Creatine Kinase Determination: A European

4 downloads 0 Views 2MB Size Report
Summary: New homogeneous enzyme immunoassays for the determination of ... The CEDIA® assays are based on the cloned enzyme donor immunoassay ...
H0rder et al.: Creatine kinase determination with Reflotron^ CK

691

Eur. J. Clin. Chem. Clin. Biochem. Vol. 29, 1991, pp. 691-696 © 1991 Walter de Gruyter & Co. Berlin · New York

Creatine Kinase Determination: A European Evaluation of the Creatine Kinase Determination in Serum, Plasma and Whole Blood with the Reflotron® System By M. Horder, P. J. Jorgensen Odense University Hospital, Odense, Denmark /. C. M. Hajkenscheid^ Sint-Radboudziekenhuis, Nijmegen, The Netherlands C. A. Carstensen Evaluierung Diagnostica Boehringer Mannheim GmbH, Mannheim, Germany C. Bachmann Laboratoire Central de Chimie Clinique, Centre Hospitalier Universitaire Vaudois, Lausanne, Switzerland K. Bauer, Christine Neuwald Kaiser-Franz-Josef-Spital der Stadt Wien, Zentrallabor, Vienna, Austria S. B. Rosalki, A. Ying Foo Royal Free Hospital and School of Mediane, Dept. of Chemical Pathology and Human Metabolism, London, Great Britain W. Vogt Deutsches Herzzentrum München des Freistaates Bayern, Institut für Klinische Chemie und Laboratoriumsmedizin, Munich, Germany (Received April 16/Jiily 23, 1991)

Summary: We evaluated a new dry-jreagent carrier System for the determination of creatine kinase (EC 2.7.3.2) activity, Reflotrpn® CK, with special attention to analytical performance with whole blood. We found a gopd withinseries imprecision. The median coefficient of Variation was 3.1% for Reflotron® CK (blood, serum and plasma) and 0.9% for the automatic analysers (serum and plasma only). The betweendays imprecision with Reflotron® CK (median CV: ^ 3%) was similar to that for the comparison method on different analysers. Fresh samples of human blood, plasma and serum were examined by Reflotron® CK and by a N-acetylcysteine activated creatine kinase method in six different clinical laboratories and in the Evaluation Department of Boehringer Mannheim GmbH. The correlation between these methods was excellent (r > 0.99), the median systematic deviation (bias) for all samples being smaller than —5%. Eur. J. Clin. Chem. Clin. Biochem. / Vol. 29,1991 / No. 10

H0rdcr et al.: Creatine kinase determination with Reflotron® CK

692

Haematocrits between 0.25 and 0.50, haemolysis up to 6 g/l haemoglobin, and icteric samples with bilirubin concentrations up to 0.2 g/l showed no interference. No drug in therapeutic concentration was found to affect the Reflotron® CK results; ascorbic acid, calcium dobesilate and sulphamethoxazole lowered the val es only when present in high concentrations. Reflotron® CK may be considered s a suitable alternative for decentralized testing sites, especially in situations where creatine kinase results are needed quickly. *r

Creatine phosphate 4-

Introduction

The diagnosis of myocardial infarction is based on clinical history, electrocardiogram and determination of enzymes released from damaged myocardial cells. Especially in patients with equivocal electrocardiographic changes, the biochemical confirmation or exclusion of a suspected myocardial infarction is of major importance (l, 2). For this purpose creatine kinase (ATP: creatine N-phosphotransferase, EC 2.7.3.2)1) is usually measured. We report here on the evaluation of Reflotron® CK, a new dry reagent carrier for the determination of the catalytic activity concentration of creatine kinase in different samples, using Reflotron®, a compact reflectance photometer (3). We have given special attention to the analytical performance of Reflotron® CK with whole blood samples, and to possible sources of interference.

Glycerol

Creatine kinase

creatme

Olycerol ki

Glycerol-3-P + O2

ADp

Glycerol phosphate oxidase Dihydroxyacetone : ~*~ phosphate + H2O2

H202 + Indicatorrcd.

