoscopic band ligation

Fig. 2. Pragmatic use of liver stiffness measurement to guide the need of further testing in patients with compensated advanced chronic liver disease according to
the Baveno VI recommendations. LSM can improve clinical decision-making in different steps of risk stratification. *This rule should not be applied to patients with
cholestatic liver disease. ACLD, advanced chronic liver disease; CLD, chronic liver disease; CSPH, clinically significant portal hypertension; HVPG, hepatic venous pressure
gradient; LSM, liver stiffness measurement; Plt, platelet count; NSBB, non-selective beta-blockers.

performance compared to LSM alone (AUROC: 0.95 patients with cholestatic liver disease and should
vs. 0.88, p \0.001). The use of two cut-off values, not be applied in this population owing to specifici-
either to rule out varices (LSPS \3.5) or to rule ties in PH in this context (i.e. the pre-sinusoidal com-
them in (LSPS [5.5), allowed correct stratification ponent that might not be properly sensed by LSM).
in 90% of cases, with a limited number of indeter- The successful development of non-invasive
minate findings.36 The higher accuracy of LSPS diagnostic criteria to reduce the number of unneces-
compared to LSM has been confirmed in two inde- sary endoscopy is a significant change in the man-
pendent studies in patients with ACLD of different agement of patients with ACLD. Further studies are
aetiologies,26,37 showing correct classification in needed to refine these criteria in order to reduce fur-
over 85% of cases using LSPS. However, the cut-off ther the number of unnecessary endoscopies and to
values used were different to those originally pub- optimise costs. A tailored individual risk approach,
lished by Kim et al.36 as postulated in a recent multicentric study (Antici-
Given that spleen size is not always available, a pate study),25 could lead to a more accurate use of

simplified combination of LSM and platelet count the LSM values in this field.
was also assessed,38,39 and good results were LSM by pSWE (VTQ, Siemens, Germany) has been
achieved in ruling out varices needing treatment tested in a limited number of studies addressing the
(low false-negative rates using thresholds of diagnosis and severity of EV. LSM by VTQ was higher
LS \20–25 kPa plus platelet count [120–150 g/L). in patients with varices, and increased in patients
As a result, experts at the Baveno VI consensus with large varices,27,28 similar to results using TE.
meeting on PH agreed that non-invasive tests could Validated cut-off values are not available yet. This
be used to identify patients with ACLD that could is also the case for 2D-SWE.
safely avoid screening endoscopy,12 however, a
conservative criterion to skip endoscopy was out-
lined, i.e. platelet count [150 g/L and LSM \20 kPa Limitations of LSM
(Fig. 2). Analyses of the performance of this crite-
rion in compensated ACLD are already avail- Liver stiffness is a mechanical property, and fibrosis
able,25,40,41 and all confirmed that about 20% of is the major determinant of LSM in ACLD; however,
endoscopies could be safely avoided, missing less several other tissue abnormalities can contribute to
than 4% of patients with varices needing treatment. increased liver stiffness, irrespective of fibrosis.
Importantly, this has not been investigated in Inflammation, infiltrative diseases, cholestasis and

404 Journal of Hepatology 2017 vol. 67 j 399–411

Table 3. Accuracy of spleen stiffness measurement using ultrasound elastography techniques for clinically significant portal hypertension and esophageal varices in
advanced chronic liver disease.

Study Year Method N included and Endpoint AUROC for the Chosen cut-off for the Sensitivity Specificity
used etiology selected endpoint selected endpoint
Stefanescu et al.51 2011 TE 174, mixed EV 0.781 46.4 kPa 83.6% 61.4%
Colecchia et al.52 2012 TE 141, HCV CSPH 0.966 Rule-out 40 kPa 98.5% 74.3%
compensated Rule-in 52.8 kPa 76.9% 97.1%

EV 0.941 Rule-out 41.3 kPa 98.1% 66.0%

Rule-in 55 kPa 71.7% 95.7%
Sharma et al.96 2013 TE 200, mixed EV 0.898 40.8 kPa 94% 76%
Calvaruso et al.54 2013 TE (modified 112, HCV LEV 0.820 54.0 kPa 80% 70%
range) compensated
Wong et al.97 2016 TE 144, HBV EV 0.685 18.9 kPa (rule-out) NPV 92.1%
PPV 56.1%
54.9 kPa (rule-in)
Vermehren et al.98 2012 pSWE (VTQ) 166, mixed LEV 0.580 3.40 m/s 87% 31%

Zykus et al.95 2015 TE 107, mixed CSPH 0.846 47.6 kPa 77.3% 79.2%
Rifai et al.99 2011 pSWE (VTQ) 100, mixed CSPH 0.680 3.29 m/s 47% 73%
Mori et al.100 2013 pSWE (VTQ) 33, HCV EV 0.800 3.41 m/s n.a. n.a.
Attia et al.27 2015 pSWE (VTQ) 78, mixed (some CSPH 0.968 2.32 m/s 96% 89%
90% CSPH; 76% EV)
Takuma et al.29 2016 pSWE (VTQ) 60, viral CSPH 0.943 3.10 m/s 97.1% 57.7%
HVPG P12 0.963 3.15 m/s 96.6% 61.3%
EV 0.937 3.36 m/s 95.8% 77.8%
LEV 0.955 3.51 m/s 93.8% 84.1%
Bota et al.101 2012 pSWE (VTQ) 140, mixed EV 0.578 2.55 m/s 96.7% 21.1%
Kim et al.102 2015 pSWE (VTQ) 125, mixed EV 0.768 3.16 m/s 87% 60.4%
LEV 0.786 3.40 m/s 78.9% 63.0%
Rizzo et al.103 2014 pSWE (VTQ) 54, mixed EV 0.959 3.10 m/s 96.4% 88.5%
+ controls
Takuma et al.37 2013 pSWE (VTQ) 340, mixed EV 0.937 (viral); 3.18 m/s 98.5% 60.1%
0.923 (others) 3.24 m/s 97.7% 65.2%

LEV 3.30 m/s 98.9% 62.9%

Ye et al.104 2012 pSWE (VTQ) 204, HBV EV 0.830 3.16 m/s 84.1% 81%
Elkrief et al.30 2015 2D-SWE 79, mixed (most CSPH 0.630 34.7 kPa 40% 100%
(SSI) decompensated; LEV 0.580 32.3 kPa 48% 71%
89% CSPH; 69%
TE Child-Pugh B-C; 34 CSPH 0.640 56.3 kPa 73% 67%
with LEV) LEV 0.650 73.5 kPa 54% 78%
Procopet et al.32 2015 2D-SWE 55, mixed CSPH 0.725 22.7 kPa (rule-out) 90% n.a.
(SSI) Compensated 40 kPa (rule-in) n.a. 90%

