Cardiac Imaging for Risk Stratification With Dobutamine-Atropine Stress Testing in Patients With Chest Pain
Echocardiography, Perfusion Scintigraphy, or Both?
Background Pharmacological stress echocardiography and myocardial perfusion scintigraphy are used frequently for risk stratification in patients with suspected myocardial ischemia. However, their relative prognostic strength has never been explored.
Methods and Results Two hundred twenty consecutive patients with chest pain (mean age, 60±12 years; 124 men, 115 with previous myocardial infarction) were studied with dobutamine-atropine stress echocardiography (ECHO) and simultaneous 99mTc sestamibi single photon emission computed tomography imaging (MIBI). Ischemia was defined as deterioration in left ventricular wall motion and reversible perfusion defects, respectively. ECHO was positive for ischemia in 76 and MIBI in 91 patients (agreement, 77%; κ=.51). During follow-up of 31±15 months, 24 patients had hard cardiac events (nonfatal myocardial infarction or cardiac death). By univariate analysis, age, history of congestive heart failure, and any abnormality or ischemia on ECHO or MIBI were associated with cardiac events. Multivariate analysis revealed that age, abnormal ECHO (odds ratio [OR], 18.9; 95% CI, 2.5 to 146.0) or MIBI (OR, 12.8; 95% CI, 1.7 to 98.3), and ischemia on ECHO (OR, 4.0; 95% CI, 1.6 to 9.9) or MIBI (OR, 3.0; 95% CI, 1.2 to 7.4) had independent predictive values. When ECHO was used as a first option, the addition of MIBI to all nonischemic ECHO studies decreased the OR from 4.0 (95% CI, 1.6 to 9.9) to 3.8 (95% CI, 1.4 to 10.2). Addition of MIBI confined to nonischemic ECHO studies in which target heart rate was not attained (nondiagnostic studies) increased the OR to a maximal 5.7 (95% CI, 2.2 to 15.0). In contrast, the addition of ECHO to nondiagnostic MIBI studies was not useful.
Conclusions Dobutamine-atropine ECHO and MIBI provide comparable prognostic information. The addition of MIBI to ECHO may be useful in patients with nondiagnostic ECHO studies.
Risk stratification of patients with known or suspected coronary artery disease is an important goal of clinical cardiology.1 In daily clinical practice, several modalities of stress testing are used for this purpose, often in conjunction with cardiac imaging.2 3 4 5 6 7 8 9 10 Among imaging modalities, myocardial perfusion scintigraphy is currently the most widely used noninvasive technique for the functional and prognostic assessment of patients with known or suspected coronary artery disease,7 8 9 10 11 in particular in combination with exercise stress. However, since up to 40% of the patients referred for noninvasive evaluation of coronary artery disease are unable to perform adequate exercise,9 there is a consensus that pharmacological stress tests are first choice in a substantial number of patients. Echocardiography is well suited to be combined with pharmacological stress,12 and there is a growing body of evidence to indicate that pharmacological stress echocardiography is as feasible and efficient as perfusion scintigraphy for diagnostic purposes.13 14 However, it is unknown which imaging modality should be preferred for risk stratification. Therefore, the present study was undertaken to assess the relative prognostic value of dobutamine-atropine ECHO, MIBI, and their combination in 220 consecutive patients with chest pain and inability to perform adequate exercise.
Over a 3-year period, between November 1990 and October 1993, 260 consecutive patients with chest pain underwent a simultaneous dobutamine-atropine stress ECHO and MIBI study. All patients were referred for pharmacological stress testing because of inadequate exercise capacity, either proven by previous nondiagnostic exercise testing or judged by the referring physician. None of these patients had prior heart transplantation, congenital or significant valvular heart disease, known primary dilated cardiomyopathy, or unstable angina. Forty patients were excluded from final analysis: 14 with technically inadequate echocardiograms at rest (11) or peak stress (3), 2 with poor-quality scintigraphic images, 8 with multiple tests (in these patients only the first test was considered), and 16 with early elective coronary revascularization within 60 days after stress testing. None of the latter patients sustained MI before coronary revascularization. The mean age of the remaining 220 patients was 60±12 years (range, 23 to 85); 124 were men (56%). One hundred fifteen patients (52%) had a previous MI and 26 patients (12%) had known coronary artery disease without MI. Fifty-two patients (24%) had typical angina, 117 (53%) had atypical angina, and 51 (23%) had nonanginal chest pain. At the time of the study, 156 patients (71%) were receiving antianginal therapy, including β-blockers in 92 (42%), either administered alone in 28 (13%) or in combination with nitrates and/or calcium channel blockers in 64 (29%).
