Author information
1AP-HP, Hôpital Avicenne, Unité d'Hépatologie, Bobigny; Université Sorbonne Paris Nord, Bobigny; Inserm U955, équipe 18, Université Paris-Est, Créteil, France. Electronic address: dominique.roulot@aphp.fr.
2AP-HP, Hôpital Avicenne, Laboratoire de Microbiologie Clinique; Université Sorbonne Paris Nord, Centre national de référence des hépatites B, C et Delta, Bobigny, Inserm U955, équipe 18, Université Paris-Est, Créteil, France.
3Université Paris-Est Créteil, INSERM, IMRB, CEpiA (Clinical Epidemiology and Ageing Unit) Team, Créteil; AP-HP, Hôpital Henri-Mondor, Unité de Recherche Clinique (URC Mondor) , Créteil, France.
4Liver Unit, University hospital, Tours, France.
5AP-HP, Hôpital Avicenne, Service d'Hépatologie, Bobigny; Université Sorbonne Paris Nord, Bobigny ; INSERM U1138, Université de Paris, France.
6AP-HP, Hôpital Beaujon, Service d'hépatologie, Clichy, France.
7Pôle hépato-digestif, University Hospital, Strasbourg ; Inserm U110, Strasbourg, France.
8AP-HP, Hôpital Henri-Mondor, Service d'hépatologie, Créteil ; Inserm U955, équipe 18, Université Paris-Est, Créteil, France.
9Liver Unit, University Hospital, Bordeaux, France.
Abstract
Background & aims: Liver biopsy remains the gold standard for fibrosis staging in patients with chronic hepatitis delta (CHD). Here we comparatively evaluated the performance of transient elastography (TE) and biomarkers for the diagnosis of liver fibrosis in patients with CHD.
Methods: 230 HDV-infected RNA-positive patients from various centers who underwent liver biopsy and liver stiffness measurements (LSM) using Fibroscan®, within a period of 6 months maximum, were investigated retrospectively. AUROC and Youden index were used to establish cut-off values of LSM. TE was compared with other noninvasive tests (NITs): APRI, Fibrosis-4 and Delta-4 fibrosis scores.
Results: Histologic fibrosis stage distribution was: 20.4% for F0-F1; 27.0% for F2; 18.7% for F3 and 33.9% for F4. TE demonstrated good diagnostic performance for detecting cirrhosis and advanced fibrosis with AUROC of 0.88 and 0.86, values, which were significantly higher than those obtained with the other NITs (P = .004 and P < .001). With a cutoff value >12 kPa for cirrhosis, sensitivity was 70.5%, specificity 86.2%, positive predictive value (PPV) 72.4% negative predictive value (NPV) 85.1% and accuracy 80.9%. Using 10.4 kPa as cut-off value for F3, sensitivity was 70.2%, specificity 83.5%, PPV 82.5%, NPV 71.7% and accuracy 76.5%. In 89% of patients with LSM ≤ 6.2 kPa, liver biopsy disclosed only absent or minimal fibrosis.
Conclusion: TE demonstrated good diagnostic performance for advanced fibrosis and cirrhosis in CHD patients. Advanced fibrosis is highly probable for LSM values ≥10 kPa. LSM values < 6 kPa almost totally exclude significant fibrosis. Between 6 and 10 kPa, liver biopsy should be discussed.