Emergency department perfusion imaging for suspected coronary artery disease: The ERASE Chest Pain Trial
James E. Udelson, M.D., and Ethan J. Spiegler, M.D.

Dr. Udelson is from the division of cardiology, New England Medical Center Hospitals, and Tufts University School of Medicine, Boston, Massachusetts, and Dr. Spiegler is from St. Agnes Health Care System, Baltimore, Maryland.

The role of perfusion imaging in the emergency department

Noninvasive imaging of myocardial perfusion in the emergency department (ED) setting is conceptually attractive because the primary pathophysiologic abnormality in patients with acute coronary syndromes is an abnormality in myocardial blood flow. The feasibility of myocardial imaging in the ED setting was demonstrated over 20 years ago, using planar thallium-201 techniques.1 These studies demonstrated that it was feasible to perform perfusion imaging in this manner, and that the performance characteristics of the imaging data for identifying patients with acute myocardial infarction (MI) was acceptable. However, despite these promising early reports, the use of thallium-201 in the ED setting was impractical. The "redistribution" characteristics of thallium-201 mandate that the imaging of the initial distribution take place relatively quickly; thus, a portable camera system is needed. These systems are not widely available, and such an approach does not lend itself to contemporary single photon emission computed tomography (SPECT) imaging. Moreover, thallium-201 is generator produced and thus it is not always readily available for acute use in the ED setting.

These hurdles were overcome in the 1990s with the introduction of Tc-99m - based agents, such as sestamibi and tetrofosmin, with their relative lack of redistribution.2 Thus, even if images are acquired 45 to 60 minutes after the time of injection, the perfusion pattern will reflect myocardial blood flow at the time of injection rather than at the time of imaging. This characteristic makes these agents more ideal for perfusion imaging of patients in the ED setting, as injection can be done at rest in the ED, and the patient then transported to the nuclear medicine department for high-quality SPECT imaging. This protocol is analogous to that often employed in the ED setting for patients with suspected pulmonary embolus. These patients are often sent directly for a ventilation/perfusion scan, then returned to the ED while the results are incorporated into the information base available to the evaluating ED physician, who makes a subsequent triage decision to admit or discharge.

Relation between imaging results and outcomes

There is now a substantial body of literature evaluating the performance of rest SPECT perfusion imaging in the ED setting. In one of the first studies, Bilodeau and colleagues3 performed rest SPECT sestamibi imaging in patients who were already hospitalized for suspected unstable angina, and injected them with the tracer at the time of an episode of spontaneous chest pain while also obtaining an electrocardiogram (ECG). The sensitivity of the SPECT sestamibi images for determining the presence of a severe coronary stenosis on subsequent angiography was 96%, while the sensitivity of the ECG was only 35%. In the setting of acute MI, Christian et al4 described a series of patients with no ST elevations who were nonetheless having an acute MI with coronary occlusion. Despite the absence of obvious ECG abnormalities, these patients had perfusion defects on acute rest sestamibi studies, which quantitatively involved an average of 20% of the left ventricular myocardium. Thus, in these early reports, acute perfusion imaging detected abnormalities in patients with acute coronary syndromes more often than the ECG.

Varetto et al5 were the first to report on the performance of acute perfusion imaging brought directly into the ED setting. In 64 ED patients with suspected acute ischemia, the SPECT sestamibi images provided important discriminatory information regarding the presence or absence of an acute MI or unstable angina. This study represented an important step forward, as it was the first to report follow-up of patients who had a normal resting SPECT sestamibi study in the ED. Among that group of patients, none ruled in for an MI, none had coronary artery disease (CAD) by angiography or stress testing, and over the 18-month follow-up, none had untoward cardiac events. Thus the negative predictive value for ruling out an acute coronary syndrome or long-term follow-up cardiac events was extremely high. This study laid the foundation for establishing both a diagnostic and a prognostic use for perfusion imaging.

Since that publication, several other studies have followed involving larger numbers of patients with essentially similar results.6-8 When one examines the published observational series regarding the relationship between SPECT perfusion imaging results in the ED setting and patient outcomes, the negative predictive value for ruling out an MI in the ED setting is greater than 99%, and the negative predictive value for ruling out an MI or any follow-up cardiac event is greater than 97% (Table 1).