Dobutamine-Atropine Stress Test
After routine preparation, a resting ECG and echocardiogram were made, intravenous access was secured, and dobutamine was administered intravenously by an infusion pump. The initial infusion rate was 10 μg/kg per minute for 3 minutes, increasing by 10 μg/kg per minute every 3 minutes up to a maximum of 40 μg/kg per minute. In patients not achieving 85% of their age- and sex-predicted maximal heart rate and without symptoms or signs of myocardial ischemia, atropine was administered in addition to the maximal dose of dobutamine, starting with 0.25 mg intravenously and repeated up to a maximum of 1.0 mg within 4 minutes with continuation of the dobutamine infusion.15 Throughout dobutamine infusion the ECG (3 leads) was continuously monitored, and a 12-lead ECG was recorded at 1-minute intervals. The level of ST-segment shift was calculated, after signal averaging, by a computer-assisted system (Cardiovit CSG/12; Schiller). Blood pressure was measured and recorded by sphygmomanometry or automatic device every 3 minutes. Reasons for interruption of the test were horizontal or downsloping ST-segment depression of >0.2 mV at an interval of 80 ms after the J-point compared with baseline, ST-segment elevation of >0.1 mV in patients without previous MI, severe angina, a symptomatic reduction in systolic blood pressure >40 mm Hg from baseline, hypertension (blood pressure >240/120 mm Hg), significant cardiac tachyarrhythmias, and any serious side effect that was regarded to be a result of dobutamine. Metoprolol was available and used to reverse the effects of dobutamine if they did not revert spontaneously and quickly.
Echocardiographic and Perfusion Scintigraphic Imaging
Echocardiographic analysis of the left ventricular wall was performed according to a 16-segment model.16 Both systolic wall thickening and inward endocardial motion were visually evaluated, and each segment was graded on a 5-point scoring system (1, normal; 2, mild hypokinesis; 3, severe hypokinesis; 4, akinesis; and 5, dyskinesis). For perfusion imaging, 370 mBq of MIBI was injected intravenously at peak stress while dobutamine infusion was continued for 1 minute. Stress scintigraphic images were acquired, on average, 1 hour after termination of the dobutamine infusion. For resting studies, 370 mBq of MIBI was injected at least 24 hours after the stress study. Image acquisition was performed with a Siemens Gammasonics single-head Rota Camera. For each study, 6 short-axis and 3 sagittal long-axis slices were analyzed. To compare the stress and rest studies, each of the 6 short-axis slices was divided into 8 equal segments. The septal part of the 2 basal slices (4 segments) was not evaluated because this region corresponds to the fibrous portion of the interventricular septum and normally exhibits reduced uptake. The apical region was assessed from the 3 sagittal cross sections. A total of 47 segments per patient were analyzed. All tomographic views were reviewed in side-by-side pairs (stress and rest) and the myocardial uptake of radiotracer was evaluated visually, with the assistance of circumferential profile analysis including the normal values, also with the use of a 5-point scoring method (1, normal; 2, minimally reduced uptake; 3, moderately reduced uptake; 4, severely reduced uptake; and 5, absence of uptake).
As depicted in Fig 1⇓, the echocardiographic and scintigraphic images were subsequently matched by regrouping the 16 echocardiographic and the 47 scintigraphic segments in 6 major myocardial regions (anterior, anterior septum, posterior septum, inferoposterior, lateral, and apical).17 A region was classified as infarcted in the case of a resting score >2 in one or more segments on ECHO18 and >2 in two or more adjacent segments on MIBI. A region was classified as ischemic on ECHO in the case of an increase in score between rest and stress in one or more segments unless an akinetic segment showed no improvement during low-dose dobutamine and became dyskinetic during high-dose dobutamine.19 On MIBI, ischemia was defined as a perfusion defect during stress that partially or totally resolved at rest in at least two contiguous segments or slices. An ECHO or MIBI study was classified as abnormal in the presence of either infarction or ischemia. All studies were reviewed by two experienced observers (A.S. and R.R. for ECHO; J.H.C. and A.R. for MIBI) unaware of all other stress test results. In cases of disagreement, a third reviewer (P.M.F.) decided on the grading of each study. Pattern (normal, ischemia alone, infarction alone, or both infarction and ischemia [mixed]) interobserver agreement, as assessed in 200 patients, was 89% for ECHO and 92% for MIBI. Interobserver agreement on ischemia was 92% for ECHO and 95% for MIBI.
