If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Department of Visceral Surgery and Medicine, Inselspital, Bern University Hospital, University of Bern, BHH D115, Bern 3010, SwitzerlandGraduate School for Health Sciences (GHS), University of Bern, Switzerland
Non-invasive tests (NITs) and liver stiffness measurement (LSM) in particular, have entered clinical practice over 20 years ago as point-of-care tests to diagnose liver fibrosis in patients with compensated chronic liver disease. Since then, NITs use has evolved thanks to a large number of studies in all major etiologies of liver disease, and they have become important tools to stratify the risk of portal hypertension and liver-related events. The Baveno VII consensus workshop provided several novel recommendations regarding the use of well-established and novel NITs in the specific setting of portal hypertension screening, diagnosis and follow-up. The Baveno VII expert panels paid special attention to summarizing the existing data into simple clinical rules able to guide clinicians in their practice. The “rule of five” for LSM is a tool to stratify the risk of liver-related events, and LSM alone or in combination with platelet count, can be used now to rule-in and rule-out compensated advanced chronic liver disease (cACLD) and clinically significant portal hypertension, as well as to rule-out high-risk varices. Use of NITs in obese subjects with non-alcoholic fatty liver disease (NAFLD) and patients with viral hepatitis C that has been successfully treated, require specific knowledge. This review will update the reader on these aspects.
The introduction of non-invasive tests (NITs) has been a key advance for staging liver fibrosis. In addition, NITs currently provide prognostic value beyond the stage of fibrosis. Portal hypertension (PH) plays a pivotal role in the progression from the compensated to the decompensated stage of chronic liver disease (CLD) and consequently holds strong prognostic value to predict clinical decompensation [
Portal hypertensive bleeding in cirrhosis: risk stratification, diagnosis, and management: 2016 practice guidance by the American Association for the study of liver diseases.
]. The gold-standard method to assess and stage portal hypertension in compensated advanced chronic liver disease (cACLD) is the measurement of hepatic venous pressure gradient (HVPG) [
]; and once the HVPG exceeds 10 mmHg PH is defined as “Clinically Significant Portal Hypertension (CSPH)”. Patients with CSPH may develop esophageal varices [
Portal hypertensive bleeding in cirrhosis: risk stratification, diagnosis, and management: 2016 practice guidance by the American Association for the study of liver diseases.
Expanding consensus in portal hypertension: report of the Baveno VI consensus workshop: stratifying risk and individualizing care for portal hypertension.
β blockers to prevent decompensation of cirrhosis in patients with clinically significant portal hypertension (PREDESCI): a randomised, double-blind, placebo-controlled, multicentre trial.
], the paradigm of treatment of portal hypertension has shifted from a bleeding-centric view (prevention of bleeding and rebleeding) to a much more comprehensive view. According to the new consensus, portal hypertension should be treated early, namely as soon as it can be proven, in order to avoid decompensation [
]. This makes the use of NITs to be used as surrogates of HVPG to assess the presence of CSPH in patients with compensated ACLD very attractive. This has been the subject of several studies over the last 10 years, leading to refined concepts on NITs use, and to the development of novel tools. In the recent Baveno VII consensus workshop, one of the sessions was devoted to assessment of PH, and the Panel on NITs extensively reviewed the literature and generated new recommendations, which will be discussed in the present review.
2. Non-invasive tools for cACLD
The term advanced chronic liver disease, ACLD refers to patients with late stages of chronic liver disease (CLD) [
]. In the Baveno VI consensus workshop on portal hypertension in 2015, the term “compensated advanced chronic liver disease” (cACLD) was promoted based on NITs [
Expanding consensus in portal hypertension: report of the Baveno VI consensus workshop: stratifying risk and individualizing care for portal hypertension.
], allowing the early identification of advanced liver disease at an asymptomatic stage. This term aimed at covering the full spectrum of patients with severe liver fibrosis (bridging fibrosis) on histology and those with compensated cirrhosis [
Expanding consensus in portal hypertension: report of the Baveno VI consensus workshop: stratifying risk and individualizing care for portal hypertension.
. In the Baveno VII consensus conference in 2021, the emphasis of the definition shifted from histological diagnosis to an even more pragmatic application, using the prognostic value of NITs to define cACLD, allowing accurate risk stratification regardless of the histological stage [
]. Accordingly, patients having liver stiffness measurement (LSM) by transient elastography (TE) >15 kPa are considered at high likelihood of cACLD in all etiologies; LSM values between 10 and 15 kPa are suggestive of cACLD, and LSM <10 kPa rules-out cACLD in the absence of other clinical/imaging signs [
]. The rationale of this definition comes from a thorough review of the most recent evidence that showed that CLD patients with LSM <10 kPa by TE have a very low 3-year risk (≤1%) of liver-related events (LRE), and the 3-year risk of LRE increases substantially between five to ten times with LSM >15 kPa, irrespective of CLD etiology [
Monitoring Occurrence of liver-related events and survival by transient elastography in patients with nonalcoholic fatty liver disease and compensated advanced chronic liver disease.
]. Meta-analyses regarding the accuracy of point shear wave elastography (pSWE) and two dimensional shear wave elastography (2D-SWE) for liver fibrosis staging in comparison to liver biopsy, including mixed etiologies, showed that these techniques had accuracy similar to TE for advanced fibrosis detection with AUROCs >0.90 [
Comparison of laboratory tests, ultrasound, or magnetic resonance elastography to detect fibrosis in patients with nonalcoholic fatty liver disease: a meta-analysis.
