Comparison of kodak amerlite FT4 and TSH-30 with T4 and TSH as first-line thyroid function tests

ClinicalBiochemistry,Vol. 29, No. 1, pp. 14, 1996 Copyright© 1996The CanadianSocietyof ClinicalChemists Printedin the USA. All rightsreserved 0009-9120/96$15.00 + .00 ELSEVIER

0009-9120(95)02009-B

Comparison of Kodak Amerlite FT 4 and TSH-30 With T4 and TSH as First-Line Thyroid Function Tests VAIJOO REGE, OLUSEGUN MOJIMINIYI, HERVEY WILCOX, and JEFFREY BARRON Department of Chemical Pathology, St Helier Hospital, Carshalton, Surrey, SM5 1AA, UK Objectives: To evaluate the effect of test automation and a change in strategy for thyroid function tests (TFT) on personnel needs and turn-around time. The first-line TFT were changed from T4 and TSH to FT4 and TSH-30. Design and Methods: Samples received for TFT from 357 randomly selected patients were analyzed by RIA for T 4, and by IRMA for TSH as first-line tests. FTz and TBG were requested as back-up tests when indicated. Patients were classified on the basis of these results and the clinical information received. All the samples were reanalyzed for FT4 and TSH on the Amerlite Processing Center, which is a batch, semiautomated immunoassay system. The thyroid status of the patients was compared using the two protocols and available clinical data. Results: There was good correlation between TSH-IRMA and TSH-30 in the 160 patients classified as euthyroid (r = 0.956; p < 0,001) and no euthyroid patient was reclassified with the new strategy. In 21 patients with borderline raised TSH-IRMA, FT4 was found to be low in only 2. All 11 patients classified as hypothyroid had TSH results greater than 10 mU/L and all except 2 patients had FT4 less than 11 nmol/L. The status of 21 hyperthyroid as well as 40 patients on carbimazole could be determined biochemically on the basis ¢,f agreement between both the FT4 and TSH-30 results. FT3 was only required if the FT4 and TSH-30 results were not in agreement. In 42 patients on T4 therapy, adequacy of replacement was assessed better using FT4 and TSH-30. No patient required backup testing with TBG to determine thyroid status using the new testing protocol. The change in TFT protocol reduced the 95% turn-around time from 3 days to 1 day. Conclusion: The introduction of FT4 and TSH-30 as first-line TFT improved the turn-around time for TFT, resulted in 25% reduction in personnel requirements, 60% reduction in FT3 assays, and discontinuation of TBG assay.

KEY WORDS: thyroid function tests; TSH; FT4; FT3; T4; t u r n - a r o u n d time.

are used as primary tests. In this laboratory, to save on staff time during sample handling, assay, test cascade, and review we used both T4 and TSH as first-line tests. This strategy for thyroid function tests (TFT) also required selected backup measurement of free triodothyronine (FT3) and thyroid binding globulin (TBG) before full diagnostic classification could be given in a number of patients. The additional tests invariably required more stafftime, increased turn-around times, and were expensive. Development of automation in immunoassay and the availability of commercial immunometric assays for free thyroxine (FT4), and especially TSH, which showed marked improvement in both sensitivity and specificity over the existing radiolabelled assays, resulted in a review of the strategy for TFT. This study examined the effect of automation on staff and turn-around times, and compared the diagnostic utility of the previous strategy (using T4 and TSH as first-line tests with selected backup testing) with the new strategy using FT4 and ultrasensitive third generation TSH assays in patients with various abnormalities of thyroid function. The use of FT 3 as a backup test in the latter strategy was also reviewed. Patients and methods

Correspondence: Dr. J.L. Barron. M a n u s c r i p t received: March 21, 1995; revised and accepted: J u n e 7, 1995. Presented at the 5th I n t e r n a t i o n a l Congress of Automation a n d New Technologies i n Clinical Laboratories held in J e r u s a l e m , J u n e 1994.

Samples received for TFT from 357 randomly selected patients were analyzed by radioimmunoassay (RIA) for T4, and by immunoradiometric (IRMA) for TSH (Immunodiagnostic Systems Ltd, Boldon, Tyne and Wear, UK) as first-line tests. FT 3 and TBG were requested as backup tests when indicated. On the basis of these results and the clinical information received, the patients were grouped as euthyroid (160 patients), hypothyroid (10 patients), hyperthyroid (25 patients), and borderline raised TSH, range 4.4-10 mU/L (21 patients). Additional backup testing to determine the biochemical thyroid status was required in 50 patients, including 9 patients with TBG abnormality (with T4 values above or below the reference range), and 41 patients with raised T4 but judged to be euthyroid on the basis of normal FT 3

CLINICAL BIOCHEMISTRY, VOLUME 29, FEBRUARY 1996

1

Introduction

here are a number of approaches to thyroid funcT tion testing; either thyroxine (T4) or thyroid stimulating hormone (TSH) or a combination of both

R E G E E T AL.

