BrHeartJ 1994;72:112-118

Contribution of creatine kinase MB mass

concentration at admission to early diagnosis ofacute myocardial infarction

Andries J Bakker, Jozef P M C Gorgels, Bob van Vlies, Mark JW Koelemay, Rina Smits,Jan G P Tijssen, Frans D M Haagen

Department ofClinical Chemistry,Klinisch ChemischLaboratorium,Leeuwarden, TheNetherlandsA J BakkerR SmitsDepartment ofCardiology, MedischCentrum Leeuwarden,Leeuwarden, TheNetherlandsF D M HaagenDepartment ofClinical Chemistry,Cardiology, andClinical Epidemiologyand Biostatistics,Academisch MedischCentrum,Amsterdam, TheNetherlandsJ P M C GorgelsM JW KoelemayJ G P TijssenB van VliesCorrespondence to:Dr A J Bakker, Departmentof Clinical Chemistry,Klinisch ChemischLaboratorium, Box 850,8901 BR Leeuwarden, TheNetherlands.

Accepted for publication5 January 1994

AbstractObjective-To assess the diagnostic valueat admission of creatine kinase MB massconcentration, alone or in combinationwith electrocardiographic changes, insuspected myocardial infarction.Design-Prospective study of all consec-utive patients admitted within 12 hoursafter onset of chest pain to a coronarycare unit for evaluation of suspectedmyocardial infarction.Setting-Large regional hospital.Patients-In 297 patients creatine kinaseand creatine kinase MB activities andcreatine kinase MB mass concentrationwere determined. Myocardial infarctionaccording to the criteria of the WorldHealth Organisation was diagnosed in154 patients and excluded in 143 patients(including 70 with unstable angina pec-toris).Results-Sensitivitylspecificity for crea-tine kinase MB mass concentration inpatients admitted within 4 hours and 4-12hours after onset of chest pain were450/6194% and 760/o179% respectively.Corresponding values for creatine kinaseactivity were 200/1890/6 and 59%1lo83%, andfor creatine kinase MB activity 161/o187%and 530%1l87%. Raised creatine kinase MBmass concentration was seen in 17% ofpatients with unstable angina pectoris.Stepwise logistic regression analysisshowed that independent predictors ofacute myocardial infarction in patientsadmitted within 4 hours after onset ofchest pain were electrocardiographicchanges and creatine kinase MB massconcentration on admission; in patientsadmitted 4-12 hours after the onset ofpain independent predictors wereelectrocardiographic changes and crea-tine kinase MB mass concentration andactivity.Conclusion-Creatine kinase MB massconcentration is a more sensitive markerfor myocardial infarction than the activ-ity of creatine kinase and its MB iso-enzyme. Electrocardiographic changeson admission in combination with crea-tine kinase MB mass concentration(instead of creatine kinase and creatinekinase MB activities) are best in diagnos-ing myocardial infarction.

(Br Heart _ 1994;72:112-118)

The diagnosis of acute myocardial infarctionis generally based on a history of chest pain,specific electrocardiographic changes, and thetypical rise and fall in the activity of cardiacenzymes. At hospital admission the diagnosticvalue of chest pain,' single electrocardio-graphic recordings,23 and single determina-tions of the activity of creatine kinase and itsMB isoenzyme-9 is limited. Serial measure-ment of creatine kinase and creatine kinaseMB activities, as a reflection of myocardialcell damage, however, plays the most impor-tant part in establishing the definitive as wellas the early diagnosis of myocardial infarc-tion,'0 because these enzymes are easily andrapidly measured with automated analysers.The immunological determination of themass concentration of creatine kinase MB is arecent advance towards greater sensitivity andspecificity.1'-'5 Creatine kinase MB mass con-centration is often raised in the early phase ofmyocardial infarction."3 16 Therefore, measur-ing creatine kinase MB mass concentrationmight be advantageous as it can be measuredrapidly'7 like creatine kinase and creatinekinase MB activity.

In this study we assessed the validity of thereference ranges for a sensitive assay for crea-tine kinase MB mass concentration and crea-tine kinase and creatine kinase MB activities.We then evaluated the diagnostic propertiesof creatine kinase MB mass concentration atadmission and compared them with those ofthe currently used methods for measuringcreatine kinase and creatine kinase MB activi-ties and electrocardiographic changes onadmission. Finally, we evaluated the predic-tive power of combining these electrocardio-graphic changes with each of thesebiochemical markers at the time of admission.

Patients and methodsThe study population consisted of all consecu-tive patients admitted over nine months to thecoronary care unit of the Medisch CentrumLeeuwarden, a large regional hospital, forevaluation of suspected myocardial infarctionwithin 12 hours after the onset of chest pain.Patients were not included when they hadbeen resuscitated or received defibrillation.Standard 12 lead electrocardiograms obtainedat admission, 24 hours after admission, and atdischarge were used for establishing the finaldiagnosis. Blood samples for the determina-tion of activities of creatine kinase, its MB

12 on F

ebruary 26, 2020 by guest. Protected by copyright.

http://heart.bmj.com

/B

r Heart J: first published as 10.1136/hrt.72.2.112 on 1 A

ugust 1994. Dow

nloaded from

Contribution of creatine kinase MB mass concentration at admission to early diagnosis of acute myocardial infarction

isoenzyme, lactate dehydrogenase, aspartateaminotransferase,, and alanine aminotrans-ferase and the mass concentration of creatinekinase MB were taken at admission, threetimes in a fixed time schedule for the first 24hours, and once- daily until discharge.

