Inverse Log-Linear Relationship between Thyroid ... - Semantic Scholar

Clinical Chemistry 57:1 122–127 (2011)

Endocrinology and Metabolism

Inverse Log-Linear Relationship between Thyroid-Stimulating Hormone and Free Thyroxine Measured by Direct Analog Immunoassay and Tandem Mass Spectrometry Hendrick E. van Deventer,1 Damodara R. Mendu,2 Alan T. Remaley,1 and Steven J. Soldin1,2,3*

BACKGROUND: Accurate measurement of free thyroxine (FT4) is important for diagnosing and managing thyroid disorders. Most laboratories measure FT4 by direct analogue immunoassay methods. The validity of these methods have recently been questioned. The inverse log-linear relationship between FT4 and thyroidstimulating hormone (TSH) is well described and provides a physiological rationale on which to base an evaluation of FT4 assays. METHODS:

The study included 109 participants for whom FT4 measurement was requested by their clinician. Samples were selected for inclusion to reflect a wide spectrum of TSH and albumin results. FT4 and TSH were measured by use of the Siemens Immulite immunoassay (IA). FT4 was also measured by liquid chromatography–tandem mass spectrometry (LC-MS/ MS) (MS-FT4).

RESULTS:

The inverse log-linear correlation coefficient between TSH and FT4 was significantly better (P ⬍ 0.0001) for MS-FT4 (0.84, 95% CI, 0.77– 0.88) than for IA-FT4 (0.45, 95% CI, 0.29 – 0.59). IA-FT4 showed a significant correlation with albumin (Spearman correlation coefficient 0.45, 95% CI, 0.29 – 0.5, P ⬍ 0.0001) and thyroxine-binding globulin (TBG) (Spearman correlation coefficient 0.23, 95% CI, 0.05– 0.41, P ⫽ 0.02). In contrast, FT4 measurement by LC-MS/MS did not show a significant correlation with albumin or TBG. CONCLUSIONS: The inverse log-linear relationship between FT4 and TSH was significantly better for FT4 measured by LC-MS/MS than by IA. The MS-FT4 method therefore provides FT4 results that agree clini-

1

Department of Laboratory Medicine, National Institutes of Health, Bethesda, MD; 2 Bioanalytic Core Laboratory, General Clinical Research Center, Georgetown University Medical Center; Departments of Pharmacology and Medicine, Georgetown University; Departments of Pediatrics and Pathology, The George Washington University School of Medicine, Washington, DC; 3 Clinical Endocrinology Laboratory, NMS Laboratories, Willow Grove, PA. * Address correspondence to this author at: Room GM12A, Preclinical Science Building, Bioanalytical Core Laboratory, Georgetown University, Washington, DC. E-mail [email protected].

122

cally with those obtained for TSH. Additionally, the significant correlation between IA-FT4 with albumin and TBG suggests that this FT4 method depends on binding protein concentrations and consequently does not accurately reflect FT4. © 2010 American Association for Clinical Chemistry

Accurate measurement of thyroxine (T4)4 and thyroidstimulating hormone (TSH) is important for the management of thyroid disorders. In serum, the majority of thyroxine circulates bound to high-concentration, low-affinity proteins, mostly albumin (ALB) and transthyretin, and to a low-concentration, high-affinity binding protein, namely thyroxine-binding globulin (TBG). Only a small percentage of total thyroxine (TT4) circulates free (1 ). It is widely accepted that it is the free thyroxine (FT4) that is biologically active and, therefore, of most interest to monitor in patients with thyroid disorders (2 ). From 1990 to 2004, the Nichols equilibrium dialysis (ED)/RIA method was regarded by endocrinologists and laboratorians alike as the state-of-the-art FT4 procedure (3 ). In addition, previous studies demonstrated an excellent comparison between the Nichols ED/RIA and ultrafiltration followed by liquid chromatography–tandem mass spectrometry (LC-MS/MS) (4, 5 ). Today the gold standard for the measurement of FT4 is considered to be ED or ultrafiltration (3, 6 ), in which free analyte is first separated from that bound to serum proteins and then measured by a highly sensitive and specific T4 assay. The imprecision of our ultrafiltration LC-MS/MS FT4 method was previously reported to provide CVs rang-

Received August 1, 2010; accepted October 28, 2010. Previously published online at DOI: 10.1373/clinchem.2010.154088 4 Nonstandard abbreviations: T4, thyroxine; TSH, thyroid-stimulating hormone; ALB, albumin; TBG, thyroxine-binding globulin; TT4, total thyroxine; FT4, free thyroxine; ED, equilibrium dialysis; LC-MS/MS, liquid chromatography–tandem mass spectrometry; IA, immunoassay; NIH-CC, NIH Clinical Center; T4-d5, deuterium-labeled L-thyroxine; MRM, multiple reaction monitoring; AUC, area under the curve.

