Main

Hereditary hemochromatosis (HHC) (OMIM 235200) is a common autosomal recessive disorder affecting adults in populations from North European descent; in Caucasians its prevalence, based on population iron loading screening, is 1 affected individual in 300 with a decrease in prevalence from the North to the South of Europe.1–4 Hereditary hemochromatosis is a disorder of iron metabolism characterized by high and uncontrolled intestinal iron absorption, progressive iron overload, and iron deposition into perenchymal tissues, which can result in organs injury. The manifestations of the iron overload develop gradually in patients. Indeed, three different stages can be distinguished as follows: (1) latency, (2) biochemical expression observed in young adults and corresponding to elevated serum iron parameters (transferrin saturation and serum ferritin concentration), and then (3) clinical expression. However, the onset of the disease is often insidious, and affected individuals display a wide variability in biochemical and clinical expression as well as in evolution of the disease. Clinical symptoms, which usually appear in the 4th decade in men and later in women, are very heterogeneous, unspecific, and range from mild symptoms, eg, fatigue and arthralgia, to life-threatening complications such as hepatic cirrhosis, diabetes mellitus, and cardiomyopathy.5,6 It is well established that clinical consequences of iron overload can be prevented by early diagnosis and iron depletion treatment through regular venesection, generally once a week till normalization of iron parameters, followed with maintenance venesections.7,8

For a long time the diagnosis of hereditary hemochromatosis has been based on clinical suspicion and biochemical assessment often confirmed on hepatic iron content evaluation from liver biopsy or retrospective assessment of total iron removed by a course of phlebotomies, i.e., more than 5 g of iron should have been removed without producing iron deficiency.9 Because the discovery of the HFE gene a reliable direct DNA-based test has become available and helps with diagnosis. The HFE gene cloned in 1996 codes a major histocompatibility complex class I-like transmembrane glycoprotein that interacts with the transferrin receptor.10–12 Knockout mice demonstrated that HFE defect is correlated to the development of severe iron overload,13 and the mutated forms of HFE account for most of the cases of the classic adult hemochromatosis. The homozygosity for the C282Y mutation (845G→A in exon 4) of HFE is associated with 60% to 90% of all cases of hereditary hemochromatosis.14–19 The C282Y homozygous patients exhibit high iron overload, and this genotype is associated with most of the severe cases of HFE-linked hemochromatosis; however, C282Y/C282Y genotype shows incomplete penetrance, predominantly in women.20 The common H63D variant (187C→G in exon 2) and the rare S65C variant (193A→T in exon 2) both display a very low penetrance and are usually associated with milder form of iron overload.2,21–25 In addition, about 10 private mutations have also been identified in the coding sequence of the HFE gene of patients affected with hemochromatosis. Despite the discovery of the HFE gene, the relation between the phenotype of hemochromatosis and HFE genotype remains complex. Incomplete penetrance of the HFE mutations and variable expressivity of the disease in patients showing an alike degree of iron overload26–28 indicate that the phenotypic expression of hemochromatosis can be, therefore, influenced by modifier gene(s) and/or environmental factors: for example, blood donation is known to reduce iron overload, whereas alcohol consumption increases it.29,30

In France, the population of Brittany shows high prevalence of hereditary hemochromatosis; the blood center of Brest has been involved in venesection treatment since the early 1970s and HFE gene test has been currently performed since the gene discovery in 1996. In this study, we report on a 6-year survey of HFE gene test-based diagnosis of hemochromatosis. The positive tests confirming hemochromatosis were 89.7% in individuals diagnosed before HFE gene testing and around 30% among the individuals subjected to HFE gene test for hemochromatosis diagnosis because of iron overloading signs. However, the HFE gene test confirmed the genetic cause of elevated iron parameters when lower iron overload was reached for the C282Y/C282Y individuals, and for a higher proportion of genotypes usually associated with low expressivity and penetrance (C282Y/H63D, H63D/H63D).

MATERIALS AND METHODS

Patients

HFE gene test was conducted on a total of 410 randomly selected subjects as controls, 478 patients with clinical diagnosis of hemochromatosis, and 3047 patients over a 5-year period because of an incidental finding of an elevated serum iron parameter and/or family history. The subjects included in the study were all from Brittany and had a Caucasian origin. Retrospective DNA analysis was performed on patients for which a clinical diagnosis of hemochromatosis had been established before HFE gene cloning and who had already undergone a venesection treatment. Diagnosis was based on classical signs and symptoms of the disease (1) elevated transferrin saturation and/or serum ferritin concentration, (2) hepatic symptoms such as unexplained elevation of serum liver enzymes, cirrhosis, liver failure, or diabetes mellitus, and (3) nonspecific compatible symptoms: fatigue, abdominal pain, joint pain, cardiac arrhythmia, and hyperpigmentation. The iron status based on transferrin saturation, ferritin concentration, and serum iron concentration was systematically determined before the start of therapeutic phlebotomies. Since the discovery of the HFE gene, patients have been prospectively tested at the time of a medical advice because of elevated iron parameter(s) above the normal values (transferrin saturation > 45%, serum ferritin > 400 μg/L, and 300 μg/L in males and females, respectively), associated or not with other symptom(s) that could suggest hereditary hemochromatosis, or on cascade family screening.

