Arising from: Kemna Erwin HJM et al. (2004) Nat Clin Pract Gastroenterol Hepatol 2: E1

In their article response, Kemna et al. stated that “...as far as we know, no sound methods and data have been published on the measurement of serum hepcidin concentrations.”1 In the past 2–3 years peer-reviewed journals have published articles in which the concentration of serum (pro)hepcidin, urine hepcidin and hepatic hepcidin mRNA have been assessed and correlated with iron status in groups of patients with various primary or secondary disorders of iron metabolism. Similar correlations between hepatic hepcidin expression and iron status have been evaluated in animal models. Table 1 in our review “Mechanisms of Disease: the role of hepcidin in iron homeostasis—implications for hemochromatosis and other disorders”2 attempts to summarize the main conclusions of all of this work, and is based on known or postulated pathogenic findings that seem to converge on the idea that serum levels of bioactive hepcidin are low in patients with anemia, high in patients with anemia of inflammation, inappropriately low in hereditary hemochromatosis patients, and so on. Table 1 was intentionally meant to describe only current pathophysiologic concepts and was not limited to clinical studies (see references in the table).

We agree that there is a general concern about the methods available to assay 'bioactive' serum hepcidin and that “...we think that we should all join forces to tackle and overcome the technical problems that prevent reliable measurements of hepcidin in serum and urine.” Nonetheless, we don't feel it is fair or correct to say that to date “no sound methods and data have been published on the measurement of serum hepcidin concentrations,” whatever this might mean, and it is certainly an overstatement to say that “...urine hepcidin measurements are currently used for clinical experiments.” Beyond any reasonable concern that is based on personal beliefs, we feel that the reliability of methods must be proved (or disproved) by studies designed specifically for this purpose, and that validation requires that the assay (whether it uses blood or urine) can be transferred to, and reproduced in, different laboratories.