1460

Clinical Chemistry 42:9

1460-1466 (1996)

Fluoroenzymometric method to measure cardiactroponin I in sera of patients with myocardial

infarctionMARTINA ZINovro,1 SARA ALTINIER,2 MKVrIA LACHIN,1 PAOLO CARRARO,1 and

MARIO PLEBANI1,2*

The aim of our study was to evaluate the clinical relevanceof serum troponin I (FnI) as a marker of ischemic myocar-

dial injury by using an automated fiuoroenzymometric

assay. The reference range for serum Tn! was establishedby measuring serum Tn! concentrations in blood from 75healthy donors. The concentration was then compared withserum creatine kinase (CK) activity, CK-MB mass, and

myoglobin concentrations in 20 patients with myocardial

infarction diagnosed according to the WHO criteria, 20

patients with chest pain of nonischemic origin, 9 patients

with unstable angina, 11 with stable angina, 11 patients withchronic muscular diseases, 6 patients with muscular trauma

without chest contusion, and 13 patients with chronic renaldisease. We found that: (a) 99% of the blood donors had

Tn! concentrations <0.26 g/L (detection limit of the assayin our study); (b) Tn! values in acute myocardial infarction(AMI) patients 4 h after onset of chest pain showed a

sensitivity of 0.769 and a specificity of 1.0 at a decisionalconcentration for Alrvfl of 1 g.tg/L, even in the presence of

severe skeletal muscle injuries or renal diseases; (c) theincrease in TnI concentrations after infarction (interquar-tile range 3.25-6 h) and the peak occurred later (interquar-tile range 11.5-24 h) than the rise found in myoglobin and

CK-MB, but the increase persisted much longer (>96 h);(d) receiver-operating characteristic curve analysis showed

the high diagnostic accuracy of Tn! in diagnosing AM! evenin patients in whom traditional biochemical markers are

adversely influenced by underlying clinical situations.

INDEXING TERMS: creatine kinase #{149}creatine kinase MB #{149}myo-

globin #{149}diagnostic efficiency #{149}biological markers

Laboratorio Centrale, Azienda Ospedaliera di Padova, Via Giustiniani 2

35128 Padova, Italy.2 Centro Regionale di Ricerca Biomedica, Castelfranco Veneto (TV), Italy.

*Author for correspondence. Fax 01139 49 663240.

Received December 13, 1995; accepted May 1, 1996.

It is widely accepted that the sensitivity and specificity of serum

enzymes and isoenzymes such as total creatine kinase (CK; EC2.7.3.2), CK-MB activity, and lactate dehydrogenase, the final

arbiters by which myocardial damage is diagnosed or excluded

up to now, are inadequatle [1, 2]. Several new biochemical

markers therefore have been investigated recently for the early

and accurate diagnosis of acute myocardial infarction (AM!)[3-6]. Because the contractile and regulatory proteins of themyocardium /7, 8] have high concentrations in myocytes and are

expressed as tissue-specific isoforms, they appear to be poten-

tially useful innovative biological markers for sensitive and

specific diagnostic tests [9-11].In particular, troponin I (TnI), the inhibitory protein of the

troponin-tropomyosin complex, has three isoforms, two skeletal

and one cardiac (sTnI and cTnI respectively), that are encodedby three distinct genes, presenting a dissimilarity of -40% in

the amino acid sequence [12]. Moreover, human cardiac TnT has

31 additional residues on its N-terminal end that are not present

in skeletal forms, thus providing a high potential for obtaining

cardiac-specific antibodies [13, 14]. Furthermore, since the mo-lecular mass of human cardiac Tnl is 24 kDa, the protein is

released fairly rapidly after AMI. This property and the unique

amino acid sequence of cardiac TnI make it a promising

candidate as a diagnostic marker for myocardial damage [15, 16].The aims of our study were therefore to evaluate the

analytical performance of a fluoroenzymometric method previ-

ously proposed for cardiac TnI measurement [17] and compare

its diagnostic accuracy with that of traditional markers of

myocardial injury in the diagnosis and monitoring of patients

with A?vH.

Materials and Methods

CLINICAL SPECIMENS

The study was carried out on 165 patients subdivided as follows:

- Nonstandard abbreviations: CK, creatine kinase; AMI, acute myocardial

infarction; and TnI (TnT), troponin I (troponin T).