]. Nonetheless, TE remains the technique for which the largest amount of evidence is available. As a prognostic indicator, the use of rule-of-five for the cut-off of LSM by TE (10-15-20-25 kPa) was recommended to estimate quickly and at the bedside the risk of decompensation and liver related-death regardless of the ACLD etiology (Fig. 1) [
. In monitoring, LSM may be repeated every year in patients with cACLD and a clinically significant decrease in LSM was defined as any decrease to a LSM <10 kPa or decrease in LSM of ≥20% accompanied by LSM <20 kPa [
]. This definition is based on recent longitudinal studies showing a minimal risk of LRE in patients with follow-up LSM <10 kPa, while the risk of LRE and mortality remains high in patients with LSM >20 kPa [
Monitoring Occurrence of liver-related events and survival by transient elastography in patients with nonalcoholic fatty liver disease and compensated advanced chronic liver disease.
. Moreover, a study showed a significant reduction in the rate of LRE in patients with a decrease of LSM compared to patients with stable and increasing of LSM (3.8% vs. 6.2% vs. 14.4%) [
Monitoring Occurrence of liver-related events and survival by transient elastography in patients with nonalcoholic fatty liver disease and compensated advanced chronic liver disease.
Fig. 1Use of NITs according to the rule of five to rule-in and rule-out cACLD, CSPH and high-risk varices. Abbreviations. ALD, alcohol-related liver disease; cACLD, compensated advanced chronic liver disease; CSPH, clinically significant portal hypertension; NASH, non-alcoholic steatohepatitis.
3. Non-invasive tools for clinically significant portal hypertension (CSPH) and varices
Simple laboratory tests and imaging tests have a limited sensitivity to rule-in and rule-out CSPH; among signs of CSPH on imaging (ultrasound, computerized tomography, magnetic resonance imaging), the presence of porto-systemic collaterals deserves to be mentioned since: (a) they are pathognomonic of CSPH in cACLD, and (b) they hold a negative prognostic significance [
]. LSM emerged as a significant advance to stratify the risk of CSPH, having a much higher accuracy than other existing NITs. The reliability of LSM by TE to identify the presence of CSPH has been assessed in patients with cACLD due to different etiologies, showing a correlation coefficient ranging between 0.55–0.82 [
]. In 2015, the Baveno VI consensus stated that a LSM >20–25 kPa can be used to identify the presence of CSPH in cACLD patients with untreated hepatitis C (HCV) or hepatitis B (HBV) [
Expanding consensus in portal hypertension: report of the Baveno VI consensus workshop: stratifying risk and individualizing care for portal hypertension.
]. Subsequently, several studies and two meta-analyses confirmed the good performance of these cut-offs for diagnosing CSPH in patients with cACLD owing to different causes [
The VITRO score (Von Willebrand factor antigen/thrombocyte ratio) as a new marker for clinically significant portal hypertension in comparison to other non-invasive parameters of fibrosis including ELF test.
Reliability of transient elastography-based liver stiffness for diagnosing portal hypertension in patients with alcoholic liver disease: a diagnostic meta-analysis with specific cut-off values.
Correlation of transient elastography with hepatic venous pressure gradient in patients with cirrhotic portal hypertension: a study of 326 patients from India.
. Then, the recent Baveno VII consensus recommend LSM ≤15 kPa plus platelets ≥150 × 10/L to rule-out CSPH in the majority of etiologies and LSM ≥25 kPa was the best cutoff to rule in CSPH (specificity and positive predictive value >90%) in alcoholic liver disease, chronic hepatitis B, chronic hepatitis C, and non-obese patients with non-alcoholic steatohepatitis (NASH). However, in obese patients with NASH, the positive predictive value was lower (62.8%) [
]. The ANTICIPATE NASH model, including LSM by TE, platelet count and body mass index (BMI), can be used in patients with NASH cACLD and obesity to predict the risk of CSPH, although more validation is required [
]. In patients with intermediate values of LSM between 15 and 25 kPa (“gray zone”), the ANTICIPATE study has provided a model to predict CSPH based on LSM and platelet count, which might be used as an additional tool to further improve risk stratification for CSPH according to the Baveno VII consensus [
The above-mentioned recommendations are in agreement with the EASL clinical practice guidelines, namely that LSM should be used to diagnose CSPH in patients with cACLD [
European Association for the Study of the L, List of panel m Easl clinical practice guidelines (Cpgs) on non-invasive tests for evaluation of liver disease severity and prognosis- 2021 update.
]. This has an important implication for clinical practice regarding to the use of nonselective beta-blockers (NSBBs) in patients with CSPH, which have been recently recommended since the findings of the PREDESCI study. This study provided evidence that the clinical decompensation was significantly lower with NSBB use versus placebo (from 27% to 17% over a median follow-up of 37 months: hazard ratio [HR] 0.51, 95% confidence interval [CI] 0.26–0.97) in patients with cACLD and CSPH with no or small varices. The main impact was a reduction in the incidence of ascites from 20% to 9% with the use of NSBB [
β blockers to prevent decompensation of cirrhosis in patients with clinically significant portal hypertension (PREDESCI): a randomised, double-blind, placebo-controlled, multicentre trial.
