Printer Friendly

Assessment of serum thyroxine binding capacity-dependent biases in free thyroxine assays.

Irrespective of the thyroid function test strategy used, a measure of thyroid hormones is required to confirm the diagnosis of thyroid disease. The most appropriate measure of thyroid hormone is the moiety of thyroxine ([T.sub.4])[4] which is not bound by the thyroid hormone-binding proteins, and it is this free form of [T.sub.4] ([FT.sub.4]) that is best correlated with thyroid status (1-3). Because the proportion of total [T.sub.4] found in the free form is extremely small (~0.02% in euthyroid subjects), its quantification in the presence of the relatively very high concentrations of [T.sub.4] bound to the binding proteins has proved exceedingly difficult. Indeed, an extensive literature exists highlighting interferences experienced with many commercially available [FT.sub.4] assays (4-10). More recently, [FT.sub.4] assays have become automated, which makes their use even more attractive to laboratories and has fueled the move from total to free thyroid hormone measurement. Although the newer [FT.sub.4] assays are substantially better than the original assays, few seem to be as analytically accurate as the direct equilibrium dialysis (ED) [FT.sub.4] method (9). Several studies have highlighted weaknesses in assay designs that produce significant biases (7,9,11,12), the magnitude of which are assay specific and related to the concentration of the protein-bound [T.sub.4] ([PBT.sub.4]). In particular, Nelson et al. (9) have shown that as the concentration of [PBT.sub.4] decreased, the magnitude of the negative bias seen was increased. Careful examination of the experimental design adopted and data presented by Nelson et al. (9) suggested that the method-specific biases observed were dependent not only on the concentration of [PBT.sub.4], but also on the serum binding capacity (sBC; calculated as the concentration x the affinity of free binding proteins). In the present study, we examine the ability of two simple experimental tests in identifying the presence and magnitude of sBC-dependent biases in an automated [FT.sub.4] assay (Vitros [FT.sub.4]). Both of these tests challenge the [FT.sub.4] assays in the clinical situations outlined recently by the National Academy of Clinical Biochemistry (13), i.e., determination of the assay-specific biases in sera having a broad spectrum of binding protein abnormalities. For the first test, we used sera from patients having a wide sBC range. To ensure that the widest possible range of sBC was challenged, the test panel included sera from pregnant women in their third trimester, ambulatory subjects, and severely ill hospitalized patients because these groups previously have been shown to have high, normal, and low sBC, respectively. For the second test, variation of sBC was achieved by the serial dilution of third trimester sera in an inert buffer.

Materials and Methods

The patient panel examined in the assays described below consisted of 26 ambulatory subjects, 18 women at the third trimester of pregnancy, and 25 severely ill patients who were admitted to an intensive care unit with a variety of illnesses, including sepsis, cardiac arrest, cardiac failure, and respiratory failure. All procedures used were in accordance with the Helsinki Declaration of 1975 (as revised in 1996). The patient sera were collected for routine analyses and were kept frozen (-20[degrees]C) until required for the hormone measurements performed as part of the present study. All assays (with the exception of the ED [FT.sub.4], which was performed in the Cardiff laboratories of Ortho-Clinical Diagnostics) were carried out at the Clinical Biochemistry Laboratory of the Royal Infirmary, Edinburgh.

The sBC and [PBT.sub.4] concentrations of three sera were reduced in vitro by serially diluting (2- to 64-fold dilutions) sera in 10 mmol/L HEPES (Sigma Chemical; cat. no. H3375) buffer solution, pH 7.4. The sera chosen were from third-trimester pregnancies because these represented sera with high sBC and [PBT.sub.4] concentrations. The undiluted and diluted samples were assayed in the two [FT.sub.4] methods using standard protocols.

The automated [FT.sub.4] method used was the Vitros Immunodiagnostic Products (Ortho-Clinical Diagnostics, Amersham UK). The manual direct ED method was the Nichols [FT.sub.4] method (Nichols Institute Diagnostics). The physicochemical principles of the assays used are fundamentally different. The Vitros [FT.sub.4] assay is a labeled antibody method (2,14), and the Nichols assay is a direct equilibrium method (15).

