Review article: the modern diagnosis and management of haemochromatosis


Dr P. C. Adams, Department of Medicine, University Hospital, 339 Windermere Rd, London, Ontario N6A 5A5, Canada.


Haemochromatosis is the most common genetic disease in populations of European ancestry. Despite estimates based on genetic testing in Caucasian populations of 1 in 227, many physicians consider haemochromatosis to be a rare disease.

The diagnosis can be elusive because of the non-specific nature of the symptoms. Of all the symptoms, liver disease has the most consistent relationship to haemochromatosis and the prognosis of haemochromatosis is most closely linked to the degree of iron overload.

With the discovery of the HFE gene in 1996, comes new insights into the pathogenesis of the disease and new diagnostic strategies. However, a growing number of new iron-related genes have been discovered and linked to other iron overload syndromes.


Haemochromatosis is the most common genetic disease in populations of European ancestry. Despite estimates based on genetic testing in Caucasian populations of one in 227,1 many physicians consider haemochromatosis to be a rare disease. The diagnosis can be elusive because of the non-specific nature of the symptoms. With the discovery of the HFE gene in 1996,2 comes new insights into the pathogenesis of the disease and new diagnostic strategies.

Clinical features of haemochromatosis

A major problem has been the attribution of clinical symptoms in haemochromatosis to iron overload. It has been assumed that the symptoms of haemochromatosis are related to iron overload causing tissue injury. However, population screening studies which have included control participants have drawn increasing attention to the non-specific nature of the clinical symptoms such as arthralgias, fatigue and even diabetes.1, 3, 4

Liver disease

Although haemochromatosis is often classified as a liver disease, it should be emphasized that it is a systemic genetic disease with multi-system involvement. The liver has a great capacity to accumulate iron within hepatocytes initially without any obvious sequelae both in terms of clinical symptoms or abnormal liver biochemistry. Since hepatic iron presumably accumulates from birth in this genetic disease one would predict a relationship between iron and age. However this may only apply serially within an individual patient. Hepatomegaly remains one of the more common physical signs in haemochromatosis but may not be present in the young asymptomatic homozygote. As patients are detected as young adults through pedigree studies or population screening studies, the prevalence of cirrhosis is much lower than reported in referred patients (Table 1). In a study of 410 referred homozygtotes from Canada and France, 22% had cirrhosis of the liver at the time of diagnosis. The mean Aspartate aminotransferase (AST) and alanine aminotransferase were within the normal range within these 410 patients. Cirrhotic patients and patients with concomitant alcohol abuse were more likely to have abnormal liver enzymes.5 A clinical presentation with marked elevations in liver enzymes and elevated iron tests should suggest an alternate diagnosis such as alcoholic liver disease, chronic viral hepatitis or non-alcoholic steatohepatitis.

Table 1.   The prevalence of cirrhosis in haemochromatosis varies according to population being studied
Population studiedNumber of liver biopsies*Cirrhosis
  1. * Liver biopsies in these studies were based on clinical judgement not study protocol.

Primary care population53 (n = 101 168)221 (5%)
Primary care population32 (n = 65 238)1494 (3%)
Primary care population54 (n = 3011)161 (6%)
Clinical health check and family members5567226 (4%)
Referred patients541190 (22%)

The effect of iron depletion therapy has usually been stabilization of the liver disease. Reversal of liver fibrosis has been described with iron depletion.6 This may account for the relatively small number of C282Y homozygotes that have required liver transplantation.7 A consistent finding with liver transplantation in the setting of iron overload, is a higher mortality post-operatively from sepsis and heart disease.8 Pre-transplant phlebotomy may improve cardiac function and is recommended if tolerated. Chelating agents have been suggested in several case reports. Recurrence of iron overload after liver transplantation has been infrequent.7 Transplantation of iron loaded livers from C282Y homozygotes into recipients has usually resulted in the mobilization of hepatic iron over time and the transplantation of a small intestine and liver into a recipient resulted in the development of iron overload in the recipient.9 These transplant experiments have been fertile ground for speculation on the pathogenesis of haemochromatosis.7

Hepatocellular carcinoma has been described in 18.5% of cirrhotic patients with haemochromatosis.10 It has rarely been described in non-cirrhotic haemochromatosis patients. The relative risk is approximately 200-fold which is similar to cirrhotic patients with chronic viral hepatitis. Screening for hepatocellular carcinoma remains a controversial topic because of the costs of the screening programme, organ shortages allowing for liver transplantation and the small number of candidates that are cured by surgical resection.

