Comprehensive study of a thyroxin-analog-based assay for free thyroxin ("Amerlex FT4").

The basic theory of thyroxin-analog-based radioimmunoassays for free thyroxin has been extended to evaluate definitively the effects arising from residual binding of the tracer analog to serum proteins. Using experimentally determined binding constants and computer simulation techniques, we studied the effects of thyroxin-analog binding to serum proteins on results of Amerlex FT4 radioimmunoassay, using this improved mathematical model. Results from computer-simulation studies were compared both directly with in vitro experimental results and indirectly with clinical studies. Agreement was good among all three approaches. The relatively weak binding of analog to thyroxin-binding globulin and prealbumin does not significantly perturb Amerlex FT4 assay results. Binding of the analog by albumin has a small but quantifiable effect on assay results, amounting to an intrinsic bias of 0.08 pmol of free thyroxin per liter per gram of albumin per liter for euthyroid serum samples. This bias is unlikely to be important for most clinical laboratory samples, but it may be significant when one is interpreting results for those rare patients with genetic albumin abnormalities such as analbuminemia or familial dysalbuminemic hyperthyroxinemia. Massively increased concentrations of nonesterified fatty acids (e.g., after treatment with heparin) will lead to a spurious increase in free thyroxin in this and most other techniques, including equilibrium dialysis.