Cassinotto et al.57 2015 2D-SWE 401, mixed (some LEV 0.80 25.6 kPa (chosen to 94% 36%
(SSI) decompensated) maximise sensitivity)

Jansen et al.59 2016 2D-SWE 158, mixed (some CSPH 0.84 26.3 kPa 79.7% 84.2%
(SSI) decompensated) 621.7 kPa rule-out
[35.6 kPa rule-in
CSPH, clinically significant portal hypertension; EV, esophageal varices; HBV, hepatitis B virus; LEV, large esophageal varices; n.a., not applicable; PPV, positive predictive
value; SSI, supersonic imagine; pSWE, point shear wave elastography; SWE, shear wave elastography; TE, transient elastography; VTQ, virtual touch quantification.

venous congestion should always be considered as Obesity used to be a limiting factor for LSM by
possible confounders of the relationship between TE;46 however, a specifically designed extra-large
LSM and portal pressure, irrespective of the elas- (XL) probe is now available, overcoming this prob-
tography method used.16 LSM (measured by any lem. Nonetheless, LSM values measured by XL probe
of the available ultrasound techniques) increases are lower than those measured by the standard M
after meal ingestion in ACLD and in patients with probe, and there is currently no data regarding XL
PH.42–45 Therefore, liver elastography should be probe cut-off values for the diagnosis of CSPH and
always performed in fasting state.17 varices. Further data on this topic are required.

Journal of Hepatology 2017 vol. 67 j 399–411 405

Technique-specific limitations for pSWE and surement range up to 150 kPa with appropriate soft-
2D-SWE include the need of a basic knowledge of ware modifications has been proposed and tested,
ultrasound to position the probe correctly, and and showed that patients with large varices, or
interpret the possible artefacts leading to non- varices which had already bled, often had an SSM
reliable results.15 well above 75 kPa.54 pSWE29,37,55 and 2D-SWE32,56
Despite a good correlation with each other, the have been used more recently, and do not have the
values measured by the different ultrasound ceiling effect of TE. Assessment of SSM using
elastography techniques and ultrasound devices 2D-SWE (SSI), like TE, is limited to patients with
are not identical, and also differ from those pro- enlarged spleen.32,56,57 However, SSM can be mea-
vided by TE.16,47,48 Therefore, technique-specific sured using pSWE (VTQ) in about 95% of cases.37
and machine-specific experience is required to gen- Higher variability has been observed in the mea-
erate appropriate results, and specific knowledge surement of this parameter compared to LSM using
and clinical experience is needed to interpret the the same technique.58
findings correctly. SSM is not yet used routinely in clinical practice
because of the limitations described above. Future
Elastography for portal hypertension beyond liver studies should clarify to what extent it adds to
Key point
stiffness measurement: Spleen stiffness measurement LSM for the diagnosis of CSPH (as proposed by a
Spleen stiffness measure- recent study59), whether it allows patients who do
ment (SSM) is closely The spleen undergoes parenchymal remodelling in not meet the Baveno criteria to safely avoid endo-
associated with portal scopy, and if it enables a2 correct risk stratification
patients with PH. This is partly attributable to pas-
hypertension. SSM is
sive congestion and increased arterial inflow, and of EV in patients with cholestatic liver disease.
increased in pre-hepatic
portal hypertension (PH), partly because of increased hyperactive splenic lym- Technical improvements, such as dedicated probes
suggesting that it could be phoid tissue and enhanced angiogenesis and fibro- for SSM using TE would facilitate these studies in
used as a surrogate for PH, genesis,49 leading to the progressive development centres using this ultrasound elastography
irrespective of its cause. of splenomegaly in most patients. Ultrasound stud- technique.
ies in the 1980s and 1990s showed that spleen vas-
cular resistance (estimated by Doppler pulsatility Prediction of clinical decompensation by liver and
and resistance indexes) is increased in patients with spleen stiffness measurement, and use of ultrasound
PH, and correlates with PH severity and complica- elastography in the follow-up of patients with portal
tions.50 Therefore, it has been postulated that spleen hypertension
stiffness measurement (SSM) by ultrasound elastog-
raphy could be an accurate non-invasive surrogate Several studies, mostly using TE, showed that LSM
for PH, and devoid of the limitations of LSM has prognostic value when examining hard
(Fig. 1). Studies comparing LSM and SSM (measured endpoints (clinical decompensation, hepatocellular
Key point by TE in an adequate left intercostal space using carcinoma and death) in patients with CLD; this
technical conditions similar to those used for LSM) was confirmed by a meta-analysis.60 In patients
SSM is a promising parame-
ter for use in predicting both showed that the spleen is substantially stiffer than with compensated ACLD undergoing both LSM and
the presence of varices, and the liver in both healthy subjects and patients with HVPG measurement and followed-up for two years
varices needing treatment. CLD.51 In some studies, SSM showed a closer corre- showed that HVPG and LSM presented a similar
lation with HVPG,52 CSPH and presence and size of accuracy for predicting a first decompensating epi-
EV when compared to LSM;51,52 in other studies, sode; all decompensating events occurred in
SSM had similar or less accuracy (Table 3). A meta- patients with LSM P21.1 kPa (using TE; a value with
analysis of the 16 studies that compared the a high specificity for identifying patients with
accuracy of SSM with LSM for use in predicting the CSPH).61 This study provided an important proof of

presence of EV (ten studies using TE, three using concept for PH-related events. In another study,
pSWE-VTQ, and three using 2D-SWE-SSI) has shown both LSM and SSM predicted clinical decompensation
that SSM was significantly superior to LSM.53 How- in patients with HCV-related cirrhosis;62 however,
ever, most of the published data were obtained in only SSM and MELD score remained associated with
heterogeneous populations of patients, with either occurrence of first decompensation in multivariate
compensated (correct target) or decompensated cir- analysis (cut-off value for discrimination: 54 kPa).
rhosis, and therefore, the superiority of SSM versus Table 4 summarises the results of longitudinal stud-
LSM for the diagnosis of PH in compensated ACLD ies using LSM or SSM to predict PH-related events.
Key point patients, has not been definitively proven. Most of the published data only include small-to-
Liver stiffness measurement TE applicability for SSM is limited to about 70% medium size cohorts with a limited follow-up.
can be used to predict of cases and, for technical reasons, it is closely Limited data are available on changes in LSM in
clinical decompensation in dependent on the presence of increased spleen size. relation to changes in portal pressure or clinical risk.
patients with compensated Additionally, measurement of stiffness by TE cur- A recommendation was formed at the Baveno con-
advanced chronic liver sensus conference on PH in 2015, suggesting that
rently reaches a maximum of 75 kPa. As the spleen
disease. Spleen stiffness
is significantly stiffer than the liver, most patients patients not requiring screening endoscopy accord-
measurement requires fur-
ther evaluation in this field. with severe PH show maximal values of SSM, above ing to non-invasive criteria on the first observation,
which, risk cannot be stratified. Widening the mea- should undergo yearly LSM and platelet count to

406 Journal of Hepatology 2017 vol. 67 j 399–411

Table 4. Liver stiffness measurement and spleen stiffness measurement using ultrasound elastography techniques for the prediction of portal hypertension-related outcomes in longitudinal studies.