DIAGNOSTIC CRITERIAElectrocardiographic findings on admissionwere classified as normal, ischaemic, or non-conclusive. Normal findings were definedaccording to the criteria of the New YorkHeart Association.'8 Electrocardiographicrcsults were classified as ischaemic if the STsegment was depressed by more than 1 mm0-08 s after the J point or the ST segment waselevated by more than 1 mm (in at least twocontiguous leads) and provided that QRScomplexes were normal and there was rightbundle branch block. All other findings inelectrocardiograms on admission were classi-fied as non-conclusive.

Patients were classed as having myocardialinfarction, unstable angina pectoris, or othercardiac disease or as having no acuteischaemic heart disease. The final diagnosiswas established by an independent cardiolo-gist who did not know the initial creatinekinase MB mass concentration. The diagnosisof myocardial infarction was established if apatient had at least two of the three classicfindings. The first was a clinical history ofischaemic chest discomfort of more than 30minutes' duration. The second was the devel-opment of typical changes in at least two leadsof the electrocardiogram. Q waves greaterthan 0-04 s or an R/S ratio greater than 1 inleads Vl and V2 was termed Q wave infarc-tion; ST segment depression of more than 1mm 0-08 s after the J point or ST segment ele-vation of more than 1 mm persisting for atleast 24 hours was termed non-Q wave infarc-tion. The third was a time dependent changein serum creatine kinase and creatine kinaseMB activities, showing an initial rise (with apeak value of creatine kinase MB activityexceeding twice the upper limit of normal val-ues) and subsequent fall.

Unstable angina pectoris was defined astypical anginal pain at rest in combinationwith reversible or persistent electrocardio-graphic changes but without a peak value inactivity of creatine kinase MB that exceededtwice the upper limit of normal values.

Patients were classed as having no acuteischaemic heart disease when there was a defi-nite alternative source of the chest pain or anatypical history with a negative result on stresselectrocardiography, myocardial scintigraphy,or coronary angiography.

ANALYTICAL METHODSCatalytic activity of creatine kinase was mea-sured at 30°C with a Hitachi 717 analyser bythe method proposed by the InternationalFederation of Clinical Chemistry6 usingcommercial reagents (product numbers475742 and 475769) supplied by BoehringerMannheim (Mannheim, Germany). Theupper limit of normal values in this assay

was 130 U/I for men and 90 U/I for women.Catalytic activity of creatine kinase MB was

measured by immunoinhibition using anappropriate antibody (product number418234; Boehringer Mannheim) and the crea-tine kinase assay method.'9 Like creatinekinase activity, creatine kinase MB activitywas measured with the Hitachi 717 analyserat 30°C. The upper limit of normal values inthis assay was 15 U/1.

Creatine kinase MB mass concentration inserum was measured according to the manu-facturer's instructions by a microparticle cap-ture enzyme immunoassay with the IMxanalyser from Abbott Laboratories (AbbottPark, Illinois; product number 2207-20). Thesensitivity of this assay was reported to be0-2 .ug/1.20 The measuring range covers con-centrations from 0-2 to 300 pug!l. The upperlimit of normal values in this assay was estab-lished in this study using serum samples from77 male and 45 female apparently healthyblood donors. The upper limit of normal(97-5th percentile) was estimated at 50 pug/laccording to the proposed guidelines ofthe International Federation of ClinicalChemistry.2'

STATISTICAL ANALYSISSensitivity, specificity, positive and negativepredictive values, relative risk, and likelihoodratios were calculated to determine the diag-nostic properties of electrocardiographic find-ings and the biochemical markers. Thelikelihood ratio of a positive test is the proba-bility of a positive test in the presence of dis-ease divided by the probability of a positivetest in the absence of disease.

Differences in baseline characteristics (age,sex, clinical history, and risk factors) betweenpatients with and without myocardial infarc-tion were determined by a X2 test or byFisher's exact test for categorical variables andby Student's t test for continuous variables (a Pvalue <0 05 was considered to be significant).A stepwise logistic regression analysis

(PROC LOGISTIC22) was performed to identifythe variables that contributed significant inde-pendent information to distinguish patientswith myocardial infarction from those with-out. At each step of the logistic regressionanalysis the variable that was most signifi-cantly associated with the presence of myocar-dial infarction was selected until no variablesmeeting the 0 05 significance level (x2 test) forentering the model were left. The probabilityof myocardial infarction in each patient givenelectrocardiographic information alone and incombination with the biochemical markerswas calculated by using the coefficientsobtained in the logistic regression analysis.

