The relationship between serum ferritin and tissue iron was investigat
ed in 26 dialysis patients (17 hemodialysis patients, 9 chronic perito
neal dialysis patients) with anemia (median hemoglobin 74 g/l, range 5
6-92 g/l). Serum ferritin ranged from 18 to 9,435 mu g/l (median 450 m
u g/l). Tissue iron was assessed in the liver biopsies of 4 hemodialys
is patients with iron overload (serum ferritin 1,150-9,435 mu g/l), in
the muscle biopsies of 5 patients with serum ferritin 170-9,435 mu g/
l, and in bone marrow aspirations (semiquantitative assessment). The m
ean liver iron concentration was 15.4+/-8.0 mu g Fe/mg protein (mean /- SD), which is similar to that previously found in patients with unt
reated idiopathic hemochromatosis. Four patients with serum ferritin 1
70-620 mu g/l had muscle iron concentrations (0.33 +/- 0.10 mu g Fe/mg
protein) similar to those found in controls (0.23 +/- 0.10, means +/-
SD). One patient with serum ferritin 9,435 mu g/l had a markedly incr
eased muscle iron concentration (1.3 mu g Fe/mg protein). The bone mar
row iron was assessed as negative in 3 patients (serum ferritin 44-85
mu g/l), positive in 8 (serum ferritin 18-379 mu g/l), increased in 11
patients (serum ferritin 222-4,210 mu g/l). and was markedly increase
d in 2 patients (serum ferritin 4,550 and 9,435 mu g/l). Bone marrow i
ron correlated significantly with serum ferritin concentrations (spear
man rank correlation coefficient rho = 0.89, p < 0.001). These results
show that in dialysis patients with a stable iron balance and unstimu
lated erythropoiesis, i.e., patients without erythropoietin treatment
and parenteral iron, serum ferritin is a useful indicator of iron stor
es. Our findings also suggest that the relationship between tissue iro
n and serum ferritin levels in end-stage renal disease is altered, i.e
., a relative increase in serum ferritin levels unrelated to iron stor
es is observed in dialysis patients.