Underestimates of serum free thyroxine (T4) concentrations by free T4 immunoassays.

Different free T4 (FT4) assays often give different FT4 measurements, and conflicting measurements have been striking in nonthyroidal illness. Because FT4 immunoassays depend upon serum protein-bound T4 (PBT4) dissociation to stabilize the FT4 concentration during assay perturbations, interassay differences in perturbations combined with variation in serum PBT4 concentrations could produce discordant FT4 measurements. This study examined the effects of PBT4 on FT4 measurements obtained by direct immunoassay methods. Standard solutions with constant FT4 levels and varying PBT4 concentrations were prepared and analyzed by direct equilibrium dialysis, two-step immunoextraction, one-step labeled T4 antibody, and one-step labeled T4 analog FT4 methods. Direct equilibrium dialysis results were independent of PBT4 concentrations and gave correct measurements of serum FT4 when the PBT4 concentration was above 8 nmol/L or 0.6 micrograms/dL, but were PBT4 dependent and underestimated serum FT4 at lower PBT4 concentrations. The other three methods were PBT4 dependent and variably underestimated serum FT4 at all levels of PBT4 up to 256 nmol/L (19.9 micrograms/dL), the highest level studied. Thus, PBT4-dependent underestimates of serum FT4 occurred with all four methods, whereas the measured FT4 level at each PBT4 concentration varied widely between methods. A serum PBT4 dependent bias causes discordant FT4 measurements and probably explains the observed underestimates of FT4 in nonthyroidal illness.