Evaluating the Relationship between FT4 and TSH

ing between 4.1% and 6.6% at concentrations between 0.66 and 2.62 ng/dL (8.5–33.8 pmol/L) (4 ). We found that the Nichols ED/RIA method, at the same concentrations, had CVs between 8% and 15%, which is similar to that found by Okabayashi et al. (7 ). Ultrafiltration for routine everyday use also has the advantage of being less time consuming (30 – 40 min vs at least 17–24 h for ED). In practice, most clinical laboratories use direct (analog) immunoassays for the measurement of FT4 (8 ). There are multiple FT4 immunoassay methods, but most are based on T4 analog binding and displacement (9 ) and rely on the measurement of FT4 in diluted serum without the preparation of protein free fractions. The validity of free thyroid hormone measurement by direct analog immunoassay is still debated and has many limitations (10 –12 ). FT4 results by immunoassay are poorly standardized (13, 14 ) and are affected by binding protein concentrations (15, 16 ). The inverse log-linear relationship between TSH and free thyroid hormone due to the negative feedback of these hormones on the pituitary is well described (17 ) and provides a physiological rationale for assessing the validity of FT4 results in a clinical setting. A recent article suggests that the correlation of FT4 with log TSH is poor on the Abbott Architect ci8200 IA platform, which uses the direct analog method for FT4 measurement (18 ). The objective of this study was to investigate the validity of a direct analog Siemens immunoassay (IA) FT4 method and a LC-MS/MS FT4 (MS-FT4) method by (a) evaluating the inverse log-linear relationship between IA-FT4 and MS-FT4 with log TSH measured with an immunoassay, and (b) assessing whether any correlation exists between IA-FT4 or MS-FT4 with ALB or TBG concentrations. Methods PARTICIPANTS

This study was a prospective study of samples received at the NIH Clinical Center (NIH-CC) from January 2010 to February 2010 for the measurement of FT4 and TSH. Samples were selected for inclusion in the study to reflect a spectrum of normal, low, and high TSH results, as well as normal and low albumin results. In total, we included 109 samples for analysis. The study was approved by the Institutional Review Board of the NIH (Clinical Protocol number 93-CC-0094). TEST METHODS

Methods performed at the NIH-CC Department of Laboratory Medicine. Blood samples were collected in plastic red-top tubes containing clot activator (Vacu-