]. Therefore, this data prompted researchers in Baveno VII to refine the recommendations. It is now suggested that in cACLD patients with NITs indicating presence of CSPH, independently of the presence of varices, treatment with NSBBs (preferably carvedilol as it is more effective at reducing HVPG) should be considered in order to prevent first clinical decompensation [
Regarding NITs for predictions of varices, a milestone recommendation of the previous Baveno Consensus (Baveno VI) regarded the use of NITs to rule-out high-risk varices, so allowing safely avoiding endoscopy. The simple combination of. platelet count >150 × 109/L and LSM <20 kPa, could be applied to identify patients with cACLD and a very low risk (<5%) of high-risk varices, [
Expanding consensus in portal hypertension: report of the Baveno VI consensus workshop: stratifying risk and individualizing care for portal hypertension.
]. Following the publication of these “Baveno VI Criteria”, a large number of studies validated this clinical rule in various etiologies. Some proposed to further refine the use of simple NITs by modeling them to better estimate individual risks, e.g. with the support of nomograms, and some additional ones suggested to expand them by increasing the LSM threshold and/or decreasing the platelet count threshold [
]. In a very recent meta-analysis of 28 studies, the Baveno VI criteria were fully validated; they displayed a pooled 99% negative predictive value for ruling out high-risk varices. Furthermore, a suboptimal performance was displayed when only 16 studies assessing the expanded Baveno VI criteria, which misclassified a significant number of patients, not fulfilling the required safety cut-off (<5% of missed high-risk varices) [
]. Therefore, the classical Baveno VI criteria remain the standard for identifying ACLD patients with very low probability of high-risk varices and who do not need screening gastroscopy. As the paradigm has shifted, and NSBB treatment is indicated for all patients with cACLD and CSPH, the Baveno VI criteria may only remain relevant in the subgroup of patients with contraindications or intolerance to NSBB, in whom endoscopy will be needed not only for screening, but also for potential therapy of high-risk varices in primary prophylaxis.
4. Spleen stiffness
Spleen stiffness measurement (SSM) has been recently proven as a more direct surrogate of PH as compared to LSM. SSM reflects augmented intrasplenic congestion and pressure due to splenic outflow obstruction, enlargement and hyperactivation of the splenic lymphoid tissue, as well as enhanced angiogenesis and fibrogenesis consequent to PH [
]. SSM is a sensitive NIT for CSPH, because while LSM mostly takes into account the fixed component of intrahepatic resistance, SSM likely additionally reproduces the increased portal flow associated with hyperdynamic splanchnic circulation [
], as well as in liver diseases with a pre-sinusoidal component, such as cholestatic liver diseases, and possibly, as recently suggested, to non-alcoholic liver disease (NAFLD) [
. A meta-analysis of 16 studies has shown that sensitivity and specificity of SSM was superior to LSM to diagnose oesophageal varices (SSM Sens 0.88, Spec 0.78 vs. LSM Sens 0.83, Spec 0.66) [
]. Regarding CSPH, in a meta-analysis of nine studies, spleen US elastography correlated well with HVPG, detecting CSPH with a sensitivity and specificity of 0.88 and 0.92 [
. In addition, SSM holds prognostic information, e.g. it may outperform LSM to predict patients who will develop a first variceal bleeding, and predicts a first clinical decompensation with better accuracy than LSM [
Measurement of spleen stiffness with acoustic radiation force impulse imaging predicts mortality and hepatic decompensation in patients with liver cirrhosis.
Some studies have underlined the beneficial use of SSM in addition to the Baveno VI criteria to further decrease the proportion of patients safely skipping screening endoscopy [
Expanding consensus in portal hypertension: report of the Baveno VI consensus workshop: stratifying risk and individualizing care for portal hypertension.
]. In a study including a prospective external validation cohort, Colecchia et al. showed that the combination of Baveno VI criteria and SSM ≤46 kPa model would have safely spared 37.4% of endoscopies, compared to 16.5% when using the Baveno VI criteria alone [
One of major limitations of SSM by TE, until recently, was that it is only applicable in about 70% of cases. The high failure rate is linked to absence of splenomegaly. Additionally, SSM by TE using the liver 50 Hz module currently reaches a maximum of 75 kPa. Broadening the range to 150 kPa with appropriate software modifications has been suggested and tested [
Modified spleen stiffness measurement by transient elastography is associated with presence of large oesophageal varices in patients with compensated hepatitis C virus cirrhosis.
. The novel 100 Hz spleen specific module was compared to the standard non-spleen specific 50 Hz module by TE. The 100 Hz SSM in combination with the Baveno VI criteria showed the best performance, sparing 38.1% of endoscopies, as compared to 26.5% of SSM 50 Hz + Baveno VI criteria and 8.1% of Baveno VI criteria alone [
In light of the growing evidence of SSM in CSPH and high-risk varices detection, the Baveno VII consensus has highlighted the role of SSM in CSPH cACLD due to viral hepatitis (untreated HCV; untreated and treated HBV). SSM can be now used routinely to rule-out and rule-in CSPH (SSM <21 kPa and SSM >50 kPa, respectively) [
Regarding treatment of CSPH, in patients who are not candidates for NSBBs fulfilling Baveno VI criteria, a SSM ≤40 kPa by TE can identify those at low probability of high-risk varices, avoiding endoscopy. Areas of further research in this field include the validation of the best cut-off using a 100 Hz specific TE-probe, as well other ultrasound elastography methods, and further validation of SSM in non-viral etiologies [
Invasive procedures such as liver biopsy and HVPG measurements are not suited to frequent use due to their costs and limitations, which limits their utility in the monitoring of cACLD. NITs, on the opposite, are repeatable and are acceptable to the vast majority of patients. NITs are being increasingly tested, not only as diagnostic, but also as prognostic biomarkers [
]. While the prognostic value of a single measurement of NITs, LSM in particular, has been well proven, less data is available on the value of changes over time (dynamic use of NITs).. A summary of available evidence is highlighted in Table 1.