The ED method is carried out by dialyzing the [FT.sub.4] from 200[micro]L of serum into 2.4 mL of dialysis buffer (at 37[degrees]C, over a 16- to 18-h incubation period). The [FT.sub.4] in the protein-free dialysate is then quantified by a sensitive solid-phase RIA for [T.sub.4]. To minimize variability (e.g., assay-to-assay variation), all samples studied were analyzed on one occasion, with the samples randomized. The package insert quotes an intraassay CV at doses falling within and above the euthyroid range as being <13%. The euthyroid [FT.sub.4] range, which was the observed range with one outlier deleted from each end, quoted in the package insert is 10.3-34.7 pmol/L.

In the Vitros [FT.sub.4] assay, 25[micro]L of sample is pipetted into microwells coated with a triiodothyronine (Ts)-protein conjugate, followed by 100[micro]L of a sheep anti-[T.sub.4] antibody labeled with horse-radish peroxidase in a 150 mmol/L phosphate buffer containing 1.0 g/L bovine gelatin and 1.0 g/L bovine [gamma]-globulin. During the 16-min incubation at 37[degrees]C, a proportion of the labeled antibody binds to the serum [FT.sub.4] and to the well surface, with the amount binding to the well surface being inversely related to the serum [FT.sub.4] concentration. The well is then washed, and a signal reagent that produces luminescence is then added; the resulting light emitted is measured in a luminometer. All procedures are carried out automatically by the Vitros ECi Immunodiagnostic system. The samples were assayed in a batch mode, with quality-control samples at three concentrations run at the beginning and end of the assay. The within-run imprecision for [FT.sub.4] values in the euthyroid and hyperthyroid range is quoted in the package insert as being <3%. The assay was calibrated using an in-house ED assay. The euthyroid range (1 and 99 percentiles) quoted in the package insert is 10-28.2 pmol/L.

In addition to the [FT.sub.4] assays, the patient sera were also analyzed in the Vitros Immunodiagnostic Products Total [T.sub.4] (TT4) and [T.sub.3] uptake (T3U) assays. The euthyroid [T.sub.4] range (2.5 and 97.5 percentiles) quoted in the package insert is 71.2-141 nmol/L. The T3U assay is calibrated in %T3U units, which are inversely related to the serum binding capacity (i.e., a serum with a high %T3U has a low binding capacity). The euthyroid %T3U range (2.5 and 97.5 percentiles) quoted in the package insert is 23.5-40.6% uptake.

All assays were performed following the manufacturers' instructions.


Analysis of data was carried out by standard methods with Microsoft Excel spreadsheets. The dependence of the [FT.sub.4] bias observed with any variable was determined by correlating (linear regression analysis performed by the least-squares method) the methodological difference (from ED [FT.sub.4]) of each patient (y-axis) against the variable studied (x-axis). The sBC was calculated by dividing the Vitros TT4 concentration by the ED [FT.sub.4] concentration (9). The concentration of [PBT.sub.4] was estimated by subtracting the concentration of [FT.sub.4], as measured by ED, from the concentration of TT4 (9). Because the [FT.sub.4] concentration relative to TT4 was small (<0.1% of the TT4), the [PBT.sub.4] concentration showed little difference from the TT4 concentration. The agreement between the Vitros [FT.sub.4] and the manual ED [FT.sub.4] was tested by Deming regression (16). A measure of the sBC was also obtained by the T3U, where the %T3U is inversely related to the binding capacity. The Student unpaired Mest was used to compare the biochemical profiles between the patient groups (within each assay format), and the paired Mest was used when the comparison was across different assay methods.


The quality-control sera used in each assay were all within expected ranges, and there was no evidence of drift in performance in any of the assays examined.

The relationship between the Vitros [FT.sub.4] and ED [FT.sub.4] assays in all patients studied is shown in Fig. 1. A significant correlation (r = 0.96; P <0.001) between the two assays was observed. The equation describing the relationship was:


Vitros [FT.sub.4] = 0.85 ([+ or -] 0.03; P <0.001) ED [FT.sub.4] + 3.67 ([+ or -] 0.61; P <0.001).

Good agreement between the [FT.sub.4] methods was achieved in all patient groups.

