Non-Invasive Prediction of High-Risk Varices in Patients with Primary Biliary Cholangitis and Primary Sclerosing Cholangitis.
Adult
Aged
Cholangitis, Sclerosing
/ blood
Cross-Sectional Studies
Elasticity Imaging Techniques
Endoscopy, Digestive System
Esophageal and Gastric Varices
/ diagnosis
False Negative Reactions
Female
Humans
Liver
/ diagnostic imaging
Liver Cirrhosis
/ blood
Liver Cirrhosis, Biliary
/ blood
Male
Middle Aged
Platelet Count
Prevalence
Retrospective Studies
Risk Assessment
Journal
The American journal of gastroenterology
ISSN: 1572-0241
Titre abrégé: Am J Gastroenterol
Pays: United States
ID NLM: 0421030
Informations de publication
Date de publication:
03 2019
03 2019
Historique:
pubmed:
14
10
2018
medline:
24
12
2019
entrez:
14
10
2018
Statut:
ppublish
Résumé
Baveno-VI guidelines recommend that patients with compensated cirrhosis with liver stiffness by transient elastography (LSM-TE) <20 kPa and platelets >150,000/mm(3) do not need an esophagogastroduodenoscopy (EGD) to screen for varices, since the risk of having varices needing treatment (VNT) is <5%. It remains uncertain if this tool can be used in patients with cholestatic liver diseases (ChLDs): primary biliary cholangitis (PBC) and primary sclerosing cholangitis (PSC). These patients may have a pre-sinusoidal component of portal hypertension that could affect the performance of this rule. In this study we evaluated the performance of Baveno-VI, expanded Baveno-VI (LSM-TE <25 kPa and platelets >110,000/mm(3)), and other criteria in predicting the absence of VNT. This was a multicenter cross-sectional study in four referral hospitals. We retrospectively analyzed data from 227 patients with compensated advanced chronic liver disease (cACLD) due to PBC (n = 147) and PSC (n = 80) that had paired EGD and LSM-TE. We calculated false negative rate (FNR) and number of saved endoscopies for each prediction rule. Prevalence of VNT was 13%. Baveno-VI criteria had a 0% FNR in PBC and PSC, saving 39 and 30% of EGDs, respectively. In PBC the other LSM-TE-based criteria resulted in FNRs >5%. In PSC the expanded Baveno criteria had an adequate performance. In both conditions LSM-TE-independent criteria resulted in an acceptable FNR but saved less EGDs. Baveno-VI criteria can be applied in patients with cACLD due to ChLDs, which would result in saving 30-40% of EGDs. Expanded criteria in PBC would lead to FNRs >5%.
Sections du résumé
BACKGROUND
Baveno-VI guidelines recommend that patients with compensated cirrhosis with liver stiffness by transient elastography (LSM-TE) <20 kPa and platelets >150,000/mm(3) do not need an esophagogastroduodenoscopy (EGD) to screen for varices, since the risk of having varices needing treatment (VNT) is <5%. It remains uncertain if this tool can be used in patients with cholestatic liver diseases (ChLDs): primary biliary cholangitis (PBC) and primary sclerosing cholangitis (PSC). These patients may have a pre-sinusoidal component of portal hypertension that could affect the performance of this rule. In this study we evaluated the performance of Baveno-VI, expanded Baveno-VI (LSM-TE <25 kPa and platelets >110,000/mm(3)), and other criteria in predicting the absence of VNT.
METHODS
This was a multicenter cross-sectional study in four referral hospitals. We retrospectively analyzed data from 227 patients with compensated advanced chronic liver disease (cACLD) due to PBC (n = 147) and PSC (n = 80) that had paired EGD and LSM-TE. We calculated false negative rate (FNR) and number of saved endoscopies for each prediction rule.
RESULTS
Prevalence of VNT was 13%. Baveno-VI criteria had a 0% FNR in PBC and PSC, saving 39 and 30% of EGDs, respectively. In PBC the other LSM-TE-based criteria resulted in FNRs >5%. In PSC the expanded Baveno criteria had an adequate performance. In both conditions LSM-TE-independent criteria resulted in an acceptable FNR but saved less EGDs.
CONCLUSIONS
Baveno-VI criteria can be applied in patients with cACLD due to ChLDs, which would result in saving 30-40% of EGDs. Expanded criteria in PBC would lead to FNRs >5%.
Identifiants
pubmed: 30315285
doi: 10.1038/s41395-018-0265-7
pii: 10.1038/s41395-018-0265-7
doi:
Types de publication
Journal Article
Langues
eng
Sous-ensembles de citation
IM
Pagination
446-452Commentaires et corrections
Type : CommentIn