The Content and Distribution of Troponin I, Troponin T, Myoglobin, and alpha-hydroxybutyric Acid Dehydrogenase in the Human Heart

Transcription

1 Clinical Chemistry / TROPONINS I AND T, MYOGLOBIN, AND HBD IN THE HUMAN HEART The Content and Distribution of Troponin I, Troponin T, Myoglobin, and alpha-hydroxybutyric Acid Dehydrogenase in the Human Heart Joost C.J.M. Swaanenburg, PhD, 1 Petra J. Visser-VanBrummen, 1 Mike J.L. DeJongste, MD, PhD, 2 and Anton T.H.M. Tiebosch, MD, PhD 1 Key Words: Troponin I; Troponin T; HBD; alpha-hydroxybutyric acid dehydrogenase; Myoglobin; Myocardial tissue content Abstract We studied the content and distribution of heartspecific markers troponin I and troponin T in relation to conventional non heart specific myoglobin and alphahydroxybutyric acid dehydrogenase (HBD) in the hearts of 34 patients who died of various causes. Tissue was obtained from the right and left ventricles, the interventricular septum, and the right and left atria. We found significant differences in the contents expressed per gram wet weight tissue in the right and left ventricles for troponin I (by 1 of the 2 methods used), troponin T, myoglobin, and HBD and no differences per gram of protein. The biochemical contents per gram wet weight tissue and per gram protein were significantly lower in the right and left atria for all studied markers compared with the right and left ventricles. No significant differences were found in biochemical contents between the right and left atria. These findings imply that estimation of myocardial damage through cardiac markers levels in serum depends on the site of injury (atrium or ventricle). Comparison of myocardial injury among individuals using marker levels in serum is not reliable because of the varied ranges of markers in tissue contents. For several years, the level of cardiac markers in blood after an acute myocardial infarction (AMI) has been used to estimate the extent of myocardial tissue damage. 1 Its relevance is indicated by the fact that the prognosis after an AMI is related to the amount of myocardial tissue loss. 2 A frequently used conventional biochemical marker for this purpose is alpha-hydroxybutyric acid dehydrogenase (HBD), 3 a marker analogous to lactate dehydrogenase isoenzyme-1. However, this marker is non heart specific, since the level also is elevated after hemolysis. Another marker that can be used is myoglobin. After an AMI, the increase in the serum concentration of myoglobin is more rapid than the increase of HBD. In addition, the myoglobin concentration normalizes faster. Both markers are non heart specific, since myoglobin and HBD are also components of skeletal muscle tissue. Moreover, because myoglobin is cleared through the kidneys and, thus, depends on the renal function of the patient, the increased serum concentration after an AMI is more difficult to interpret than increased HBD activity. More specific biochemical markers for myocardial tissue damage have been reported, including the cardiac troponins. 4,5 The troponins are components related to the actin-myosin complexes of the thin filaments of striated muscle. These complexes are involved in muscle contraction and relaxation. Three forms of troponin have been described: troponin I, troponin C, and troponin T. Each troponin has its own function. Troponin C binds calcium, which activates muscle contraction. Troponin I inhibits the adenosinetriphosphatase activity of the actin-myosin complex, resulting in muscle relaxation. Troponin T binds the troponin I troponin C complex to tropomyosin. In contrast with troponin C, the amino acid sequences of troponin I and troponin T in skeletal muscle differ from those in myocardial muscle. To separate them immunologically, 77 Am J Clin Pathol 21;115: American Society of Clinical Pathologists