Follow-up data were obtained over a 31±15-month period (range, 12 to 48 months) by outpatient clinic assessment, review of case notes, and contact with the patient, general practitioner, or other hospitals when necessary. Outcome events were cardiac death, nonfatal MI, and revascularization (coronary bypass surgery or percutaneous transluminal coronary angioplasty). Cardiac death was defined as a death temporally associated with a known or suspected acute MI, life-threatening arrhythmia, or pulmonary edema. Unexpected death without an identified noncardiac cause also was considered to be cardiac death. Occurrence of an acute MI was confirmed with the use of standard clinical and ECG criteria and when total creatine kinase enzyme levels exceeded twice the normal values. Hard cardiac events (cardiac death and nonfatal MI) and all cardiac events (hard events or revascularization) were separately analyzed as end points. Patients with revascularization procedures were censored at the time of intervention, so only the first event in each patient was considered.
Values were expressed as mean±SD unless specified. Comparison of patients with and without cardiac events was performed with the Student’s t test for continuous variables and χ2 test for discrete variables. Differences of P<.05 were considered significant. Variables correlated with cardiac events at a significance level of P<.10 by univariate analysis and selected variables with P=.10 to .20 were further analyzed with the use of stepwise logistic regression. ORs and 95% CIs were calculated for variables used in the multivariate model. Kaplan-Meier life table estimates of infarction-free survival (survival without cardiac death or nonfatal MI) and event-free survival (survival without cardiac death, nonfatal MI, or revascularizations) were used to summarize the follow-up results. Comparison of life tables was performed with the use of the log-rank test.
Dobutamine-Atropine Stress Test
Hemodynamic Results, End Points, and Side Effects
The maximal dobutamine dose used was 10 μg/kg per minute in 1 patient, 20 μg/kg per minute in 4 patients, 30 μg/kg per minute in 30 patients, and 40 μg/kg per minute in 185 patients. Atropine was added in 83 patients (38%) and was more often used in patients on β-blockers (58 of 92 on versus 25 of 128 off β-blockers, P<.0001). Heart rate increased by 63±15 bpm up to 134±17 bpm at peak stress, systolic blood pressure by 13±27 mm Hg up to 151±31 mm Hg, and the double product (heart rate times systolic blood pressure) by 10 382±3809 up to 20 108±4825 beats×mm Hg per minute.
Target heart rate (85% of maximum for age and sex) was not reached in 46 tests (21%) either after maximal-dose dobutamine-atropine had been given in 18 patients (8%) or premature stopping of the test in 28 patients (13%). The test was prematurely stopped because of angina in 19 patients, ST-segment changes in 2, hypertension in 1, symptomatic hypotension in 4, nonsustained ventricular tachycardia in 1, and anxiety in 1. Most patients not reaching their target heart rate despite maximal-dose dobutamine-atropine were on β-blockers (15 of 18 on versus 3 of 18 off β-blockers, P<.001). Side effects usually were minor, and significant tachyarrhythmias (paroxysmal supraventricular or nonsustained ventricular) were encountered during dobutamine infusion or recovery in 14 patients (6%).
Distribution of Imaging Patterns
As shown in Fig 2⇓, the distributions of ECHO and MIBI patterns were, respectively, normal in 86 (39%) and 70 (32%), infarction alone in 58 (26%) and 59 (27%), ischemia alone in 31 (14%) and 32 (15%), and both ischemia and infarction in 45 (20%) and 59 (27%). Therefore, abnormal patterns were present in 134 (61%) and 150 (68%) patients, infarction patterns in 103 (47%) and 118 (54%), and ischemic patterns in 76 (35%) and 91 (41%), respectively. Pattern agreement (normal, ischemia alone, infarction alone, and mixed patterns) between ECHO and MIBI was 66% (κ=.53). The agreement for ischemia was 77% (κ=.51).
Twelve patients (5%) had “incomplete” follow-up, 7 because of noncardiac death (cancer in 3, pneumonia in 2, AIDS in 1, and myelodysplasia in 1) and 5 because of geographic relocation. During follow-up, 11 patients had a nonfatal MI and 13 died as the result of a cardiac event. Thirty patients underwent a late or nonelective revascularization procedure. In total, 24 patients had a “hard” cardiac event (cardiac death or nonfatal MI) and 54 patients had “any” cardiac event (cardiac death, nonfatal MI, or revascularization).
