Printer Friendly

Quantifying spurious free [T.sub.4] results attributable to thyroxine-binding proteins in serum dialysates and ultrafiltrates.

Direct equilibrium dialysis and direct ultrafiltration free thyroxine ([T.sub.4]) methods use semipermeable membranes to separate free [T.sub.4] from [T.sub.4]-binding serum proteins. Nonanalog free [T.sub.4] RIAs measure this free [T.sub.4]. These measurements are based on [[sup.125.I]][T.sub.4] binding to anti-[T.sub.4] and its displacement from anti-[T.sub.4] by unlabeled free [T.sub.4] (1-4). When [T.sub.4]-binding serum proteins ([T.sub.4]BSPs) are present they bind [[sup.125.I]][T.sub.4] and displace [[sup.125.I]][T.sub.4] from anti-[T.sub.4], yielding spurious results from these free [T.sub.4] assays (see Figure 1 in the Data Supplement that accompanies the online version of this Technical Brief at The spurious free [T.sub.4] results caused by [T.sub.4]BSPs and the protein concentrations at which they occur have not been reported. This study used [T.sub.4]-depleted serum proteins to obtain spurious free [T.sub.4] results and to determine the protein concentrations at which they occurred.

A nonanalog free [T.sub.4] RIA (Nichols Institute Diagnostics) and a total [T.sub.4] RIA (Coat-A-Count, Diagnostic Products) were used. A pool of normal human serum with thyroxine-binding globulin (TBG; 17 mg/L), transthyretin (0.27 g/L), albumin (42 g/L), total protein (78 g/L), total [T.sub.4] (94 nmol/L, 73 [micro]g/L), and dialyzable (free) [T.sub.4] (12.9 pmol/L, 10 ng/L) was obtained from Equitech-Bio. Anti-[T.sub.4], anti-[T.sub.3], anti-IgG, and salicylates were undetectable. Adapting the method of Grundy et al. (5), we used Amberlite IRA-410 anion exchange resin (Alfa Aesar) to strip 1 serum aliquot of [T.sub.4]. Another serum aliquot was ultrafiltered. Concentrations of these [T.sub.4]-depleted serum total proteins varied from 5 x [10.sup.-7] to 0.05 g/L. In addition, proteins concentrated by ultrafiltration were diluted with the corresponding ultrafiltrate to concentrations that varied from 40-160 g/L. Highly purified human serum albumin (>99%, Sigma-Aldrich) and TBG (>95%, Cortex Biochem) were dissolved in [T.sub.4]-depleted human serum dialysate. Concentrations of albumin varied from 2 X [10.sup.-2] to 200 g/L and of TBG from 6 x [10.sup.-5] to 60 mg/L. Total and free [T.sub.4] were undetectable in these protein solutions.

The dialysis devices tested were a vertical membrane device using 10 mL of retentate and 10 mL of dialysate (Fisher Scientific) and a horizontal membrane device using 200 [micro]L of retentate and 2400 [micro]L of dialysate (Nichols Institute Diagnostics). According to the manufacturers, both devices use a regenerated cellulose membrane with a 12-14 kDa molecular weight cutoff (MWCO; Spectra/Por, Fisher Scientific). Membranes were washed with deionized water before use. The dialysis buffer used was reported previously (1). To test for proteins in dialysates, [T.sub.4]-depleted human serum (40-160 g/L) was dialyzed for 18 h at 37 [degrees]C. During dialysis, mean (SD) pH was controlled to 7.4 (0.1) at 37[degrees]C by the HEPES contained in the dialysis buffer (6). The final HEPES ion concentration was 54 mmol/L (1).

The ultrafiltration devices tested were the Centricon YM-10, Centricon YM-30, and Amicon Ultra-4 (Millipore). According to the manufacturer the regenerated cellulose membranes in these devices had MWCOs of 10 kDa, 30 kDa, and 10 kDa, respectively.

All devices were prerinsed with deionized water. Centrifugation was carried out in a temperature-controlled (37 [degrees]C), fixed-angle rotor centrifuge at 30008. Before ultrafiltration, [T.sub.4]-depleted serum pH was controlled to 7.4 (0.1) at 37[degrees]C, by addition of 40 [micro]L of 1200 mmol/L HEPES acid (Fisher Biotech) per milliliter of serum. Serum pH stability was obtained by 15 min of vortex-mixing at room temperature while a continuous stream of moist air passed across the serum. The final HEPES ion concentration was 54 mmol/L (1).

A 4th ultrafiltration device, the Centricon YM-100 device (Millipore), was used as a positive control. This device had a regenerated cellulose membrane with a MWCO of 100 kDa. [T.sub.4] binding serum proteins ([T.sub.4]BSPs) are expected to enter the ultrafiltrates obtained with this device because their molecular weights are 54-66 kDa. The dialysate of [T.sub.4]-depleted human serum was used as the negative control.

