Chapter Contents
Iron metabolism and homeostasis 275
Hereditary iron overload 279
Hereditary haemochromatosis 279
Ferroportin disease 285
Hereditary hyperferritinaemia 286
Atransferrinaemia and hypotransferrinaemia 287
Acaeruloplasminaemia 287
Miscellaneous iron overload 287
Acquired and secondary iron overload 289
Thalassaemia 289
Sideroblastic anaemias 290
Porphyria cutanea tarda 290
Parenteral iron overload 291
Anaemia of inflammation 291
Iron overload in other chronic liver diseases 292
Role of liver biopsy in iron overload 297
Iron is by mass the most abundant trace element on Earth and an essential micronutrient for most living organisms. Under an oxygen-free atmosphere, the chemistry of early life on our planet was founded on the capacity of water soluble ferrous iron (Fe 2+ ) to exchange electrons with water insoluble ferric iron (Fe 3+ ). In an oxygen-rich environment, however, the same chemistry limits its use and sets the basis for its toxicity. Iron enables fundamental reactions that are essential for life, including oxygen transport and delivery in haemoglobin and myoglobin, DNA synthesis, oxidative phosphorylation and host defence, but in the presence of oxygen, unchelated ‘free iron’ catalyzes the generation of noxious reactive oxygen intermediates that damage macromolecules such as proteins, lipids and nucleic acids. Therefore since humans have not developed active mechanisms to dispose of excess iron, blood and tissue iron levels must be kept within a very narrow range through fine-tuned regulatory systemic and local mechanisms. Moreover, a number of plasma and tissue proteins have evolved to absorb, transport and store highly reactive iron (see next section).
Based on these premises, it is no surprise that both iron deficiency and iron overload may represent serious threats to human health. A variety of abnormal states may arise from the disruption of iron homeostasis, ranging from iron deficiency anaemia to genetic iron overload, as typified by hereditary haemochromatosis. Together, these reflect two of the more common disease states of humans. In recent years, there have been dramatic advances in understanding the handling of iron in the body and its homeostatic regulation. The liver, the main body iron store, is also the main site of iron toxicity during iron excess. However, the recent discovery that the liver is the main source of the iron hormone, hepcidin, has shed new light on its central role in both the regulation of body iron homeostasis and the pathogenesis of numerous human diseases related to iron excess.
Iron metabolism and homeostasis
Total body iron content ranges from 3 to 5 g, mostly in haemoglobin or as storage iron in the liver and spleen, with smaller amounts in myoglobin and various enzymes. The daily requirement for iron in the adult is 20–25 mg, mainly for erythropoiesis, but only 1–2 mg of dietary iron is required daily in a healthy individual to compensate the minimal loss (1–2 mg daily), because most iron is recycled through the reticuloendothelial system by the efficient phagocytosis of effete (weak) red blood cells (RBCs).
Iron absorption and transport
Under physiological conditions, the only regulated step of iron metabolism is iron absorption by the small intestine. Dietary iron is digested in the acidic environment of the stomach and taken up by the apical brush border of the small intestine, principally the duodenum. The ingested ferric iron must be reduced to its ferrous state through the action of a brush border ferrireductase, DcytB. It is then actively transported through the apical membrane of the duodenal enterocyte by the divalent metal transporter DMT1, formerly Nramp2 or DCT1, which is also capable of transporting other divalent metals. Dietary haem, after its dissociation from globin, is transported across the brush border and released within the enterocyte by haem oxygenase or transported to plasma as an intact porphyrin, but the transporter responsible for haem uptake at the apical membrane has not yet been conclusively identified.
The absorbed iron, from either ionic or haem pathways, may be retained within the enterocyte as ferritin and eventually lost through desquamation, or transported actively through the basolateral membrane. The only iron-exporter protein currently identified in mammals is ferroportin, a multipass transmembrane iron channel also known as Ireg1 or MTP1.
The export is of iron facilitated by an accessory protein, hephaestin, a ferroxidase and caeruloplasmin homologue, thus returning the iron to the ferric state and promoting its binding to transferrin and caeruloplasmin, also required for the export of iron from nonintestinal cells. The crucial relationship between ferroportin and the regulatory peptide hormone hepcidin is discussed later.
Transferrin (Tf) is the major iron-binding glycoprotein in plasma. The Tf molecule allows for high-affinity binding of two Fe 3+ atoms with Tf iron-binding sites and is normally about 30% saturated in the physiological state. Given the potentially toxic nature of iron, Tf allows the iron to remain soluble and nonreactive in the aqueous environment and facilitates the cellular importation of iron into cells. The latter occurs through the Tf cycle, whereby the two diferric Tf molecules bind with a high-affinity transferrin receptor (TfR1) that is ubiquitously expressed on most cell surfaces. All nucleated cells have TfR1 on their cell surfaces, most importantly the developing RBCs in bone marrow, syncytiotrophoblasts and both Kupffer cells and hepatocytes, which receive first-pass blood from the portal circulation. The Tf–TfR1 complex is internalized, and after endocytosis the endosome is acidified by proton pumps, and the iron is released from Tf into the cytoplasm by a process that also requires DMT1. The unbound Tf, apotransferrin and TfR1 recycle to the cell membrane to participate again in the transport pathway.
During iron overload states, non-transferrin-bound forms of iron appear in the blood that are not tightly associated with Tf and whose labile fraction (‘labile plasma iron’) has high propensity for oxygen-reduction (redox) activities. Labile plasma iron is rapidly taken up by hepatocytes through unregulated mechanisms, likely involving voltage-dependent calcium ion (Ca 2+ ) channels, zinc ion (Zn 2+ ) transporter ZIP14 and others, and enters the cytosolic transit iron pool (‘labile cell iron’), where it can favour uncontrolled production of reactive oxygen intermediates, particularly the highly reactive hydroxyl (OH − ) radicals through Fenton chemistry.
Once in the cytosol, iron is rapidly directed to the functional utilization sites, mainly in cell organelles, where it catalyzes enzymatic and nonenzymatic reactions essential for life. Unneeded iron is safely stored within the core of a multimeric protein, ferritin or, much less efficiently, in haemosiderin. Ferritin is the most important mechanism by which cells store iron. This large and complex protein comprises 24 similar subunits that sequester up to 4500 iron atoms in a central core, thus isolating the metal from the cellular environment. Ferritin is highly conserved across all organisms, thus indicating a critical function in iron homeostasis. Human ferritin comprises H (active) and L (inactive) subunits whose relative proportions vary according to cell-specific iron and oxygen homeostasis, tissue type, development and, in animal models, iron overload. The H subunit carries a ferroxidase site that allows rapid oxidation of Fe 2+ with the formation of diferroxo-mineral precursors at the expense of generating hydrogen. Although the ferritin system is very efficient in iron cycling and storage, degradation of ferritin leads to the accumulation of haemosiderin that is, by contrast, a potentially pathological nonhomogeneous conglomerate of iron, protein and membrane breakdown products, one that is mobilized poorly if at all.
Under normal physiological conditions, hepatocytes and macrophages store iron to an amount of 0.5–1 g. The hepatocyte is the major storage facility for iron, whereas tissue macrophages are primarily responsible for scavenging and phagocytosing effete erythrocytes.
After the RBC is degraded, iron is released from haem by haem oxygenase and may be stored as ferritin or efficiently mobilized and exported as required to the bone marrow for purposes of erythropoiesis. The export process is incompletely understood, but apparently it again occurs through the action of ferroportin and hephaestin. This in turn is controlled by the peptide hormone hepcidin, as reviewed later.
Regulation of iron homeostasis
As previously noted, iron is a critical nutrient but also a potentially toxic agent, and no physiological pathway for iron excretion exists in humans. Thus there must be meticulous coordination of the uptake, storage and utilization of the metal, both at the cellular level and systemically.
