Hemochromatosis, Type 1

A number sign (#) is used with this entry because hemochromatosis type 1 (HFE1) is caused by homozygous or compound heterozygous mutation in the HFE gene (613609) on chromosome 6p22.

Description

Hereditary hemochromatosis is an autosomal recessive disorder of iron metabolism wherein the body accumulates excess iron (summary by Feder et al., 1996). Excess iron is deposited in a variety of organs leading to their failure, and resulting in serious illnesses including cirrhosis, hepatomas, diabetes, cardiomyopathy, arthritis, and hypogonadotropic hypogonadism. Severe effects of the disease usually do not appear until after decades of progressive iron loading. Removal of excess iron by therapeutic phlebotomy decreases morbidity and mortality if instituted early in the course of the disease. Classic hemochromatosis (HFE) is most often caused by mutation in a gene designated HFE on chromosome 6p21.3.

Adams and Barton (2007) reviewed the clinical features, pathophysiology, and management of hemochromatosis.

Genetic Heterogeneity of Hemochromatosis

At least 4 additional iron overload disorders labeled hemochromatosis have been identified on the basis of clinical, biochemical, and genetic characteristics. Juvenile hemochromatosis, or hemochromatosis type 2 (HFE2), is autosomal recessive and is divided into 2 forms: HFE2A (602390), caused by mutation in the HJV gene (608374) on chromosome 1q21, and HFE2B (613313), caused by mutation in the HAMP gene (606464) on chromosome 19q13. Hemochromatosis type 3 (HFE3; 604250), an autosomal recessive disorder, is caused by mutation in the TFR2 gene (604720) on chromosome 7q22. Hemochromatosis type 4 (HFE4; 606069), an autosomal dominant disorder, is caused by mutation in the SLC40A1 gene (604653) on chromosome 2q32. Hemochromatosis type 5 (HFE5; 615517) is caused by mutation in the FTH1 gene (134770) on chromosome 11q12.

Clinical Features

Muir et al. (1984) recognized 4 different types of hereditary hemochromatosis which 'bred true' in families, suggesting that more than one genetic lesion in iron metabolism can lead to hereditary hemochromatosis. Group I was termed the classic form with elevated transferrin (190000) saturation, serum ferritin levels, and liver iron content; group II was characterized by severe iron overload and accelerated disease manifesting at an early age; group III was characterized by elevated total body iron stores, normal transferrin saturation and serum ferritin levels; and group IV was characterized by markedly elevated findings on serum biochemical tests, i.e., transferrin saturation and serum ferritin, with minimal elevation in total body iron stores. Milman et al. (1992) found no relationship between genetic subtypes of transferrin and the expression of disease in hemochromatosis patients.

Edwards et al. (1980) identified 35 hemochromatosis homozygotes through pedigree studies, using the close linkage to HLA-A (142800) in the identification. Thirteen were asymptomatic. Arthropathy was present in 20, hepatomegaly in 19, transaminasemia in 16, skin pigmentation in 15, splenomegaly in 14, cirrhosis in 14, hypogonadism in 6, and diabetes in 2. None had congestive heart failure. Only 1 had the triad of hepatomegaly, hyperpigmentation, and diabetes. Serum iron was increased in 30 of 35, transferrin saturation was increased in all 35, serum ferritin in 23 of 32, urinary iron excretion after deferoxamine in 28 of 33, hepatic parenchymal cell stainable iron in 32 of 33, and hepatic iron in 27 of 27. Iron loading was 2.7 times greater in men than in women. No female had hepatic cirrhosis.

By studying 1,058 individuals who were heterozygous for the HLA-linked hemochromatosis mutation, Bulaj et al. (1996) found that the mean serum iron concentrations and transferrin-saturation values were higher in heterozygotes than in normal subjects and did not increase with age. Initial transferrin-saturation levels exceeding the threshold associated with the homozygous genotype were found in 4% of males and 8% of female heterozygotes. The geometric mean serum ferritin concentration was higher in heterozygotes than in normal subjects and increased with age. Higher-than-normal values were found in 20% of males and 8% of female heterozygotes. The clinical and biochemical expression of hemochromatosis was more marked in heterozygotes with paternally transmitted mutations than in those with maternally transmitted mutations. Liver biopsy abnormalities were generally associated with alcohol abuse, hepatitis, or porphyria cutanea tarda. Bulaj et al. (1996) concluded that complications due to iron overload alone in hemochromatosis heterozygotes are 'extremely rare.' This was the first description of parent-of-origin effects in hemochromatosis.

Escobar et al. (1987) established the diagnosis of hemochromatosis in a 7-year-old boy and his 29-month-old brother. These were said to be the youngest children with primary hemochromatosis reported to that time. They were members of a family in which 3 generations had affected individuals. Data from the literature on values of serum iron, serum ferritin, transferrin saturation, and hepatic iron were reviewed. Kaikov et al. (1992) described hemochromatosis in asymptomatic sibs in whom the diagnosis was made after an unexpected finding of elevated serum iron concentrations. The sibs were 7, 6, and 4 years of age. Elevated red cell mean corpuscular volume (MCV) was elevated in all 3, at 90 to 92 fL. In their review of the literature, they found 16 cases of symptomatic homozygous children at ages ranging from 4 to 19 years at the time of diagnosis. They suggested that normalization of the MCV may be an indirect index of adequate phlebotomy. The cases of Escobar et al. (1987) and Kaikov et al. (1992) may have been juvenile hemochromatosis (602390).