Table 1Significance of dynamic changes in non-invasive tests in chronic liver disease.
Author
Year
Study design
Study population
N
Non-invasive tests
Follow-up
Key points
NAFLD
Pons et al.
2016
Retrospective cohort
Patients with cACLD, with baseline LSM ≥10 kPa, Child-Pugh score 5 and without previous decompensation
94
LSM by TE
43.6 months (median)
Both baseline LSM (OR 1.12, p < 0.01) and increase in LSM (OR 1.02, P < 0.05) were independent predictors for death, liver decompensation and impairment in at least 1 point in Child-Pugh score during follow-up. High-risk population defined by baseline LSM ≥21 kPa and increase in LSM ≥10% (OR 47.1%, 95% CI: 23–71%).
Hagström et al.
2020
Retrospective analysis of a prospective cohort
Participants with two FIB-4 measurements in a population-based Swedish cohort
40,729
FIB-4
2.4 years (mean)
Increase of 1 unit in FIB-4 associated with elevated risk of severe liver disease (aHR 1.81; 95% CI 1.67–1.96). Transitioning from low or intermediate to a high-risk group during follow-up associated with increased risk of severe liver disease (aHR 7.99 and 8.64, respectively), compared to a consistently low-risk group.
Petta et al.
2021
Multicenter retrospective cohort
NAFLD patients and histologically confirmed F3–F4 fibrosis and/or LSM by TE>10 kPa
533
LSM by TE
35 months (median)
Increase in LSM independently associated with elevated risk of hepatic decompensation (HR, 1.56; 95% CI 1.05–2.51), HCC (HR 1.72, 95% CI 1.01–3.02), overall mortality (HR 1.73, 95% CI 1.11–2.69), and liver-related mortality (HR 1.96, 95% CI 1.10–3.38).
Siddiqui et al.
2019
Retrospective cohort
NAFLD patients with 2 biopsies and accompanying laboratory data
292
FIB-4, NFS, APRI, FIB-4,AST/ALT ratio
2.6 years (median)
Changes in FIB-4 (c-statistics 0.81, 95% CI 0.73–0.81), APRI (0.82, 95% CI 0.74–0.89), and NFS (0.80, 95% CI 0.71–0.88 can detect progression to advanced fibrosis in patients with NAFLD.
Younossi et al.
2021
Nested prospective analysis in 4 randomized controlled trials
Patients with advanced NASH (NASH Clinical Research Network stage F3 or F4) from 4 multinational clinical trials of simtuzumab and selonsertib.
2154
ELF, NFS, FIB-4, LSM by TE, Fibrotest
16 months (median)
Increase in all NIT associated with elevated risk of histologic progression to cirrhosis or liver-related events in the F3 group (p < 0.01). Increase in ELF, NFS, FIB-4, and LSM associated with an increased risk of liver-related events in the F4 group (p < 0.01).
Cholestatic and Autoimmune liver disease
Corpechot et al.
2014
Prospective cohort
Patients with PSC with any fibrosis stage
142
LSM by TE
3.9 years (mean)
Increase in LSM independently associated with elevated risk of death, liver transplant, ascites, hepatic encephalopathy, gastrointestinal bleeding related to portal hypertension, cholangiocarcinoma, or HCC (large-duct PSC: HR 7.3, 95% CI 2.9–18.1, overall population including PSC-AIH overlap: HR 11.9,95% CI, 5.2–27.4).
Corpechot et al.
2012
Prospective cohort
Patients with PBC with any fibrosis stage
150
LSM by TE
2.6 years (mean)
Progression of liver stiffness in PBC is predictive of death, liver transplant, or liver decompensation including ascites, variceal bleeding, hepatic encephalopathy, HCC, doubling of total serum bilirubin level above 6 mg/dL, or minimal criteria for liver transplant (HR: 1.3; 95% CI: 1.2–1.5).
Hartl et al.
2018
Prospective cohort
Patients with AIH with any fibrosis stage
125
LSM by TE
2.7 years (mean)
Complete biochemical remission is a reliable predictor of a good prognosis in AIH and leads to fibrosis regression that can be monitored by LSM. Patients in complete biochemical remission of AIH showed a considerable decrease in LSM (7.5%/year; 95% CI 11% to 2.0%; p < 0.01), whereas patients without complete biochemical remission no statistically significant change in LSM.
HCV before and after sustained virological response
Mandorfer et al.
2016
Retrospective cohort
HCV patients with CSPH at baseline prior to DAA therapy
60
LSM by TE
217 days (median)
Excellent diagnostic accuracy for CSPH at a cut-off of 25.3 kPa follow-up LSM (AUROC 0.93, 95% CI 0.90–1.00), at a cut-off of 27.2 kPa or baseline TE (AUROC 0.90, 95% CI 0.82–0.98). Absolute and relative LSM changes from baseline not accurate enough. Optimized baseline and follow-up LSM cut-offs were 18.8 and 12.4 kPa to rule-out CSPH after DAA therapy.
Pons et al.
