Evaluation of a commercial immunoenzymometric assay kit for creatine kinase MB isoenzyme determination using monoclonal antibodies

A simultaneous two-site immunoenzymometric assay for creatine kinase MB determination (Hybritech Tandem-E CK-MB) using monoclonal antibodies was evaluated and compared with cellulose acetate electrophoresis using fluorometric scanning densitometry. The assay has satisfactory precision (between-day analysis gives a coefficient of variation between 2.1 and 9.4%) and is not susceptible to interference by concentrations of creatine kinase MM up to 5000 micrograms/l (3400 U/l) and creatine kinase BB up to 1000 micrograms/l (1085 U/l). The upper limit of MB isoenzyme concentration in 250 apparently healthy people was 5.5 micrograms/l. Comparison between the immunoenzymometric assay (y) and electrophoresis (x) yielded the following linear regression equation: y = 0.37x + 1.9, with a correlation coefficient of 0.828. The characteristics of the temporal kinetics of MB isoenzyme, calculated by two methods, in 49 patients with acute myocardial infarction, were nearly identical in terms of the rate of creatine kinase MB release and the time at which the peak value is obtained, but not in terms of the rate of elimination of the isoenzyme. The fractional disappearance rate of MB isoenzyme from the circulation was significantly higher if calculated with Tandem-E results rather than with electrophoresis results (-0.035 vs -0.028, p less than 0.001). Whereas in the first day after infarction immunoenzymometric assay and electrophoresis had the same clinical sensitivity for identifying patients with acute myocardial infarction, in specimens collected more than 24 hours after the onset of the chest pain, the clinical sensitivity of the immunoenzymometric method was lower. Our results show that it is still premature to draw definitive clinical conclusions from the immunoassay results.     

A "column-batch" method for separating MB and LD1 in a single fraction

In this "column-batch" method for separating the MB and BB isoenzymes of creatine kinase and the LD1 isoenzyme of lactate dehydrogenase, one can, alternatively, separate MB from BB or obtain a combined fraction containing MB, BB, and LD1. The principal advantage is that the resulting fractions are twofold as concentrated as was the applied sample. Thus, activity can be measured by conventional automated methods, with no need for the modifications to compensate for diluted fractions that are required by other ion-exchange methods. Another advantage is the total absence of interference by the MM isoenzyme. A strong anion exchanger (AG-MP1, Bio-Rad) is used in the acetate form at pH 6.3. There is no retention of MM; retained MB, BB, and LD1 are eluted with a solution of magnesium acetate. Results are compared with those obtained for subunit B and LD1 by immunoinhibition. Results with patients are considered consistent with myocardial infarction if MB exceeds 20 U/L and 3% of the total CK and LD1 exceeds 130 U/L or 28% of the total LD activity.     

A sensitive bioluminescent immunoinhibition test for CK-B subunit activity and a CK-MB specific ELISA compared: correlation with agarose electrophoresis and influence of CK-isoenzyme profile on results

Searching for alternatives to the imprecise spectrophotometric tests for low-concentration creatine kinase (EC 2.7.3.2) isoenzyme MB (CK-MB), we investigated the analytical performance of two potentially superior approaches--a bioluminescent immunoinhibition assay (I, LKB-Wallac) and an ELISA (enzyme-labeled immunosorbent assay) technique (II, Hybritech)--in comparison with an electrophoretic method (III, Beckman). Only I showed good between-day precision (CV 8.3%) at the upper reference limit, allowing reproducible assay of CK-B subunit activity down to at least 3 U/L. In conditions where CK isoenzyme assays remained unaffected by CK-MM concentrations, test results were proportional to the amount of CK-MB in the sample up to at least 50 U/L for I, 120 micrograms/L for II, and 100 U/L for III (r greater than 0.998 by linear regression analysis). For CK-MB-positive samples, the data by I correlated more closely with values by III (n = 24; r = 0.994) than did results by II (n = 15; r = 0.909), but both methods were equally effective in discriminating between samples with or without electrophoretically supranormal CK-MB activity (93% sensitivity). II was entirely CK-MB specific, whereas CK-B activity by I was consistently (18/18) increased in CK-MB-negative samples containing CK-BB (n = 6; r = 0.996) or macro CK, types 1 or 2 (n = 12; r = 0.930). I is highly sensitive for screening for increased non-MM CK activity, the nature of which should be subsequently clarified by electrophoresis.     

Detection of cardiac-specific creatine kinase isoenzyme in sera with normal or slightly increased total creatine kinase activity.