ClinicalChemistry 42, No. 9, 1996 1461

Group 1. Seventy-five healthy subjects (36 men and 39 women),

ages 18 to 73 years (median 45 years), with no history of

cardiovascular disease, as controls to obtain a reference interval.

Group 2. Twenty patients (13 men and 7 women, ages 46 to 87

years, median 67 years) with AM! diagnosed according to the

‘WHO criteria (characteristic chest pain, electrocardiographic

findings, serial increases of total CK activity and CK-MB

concentrations with peak values of twice the upper limit of the

reference interval): 16 with Q-wave and 4 with non-Q-wave

infarctions admitted to the coronary care unit within 9 h after

the onset of symptoms (range 0.5-9 h; median 3.5 h); 14 given

a thrombolytic treatment while 6 were not treated, depending

on clinical circumstances, and 12 of these classified as reperfused

on the basis of clinical pictures (symptoms resolved rapidly,

ST-segment changes improved, presence of reperfusion ar-

rhythmias) [18].

Group 3. Forty patients, 20 with chest pain of nonischemic

origin, 9 with unstable angina (five patients assigned to class III,

two patient to class II, and two to class I according to Braun-

wald’s classification[191), and 11 with stable angina, to evaluate

the usefulness of new marker for ruling out AMI in emergency

department patients.

Group 4. Patients with chronic skeletal muscle diseases (11),

multiple traumas without chest contusion (6), and with chronic

renal failure (13), to compare the clinical specificities of the tests

in a worst-case situation. In these patients the concomitant

myocardial injury was excluded by the usual criteria,i.e.,no

history of chest pain and no electrocardiographic signs of

ischemia within the previous 3 weeks.

BLOOD SAMPLING

Only one blood sample was drawn from groups 1 and 4 patients.

Blood samples were collected from AMI patients (group 2)

immediately upon admission to the coronary care unit, and then

hourly for the next 6 h, every 3 h until 24 h, and then every 6 h

until the fourth day after admission. In patients with unstable

angina (group 3), the blood samples were collected at 4-h

intervals during the first 24 h and at 6-h intervals until 2 days

after hospitalization, whereas in patients with chest pain and

stable angina (group 3), one blood sample only was drawn. All

patients gave their informed consent for extra blood samples to

be drawn.

In serum samples obtained from group 1 patients, TnI only

was assayed, whereas in each serum sample from groups 2, 3, and

4 patients, TnI, total CK activity, CK-MB (mass concentration),

and myoglobin were measured.

TNI

TnT concentrations were assayed by a commercially available

automated two-site immunoassay (Stratus Cardiac Troponin I;

Baxter Dade, Milano, Italy) [20] that includes two monoclonal

antibodies specific for cardiac isotype of the protein. The

measuring range of the method was from 0.35 (analytical

sensitivity declared by the manufacturer) to 50 j.tg/L, the assay

generating Tn! results within 10 mm.Three control samples containing different concentrations of

TnI (Baxter Dade Stratus Troponin I controls) and a pooi of

patients’ sera were used in each run to verify the imprecision of

the method, whereas the detection limit of the assay (mean

absorbance + 3 SD of the zero calibrator) was verified by

running 10 replicates of the zero calibrator in different runs.

TOTAL CK ACTIVITY

The catalytic concentration of CK was determined according to

the method recommended by the IFCC [21] at 37 #{176}C(Bracco,

Milano, Italy). The decisonal concentrations for AIvII used in

this study were >160 U/L for women and >190 U/L for men.

CK-MB

The mass concentration of CK-MB was measured with a

commercially available immunoabsorbant assay (Stratus CK-

MB; Baxter Dade) with a sensitivity of 0.4 g/L. The discrim-

inator value for AMI was 5 j.tg/L [22] and 2.5 for the relative

index [CK-MB (.tg/L)/total CK (U/L)] X 100].

MYOGLOBIN

The serum myoglobin concentration was assayed with the

immunological latex method (Istituto Behring; Scoppito,

L’Aquila, Italy), and a discriminator value for AM! of 50 j.tg/L

was chosen [23].

STATISTICAL EVALUATION

To compare the results from cases in which the time elapsed

between onset of chest pain and admission to the coronary care

unit was different, we added this individual time to that of each

blood sampling and produced 24 time series of 1 (until 6 h), 3

(until 30 h), and 5 (until 4 days) h. Each result was thus included

in the appropriate time series.

