CORRESPONDENCE Prevalence of liver fibrosis and cirrhosis in screening-detected C282Y homozygous subjects ARNE A ˚ SBERG 1 , KRISTIAN HVEEM 2 , TORE B. HALVORSEN 3 & HANNE-BRIT G. SMETHURST 3 1 Department of Clinical Chemistry, Trondheim University Hospital, Trondheim, Norway, 2 Department of Medicine, Innherred Hospital, Levanger, Norway, and 3 Department of Pathology, Trondheim University Hospital, Trondheim, Norway TO THE EDITOR: Most patients with clinically manifest hereditary haemochromatosis are homozy- gous for the C282Y-mutation in the HFE -gene [1]; however, most C282Y homozygous subjects never become ill [2,3]. In a recent article, Powell et al. published their results from clinical evaluation in- cluding liver biopsy in 672 essentially asymptomatic C282Y homozygous subjects detected by screening in Australia [4]. Liver biopsy is not commonly done in asymptomatic subjects nowadays [5]. The Powell et al. article [4] is important because the investiga- tors document a considerable prevalence of liver fibrosis and cirrhosis in C282Y homozygous sub- jects. Their findings are in contrast to our findings in 297 C282Y homozygous subjects detected in a large, primarily phenotypic screening in Norway [6]. However, very few data specific for the C282Y homozygous population were given in that article [6]. Here we provide such information, tabled in a format comparable with Table I in the article by Powell et al. [4]. Data on participants and methods are presented in detail elsewhere [6]. In short, when screening 65,238 adult Norwegians for hereditary haemochromatosis (as a part of The HUNT Study) we detected 297 C282Y homozygous subjects in 559 persons with confirmed high serum transferrin saturation. Most of the C282Y homozygous subjects were examined clinically by a gastroenterologist and were offered a liver biopsy. Liver biopsy was done in 63% of male and 36% of female C282Y homozygous subjects. Statistically, this is not significantly different from 62% and 41%, respectively, in the Powell article (p  /1.00 and p  /0.33, respectively (the Fisher exact test)). We did not obtain a liver biopsy according to protocol [6] in 27 male and 47 female subjects. Liver biopsy was not done owing to protocol violation in 16 males and 26 females, and was refused by 21 male and 8 female patients. In these non-biopsy groups, the serum ferritin concentration was lower than that in the corresponding biopsy group, statistically signifi- cantly so in all cases, with the exception of the patient refusal groups (two-sided Mann-Whitney test). We take this to mean that the observed prevalence of pathologic findings in the liver biopsies represents the maximum prevalence in this C282Y homozygous population. Compared to the findings of Powell et al. [4] our cases were slightly older and the prevalence of fibrosis and cirrhosis was lower. However, cirrhosis in 4 (4%) out of 92 specimens from male patients is not statistically significantly different from 20 (9%) out of 222 specimens in the Powell material (p  /0.24, (the Fisher exact test)). Fibrosis in 11 (11%) out of 96 male specimens is marginally significantly differ- ent from 46 (21%) out of 222 (p  /0.056, (the Fisher exact test)). In females, the prevalence of fibrosis and cirrhosis is less in the Norwegian material, but not statistically significantly different. The percentage of male and female subjects with Grade 4 liver iron is also less than the corresponding percentage in the Powell et al. material [4], but again the p -values do not reach statistical significance. Correspondence: Arne A ˚ sberg, MD, PhD, Department of Clinical Chemistry, Trondheim University Hospital, NO-7006 Trondheim, Norway. Tel:  /47 99 629 819. Fax:  /47 72 576 420. E-mail: arne.asberg@stolav.no Scandinavian Journal of Gastroenterology, 2007; 42: 782  783 ISSN 0036-5521 print/ISSN 1502-7708 online # 2007 Taylor & Francis DOI: 10.1080/00365520601076058