tainer®; Becton Dickinson). Samples were processed according to usual laboratory procedures for TSH, FT4, and albumin. Serum was removed and stored in cryogenic vials (Corning) at ⫺80 °C until LC-MS/MS analysis. We measured FT4 by a direct (analog) immunoassay method on a Siemens Immulite 2500 analyzer (Diagnostic Products; Siemens Healthcare Diagnostics) (reference interval 10.3–19.4 pmol/L). TSH (reference interval 0.40 – 4.00 mIU/L), and TBG (reference interval 241–271 nmol/L) were measured on the same analyzer. We measured albumin (reference interval 34 –50 g/L) on the Dimension Vista (Siemens Healthcare Diagnostics) using a bromcresol purple dyebinding method. All immunoassays and protein measurements were performed at the NIH on the day of sample collection. Methods performed at the Bioanalytical Core Laboratory, Georgetown University. MS-FT4 measurements were performed as described (4, 19 –21 ) with minor modifications. Briefly, samples frozen at ⫺80 °C were thawed at room temperature, and 400 ␮L serum was placed in a 30-kDa ultrafiltration device (Centrifree YM-30, Millipore) and centrifuged in an Eppendorf temperature-controlled centrifuge at 1113g for 30 min at 37 °C. We then added 150 ␮L ultrafiltrate to 450 ␮L methanol containing deuterium-labeled L-thyroxine (T4-d5) from IsoSciences used as internal standard (⬎96.2% pure), vortex-mix mixed, and centrifuged. We diluted 500 ␮L of the supernatant with 600 ␮L water and injected 600 ␮L onto an Agilent SB C-18 (2.1 mm ⫻ 50 mm, 3.5 ␮m ID) chromatographic column. The HPLC system consisted of 3 Shimadzu LC-20AD pumps, a Shimadzu SIL-HTA autosampler, and a Shimadzu DGU-20A5 degasser. The procedure involved an online extraction step followed by activation of a built-in Valco switching valve and subsequent sample introduction into the mass spectrometer. After a 3-min wash with 20% (vol/vol) methanol in 0.01% acetic acid at a flow rate of 1.0 mL/min, the switching valve was activated and the analytes of interest were eluted from the column and introduced into the mass spectrometer with a water/methanol gradient (see Supplemental Table 1, which accompanies the online version of this article at www.clinchem.org/content/ vol57/issue1). We used an API-5000 tandem mass spectrometer (Applied Biosystems/MDS Sciex) equipped with TurboIonSpray source, operated in the negative ionization multiple reaction monitoring (MRM) mode. The MRM transitions monitored for FT4 and T4-d5 were 775.6/126.9 and 780.8/126.9, respectively. Compounddependent and instrument-dependent parameters of the mass spectrometer were collision gas 11, curtain gas 30, ion source gas1 15, ion source gas2 45, ionspray voltage Clinical Chemistry 57:1 (2011) 123

4200 V, probe temperature 670 °C, and dwell time 250 ms. Declustering potential was ⫺120 for FT4 and ⫺173 for T4-d5, and collision energy was ⫺62 for FT4 and ⫺82 for T4-d5. Data were acquired and processed by Analyst 1.4.1 software package. L-Thyroxine for preparation of the calibration curve was obtained from Sigma. HPLC-grade methanol was from Fisher Scientific. Differences between this and previously published methods included (a) ultrafiltration performed at 37 °C [previous studies compared 25 °C to 37 °C, see Jonklaas et al. (20 )], (b) chromatography performed on an Agilent C-18 column, [previous studies used a Supelco LC-18-DB (3.3 cm ⫻ 3.0 mm, 3.0 ␮m ID)], and (c) introduction of a gradient instead of isocratic mobile phase. STATISTICAL METHODS

We conducted statistical analysis using Analyze-it for Microsoft Excel. We used the Shapiro–Wilk test to test for normality, and we used correlation coefficient, Bland–Altman difference plots, and ordinary least square regression analysis to evaluate the methods. Results METHOD VALIDATION

Pearson correlation coefficient between the Agilent SB C-18 (2.1 mm ⫻ 50 mm, 3.5 ␮m ID) and Supelco LCC-18-DB (3.0 mm ⫻ 33 mm, 3.0 ␮m ID) columns, with ultrafiltration performed at 37 °C, was R ⫽ 0.99 (95% CI, 0.98 –1.00), and the residual standard deviation (Sy|x) was 0.28 pmol/L. The slope was 0.86 (95% CI, 0.73– 0.99), and the intercept was 0.68 pmol/L (95% CI, ⫺2.36 to 3.72) calculated using weighted Deming regression analysis (n ⫽ 15, range: 2.58 –387 pmol/L). Comparison of modified LC-MS/MS procedure with LC-MS/MS method at NMS Laboratories gave a slope of 0.95 (95% CI, 0.84 –1.06), an intercept of 2.67 (95% CI, 2.52–2.82) pmol/L, a R ⫽ 0.99 (95% CI, 0.98 –1.00), and Sy|x of 0.22 pmol/L (n ⫽ 23, range 1.29 –343.14 pmol/L). PARTICIPANTS AND TEST RESULTS

The study included 109 samples. The median TSH was 2.00 mIU/L and ranged between ⬍0.02 mIU/L and 215 mIU/L. Twenty-five samples (23%) had TSH ⬍0.04 mIU/L, 47 (43%) had TSH between 0.04 and 4.00 mIU/L, and 37 (34%) had TSH ⬎4.00 mIU/L. The median ALB was 37 g/L and ranged between 14 and 45 g/L. Thirty-one samples (28%) had ALB ⬍34 g/L. The mean TBG was 348 nmol/L (95% CI, 331–365) and ranged between 135 and 579 nmol/L. 124 Clinical Chemistry 57:1 (2011)

Fig. 1. Inverse log-linear relationship between TSH and IA-FT4 (A) and MS-FT4 (B).

