Journal zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA of Hepatology 1996; 25: 633438 Printed in Denmark All rights reserved M unksgaard Copenhagen Copyright Q European Association for the Study of the Liver 1996 .~ Journal of Hepatology ISSN 0168.8278 zyxwvutsrqponmlk Threshold effect of liver iron content on hepatic inflammation and fibrosis in hepatitis B and C Nele K. Beinker’, Michael D. Voigt’, Michael Arendse*, Julian Smit3, Ilse A. Stander4 and Ralph E. Kirsch’ ‘MRCAJCT Liver Research Centre, 2UCT Department of Pathology and Vimlogy, 3Surveying, 4MRC Centre for Epidemiological Research in SA Background/Aims: In hepatitis C, iron depletion may improve serum aminotransferases and the response to interferon, but it is not known whether inflammation and fibrosis correlate with hepatic iron content. Our aim was to establish whether hepatic iron content correlates with histological and serum indices of hepatic inflammation and fibrosis in hepatitis B and C. Methods: Total hepatic iron was measured using computerized histomorphometry, and hepatic inflammation and fibrosis using a modified Knodell score, on histological slides from 31 patients with chronic hepatitis B and 38 with hep- atitis C. Results: Total hepatic iron was similar in the hepa- titis B and C groups (0.82+1.72% and 0.56+1.12%; mean&SD). No iron was detectable in 11 patients with hepatitis B and 13 with hepatitis C. Alanine aminotransferase (85.96k67.1 vs 44.2k39.7 p<O.O5), aspartate aminotransferase UBJECTS zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA S with hepatitis B and C viral infection often have raised serum ferritin and transferrin satu- ration (l-5) which may be associated with accumu- lation of hepatic iron (6,7). Iron overload may facilitate viral replication (8) increase the risk of chronicity (8) and may be asso- ciated with increased mortality in subjects with hep- atitis B (9). Increased hepatic iron content has been shown to reduce responsiveness to Interferon-a in hepatitis C (lO-12), while iron removal improves therapeutic responsiveness (10,13). The mechanism Received 16 November 1995: revised 18 Marrh; accepted 19 Marrh 1996 Correspondence: Michael D. Voigt, MRC/UCT Liver Re- search Centre, K 46.51, K Floor, OMB, Groote Schuur Hospital, 7925 Observatory, Cape Town, South Africa. Tel. (21) 4066394. Fax:(21) 4486815. Email: Mikev@Liver.uct.ac.za. (93.8’r75.6 vs 47ti3.5 IU/mlp<O.O5) and histologi- cal inflammatory score (9.33fi.51 vs 7.79E3.3 p=O.O7) were increased in those with stainable hepatic iron compared to those without. However, where iron was present, no association was found between the amount of hepatic iron and inilam- matory or fibrosis scores. In hepatitis C, fibrosis was minimal in 77% of patients if iron was absent vs 24% with iron present, while marked fibrosis was present in 56% with iron vs 15% without iron (~~0.01, Fisher’s exact test). Conclusion: Hepatic iron is associated with increased hepatic inflammation in chronic hepati- tis B and hepatitis C and with high fibrosis scores in hepatitis C. There is a threshold effect, and once present, increasing iron does not correlate with increasing inflammation or fibrosis. Key words: Inflammation; Fibrosis; Hepatitis B; Hepatitis C; Iron. by which iron affects viral hepatitis and its therapy has not been established. Indirect evidence based on serum measures of iron stores, suggests that excess iron may aggravate hepatic necro-inflammatory activity in chronic viral hepatitis (14,15), and that fibrosis may be increased by iron overload (16-18). However, serum measures of iron stores, as used in these studies, correlate poorly with hepatic iron con- tent in the presence of hepatitis or inflammation (4, 19-21). The relationship between hepatic iron con- tent and histological indices. of hepatic inflammation has not previously been studied in patients with hep- atitis B or C. We therefore performed computerized measure- ments of the hepatic iron content on biopsies of patients with hepatitis B and C, and compared these to the necro-inflammatory and fibrotic activity on the same biopsies, using Knodell’s (22) semi-quantitative 633