Medians and percentiles were calculated to describe contin-

uous variables.

To evaluate the efficiency of the markers, we used a PCprogram called “Test Evaluation” supplied by W. Gerhardt

(Department of Clinical Chemistry and Section of Cardiology,

Lasarettet, Helsingborg, Sweden) and applied it to an IBM

computer [24]. The test produces comparative graphs by making

double histograms with overlapped sensitivity, specificity, and

efficiency curves. Then, by plotting the sensitivity vs 1 -

specificity, receiver-operating characteristic (ROC) curves were

constructed to compare the discriminative power of the methods

tested. The diagnostic efficiency was ascertained on the basis of

the area under the ROC curve. Areas under the curves and

differences between ROC curves were calculated according to

the method proposed by Hanley and McNeil [25, 26], with the

Graph ROC for windows program by Veli Kairisto and Allan

Poola (Department of Clinical Chemistry, University of Turku,

Turku, Finland). Confidence intervals (95%) were calculated by

Confidence Interval Analysis microcomputer program [27].

:::::i:::]?A

1.0

0.,

0.S

0.7

E1.6

0.5

0.4

0.3

0.2

0.1

0.0

1.0

0.,

B0.S

0.5

0.5

0.4

0.3

0.2

0.1

0.0

6 0.00 0.20 0.30 Fig. 1. Frequency distributions of ml values in AMI patients: number ofcases are plotted as a function of test values (abscissa).

13 0.00 0.00 0.00 Sensitivity, specificity, and efficiency curves are plotted as a function of testvalues. On the basis of these curves, a discriminator value of 1.0 eg/L yielded20 0.00 0.00 0.20a sensitivity of 0.769 and a specificity of 1.0 (values directly off the scale on the

9 0.00 0.40 0.78 ordinate of the graph). (A) Upper field: frequency distribution of ml values in AMI

11 0.00 0.00 0.24 patients4 h after the onset of chest pain; lower fiekt frequencydistributionofml values in group 3 patients (chest pain, stable and unstable angina). (B) Upper

20 1.00 2.90 22.88 field frequency distribution of ml values in AMI patients 4 h after the onset of

chest pain; lower fielct frequency distribution of Tnl values in groups 3 and 4(chronic skeletal muscle diseases, multiple traumas without chest contusion.

_______________________________________ chronic renal failure) patients.

1462 Zaninotto et al.: Fluoroenzymometric measurement of cardiac TnT

Control sample 1

Assigned value

3.8

Table 1. AnalytIcal ImpWithin run (n = 10)

reclslonBetween run (n = 20)

x, Lg/L

3.97

SD, gtg/L

0.20

CV, %

5.04

x, ag/L

3.98

SD, gJL

0.07

CV, %

1.88Controlsample 2 16.7 16.5 0.23 1.39 16.11 0.44 2.76Control sample 3 31.6 30.4 0.86 2.83 29.2 1.06 3.63Pool 1.28 0.05 3.90 1.27 0.07 5.51

ResultsThe TnI assay evaluated in our study (Table 1) has a good

precision, with CVs ranging from 1.39% (. = 16.5 .tgfL) to

5.51% ( = 1.27 j.tg/L); the lowest detection limit (analytical

sensitivity) of the measuring range in our study was 0.26 .tg/L.Furthermore, we found that the stability of the reagents (8 h at

room temperature) and of the calibration curve (at least 1

month) was satisfactory.

In healthy subjects TnI concentrations had a skewed distri-

bution, ranging from 0 to 0.3 p.gfL; >99% of values were lowerthan the detection limit of the assay.

The TnI concentrations observed in all patients studied andexpressed as median, 10%, and 90% are reported in Table 2.

With the exception of AM! and unstable angina patients, cardiac

TnT could not be detected by this assay in other patients studied.

In seven patients with unstable angina (classIII and class II),

detectable amounts of TnT without significant variations during

monitoring were observed. In our patients, no cardiac events

during the hospitalization and during the monitoring were

clinically diagnosed, even in the patient with the highest value of

TnT (0.9 .tgfL).