We describe a spectrophotometric kinetic assay for detecting creatine kinase MB isoenzyme activity in the 1 to 10 U/liter range. The MB isoenzyme was isolated [Clin. Chem. 20, 36 (1974)] and assayed (Rosalki method) with an Abbott ABA-100. Good reproducibility was demonstrated for MB isoenzyme activities near 1 U/liter (CV = 2.6%). Sera with normal or slightly increased total creatine kinase activity were evaluated. Sera of 14 patients with acute myocardial infarction contained, per liter, 84 to 236 U of total creatine kinase activity and 4.6 to 28.0 U of isoenzyme MB activity; corresponding ranges for sera from healthy lab technicians and patients with noncardiac disease were 36 to 277 and 0 to 2.6 U. MB isoenzyme activity for infarction patients rose and fell sharply within three days after the infarction. Atypical time-course patterns, MB isoenzyme activity remaining abnormally great for five days, were observed in serum from patients with prolonged atrial fibrillation and congestive heart failure or cardiomyopathy; the BB isoenzyme (1 to 5 U/liter) was also detected in sera of such patients but was absent in sera from infarcation patients. Quantification of column-isolated MB by the assay described is rapid, easy, specific, and extremely sensitive for measuring MB in the 1 to 10 U/liter range.     

Creatine kinase isoenzyme MB determination on the ACA: dependence on serum matrix and other effectors

Creatine kinase isoenzyme MB catalytic activities in human serum, determined by ACA ion exchange chromatography and immunoinhibition, differ significantly, the correlation coefficient being 0.88. The reasons for this variation are interference of antibodies with the creatine kinase B subunit in the immunoinhibition assay, nonreproducible elution of creatine kinase isoenzyme MB from the ion exchange resin in the ACA pack, due to varying protein concentrations in the serum samples and increasing elution of creatine kinase isoenzyme MM from the ion exchange column caused by a preceding partial inactivation of creatine kinase isoenzyme MM. Pretreatment of serum samples with a solution containing magnesium sulphate, maleate and 2-oxoglutarate (solution A) prior to determination of creatine kinase isoenzyme MB catalytic activities on the ACA significantly improves the sensitivity and specificity of the method; the correlation coefficient for the values from the ACA and immunoinhibition then becomes 0.92. Dilution of serum samples with bovine serum albumin solution is now practicable.     

Modification of fully automated total iron-binding capacity (TIBC) assay in serum and comparison with dimension TIBC method

BACKGROUND: We previously reported the development of a fully automated assay for total iron-binding capacity (TIBC) in serum, using a multipurpose automated analyzer. However, this method requires four different reagents and is thus useful only with a limited number of available analyzers. We simplified our original assay and compared the analytical performance of the modified method with that of a commercial, fully automated TIBC assay (Dimension TIBC assay). METHODS: We simplified our original method to require only three reagents. Calibration was also altered and was performed with human transferrin standard solutions. An advantage of this method is that it does not require separation of excess unbound iron after the first step of transferrin saturation. Unbound iron is eliminated by formation of a complex with the chromogenic reagent ferrozine in the second step. Iron dissociated from transferrin by acidic pH reacts with ferrozine to form a colored complex in the final step, and the increase in absorbance at 570/660 nm is directly proportional to the TIBC measured. TIBC values were determined for 49 healthy individuals and 148 patients with this modified TIBC assay and with a commercial, fully automated TIBC method (Dimension clinical chemistry system), and calculation of TIBC based on the sum of the serum iron and unsaturated iron-binding capacity was performed for 97 patients. RESULTS: The within-run CVs for the modified TIBC assay and the Dimension TIBC assay were <4.8% and <2.4%, and the between-run CVs were 1.2% and 1.7%, respectively. The dilution curves were linear for TIBC values up to at least 180 micromol/L with both methods. TIBC values obtained by our method were linearly correlated with serum transferrin concentrations (r = 0.984; S(y/x) = 3.18 micromol/L; P <0.001). The correlation between the values obtained with the present method (y) and those obtained with the Dimension TIBC method (x) was y = 1.04x + 1.19 micromol/L (r = 0.985; S(y/x) = 2.47 micromol/L), and with the calculation method (x) was y = 1.18x + 2.62 micromol/L (r = 0.976; S(y/x) = 3.27 micromol/L). CONCLUSIONS: Our modified, fully automated TIBC assay performed similarly to the Dimension TIBC assay and is adaptable for use with many multipurpose automated analyzers.     