Intracellular iron homeostasis: the iron regulatory proteins
To orchestrate intracellular iron traffic, cells contain proteins that have the ability to register the presence or absence of iron, identified in mammalian cells as two homologous iron regulatory proteins, IRP1 and IRP2. These discern cytosolic iron levels and alter the expression of proteins involved in iron movements, such as TfR1 (but not TfR2), DMT1 and ferroportin, or iron storage, such as ferritin. IRP1 and IRP2 share extensive sequence homology, but significant differences exist between the two. Sensing of iron levels is coupled to availability of oxygen and other oxidants. Many proteins involved in iron metabolism have motifs called iron-responsive elements (IREs), which are stem-loop structures in their 3′ or 5′ untranslated regions. When, for example, intracellular iron is scarce, the IRP binds to the 5′ untranslated region of the ferritin mRNA transcript, and ferritin synthesis is halted, to prevent iron storage, while IRP binding at the 3′ untranslated region of TfR1 stabilizes its mRNA, leading to higher TfR1 protein expression and increased iron uptake. It has been suggested that the molecular basis for the iron-sensing process in mammalian cells relies on stabilization or degradation of an iron- and oxygen-binding protein that target IRPs to the ubiquitin (Ub) proteasome.
Systemic iron homeostasis: hepcidin, the iron hormone
Systemic iron homeostasis is also tightly regulated to avoid iron starvation or excess, both responsible for disease states. The nature of the iron regulatory signals, mediators and targets are now being uncovered. Storage iron in hepatocytes and tissue macrophages must be mobilized in response to need, and the concept of store and erythroid ‘regulators’ was originally conceived by Finch. The store regulator would control duodenal iron uptake by means of a tightly regulated feedback mechanism to prevent iron overload. The erythroid regulator would enhance intestinal absorption in response to erythroid demand when there is an increased requirement for iron but the capacity of storage cells is insufficient. In addition, iron could be regulated at the level of the duodenum by the amount of iron recently consumed, thus invoking a dietary regulator, probably resulting from the buildup of intracellular iron. Even in systemic iron deficiency, a bolus of iron may cause a so-called mucosal block. Furthermore, iron homeostasis is also modified in conditions of hypoxia, by a humoral hypoxia regulator, and cellular iron may be retained and its absorption interrupted by infection, through the inflammatory regulator, presumably to withhold iron from the pathogen.
Such ‘regulators’ were largely theoretical until the discovery of hepcidin (hepatic bactericidal protein), or LEAP1 (liver-expressed antimicrobial peptide 1), the iron regulatory hormone. Hepcidin functions at the point of convergence of all humoral iron regulators, including the erythroid, the store and the inflammatory regulators. Originally identified as a type II acute-phase protein produced by the liver, hepcidin is a 25-amino acid peptide containing four disulphide bonds that forms a hairpin shape stabilized by these bonds. Hepcidin is synthesized as an 84-amino acid prepropeptide containing a typical N-terminal 24-amino acid, endoplasmic reticulum-targeting signal sequence and a consensus furin cleavage site immediately preceding the C-terminal 25-amino acid bioactive peptide. In addition to prohepcidin and hepcidin-25, carboxy-terminal 22- and 20-amino acid forms of hepcidin are found in the circulation and urine, but hepcidin-25 is the bioactive form. Hepcidin is a member of the cysteine-rich, cationic, antimicrobial peptide family and, in fact, has retained some antibacterial and antifungal effects in vitro . Of all known antimicrobial peptides involved in innate immunity, however, it seems hepcidin has evolved the unique capability of fighting pathogens by restraining serum iron that is necessary for their growth and proliferation during infection. This task is accomplished by binding to ferroportin. As a result of its interaction with circulating hepcidin, ferroportin is internalized and degraded, thereby diminishing the cells’ ability to transfer iron to the plasma compartment. Therefore low serum levels of hepcidin leads to increased iron flux from enterocytes and macrophages into the circulatory iron pool, whereas high serum hepcidin leads to low circulatory iron due to inhibition of its intestinal absorption or release from macrophages.
Hepcidin has evolved as part of the innate immune defence. As such, hepcidin is induced by infection and inflammation and plays a central role in the anaemia and hypoferraemia of chronic disease, alternatively called ‘anaemia of inflammation.’ Hepcidin responds to a variety of inflammatory signals and mediators, particularly interleukin-6 (IL-6), IL-1, IL-22, and activin B. However, hepcidin is in essence an acute-phase protein, and it senses, beyond inflammation, a number of intracellular and extracellular stress signals ( Fig. 4.1 A ). Endoplasmic reticulum (ER) stress is primarily associated with disruption of ER homeostasis and accumulation of unfolded or misfolded proteins in the ER. ER stress has been also involved in a number of pathophysiological states, including inflammatory response, nutrient disorders and viral infection. Exogenous and endogenous ER stressors can trigger hepcidin transcription through the cyclic adenosine monophosphate (cAMP) response element binding protein 3–like 3 (CREB3L3, also known as CREBH) and leads to perturbation of iron homeostasis in vivo . Also, metabolic stress can turn on hepcidin transcription. It has been recently reported that PPARGC1A, a transcriptional coactivator, cooperates with CREBH to activate hepcidin and regulate iron traffic in vivo during food deprivation. The latter finding has important implications during human disorders associated with insulin resistance and induced gluconeogenesis, such as obesity, diabetes and nonalcoholic fatty liver disease.
Hepcidin induction in the presence of stress factors that perturb the internal homeostasis—our legacy to the innate immune response to pathogens—aims at preserving and retaining in the body (or within the cells) the precious iron needed for vital energy production.
During iron deficiency and anaemia, hepatic hepcidin transcription must be turned down so that more iron can be transferred from the intestine and storage sites to serum transferrin and to the bone marrow. Hypoxia and erythropoietin inhibit hepcidin synthesis and increase iron absorption. Circulating factors derived from maturing erythroblasts in the bone marrow have been reported to downregulate hepcidin transcription in the liver, including growth differentiation factor 15 (GDF15), a member of the transforming growth factor beta (TGF-β) superfamily; twisted gastrulation protein, a BMP-binding protein, and more recently erythroferrone, which mediates hepcidin suppression during stress erythropoiesis.
