August 2001
Corresponding author: A J Stanley:
e mail: adrian.stanley@northglasgow.scot.nhs.uk
Genetic haernochromatosis (GH) is the commonest inherited disorder in Caucasian populations and typically affects individuals of North European or Celtic descent. The prevalence has been reported to be between 1 in 200 to 1 in 400 in Western countries,1?3 and an autosomal recessive inheritance has been recognised for more than 25 years.4 It leads to excessive absorption of dietary iron with resulting iron overload in the liver, heart, pancreas, joints, gonads and skin. In 1996, the common gene defect that leads to the development of this condition was identified. A missense mutation (C282Y) leads to the substitution of tyrosine for cysteine at the 282 amino acid position of the protein product of the HFE gene which is located on the short arm of chromosome six.5 A second, more common mutation (H63D) does not appear to cause clinical sequelae unless associated with a single C282Y mutation (“compound heterozygotes”), in which case affected individuals may develop iron overload.
The early population screening studies in Western countries established a prevalence of GH based on the biochemical expression of the disorder as indicated by elevated transferrin saturation and serum ferritin.6?9 In some studies, HLA?typing was also used, followed by a confirmatory liver biopsy. However, these studies were undertaken prior to the availability of accurate gene mutation analysis. In Britain, over 90% of individuals with GH are homozygous for the C282Y mutation whilst another 4% are compound heterozygotes.10 The largest study examining the mutation frequency in Britain reported a frequency of the C282Y mutation of 8.23% (1 in 12) amongst 10,500 South Wales blood donors.11 Seventy?two donors were homozygous for C282Y (1 in 147) and 253 were compound heterozygotes for C282Y and H63D (1 in 42).
Olynk et al conducted a population?based study on 3,011 unrelated, white adults in Busselton, Australia and found that 1 in 200 were homozygous for the C282Y mutation.3 A retrospective study was recently conducted in Scandinavia examining anonymised dried blood of newborns.12 It reported that Denmark had an 8.2% carrier rate of the C282Y mutation. Although no direct prevalence studies have been undertaken in Scotland, several other population studies support the hypothesis that GH has arisen from an ancient Celtic race.13 ?16 Therefore the prevalence of the condition in the Scottish population is likely to be high. If so, the important question is: where are all the patients? The problem is that despite the high prevalence of genetic haemochromatosis, the clinical expression of the disorder (phenotypic haemochromatosis) appears much less common. Early reviews relying on the classical features of cirrhosis, pigmentation, diabetes and arthropathy for diagnosis suggested a frequency of 1 in 20,000 hospital admissions in the United States.17 However, post?mortem studies suggest that this figure is a gross underestimate, reporting a frequency of 1 or 2 per 1000.18,19 This highlights the important fact that whilst the prevalence of GH is extremely common in certain populations, its clinical presentation and diagnosis is relatively uncommon. Whether this is due to a lack of penetrance or simple under?diagnosis is unclear.
No study has prospectively followed up healthy homozygotes without venesection, to ascertain the proportion who subsequently develop symptoms or complications. A study which shed some light on this problem was that by Powell et al, who reported that 96% of homozygotes expressed the disease biochemically as indicated by increased hepatic iron concentration over an eight years period.20 Other studies have suggested that approximately half of male and a quarter of female homozygotes will develop life?threatening complications of this disorder.21,22 The uncertainty about the clinical expression of GH may be resolved in the near future by population screening and cohort studies which are already in progress in over 15 countries. However, considering the estimated prevalence of the common mutations, there is likely to be a large number of patients in the Scottish population. Yet from our (and other centres) anecdotal experience, the number of patients with known haemochromatosis remains relatively small. In addition, many of those with known GH were diagnosed following screening family members of an affected individual. Therefore, there appears to be a strong argument to improve diagnosis of this condition and identify individuals with undiagnosed haemochromatosis. Indeed, there have been calls to introduce population screening for GH and the Centre for Disease Control has recommended screening by transferrin saturation for the entire US population.23 The condition fulfils most of the accepted criteria for screening. It has an effective treatment in the form of phlebotomy, which if implemented before evidence of end?organ damage, results in normal life expectancy.24 We also now have specific and sensitive screening methods. Biochemical markers of iron overload such as a transferrin saturation level >50% for women and >60% for men have a sensitivity of 0.92 and a specificity of 0.93 with a positive predictive value of 86%.25,26 These can be used prior to confirmation of the disorder by the more expensive genetic test. Modelling studies have also suggested that population screening for GH is cost effective.27 However, until more is known about the disease penetrance and natural history, there are major problems with the introduction of population screening. Labelling individuals with a genetic “disease” without clinical expression is a major concern. It can lead to difficulties with insurance policies, adoption and family relationships. In addition the psychological effect of such information on an asymptomatic individual is unknown. Although a number of bodies in this country are currently drawing up guidelines for genetic diagnosis and research, the laws governing this issue are currently unclear.28 Due to these problems, it is generally accepted that population screening for GH should not be introduced at present.29
In the meantime, it seems a reasonable compromise to screen high risk groups for GH.30,31 Most patients attending liver clinics for investigation of liver dysfunction are already assessed for evidence of iron overload. Interestingly, this approach has not led to particularly high rates of GH diagnosis.32 Two other groups of particular interest are diabetics and patients attending rheumatology clinics. To date, five studies have examined the prevalence of haemochromatosis in a diabetic population.33-37 Although the data is not fully consistent, there appears to be a slightly increased prevalence of GH in this patient group, with a carrier rate of between 1 in 100?200 confirmed by liver biopsy. Arthralgia is a common presenting symptom of haemochromatosis, and the highest prevalence of previously unrecognised GH (1.5%) was reported from screened patients attending an Australian rheumatology clinic.38
No study has examined the prevalence of haemochromatosis amongst patients with cardiac failure or chronic fatigue, although these may be other potential groups for screening. In general, targeted screening of selected sub?groups such as these should result in identification of greater numbers of patients with GH, and allow screening of family members. However, it is worth noting that many of those identified using this approach will already have cirrhosis or evidence of other end?organ damage, therefore further research is required to assess the benefits of this approach.
In summary, due to the Celtic preponderance of GH, the prevalence of the condition in Scotland is probably amongst the highest in the world. However it remains underdiagnosed and limited research on GH has been undertaken in this country. Until further data are available regarding its penetrance and clinical progression, population screening of this disorder is not recommended. In the meantime, screening high?risk groups such as those attending diabetic, rheumatology or cardiology clinics appears to be a sensible strategy. Although further data are awaited to confirm which groups benefit most from testing, this approach should increase rates of diagnosis of GH, allow appropriate intervention, and enable predictive genetic testing of family members to be undertaken. In addition we must increase awareness of this treatable disorder amongst health professionals and the public and lower the threshold for diagnostic investigations.