C. Bordat et al., IRON DISTRIBUTION IN THALASSEMIC BONE BY ENERGY-LOSS SPECTROSCOPY ANDELECTRON SPECTROSCOPIC IMAGING, Calcified tissue international, 53(1), 1993, pp. 29-37
Iron overload occurs frequently in thalassemia as a consequence of reg
ular blood transfusions, and iron has been found to accumulate in bone
, but skeletal toxicity of iron is not clearly established. In this st
udy, hone biopsies of thalassemic patients were investigated by light
(n = 6) and electron microscopy (n = 8) in order to analyze iron distr
ibution and possible iron-associated cellular lesions. Sections (5 mum
thick) were used for histomorphometry and iron histochemistry. Ultrat
hin sections were examined with an energy filtering transmission elect
ron microscope. Iron was identified by electron energy loss spectrosco
py (EELS), and iron distribution was visualized by electron spectrosco
pic imaging (ESI) associated with computer-assisted treatment (two-win
dow method). This study shows that EELS allows the detection of 4500-9
000 iron atoms, and that computer-assisted image processing is essenti
al to eliminate background and to obtain the net distribution of an el
ement by ESI. This study shows also that stainable iron was present al
ong trabecular surfaces, mineralizing surfaces, and on cement lines in
the biopsies of all patients. Moreover, iron was detected by EELS in
small granules (diffusely distributed or condensed in large clusters),
in osteoid tissue, and in the cytoplasm of bone cells, but not in the
mineralized matrix. The shape and size (9-13 nm) of these granules we
re similar to those reported for ferritin. As for iron toxicity, all p
atients had osteoid volume and thickness and osteoblast surface in the
normal range. Stainable iron surfaces did not correlate with osteobla
st surfaces, plasma ferritin concentrations, or the duration of transf
usion therapy. Numerous osteoblasts contained damaged mitochondria, an
d impaired osteoblast activity can therefore not be excluded.