Interference in Thyroid-Function Tests in Postpartum ... - CiteSeerX

of methods for the estimation of plasma low- and very-low-density lipoprotein cholesterol: the Lipid Research Clinice prevalence study. J Am Med Assoc 1986256:2372-7. 10. Hata Y, Nakajima K. Application of Friedewald’s LDL-cholesterol estimation formula to serum lipids in the Japanese population. Jpn Circ J 1986;50:1191-200. 11. Kuba K, Frantz ID. Validity of the Friedewald equation for the estimation of plasma low-density lipoprotein cholesterol (LDL-C) at various concentrations of triglycerides (Abstract]. Clin comparison

Chem 1987;33:896. 12. Lippi U, Graziani MS, Manzato F, Schinella M. The Friedewald formula: statistical acrobatics? (Letter]. Cliii Chem 1986;32:909-10. 13. Fujimoto WY. Friedewald’s LDL-cholesterol estimation formula in a Japanese American population. Jpn Circ J 1988;52: 604-6. 14. Rao A, Parker All, El-Sherom NA, Babelly MM. Calculation of low-density lipoproteun cholesterol with use of triglyceride/ cholesterol ratios in lipoproteins compared with other calculation methods. Cliii Chem 1988;34:2532-4. 15. Kurschinski DT, Dennen DA, Garcia M, Scanu AM. Plasma lipoprotein(a) and the Friedewald formula [Letter]. Clin Chem 1989;35:2156-7. 16. Lippi U, Graziani hIS, Schinella M. Assessing hypercholesterolemias by total and non-high-density lipoprotein cholesterol [Letter]. Cliii Chem 1990;36:1391. 17. G#{243}mez Gerique JA, Pedreflo J, Franco M. Valoraci#{243}n de Ia inexactitud de Ia estimaci#{244}n del colesterol de las lipoprotelnas de baja densidad mediante la formula de Friedewald. Clin Invest Arteriosclerosis 1990;2:59-62. 18. Gonzalez G”-Estrada M, Rodriguez Ferrer CE. Recalde Astarlea I, Montalvo Lahera E. Use of serum cholesterol/triglyceride ratio to discern for which individuals the Friedewald formula can be used confidently. Cliii Chem 1990;36:1673-5.

19. Krauss EM. Relationship of intermediate and low-density lipoprotein subspecies to risk of coronary artery disease. Am Heart J 1987;113:57-82. 20. Krauss RM, Lundgren FF, Williams EM, et al. Intermediatedensity lipoproteins and progression of coronary artery disease in hypercholesterolaemic men. Lancet 1987;ii:62-6. 21. Steinert G, Schwartz L, Shumak S, Poapat M. The association of increased levels of intermediate-density lipoproteins with smoking and with coronary artery disease. Circulation 1987;75:124-30. 22. Sentf M, Noguea X, Pedro-Botet J, Rubi#{233}s-Prat J. Intermediate density lipoproteins (IDL) in men with peripheral vascular disease [Abstract]. J Intern Med 1990;228(Suppl 733):65. 23. Nestel PJ,Fidge NH, Tan MR. Increased lipoprotein-remnant formation in chronic renal failure. N EngI J Med 1982;307:329-.33. 24. Bachorik PS. Collection of blood samples for lipoprotein analysis. Clin Chem 1982;28:1375-8. 25. Fontanals-Ferrer N, Serrat-Serrat J, Sorribas-Vivas A, Gonz#{225}lez-Garcf a C, Gonz#{225}lez-SastreF, GOmez-Gerique J. Quick method of determining lipoproteins, including those of intermediate density, in serum. Clin Chem 1988;34:1753-7. 26. Lopes-Virella MF, Stone P, Ellis 5, CoIwell JA. Cholesterol

determination in high-density lipoproteins separated by three

different methods. Cliii Chem 197723:882-4. 27. Warnick GE. Cheung MC, Albers JJ.Comparison of current methods for high-density lipoprotein cholesterol quantitation. Clin Chem 197925:596-604. 28. Ellefson RD. Variability in the VLDL-C/VLDL-T ratio and related errors in the estimation of LDL cholesterol (LDL-C) [Abstract]. Clin Chem 1985;31:948. 29. Demacker PN, Hi,jmans AG, Brenninkmeujer BJ, Jansen AdP, van’t Laar A. Five methods for determining low-density lipoprotein cholesterol compared. Cliii Chem 1984;30:1797-.800. 30. Lish HJ, Sailer S. Lipoprotein patterns in diet, sulphonylurea and insulin treated

diabetics.

Diabetologia 1981;20:118-22.