ResultsOur study included 297 patients (148 of themadmitted within 4 hours after the onset ofanginal chest pain). Acute myocardial infarc-tion was confirmed in 154 patients (75 Qwave); 50 of them (32%) received thrombo-lysis. In 143 patients acute myocardial

113 on F

ebruary 26, 2020 by guest. Protected by copyright.

http://heart.bmj.com

/B

r Heart J: first published as 10.1136/hrt.72.2.112 on 1 A

ugust 1994. Dow

nloaded from

Bakker, Gorgels, van Vlies, Koelemay, Smits, Tijssen, Haagen

Table 1 Baseline characterstics ofpatients admitted within 12 hours after onset of chespain. Values are numbers (percentages) ofpatients unless stated othervise

Patients with Patients withoutacute myocardial infarction acute myocardial infarcti(n=154) (n=143)

Men 111 (72) 96 (67)Mean (SD) age (years) 65 (11.9) 62 (12-2)History of angina 60 (39) 76 (53)*Previous:

Acute myocardial infarction 39 (25) 50 (35)Percutaneous transluminal

coronary angioplasty 3 (2) 10 (7)Coronary artery bypass grafting 7 (5) 7 (5)

Diabetes 20 (13) 11 (8)Hypertension 46 (30) 48 (34)Hyperlipidaemia 70 (46) 60 (42)Smoking 63 (41) 53 (37)

*Significant difference between patients with and without myocardial infarction group(P=0-023).

infarction was excluded: 70 patients weclassed as having unstable angina pecto:(three patients received thrombolysis) and ]as having no acute ischaemic heart disea(one patient received thrombolysis). 0thcardiac diseases were diagnosed in 11 patier(one patient received thrombolysis), fopatients had no definitive diagnosis. No sinificant differences in baseline characteristi

100 nNo acute ischaemic heart disease

Creatine kinase activity(men, 130 U/I; women, 90 U/I)

Xlt * Creatine kinase MB activity(15 UAl)

0 Creatine kinase MB massconcentration (5 [tg/l)

172 4 6 8

Normalised concentration(multiples of upper limit of normal values)

Unstable angina pectoris

4 6 82

Normalised concentration(multiples of upper limit of normal values)

Frequency of biochemical markers at admission above normalised concentrations inpatients with no acute ischaemic heart disease and in those with unstable angina pectorisVahles in parentheses after biochemical markers are upper limit ofnormal values.

(age, sex, clinical history, and risk factors)with the exception of a history of angina werefound between patients with and withoutmyocardial infarction (table 1).

VERIFICATION OF NORMAL RANGES INPATIENTS WITHOUT ACUTE ISCHAEMIC HEARTDISEASEThe upper limit of normal values for the threemarkers was verified using the resultsobtained at admission for the patients (n =58) classed as having no acute ischaemic heartdisease. For creatine kinase MB mass concen-tration 95% (in estimating the normal rangeusually the 95% range is calculated) of theresults obtained in the admission sampleswere less than or equal to the upper limit ofnormal values, while only 79% and 84% of

re the results for creatine kinase and creatineris kinase MB activity were within the normal58 range (figure). The 5% limit for creatinese kinase MB mass concentration was 40 ,ug/l,er for creatine kinase MB activity 41 U/1, and forits creatine kinase 1-56 times the upper limit ofur normal values.ig- Further analysis of the data showed thatics most of the raised creatine kinase and creatine

kinase MB activities were false positive resultswith respect to the diagnosis of myocardialinfarction. These activities were raised in sixpatients because of muscle injury as the ratioof creatine kinase MB to total creatine kinaseactivity was normal (a ratio <6% excludesmyocardial infarction), in three patientsbecause of the presence of a macromolecularcreatine kinase (creatine kinase in two ofthemand creatine kinase MB in all three), and inseven patients because haemolysis freedadenylate kinase from erythrocytes (creatinekinase in four patients, and creatine kinaseMB in all seven). Two patients had a raisedcreatine kinase MB concentration. One hadnon-specific thoracic pain and a macromolec-ular creatine kinase with a creatine kinase MBmass concentration of 75,ug/l; the other had

10 rheumatoid arthritis and hypertension and acreatine kinase MB mass concentration of 6-0jug/l, but total creatine kinase and creatinekinase MB activities were low (creatine kinase20 U/1; creatine kinase MB 3 U/1). The raisedcreatine kinase MB mass concentration couldnot be explained in either patient. Electro-cardiographic findings on admission wererated ischaemic in six (10%) patients, non-conclusive in seven (12%), and normal in 45(78%) patients with no acute ischaemic heartdisease.