2020
Prospective cohort
Patients with HCV and cACLD after DAA treatment
572
LSM by TE, albumin (serum)
2.8 years (median)
Baseline LSM ≥20 kPa with no LSM decrease during follow-up associated with increased risk of liver decompensation (HR 39.7; 95% CI 4.4–355.4). Albumin levels at follow-up (HR 0.08, 95% CI 0.02–0.25) and LSM <10 kPa at follow-up (HR 0.33, 95% CI 0.11–0.96) independently associated with a decreased HCC risk.
Vergniol et al.
2014
Prospective cohort
Patients with chronic HCV of any fibrosis stage (44% F2-F4)
1025
LSM by TE, APRI, FIB-4
38 months (median)
LSM/FIB-4 at baseline, change in LSM/FIB-4 and SVR independently predicted survival after DAA treatment.
Vutien et al.
2020
Retrospective cohort
HCV patients with at least one liver stiffness before (n = 492) or after therapy
124 (longitudinal)
LSM by TE
27.3 months (median)
Post-treatment LSM >20 kPa, associated with increased risk of decompensated cirrhosis (adjusted HR 3.85, 95% CI 1.29–11.50) and the composite outcome of death, liver transplant, decompensated cirrhosis or HCC (adjusted HR 1.95, 95% CI: 1.07–3.56), compared to ≤12.5 kPa. Increasing or stable LSM post-treatment associated with significant association with death or liver transplant (adjusted HR 7.93, 95% CI 1.59–39.47) and the composite outcome (adjusted HR 4.83, 95% CI 1.12–20.86). No significant associations between pre-treatment liver stiffness and any outcomes on multivariable analysis.
Ravaioli et al.
2018
Retrospective cohort
DAA-cured HCV patients with cirrhosis
139
LSM by TE
15 months (median, after end of DAA)
Decrease in LSM significantly lower in patients with new HCC (−18.0% vs. −28.9% p < 0.05) than in controls. Change in LSM < −30% independently associated with HCC development.
Lens et al.
2017
Prospective cohort
DAA-cured HCV patients with cirrhosis and CSPH
226
LSM by TE
Maximum 6 months prior to, follow-up 24 weeks after treatment
1/3 of patients with LSM< 13.6 kPa after SVR with ongoing CSPH. Cut-off at 13.6 kPa is not reliable enough for ruling out CSPH after SVR (Sensitivity 88%, Specificity 54%, positive predicate value 87%, negative predictive value: 57%). Higher baseline HVPG and a lower decrease in LSM after treatment associated with persisting CSPH after SVR. Changes in TE do not correlate with HVPG and
One year after SVR, LSM and ELF showed an excellent diagnostic accuracy to rule out advanced fibrosis (LSM >10.6 kPa, AUROC 0.90, 95% CI 0.84–0.0.96; ELF >10.83, AUROC 0.88, 95% CI 0.79–0.98) and CSPH (LSM > 11.3 kPa, AUROC 0.88, 95% CI 0.80–0.98). ELF showed a fair diagnostic accuracy at a cut-off > 10.25 for CSPH
Piedade et al.
2021
Retrospective cohort
HCV patients with TE ≥ 10 kPa at baseline and serial measurements before DAA and after SVR
456
LSM by TE
2.3 years (median)
LSM decrease ≥ 20% after SVR decreases the risk of liver-related events or death (HR = 0.45, 95% CI 0.21–1.02)
HBV
Kim et al.
2013
Prospective cohort
Patients with histologically F3 or F4 fibrosis due to HBV receiving antiviral therapy
103
LSM by TE
6 months
Changes in LSM significantly correlated with liver-related events (ascites, variceal bleeding, hepatic encephalopathy, spontaneous bacterial peritonitis, hepatorenal syndrome, HCC and liver–related death). Increased LSM > 11.6 kPa at both baseline and follow-up showed the highest incidence (11.05% per person-year) and those with consistent LSM < 11.6 kPa the lowest incidence (1.22% per person-year)
Ye et al.
2021
Prospective cohort
Treatment-naive HBV patients with decompensated cirrhosis awaiting antiviral treatment at baseline
149
LSM by 2D-SWE
34.8 months
Last follow-up LSM (HR 1.11, 95% CI 1.04–1.18) was the only independent risk factor for the occurrence of liver-related events (spontaneous bacterial peritonitis, variceal bleeding, hepatorenal syndrome, hepatopulmonary syndrome, and HCC) rather than pre-treatment or dynamic changes in LSM.
Kim et al.
2018
Prospective cohort
Patients with HBV-related advanced fibrosis or cirrhosis
209
LSM by TE
2 years
LSM < 11.6 kPa after two years of antiviral therapy was independently associated with a lower risk of HCC development (HR = 0.49, 95% CI 0.23–0.92)
Liu et al.
2020
Retrospective cohort
Patients with HBV-related HCC awaiting operation
158
LSM by TE
12 months
LSM changes were independent factors associated with overall survival (HR 1.89). LSM changes were independent factors for HCC-free survival (HR 1.52).
Both baseline LSM and an increase in LSM by TE were found to be independent predictors of death, liver decompensation and increase of at least 1 point in Child-Pugh score in a retrospective cohort with cACLD of any etiology (OR 1.12 and 1.02, respectively, both p < 0.05) [
]. Interestingly, the combination of baseline LSM ≥ 21 kPa and a LSM increase ≥10% was associated with a 47-fold risk increase for disease progression in the same study [
Furthermore, changes in FIB-4 were associated with an increased risk of future severe liver disease in a population-based Swedish cohort including 40,729 measurements [
]. Increases from low-intermediate to high-risk FIB-4 categories were associated with a substantially increased risk of progression to cirrhosis (adjusted HR 7.99 and 8.64, respectively), and belonging to a high-risk FIB-4 category both at baseline and on follow-up further increased the magnitude of this observation (adjusted HR 17.04; 95% CI 11.67–24.88) [
In a multicenter retrospective analysis of 533 patients with cACLD in NAFLD with a median follow-up of 35 months, an increase in LSM of at least 20% was independently associated with a 56–96% relative risk increase of hepatic decompensation, HCC, liver-related and all-cause mortality [
Monitoring Occurrence of liver-related events and survival by transient elastography in patients with nonalcoholic fatty liver disease and compensated advanced chronic liver disease.