Effects of frusemide and captopril on the relationship between biologically and immunologically active renin in human plasma

1. The plasma renin concentration (PRC) and active renin concentration determined by radioimmunometric assay (ARC) were measured before and after administration of frusemide or captopril to normal volunteers. 2. Injection of 40 mg of frusemide followed by 1 h in the upright position significantly increased both PRC (P less than 0.001) and ARC (P less than 0.001). Oral administration of 50 mg of captopril also increased PRC (P less than 0.05) and ARC (P less than 0.05). 3. ARC and PRC were linearly correlated in the basal supine position [y = 0.635x - 0.048 (y = PRC, x = ARC), r = 0.958, P less than 0.001], after frusemide injection and standing (y = 0.800x + 0.359, r = 0.793, P less than 0.02) and after captopril administration in the supine position (y = 0.938x - 0.555, r = 0.998, P less than 0.001). 4. A cross-calibration study in which pure human renin was added to pooled plasma and PRC and ARC were measured showed linearity between the values obtained by the two methods. 5. The regression line for values of PRC and ARC in the supine position after captopril administration had a significantly greater slope (P less than 0.001) and that for values after frusemide injection and standing was significantly elevated (P less than 0.001) compared with the regression line for basal values in the supine position. 6. These results show that the biological activity of renin may be increased by various acute stimulations of renin to inappropriately high levels compared with the immunological activity.(ABSTRACT TRUNCATED AT 250 WORDS)     

Clinical evaluation of an automated chemical inhibition assay for lactate dehydrogenase isoenzyme 1

We evaluated an automated assay for lactate dehydrogenase (LD; EC 1.1.1.27) isoenzymes, supplied by Boehringer Mannheim Diagnostics (BMD) and based on selective chemical inhibition of non-LD-1 isoenzymes by guanidine thiocyanate. Results were compared with the Roche Isomune LD-1 method. The Hitachi 717 analyzer was used to measure enzyme activity for both procedures in 229 serum samples. One hundred specimens were also analyzed by the Helena rapid electrophoresis (REP) method. We determined the limit of linearity of the BMD method to be about 1200 U of LD-1 per liter. The analytical correlation of BMD (y) with Isomune (x) yielded y = 1.0x + 0.5 U/L, r = 0.997, Sy/x = 16.9 (range 20-1397 U/L). The regression equation for BMD vs REP was y = 1.1x + 7.2% (r = 0.800, Sy/x = 7.4, range 14-83%). Average values for within-run precision for low (38 U/L), medium (180 U/L), and high (865 U/L) controls were 4.1%, 1.0%, and 0.5%, respectively (16 trials of six each). The average values for run-to-run precision were 4.1%, 1.7%, and 1.1%, respectively, for these controls (n = 16). We used receiver-operating characteristic curves to determine optimum decision limits. Using an LD-1 cutoff of 40% of total LD, we obtained a clinical sensitivity of 97-100% and a specificity of 95% when blood was collected during the optimum interval, 24-48 h after the onset of chest pain. We conclude that the BMD LD-1 assay is equivalent to the immunochemical and electrophoretic assays for measuring the LD-1 isoenzyme.     

Stratus automated creatine kinase-MB assay evaluated: identification and elimination of falsely increased results associated with a high-molecular-mass form of alkaline phosphatase

We compared the performance of an automated assay of creatine kinase MB isoenzyme (CK-MB) mass (Stratus) with that of a CK-MB enzymatic assay routinely used at our institutions. Both of these assays use the same CK-MB-specific monoclonal antibody to immunocapture CK-MB, thus providing a direct means of comparing a mass assay with an activity assay. Routine CK-MB measurements for 206 samples within the analytical range of both assays revealed the following relationship: Stratus (micrograms/L) = 0.67 (activity U/L) + 0.18 (r = 0.95, Sx.y = 4.45). The linearity, sensitivity, and precision of the Stratus assay were acceptable for routine clinical use. Icteric, lipemic, and hemolyzed samples do not interfere with the assay. During our evaluation we identified a single, clinically significant false-positive sample. Because this patient had alkaline phosphatase values greater than 1100 U/L, we investigated additional samples with increased activities of alkaline phosphatase and found that samples from 12 of 23 patients selected for alkaline phosphatase values greater than 460 U/L produced falsely increased CK-MB values. We determined that a membrane-associated, high-molecular-mass form of alkaline phosphatase was a cause of these falsely increased values and instituted an approach to identify falsely increased Stratus CK-MB values. Samples from 23 of 1933 patients were falsely increased, the increase being clinically significant in samples from 14 of these patients. Consultation with the manufacturer resulted in the successful reformulation of the substrate/wash solution to minimize interferences from high-molecular-mass forms of alkaline phosphatase.     