Endpoint Study Year Parameter N included and etiology Cumulative AUROC for the Chosen cut-off Notes
and method incidence of selected for the
the endpoint endpoint selected
Clinical decompensation Robic 2011 LSM by TE 96 patients with cirrhosis 41% 0.815 21.1 kPa 100% Sens and NPV; 41% Spec
et al.61 (mixed etiology) followed up HVPG based AUROC: 0.837
in mean for 245 ± 244 days
Kitson 2015 LSM by TE 95 patients with cirrhosis 29.5% 0.730 34.5 kPa LSM: Sens 75.0%, Spec 69.4% HVPG:
et al.93 followed up in mean for Sens 100%, Spec 40.3%
15.1 months
Merchante 2012 LSM by TE 239 HIV/HCV coinfected 12.9% n.a. 40 kPa LSM had borderline significance as
et al.105 patients with cirrhosis predictor (HR 1.03; p = 0.08)
followed up in median for
20 months
Journal of Hepatology 2017 vol. 67 j 399–411

Colecchia 2014 SSM 92 patients with HCV cirrhosis 32.6% 0.678 54 kPa NPV 97.5%
et al.62 by TE followed up in mean for HVPG based AUROC: 0.830
24 months
Grgurevic 2015 LSM and SSM 44 patients with compensated 40.9% n.a. LSM: 21.5 kPa Risk of decompensation 3.4-fold
et al.106 by 2D-SWE cirrhosis followed up in mean higher over the cut-off value
(SSI) for 28 months SSM: 31.7 kPa Both LSM and SSM were associated to
decompensation, but LSM was
Clinical worsening defined as Wang 2014 LSM by TE 220 patients with viral 12.8% at Progression of 17 kPa LSM measured every 6–12 months, but
progression of endoscopic et al.107 cirrhosis; median follow-up 3 years varices: 0.744 unclear if changes improve the
signs, decompensation or HCC 36.9 months prediction
Clinical 21.1 kPa

No value for HCC
Variceal bleeding Kim 2011 LSM by TE 577 HBV cirrhosis patients; 4.3% (16.7% of 0.929 LSPS [6.5 LSPS independent predictor, alongside
et al.108 + spleen size 150 with VNT, followed up for those with large variceal size and Child-Pugh B/C
and Plt (LSPS) in median 29 months VNT)
Merchante 2016 LSM by TE 446 HIV/HCV coinfected 3.4% n.a. 21 kPa NPV 100%
et al.109 patients with cirrhosis
followed up in median for
49 months
Takuma 2016 SSM by pSWE 446 patients with cirrhosis 7.4% 0.857 overall 3.64 m/s MELD score, red colour signs on
et al.55 (VTQ) followed up in mean for overall varices and SS independent predictors
32.7 months of EV bleeding; SSM best parameter
0.911 in 3.48 m/s in
compensated compensated
HCC, hepatocellular carcinoma; LSM, liver stiffness measurement; LSPS, LSM-spleen diameter to platelet ratio score; MELD, model for end-stage liver disease; n.a., not applicable; NPV, negative predictive value; pSWE,
point shear wave elastography; Sens, sensitivity; Spec, specificity; SSI, supersonic imagine; SSM, spleen stiffness measurement; SWE, shear wave elastography; TE, transient elastography; VNT, varices needing treatment;
Plt, platelet count; VTQ, virtual touch quantification.


Site of increased resistance to portal flow Liver stiffness Spleen stiffness HVPG
Pre-hepatic PH Normal Increased Normal
Pre-sinusoidal PH Normal or slightly Increased Normal or slightly
(idiopathic PH) elevated elevated (<10 mmHg)
(<12 kPa) Often veno-venous
Intrahepatic PH communications
Sinusoidal PH Increased >13.6 kPa Increased Increased:
(cirrhosis; others) LSM ≥21 kPa: HVPG ≥10 mmHg
Spec >90% for CSPH defines CSPH
Post-hepatic PH No published data; No published data; Normal or increased,
(Budd-Chiari syndrome) likely increased due likely increased due with high free hepatic
to venous congestion to PH venous P

Fig. 3. Findings at liver and spleen stiffness measurements, and hepatic venous pressure gradient measurement in patients
with portal hypertension, classified according to the increased site of resistance to portal blood flow. The combined use of liver
and spleen stiffness measurements can improve the characterisation of patients with portal hypertension of unknown cause. The
final diagnosis remains, however, to be confirmed by liver biopsy. CSPH, clinically significant portal hypertension; EHPVO,
extrahepatic portal vein obstruction; HVPG, hepatic venous pressure gradient; LSM, liver stiffness measurement; P, pressure; PH,
portal hypertension.

pin-point the optimum time for first endoscopy.12 detected (e.g. idiopathic PH, schistosomiasis, amyloi-
This recommendation awaits validation. dosis, lymphoma, tuberculosis, hepatic myeloid
LSM using TE did not reflect the HVPG response metaplasia due to extramedullary haematopoiesis in
to NSBB in one study.20 Conversely, another study patients with myeloproliferative neoplasms). Liver
using 2D-SWE (SSI) in a small group of patients, biopsy remains crucial in patients without evident
showed a strong correlation between changes in causes of ACLD, although LSM and SSM may help the
LSM and HVPG (r = 0.863).63 clinician in this initial assessment (Fig. 3). LSM is usu-
Data regarding SSM in relation to NSBB therapy ally only moderately increased in idiopathic PH (mean
are lacking. However, after transjugular intrahep- value of 8.4 ± 3.3 kPa),67 showing a clear mismatch
atic portosystemic shunt (TIPS) placement, SSM with the values expected in patients with cirrhosis;
using pSWE (Philips iU22 in one study; VTQ in one however, SSM in this population is elevated to values
study) decreased,64,65 suggesting that this parame- similar or even higher than those observed in patients
ter might have potential for monitoring the effects with cirrhotic PH.68 Cirrhotic and idiopathic PH often
of therapy on PH, and deserves further investiga- appear similar when imaged, and therefore, the ratio
tion. There is little evidence to suggest whether between LSM and SSM could improve the clinicians’
LSM or SSM could be useful in identifying patients ability to identify idiopathic PH and avoid an incorrect
remaining at risk of PH-related complications after diagnosis of cryptogenic cirrhosis.
successful treatment of HCV in patients with cirrho- In patients with extrahepatic portal vein obstruc-
sis. Nevertheless, one recent study showed that LSM tion (EHPVO) spleen stiffness is increased,67,69 and
rapidly decreased after virus clearance, but this was SSM values are higher in patients who had already
probably the result of decreased liver inflamma- bled from varices, versus patients whose varices
tion,66 thus making interpretation of post-therapy had not bled.69 Therefore, SSM might be a valuable