]. In addition, increases in FIB-4, APRI, and NAFLD Fibrosis Score were found to accurately predict progression to advanced liver fibrosis in a retrospective cohort study, as compared to serial liver biopsies [
The association of histologic and noninvasive tests with adverse clinical and patient-reported outcomes in patients with advanced fibrosis due to nonalcoholic steatohepatitis.
], increases in all NITs (including Enhanced Liver Fibrosis (ELF) score and LSM) were associated with histologic progression to cirrhosis and/or liver-related events in F3 patients. All NITs, but FibroTest, were associated with development of liver-related events in patients showing cirrhosis at baseline. Furthermore, in a prospective cohort study including 142 patients with primary sclerosing cholangitis (PSC), and serial LSM measurements, an increase in LSM was independently associated with a 7 to almost 12-fold increased risk of liver transplant, death, variceal bleeding, hepatic encephalopathy after a median follow-up of 3.4 years among patients with large-duct and any PSC with or without overlap, respectively [
Baseline values and changes in liver stiffness measured by transient elastography are associated with severity of fibrosis and outcomes of patients with primary sclerosing cholangitis.
]. Similarly, Corpechot et al. found increases in LSM to be associated with a 30% risk increase in all-cause mortality, liver transplant, or hepatic decompensation in a prospective cohort of 150 patients with primary biliary cholangitis (PBC) [
Decreases in LSM below the threshold of 12 kPa after effective antiviral therapy are associated with resolution of CSPH, and the current Baveno VII recommendations take into consideration this data (Fig. 2).
Fig. 2Comparison of the Baveno VI and VII milestones regarding noninvasive tools for compensated advanced liver disease and portal hypertension. Abbreviations. cACLD, compensated advanced chronic liver disease; CSPH, clinically significant portal hypertension; HVPG, hepatic venous pressure gradient; NASH, non-alcoholic steatohepatitis; NSBB, Nonselective beta-blockers; LSM, Liver stiffness measurement; TE, transient elastography.
In conclusion, measuring changes in NITs over time (dynamic use), seems to refine the prognostic ability of single measurements in cACLD and may help better stratify the risk of liver-related complications and liver-related and all-cause mortality. Therefore, recent Baveno VII guidelines recommend monitoring LSM every 12 months [
]. Furthermore, a decrease in LSM by ≥20% was defined as clinically significant due to a substantially reduced risk of decompensation and liver-related death.
6. Etiology specific aspects
6.1 Non-alcoholic steatohepatitis (NASH)
The Baveno VII statement acknowledges that NASH, compared to other etiologies of liver disease, presents important differences with regard to non-invasive diagnostic and prognostic assessment [
]. Presence of CSPH at baseline, defined as an HVPG ≥10 mmHg, was found to be associated with a higher rate of liver-related events during a 24 month follow-up period in NASH patients with bridging fibrosis and compensated cirrhosis compared to patients without CSPH (HR, 2.83; 95% CI, 1.33–6.02; p = 0.007) [
Decompensation in advanced nonalcoholic fatty liver disease may occur at lower hepatic venous pressure gradient levels than in patients with viral disease.
]. Previous research also indicates that NASH patients show lower overall wedged hepatic venous pressure (WHVP) and HVPG measurements across fibrosis stages, compared to other etiologies (3.4 ± 2.4 vs. 7.5 ± 11 mm Hg/stage; p = 0.01) [
]. These findings support the presence of more severe liver disease despite detection of comparatively low HVPG values in NASH-cirrhosis. Conversely, HVPG values in line with presence of CSPH (≥10 mmHg) have been described in individuals with NASH without histological fibrosis [
]. For one, it is hypothesized that HVPG may be less accurate in assessing portal pressure in NASH. In a case-control study of decompensated cirrhotic patients undergoing Transjugular intrahepatic portosystemic shunt (TIPS), WHVP underestimated portal pressure (PP) in the NASH group compared to matched patients with alcohol- or HCV-related cirrhosis [
. In a cohort of patients with different etiologies of cACLD defined according to the Baveno VI criteria (LSM ≥10 kPa), undergoing both LSM by TE and HVPG assessment, overall prevalence of CSPH was found to be markedly lower among obese NASH patients [
]. The derived ANTICIPATE-NASH model, including LSM, BMI and platelet count, to predict CSPH in NASH is newly recommended in the Baveno VII statement [
. Taking into consideration, however, that HVPG may not be an ideal reference standard for PP assessment in NASH, further validation of the ANTICIPATE-NASH as a prognostic score may be warranted.
Besides LSM, blood-based markers have demonstrated impaired test performance in sub-populations of NASH. The performance of several blood-based NITs for detection of different fibrosis stages has been shown to vary by degree of obesity [
Overall, risk stratification with commonly used non-invasive tools according to established cut-offs may be less reliable in the NASH population, especially those with obesity and diabetes.