FREQUENCY OF RAISED RESULTS IN PATIENTSWITH tJNSTABLE ANGINA PECTORISThe results at admission for patients withunstable angina pectoris (n = 70) were alsocompared with the upper limit of normal

4 values (figure). In this group 95% of the10 results for the activities of creatine kinase and

creatine kinase MB was within the normalrange. For creatine kinase MB mass concen-tration, however, 17% (12/70) of the resultswere above the upper limit of normal values,and 5% of the results were > 12 ug/l.

e-

0

'5 80

0

60A(a

0

4) 40

c 200

03

LLv0

100

80

60

40

20

0

0

3A

)0

01)

n

0

c

LL

114 on F

ebruary 26, 2020 by guest. Protected by copyright.

http://heart.bmj.com

/B

r Heart J: first published as 10.1136/hrt.72.2.112 on 1 A

ugust 1994. Dow

nloaded from

Contribution of creatine kinase MB mass concentration at admission to early diagnosis of acute myocardial infarction

Table 2 Diagnostic sensitivity, specificity, and positive and negative predictive values (%) at admission for biochemicalmarkers ofmyocardial damage

Time between onset ofpain and admission (h)

<4 4-12

Positive Negative Positive Negativepredictive predictive predictive predictive

Marker Sensintvity Specificity value value Sensitivity Specificity value value

Creatine kinase activity 20 89 66 51 59 83 79 65Creatinekinase MB activity 16 87 57 50 53 87 82 63Creatine kinase MB mass

concentration 45 94 89 62 76 79 80 75

Electrocardiographic findings on admissionwere rated ischaemic in 15 (21%) of thepatients, non-conclusive in 20 (29%), andnormal in 35 (50%).

FREQUENCY OF RAISED RESULTS IN PATIENTSWITH ACUTE MYOCARDIAL INFARCTIONIn patients with acute myocardial infarctionraised results at admission were encounteredmore frequently for creatine kinase MB massconcentration (60%) than for creatine kinaseand creatine kinase MB activities (40% and34%). Electrocardiographic findings onadmission were rated ischaemic in 66 (43%)patients, non-conclusive in 60 (39%) patients,and normal in 28 (18%) patients. On admis-sion 13 patients with acute myocardial infarc-tion (8%) had normal electrocardiographicfindings combined with a diagnostic creatinekinase MB activity. One patient receivedthrombolysis because of electrocardiographicchanges, but subsequent rises in creatinekinase and creatine kinase MB activities werenot observed. In this patient creatine kinase

Table 3 Distribution of results and statis4cal characteristics of diagnostic tests formyocardial infarction

Patients Patientswith withoutacute acute Relative nsk Likeli-myocardial myocardial (95% confidence hoodinfarction infarction interval) ratio

Admission within 4 h after onset of chest painElectrocardiographic findings:

Ischaemic (n = 47) 35 12 3-41 (2-22 to 5-22) 2-76Non-conclusive (n = 46) 29 17 2 89 (2-00 to 4-18) 1-62Normal (n= 55) 12 43 0-26

Creatine kinase*:Normal (n = 125) 61 64 0 90Raised (n = 23) 15 8 1 34 (0 90 to 1 99) 1-78

Creatine kinase MB:Activity:Normal (< 15 U/1) (n = 127) 64 63 0 96Raised(>15U/l) (n=21) 12 9 1 13 (075to 171) 1 26

Mass concentration:Normal (<5pg/l) (n = 110) 42 68 0 59Raised (>5 ug/l) (n= 38) 34 4 2-34 (1-72 to 3-16) 8-00

Admission 4-12 h after onset of chest painElectrocardiographic findings:

Ischaemic (n = 45) 31 14 2-54 (1-76 to 3 66) 2 02Non-conclusive (n = 45) 31 14 2-54 (1-76 to 3 66) 2-02Normal (n = 59) 16 43 0 34

Creatine kinase*:Normal (n = 91) 32 59 0-49Raised (n= 58) 46 12 2-26 (1-67 to 3 06) 3-48

Creatine kinase MB:Activity:Normal (< 15 U/1) (n = 99) 37 67 0 54Raised (>15 U/1) (n = 50) 41 9 2-19 (1-63 to 2-95) 4-15

Mass concentration:Normal (<5 pg/l) (n = 75) 19 56 0-31Raised (>5jug/l) (n = 74) 59 15 3-15 (2-24 to 4-41) 3-58

*Upper limit of normal values 130 U/l for men and 90 U/I for women.

MB mass concentration was slightly raised inthe sample at admission (7 3 Rg/l) but notthereafter.

PATIENTS WITH OTHER, NON ISCHAEMIC,CARDIAC DISEASEFive patients had pericarditis, four arrhyth-mias, and two congestive heart failure. Onepatient with pericarditis had raised results forthe three markers (creatine kinase activity 419U/1; creatine kinase MB activity 37 U/1; crea-tine kinase MB mass concentration 42-1 pug/l),while the other four patients had all threemarkers within the normal range. Only onepatient with arrhythmia had a raised creatinekinase MB mass concentration (5-3 ug/l); onepatient with congestive heart failure had araised creatine kinase MB activity (21 U/1)and creatine kinase MB mass concentration(8-5 pug/l), while the other had all resultswithin the normal range. Electrocardiographicfindings on admission were ischaemic in onepatient with pericarditis (the patient withraised values for the three markers), in onepatient with arrhythmia, and in one patientwith congestive heart failure. Electrocardio-graphic findings were non-conclusive in onepatient with pericarditis and in three patientswith arrhythmia.