Roldan et al. (1998) described acute liver failure after iron supplementation in a 29-year-old woman with unrecognized hemochromatosis.

Roy and Andrews (2001) reviewed disorders of iron metabolism, with emphasis on aberrations in hemochromatosis, Friedreich ataxia (FRDA; 229300), aceruloplasminemia (604290), and other inherited disorders.

McDermott and Walsh (2005) assessed the prevalence of hypogonadism in a large group of patients with hemochromatosis diagnosed in a single center over a 20-year period. Abnormally low plasma testosterone levels, with low luteinizing hormone (LH; see 152780) and follicle-stimulating hormone (FSH; see 136530) levels, were found in 9 of 141 (6.4%) male patients tested. Eight of nine (89%) had associated hepatic cirrhosis; 3 of 9 (33%) had diabetes. Inappropriately low LH and FSH levels were found in 2 of 38 females (5.2%) in whom the pituitary-gonadal axis could be assessed. McDermott and Walsh (2005) concluded that patients with lesser degrees of hepatic siderosis at diagnosis are unlikely to develop hypogonadism.

Liver Cirrhosis and Liver Cancer

Deugnier et al. (1993) analyzed the occurrence of primary liver cancer in hemochromatosis; there was 1 instance of cholangiocarcinoma and 53 instances of hepatocellular carcinoma (HCC; 114550). Of the 54 patients, 32 were untreated and 22 had been 'de-ironed.' Three of the patients had hepatocellular carcinoma in noncirrhotic but only fibrotic liver. Chronic alcoholism and tobacco smoking was higher in patients with hepatocellular carcinoma than in matched hemochromatosis patients without carcinoma.

A common manifestation of tissue damage caused by iron accumulation in hereditary hemochromatosis is hepatic cirrhosis that may lead to hepatocellular carcinoma. Willis et al. (2000) determined the risk of developing such disease manifestations in individuals with HFE mutations in Norfolk, UK. The frequency of mutant HFE alleles in archived liver tissue blocks from patients with cirrhosis or liver cancer was compared with that in 1,000 control blood samples. This control group was derived from a number of sources; no sample was from an individual with diagnosed HH. Of 34 cases of liver cancer, 3 (8.8%) were homozygous for the C282Y (613609.0001) mutation (2 hepatocellular carcinomas, 1 undifferentiated liver carcinoma). None of these patients had been given a diagnosis of HH prior to the diagnosis of liver cancer. None were C282Y/H63D (613609.0002) compound heterozygotes. Five of 190 cirrhosis samples (2.6%) were homozygous for C282Y; 4 of these patients had been given a clinical diagnosis of HH at the time of biopsy, and the remaining case fell also into the liver cancer group. Six cirrhosis samples were from C282Y/H63D compound heterozygotes; none had been given a clinical diagnosis of HH. The frequency of C282Y homozygotes in the control group was 1 in 230, and of C282Y/H63D compound heterozygotes was 1 in 108. HFE mutations were significantly more common in disease than in control specimens. Willis et al. (2000) calculated that, in their population, 2.7% of C282Y homozygotes and 1% of C282Y/H63D compound heterozygotes develop liver disease at some point in their lives.

Both Wilson disease (WND; 277900) and hemochromatosis, characterized by excess hepatic deposition of iron and copper, respectively, produce oxidative stress and increase the risk of liver cancer. Because the frequency of p53 mutated alleles (191170) in nontumorous human tissue may be a biomarker of oxyradical damage and identify individuals at increased cancer risk, Hussain et al. (2000) determined the frequency of p53 mutated alleles in nontumorous liver tissue from WND and hemochromatosis patients. When compared with the liver samples from normal controls, higher frequencies of G:C to T:A transversions at codon 249, and C:G to A:T transversions and C:G to T:A transitions at codon 250 were found in liver tissue from WND cases, and a higher frequency of G:C to T:A transversions at codon 249 was also found in liver tissue from hemochromatosis cases. Sixty percent of WND and 28% of hemochromatosis cases also showed a higher expression of inducible nitric oxide synthase in the liver, which suggested nitric oxide as a source of increased oxidative stress. The results were consistent with the hypothesis that the generation of oxygen/nitrogen species and unsaturated aldehydes from iron and copper overload in hemochromatosis and WND causes mutation in the p53 tumor suppressor gene.

Other Features

Chromium, an essential trace mineral required for normal insulin function, is transported bound to transferrin and competes with iron for that binding. Sargent et al. (1979) found that less chromium is retained in patients with hemochromatosis than in controls, and suggested that the diabetes of hemochromatosis may be due in part to chromium deficiency.