HFE gene test

The HFE gene test was performed on DNA extracted from peripheral blood leukocytes by standard method. The C282Y, H63D, and S65C mutations were analyzed as previously described.19,25 Briefly, part of exons 4 or 2 was amplified using the primers 5′-GGAAGAGCAGAGATATACGT-3′ and 5′-TACCTCCTCAGGCACTCCT-3′ and 5′-ACATGGTTAAGGCCTGTTGC-3′ and 5′-GCCACATCTGGCTTGAAATT-3′, respectively. Then 10 μL of the PCR products were digested to completion with 5U of RsaI, BclI, or HinfI restriction enzyme for C282Y, H63D, and S65C mutations, respectively, and finally resolved on a 3% Nusieve 3:1 agarose gel (FMC BioProducts, Rockland, ME).

RESULTS

To evaluate the prevalence of the genotypes associated with hemochromatosis in Brittany 410 unrelated and randomly selected individuals were genotyped for C282Y, H63D, and S65C mutations (Table 1). The C282Y allele frequency observed in the control population is 7.7% with 95% CI [0.059–0.095], and the carrier prevalence is 12% (1/8.3); the proportion of individuals carrying at least a C282Y allele, and thus presenting a risk of transmission of the allele associated with hemochromatosis disease, is 1/7 (14.9%). The expected prevalence of the C282Y homozygous genotype in the control population is 1/170 (0.6% with 95% CI [0–0.0116]), and this represents the frequency of potentially affected individuals in the overall population when assuming complete penetrance of the C282Y/C282Y genotype. The H63D variant accounts for 14% with 95% CI [0.12–0.16] of the chromosomes, the carrier prevalence is 23.6% (1/4.2), and then the prevalence of individuals carrying at least a H63D allele is 1/3.5 (27.3%). The genotypes most often associated with hemochromatosis phenotype, i.e., C282Y homozygous and C282Y/H63D compound heterozygous, have an expected prevalence of 0.6% and 2.2%, respectively, thus 2.8% of the population or 1/36. The S65C variant accounts only for 2% with 95% CI [0.01–0.029] of the total alleles in the control population, thus 3.9% of individuals carry the S65C allele and the expected frequency of the genotypes (S65C/S65C, S65C/C282Y, and S65C/H63D) that may lead to hemochromatosis is 0.88%.

Table 1 HFE genotype frequencies in control population
Full size table

Genotyping for the three main mutations of the HFE gene, C282Y, H63D, and S65C, was performed on patients diagnosed before HFE gene testing as suffering from hemochromatosis and currently on venesection treatment at the Brest blood center (Table 2). The HFE gene test confirmed the existence of a mutation on both chromosomes for 89.7% of the 478 studied probands. Among these positive probands 90.4% with 95% CI [0.876–0.932] were homozygous for the C282Y mutation, 7.5% with 95% CI [0.0497–0.0994] were compound C282Y/H63D heterozygotes, 1.4% with 95% CI [0.00287–0.02509] were H63D homozygotes, and 0.7% with 95% CI [0–0.01487] were compound heterozygotes with S65C mutation. After HFE gene discovery, a direct gene test was systematically performed on individuals under clinical suspicion of hemochromatosis because of symptoms or signs indicative of iron overload evidenced at the time of a medical advice or on a cascade family screening. From 1997 to mid-2002, 3047 individuals were prospectively tested mainly because of incidental finding of an elevated serum iron parameter or family history. The number of requested gene tests regularly increased in 1997 and 1998 to stabilize at a constant level since 1999 (Table 3). From early 1997 to mid 1998, 935 tests were performed, and 42.8% were positive, i.e., homozygotes or compound heterozygotes for HFE mutation; the positive tests consisted of 64.7% of C282Y/C282Y, 7.95% of H63D/H63D, and 27.35% of compound heterozygotes. Between mid-1998 and mid-2002, about 530 tests were conducted every year and 29.5% of them, on average, were positive. They consisted of 48.3% of C282Y/C282Y, 11.9% of H63D/H63D, and 39.8% of compound heterozygotes. Thus, despite the increase in the number of performed HFE gene tests, the proportion of positive ones decreased after the start of HFE gene testing. However, the overall frequencies of the genotypes with mutation on both chromosomes were significantly higher than in the control ones. From 1997 to mid-1998, the C282Y/C282Y genotype frequency was 55-fold greater than the control one; then from mid-1998 to mid-2002 it was 29-fold the control one. On average, the H63D/H63D, C282Y/H63D, and C282Y/S65C displayed 4.7-, 4.6- and 3.4-fold increases, respectively, compared to controls, whereas the H63D/S65C was not significantly different from controls. The heterozygous C282Y/wt genotype was also 1.7-fold more frequent than in controls, whereas the H63D/wt and S65C/wt were respectively found at 1.5- and 2-fold lower frequencies. Nevertheless, H63D homozygotes and all compound heterozygotes and carriers of a mutation were significantly more frequent in this group of prospectively tested subjects than in the clinically diagnosed hemochromatosis patients.

