Methodological Causes of Discrepancies in Radioimmunoassay for Digoxin in Human Serum
Abstract In our series of 73 patients, definite indications of toxicity were always associated with a serum digoxin concentration of 2.0 ng/ml or higher, while in the possibly intoxicated and nonintoxicated patients only 24% and 10%, respectively, had a concentration this high. The nonintoxicated patients included significantly more with subnormal serum albumin concentrations. In vitro addition of radiolabeled digoxin and antibody showed that sera from many patients with hypoalbuminemia will enhance the binding of the label, which could result in a digoxin radioimmunoassay value as much as threefold low. These results were unrelated to other serum constituents, including the associated hyperglobulinemia, or to methodologic problems in the immunoassay. The proportion of digoxin bound to serum proteins was not significantly different between the normal (38% ± 3%) and hypoalbuminemic (31% ± 3%) individuals, and so this could not explain the difference in the assay results. We conclude that patients whose serum albumin concentrations are low may show erroneously low radioimmunoassay values for serum digoxin, and that this may explain why some patients receiving high doses of digoxin are found to have unexpectedly low apparent concentrations of digoxin in their serum.