In genetic hemochromatosis (HC), the intestinal absorption of iron is high and multi-organ iron overload with organ failure may occur over a period of decades. Myocardial iron loading is well known as a possible complication, but as far as we are aware, this is the first study of myocardial iron loading in newly presenting patients with possible HC. This is now feasible because of the non-invasive nature of T2* CMR as myocardial biopsy to assess iron loading is restricted to highly selected cases. Analysis of possible predictors of myocardial iron loading as assessed by T2* showed that all genetically confirmed HFE-HC patients with myocardial siderosis had a presenting serum ferritin value of >1000 μg/L, a value that is also associated with an increased risk of developing liver pathology. Above this threshold, myocardial siderosis was present in 6/18 (33%) patients compared with 0/13 (0%) patients with presenting ferritin values <1000 µg/L (p=0.028). Whilst the lowest serum ferritin measured in genetically confirmed HFE-HC patients with myocardial siderosis was 1300 ug/L, a more conservative threshold of 1000 ug/L would appear to be a more appropriate level for clinical practice, because the patient with the lowest serum ferritin of 1300 ug/L had a myocardial T2* of 7.4 ms indicating severe iron loading. The limited sample size of this study may therefore not have identified patients with a milder degree of myocardial iron loading associated with a lower serum ferritin. As this lower serum ferritin threshold would not miss patients with myocardial siderosis, it would seem reasonable on this basis of the data presented here to recommend T2* CMR on all confirmed HC patients presenting with a serum ferritin >1000 µg/L.
Another consideration in HC, is the development of heart failure. Dilated cardiomyopathy (DCM) is well recorded in HC, and reports recommend that serum ferritin should be measured in newly presenting cases of DCM [28
]. In thalassemia, there is unequivocal evidence showing that iron causes the LV dysfunction, because severe myocardial siderosis is present in patients with heart failure [16
], and LV dysfunction is reversible with iron chelation during which time the myocardial siderosis improves [18
]. The relation between myocardial siderosis and LV dysfunction has never been shown in-vivo in HC. In addition the cause of LV dysfunction in HC is controversial. Apart from myocardial iron deposition, there have been 3 alternative proposed explanations.
The first alternative explanation for LV dysfunction in HC is that myocardial damage results from coronary artery disease (CAD) exacerbated by chronically increased iron levels. A correlation has been shown between high ferritin levels, LDL cholesterol and risk of myocardial infarction, with a plausible mechanism of augmented lipid peroxidation [32
]. Iron could also promote atherogenesis and post-ischemic myocardial injury and perhaps arrhythmias [33
]. Supportive data for this view includes the lower incidence of coronary artery disease in women due to lower iron levels through life resulting from menstruation, and the increased incidence of CAD in a cohort of 2,873 Framingham women who had a natural or surgical menopause. However, others have not found a relation between storage iron and myocardial infarction [34
]. An additional issue in HC is the possible interaction of the genetic abnormality and risk of coronary disease. There has been a report of an increased risk of myocardial infarction in men who are carriers of the HFE C282Y mutation, which may suggest that that the mutation might confer an increased risk of CAD [35
]. A study of 20,555 post menopausal women also showed that women with CAD risk factors who were heterozygous for the HFE C282Y mutation developed more CAD in comparison with matched women with homozygous wild-type HFE genotype [36
]. These studies suggest that the HFE mutation in itself may promote CAD that could explain LV dysfunction.
The second alternative hypothesis is that the HFE gene is directly causative for DCM, with an increased frequency of the H63D mutation, but not the C282Y mutation [37
]. The survival rate in this study was similar between the H63D carriers and non-carriers, but patients with the C282Y mutation had a shorter duration of illness before presentation, less LV volume dilatation and better fractional shortening. This is somewhat surprising because H63D has less of an effect on iron metabolism compared with C282Y, which suggests that factors other than just direct iron damage may predispose the patients to DCM. However, other studies have not observed differences in frequency between the DCM population with C282Y or H63D mutations and non carriers [3
], including the CARDIGENE study in which no difference in HFE carrier frequency was found between DCM patients and controls [38
]. If genetic predisposition is a cause of DCM in HC [39
], there is limited understanding of the mechanism.
A third explanation for LV dysfunction in HC is the possibility of predisposition to autoimmune disease, as the HFE gene is located close to the human leucocyte antigen (HLA) locus. An autoimmune mechanism for DCM has been suggested in subjects where there is an abnormal response of the immune system to different insults [37
]. The low CD8+ T cell count in DCM patients have been stated to support this hypothesis [41
Therefore the four possible explanations for the LV dysfunction in HC, either alone or in concert, remain unresolved in relative importance. Our data from the current study substantially clarify this debate. We found 2 groups of genetically confirmed HFE-HC patients with LV dysfunction. The first group were those patients with myocardial siderosis (T2* <20 ms). The prevalence of myocardial siderosis at first presentation of HC in patients with serum ferritin >1000 ug/L was high at 33%. Of these genetically confirmed HFE-HC patients with myocardial siderosis, 83% had associated LV dysfunction and all had heart failure. Myocardial siderosis was therefore the commonest cause of LV dysfunction. A second group of genetically confirmed HFE-HC patients in our study with LV dysfunction had no myocardial iron loading but had well documented existing causes, namely anterior myocardial infarction, or substance abuse. Two other cases of LV dysfunction occurred in the genetically unconfirmed HC patients, which were caused by myocarditis or cancer treatment.
It was notable that evident coronary artery disease was not common in this predominantly middle-aged to elderly population, and there was little evidence of unexplained cardiomyopathy to support the other possible etiologies (direct genetic effects or autoimmunity) which have been previously postulated for LV dysfunction in HC. This indicates that myocardial T2* is required to definitively diagnose myocardial siderosis in newly presenting patients with HC. This has the added merit in patients with impaired or borderline LV function of confidently excluding myocardial iron loading as the cause where possible alternative diagnoses are known or suspected, such as coronary disease or the prior use of cardiotoxic drugs, as long-term specific treatment for the cause of LV dysfunction is required.
The limitations of this research include the relatively small sample size, which relates to the difficulty of recruiting newly-presenting HC patients. Late gadolinium enhancement was performed only in 7 patients in whom there was a possible history of cardiac disease and therefore there is limited data on cardiac fibrosis. Of the 5 genetically confirmed
HFE-HC patients who presented in heart failure with a low cardiac T2*, 2 had LGE imaging with one having no LGE and one having mid-wall LGE with no prior history of cardiac pathology. It is not clear whether the latter patient’s LGE could be attributed to myocardial siderosis. Only 2 patients were referred to be rescanned for clinical reasons, and therefore the response of cardiac T2* to venesection is not known for all patients. However, in the 2 that were rescanned, there was an impressive improvement in myocardial T2* and EF with venesection over nearly 3 years in both cases. We did not test for nutritional factors including thiamine and vitamin D [42
], which may exacerbate heart failure in other iron overload conditions such as thalassemia major, although there is little evidence for derangement in HC. The patient cohort comprised of patients with a clinical diagnosis of HC and subsequent genetic analysis was only able to confirm the genetic diagnosis in 31 of 41 patients studied. We present the data of these 31 genetically confirmed
HFE-HC patients separately to ensure comparability with future studies. Further detailed genetic analysis of the 10 genetically unconfirmed
HC patients was not available.