METHOD COMPARISON

The Pearson correlation coefficient between IA-FT4 and LC-MS/MS was 0.45 (95% CI 0.29 – 0.59). Bland– Altman difference plots between IA-MS and LCMS/MS showed 95% limits of agreement to be between 32.61 and ⫺42.43 pmol/L (see online Supplemental Fig. 1). The inverse log-linear Pearson correlation between MS-FT4 and log TSH, 0.84 (95% CI, 0.77– 0.88), was significantly better (P ⬍ 0.0001) than between IAFT4 and log TSH, 0.45 (95% CI, 0.29 – 0.59) (Fig. 1). For patients with TSH ⬍0.40 mIU/L and TSH ⬎4.00 mIU/L, the inverse log-linear Pearson correlation between MS-FT4 and log TSH, 0.86 (95% CI, 0.77– 0.91),

Evaluating the Relationship between FT4 and TSH

was significantly better (P ⬍ 0.0001) than between IAFT4 and log TSH, 0.50 (95% CI, 0.28 – 0.66). For patients with TSH 0.40 mIU/L to 4.00 mIU/L, the Pearson correlation between MS-FT4 and log TSH of 0.72 (95% CI, 0.55– 0.83) was significantly better (P ⫽ 0.0009) than between IA-FT4 and log TSH, 0.20 (95% CI, 0.09 – 0.46). FT4 measurement by immunoassay showed a significant correlation with albumin concentration (Spearman correlation coefficient 0.45; 95% CI, 0.29 – 0.59; P ⬍ 0.0001) and also with TBG concentration (Spearman correlation coefficient 0.23; 95% CI, 0.05– 0.41; P ⫽ 0.02). In contrast, FT4 measurement by LCMS/MS did not show a significant correlation with albumin concentration (Spearman correlation coefficient 0.05; 95% CI, ⫺0.14 – 0.24; P ⫽ 0.59) or TBG concentration (Spearman correlation coefficient 0.13; 95% CI, ⫺0.06 – 0.31; P ⫽ 0.19) (Fig. 2). The area under the curve (AUC), calculated using ROC curves, for the prediction of TSH ⬎4.0 mIU/L was significantly better for MS-FT4 (0.89, 95% CI, 0.83– 0.95) than for IA-FT4, (0.62, 95% CI, 0.51– 0.73) (P ⬍ 0.0001). The AUC for the prediction of TSH ⬍0.4 mIU/L was significantly better for MS-FT4 (0.97, 95% CI, 0.95–1.00) than for IA-FT4 (0.76, 95% CI, 0.62– 0.89) (P ⫽ 0.002). Discussion In this study, the IA-FT4 and MS-FT4 methods yielded widely discrepant results; however, the MS-FT4 method correlated significantly better with log TSH than IA-FT4. This suggests that this MS-FT4 method more accurately reflects true FT4 concentrations. One possible reason, demonstrated in this study, is that the IA-FT4 concentration still remains dependent on both ALB and TBG concentration and therefore may not be a true reflection of free hormone concentration. The ultrafiltration device used in this MS-FT4 method effectively removes proteins (22 ) and therefore removes protein-bound T4 from the sample and replaces the dialysis step of the classic equilibrium dialysis method. The protein concentration in the ultrafiltrate is dependent on the filtration device used, temperature, centrifugal force, and time of centrifugation. The fact that no significant correlation exists in this study between FT4 measured by ultrafiltration LC-MS/MS and albumin or TBG indicates that protein leakage does not play a role in our method. The poor correlation between IA-FT4 and log TSH, as well as the significant correlation between IAFT4 and ALB and TBG, highlights some of the limitations of current immunoassays for the measurement of FT4. It is important that clinicians be aware of these

Fig. 2. Correlation between albumin and IA-FT4 (A) and MS-FT4 (B).

