|
Journal Publication |
||||||||||||
Hereditary
Hemochromatosis: Perspectives of Public Health, Medical Genetics, and Primary
Care
Giuseppina Imperatore, MD, PhD 1; Linda E. Pinsky, MD 2; Arno Motulsky, MD 2; Michele Reyes, PhD 1; Linda A. Bradley, PhD 4; Wylie Burke, MD, PhD 3 Hereditary hemochromatosis (HHC) is a condition characterized by excess iron in body tissues, resulting in complications such as cirrhosis, cardiomyopathy, diabetes, and arthritis. These complications usually manifest during adulthood. Two methods of screening for the detection of early stage of HHC are available: serum iron measures and molecular testing to detect mutations in the HFE gene. These phenotypic and genotypic screening tests are of particular interest because a simple treatment-periodic phlebotomy-can be used to prevent iron accumulation and clinical complications. HHC might represent the first adult-onset genetic disorder for which universal population-based screening would be appropriate. Therefore, HHC has been proposed as a paradigm for the introduction of adult genetic diseases into clinical and public health practice. However, universal screening for HHC has not been recommended because of the uncertainty about the natural history of the iron overload or HHC and, in particular, uncertainty about the prevalence of asymptomatic iron overload and the likelihood that it will progress to clinical complications. If universal screening is not appropriate based on current data, what other measures might reduce the disease burden of iron overload? New studies provide more systematic information about the penetrance of the HFE C282Y mutation and shed further light on the natural history of the disorder. The authors review these data and consider their implications for public health, medical genetics, and primary care. Key Words: hereditary
hemochromatosis; screening; genetics; HFE; gene; iron overload
From the 1Centers for Disease Control and Prevention, Atlanta, Georgia; Departments of 2 Medicine and 3 Medical History and Ethics, University of Washington, Seattle, Washington; and 4 Clinigene Laboratories, Hauppauge, New York. Received:
August 30, 2002. Giuseppina Imperatore, MD, PhD, Division of Diabetes Translation, National Center for Chronic Disease Prevention and Health Promotion, Centers for Disease Control and Prevention, 4770 Buford Highway, NE, Mailstop K-10, Atlanta, GA 30341.
Hereditary hemochromatosis (HHC) has been proposed as a paradigm for the introduction of adult genetic diseases into clinical and public health practice. This disorder, inherited as an autosomal recessive condition, can result in the accumulation of iron in body tissues, resulting in complications such as cirrhosis, cardiomyopathy, diabetes, and arthritis. 1 Two methods of screening for early detection of the disease are available: serum iron measures and molecular testing to detect mutations in the HFE gene. 2 These phenotypic and genotypic screening tests are of particular interest because a simple treatment-periodic phlebotomy-can be used to prevent iron accumulation and clinical complications. These characteristics suggest that HHC might represent the first adult-onset genetic disorder for which universal population-based screening would be appropriate. When policy-makers began to consider this screening option, however, important knowledge gaps were identified. 2-5 Little was known about the natural history of iron overload or HHC and, in particular, about the prevalence of asymptomatic iron overload or the likelihood that it would progress to clinical complications. No population-based data were available to address these questions. Studies of patients seen in referral centers provided only partial insight into the natural history of HHC. Some studies suggested that iron accumulation occurred progressively over time, but that the rate was highly variable. Yet the answer to the most important question, the proportion of people with a mild degree of iron overload who would progress to clinically apparent disease, remained uncertain. The discovery of the HFE gene provided a new tool to evaluate these questions. 