A graphic test evaluation of the TnI assay in AM! is

presented in Fig. 1. In Fig. 1A, the double histogram compares

the frequency distributions of TnT values observed 4 h after

onset of chest pain in AM! cases (group 2, upper field) with

those obtained in group 3 subjects (lower field), arranged in 10

nonequidistant classes,ranging from 0 to 30 .tg/L. The sensi-

tivity and specificity curves are a function of the test value and

reflect different distributions. On the basis of these curves, a

discriminator value of 1 .tg/L yields a sensitivity of 0.769 and aspecificity of 1.0 with a diagnostic efficiency of 0.943. For

evaluating the clinical specificity under worst-case situations, we

replotted the TnI data of AMI cases (Fig. 1B, upper field) in

Table 2. Tnl concentratIons (tg/L) In patients studIedDiagnosis n 10% 50% 90%

Chronic skeletal muscle 11 0.00 0.20 0.30diseases

Multiple traumas withoutchest contusion

Chronic renal failureChest pain

Unstable angina*Stable anginaAMI**

* First sample available.**4 h after onset of chest pain.

Cutoff

1 g/L

160 U/L**

CK-MB 5 pg/L

Myoglobin 50 .g/L

ClinicalChemistry 42, No. 9, 1996 1463

*Obtained by adding to group 3 group 4 patients.

**All patients (men and women).

comparison with those observed in groups 3 and 4 patients (Fig.

1B, lower field). No difference in the clinical specificity of the

TnT assay was observed in two clinical conditions considered.

The performances of all markers considered in our study were

calculated in the described population with the same approach

(Table 3). During the first 4 h after onset of chest pain,

myoglobin showed the highest sensitivity, followed by CK-MB

and TnI, whereas total CK was significantly less sensitive. On

the other hand, during - 3 days of monitoring, we observed for

all tests a time-dependent sensitivity (Fig. 2). In particular,

during the first 2 h after the onset of symptoms, myoglobin

showed the highest sensitivity (1.0), while those of total CK,

CK-MB, and Tn! were -0.60. The maximum value of sensitiv-

ity (1.0) was observed from 7 to 24 h for CK-MB and from 10

to 21 h for total CK. Tn! showed a constantly high sensitivity

(1.0) from the 10th hour forward, while, as expected, those of

the other markers progressively decreased, being 0.20 (myoglo-

bin), 0.37 (CK-MB), and 0.62 (total CK) 70 h after the onset ofsymptoms. Furthermore, the clinical specificity of all markers

studied, except that of TnT, decreased markedly if the evaluationwas carried out in a worst-case situation, i.e., in patients with

concomitant skeletal or renal diseases.In another approach for assessing the diagnostic efficiency of

TnT compared with that of the traditional biochemical AM!

markers considered in our study, ROC curves were generated

(Fig. 3). Despite minor differences, the area under the TnT ROC

curve (0.932; 95% confidence interval 0.83-1.0) was almost the

same as the unit area under the CK-MB curve (0.956; 95%

confidence interval 0.87-1.0) and total CK (0.920; 95% confi-

dence interval 0.81-1.0), showing no significant difference in

comparison with the ROC curve for myoglobin (0.960; 95%

confidence interval 0.88-1.0), the last being the most useful

marker in diagnosing myocardial infarction within 4 h after the

onset of chest pain (Fig. 3A). On the other hand, in patients with

concomitant skeletal muscle damage or with renal diseases (Fig.

3B), the ROC curve for TnI (0.926; 95% confidence interval

0.82-1.0) was greater than that for CK-MB (0.830; 95% confi-

dence interval 0.68-0.97) and significantly different from those

for total CK (0.685; 95% confidence interval 0.5 1-0.86;

z-score = 2.604) and myoglobin (0.485; 95% confidence inter-

val 0.11-0.86; z-score = 2.2 36). In these patients the relative

index improves the clinical specificity of CK-MB assay, the area

under the ROC curve being 0.917 (95% confidence interval

0.81-1.0).

The release kinetics, and the relative increase in values of

markers studied in AM! patients, are summarized in Table 4.

The first increased value in Tn! concentrations after infarction

(interquartile range 3.25-6.0 h) and the peak (interquartile

range 11.5-24.0 h) occurred later than in myoglobin and in

CK-MB concentrations, but increased serum TnI concentra-tions persisted much longer (>96 h). Furthermore, serum TnT

concentrations exceeded the discriminator limit by >158-fold.