There are other symptoms in haemochromatosis including diabetes, arthralgia, fatigue, pigmentation and endocrine problems. Some of these symptoms may be sequelae of cirrhosis and many of these symptoms have not been shown to differ between C282Y homozygotes and a matched control population.

Diagnosis of haemochromatosis

A paradox of genetic haemochromatosis is the observation that the disease is underdiagnosed in the general population, and overdiagnosed in patients with secondary iron overload. A case definition that appeals to all experts has been elusive. A minimum criterion for the diagnosis of haemochromatosis is increased iron stores in the absence of a cause for secondary iron overload. Genetic testing for HFE mutations (C282Y and H63D) has been a major advance with a single mutation explaining most typical cases but there are a growing number of new genetic mutations in other genes that make a HFE specific case definition unsatisfactory.11 The original descriptions of 311 cases of iron overload described by Sheldon in 1935, were likely C282Y related, and many experts continue to use the presence of homozygosity for the C282Y mutation of the HFE gene as the cornerstone of the diagnosis of haemochromatosis.12 A more broad case definition dependent on a degree of iron overload in the liver or as iron mobilized by phlebotomy allows for the consideration of an expanding number of newer genetic mutations which may result in iron overload. An increasing awareness that iron overload can be a sequelae of a wide range of chronic liver diseases has led to many previous cases of ‘haemochromatosis’ being re-classified as iron overload secondary to cirrhosis. From a practical perspective, the clinician should not get too immersed in the debate about whether an individual case has ‘haemochromatosis’ as this is largely a semantic debate. The focus should be on identifying treatable causes of iron overload and initiating the appropriate diagnostic tests and therapies in the patient and other affected family members. It should also be noted that different genetic mutations causing iron overload may also have different clinical consequences and response to iron depletion therapy.

Diagnostic tests for haemochromatosis

A number of diagnostic algorithms based on laboratory tests have been proposed for the diagnosis of haemochromatosis13, 14 (Figure 1). These should be used as guidelines for the clinician and do not replace clinical judgement based on history and physical examination, imaging studies, pathology and pedigree studies.

Figure 1.

 A proposed diagnostic algorithm for patients with suspected haemochromatosis.

Transferrin saturation

The transferrin saturation is often elevated in patients with HFE-linked haemochromatosis. The transferrin saturation has been reported to have a sensitivity of >90% for haemochromatosis. However, this has previously been part of the diagnostic criteria. The sensitivity and specificity of transferrin saturation has usually been established at referral centres where most of the cases have HFE-linked haemochromatosis. The sensitivity of transferrin saturation is lower in population screening studies designed to detect C282Y homozygotes and was only 52% (threshold >50%) in a large screening study from San Diego which included a significant number of cases with a normal serum ferritin.15 The transferrin saturation continues to have a high sensitivity for the detection of an iron loaded haemochromatosis patient. A fasting value has even greater predictive value but may not always be practical. The fasting transferrin saturation is most useful in excluding false positive cases. The transferrin saturation is often elevated in young adults with haemochromatosis before the development of iron overload and a rising ferritin. The threshold to pursue further diagnostic studies has varied from 45  to 62% in previous studies. A lower threshold picks up more patients with haemochromatosis but also leads to more investigations in patients without haemochromatosis. A higher threshold leads to fewer investigations overall with a greater possibility of missing some patients. A common threshold used in screening studies is >45% in women and >50% in men. An elevated transferrin saturation in the presence of a normal serum ferritin rarely indicates significant iron overload but may be a marker that iron overload may develop over time in that patient. Newer genetic mutations may not share this typical pattern of an elevated transferrin saturation and ferritin. A marked elevation (>1000 μg/L) in serum ferritin in the presence of a normal transferrin saturation may still represent significant iron overload and further investigations may be indicated to differentiate iron overload from an inflammatory elevation in ferritin.