2 Clinical Chemistry / ORIGINAL ARTICLE specific antibodies directed against the cardiac troponins I and T were developed. These antibodies form the basis of the techniques for measuring concentrations of cardiac troponins I and T in serum. Because the amino acid composition of troponin C in skeletal muscle equals that in myocardial muscle, no cardiac troponin C method can be developed to specifically measure myocardial troponin C. To relate the extent of myocardial tissue loss to increased cardiac troponin concentrations in serum, the troponins should be distributed homogeneously in the myocardium and should have similar myocardial contents. 6 So far, a limited number of studies have been reported about the contents of cardiac markers in the heart However, these studies included a limited number of patients, studied hearts of nonhuman origin, focused on troponin I or T, analyzed only the contents of the ventricles, or expressed the contents per gram wet weight tissue or per gram of protein. Our aim was to study the distribution and content (expressed per gram wet weight tissue and per gram of protein) of cardiac troponin I (measured by 2 methods) and cardiac troponin T in the human myocardium in relation to the distribution and the content of the more conventional markers myoglobin and HBD. Materials and Methods Postmortem tissue was obtained from 34 patients at 7 sites in the heart and as negative control from the psoas muscle. Care was taken that all tissue samples were free of scars and necrosis. The samples were taken from the right ventricle (RV); the left ventricle (LV); the posterior, lateral, and anterior walls of the LV; the interventricular septum of the LV; the right atrium (RA); and the left atrium (LA). Immediately after harvesting, samples were weighed and frozen in liquid nitrogen. Subsequently, they were destructed in liquefied nitrogen using the Micro-dismembrator U (Braun Biotech International, Salm and Kipp, Breukelen, the Netherlands) at 2, rpm for 1 minute. After completing the procedure, 5 ml of modified phosphate-buffered saline (ph 7.) was added. The solutions were frozen and stored at 2 C. Subsequently, analysis was performed of the biochemical markers troponin I, troponin T, myoglobin, and HBD, and the protein concentrations were determined. Troponin I was measured using 2 analyzers. One analyzer was the AxSYM (Abbott Diagnostic Division, Hoofddorp, the Netherlands). 13 Instructions of the manufacturer were followed. The coefficients of variation (CVs) at 3 different levels were determined using control serum samples. The CVs were 6.8%, 5.2%, and 5.8%, respectively, for the levels 2.9, 7.6, and 29. ng/ml (2.9, 7.6, and 29. µg/l). The second analyzer was the Access (Beckman, Mijdrecht, the Netherlands) according to the instructions of the manufacturer. 14 The CVs for this assay were 8.9%, 8.2%, and 7.4%, respectively, for the levels.22, 5.88, and 25.4 ng/ml (.22, 5.88, and 25.4 µg/l). Troponin T measurements were performed with an Elecsys 21 analyzer (Roche, Almere, the Netherlands). For these measurements, the third-generation troponin T reagent was used. This reagent is calibrated with standards of human origin, whereas the second generation uses standards of bovine origin. Furthermore, the third generation uses a 1 amino acid modified signal antibody. These modifications result in a better linearity than that of the second-generation troponin T reagent. 15 Myoglobin was measured using the BNII nephelometer (Dade-Behring, Leusden, the Netherlands) according to the instructions of the manufacturer. HBD activities were measured at 37 C using HBD reagent (product number , Roche-Boehringer Mannheim, Almere, the Netherlands) on a Mega analyzer (Merck, Amsterdam, the Netherlands). For all methods, the linearity was validated by using various dilutions of different tissue homogenate solutions. No blank reaction was observed for any assay methods using the modified phosphate-buffered saline. The protein measurements were performed using the pyrogallol red method (product number A1217, Biotrol Diagnostic, Chennevières les-louvres, France, manufactured by Merck) on a Mega analyzer. The instructions of the manufacturer were modified by using 3 µl of homogenate solution and 25 µl of reagent. This modification resulted in an enhancement of the linearity up to 4 g/l. The imprecision was determined using 2 controls. The mean value of control 1 was.78 g/l (SD.4 g/l), and that of control 2 was 2.98 g/l (SD.11 g/l). The measured concentrations of the tissue homogenate solutions were between 1 and 3 g/l. Biochemical results were expressed as micrograms for both troponin I methods and for troponin T, milligrams for myoglobin, and units for HBD per gram wet weight myocardial tissue and per gram of protein. Statistical Analysis Differences of the biochemical markers among the sampled locations were tested by using the Friedman test. If significant differences were found with the Friedman test, the Wilcoxon signed rank test was used as the post hoc test between the individual sampled locations using the Bonferroni correction method. Differences in heart weight were considered by testing differences in content of biochemical markers for the sampled locations in normal (heart weight, 45 g or less) and hypertrophic (heart weight, more than 45 g) hearts using the Mann-Whitney U test. P values of.5 or less were considered statistically significant. Sex dependency also was considered. American Society of Clinical Pathologists Am J Clin Pathol 21;115:

