Digoxin-like immunoreactivity eliminated from serum by centrifugal ultrafiltration before fluorescence polarization immunoassay of digoxin.

Digoxin determined in the Abbott "TDx" by fluorescence polarization immunoassay by the manufacturer's recommended method involving precipitation of protein with 5-sulfosalicylic acid (SSA) is subject to interference from endogenous compounds having digoxin-like immunoreactivity. Guided by the work of Graves et al. (Clin Chem 1986;32:1506-9), we eliminated interference caused by digoxin-like immunoreactivity by substituting ultrafiltration for precipitation with SSA to remove protein. Using the manufacturer's method, we quantified digoxin in serum from 53 patients in three clinically defined groups who were receiving no digoxin, finding apparent digoxin in excess of the 200 ng/L detection limit in 24% of the 17 pregnant women, 59% of the 17 renal-dialysis patients, and all of 19 neonatal cord-blood samples examined. No measurable digoxin immunoreactivity was observed by fluorescence polarization immunoassay for any of the 53 clinically defined patients after removal of protein by ultrafiltration. For 22 men for whom digoxin was prescribed, digoxin measurement after protein removal by SSA and by ultrafiltration correlated well (r = 0.98), with good proportionality (slope = 1.04). Analytical recovery of added digoxin from adulterated serum was 115% after SSA, but 100% after ultrafiltration. Thus, before this assay procedure, we recommend ultrafiltration, to remove digoxin-like interference.