|Home | About | Journals | Submit | Contact Us | Français|
Iron accumulation is a well‐known risk factor for the progression of chronic hepatitis C (CHC) to fibrosis. However, the profibrogenic role of the genes controlling iron homeostasis is still controversial.
To evaluate the relative role of haemachromatosis (HFE), ferroportin and β‐globin gene mutations in promoting iron accumulation and fibrosis in patients with CHC.
Genetic analysis was performed together with the assessment of hepatic iron content and histology in 100 consecutive HIV‐antibody and hepatitis B surface antigen‐negative patients with biopsy‐proven CHC.
Among the patients investigated, 12 were heterozygous for various β‐globin gene mutations (39[C→T], IVS1.1[G→A], 22 7 bp deletion and IVS1.6[T→C]) and 29 carried HFE (C282Y, H63D and S65C) gene mutations. One further patient was heterozygous for both HFE (H63D) and β‐globin (39[C→T]) variants, whereas 58 had the wild‐type alleles of both the genes. Hepatic iron concentration (HIC) and hepatic stainable iron were significantly higher (p<0.05) in patients with CHC carrying β‐globin mutations than in those with HFE mutations or the wild‐type alleles. Multivariate analysis confirmed that the presence of β‐globin mutations was independently associated with both HIC (p=0.008) and hepatic‐stainable iron (odds ratio (OR) 6.11; 95% CI 1.56 to 23.92; p=0.009). Moderate/severe fibrosis or cirrhosis (Ishak's score >2) was observed in 48 of 100 patients. Logistic regression demonstrated that age (OR 1.05; 95% CI 1.02 to 1.09; p<0.005) and β‐globin mutations (OR 4.99; 95% CI 1.22 to 20.3; p=0.025) were independent predictors of the severity of fibrosis.
Heterozygosis for β‐globin mutations is a novel risk factor for both hepatic iron accumulation and the progression to fibrosis in patients with CHC.
Chronic hepatitis C (CHC) is characterised by a high variability in the rate of the disease progression with some patients showing minimal evolution of hepatic injury, and others rapidly developing fibrosis that leads to cirrhosis and hepatocellular carcinoma.1,2,3 A number of viral and host‐related factors, including older age at infection, alcohol consumption, male gender, coinfections with hepatitis B virus (HBV) or HIV, overweight, and hepatic iron status have been associated with an increased risk of CHC progression.2,3 Among these, the interaction of hepatitis C virus (HCV) and iron has received increasing attention.4 Mild iron overload is a common finding in liver biopsy specimens from patients with CHC, and HCV‐infected patients with increased hepatic iron content show increased liver fibrosis as compared with patients without iron accumulation.4,5 Moreover, iron‐reducing treatments have been reported to improve liver damage in untreated patients with CHC, as well as to enhance the rate of response to interferon treatment in interferon‐resistant subjects.5 However, the mechanisms responsible for iron accumulation during HCV infection are not fully understood. It has been proposed that necroinflammatory events caused by the virus may lead to alterations in the tissue distribution of the metal.5 Nonetheless, the role of genetic factors that influence iron trafficking cannot be disregarded. In this context, the broad distribution of polymorphisms of the haemachromatosis (HFE) gene and their importance in causing hereditary HFE6 has led to the search for a possible role of HFE mutations in the progression of CHC. However, while some studies have shown an association between mutated HFE genotypes and both iron overload and advanced fibrosis or cirrhosis, others have failed to document any significant involvement of HFE mutations in promoting hepatic iron accumulation and CHC evolution.4,7 Iron trafficking is affected by other genetic defects such as the ferroportin 1 (FPN1) and transferrin receptor 2 (TFR2) mutations, which are involved in non‐classic hereditary HFE,8,9 as well as by the haemoglobin mutations responsible for iron‐loading anaemia.10 Several studies have documented the rapid evolution of hepatic fibrosis in patients with thalassaemia major and HCV infection.11 Although heterozygous carriers for thalassaemic mutations account for about 1.67% of the world population,10 the possible contribution of thalassaemic traits in favouring iron accumulation in CHC has not been characterised. The main purpose of this study was to explore the possible associations between different mutations in the β‐globin, HFE, FPN1 and TFR2 genes, and both hepatic iron content and the severity of liver fibrosis among Italian patients with CHC.