Assay of pancreatic amylase with use of monoclonal antibodies evaluated

To evaluate a new method for measuring pancreatic amylase in serum, in which the salivary isoenzyme is inhibited with a specific monoclonal antibody, we determined the activity of pancreatic and salivary amylase in sera from 103 healthy subjects and from 114 hospitalized patients having a wide range of total amylase activities. CVs for the proposed method ranged from 0.8% to 5.1% (within day) and from 2.3% to 6.6% (day to day). Results correlated well with those obtained by the wheat-germ inhibition method (r = 0.998) and by electrophoresis on cellulose acetate. Analytical-recovery studies confirmed the good specificity of the monoclonal antibody for salivary amylase (97%) and its low cross-reactivity (0.6%) toward pancreatic amylase. The assay procedure presents a wide range of linearity (141-1817 U/L) and can easily be adapted to an automated kinetic system. We found the proposed method suitable for routine determinations of pancreatic amylase.     

Concordance of creatine kinase-MB activity and mass

The recent availability of monoclonal antibodies that are highly specific for creatine kinase (CK; EC 2.7.3.2) MB isoenzyme should allow for the development of rapid, sensitive, and specific assays of CK-MB mass and activity. However, the relationship between the mass concentration of CK-MB and its activity in plasma has previously been thought by some to be variable. To determine the extent to which discrepancies of potential clinical significance might arise between measurements of activity and mass in plasma, we compared CK-MB activity and concentration in 1298 samples obtained from 226 patients admitted to the cardiac-care unit. CK-MB activity concentration was determined with an immunoadsorption assay, and mass concentration was measured by an automated "sandwich" assay (Magic Lite; Ciba Corning Diagnostics). Both of these assays are based on specific monoclonal antibodies for CK-MB. Values obtained with these assays correlated well (r = 0.94). Normal and abnormal values with the two assays were concordant in 96% of the samples. In all but three instances, differences occurred late after myocardial infarction and were characterized by minimal increases as determined by one method vs values at the upper limit of normal as determined with the other. Thus, measurements of CK-MB mass and activity concentrations in plasma with assays based on these specific monoclonal antibodies are comparable for the detection or exclusion of acute myocardial infarction.     

Electroimmuno- and immunonephelometric assays of apolipoprotein A-I by using a mixture of monoclonal antibodies

The use of mixtures of well-defined monoclonal antibodies may represent a step forward in the standardization of immunochemical assays. We developed and optimized working conditions for using such a mixture to determine apolipoprotein A-I in human sera by two independent techniques (electroimmuno- and immunonephelometric-assays). Six monoclonal antibodies, each addressed to distinct epitopes located at the surface of apolipoprotein A-I, were used in combination to permit a reproducible measurement of the protein, without prior delipidation of samples. Parallel standard curves for a high-density lipoprotein subfraction (HDL3, the primary standard) and a reference serum (the secondary standard) were obtained. Within- and between-run coefficients of variation were acceptable for both methods. Apolipoprotein A-I concentrations, as measured in 60 subjects selected to present a large range of apolipoprotein content by electroimmunoassay (y1) and immunonephelometric assay (y2) with monoclonal antibodies, compared well with those measured by the same techniques but with polyclonal antibodies (x): r1 = 0.96, r2 = 0.99; y1 = 1.19x - 0.11 g/L, y2 = 0.98x. Comparison of results obtained by electroimmunoassay and immunonephelometric assay performed with monoclonal antibodies was also good: r = 0.96; y2 = 1.08y1 + 0.13 g/L.     

Measurement of calcitonin by immunoassay analyzers.

