Elsevier

Journal of Hepatology

Volume 65, Issue 5, November 2016, Pages 899-905
Journal of Hepatology

Research Article
Validation of the Baveno VI criteria to identify low risk cirrhotic patients not requiring endoscopic surveillance for varices

https://doi.org/10.1016/j.jhep.2016.06.021Get rights and content

Background & Aims

The Baveno VI guidelines propose that cirrhotic patients with a liver stiffness measurement (LSM) <20 kPa and a platelet count >150,000/μl can avoid screening endoscopy as their combination is highly specific for excluding clinically significant varices. The aim of the study was to validate these criteria.

Methods

Transient elastography data was collected from two institutions from 2006–2015. Inclusion criteria were a LSM ⩾10 kPa and an upper gastrointestinal endoscopy within 12 months, with a diagnosis of compensated chronic liver disease. Exclusion criteria were porto-mesenteric-splenic vein thrombosis and non-cirrhotic portal hypertension. Varices were graded as low risk (grade <2) or high risk (grade ⩾2).

Results

The study included 310 patients (169 (55%) hepatitis C, and 275 (89%) Child-Pugh A). Varices were present in 23% cases, with 5% prevalence of high risk varices. Overall 102/310 (33%) met the Baveno VI criteria. Within this group 11% had varices and 2% had high risk varices, representing 2/15 (13%) of all high risk varices. The Baveno VI criteria gave a sensitivity 0.87, specificity 0.34, positive predictive value 0.06, negative predictive value 0.98, positive likelihood ratio 1.31 and negative likelihood ratio 0.39. The AUROC for LSM and platelet count combined was 0.746.

Conclusions

The Baveno VI criteria performed well correctly identifying 98% of patients who could safely avoid endoscopy.

Lay summary

This study examines the effectives of a recent set of guidelines published by the Baveno VI conference, which states that patients with chronic liver disease and a low liver stiffness (<20 kPa) and high platelet count (>150) are at low risk of having varices and do not need a screening endoscopy. Varices are a complication of cirrhosis, confer a risk of serious bleeding, and can be diagnosed and treated by endoscopy. Our study reviewed the clinical records of patients who have had liver stiffness scans and endoscopy over a 9-year period at two hospitals. The results show that only about 2% of patients who meet the Baveno VI criteria will be miss-classified as not having varices.

Introduction

Gastroesophageal varices occur as a consequence of portal hypertension and are a major cause of morbidity and mortality due to the risk of haemorrhage. In cirrhosis raised portal pressures initially develop as a result of advanced fibrosis and deranged liver architecture, but as liver disease progresses additional haemodynamic factors, such as splanchnic vasodilatation and hyperdynamic circulation, become increasingly important [1]. Portal pressures have traditionally been measured using hepatic venous pressure gradient (HVPG), and an HVPG ⩾10 mmHg confers increased risk of developing gastroesophageal varices [2]. HVPG has been shown to correlate well with the presence and size of varices [3], however measuring portal pressures by HVPG is invasive and limited to the centres with the relevant expertise.

Over the last decade transient elastography (TE) has become a widely used, non- invasive measure of liver stiffness and fibrosis. Following initial studies showing its accuracy in diagnosing significant fibrosis its clinical applications have been widened. The use of TE as a surrogate marker of portal hypertension has been demonstrated by liver stiffness measurement (LSM) correlating well with portal pressures up to a HVPG of 10–12 mmHg [1], [4]. Subsequent data has shown that TE is of potential benefit in the non- invasive diagnosis of varices, especially when TE is combined with other markers such as platelet count and spleen size [5].

A major limitation to implementing these tests into clinical practice for diagnosing gastroesophageal varices has been an inadequate specificity. As a result the diagnostic strength of non- invasive investigations such as TE have not yet been sufficient to replace endoscopy in the diagnosis of varices [6] (EASL [7]), and all patients with cirrhosis currently require routine surveillance with frequent oesophagogastroduodenoscopy (OGD).

The promising sensitivity and negative predictive value of TE, especially in combination with other non- invasive markers, means these investigations may be more effective tools at identifying low risk cirrhotic patients who can be safely ‘ruled out’ of needing an endoscopy. The recent Baveno VI guidelines acknowledge this application and recommend that in patients with compensated advanced chronic liver disease (cACLD) a LSM <20 kPa and a platelet count >150,000 cells/μl have a very low risk of having varices requiring treatment and therefore do not require screening endoscopy. They advise longitudinal follow-up of such patients by annual repetition of TE and platelet count with the guidance that if liver stiffness increases or platelet count declines to within the recommended values, these patients should undergo screening OGD [8].

