LOW SERUM ALUMINUM VALUES IN DIALYSIS PATIENTS WITH INCREASED BONE ALUMINUM LEVELS

Background. Using an HPLC/ETAAS hybrid speciation technique we previou sly demonstrated iron to have a multifold effect on the binding of alu minum to transferrin by limiting the number of available binding sites and decreasing the affinity of transferrin for aluminum. Theoreticall y, at a 60% iron-transferrin saturation the aluminum-transferrin fract ion in serum should not exceed 30 mu g/l. In the present study previou s experimental data were confronted with recent clinical observations in patients with either normal iron status or iron overload. Patients and results. Serum aluminum levels and iron overload: In 38 dialysis p atients with a normal iron status and of whom 63% received Al(OH)(3) f or phosphate binding 26 (68%) had a serum aluminum level >30 mu g/l. O n the other hand out of 28 transfusional iron overloaded patients; 68% of them taking Al(OH)(3), only 1 subject (4%) had a serum aluminum va lue in excess of 30 mu g/l. Taking patients of both groups receiving A l(OH)(3) together a significant (p = 0.001) negative correlation (r = -0.5017) was found between the iron-transferrin saturation and the ser um aluminum levels. Iron status and parenteral aluminum loading: Also could a significant (p = 0.001) negative correlation (r = -0.6383) bet ween these parameters be found in an independent group of 44 patients which were acutely intoxicated by the use of aluminum-contaminated dia lysis fluids. Since in this population aluminum loading occurred paren terally and not via the gastrointestinal tract, a direct effect of iro n on the transferrin binding of aluminum rather than on the element's gastrointestinal absorption must have been responsible for the inverse relationship. Bone aluminum and iron overload: Out of 22 patients wit h a normal iron status (mean +/- SD serum ferritin: 216 +/- 245 mu g/l ; iron-transferrin saturation 20.4 +/- 9.6%), all of them having alumi num overload (bone aluminum level >15 mu g/g and/or positive Aluminon staining) none of them presented with a serum aluminum <30 mu g/l (mea n +/- SD: 82.2 +/- 51.6 mu g/l). On the other hand out of 13 iron over loaded patients (serum ferritin >800 mu g/l; iron-transferrin saturati on 61.4 +/- 17.6%) 10 (77%) presented the proposed criteria of aluminu m overload in the presence of a serum aluminum level <30 mu g/l. Concl usions. Our data indicate that in dialysis patients with iron overload (iron-transferrin saturation >60%; serum ferritin >800 mu g/l) serum aluminum levels are low (<30 mu g/l) despite exposure to aluminum by t he intake of Al(OH)(3) or the use of aluminum-contaminated dialysis fl uids. Low serum aluminum nevertheless may be associated with aluminum overload and even aluminum-related bone disease. An effect of iron on the serum aluminum speciation may at least in part explain our observa tions. Our findings allow a more accurate interpretation of baseline s erum aluminum values.