Prediction of Events From Clinical and Stress Test Results
The clinical and the stress test variables in patients with and without hard cardiac events are summarized in Tables 1⇓ and 3⇓. Clinical variables associated with hard cardiac events were age and history of CHF. Of the stress test variables, peak systolic blood pressure and double product were lower in patients with subsequent events. However, peak heart rate, angina, or ST-segment changes during the test were not associated with increased rate of hard cardiac events. Imaging patterns associated with hard cardiac events were the presence of any abnormality on ECHO or MIBI, infarction on MIBI, and ischemia on ECHO or MIBI. Associated with all cardiac events were the clinical variables of male sex, a history of typical angina, MI, or revascularization, and all abnormal imaging patterns on ECHO or MIBI (Tables 2⇓ and 4⇓).
Infarction-Free Survival Curves
The infarction-free survival curves in patients with the different ECHO and MIBI patterns are depicted in Fig 3A⇓ and 3B⇓. A normal study was associated with a favorable prognosis over the follow-up period, with a negligible annual cardiac event rate of 0.4% by ECHO and 0.5% by MIBI. In contrast, patients with infarction alone, ischemia alone, and mixed patterns had significantly increased cardiac event rates of, for ECHO versus MIBI, respectively, 7.2% (P<.0005) versus 6.4% (P<.005), 10.8% (P<.0001) versus 6.1% (P<.005), and 7.2% (P<.0005) versus 7.5% (P<.005). For event-free survival (Fig 4A⇓ and 4B⇓), annual event rates were, for the different patterns on ECHO versus MIBI, respectively, 1.8% versus 1.8%, 14.3% (P<.0001) versus 11.3% (P<.005), 10.8% (P<.0001) versus 13.3% (P<.0005), and 26.7% (P<.0001) versus 18.6% (P<.0001).
Annual Hard Event Rates According to Extent of Abnormalities
As seen in Fig 5A⇓, patients with normal segments only had annual event rates of 0.4% (n=86) for ECHO and 0.5% (n=70) for MIBI. For patients with one or two and more than two abnormal segments, annual event rates were, for ECHO and MIBI, respectively, 4.9% (n=74) versus 5.7% (n=98) and 13.6% (n=60) versus 9.6% (n=52). As seen in Fig 5B⇓, patients without infarcted segments had annual event rates of 2.6% (n=117) for ECHO and 2.0% (n=102) for MIBI. For patients with one and more than one infarcted segment, annual event rates were, for ECHO and MIBI, respectively, 2.5% (n=35) versus 5.6% (n=56) and 10.4% (n=68) versus 8.9% (n=62). As seen in Fig 5C⇓, patients without ischemic segments had annual event rates of 2.3% (n=144) for ECHO and 2.6% (n=129) for MIBI. For patients with one and more than one ischemic segment, annual event rates were, for ECHO and MIBI, respectively, 6.0% (n=35) versus 5.6% (n=51) and 12.3% (n=41) versus 9.3% (n=40).
Annual Event Rates According to Combination of ECHO and MIBI Results
As seen in Fig 6A⇓, patients with both a negative ECHO study and a negative MIBI study had annual hard event rates of 0% (n=60) if any abnormality was considered and 2.0% (n=111) if ischemia was considered. For patients with a negative ECHO study and a positive MIBI study, these numbers were 1.3% (n=10) and 3.7% (n=18); for patients with a positive ECHO study and a negative MIBI study, these numbers were 3.0% (n=26) and 8.2% (n=33); and for patients with both a positive ECHO study and a positive and MIBI study, these numbers were 9.2% (n=124) and 9.9% (n=58), respectively. Similarly, for all cardiac events, the respective numbers were 0.5% and 4.5%, 5.5% and 10.9%, 16.5% and 14.3%, and 23.4% and 31.8% (Fig 6B⇓).