DIAGNOSTIC PROPERTIES OF MARKERS FORMYOCARDIAL DAMAGEBecause the release of markers after myocar-dial infarction depends on the time elapsedsince the onset of necrosis and becausethrombolysis is most effective when startedwithin four hours after the onset of chest pain,we related the diagnostic properties of themarkers to the time between the onset ofchest pain and admission (table 2). Table 3shows the frequency of normal and abnormalresults at admission on electrocardiographyand for the biochemical markers, togetherwith the relative risk and likelihood ratios.

COMBINATION OF ELECTROCARDIOGRAPHICFINDINGS WITH ONE BIOCHEMICAL MARKER INIDENTIFYING MYOCARDIAL INFARCTIONTo determine the effect of combining thediagnostic tests, the electrocardiographicresults, activities of creatine kinase and crea-tine kinase MB activities, and creatine kinaseMB mass concentration were entered into amodel for stepwise logistic regression analysis.In the group admitted within 4 hours afteronset of chest pain the independent predictors

115 on F

ebruary 26, 2020 by guest. Protected by copyright.

http://heart.bmj.com

/B

r Heart J: first published as 10.1136/hrt.72.2.112 on 1 A

ugust 1994. Dow

nloaded from

Bakker, Gorgels, van Vlies, Koelemay, Smits, Tijssen, Haagen

Table 4 Independent predictors of myocardial infarction at admission indentified bystepwise logistic regression analysis

95%Odds confidence

Variable Coefficient SE ratio* interval

Admission within 4 h after onset of chest painCreatine kinase MB mass concentration 2-60 0-61 13-4 4 0 to 44-4Electrocardiographic findings:

Ischaemic 2-38 0-51 10-8 3-9 to 29-9Non-conclusive 1-59 0 50 4-9 1-8 to 13-3

Constant -1-75 0-38

Admission 4-12 h after onset of chest painCreatine kinase:

Activity 1-66 0-57 5-2 1-7 to 16-2Mass concentration 1-55 0-46 4-7 1-9 to 11-8

Electrocardiographic findings:Ischaemic 1-78 0 55 5 9 20 to 17-5Non-conclusive 1-95 0-55 7-1 2-4 to 21-1

Constant -2-28 0-45

*Odds of disease given a positive test divided by the odds of disease given a negative test. Thismeans, for example, that odds for myocardial infarction in patients with raised creatine kinaseMB mass concentration is 13-4 times higher than that in patients with normal creatine kinaseMB mass concentration.

of myocardial infarction were electrocardio-graphic findings on admission and creatinekinase MB mass concentration (table 4).Creatine kinase MB mass concentrationstrongly enhanced discrimination betweenpatients with and without myocardial infarc-tion when electrocardiographic findings werenormal or non-conclusive. Table 5 shows theprobabilities of myocardial infarction givenadmission electrocardiographic findings aloneand in combination with one of the biochemi-cal markers (the respective probabilities werecalculated in a logistic regression model con-sidering the combination of electrocardio-graphic findings and each marker separately).

In the patients admitted between four and12 hours after onset of chest pain the activity ofcreatine kinase MB was an independent pre-dictor of myocardial infarction in addition to

the electrocardiographic findings and creatinekinase MB mass concentration.

DiscussionEstablishing the diagnostic value of testsrequires proper validation of normal ranges,which are usually estimated in serum samplesfrom healthy blood donors or laboratory staff;these subjects are young and healthy com-

pared with the patients the test is meant for.We therefore verified the upper limit of nor-

mal values for the cardiac markers in patientswho were admitted for evaluation of possiblemyocardial infarction, and classed as havingno acute ischaemic heart disease at dischargefrom hospital. Falsely raised results for crea-

tine kinase and creatine kinase MB activitieswere found as a result of known causes suchas muscle injury,5 the occurrence of macro-molecular creatine kinase,7-9 and the release ofadenylate kinase from haemolysed erythro-cytes.56 Raised results in patients with no evi-dence of acute ischaemic heart disease were

encountered to a far lesser extent for creatinekinase MB mass concentration. Although cre-

atine kinase MB mass concentration may riseabove the upper limit of normal values afterheavy physical exercise,23 no evidence ofmuscular damage was found in the firstpatient with a raised creatine kinase MB mass

concentration. In the other patient with a

raised creatine kinase MB mass concentra-tion, creatine kinase and creatine kinase MBactivities were raised as a result of thepresence of a macromolecular creatine kinase,so muscular damage could not be excludedwith certainty. For creatine kinase MB massconcentration the upper limit of normalvalues found in our study was supported notonly by the findings in the patients with noacute ischaemic heart disease but also by the

Table 5 Probabilities of myocardial infarction obtained by logistic regression analysis, given electrocardiographic resultson admission alone and in combination with three biochemical markers

Electrocardiographic results combined with:

Creatine kinase Creatine kinaseElectrocardiographic MB mass MB Creatine kinaseresults on admission Test result concentration activity activity