The main humoral signal for hepcidin synthesis is iron itself ( Fig. 4.1 B ). The iron-sensing system resides within the liver, and its disruption is usually responsible for human haemochromatosis. Iron-sensing involves transferrin-iron (i.e. the extent of Tf saturation) and a class of ligands of the TGF-β superfamily, the bone morphogenetic proteins (BMPs), which normally play a crucial role in embryonic development and in fundamental processes during postnatal life. Tf-iron takes part of a multiprotein complex at the hepatocyte plasma membrane made by BMPs, BMP receptors, a BMP co-receptor (haemojuvelin, HJV) and a number of ancillary proteins (including HFE and the second transferrin receptor, TfR2) ( Fig. 4.1 ). This interaction triggers the phosphorylation of SMAD1,5,8 complex (receptor-associated SMADs, R-SMADs), its subsequent binding to SMAD4 (common-partner SMAD, co-SMAD), the translocation of the SMAD complex to the nucleus and activation of hepcidin transcription. The BMP co-receptor HJV, which is present in either a soluble or a cell-associated form, provides specificity to the iron signal in the liver and functions as an enhancer for iron signalling to hepcidin. At least two other proteins are required for normal signalling of iron status to hepcidin via the BMP6/SMAD1,5,8 pathway, HFE and TfR2. In fact, functional loss of HFE in mice and humans leads to low hepcidin and haemochromatosis. Functional loss of TfR2 in mice and humans is also associated with blunted hepcidin expression and iron overload. The details of HFE and TfR2 function in the context of the BMP/SMAD signalling pathway are still not completely understood. HFE is a major histocompatibility complex (MHC) class I-like protein that interacts with TfR1, the receptor for serum Tf that mediates uptake of Tf-bound iron. The C282Y HFE mutation associated with human haemochromatosis disrupts a disulphide bond required for HFE binding to β2-microglobulin and transport to the cell surface and endosomal membranes, where it interacts with TfR1. The H63D mutation, a common HFE polymorphism, does not impair HFE-TfR1 interaction. HFE is not required for transcriptional regulation of BMP6 in response to dietary iron, but loss of HFE reduces BMP6 signalling, in vitro and in vivo . HFE interacts with the BMP type I receptor ALK3 to stabilize it and increase its expression, thereby inducing hepcidin expression. TfR2 mediates the uptake of Tf-bound iron by hepatocytes in vitro, but its in vitro affinity for Tf is much lower than that of TfR1. It has been postulated that also TfR2 interacts with HFE, forming a unique iron-sensing complex with TfR1-HFE that modulates hepcidin expression in response to Tf-iron. However, human studies in patients with combined TfR2 and HFE mutations and studies in HFE/TfR2 double-knockout mice have shown that the contemporary loss of HFE and TfR2 has additive phenotypic effects. This suggests that HFE and TfR2 regulate hepcidin and iron metabolism in an independent manner. Both proteins seem to be important for BMP signalling and necessary for an optimal response to BMPs.
The BMP-SMAD pathway is likely counter-regulated by a number of feedback regulatory signals that prevent overshooting of hepcidin expression in response to positive stimuli. A key role is likely played by a membrane serine protease 2 (also called TMPRSS6) that functions by inhibiting the BMP pathway, possibly by cleaving HJV, although this aspect is still debated. Loss of TMPRSS6 in humans causes a syndrome known as ‘iron deficiency, iron-refractory anaemia’ (IDIRA) characterized by hyperhepcidinaemia and microcytic iron deficiency anaemia resistant to oral iron supplementation. A second BMP-pathway inhibitor is SMAD7, induced by chronic dietary iron loading, and BMP-SMAD signalling pathway activity in the liver. Another player in the BMP/SMAD pathway is neogenin, a member of the DCC (deleted in colorectal cancer) family of tumour suppressor molecules. Neogenin appears to interact with HJV, but its role in HJV biology and hepcidin synthesis is still controversial. Interestingly, the livers of neogenin-mutant mice exhibit reduced BMP signalling, low levels of hepcidin and iron overload. BMPs and the BMP/SMAD pathway are therefore central in hepcidin transcription in the liver. BMP6 in particular seems to play a key role in this process. BMP6 is largely produced by hepatic sinusoidal cells and nonparenchymal cells (see Fig. 4.1 B ) and is directly induced by excess iron. In fact, blocking BMP6 in vivo inhibits hepcidin expression and increases serum iron, whereas genetic BMP6 ablation in mice leads to low hepcidin expression and haemochromatosis, indicating that BMP6 is an endogenous regulator of hepcidin expression and iron metabolism in vivo.
In summary, iron homeostasis requires specific transportation across membranes and meticulous intracellular storage. Diminution of iron stores caused by dietary deficiency, iron loss or infection gives rise to eventual iron restriction and progression to anaemia. The converse may arise, however, in which the large ionic iron molecule, in excess and in solution with oxygen, may generate free-radical formation through Fenton and Haber–Weiss chemistry, with hydrogen peroxide (H 2 O 2 ) being changed into its radical (HO − ). This leads to consequent damage to DNA, proteins and membranes.
The classic example of iron overload in human pathology is HFE -hereditary haemochromatosis, but numerous other entities are now identified to cause pathological iron overload. Some are well described and have a verified hereditary basis, whereas in others the genetic and hereditary basis is still speculative. Still others reflect iron overload on an inflammatory or infectious basis or as a reactive response to systemic or as-yet unknown disease processes.
Controversy surrounds nomenclature and classification of these iron overload states. Haemochromatosis (HC), as identified in the liver, has previously been related to iron loading in primary liver cells, hepatocytes and biliary epithelium, whereas haemosiderosis, iron staining in tissues, was employed by some as the terminology of choice when the overload was predominantly within Kupffer cells. Many overload states, however, may show elements of both HC and haemosiderosis. A further caveat is that current classifications, including the Online Mendelian Inheritance in Man (OMIM) database, are founded on the basis of single-gene defects and have significant shortcomings in that numbers of atypical cases have been identified that relate to multiple gene mutations. It appears now that failure to produce adequate amounts of the iron hormone hepcidin or its impaired activity, mostly from genetic changes but also from acquired factors, results in a syndromic entity that resembles the historical definition of HC. Thus hereditary HC can be defined and classified phenotypically, regardless of the underlying gene defect, as an inherited disorder resulting from an inborn error of iron metabolism causing hepcidin deficiency that leads to progressive loading of parenchymal cells of the liver, pancreas, heart and endocrine organs and leads to damage and disease states.
Hereditary iron overload
Hereditary haemochromatosis
Historical perspective
Trousseau first described a diabetic patient with an ‘almost bronzed’ appearance in 1865. In 1871, Emile Troisier detailed the first autopsied case of a diabetic patient with cirrhosis and a red-brown liver containing clumps of pigment. Von Recklinghausen later termed the condition ‘haemochromatosis’. Two main theories emerged from subsequent reports: the primary disease was diabetes that then gave rise to cirrhosis and pigmentation, or alternatively, the pigment was primarily derived from the blood. After years of dispute as to its pathogenesis, Sheldon consolidated the existing knowledge of HC with its classic triad of diabetes, cirrhosis and melanin-based pigmentation. He described the entity as ‘an inborn error of metabolism, which has an overwhelming incidence in males and which at times has a familial incidence.’
The autosomal recessive inheritance pattern of classic hereditary HC was thereafter established, but it was only in 1976 that Simon et al. confirmed an association between the disorder and the human leukocyte antigens (HLAs) A3 and B14. The subsequent discovery that β2-microglobulin knockout mice developed iron overload analogous to human HC raised the postulate that the defective gene would be within an MHC molecule. Positional cloning experiments using linkage disequilibrium and haplotype analysis allowed the identification of the affected gene, an atypical class I HLA molecule originally designated HLA-H and subsequently redefined as HFE . Subsequent knockout of the mouse Hfe gene confirmed iron overload, thus confirming HFE to be the defective gene in classic hereditary HC.
As genetic testing for HFE mutations became more widespread, it rapidly became clear that the situation was more complicated than initially thought. Other iron genes were discovered whose mutations were associated with hereditary iron overload syndromes with some, or many, or apparently even all, of the phenotypic features of classic HC ( Table 4.1 ): ferroportin-associated iron overload ( FPN ), transferrin receptor 2-associated HC ( TfR2 ), hepcidin-associated ( HAMP ) and haemojuvelin-associated ( HJV ) ‘juvenile’ HC.