Table 2 Retrospective HFE genotype in hemochromatosis patients selected before gene discovery
Full size table
Table 3 Genotype frequencies in prospectively HFE tested subjects after gene discovery
Full size table

The age and serum iron parameters displayed at diagnosis and before the first venesection by patients who had been clinically diagnosed hemochromatosis before the HFE gene test availability were compared to those of patients subjected to prospective gene test (Table 4). For all genotypes, no significant difference was found between the mean age, at diagnosis, of retrospectively tested hemochromatosis subjects and the one of prospectively tested subjects. The assessment of serum iron parameters evidenced that, in C282Y homozygotes, transferrin saturation was not significantly different, whereas the patients prospectively subjected to HFE gene test showed a serum ferritin concentration significantly lower than the clinically diagnosed hemochromatosis patients, 1010 ± 961 versus 1754 ± 1668 μg/L (P = 3.10−12). In compound, heterozygotes (C282Y/H63D, C282Y/S65C, H63D/S65C) and H63D/H63D transferrin saturation tended to be lower (P = 0.03 and 0.07) in patients prospectively subjected to HFE gene test. These data indicate that the HFE gene-based test performed because of medical advice or cascade family screening allows to identify a genetic cause of elevated iron parameter in, at least, 29% of subjects. However, among the individuals subjected to prospective gene-test, a large proportion consists in non-C282Y homozygote patients with high serum ferritin concentration and wherein secondary iron overload cannot be excluded. Furthermore, the male-to-female sex ratio in subjects positive for HFE gene-test, and who showed at least an effective elevated transferrin saturation value, was also affected: it passed from 4:1 in patients diagnosed before the HFE gene discovery to 2:1 after; one should note that it was roughly 1:1 among the C282Y homozygotes. The association of HFE genotyping with iron parameters evaluation allows an early diagnosis of hemochromatosis for patients, like women, lacking of clear clinical symptoms, but displaying mild iron overload.

Table 4 Age and iron parameters at diagnosis in hemochromatosis patients tested for HFE gene mutations and propectively tested subjects
Full size table

DISCUSSION

Hemochromatosis is a common recessive genetic disorder in Northern Europe, and it is one among the rare cases of genetic diseases that can be treated. The treatment of the disease by iterative phlebotomy is simple, safe and effective in the prevention of irreversible damages of iron overload when started early. However, before HFE gene discovery early diagnosis was difficult because clinical history was associated with non specific symptoms; moreover the biochemical tests based on measured serum iron parameters, i.e., serum transferrin saturation and serum ferritin concentration, may have given false-positive results. The definitive diagnosis of hemochromatosis was traditionally based on an evaluation of hepatic iron content from liver biopsy or quantitative venesection, but failed in the complete identification of patients in the early stages of iron accumulation. The strategy for the diagnosis of hereditary hemochromatosis evolved in 1996, further to the availability of the HFE gene test. Testing for HFE gene mutations now plays a key-role in the confirmation of hemochromatosis diagnosis. Since 1997, most of the patients tested for HFE gene have been prospectively tested because of incidental finding of an elevated serum iron parameter and/or a family history of the disease. Thus, the classical clinical features of hemochromatosis had not yet occurred at the time of prospective HFE gene test. Nowadays, in most cases, invasive liver biopsy is no longer required to establish a confident diagnosis of the disease. It has been estimated that liver biopsy needs to be performed only as a prognosis indicator of fibrosis whenever the serum ferritin level is above 1000 μg/L.31 In addition, after the HFE gene discovery, patients were commonly identified as subject for a DNA-based test when their serum iron parameters were lower than those of patients diagnosed hemochromatosis before gene testing. HFE gene test can distinguish individuals with disease-related genotype from those with other causes of abnormal iron values or clinical symptoms; thus, the HFE gene test performed on individuals with suspicion of iron overload allows one to detect HFE-associated mutations in individuals at an early stage of iron overload, when biochemical expression is reached.

