We thank Dr. Castiella and his colleagues for their interest in our recent study of noninvasive markers of fibrosis in patients with C282Y hemochromatosis.1 We agree that studying the clinical role of noninvasive markers in this population is of great interest with significant potential implications for patient care. Dr. Castiella and his colleagues are correct to point out that none of the patients in our study with cirrhosis had a platelet count > 200 × 109/L and normal aspartate aminotransferase; this supports their own contention that these characteristics confer a high negative predictive value for cirrhosis. We agree that these variables should be taken into account when liver biopsy is considered in affected patients. However, in our study, only 10 of the 26 patients with serum ferritin > 1000 μg/L had cirrhosis, and this indicates the need for better noninvasive markers of cirrhosis in this population. In our population, a serum hyaluronic acid concentration > 46.5 ng/mL was 100% sensitive and 100% specific in identifying the C282Y hemochromatosis patients with cirrhosis. Therefore, we concluded that a combination of serum hyaluronic acid and serum ferritin concentration is an accurate method for the diagnosis of cirrhosis and that this strategy is likely to be very cost-effective.
We found their own report very interesting because they describe three of four patients with cirrhosis having a serum ferritinconcentration less than 1000 μg/L.2 This is particularly unusual and in contrast to data obtained in the French, Canadian,3 and Australian populations.1 In our article, we highlighted the remarkable consistency between these three populations with respect to the frequency with which combinations of different laboratory variables occurred in patients with cirrhosis due to C282Y hemochromatosis. For example, the percentages of the cohorts with serum ferritin > 1000 μg/L and cirrhosis in Canadian, French, and Australian C282Y populations were 36%, 39.7%, and 38.5%, respectively. The three centers from which these data were reported are major referral institutions for hemochromatosis, and this could introduce referral bias into the observations. It would be of particular interest to determine whether similar results are reproduced in studies of patients reported from centers in which referral bias may not be an operative factor. Whether there are significant geographical and/or ethnic factors that alter the role of laboratory predictors of cirrhosis is unknown. We would encourage Dr. Castiella and his colleagues to extend their studies to determine if there are differences between populations with respect to the clinical utility of markers of fibrosis in C282Y hemochromatosis and particularly whether measurements of the serum hyaluronic acid concentration are also reliable for predicting cirrhosis in their population.