Hereditary (gene locus) | Acquired | Miscellaneous |
---|---|---|
Hereditary haemochromatosis | Transfusion-dependent iron overload (thalassaemia major, sideroblastic anaemia, etc.) | Neonatal haemochromatosis a |
Adult onset | Enteral iron overload | African American iron overload b |
HFE (6p21.3) | Parenteral iron overload | |
TfR2 associated (7q22) | Porphyria cutanea tarda | |
FPN associated (2q32) | Anaemia of inflammation | |
Juvenile-onset | Chronic liver diseases | |
HJV associated (1q21) | Hepatitis B and C viruses c | |
HAMP associated (19q13) | Alcoholic siderosis c | |
TfR2/HFE associated | Insulin resistance (NAFLD/NASH) | |
HFE/HJV-HAMP associated | End-stage liver disease c | |
Ferroportin disease (2q32) | Portocaval shunt | |
Atransferrinaemia (3q21) | ||
Acaeruloplasminaemia (3q23–24) | ||
DMT1 deficiency (12q13) | ||
Hereditary iron-loading anaemias with inefficient-erythropoiesis |
a Unknown, but variable inheritance from recessive to non-Mendelian. Alloimmunity queried.
b Possible environmental and genetic (ferroportin polymorphism) influences.
c Direct inhibition of hepcidin transcription in the liver is also postulated as a cofactor for hepatic iron overload.
Sheldon almost certainly described patients having the characteristics of ‘juvenile’ HC. However, the entity was only formally recognized after the case study and literature review of Lamon et al. They described a 36-year-old woman with heart failure, diabetes mellitus, hepatomegaly and secondary amenorrhoea and reviewed another 52 cases published over the previous 85 years. Essential differences between these cases and classic hereditary HC are notable compared with the large series of Finch and Finch, namely, that it is a rare form of iron overload with rapid and severe progression of disease leading to significant complications before 30 years of age. Lamon et al. also noted the inherited nature of this rare disease and favoured an autosomal recessive inheritance. Subsequent genotyping identified the disorder to be distinct from HFE-linked iron overload. A high frequency of consanguinity is observed in patients with juvenile HC, with the disease occurring in siblings but not parents. A human-genome search demonstrated linkage between the disease and numerous markers on the long arm of chromosome 1q. Subsequent analysis of Greek, Canadian and French families carrying the disorder identified multiple deleterious mutations at LOC148738, whose protein product is identified as haemojuvelin (HJV), previously HFE2. In addition, the investigation of juvenile HC identified a second cohort that is even rarer, demonstrating a mutation, C70R, on 19q13, the gene encoding hepcidin. The concept and spectrum of juvenile HC has been further extended by the recent identification of two young patients presenting with typical features of the disease, severe endocrinopathy and cardiomyopathy, but testing positively for combined mutations for C282Y/H63D and TfR2 . In addition, there is evidence that selected HJV and HAMP mutations, when carried simultaneously with mutant HFE genes, may influence iron status in older patients being evaluated for a molecular diagnosis of HC because of increased transferrin saturation and/or serum ferritin. This further challenges the current classification system for hereditary HC in that ‘juvenile HC’ does not represent a distinct monogenic disorder but may be linked to the adult-onset form of hereditary HC.
Definition
The present definition of hereditary HC embraces the classic disorder related to HFE C282Y homozygosity (the prototype for this syndrome and by far the most common form) and the rare disorders more recently attributed to loss of TfR2, HAMP and HJV or, in very rare cases, FPN ( Table 4.1 ). This concept of HC stems from the idea that, beyond their genetic diversities, all known haemochromatoses belong to the same clinicopathological entity: they all originate from the failure to prevent unneeded iron from entering the circulatory pool as a result of genetic changes compromising the synthesis or activity of hepcidin, the iron hormone ( Fig. 4.2 ). Depending on the gene involved and its role in hepcidin regulation, the phenotype of HC varies, ranging from the rare HJV and HAMP juvenile forms characterized by massive iron loading with severe health and endocrine disorders, to the adult-onset phenotype dominated by liver disease and usually associated with HFE, TfR2 and rarely FPN mutations. In brief, the features that distinguish hereditary HC from all other iron-loading disorders listed in Table 4.1 are hereditary nature (usually autosomal recessive); early and progressive expansion of the plasma iron compartment (increasing transferrin saturation); progressive parenchymal iron deposits with potential for severe damage and disease that may involve liver, endocrine glands, heart and joints; nonimpaired erythropoiesis and optimal response to therapeutic phlebotomy; and inadequate hepcidin synthesis/activity.
Epidemiology
An autosomal recessive disease, classical HFE -associated hereditary HC is the most common monoallelic inherited disorder in Western society. The dominant missense polymorphism identified in 80% of cases, C282Y, is characterized by a single nucleotide change, G to A, resulting in a Cys→Tyr at position 282 of the unprocessed protein, occurring in a highly conserved region involved in the disulphide bridge in the MHC class I protein. A less common polymorphism is a C-to-G change resulting in a Hys→Asp substitution at amino acid 63, showing limited clinical effects, although compound heterozygosity for C282Y and H63D may cause disease expression. Of several other uncommon polymorphisms, S65C, in which cysteine replaces serine at position 65, has been associated with compound heterozygosity for either C282Y or H63D. It has been related to the development of mild to moderate hepatic iron overload but without clinical manifestations. Its significance is still controversial.
The C282Y polymorphism is most prevalent among populations of northern European origin, with more than 90% of patients with clinically penetrant HC in the United Kingdom being homozygous for C282Y. Worldwide allele frequencies are calculated at 1.9% for C282Y and 8.1% for H63D, with highest frequencies of 10% for C282Y in Irish chromosomes and 30.4% for H63D in Basque chromosomes. The origin of the genetic mutation was thought to represent a unique event in chromosome HLA-A3 and -B7 originating from the Celts in central Europe between 65 and 70 generations ago and spreading north and west by population migration. However, an alternative Viking origin has been postulated, and it is recently proposed that the initial C282Y polymorphism occurred in mainland Europe before 4000 bce , earlier than both the Celtic and the Viking period.
The genetic defect, which caused no serious obstacle to reproduction and may even have conferred some advantages, was passed on and spread through population migration. The distribution of the C282Y mutation coincides with its northern origin, with frequencies ranging from 12.5% in Ireland to 0% in southern Europe. In addition to C282Y, H63D, the ‘minor’ HFE polymorphism, has been found more frequently in HC patients than in the control population. The frequency of the H63D polymorphism shows less geographic variations with an average allelic frequency of 14 %, but its clinical impact appears to be limited. An additional HFE polymorphism is S65C, which can be associated with HC when inherited in trans with C282Y on the other parental allele. The prevalence of C282Y homozygosity among patients with liver disease is about 5%, 10-fold higher than the reported prevalence in the general population. This figure increases if patients with liver disease are preselected for increased transferrin saturation. Even higher C282Y frequencies can be found in patients with hepatocellular carcinoma, a known complication of HC.
Haemochromatosis is associated with homozygosity for the C282Y HFE mutation in approximately 80% of clinically characterized patients of European ancestry. Therefore almost 20% of such patients have the disease in absence of C282Y. Although compound heterozygosity (H63D/C282Y) appears to be disease associated, usually cofactors are implicated in the clinical expressivity.
Other genes associated with clinical HC are TfR2 , HJV , HAMP and FPN (see Table 4.1 ). None of these non- HFE HC appears to be restricted to Northern Europeans. Although the TfR2 mutation is rare, at least nine individual TfR2 mutations have been identified. With the exception of one patient of Portuguese descent with a c2069 A→C, Q690P mutation in the TfR2 gene mutation and two affected homozygous female siblings, most represent largely inbred families of Italian extraction with various mutations of the TfR2 gene and consanguinity a common feature. However, the AVAQ 594–597 deletion, originally described by Girelli et al. in their Italian cohort, has also been identified in three members of a Japanese family. Subsequent investigation of nine other, unrelated Japanese patients with HC of unknown origin described two more novel TfR2 mutations, L490R and V561X. The patient with the V561X mutation, age 58, was a member of a consanguineous family. Both these patients had cirrhosis and diabetes, one with associated skin pigmentation. HC is rare in the Far East, particularly among the Japanese. It would thus appear that TfR2 -hereditary HC is the most common form of hereditary iron overload among Japanese patients, although iron loading secondary to acaeruloplasminaemia is also identified in this population. However, in 2001, a Y231del mutation in the HFE gene was found in the Huh-7 hepatoma cell line (obtained from a Japanese donor) and was shown to prevent the translocation of HFE to the cell surface, similar to the C282Y mutation. More recently, this same HFE mutation has been reported in a Japanese pedigree affected by HC, indicating the occurrence of a HFE -related form of haemochromatosis also in Asian populations.