The relationship between the Vitros [FT.sub.4] to ED [FT.sub.4] in (a) the ambulatory and pregnant subjects (groups combined, n = 44), and (b) the hospitalized patients (n = 25) is described by the following equations:

(a) Vitros [FT.sub.4] = 0.9 ([+ or -] 0.09) ED [FT.sub.4] + 3.09 ([+ or -] 1.19); r = 0.83;P <0.001

(b) Vitros [FT.sub.4] = 0.89 ([+ or -] 0.07) ED [FT.sub.4] + 2.33 ([+ or -] 1.9); r = 0.94; P <0.001

Table 1 shows the mean ([+ or -] SD) and observed ranges for the two [FT.sub.4] methods under investigation ([FT.sub.4] in pmol/L), the TT4 (in nmol/L), %T3U, and sBC (in nmol/ pmol) in the ambulatory, pregnant, and hospitalized patients. The [FT.sub.4] concentration, as measured by both [FT.sub.4] methods, in the hospitalized group was significantly (P <0.001) higher than the [FT.sub.4] concentration in the ambulatory group. None of the patients in the hospitalized group had [FT.sub.4] concentrations below the corresponding euthyroid range of the two [FT.sub.4] assays examined. The calculated sBC in the hospitalized group was significantly lower (P <0.001) than that found in the ambulatory group, whereas the corresponding %T3U value was significantly higher (P <0.001). Thus, both measures of serum [T.sub.4] binding support the view that hospitalized patients have decreased serum [T.sub.4] binding capacities. The TT4 concentration in the hospitalized group was significantly lower (P <0.001) than that obtained in the ambulatory group, and 12 of the 25 hospitalized patients had TT4 concentrations below the euthyroid range. The [FT.sub.4] concentration in the pregnancy group was significantly lower (P <0.001) than in the ambulatory group with both [FT.sub.4] methods, whereas the sBC and TT4 concentrations were increased (P <0.001). As expected, the %T3U was significantly (P <0.001) reduced.

The [FT.sub.4] concentrations obtained in the pregnancy group by the Vitros method were higher than the corresponding concentration obtained with the ED method. These differences, although small (mean difference, 2.6 pmol/L; range of difference in 95% confidence intervals, 2.1-3.2 pmol/L), were statistically significant (P <0.001). Similar differences from ED [FT.sub.4] were seen in the ambulatory group (mean difference, 1.6 pmol/L; range of differences in 95% confidence intervals, 0.7 to 2.6 pmol/L; both P <0.001). The Vitros [FT.sub.4] concentration in the hospitalized group was not significantly different (P >0.05) from the mean concentration obtained in the ED method.

Fig. 2 depicts the relationship between the Vitros [FT.sub.4] bias and sBC in all the patient samples. The regression equations derived in each individual category of subjects are shown below:

Ambulatory group: Vitros [FT.sub.4] bias = 0.77 ([+ or -] 0.43) sBC - 3.01 ([+ or -] 2.62); r = 0.34; P >0.05

Pregnant group: Vitros [FT.sub.4] bias = 0.167 ([+ or -] 0.077) sBC - 0.079 ([+ or -] 0.167); r = 0.475; P <0.05

Hospitalized group: Vitros [FT.sub.4] bias = -0.31 ([+ or -] 0.665) sBC + 0.29 ([+ or -] 0.182); r = 0.098; P >0.05

The slopes and intercepts of all the relationships did not differ significantly from each other (P >0.05). The regression equation describing the relationship of the Vitros [FT.sub.4] bias and sBC (the only individual patient category that yielded a significant slope and r value, both at P <0.05, was the pregnancy group) indicates that at the highest observed sBC (23.9 nmol/pmol), the Vitros [FT.sub.4] will be positively biased by 4 pmol/L (95% confidence interval, 2.6-5.4 pmol/L).


There was no apparent relationship (P >0.05) between the [PBT.sub.4] concentrations and the Vitros [FT.sub.4] concentrations in any of the patient groups studies.

Fig. 3 depicts the relationship between the Vitros [FT.sub.4] bias and %T3U in all patients studied. The equations derived in each of the individual patient categories are summarized below:

Ambulatory group: Vitros [FT.sub.4] bias = -0.04 ([+ or -] 0.22) %T3U + 2.75 ([+ or -] 6.46); r = 0.035; P >0.05

Pregnant group: Vitros [FT.sub.4] bias = 0.003 ([+ or -] 0.199) %T3U + 2.61 ([+ or -] 3.78); r = 0.003; P >0.05



Hospitalized group: Vitros [FT.sub.4] bias = 0.06 ([+ or -] 0.066) %T3U - 3.37([+ or -] 3.18); r = 0.19; P > 0.05

None of these relationships reached statistical significance.