Magnetic resonance elastography is being increasingly used in clinical trials, and has recently proven prognostic value for the prediction of clinical decompensation [
Liver stiffness on magnetic resonance elastography and the MEFIB index and liver-related outcomes in nonalcoholic fatty liver disease: a systematic review and meta-analysis of individual participants.
]. Its use in European centers is still limited by its cost and suboptimal availability.
6.2 Chronic hepatitis after removal of the etiologic agent
There is increasing knowledge on the role of NITs in patients who achieved viral suppression (HBV) or cure of the underlying viral infection (HCV) after antiviral therapy, and this topic has been discussed in detail in the Baveno VII consensus.
6.3 Chronic hepatitis C after DAA treatment
Most longitudinal studies agree on a significant decrease in LSM, FIB-4 and APRI after successful DAA treatment in the vast majority of patients achieving sustained virological response (SVR), while the amplitude and definition of predictors and sub-population at risk for progression of fibrosis, development of liver-related events/death and all-cause mortality is quite heterogeneous [
Non-invasive assessment of fibrosis regression and portal hypertension in patients with advanced chronic hepatitis C virus (HCV)-associated liver disease and sustained virologic response (SVR): 3 years follow-up of a prospective longitudinal study.
Dynamics of liver stiffness by transient elastography in patients with chronic hepatitis C virus infection receiving direct-acting antiviral therapy-results from the German hepatitis C-registry.
Portal pressure and liver stiffness measurements in the prediction of fibrosis regression after sustained virological response in recurrent hepatitis C.
Longitudinal assessment of hepatic fibrosis in responders to direct-acting antivirals for recurrent hepatitis C after liver transplantation using noninvasive methods.
The predictive value of NITs changes for HVPG variations and for clinical outcomes deserves special attention (Table 1).
Dynamic LSM values did not correlate with changes in HVPG in 226 patients with HCV cirrhosis 24 weeks after successful DAA treatment. Importantly, one third of patients, with a reduction in LSM to below 13.6 kPa after SVR, still had CSPH at this time point [
]. In a second retrospective study from a different center, however, in 226 patients with HCV cirrhosis and CSPH assessed by serial HVPG measurements, both follow-up and baseline LSM showed an excellent predictive value for persistent CSPH after SVR. Here again, absolute and relative changes in LSM from baseline were not accurate enough to rule out CSPH [
In HCV-infected liver transplant recipients undergoing antiviral treatment, both LSM and ELF at 1 year after SVR12 showed an excellent diagnostic accuracy to rule out advanced fibrosis (TE < 10.6 kPa, ELF < 10.83), but only LSM reliably ruled out CSPH (TE < 11.3 kPa) [
Portal pressure and liver stiffness measurements in the prediction of fibrosis regression after sustained virological response in recurrent hepatitis C.
]. Of note, SSM unlike LSM by TE and acoustic radiation force impulse (ARFI) did not significantly decrease in a prospective cohort including 54 patients with HCV-associated cirrhosis after a follow-up of 3 years after treatment [
Non-invasive assessment of fibrosis regression and portal hypertension in patients with advanced chronic hepatitis C virus (HCV)-associated liver disease and sustained virologic response (SVR): 3 years follow-up of a prospective longitudinal study.
In light of evidence showing that CSPH continues to decrease over time after SVR, and might stabilize only in the medium-long term, the above-mentioned, apparently contradictory data have been re-assessed at the Baveno VII consensus. An individual patient data meta-analysis was recently published [
]. The authors showed that among cACLD patients, the prevalence of CSPH decreased from 80% to 54%, and that the correlation between LSM and HVPG improves after SVR (r = 0.60 vs. 0.45 pre-treatment); the correlation between platelet count and HVPG remained unchanged. Combining post-treatment LSM/platelet count yielded a high diagnostic accuracy for post-treatment-CSPH (AUC: 0.884; 95%CI: 0.843–0.926). Post-treatment-LSM<12 kPa & platelet count >150 × 109/L excluded CSPH (sensitivity: 99.2%), while LSM ≥ 25 kPa was highly specific for CSPH (93.6%).
The Baveno VII recommendations hence suggest that cured HCV patients with LSM < 12 kPa and platelets >150 × 109/L post treatment could be discharged from follow-up for CSPH owing to the negligible risk of high-risk varices and hepatic decompensation [
Regarding clinical outcomes, only LSM > 20 kPa after SVR 12 was associated with increased risks for decompensated cirrhosis (adjusted HR 3.85 vs ≤12.5 kPa) and the composite endpoint of the latter, death, need for liver transplant, and HCC (adjusted HR 1.95) after 27.3 months in a retrospective analysis [
]. Another large prospective study identified LSM/FIB-4 at baseline, changes in LSM/FIB-4 and SVR as independent predictors for survival after DAA treatment [
]. Furthermore, baseline LSM ≥20 kPa with no improvement during follow-up was associated with a 40-fold increased risk of decompensated liver cirrhosis in a prospective cohort study [
6.4 Suppressed HBV infection following antiviral treatment
Evidence on the role of NIT after adequate HBV suppression has also been accumulating during the last years, which is reflected by recently published guidelines in the field [
], the dynamic use of NITs and its prognostic accuracy for liver-related events and mortality deserve further investigation in this population (Table 1).
