Background: The monitoring of phlebotomies in hemochromatosis patients depends on iron status measured by ferritin and transferrin saturation (TS). However, in the presence of inflammation or liver injury, soluble transferrin receptor (sTfR) determination was proposed to replace ferritin for diagnosing iron deficiency (ID). The present study evaluated performances of sTfR for the prediction of iron deficiency in a large number of hemochromatosis patients under phlebotomy therapy. Methods: We studied 52 patients undergoing therapeutic phlebotomies and obtained 2 samples from 37 patients. Biological parameters were determined before each phlebotomy began. Performances of sTfR and TS in the diagnosis of iron deficiency were compared, according to ferritin levels under 12 mug/l. Results: Ferritin and TS were correlated with removed iron (r=0.473, p<0.005 and r=0.345, p<0.05, respectively) and sTfR was correlated with the decrease in hemoglobin levels induced by phlebotomies (r=0.678, p<0.0001). Areas under Receiver Operating Characteristics (ROC) curves for sTfR and TS were not statistically different for prediction of iron deficiency and sensitivity/specificity of sTfR at 1.64 mg/l were 67/86%. Conclusions: sTfR determination could be used to predict iron depletion induced by phlebotomies when ferritin is of limited interest, to avoid the appearance of anemia. (C) 2004 Elsevier B.V. All rights reserved.