The results of the serum dilution experiment performed on the pregnancy sera are shown in Fig. 4. The theoretically derived [FT.sub.4] concentrations suggest that the decrease after a 32-fold dilution will be <1%. Similarly, the [FT.sub.4] concentration as measured by ED was only slightly affected by serum dilution. At dilutions of 1:16 and 1:32 (i.e., dilutions that are expected to decrease the sBC to levels usually seen in some hospitalized patients) the ED [FT.sub.4] was within 5% of the concentration obtained in the undiluted sample. The Vitros [FT.sub.4] concentrations at these dilutions did not vary significantly (<5%) from the [FT.sub.4] concentration obtained in the undiluted sample.


It is now accepted that [FT.sub.4] shows a better correlation to the thyroid status than total [T.sub.4] (1-3); thus it is the free hormone fraction that should be used to confirm the diagnosis of thyroid disease (17-19). The direct measurement of [FT.sub.4] has, however, proved to be exceedingly difficult. Over the last two decades, numerous methodologies have been developed, and their limitations have been the subject of many publications and heated discussions in the literature (3, 7,11,20-28). Several limitations (e.g., "albumin" effects) of the early [FT.sub.4] assays have been eliminated or minimized, but as shown recently, large methodological differences are still present (9). In their study, Nelson et al (9) measured [FT.sub.4] (using different [FT.sub.4] RIAs) in serum preparations in which the concentration of [PBT.sub.4] (and sBC) was altered, whereas the ED [FT.sub.4] concentration was kept constant. The results showed that the three [FT.sub.4] methods studied were significantly, but to varying degrees, influenced by the serum [PBT.sub.4]. This method-specific dependency on [PBT.sub.4] was proposed as a possible explanation for the discordant [FT.sub.4] measurements seen in nonthyroidal illness. In an additional study, this group of investigators (10) suggested that the method-specific [PBT.sub.4] dependency was a direct result of increased sequestration of [T.sub.4] by the assay reagents.

We have used two simple tests to examine the bias of an automated (Vitros) [FT.sub.4] method. The first test involved the comparison of the [FT.sub.4] results obtained in the Vitros method in patients having a wide range of binding capacities against those obtained in a commercial direct ED system. The main reason for choosing ED as the reference method is that it is one of the methods, in addition to ultrafiltration, generally considered as the "gold standard" method for free thyroid hormone measurement. However, even these gold standards have their own inconsistencies and technical weaknesses (15,29-32). Thus, the elevation of this particular ED method as the definitive gold standard method is arguable. Nonetheless, the validity of the ED method chosen has been well documented (15), although even this ED may be biased in some nonthyroidal illness sera because of dilution effects during the dialysis step (33). The ED method we used produces a 13-fold dilution of dialyzable substances in sera. The patient categories included in the study have been documented to have high (i.e., third-trimester pregnancy) and low (i.e., hospitalized patients) binding capacities (13). This binding capacity profile has been confirmed in the present study by two independent methods (the sBC, derived by dividing the TT4 result by the ED [FT.sub.4] result, and the %T3U). The second test used examined the effect of serum dilution on the [FT.sub.4] concentration obtained by the methods under investigation (Vitros and ED [FT.sub.4]). The sample dilution test was chosen for several reasons: (a) one can alter the sBC in a predictable fashion (e.g., a twofold dilution will reduce the sBC by ~50%); (b) one can readily predict the ideal performance of an assay and thus eliminate the need to compare results with those obtained by a reference method; (c) one can "manufacture" serum dilution pools whose sBCs mimic the range of sBCs found in patients [e.g., the sBC concentration range in patients undergoing thyroid testing has been shown to be ~30-fold (13); thus dilution of a pregnant serum by this factor will encompass the sBC likely to be experienced by a laboratory]; and (d) this test can easily be performed by any user.