values difficult. In the same study SSM did not tool used to stratify the severity of PH in patients
change significantly, possibly indicating persistence with EHPVO, in whom HVPG is not reliable (pre-
of PH in this population.66 Further studies with long hepatic PH).
follow-up and longitudinal assessment of these No data are available in patients with BCS;
parameters are needed in this field. however, in this author’s experience, LSM is often
markedly increased in patients with BCS, likely due
Elastography in patients with suspected portal to hepatic congestion.70 In a study using a mouse
hypertension of unknown cause, and in patients model of congestive hepatopathy,71 chronic venous
with non-cirrhotic portal hypertension2 stasis led to an increase in LSM (measured by mag-
netic resonance elastography) in the absence of
When PH is suspected, the first step towards its fibrosis, probably because of the mechanical tension
correct characterisation is to identify its cause. This generated by vascular strain in congestion. Using an
requires excluding the presence of thrombosis in in vitro approach, the authors demonstrated that
the portal vein and/or the hepatic veins by ultra- mechanical cyclic strain increases fibronectin
sound, so that the differential diagnosis is restricted secretion from hepatic stellate cells, and increased
to intrahepatic forms of PH. ACLD accounts for over extracellular matrix fibril assembly, suggesting
90% of cases, but other uncommon forms can be that mechanical forces, such as sinusoidal stretch

408 Journal of Hepatology 2017 vol. 67 j 399–411

secondary to congestion lead to early steps of LSM cannot replace HVPG for monitoring the
matrix development.71 This may mean that the haemodynamic response to NSBB. Further research
prognostic value of LSM in ACLD might be partially is needed to establish whether the dynamics of
ascribed to the accelerated fibrogenesis in patients SSM over time or in response to treatment could
with stiffer livers. be a better indicator of HVPG changes.
Data regarding the pediatric population with PH The field of ultrasound elastography is rapidly
due to biliary atresia, suggest that liver and spleen evolving, and newer techniques are becoming
elastography (measured either by TE or pSWE) widely available; their diagnostic performance for
could be valuable tools to help predict outcomes PH remains to be established, but it is likely to be
before surgery, and might be used after the Kasai similar to TE. Clinical hepatologists are required to
operation to monitor liver disease and PH.72,73 understand the specificities of each method, and
have the technical skills and knowledge of the val-
ues, pitfalls and limitations, which is a challenge.
Head-to-head comparisons to select the best
method for each clinical scenario in PH is certainly
Ultrasound elastography is now an established and
a field for future research.
useful tool for hepatologists. It enables a point-of-
care and quantitative assessment of liver and
spleen stiffness, which relate to PH and its compli-
cations, thus allowing rapid risk stratification and Financial support
identification of patients requiring further testing
(i.e. invasive assessment). TE remains the most val- Interdisciplinary Grant 2015 of the University of
idated technique. LSM using TE cannot provide the Bern (UniBe-ID 2015).
exact value of HVPG, nor identify with high cer-
tainty which patients carry EV, but it can confirm
the presence of CSPH. Even more importantly, Conflict of interest
LSM combined with platelet count has now entered
in the decision algorithm for patients with compen- The author declared that she does not have anything
sated ACLD, allowing patients at very low risk of to disclose regarding funding or conflict of interest
varices needing treatment, to avoid endoscopy. with respect to this manuscript.

References [14] Shiina T, Nightingale KR, Palmeri ML, Hall TJ, Bamber JC, Barr RG, et al.
WFUMB guidelines and recommendations for clinical use of ultrasound
[1] Bosch J, Abraldes JG, Berzigotti A, Garcia-Pagan JC. The clinical use of HVPG elastography: Part 1: basic principles and terminology. Ultrasound Med Biol
measurements in chronic liver disease. Nat Rev Gastroenterol Hepatol 2015;41:1126–1147.
2009;6:573–582. [15] Piscaglia F, Marinelli S, Bota S, Serra C, Venerandi L, Leoni S, et al. The role of
[2] Valla D. Splanchnic Vein Thrombosis. Semin Thromb Hemost 2015;41: ultrasound elastographic techniques in chronic liver disease: current status
494–502. and future perspectives. Eur J Radiol 2014;83:450–455.
[3] Bosch J. Vascular deterioration in cirrhosis: the big picture. J Clin Gastroen- [16] Friedrich-Rust M, Poynard T, Castera L. Critical comparison of elastography
terol 2007;41:S247–S253. methods to assess chronic liver disease. Nat Rev Gastroenterol Hepatol
[4] Berzigotti A, Seijo S, Reverter E, Bosch J. Assessing portal hypertension in liver 2016;13:402–411.
diseases. Expert Rev Gastroenterol Hepatol 2013;7:141–155. [17] European Association for Study of Liver, Asociacion Latinoamericana para el
[5] Tsochatzis EA, Bosch J, Burroughs AK. Liver cirrhosis. Lancet 2014;383: Estudio del Higado. EASL-ALEH Clinical Practice Guidelines: Non-invasive
1749–1761. tests for evaluation of liver disease severity and prognosis. J Hepatol
[6] Abraldes JG, Tarantino I, Turnes J, Garcia-Pagan JC, Rodes J, Bosch J. 2015;63:237–264.
Hemodynamic response to pharmacological treatment of portal hyperten- [18] Carrion JA, Navasa M, Bosch J, Bruguera M, Gilabert R, Forns X. Transient
sion and long-term prognosis of cirrhosis. Hepatology 2003;37:902–908. elastography for diagnosis of advanced fibrosis and portal hypertension in