Murphy (1987) noted that a considerable proportion of the patients who develop Vibrio vulnificus septicemia are persons with hemochromatosis. This organism thrives in an environment with abundant iron. It occurs naturally in many warm coastal waters and sometimes contaminates shellfish harvested from these areas. The organism can cause infection when ingested in raw or improperly cooked contaminated shellfish or when introduced into the open wounds of persons who handle contaminated seafood or bathe in contaminated waters. Bacteremia due to V. vulnificus in patients with hemochromatosis may be related to the availability of iron for microbial metabolism or to the presence of hepatic cirrhosis (Bullen et al., 1991) and is often fatal.

Diamond et al. (1989) studied the prevalence and pathogenesis of osteopenia in 22 men with hemochromatosis. They concluded that a significant decrease in bone density is observed in this condition, particularly when hypogonadism is present. They speculated that low serum free-testosterone concentrations, rather than calciotrophic hormones, determine bone mass in this disorder.

Barton et al. (1994) demonstrated that hemochromatosis homozygotes and, to a lesser extent, heterozygotes, both male and female, have increased blood levels of lead. In contrast, mean blood lead of subjects with transfusion-induced iron overload did not differ significantly from that of normal controls. The findings in homozygotes could not be related to age, presence or absence of iron loading, or the extent of therapeutic phlebotomy. Increased absorption of iron and cobalt, which may have the same absorptive pathway, had previously been documented in homozygotes; the new findings were interpreted as indicating increased absorption of lead as well. The findings suggested that patients with hemochromatosis, like children with iron deficiency, are more susceptible to lead poisoning.

Anand et al. (1983) and Eriksson et al. (1986) described cases suggesting a possible relationship between alpha-1-antitrypsin deficiency (613490) and hemochromatosis. In a series of 15 patients referred to a liver transplantation center in the U.S., Rabinovitz et al. (1992) found a significant correlation between heterozygous PiZ (107400.0011) alpha-1-antitrypsin deficiency and hemochromatosis. Other studies, however, failed to show a relationship between the 2 inborn errors of metabolism. To investigate the matter further, Elzouki et al. (1995) used a monoclonal antibody against the PiZ variant of AAT in 67 consecutive patients with genetic hemochromatosis seen in 2 Swedish hospitals. In 3 of the patients with hemochromatosis, homozygosity for the PiZ variant was found. Liver biopsy was performed in 65 of the 67 patients; 2 of the 3 PiZ homozygotes were found to have cirrhosis, compared to 10% (6 of 59) of the noncarriers of the PiZ variant. None of the homozygous or heterozygous AAT-deficient patients had developed hepatocellular carcinoma compared with 2 of 59 of the non-PiZ gene carriers. Severe emphysema developed in 2 of the patients with the homozygous phenotype. Elzouki et al. (1995) concluded that the data suggested that the presence of the PiZ allele in double dose when associated with genetic hemochromatosis contributes to the earlier onset of cirrhosis, although it may not increase the risk of hepatocellular carcinoma.

Grove et al. (1998) examined the hypothesis that mutations in the HFE gene determine hepatic iron status in alcoholics and predispose to advanced alcoholic liver disease. The sample population was derived from the northeast of England and consisted of 257 individuals with alcoholic liver disease and 117 controls from the local population. No significant excess of C282Y (613609.0001) or H63D (613609.0002) alleles was demonstrated in alcoholics with advanced liver disease compared to those with no liver disease. There was no difference in age at biopsy or presentation. No difference in allele distribution was noted between alcoholics and controls. No relationship between allele frequency and histologic evidence of iron overload was noted. The authors commented that HFE mutations did not predispose to advanced liver disease in alcoholics.

Because ceruloplasmin (CP; 117700) seems to be involved in iron mobilization, Cairo et al. (2001) measured serum CP levels in 35 patients with hereditary hemochromatosis, 12 patients with acquired iron overload, and 36 healthy subjects. Ceruloplasmin was lower in HH patients than in controls; no difference was found between untreated HH patients and those on a phlebotomy program and between HH patients carrying the normal and mutated alleles of the HFE gene. CP levels in patients with acquired iron overload were significantly higher than in HH patients and similar to those of controls. No differences in albumin, alpha-1-acid glycoprotein, or copper serum levels were observed in the 3 groups.

Cippa and Krayenbuehl (2013) hypothesized that sustained enhanced iron absorption in patients with HFE hemochromatosis may have a beneficial effect on growth. They assessed the height in a cohort of 176 patients with HFE hemochromatosis at the University Hospital Zurich. Homozygous C282Y (613609.0001) mutations were found in 93% of patients, whereas compound heterozygosity for H63D (613609.0002) and C282Y mutations was found in 7%. Height in patients with hemochromatosis was compared with that in an age- and sex-matched Swiss reference population, with the use of data reported in the registry of military conscription and by the Swiss Federal Statistical Office. The mean height in men with hemochromatosis (120) was 178.2 cm, versus 173.9 cm in controls (458,322), a difference of 4.3 cm (95% CI, 3.0 to 5.5; p less than 0.001). The mean height in women with hemochromatosis (56) was 167.1 cm, versus 163.8 cm in controls (10,260), a difference of 3.3 cm (95% CI, 1.3 to 5.3; p less than 0.001). Cippa and Krayenbuehl (2013) speculated that patients with HFE hemochromatosis may benefit in their first 2 decades from constantly enhanced iron absorption, providing a steadily sufficient supply of iron during physical development.