BACKGROUND: Since Nichols Institute Diagnostics (NID) ended production of their automated calcitonin immunoassay, evaluation of an alternative calcitonin assay was necessary. METHODS: Calcitonin measured by the NID procedure was compared with a test from Diagnostic Products Corporation (DPC). Calcitonin was also measured by assays from DiaSorin (DS) and Scantibodies (SC). RESULTS: In 187 samples, detection failed either by NID (39%) or DPC (96%) assay; in 35% of samples, the tests agreed for non-measurable calcitonin. The regression line DPC=0.78 x NID-0.3 (r=0.998) fitted results from 236 samples with detectable calcitonin. A linear relationship, albeit with scattering at low concentrations, was observed. Importantly, stimulation by pentagastrin above basal calcitonin concentrations yielded similar results with both assays. Comparisons (DPC=0.80 x DS-3.4 and DPC=0.81 x SC-1.1) confirmed an aberrant calibration of the DPC test. To overcome the method bias, we propose a multiplication factor of 0.8 to convert NID to DPC results. CONCLUSIONS: Apparently due to non-specific effects, the DS and SC assays produced calcitonin results in samples from completely thyroidectomized patients, while the DPC assay correctly failed to detect calcitonin. Thus, the DPC assay on an Immulite 2000 analyzer may be used as a more accurate substitute for NID if the different calibration is noted.     

A comparison of two creatine kinase MB determinations: a chromatographic versus an immunological method.

Determinations of the MB isoenzyme of creatine kinase by ion-exchange chromatography and by an immunological method show good overall correlation in plasma of patients with proven infarctions, r=0.98. In analyzing myocardial tissue samples, unexpected discrepancies are found between these two methods. An as-yet-unidentified creatine kinase, in chromatography behaving as MM and immunologically undistinguishable from MB, is found.     

Creatine kinase isoenzyme variants in human serum

In serum from about 800 patients, total creatine kinase and its subunit B activities were determined by the recommended Scandinavian creatine kinase method in the absence and presence of a creatine kinase M subunit inhibitory antibody. Eight patients had supranormal subunit B activities, but normal or near-normal values for total creatine kinase activity. Electrophoresis of sera from these eight patients showed, in addition to the normally migrating isoenzyme MM, one or two abnormally migrating creatine kinase isoenzyme bands, located between normally migrating isoenzymes MM and MB. Experimental data suggest that these abnormal bands may be isoenzyme BB with changed electrophoretic mobility. The eight patients had no particular disorder in common.     

Conversion of MM creatine kinase isoforms in human plasma by carboxypeptidase N

This study was undertaken to identify the carboxypeptidase(s) (CPase) in plasma mediating sequential conversion of the tissue isoform of the MM isoenzyme of creatine kinase (MM3 CK) to MM2 and MM1 isoforms and to elucidate relationships between CPase activity measured in plasma and observed rates of isoform conversion in vitro. Purified MM3 was incubated at 37 degrees C in plasma from normal subjects and patients with acute myocardial infarction. Isoforms were quantified by chromatofocusing. Preincubation with antiserum to CPase N prevented conversion of added MM3 to MM2 and MM1. Isoform conversion rates in the absence of antibody were proportional to plasma CPase N activity assayed spectrophotometrically by hydrolysis of furylacryloyl-L-alanyl-L-lysine substrate (r = 0.89, n = 8). Plasma CPase N activity varied by nearly 300% among individuals, but average activity was similar in samples from normal subjects (267 +/- 45 [SD] U/L, n = 18), those from outpatients with angina (289 +/- 43 U/L, n = 9), and those obtained at hospital admission from patients with acute infarction (Q wave: 279 +/- 70 U/L, n = 16; non-Q wave: 272 +/- 61 U/L, n = 14) or unstable angina (280 +/- 71 U/L, n = 11). In patients with Q wave infarction, CPase N activity increased by 43% +/- 25% between 48 hours and 72 hours (P less than 0.005 compared with admission) with a concomitant change in the rate of conversion of isoforms. Thus, the rate of conversion of isoforms in individual subjects can be estimated by assay of CPase N activity in plasma.(ABSTRACT TRUNCATED AT 250 WORDS)     