3 Swaanenburg et al / TROPONINS I AND T, MYOGLOBIN, AND HBD IN THE HUMAN HEART Results The patient characteristics (age, sex, cause of death, heart weight, and interval between the death and autopsy) are given in Table 1. There were 18 women, mean age 62 years (range, years). Their mean heart weight was 365 g (range, g), and the mean interval between death and autopsy was 26 hours (range, 4-48 hours). For the 16 men included in the study, mean age was 66 years (range, years). The mean heart weight was 441 g (range, g), and the mean interval between death and autopsy was 15 hours (range, 6-24 hours). The patients died of a variety of causes: pulmonary disease, 1; (metastatic) cancer, 7; cerebral disease, 5; infectious disease, 4; heart disease, 4; vascular cause, 2; or miscellaneous causes, 2. Table 2 shows the content of the studied biochemical markers in the different parts of the LV. Table 2 shows that within the LV, there was no statistically significant difference between the sampled areas: posterior, lateral, and anterior walls and the interventricular septum. Therefore, mean values of the LV samples will be discussed. Figure 1 gives the ranges of the contents expressed per gram wet weight tissue for the studied biochemical markers per myocardial site and the ranges of the contents of the Table 1 Patient Characteristics Case No./Sex/Age (y) Cause of Death Heart Weight (g) Time Between Death and Autopsy (h) 1/F/86 Heart failure /M/56 Heart failure /M/91 Pneumonia /F/64 Amyloidosis, left ventricular hypertrophy /F/75 Cardiomyopathy /M/6 Intracranial bleeding 4 9 7/M/67 Metastatic lung cancer /M/62 Pulmonary embolism /M/52 Pulmonary embolism /F/52 Metastatic endometrial cancer /F/91 Respiratory insufficiency in interstitial lung disease /F/64 Metastatic endometrial cancer /M/73 Sepsis and multiorgan failure /F/47 Respiratory insufficiency /M/87 Bronchopneumonia, Wegener disease /F/4 Encephalitis /F/56 Meningitis /F/34 Sudden death, right ventricular dysplasia /F/76 Pneumonia 6 1 2/M/54 Metastatic lung cancer /F/29 Metastatic breast cancer /F/68 Pneumonia /M/66 Intracranial bleeding /M/7 Aortic aneurysm /M/6 Pulmonary embolism /M/63 Septic shock /F/49 Metastatic pancreatic cancer /M/61 Pancreatitis /M/59 Pulmonary embolism /F/53 Hemochromatosis /F/71 Abdominal aneurysm /M/74 Pancreatitis, adult respiratory distress syndrome /F/74 Cerebral infection /F/81 Metastatic cancer, unknown primary site Table 2 Content of Various Biochemical Markers at Different Places in the Left Ventricle of the Human Heart * Biochemical Marker Posterior Wall Lateral Wall Anterior Wall Septum Overall Troponin I, Access (µg) 77 (.2-543) 69 (.4-62) 68 (.1-459) 67 (.1-491) 71 (.4-62) Troponin I, AxSYM (µg) 2,72 ( ,52) 2,338 (8.4-22,374) 3,182 (8.6-15,756) 2,787 (6.5-1,678) 2,453 (6.5-22,374) Troponin T (µg) 85 (17-234) 117 (18-213) 94 (24-244) 17 ( ) 95 ( ) Myoglobin (mg) 2.4 ( ) 2.5 ( ) 2.2 ( ) 2.4 ( ) 2.4 ( ) HBD (U) 143 (38-247) 151 (13-27) 138 (62-196) 135 (5-183) 14 (5-27) HBD, alpha-hydroxybutyric acid dehydrogenase. * Data are given as median (range) per gram wet weight myocardial tissue. No statistically significant differences were detected. Access analyzer, Beckman, Mijdrecht, the Netherlands; AxSYM analyzer, Abbott Diagnostic Division, Hoofddorp, the Netherlands. 772 Am J Clin Pathol 21;115: American Society of Clinical Pathologists