Inheritance

Debre et al. (1958) concluded that the biochemical defect of idiopathic hemochromatosis is present in heterozygotes and that whether the disease develops is dependent on other influences on iron metabolism. They suggested that juvenile hemochromatosis resulting from consanguineous marriages may represent the homozygous state of the gene.

Bothwell et al. (1959), Debre et al. (1958), and several others concluded that 1 form of hemochromatosis is inherited as an autosomal dominant disorder with incomplete penetrance in females because of loss of blood in menstruation and pregnancy. Saddi and Feingold (1974) reported a study of 96 pedigrees which, they concluded, supported autosomal recessive inheritance. Consanguinity was increased among the parents. No parent or offspring was affected. Segregation analysis was consistent with autosomal recessive inheritance if reduced penetrance in females was assumed.

Simon et al. (1977) concluded that idiopathic hemochromatosis is recessive, although polygenic (probably oligogenic) inheritance could not be excluded.

Bassett et al. (1982) provided evidence that clarified some of the previous confusion of whether hemochromatosis is a recessive or a dominant. They observed 5 families with hemochromatosis in 2 successive generations. HLA typing of the subjects indicated that a homozygous-heterozygous mating almost certainly had occurred in 4 of the 5 families, resulting in homozygous offspring. Powell et al. (1987) restudied a family reported by Bassett et al. (1982) in which 2 children apparently homozygous for hemochromatosis did not manifest overt disease; alternative explanations such as dominant inheritance were postulated. Subsequent studies provided the correct explanation (pseudodominant inheritance) and added further evidence for the tight linkage of HFE to HLA-A.

Borecki et al. (1989) performed a segregation analysis on 147 HH pedigrees from Brittany, France, indexed by the measurement of latent capacity of transferrin. No evidence for heterozygous expression was observed, either in the biochemical domain of latent capacity of transferrin, or in increased liability to overt disease. The analysis allowed clear resolution of the recessive single gene inheritance pattern in these families. Borecki et al. (1990) concluded that the hemochromatosis gene is completely recessive with respect to both clinical manifestations and serum iron abnormalities, with significant differences in expression by sex. Clinical manifestations were present in all male homozygotes, suggesting that the recessive hemochromatosis genotype is fully penetrant at all ages in males. This was not the case for younger females, however.

Mapping

Simon et al. (1976) found HLA-A3 in 78.4% of hemochromatosis patients and 27% of controls; HLA-B14 was found in 25.5% of cases and 3.4% of controls. Among sibs with hemochromatosis, Simon et al. (1977) found a highly significant association between hemochromatosis and possession of the same 2 haplotypes. For 6 families a lod score of 2.239 at a recombination fraction of 0.005 supported linkage of HLA and hemochromatosis.

Stevens et al. (1977) concluded that a gene for hemochromatosis may be on chromosome 6 close to the HLA-A locus in linkage disequilibrium with high frequency of A3 in patients with hemochromatosis.

Cartwright et al. (1978) obtained lod scores well above the 3.0 for the HLA-hemochromatosis linkage. That the high lod score is not an artifact due to A3, B7 and B14 associations was supported by the finding of a lod score of 4.14 at theta 0.00 in 5 pedigrees in which these antigens were not present in the probands (Dadone et al., 1982). Skolnick (1983) contended that linkage disequilibrium cannot explain the HLA-hemochromatosis association because the association is with a haplotype, either A3-B7 or A3-B14.

Edwards et al. (1985, 1986) presented the first known example of recombination between the HLA-A and hemochromatosis loci and proposed that the (or at least a) hemochromatosis locus lies between the HLA-A and HLA-B loci.

David et al. (1986, 1987) studied an exceptional recombinant family with 3 HLA-identical sibs: 1 had hemochromatosis, whereas the other 2 were free of any clinical or biologic signs of the disease. The study of restriction patterns using 2 MHC class I probes showed 2 differences between the proband and his sibs which were attributed to an unbalanced crossover or a genetic conversion. The absence of a 7.7-kb HindIII fragment in the proband suggested that this segment is the location of at least part of the hemochromatosis gene. Furthermore, it appeared that the hemochromatosis gene lies telomeric to the HLA-A locus. Lucotte and Coulondre (1986) found that a specific PvuII restriction fragment correlates absolutely with the HLA-A3 serologic allele and with the hemochromatosis allele.

Using pulsed field gel electrophoresis in conjunction with probes that map within, or in the vicinity of, the HLA class I region, Lord et al. (1990) did not detect any disease-specific differences in affected members of 3 HH pedigrees or in 6 unrelated patients with the disorder. The authors concluded that the lesion responsible for HH lies beyond the resolution of this technique and does not involve large structural deletions or extensive rearrangements.