The law of mass action (2, 3) dictates that the concentration of [FT.sub.4] in serum depends on the equilibrium that exists between the [PBT.sub.4] and the concentration and affinity (i.e., the sBC) of the free binding sites:

[FT.sub.4] = [PBT.sub.4]/sBC

In estimating the sBC (i.e., [PBT.sub.4]/[FT.sub.4]) we made the same assumption as Nelson and co-workers (9,13), which is that the [FT.sub.4] as measured by ED gives an unbiased estimate of the true [FT.sub.4]. As discussed previously, this assumption is arguable. The highest and lowest individual sBC values in our study population were 23.9 and 1.2 nmol/pmol, which represent a range of 19.9-fold. Thus, it is clear from this and other studies (9,13) that patients undergoing thyroid function tests possess a very wide range of serum [T.sub.4] binding (and consequently wide ranges of [PBT.sub.4]). The results presented here show that the [FT.sub.4] concentrations obtained by the Vitros assay (relative to ED), were not dependent on [T.sub.4] binding (sBC or T3U) or [PBT.sub.4]. The small positive bias (from ED) seen in pregnancy is likely because of calibration differences. This is supported by the fact that the corresponding [FT.sub.4] concentrations in the ambulatory group were also positively biased to a similar degree.

In the second test, we examined the [FT.sub.4] biases of the two methods (Vitros and ED [FT.sub.4]) by analyzing sera whose protein concentrations were decreased in vitro. Lowering of the sBC by decreasing the concentration of the binding proteins was accomplished by diluting sera in an inert buffer (HEPES). The [FT.sub.4] concentrations as measured by both the ED and Vitros methods were not reduced by the serum dilution.

The estimation of serum [FT.sub.4] by immunoassay, irrespective of the methodology used will invariably disturb the normal equilibrium between [PBT.sub.4] and sBC (2, 3). This will come about as a result of the addition of the antibody and the dilution of the sample in assay reagents that may also contain [T.sub.4] binders such as albumin or animal sera, which are often added to protect assays from interferences from heterophilic antibodies. These additions will lead to the establishment of a new equilibrium and a new "in vitro" [FT.sub.4] concentration. The concentration of this in vitro [FT.sub.4] will be dictated by the in vitro bound [T.sub.4] and unbound binding site concentrations. The in vitro bound [T.sub.4] will be the net sum of [PBT.sub.4] and [T.sub.4] bound by the immunoassay reagents (i[PBT.sub.4]), and the in vitro unbound binding sites will include, in addition to sBC, the free antibody-binding sites and the free binding sites of other binders included in the reagents [i.e., immunoassay binding capacity (iBC), which will be equal to the affinity x concentration of the free immunoassay binding sites]. Thus, the in vitro [FT.sub.4] will be equal to ([PBT.sub.4] + [iPBT.sub.4])/ (sBC + iBC). From this formula, one can predict that a high iBC will cause biases that will be dependent on the magnitude of the endogenous sBC. Thus, as the sBC decreases (as seen in hospitalized patients and mimicked by serum dilution), the assay will yield increasingly negative results. The results presented show a lack of sBC-dependent biases in both the Vitros and the ED [FT.sub.4] assays, suggesting that the in vitro disturbance of the [T.sub.4]/protein equilibrium induced by both assays is negligible.

The tests described in the present study can be used to assess the presence and magnitude of sBC-dependent biases of other commercial [FT.sub.4] methods.


(1.) Robbins J, Rail JE. The interaction of thyroid hormones and protein in biological fluids. Recent Prog Horm Res 1957;13:161-208.

(2.) Ekins R. Measurement of free hormones in blood. Endocrinol Rev 1990;11:5-46.

(3.) Ekins R. The free hormone hypothesis and measurement of free hormones. Clin Chem 1992;38:1289-93.

(4.) Gow SM, Kellett HA, Toft AD, Beckett GJ. Accuracy and precision of five analog radioimmunoassays for free thyroxin compared. Clin Chem 1985;31:1888-92.

(5.) Csako G, Zweig MH, Benson C, Ruddel M. On the albumin dependence of measurements of thyroxin. I. Technical performance of seven methods. Clin Chem 1986;32:108-15.

(6.) Wilke TJ. Five kits for estimating free thyroxine concentration in serum evaluated, and correlated with other indices to thyroid status. Clin Chem 1982;28:2051-6.

(7.) Helenius T, Liewendahl K. Improved dialysis method for free thyroxin in serum compared with five commercial radioimmunoassays in nonthyroidal illness and subjects with abnormal concentrations of thyroxin-binding globulin. Clin Chem 1983;29:816-22.