4 Clinical Chemistry / ORIGINAL ARTICLE A B 7 25, 6 2, 5 Troponin I 4 3 Troponin I 15, 1, 2 1 5, C D Troponin T 3 2 Myoglobin E HBD Figure 1 The ranges of contents expressed per gram wet weight myocardial tissue for several cardiac markers sampled at different locations in the heart and the ranges of contents per gram wet weight skeletal tissue. A, Troponin I measured by the Access analyzer, Beckman, Mijdrecht, the Netherlands. B, Troponin I measured by the AxSYM analyzer, Abbott Diagnostic Division, Hoofddorp, the Netherlands. C, Troponin T (µg/g wet weight tissue). D, Myoglobin (mg/g wet weight tissue). E, alpha-hydroxybutyric acid dehydrogenase (HBD; U/g wet weight tissue). The contents of the RV and the LV were significantly higher than those of the RA and the LA for all markers (P <.5); for troponin I by the Access analyzer, troponin T, myoglobin, and HBD, the contents of the LV were significantly higher than those of the RV (P <.5). LA, left atrium; LV, left ventricle; RA, right atrium; RV, right ventricle; SM, skeletal muscle. American Society of Clinical Pathologists Am J Clin Pathol 21;115:

5 Swaanenburg et al / TROPONINS I AND T, MYOGLOBIN, AND HBD IN THE HUMAN HEART skeletal muscle tissue. Figure 2 shows these contents per gram of protein. The contents of troponin I by the Access method, troponin T, myoglobin, and HBD per gram wet weight myocardial tissue of the RV were significantly lower (P <.5) than those of the LV overall. In contrast, there was no difference in the contents per gram of protein. Furthermore, the contents per gram wet weight tissue and per gram of protein of all studied markers in the RV and the LV were significantly higher than those in the RA and the LA. The contents of biochemical markers in the RA did not differ significantly (P >.5) from those in the LA. For all sampled sites, we observed lower results with the Access method than with the AxSYM method. The contents of the studied troponins were at least 5 times less in skeletal muscle than in the various sites of the heart. The measured values of the skeletal muscle can be explained by nonspecific binding. This was confirmed by using immunoblot technology: in skeletal muscle tissue homogenate, no cardiac troponin could be demonstrated. The content of HBD in skeletal muscle was 1.5 to 5 times less than in the studied heart sites. The content of myoglobin in skeletal muscle was 1.5 to 4 times higher than in heart muscle tissue. Table 3 gives the median contents per gram wet weight myocardial tissue and per gram of protein for sampled sites (RV, LV overall, RA, and LA) and of the skeletal muscle tissue for both cardiac troponin I methods, cardiac troponin T, myoglobin, and HBD. We observed no heart-weight dependency of contents of the studied biochemical markers in any sampled locations between normal (n = 25) and hypertrophic (n = 9) hearts (all P values >.5). Furthermore, no differences were found in biochemical contents between the hearts of male and female patients (all P values >.5). Discussion We studied the contents of various cardiac markers per gram wet weight tissue and per gram of protein. In addition, we analyzed the distributions of cardiac troponin I, cardiac troponin T, HBD, and myoglobin in the human heart. The person-to-person variety of the contents was high (the contents differed for some persons more than 5 times). Furthermore, the distribution of the markers through the heart was not homogeneous. The biochemical contents in the RA and the LA were significantly lower than those in the RV and the LV, whereas the contents in the RA and the LA did not differ significantly. Furthermore, the contents expressed per gram wet weight tissue in the RV were, in contrast with the contents expressed per gram of protein, significantly lower than those in the LV for troponin I by the Access method, troponin T, myoglobin, and HBD. The contents within the LV showed no differences for all studied biochemical markers in the various LV sites. We also studied sex and heart-weight dependency. Although the mean heart weight of women (365 g) was lower than that of men (441 g), we found no statistically significant differences between the contents of the studied markers in female and male hearts. Furthermore, we found no statistically significant differences between the contents of normal and hypertrophic hearts. In clinical practice, the extent of myocardial tissue injury is based on levels of cardiac markers in serum. However, the nonhomogeneous distribution of cardiac markers in the heart implies the necessity of knowing the site of the damage for determining the loss of myocardial tissue: the atria should be analyzed apart from the ventricles. A consequence of the reported person-to-person variety in