Boretto et al. (1992) reported linkage studies with restriction polymorphisms which were consistent with location of the hemochromatosis locus either less than 100 kb centromeric to the HLA-A locus or on its telomeric side.

Jazwinska et al. (1993) found a maximum lod score of 9.90 at theta = 0.0 for HLA-A and 8.26 at theta = 0.0 for a microsatellite marker at D6S105. No recombination was observed with either marker. Other markers were separated from the hemochromatosis locus by recombination, thereby defining the centromeric and telomeric limits for the HFE gene as HLA-B and D6S109, respectively. A multipoint map indicated that hemochromatosis locus is located in a region less than 1 cM proximal to HLA-A and less than 1 cM telomeric of HLA-A.

In a single family with hemochromatosis, Calandro et al. (1995) identified 2 recombinant individuals confirmed by analysis of 16 polymorphic markers located near HLA-A and D6S105. One of the recombinants provided evidence that the HH gene is telomeric to the 5-prime end of the HLA-F locus. The HLA-F locus was placed approximately 0.027 cM distal to HLA-A, which in turn was 0.01 cM distal of HLA-B. Raha-Chowdhury et al. (1996) showed that a highly polymorphic polypurine tract in the 5-prime untranslated region of HLA-F is as strongly associated with hemochromatosis as HLA-A3 or D6S105-8. The observed frequency of heterozygosity at the HLA-F polymorphism was 95% and the locus was found to be informative in pedigrees that are not informative at HLA-A and D6S105.

By fluorescence in situ hybridization analysis, Hashimoto et al. (1995) mapped the HFE gene to chromosome 6p22.

Heterogeneity

Edwards et al. (1981) suggested that 2 families reported by Wands et al. (1976) and Rowe et al. (1977) may have had a rare distinct form of hemochromatosis. In these families, neither serum ferritin concentration nor transferrin saturation was a reliable indicator of hepatic siderosis and fibrosis. Hepatic fibrosis was observed in some individuals with a very modest increase in hepatic iron and in a few individuals with normal hepatic iron content. The disorder appeared to be transmitted as an autosomal dominant. No HLA data were reported in these families.

In Australia, Jazwinska et al. (1996) found that all patients of northern European origin with hemochromatosis were homozygous for the cys282-to-tyr mutation (C282Y; 613609.0001). The frequency was greater than 90% in Brittany (Jouanolle et al., 1996). However, in Italy, Carella et al. (1997) performed mutation analysis on the HFE gene in patients from families with the 6p-linked disease but without the C282Y mutation and failed to find nucleotide abnormalities in coding sequences and intron/exon boundaries that could account for the disorder. The negative findings of RNA-SSCP were supported by the absence of mutations in the HFE gene by direct sequencing. Major deletions or rearrangements of the gene were excluded by Southern blotting. Carella et al. (1997) concluded that hemochromatosis in Italy appears to be more heterogeneous than reported in northern Europe, and suggested abnormalities in unexplored portions of introns, RNA untranslated regions, regulatory elements, or another tightly linked locus as alternative possibilities for the cause of the disorder. Studies by Carella et al. (1997) and Piperno et al. (1998) indicated that only 64% of patients with hemochromatosis in Italy were homozygous for the C282Y mutation.

In commenting on the report of Carella et al. (1997), Beutler (1997) pointed to the 0.01 gene frequency in the Italian population, which is considerably lower than in persons of European ancestry who have been studied in the United States and in northern Europe. In agreement with the data from this southern European population, Beutler and Gelbart (1997) found that among nearly 400 Ashkenazi Jews the gene frequency of the C282Y mutation was only 0.013, compared with 0.07 in the non-Jewish American white population. These findings and those of Carella et al. (1997) seem consistent with the putative Celtic origin of the C282Y mutation (Jazwinska et al., 1995).

Molecular Genetics

In patients with hereditary hemochromatosis, Feder et al. (1996) identified 2 mutations in the HFE gene (C282Y; 613609.0001 and 613609.0002). The C282Y mutation was detected in 85% of all HFE chromosomes, indicating that in their population 83% of hemochromatosis cases are related to C282Y homozygosity.

Beutler et al. (1997) pointed out that calreticulin (CALR; 109091), like beta-2-microglobulin (B2M; 109700), associates with class I HLA proteins and appears to be identical to mobilferrin, a putative iron transport protein. Thus these 2 proteins were considered candidates for mutations in patients with hemochromatosis. The investigators sequenced the coding region and parts of introns of the HFE gene (called by them HLA-H), the B2M gene, and the CALR gene in 10, 7, and 5 hemochromatosis patients, respectively, selecting those who were not homozygous for the common C282Y mutation. No additional mutations were found in the HLA-H gene and no disease related mutations in the other 2 genes. The authors noted that the basis for hemochromatosis in more than 10% of European patients and in most Asian patients awaits explanation. Beutler et al. (1997) speculated that the finding of some effects in heterozygotes (Bulaj et al., 1996) and the rarity of mutations other than C282Y and his63 to asp (H63D; 613609.0002) may point to a gain-of-function consequence of these mutations, similar, they suggested, to sickle cell anemia, which is caused by only 1 type of mutation (see 141900.0038) and represents in effect a gain-of-function mutation. The unique mutation causing achondroplasia, gly380 to arg (G380R; 134934.0001), might also be cited.

