Assessment of iron status by erythrocyte ferritin in uremic patients with or without recombinant human erythropoietin therapy.

Erythrocyte ferritin may be a better estimator of iron bioavailability than the conventional markers of iron stores (serum ferritin and transferrin saturation). To investigate the accuracy of these conventional markers in uremic patients compared with erythrocyte ferritin, we studied 29 chronic hemodialysis patients on erythropoietin (EPO) therapy, 18 without EPO therapy, and 22 healthy control subjects. Apart from the red blood cell indices, serum ferritin, transferrin saturation, and erythrocyte ferritin, the analytical study included red blood cell protoporphyrin and plasma aluminum levels. The control group showed erythrocyte ferritin concentrations between 8.3 and 12.5 attograms/cell (95% confidence interval). In the EPO group, red blood cell protoporphyrin correlated negatively with erythrocyte ferritin, but not with serum ferritin or transferrin saturation. In the non-EPO group, serum ferritin, erythrocyte ferritin, and transferrin saturation did not correlate with red blood cell protoporphyrin. Even though erythrocyte ferritin correlated well with serum ferritin in the EPO group (r = 0.61, P = 0.0003), the sensitivity of normal serum ferritin levels (30 to 300 ng/mL) to discard a low erythrocyte ferritin concentration (erythrocyte ferritin less than 7 ag/cell) was 0.53, while the sensitivity of serum ferritin at levels less than 30 ng/mL to indicate an absolute iron deficiency expressed as a low erythrocyte ferritin concentration was 0.28. Only values of serum ferritin and transferrin saturation greater than 300 ng/mL and 35%, respectively, could rule out a relative iron deficiency expressed as a low erythrocyte ferritin and high red blood cell protoporphyrin concentration. Plasma aluminum levels did not correlate with red blood cell protoporphyrin or erythrocyte ferritin levels in either uremic group.(ABSTRACT TRUNCATED AT 250 WORDS)