(8.) Perdrisot R, Bounaud MP, Bounaud JY, Jallet P. Four nonisotopic immunoassays of free thyroxin evaluated. Clin Chem 1989;35: 115-20.

(9.) Nelson JC, Weiss RM, Wilcox RB. Underestimates of serum free thyroxine ([T.sub.4]) concentrations by free [T.sub.4] immunoassays. J Clin Endocrinol Metab 1994;79:76-9.

(10.) Nelson JC, Nayak SS, Wilcox RB. Variable underestimates by serum free thyroxine ([T.sub.4]) immunoassays of free [T.sub.4] concentrations in simple solutions. J Clin Endocrinol Metab 1994;79:1373-5.

(11.) Czako G, Zweig MH, Benson C, Ruddel M. On the album-independence of measurements of free thyroxin. II. Patients with nonthyroidal illnesses. Clin Chem 1987;33:87-92.

(12.) Wong TK, Pekary AE, Hoo GS, Bradley ME, Hershman JM. Comparison of methods for measuring free thyroxine in nonthyroidal illness. Clin Chem 1992;38:720-4.

(13.) Nelson JC, Wilcox RB. Analytical performance of free and total thyroxine assays. Clin Chem 1996;42:146-54.

(14.) Christofides ND, Sheehan CP. Enhanced chemiluminescence labeled-antibody immunoassay (Amerlite MAB) for free thyroxine: design, development, and technical validation. Clin Chem 1995; 41:17-23.

(15.) Nelson JC, Tomei RT. Direct determination of free thyroxine in undiluted serum by equilibrium dialysis/radioimmunoassay. Clin Chem 1988;34:1737-44.

(16.) Combleet PJ, Gochman N. Incorrect least squares regression coefficients in method-comparison analysis. Clin Chem 1979;25: 432-8.

(17.) Stockigt JR. Guidelines for diagnosis and monitoring of thyroid disease: nonthyroidal illness. Clin Chem 1996;42:188-92.

(18.) Becker DV, Bigos ST, Gaitan E, Morris JC, Rallison ML, Spencer CA, et al. Optimal use of blood tests for assessment of thyroid function [Letter]. DAMA 1993;269:2736.

(19.) Braverman LE. Evaluation of thyroid status in patients with thyrotoxicosis. Clin Chem 1996;42:174-8.

(20.) Bayer MF. Free thyroxin results are affected by albumin concentration and nonthyroidal illness. Clin Chim Acta 1983;130:391-6.

(21.) Midgley JEM, Moon CR, Wilkins TA. Validity of analog free thyroxin immunoassays. Part II [Opinion]. Clin Chem 1987;33:2145-52.

(22.) Ekins RP. Validity of analog free thyroxin immunoassays [Opinion]. Clin Chem 1987;33:2137-52.

(23.) Chan DW. Nonthyroidal illness: performance of free [T.sub.4] methods [Review]. J Clin Immunoassay 1984;7:206-11.

(24.) Liewendahl K, Tikanoja S, Helenius T, Valimaki M. Discrepancies between serum free triiodothyronine and free thyroxin as measured by equilibrium dialysis and analog radioimmunoassay in nonthyroidal illnesses. Clin Chem 1984;30:760-2.

(25.) Stockigt JR, deGaris M, Csicsmann J, Barlow JW, White EL, Hurley DM. Limitations of a new free thyroxine assay (Amerlex [FT.sub.4]). Clin Endocrinol (Oxf) 1981;15:313-8.

(26.) Ooi G, Mahadevan MS, Greenway DC, Gertler SZ. Evaluation of four commercially available assays for free thyroxin. Clin Chem 1988;34:2302-7.

(27.) Czako G, Zweig MH, Glickman J, Ruddel M, Kestner J. Direct and indirect techniques for free thyroxin compared in patients with nonthyroidal illness. II. Effect of prealbumin, albumin and thyroxin-binding globulin. Clin Chem 1989;35:1655-62.

(28.) Wilkins TA. Albumin in analog [FT.sub.4] assay reagents: the facts [Letter]. Clin Chem 1987;33:1293.