For this study, 100 consecutive HIV antibody‐ and hepatitis B surface antigen‐negative patients with CHC (54 men and 46 women; mean age 52 years) were recruited from January 2002 to February 2004. All of them had untreated CHC, defined as having a liver biopsy assessed by a pathologist, a second‐generation positive enzyme‐linked immunosorbent assay and the presence of HCV‐RNA in the serum evaluated by nested PCR. The criteria of exclusion were current drug addiction, hepatitis B surface antigen or anti‐HIV positivity, every other known liver disease, and present or previous antiviral treatment. All subjects gave informed consent to the analysis. The study was planned according to the guidelines of the local ethical committee in conformity with the 1975 Declaration of Helsinki. Anti‐HCV status was assessed by second‐generation enzyme‐linked immunosorbent assay (Ortho Diagnostic System, Milan, Italy). HCV‐RNA was detected in the serum by the Amplicor HCV Kit (Roche Diagnostic System, Branchburg, New Jersey, USA) following the manufacturer's instructions. The lifetime mean daily alcohol consumption was assessed at the time of admission with a standardised questionnaire, presented as part of a survey on life habits. Alcohol intake >50 g/day for men and >20 g/day for women in the 5 years that preceded the hospital admission were considered.
DNA was extracted from peripheral blood mononucleated cells using QIAamp DNA extraction kit (QIAGEN, Valencia, California, USA) according to the manufacturer's instructions.
The polymorphisms of the β‐globin, HFE, TFR2 and FPN1 genes were assessed using the β‐Globin StripAssay and Haemochromatosis StripAssay A (ViennaLab Labordiagnistika GmbH, Vienna, Austria), consisting of PCR amplification of the sequences of interest using biotinylated primers, followed by the hybridisation of the amplification products to a test strip containing immobilised allele‐specific oligonucleotide probes. Bound biotinylated sequences are detected using streptavidin–alkaline phosphatase and colour substrates. The following 22 mutations of the β‐globin gene were investigated: –87 [C→G], –30[T→A], codon 5 [−CT], Hb C, Hb S, codon 6 [−A], codon 8 [−AA], codon 8/9 [+G], codon 22 [7 bp del], codon 30 [G→C], IVS1.1 [G→A], IVS1.2 [T→A], IVS1.5 [G→C], IVS1.6 [T→C], IVS1.110 [G→A], IVS1.116 [T→G], IVS1–25 [25 bp del], codon 36/37 [−T], codon 39 [C→T], codon 44 [−C], IVS2.1 [G→A], IVS2.745 [C→G]. The HFE gene was analysed for 11 mutations as follows: V53M, V59M, H63D, S65C, Q127H, P160delC, E168Q, E168X, W169X, C282Y and Q283P9. The presence of E60X, M172K, Y250X and AVAQ594‐597del mutations of TFR2 and N144H, V162del of the FPN1 genes was also evaluated by the same procedure. The accuracy of the assays was confirmed by sequencing randomly selected DNAs.
Liver biopsy was performed at the time of the admission using a modified Menghini procedure. On a portion of each sample, hepatic iron concentration (HIC) was measured by atomic absorption spectroscopy and expressed as μmol/g dry weight. The cut‐off value for HIC was fixed at 25 μmol/g. Hepatic iron index was calculated as reported previously.12 The remainder of the sample was immediately fixed in formalin and embedded in paraffin wax. Sample sections of thickness 5 μm were stained with haematoxylin and eosin, Masson's trichrome and periodic acid‐Schiff after diastase digestion, as well as with the Gomori's method for reticulin and the Perls's method for iron staining. The grading and the staging of chronic hepatitis were scored according to Ishak's criteria.13 Histological assessment of hepatic iron stores was performed using the score system proposed by Searle et al.14 All histological evaluations were performed double‐blind by two experienced pathologists. In the case of discordant opinions, the two examiners analysed the discrepancies to reach a consensus.
Stata Statistical Software V.9.0 was used in all the statistical analyses. Each variable predictive of or associated with the presence of hepatic fibrosis, HIC and hepatic stainable iron was analysed with univariate linear regression (when indicated) and/or with univariate logistic regression. The variables selected by every univariate analysis were entered into linear and/or logistic regression models using a forward stepwise elimination algorithm (terms with p>0.05 were eligible for removal). The analysis of variance model was used to compare the continuous variables among groups, and multiple comparison tests were performed with the Bonferroni's correction. The Kruskal–Wallis model and the Wilcoxon rank‐sum test were used to compare discrete variables.