Peroxidase

» Indicatorox. + H2O

Reflotron® CK is calibrated with the Niacetylcysteine-activated creatine kinase method, the calibration procedure being similar to that used with other enzymatic Reflotron® pafameters (5). The structure of the reagent carrier is shown in figure l . Protective mesh —·

Indicator film Magnetic code 1

— Separating layer — Transport layer

1

• Creatine phosphate/ N-acetylcysteine tissue

Materials and Methods Procedure The Reflotron® System (Boehringer Mannheim GmbH, Mannheim, Germany) has been described previously (3). The instruments were operated according to the manufacturer's instructions. In the Reflotron® CK assay (4), creatine kinase1) is activated by N-acetylcysteine in a preliminary reaction (within 80 s). The activated creatine kinase transfers the phosphate from creatine phosphate to adenosine diphosphate. In a second reaction Step, the adenosine triphosphate formed reacts with glycerol in the presence of glycerol kinase1) to give glycerol 3-phosphate. In the presence of glycerophosphate oxidase1), the latter reacts with oxygen to generate dihydroxyacetone phosphate and hydrogen peroxide. The resulting hydrogen peroxide converts the colourless form of the diarylimidazole indicator into the blue form, provided horse-radish peroxidase1) is present. The formation of the dye is monitored by a six-point measurement at 642 nm. The result is displayed approx. 190s after sample application.

Enzymes: Adenylate kinase Creatine kinase Glycerol kinase Glycerol phosphate oxidase Peroxidase

— Carrier foil Fig. 1. Structure of Reflotron® CK. Reagents contained in the test area are: creatine phosphate 116 μ& glycerol kinase > 1.0 ; glycerol 4.4 μg glycerophosphate oxidase > 0.09 U peroxidase > 1.7 U ascorbate oxidase > 0.41 U; ADP 4.0 μ& diadenosine pentaphosphate 1.0 μg; 2-(3,5-di-^r/-butyl-4-hydroxyphenyl)-4-(5)-(9-juiolidino)-5=(4)methyl-(lH)-imidazole (indicator) 9.4 μ^ N-acetylcysteine 11.3 μg; EGTA 16 mg; KH2PO4/Na2HPO4 buffer, pH 6.2; magnesium aspartate 29 μ^ Comparison method

(EC (EC (EC (EC (EC

2.7.4.3) 2.7.3.2) 2.7.1.30) 1.1.3.21) 1.11.1.7)

The comparison method t ized was the optimized Standard method according to the recommendations of the German Society for Clinical Chemistry, CK NAC-aetivated .using reagents from Boehringer Manniheim GmbH, Mannheim, and E. Merck, Darmstadt, Germany (6—*8); unless otherwise stated, Eur. J, Clin. Chem. Clin. Biochem. / Vol. 29,1991 / No. 10

Horder et al.: Creatine kinase determination with Reflotron^ CK the reaction lemperature was 37 °C. Measurements were made according to Ihe manufacturer's instructions, by either manual or automated procedures (tab. 3). Evalualion protocol Imprecision studies and method comparison experiments were carried out collaboratively in six elinical laboratories. In addition, interference testing was carried out at Ihe Evahiation Department of Boehringer Mannheim GmbH. For the within-series imprecision study, each Jaboratory performed 18 series of tenfold determinations with both control sera (Precinorm® U and Precipath® U), 8 series with heparin blood, 10 series with heparin plasma and two series with serum. For the between-day imprecision study, the two control sera were examined daily for inore than 7 days in each laboratory. Serum and plasma measurements were made with both analytical Systems. Blood was measured only with Reflotron® CK.

693

3%. Remarkably, different creatine kinase activity concentrations had no influence on the corresponding CV. The automated CK-NAC comparison methods showed median CVs of about \ %. The between-day imprecision of the reagent carrier System showed a median CV of < 4.2%, a value sirhilar to that of the automated CK-NAC comparison methods (median CV < 2.6%) (tab. 2). Method comparison

The method comparison experiments were carried out at all seven evaluation sites; six of these used fresh heparin-treated venous blood, plasma or serum frona hospitalized patients, while one evaluation site used fresh capillary blood for the Reflotron® CK and heparin-treated capillary plasma for the comparison method. The capillary blood was drawn frpm Sports students after intensive training. For the collection, Microtainer® Tubes (Becton Dickinson, Rutherford, NJ) containing lithium heparinate äs an anticoagulant were used. After skin puncture of the flngertip (9), the first drop of blood was discarded, then about 300 of blood were taken.

In table 3, the correlation results of the method comparison experiments are summarized. In all comparisons using blood samples for Reflotron® CK and plasma for CK-NAC method, the correlation coefficients exceeded 0.994. Also, for the method comparison with capillary blood samples, the correlation coefficient (0.998) was äs high äs for the venous blood 2500 r

2000

All linear regressions were calculated according to the recornmendations of Passing & Bablok (10). For the haematocrit interfererice study 4 laboratories measured creatine kinase with the Reflotron® in a total of 434 heparin blood samples and the corresponding plasma samples. Analytical recovery of the Reflotron® CK for blood samples was calculated relative to Reflotron® CK results for plasma. Recovery was assessed for 7 classes of haematocrit, each differing by an interval of 0.05. The protocols for the studies of interference by bilirubin, haemoglobin and drugs have been described elsewhere (13).

Results and Discussion

Imprecision

500 1000 1500 2000 Creatine kinase (N-acetylcysteine activated, 37°C, capillary heparin plasma) [U/l]

The within-series impreeision for Reflotron® CK with control sera (Precinorm® and Precipath®) and with blood, serum and plasma samples is shown in table l. For all materials we föünd median CVs of about

2500

Fig. 2. Method comparison: Reflotron® CK vs CK NAC-activated determinations for 106 blood samples: y = 0.964 -8.7 [U/l], r = 0.998

Tab. 1. Results of imprecision study: within-series imprecision (N = 10) Reflotrpn® CK CV [%]1 Assigned value Min. [U/l]

Sample material Precinorm®, Precipath®, Heparin blood, Heparin plasma, Serum*, a

18 series 18 series 8 series 10 series 2 series

175 447 75-1303 78-1292 56- 544

1.3 1.6 1.8 1.5 3.1

25 °C

Eur. J. Clin. Chem. Clin. Biochem,/ Vol. 29,1991 / No. 10

Median

Max.

2.9 2.9 2.7 3.4 3.5

5.2 4.8 3.8 5.1 5.0

CK-NAC method CV [%]1 Assigned vaiue Min. [U/l] 198 491

— 85-1189 53- 493

0.3 0.3 — 0.3 0:5

Median

Max.

1.0 0.9 — 0.9 1.0

2.2 1.8 — 2.8 1.0

Herder et al.: Creatine kinase determination with Reflotron* CK

694

Tab. 2. Quality control. Belween-day imprecision with control sera Comparison method Bvaluator (syslem) No. ofdays

Target value [U/l] 37 °C [U/l] 25 °C

Precinorm® U 198 81.2

Precipath® U 491 201

Bias [%] Precinorm® U

Precipath® U

1. (Encore®) n = 10 d

x [U/l] CV [%]

187 5.0

489 1.5

-^5.6 —

- 0.4 —

2. (Hitachi 704) n = 7d

χ [U/l] CV [%]

188 2.0

463 1.5

-5.1 —

- 5.7 —

3. (Hitachi 717) n = 8d

χ [U/l] CV [%]

180 2.5

439 1.7

-9.1 —

-10.6 —

4. (Cobas® Βίο) n = 37 d

χ [U/l] CV [%]

187 2.4

459 2.6

-5.6 ·=-

- 6.5 -~

5. (Multistat®) n = 9d

χ [U/l] CV [%]

208 3.4

501 3.1

+ 5.1 -^

+ 2.0 —

6. (Greiner® 450) n = 10 d, 25 °C

χ [U/l] CV [%]

78.3 4.9

189 3.4

-3.5 —

- 6.1 —

Reflotron® CK

Evaluator No. of days1

Target value [U/l] 37 °C [U/l] 25 °C

Precinorm® U 175 71.7

Precipath® U 447 182

Bias [%] Precmonn®U

Precipath® U

1. n = 10 d

x[U/l] CV [%]

179 2.6

451 3.3

4-2.3 —

+0.9 —

2. n = 7 d

χ [U/l] CV [%]

181 3.4

453 1.9

+ 3.4 -=-

+ 1.3 -

3. n =

8d

χ [U/l] CV [%]

177 3.7

447 5.6

+ 1.1 —

+0 —

4. n = 37 d

χ [U/l] CV [%]

173 4.0

430 4.3

-1.1 —

-3.8 —

5. n = 9 d

χ [U/l] CV [%]

180 3.6

454 4.2

+ 2.9 —

+ 1.6 —

6. n = 10 d, 25 °C

χ [U/l] CV [%]

73.0 4.4

182 4.2

+ 1.7 —

-0.7 —

Tab. 3. Correlation of method comparison results Comparison method xa CK, NAC activated Instrument, sample

Sample material, method y

Activity r nge [U/l]

No. of samples

Slope

y-Intercept [U/l]

r

Encore®0, Plasma Encore®c, Plasma Hitachi 704d, Plasma Hitachi 71 7d, Plasma Cobas® Bioc, Plasma Multistat®f, Plasma Greiner® 4508, Serumh Uvikon® 61 01, heparinized capillary plasma

Blood Plasma Blood Plasma Blood Blood Serum Native capillary blood

21-1650 21-1650 26-1650 26-1760 25-1860 26-1700 7- 780 60-2120

100 100 93 125 141 100 111 106

0.957 0.950 0.981 1.088 0.944 0.905 1.029 0.964

3.7 5.4 - 2.6 -11.7 - 1.1 - 0.9 - 2.8 - 8.7

0.998 0.999 0.996 0.988 0.994 0.996 0.988 0.998

a b 0 d c f h

'"

Method y = Reflotron® CK As determined by method χ Baker Instruments, Allentown, PA 18001, USA Boehringer Mannheim GmbH, Mannheim, Germany Hoffmann-La R che & Co. AG, Basel, Switzerland IL, Instrumentation Laboratories, Lexington MA 02173-3190, USA Greiner, Langenthal, Switzerland Measuring temperature 25 °C Kontron/Tegimenta AG, Rotkreuz, Switzerland

• « * >

Eur. J. Clin. Chem. Clin. Biochern. / VoL 29,1991 / No. 10

H0rder et al.: Creatine kinase determination with Reflotron® CK

695

samples, proving that capillary blood is an excellent sample material for Reflotron® CK (fig. 2). The ability to measure creatine kinase activity on capillary blood is an advantage, especially in emergency cases where quick results are needed.

a reaction temperature of 37 °C. A temperature conversion factor of 0.41 can be used to convert the Reflotron® results for a 25 °C reaction temperature. The validity of this factor is confirmed by the good agreement for the method comparison at 25 °C.

The regression analysis for the method comparisons with blood samples yielded negligible intercepts, and the slopes ranged from 0.905 to 0.981 with a median of 0.96 (tab. 3).

Haematocrit interference

With plasma and serum the slopes of the regression equations where slightly higher, ranging from 0.950 to 1.088 (tab. 3). In total we examined 540 blood samples, of which 284 had an elevated creatine kinase activity [> 150 U/l (37 °C)]. All samples with elevated activities showed relative biases, compared with the comparison method, in which less than 20% showed a bias, and 90% of all blood samples with elevated activities had a bias smaller than 15%. In accordance with the results on quality control material the results of the method comparisons confirm the good accuracy of Reflotron® CK.

Table 4 shows the effect of haematocrit Variation on creatine kinase recovery for all 534 blood samples. Variation of haematocrit values between 0.25 and 0.50 is without effect on the measured creatine kinase activity on the Reflotron® CK. Blood samples with haematocrit values higher than 0.50 are rare, and in such cases the plasma should be used. Haemolysis It is known that adenylate kinase (myokinase, EC 2.7.4.3) catalyses the reversible conversion of ADP to ATP and AMP. The activity of adenylate kinase in Tab. 4. Results of haematocrit interference study Haematocrit3

No. of samples

Recovery6

These results show that Reflotron® CK gives values identical to the CK-NAC activated method, irrespective of the source of the blood samples with their isoenzyines. We obtained samples from general hospitals, heart centres and a College of physical education.

0.25 0.30 0.35 0.40 0.45 0.50

14 27 108 142 101 40

100 99 98 100 98 97

Measuring r nge

a b

(110)

0.55

In the method comparison experiments, sample with a catalytic concentration > 1500 U/l were marked on the display of the Reflotron® Instrument with an asterisk, indicating a non-line r measurement. For samples with > 1900 U/l Reflotron® usually displays no result, but reepmmends a dilution of the sample (according to the manufact rer 1 + 1 by a serum with low creatine kinase activity). In cases where a result was displayed (with pr without asterisk) a good agreement between Reflotrpn® CK and the CK-NAC method was obtained. A dilutipft with NaCl (153 mmol/1) is not possible, because the changed matrix leads to elevated values. Temperature conversion factor One evaluation site (tab. 3) compared the Reflotron® CK and CK-NAC method using a reaction temperature of 25 °C for the CK-NAC method. With the Reflotron® System all measurements are performed at Eur. J. Clin. Chem. Clin. Biochem. / Vol. 29,1991 / No. 10

Average of class. Mean recovery of class.

Tab. 5. Results of haemoglobin and bilirubin interference studies Interferent Haemoglobin [g/l] 1.0 3.0 4.5 6.0 Bilirubin [g/l] 0.008 0.02 0.04 0.1 0.2

Recovery51 ΓΟΛΊ 99 98 98 98 100 99.5 98 103 98

For haemoglobin interference: [(Reflotron® CK activity measured — original Reflotron® CK activity)/original Reflotron® CK activity] · 100. For bilirubin interference: [(Reflptron® CK activity measured - CK-NAQ/CK-NAC] •100.

696

H0rder et aL: Creatine kinase detcrmination with Reflotron® CK

serum lies between 0 — 50 U/l (25 °C) and is markedly greater in haemolysed specimens. This could result in an increase of apparent crea ne kinase activity proportionale to the the degree of haemolysis. Adenlyate kinase may be inhibited by several substances (6). For Reflotron® CK diadenosine pentaphosphate was chosen, and s shown in table 5 it completely inhibited the adenlyate kinase of haemolysed samples; in the presence of the inhibitor, haemolysis up to a haemoglobin concentration of 6 g/l did not affect the Reflotron® CK results. The findings of Szasz (11, 12) seem to be valid only for "wet chemistry". Bilirubin interference There was no interference by bilirubin up to the maximum concentration tested (0.2 g/l, 340 μηιοΐ/ΐ, tab. 5).

Drug interference In the in vitro drug interference study, 32 drugs (13) were examined in therapeutic and toxic concentrations. Only the high concentrations of ascorbic acid (recovery 70%), calcium dobesilate (85%) arid stilphamethoxazole (80%) showed an influence on the recovery with Reflotron® CK. Therapeutic drug concentrations did not affect the results. Conclusions

Reflotron® CK allows rapid and reliable measurement of creatine kinase in serum, plasma and blood samples. Precision, accuracy and interference testing show its suitability s a diagnostic tpol for creatine kinase determination, especially in emergencies when an immediate result is required.

References 1. Collinson, R O., Rosalki, S. B., Fiather, M., Wolman, R. & Evens, T. (l 988) Early diagnosis of myocardial infarction by times sequential enzyrae measurements. Ann. Clin. Biochem. 25, 376-382. 2. Matthews, S. B. & Davies, J. (1987) Clinical detection of myocardial infarction using CK-MB compared with total CK and AST. Ann. Clin. Biochem. 24, 233-235. 3. Steinhausen, R. L. & Price, C. R (1985) Principle and practice of dry chemistry Systems. In: Recent advances in clinical biochemistry. Vol. 3. (Price, C. P. & Alberti, K. G. M., eds.) Edinburgh: Churchill Livingstone, pp. 273-296. 4. Braun, H. R, Deneke, U, Rittersdorf, W. & Bartl, K. (1990) Construction and function of a new Reflotron® test for the determination of creatine kinase. Clin. Chem. 36, 1129. 5. Price, C. P. & Koller, R U. (1988) A multicenter study of the new Reflotron® System for the measurement of urea, glucose, triacylglycerols, cholesterol, gamma-glutarnyltransferase and hemoglobin. J. Clin. Chem. Clin. Biochem. 2(5,233-250. 6. Szasz, G., Gruber, W. & Bernt, E. (1976) Creatine Kinase in Serum: 1. Determination of Optimum Reaction Conditions. Clin. Chem. 22, 650-656. 7. Gruber, W. (1978) Inhibition of creatine kinase activity by Ca2+ and reversing effect of EDTA. Clin. Chem. 24, 177178.

8. Recommendations of the German Society for Clinical Chemistry: Standard Method for the Determination of Creatine Kinase Activity. (1977) J. Clin. Chem, Clin. Biochem. 75, 255-260. 9. National Committee for Clinical Laboratory Standards. Procedures for the collection of diagnostic blood specimens by skin puncture — second edition; Approved Standard. NCCLS publication H4-A2. Villanova, Pa.: NCCLS; 196. 10. Passing, H. & Bablok, W. (1983) A New Biometrical Pror cedure for Testing the Equality of Measurements froih Two Different Analytical Methods. J. Cliri. Chem. Clin. Biochem. 27, 709-720. 11. Szasz, G., Gerhardt, W., Gruber, W. & Bernt, E. (1976) Creatine kinase in serum; 2. Interference of adeiiylate kinase with the assay. Clin. Chem. 22, 1806—1811. 12. Szasz, G., Gerhardt, W. & Gruber, W. (1977) Creatine kinase in serum: 3. Further study of adenylate kinase ihhibitors. Clin. Chem. 23, 1888-1892. 13. Koller, P. U, Tritschler, W. & Carstensen, C. A. (1989) Interference studies on the Reflotron® System. Lab. Med. 73, 399-402. Carsten A. Carstensen Boehringer Mannheim GmbH Sandhofer Stra e 116 W-6800 Mannheim 31 Germany

Eur. J. Clin. Chem. Clin. Biochem. / Vol. 29,1991 / No. 10

Hörn et al.: Thyroxine EIA on Hitachi analysis Systems

697

Eur. J. Clin. Chem. Clin. Biochem. Vol. 29, 1991, pp. 697-703 © 1991 Walter de Gruyter & Co. Berlin · New York

The Determination of Thyroxine and Thyroxine Uptake with New Homogeneous Enzyme Immunoassays Using Boehringer Mannheim/Hitachi Analysis Systems By K. Hörn1, M. J. Castineiras2, J. Ortolä2, R. Kock\ F. C. Perriard4, S. Bittner5, J. V. Pairet J. Boulanger*, S. Zeidner1, R. Maier*, F. Boege9, H. Dubois™, M. McGovern10 and /. Opitz™ 1 2 3 4 5 6 7 8 9 10

. Ers\

Medizinische Klinik Innenstadt, Universität München, München, Germany Servei de Bioquimica Clinica, Hospital Princeps d'Espanya, Barcelona, Spain Klinikum der Rheinisch-Westfälischen Technischen Hochschule, Zentrallabor, Aachen, Germany Laboratoire central, Hopital cantonal, Fribourg, Switzerland Abteilung för Klinische Chemie der Medizinischen Klinik der Universität Eppendorf, Hamburg, Germany Hospital St. Ode, Baconfoy, Belgium Gemeinschaftslabor Dr. Spranger lDr. Kloß, Ingolstadt, Germany Laborgemeinschaft Niederbayrischer Ärzte, Straubing, Germany Medizinische Poliklinik der Universität, Würzburg, Germany Boehringer Mannheim GmbH, Mannheim, Germany

(Received January 27/June 22, 1991)

Evaluators - Prof. Dr. K. Hörn Medizinische Klinik Innenstadt, Universität München München, Germany — Prof. Dr. A. Torralba, Dr. X. Fuentes-Arderiu, Dr. M. J. Castineiras, Dr. J. Ortolä, Dr. M. A. Navarro Servei de Bioquimica Clinica, Hospital Princeps d'Espanya Barcelona, Spain - Prof. Dr. Dr. H. Greiling, Dr. R. Kock, Mr. B. Delvoux Klinikum der Rhein.-Westf. Technischen Hochschule, Zentrallabor Aachen, Gennany - Dr. F. C. Perriard Laboratoire central, Hopital cantonal Fribourg, Switzerland - Prof. Dr. C. Wagener, Prof. Dr. K. Wielkens, Ms. S. Bittner Abt. f. Klinische Chemie der Medizinischen Klinik der Universität Eppendorf Hamburg, Gernmny

Dr. J. V. Pairet, Dr. P. Ers, Mr. J. Boulanger Hospital St. Ode Baconfoy, Belgium Dr. S. Zeidner Gemcinschaftslabor Dr. Spranger/Dr. Kloß Ingolstadt, Germany Mr. R. Maier Laborgemeinschafl Niederbayrischer Ärzte Straubing, Germany Dr. F. Boege, Mr. H. Schraml Medizinische Poliklinik der Universität Würzburg, Germany Mrs. E. Wulff Burgwedel, Germany Mrs. M. N. Duverger Boehringer Mannheim France Meylan, France

Summary: New homogeneous enzyme immunoassays for the determination of thyroxine and thyroxine uptake have been developed. The CEDIA® assays are based on the cloned enzyme donor immunoassay technology, which involves fragments of ß-galactosidase prepared by genetic engineering. The assays have been adapted for Boehringer Mannheim/Hitachi analysers. The CEDIA® T4/T Uptake assays were evatuated in eleven clinical chemistry laboratories on various Boehringer Niannheim/Hitachi analysis Systems, using a 2-point calibration. The analytical ränge of the T4 test was 10 to 258 nmol/1 thyroxine. The T uptake test had a measuring ränge between 20-50%. Depending on the concentration of the analyte (samples from hypo-, eu- or hyperthyroid patients), mean coefficients of Eur. J. Clin. Chem. Clin. Biochem. / Vol. 29,1991 / No. 10

Hörn et al.: Thyroxine EIA on Hitachi analysis Systems

698

Variation ranged from 1.8 to 4.8% within-run and from 4.1 to 6.5% between-run for the T4 assay. Even better coefficients of Variation were obtained for the T uptake assay (1.4 to 2.3% within-run, 2.8 to 3.3% between run). The relative inaccuracy of the CEDIA® assays with respect to values assigned by other tests was satisfactory in various control sera. The T4 assay was compared with one radioimmunoassay, one enzyme immunoassay and one fluorescence Polarisation immunoassay. Slopes ranging from 0.9 to 1.1 and intercepts ranging from —10 to +10 nmol/1 thyroxine were obtained with two exceptions. The resülts of the T uptake test correlated reasonably with those of other thyroxine-binding methods. No interference was observed with icteric and lipaemic sera. Haemoglobin up to 4 g/l had no significant influence. Resülts of the CEDIA® T Uptake test are mainly used for calculation of the free thyroxine index, in which the thyroxine value is corrected for variations of thyroxine-binding protein concentrations. The free thyroxine index is related to the concentration of free T4 determined by enzyme immunoassay. In conclusion, the CEDIA® T4/T Uptake assays are convenient and reliable methods which offer an alternative in thyroid diagnostics in combination with routine clinical chemistry on Boehringer Mannheim/Hitachi analysers.

Introduction

There are several types of procedure for estimating the concentration of the free thyroxine hormone: direct methods like equilibrium dialysis which is not suitable for routine use, two-step assays for free thyroxine, and single Step "thyroxine analogue" methods or indirect methods like the free thyroxine index (FTI) or the thyroxine/thyroxine-binding globulin (T4/TBG) ratio (1). Controversial opinions exist on the advantages and disadvantages of these techniques. New methods are still being developed. For routine measurements, rapid and automated procedures are superior to other methods. For this reason, homogeneous enzyme immunoassays play a major role in the clinical laboratory. The CEDIA® assays are homogeneous enzyme immunoassays based on the cloned enzyme donor immunoassay technology (2), which was first used for the determination of serum thyroxine and thyroxine uptake (3). CEDIA® assays have been optimized for automated analysers used in the routine clinical laboratory. It was the aim of the present study to evaluate the CEDIA® T4 and T Uptake assays on Boehringer Mannheim/Hitachi1) analysers. Eleven different centres participated in the multicentre evaluation. Four of these centres performed only the imprecision studies. The evaluation resülts are combined in this report. Materials and Methods Sera For analysis of imprecision, reagent stability, calibration stability and relative inaccuracy, Lyphochek® 3 level control sera ( -Rad, München, Germany, cat. No. C-370-5) or Preci') abbreviated to BM/Hitachi in the following text

norm® U control sera (Boehringer Mannheim GmbH, Mannheim, Germany, cat. No. 171743) were used. Other studies were performed with fresh human sera. Calibration material was included in the respective test kits. Reagents The CEDIA® T4/T Uptake assays were from Boehringer Mannheim GmbH (Mannheim, Germany, cat. No. T4 assay kit 1179667; cat. No. T Uptake assay kit 1179675). The CEDIA® T4 test was compared with one radioimmunoassay (RIA; Amerlex-MT4 RIA kit, Amersham, Braunschweig, Germany, cat. No. IM 3014), one enzyme immunoassay (EIA; Enzymun-Test® T4, Boehringer Mannheim GmbH, Mannheim, Germany, cat. No. 204510) and one fluorescence Polarisation immunoassay (FPIA; TDx® T4 plus, Abbott, North Chicago, USA, cat. No. 9113-20). The CEDIA® T Uptake test was compared with the TDx® T Uptake kit (Abbott, North Chicago, USA, cat. No. 9117-20) and the EIA Enzymun-Test® TBK (Boehringer Mannheim GmbH, Germany, cat. No. 249416). FTI calculated by the CEDIA® assays was cpmpared with free thyroxine determined by an EIA (Enzymun-Test® fT4, Boehringer Mannheim GmbH, Mannheim, Germany, cat. No. 1007955). Instruments In eight of the eleven evaluation centres, a BM/Hitachi 717 was used for quantifying the CEDIA® assays. Two centres used the BM/Hitachi 704 and one used a BM/Hitachi 705. The instrument settings are given in table 1. Resülts of Enzyrmm-Test® T4, thyroxine-binding capacity (TBK) and free thyroxine (FT4) were obtained using the fully automated ES 600 System (4). Measurements of the FPIA assay were perforrned on an Abbott TDx® analyser. Principle of the CEDIA® T 4 /T Uptake assays The CEDIA® T4/T Uptake assays involve two separate fragments of ß-galactosidase2) which are produced by recombinant DNA technology (2). The mdividual fragments, enzyme donor and enzyme acceptor are inactive, but spontaneously recombine to form active ß-galactosidase. Thyroxine is conjugated tp the enzyme donor peptide in such a way that binding does not interfere with reassociation of the enzyme fragments (flg. 1). However, if thyroxine-specific antibodies are bound to the thyroxine-enzyme donor complex, formation of analytically 2

) ß-Galactosidase (EC 3.2.1.23). * Eur. J. Cliii. Chem. Clin. Biochem. / Vol. 29,1991 / No. 10

699

Hörn et al.: Thyroxine EI A on Hitachi analysis Systems Tab. l. Instrument settings for the determination of thyroxine and thyroxine uptake in CEDIA® assays Assay CED1A® T4

Parameter Sample volume ( ) Reagent 1 volume ( ) Reagent 2 volume ( ) Wavelength (nm) Preincubation/ measuring interval

CEDIA® T Uptake

Sample volume ( ) Reagent 1 volume ( ) Reagent 2 volume ( ) Wavelength (nm) Preincubation/ measuring interval

Instrument BM/Hitachi 704

BM/Hitachi 705

BM/Hitachi 717

12 235 135 660/415 (sub/main) 5 min/ 9-10 min

12 235 135 660/415 (sub/main) 5 min/ 9—10 min

10 195 110 660/405 (sub/main) 5 min/ 9 -10 min

20 275 55 660/415 (sub/main) 5 min/ 9-10 min

20 275 55 660/415 (sub/main) 5 min/ 9—10 min

10 250 50 660/405 (sub/main) 5 min/ 9-10 min

Enzyme acceptor ~ (inactive) EA)

Enzyme donor labelled with anatyte

j donor(Cömpiementatton Antibody to the active enzyme sterically hindered)

Active ß-galactosidase

—Sample o

< =

analyte

Antibpdy-anatyte

Fig. 1. Principle of the CEDIA® T4 assay. Enzyme donor (ED) and enzyme acceptor (EA) are fragments of ß-galactosidase (EC 3.2.1.23) obtained by genetic engineering.

serum provided by the manufacturer. This serum was measured in 60 independent runs (21 replicates per run). The lower limit of detection, defmed äs the +3 Standard deviation (99th percentile) of the mean, was 10 nmol/1 thyroxine. To study the analytical ränge, aliquots of a human serum containing a high concentration of thyroxine (303 nmol/1) were mixed in various ratios with human serum containing a low concentration of thyroxine (33.5 nmol/l). Unlike most homogeneous enzyme immunoassays, the CEDIA® T4 assay has a linear dose response relationship, which enables a 2-point calibration (flg. 2). Linearity was obtained up to 258 nmol/1 thyroxine.

active ß-galactosidase is inhibited. Thyroxine in the sample and the thyroxine-enzyme donor complex compete for a limited amount of antibodies, thereby reguläting the amount of enzyme that is fonned. Thus, the concentration of thyroxine in the sample is directly proportional to ß-galactosidase activity. In the T uptake assay, the thyroxine-enzyme donor complex binds directly to thyroxine-binding proteins in the sample and is no longer available for reassociation with the enzyme aeceptor fragment. Thus, thyrpxine-binding proteins regulate the amount of ß-galactosidase fonned, and enzyme activity is inversely proportional to the concentration of unoccupied thyrpxine biliding sites. Procedures The CEDIA® T4/T Uptake assays, äs well äs the assays with test kits for method comparison, were performed according to the manufacturer's mstrüctions.

Results Analytical ränge The lower limit of detection of the T4 assay was determined using an almost thyroxine-free control Eur. J. Clin. Chem. Clin. Biochem. / Vol. 29,1991 / No. 10

100

200

Thyroxine (expected) [nmol/1]

300

Fig. 2. Analytical ränge of the CEDIA® T4 test in human serum.

700

H rn et al.: Thyroxine EIA on Hitachi analysis Systems

A 2-point calibration can also be used for the CEDI A® T Uptake test, which has an analytical r nge between 20 and 50%.

300 -

Imprecision The determination of imprecision was performed at three different concentrations with samples from hypo-, eu- or hyperthyroid patients. Either Lyphochek® 3 level control sera or human serum pools were used. Within-run imprecision was measured in 60 independent runs (21 replicates per ran). Mean CV values for the T4 assay ranged from 1.8 to 4.8% and for the T uptake assay from 1.4 to 2.3% (fig. 3). Between-run imprecision was evaluated in eleven separate runs on 20 days. Mean CV values for the CEDIA® T4 assay are between 4.1 and 6.5% and for the T Uptake test between 2.8 and 3.3% (fig. 3).

50

100

150

200

250

300

50

100 150 200 250 Thyroxine (EIA) [nmol/l]

300

50

100 150 200 Thyroxine (FPIA) [nmol/l]

300

Thyroxine (RIA) [nmol/l]

300

a

Hypothyroid 39-Ό4 nmol/l

' \

Euthyroid 91 -129 nmol/l \

Hyperthyroid 18 -23Z nmol/l C)

'

YW/y&yfr^/s/Wpffii'tfy

250·-

'

\ \ Y/////s//SJ(/'!ffiff

\

\

Y/fiuS/ri/f

/tf$fy///Mi

2

/,

CV[%]