Myocardial Salvage in Patients With Non–ST-Elevation Myocardial Infarction Determined by Myocardial Perfusion Imaging Michael C. Kontos, MD, Karen A. Kurdziel, MD, Joseph P. Ornato, MD, Robert L. Jesse, MD, PhD, and James L. Tatum, MD We compared acute and late myocardial perfusion imaging using technetium-99m isotopes in 69 pa- tients who had non–ST-elevation myocardial infarc- tion. Among these patients, we found that the isch- emic risk area was often large (19% of the left ventricle), and that 67% had significant myocardial salvage, defined as a >25% decrease in risk area, which was associated with an improved ejection fraction. 2005 by Excerpta Medica Inc. (Am J Cardiol 2005;95:398 – 401) S erial myocardial perfusion imaging (MPI) in pa- tients with ST-elevation myocardial infarction (MI) has shown successful therapeutic results for sub- stantial myocardial salvage. 1 Similar studies have not been performed in patients who have non–ST-eleva- tion MI. We report findings of serial MPI performed in a cohort of patients who had non–ST-elevation MI. ••• The evaluation and triage strategy for patients who have chest pain at our institution has been reported in detail previously. 2 All patients who present to the emergency department with complaints suggestive of myocardial ischemia undergo an evaluation by house staff and attending physicians that includes a medical history, physical examination, and electrocardiogram. Patients at high risk (e.g., ischemic ST-segment changes or known coronary disease with typical symptoms) are admitted directly to the coronary care unit. Patients who have a nonischemic electrocardio- gram and are initially considered at low or moderate risk for acute coronary syndrome undergo further risk stratification according to a routine clinical protocol that uses MPI at rest. 2 Those who have positive find- ings on MPI are admitted to the coronary care unit and undergo serial sampling of cardiac markers (creatinine kinase, creatinine kinase-MB, and troponin I). Deci- sions regarding additional diagnostic evaluations are made by the attending cardiologist based on clinical variables, myocardial markers, and test results. We define MI as a creatinine kinase-MB level 5.0 ng/ml in association with a typical increase in markers and symptoms consistent with myocardial ischemia. After January 1996, an elevated level of troponin I was also required. Patients were included if they had MI confirmed by increased markers and repeat MPI, which was per- formed for clinical reasons (n = 44) or as part of a clinical research protocol (n = 25) approved by the institutional review board. MPI and interpretation were performed as described previously. 2 Patients were injected with 20 to 30 mCi of sestamibi or tetrofosmin and underwent gated single-photon emis- sion computed tomographic imaging. Data were pro- cessed according to predefined commercial protocols to generate short- and long-axis static reconstructions and multilevel gated cines for visual interpretation. Ejection fraction (EF) was determined quantitatively with a previously validated automated algorithm 3 and was considered normal at 50%. Defect size quantitation was performed as de- scribed previously. 4 Compared with a commercially available phantom, a threshold value of 50% of peak counts resulted in the highest correlation be- tween calculated and known defect sizes (r = 0.99). Each pixel was normalized to the peak value for the study. After application of the 50% threshold, short- axis images were saved in a Tiff format and ex- ported into a commercial software program (Adobe PhotoShop 5.0, Schaumburg, Illinois). A circular region of interest was overlaid on each short-axis slice, and obvious extracardiac activity was manu- ally removed. From each short-axis slice, the area of nonwhite pixels, which represented areas with 50% uptake, was selected and quantitated. A sec- ond set of images was created by manually filling in the visualized defect (if any). This represented the normally perfused heart. The difference between the 2 represented the defect size. Repeat MPI was used to determine final infarct size. No patient had repeat MI between the 2 sets of images. Myocardial salvage was calculated as (initial defect size - final defect size)/initial defect size. We defined “early revascularization” as revascularization 9 to 12 hours after presentation and “late revascularization” as revascularization 12 hours after presentation but before repeat MPI at rest. Results are presented as mean  1 SD. Continuous and categorical data were compared with Student’s t test and chi-square analysis, respectively. A p value 0.05 was considered statistically significant. Initial and final defect sizes and EF were compared with the paired t test. During the study, 69 patients had acute MPI in the emergency department, were diagnosed with MI, and From the Departments of Internal Medicine, Cardiology Division, Emergency Medicine, and Radiology, Virginia Commonwealth Uni- versity, Richmond, Virginia. Dr. Kontos’s address is: Room 7-074, Heart Station, North Hospital, P.O. Box 980051, Medical College of Virginia, 1300 East Marshall Street, Richmond, Virginia 23298- 0051. E-mail: mkontos@hsc.vcu.edu. Manuscript received July 28, 2004; revised manuscript received and accepted September 29, 2004. 398 ©2005 by Excerpta Medica Inc. All rights reserved. 0002-9149/05/$–see front matter The American Journal of Cardiology Vol. 95 February 1, 2005 doi:10.1016/j.amjcard.2004.09.042