Table 11 summarises the demographic, clinical and laboratory characteristics of the patients with CHC included in this study. The probable source, the duration of the infection and the age at infection were available for 41 patients, who received blood transfusions before 1990 (31 patients) or shared syringes while using illicit drugs (10 patients). Genetic characterisation of all the 100 patients with CHC demonstrated that 12 were heterozygotic for one of the different β‐globin gene mutations (codon 39 C→T in seven cases, codon IVS 1.1 G→A in three cases, codon 22 7 bp deletion in one case and codon IVS 1.6 T→C in one case) affecting haemoglobin synthesis, and that 29 carried HFE gene mutations (heterozygotic C282Y in three cases, homozygotic H63D in three cases, heterozygotic H63D in 22 cases and S65C heterozygous in one case). One further patient was heterozygotic for both HFE (H63D) and β‐globin (codon 39 C→T) variants. In all, 58 patients had the wild‐type alleles of both the genes (table 22).). None of the patients carried mutations of the FPN1 and TFR2 genes (table 22).
As expected, haemoglobin was lower in patients with CHC and mutated β‐globin (11.8 (1.95) g/dl; 95% CI 10.6 to 13.0; p<0.001) than in those with HFE variants (14.4 (1.5) g/dl; 95% CI 13.8 to 15.0) or the wild‐type alleles (14.4 (1.4) g/dl; 95% CI 14.0 to 14.8). Intrahepatic accumulation of iron, as estimated by both HIC and iron staining by the Perls's method, was evident in, respectively, 22 of 96 (23%) and 40 of 98 (41%) subjects investigated. HIC and hepatic iron staining were significantly higher (p<0.05 and p<0.01, respectively) in patients with CHC with β‐globin mutations than in those with HFE mutations or the wild‐type alleles (table 33).). The carriers of β‐globin variants also had higher serum ferritin levels (p<0.05) than the patients with mutant HFE (table 33).). Transferrin saturation was increased in patients with CHC with β‐globin mutations as compared with those with the wild‐type alleles (table 33).). No difference was appreciable in the serum iron content between the three groups (table 33).
In both univariate and multivariate analyses, the presence of β‐globin mutations was the only variable found to be associated (p=0.008) with HIC (table 44).). Male gender (odds ratio (OR) 2.55; 95% CI 1.10 to 5.92; p=0.029), necroinflammatory grading (OR 1.25; 95% CI 1.02 to 1.53; p=0.029) and mutated β‐globin alleles (OR 6.11; 95% CI 1.56 to 23.92; p=0.009) were also associated with the detection of hepatic stainable iron (Searle's score >0) (table 44).). Nonetheless, following multivariate logistic regression analysis, the heterozygosis for the different β‐globin mutations was the only risk factor (OR 6.11; 95% CI 1.56 to 23.92; p=0.009) for the histological detection of iron deposition.
Histology revealed the presence of moderate/severe liver fibrosis (staging >2) in 48 of 100 patients with CHC and cirrhosis (staging >4) in 14 patients (table 11).). The scores for both histological grading and staging were significantly (p<0.05) higher in the patients with β‐globin mutations as compared with those with the HFE variants or the wild‐type alleles (table 33).). Logistic regression analysis taking the presence of moderate/severe fibrosis (staging score >2) as dependent variable revealed that age (OR 1.05; 95% CI 1.02 to 1.09; p<0.005) and β‐globin mutations (OR 4.99; 95% CI 1.22 to 20.3; p=0.025) were the only independent predictors of moderate/severe fibrosis or cirrhosis in the patients investigated (table 55).). Similar results (age OR 1.04; 95% CI 1.01 to 1.09; p=0.008; β‐globin mutations OR 5.0; 95% CI 1.24 to 20.2; p=0.024) were obtained after correction for the possible confounding action of the high alcohol intake in seven of the patients investigated.
In 1992, Di Bisceglie et al15 first reported that a high proportion of patients with CHC had abnormal serum iron values. Such an observation has been confirmed by subsequent studies showing a correlation between the extent of hepatic stainable iron and the severity of hepatitis.5 Moreover, we have reported that menstruating women experienced a milder CHC than men of the same age, in relation to the lower HIC due to blood losses.16 Consistently, a number of studies have shown that lowering the iron stores by phlebotomy or other iron‐reducing treatments improves the evolution of CHC and enhances the response to interferon treatments.4,5 The mechanisms responsible for the worsening of HCV infection by iron have not been characterised in detail. Recent observations suggest that iron might enhance HCV replication by stimulating the expression of translation initiation factor 3.17 However, other evidence points to the interference of iron with the virus replication processes.18 Nonetheless, the notions that HCV proteins stimulate the formation of reactive oxygen species within the hepatocytes and that iron exacerbates oxidative damage suggest that iron accumulation might amplify oxidative stress‐mediated events, leading to both hepatocellular damage and fibrogenic processes.19 Supporting this view, Furutani et al20 have recently shown that mild iron accumulation in the liver of transgenic mice overexpressing HCV polyprotein enhances lipid peroxidation and promotes the development of hepatocellular carcinomas.