6 Multiple studies in clinical centers confirmed that the majority of people with a diagnosis of HHC are homozygous for the C282Y mutation of the HFE gene.7 A second HFE mutation, H63D, contributes to risk for iron overload as well, when in the homozygous state or as a compound heterozygotes with C282Y. 8 However, genotyping also raised questions about the natural history of the disease. Families were described wherein some siblings with C282Y homozygosity had classical symptoms of HHC while other siblings with the same genotype remained asymptomatic into old age. 1, 2 On the basis of these data, many policy-makers concluded that additional information was needed before universal screening for HHC could be considered. 2-5 If universal screening is not appropriate based on current data, what other measures might reduce the disease burden of iron overload? New studies provide more systematic information about the penetrance of the HFE C282Y mutation and shed further light on the natural history of the disorder. In this article we review these data and consider their implications for public health, medical genetics, and primary care. NATURAL HISTORY OF HHCInsights from prevalence studies using serum iron measuresThe prevalence of HHC and iron overload can be estimated by using serum iron measures. Most studies have used transferrin saturation (TS = serum iron/total iron binding capacity × 100) and serum ferritin to identify individuals at risk, 1, 9 followed by liver biopsy or quantitative phlebotomy to determine those who are iron-overloaded. However, data comparisons are complicated by differences in screening protocols (e.g., selection of initial screening test, screening thresholds for TS and serum ferritin), populations selected for study, and definitions of iron overload. Most studies have been done in populations of predominantly European descent. Taken together, these studies provide a rough estimate of the prevalence of iron overload in this population. Iron overload has also been observed in African Americans and in African populations, but it is not well characterized clinically; estimates of prevalence in this population are not available. 10-12 Some observations indicate that the predominant mechanism of iron overload is different: TS levels may be lower in people of African descent with iron overload, with iron accumulating primarily in Kuppfer cells, 13 as compared to hepatocytes in HHC. As with HHC, family studies suggest a causative genetic factor. Elevated TS represents the earliest phenotypic finding in HHC. Recommendations for the threshold definition of elevated TS have ranged from 45% to 62%. 14 In the US adult population, the prevalence of elevated TS on random blood draw ranges between 1% and 6%, depending on the value used to define an elevated level. 14 Among persons with an initial elevated TS, repeated testing (usually done on a fasting specimen) has revealed a persistently elevated level in 15% to 54% of persons (Table 1). 15-22 On the basis of these data, the estimated population prevalence of persistently elevated TS varies from 0.3% to 2.0% (Table 1). 15-22 These studies also reported that elevated serum ferritin levels were found in 0% to 61% of those with persistently elevated TS. 15, 16, 19, 21 Confirmatory testing provided evidence of iron overload in about half of subjects with persistently elevated TS 17, 20 and in 60% to 100% of subjects with elevations in both TS and serum ferritin. 16, 23 The wide range of estimates reported in these studies reflects differences in study characteristics noted previously; however, all observe a drop-off in subjects with positive test results with each step in the testing pathway (Table 1). 24 Few studies, however, have assessed the proportion of persons with iron overload who have clinical disease. Bradley et al. 25 evaluated this question in a review of population-based studies. The studies were included if they confirmed iron overload by liver biopsy or quantitative phlebotomy and collected information on clinical complications, including liver fibrosis, liver cirrhosis, cardiomyopathy, arthropathy, diabetes, abdominal pain, and hepatomegaly. At the time these studies were conducted, HFE genotyping was not available. In the pooled analysis, 54% of subjects with iron overload (58% of men and 44% of women) had one or more of these complications. For some, the only finding was asymptomatic liver fibrosis; clinical symptoms were reported in 27% of men (half of those with clinical findings) and 33% of women (75% of those with clinical findings). However, these studies of symptomatic patients did not use control groups; as a result, it is difficult to estimate the proportion of clinical findings attributable to iron overload. This proportion could be very low: for example, studies of patients with end-stage complications of HHC have consistently found fewer affected women than men, 1, 26, 27 suggesting that at least some clinical findings among women identified in prevalence studies are due to other causes. Similarly, a population-based study using TS level as the primary screening method found fewer women than men with significant morbidity. 28 Bradley et al. 25 calculated the weighted average for prevalence of iron overload to be 0.25%; correcting for missed diagnoses and compliance, the authors estimated actual prevalence of iron overload to be 0.5% in men and 0.6% in women. 25 Other studies suggest a prevalence of iron overload ranging from 0.06% to 0.4%. 24 Assuming that one third to one half of people with iron overload have clinical symptoms, these figures suggest that the prevalence of clinical disease due to iron overload is in the range of 0.02% to 0.25%. By comparison, autopsy studies have reported pathological evidence of iron overload in the liver in 0.09% to 0.19% of specimens. 15, 29, 30 The prevalence of recognized HHC as a cause of death is lower, ranging between 0.017% and 0.032%. 31 Estimates for the prevalence of iron overload and associated clinical symptoms based on serum iron measures vary over a 10-fold range. At the upper end (0.4-0.5%), the estimates are in the same range as the prevalence of the C282Y/C282Y genotype. Insights from prevalence studies using HFE genotypeMany studies have evaluated the prevalence of the two HFE mutations associated with iron overload, C282Y and H63D, in different populations around the world. These studies (reviewed by Hanson et al. 7) document that the C282Y mutation and C282Y homozygosity are most prevalent in populations of European descent. Because the C282Y homozygous genotype accounts for the majority of clinically diagnosed cases of HHC, these data support congruence between iron overload in people of European descent and HHC. This overlap is likely to be most complete in populations of Northern European descent. Other rare genetic causes of iron overload have been described in European populations. In Southern Europe, families with a clinical entity indistinguishable from HHC have been described, but genetic studies document causative mutations in two other genes, TFR2 and SLC11A3. 32-38 Another rare disorder, juvenile hemochromatosis, is inherited as an autosomal recessive disorder linked to chromosome 1 and results in severe iron overload by the second decade of life; males and females are equally affected. 36, 39 These genetic entities appear to account for only a small proportion of persons with iron overload. A pooled analysis of population studies of HFE genotypes found a prevalence of 0.4% (95% confidence interval [CI] 0.3-0.5%) for the C282Y/C282Y genotype, 1.6% (95% CI 1.4-1.9%) for the C282Y/H63D genotype, and 1.9% (95% CI 1.6-2.1%) for the H63D/H63D genotype. 7 Similarly, the three largest population-based genotype frequency studies in the United States estimated a prevalence for C282Y homozygosity of 0.3% (95% CI 0.1-0.5), 0.4% (95% CI 0.2-0.9), and 0.5% (95% CI 0.3-0.6) (Table 2). 21, 40, 41 These data are relatively consistent; observed differences are likely be due to differences in racial/ethnic distributions in the populations studied. The pooled analysis of HFE genotypes found that 77.5% (95% CI 75.9-78%) of HHC cases have the C282Y/C282Y genotype. 7 A range of other HFE genotypes including C282Y/H63D (5.3%, 95% CI 4.5-6.2%) and H63D/H63D (1.5%, 95% CI 1.1-2.1%) is found in the remaining cases; some cases carry either a single HFE mutation or none at all. 7 Penetrance of the C282Y/C282Y genotypeCalculation of HFE genotype penetrance-that is, the proportion of persons with the genotype who have (or will develop) clinical disease-represents another way to estimate the prevalence of clinical complications of HHC. Four studies have estimated the penetrance of C282Y homozygosity, the HFE genotype conferring the highest risk. In a Utah study, unselected relatives of HHC patients were tested for iron status and clinical symptoms. 42 Among 214 relatives, HFE mutation testing was done in 158; 87% of these had the C282Y/C282Y genotype. The study used a low threshold for iron overload: serum ferritin >325 μg/L in males or 125 μg/L in females, or a liver biopsy revealing at least 25 μmol of iron/g dry weight (for a man aged 50 years, this level of liver iron would result in a hepatic index of 0.5, compared with the usual threshold of 1.9 cited for diagnosis of HHC). 43 By these criteria, most relatives had iron overload. However, evidence of liver disease was relatively infrequent: Among all male subjects, 12% had cirrhosis (95% CI 7-20%) and 12% had liver fibrosis (95% CI 6-19%), while for females the comparable figures were 2% (95% CI 0.2-7%) and 4% (95% CI 1-10%). A small proportion of patients also had HHC-related arthropathy as detected by radiologic examination. More than 90% of these subjects reported nonspecific symptoms that could be caused by iron overload, including arthralgia, weakness, and abdominal pain, but no comparison group was evaluated. As a result, the proportion of these symptoms attributable to HHC could not be determined. Similarly, a population-based study in Australia (N = 3,011) 28 identified 16 C282Y homozygotes (for a population prevalence of 0.5%), of which 8 had clinical findings: hepatomegaly in 3 [43% (3 of 7) (95% CI 10-82%) of males and none (0 of 9) of females (95% CI 0-34%)] and skin pigmentation and/or arthritis in the other 5. No comparison group was included. A screening study at a health maintenance organization in southern California provided additional information about the clinical status of persons homozygous for C282Y. 44 TS was ≥50% in 75% of males and 40% of females and serum ferritin was ≥250μg/L in 76% of men and >200 μg/L in 54% of females. Compared with control subjects, persons homozygous for C282Y were more likely to have a history of "liver and hepatitis problems" (8% vs. 4%), elevated serum aspartate aminotransferase (8% vs. 4%), and elevated plasma collagen IV, a measure associated with hepatic fibrosis (26% vs. 11%). However, C282Y homozygotes were no more likely to have a history of fatigue, joint pain, impotence, skin pigmentation, or diabetes. Among all 152 subjects with the C282Y/C282Y genotype, only 1, an alcoholic, had a clinical history of end-stage HHC. Two of 119 with complete data had markedly abnormal laboratory values suggestive of severe liver fibrosis. The authors estimated the penetrance of significant clinical disease in persons with the C282Y/C282Y genotype at about 1%. In addition, the California study found a similar prevalence for the C282Y/C282Y genotype among older and younger subjects. 44 A United Kingdom study obtained a similar result: the prevalence of the C282Y/C282Y genotype was 0.67% among elderly men, suggesting no loss of this genotype in men at older ages. 45 These observations also argue for low penetrance of the C282Y/C282Y genotype, because clinical complications resulting in early mortality would be expected to reduce the prevalence of the genotype at older ages. A fourth study assessed clinical disease in a population-based sample of 65,238 adults above age 20 in Norway (median age 49). 46 HHC was diagnosed on the basis of persistently elevated TS, elevated serum ferritin, and the absence of other medical explanations for these abnormalities. Among 92 women and 177 men with HHC diagnosed as a result of the screening study, 85% had the C282Y/C282Y genotype. A third of persons with HHC had elevated serum transaminase levels, and 4% had diabetes. Fatigue was reported by 16% of women and 14% of men, joint pain by 13% of women and 20% of men; 3% of men reported impotence. No comparison group was evaluated. In addition, no clinical information was reported on three women and six men with HHC diagnosed prior to the screening study. These studies have important limitations. The Utah, Australian, and Norwegian studies failed to include comparison groups. In addition, the Utah study evaluated relatives of patients with clinical disease due to HHC, a group that might be more likely to develop clinical disease due to shared environment or genetic background. Conversely, the California study drew subjects from a prevention clinic, thus potentially selecting against patients with clinical disease. In addition, this study could not fully evaluate 28 of the 152 C282Y homozygotes; the authors state that these subjects, for whom questionnaire data were not available, were diagnosed on the basis of screening. Despite the potential biases, these studies all suggest that only a minority of C282Y homozygotes develop clinical disease attributable to iron overload. However, the actual proportion with clinical symptoms remains uncertain. Considering these studies and those based only on iron measures, the clinical penetrance of HHC could range from 1% to 50%.
|