Multivariate Analysis: Addition of ECHO or MIBI to Clinical Data
Table 5⇓ summarizes the results of univariate and multivariate (stepwise logistic regression) analyses of clinical data and stress test imaging results as predictors of subsequent cardiac events. The results of the addition of ECHO or MIBI to clinical data were analyzed separately (clinical data + ECHO and clinical data + MIBI). Furthermore, each analysis was performed twice; in the first model (Table 5⇓, model I) the only imaging pattern variable entered was the presence of an abnormal pattern (any abnormality), and in the second model (Table 5⇓, model II) the presence of an ischemic or infarcted pattern was separately included. As shown in Table 5⇓, age and an abnormal pattern on ECHO (OR, 18.9; 95% CI, 2.5 to 146.0) or MIBI (OR, 12.8; 95% CI, 1.7 to 98.3) in model I and an ischemic pattern on ECHO (OR, 4.0; 95% CI, 1.6 to 9.9) or MIBI (OR, 3.0; 95% CI, 1.2 to 7.4) in model II were independent predictors of subsequent hard cardiac events. Infarcted patterns on ECHO or MIBI were not independent predictors for hard cardiac events. However, when the presence of an infarcted pattern in 3 or more segments was forced into multivariate model II, independent predictors were age, an infarcted pattern on ECHO (OR, 6.2; 95% CI, 2.3 to 16.7) or MIBI (OR, 3.9; 95% CI, 1.4 to 10.7), and an ischemic pattern on ECHO (OR, 4.8; 95% CI, 1.8 to 12.5) or MIBI (OR, 3.1; 95% CI, 1.2 to 7.8). For all cardiac events (Table 6⇓), age, typical angina, a history of MI or revascularization, and an abnormal pattern on ECHO (OR, 8.9; 95% CI, 3.3 to 23.8) or MIBI (OR, 8.8; 95% CI, 2.9 to 26.6) in model I and an ischemic pattern on ECHO (OR, 4.0; 95% CI, 2.0 to 7.9) or MIBI (OR, 3.9; 95% CI, 1.9 to 7.8) in model II were independent predictors of events.
Multivariate Analysis: Different Additions of MIBI to Clinical and ECHO Data
Since MIBI is thought to be slightly more sensitive (but less specific) for the detection of ischemia20 21 and more informative in submaximal stress22 than ECHO, we tried to assess if, to what extent, and in which patients MIBI could provide additional prognostic information in addition to ECHO. In this multivariate regression analysis, MIBI was added to nonischemic ECHO studies according to four different strategies (Table 7⇓). In each strategy, a patient was considered to be at risk for future events when either one of the two techniques revealed ischemia.
In the first strategy (A), all 220 patients underwent only an ECHO study. On the basis of an ischemic response, 76 patients were considered to be at risk, and 15 of them had a hard cardiac event. The predictive OR for cardiac events was 4.0 (95% CI, 1.6 to 9.9) (as reported in Table 5⇑, model II).
In the second strategy (B), MIBI was added to all patients without ischemia on ECHO. This strategy required 144 MIBI scans, and, by definition, yielded a higher number of patients at risk compared with strategy A (109 versus 76 patients). Eighteen of the patients at risk had a hard cardiac event. Because of a decrease in specificity, the OR of an ischemic response for the prediction of events was lower compared with ECHO used alone (3.8; 95% CI, 1.4 to 10.2).
In the third strategy (C), MIBI was added to ECHO only in patients with a nondiagnostic ECHO study (a submaximal test, without ischemia on ECHO). This strategy required 28 MIBI studies; myocardial ischemia was detected in 85 patients, and a hard cardiac event occurred in 17 of them. This strategy resulted in an OR of 5.3 (95% CI, 2.0 to 14.0).
In the last strategy (D), the addition of MIBI was limited to the 12 patients with a nondiagnostic study in which the ECHO test was interrupted prematurely because of other potential signs or symptoms of ischemia such as angina, ST-segment changes, or ventricular tachyarrhythmias but without ischemia on ECHO (studies in which the probability of a false-negative ECHO study is highest). By this strategy, 81 patients at risk could be identified; 17 had a hard cardiac event, and, with the use of the least additional MIBI studies, the OR was improved to 5.7 (95% CI, 2.2 to 15.0). By considering all patients in this last strategy to be at risk without addition of any MIBI study, 88 patients were classified to be at risk, 17 had a hard cardiac event, and the OR was 5.1 (95% CI, 2.0 to 13.5). In a similar fashion, for the prediction of any event, the addition of MIBI was also most useful in strategy D (Table 7⇑, bottom). A similar analysis with reversed strategies (addition of ECHO to MIBI) is presented in Table 8⇓. The results of this analysis show that the addition of ECHO to nondiagnostic MIBI did not add to the information provided by MIBI alone.