Admission within 4 h after after onset on chest painRaised 0 70 0-27 0 35

Normal (n=55) 0 22Normal 0-15 0-21 0-20

Raised 0-92 0 70 0 770-51 Non-conclusive (n=46) 0-63

-Normal 0-46 0-62 0-60

Raised 0-96 0 79 0-85Ischaemic (n=47) 0 74

- Normal 0-65 0 73 0-72

Admission 4-12 h after onset of chest painRaised 0 59 0-62 0-57

Normal (n=59) 0-27Normal 0-13 0-12 0-14

Raised 0-87 0 94 0-900-52 Non-conclusive (n=45) 0069_

Normal 0-42 0-56 0 53

Raised 0-86 0 93 0-89Ischaemic (n=45) 0-69

Normal 0-38 0 54 0-51

116 on F

ebruary 26, 2020 by guest. Protected by copyright.

http://heart.bmj.com

/B

r Heart J: first published as 10.1136/hrt.72.2.112 on 1 A

ugust 1994. Dow

nloaded from

Contribution of creatine kinaseMB mass concentration at admission to early diagnosis of acute myocardial infarction

results reported in other studies.'3 24 25 Theaccidents causing raised creatine kinase andcreatine kinase MB activities therefore do notseem to have any effect on creatine kinase MBmass concentration.

Verifying the upper limit of normal valuesby results from patients with a final diagnosisof unstable angina pectoris does not seemvalid. Whereas in the patients with unstableangina pectoris the 95% limits for creatinekinase and creatine kinase MB activity are notdifferent from the upper limit of normal val-ues, creatine kinase MB mass concentration israised in a fair proportion of these patients(17%). No myocardial damage was estab-lished in these patients according to theWorld Health Organisation's criteria for diag-nosing myocardial infarction (based on theinsensitive methods for measuring enzymeactivity), but this does not imply that minormyocardial damage was excluded. This sup-position is supported by two recent studies,which show that patients with unstable anginapectoris and a raised creatine kinase MB massconcentration had a significantly higher risk ofdeveloping acute myocardial infarction orrequiring revascularisation.26 As a conse-quence, minor myocardial damage, whichremained undetectable with the less sensitiveassays for creatine kinase and creatine kinaseMB activities used so far, is probably presentin patients with unstable angina pectoris and araised creatine kinase MB concentration. Toadapt the upper limit of normal values for cre-atine kinase MB mass concentration on thebasis of the results in patients with unstableangina pectoris is not therefore justified.

In patients presenting with acute chest painshortly (<4 hours) after onset accurate earlydiagnosis is an important requirement toassure proper allocation of expensive coronarycare beds, to identify patients with acutemyocardial infarction who may benefit frominterventions aimed at limiting the size of theinfarct,3 2831 and to avoid premature dischargefrom hospital of patients with myocardialinfarction. Selection of therapeutic measuresis guided by clinical history, age of the patient,coagulation state, and the diagnostic facilitiesavailable to the physician. Objective biochem-ical data, although desirable, are not currentlyavailable for the decision making processbecause the currently used markers, creatinekinase and creatine kinase MB activities, havea low diagnostic sensitivity at admission (crea-tine kinase 20%, creatine kinase MB 16%).Creatine kinase MB mass concentrationseems to have a significantly better diagnosticsensitivity (45%) shortly after onset of chestpain.121316 Moreover, this biochemical assaycan be performed rapidly.'7 The electrocar-diogram on admission is another objectivemeasure for early detection of myocardialinfarction. ST segment changes were found tobe the most sensitive criterion.232-34 SpecificST changes may, however, be absent in theearly phase of acute myocardial infarction in40-60% of patients (in our study, 57%).Although electrocardiographic recordings canbe performed easily and rapidly, the diagnostic

accuracy shortly after onset of chest pain islimited.To improve the diagnostic possibilities at

admission shortly after onset of chest pain, weinvestigated the diagnostic power of combin-ing the electrocardiographic findings onadmission with creatine kinase MB massconcentration on admission. With this combi-nation of tests, which both can be performedrapidly, the diagnostic accuracy increasesconsiderably in comparison with electro-cardiographic findings, alone or in combina-tion with creatine kinase or creatine kinaseMB activity. In patients admitted betweenfour and 12 hours after the onset of chest painthe results are less convincing for creatinekinase MB mass concentration as a conse-quence of the high false positive rate inpatients with unstable angina pectoris. Thehigher sensitivity of creatine kinase MB massconcentration for myocardial cell necrosis isaccompanied by a lower specificity becausepatients with minor myocardial damage and araised creatine kinase MB mass concentrationwill be called false positive according to theWHO criteria. Such a classification, however,does not account for the significance ofdetecting minor myocardial cell damage inpatients not yet fulfilling the WHO criteria fordiagnosis of myocardial infarction. Thesepatients carry an increased risk of dying orhaving a major myocardial infarction,2627 andthey may be detected by measuring creatinekinase MB mass concentration. Thereby thesepatients constitute a subgroup, which mightbenefit from a more aggressive treatment-that is, coronary angiography followed by sub-sequent revascularisation.