tissue contents of cardiac markers is that an individual patient with higher cardiac marker levels in serum after an AMI does not necessarily have more myocardial injury than another patient who has lower cardiac marker levels in the serum at the same time interval after an AMI. In contrast, groups of patients (for example, in clinical trials) could be compared for estimations of myocardial damage using cardiac marker levels in serum, if the following conditions have been fulfilled: similar distributions of baseline characteristics and sufficiently large sample sizes (type B error). The difference between the results of the 2 troponin I measurement methods is related to the nonuniformity of the troponin I analysis, which is, on the one hand, due to lack of consensus about the use of the antibody required to detect troponin I. On the other hand, there is no consensus about the composition of the gold standard for uniform troponin assay calibration. The results of the various troponin I methods will become more comparable if such a gold standard is accepted and introduced by the different assay manufacturers. However, realization of such a standard is complicated, as troponin is present in the circulation in various forms, including free and several complexes such as troponin I troponin C, troponin I troponin C troponin T, and troponin I troponin T. 16 Nevertheless, an American Association for Clinical Chemistry standardization subcommittee is working on this subject and probably will provide guidelines in the near future. 17 The influence of possible changes of the conformation and composition of troponin-bound complexes in the interval between death and autopsy was studied according to the procedure reported by Kragten et al. 9 The intervals between death and autopsy of the studied patients were related to the tissue contents for all studied cardiac markers. We found no significant (negative) correlation between the intervals and any of the studied cardiac markers. From this finding, we conclude that during the interval between death and autopsy, 774 Am J Clin Pathol 21;115: American Society of Clinical Pathologists

6 Clinical Chemistry / ORIGINAL ARTICLE A 2 B Troponin I 1 Troponin I C D Troponin T 5 Myoglobin E HBD Figure 2 The ranges of contents expressed per gram of protein for several cardiac markers sampled at different sites in the heart and the ranges of contents per gram of protein in skeletal muscle are shown. A, Troponin I measured by the Access analyzer. B, Troponin I measured by the AxSYM analyzer. C, Troponin T (µg/g protein). D, Myoglobin (mg/g protein). E, alpha-hydroxybutyric acid dehydrogenase (HBD; U/g protein). For all markers, the contents of the RV and LV were significantly higher than those of the RA and the LA (P <.5); there were no significant differences between the RV and the LV. LA, left atrium (P >.5); For explanation of abbreviations and proprietary information, see the legend for Figure 1. American Society of Clinical Pathologists Am J Clin Pathol 21;115:

7 Swaanenburg et al / TROPONINS I AND T, MYOGLOBIN, AND HBD IN THE HUMAN HEART Table 3 Content of Various Biochemical Markers at Different Places in the Human Heart and in Skeletal Muscle Tissue * Right Ventricle Left Ventricle Right Atrium Left Atrium Skeletal Muscle Biochemical Marker WW Protein WW Protein WW Protein WW Protein WW Protein Troponin I, Access (µg) Troponin I, AxSYM (µg) 2, , , , Troponin T (µg) Myoglobin (mg) HBD (U) HBD, alpha-hydroxybutyric acid dehydrogenase. * Data are given per gram wet weight (WW) of myocardial tissue and per gram of protein. For proprietary information, see Table 2. no significant changes take place in the conformation and composition of the troponin-bound complexes. Although for several years increased concentrations in blood of cardiac markers have been used to estimate the extent of myocardial necrosis, only a few publications address the distribution and content of biochemical parameters in the human heart. van der Laarse et al 1 reported no difference between the enzyme content expressed per gram wet weight tissue in the RV and the LV of tissue obtained by biopsy during open heart surgery. They reported a mean content of 12 U/g wet weight myocardium for HBD. In contrast with our experiments, these HBD measurements were performed at 25 C. Furthermore, these results were based on LV tissue obtained by biopsy from 6 patients and RV tissue obtained by biopsy from 9 patients. It is unclear whether these tissues were derived (partly) from the same patients. In contrast with the results of van der Laarse et al, 1 we observed statistically significant differences among the contents (per gram wet weight tissue) in the RV and the LV. In addition to the previously described discrepancy in the number of patients from whom tissue was harvested from the RV and LV (in our study, n = 34 vs van der Laarse et al, n = 6 [LV] and n = 9 [RV]), the collection procedure was different (autopsy vs biopsy). Finally, we used the Friedman and Wilcoxon signed rank tests to compare the results of the contents in the RV and the LV; van der Laarse et al 1 used a t test to test the significance of differences between means of 2 groups and the Fisher F test to test the significance of differences between variances of 2 groups. Kragten et al 9 reported the content per gram wet weight tissue for HBD and for troponin T from only the LV tissue derived from 17 patients at autopsy. They reported a mean HBD content of 156 U/g wet weight, which is in agreement with our findings. The mean troponin T content was reported as 234 µg/g wet weight. This is more than twice the value we observed. However, Kragten et al 9 used the second-generation troponin T reagent, whereas we used the third-generation reagent. As reported, 15 this third-generation troponin T assay is more accurate, demonstrates improved linearity, and shows lower results beyond values of.2 µg/l compared with the second-generation troponin T assay. Between. and.2 µg/l there is no difference in results between the second- and the third-generation assays. The difference in results starts at a level of.2 µg/l. At a level of 1 µg/l, the results of the third-generation reagent are 5% of those determined by the second-generation reagent. At a level of 25 µg/l, the results with the third-generation reagent are 4% of those with the second-generation reagent. 7 All our measurements in the tissue homogenates were performed in diluted samples because the undiluted tissue-homogenate solutions would be out of the analyzer range. Homogenate solutions were diluted to concentrations between 1 and 15 µg/l. As stated, the difference between the second- and third-generation troponin T reagents is, at this level, a factor between 2 and 2.5. This might be a reasonable explanation for the difference in troponin T contents in our findings and those reported by Kragten et al. 9 Katus et al 11 reported only free troponin T contents in the LVs of hearts from 3 patients undergoing heart transplantation. These patients are different from our population, because most of the patients in our study had no history of heart disease, and none of the patients underwent heart transplantation. Bleier et al 12 reported the content per gram wet weight tissue of myoglobin, troponin I, and troponin T in the RA from 11 patients undergoing heart surgery. They reported a mean myoglobin content of.97 mg/g wet weight, which is in agreement with our findings for the RA. The lack of standardization of the troponin analysis and the use of other troponin I and troponin T methods may explain the different outcomes of the troponin contents. Voss et al 7 reported the content per gram of protein for cardiac troponin T and for myoglobin in the LV and the RV in hearts of 3 healthy and 3 diseased persons. The cardiac troponin T measurements were not performed with the thirdgeneration reagent, so these results cannot be compared with our findings. The mean content of myoglobin of the LV for the healthy persons was 18.4 mg/g and for diseased persons, 776 Am J Clin Pathol 21;115: American Society of Clinical Pathologists

8 Clinical Chemistry / ORIGINAL ARTICLE 49.8 mg/g (overall mean for these persons was 34.1 mg/g). For the RV, these contents were reported as 28.8 and 66.4 mg/g, respectively (overall mean, 47.6 mg/g). We report similar median values of 46.1 mg/g for the LV and 45.1 for the RV. We conclude that there are statistically significant differences among the contents per gram wet weight tissue in the RV and the LV for the biochemical markers troponin I by the Access method, troponin T, myoglobin, and HBD. The biochemical contents per gram wet weight tissue and per gram of protein in the RA and the LA were significantly lower for all studied biochemical markers compared with those in the RV and the LV. There were no statistically significant differences between the biochemical contents in the RA and the LA. Furthermore, within the LV, there was no statistically significant difference in the posterior, lateral, and anterior wall and interventricular septum sites. For clinical practice, these findings imply that for estimation of myocardial damage using cardiac marker levels in serum, it is useful to know the site (atrium or ventricle) of damage in the heart. Furthermore, comparison of myocardial damage among individuals is not reliable using cardiac marker levels in serum because of the variety in ranges of tissue contents. In contrast, comparison among groups of patients (clinical trials) is possible, if the composition of these groups is similar (eg, the same person-to-person variety) and if the group sizes are large enough. From the Departments of 1 Pathology and Laboratory Medicine and 2 Cardiology, University Hospital Groningen, Groningen, the Netherlands. Address reprint requests to Dr Swaanenburg: Sint Jans Gasthuis, Postbus 29, 6 AA Weert, the Netherlands. References 1. Witteveen SAGJ, Hemker HC, Hollaar L, et al. Quantitation of infarct size in man by means of plasma enzyme levels. Br Heart J. 1975;37: Fioretti P, Slavo M, Brower RVV, et al. Prognosis of patients with different peak serum creatine kinase levels after first myocardial infarction. Eur Heart J. 1985;6: van der Laarse A, Hermens WT, Hollaar L, et al. Assessment of myocardial damage in patients with acute myocardial infarction by serial measurement of serum alphahydroxybutyrate dehydrogenase levels. Am Heart J. 1984;17: Bodor GS. Cardiac troponin I: a highly specific biochemical marker for myocardial infarction. J Clin Immunoassay. 1994;17: Wu AH. Cardiac troponin T: biochemical, analytical, and clinical aspects. J Clin Immunoassay. 1994;17: Adams JE, Abendschein DR, Jaffe AS. Biochemical markers of myocardial injury: is MB creatine kinase the choice for the 199s? Circulation. 1993;88: Voss EM, Sharkey SW, Gernert AE, et al. Human and canine cardiac troponin T and creatine kinase-mb distribution in normal and diseased myocardium. Arch Pathol Lab Med. 1995;119: Apple FS. Tissue specificity of cardiac troponin I, cardiac troponin T and creatine kinase-mb. Clin Chim Acta. 1999;284: Kragten JA, Hermens WT, VanDieijen-Visser MP. Quantification of cardiac troponin T release into plasma after acute myocardial infarction: only fractional recovery compared with enzymes. Ann Clin Biochem. 1996;33: van der Laarse A, Dijkshoorn NJ, Hollaar L, et al. The (iso)enzyme activities of lactate dehydrogenase, alphahydroxybutyrate dehydrogenase, creatine kinase and aspartate aminotransferase in human myocardial biopsies and autopsies. Clin Chim Acta. 198;14: Katus HA, Remppis A, Scheffold T, et al. Intracellular compartmentation of cardiac troponin T and its release kinetics in patients with reperfused and nonreperfused myocardial infarction. Am J Cardiol. 1991;67: Bleier J, Vorderwinkler KP, Falkensammer J, et al. Different intracellular compartmentations of cardiac troponins and myosin heavy chains: a causal connection to their different early release after myocardial damage. Clin Chem. 1998; 44: Apple FS, Maturen AJ, Mullins RE, et al. Multicenter clinical and analytical evaluation of the AxSYM Troponin-I immunoassay to assist in the diagnosis of myocardial infarction. Clin Chem. 1999;45: Christenson RH, Apple FS, Morgan DL, et al. Cardiac troponin I measurement with the ACCESS immunoassay system: analytical and clinical performance characteristics. Clin Chem. 1998;44: Klein G, Baum H, Gurr E, et al. Multicenter evaluation of two new assays for myoglobin and troponin T on the Elecsys 21 and 11 analyzers [abstract]. Clin Chem. 1999;45:A Katrukha AG, Bereznikova AV, Esakova TV, et al. Troponin I is released in bloodstream of patients with acute myocardial infarction not in free form but as complex. Clin Chem. 1997;43: Wu AHB, Feng Y, Moore R, et al, for the American Association for Clinical Chemistry Subcommittee on ctni Standardization. Characterization of cardiac troponin subunit release into serum after myocardial infarction and comparison of assays for troponin T and I. Clin Chem. 1998;44: American Society of Clinical Pathologists Am J Clin Pathol 21;115: