Serum Free Thyroxine in Thyroidal and Nonthyroidal Illnesses: A Comparison of Measurements by Radioimmunoassay, Equilibrium Dialysis, and Free Thyroxine Index*

Abstract

The present study was undertaken to examine serum concentrations of free T4 in various functional states of the thyroid by a two-tube (A and B) solid phase RIA (Corning Immophase free T4 RIA) using two methods of calculation (I and II). Method I employed the ratio of radioactive T4 bound to anti-T4 antibody in tube A without thiomerosal to that in tube B containing thiomerosal as an inhibitor of serum binding of T4 or A/B as an index of free T4. Method B employed the factor (A/Tc × total T4, in which A represents radioactive T4 bound to anti-T4 in tube A and Tc is the total radioactivity) as an indicator of free T4. Both methods yielded clearly subnormal values in hypothyroidism and clearly supranormal values in hyperthyroidism. However, method I yielded low free T4 concentration values in euthyroid subjects with high T4-binding globulin (TBG) and high values in euthyroid subjects with low TBG. Therefore, it did not appear to be of value in routine clinical work. On the other hand, RIA (method B) was influenced minimally by alterations in TBG concentration; it gave clearly normal values in euthyroid subjects with high TBG and normal values or only borderline values in euthyroid subjects with low TBG. However, in clinically euthyroid subjects with nonthyroid illnesses (NTIs), the mean free T4 concentration estimated by RIA (method II) was subnormal (1.0 vs. 1.66 ng/dl; P < 0.01) whether serum total T4 was low (n = 11) or normal (n = 24; 1.37 us. 1.66 ng/dl; P < 0.01). The low mean free T4 RIA (method II) in NTI patients with low T4 levels was compatible with the low mean free T4 index but not with the high mean free T4 concentration estimated by the dialysis method. Similarly, a subnormal mean free T4 concentration, measured by RIA (method II) in NTI patients with normal T4 levels, was observed at a time when the mean free T4 index was clearly normal and the mean free T4, measured by the dialysis method, was clearly supranormal. When examined by the least mean square method, there was a good correlation between free T4 estimates, determined by RIA (method II), and the free T4 index (r = 0.89; n = 62; P < 0.001) and equilibrium dialysis (r = 0.84; n = 53; P < 0.001) in euthyroid subjects with or without abnormalities in TBG concentration and hypothyroid and hyperthyroid patients. However, in patients with NTI, estimates of free T4 by RIA (method II) correlated significantly only with the free T4 index (r = 0.54; n = 35; P < 0.001) and not with free T4 by the dialysis method (r = 0.32; n = 35; P > 0.05). When data from all groups of subjects were analyzed together, the correlation coefficient (r) between free T4 (measured by RIA method II) and the free T4 index (0.89; n = 97; P < 0.001) was higher than that between free T4 (measured by RIA method II) and free T4 by the dialysis method (0.49; n = 88; P < 0.01). The various data suggest that 1) free T4 measured by RIA (method I) is influenced by alterations in TBG concentration too severly to be useful for routine clinical work: 2) free T4 measured by RIA (method II) is a rapid, convenient, and reliable method for quantifying the free T4 concentration in various thyroidal states in the absence of NTI; 3) free T4 measured by RIA (method II) yields subnormal values in several clinically euthyroid patients with NTI; 4) in the presence of NTI, the results of free T4 measured by RIA (method II) correlate better with those of the free T4 index than free T4 by the equilibrium dialysis method; 5) the true concentration of free T4 in NTI continues to exhibit uncertainties because differing values (low or normal by RIA method II), low or normal by the index method, and normal or high by equilibrium dialysis) may be obtained in the same patients using different analytical techniques.