The overall frequencies of HFE C282Y and H63D alleles, 7.7% and 14%, respectively, and genotypes among the Brittany population are comparable to other data available for the European population indicating 5% to 10% and 10% to 20% of C282Y and H63D allele frequencies, respectively.32,33 The frequencies of the genotypes based on the detection of 478 probands phenotypically diagnosed before gene discovery from Brittany population agree with those reported about European descent, with 82.3% to 100% C282Y/C282Y and 4% to 5.44% C282Y/H63D.14–18 In the population of Brittany, the carrier prevalence of C282Y and H63D alleles, 12%, and 23.6%, respectively, underlined the high rate of transmission of alleles potentially associated with iron overload. Our set of data about the numerous HFE gene tests conducted within the blood center in Brest on relatives of identified hemochromatosis-affected patients and subjects with evidence of modified iron parameters, i.e., elevated transferrin saturation or high serum ferritin level, clearly demonstrates that gene testing was helpful in diagnosis clarification and to identify patients at risk of developing hemochromatosis. Indeed, although prospectively tested patients had been phenotypically and genotypically more heterogeneous than those diagnosed hemochromatosis before HFE gene discovery, we found high frequencies of disease-related genotypes among these patients selected upon nonstrict hemochromatosis criteria after HFE gene discovery compared to controls. A large proportion of these individuals may never develop clinical symptoms due to incomplete penetrance of the HFE mutations.28,34 Family studies as well as population screening revealed not only the low penetrance of the H63D/H63D and compound heterozygous genotypes C282Y/H63D, C282Y/S65C, and H63D/S65C, but also the reduced penetrance of the C282Y/C282Y genotype. Numerous studies have described nonexpressing C282Y homozygous individuals although the most severe cases of iron overload in classic hemochromatosis are associated with this genotype: as a whole, 33% of women and 6.7% of men homozygous for C282Y mutation do not show significantly elevated iron stores,20 and many C282Y homozygotes, between 25% and 81%, have been found to display normal serum ferritin levels.3,35,36 However, on condition to appropriately inform patients, the HFE gene test should allow a biochemical follow-up of these individuals at risk of developing severe iron overload and serious irreversible damage. It should also make phlebotomy treatment more beneficial through the prevention of further complications together with a reduced morbidity for the most severe penetrant cases.

The patients prospectively tested showed a lower proportion of positive HFE gene test, 42.8% for 1997 to mid-1998 and 29.5% for mid-1998 to mid-2002, than those diagnosed hemochromatosis before 1996 and retrospectively tested (89.7%). The higher proportion over the 1997 to mid-1998 period can be attributed to hemochromatosis patients still retrospectively tested. Despite the decrease in the number of positive HFE gene test, it still remains higher than the control one; thus, prospective testing based on iron loading signs is sufficient to identify individuals at risk of developing HFE-associated hemochromatosis. In addition, the genotypes of positive-tested subjects differed significantly between groups. The frequencies of H63D homozygotes, all compound heterozygotes, and carriers were higher in prospectively tested subjects than in clinically diagnosed hemochromatosis patients. This likely results from the high frequency of the most penetrant C282Y/C282Y genotype group within clinically diagnosed patients that leads to a selection against other HFE genotypes associated with hemochromatosis. In addition, serum ferritin concentration was significantly lower in prospectively tested C282Y/C282Y subjects than in those from clinical hemochromatosis group; this parameter being gender- and age-related, thus the sex-ratio in these two groups could explain this difference. Symptoms of hemochromatosis are more frequent in males than in females probably due to the protection induced by menstrual blood loss and pregnancies. Before gene test availability, the male-to-female sex ratio of hemochromatosis individuals was 4:1, which is far from the expected 1:1 ratio for a recessive trait. It is noteworthy that the male-to-female sex ratio of hemochromatosis patients since HFE gene test-based diagnosis has become 2:1 and is about 1:1 among C282Y homozygotes. It seems that, at the time of no available HFE gene-test, women were under-diagnosed because of their lower iron parameter values, and compared to men who usually present with cirrhosis and diabetes, they showed other and milder symptoms such as fatigue and pigmentation.37 These data indicate that HFE gene test allows a diagnosis based on lower serum iron parameters and/or mild symptoms, which in the future will be beneficial for the female population. The iron loading in hemochromatosis is sex- and age-related, and the incomplete penetrance of the HFE genotypes demonstrates uncertainty about genotype-phenotype correlations. However, considering the HFE gene-test as a diagnosis criterion, it helps to confirm a genetic cause of iron overload, and gives the opportunity to ascertain a diagnosis based on abnormal iron values or unspecific signs under strict phenotypic hemochromatosis diagnosis, it is thus appropriate for a biochemical follow-up in people with disease-related genotypes.