The juvenile form of HC is rare. Most cases are caused by mutations of HJV . One common HJV mutation, G320V, has been identified in all reported studies. It is present in half of juvenile HC families. Numerous other mutations have subsequently been identified. More recent analysis of patients from the central and northern parts of Europe, specifically Germany, Slovakia, Croatia and Ireland, has also identified the G320V mutation in the majority. A small proportion of patients with the juvenile form of HC carry mutations in the gene encoding hepcidin.
Although most FPN mutations give rise to a distinct form of hereditary iron overload called the ‘ferroportin disease’, unusual FPN mutations cause rare forms of HC similar to HFE -HC.
Pathogenesis
The first biochemical manifestation of HC is an increase of transferrin saturation that reflects an uncontrolled influx of iron into the bloodstream from enterocytes and macrophages. Except for menstruation, the body has no effective means of significantly reducing plasma iron levels. Without therapeutic intervention, overload in the plasma compartment will lead to the progressive accumulation of iron in the parenchymal cells of key organs, creating a distinct risk for oxidative damage. This stage is reflected in increasing serum ferritin levels. The time of onset and pattern of organ involvement in HC vary depending on the rate and magnitude of plasma iron overloading, which in turn depend on the underlying genetic mutation. For this reason, milder adult-onset forms (e.g. HFE and TfR2 related) and more severe juvenile-onset forms (e.g. HJV and HAMP related) are recognized. The extent of iron release from enterocytes and macrophages into the bloodstream in humans is under the control of the hepcidin–ferroportin axis (see Fig. 4.1 ). It now seems that HFE and TfR2 play a role in conveying the iron signal to hepcidin in the hepatocyte, although the details of this process are not fully uncovered.
In view of these findings, HC should be seen as a genetically heterogeneous disease that results from the complex interaction between genetic and acquired factors. If the altered gene plays a dominant role in hepcidin synthesis (e.g., HAMP itself or HJV ), circulatory iron overload occurs rapidly and reaches high levels (see Fig. 4.2 ). In these cases the modifying effects of acquired environmental and lifestyle factors will be negligible, and the clinical presentation will invariably be dramatic, with early onset (first to second decade) of a full-blown organ disease. In contrast, C282Y- HFE homozygosity results in a genetic predisposition that requires the concurrence of host-related or environmental factors to produce disease (see Fig. 4.2 ). Co-inherited mutations in other HC genes, such as HAMP and HJV , may have a role in disease penetrance of HFE -HC, but they are rare.
Clinical aspects
HFE -associated haemochromatosis
The clinical presentation in patients who are homozygous for C282Y is highly variable, with general signs of weakness (60% of patients), arthralgia/arthritis (30–40%), hepatomegaly/cirrhosis (13–60%), diabetes mellitus (10–30%), sexual dysfunction (10–40%) and cardiac symptoms with arrhythmia (20–29%) and cardiac failure (15–35%). Skin pigmentation is estimated to occur in 47% of proband cases. However, extrahepatic disease tends to be less prominent in modern studies, including diabetes mellitus, arthritis and heart disease. The initial presentation is often vague and nonspecific in many patients, and a high index of suspicion is required to diagnose the condition.
Furthermore, whereas patients with defined HFE -HC will test positively for the C282Y mutation, a significant proportion do not develop significant hepatic iron overload. Gene penetrance is not obligatory and recent large population-based studies confirm that penetrance is low in HFE -hereditary HC, implying that C282Y homozygosity is necessary but not sufficient on its own to cause clinically manifesting disease. Ethnicity, environmental factors such as blood loss (important modifier in females), dietary intake, alcohol and other, as-yet unknown genetic modulators may play a part in this differential expression. It is likely that these disease frequencies will change further with increased awareness of the disease and the development of screening techniques.
In proband patients in whom the diagnosis is suspected, serum testing for iron overload forms the cornerstone of initial identification: transferrin saturation, unbound iron-binding capacity (UIBC) and serum ferritin. Consensus is that Tf saturation >45%, UIBC <28 µmol/L and ferritin >300 µg/L (>200 in women) identifies hereditary HC, although these values may not be appropriate for all populations. Ferritin in isolation is highly sensitive and its normality rules out iron overload, but it is an acute-phase protein that is also elevated in inflammatory states.
Before the discovery of HFE gene mutations, liver biopsy confirmed diagnosis, identifying the typical histological features and allowing for grading and determination of the hepatic iron index (ratio of hepatic iron concentration divided by age), as discussed later. However, subsequent to the development of genotyping, the necessity for liver biopsy now depends on clinical status. If inflammation and other confounding factors are excluded, serum ferritin correlates well with the level of iron excess. Magnetic-susceptibility measurement or magnetic resonance imaging (MRI) may act as a second estimation of iron overload. In C282Y HFE homozygotes, if iron load is moderate and other diagnostic parameters are normal (e.g. serum transaminases, fasting blood glucose, hormonal evaluation, cardiogram, joint/bone x-ray films), therapy by venesection (phlebotomy) may commence without liver biopsy. In cases of heavy iron overload, when one or more of these parameters are abnormal, liver biopsy is mandatory, in particular to evaluate the presence of cirrhosis, other hepatic pathology or iron-free foci, and at this stage the biopsy is performed as a prognostic marker. The current gold standard in the case of hereditary HC is thus genetic testing. The relevance of the H63D mutation as well as C282Y/H63D compound heterozygosity is still uncertain because a small percentage may develop clinically significant HC, normally in the presence of damaging cofactors.
The macroscopic and microscopic features of penetrant HFE -HC are highly characteristic in the liver. The classic description is one of a brown to rusty colour, staining intense blue with Perls ( Fig. 4.3 ), with or without the gross appearance of cirrhosis. Histologically, the progression of the disease is again characteristic. Iron deposition begins in the periportal hepatocytes (zone 1) ( Fig. 4.4 ) and extends progressively to involve all zones of the liver ( Fig. 4.5 ). The iron has a characteristic pericanalicular pattern when observed under high power ( Fig. 4.6 ). With progression of the disease, there can be deposition of iron in the biliary epithelium ( Fig. 4.7 ) and transfer into Kupffer cells and portal macrophages with sideronecrosis ( Fig. 4.8 ). Progressive portal-based fibrosis may evolve to cirrhosis ( Fig. 4.9 ). Of importance is the identification of iron-free foci in histologically advanced liver disease, which may foretell the development of hepatocellular carcinoma, as discussed later.
Venesection forms the cornerstone of successful therapy for hereditary HC. Ideally, this entails weekly phlebotomy (400–500 mL) until ferritin <20–50 µg/L and Tf saturation <30%, after which maintenance venesections are done to keep serum ferritin normal. If biopsy has been performed, the hepatic iron index can be calculated as well as a ‘body iron ratio’, which estimates total iron removed based on phlebotomies and related to age. In addition, the biopsy may be graded for degree of iron overload. Abstinence from iron-rich foods, supplements and vitamins is recommended. Chelation therapy is not warranted unless there are contraindications to venesection.