In this retrospective cross-sectional cohort study, we reviewed all patients over a nine-year period at two centres who have undergone clinical, laboratory, TE and endoscopic evaluation of portal hypertension. The primary aim was to validate the recently proposed Baveno VI criteria and assess their sensitivity at accurately identifying those patients who can safely avoid screening endoscopy. Secondary aims were to assess if the criteria had similar sensitivities across all aetiologies of chronic liver disease, given the majority of published data is from patients with viral hepatitis, and to identify if alternative LSM or platelet parameters should be recommended.

Section snippets

Study population

This is a retrospective cohort study. TE data collected from two institutions from November 2006 – September 2015 were analysed. All patients with a LSM ⩾10 kPa were selected. Additional inclusion criteria were an OGD within 12 months of TE, and a diagnosis of chronic liver disease. Exclusion criteria were decompensated disease (defined as Child-Pugh C disease or Child-Pugh B with evidence of ascites, encephalopathy or previous variceal haemorrhage), current use of non-selective beta-blockers,

Study population

Over the study period 12331 transient elastography (TE) scans were performed. After excluding inadequate scans, values <10 kPa, and multiple scans on the same patient and scans without an OGD within 12 months, 391 cases remained. Of these, a further 81 were excluded: n = 10 non-cirrhotic portal hypertension, n = 5 portal/mesenteric/splenic vein thrombosis, n = 13 current use of non- selective beta-blockers, n = 53 decompensated disease. In total 310 patients were included in the study (Fig. 1).

Demographic data

Of the 310

Discussion

In this large dual centre cross- sectional cohort study we validate the recently published Baveno VI guidelines for using non-invasive criteria in patients with cACLD to identify patients who are at low risk of clinically significant varices and thus can safely avoid screening endoscopy. We have demonstrated that applying such criteria will reduce the number of surveillance endoscopies by about 30%, but could incorrectly classify 2% of patients. Thus adherence to these criteria may delay

Conflict of interest

The authors who have taken part in this study declared that they do not have anything to disclose regarding funding or conflict of interest with respect to this manuscript.

Authors’ contributions

JM: collected data, analysed data, wrote the manuscript, approved final manuscript.

EB, FA, AN, HP, SK: collected data and contributed to the drafting and final approval of the manuscript. AD, JO’B, DP, MP, RM, ET: provided data and contributed to the drafting and final approval of the manuscript. RW: Provided overall oversight of the study, analysed the data, contributed to the drafting and final approval of the manuscript.

Acknowledgements

We would like to thank Talay Hakan, Heather Lewis and Louise Campbell for their help in obtaining extensive transient elastography data in the two institutions. This study received no external funding.

References (16)

  • R. de Franchis

    Expanding consensus in portal hypertension

    J Hepatol

    (2015)
  • S. Augustin

    Detection of early portal hypertension with routine data and liver stiffness in patients with asymptomatic liver disease: A prospective study

    J Hepatol

    (2014)
  • F. Vizzutti

    Liver stiffness measurement predicts severe portal hypertension in patients with HCV-related cirrhosis

    Hepatology

    (2007)
  • R.J. Groszmann

    Beta-blockers to prevent gastroesophageal varices in patient with cirrhosis

    N Engl J Med

    (2005)
  • M. Wadhawan

    Hepatic venous pressure gradient in cirrhosis: correlation with the size of varices, bleeding, ascites, and child’s status

    Dig Dis Sci

    (2006)
  • C. Bureau

    Transient elastography accurately predicts presence of significant portal hypertension in patients with chronic liver disease

    Aliment Pharmacol Ther

    (2008)
  • A. Berzigotti

    Elastography, spleen size, and platelet count identify portal hypertension in patients with compensated cirrhosis

    Gastroenterology

    (2013)
There are more references available in the full text version of this article.

Cited by (172)

  • Spleen stiffness: a new tool to predict high-risk varices in cirrhotic patients

    2023, Journal of Gastroenterology and Hepatology (Australia)
View all citing articles on Scopus
View full text