By sequence analysis of exons 2, 3, 4, and 5, and portions of introns 2, 4, and 5 of the HFE gene, Barton et al. (1999) identified novel mutations in 4 of 20 hemochromatosis probands who lacked C282Y homozygosity, C282Y/H63D compound heterozygosity, or H63D homozygosity. Probands 1 and 2 were heterozygous for the previously undescribed mutations ile105 to thr (I105T; 613609.0009) and gly93 to arg (G93R; 613609.0010). Probands 3 and 4 were heterozygous for the previously described but uncommon HFE mutation ser65 to cys (S65C; 613609.0003). Proband 3 was also heterozygous for C282Y and had porphyria cutanea tarda (see 176100), and proband 4 had hereditary stomatocytosis (185000). Each of these 4 probands had iron overload. In each proband with an uncommon HFE coding region mutation, I105T, G93R, and S65C occurred on separate chromosomes from those with the C282Y or H63D mutations. Neither I105T, G93R, nor S65C occurred as spontaneous mutations in these probands. In 176 normal control subjects, 2 were heterozygous for S65C, but I105T and G93R were not detected.

Griffiths and Cox (2000) reviewed the molecular pathophysiology of iron metabolism.

Pietrangelo (2004) reviewed the various forms of hemochromatosis. In a useful diagram, he illustrated the polygenic nature and phenotypic continuum of hereditary hemochromatosis. The continuum involves age at onset, clinical severity, and contribution of host or environmental factors to expressivity. Intermediate phenotypes can result from combined heterozygous mutations (compound heterozygosity) or homozygous mutations of more than 1 hemochromatosis gene. For instance, the relatively mild phenotype associated with homozygous mutation of HFE can be aggravated and accelerated by a coexisting heterozygous mutation in a gene associated with a juvenile form of the disease, such as HAMP. The latter mutation, combined with a normally silent heterozygous HFE mutation, can also result in unexpected expression of disease.

Lee et al. (2004) identified a patient with adult-onset hemochromatosis who was compound heterozygous for mutations in the HJV gene (G320V, 608374.0001; 608374.0007).

Genetic Modifiers

In patients with 'atypical' hemochromatosis, defined as having a discordant iron phenotype despite having the same HFE genotype, Hofmann et al. (2002) performed mutation analysis of the transferrin receptor-2 gene (TFR2), which is mutated in HFE3. Sib pairs homozygous for HFE C282T had a discordant phenotype in serum transferrin concentration and/or significant differences in liver fibrosis and liver enzyme levels. Also included were individuals who were not homozygous for C282Y, but who had evidence of iron excess. In a pair of brothers homozygous for the C282Y mutation, Hofmann et al. (2002) found a mutation in the TFR2 only in the brother with liver fibrosis, suggesting that TFR2 functions as a modifier for penetrance of the hemochromatosis phenotype when present with homozygosity for C282Y. The screening for mutations in all 18 exons indicated that mutations of the TFR2 gene are rare.

Merryweather-Clarke et al. (2003) described 2 families who exhibited digenic inheritance of hemochromatosis. In family A, the proband had a JH phenotype and was heterozygous for the C282Y mutation in the HFE gene as well as a frameshift mutation in the HAMP gene (606464.0003). The proband's unaffected mother was also heterozygous for the HAMP frameshift mutation, but lacked the HFE C282Y mutation and was heterozygous for the HFE H63D mutation (613609.0002). In family B, there was a correlation between severity of iron overload, heterozygosity for a HAMP G71D mutation (606464.0004), and heterozygosity or homozygosity for the HFE C282Y mutation. The authors proposed that the phenotype of C282Y heterozygotes and homozygotes may be modified by heterozygosity for mutations which disrupt the function of hepcidin in iron homeostasis, with the severity of iron overload corresponding to the severity of the HAMP mutation.

Among 310 C282Y homozygous HFE patients, Le Gac et al. (2004) found 9 patients with an additional heterozygous HJV mutation, including the L101P (608374.0006) and G320V mutations. Iron indices of 8 of these patients appeared to be more severe than those observed in sex- and age-matched C282Y homozygotes without an HJV mutation. Mean serum ferritin concentrations of the 6 males with an HJV mutation were significantly higher than those of C282Y homozygous males without an HJV mutation.

Using pretherapeutic serum ferritin levels in C282Y homozygotes as a marker of penetrance, Milet et al. (2007) found an association between a common T/C SNP in the 3-prime region of the BMP2 gene (112261), rs235756, and hemochromatosis penetrance. Mean ferritin level, adjusted for age and sex, was 655 ng/ml among TT genotypes, 516 ng/ml in TC genotypes, and 349 ng/ml in CC genotypes. The subjects studied were all homozygous for the common C282Y mutation. The results further suggested an interactive effect on serum ferritin level of rs235756 in BMP2 and a SNP in HJV (608374), with a small additive effect of a SNP in BMP4 (112262).