The pattern of serum TnT release in AM! patients shows a

monophasic time-dependent kinetic (Fig. 4, upper field). A

marked increase in serum TnI was observed on day 1, resultingin a median peak value at -40 times the discriminator limit at a

median time of 18 h after the onset of pain. The initially releasedTnT was rapidly eliminated from the circulation, and a long

plateau effect was observed from the second to the fourth day

after onset of symptoms when TnT concentrations were persis-tently -threefold the discriminator limit. Changes in TnT were

similar to those of CK-MB (Fig. 4, lower field), the most strikingdifference being the long release of TnT, the concentrations of

which remained increased >96 h after onset of chest pain. Theprolonged serum increase of this marker provided a diagnostic

window from 10 to at least 96 h after onset of pain, during which

TnT was detectable in all serum samples of all patients. In our

cases the pattern of its release in AM! patients was not influ-

enced by reperfusion of the infarct-related artery, being

monophasic in all patients.

rnscussionOur findings demonstrate that the fluoroenzymometric method

for TnT assay has a good analytical performance. The evaluated

assay, commercially available, uses the same antibodies origi-

nally developed by Bodor et al. [17], and extensively reported inexperimental studies [28, 29]. The relation between the current

assay and the research method calculated by the weighted

least-squares equation is 0.211 (research assay)-0.0l9 (Laden-

son JH, Division of Laboratory Medicine, Washington Univer-

sity, St. Louis, MO, personal communication). Thus, previously

reported cutoff values cannot be taken into consideration using

the commercial procedure.

We found that Tn! has several advantages over traditional

Table 3. DIagnostIc performances (and 95% confidence Intervals) of bIochemIcal markers for AMI 4 h after onset of chestpaIn.

Sensitivity

0.769 (0.46-0.95)

0.692 (0.39-0.91)

0.923 (0.64-1.00)

1.000 (0.75-1.00)

Specificity

1.00 (0.95-1.00)1.00* (0.91-1.00)0.923 (0.75-0.99)0.660* (0.52-0.78)

0.961 (0.80-1.00)0.785* (0.66-0.88)

0.923 (0.75-1.00)0.464* (0.33-0.60)

Efficiency

0.943 (0.90-0.99)0.956* (0.84-0.99)0.846 (0.76-0.97)0.666* (0.53-0.78)0.948 (0.83-0.99)0.811* (0.70-0.90)0.984 (0.83-0.99)0.565* (0.44-0.68)

TnI

N.,.. CK

1.0

0.8

0.6

.

0.4

0.2

0.0

#{163}

-

degradation, the content of free cytoplasmic pool of TnT being

lower than that reported for TnT (-3% vs 6%, respectively)

[28]. Nevertheless, this finding must be confirmed in a larger

population because the results reported in the literature by

different authors are contradictory [31, 35]. In a larger popula-

tion, the diagnostic potential of the TnI assay for detecting

minor myocardial damages should also be verified; no conclu-

1.0

0---C

CK-MB

b

- -6- -6- -6- -4

Myoglobin

:

V

0.8

0.6

0.4

0.2

0.0

A

-.-.-.-#{149}-.. CK-MB#{149} Tnl

.-.-*-- Myoglobin

0 5 10 15 20 25 30 35 40 45 50 55 60 65 70

Hours after the onset of chest pain

Ag. 2. Diagnostic sensitivity of mnl, total CK activity, CK-MB massconcentration, and myoglobin in relationto hours after onset of AMIsymptoms.

serological markers of myocardial injury, the greatest one being

its cardiospecificity[30]. TnI measurement is therefore partic-

ularly helpful in the assessment of patients with myocardial

ischemia and skeletal muscle damage [31, 32]. Tn! results are

also easy to interpret because serum concentrations are usuallyvery low (between 0 to 0.3 j.g/L) and increase substantially (up

to 158 times the discriminator) in response to an acute infarc-

tion. At the decisional concentration for AM! chosen in our

study, 1 g/L, the assay shows a good diagnostic efficiency

(0.943) with an excellent clinical specificity (1.0), even in

patients in whom traditional biochemical markers of AMI are

adversely influenced by diverse underlying clinical situations

[32, 33]. Tn fact, in clinical pratice it is necessary to rule out AM!

not only in patients with angina and other heart diseases, but

also in patients with muscular or renal diseases admitted to the

hospital with aspecific symptoms.