Performance evaluation of a new chemiluminescent cardiac troponin I assay

Cardiac troponin I is a marker for diagnosis of myocardial damage. Several immunoassays are currently available for determination of concentrations of troponin I in serum. We evaluated a chemiluminescent assay for troponin I using ACS:180 automated analyzer (Bayer Diagnostics). We compared our results with two other immunoassays using the OPUS Magnum (OPUS troponin I assay, Dade Behring) and AxSYM (microparticle enzyme immunoassay, Abbott) analyzers. The within-run and between-run CVs were less than 5% for all three levels of controls. The chemiluminescent assay for troponin I was linear up to a serum troponin I concentration of 50 ng/mL and the detection limit was 0.1 ng/mL of troponin. A good correlation between troponin I concentration measured by the chemiluminescent assay (y axis) and the microparticle enzyme immunoassay (MEIA) (x axis) was observed, although the concentrations of troponin I in individual specimens were approximately four times higher, when measured by the MEIA assay, than those measured by chemiluminescent assay. The correlation coefficient was 0.98 with the regression equation y = 0.22x + 1.125. We also observed a good correlation in troponin I concentrations obtained by the chemiluminescent assay (y axis) and OPUS troponin I assay (x axis). The correlation coefficient was 0.96 and the regression equation was y = 0.79x - 0.52. The correlation coefficient was 0.93 when we compared troponin I concentrations obtained by the OPUS assay (x axis) with the corresponding concentrations obtained by the MEIA assay (y axis). The corresponding regression equation was y = 0.25x + 3.5. We conclude that the chemiluminescent troponin I assay showed good analytical performance.     

Abbott IMx evaluated for assay of prostate-specific antigen in serum.

We evaluated a new fully automated procedure for quantitative measurement of prostate-specific antigen (PSA) by the Microparticle Enzyme Immunoassay (MEIA) technology developed for the Abbott IMx automated immunoassay system. The performance characteristics of the Abbott IMx PSA assay (y) were evaluated and compared with those of the Hybritech Tandem-E PSA assay (x), a solid-phase two-site immunoenzymometric assay. PSA values for both assays were well correlated (r = 0.99); regression analysis yielded the equation y = 0.92x - 0.23 micrograms/L. The Abbott assay proved reliable and reproducible, as shown by the intra- and interassay coefficients of variation (2.0-3.4% and 3.1-4.7%, respectively). The assay gave a linear standard curve up to 100 micrograms/L and was very sensitive (detected PSA < 0.1 microgram/L). This analytical sensitivity was comparable with that of the Tandem-E PSA assay. Overall, the IMx PSA assay demonstrated the accuracy, precision, linearity, and intermethod correlation required for monitoring patients with prostate cancer.     

Bioluminescence assay of creatine kinase and its isoenzymes in serum and cerebrospinal fluid.

We examined the sensitivity of bioluminescence for the determination of very low concentrations of creatine kinase brain-type subunit (CK-BB) in serum and in cerebrospinal fluid. To optimize the sensitivity of CK-isoenzyme assays and eliminate possible sources of error, we separated the isoenzyme fractions by using inhibiting anti-MM and precipitating anti-MM and anti-BB antibodies. The results with the bioluminescence assay correlated with spectrophotometric values such that r = 0.97 for the total CK activity and r = 0.98 for the CK-B activity. The reproducibility of the present method was comparable with the spectrophotometric method and was even better at low enzyme activities. The within-series precision for assay of total CK activity at 2 U/L corresponded to a CV of 9%; at 13 U/L the CV was 5.8%. All the assays were carried out at 25 degrees C. Even at this low temperature, CK activities as low as 0.2 U/L could be determined. In eight patients without any evidence of cerebral cell damage, total CK activity in cerebrospinal fluid was x = 1.05 +/- 0.6 U/L, and CK-BB activity was x = 0.7 +/- 0.4 U/L. In sera of these patients CK-BB activity was x = 0.6 +/- 0.5 U/L. Differences in CK and CK-BB activities in four patients with transient or progressive brain-cell damage are discussed.     

Isoforms of creatine kinase MM and MB in acute myocardial infarction: a clinical evaluation

Serum creatine kinase (CK, EC 2.7.3.2) isoenzymes MM and MB were resolved, respectively, into three (MM1, MM2, MM3) and two (MB1, MB2) isoforms (subforms derived from the same isoenzyme) by electrophoresis and the isoform patterns were determined in multiple sequential serum samples, timed from the onset of chest pain, from 58 patients with acute myocardial infarction (AMI). During the first 3 h after the onset of chest pain, the serum isoform activity resembled the pattern seen in normal volunteers. Specimens obtained 6 h after AMI showed predominantly MM3 and MB2 (45% and 11% of the total CK activity, respectively). Between 10 and 72 h, there was a gradual shift in which MM3, MM2 and MB2 decreased, while MM1 and MB1 increased. MB2 and MB1 disappeared from the pattern for samples collected after 24-48 h, while MM1 was always the most prominent band at the end of the observation period (66%, range 41-77%, at 48 h). These data suggest that a single determination of CK isoform pattern, drawn between 6 and 48 h after AMI, may provide an effective means of predicting the time of onset of necrosis. There were no significant differences in the CK isoform patterns according to infarct location and functional status of patients.