[7] Garcia-Tsao G, Abraldes JG, Berzigotti A, Bosch J. Portal hypertensive bleeding patients with hepatitis C recurrence after liver transplantation. Liver Transpl
in cirrhosis: Risk stratification, diagnosis, and management: 2016 practice 2006;12:1791–1798.
guidance by the American Association for the study of liver diseases. [19] Vizzutti F, Arena U, Romanelli RG, Rega L, Foschi M, Colagrande S, et al. Liver
Hepatology 2016. stiffness measurement predicts severe portal hypertension in patients with
[8] Rosselli M, MacNaughtan J, Jalan R, Pinzani M. Beyond scoring: a modern HCV-related cirrhosis. Hepatology 2007;45:1290–1297.
interpretation of disease progression in chronic liver disease. Gut [20] Reiberger T, Ferlitsch A, Payer BA, Pinter M, Homoncik M, Peck-Radosavljevic
2013;62:1234–1241. M, et al. Non-selective beta-blockers improve the correlation of liver stiffness
[9] Saffioti F, Pinzani M. Development and regression of cirrhosis. Dig Dis and portal pressure in advanced cirrhosis. J Gastroenterol 2012;47:561–568.
2016;34:374–381. [21] Bureau C, Metivier S, Peron JM, Selves J, Robic MA, Gourraud PA, et al.
[10] Patel K, Bedossa P, Castera L. Diagnosis of liver fibrosis: present and future. Transient elastography accurately predicts presence of significant portal
Semin Liver Dis 2015;35:166–183. hypertension in patients with chronic liver disease. Aliment Pharmacol Ther
[11] de Franchis R. Cirrhosis: screening for esophageal varices. Nat Rev Gastroen- 2008;27:1261–1268.
terol Hepatol 2009;6:449–450. [22] Castera L, Pinzani M, Bosch J. Non invasive evaluation of portal hypertension
[12] de Franchis R, Baveno VIF. Expanding consensus in portal hypertension: using transient elastography. J Hepatol 2012;56:696–703.
Report of the Baveno VI Consensus Workshop: stratifying risk and individ- [23] You MW, Kim KW, Pyo J, Huh J, Kim HJ, Lee SJ, et al. A meta-analysis for the
ualizing care for portal hypertension. J Hepatol 2015;63:743–752. diagnostic performance of transient elastography for clinically significant
[13] Bamber J, Cosgrove D, Dietrich CF, Fromageau J, Bojunga J, Calliada F, et al. portal hypertension. Ultrasound Med Biol 2017;43:59–68.
EFSUMB guidelines and recommendations on the clinical use of ultrasound [24] Llop E, Berzigotti A, Reig M, Erice E, Reverter E, Seijo S, et al. Assessment of
elastography. Part 1: Basic principles and technology. Ultraschall Med portal hypertension by transient elastography in patients with compensated
2013;34:169–184. cirrhosis and potentially resectable liver tumors. J Hepatol 2012;56:103–108.

