Epidemiology, clinical spectrum and prognosis of hemochromatosis

Eleven prospective epidemiological studies from various countries have as yet evaluated the gene prevalence of HLA-linked hemochromatosis. The estimated frequency ranged from 0.027-0.107, the frequency of homozygotes from 0.00074-0.0116, and the frequency of heterozygotes from 0.052-0.191. In a meta-analysis of the eleven surveys the frequency is 0.0016 (106/64758 subjects) for homozygotes which corresponds to a gene frequency of 0.041 and a frequency of heterozygotes of 0.078. Further analyses showed that some of these studies have probably underestimated the prevalence which in reality is probably two- to threefold higher than estimated by the meta-analysis.

In the total group of 251 patients diagnosed with hemochromatosis in the University of Düsseldorf from 1959-1992, abnormality in liver function tests (75%), weakness and lethargy (74%), skin hyperpigmentation (70%), diabetes mellitus (48%), arthralgia (44%), impotence (45% in males), and ECG abnormalities (31%) were the most frequent findings and symptoms at diagnosis. In recent years about 50% of patients were detected without having liver cirrhosis and 20% of patients did not have any symptoms and pathology except iron overload.

Survival analysis in the 251 patients showed that in the absence of cirrhosis and diabetes iron removal by phlebotomy therapy prevents further tissue damage and guarantees a normal life expectancy. Sex and presence of arthropathy did not predict prognosis. However, patients with massive and long-lasting iron overload had a worse prognosis than patients with less severe iron excess. Iron removal in general ameliorated liver disease, weakness and cardiac abnormalities, and also prevented the progression of endocrine alterations. Therapy, however, did not influence arthropathy which even got worse in several patients. Iron removal also failed to reverse insulin-dependent diabetes. During a mean followup of 13.4 years 69 deaths occurred in the 251 patients. In 19 patients death was due to liver cancer, in 14 due to liver cirrhosis, in 5 due to cardiomyopathy, and in 3 due to diabetes mellitus (all causes significantly more frequent than expected for the general population). The other causes of death were as frequent as expected including extrahepatic malignancies. All liver cancers were seen in cirrhotic livers, but often occurred many years or even decades after complete iron removal. Further strategies have to evaluate the design of screening programs in order to diagnose more patients in the precirrhotic and asymptomatic stage.