Sodium levothyroxine (for injection) was obtained in 500-[micro]g vials (Bedford Laboratories) and dissolved in 5 mL of sodium chloride, USP, 9 g/L (Abbott Laboratories). This [T.sub.4] solution (125 /,mol/L, 100 mg/L) was diluted with [T.sub.4] depleted human serum dialysate to concentrations ranging from 2.6 x [10.sup.-5] to 26 nmol/L (2 x [10.sup.-5] to 20 [micro]g/L).

Free [T.sub.4] is often adsorbed from aqueous solutions onto solid surfaces. This adsorption was determined using the method of Holm et al. (7) for the test tubes (borosilicate glass) and screw-capped vials (borosilicate glass) used in the present study (Fisher Scientific). [[sup.125.I]][T.sub.4] (Perkin-Elmer Life Sciences) was purified using Sephadex G-25 (Sigma-Aldrich) column chromatography (8). The Sephadex columns were equilibrated to 0.01 mol/L PBS (1 PBS tablet dissolved in 200 mL of water to obtain: 10 mmol/L phosphate buffer, 2.7 mmol/L potassium chloride, and 137 mmol/L sodium chloride; Sigma-Aldrich), at pH 5.4 and room temperature. A stock solution of [[sup.125.I]][T.sub.4] was added to the column and eluted with 100 mmol/L sodium hydroxide (Sigma-Aldrich). Fractions were collected in 13 x 100 mm glass test tubes using an automated fraction collector (LKB). Gamma radiation was quantified using a multiwell, automated gamma counter (Gamma 4000, Beckman-Coulter). Adsorption was <0.4% in test tubes and vials.

A theoretical lower limit for the detection of displacements was calculated as the mean minus 2 SD of the zero control data. This limit was 0.001 mg/L for [T.sub.4]-depleted TBG, 0.025 g/L for [T.sub.4]-depleted albumin, and 0.0015 g/L for [T.sub.4]-depleted total protein, compared to 0.8 pmol/L (0.6 ng/L) for free [T.sub.4] (see Fig. 2 in the online Data Supplement).

As a percentage of normal serum concentration, [T.sub.4]depleted serum total protein concentrations and [T.sub.4]-depleted TBG concentrations were similarly effective in displacing [[sup.125.I]][T.sub.4], and both were more effective than [T.sub.4]-depleted albumin (Table 1). We used [T.sub.4]-depleted serum total proteins, which contain all [T.sub.4]BSPs, to test for spurious free [T.sub.4] results in dialysates and ultrafiltrates.

When [T.sub.4]-depleted serum total proteins at concentrations ranging from 40 to 160 g/L were applied to the 2 dialysis devices tested, no spurious free [T.sub.4] determinations were obtained. One of the 3 ultrafiltration devices yielded spurious free [T.sub.4] results (Fig. 1), which correlated with the concentrations of serum proteins in retentates (r = 0.94; P = 0.047). This result was unexpected because the designated MWCO was 10 kDa and the molecular weights of [T.sub.4]13SPs are 54-66 kDa. Spurious free [T.sub.4] determinations were obtained with the positive control, as expected (Fig. 1).

Previous studies have measured albumin in serum dialysates or ultrafiltrates. Weeke et al. (9, 10) measured albumin in serum ultrafiltrates with a microalbuminuria RIA with a sensitivity of 0.001 g/L. They calculated that a leakage of 0.0015% of undiluted serum albumin would induce an error of ~5% in free [T.sub.4] determinations. With the free [T.sub.4] RIA method, the presence of 0.06% (0.025 g/L) serum albumin resulted in spurious free [T.sub.4] values of ~3% (0.5 pmol/L).


Tikanoja et al. (11) measured albumin in serum ultrafiltrates using an albumin RIA with a detection limit of 8.0 x [10.sup.-4] g/L. They used a cutoff value for protein leakage of 0.005% of serum proteins, citing Weeke et al. (9). Ultrafiltration devices that allowed <0.005% (0.002 g/L) of albumin into ultrafiltrates were regarded as acceptable. Only 1 device of 4 met this criterion. Again, this compares to the free [T.sub.4] RIA method for which 0.06% (0.025 g/L) of serum albumin resulted in spurious free [T.sub.4] values of ~3% (0.5 pmol/L).

Holm et al. (12) measured albumin in both serum dialysates and serum ultrafiltrates using a double antibody sandwich ELISA with a detection limit of 2.8 x [10.sup.-6] g/L. They found no detectable albumin in serum dialysates but found detectable albumin in the ultrafiltrates obtained with each of 5 different devices.

The previous studies measured protein concentrations by use of albumin assays that were more sensitive for detecting albumin than the free [T.sub.4] RIA method used in the present study (0.001 g/L, 8.0 x [10.sup.-4] g/L, and 2.8 x [10.sup.-6] g/L, compared to 0.025 g/L). The present method, however, detects the spurious free [T.sub.4] values caused by any and all [T.sub.4]BSPs as a consequence of binding radiolabeled [T.sub.4].

Uncertainty about spurious free [T.sub.4] determinations as a consequence of [T.sub.4]13SP interference is unnecessary. The laboratory that applies such a free [T.sub.4] RIA to the quantification of [T.sub.4] in serum dialysates or serum ultrafiltrates can use the same RIA to rule in or rule out these spurious free [T.sub.4] values.