limitations to make informed clinical decisions. These limitations are important in conditions that affect binding protein concentration, such as acute illness, pregnancy (23 ), and hereditary variants in the structure of TBG, ALB, or transthyretin (24 ). It is doubtful that standardization of direct analog immunoassays, without addressing the dependence of FT4 measured on binding protein concentration, will result in methods with acceptable performance, especially for diseased individuals. The increased availability of mass spectrometry in clinical laboratories coupled with a rapid ultrafiltration procedure described in this and Clinical Chemistry 57:1 (2011) 125

previous studies shows that it is possible for ultrafiltration LC-MS methods to be routinely adopted. Indeed, this is already the case at Children’s National Medical Center, where such an approach has now been used routinely for ⬎4 years. Ultrafiltration was performed at 37 °C instead of the previously published 25 °C (19 ). Switching from 25 °C to 37 °C increases FT4 results by a factor of 1.5 (21 ). As stated previously, the correlation of LCMS/MS FT4 with log TSH was excellent at both 25 °C and 37 °C (20, 25 ). A limitation of this study is that TSH was measured using an immunoassay. Although TSH measurement is generally considered reliable (8 ), TSH is not well defined, and no reference measurement procedure is available. TSH results differ among immunoassay manufacturers (26 ). TSH is a glycoprotein that has various glycoforms in blood, and glycosylation patterns may differ with different thyroid disease states (27 ). A further limitation is that no clinical information is available; the study may therefore include patients taking thyroid medication or with nonthyroidal illnesses, since these patients could not be excluded. These are, however, patients in which regular monitoring of TSH and FT4 is routinely requested. In conclusion, IA-FT4 and MS-FT4 yielded widely discrepant results. Ultrafiltration MS-FT4, however, correlated significantly better with log TSH. IA-FT4 measurements still remained depen-

dent on binding protein concentration. Clinicians need to be aware of current limitations of immunoassays for the measurement of FT4. The validity of current immunoassays for the measurement of FT4 needs to be reexamined, and the more widespread adoption of ultrafiltration LC-MS/MS for the measurement of FT4 may be warranted.

Author Contributions: All authors confirmed they have contributed to the intellectual content of this paper and have met the following 3 requirements: (a) significant contributions to the conception and design, acquisition of data, or analysis and interpretation of data; (b) drafting or revising the article for intellectual content; and (c) final approval of the published article. Authors’ Disclosures or Potential Conflicts of Interest: Upon manuscript submission, all authors completed the Disclosures of Potential Conflict of Interest form. Potential conflicts of interest: Employment or Leadership: None declared. Consultant or Advisory Role: S.J. Soldin, NMS Laboratories. Stock Ownership: None declared. Honoraria: S.J. Soldin, Siemens. Research Funding: Tests performed at NIH (FT4 by immunoassay and the TBG, albumin, and TSH tests) were supported in part by the Intramural Research Program of the NIH, Warren Grant Magnuson Clinical Center. S.J. Soldin is partially supported by NIH Clinical and Translational Science Awards grant 1UL1RR031975-01. Expert Testimony: None declared. Role of Sponsor: The funding organizations played no role in the design of study, choice of enrolled patients, review and interpretation of data, or preparation or approval of manuscript.

References 1. Schussler GC. Thyroxine-binding proteins. Thyroid 1990;1:25–34. 2. Midgley JE. The free thyroid hormone hypothesis and measurement of free hormones. Clin Chem 1993;39:1342– 4. 3. Soldin OP, Soldin SJ. Thyroid hormone testing by tandem mass spectrometry. Clin Biochem [Epub ahead of print 2010 Aug 4]. 4. Soldin SJ, Soukhova N, Janicic N, Jonklaas J, Soldin OP. The measurement of free thyroxine by isotope dilution tandem mass spectrometry. Clin Chim Acta 2005;358:113– 8. 5. Kahric-Janicic N, Soldin SJ, Soldin OP, West T, Gu J, Jonklaas J. Tandem mass spectrometry improves the accuracy of free thyroxine measurements during pregnancy. Thyroid 2007;17:303– 11. 6. 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. 7. Okabayashi T, Takeda K, Kawada M, Kubo Y, Nakamura S, Chikamori K, et al. Free thyroxine concentrations in serum measured by equilibrium dialysis in chronic renal failure. Clin Chem 1996; 42:1616 –20. 8. Baloch Z, Carayon P, Conte-Devolx B, Demers LM, Feldt-Rasmussen U, Henry JF, et al. Laboratory medicine practice guidelines: laboratory support

126 Clinical Chemistry 57:1 (2011)

9.