CLIN.CHEM.37/8, 1397-1400 (1991)

Interference in Thyroid-Function Tests in Postpartum Thyroiditis Rhys John,’

Sakinah

Othman,2

Arthur

B. Parkes,2 John H. Lazarus,2

Three women are described from a study of patients with postpartum thyroiditis whose sera gave spuriously increased concentrations of free thyroid hormone because of antibody binding of radiolabeled thyroxin (T4) and triiodothyronine (T3) analogs. All of the women showed increased serum concentrations of thyroid autoantibodies. The antibody binding of radiolabeled analogs and its effect on free T4 and free T3 assays disappeared by 48 weeks postpartum. Postpartum women who develop thyroid autoantibodies have -2% prevalence of increased binding of radiolabeled analogs, which can result in an interference in thyroid hormone assays involving T4 and T3 analogs. AdditIonal Keyphraeee: autoantibodies hormoneanalogs

radiolabeled thyroid

analytical error

Departments of’Medical Biochemistry and2 Medicine, University of Wales College of Medicine, Heath Park, Cardiff CF4 4XN,

U.K. Received January

23, 1991; accepted

May 28, 1991.

and ReginaLd Hall2

Since the first report of a patient whose serum contained a thyroxin (T4)-binding immunoglobulin (1), many patients have been identified with T4- or triiodothyronine (T3)-binding immunoglobulins that caused interference in radiounmunoassays of T4 or T3 (2). Spuriously high or low estimates of total thyroid hormone were observed, depending on the method used to separate antibodybound and free hormone concentrations in the RIA technique. The reported prevalence of T4- or T3-binding antibodies varies widely, depending on the patients studied and the methods for detection. Thus, although the interference in total T4 and total T3 RIAs is low (3), the incidence of interference is much higher in assays for free thyroxin (Fr.) and free triiodothyronine involving radiolabeled T4 or T3 analogs. In samples submitted to a routine clinical chemistry department for thyroid-function testing, the incidence of antibody interference in an

(Fr3)

3Nonstandard abbreviations: T4,thyroxin; T3, triiodothyronine; thyroxin; FT, free triiodothyronine; and H, thyrotropin.

Fr4, free

CLINICALCHEMISTRY,Vol.37, No.8, 1991 1397

analog Fr4 assay was 1 in 2460 (4). We report here the antibody interference we observed in three subjects who were part of a study of postpartum thyroiditis. Interference in an FT3 analog assay has been reported previously in an infant and her mother postpartum (5), but neither was part of the postpartum study reported here. Materials and Methods Subjects All women who attended the antenatal booking clinic at Caerphilly District Miners’ Hospital inMid Glamorgan between October 1987 and January 1990 were screened for the presence of thyroid autoantibodies as previously described (6). A study group of 148 women positive for thyroid antibody and 261 women negative for thyroid antibody were seen every four weeks for up to 12 months postpartum. Methods Concentrations of circulating FT4 and FT3 were measured by Amerlex-M (Amersham International plc, Amersham, Bucks., U.K.),analog methods with 1251-labeled analogs of T4 or T3;thyrotropin (TSH) was measured by an immunoradiometric assay (7). Both of the free thyroid hormone assays and the TSH assay involved the Kemtek Model 1000 automated sample processor and the Model 4000 automated RIA analyzer (both from Kemble Instrument Co., Ltd., Burgess Hill, Sussex, U.K.). We investigated antibody binding of labeled analogs as follows: for Amerlex-M FT4 and assays, we incubated 50 ML of serum with 250 pL of radiolabeled analog of T4 or T3 at 37 #{176}C for 1 h, then added 1 mL of a 200 g/L solution of Polyethylene Glycol 6000 (BDH Chemicals, Ltd., Poole, U.K.) and mixed the solution. After the samples stood for 10 mm at room temperature, we centrifuged them for 15 mm at 1000 x g, decanted the supernate, and counted the radioactivity in the antibody-bound fraction with a multidetector gamma system counter (Canberra Packard, Pangbourne, Berks., U.K.)for 60 s. Normal bindings of radiolabeled T4 and T3 analogs were 50 milli-int. unitsfL. By 28 weeks postpartum, serum TSH concentration was decreasing and we observed antibody interference in her VP4 measurement at this clinic visit. The data on antibody concentration and binding of radiolabeled T4 and T3 analogs are not complete for this patient because of insufficient serum samples (Figure 3). However, the available data showed an increase in concentration of thyroid microsomal antibody with only a slight increase in concentration of thyroglobulin antibody. At 28 weeks postpartum, we observed increased binding of radiolabeled T4 analog but not of radiolabeledT3 analog. PatIent 3

30

I

I 26

20 IS

10

I

20

20

15

IS

10

10

b

p

6

0

0

6

0 0

4

8

12

16

20 Weeks

24

28

32

36

postpartum

40

44

48

52

Fig. 2. Thyroidautoantibodles and binding of radlolabeled T4andT3 analogswith % radioactiveanalogprecipitated Symbols as In Fig. 1, plus bindingof radlolabeled T4 analog (C)

0

4

8

12

16

20

24

28

32

36

40

44

48

62

Weeks postpartum

Fig.3. Thyroidautoantibodies and bindingof radlolabeled T4and T3 analogs with % radioactive analogprecipitated Symbols as in Figs. 1 and 2