Despite its structural localization, in AM!, TnT appears in thecirculation within a few hours after the onset of chest pain [34],and high concentrations persist for at least 4 days [35]. Thissustained increase in the serum concentrations probably reflects

a continuing release of this protein from disintegrating myofila-ments; consequently there is a long diagnostic window, from 10

to at least 96 h after onset of pain, during which the TnT

concentration exceeded the discriminator limit for AM! in all

serum samples of our patients. This release pattern, in commonwith that of other contractile proteins, increases the likelihood

of a positive test result, particularly in the subacute phase of

infarction [8, 9].The serum release kinetic observed for Tn! in our study was

monophasic in all patients studied, irrespective of the outcome

of thrombolytic treatment. This behavior, different from that

observed in some studies for other structural proteins such as

troponin T (TnT) [9, 11], may suggest that the rapid loss of the

cytoplasmic pool was superimposed on prolonged myofibrillar

0.0 0.2 0.4 0.6 0.8 1.0

FP (1-specificity)

0.6

0.4’I!

/ -.-#{149}-.-.-.. CK-MB

/ , #{149}0.2’ y

/ 4 -.-*---. MyoglobinI-...-

/0.0 ‘ I I I I

0.0 0.2 0.4 0.6 0.8 10FP (1-specificity)

Fig.3. ROCcurves for total CK activity, CK-MB mass concentration, Tnl,and myoglobin 4 h after the onset of chest pain.(A) Values of specificity obtained by using group 3 patients (chest pain, stableand unstable angina) as comparison group. (B) Valuesofspecificityobtainedbyusingtogethergroups 3 and 4 patients (chronic skeletal muscle diseases,multiple traumas without chest contusion, chronic renal failure) as comparisongroup.

1464 Zaninotto et al.: Fluoroenzyrnometric measurement of cardiac Tn!

25%-‘-- 50*

b 75%

Hours after onset of chest pain96

--‘#{149} 25%-‘- 50%“b 75%

Clinical Chemistry 42, No. 9, 1996 1465

Fig.4. Serum concentration changes of cardiac Tnl (upper field) andCK-MB mass concentration (lower field) in 20 AMI patients.

Table 4. Release kInetIcs for ml, CK-MB (mass concentration), total CK actIvIty, and myoglobln Inmyocardlal Infarction.ml CK-MB CK Myoglobin

Cutoff I p.g/L 5 JL 190 U/L (M)

160 U/L (F)

50 p.g/L

Time (h) to first increased value (50%) 4.50 4.50 4.50 3.50

(25%, 75%) 3.25-6.00 3.25-4.87 3.25-6.00 2.50-4.50

Time (h)to peak (50%) 18.00 15.00 20.00 6.00

(25%, 75%) 11.50-24.00 10.00-20.00 14.25-27.00 6.00-10.00

Time (h)of normalization(50%) >96 70 90 24

Multiplesof the discriminatorlimitforAMI 158 110 37 53

sion, in fact, can be drawn from our data, being too limited in

the number of patients with angina.

Our results confirm that TnT measurement is a substantial

advance in the laboratory diagnosis and monitoring of AM!, and

is of particular value because a rapid and automated Tn!

determination is now available. The high analytical sensitivity of

the current method, its turnaround time, and the practicability

enabled us to detect cardiac TnT very early after A1vII, in an

emergency situation.

a.

C

aI-

60

50

40

30

20

l0

0

06121824303642485460667278849096

Hours after onset of chest pain

The unique properties of this protein and, in particular, its

high diagnostic specificity (1.0) and sensitivity in the early phase

of AMI (0.769 during 0-4 h after onset of pain), the high

concentration gradient between myocardial cells and normal

blood, and the stability and rate of elimination suggest that theTn! measurements could become a valuable differential diag-

nostic tool to investigate myocardial damage in emergency room

patients, even in cases of concomitant skeletal muscle injury, as

occurs in polytraumatized patients. Tn fact, as reported in our

study, the measurement of the traditional markers in these

patients is ineffectual because of its lack of specificity. In our

experience myoglobin remains the more sensitive marker in the

earlier phases of AM!. Differences in the sensitivity of this

marker described in the literature [34, 36, 37] can be attributed

to the lack of standardization and optimization of methods used

for its measurement.