Le Gac et al. (2008) reported a 47-year-old woman of Sardinian descent who presented with mild hemochromatosis. Genetic analysis showed that she was homozygous for a deletion involving the entire HFE gene; however, her phenotype was relatively mild and similar to that of women homozygous for the common lower-penetrance C282Y mutation. The report indicated that additional genetic and environmental factors must play a role in the pathogenesis of the disease.

Genotype/Phenotype Correlations

Dadone et al. (1982) found saturation of transferrin above 62% to be the best simply measured indicator of genotype: homozygosity was accurately predicted in 92% of cases. The logarithmic scale of serum ferritin concentration was only 71% accurate. The frequency of the hemochromatosis gene was estimated at 0.069 +/- 0.020, corresponding to a heterozygote frequency of 0.13 and a homozygote frequency of 0.005.

Barton et al. (1999) studied the phenotype-genotype correlation in 150 family members (72 males and 78 females) of 61 Caucasian American probands. Thirty-four of the family members had an HFE phenotype. Genotyping was limited to the 2 major alleles, C282Y and H63D. Among the family members, 92% of C282Y homozygotes, 34.5% of C282Y/H63D compound heterozygotes, and none of the H63D homozygotes had the HFE phenotype. In contrast, a few individuals heterozygous for one or the other allele had iron overload. Pseudodominant patterns of inheritance were not infrequently observed. Hence, phenotyping and genotyping are complementary in screening for hemochromatosis among family members of probands.

Mura et al. (2001) studied 545 probands who were homozygous for the C282Y mutation (613609.0001), showed various signs of clinical hemochromatosis, and had been referred for treatment by phlebotomy. Iron loading was found to be significantly lower in females than in males and to be correlated with increasing age in both males and females. A study of 18 same-sex sib pairs showed no correlation of iron marker status between HH sibs and other sibs, indicating a variable phenotypic expression of iron loading independent of the HFE genotype. Mura et al. (2001) also found that transferrin saturation percentage was the best indicator of the hereditary hemochromatosis phenotype in young subjects, and serum ferritin concentration was the best marker of iron overload in these patients.

The superoxide dismutase-2 (SOD2; 147460) val16 allele (147460.0001) has 30 to 40% lower enzyme activity and increases susceptibility to oxidative stress. Valenti et al. (2004) found a significantly increased frequency of the val16 allele among 217 unrelated patients with hereditary hemochromatosis who developed dilated or nondilated cardiomyopathy compared to HH patients without cardiomyopathy and controls (frequencies of 0.67, 0.45, and 0.52, respectively). The val/val genotype conferred a 10.1-fold increased risk for cardiomyopathy in the HH patients. The association was independent of cirrhosis, diabetes, arthropathy, and hypogonadism, and did not apply to ischemic heart disease. Valenti et al. (2004) concluded that the val16 allele increased the risk of cardiomyopathy due to iron overload toxicity and oxidation in HH patients as a result of decreased activity of the SOD2 enzyme.

To test whether common HFE mutations that associate with this condition and predispose to increases in serum iron indices are overrepresented in diabetic populations, Halsall et al. (2003) determined the allele frequencies of the C282Y (613609.0001) and H63D (613609.0002) HFE mutations among a cohort of 552 patients with typical type 2 diabetes mellitus. There was no evidence for overrepresentation of iron-loading HFE alleles in type 2 diabetes mellitus, suggesting that screening for HFE mutations in this population is of no value.

Diagnosis

Early diagnosis of hemochromatosis by clinical features is difficult, but important because organ damage can be prevented by early therapy. Hepatic iron is the most sensitive index of preclinical disease; of noninvasive tests, serum ferritin is unreliable, whereas transferrin saturation correlates with hepatic iron content (Rowe et al., 1977; Edwards et al., 1977). Unexplained elevation of transferrin saturation should prompt study for hemochromatosis, and elevated serum iron is a diagnostically valuable finding which can be sought in relatives of full-blown cases.

On the basis of data generated by an ongoing study of hemochromatosis in Brittany, France, Borecki et al. (1990) concluded that percent transferrin saturation is a reliable indicator of the homozygous state but that, contrary to previous studies, there is no evidence for partial expression of this value in heterozygotes.

Phatak et al. (1998) reported that the prevalence of clinically proven and biopsy-proven hemochromatosis combined was 4.5 per 1,000 in a total sample of 16,031 primary care patients and 5.4 per 1,000 in white persons in the sample. The prevalence was higher in men than in women. Diagnosis was achieved by serum transferrin saturation, followed by the same test under fasting conditions and supplemented by serum ferritin levels. Patients with a fasting serum transferrin saturation of 55% or more and a serum ferritin level of 200 micro g/L or more with no other apparent cause were presumed to have hemochromatosis and were offered liver biopsy to confirm the diagnosis.