Grant/funding support: This work was partially supported by the Loma Linda University School of Medicine and Mortensen Chair. No extramural funds were used for this study.

Financial disclosures: Jerald C. Nelson is a consultant to Antech Diagnostics.

Acknowledgements: We thank Nichols Institute Diagnostics for providing some of the test kits and reagents used in this study.

DOI: 10.1373/clinchem.2007.085316


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

(2.) 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.

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

(4.) Weeke J, Orskov H. Measurement of free thyroid hormones by dialysis and radio-immunoassay. Intl Symposium on Free Thyroid Hormones. 1978, Venice, Italy.

(5.) Grundy EN, Smith AM. Preparation of ([T.sub.3] + [T.sub.4])-free pooled plasma. Ann Clin Biochem 1986;23:494-5.

(6.) Wilcox RB, Nelson JC. Time course of pH regulation in free thyroxin determinations in serum. Clin Chem 1991;37:1298-300.

(7.) Holm SS, Hansen SH, Faber J, Staun-Olsen P. Reference methods for the measurement of free thyroid hormones in blood: evaluation of potential reference methods for free thyroxine. Clin Biochem 2004;37:85-93.

(8.) Bauer R. Use of alkaline Sephadex G-25 for extraction and measurement of thyroxine. Clin Chem 1974;20:917-8.

(9.) Weeke J, Boye N, Orskov H. Ultrafiltration method for direct radioimmunoassay measurement of free thyroxine and free tri-iodothyronine in serum. Scand J Clin Lab Invest 1986;46:381-9.

(10.) Christensen C, Orskov C. Rapid screening PEG radioimmunoassay for quantification of pathological microalbuminuria. Diabetic Nephropathy 1984;3:242-50.

(11.) Tikanoja S. Ultrafltration devices tested for use in a free thyroxine assay validated by comparison with equilibrium dialysis. Scand J Clin Lab Invest 1990;50:663-9.

(12.) Holm SS, Andreasen L, Hansen SH, Faber J, Staun-Olsen P. Influence of adsorption and deproteination on potential free thyroxine reference methods. Clin Chem 2002;48:108-14.

Kristofer S. Fritz, [1] R. Bruce Wilcox, [1] and Jerald C. Nelson [2] *

[1] Departments of Biochemistry and [2] Internal Medicine and Pathology, Loma Linda University School of Medicine, Loma Linda, CA;

* Address correspondence to this author at: Loma Linda University Medical Center, 11234 Anderson St., Rm. 1568, Loma Linda, CA 92354; fax 909-558-0490; e-mail
Table 1. Spurious free [T.sub.4] values due to [T.sub.4]BSPs.

 Spurious free
 [sup.125][I-T.sub.4] [T.sub.4] values,
 displacements, % pmol/L (a)

TBG 5 1.14
 25 9.8
 50 33
 75 112
Albumin 5 1.14
 25 9.8
 50 33
 75 112
Serum total proteins 5 1.14
 25 9.8
 50 33
 75 112

 [T.sub.4]BSP as
 [T.sub.4]BSP portion of normal
 concentrations concentration, %

TBG 0.001 mg/L 0.006
 0.008 mg/L 0.05
 0.027 mg/L 0.16
 0.1 mg/L 0.6
Albumin 0.036 g/L 0.09
 0.24 g/L 0.6
 0.76 g/L 1.8
 2.4 g/L 6.0
Serum total proteins 0.002 g/L 0.002
 0.01 g/L 0.01
 0.04 g/L 0.05
 0.14 g/L 0.16

(a) To convert [T.sub.4] pmol/L to ng/dL, divide by 12.9.
COPYRIGHT 2007 American Association for Clinical Chemistry, Inc.
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2007 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:Technical Briefs
Author:Fritz, Kristofer S.; Wilcox, R. Bruce; Nelson, Jerald C.
Publication:Clinical Chemistry
Date:May 1, 2007
Previous Article:Perioperative activin a concentrations as a predictive marker of neurologic abnormalities in children after open heart surgery.
Next Article:Detection of unexpected isoforms of human chorionic gonadotropin by qualitative tests.

Related Articles
A direct free thyroxine ([T.sub.4]) immunoassay with the characteristics of a total [T.sub.4] immunoassay.
Unequal concentrations of free [T.sub.3] and free [T.sub.4] after ultrafiltration.
Direct and indirect free thyroxine assay methods: theory and practice.
Heparin-induced release of protein-bound solutes during hemodialysis is an in vitro artifact.
Recombinant human intrinsic factor expressed in plants is suitable for use in measurement of vitamin [B.sub.12].
Intermethod discordant free thyroxine measurements in bone marrow-transplanted patients.
Homogeneous enzyme immunoassay for triiodothyronine in serum.
Influence of adsorption and deproteination on potential free thyroxine reference methods.
Effects of uremic toxins and fatty acids on serum protein binding of furosemide: possible mechanism of the binding defect in uremia.
Dysprealbuminemic hyperthyroxinemia in a patient with hyperthyroid graves disease.

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