In conclusion, the determination of TnT in combination with

myoglobin, the earlier marker of myocardial damage, allows areliable and timely diagnosis to be made; myocardial necrosis to

be ruled out in patients with renal failure [33] or with skeletal

muscle myopathies; and the course of myocardial damage to be

monitored during therapy.

We are indebted to R. Brandolese (Servizio di Anestesia eRianimazione, Padova, Italy), C. Angelini (Clinica Neurologica,

Padova, Italy), and A. Piccoli (!stituto di Medicina Clinica,

Padova, Italy) for the valuable cooperation in collecting sera and

data from patients.

References1. Bhayana V, Henderson AR. Biochemical markers of myocardial

damage. Clin Biochem 1995;28:1-29.2. Adams JE, Abendschein DR, Jaffe AS. Biochemical markers of

myocardial Injury. Is MB creatine kinase the choice for the1990’s? Circulation 1993:88:750-63.

3. Mair J, Artner-Dworzak E, Dienstl A, Lechleitner P, Morass B, SmidtJ, et al. Early detection of acute myocardial infarction by measure-ment of mass concentration of creatine kinase-MB. Am J Cardiol1991;68:1545-50.

4. Puleo PR, Meyer 0, Wathen C, Tawa CB, Wheeler 5, Hamburg Ri,

et a). Use of a rapid assay of subforms of creatine kinase MB to

diagnose or rule out acute myocardial infarction.N Engi J Med

1994;331:561-6.

1466 Zaninotto et al.: Fluoroenzymometric measurement of cardiac TnI

5. Ohman EM, Casey C, Bengtson JR, Pryor D, Tormey W, Horgan JH.Early detection of acute myocardial infarction: additional diagnos-tic information from serum concentrations of myoglobin in patientswithout ST elevation. Br Heart J 1990;63:335-8.

6. Rabitzsch G, Mair i,LechleitnerP, NollF, Hofmann U, Krause EG,

et al.lmmunoenzymometric assay of human glycogen phosphory-

lase isoenzyme BB in diagnosis of ischemic myocardial injury. Clin

Chem 1995;41:966-78.7. Katus HA, Looser S, Hallermayer K, Remppis A, Scheffold T,

Borgya A, et al. Development and in vitro characterization of a newimmunoassay of cardiac troponin 1. Clin Chem 1992;38:386-93.

8. Katus HA, Yasuda 1, Gold HK, Leinbach R, Strauss HW, Waksmon-skiC, et al.Diagnosis of acute myocardial infarctionby detection

of circulatingcardiac myosin lightchains. Am J Cardiol1984;54:

964-70.9. Burlina A, Zaninotto M, Secchiero 5, Rubin D, Accorsi F. Troponin

I as a marker of ischemic myocardial injury. Clin Biochem 1994;27:113-21.

10. Donnely R, Hillis WS. Cardiac troponin I [Commentaryl. Lancet1993;34:410-1.

11. Katus HA, Remppis A, Scheffold T, Diederich KW, Kubler W.Intracellular compartmentation of cardiac troponin I and itsrelease kinetics in patients with reperfused and nonreperfusedmyocardial infarction. Am J Cardiol 1991;67:1360-7.

12. Wilkinson JM, Grand RJA. Comparison of amino acid sequence oftroponin I from differentstriatedmuscles. Nature 1978;27I:31-5.

13. Vallins WJ, Brand NJ, Dabhade N, Butler-Browne G, Yacoub MH,Barton PJR. Molecular cloningof human cardiactroponin Iusing

polymerase chain reaction.FEBS Lett1990;270:57-61.

14. SagginL,Gorza L,Ausoni S, SchiaffinoS. Troponin Iswitchingin

the developing heart. J Biol Chem 1989;264:16299-302.15. Katagiri I, Kobayashi Y, Sasai Y,Toba K, Niitani H. Alterations in

cardiac troponin subunits in myocardial infarction. Jpn Heart J1981;22:653-64.

16. Cummins B, Aukland ML, Cummins P. Cardiac-specific troponin-lradioimmunoassay in the diagnosis of acute myocardial infarc-tion. Am Heart i 1987:113:1333-44.

17. Bodor GS, Porter S, Landt Y, Ladenson JH. Development ofmonoclonal antibodies for an assay of cardiac troponin-l andpreliminary results in suspected cases of myocardial infarction.Clin Chem 1992:38:2203-14.