Fully penetrant hereditary HC carries with it significant potential for morbidity and mortality, including cirrhosis and hepatocellular carcinoma, diabetes mellitus, cardiac failure and impotence. The prevalence of alcohol abuse and chronic hepatitis C virus infections appears increased in patients with HC (see later), but sepsis secondary to bacterial infection does not appear to be a significant problem, with the exception of cases of Yersinia infections. Hepatocellular carcinoma is identified as a significant cause of mortality in HFE -hereditary HC, although its prevalence is not as high as was previously thought. The risk of developing other extrahepatic cancers is still debated. This anticipates the potential role of iron as a putative carcinogen.
However, quantification of these factors is confounded by the prevalence of the C282Y mutation when measured against the penetrance of the gene. Furthermore, many common entities classically identified in patients with hereditary HC, including diabetes and arthritis, also frequently occur in the general population. Earlier series demonstrated significant morbidity and mortality in both untreated and treated patients. Bomford and Williams identified poor survival amongst untreated patients of 18% and 6% at 5 and 10 years, respectively. Similarly, for treated cases, Niederau et al. identified a cumulative survival of 93% at 5 years and 77% at 10 years, significantly reduced compared with the expected survival rates for a normal population. However, more modern series identify that, although most C282Y homozygotes have increased levels of Tf saturation and serum ferritin, symptoms associated with iron overload are no greater in C282Y homozygotes compared with gender- and age-matched controls, and life expectancy is not significantly curtailed.
Ernest Beutler was the first to draw attention to the low clinical expressivity of the C282Y HFE mutation. Subsequent studies in which patients were followed for more than 20 years have then shown that disease progresses in only a minority of untreated C282Y homozygotes. As many as 38–50% of patients homozygous for HFE C282Y will develop iron overload, and 10–33% will eventually develop HC-associated morbidity. Penetrance is usually higher among male patients homozygous for HFE C282Y than female patients, probably because of menstruation and pregnancies. Apart from male gender, and obvious causes of blood loss, alcohol abuse is likely an important ‘modifier’ associated with HC-related cirrhosis. Combinations of mutations in HAMP , HJV and TfR2 have been associated with more severe phenotype in rare patients. Polymorphic variants of BMP2 are also reportedly associated with higher penetrance of HFE -HC. In recent years, novel loci affecting iron homeostasis in individuals at risk for HC have been reported, with some including known iron-related genes. Single-nucleotide polymorphisms (SNPs) at ARNT1, transferrin and TfR2 affect iron markers in HFE C282Y homozygotes at risk for HC. A genome-wide association study (GWAS) using DNA collected from 474 unrelated C282Y homozygotes indicated a link between the rs3811647 polymorphism in the transferrin gene and the phenotypic presentation of HFE-HH. More recently, an exome sequencing study reported that a p.D519G variant in the glyceronephosphate O -acyltransferase ( GNPAT ) gene showed the most significant association with severe iron overload, but the data have been challenged by others.
TfR2 -associated haemochromatosis
The clinical appearance of patients with TfR2 -associated HC mimics that of HFE -hereditary HC: high Tf saturation and serum ferritin levels and low penetrance in premenopausal women. The age range is somewhat younger, but with slow progression of iron overload, thus differing it from juvenile HC.
Furthermore, analysis of a family of southern Italian lineage with features of juvenile HC has recently demonstrated combined mutations for C282Y/H63D (compound heterozygosity) and TfR2 (Q317X) homozygosity while failing to show mutations for either HJV or hepcidin. Two siblings, male and female, diagnosed with severe endocrinopathy and cardiomyopathy at age 24 and 25 years, respectively, both showed cirrhosis with massive loading of iron in hepatocytes. A younger brother age 21 years, however, demonstrated a milder phenotype resembling the iron distribution of classic adult-onset hereditary HC but carrying only the Q317X serum TfR2 homozygote mutation.
Although relatively few TfR2 -HC cases have been documented, the liver pathology is again described as strongly resembling HFE -HC, with early iron deposition in periportal hepatocytes ( Fig. 4.10 ). Most demonstrate a milder degree of iron overload than those with HFE -hereditary HC, although there is progression to cirrhosis in some published cases.
Treatment, as for classic hereditary HC, is by venesection, although there is often a persistence of Tf saturation after phlebotomy. The diagnosis is made by clinical presentation, serum iron indices and elimination of the HFE genotype. Although TfR2- hereditary HC is even rarer than juvenile HC, both stand to be of considerable importance in the understanding of the molecular pathogenesis of iron regulation.
HJV- and HAMP- associated haemochromatosis
HJV -associated HC and, more rarely, HAMP -associated HC account for almost all cases of ‘juvenile HC’. Juvenile HC otherwise differs considerably from HFE -hereditary HC with respect to age, an almost equal ratio between genders, greater frequency of cardiac and endocrine disturbances and lower frequency of diabetes and hepatic involvement. The patient usually presents in the second decade, typically with hypogonadism that manifests as primary infertility in the female. A dilated cardiomyopathy that often becomes refractory to treatment is a common complication, and the untreated patient usually dies of cardiac disease by the 30th year. The hepatic complications of iron overload in juvenile HC are not as common as in the case of hereditary HC, despite that iron indices are usually much higher in these younger patients. The hepatic pathology may be profound, however, with histologically diagnosed cirrhosis developing even at a young age in up to 40% of patients. However, the clinical diagnosis of juvenile HC is often coincidental, relating to investigation of endocrine or cardiac abnormalities, including cardiac shock. Glucose intolerance is manifest in almost two-thirds of patients, and the presentation may involve arthropathy or skin changes. Of note, iron indices are often considerably higher than in patients with hereditary HC.
The histological features of juvenile HC are similar to those of hereditary HC in that there is progressive iron loading of parenchymal cells with, at least initially, sparing of the reticuloendothelial system ( Fig. 4.11 ). Untreated patients, even if clinical picture is dominated by hypogonadotropic hypogonadism and cardiomyopathy, often present with advanced liver fibrosis.
As with hereditary HC, aggressive venesection remains the cornerstone of therapy for HJV – or HAMP -associated HC. Depending on the extent of progression of the disease, there may be a place for chelating therapy, with successful cardiac transplants performed in some patients.
FPN- associated haemochromatosis
Most mutations in FPN that result in an iron overload syndrome are associated with ferroportin disease, which is pathogenically distinct from HC (see next section). However, distinct missense heterozygote mutations of ferroportin can result in a haemochromatosis syndrome—the only known form of ‘classic’ HC with an autosomal dominant trait—caused by resistance of ferroportin to hepcidin inhibition. Even though hepcidin is produced at normal or even higher levels, the mutations in FPN cause hyperabsorption of iron from the diet and hepatocellular iron overload, as in classic HC. In FPN -HC the mutant FPN localizes to the membrane but is resistant to hepcidin. The mutations likely affect domains required for internalization and degradation; the lack of ferroportin regulation by hepcidin thus mimics hepcidin deficiency and causes a phenotype similar to classic HC.
Patients with FPN -HC present clinical manifestations identical to HFE -HC (or TfR2 -HC) with high Tf saturation and serum ferritin levels, predominant hepatic parenchymal iron overload and cirrhosis and organ failure in advanced cases. In the original report by Njajou et al., iron accumulation in patients carrying the NI44H FPN mutation was in hepatocytes, as in classic HC. In a family of European heritage described by Sham et al., with a unique mutation, 977 GYC (Cys326Ser), there was hepatocyte rather than Kupffer cell accumulation. Also, in patients first described as having an autosomal dominant form of HC, from the Solomon islands, the histology resembles the gradient pattern of hepatocyte described for classic hereditary HC. As expected, and similar to the other forms of HC, phlebotomy appears to be well tolerated and effective for FPN -HC patients.