10.

11.

12.

13.

14.

for the diagnosis and monitoring of thyroid disease. Thyroid 2003;13:3–126. Nelson JC, Wang R, Asher DT, Wilcox RB. The nature of analogue-based free thyroxine estimates. Thyroid 2004;14:1030 – 6. Midgley JE, Christofides ND. Point: Legitimate and illegitimate tests of free-analyte assay function. Clin Chem 2009;55:439 – 41. Wilcox RB, Nelson JC. Counterpoint: Legitimate and illegitimate tests of free-analyte assay function: we need to identify the factors that influence free-analyte assay results. Clin Chem 2009;55:442– 4. d’Herbomez M, Forzy G, Gasser F, Massart C, Beaudonnet A, Sapin R. Clinical evaluation of nine free thyroxine assays: persistent problems in particular populations. Clin Chem Lab Med 2003; 41:942–7. Thienpont LM, Van Uytfanghe K, Beastall G, Faix JD, Ieiri T, Miller WG, et al. Report of the IFCC Working Group for Standardization of Thyroid Function Tests; part 2: Free thyroxine and free triiodothyronine. Clin Chem 2010;56: 912–20. Steele BW, Wang E, Klee GG, Thienpont LM, Soldin SJ, Sokoll LJ, et al. Analytic bias of thyroid function tests: analysis of a College of American Pathologists fresh frozen serum pool by 3900 clinical laboratories. Arch Pathol Lab Med 2005; 129:310 –7.

15. Bayer MF. Free thyroxine results are affected by albumin concentration and nonthyroidal illness. Clin Chim Acta 1983;130:391– 6. 16. Csako 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. 17. Larsen PR. Thyroid-pituitary interaction: feedback regulation of thyrotropin secretion by thyroid hormones. N Engl J Med 1982;306:23–32. 18. Soldin SJ, Cheng LL, Lam LY, Werner A, Le AD, Soldin OP. Comparison of FT4 with log TSH on the Abbott Architect ci8200: pediatric reference intervals for free thyroxine and thyroidstimulating hormone. Clin Chim Acta 2010; 411:250 –2. 19. Gu J, Soldin OP, Soldin SJ. Simultaneous quantification of free triiodothyronine and free thyroxine by isotope dilution tandem mass spectrometry. Clin Biochem 2007;40:1386 –91. 20. Jonklaas J, Kahric-Janicic N, Soldin OP, Soldin SJ. Correlations of free thyroid hormones measured by tandem mass spectrometry and immunoassay with thyroid-stimulating hormone across 4 patient populations. Clin Chem 2009; 55:1380 – 8. 21. Soldin OP, Jang M, Guo T, Soldin SJ. Pediatric reference intervals for free thyroxine and free

Evaluating the Relationship between FT4 and TSH

triiodothyronine. Thyroid 2009;19:699 –702. 22. McMillin GA, Travis JJ, Hunt JW. Direct measurement of free copper in serum or plasma ultrafiltrate. Am J Clin Pathol 2009;131:160 –5. 23. Pop VJ, Brouwers EP, Vader HL, Vulsma T, van Baar AL, de Vijlder JJ. Maternal hypothyroxinaemia during early pregnancy and subsequent child development: a 3-year follow-up study. Clin Endocrinol (Oxf) 2003;59:282– 8.

24. Gillam MP, Kopp P. Genetic defects in thyroid hormone synthesis. Curr Opin Pediatr 2001;13: 364 –72. 25. Jonklaas J, Soldin SJ. Tandem mass spectrometry as a novel tool for elucidating pituitarythyroid relationships. Thyroid 2008;18:1303– 11. 26. Thienpont LM, Van Uytfanghe K, Beastall G, Faix JD, Ieiri T, Miller WG, et al. Report of the IFCC

Working Group for Standardization of Thyroid Function Tests; part 1: Thyroid-stimulating hormone. Clin Chem 2010;56:902–11. 27. Donadio S, Pascual A, Thijssen JH, Ronin C. Feasibility study of new calibrators for thyroidstimulating hormone (TSH) immunoprocedures based on remodeling of recombinant TSH to mimic glycoforms circulating in patients with thyroid disorders. Clin Chem 2006;52:286 –97.

Clinical Chemistry 57:1 (2011) 127