18. Pasternak RC, Braunwald E, Sobel BE. Acute myocardial infarc-tion. In: Braunwald E, ed. Heart disease. Philadelphia: WB Saun-ders, 1988:1222-313.

19. Braunwald E. Unstable angina: a classification. Circulation 1989;80:410-4.

20. RuggJA, Rigi CT, Leung K, Lamar SL, Welsh M, LeBlanc M, et al.Radialpartitionimmunoassay appliedto automated quantification

of human choriogonadotropin with use of two monoclonal antibod-ies. Clin Chem 1986;32:1844-8.

21. International Federation of Clinical Chemistry. IFCC methods forthe measurement of catalytic concentration of enzymes; Part 7.IFCC method for creatine kinase (ATP: creatine N-phosphotrans-ferase, EC 2.7.3.2). IFCC recommendation. Clin Chim Acta 1990;190:S4-S40.

22. Chapelle JP, El Allaf M. Automated quantification of creatinekinase MB isoenzyme in serum by radial partition immunoassay,with use of the Stratus analyzer. Clin Chem 1990;36:99-101.

23. Delanghe JR, Chapelle JP, Van derschueren SC. Quantitativenephelornetric assay for determining myoglobin evaluated. ClinChem 1990;36:1675-8.

24. Gerhardt W, Keller H. Evaluationoftestdata from clinicalstudies.

I. Terminology, graphic interpretations, diagnostic strategies andselection of sample groups. II. Critical review of the concepts ofefficiency,receiver operating characteristic (ROC) and likelihoodratios.Scand J Lab Invest1986;181(Suppl):11-74.

25. Hanley JA, McNeil BJ. The meaning and use of the area under areceiveroperating characteristic(ROC) curve. Radiology 1982;

143:29-36.

26. Hanley JA, McNeil BJ. A method of comparing the area underreceiver operating characteristic curves derived from the samecases. Radiology 1983;148:839-943.

27. Gardner Mi, Gardner SB, Winter PD. Confidence intervalanalysis

microcomputer program. London: British Medical Journal Publish-ing Group, 1989:77pp.

28. Adams JE III, Schechtman KB, Landt Y, Ladenson JH, Jaffe AS.Comparable detection of acute myocardial infarction by creatinekinase MB isoenzyme and cardiac troponin I. Clin Chem 1994;40:1291-5.

29. Apple FS, Sharkey SW, Henry ID. Early serum cardiac troponin Iand I concentrations after successful thrombolysis for acute

myocardial infarction [Tech Brief]. Clin Chem 1995;41:1197-8.

30. Bodor GS, Porterfield 0, Voss EM, Smith S, Apple FS. CardiactroponinIisnot expressed infetaland healthyor diseased adult

human skeletalmuscle tissue.ClinChem 1995;41:1710-5.

31. Larue C, CalzolariC, BertinchantJP,LeclercqF,GrolleauR, Pau B.Cardiac-specificimmunoenzymometric assay of troponin I inthe

early phase of acute myocardial infarction. Clin Chem 1993;39:972-9.

32. Adams JE Ill, Bodor GS, D#{224}vila-RomanVG, Delmez JA, Apple FS,Ladenson JH, et al. Cardiac troponin I. A marker with highspecificity for cardiac injury. Circulation 1993;88:lOi.-6.

33. Li D, Jialal I, Keffer J. Greater frequency of increased cardiactroponin I than increased cardiac troponin I in patients withchronic renal failure [Letter]. Clin Chem 1996;42:114-5.

34. Mair J, Morandell 0, Genser N, Lechleitner P, Dienstl F, Puschen-dorf B. Equivalent early sensitivities of myoglobin, creatine kinaseMB mass, creatine kinase isoform ratios, and cardiac troponins Iand T for acute myocardial infarction. Clin Chem 1995;41:1266-72.

35. Mair I, Wagner I, Puschendorf B, Mair P, Lechleitner P, Dienstl F,et al. Cardiac troponin I to diagnose myocardial injury. Lancet1993;341:838-9.

36. Kilpatrick WS, Wosornu 0, McGuinness JB, Glen ACA. Earlydiagnosis of acute myocardial infarction: CK-MB and myoglobincompared. Ann Clin Biochem 1993;30:435-8.

37. Montague C, Kircher I. Myoglobin in the early evaluation of acutechest pain. Am I Clin Pathol 1995;104:472-6.