BACKGROUND: The METAVIR score, which is the most widely used score in France, was specifically elaborated and evaluated in chronic hepatitis C and has never been validated in HIV-hepatitis virus B (HBV) co-infected patients. AIMS: To validate the use of the METAVIR scoring system for activity and fibrosis on liver biopsies in co-infected HIV-HBV patients. METHODS: METAVIR scoring for activity and fibrosis was first conducted on both original and virtual slides by one pathologist for comparison. Then 55 biopsies turned into virtual slides were scored by three pathologists independently. RESULTS: The scoring comparison between virtual slides and glass slides showed an almost perfect agreement for fibrosis (weighted kappa (kappa(w)) 0.8) and a substantial agreement for activity (kappa(w) 0.68). The inter-observer agreement on virtual slides was moderate to almost perfect (kappa(w) 0.52 to 0.84) for fibrosis and was dependent on the pair of pathologists considered. The best agreement was obtained in scoring advanced fibrosis and cirrhosis versus significant fibrosis versus no or mild fibrosis (kappa(w) 0.70 to 0.84). The agreement for cirrhosis was rated moderate to substantial (kappa(w) 0.54 to 0.79). Agreement for activity was substantial (kappa(w) 0.66 to 0.8). CONCLUSIONS: This study validates the use of virtual slide technology to assess fibrosis and activity on liver biopsies. It also validates the use of the METAVIR score in co-infected HIV-HBV patients and illustrates the challenges in establishing the histological diagnosis of cirrhosis in the HIV-HBV context.
BACKGROUND: The METAVIR score, which is the most widely used score in France, was specifically elaborated and evaluated in chronic hepatitis C and has never been validated in HIV-hepatitis virus B (HBV) co-infectedpatients. AIMS: To validate the use of the METAVIR scoring system for activity and fibrosis on liver biopsies in co-infected HIV-HBVpatients. METHODS: METAVIR scoring for activity and fibrosis was first conducted on both original and virtual slides by one pathologist for comparison. Then 55 biopsies turned into virtual slides were scored by three pathologists independently. RESULTS: The scoring comparison between virtual slides and glass slides showed an almost perfect agreement for fibrosis (weighted kappa (kappa(w)) 0.8) and a substantial agreement for activity (kappa(w) 0.68). The inter-observer agreement on virtual slides was moderate to almost perfect (kappa(w) 0.52 to 0.84) for fibrosis and was dependent on the pair of pathologists considered. The best agreement was obtained in scoring advanced fibrosis and cirrhosis versus significant fibrosis versus no or mild fibrosis (kappa(w) 0.70 to 0.84). The agreement for cirrhosis was rated moderate to substantial (kappa(w) 0.54 to 0.79). Agreement for activity was substantial (kappa(w) 0.66 to 0.8). CONCLUSIONS: This study validates the use of virtual slide technology to assess fibrosis and activity on liver biopsies. It also validates the use of the METAVIR score in co-infected HIV-HBVpatients and illustrates the challenges in establishing the histological diagnosis of cirrhosis in the HIV-HBV context.