Hereditary hemochromatosis may lead to hepatic cirrhosis, cardiomyopathy, diabetes, arthritis, and impotence (1, 2). In the Caucasian population, HFE gene mutations (C282Y and H63D) are present in the majority of patients demonstrating phenotypic expression (3–6). Conversely, the clinical penetrance in mutation carriers is low (7).
In the precirrhotic stage, ∼20% of hemochromatosic patients demonstrate hyperglycemia, with the prevalence increasing to >70% in the presence of liver cirrhosis (8). Two mechanisms contribute to the development of hyperglycemia and diabetes. Liver iron overload leads to insulin resistance, and the pancreatic β-cell iron accumulation results in cell damage and diminished insulin secretion (1). The prevalence of genotypic and phenotypic hemochromatosis is higher in diabetic versus nondiabetic populations (9–11).
In an attempt to distinguish patients with hemochromatosis-associated secondary diabetes from other forms of diabetes, we applied a screening program to diabetic patients aged ≥40 years. The present report summarizes our first 3 years’ experience.
From the year 2000, in-hospital patients with diabetes have been invited to participate in the screening program (12). The initial phenotypic screening consisted of serum iron and transferrin analyses and calculation of the transferrin saturation. Elevated transferrin saturation defined as >45% was confirmed by an independent second analysis, which included determinations of serum ferritin and liver enzymes (alanine aminotransferase, aspartate aminotransferase, and γ-glutamyl transpeptidase). Patients with a phenotype indicative of iron overload were genotyped for the presence of aberrant HFE gene variants (C282Y, H63D, and S65C) (13, 14).
Among 3,500 diabetic patients, elevated transferrin saturation at two occasions was observed in 22 male (aged 52.9 ± 11.6 years) and in 12 female (aged 49.1 ± 14.0 years) subjects, resulting in an iron overload phenotype in 0.97% of the patients. Mutations in the HFE gene were identified in seven patients (20.6%), with five of them demonstrating homozygosity for the variant allele at position 282. In addition, one case of compound heterozygosity was documented (C282Y/H63D). The new splice site HFE gene mutation (IVS5 + 1 G/A) in one patient of Vietnamese origin has been reported separately (15). Serum ferritin concentration was increased in five of seven patients, whereas pathological liver enzyme activities were noted in four patients. The remaining 27 patients were characterized by the absence of the three most common HFE gene mutations. Among them, 13 patients demonstrated elevated serum ferritin concentrations and 12 patients pathological liver enzyme activities.
Phenotype- and genotype-based screening programs have variably been used to investigate the relationship between iron overload, hemochromatosis, and its clinical manifestations including diabetes (16). Using transferrin saturation as a clinically useful initial screening parameter (12, 16) to detect iron overload, we obtained evidence for iron overload in only 1% of a large number of diabetic patients. Approximately half of these patients demonstrated liver damage as reflected by increased liver enzyme activities. Genotyping for the most common HFE gene mutations revealed hereditary hemochromatosis in only one in seven of these patients, equal to 0.2% of the overall diabetic population investigated in the present study.
The screening program can be performed at a reasonable price and is considered efficient in detecting iron overload in diabetic patients. Direct laboratory costs amount to $2.50 U.S. per patient for the laboratory screening profile (transferrin saturation and liver enzymes) and $5.00 U.S. for the analysis of serum ferritin. The genotyping approach (HFE C282Y, H63D, and S65C) results in $40.00 U.S. of additional costs in selected patients. The cost-effectiveness of screening programs for hereditary hemochromatosis was previously assessed. A study (17) of the cost-effectiveness of a population screening approach concluded that screening was an effective strategy for asymptomatic subjects if the prevalence of hemochromatosis was at least 0.3%, accompanied by a probability of disease manifestation >0.4 and test costs <$12.00 U.S. In another study (18), the cost-effectiveness of genetic testing as a screening method was evaluated using a decision model. The model estimated that initial genetic testing would be cost-effective if the cost per test was <$28.00 U.S. Based on the assumption that diabetes is the manifestation of hereditary hemochromatosis, we conclude that a stepwise laboratory diagnostic approach is a cost-effective method for the screening of hereditary hemochromatosis among diabetic patients. In hemochromatosic patients with hyperglycemia/diabetes, it may open the possibility to treat iron accumulation with the aim to prevent further damage to the liver and pancreatic β-cells, which in turn may be expected to contribute to preservation of the remaining insulin secretion, which even at low concentrations may delay the development of diabetes complications (19). Interventional trials should be performed to test the clinical outcome of a combined screening/treatment approach.