CLINICALCHEMISTRY,Vol.37, No.8, 1991 1399

Discussion Of the study group of 148 women positive for thyroid antibody, three (2%) developed antibody interference in analog methods for the measurement of VP4 and FT3 during the postpartum period. None of the women negative for thyroid antibody showed any interference in measurements of free thyroid hormone. In all three women studied, thyroid hormone antibodies were present in their serum by eight weeks postpartum. An increasing and then decreasing antibody concentration was observed during the study period. The increased serum binding of thyroid hormone analogs in patients 1 and 2 closely paralleled the changes in concentrations of thyroid antibody; for patient 3, only limited data were available. It is tempting

to speculate that thyroid hormone antibodies were responsible for the increased binding of radiolabeled ‘P4 or T3 analog in these patients. The reason why some patients, but not others, with increased concentrations of thyroid microsomal and thyroglobulin antibodies bind radiolabeled analogs is not at all obvious. Presumably, epitopes on some microsomal and thyroglobulin antibodies can be recognized by thyroid hormone analogs, so binding can occur. It is becoming well recognized that there is a relationship between antibodies to thyroid hormone and to thyroglobulin, which has been summarized recently by Sakata (9). In a patient he studied with Hashimoto thyroiditis who had antibodies to T4, the [IJT4 that bound to T4 antibody was displaced not only by unlabeled T4 but also by thyroglobulin from different species. However, similar results could be obtained in a rabbit serum containing T4 antibodies that had been raised by immunization with a T4-bovine serum albumin conjugate. Thyroxin and T3 are haptens and are thus unable to exert an antibody response in their free form; however, when they are coupled to thyroglobulin or albumin, antibodies to thyroid hormone are produced. In some cases there must be a very close similarity between the antibodies produced against thyroglobulin and those produced against T4 or T3 bound to albumin. The transient

increases labeled analogs mirrored

in antibody binding of radiothe transient nature of the

postpartum thyroiditis. Two of the three patients reported here had postpartum thyroiditis. Patient 1 became transiently hypothyroid, and patient 3, after a transient hyperthyroid phase, became transiently hypothyroid. Thyroxin treatment was started in patient 3 because she was thought to have more permanent hypothyroidism. Patient 2 did not have postpartum thyroiditis, but remained biochemically euthyroid despite

1400

CLINICALCHEMISTRY,Vol.37, No.8, 1991

the sometimes anomalous estimate of her VP4 and FT3 concentrations. The three patients, besides displaying three different biochemical states during their postpartum period, also displayed three different patterns of antibody interference. Patient 1, who was transiently hypothyroid, had interference in the estimate of her VP3 concentrations. Patient2,who remained euthyroid, had interference in the estimate of her VP4 and VP3 concentrations. Patient 3, who had hyperthyroidism followed by hypothyroidism, had interference in the estimate of her VP4 concentrations alone. Diagnostic confusion with inappropriate treatment can arise if antibody interference causing a spuriously high estimate of VP4 or VP3 concentrations is not recognized. In patient 1, the spuriously high estimate of FT3 concentration would probably have gone unnoticed because the woman became hypothyroid, and thus VP3 measurement would not normally have been requested. In patient 2, confusion may very well have been caused because VP4 and VP3 concentrations were both increased; if her TSH concentration had not been measured, a diagnosis of hyperthyroidism could well have been made. Because of the transient nature of the thyroid disorder and of antibody interference, it is important that women with postpartum thyroiditis receive full thyroid-function testing and be checked for possible antibody binding of analog tracers if inappropriately increased concentrations of free thyroid hormones are 5-.

References 1. Robbins J, Rail JE, Rawson RW. An unusual instance of thyroxine-binding by human serum gamma globulin. J Clin Endocrinol Metab 1956;16:573-9. 2. Sakata 5, Nakamura S, Miura K. Autoantibodies against thyroid hormones or iodothyronine. Ann Intern Med 1985; 103:579-89. 3. Li Caizi L, Benvenga 5, Battiato S, Santini F, Trimarchi F. Autoantibodies to thyroxin and triiodothyronine in the immunoglobulin G fraction of serum. Clin Chem 1988;34:2561-2. 4. John H, Henley H, Shankland D. Concentrations of free thyroxin and free triiodothyronine in serum of patients with thyroxinand triiodothyronine-binding autoantibodies. Clin Chem 1990;36:

470-3. 5. John R, Hughes I, Hall H. Congenital hypothyroidism with spuriously increased FF3 concentrations in infant and mother. Ann Clin Biochem

1990;27:85-7.

Fung HYM, KologluM, Collinson K, et al. Postpartum thyroid dysfunction in Mid Glamorgan. Br Med J 1988296:241-4. 7. John H, Jones MK. An automated immunoradiometric assay for 6.

human thyrotropin. Clin Chem 1984;30:1396-8. 8. Schardt CW, McLachlan SM, Matheson J, Rees Smith B. An enzyme-linked immunoassay for thyroid microsomal antibodies. J Immunol Methods 1982;55:155-68. 9. Sakata S. On the role of thyroglobulin as an antigen of thyroid hormone autoantibodies. Clin Chim Acta 1990;191:119-20.