Automated Creatine Kinase-MB Estimationby lmmuno-lnhibition: A ...

2 downloads 0 Views 1MB Size Report
of Pathology, Henry Ford Hospital, Detroit, MI. 48202. I Nonstandard abbreviations: CK, creatine kinase; MI, myocardial infarction;. LD, lactate dehydrogenase.
CLIN. CHEM.26/5, 568-572 (1980)

Automated Creatine Kinase-MB Estimationby lmmuno-lnhibition: A Clinical

Evaluation Thomas J. Delahunty and Craig C. Foreback We detected the presence of creatine kinase (EC 2.7.3.2) MB isoenzyme (CK-MB) in serum by using an immunoinhibition technique to inactivate the creatine kinase isoenzyme of muscle origin (CK-MM). The GEMSAEC centrifugal analyzer is ideally suited for this procedure because it rapidly mixes reagents and has a throughput of 60 samples/h. The within-run CV was acceptable when the CK-MB value was 7.5 U/L or more. CK-MM was totally inactivated up to 5000 U/L, a value well above that usually seen in myocardial-infarct patients. We assessed the predictability of the assay for detecting myocardial infarction among 120 coronary-care patients when 2% of total CK activity was considered diagnostic and compared the results with those obtained with an electrophoresis technique. The positive predictive value of the immunoinhibition assay was 97.8%, as compared to 92% with the CK electrophoresis technique, when lactate dehydrogenase (EC 1.1.1.27) isoenzyme results were included in both cases. The optimal negative predictive value was 95.1% with the immunoinhibition assay vs 98.2% with electrophoresis. We conclude that the immunoinhibition technique for estimating CK-MB can be automated to assess myocardial status rapidly, precisely, inexpensively, and sensitively. AddItIonal Keyphrases:

enzyme activity isoenzymes electrophoresis on agarose gel centrifugal analyzer myocardial infarction “kit” methods heart disease monitoring therapy effects of lipids, hemoglobin, myokinase, bilirubin .

.

.

The activity of the cardiac isoenzyme (CK-MB) of creatine kinase (EC 2.7.3.2) in serum is now widely measured (1-4) to detect and monitor myocardial infarction (MI).1 The isoenzyme CK-MB, which is highly specific for the myocardium, is usually detected electrophoretically (1). However, electrophoresis is time consuming and quantitation laborious (5). An immunochemical procedure has recently become available in which CK derived from muscle (CK-MM) is inactivated with anti-CK-MM rabbit serum (6, 7). The predictive value of this technique in detecting MI has also been reported (8). We have adapted this antibody method as commercially available so that multiple serum samples can be rapidly assessed for CK-MB activity with the GEMSAEC centrifugal analyzer. Its reliability for the detection of MI was compared with the electrophoretic techniques currently in use in our laboratory. A preliminary report of this study has been published (9). Department 48202. I

Nonstandard

infarction;

Received 568

of Pathology, abbreviations:

Henry

Ford

CK, creatine

dehydrogenase. Aug. 20, 1979; accepted Jan.

Hospital, kinase;

LD, lactate

8, 1980.

CLINICALCHEMISTRY, Vol. 26, No. 5, 1980

Detroit,

MI

MI, myocardial

Materials and Methods Patients One hundred and twenty patients from the Coronary Care Unit who were suspected of having had an MI were sequentially chosen for this study. All of these patients had at least three

serial

blood

onset of symptoms. or non-MI

group

samples

drawn

A patient on the

basis

by venipuncture

was assigned of the final

after

into the acute discharge

the

MI

summary

statement issuedindependently of thisstudy by the clinician responsible for the patient. Both transmural and sub-endocardial infarcts were included in the MI group.

Procedures The

blood

samples

were

analyzed

for

isoenzymes

by

methods identical to those routinely used in our laboratory, as follows. After centrifugation to sediment erythrocytes, the blood samples for CK analysis were stored at 4 #{176}C while those for LD (lactate dehydrogenase, EC 1.1.1.27) assay were kept at room temperature. The enzyme analysis was always performed within 24 h after phlebotomy. Total serum CK was routinely measured by the method of Rosalki (10) with commercially obtained reagents (Calbiochem-Behring Corp., LaJolla, CA 92037). Serum LD was routinely determined as described by Wacker et al. (11), with the Statzyme#{174} reagents (Worthington Diagnostics, Freehold, NJ 94306). CK isoenzymes were separated by electrophoresis on “Special” agarose gel plates (Corning, Palo Alto, CA 94306). When present, CK-MB was detected by viewing the plates under ultraviolet light after incubation with the Corning substrates at 37 #{176}C for 20 mm. The presence of intense or trace fluorescent CK-MB bands was recorded. LD isoenzymes were separated on Corning “Universal” agarose gels with use of the Pol-EStrat#{174} reagents (Pfizer Inc., New York, NY 10017). After color development with p-nitroblue tetrazolium, the plates were scanned at 575 nm with an ACD-15 scanning densitometer (Gelman Instrument Co., Ann Arbor, MI 48106). The relative amounts of each LD isoenzyme were then estimated. A “flipped” pattern was assigned to a patient’s serum if the LD1 isoenzyme was greater than or equal to the LD2 isoenzyme content (12). The CK-MB was estimated by the Cardiozyme#{174}immunoinhibition technique (Dade Division, American Hospital Supply Corp., Miami, FL 33152). The Cardiozyme kit contains a specific antibody raised in goats against the M subunit of CK combined with the usual reagents for determining CK by the Rosalki technique (10). Each vial containing lyophilized reagents was reconstituted with 1.5 mL of Cardiozyme buffer. Aliquots of this reagent (0.33 mL) were pipetted into eaoh C compartment of a GEMSAEC rotor (Electro-Nucleonics, Inc., Fairfield, NJ 07006). The serum samples (40 zL) were pipetted into each B compartment with 100 L of buffer. In this way a maximum of 15 serum specimens could be simultaneously analyzed.

