Rapid assessment of patients with non—ST-segment elevation acute chest pain: Troponins, inflammatory markers, or perfusion imaging?

BACKGROUND: Perfusion imaging during or soon after pain has been shown to provide diagnostic and prognostic information in patients with suspected angina. Measurement of troponin I (TnI) and troponin T (TnT) provides similar information but only several hours after onset of pain. The role of inflammatory markers in this setting is less clear. METHODS AND RESULTS: We prospectively studied 80 nonconsecutive patients using gated technetium 99m sestamibi single photon emission computed tomography (MIBI), TnT, TnI, C-reactive protein, IL-6, and tumor necrosis factor alpha. Of these patients, 50 (63%) had abnormal MIBI, 13 (17%) had elevated TnT, 17 (21%) had elevated TnI, and C-reactive protein, IL-6, and tumor necrosis factor alpha were raised in 46 (58%), 14 (18%), and 29 (37%), respectively. Myocardial infarction was the presenting event in 13 patients (16%), and 23 (34%) of those without index myocardial infarction sustained a cardiovascular event during follow-up. MIBI, TnT, TnI, and electrocardiogram all had similar negative predictive values for index myocardial infarction (97%, 97%, 95%, and 97%, respectively). However, only MIBI had a high negative predictive value for the prediction of subsequent events during follow-up (86%). TnT and MIBI were the only independent predictors of all events. Inflammatory markers provided no useful additional prognostic information. CONCLUSIONS: The combination of TnT and MIBI is the best model for early prediction of cardiac events in patients with acute chest pain.     

Detection and size estimation of acute myocardial infarction using 99Tc-glucoheptonate.

Twenty-seven patients with suspected acute myocardial infarction were studied by precordial scanning after intravenous administration of 99mTc-gluchoheptonate 2-48 hr after the onset of chest pain. Fifteen of the patients had clinically documented acute myocardial infarctions. Twelve of these 15 (80%) had areas of distinctly increased tracer uptake in the region of the heart. The three infarctions not identified by scan had peak serum CPK values of less than 300. In seven patients without infarction, no distinct areas of increased tracer uptake were found in the region of the heart. Five patients could not be classified as to whether infarction had or had not occurred. Three had abnormal scans, the significance of which is uncertain. Infarct size was estimated from the area of increased 99mTc-glucoheptonate concentration on scan and compared to peak serum CPK values. A linear correlation with a correlation coefficient of 0.77 was found. Technetium-99-m-glucoheptonate scanning was useful for the identification and size estimation of moderate- to large-size transmural and nontransmural acute myocardial infaractions.     

Performance of rest myocardial perfusion imaging in the management of acute chest pain in the emergency room in developing nations (PREMIER trial)

Background

Rest myocardial perfusion imaging (MPI) is effective in managing patients with acute chest pain in developed countries. We aimed to define the role and feasibility of rest MPI in low-to-middle income countries.

Methods and Results

Low-to-intermediate risk patients (n?=?356) presenting with chest pain to ten centers in eight developing countries were injected with a Tc-99m-based tracer, and standard imaging was performed. The primary outcome was a composite of death, non-fatal myocardial infarction (MI), recurrent angina, and coronary revascularization at 30?days. Sixty-nine patients had a positive MPI (19.4%), and 52 patients (14.6%) had a primary outcome event. An abnormal rest-MPI result was the only variable which independently predicted the primary outcome [adjusted odds ratio (OR) 8.19, 95% confidence interval 4.10-16.40, P?=?.0001]. The association of MPI result and the primary outcome was stronger (adjusted OR 17.35) when only the patients injected during pain were considered. Rest-MPI had a negative predictive value of 92.7% for the primary outcome, improving to 99.3% for the hard event composite of death or MI.

Conclusions

Our study demonstrates that rest-MPI is a reliable test for ruling out MI when applied to patients in developing countries.     

Impact of myocardial perfusion imaging on clinical management and the utilization of hospital resources in suspected acute coronary syndromes

Recent advances in the treatment of acute coronary syndromes has raised awareness in the community that prompt presentation for chest pain may be life saving. Each year in the United States, more than 6 million people present to the hospital with an acute chest pain, making this the most common presenting chief complaint second only to abdominal pain. Most patients presenting with chest discomfort have a non-ischaemic electrocardiogram on presentation. However, these patients are routinely admitted to hospital due to diagnostic uncertainty for occult myocardial infarction or ischaemia. As an approach to this dilemma, many hospitals have created protocols as a means of facilitating the identification of infarction and ischaemia and the safe and effective triage of patients with a chief complaint of chest pain. Myocardial perfusion imaging at rest has been shown to be highly sensitive for the detection of acute myocardial infarction, and can be supplemented with provocative testing after infarction has been excluded. Diagnostic strategies that utilize myocardial perfusion imaging for the evaluation of acute chest pain have successfully improved the triage of these patients by avoiding inadvertent discharge of patients with myocardial infarctions, and reducing unnecessary hospital admissions and overall cost and expenditure.     

Comparison of Tc-99m pyrosphosphate and Tc-99m hydroxymethylene diphosphonate in acute myocardial infarction: concise communication

A clinical comparison between a new bone seeking radiopharmaceutical, Tc-99m hydroxymethylene diphosphonate (TcHMDP) and the standard agent, Tc-99m pyrophosphate (TcPPi), was performed in 18 patients with acute myocardial infarction. Each patient was imaged initially with either TcHMDP or TcPPi, and imaged 24 hr later with the other tracer. All 18 patients had images positive for acute myocardial infarction with TcPPi, whereas 16 of 18 patients (89%) had positive studies with TcHMDP. The TcPPi images were graded significantly superior to those obtained with TcHMDP in 61% of the patients, and they were equal in 33%. In only one patient (6%) was TcHMDP better. The results indicate that compared with TcHMDP, TcPPi not only has a superior sensitivity for acute myocardial infarction but also has a significantly increased intensity of uptake in positive areas. TcPPi remains the agent of choice for myocardial infarct imaging.     

Indium-111-antimyosin antibody imaging for detecting different stages of myocardial infarction: comparison with technetium-99m-pyrophosphate imaging

The diagnostic value of 111In-antimyosin (AM) imaging for identifying myocardial infarction was evaluated in comparison with 99mTc-pyrophosphate (PPi) imaging. Twenty-four patients with various stages of myocardial infarction, ranging from three days to nine months after the onset of infarction, underwent both AM and PPi scans. Of 26 infarct lesions AM scan identified 22 (85%), while PPi scans detected 10 (38%) (p less than 0.01). When less than a week had passed since the onset both scans demonstrated all infarct lesions. For seven subacute lesions studied within one to two weeks of onset, AM scans detected (100%), while PPi scans identified only 2 (29%). Furthermore, AM scans showed discrete myocardial uptake in 7 (64%) of those studied more than two weeks after onset. The intensity of AM uptake in the infarcts studied more than seven days after onset was less than that in acute infarcts studied within seven days of onset (p less than 0.05). These preliminary data indicate that the abnormal myocardial uptake of AM persists beyond the first two weeks when PPi no longer accumulates. Thus, AM scans can be considered to provide a sensitive diagnosis of subacute as well as acute myocardial necrosis.     

Early isonitrile SPECT in acute myocardial infarction: feasibility and results before and after fibrinolysis

Early 99Tcm-labelled methoxy-isobutyl-isonitrile (MIBI) SPECT was performed in 14 patients with suspected acute myocardial infarction (AMI). The radiopharmaceutical was administered immediately upon admission to the intensive care unit and before any diagnostic confirmation. Then, if decided, thrombolytic therapy was started. Cardiac imaging was performed 1 h later, and as there is no significant re-distribution, the pictures still showed the pre-treatment MIBI uptake. In three cases acute myocardial infarction was not confirmed. For one of them, the result was normal and this patient was ultimately considered to have had a transient ischaemic event. The two other cases had acute chest pain with a previous history of myocardial infarction (MI) and a pathological MIBI SPECT. In the 11 cases with confirmed first AMI significant perfusion defect was seen. For every patient a new MIBI injection with a control SPECT was repeated 72 h after admission. Eight patients were seen 1 h 15 min to 3 h 15 min after the onset of chest pain and had thrombolytic therapy. Defects were always in agreement with coronary angiography and 2D echocardiography performed in the same period. After thrombolysis, control SPECT showed no recovery in three cases, partial recovery in four, and nearly complete recovery in one. Using this technique, it was then possible to get high quality myocardial perfusion imaging without delaying treatment of AMI. This preliminary series suggests that MIBI SPECT may be useful in accurately showing the size and location of the immediate perfusion defect, and in assessing the response to emergency therapy of AMI, especially thrombolysis.     

Myocardial perfusion imaging versus biochemical markers in acute coronary syndromes

The assessment and appropriate clinical management of patients with acute chest pain and non-diagnostic electrocardiograms remain a continuing clinical problem. Accordingly, there is considerable interest in evaluating new strategies to improve early diagnostic accuracy in patients with possible acute myocardial ischaemia. Cardiac troponins (T and I) and acute rest myocardial perfusion imaging have similar sensitivities for detecting acute myocardial infarction. Whereas cardiac markers require 6-12 h to become positive, acute rest myocardial perfusion imaging immediately reflects the status of regional myocardial blood flow at the time of radiopharmaceutical injection. The measurement of cardiac troponins is particularly useful in the diagnosis and estimation of the degree of myocardial injury in those patients with a high likelihood of coronary artery disease and myocardial necrosis and for prognostication of adverse cardiac events in those patients with unstable angina. In contrast, the most appropriate use of acute rest myocardial perfusion imaging is in the setting of patients with acute ischaemic symptoms, non-diagnostic electrocardiogram and a low likelihood of myocardial necrosis, in which early imaging will assist in effective triage decisions.     

Technetium-99m antimyosin antibody (3–48) myocardial imaging: human biodistribution,safety and clinical results in detection of acute myocardial infarction

Technetium-99m antimyosin (^99mTc-AM) antibody imaging may have significant advantages over indium-111 antimyosin in clinical practice. The purpose of this study was to determine the human biodistribution, the safety profile and the sensitivity of^99mTc-AM (3–48) imaging in the detection of both Q-wave and non-Q-wave myocardial infarction (MI). Biodistribution and safety parameters were mainly determined in 12 normal healthy volunteers while 40 patients with proven MI (22 Q-wave, 18 non-Q-wave) were injected with^99mTc-AM (20–25 mCi) between 5 h and 7 days after the onset of acute chest pain. Three standard planar views were performed at 6 h and at 24 h post injection. Both sets of images were completed in 33 patients while two patients were imaged only at 6 h, three patients only at 18 h and one at 18 and 24 h. One patient was not imaged. Vital signs and ECG were recorded and blood samples for haematology, biochemistry and human antimurine antibodies (HAMA) and urinalysis were obtained in all volunteers and patients. No serious adverse reactions or side-effects attributable to^99mTc-AM have been reported. No volunteers or patients developed allergic reactions or significant increases in HAMA titres. Reading of^99mTc-AM imaging was performed by two blinded experienced observers. The sensitivity of^99mTc-AM in the detection of MI was 100% (21/21) for Q-wave and 83.3% (15/18) for non-Q-wave infarctions. The overall sensitivity was 92.3% (36/39). The three false-negative cases were inferoposterior MI. A certain degree of uptake focalization was seen in 26 out of 35 (74.2%) at 6 h. At 24 h, two patients (5.8%) did not show^99mTc-AM uptake while 22 (64.7%) showed intense focal uptake, seven (20.6%) moderate uptake and 3 (8.9%) slight uptake. It is concluded that^99mTc-AM (3–48) imaging is safe and shows high sensitivity in the detection of both Q-wave and non-Q-wave MI even with early imaging (6 h post injection). These promising results warrant further clinical investigation.     

Prognosis in patients with spontaneous chest pain, a nondiagnostic electrocardiogram, normal cardiac enzymes, and no evidence of severe resting ischemia by quantitative technetium 99m sestamibi tomographic imaging

Background  There are limited data addressing the outcome of patients with normal or near normal myocardial perfusion during chest pain at rest. The purpose of this study was to determine the prognosis of patients with spontaneous chest pain, a normal or nondiagnostic electrocardiogram, no enzymatic evidence of myocardial infarction, and no evidence of severe resting ischemia by quantitative technetium 99m (^99mTc) sestamibi imaging. Methods  In the study, 111 patients who fulfilled the above criteria were injected with ^99mTc sestamibi during resting chest pain and were followed for a median 2.7 years. Of the patients in the study group, 58% had coronary artery disease that was documented by clinical history or coronary angiography. Tomographic ^99mTc perfusion images were interpreted with a quantitative threshold technique initially developed to detect severely hypoperfused myocardium. The images were also interpreted qualitatively to detect patients with milder degrees of hypoperfused myocardium. Results  During follow-up 3 patients had cardiac deaths, 5 had nonfatal myocardial infarctions, and 21 underwent revascularization procedures (13 within 3 months and 8 more than 3 months after the sestamibi study). At 3 years, survival free of cardiac death was 97%, survival free of cardiac death or myocardial infarction was 91%, and survival of cardiac death, myocardial infarction, or late revascularization was 82%. Quantitative analysis of the scans revealed that 100% of patients without fixed defects had 3-year survival free of cardiac death versus 76% of patients who had fixed defects (p<0.001). Mild to moderate resting ischemia by qualitative interpretation of the scans was present in 20% of patients, but this did not predict outcome. Conclusions  Patients with spontaneous chest pain and nonischemic quantitative ^99mTc sestamibi images were at reasonably low risk for hard cardiac events although some patients (18%) required revascularization. Supported in part by the DuPont Merck Pharmaceutical Company, N. Billerica, Mass.     

Rectilinear scanning in the detection of acute myocardial infarction.

Myocardial scintigraphy with phosphate bone scanning agents is a new and very useful development in the detection of acute myocardial infarction. Initial experience using a rectilinear scanner is described in 50 consecutive patients admitted to hospital because of suspected myocardial infarction. The routine dose was 15 mCi/5 mg, 99Tcm stannous pyrophosphate intravenously with anterior, left anterior oblique, and lateral scans obtained 45-90 minutes after injection. There was only one false negative scan in 17 patients with proven acute myocardial infarction and that was 16 days after onset. There was no proven false positive investigation in seven patients in which fresh myocardial infarction could be definitely excluded, or in a further 11 cases in which it could be excluded with reasonable confidence. Myocardial scintigraphy was considered helpful in resolving the diagnostic problem in 27 out of 29 equivocal cases. It is concluded that myocardial scintigraphy with a rectilinear scanner is a highly accurate and safe procedure in the detection of acute myocardial infarction. The optimum scanning time is two to six days after onset of suspected infarction, when a diagnostic accuracy approaching 100 per cent can be expected.     

The role of cardiovascular imaging techniques in the assessment of patients with acute chest pain

Chest pain is the most common presenting symptom of coronary artery disease. The assessment and appropriate management of patients with acute chest pain and non-diagnostic electrocardiograms (ECGs) remain a continuing clinical problem, with major logistic and financial implications for health-care providers. Cardiovascular imaging is at the forefront of health care, experiencing rapid changes over the recent years, particularly with the use of advanced medical technologies. Imaging techniques like acute rest myocardial perfusion imaging (MPI), echocardiography, electron beam computed tomography (CT), cardiac magnetic resonance imaging (MRI) and multi-detector CT (MDCT) have been used recently in the evaluation and triage of patients with chest pain in addition to the conventional investigations such as ECGs and cardiac biomarkers in the chest pain units. The annual potential cost savings, by incorporating the routine use of acute rest MPI in patients with low-to-moderate risk and non-diagnostic ECGs are substantial. The high negative predictive value of a normal resting MPI in patients with chest pain for myocardial infarction and future cardiac events is well established. Echocardiography is also considered to be useful but the technique is operator dependent and at present there is insufficient data to support its use. Cardiac MRI is expensive and time consuming and there is insufficient diagnostic and prognostic data to make it suitable for chest pain patients at present. There has been increasing interest in MDCT recently, especially with the advent of 64-slice CT but the sensitivity and specificity in chest pain patients are no better than MPI so far.     

MRI for acute chest pain: Current state of the Art

This article reviews the magnetic resonance imaging (MRI) and angiography (MRA) techniques, imaging findings, and evidence for evaluating patients with acute chest pain due to acute pulmonary embolus (PE), aortic dissection (AD), and myocardial infarction (MI). When computed tomographic angiography (CTA) is contraindicated, MRI and MRA are important alternative imaging modalities for diagnosis and management of patients with acute PE, AD, and MI. Familiarity with the techniques, imaging findings, and evidence is critical to safely and appropriately managing patients presenting with acute chest pain. J. Magn. Reson. Imaging 2013;37:1290–1300. © 2013 Wiley Periodicals, Inc.     

Non-invasive in vivo imaging of myocardial apoptosis and necrosis

Myocardial necrosis plays an important role in the pathogenesis of various cardiovascular disorders and can result from different myocardial insults. Its non-invasive identification and localisation therefore may help in the diagnosis of these disorders, as well as in prognosis and assessment of treatment response. Apoptosis, or programmed cell death, is important in the spectrum of myocardial damage since it is gradually becoming more apparent that cell death may begin as apoptosis and not as necrosis. First attempts to directly visualise the area of myocardial necrosis were based on recognition of myocardial infarction with "hot spot imaging agents" in patients with chest pain. Since then, the study of myocardial necrosis with gamma imaging agents has gone beyond the detection of myocardial infarction, and attempts have been made to diagnose other cardiovascular disorders associated with cardiac cell death such as heart transplant rejection, myocarditis, cardiotoxicity and cardiomyopathies. Traditionally, two hot spot imaging agents have been used for the detection of myocardial necrosis, (99m)Tc-pyrophosphate and (111)In-antimyosin. In addition, preliminary studies have demonstrated promising results with (99m)Tc-glucarate. Recently, (99m)Tc-annexin V has been successfully used for non-invasive gamma imaging of apoptosis after acute myocardial infarction, acute myocardial ischaemia, acute cardiac allograft rejection and malignant intracardiac tumours. This review article focusses on the characteristics of these different myocardial necrotic and apoptotic markers and compares their role in the assessment of myocardial damage.     

Impact of acute chest pain Tc-99m sestamibi myocardial perfusion imaging on clinical management

BACKGROUND: Previous studies have demonstrated the high sensitivity and specificity of technetium 99m sestamibi scintigraphy during acute chest pain myocardial perfusion imaging. However, no study has shown that this technique would alter clinical management in practice. METHODS AND RESULTS: One hundred twenty consecutive patients were injected with Tc-99m sestamibi (22 mCi) during pain; single photon emission computed tomography was performed 1 to 6 hours later. The population included inpatients and those who arrived at the emergency department with chest pain deemed to be at intermediate risk for myocardial ischemia. The requesting physician completed a questionnaire before the study, indicating the likelihood of cardiac disease and proposed management had the test not been available. Follow-up management was evaluated from medical records. There was a 34% reduction in total admissions and 59% in planned admissions to the coronary care unit (P <.001). Conversely, 7 patients had discharge cancelled and 17 required coronary care purely because of abnormal acute rest myocardial perfusion imaging results. Coronary angiography was reduced by 40% in a selected subgroup. CONCLUSIONS: In our population, acute rest myocardial perfusion imaging reduced total admissions and altered resource utilization. This may result in more appropriate triage of individual patients in the management algorithm, as well as potential cost savings.     

Clinical trial of 111In-antimyosin antibody imaging: (2). Imaging of myocardial infarction and myocarditis

A new scintigraphic method to detect myocardial necrosis has been developed using antimyosin monoclonal antibody F ab labeled with indium-111 (111In-antimyosin). We investigated 111In-antimyosin scintigraphy in 35 patients with myocardial infarction, 5 patients with myocarditis and 3 patients with angina pectoris. 111In-antimyosin F ab was administered iv and antimyosin images were recorded by planar and single photon emission computed tomography (SPECT) 48-72 hrs after injection. Planar images showed discrete localization of 111In-antimyosin in 26 of 27 patients within 16 days after the onset of acute myocardial infarction in 14 of whom creatine kinase, glutamic oxaloacetic transaminase and lactic dehydrogenase had already normalized. In addition, positive scans were also obtained in 4 of 8 patients 1 to 9 months after the onset of the disease. Three patients with acute myocarditis (two of whom were biopsy-proven) had positive scans 2 and 4 weeks after the onset of the disease. Although mechanism of persistent positive anti-myosin images in the chronic stage remains to be clarified, 111In-antimyosin scintigraphy holds potential promise as a noninvasive method for the detection of myocardial injury.     

Myocardial imaging. Coxsackie myocarditis

A 3-week-old male neonate with heart failure associated with Coxsackie virus infection was imaged with Tc-99m PYP and TI-201. The abnormal imaging pattern suggested myocardial infarction. Autopsy findings indicated that the cause was myocardial necrosis secondary to an acute inflammatory process. Causes of abnormal myocardial uptake of Tc-99m PYP in pediatrics include infarction, myocarditis, cardiomyopathy, bacterial endocarditis, and trauma. Myocardial imaging cannot provide a specific cause diagnosis. Causes of myocardial infarction in pediatrics are listed in Table 1.     

Diagnosis of acute myocardial infarction in patients undergoing open heart surgery: a comparison of serial myocardial imaging with cardiac enzymes, electrocardiography, and vectorcardiography.

In 44 consecutive patients undergoing elective open heart surgery (OHS), serial electrocardiograms (ECG), vectorcardiograms (VCG), serum CPK, cardiac isoenzymes (CPKMB), and myocardial images using Tc-99m pyrophosphate were obtained, before and after the operation, for the detection of acute myocardial infarction (AMI). Twenty-nine patients developed one or more positive tests postoperatively. Two patients had positive myocardial scintiscans; both had other evidence of infarction. Conversely, the appearance of CPKMB, or new ECG and VCG changes, occurred frequently without evidence of infarction, and were not associated with the development of a positive scintiscan. The results show that false-negative results are infrequent in patients imaged early after OHS, and that cardiac surgical procedures do not cause a high incidence of false-positive scintigrams. Consequently, radionuclide imaging for AMI offers an important adjunct for excluding acute infarction following open heart surgery.     

The role of antimyosin antibodies in acute myocardial infarction

Antimyosin is an Fab fragment of a monoclonal antibody that binds with human myosin exposed in myocytes irreversibly damaged by an ischemic event. Labeled with 111In, the antibody is taken up into acutely necrotic tissue and can be imaged by planar or single photon emission computed tomography (SPECT) techniques. A large, multicenter clinical trial has demonstrated a high degree of both sensitivity for detecting infarction and specificity for excluding a recent ischemic event in patients admitted with chest pain syndrome. No allergic reactions to antibody injection have occurred, nor have there been documented significant increases in human antimouse antibody titers postinjection. Due to relatively slow blood clearance, the optimal imaging time is 24 to 48 hours post-injection. Between 13% and 21% of 24-hour scans are nondiagnostic due to persistent blood pool activity. In two thirds of these patients, 48-hour scans confirm negative tracer uptake. Moderate to intense cardiac uptake occurs in greater than 80% of scans. Faint tracer uptake, which occurs in a small minority of patients, is associated with inferoposterior infarct location and an occluded infarct vessel. Potential clinical uses include both diagnostic and prognostic areas. A negative scan in a patient with chest pain syndrome and no ECG changes rules out a recent significant ischemic event. The extent of antimyosin uptake (infarct size), measured semiquantitatively from planar scans or quantitatively from SPECT reconstructions, has been shown to correlate with future cardiac events. Relative patterns of distribution of indium-antimyosin and 201TI on simultaneous dual isotope SPECT reconstructions may identify patients with residual myocardium at further ischemic risk.     

Acute myocardial infarction imaged with 99mTc-stannous pyrophosphate and 201Tl: a clinical evaluation.

Twenty-six patients suspected of having acute myocardial infarction (AMI) underwent myocardial scintigraphy sequentially with 201Tl and 99mTc-stannous pyrophosphate (99mTc-PPi). Of the 26 patients, 24 had AMI documented by enzyme and electrocardiographic changes. Nineteen had transmural and five had subendocardial myocardial infarctions. The remaining two patients had "unstable angina pectoris." The mean time from onset to imaging was 4 days. Of the 24 patients with AMI, 22 had positive 99mTc-PPi scintigrams. In 20 the area of acute myocardial damage appeared to be the same by 99mTc-PPi scintigram as by ECG; in two, the location could not be precisely determined. The two patients with negative 99mTc-PPi scintigrams at the time of combined myocardial imaging had had positive 99mTc-PPi images previously. In all 24 patients, the 201Tl images were abnormal in at least the location suggested by the electrocardiogram. In seven patients, the area of decreased 201Tl activity was grossly equal to the positive area on the 99mTc-PPi images; in 15, the 201Tl defect was definitely larger; and in two, the 201Tl defect appeared slightly smaller. Although the 99mTc-PPi and 201Tl myocardial images provide different information, both are valuable in determining the overall integrity of the myocardium in patients with ischemic heart disease.