Usually, the presence of specific STchanges has such a high positive predictivevalue for the diagnosis of myocardial infarc-tion2 34 that little further information is neededwhen this electrocardiographic finding is pre-sent, especially in conjunction with appropri-ate symptoms. Unfortunately, many patientswith myocardial infarction will have otherfindings on an admission electrocardiogram.Many patients presenting with suggestivesymptoms and without characteristic findingson an admission electrocardiogram will haveunstable angina or non-cardiac diagnoses,while others will have myocardial infarction.Our results show that creatine kinase MBmass concentration constitutes a majorimprovement in the diagnostic process.Further improvement probably can beobtained by using creatine kinase MB sub-forms,35 but such measurements are not yetpracticable because labour and time consum-ing electrophoretic methods are needed. Inpatients with non-conclusive electrocardio-graphic findings-that is, patients with a leftbundle branch block-and clinical suspicionof myocardial infarction creatine kinase MBmass concentration might be helpful in decid-ing whether thrombolysis is needed.

In conclusion, we have four main findings.(1) Creatine kinase MB mass concentration

is more sensitive in detecting patients withacute myocardial infarction than creatine

117 on F

ebruary 26, 2020 by guest. Protected by copyright.

http://heart.bmj.com

/B

r Heart J: first published as 10.1136/hrt.72.2.112 on 1 A

ugust 1994. Dow

nloaded from

Bakker, Gorgels, van Vies, Koelemay, Smits, Tijssen, Haagen

kinase and creatine kinase MB activities.(2) Creatine kinase MB mass concentration

in combination with electrocardiographicfindings on admission best distinguishespatients with myocardial infarction from thosewithout when admission is within 4 hoursafter the onset of chest pain.

(3) In patients with a final diagnosis of noacute ischaemic heart disease considerablyless false positive results were encounteredusing creatine kinase MB mass concentrationinstead of creatine kinase and creatine kinaseMB activities.

(4) In patients with unstable angina pec-

tons creatine kinase MB mass concentrationwas raised in a considerable number ofpatients in contrast with creatine kinase andcreatine kinase MB activities, thereby allow-ing detection of a subgroup of patients withpossibly an increased risk for major cardiacevents.We conclude that for early diagnosis of

myocardial infarction rapid assays for thedetermination of creatine kinase MB mass

concentration should replace the measure-

ments of creatine kinase and creatine kinaseMB activities.

The reagents for the immunological determination of creatinekinase MB mass concentration were given by Abbott.

1 Lee TH, Rouan GW, Weisberg MC, et al. Sensitivity ofroutne clinical criteria for diagnosing myocardial infarc-tion within 24 hours of hospitalization. Ann Intern Med1987;106: 181-6.

2 Rude RE, Poole WK, Muller JE, et al. Electrocardio-graphic and clinical criteria for recognition of acutemyocardial infarction based on analysis of 3697 patients.Am Y Cardiol 1983;52:936-42.

3 Bar FW, Vermeer F, De Zwaan C, et al. Value of admis-sion electrocardiogram in predicting outcome of throm-bolytic therapy. Am

_Cardiol 1987;59:6-13.

4 Lee TH, Weisberg MC, Cook EF, Daley K, Brand DA,Goldman L. Evaluation of creatine kinase and creatinekinase MB for diagnosing myocardial infarction. ArchIntn Med 1987;147:115-21.

5 Moss DW, Henderson AR, Kachmar JF. Enzymes. In:Tietz NW, ed. Textbook ofdinical chemisty. Philadelphia:Saunders, 1986:619-774.

6 Horder M, Elser RC, Gerhardt W, Mathieu M, SampsonEJ. Approved recommendation on IFCC methods forthe measurement of catalytic concentrations of enzymes.Part 7. IFCC method for creatine kinase. EuropeanJournal of Clinical Chemisty and Clinical Biochemisty1991;29:435-56.

7 Griffiths PD. CK-MB a valuable test? Ann Clin Biochem1986;23:238-42.

8 Stein W, Bohner J, Bahlinger M. Analytical patterns andbiochemical properties of macro creatine linase type 2.Clin Chem 1985;31:1952-8.

9 Rogalsky VY, Koven IH, Miller DR, Pollard A. Macrocreatine kinase type 2 in human colonic tissues. ClinBiochem 1985;18:338-41.

10 Lee TH, Goldman L. Serum enzyme assays in the diagno-sis of acute myocardial infarction: recommendationsbased on a quantitative analysis. Ann Intern Med 1986;105:221-33.

11 Delanghe JR, de Mol AM, de Buyzere ML, de ScheerderIK, Wieme RJ. Mass concentration and activity concen-tration of creatine kinase isoenzyme MB compared inserum after acute myocardial infarction. Clin Chem1990;36: 149-53.

12 Coilinson PO, Rosalki SB, Kuwana T, et al. Early diagnosisof acute myocardial infarction by CK-MB mass mea-surements. Ann COin Biochem 1992;29:43-7.

13 Mair J, Artner-Dworzak E, Diensti A, et al. Early detectionof acute myocardial infarction by measurement of mass

concentration of creatine kinase MB. Am J Cardiol1991;68:1545-50.

