Effect of antibody specificity on results of selected digoxin immunoassays among various clinical groups

We examined the specificity of three automated digoxin immunoassays (Abbott TDxFLx Digoxin II assay, Baxter-Dade Stratus II Digoxin assay, and Ciba Corning ACS Digoxin assay) applied without modification to (a) sera from 229 digoxin-free patients in 12 cohorts associated with nonspecific or endogenous digoxin-like immunoreactive factor (DLIF) interference, and (b) drug-free serum supplemented with the major metabolites and analogs of digoxin. We observed three patterns of apparent digoxin results among the DLIF samples: one common to kidney and liver failure patients, where TDx and Stratus assays showed significant positive results; one common to newborns and cord blood, where only the TDx assay had significant interference; and one from cardiac surgery patients, where the Stratus assay alone showed interference. Of the three assays, the ACS had the least interference from DLIF. The assays also behaved differently with respect to cross-reactivity with digoxin metabolites, digitoxin, and digitoxin metabolites. The ACS assay again had the least analog or metabolite cross-reactivity. The three methods agreed well on digoxin-positive specimens, with a mean bias of <0.15 microgram/L digoxin for each and discrepancies (defined as >3 SD between the assay pairs compared) of only 3-5%.