The serum was rapidly mixed with reagents by centrifugation in the GEMSAEC and a 7-mm pre-incubation time was allowed before readings were taken. A total of seven absorbance readings were then taken at 30-s intervals. Results were not doubled to compute CK-MB because serum aliquots were twice the manufacturer’s recommended volume. Use of the GEMSAEC enabled us to discriminate between a positive reaction rate ascribable to the presence of CK-MB and a negative reaction rate resulting from the oxidoreductases that occasionally are present in serum. In the latter case, a CK-MB of zero was assigned to the serum. When only total CK was assayed, 0.5 mL of Calbiochem-Behring reconstituted reagent (1 vial with 42 mL of water) was placed in the C compartment of a rotor and 20 fiL of serum was aspirated into the B compartment with 80 tL of water. The pre-incubation period was 2 mm and 1.5 mm for the total CK and LD determinations, respectively. All GEMSAEC determinations were performed at 30#{176}C, but we used factors of 1.4 and 1.7 to convert CK and LD results to the expected values at 37 #{176}C, respectively. Myokinase (EC 2.7.4.3) activity was estimated by using the Ultrazyme Plus#{174} CK reagents (Harleco Division, American Hospital Supply Corp., Gibbstown, NJ 08027), which did not contain the phosphocreatine substrate. We performed the assays in the GEMSAEC, using the usual instrumental arrangements for CK. Rabbit myokinase (1000 kU/g of protein) was used to standardize the assay (Sigma Chemical Co., St. Louis, MO 63178). Plasma hemoglobin was measured by the method of Crosby and Furth (13). The Triton X-100 surfactant used to hemolyze blood was a product of Technicon Instruments Corp., Tarrytown, NY 10591. Manual absorbance readings were performed with a Perkin-Elmer (Coleman 124) recording spectrophotometer supplied by Hitachi, Tokyo, Japan. The sensitivity, specificity, and positive predictive values were calculated according to the formulas described by Galen and Gambino (14). We separated CK-MB from CK-MM by ion-exchange chromatography with use of the Harleco Ultrazyme Plus kit. Lyophilized controls I and II for the day-to-day imprecision study were supplied by Dade. “Level III” was the CK/LD isoenzyme control supplied by Helena Laboratories, Beaumont, TX 77704 (reconstituted daily with 10 mL of 9 g/L NaC1).

Results Table 1 shows the within-run coefficients of variation (CV) obtained when sera from six myocardial infarct patients were analyzed for total CK and CK-MB by the immunoinhibition technique. The patients had CK values ranging from 2503 to 423 U/L, with a CV never exceeding 3%. When we performed the CK-MB assay on the same serum samples, we obtained CK-MB values ranging from 198 U/L to 7.5 U/L. It can be seen that above 7.5 U/L the CV of the CK-MB assay was at an acceptable level (0.05) was obtained (Dade reports a value of 0.955 for 63 samples). When CK-MB was assessed manually as described in the Cardiozyme insert sheet, the Dade control serum II gave a CK-MB of 38.8, SD 1.8, UIL (n = 20) at 37 #{176}C, a value that was 86% of that obtained with the centrifugal analyzer (45.3 U/L). Figure 1 depicts an experiment in which we tested the ability of the Cardiozyme to inhibit CK of muscle origin (CK-MM), by assaying solutions with an increasing concen-

Table 1. CoeffIcients of Variation for Serum Total CK and CK-MB from SIx Selected Patients Total CIC a U/L

2503 ± 1518 ± 944 ± 475 ± 354 ± 423 ± a

CK-MB U/L

CV, %

60 46 26 5.9 8.3 5.9

2.4 3.0

-

2.7 1.2 2.3 1.4

a CV,%

198 ± 128 ± 57.3 ± 26.1 ± 14.8 ± 7.5 ±

2.9 8 1.8 1.1 1.1 0.7

%MB

1.45 2.3 3.1 4.35 7.3 9.0

7,9 8.4 6.1 5.5 4.2 4.8

CK, UIL, are expressed as the mean ±SD (n = 12).

Table 2. Day-to-Day Imprecision of Cardiozyme CK-MB Assay CK-MB, u/La

14.4 ± 1.9

45.3 ± 3.2

13.6

7.1

CV, % a

Mean CK-MB actMty (and SD) dtilng 20 consecutive

103 ± 7.5 7.4

days.

tration of CK. We prepared the samples using serum containing a total CK activity of 16 kU/L. As indicated in the diagram, the CK-MB activity responded in a linear fashion to approximately 100 U/L (total CK = 5000 U/L), beyond which uninhibited CK-MM was apparently influencing the apparent content of CK-MB. We assessed 120 patients in the Coronary Care Unit over a four-month period, by the CK-MB assay procedure. Of these patients, 54 were judged by the clinicians to have had a myocardial infarction near the time the blood samples were drawn. The mean peak serum CK activity for the MI group was 1083 (SD 757) U/L, a value significantly different from the CK of the non-MI group (470, SD 956, U/L; p