Accuracy and Standardisation of Serum Thyroxine Assays

Serum thyroxine was assayed concurrently by at least two of three different radio-displacement analysis kits—‘Tetralute’ (Ames), ‘Tetrasorb’ (Abbott) and ‘Thyopac-4’ (Radiochemical Centre). Repeated and careful cross checking by each of the methods was carried out on the ‘standard’ sera supplied with the kits and on other commercially available control sera. Differences of −3 to + 32% in apparent T-4 concentration were found. An unacceptable variability in T-4 content amongst batches of the same control sera was also evident which can be explained only by a lack of constancy in the commercial serum T-4 ‘standards’.

The molar extinction coefficients, water and tri-iodothyronine contents of commercial ‘pure’ thyroxine preparations indicate purities of 89–99 %. An examination of published ‘euthyroid’ ranges and those found by the ‘kit’ radiometric methods suggests that many normal serum T-4 values are based on the use of an impure primary or doubtful secondary standard. A primary standardisation of the assay with thyroxine of known purity and of known concentration, as determined from its molar extinction at 324 nm is advocated. Using this procedure, the normal euthyroid range for serum T-4 is 55–148 nmol T4/1 (2.8–7.5 μg T-4I/100 ml) or close to it—in agreement with the earlier findings of Howarth and Maclagan (1969) and Ekins et al. (1969).

During the period studied, ‘Tetralute’ produced the most accurate results.

Minimal effects of free fatty acids on the competitive protein-binding assay of serum thyroxine on reusable Sephadex columns

In rats; following oral administration of cream and intravenous heparin, plasma free fatty acids and triiodothyronine Sephadex uptake were elevated. Simultaneously apparent serum thyroxine, measured by the Tetrasorb-E kit, was increased, but not when measured using reusable Sephadex columns. Addition of oleate to a human serum albumin solution or to human serum produced similar effects; appreciable rises in apparent thyroxine were noted in the kit assay but only minor changes occurred when Sephadex columns were used. In the latter procedure almost all [14C]oleic acid added to serum was removed in the first barbital buffer wash, before application of the thyroxine-binding globulin solution. In the serum thyroxine assay on reusable Sephadex columns, there is minimal interference by free fatty acids because they are removed by the first barbital buffer wash.

Serum triiodothyronine determination in clinical use

Two radioimmunassays for the determination of serum triiodothyronine (T3) were developed. The assay of T3 in unextracted serum had several advantages over the assay on extracted serum and was chosen for the routine determination of T3 in serum from 117 patients requiring assessment of their thyroid status. In 53 subjects considered retrospectively not to have thyroid dysfunction nor to have been on steroid contraceptives or therapy, the pooled mean serum T3 concentration was 1-92 (actual range 0-88-2-62) nmol/l. A significant inverse relationship was observed between the serum T3 level and the age of the subject. Serum total T3 levels discriminate clearly between hypo-,eu- and hyperthyroid patients and provide a rather more sensitive index of hyperthyroid function than total serum T4. In the face of normal serum T4 the T3 level was depressed in five patients with marked hypoproteinaemia and elevated in two patients taking heroin.

Assay of serum thyroxine by competitive protein binding: reuse of disposable sephadex columns

A method for the assay of total serum thyroxine on reusable Sephadex G-25 fine columns, obtained from the Ames Tetralute kit, and competitive protein binding with TBG is described. The columns are regenerated with diluted plasma and may be used for at least 3 months. The procedure is rapid and simple and requires no extraction, centrifugation or evaporation. Recovery of added thyroxine from serum is essentially complete. The between-assay and within-assay coefficients of variation are about 5.5%. Correlation with values obtained using the Tetralute kit was good (r = 0.988). Values in 75 euthyroid subjects were normally distributed and therefore the normal range was taken as the mean +/- 2 SD or 4.6-10.4 mug/100 ml.

Stability of thyroxine and triiodothyronine in biological fluids

The stability of thyroxine and triiodothyronine in serum has been investigated. Apparent levels of total thryroxine, as determined by two different protein-binding assays employing thyroxine-binding globulin as the binding protein, increased significantly in serum and plasma samples stored at room temperature and were signficantly lower in haemolysed samples. Values did not change significantly in samples stored at 4 degrees C, nor in samples stored at room temperature when determined by radioimmunoassay. Total triiodothyronine levels, as determined by radioimmunoassay, fell slightly on storage. Failure to appreciate the effect of storing samples at room temperature on apparent levels of total thyroxine, as determined by some protein-binding assays, could lead to an incorrect assessment of thyroid status.

Iodine contamination in iodometric measurements of thyroid hormones

The incidence of iodine contamination in serum protein-bound iodine requests was assessed by comparison of results with those of a method using anion-exchange columns to improve specificity. The precision, accuracy, and specificity of the column method was found to be adequate, abnormal elution patterns generally warning where results were affected by contamination. The proportion of discrepancies in results and diagnoses between the two tests showed the protein-bound iodine to be unsatisfactory as a sole test of thyroid hormone levels, or in hypothyroidism, even as a screening test.

Adaptation of the Analmatic for Use in Determination of Serum Thyroxine

P.B.I. determinations are now being replaced in many laboratories by thyroxine determinations, but the latter have previously had the disadvantage of not being automated. The Analmatic preparation unit has been adapted for the large scale determinations of serum thyroxine using the Radiochemical Centre's Thyopac-4 kits. There is satisfactory correlation between the manual and automated techniques. Both the between-batch and within-batch precision were found to be better than that for the manual method.