RESUMO
A recently developed enzyme-linked immunosorbent assay (EIAZ, ELISA) using two murine monoclonal anti-erythropoietin antibodies was compared with a radioimmunoassay (RIA) and a commercial in-vitro bioassay, EPOS, for measuring serum erythropoietin (Epo) in humans. Specificity and validity for Epo-EIA and the other two assays were examined. The serum Epo in normal subjects was 18 +/- 12 mU/ml (mean +/- SD, n = 80) for EIA compared with 22.5 +/- 18.5 mU/ml (n = 20) for RIA and 136 +/- 132 mU/ml (n = 14) for the bioassay. The serum Epo concentrations in normals and patients were highly comparable between EIA and RIA for Epo (P less than 0.01, r = 0.95). Epo concentrations by the EIA for normal female and male subjects were 20.5 +/- 13 and 16.5 +/- 10 mU/ml, respectively. Epo levels in patients with secondary polycythaemia or autoimmune haemolytic anaemia were significantly higher than normal subjects by the three methods. Epo levels in patients with chronic renal failure were within the normal range. By the EPOS bioassay, the Epo concentrations of normals and patients with renal failure were significantly higher than expected (136 +/- 132 and 447 +/- 273, respectively). Due to its inherent design, the EPOS bioassay possibly measures bone marrow proliferative activity in response to other serum growth regulators besides erythropoietin and was found to be unsuitable for clinical assessment of Epo. We concluded that the new EIA and RIA were similarly sensitive, reliable and accurate for measurement of serum Epo. The EIA method has the advantage of being less time consuming, more convenient and avoids the use of a radioisotope.