Is Free Thyroxine Accurately Measurable at Room Temperature?

Two recently developed two-step, or "back-titration" assay kits for free thyroxine (FT4)--one based on a europium-labeled derivative of T4, the other on conventional radiolabeled T4--were compared with both symmetrical dialysis and an indirect FT4 radioimmunoassay. Discrepancies between the two-step and the other two methods were observed, particularly in samples that had very low to zero thyroxine-binding globulin (TBG) content. In those instances the two-step methods gave values almost twice as high as the other methods. This effect could be largely reversed in one of the two-step assays and completely reversed in the other by performing the first incubation step at 37 degrees C rather than at room temperature, as prescribed by the manufacturers. When symmetrical dialysis is performed at both temperatures, FT4 at room temperature is about 40% of the amount determined at 37 degrees C, except in zero-TBG samples, where it averages almost 80% of the value at 37 degrees C. Moreover, we demonstrated that the affinity of TBG for T4 is much more temperature-dependent than the affinity of transthyretin and albumin for T4, so that the net temperature effect on the FT4 in a sample depends on the relative contribution of TBG to total binding. We conclude that performing FT4 assays at room temperature is principally incorrect and leads to falsely increased values when samples with very low TBG concentrations are analyzed.