Quantification of myocardial infarction: Template model for serial creatine kinase analysis

Abstract

A self-modeling procedure was used to develop a template g(z) from the serial creatine kinase (CK) release in 32 patients with acute myocardial infarction. An additional 16 patients were used as an extrinsic test of the template model. For a given patient the fitted CK curve γ(t) is related to the template g(z) by the expression γ(t) = β1g (t-β2/β3) + β4, where β1 is equal to the peak height of the CK transient above background, β4, β2 is the time at which CK begins to rise, β3 is the time taken for CK to rise to its maximum value. Calculations of infarct size using the template and a numerical estimate yielded values of 34.6 g-Eq and 33.5 g-Eq, respectively, with good agreement (r = 1.00). Comparisons of all point numerical estimations of infarct size with early point predictions revealed that the template and lognormal models performed equally well with 7 and 6 points; however, the template model was superior with 5 and 4 points. The template model also provides insight into the CK time activity curve. In particular, total CK activity and hence infarct size, are shown to be proportional to the peak of the excess CK curve, and the time course of CK appearance is revealed by the appearance and cumulative appearance functions. We found a high correlation between all point estimates of completed infarct size and a linear estimate obtained by fitting a straight line to the ascending portion of the CK curve.