Unsaturated iron binding capacity

The unsaturated iron binding capacity (UIBC) is a one step colorimetric assay that has been used in many reference laboratories to calculate the transferrin saturation. It is an inexpensive test compared with transferrin saturation and has been demonstrated to be a promising initial screening test for haemochromatosis.16, 17 A UIBC < 26 μmol/L in men, and <33 μmol/L in women has been proposed as a screening threshold for HFE-linked haemochromatosis with a similar sensitivity and specificity to transferrin saturation.18 The major value is the lower cost and ease of automation and it may be the ideal initial test for high volume population screening studies.

Serum ferritin

The relationship between serum ferritin and total body iron stores has been clearly established by strong correlations with hepatic iron concentration and amount of iron removed by venesection. However, ferritin can be elevated secondary to obesity, chronic alcohol consumption, steatohepatitis, chronic inflammation including viral hepatitis and histiocytic neoplasms. A major diagnostic dilemma in the past was whether the serum ferritin is related to haemochromatosis or another underlying liver disease such as alcoholic liver disease, chronic viral hepatitis or non-alcoholic steatohepatitis (with or without insulin resistance). It is likely that some of these difficult cases will now be resolved by genetic testing. As the ferritin increases the risk of significant liver disease also increases.

Elevations of serum ferritin in the range of 1000–5000 μg/L can be associated with clinical haemochromatosis but a careful investigation for hepatitis B and C, alcoholic liver disease and non-alcoholic steatohepatitis should be considered.19 If another liver disease is the predominant clinical diagnosis, it is more likely that the ferritin elevation is secondary to that disease rather than the concomitant presence of genetic haemochromatosis. In this range of serum ferritin, liver biopsy is often recommended and the review of the liver pathology, iron staining and liver iron concentration is often diagnostic.

The most common problem is the assessment of mild elevations in serum ferritin in the range of 300–1000 μg/L. The prevalence of an increased serum ferritin is so common in men that the origin and appropriateness of a reference range of up to 300 μg/L has been questioned. Genetic testing can be useful to detect early haemochromatosis but most cases will have normal HFE testing, and heterozygosity for a HFE mutation is an unlikely explanation for an elevated ferritin.20 Although it had been considered that heterozygotes could have mild iron overload, since genetic testing has become available it has become apparent that most of these apparent heterozygotes with mild iron overload were actually compound heterozygotes (C282Y/H63D).21 The epidemic of obesity has likely contributed to this high prevalence of ferritin elevations as fatty liver may be the most common cause of an elevated serum ferritin. This is assumed to be related in most cases to inflammation secondary to steatohepatitis and not to iron overload. However, many liver biopsies do not show large amounts of inflammatory cells and correlations with inflammatory markers such as C-reactive protein or ESR have been inconsistent. A liver biopsy is often unappealing to the physician and the patient in the setting of a mild elevation in ferritin. Serial monitoring is often done and a ferritin rising >1000 μg/L is an indication for more investigations or empirical phlebotomy therapy. There are patients with marked elevations in serum ferritin without iron overload. The presence of cataracts at a young age is a clue to the diagnosis of the hyperferritinaemia-cataract syndrome.22 Still's disease and HIV disease are two other conditions in which marked elevations of serum ferritin without iron overload can occur.

Liver biopsy

Liver biopsy has previously been the gold standard diagnostic test for haemochromatosis. Liver biopsy has shifted from a major diagnostic tool to a method of estimating prognosis and concomitant disease. The need for liver biopsy seems less clear now in the young asymptomatic C282Y homozygote where there is a low clinical suspicion of cirrhosis based on history, physical examination and liver biochemistry. A large study conducted in France and Canada suggested that C282Y homozygotes with a serum ferritin of <1000 μg/L, a normal AST, and without hepatomegaly have a very low risk of cirrhosis.23 C282Y homozygotes with a ferritin >1000 μg/L, an elevated AST, and a platelet count <2.0 × 109 had a 77–81% chance of having cirrhosis.24 Patients with cirrhosis have a 5.5-fold relative risk of death compared with the non-cirrhotic haemochromatosis patients.25, 26

Liver biopsy has not been widely accepted by haematologists, public health physicians, geneticists and some haemochromatosis patients, and should always be considered optional. Liver biopsy is considered in typical C282Y homozygotes with liver dysfunction, however, it is most useful in the patient without HFE mutations as it may demonstrate that iron overload is not present and therefore phlebotomy therapy is not required. The distribution of iron within the liver can still provide clues to the cause of the iron overload. Typical C282Y-linked haemochromatosis have a portal to central vein gradient of iron distribution within hepatocytes. A predominance of iron within macrophages in the absence of transfusions may suggest a ferroportin mutation. If the liver biopsy suggests another diagnosis such as alcoholic hepatitis, or chronic viral hepatitis with patchy iron distribution in macrophages, the iron is likely secondary to the primary disease.

Simple C282Y heterozygotes, compound heterozygotes (C282Y/H63D) and patients with other risk factors (alcohol abuse, chronic viral hepatitis) with moderate to severe iron overload (ferritin > 1000 μg/L) may be considered for liver biopsy.

Post-treatment liver biopsies are not routinely done or recommended but reversal of liver fibrosis has been reported after iron depletion.6

Hepatic iron concentration and hepatic iron index

Hepatic iron concentration can be measured using atomic absorption spectrophotometry. This can be done on a piece of paraffin embedded tissue so special preparation is not required at the time of the biopsy. The normal reference range for hepatic iron concentration is 0–35 μmol/g dry weight (<2000 μg/g). The hepatic iron concentration (μmol/g) divided by age (years) is the hepatic iron index. This was demonstrated to be a useful test in differentiating the patient with genetic haemochromatosis from the patient with alcoholic siderosis. Early diagnosis in population screening and pedigree studies have demonstrated many homozygotes with a hepatic iron index <1.9.5 Increasing awareness of the concept of moderate iron overload in cirrhosis of any aetiology has demonstrated many patients without haemochromatosis with a hepatic iron index >1.9. The hepatic iron index has become less useful with the advent of genetic testing. The comment on liver biopsy reports that the hepatic iron index is >1.9 confirming a diagnosis of genetic haemochromatosis should be strongly discouraged. It will remain a tool to aid the clinician with their clinical judgement about an individual case.

Imaging studies of the liver

Magnetic resonance imaging (MRI) can demonstrate moderate to severe iron overload of the liver. Proponents of MRI to alleviate the need for liver biopsy have emphasized the non-invasive nature of the test for the diagnosis. In the study of 174 patients, Gandon et al.27 demonstrated that a simple MRI protocol could detect hepatic iron overload >60 μmol/g (normal range 0–36) with a sensitivity of 89%. A modification of this protocol was used by a second group in the study of 112 patients in which a correlation coefficienct of 0.94 was demonstrated between liver iron concentration and MRI estimation of liver iron concentration.28 These techniques are improving and may be ideally suited to exclude iron overload in a patient in whom an inflammatory condition may be responsible for the elevations in iron tests such as severe alcoholic hepatitis.29 The MRI can also demonstrate the clinical features of cirrhosis such as nodularity of the liver, ascites, portal hypertension and splenomegaly as well as hepatocellular carcinoma, but these features can be more readily assessed by abdominal ultrasound at a lower cost. Magnetic susceptibility has been studied using a SQUID machine to estimate liver iron concentration, but the technology is not widely available (Table 2).30

Table 2.   Differential diagnosis of iron overload
HFE-related haemochromatosis
 C282Y homozygotes (95%)
 C282Y/H63D compound heterozygotes (4%)
 H63D homozygotes (1%)
Non-HFE related haemochromatosis
 Ferroportin disease
 Transferrin receptor 2 mutation
 Juvenile haemochromatosis (young adults with cardiac and endocrine dysfunction)
Miscellaneous iron overload
 African-American iron overload
 African iron overload
 Transfusional iron overload
 Insulin resistance related iron overload
 Alcoholic siderosis
 Iron overload secondary to end stage cirrhosis
 Porphyria cutanea tarda
 Postporta-caval shunt

Genetic testing for haemochromatosis

A major advance that stems from the discovery of the haemochromatosis gene (HFE), is the use of a diagnostic genetic test. Clinical studies in well-defined haemochromatosis pedigrees reported that 90–100% of typical haemochromatosis patients were homozygous for the C282Y mutation of the HFE gene. The presence of a single mutation in most patients is in marked contrast to other genetic diseases in which multiple mutations were discovered (cystic fibrosis, Wilson disease, alpha-1-anti-trypsin deficiency). A second minor mutation, H63D, was also described in the original report.2 This mutation does not cause the same intracellular trafficking defect of the HFE protein. Compound heterozygotes (C282Y/H63D) and less commonly H63D homozygotes31 may resemble C282Y homozygotes with mild to moderate iron overload. These genotypes are much more common than C282Y homozygotes in the general population yet are not commonly reported in large series of typical haemochromatosis patients. Large population studies have demonstrated that most patients with C282Y/H63D or H63D/H63D have normal iron studies1, 32 (Figure 6).4 Other haemochromatosis HFE mutations have not been clearly established in large studies to explain iron overload in non-C282Y homozygotes. It is likely that more mutations will be found but they will only to relevant to a minority of patients. The interpretation of the test in several settings is shown in Table 3.

Table 3.   Interpretation of genetic testing for haemochromatosis
C282Y homozygote: This is the classical genetic pattern which is seen in >90% of typical cases. Expression of disease ranges from no evidence of iron overload to massive iron overload with organ dysfunction. Siblings have a one in four chance of being affected and should have genetic testing. For children to be affected the other parent must be at least a heterozygote. If iron studies are normal, false positive genetic testing or a non-expressing homozygote should be considered
C282Y/H63D – Compound heterozygote: This patient carries one copy of the major mutation and one copy of the minor mutation. Most patients with this genetic pattern have normal iron studies. A small percentage of compound heterozygotes have been found to have mild to moderate iron overload. Severe iron overload is usually seen in the setting of another concomitant risk factor (alcoholism, viral hepatitis)
C282Y heterozygote: This patient carries one copy of the major mutation. This pattern is seen in about 10% of the Caucasian population and is usually associated with normal iron studies. In rare cases the iron studies are high in the range expected in a homozygote rather than a heterozygote. These cases may carry an unknown haemochromatosis mutation and liver biopsy is helpful to determine the need for venesection therapy
H63D homozygote: This patient carries two copies of the minor mutation. Most patients with this genetic pattern have normal iron studies. A small percentage of these cases have been found to have mild to moderate iron overload. Severe iron overload is usually seen in the setting of another concomitant risk factor (alcoholism, viral hepatitis).
H63D heterozygote: This patient carries one copy of the minor mutation. This pattern is seen in about 20% of the Caucasian population and is usually associated with normal iron studies. This pattern is so common in the general population that the presence of iron overload may be related to another risk factor. Liver biopsy may be required to determine the cause of the iron overload and the need for treatment in these cases
No HFE mutations: There are other iron overload diseases associated with mutations in other iron-related genes (transferrin receptor 2, ferroportin, HJV). Genetic testing is not widely available for these conditions. There will likely be other haemochromatosis mutations discovered in the future. If iron overload is present without any HFE mutations, a careful history for other risk factors must be reviewed and liver biopsy may be useful to determine the cause of the iron overload and the need for treatment. Most of these cases are isolated, non-familial cases

Genetic discrimination has been a major concern with the widespread use of genetic testing. However, several population screening studies have suggested that this is rare.33–37 In the case of haemochromatosis, the advantages of early diagnosis in young adulthood of a treatable disease outweigh the disadvantages of genetic discrimination.

The widespread use of genetic testing for haemochromatosis has also led to misinterpretation of the test results by the patient and physician. For example, a H63D heterozygote that is seen in one in five of the Caucasian population may be interpreted as evidence of haemochromatosis. This can occur because of the complexity of the genetic test report, and is also commonly seen when the patient would prefer to attribute their lifestyle induced liver disease to a genetic problem. In this setting, the patient often attributes every symptom from head to toe to this genetic test result and may be seeking disability benefits. This is clearly an adverse effect of genetic testing. Genetic testing is not recommended for children as organ damage is not typically seen in childhood and early detection may lead to insurance discrimination, labelling or stigmatization.

Most cases of familial iron overload are C282Y homozygotes but there are a growing number of less common genetic mutations that may result in iron overload38, 39 (Table 3) Some of these conditions have been labelled HFE2, HFE3, HFE4 but they do not refer to mutations in the HFE gene so this nomenclature is not recommended. In an isolated case with hepatic iron overload, consideration should first be given to a secondary cause of iron overload. The identification of an iron loaded family member is a powerful clinical clue to the presence of a genetic disorder in iron metabolism (Table 4).

Table 4.   Genetic diseases of iron metabolism
Clinical featureHFE linked haemochromatosisFerroportin diseaseJuvenile haemochromatosisTransferrin receptor 2Aceruloplasminaemia
Gene location6p21.32q321q21 (HJV), 19q13.1 (hepcidin)7q223q25
Inheritance patternAutosomal recessiveAutosomal dominantAutosomal recessiveAutosomal recessiveAutosomal recessive
Organs involvedLiver, endocrine, heartLiver, spleenHeart, endocrine, liverLiver, endocrine, heartRetina, basal ganglia, pancreas, liver
Hepatic iron distributionHepatocytesMacrophagesHepatocytesHepatocytesHepatocytes
Response to phlebotomyExcellentAnaemiaExcellentExcellentAnaemia

Genetic testing for non-HFE iron overload

The genetic tests for mutations in ferrportin, hepcidin, haemojuvelin, transferrin receptor 2, are not likely to become widely available commercially because of the low prevalence of these mutations. As more polymorphisms are described, the genotypic–phenotypic correlations have been difficult to establish and it is important to establish the prevalence of these new polymorphisms in larger populations of affected and unaffected cases. Newer advances in microarrays and gene chips may allow for the simultaneous identification of multiple iron genes.40

Non-expressing homozygotes

As genetic testing becomes more widespread there are an increasing number of subjects who have been found with the haemochromatosis gene without iron overload. This includes siblings within well-defined haemochromatosis pedigrees.41 The term ‘non-expressing’ homozygote has been used for C282Y homozygotes with a normal ferritin and transferrin saturation, a normal transferrin saturation and an elevated ferritin, and in asymptomatic homozygotes with elevated iron tests. Patients who are homozygous for the C282Y mutation should be considered at risk of developing iron overload, but if there are no abnormalities in transferrin saturation or ferritin in adulthood, it seems more likely that they are a non-expressing homozygote rather than a patient who will develop iron overload later in life. The follow-up of adult C282Y homozygotes with a normal serum ferritin has not demonstrated a marked increase in the serum ferritin.42–45 It seems appropriate at the present time, to repeat the serum ferritin and transferrin saturation every 5 years in non-iron loaded C282Y homozygotes to understand more about their natural history. The study of the non-expressing homozygote may provide additional information about new modifying genes that counteract the effect of the haemochromatosis gene.

Family studies in haemochromatosis

Once the proband case is identified and confirmed with the genetic test for the C282Y mutation, family testing is strongly recommended.46 Siblings have the highest chance of carrying the gene and should be screened with the genetic test (C282Y and H63D mutation), transferrin saturation and serum ferritin. Phenotypic expression can vary widely between siblings suggesting that environmental factors are contributory. Patients are usually very concerned about their children and may have difficulty with the concept of autosomal recessive transmission. The risk to a child is dependent on the prevalence of heterozygotes in the community and is probably greater than one in 20 and much lower if the spouse is non-Caucasian.47 A cost-effective strategy now possible with the genetic test is to test the spouse for the C282Y mutation to assess the risk in the children. If the spouse is not a C282Y heterozygote or homozygote, the children will be obligate heterozygotes. This assumes paternity and excludes another gene or mutation causing haemochromatosis. This strategy is particularly advantageous where the children are geographically separated or may be under a different healthcare system.

If an isolated heterozygote is detected by genetic testing, it is recommended to test siblings. Extended family studies are less revealing than a family study with a homozygote but more likely to uncover a homozygote than random population screening.

Genetic counselling for an autosomal recessive disease like HFE-linked haemochromatosis has previously been within the realm of the gastroenterologist or hepatologist. However as additional mutations are discovered, new polymorphisms are identified that may or may not contribute to iron overload, and multiple genes are tested simultaneously, genetic counselling becomes far more complex and may require additional support from medical geneticists.

Treatment of HFE-linked haemochromatosis

The therapy of haemochromatosis continues to use the mediaeval therapy of periodic bleeding. The goal of therapy is to remove excess iron to prevent any further tissue damage. Phlebotomy therapy has never been subjected to a randomized clinical trial and this has hindered our understanding of the natural history of untreated disease. Although most experts believe that iron depletion can stabilize liver disease, improve cardiac function and dyspnoea, and reduce skin pigmentation, there are still skeptics that have suggested that there is no evidence to support phlebotomy therapy.43 At our centre, patients attend an ambulatory care facility and the venesections are performed by a nurse using a kit containing a 16 gauge straight needle and collection bag (Blood Pack MR6102; Baxter, Deerfield, IL, USA) Blood is removed with the patient in the reclining position over 15–30 min. A haemoglobin count is performed at the time of each venesection. If the haemoglobin decreased to <100 g/L the venesection schedule is modified to 500 mL every other week. The concomitant administration of a salt containing sport beverage (e.g. Gatorade) is a simple method of maintaining plasma volume during the venesection. Maintenance venesections after iron depletion of three to four venesections per year are done in most patients although the rate of iron reaccumulation is highly variable.48, 49 The transferrin saturation will remain elevated in many treated patients and will not normalize unless the patient becomes iron deficient. In some cases, a component of the serum ferritin elevation is related to inflammation rather than iron overload, so the ferritin does not decrease with treatment and the patient becomes anaemic. In these cases, a careful review of the liver biopsy including hepatic iron concentration may be helpful in deciding to discontinue treatment or decrease the frequency of phlebotomies. There are different ways to follow the progress of phlebotomy therapy. At our centre, patients are treated until the serum ferritin is approximately 50 μg/L. This is at the low end of the normal range and allows some room for iron reaccumulation into the normal range. Patient support groups have advocated for more intensive phlebotomy but fatigue begins to intervene as iron deficiency is approached. The transferrin saturation may not decrease until the patient is on the brink of iron deficiency and therefore, we will discontinue phlebotomy therapy in some patients with a low ferritin but an elevated transferrin saturation. Other approaches to monitoring phlebotomy therapy include the continuation of weekly phlebotomy until anaemia develops (haemoglobin < 100 g/L) or monitoring of the MCV during therapy.50 Many patients enjoy the concept of maintenance therapy particularly if they can be voluntary blood donors.51 The evidence supporting the need for maintenance therapy is lacking48 and it may be useful to repeat a serum ferritin in 6 months following cessation of therapy to estimate the risk of iron reaccumulation.

Chelation therapy with desferrioxamine is not recommended for haemochromatosis. The therapy is expensive, inefficient, cumbersome and potentially toxic. Oral chelators such as deferasirox have not been well studied in haemochromatosis. Erythrocytophoresis has been used but is more expensive than simple phlebotomy therapy.

Patients are advised to avoid oral iron therapy and alcohol abuse but there are no proven dietary restrictions. Patient support groups are discouraged by the practice of iron fortification of foods, but much of this iron is in an inexpensive form with poor bioavailability. Iron fortification has been removed from food in Sweden and a decrease in the mean serum ferritin has been demonstrated in the normal population.52 Tea consumption has been shown to decrease intestinal iron absorption. Many patient support groups recommend avoidance of shellfish because of the increased risk of vibrio infections in iron overload.


The clinical profile of a typical haemochromatosis patient has not changed significantly since the scholarly descriptions of Sheldon in 1935. However, the inclusion of control patients in population-based studies has brought into question the attribution of symptoms to haemochromatosis and iron overload. Of all the symptoms, liver disease has the most consistent relationship to haemochromatosis and the prognosis of haemochromatosis is most closely linked to the degree of iron overload.

The discovery of the HFE gene in 1996 was a major advance in the field, and most Caucasian patients with typical haemochromatosis can be diagnosed with a commercially available genetic test. However, a growing number of new iron-related genes have been discovered and linked to other iron overload syndromes.