At present, the mechanisms responsible for hepatic iron deposition during HCV infection are not fully understood. It has been proposed that necroinflammation sustained by the virus along with genetic factors that modify iron trafficking might lead to the alterations in iron homeostasis.5 In humans, the only known regulated step in iron trafficking is the intestinal absorption.9 Hepcidin, a 25 amino acid peptide synthesised in the liver, plays a key role in such a process by inhibiting iron efflux from the enterocytes as well as the recycling of the metal by the macrophages.21 Recently, Aoki et al22 have reported that, in patients with CHC, liver hepcidin mRNA expression is not influenced by necroinflammation and fibrosis and increases in relation with hepatic iron concentration. This suggests that iron stores regulate hepcidin expression in patients with CHC, and that iron accumulation is not due to inappropriate hepcidin expression. However, other hepatic proteins essential for the normal iron homeostasis, including HFE, TFR2 and haemojuvelin, have been proposed to modulate the expression of hepcidin.9,21 Indeed, hepcidin production is decreased in patients with hereditary HFE and in HFE‐knockout mice.9 A number of studies have addressed the role of the common HFE mutations in favouring hepatic iron accumulation during CHC, obtaining conflicting results.4,7 These discrepancies might be due to the failure to correct for confounding factors.7 Moreover, as pointed out by Tung et al,23 the association between HFE mutations and the severity of both iron accumulation and fibrosis is more evident in patients with compensated disease after controlling for the duration of infection, and disappears in patients with end‐stage liver disease. Although the patients included in the present study all have compensated CHC, we failed to see a contribution of HFE mutations in causing hepatic iron accumulation and disease progression. This might be ascribed to the predominance of the HFE H63D variant among our patients, which affects iron homeostasis in CHC to a lower extent than the C282Y variant.24,25 Moreover, the presence of other, still poorly characterised, factors that interfere with the phenotypical expression of mutated HFE genes cannot be excluded.26
Thalassaemia represents a common cause of iron overload in many countries of the Mediterranean, Middle East and South Asia, and its frequency is rapidly increasing in the US.27 The combination of HCV infection and iron overload is a well‐established risk factor for hepatic cirrhosis and hepatocellular carcinoma among patients with β‐thalassaemia.28,29 However, in spite of the high frequency of thalassaemic trait carriers worldwide, the contribution of heterozygotic thalassaemic variants to the evolution of hepatitis C has so far been neglected. Our observation that patients with CHC heterozygotic for different β‐globin mutations have a 5–6‐fold increased risk of both iron overload and extensive fibrosis/cirrhosis adds new emphasis to the role of the haemoglobin defects in the progression of HCV infection. It is noteworthy that the frequency of β‐thalassaemic traits among our patients with CHC is higher (13%) compared with that in the general population from the same area of north Italy (about 4%),30 probably because of a clustering of the mutant β‐globin alleles among the subjects with active hepatitis C. In patients with β‐thalassaemia, major ineffective erythropoiesis and anaemia are important stimuli for intestinal iron absorption, despite a massive iron overload.31 Consistently, urinary hepcidin levels are very low or undetectable in patients with thalassaemia major or intermedia.32 A low hepatic expression of hepcidin mRNA is also detectable in the mouse models of the different forms of thalassaemia.33 Impaired erythropoiesis and increased iron absorption are evident in the carriers of β‐thalassaemic traits,34 suggesting a decrease in hepcidin production. Thus, it is possible to speculate that, even though HCV infection by itself does not influence hepatic hepcidin synthesis, patients with CHC heterozygotic for β‐globin mutations experience hepatic iron accumulation because of the predominance of the signals from the erythroid compartments over those coming from the iron stores in down‐modulating hepcidin production. Accordingly, the incubation of HepG2 cells with the sera from patients with thalassaemia decreases hepcidin mRNA expression.35
In conclusion, our study identifies the heterozygosis for β‐globin gene mutations as a new independent risk factor for iron accumulation and the progression of liver fibrosis in patients with CHC, and suggests the presence of thalassaemic traits as a possible contributor to the alterations in iron homeostasis associated with chronic liver diseases.
This work has been supported by grants from the University of East Piedmont, Regional Government of Piedmont, and from the Italian Ministry for University and Scientific and Technological Research (Research Program: 200401213 Epatopatia steatosica non alcolica; studio dei fattori metabolici, genetici e virologici responsabili del danno epatico e della sua evoluzione).
CHC - chronic hepatitis C
FPN1 - ferroportin 1
HCV - hepatitis C virus
HFE - haemachromatosis
HIC - hepatic iron concentration
TFR2 - transferrin receptor 2
Competing interests: None.