Journal of Hepatology 2017 vol. 67 j 399–411 409

[25] Abraldes JG, Bureau C, Stefanescu H, Augustin S, Ney M, Blasco H, et al. [47] Ferraioli G, Filice C, Castera L, Choi BI, Sporea I, Wilson SR, et al. WFUMB
Noninvasive tools and risk of clinically significant portal hypertension guidelines and recommendations for clinical use of ultrasound elastography:
and varices in compensated cirrhosis: The ‘‘Anticipate” study. Hepatology Part 3: liver. Ultrasound Med Biol 2015;41:1161–1179.
2016. [48] Cosgrove D, Piscaglia F, Bamber J, Bojunga J, Correas JM, Gilja OH, et al.
[26] Berzigotti A, Seijo S, Arena U, Abraldes JG, Vizzutti F, Garcia-Pagan JC, et al. EFSUMB guidelines and recommendations on the clinical use of ultrasound
Elastography, spleen size, and platelet count identify portal hypertension in elastography. Part 2: clinical applications. Ultraschall Med 2013;34:238–253.
patients with compensated cirrhosis. Gastroenterology 2013;144:e101. [49] Mejias M, Garcia-Pras E, Gallego J, Mendez R, Bosch J, Fernandez M.
[27] Attia D, Schoenemeier B, Rodt T, Negm AA, Lenzen H, Lankisch TO, et al. Relevance of the mTOR signaling pathway in the pathophysiology of
Evaluation of liver and spleen stiffness with acoustic radiation force impulse splenomegaly in rats with chronic portal hypertension. J Hepatol 2010;52:
quantification elastography for diagnosing clinically significant portal 529–539.
hypertension. Ultraschall Med 2015;36:603–610. [50] Piscaglia F, Donati G, Cecilioni L, Celli N, Stagni B, Pini P, et al. Influence of the
[28] Salzl P, Reiberger T, Ferlitsch M, Payer BA, Schwengerer B, Trauner M, et al. spleen on portal haemodynamics: a non-invasive study with Doppler
Evaluation of portal hypertension and varices by acoustic radiation force ultrasound in chronic liver disease and haematological disorders. Scand J
impulse imaging of the liver compared to transient elastography and AST to Gastroenterol 2002;37:1220–1227.
platelet ratio index. Ultraschall Med 2014;35:528–533. [51] Stefanescu H, Grigorescu M, Lupsor M, Procopet B, Maniu A, Badea R. Spleen
[29] Takuma Y, Nouso K, Morimoto Y, Tomokuni J, Sahara A, Takabatake H, et al. stiffness measurement using Fibroscan for the noninvasive assessment of
Portal hypertension in patients with liver cirrhosis: diagnostic accuracy of esophageal varices in liver cirrhosis patients. J Gastroenterol Hepatol
spleen stiffness. Radiology 2016;279:609–619. 2011;26:164–170.
[30] Elkrief L, Rautou PE, Ronot M, Lambert S, Dioguardi Burgio M, Francoz C, et al. [52] Colecchia A, Montrone L, Scaioli E, Bacchi-Reggiani ML, Colli A, Casazza G,
Prospective comparison of spleen and liver stiffness by using shear-wave and et al. Measurement of spleen stiffness to evaluate portal hypertension and
transient elastography for detection of portal hypertension in cirrhosis. the presence of esophageal varices in patients with HCV-related cirrhosis.
Radiology 2015;275:589–598. Gastroenterology 2012;143:646–654.
[31] Kim TY, Jeong WK, Sohn JH, Kim J, Kim MY, Kim Y. Evaluation of portal [53] Ma X, Wang L, Wu H, Feng Y, Han X, Bu H, et al. Spleen stiffness is superior to
hypertension by real-time shear wave elastography in cirrhotic patients. liver stiffness for predicting esophageal varices in chronic liver disease: a
Liver Int 2015;35:2416–2424. meta-analysis. PLoS One 2016;11:e0165786.
[32] Procopet B, Berzigotti A, Abraldes JG, Turon F, Hernandez-Gea V, Garcia- [54] Calvaruso V, Bronte F, Conte E, Simone F, Craxi A, Di Marco V. Modified
Pagan JC, et al. Real-time shear-wave elastography: applicability, reliability spleen stiffness measurement by transient elastography is associated with
and accuracy for clinically significant portal hypertension. J Hepatol presence of large oesophageal varices in patients with compensated hepatitis
2015;62:1068–1075. C virus cirrhosis. J Viral Hepat 2013;20:867–874.
[33] Jansen C, Bogs C, Verlinden W, Thiele M, Möller P, Görtzen J, et al. Algorithm [55] Takuma Y, Nouso K, Morimoto Y, Tomokuni J, Sahara A, Takabatake H, et al.
to rule out clinically significant portal hypertension combining Shear-wave Prediction of oesophageal variceal bleeding by measuring spleen stiffness in
elastography of liver and spleen: a prospective multicentre study. Gut 2016. patients with liver cirrhosis. Gut 2016;65:354–355., [Epub ahead of print]. [56] Garcia-Pagan JC, Di Pascoli M, Caca K, Laleman W, Bureau C, Appenrodt B,
[34] Choi S-Y, Jeong WK, Kim Y, Kim J, Kim TY, Sohn JH. Shear-wave elastography: et al. Use of early-TIPS for high-risk variceal bleeding: results of a post-RCT
a noninvasive tool for monitoring changing hepatic venous pressure surveillance study. J Hepatol 2013;58:45–50.
gradients in patients with cirrhosis. Radiology 2014;273:917–926. [57] Cassinotto C, Charrie A, Mouries A, Lapuyade B, Hiriart JB, Vergniol J, et al.
[35] Shi KQ, Fan YC, Pan ZZ, Lin XF, Liu WY, Chen YP, et al. Transient elastography: Liver and spleen elastography using supersonic shear imaging for the non-
a meta-analysis of diagnostic accuracy in evaluation of portal hypertension invasive diagnosis of cirrhosis severity and oesophageal varices. Dig Liver Dis
in chronic liver disease. Liver Int 2013;33:62–71. 2015;47:695–701.
[36] Kim BK, Han KH, Park JY, Ahn SH, Kim JK, Paik YH, et al. A liver stiffness [58] Balakrishnan M, Souza F, Munoz C, Augustin S, Loo N, Deng Y, et al. Liver and
measurement-based, noninvasive prediction model for high-risk esophageal spleen stiffness measurements by point shear wave elastography via acoustic
varices in B-viral liver cirrhosis. Am J Gastroenterol 2010;105:1382–1390. radiation force impulse: intraobserver and interobserver variability and
[37] Takuma Y, Nouso K, Morimoto Y, Tomokuni J, Sahara A, Toshikuni N, et al. predictors of variability in a US population. J Ultrasound Med 2016;35:
Measurement of spleen stiffness by acoustic radiation force impulse imaging 2373–2380.
identifies cirrhotic patients with esophageal varices. Gastroenterology [59] Jansen C, Bogs C, Verlinden W, Thiele M, Moller P, Gortzen J, et al. Shear-wave
2013;144:e102. elastography of the liver and spleen identifies clinically significant portal
[38] Ding NS, Nguyen T, Iser DM, Hong T, Flanagan E, Wong A, et al. Liver stiffness hypertension: a prospective multicentre study. Liver Int 2016.
plus platelet count can be used to exclude high-risk oesophageal varices. [60] Singh S, Fujii LL, Murad MH, Wang Z, Asrani SK, Ehman RL, Kamath PS, et al.
Liver Int 2016;36:240–245. Liver stiffness is associated with risk of decompensation, liver cancer, and
[39] Augustin S, Millan L, Gonzalez A, Martell M, Gelabert A, Segarra A, et al. death in patients with chronic liver diseases: a systematic review and meta-
Detection of early portal hypertension with routine data and liver stiffness in analysis. Clin Gastroenterol Hepatol 2013;11:1573–1584, e1571–1572; quiz
patients with asymptomatic liver disease: a prospective study. J Hepatol e1588–1579.
2014;60:561–569. [61] Robic MA, Procopet B, Metivier S, Peron JM, Selves J, Vinel JP, et al. Liver
[40] Maurice JB, Brodkin E, Arnold F, Navaratnam A, Paine H, Khawar S, et al. stiffness accurately predicts portal hypertension related complications in

Validation of the Baveno VI criteria to identify low risk cirrhotic patients not patients with chronic liver disease: a prospective study. J Hepatol
requiring endoscopic surveillance for varices. J Hepatol 2016;65:899–905. 2011;55:1017–1024.
[41] Marot A, Trepo E, Doerig C, Schoepfer A, Moreno C, Deltenre P. Liver stiffness [62] Colecchia A, Colli A, Casazza G, Mandolesi D, Schiumerini R, Reggiani LB, et al.
and platelet count for identifying patients with compensated liver disease at Spleen stiffness measurement can predict clinical complications in compen-
low risk of variceal bleeding. Liver Int 2016. sated HCV-related cirrhosis: a prospective study. J Hepatol 2014;60:
[42] Arena U, Lupsor Platon M, Stasi C, Moscarella S, Assarat A, Bedogni G, et al. 1158–1164.
Liver stiffness is influenced by a standardized meal in patients with chronic [63] Choi SY, Jeong WK, Kim Y, Kim J, Kim TY, Sohn JH. Shear-wave elastography:
hepatitis C virus at different stages of fibrotic evolution. Hepatology a noninvasive tool for monitoring changing hepatic venous pressure
2013;58:65–72. gradients in patients with cirrhosis. Radiology 2014;273:917–926.
[43] Berzigotti A, De Gottardi A, Vukotic R, Siramolpiwat S, Abraldes JG, Garcia- [64] Novelli PM, Cho K, Rubin JM. Sonographic assessment of spleen stiffness
Pagan JC, et al. Effect of meal ingestion on liver stiffness in patients with before and after transjugular intrahepatic portosystemic shunt placement
cirrhosis and portal hypertension. PLoS One 2013;8:e58742. with or without concurrent embolization of portal systemic collateral veins
[44] Popescu A, Bota S, Sporea I, Sirli R, Danila M, Racean S, et al. The influence of food in patients with cirrhosis and portal hypertension: a feasibility study. J
intake on liver stiffness values assessed by acoustic radiation force impulse Ultrasound Med 2015;34:443–449.
elastography-preliminary results. Ultrasound Med Biol 2013;39:579–584. [65] Ran HT, Ye XP, Zheng YY, Zhang DZ, Wang ZG, Chen J, et al. Spleen stiffness
[45] Gersak MM, Badea R, Lenghel LM, Vasilescu D, Botar-Jid C, Dudea SM. and splenoportal venous flow: assessment before and after transjugular
Influence of food intake on 2-D shear wave elastography assessment of liver intrahepatic portosystemic shunt placement. J Ultrasound Med 2013;32:
stiffness in healthy subjects. Ultrasound Med Biol 2016;42:1295–1302. 221–228.
[46] Castera L, Foucher J, Bernard PH, Carvalho F, Allaix D, Merrouche W, et al. [66] Verlinden W, Francque S, Michielsen P, Vanwolleghem T. Successful antiviral
Pitfalls of liver stiffness measurement: a 5-year prospective study of 13,369 treatment of chronic hepatitis C leads to a rapid decline of liver stiffness
examinations. Hepatology 2010;51:828–835. without an early effect on spleen stiffness. Hepatology 2016;64:1809–1810.

