Validation of Baveno VII criteria and other non-invasive diagnostic algorithms for clinically significant portal hypertension in hepatitis delta

ABSTRACT

Background and Aims

Non-invasive tests (NIT) for clinically significant portal hypertension (CSPH) in compensated advanced chronic liver disease (cACLD) lack validation in patients infected with hepatitis D virus (HDV).

Methods

HDV-cACLD patients (LSM ≥10 kPa or histological METAVIR F3/F4 fibrosis) who underwent paired HVPG and NIT assessment at Medical University of Vienna or Hannover Medical School between 2013 and 2023 were retrospectively included. Liver stiffness measurement (LSM), von Willebrand factor to platelet count ratio (VITRO), and spleen stiffness measurement (SSM) were assessed. Individual CSPH risk was calculated according to previously published models (ANTICIPATE, 3P/5P). The diagnostic performance of Baveno-VII criteria and refined algorithms (Baveno-VII-VITRO, Baveno-VII-SSM) was evaluated. The prognostic utility of NIT was investigated in the main and an independent, multicenter validation cohort.

Results

Fifty-one patients (HVPG ≥10 mmHg/CSPH prevalence: 62.7%, varices: 42.2%) were included. LSM (25.8 [17.2-31.0] vs. 14.0 [10.5-19.8] kPa; p<0.001), VITRO (n=31, 3.5 [2.7-4.5] vs. 1.3 [0.6-2.0] %/[G/L]; p<0.001), and SSM (n=20, 53.8 [41.7-75.5] vs. 24.0 [17.0-33.9] kPa; p<0.001) were significantly higher in CSPH patients. Composite CSPH risk models yielded excellent AUROC (ANTICIPATE: 0.885, 3P: 0.903, 5P: 0.912). Baveno-VII criteria ruled out CSPH with 100% sensitivity and ruled in CSPH with 84.2% specificity. The Baveno-VII 'grey zone' (41.1%) was significantly reduced by Baveno-VII-VITRO or Baveno-VII-SSM, while maintaining diagnostic accuracy. Hepatic decompensation within two years occurred only in patients who had CSPH or met Baveno-VII rule-in criteria. The prognostic value of NIT was confirmed in the validation cohort comprising 92 patients.

Conclusions

Standalone and composite NIT/ diagnostic algorithms are useful for CSPH diagnosis in HDV-cACLD patients. Thus, NIT may be applied to identify and prioritize patients with CSPH for novel antiviral treatments against CHD.

Impact and implications

Non-invasive tests (NIT) for clinically significant portal hypertension (CSPH) have been developed to identify compensated advanced chronic liver disease (cACLD) patients at risk for decompensation, but conflicting data has been published regarding the accuracy of liver stiffness measurement (LSM) for the staging of fibrosis in patients infected with hepatitis D virus (HDV). In our study including 51 HDV-cACLD patients, NIT, i.e., most importantly, the ANTICIPATE model based on LSM and platelet count, but also lab-based approaches, i.e., 3P/5P model and the von Willebrand factor to platelet count ratio (VITRO), and spleen stiffness measurement (SSM) yielded high AUROC for CSPH. Moreover, only patients with CSPH or high non-invasively assessed CSPH-risk were at risk for decompensation within two years, and the prognostic value of NIT was confirmed in a validation cohort. Thus, NIT should be applied and updated in yearly intervals in clinical routine to identify HDV-cACLD patients at short-term risk and may guide prioritization for novel antiviral treatment options.