In the news

A major obstacle to assessing the desirability of population screening for hereditary haemochromatosis has been uncertainty about the penetrance of haemochromatosis-associated mutations in the HFE gene (see Haemochromatosis page for background information). Willis et al have attempted to estimate penetrance by performing a retrospective analysis of HFE genotypes in liver biopsy specimens from patients with liver disease (cirrhosis or liver cancer) over a 20 year period among the population of approximately 250,000 served by the Norfolk and Norwich district hospital in the UK [Willis, G et al. (2000) Gut 46, 401-404 (Abstract)]. They estimated that fewer than 2% of people with haemochromatosis-related genotypes have been treated by phlebotomy during this period. The percentage of liver disease patients with these genotypes (particularly homozygosity for the C282Y mutation) was significantly higher than the frequency of these genotypes in the population as a whole. However, when the number of C282Y homozygotes among liver disease patients was compared with the total number of births or deaths of  C282Y homozygotes expected within the total population over 20 years, it appeared that only about 2.5% of C282Y homozygotes in this population could expect a diagnosis of liver disease within their lifetime. Assuming that liver disease accounts for about half the phenotypic manifestation of haemochromatosis, Willis et al suggest that the penetrance is only about 5%.

Comment: The work of Willis et al does not lend support to calls for population screening for haemochromatosis. Although their study has some limitations, even a three- or four-fold underestimate would still suggest a relatively low penetrance and an unfavourable cost/benefit ratio. In this connection, Hickman et al, in a report in the same issue of Gut, suggest that measurement of unsaturated iron-binding capacity, following by genetic testing for those with an elevated measurement, is a suitable automatable method for population screening [Hickman, P.E. (2000) Gut 46, 405-409 (Abstract)]. They calculate the cost of detecting one "case" as about AUS$2300 (around £1000), but no counselling costs are included in this estimate, and the definition of a "case" must be questioned if iron overload is not always associated with disease. In an editorial, Adams suggests that the way forward in the haemochromatosis debate is for both haemochromatosis physicians and public health specialists to become more aware of the the contribution of each other's expertise: physicians advocating population screening need to understand the importance of knowing the penetrance of the mutations and the population attributable risk, while public health specialists need to appreciate the morbidity associated what has almost certainly been an underdiagnosed disease [Adams, P. (2000) Gut 46, 301-303]. Greater awareness could make screening unnecessary.