Feder et al. (1996) viewed hemochromatosis as a model disorder for genetic testing since it is a frequent disorder and effective intervention, namely therapeutic phlebotomy, is available. Cox (1996) discussed the importance of their simple PCR-based test to detect homozygosity for the mutant hemochromatosis gene. Powell et al. (1998) pointed out that a DNA-based test for the HFE gene was commercially available, but its place in the diagnosis of hemochromatosis was still being evaluated.

Screening for Hemochromatosis

From a screening of 1,968 employees of 2 large corporations, Leggett et al. (1990) concluded that the prevalence of significant iron overload due to homozygous hemochromatosis warranting treatment is approximately 1 in 300 among Australians (predominantly Caucasians). They suggested that transferrin saturation should be included in adult health screening programs. Worwood et al. (1991) urged that a regular program be instituted for identifying homozygotes for hemochromatosis on the basis of ferritin concentrations and inviting these individuals to donate frequently to keep the ferritin concentration toward the lower end of the normal range. Such a program would be beneficial both to persons with this common disease and to the blood supply.

In a discussion of the research priorities in hereditary hemochromatosis, Brittenham et al. (1998) commented on anticipating impediments for implementation of a screening program for the disorder: the risk that the genetic information resulting from screening might be used by insurers, employers, or others to deny health care coverage or services to persons identified as being at risk for iron overload; and concern that the diagnosis of hereditary hemochromatosis would lead to changes in self perception, family interactions, and risk-taking behaviors. Because of these considerations, education, counseling, and obtaining informed consent are all important.

Looker and Johnson (1998) did a study to determine the prevalence of an initially elevated serum transferrin saturation and the prevalence of concurrently elevated serum transferrin saturation and serum ferritin levels in the adult population of the United States. They examined 15,839 men and nonpregnant women 20 years of age or older. Depending on the cut-off values used to determine serum transferrin saturation, the prevalence of initially elevated values ranged from 1 to 6%. Approximately 11 to 22% of those with elevated serum transferrin saturation had concurrently elevated serum ferritin levels. Looker and Johnson (1998) concluded that a hemochromatosis screening program that used a cut-off value of greater than 60% to define elevated serum transferrin saturation would identify 1.4 to 2.5 million U.S. adults for further testing.

Hickman et al. (2000) noted that the measurement of transferrin saturation was not suitable for large-scale, automated population screening for HH. The authors developed an automated measurement of unsaturated iron binding capacity and screened 5,182 consecutive blood samples received by a hospital chemical pathology department over 28 consecutive days. Six hundred ninety-seven samples had a value of less than 30 micromoles/liter, the cutoff value for this study. In these samples, measurement of transferrin saturation identified 294 samples for further analysis. HFE C282Y genotyping was possible in 227 of these and identified 9 C282Y homozygotes and 44 C282Y heterozygotes. A clinical diagnosis of HH had been made independently in 2 of the 9 homozygotes. Hickman et al. (2000) concluded that this technique provided a cost-effective screening tool.

Bulaj et al. (2000) examined the usefulness of genetic screening of relatives of probands with hemochromatosis. They studied 291 probands homozygous for mutations in the HFE gene who had presented to a clinic with signs or symptoms of hemochromatosis or who had elevated transferrin-saturation values. They identified 214 homozygous relatives of these 291 homozygous probands. Of the 113 male homozygous relatives (mean age, 41 years), 96 (85%) had iron overload, and 43 (38%) had at least 1 disease-related condition. Of the 52 men over 40 years of age, 27 (52%) had at least 1 disease-related condition. Of the 101 female homozygous relatives (mean age, 44 years), 69 (68%) had iron overload, and 10 (10%) had at least 1 disease-related condition. Of the 43 women over 50 years of age, 7 (16%) had at least 1 disease-related condition. If the proband had a disease-related condition, male relatives were more likely to have morbidity than if the proband had no disease-related condition. Bulaj et al. (2000) concluded that a 'substantial number' of homozygous relatives of patients with hemochromatosis, more commonly men than women, have conditions related to hemochromatosis that had not previously been detected clinically.

Clinical Management

Niederau et al. (1985) concluded that HH patients diagnosed in the precirrhotic stage and treated with therapeutic phlebotomy have a normal life expectancy, whereas cirrhotic patients have a shortened life expectancy and a high risk of liver cancer even when complete iron depletion has been achieved. Siemons and Mahler (1987) found that phlebotomy conducted over a 16-month period restored fertility and normal endocrinologic findings in a 37-year-old man with severe hypogonadotropic hypogonadism due to hemochromatosis.

Barton et al. (1998) recommended that therapeutic phlebotomy to remove excess iron be initiated in men with serum ferritin levels of 300 micrograms/L or more and in women with serum ferritin levels of 200 micrograms/L or more, regardless of the presence or absence of symptoms. Typically, therapeutic phlebotomy consists of removal of 450 to 500 mL of blood weekly until the serum ferritin level is 10