Comparison with enzymatic markers of acute coronary syndromes

There are several distinctions between the information provided by SPECT myocardial perfusion imaging in this setting and that provided by serial cardiac enzyme analysis. The various cardiac enzymes that are generally evaluated mark the presence of myocardial necrosis and will be positive in most patients with acute MI. Recent data also suggest that among patients with what is traditionally regarded as an unstable anginal syndrome, approximately 30% will have positive troponin markers.9 This subgroup of patients with unstable angina is at higher risk for adverse cardiac events over time compared with patients with an unstable angina syndrome but normal troponin results.9,10 Moreover, the subset of unstable angina patients with elevated troponin markers appears to be the group of patients who gain the most benefit from aggressive anti-platelet therapy with platelet GP IIbIIIa inhibitors.10,11 However, most patients with an unstable angina syndrome will have enzyme marker levels within the normal range. In contrast, myocardial perfusion imaging should theoretically be abnormal in any patient with an abnormality in myocardial blood flow, including patients with both an acute MI and unstable angina. The high sensitivity and negative predictive value of the published studies support the concept that perfusion imaging can provide powerful discriminatory value in this setting.

Another important distinction between perfusion imaging and enzyme results is the significant difference in the time course during which results are abnormal in patients with acute coronary syndromes. Perfusion imaging data should be abnormal almost immediately after an abnormality in myocardial blood flow is established. Cardiac enzyme markers, and particularly the troponins, may begin to show abnormal results 4 to 8 hours after symptom onset with the peak abnormality being evident at 12 to 18 hours after symptom onset. In a multicenter study of cardiac enzyme markers, optimal sensitivity for troponins T or I to detect acute MI occurred 18 hours after symptom onset.12 Similarly, the most powerful prognostic value for adverse cardiac events using troponin assays also occurs at approximately 18 hours after symptom onset.13

Consistent with these data is a study from Kontos and colleagues14 of a large group of patients seen in the Chest Pain ED at the Medical College of Virginia. In this study, SPECT sestamibi imaging performed in the ED was 92% sensitive for detecting acute MI, while cardiac troponin I values drawn at the same time had a sensitivity of only 30%. Subsequently, the maximum troponin I over the first 24 hours had sensitivity similar to that of the acute rest sestamibi imaging, but at a distinctly later time point. Thus, acute perfusion imaging has the potential to identify acute coronary syndromes much earlier in their evolution than enzyme markers. This could potentially translate into earlier application of aggressive reperfusion or antiplatelet therapies in the appropriate setting.

Cost-effectiveness

While the incorporation of myocardial perfusion imaging into an ED strategy engenders an added cost, the potential for reduction of inappropriate hospital or observation unit admissions might more than offset the additional costs of imaging. Weissman et al15 found that over 50% of the physicians' decisions were affected by the perfusion imaging results and estimated a potential cost savings of approximately $900 per patient. Similar results were calculated in a decision analytic model by Radensky and colleagues.16

In a report of several thousand patients evaluated in a chest pain center incorporating myocardial perfusion imaging, Ziffer et al17 reported that following incorporation of perfusion imaging into the algorithm, the missed MI rate (that is, the proportion of patients sent directly home from the ED but who were actually having an MI) dropped from 1.8% to 0.1%. Unnecessary hospital admissions were also reduced. They estimated savings (charges) of approximately $1,900 per patient.

Unanswered questions

Thus, the observational data on the relation between perfusion imaging results in the ED setting and outcomes were strong, and several studies suggested that perfusion imaging in this setting may be cost-effective. However, important unanswered questions remained in place despite this robust literature. In none of these studies were the perfusion imaging results allowed to actually affect clinicians' decision making in the ED. Thus, it was unclear whether ED or other physicians evaluating such patients could confidently incorporate the imaging data into their own decision making and actually reduce unnecessary admissions in practice. Moreover, the cost-effectiveness data were uncontrolled and often based on modeling or assumptions regarding decisions that might be made given the perfusion data.

A review of the various technologies available to aid ED physicians in this setting to help optimize the detection of acute ischemia by the National Heart Attack Alert Program Technology Assessment Working Group18 found that very few diagnostic modalities in this setting have been subject to rigorous evaluation in a randomized prospective clinical trial format. Thus, there was a need for a more rigorous evaluation of perfusion imaging in this setting before such policy and guideline-setting groups could recommend widespread application.