Early LSM changes within 6 months of initiation of antiviral therapy were assessed in a prospective cohort including 103 HBV patients with cACLD. Using a threshold of 11.6 kPa at baseline and follow-up, the lowest risk for liver-related events was found in patients with consistently low LSM or decreasing LSM during follow-up, whereas the highest risk category was found in serial TE ≥ 11.6 kPa and those who changed from values below to above this threshold after antiviral treatment [
Repeated measurements of LSM using 2D-SWE predicted liver-related outcomes in a prospective cohort of 149 patients with decompensated HBV cirrhosis after a follow-up of 34.8 months [
]. There is very scarce data on the correlation of HVPG and LSM in the context of treated HBV, and this is a field for future research.
7. Conclusion
NITs use in the context of cACLD and portal hypertension continues to grow. The use of LSM as a simple tool to trigger NSBB initiation in patients with cACLD could potentially prevent a large number of decompensating events, if widely used. This is an attractive field for clinical research. However, several areas require further work. About 40–50% of patients belong to the “gray zone” of LSM 15–25 kPa, in which a precise estimation of the risk of CSPH is not possible. Whether SSM can be used to reduce the proportion of patients in this indeterminate group is currently matter of research, and pilot data suggest that this might be the case [
A combined Baveno VII and spleen stiffness algorithm to improve the non-invasive diagnosis of clinically significant portal hypertension in patients with compensated advanced chronic liver disease.
]. Further refining NITs in emerging etiologies such as NAFLD in obese subjects and patients with mixed etiologies is urgently needed; while combining unrelated NITs (e.g. FIB-4 and LSM, or LSM and SSM) might reduce misclassification of patients [
], a precise discrimination of patients with and without CSPH is still not achievable. How to exactly account for co-factors of progression of liver disease in the context of patients cured from their primary etiology remains an open field for research. Finally, novel, blood-based markers of CSPH to be used alone or combined to LSM (and SSM) to guide treatment, use of dynamic NIT values, and potential use of SSM to predict treatment response in the emerging etiologies of liver disease remain unmet needs in this field.
Declaration of Competing Interest
None declared.
Grant support
Yuly Mendoza and Naomi Lange received financial support from the Stiftung für Leberkrankheiten Bern - SwissLiver. Naomi Lange further received financial support from the Gottfried and Julia Bangerter-Rhyner Foundation and the Swiss Academy of Medical Sciences (SAMS).
All authors have commented on the manuscript and approved the final version.
References
Garcia-Tsao G.
Abraldes J.G.
Berzigotti A.
Bosch J
Portal hypertensive bleeding in cirrhosis: risk stratification, diagnosis, and management: 2016 practice guidance by the American Association for the study of liver diseases.
Expanding consensus in portal hypertension: report of the Baveno VI consensus workshop: stratifying risk and individualizing care for portal hypertension.
β blockers to prevent decompensation of cirrhosis in patients with clinically significant portal hypertension (PREDESCI): a randomised, double-blind, placebo-controlled, multicentre trial.
Monitoring Occurrence of liver-related events and survival by transient elastography in patients with nonalcoholic fatty liver disease and compensated advanced chronic liver disease.
Comparison of laboratory tests, ultrasound, or magnetic resonance elastography to detect fibrosis in patients with nonalcoholic fatty liver disease: a meta-analysis.
The VITRO score (Von Willebrand factor antigen/thrombocyte ratio) as a new marker for clinically significant portal hypertension in comparison to other non-invasive parameters of fibrosis including ELF test.
Reliability of transient elastography-based liver stiffness for diagnosing portal hypertension in patients with alcoholic liver disease: a diagnostic meta-analysis with specific cut-off values.
Correlation of transient elastography with hepatic venous pressure gradient in patients with cirrhotic portal hypertension: a study of 326 patients from India.
Measurement of spleen stiffness with acoustic radiation force impulse imaging predicts mortality and hepatic decompensation in patients with liver cirrhosis.
Expanding consensus in portal hypertension: report of the Baveno VI consensus workshop: stratifying risk and individualizing care for portal hypertension.
Modified spleen stiffness measurement by transient elastography is associated with presence of large oesophageal varices in patients with compensated hepatitis C virus cirrhosis.
The association of histologic and noninvasive tests with adverse clinical and patient-reported outcomes in patients with advanced fibrosis due to nonalcoholic steatohepatitis.
Baseline values and changes in liver stiffness measured by transient elastography are associated with severity of fibrosis and outcomes of patients with primary sclerosing cholangitis.
Decompensation in advanced nonalcoholic fatty liver disease may occur at lower hepatic venous pressure gradient levels than in patients with viral disease.
Liver stiffness on magnetic resonance elastography and the MEFIB index and liver-related outcomes in nonalcoholic fatty liver disease: a systematic review and meta-analysis of individual participants.
Non-invasive assessment of fibrosis regression and portal hypertension in patients with advanced chronic hepatitis C virus (HCV)-associated liver disease and sustained virologic response (SVR): 3 years follow-up of a prospective longitudinal study.
Dynamics of liver stiffness by transient elastography in patients with chronic hepatitis C virus infection receiving direct-acting antiviral therapy-results from the German hepatitis C-registry.
Portal pressure and liver stiffness measurements in the prediction of fibrosis regression after sustained virological response in recurrent hepatitis C.
Longitudinal assessment of hepatic fibrosis in responders to direct-acting antivirals for recurrent hepatitis C after liver transplantation using noninvasive methods.
A combined Baveno VII and spleen stiffness algorithm to improve the non-invasive diagnosis of clinically significant portal hypertension in patients with compensated advanced chronic liver disease.
00743-5&id=gr1.jpg){kind=link}
00743-5&id=gr2.jpg){kind=link}