and TSH-IRMA results. Patients with existing thyroid disease on treatment included 40 patients taking carbimazole and 51 patients on T4 replacement. The samples for all the patients were reanalyzed for FT 4 and TSH on the Amerlite Processing Centre (Kodak Ltd, UK) which is a batch, semiautomated immunoassay system. TSH-30 requires a 30-min incubation, and is a third generation assay with an analytical sensitivity of 0.005 mU/L (1). The status of the patients was reviewed and the diagnostic information compared using the previous and new testing strategies. Results

There was good correlation between TSH-IRMA and TSH-30 in the patients classified as euthyroid (r = 0.956; p < 0.001). However, the TSH-30 assay gave consistently lower results resulting in a change in the reference range from 0.2-4.3 mU/L (TSHIRMA) to 0.15-3.5 mU/L (TSH-30). No euthyroid patient was reclassified on testing with the new strategy. In 21 patients with borderline raised TSHIRMA (4.4-10 mU/L), the range of TSH-30 values was 2.7-8.9 mU/L and, in 2 of these patients, the FT 4 was low. All patients classified as hypothyroid had TSH-30 and TSH-IRMA results greater than 10 mU/L and all except 2 patients (FT4 = 14 and 16 nmol/L) had FT4 less than 11 nmol/L. Four of the patients classified as hypothyroid, on the basis of the TSH and clinical information, had normal T4 values. Two patients with high T4 had high TBG and 7 patients with low T4 had low TBG, but all 9 patients had normal FT 4 and TSH-30. Figure 1 shows the good correlation between TSH-30 and TSH-IRMA in the 41 patients with high T4 whose euthyroid status was judged on the basis of normal TSH and FT 3 results. Thirty-six of these 41 patients had normal FT 4 and TSH-30 values; the FT4 was normal in 3

patients with low TSH-30 and high in 2 with normal TSH-30. All 25 patients classified as hyperthyroid had raised FT 4 and FT~ and all except 2 had TSH-30 less than 0.05 mU/L, but the TSH-IRMA was less than 0.08 mU/L in only 3 patients. Table 1 summarizes the performance characteristics of each test and the numbers of euthyroid, hypothyroid, and hyperthyroid patients correctly classified. Of 40 patients taking carbimazole, 29 had their thyroid status assigned biochemically on the basis of agreement between both the FT 4 and TSH-30 results (euthyroid 17, hypothyroid 5, hyperthyroid 4). In the remaining 11 patients, there was disagreement between the FT 4 and TSH-30 results, but the FT3 and FT4 results were in agreement. Using the previous testing protocol for T 4 and TSH, all patients taking carbimazole required FT 3 to assist in the biochemical assignment of thyroid status. Patients taking T 4 replacement therapy were assigned by T4 and TSH into 3 groups: adequate T 4 replacement (n = 33); TSH not normalized, suggest review dose or compliance (n = 11); biochemically over-replaced, suggest review T 4 dose (n = 6). There was no reclassification with the new strategy. Using the previous TFT strategy, the turn-around times for TFT were: 1 day (17%), 3 days (96%), and 15 days (100%); these were reduced to 1 day (95%) and 5 days (100%) using FT4, TSH-30, and FT 3. With the introduction of FT4 and TSH-30 as firstline TFT, the turn-around time has improved with 25% reduction in personnel, 60% reduction in FT 3 assays, and discontinuation of the TBG assay. Discussion

The increased awareness of the prevalence of thyroid disease with the need to investigate, especially in the elderly, has increased the workload in clinical laboratories. There is increasing pressure on labo-

3.5 3 2.5

.< 2 1.5 151 ["

l 0.5 0 0

0.5

1

1.5

2

2.5

TSH-30 mU/L Figure 1 -- Correlation between TSH-30 and TSH-IRMA in patients (n = 41) with total T 4 judged to be euthyroid on the basis of TSH, FT3, and available clinical data. 2

CLINICAL BIOCHEMISTRY, V O L U M E 29, F E B R U A R Y 1996

COMPARISON OF KODAK AMERLITE FT4 AND TSH-30 WITH T4 AND TSH TABLE 1 Perform:race Characteristics of Tests of Thyroid Function in Terms of Sensitivity and Specificity; Patients on Therapy for Thyroid Disease are Excluded Clinical Status Euthyroid True-positive True-negative False-positive False-negative Sensitivil;y Specificity Hypothyroid True-positive True-negative False-positive False-negative Sensitivity Specificity Hyperthyroid True-positive True-negative False-positive False-negative Sensitivity Specificity

TSH-IRMA

T4

TSH-30

FT4

210 29 6 21 91% 83%

181 29 6 50 77% 89%

2O9 35 0 22 92% 100%

227 33 2 4 98% 94%

10 235 21 0 100% 92%

6 249 7 4 60% 97%

10 237 19 0 100% 93%

8 254 2 2 80% 99%

19 241 0 6 76% (TSH <0.2) 12% (TSH <0.08) 100% (TSH <0.2)

23 198 43 2 92%

25 238 3 0 100% (TSH <0.15) 92% (TSH <0.05) 99% (TSH <0.15)

25 239 2 0 100%

82%

99%

ratories to produce accurate diagnostic information at minimum cost and shorter turn-around times. Various strategies for TFT have been proposed ( 2 4) and the availability of sensitive TSH assays has led to adoption of this analyte as a first-line test of thyroid function in some centers (2,5). The problems associated with the adoption of this latter strategy include low concentrations of TSH in patients with nonthyroidal illness and pituitary disease (6-8) and the fact that a single laboratory test cannot be used to categorize patients with thyroid disease (9). Table 1 shows that there is no one test with 100% sensitivity and specificity. Adopting a new TFT .,~trategy using FT 4 and TSH30 provided a rapid assessment of thyroid status in the different groups of patients in this study. Fewer samples required second-line testing to establish the thyroid status. For example, 9 patients with abnormalities in TBG concentration had normal FT4 and TSH-30 results and were classified as euthyroid on the basis of these results, without the need to do an additional test (TBG or FT 3) to investigate the cause of high or low T4 results. TSH-30 was more useful for the diagnosis of hyperthyroidism especially in the early stages, because all but 2 of the :patients had TSH-30 <0.05 mU/L. In contrast, only 3 patients in this group had TSH-IRMA <0.08 mU/L. The analytical sensitivity of the TSH-30 assay and its ability to reliably distinguish between euthyroid and hyperthyroid patients (1) makes it ideal for rapid screening of patients with clinical suspicion of hyperthyroidism. The good agreement between FT3 and FT 4 indicates that either assay could be used with resultant reduction in the FT s workload by 60%. All patients

taking carbimazole therapy had F T s estimations using T4 and T S H - I R M A as first-linetests, but the new strategy shows that 74% of patients could have their therapy monitored on the basis of agreement between FT 4 and TSH-30 results without the need for FT 3 assays. However, when the TSH-30 and F T 4 are not in agreement, an additional F T 3 is useful in deciding whether or not to alter dosage. In monitoring patients on thyroxine replacement therapy, all patients were classified to the same group by both T F T strategies.There was good agreement between T S H - I R M A and TSH-30 in 11 patients in w h o m replacement was not adequate and in those with suppressed TSH. Five patients in the latter group had FT4 well above the normal range, were considered overtreated (10) and the report suggested a review of the dosage. In conclusion, automation of F T 4 and TSH-30 on the Amerlite Processing Centre using a combination of the 2 tests has simplified the strategy of TFT, reduced the turnaround time and led to a reduction in personnel. As the TSH-30 assay requires only 30rain incubation, its use could provide a rapid assessment of thyroid status in centers where T S H is used as the first-linetest. However, as a further test is often required in the diagnosis and treatment of hyperthyroidism, hypothyroidism, and patients with nonthyroid illness,a combination of FT4 and TSH30 is recommended as first-linetests.

CLINICAL BIOCHEMISTRY, V O L U M E 29, FEBRUARY 1996

3

References

i. Wilkinson E, Rae WH, Thomson KJT, Tort AD, Spencer CA, Beckett GJ. Chemiluminescent third-generation assay (Amerlite TSH-30) of thyroid-stimu-

REGE E T

lating hormone in serum or plasma assessed. Clin Chem 1993; 39: 2166-73.

2. Caldwell G, Gow SM, Sweeting VM, Kellet HA, Beckett GJ, Seth J, Toft AD. A new strategy for thyroid function testing. Lancet 1985; I: 1117-9. 3. Klee GG, Hay ID. Assessment of sensitive thyrotropin assays for an expanded role in thyroid function testing: Proposed criteria for analytic performance and clinical utility. J Clin Endocrinol Metab 1987; 64: 461-71. 4. de los Santos ET, Starich GH, Mazzaferri MD. Sensitivity, specificity, and cost-effectiveness of the sensitive thyrotropin assay in the diagnosis of thyroid disease in ambulatory patients. Arch Intern Med 1989; 149: 526-32. 5. Bannister P, Shapiro L, Faye S, Bolton R. A simplified scheme for the investigation of thyroid function in chronic liver disease. A n n Clin Biochem 1988; 25: 373-5.

4

AL.

6. Bayer MF, Macoviak JA, McDougal IR. Diagnostic performance of sensitive measurements of serum thyrotropin during severe non-thyroidal illness: their role in the diagnosis of hyperthyroidism. Clin Chem 1987; 33: 2178-84. 7. Ratnaike S, Goodwin M, Deam D. Anomalous thyrotropin values. Clin Chem 1987; 33: 1213-4. 8. Spencer C, Elgen A, Shen D et al. Specificity of sensitive assays of thyrotropin (TSH) used to screen for thyroid disease in hospitalized patients. Clin Chem 1987; 33: 1391-6. 9. Taimela E, T/ihtel~i R, Koskinen Pet al. Ability of two new thyrotropin (TSH) assays to separate hyperthyroid patients from euthyroid patients with low TSH. Clin Chem 1994; 40: 101-5. 10. Surks MI, Chopra IJ, Mariash CN, Nicoloff JT, Solomon DH. American thyroid association guidelines for use of laboratory tests in thyroid disorders. J A M A 1990; 263: 1529-32.

CLINICALBIOCHEMISTRY,VOLUME29, FEBRUARY 1996