(29.) Tikanoja SH, Liewendahl K. New ultrafiltration method for free thyroxin compared with equilibrium dialysis in patients with thyroid dysfunction and nonthyroidal illness. Clin Chem 1990;36:800-4.

(30.) Docter R, van Toor H, Krenning EP, de Jong M, Hennemann G. Free thyroxine assessed with three assays in sera of patients with nonthyroidal illness and of subjects with abnormal concentrations of thyroxine-binding proteins. Clin Chem 1993;39:2805-11.

(31.) Shannon N, Woolf PD. Determination of free thyroxin in serum by ultrafiltration: validation of a method and preliminary results. Clin Chem 1984;30:1770-3.

(32.) Wang Y-S, Hershman JM, Pekary AE. Improved ultrafiltration method for simultaneous measurement of free thyroxine and free triiodothyronine in serum. Clin Chem 1985;31:517-22.

(33.) Nelson JC, Weiss RM. The effect of serum dilution on free thyroxine ([T.sub.4]) concentration in the low [T.sub.4] syndrome of nonthyroidal illness. J Clin Endocrinol Metab 1985;61:239-46.


[1] Research and Development, Ortho-Clinical Diagnostics, Cardiff Laboratories, Whitchurch, Forest Farm Estate, Cardiff CF4 7YT, Wales, UK.

University Departments of [2] Clinical Biochemistry and [3] Anesthetics, The University of Edinburgh, Royal Infirmary, Lauriston Place, Edinburgh, Scotland, UK.

[4] Nonstandard abbreviations: [T.sub.4], thyroxine; [FT.sub.4], free thyroxine; ED, equilibrium dialysis; [PBT.sub.4], protein-bound thyroxine; sBC, serum binding capacity; T3, triiodothyronine; TT4, total thyroxine; T3U, triiodothyronine uptake; iPBT4, protein-bound thyroxine in the immunoassay; and iBC, immunoassay binding capacity.

* Author for correspondence. Fax 44 (0)1222526635; e-mail

Received May 28, 1998; accepted January 13, 1999.
Table 1. Measured and calculated values in the
three patient groups studied.

 Pregnant Ambulatory
ED [FT.sub.4],
 pmol/L 9.0 [+ or -] 1.7 (a) 14.3 [+ or -] 2.7
 Range 5.8-11.6 9.7-23.7

Vitros [FT.sub.4],
 pmol/L 11.6 [+ or -] 1.7 (a) 15.9 [+ or -] 2.8
 Range 8.5-14.4 11.7-25.8
Vitros [TT.sub.4],
 nmol/L 136.2 [+ or -] 23.1 (a) 84.7 [+ or -] 11.9
 Range 90.8-172.0 62.2-116.0

Vitros %T3U 18.9% [+ or -] 1.3% (a) 29.3% [+ or -] 2.18%
 Range 16.9-22.2% 25.1 [+ or -] 33.3%

 sBC, nmol/pmol 15.5 [+ or -] 3.1 (a) 6.1 [+ or -] 1.1
 Range 10.7-23.9 4.2-8.5

ED [FT.sub.4],
 pmol/L 25.7 [+ or -] 9.1 (a)
 Range 14.2-48.6

Vitros [FT.sub.4],
 pmol/L 25.2 [+ or -] 8.1 (a)
 Range 13.2-48.3
Vitros [TT.sub.4],
 nmol/L 62.2 [+ or -] 24.1 (a)
 Range 23.2-111.0

Vitros %T3U 47.5% [+ or -] 10.1% (a)
 Range 31.3-72.9%

 sBC, nmol/pmol 2.6 [+ or -] 1.0 (a)
 Range 1.2-4.5

(a) Significant (P <0.001) difference from the corresponding
ambulatory concentration.
COPYRIGHT 1999 American Association for Clinical Chemistry, Inc.
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1999 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:Endocrinology and Metabolism
Author:Christofides, Nicos D.; Wilkinson, Ewan; Stoddart, Mary; Ray, Avid C.; Beckett, Geoff J.
Publication:Clinical Chemistry
Date:Apr 1, 1999
Previous Article:A practical approach to determine cutoff concentrations for opiate testing with simultaneous detection of codeine, morphine, and 6-acetyl morphine in...
Next Article:Immunoradiometric assay for intact human osteocalcin(I-49) without cross-reactivity to breakdown products.

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters