Professional Documents
Culture Documents
Mller Bardorff2000
Mller Bardorff2000
Mller Bardorff2000
net/publication/12335529
CITATIONS READS
56 460
14 authors, including:
Some of the authors of this publication are also working on these related projects:
All content following this page was uploaded by Rainer Zerback on 12 September 2014.
Margit Müller-Bardorff1, Christer Sylvén2, Gundars nificant analytical interference was detected for any of
Rasmanis2, Bo Jørgensen3, Paul O. Collinson4, Ulla the assays in investigations with biotin (up to 100
Waldenhofer5, Michael M. Hirschl5, Anton N. µg/l), hemoglobin (up to 0.125 mmol/l), hematocrit (26
Laggner5, Willie Gerhardt6, Gerd Hafner7, Irene to 52%), bilirubin (up to 340 µmol/l), triglycerides (up
Labaere8, Robert Leinberger8, Rainer Zerback8 and to 5.0 mmol/l), and 18 standard drugs.
Hugo A. Katus1 With the Cardiac Reader reliable quantitative results
1
can be easily obtained for both cardiac markers. The
Medizinische Klinik II, Medizinische Universität zu Lübeck,
system is, therefore, particularly suitable for use in
Lübeck, Germany
2 Department of Cardiology, Huddinge Hospital, Huddinge, emergency rooms, coronary care units and small hos-
Sweden pitals.
3 Department of Clinical Biochemistry, Ålborg Hospital,
Key words: Troponin T; Myoglobin; Cardiac enzymes;
Ålborg, Denmark
4 Point-of-care testing; Myocardial infarction; Unstable
Department of Pathology, Mayday University Hospital,
London, United Kingdom angina pectoris.
5
Notfallaufnahme, Allgemeines Krankenhaus – Universitäts- Abbreviations: cTnT cardiac troponin T; sTnT skeletal
kliniken, Wien, Austria troponin T.
6 Department of Clinical Chemistry, Helsingborg Hospital,
Helsingborg, Sweden
7 Institut für Klinische Chemie, Johannes-Gutenberg-Univer-
since it is only through an evaluation of its release ki- Calibration of the Cardiac Reader
netics that a reliable infarct diagnosis can be made (5, CARDIACM was calibrated identically as CARDIACT Quantita-
8–10, 33, 34). Release kinetics are also required for as- tive. Parallel samples of heparinized blood and serum were
sessment of reperfusion. collected from over a hundred patients with suspected acute
We have therefore developed a system, based on the coronary syndromes. Heparinized blood samples were used
existing troponinT rapid test, that permits a quantita- to determine the reflectance with CARDIACM, and serum
tive determination of troponin T. A completely new samples to measure the myoglobin concentration with Tina-
quantitative myoglobin test was also developed for quant Myoglobin (Roche Diagnostics)1. The resulting pairs of
this system. We describe the operating principle of this values were used to calibrate this lot of CARDIACM by regres-
sion analysis. This lot served as a master lot for calibrating fur-
new system for rapid determination of troponinT and
ther lots. Using this master lot, and a new lot to be calibrated,
myoglobin in heparinized blood and present the results
measurements were taken on 18blood samples spiked with
of a multicenter evaluation. various concentrations of myoglobin. The concentrations ob-
tained with the master lot served as target concentrations for
the new lot; the corresponding reflectance values were ob-
Materials and Methods tained directly from measurements. The resulting pairs of val-
ues (target concentration/reflectance) were then used to es-
Test principle of the Cardiac Reader tablish the calibration curve for the new lot of CARDIACM.
The Cardiac Reader (Roche Diagnostics, Mannheim, Ger- The lot-specific calibration curve is communicated to the
many) is a system for quantitative bedside determinations of reader via a code chip. The agreement of code and test strip is
checked by the bar code (Figure 1) on the bottom of the test
myoglobin (CARDIACM) and troponinT (CARDIACT Quantita-
tive). The chemical reaction principle of the test strips has strip.
been described previously (20, 21). The Cardiac Reader is a
Imprecision and between-instrument variability
camera with a charged coupled device that optically records
the reflectance signal from the detection zone of the immuno- Investigations of within-series imprecision and of inter-in-
chemical test strips (Figure1). Signal and control lines are strument variability were carried out on heparinized blood
identified by a pattern recognition algorithm. The intensity of samples collected from healthy volunteers and spiked with
the signal line is proportional to the concentration of the ana- troponin T-containing sera or purified myoglobin and with
lytes troponin T and myoglobin. The optical signal is con- control material (controls of the Cardiac Reader system and
verted into concentration via a lot-specific calibration curve calibrators of the reference methods). Hematocrit of all spiked
(see below), which is stored in a code chip. samples was readjusted to the initial value of the original
The tests require a 150µl sample of heparinized blood and donor blood sample. Each sample was measured 10times at
the reaction times are approximately 8min for myoglobin and 20different instruments using one lot. For each series (n = 10)
approximately 12min for troponinT determination. at each instrument, mean value and coefficient of variation
For CARDIACT Quantitative the quantitative measurement (CV) were calculated. The CVs for the within-series impreci-
range is 0.1 to 3µg/l troponin T; the display shows values be- sion given in the result section were calculated as mean val-
low 0.05µg/l as “negative”, values between 0.05 and 0.1µg/l ues of the 20 CVs with 20 instruments. The between-instru-
as “low”, and values above 3.0µg/l as “high”. For CARDIAC ment CVs were calculated as the CVs of the serial mean values
M, the quantitative measurement range is from 30 to 700µg/l from 20 instruments. Serial measurements were done by one
myoglobin; values below this range are displayed as “nega- laboratory technician.
tive” and values above as “high”.
Method comparisons
n.d. = not determined. Concentration given as final sample concentration. The percentage
recovery compared to reference is shown.
globin level 1 and 2 (Roche Diagnostics), respectively, at each solution (for final concentrations see Table2; for concentra-
center. The quality control of cTnT ELISA and Tina-quant myo- tions of the drugs see Table 3 and ref. 37) and with a human
globinwas performed at each center with the respective pack- serum sample containing troponinT in a high concentration.
age controls. The influence of triglycerides was investigated using human
lipemic serum samples spiked with troponinT sera. To deter -
Lot-to-lot comparison mine the influence of hematocrit, samples collected from pa-
tients with coronary artery disease were used. Volume depen-
To check the reproducibility of the calibration, five lots of CAR-
dence was investigated using heparinized blood samples
DIACM (226007, 226008, 226009, 2260012-20, 2260012-30)
from healthy donors spiked with a human serum sample con-
and six lots of CARDIACT Quantitative (226003, 226005,
taining troponinT in a high concentration.
226013, 226014, 226026, 226031) were investigated using
Cross-reactivity of CARDIACT Quantitative with sTnT was
fresh heparinized blood collected from patients with sus-
calculated according to the following formula:
pected acute coronary syndromes. The numbers of samples
are listed in Table1. recovered sTnT concentration
—————————————— x 100%
original sTnT concentration
Interference studies
Comparisons of recoveries in the hematocrit study were per-
To check for interfering factors, heparinized blood or plasma formed by a h-test according to Kruskal-Wallis with a level of
was spiked with potential interfering substances (sTnT), bi- significance of p < 0.05.
otin, hemoglobin, bilirubin, 18standard drugs) from a stock
570 Müller-Bardorff et al.: Point-of-care-system for troponin T and myoglobin
n.d. = not determined. Concentrations given as final sample Fig. 2 Within-series imprecision of CARDIAC M(a) and CAR-
concentration. The percentage recovery compared to refer- DIAC T Quantitative (b). j blood samples, m controls, n = 10
ence is shown. replicates.
Müller-Bardorff et al.: Point-of-care-system for troponin T and myoglobin 571
Lot-to-lot variability
In lot-to-lot comparisons, the lots of CARDIACT Quan-
titative investigated showed good agreement with one
another, as did the lots of CARDIACM (Table1). For
Fig. 4 Multicenter method comparison of CARDIAC M vs. CARDIACT Quantitative, the correlation coefficients r
Tina-quant Myoglobin. y= 0.92x + 1.6; r = 0.98. HVB = he- were between 0.92 and 0.97 and the accuracy differ-
parinized venous blood. ences were between –6% and +5%. CARDIAC M
showed correlation coefficients between 0.97 and 0.99
with accuracy differences ranging from 0% to +11%.
The same analysis for CARDIACT Quantitative with
the cTnTELISA reference method showed a bias of 7%
Interference
for the whole test range (Figure 5).
When the results obtained with CARDIACT Quantita- In investigations with biotin (up to 100 µg/l), hemoglo-
tive were compared with the reference method cTnT bin (up to 2000 mg/l), bilirubin (up to 200 mg/l), and
572 Müller-Bardorff et al.: Point-of-care-system for troponin T and myoglobin
Tab. 4 Summary of the results of the daily quality control in the multicenter evaluation.
Tab. 5 The effect of the hematocrit on the relative recovery such as troponin T rather than simple ruling in or out of
of the troponin T concentration (cTnT ELISA) by CARDIAC T myocardial infarction based on classic WHO criteria.
Quantitative and of the myoglobin concentration (Tina-quant A crucial point for the clinical utility of a marker is the
Myoglobin) by CARDIAC M. time from the first investigation of the patient and
blood sampling until the time a result is available
Hematocrit (%) CARDIAC T CARDIAC M
which is then used for clinical and therapeutic decision-
Quantitative
making by the physician. In this setting, point-of-care
Recovery (%) n Recovery (%) n systems will offer a more rapid result.
It is now possible to test highly specific troponin T
25–34 98 22 103 14 and the early marker myoglobin quantitatively with a
35–44 94 132 106 66 turnaround time of less than 15 minutes at the bedside.
45–54 87 40 106 12 No significant analytical interference was found
when various standard drugs, triglycerides, bilirubin
The medians for different hematocrit ranges are shown. and biotin were tested. Underdosing or overdosing a
blood sample by 15 µl (standard volume 150 µl) influ-
enced the test result by no more than 10%. The cross-
triglycerides (up to 5000 mg/l), no significant analytical reactivity with skeletal troponin T was calculated to be
interference was detected, i.e. all deviations from ex- approximately 0.003%. No significant correlation was
pected values were ≤ 10%. Overdosing or underdosing found between the hematocrit and the concentration-
by 15µl affected the test result by no more than 10% dependent recovery with both quantitative bedside as-
(Table2). says.
Appreciable cross-reactivity with sTnT was only ob- Addition of all the errors (imprecision, lot-to-lot vari-
served at concentrations from 1000µg/l. Cross-reactiv- ability, accuracy in the method comparison) yielded a
ity was, therefore, calculated to be approximately total error range of ±25% for CARDIACT Quantitative
0.003% (Table2). and ±30% for CARDIACM. The variation of a test result
Various standard drugs did not influence the recov- is of most importance in the range near the cut-off
ery with CARDIACT Quantitative and CARDIACM by level. At the cut-off for troponinT (0.1µg/l), CARDIAC T
more than 26% when toxic concentrations of the drugs Quantitative, therefore, has a maximum range of varia-
were used, and by no more than 12% when therapeutic tion between 0.075 and 0.125µg/l, after taking into ac-
concentrations were used (Table 3). count the differences between lots and the comparison
No significant correlation was found between hema- with the reference method. At the cut-off for myoglobin
tocrit and the recovery of the reference method con- (70µg/l), CARDIACM has a calculated maximum range
centrations by the Cardiac Reader tests (CARDIAC T of variation from 49 to 91µg/l. However, in serial mea-
Quantitative: r = 0.12, CARDIACM: r = –0.05). The surements with CARDIACM, which is the primary need
slightly lower recovery with CARDIAC M at higher for a quantitative myoglobin value, only imprecision
hematocrits is not significant (p > 0.05) in the h-test ac- remains as a source of variation, so that the total error
cording to Kruskal-Wallis (Table 5). range diminished to ±15% (59 to 81µg/l at the cut-off).
When using the clinical cut-off of 0.1 µg/l, there were
only 1% discordant results between CARDIACT Quan-
Discussion titative and the reference method cTnT ELISA. Compar-
ing the two myoglobin tests with the cut-off level of 70
Newer cardiac markers like the troponins have con- µg/l, we found 6% discordant results. These discrepan-
tributed substantially to the diagnostic workup of pa- cies of 1% with troponin T and 6% with myoglobin are
tients with suspected myocardial infarction. Risk strati- probably caused by imprecision of both the tested
fication of patients with unstable angina and suspected method and the reference method.
minor myocardial damage is possible with a marker The Cardiac Reader was designed to allow a rapid
Müller-Bardorff et al.: Point-of-care-system for troponin T and myoglobin 573
point-of-care determination of two essential cardiac 5. Brogan GX, Friedman S, McCuskey C, Cooling DS, Berruti
markers. The clinical performance of the quantitative L, Thode HC, et al. Evaluation of a new rapid quantitative
bedside assays for troponin T and myoglobin is being immunoassay for serum myoglobin versus CK-MB for rul-
ing out acute myocardial infarction in the emergency de-
further investigated in clinical studies ongoing at pre-
partment. Ann Emerg Med 1994; 24:665–71.
sent. Preliminary results from these studies show the
6. Castaldo AM, Ercolini P, Forino F, Basevi A, Vrenna L,
gain in time by using the Cardiac Reader compared to Castaldo P, et al. Plasma myoglobin in the early diagnosis
conventional laboratory testing (72 minutes, p < of acute myocardial infarction. Eur J Clin Chem Clin
0.0001) (Collinson, unpublished data). Biochem 1994; 32:349–53.
The design of the Cardiac Reader meets the require- 7. De Winter RJ, Koster RW, Sturk A, Sanders GT. Value of
ments on the use of cardiac markers in point-of-care myoglobin, troponin T, and CK-MB mass in ruling out an
testing as recommended by two scientific societies. acute myocardial infarction in the emergency room. Circu-
The International Federation of Clinical Chemistry and lation 1995; 92:3401–7.
Laboratory Medicine (IFCC) Guidelines on biochemical 8. Montague C, Kircher T. Myoglobin in the early evaluation
of acute chest pain. Am J Clin Pathol 1995; 104:472–6.
markers in coronary artery disease (39), which are in
9. Tucker JF, Collins RA, Anderson AJ, Hess M, Farley IM,
line with the National Academy of Clinical Biochem-
Hagemann DA, et al. Value of serial myoglobin levels in the
istry (NACB) Guidelines (40), recommend the routine
early diagnosis of patients admitted for acute myocardial
use of cardiac troponins in acute coronary syndromes. infarction. Ann Emerg Med 1994; 24:704–8.
The value of myoglobin as an early marker of cardiac 10. Woo J, Lacbawan FL, Sunheimer R, LeFever D, McCabe JB.
damage has been acknowledged in the IFCC Guide- Is myoglobin useful in the diagnosis of acute myocardial
lines on cardiac markers as well as in the NACB Guide- infarction in the emergency department setting? Am J Clin
lines. The importance of point-of-care testing has been Pathol 1995; 103:725–9.
emphasized both in the IFCC Guidelines and the NACB 11. Klootwijk P, Cobbaert C, Fioretti P, Kint PP, Simoons ML.
Guidelines. Moreover the NACB Guidelines recom- Noninvasive assessment of reperfusion and reocclusion
mend a turnaround time of one hour or less for bio- after thrombolysis in acute myocardial infarction. Am J
Cardiol 1993; 72:75G–84G.
chemical testing.
12. Abe J, Yamaguchi T, Isshiki T, Naka H, Taguchi J, Ishizaka
N, et al. Myocardial reperfusion can be predicted by myo-
globin/creatine kinase ratio of a single blood sample ob-
Conclusions tained at the time of admission. Am Heart J 1993;
126:279–85.
The Cardiac Reader represents an analytically reliable 13. Ellis AK, Little T, Masud ARZ, Liberman HA, Morris DC,
and easy method that further extends the spectrum of Klocke FJ. Early noninvasive detection of successful reper-
applications of the visual troponin T rapid assay test in fusion in patients with acute myocardial infarction. Circu-
early infarct diagnostics, in reperfusion control, and in lation 1988; 78:1352–7.
cardiac risk assessment. With these attributes, it is, 14. Ishii J, Nomura M, Ando T, Hasegawa H, Kimura M,
Kurokawa H, et al. Early detection of successful coronary
therefore, suited for adoption by emergency depart-
reperfusion based on serum myoglobin concentration:
ments, coronary care units and small hospitals with
comparison with serum creatine kinase isoenzyme MB ac-
few cardiac patients where investment in larger and tivity. Am Heart J 1994; 128:641–8.
more expensive equipment is not justified. 15. Jurlander B, Clemmensen P, Ohman EM, Christenson R,
Wagner GS, Grande P. Serum myoglobin for the early non-
invasive detection of coronary reperfusion in patients with
Acknowledgements acute myocardial infarction. Eur Heart J 1996; 17:399–406.
16. Lavin F, Kane M, Forde A, Gannon F, Daly K. Comparison of
This study was funded by Roche Diagnostics (Mannheim, Ger- five cardiac markers in the detection of reperfusion after
many). Dr. Katus is a consulstant for Roche Diagnostics and thrombolysis in acute myocardial infarction. Br Heart J
holds a patent for troponin T jointly with Roche Diagnostics. 1995; 73:422–7.
17. Miyata M, Abe S, Arima S, Nomoto K, Kawataki M, Ueno
M, et al. Rapid diagnosis of coronary reperfusion by mea-
References surement of myoglobin level every 15 min in acute my-
ocardial infarction. J Am Coll Cardiol 1994; 23:1009–15.
1. Dean KJ. Cardiac troponin T as a marker of myocardial in- 18. Zabel M, Hohnloser SH, Köster W, Prinz M, Kasper W, Just
jury. In: Wu AHB, editor. Pathology and laboratory medicine H. Analysis of creatine kinase, CK-MB, myoglobin, and tro-
3: cardiac markers. Totowa: Humana Press, 1998: 205–27. ponin T time-activity curves for early assessment of coro-
2. Gerhardt W, Ljungdahl L. Troponin T. A sensitive and spe- nary artery reperfusion after intravenous thrombolysis.
cific diagnostic and prognostic marker of myocardial dam- Circulation 1993; 87:1542–50.
age. Clin Chim Acta 1998; 272:47–57. 19. Antman EM, Grudzien C, Sacks DB. Evaluation of a rapid
3. Rottbauer W, Greten T, Müller-Bardorff M, Remppis A, Ze- bedside assay for detection of serum cardiac troponin T. J
helein J, Grünig E, et al. Troponin T: a diagnostic marker for Am Med Assoc 1995; 273:1279–82.
myocardial infarction and minor cardiac cell damage. Eur 20. Collinson PO, Gerhardt W, Katus HA, Müller-Bardorff M,
Heart J 1996; 17:Suppl F:3–8. Braun S, Schricke U, et al. Multicentre evaluation of an im-
4. Bakker AJ, Koelemay MJ, Gorgels JP, van Vlies B, Smits R, munological rapid test for the detection of troponin T in
Tijssen JG, et al. Troponin T and myoglobin at admission: whole blood samples. Eur J Clin Chem Clin Biochem 1996;
value of early diagnosis of acute myocardial infarction. Eur 34:591–8.
Heart J 1994; 15:45–53. 21. Müller-Bardorff M, Freitag H, Scheffold T, Remppis A,
574 Müller-Bardorff et al.: Point-of-care-system for troponin T and myoglobin
Kübler W, Katus HA. Development and characterization of 31. Hamm CW, Heeschen C, Goldmann BU, Barnathan E,
a rapid assay for bedside determinations of cardiac tro- Simoons ML, for the CAPTURE investigators. Value of tro-
ponin T. Circulation 1995; 92:2869–75. ponins in predicting therapeutic efficacy of abciximab in
22. Christenson RH, Fitzgerald RL, Ochs L, Rozenberg M, patients with unstable angina [abstract]. J Am Coll Cardiol
Frankel WL, Herold DA, et al. Characteristics of a 20-minute 1998; 31:185A.
whole blood rapid assay for cardiac troponin T. Clin 32. Lindahl B, Venge P, Wallentin L, for the Fragmin in Unsta-
Biochem 1997; 30:27–33. ble Coronary Artery Disease (FRISC) Study Group. Tro-
23. Gerhardt W, Ljungdahl L, Collinson PO, Lovis C, Mach F, ponin T identifies patients with unstable coronary artery
Sylvén C, et al. An improved rapid troponin T test with a disease who benefit from long-term antithrombotic pro-
decreased detection limit: a multicentre study of the ana- tection. J Am Coll Cardiol 1997; 29:43–8.
lytical and clinical performance in suspected myocardial 33. Mastropaolo W. Use of delta one-hour myoglobins to rule
damage. Scand J Clin Lab Invest 1997; 57:549–58. out myocardial infarction less than four hours after onset
24. Gambke B, Herkner H, Hirschl MM, Laggner AN, Sylvén C, of symptoms [abstract]. Clin Chem 1996; 42:S95.
Rasmanis G, et al. Analytical and clinical evaluation of an 34. Sunheimer R, Bushnell A, McCabe J, Woo J. Myoglobin as
improved qualitative troponin T rapid test (T ROPT sensi- an early marker for detecting acute myocardial infarction
tive) for whole blood samples [abstract]. The Official Pub- in the emergency department. Clin Chem 1996; 42:S105.
lication of MEDLAB 97 Abstracts. 12th European Congress 35. Müller-Bardorff M, Hallermayer K, Schröder A, Ebert C,
of Clinical Chemistry, Basel Switzerland, 17–22 Aug Borgya A, Gerhardt W, et al. Improved troponin T ELISA
1997:118. specific for cardiac troponin T isoform: assay development
25. Hirschl MM, Herkner H, Laggner AN, Sylvén C, Rasmanis and analytical and clinical validation. Clin Chem 1997;
G, Collinson PO, et al. Analytical and clinical performance 43:458–66.
of an improved qualitative troponin T rapid test in labora- 36. Panteghini M, Altinier S, Dolci A, Pagani F, Pergolini P, Ver-
tories and critical care units. Arch Pathol Lab Med 2000; nocchi A, et al. Multicenter evaluation of five methods for
124(4):583–7. myoglobin determination. Clin Chem 1998; 44:A121.
26. Sylvén C, Lindahl S, Hellkvist K, Nyquist O, Rasmanis G. 37. Breuer J. Report on the symposium “Drug effects in clini-
Excellent reliability of nurse-based bedside diagnosis of cal chemistry methods”. Eur J Clin Chem Clin Biochem
acute myocardial infarction by rapid dry-strip creatine ki- 1996; 34:385–6.
nase MB, myoglobin, and troponin T. Am Heart J 1998; 38. Panteghini M, Apple FS, Christenson RH, Dati F, Mair J, Wu
135:677–83. AH. Use of biochemical markers in acute coronary syn-
27. Schouten Y, de Winter RJ, Gorgels JPMC, Koster RW, dromes. IFCC Scientific Division, Committee on Standard-
Adams R, Sanders GT. Clinical evaluation of the CARDIAC ization of Markers of Cardiac Damage. International Feder-
STATus, a rapid immunochromatographic assay for simul- ation of Clinical Chemistry. Clin Chem Lab Med 1999;
taneous detection of elevated concentrations of CK-MB 37:687–93.
and myoglobin in whole blood. Clin Chem Lab Med 1998; 39. Wu AH, Apple FS, Gibler WB, Jesse RL, Warshaw MM,
36:469–73. Valdes R. National Academy of Clinical Biochemistry Stan-
28. Brimhall BB, Johnson P, Goldman L, Lee T, Sacks D. Diag- dards of Laboratory Practice: recommendations for the
nostic and prognostic value of cardiac troponin T in pa- use of cardiac markers in coronary artery diseases. Clin
tients with acute chest pain [abstract]. Clin Chem 1995; Chem 1999; 45:1104–21.
41:S167.
29. Lindahl B, Venge P, Wallentin L, for the FRISC Study Group. Received 5 January 2000; revised 8 March 2000;
Relation between troponin T and the risk of subsequent accepted 9 March 2000
cardiac events in unstable coronary artery disease. Circu-
lation 1996; 93:1651–7. Corresponding author: Prof. Dr. Hugo A. Katus, Medizinische
30. Ohman EM, Armstrong PW, Christenson RH, Granger CB, Universität zu Lübeck, Medizinische Klinik II, Ratzeburger
Katus HA, Hamm CW, et al. for the GUSTO-IIA Investiga- Allee 160, D-23562 Lübeck, Germany
tors. Cardiac troponin T levels for risk stratification in acute Tel.: +49-451-500-2501, Fax: +49-451-500-6437
myocardial ischemia. N Engl J Med 1996; 335:1333–41. Email: katus@medinf.mu-luebeck.de