The ERASE Chest Pain Trial

To address the issue, a prospective multicenter randomized controlled clinical trial of perfusion imaging in patients with suspected acute ischemia in the ED setting was performed.The Emergency Room Assessment of Sestamibi for Evaluating (ERASE) Chest Pain Trial took place at a diverse group of seven hospital EDs.19 This study was designed specifically to reflect real-life practice, that is, as an effectiveness trial. Over 2,500 patients with symptoms suggestive of acute cardiac ischemia to the evaluating ED physician but with a nondiagnostic ECG for acute ischemia or infarction were randomized to one of two evaluation strategies. The control strategy was the usual ED strategy for evaluating such patients, which generally included enzyme analysis. The second strategy was the usual ED strategy supplemented by information from acute rest SPECT sestamibi imaging. Patients randomized to the scan strategy were injected with sestamibi in the ED and then transported to the nuclear medicine department for SPECT imaging. The results were called immediately to the ED physician, who incorporated the imaging information into his or her triage decision. The primary end point of the trial was the appropriateness of the decision to hospitalize or discharge directly from the ED. All patients, whether admitted or discharged directly from the ED, were followed up with subsequent enzyme analysis and stress testing, and all information (including catheterization when performed) was used to partition the study subjects into those with an acute cardiac ischemic syndrome (positive enzymes, positive stress test, or catheterization evidence of significant coronary disease) or those without acute coronary ischemia (all negative follow-up studies with negative stress testing and no events at 30 days).

Among patients with an acute ischemic syndrome, both the scan and no-scan randomization groups had a very high and appropriate admission rate to the hospital from the ED. However, among patients ultimately found to not have an acute ischemic syndrome, those initially randomized to the scan strategy had a highly significant reduction in unnecessary hospitalizations. There was a 20% reduction in the relative risk of being hospitalized among those randomized to the scan strategy who were ultimately found to be free of acute cardiac ischemia (P <0.001). this reduction in the unnecessary hospitalization rate was seen in all age groups, in the presence or absence of risk factors, in men and in women, in hospitals with high- or low-volume eds, and in hospitals with or without previous experience with imaging in this setting. in a multivariate analysis, the imaging data were among the most powerful factors associated with the decision to appropriately discharge the patient from the ed. thus the data were robust and potentially generalizable to a much wider setting.

setting up an ed imaging service

Setting up an ED perfusion imaging protocol requires significant contribution and cooperation from many stakeholders, including ED physicians and support personnel, cardiologists who are often called upon to assist in the decision making, and nuclear medicine and/or nuclear cardiology physicians who must perform and interpret the imaging with confidence and quality. While a reduction in unnecessary hospitalization with potential associated cost savings is a worthy goal and outcome, in some provider settings there are paradoxical incentives, which may work against such a program. For instance, short observation or telemetry unit admissions to rule out MI can in some settings be profitable for hospitals based on the local payment system. Thus, sending such patients home directly from the ED may be an important source of lost income for an institution. Adoption of this proven technology into such a setting will often require realigning incentives among payers and providers to allow incorporation of such a potential cost-effective algorithm into the ED setting. When used properly and with expertise, a reduction in unnecessary hospitalization is obviously favorable for patients and for the health care system.

It is also important to define the ideal patients eligible for an imaging protocol. Patients who report to the ED with symptoms consistent with an acute coronary syndrome and an ECG diagnostic for acute ischemia or infarction do not benefit from acute imaging. Such patients are triaged based on the presence of ST elevation or ST depression into appropriate algorithms for acute MI or unstable angina/non " ST- ST-segment elevation infarction, respectively.20 Ideal patients for imaging are those with nondiagnostic or normal electrocardiograms and symptoms suspicious for acute ischemia. Ideal patients also have no history of infarction or significant Q waves on the ECG. Such patients will often have a perfusion defect representative of the old infarction, and thus the data will not be as helpful for discrimination of a new acute ischemic syndrome.

Future directions

While the data from the ERASE Chest Pain Trial provide compelling evidence for incorporation of acute SPECT myocardial perfusion imaging into ED evaluation strategies for patients with suspected acute ischemia, future studies will refine the patient population likely to benefit most from such a procedure. There are likely certain clinical characteristics of such patients, which may on a larger scale be identified by ECG predictive instruments for example, in which the pretest probability of acute ischemia is so low that imaging is not beneficial. The optimal combination of imaging data, enzyme data and stress testing, and the temporal distribution of such testing also needs to be better defined.

However, given the difficulty that many physicians have evaluating such patients with confidence, reflected by the very high rate of admission for observation or evaluation from the ED, the incorporation of perfusion imaging has the potential to reduce unnecessary hospitalizations significantly, with potential associated cost savings.