Ferroportin disease
After the identification of the HFE gene, it rapidly became apparent that significant numbers of iron overload disorders could not be explained by HFE mutations, particularly in Europe, where C282Y homozygosity was responsible for 90% of cases in the United Kingdom and Brittany but only 64% and 30% in the southern European countries of Italy and Greece, respectively. Occasional cases could be explained by mutation of TfR2 or HJV / HAMP, but a more common disease was related to abnormalities of the iron-exporter ferroportin.
Previous studies had identified an autosomal dominant iron overload condition in a large family from Italy. In the wake of the discovery of ferroportin, genome-wide screening procedures confirmed that these same patients were affected by a candidate gene on 2q32, SLC40A1, previously named SLC11A3. All were heterozygous for a c.230 C→A substitution resulting in the replacement of alanine 77 with aspartate. This was subsequently known as ferroportin disease.
Ferroportin encodes a transmembrane transporter with iron-responsive element (IRE) function that acts as an iron exporter. The pathogenesis is thus quite different from that of hereditary HC. Ferroportin is directly involved in the release of iron from macrophages, and mutations of SLC40A1 represent the mechanism leading to the ferroportin disease. Various mutations of the gene have been identified with similar outcomes, leading to the conclusion that the basic mechanism is a net loss of protein function.
All mutations are of missense type and, depending on the mutation, the mutant ferroportin can affect the cellular location of the wild-type protein. Loss of iron export function is likely caused by mislocalization of the mutant protein, with the protein localizing in an intracellular distribution, as opposed to the normal membranous display. In this situation, there is no binding to hepcidin or hepcidin-induced degradation. The resultant reduction in iron efflux due to haploinsufficiency does not appear to limit iron transfer in cells exposed to low iron traffic (e.g. enterocytes). However, it causes a bottleneck in macrophages, which generate the largest iron flow, resulting in iron accumulation in Kupffer cells and macrophages with high ferritin levels and low to normal Tf saturation.
Numerous mutations of the ferroportin gene have been identified, with divergent findings with respect to the pattern of ferritin/transferrin dissociation in probands of French-Canadian, Melanesian, Thai and European heritage, which explains the in vitro findings previously discussed. Clinical presentation thus appears heterogeneous. Ferroportin disease, as originally described, may manifest with a milder phenotype than classic HC. The associated liver disease is usually not as severe. Depending on the mutation, serum ferritin increases early in the disease despite low to normal Tf saturation, the opposite picture to classic HC. Hypochromic anaemia is common and may require iron supplementation, which may further exacerbate the iron overload.
In the ferroportin disease originally described, the early stage demonstrates Kupffer cell iron overload, increasing over time with often large and coalescent deposits in Kupffer cells and macrophages and some deposition in hepatocytes ( Fig. 4.12 A ). In FPN -associated HC (erroneously classified as ‘ferroportin disease type B’ by OMIM), in contrast, there is evidence of primary hepatocyte loading ( Fig. 4.12 B ).
Although venesection is again the cornerstone of therapy, it may not be tolerated equally in all patients, and low Tf saturation with anaemia may be rapidly established despite serum ferritin still being elevated. If phlebotomy is discontinued, there is a rapid increase in the ferritin level, and both oral chelation and erythropoietin may be of some benefit. Ferroportin disease must be suspected in any individual with unexplained hyperferritinaemia and investigated with serum iron studies and genetic testing, if available, of the immediate family.
Hereditary hyperferritinaemia
Elevation of serum ferritin results from a variety of causes, acquired or hereditary. Acquired causes include a plethora of conditions, such as macrophage activation syndrome, inflammation, malignancy, alcohol, insulin resistance and other primary liver diseases. Hereditary causes encompass the various types of hereditary HC. After eliminating the former, most variants of HC are confirmed by symptomatology, a biochemical profile including ferritin and Tf saturation, visceral iron overload and genotyping. A high serum ferritin in the face of high Tf saturation would suggest hereditary HC that may be HFE related, juvenile or the TfR2 variant. Conversely, a high serum ferritin without increased Tf saturation raises the possibility of ferroportin disease or African iron overload. However, in the absence of an identifiable cause, the presence of a high to very high serum ferritin with normal to only moderately raised Tf saturation may suggest a hereditary hyperferritinaemia.
The hereditary hyperferritinaemias, excluding ferroportin disease, are mostly related to mutations in the ferritin light-chain gene FTL , which causes the hereditary hyperferritinaemia cataract syndrome, or the caeruloplasmin gene, as discussed later. The hereditary hyperferritinaemia cataract syndrome is characterized by an elevation of serum ferritin with early onset cataract formation. The hyperferritinaemia is caused by the aberrant translation of the light chain of ferritin, which is caused by mutations throughout the 5′-untranslated region of the IRE of the FTL gene. This rare condition is not associated with visceral iron overload. Although liver histology data are limited, there typically is no iron accumulation. FTL mutations also lead to hereditary neuroferritinopathy, a neurodegenerative disorder associated with haemosiderin and ferritin accumulation in the central nervous system (CNS), particularly the basal ganglia. Iron does not appear to accumulate in the liver, although data are sparse. However, the hepatocyte nuclei in one study were found to have large, pale nuclear inclusions that stained light blue on Perls iron stain.
In contrast to light-chain mutations, mutations of the FTH1 gene, encoding the ferritin heavy chain, are associated with both hyperferritinaemia and concomitant liver iron overload, as described in four of seven members of a Japanese family. It is a dominantly inherited iron overload with a heterozygous single point mutation (A49U) in the IRE motif of H-ferritin mRNA. The proband, a 56-year-old woman, was coincidentally identified during workup for gastric cancer and had a serum ferritin of 1654 µg/L and a moderately increased Tf saturation of 58%. Liver biopsy demonstrated a pattern of iron overload similar to hereditary HC, in addition to heavy loading of iron in splenic macrophages. No further cases have been identified since the original series, and the results are not reproduced in the heterozygous H-ferritin knockout mouse that demonstrates iron overload in the spleen but not the liver. This entity has been called ‘haemochromatosis type 5’, but this terminology was not adopted in the OMIM database.
Atransferrinaemia and hypotransferrinaemia
Hypotransferrinaemia may be acquired in cases of infection, primary liver disease and malignant tumours. An extremely rare hereditary disorder, atransferrinaemia was first described in a young girl with severe hypochromic anaemia and marked, generalized iron overload and has since been described in <10 families worldwide, with molecular characterization in several individuals. An autosomal recessive inheritance is postulated, with possible consanguinity identified in one patient. Data also suggest the mutation may be slightly ‘leaky’, with a small amount of gene product being produced in mouse models and some patients.
Transferrin delivers iron to the erythroid precursors, and the defect leads to decreased haemoglobin synthesis resulting in a severe microcytic, hypochromic anaemia. However, this in turn leads to increased intestinal absorption of iron. This excess iron is inefficiently handled in the plasma, being imported by parenchymal cells, often leading to severe parenchymal iron overload at sites including liver, myocardium, pancreas and thyroid. Histological descriptions are scant, but the liver biopsy specimen can show marked iron deposition in both hepatocytes and Kupffer cells. Mild fibrosis has also been reported.
The clinical presentation and features of the deficiency include pallor and fatigue with high serum ferritin, high serum iron and decreased total iron-binding capacity (TIBC) with absent to low Tf saturation. Treatment may be relatively effective, at least in some patients, with a combination of infusion of fresh-frozen plasma and subsequent phlebotomy or chelation therapy. A spontaneous form of hypotransferrinaemia has been identified in mice, resulting in circulating Tf levels at about 1% of normal. Animals survive for only short periods unless treated with transfusions and transferrin replacement. Development is then normal apart from iron overload in multiple organs and subtle architectural changes in the CNS.
Acaeruloplasminaemia
Caeruloplasmin is a serum glycoprotein and a copper-containing ferroxidase encoded by the CP gene, located on chromosome 3q21–24.7. Caeruloplasmin is synthesized by hepatocytes and catalyzes the oxidation of ferrous to ferric iron, which is necessary for the release of iron to plasma transferrin. Recognition of this role came with the identification of patients with acaeruloplasminaemia, an extremely rare autosomal recessive disease described mainly in Japanese patients. Loss-of-function mutations in the caeruloplasmin gene result in iron overload in the liver and pancreas and progressive neurodegeneration.
Patients develop diabetes mellitus, retinal degeneration, ataxia and dementia late in life. A mild to moderate degree of anaemia with low serum iron and elevated serum ferritin is a constant feature, and the pattern of hepatic iron overload is reminiscent of hereditary HC. As noted by Bosio and colleagues, although rare in Caucasians, acaeruloplasminaemia is sometimes identified and should be included in the differential diagnosis of anaemia with high serum ferritin level.
The liver biopsy in most reported cases has shown marked, predominantly hepatocellular iron deposition, with a panlobular distribution and pericanalicular pattern. Kupffer cell iron accumulation can be seen but is limited compared to the striking hepatocellular iron deposits. Inflammation is absent or minimal. Fibrosis has been absent or minimal in cases reported to date, despite the striking iron accumulation.
Miscellaneous iron overload
Neonatal haemochromatosis
Neonatal haemochromatosis is a disorder characterized by stillbirth or hepatic failure during the first days to weeks of life, associated with systemic iron loading, including iron accumulation in the liver. The majority of neonatal HC cases result from a poorly understood alloimmune gestational disease in which maternal antibodies cross the placenta and attack the fetal liver in utero. There is a high rate of recurrence if a mother has a previous pregnancy complicated by neonatal HC. In fact, the recurrence rate approaches 90%, higher than predicted for simple Mendelian autosomal recessive inheritance, an early clue that this disease was different than other hereditary iron overload syndromes.
The newborn with neonatal HC is often stillborn or premature, exhibiting intrauterine growth retardation and often associated placental oedema and either oligohydramnios or polyhydramnios. The overriding presentation is one of liver failure in the antenatal or early neonatal period, usually within hours to days after delivery. Jaundice with coagulopathy, hypoglycaemia and hypoalbuminaemia are common with low Tf saturation, low TIBC and high serum ferritin. The latter may reflect nonspecific liver disease or inflammation. The diagnosis is made after exclusion of other causes of liver failure and may be confirmed by oral cavity salivary gland biopsy, which is informative when excess iron is demonstrated. However, a negative biopsy does not exclude neonatal HC, even if there is adequate sampling of the salivary glands, because iron deposition can be patchy. MRI can also help confirm the diagnosis, demonstrating iron deposition in liver, pancreas and heart, but with sparing of the spleen.
Autopsy findings vary considerably depending on the length of survival and can range from an oedematous, enlarged liver, secondary to inflammation and cholestasis, to a shrunken liver with cirrhosis. On histological examination, the findings also range from extensive necrosis with almost no residual hepatocytes to a greatly inflamed liver with giant cell transformation, cholestasis and postnecrotic regenerative nodules. The hepatocytes stain positive for the C5b-9 complex, the terminal complement cascade that is formed in the assembly of the membrane attack protein complex, consistent with an alloimmune injury. An iron stain can demonstrate haemosiderin deposition in hepatocytes, reticuloendothelial system and sometimes in bile duct cells ( Fig. 4.13 ). Interestingly, the grade of iron deposition often correlates inversely with fibrosis, and livers with fully established cirrhosis tend to have the least iron accumulation. Regenerative nodules may show cytological atypia. Extensive haemosiderin deposition is also typically identified in many other organs, including myocardium, pancreas, gastric mucosa, thymus, thyroid and salivary glands.
As indicated earlier, most cases of neonatal HC result from gestational alloimmune injury to the fetal liver injury. Importantly, neonatal HC can be treated in the antenatal period with intravenous immune globulin (IVIG), starting at 14 weeks’ gestation, which prevents the development of neonatal HC. Although spontaneous remissions have been reported in rare cases, the prognosis of neonatal HC without prenatal treatment has historically been poor. Postnatal therapy now focuses on exchange transfusion and IVIG; in one study, a good outcome was achieved in 75% of cases, compared to 17% of historical controls. Antioxidant therapy and chelation are of limited value.
It is important to note that the liver in healthy infants can have physiological iron deposits in the periportal hepatocytes during the first 10 days of life. This transient increase in hepatocellular iron deposits may raise the differential diagnosis of neonatal HC when these patients present with severe liver disease. The iron deposits may be prominent and prolonged in relation to nonspecific injuries, as emphasized in numerous studies of metabolic disorders, including tyrosinaemia, Zellweger syndrome, Gaucher disease, mitochondriopathies, tyrosinaemia and galactosaemia. This general pattern of liver injury with increased liver iron can also be seen in patients with other metabolic and infectious disorders, including Down syndrome, trichohepatoenteric and GRACILE syndromes, cytomegalovirus and ‘non-A, non-B’ hepatitis. These observations underscore the importance of extrahepatic iron deposits in making the diagnosis of neonatal haemochromatosis.
African (American) iron overload
Strachan originally identified iron overload in black African patients in 1929, based on a necropsy study of 876 individuals. These patients were from several parts of southern and central Africa, dying in Johannesburg, South Africa, from 1925 to 1928. He noted its commonality, postulated a dietary origin and identified changes ranging from mild pigmentation to established cirrhosis and ‘bronzed diabetes’. Individuals with African iron overload are mostly male patients, and the levels of iron accumulation can reach that of hereditary HC. Although declining in urbanized communities in South Africa, African iron overload, formerly called ‘Bantu siderosis’, is still an important pathology in urban and in rural societies, where up to 5% and 15% of adults, respectively, may be affected.
In southern Africa, more than two-thirds of the rural adult population consumes traditional beverages. Traditional teaching is that African iron overload is the consequence of consumption of large quantities of traditional beer prepared in iron pots or drums. During cooking or brewing in these iron pots, the pH falls to acidic levels, thus leeching an ionized and bioavailable iron into the food or beverage. African iron overload was thus originally thought to result exclusively from surplus dietary iron. However, more recent research in the field indicates that not all those who consume large quantities of traditional beer acquire the disease, and some individuals who consume no traditional beer still develop the disease. This has led to a revised model in which African iron overload may have both dietary and genetic components. There is no evidence that African iron overload results from HFE mutations. There has also been interest in iron overload among African American patients who present with mild anaemia and possible tendency to iron loading, thought to be underdiagnosed in this population. Its relationship to African iron overload is unclear, although a mutation of the ferroportin gene (Q248H) has recently been identified in a minority of Africans and African-Americans with iron overload. However, its significance in the pathogenesis of these disorders is currently unknown, and the data overall do not strongly support a major role for Q248H mutations in African iron overload.
The liver pathology in African iron overload has traditionally been described as a primary reticuloendothelial disease with secondary involvement of the hepatocytes. However, a review of patients in later years has identified a range of morphological changes, from primary deposition in hepatocytes ( Figs 4.14 and 4.15 ), through a mixed pattern, to the predominant reticuloendothelial pattern described by earlier workers ( Fig. 4.16 ). This bears a resemblance to the patterns of alternate Kupffer cell and hepatocyte dominance identified in ferroportin-linked HC. Despite the alcoholic connotation of African iron overload, there is very rarely evidence of acute or chronic alcoholic liver disease. Steatosis, alcoholic hepatitis and perivenular fibrosis are not identified, and the pattern of fibrosis is characteristically progressive portoportal linking, with normal hepatic veins not identified until late in the disease after cirrhosis has intervened.