410 Journal of Hepatology 2017 vol. 67 j 399–411

[67] Seijo S, Reverter E, Miquel R, Berzigotti A, Abraldes JG, Bosch J, et al. Role of [89] Sanchez-Conde M, Miralles P, Bellon JM, Rincon D, Ramirez M, Gutierrez I,
hepatic vein catheterisation and transient elastography in the diagnosis of et al. Use of transient elastography (FibroScan(R)) for the noninvasive
idiopathic portal hypertension. Dig Liver Dis 2012;44:855–860. assessment of portal hypertension in HIV/HCV-coinfected patients. J Viral
[68] Furuichi Y, Moriyasu F, Taira J, Sugimoto K, Sano T, Ichimura S, et al. Hepat 2011;18:685–691.
Noninvasive diagnostic method for idiopathic portal hypertension based on [90] Reiberger T, Ferlitsch A, Payer BA, Pinter M, Schwabl P, Stift J, et al.
measurements of liver and spleen stiffness by ARFI elastography. J Gas- Noninvasive screening for liver fibrosis and portal hypertension by transient
troenterol 2013;48:1061–1068. elastography–a large single center experience. Wien Klin Wochenschr
[69] Sharma P, Mishra SR, Kumar M, Sharma BC, Sarin SK. Liver and spleen 2012;124:395–402.
stiffness in patients with extrahepatic portal vein obstruction. Radiology [91] Hong WK, Kim MY, Baik SK, Shin SY, Kim JM, Kang YS, et al. The usefulness of
2012;263:893–899. non-invasive liver stiffness measurements in predicting clinically significant
[70] Colli A, Pozzoni P, Berzuini A, Gerosa A, Canovi C, Molteni EE, et al. portal hypertension in cirrhotic patients: Korean data. Clin Mol Hepatol
Decompensated chronic heart failure: increased liver stiffness measured by 2013;19:370–375.
means of transient elastography. Radiology 2010;257:872–878. [92] Schwabl P, Bota S, Salzl P, Mandorfer M, Payer BA, Ferlitsch A, et al. New
[71] Simonetto DA, Yang HY, Yin M, de Assuncao TM, Kwon JH, Hilscher M, et al. reliability criteria for transient elastography increase the number of accurate
Chronic passive venous congestion drives hepatic fibrogenesis via sinusoidal measurements for screening of cirrhosis and portal hypertension. Liver Int
thrombosis and mechanical forces. Hepatology 2015;61:648–659. 2015;35:381–390.
[72] Colecchia A, Di Biase AR, Scaioli E, Predieri B, Iughetti L, Reggiani ML, et al. [93] Kitson MT, Roberts SK, Colman JC, Paul E, Button P, Kemp W. Liver stiffness
Non-invasive methods can predict oesophageal varices in patients with and the prediction of clinically significant portal hypertension and portal
biliary atresia after a Kasai procedure. Dig Liver Dis 2011;43:659–663. hypertensive complications. Scand J Gastroenterol 2015;50:462–469.
[73] Uchida H, Sakamoto S, Kobayashi M, Shigeta T, Matsunami M, Sasaki K, et al. [94] Cho EJ, Kim MY, Lee JH, Lee IY, Lim YL, Choi DH, et al. Diagnostic and
The degree of spleen stiffness measured on acoustic radiation force impulse prognostic values of noninvasive predictors of portal hypertension in
elastography predicts the severity of portal hypertension in patients with patients with alcoholic cirrhosis. PLoS One 2015;10:e0133935.
biliary atresia after portoenterostomy. J Pediatr Surg 2015;50:559–564. [95] Zykus R, Jonaitis L, Petrenkiene V, Pranculis A, Kupcinskas L. Liver and spleen
[74] Garcia-Tsao G, Groszmann RJ, Fisher RL, Conn HO, Atterbury CE, Glickman M. transient elastography predicts portal hypertension in patients with chronic
Portal pressure, presence of gastroesophageal varices and variceal bleeding. liver disease: a prospective cohort study. BMC Gastroenterol 2015;15:183.
Hepatology 1985;5:419–424. [96] Sharma P, Kirnake V, Tyagi P, Bansal N, Singla V, Kumar A, et al. Spleen
[75] Groszmann RJ, Garcia-Tsao G, Bosch J, Grace ND, Burroughs AK, Planas R, stiffness in patients with cirrhosis in predicting esophageal varices. Am J
et al. Beta-blockers to prevent gastroesophageal varices in patients with Gastroenterol 2013;108:1101–1107.
cirrhosis. N Engl J Med 2005;353:2254–2261. [97] Wong GL, Kwok R, Chan HL, Tang SP, Lee E, Lam TC, et al. Measuring spleen
[76] Ripoll C, Groszmann R, Garcia-Tsao G, Grace N, Burroughs A, Planas R, et al. stiffness to predict varices in chronic hepatitis B cirrhotic patients with or
Hepatic venous pressure gradient predicts clinical decompensation in without receiving non-selective beta-blockers. J Dig Dis 2016;17:538–546.
patients with compensated cirrhosis. Gastroenterology 2007;133:481–488. [98] Vermehren J, Polta A, Zimmermann O, Herrmann E, Poynard T, Hofmann WP,
[77] Berzigotti A, Reig M, Abraldes JG, Bosch J, Bruix J. Portal hypertension and the et al. Comparison of acoustic radiation force impulse imaging with transient
outcome of surgery for hepatocellular carcinoma in compensated cirrhosis: a elastography for the detection of complications in patients with cirrhosis.
systematic review and meta-analysis. Hepatology 2015;61:526–536. Liver Int 2012;32:852–858.
[78] Bruix J, Castells A, Bosch J, Feu F, Fuster J, Garcia-Pagan JC, et al. Surgical [99] Rifai K, Cornberg J, Bahr M, Mederacke I, Potthoff A, Wedemeyer H, et al. ARFI
resection of hepatocellular carcinoma in cirrhotic patients: prognostic value elastography of the spleen is inferior to liver elastography for the detection of
of preoperative portal pressure. Gastroenterology 1996;111:1018–1022. portal hypertension. Ultraschall Med 2011;32(Suppl 2):E24–30.
[79] Ripoll C, Groszmann RJ, Garcia-Tsao G, Bosch J, Grace N, Burroughs A, et al. [100] Mori K, Arai H, Abe T, Takayama H, Toyoda M, Ueno T, et al. Spleen stiffness
Hepatic venous pressure gradient predicts development of hepatocellular correlates with the presence of ascites but not esophageal varices in chronic
carcinoma independently of severity of cirrhosis. J Hepatol 2009;50: hepatitis C patients. Biomed Res Int 2013;2013:857862.
923–928. [101] Bota S, Sporea I, Sirli R, Focsa M, Popescu A, Danila M, et al. Can ARFI
[80] Merkel C, Bolognesi M, Bellon S, Zuin R, Noventa F, Finucci G, et al. Prognostic elastography predict the presence of significant esophageal varices in newly
usefulness of hepatic vein catheterization in patients with cirrhosis and diagnosed cirrhotic patients? Ann Hepatol 2012;11:519–525.
esophageal varices. Gastroenterology 1992;102:973–979. [102] Kim HY, Jin EH, Kim W, Lee JY, Woo H, Oh S, et al. The role of spleen stiffness
[81] Berzigotti A, Rossi V, Tiani C, Pierpaoli L, Zappoli P, Riili A, et al. Prognostic in determining the severity and bleeding risk of esophageal varices in
value of a single HVPG measurement and Doppler-ultrasound evaluation in cirrhotic patients. Medicine (Baltimore) 2015;94:e1031.
patients with cirrhosis and portal hypertension. J Gastroenterol 2011;46: [103] Rizzo L, Attanasio M, Pinzone MR, Berretta M, Malaguarnera M, Morra A,
687–695. et al. A new sampling method for spleen stiffness measurement based on
[82] Merkel C, Marin R, Sacerdoti D, Donada C, Cavallarin G, Torboli P, et al. Long- quantitative acoustic radiation force impulse elastography for noninvasive
term results of a clinical trial of nadolol with or without isosorbide assessment of esophageal varices in newly diagnosed HCV-related cirrhosis.
mononitrate for primary prophylaxis of variceal bleeding in cirrhosis. Biomed Res Int 2014;2014:365982.
Hepatology 2000;31:324–329. [104] Ye XP, Ran HT, Cheng J, Zhu YF, Zhang DZ, Zhang P, et al. Liver and spleen
[83] Stanley AJ, Robinson I, Forrest EH, Jones AL, Hayes PC. Haemodynamic stiffness measured by acoustic radiation force impulse elastography for