References

1. Wackers FJ, Lie KI, Liem KL, et al. Potential value of thallium-201 scintigraphy as a means of selecting patients for the coronary care unit. Br Heart J 1979;41:111-1117.
2. Beller GA. Radiopharmaceuticals in nuclear cardiology. In: Beller GA, ed. Clinical Nuclear Cardiology. Philadelphia: Saunders; 1995:37-81.
3. Bilodeau L, Theroux P, Gregoire J, et al. Technetium-99m sestamibi tomography in patients with spontaneous chest pain: correlations with clinical, electrocardiographic and angiographic findings. J Am Coll Cardiol 1991;18:1684-1691.
4. Christian TF, Clements IP, Gibbons RJ. Noninvasive identification of myocardium at risk in patients with acute myocardial infarction and nondiagnostic electrocardiograms with technetium-99m-sestamibi. Circulation 1991;83:1615-1620.
5. Varetto T, Cantalupi D, Altieri A, et al. Emergency room technetium-99m sestamibi imaging to rule out acute myocardial ischemic events in patients with nondiagnostic electrocardiography. J Am Coll Cardiol 1993;22:1804-1808.
6. Hilton TC, Thompson RC, Williams HJ, et al. Technetium-99m sestamibi myocardial perfusion imaging in the emergency room evaluation of chest pain. J Am Coll Cardiol 1994;23:1016-1022.
7. Tatum JL, Jesse Rl, Kontos MC, et al. Comprehensive strategy for the evaluation and triage of the chest pain patient. Ann Emerg Med 1997;29:116-125.
8. Heller GV, Stowers SA, Hendel RC, et al. Clinical value of acute rest technetium-99m tetrofosmin tomographic myocardial perfusion imaging in patients with acute chest pain and nondiagnostic electrocardiograms. J Am Coll Cardiol 1998;31:1011-1017.
9. Antman EM, Tanasijevic MJ, Thompson B, et al. Cardiac-specific troponin I levels to predict the risk of mortality in patients with acute coronary syndromes. N Engl J Med 1996;335:1342-1349.
10. . Hamm CW, Heeschen C, Goldmann B, et al. Benefit of abciximab in patients with refractory unstable angina in relation to serum troponin T levels. c7E3 Fab Antiplatelet Therapy in Unstable Refractory Angina (CAPTURE) Study Investigators. N Engl J Med 1999;340:1623-1629.
11. Heeschen C, Hamm CW, Goldmann B, et al. Troponin concentrations for stratification of patients with acute coronary syndromes in relation to therapeutic efficacy of tirofiban. PRISM Study Investigators. Platelet Receptor Inhibition in Ischemic Syndrome Management. Lancet 1999;354:1757-1762.
12. Zimmerman J, Fromm R, Meyer D, et al. Diagnostic marker cooperative study for the diagnosis of myocardial infarction. Circulation 1999;99:1671-1677.
13. Newby LK, Christenson RH, Ohman EM, et al. Value of serial troponin T measures for early and late risk stratification in patients with acute coronary syndromes. The GUSTO-IIA Investigators. Circulation 1998;98:1853-1859.
14. Kontos MC, Jesse RL, Anderson FP, et al. Comparison of myocardial perfusion imaging and cardiac troponin I in patients admitted to the emergency department with chest pain. Circulation 1999;99:2073-2078.
15. Weissman IA, Dickinson CZ, Dworkin HJ, et al. Cost-effectiveness of myocardial perfusion imaging with SPECT in the emergency department evaluation of patients with unexplained chest pain. Radiology 1996;199:353-357.
16. Radensky PW, Hilton TC, Fulmer H, et al. Potential cost effectiveness of initial myocardial perfusion imaging for assessment of emergency department patients with chest pain. Am J Cardiol 1997;79:595-599.
17. Ziffer J, Nateman DR, Janowitz WR, et al. Improved patient outcomes and cost-effectiveness of utilizing nuclear cardiology protocols in an emergency department chest pain center. J Nucl Med 1998;39:139P
18. Selker HP, Zalenski RJ, Antman EM, et al. An evaluation of technologies for detecting acute cardiac ischemia in the emergency department: a report from a National Heart Attack Alert Program Working Group. Ann Emerg Med 1997;29:1-87.
19. Udelson JE. The ERASE Chest Pain Trial. Presented at "Special Session: Clinical Trials" at the 72nd Scientific Sessions of the American Heart Association. Atlanta, Ga. November 1999.
20. Braunwald E, Antman EM, Beasley JW, et al. ACC/AHA guidelines for the management of patients with unstable angina and non-ST-segment elevation myocardial infarction. A report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (Committee on the Management of Patients With Unstable Angina). J Am Coll Cardiol 2000;36:970-1062.