Evaluation of a direct assay for atrial natriuretic peptide.

A direct radioimmunoassay for the assessment of human atrial natriuretic peptide (ANP) in plasma, using a highly specific antibody and a well-defined monoiodotyrosyl-tracer was developed and evaluated by concurrent application of an extraction method. Sensitivity was 13.4 pg/ml; intra- and interassay variations were 3.1 and 5.5%, respectively; recovery averaged 99%; normal values ranged from 15-111 pg/ml (mean +/- SD = 59 +/- 25 pg/ml, n = 41). The results, including the effect of exercise, of the two methods correlated well. Pooling ANP values in normal subjects and patients with congestive heart failure gave a good correlation (p less than 0.01). However, due to processes in unextracted plasma, in the lower range the results from the direct method were erratic. Velocity and duration of centrifugation changed the number of platelets, but no effect on ANP levels, whether assessed by the direct or by the extraction method, was observed. Although the direct method is considerably less laborious than the extracted method its lack of reliability disqualifies it for most purposes.