parameters predicting variceal haemorrhage and survival in alcoholic noninvasive assessment of liver fibrosis and esophageal varices in patients
cirrhosis. QJM 1998;91:19–25. with chronic hepatitis B. J Ultrasound Med 2012;31:1245–1253.
[84] Moitinho E, Escorsell A, Bandi JC, Salmeron JM, Garcia-Pagan JC, Rodes J, et al. [105] Merchante N, Rivero-Juarez A, Tellez F, Merino D, Jose Rios-Villegas M,
Prognostic value of early measurements of portal pressure in acute variceal Marquez-Solero M, et al. Liver stiffness predicts clinical outcome in human
bleeding. Gastroenterology 1999;117:626–631. immunodeficiency virus/hepatitis C virus-coinfected patients with compen-
[85] Abraldes JG, Villanueva C, Banares R, Aracil C, Catalina MV, Garci APJC, et al. sated liver cirrhosis. Hepatology 2012;56:228–238.
Hepatic venous pressure gradient and prognosis in patients with acute [106] Grgurevic I, Bokun T, Mustapic S, Trkulja V, Heinzl R, Banic M, et al. Real-time
variceal bleeding treated with pharmacologic and endoscopic therapy. J two-dimensional shear wave ultrasound elastography of the liver is a
Hepatol 2008;48:229–236. reliable predictor of clinical outcomes and the presence of esophageal
[86] Rincon D, Lo Iacono O, Ripoll C, Gomez-Camarero J, Salcedo M, Catalina MV, varices in patients with compensated liver cirrhosis. Croat Med J
et al. Prognostic value of hepatic venous pressure gradient for in-hospital 2015;56:470–481.
mortality of patients with severe acute alcoholic hepatitis. Aliment Pharma- [107] Wang JH, Chuah SK, Lu SN, Hung CH, Kuo CM, Tai WC, et al. Baseline and
col Ther 2007;25:841–848. serial liver stiffness measurement in prediction of portal hypertension
[87] Serste T, Bourgeois N, Lebrec D, Evrard S, Deviere J, Le Moine O. Relationship progression for patients with compensated cirrhosis. Liver Int 2014;34:
between the degree of portal hypertension and the onset of spontaneous 1340–1348.
bacterial peritonitis in patients with cirrhosis. Acta Gastroenterol Belg [108] Kim BK, Kim DY, Han KH, Park JY, Kim JK, Paik YH, et al. Risk assessment of
2006;69:355–360. esophageal variceal bleeding in B-viral liver cirrhosis by a liver stiffness
[88] Lemoine M, Katsahian S, Ziol M, Nahon P, Ganne-Carrie N, Kazemi F, et al. measurement-based model. Am J Gastroenterol 2011;106(1654–1662):1730.
Liver stiffness measurement as a predictive tool of clinically significant [109] Merchante N, Rivero-Juarez A, Tellez F, Merino D, Rios-Villegas MJ, Ojeda-
portal hypertension in patients with compensated hepatitis C virus or Burgos G, et al. Liver stiffness predicts variceal bleeding in HIV/HCV-
alcohol-related cirrhosis. Aliment Pharmacol Ther 2008;28:1102–1110. coinfected patients with compensated cirrhosis. AIDS 2016.

Journal of Hepatology 2017 vol. 67 j 399–411 411