14 Wolfson D, Lindberg E, Su L, Farber SJ, Dubin SB.Three rapid immunoassays for the detennination ofcreatine kinase MB: an analytical, clinical and interpre-tve evaluation. Am HeartJ 1991;122:958-64.

15 GiblerWB, Lewis LM, Makens PK, et al. Early detection ofacute myocardial infarction in patients presenting withchest pain and nondiagnostic ECGs: serial CK-MBsampling in the emergency departnent. Ann Emerg Med1990;19: 1359-66.

16 Bakker AJ, Gorgels JPMC, van Vlies B, Haagen FDM,Smits R. The mass concentrations of serum troponin Tand creatine kinase MB are elevated before creatinekinase and creatine kinase MB activities in acutemyocardial infarction. European Journal of ClinicalChemtsy and Chnical Biochemisty 1993;31:715-24.

17 Apple FS, Preese LM, Riley L, Gerken KL, van Lente F.Financial impact of a rapid CK-MB-specific immunoas-say on the diagnosis of myocardial infarction. ArchPathol Lab Med 1990;114:1017-20.

18 Criteria Committee of the New York Heart Association.Nomenclature and criteria for diagnosis of diseases of theheart and great vessels. 8th ed. Boston: Little, Brown,1979.

19 Stein W. Laboratory diagnosis of acute myocardial infarc-tion using CK and CK-MB by inmmunoinhibition.Darmstadt: GIT Verlag, 1988:20-7.

20 Brandt DR, Gates RC, Eng KK, et al. Quantifying the MBisoenzyme of creatine kinase with the Abbott "IMx"immunoassay analyzer. Clin Chem 1990;36:375-8.

21 Solberg HE. The theory of reference values, part 5.Statistical treatment of collected reference values.Determination of reference limits. J Clin Chem ClinBiochem 1983;21:749-60.

22 SAS Institute. SAS technical report P-200, SASISTAT soft-ware: Calis and Logistc procedures. Release 6-04. Cary,NC: SAS Institute, 1990:175-230.

23 Ord6fiez-Llanos J, Serra-Grima JR, Merce-Muntafiola J,Gonzklez-Sastre P. Ratio of creatine kinase 2 mass con-centration to total creatine kinase activity not altered byheavy physical exercise. Clin Chem 1992;38:2224-7.

24 Wu AHB, Gornet TG, Harker CC, Chen H-L. Role ofrapid immunoassays for urgent ("stat") determinationsof creatine kinase isoenzyme MB. Clin Chem 1989;35:1752-6.

25 Chapelle JP, Allaf ME. Automated quantification ofcreatine kinase MB isoenzyme in serum by radial parti-tion immunoassay with use of the Stratus analyzer. ClinChem 1990;36:99-101.

26 Ravkilde J, Bo Hansen A, Horder M, Jorgensen PJ,Thygesen K. Risk stratification in suspected acutemyocardial infarction based on a sensitive immunoassayfor serum creatine kinase isoenzyme MB. Cardiology1992;80:143-51.

27 Markenvard J, Dellborg M, Jagenburg R, Swedberg K.The predictive value of CKMB mass concentration inunstable angina pectoris: preliminary report. J InternMed 1992;231:433-6.

28 Lee Th, Rouan GW, Weisberg MC, et al. Clinical charac-teristics and natural history of patients with acutemyocardial infarction sent home from the emergencyroom. Am J Cardiol 1987;60:219-24.

29 Goldman L, Cook EF, Brand DA, et al. A computer proto-col to predict myocardial infarction in emergencydepartment patients with chest pain. NEnglJMed 1988;318:797-803.

30 Lee TH, Juarez G, Cook EF, et al. Ruling out acutemyocardial infarction. A prospective multicentre valida-tion of a 12-hour strategy for patients at low risk. NEnglJ Med 1991;324:1239-46.

31 Pozen MW, D'Agostino RB, Selker HP, Sytkowski PA,Hood WB. A predictive instrument to improve coronarycare unit admission practices in acute ischemic heartdisease. NEnglMed 1984;310:1273-8.

32 Ohman EM, Casey C, Bengtson JR, Pryor D, Tormey W,Horgan JH. Early detection of acute myocardial infarc-tion: additional diagnostic information from serumconcentrations of myoglobin in patients without ST-elevation. BrHearJ 1990;63:335-8.

33 Gibler WB, Young (P, Hedges JR, et al. Acute myocardialinfarction in hesit pain patents with nondiagnosticECGs: serial C1-MB sampling in the emergencydepartment. Ann EiiXrg Med 1992;21:504-12

34 Yusuf S, Pearson M, Srry H, et al. The entry ECG in theearly diagnosis and pibgnostic stratification of patientswith suspected acute myocardial infarction. Eur Heart Y1984;5:690-6.

35 Puleo PR, Guadagno PA, Roberts R, et al. Early diagnosisof acute myocardial infarction based on an assay forsubforms of MB creatine kinase. Circulation 1990;82:759-64.

118 on F

ebruary 26, 2020 by guest. Protected by copyright.

http://heart.bmj.com

/B

r Heart J: first published as 10.1136/hrt.72.2.112 on 1 A

ugust 1994. Dow

nloaded from