摘要：

results eliminates the principal source of error which has given rise to problems encountered with the Corning kit, and the manufacturers have now incorporated this change into their recommended protocol.4 Nevertheless other, albeit relatively minor, TBG-dependent effects (for example, the "non-specific" binding of radioactive T4 to the solid-supported antibody used in this kit) may influence free T4 determinations made on the basis of the recommended Corning assay procedure, and thereby distort free T4 values in subjects in whom thyroxine-binding protein concentrations are significantly altered. The physiological significance of small changes observed in pregnant subjects in serum free T4 (as measured using the Corning technique) may in consequence be questionable, although the use of this kit in a diagnostic role may not be significantly compromised by any minor residual sources of error that it may contain. Meanwhile a reduction in free T4 concentration during pregnancy has also been reported by a number of other workers (for example, Avruskin et al} and Arango et al') using a variety of conventional free T4 assay techniques (albeit the magnitude of the effect has been less than that reported by Boss et al). Many of these techniques have likewise proved to be methodologically suspect, yielding apparent frce T4 values significantly higher than those observed with modern "direct" methods (see, for example, Ekins7). Were the observation of a reduced free thyroid hormone concentration in pregnancy to be confirmed, however, speculation must arise regarding the physiological role of the free hormone moiety and the nature of mechanisms governing thyroid hormone transport and control. In attempting to establish the validity of the reported changes in pregnancy, we have used a new "direct" method for free T4 measurement developed in this laboratory. In brief, the method involves the incubation, at 37 C for one hour, of serum with T4 antiserum linked to a particulate solid support (Sephadex) followed by separation of the antibody from the serum; subsequently the amount of T4 bound to antibody is estimated by "back titration" with '251-labelled T4. Serum standards for the system have been calibrated by an equilibrium dialysis technique relying on direct radioimmunoassay of T4 in the dialysate.8 (This method represents an entirely novel approach to frec hormone measurement, the theoretical and methodological basis of which will be presented in detail clsewhere.) Serum samples were collccted from normal women attending a gynaccological clinic for contraceptive advice (none were on oral contraceptivcs) and also from pregnant patients attending the antenatal clinic. To minimise any possible methodological bias normal and pregnancy sera were placed as alternate samples within each assay batch. The normal samples yielded a mean free 14 concentration of 15 12 pmol/l (1 17 ng/100 ml) (n 100, SE of mean 0 32 (0 02)) with a range of 9 8-22 4 pmol/l (0 76-1-74 ng/100 ml). For the pregnancy samples the mean concentration was reduced to 13-6 pmolIl (1 06 ng/100 ml) (n 123, SE of mean 0-26 (0-02); t 3 68). The results of the 101 patients whose stage of pregnancy was known are shown in the figure. The free T4 concentration appeared lower in the second half of pregnancy, with a mean value of 12 2 pmol/l (0-95 ng/100 ml) (n45, SE of mean 0 33 (0-02): in the first half of pregnancy the results were near normal (mean concentration 14 4 pmol/l (1-12 ngj 100 ml) (n 56, SE of mean 0-4 (0 03)).

展开