In daily clinical practice, stress-induced transient perfusion defects and wall motion abnormalities are used as myocardial ischemia markers.7 8 9 10 11 12 13 14 However, their relative prognostic information in a heterogeneous population, such as that referred to a busy cardiac stress imaging laboratory, is unknown. Therefore, we initiated this study to make a head-to-head comparison between the prognostic value of these different ischemic markers from high-dose dobutamine-atropine stress testing in 220 patients unable to perform an adequate exercise test. The main findings of our study are (1) an ischemic pattern on dobutamine-atropine stress ECHO or MIBI provides comparable, independent prognostic information in addition to clinical data, (2) an increased number of ischemic segments is directly related to a worse prognosis for both ECHO and MIBI, and (3) if ECHO is selected as the imaging modality of first choice, the addition of MIBI to clinical and ECHO data can be useful but should be limited to the minority of patients with a nondiagnostic ECHO study, whereas the reverse addition of ECHO to MIBI seems less useful, certainly from a cost-effective point of view.
Stress Test Technique
In the present study, high-dose dobutamine-atropine was used as the stressing agent. High-dose dobutamine, up to 40 μg/kg per minute, eventually in combination with atropine, has been used widely for the diagnosis of coronary artery disease in conjunction with echocardiography and, although less frequently, also with perfusion scintigraphy.11 13 14 20 Dobutamine is a predominant β1-agonist that causes an increase of myocardial oxygen demand mainly resulting from increased contractility and heart rate, providing hemodynamic changes partially similar to exercise.23 In the case of significant coronary stenoses, dobutamine induces a maldistribution of flow and eventually a worsening of regional wall thickening that can be detected by perfusion SPECT imaging and echocardiography, respectively. In echocardiographic studies, the addition of atropine to dobutamine has been shown to improve its diagnostic accuracy, especially in patients receiving β-blockers.15 24
As shown in other studies,25 26 27 dobutamine-atropine stress is a safe and feasible stress method. Consistently, in the present series there were no major side effects such as sustained ventricular tachycardia, ventricular fibrillation, MI, or death. The feasibility of the test was also high, since in only 18 tests (7%) the maximal dose of dobutamine-atropine was insufficient to attain 85% of predicted maximal heart rate and there were only a few nonischemia-related, limiting side effects. Apart from 14 patients (5%) with inadequate acoustic echocardiographic windows for the assessment of all ventricular regions and 2 patients (1%) with scintigraphic images that could not be interpreted completely, only 28 of all ECHO studies (13%) and 23 of all MIBI studies (10%) were nondiagnostic.
Image Pattern Distribution
Ischemic segments were relatively more common on MIBI than on ECHO. Of note, ischemia was especially more frequently detected in patients with infarct patterns. These findings are not surprising, since it is known that according to the “ischemic cascade” theory,28 perfusion abnormalities are expected to precede the development of true ischemia, eventually resulting in wall motion abnormalities. Furthermore, in segments with resting myocardial dysfunction, the detection of ischemia on ECHO can be problematic.29 Abnormal echocardiographic images in the presence of normal perfusion images are hard to explain according to a pathophysiological mechanism. In these 10 patients, 2 had a moderately dilated left ventricle with diffuse hypokinesis but “normal” perfusion, possibly the result of cardiomyopathy. Unfortunately, coronary angiography was not available in these patients. The other 8 patients all had their echocardiographic wall motion abnormalities in basal inferoposterior segments, although in most patients the mid part of the wall was also involved. These regions of the myocardium are known for their tendency to cause false-positive results.30
This is the first study conducted as a head-to-head comparison of the prognostic information of stress echocardiography and perfusion scintigraphy in patients with known or suspected coronary artery disease and suspected myocardial ischemia. Univariate and multivariate analyses, in which clinical and stress test variables were incorporated, confirmed the prognostic value of well-known parameters such as age, a history of CHF, and any abnormality detected by stress ECHO and MIBI.9 10 Dobutamine-atropine stress–induced myocardial ischemia (whether detected by ECHO or MIBI) also carried independent prognostic information in addition to clinical data. This was not as strong as that of any perfusion or wall motion abnormalities, including fixed defects corresponding to myocardial scarring. However, the additional prognostic value of stress-induced ischemia is clinically relevant for its potential to be relieved by medical treatment or revascularization procedures.
Several reports on comparable populations have been published on the individual prognostic value of stress echocardiography and myocardial perfusion scintigraphy.4 7 10 Although different stress modalities were used, these studies reported similar figures in terms of follow-up results and predictive value of the tests. Therefore, our findings on the prognostic value of dobutamine-atropine ECHO and MIBI are not very surprising. However, this study is unique in the assessment of the relative prognostic value of stress echocardiography and myocardial perfusion scintigraphy applied simultaneously in the same population.
Because both imaging modalities seem to have similar prognostic values, the choice should be made on the basis of cost aspects, availability, and (most importantly) local experience and skill. Dobutamine-atropine perfusion scintigraphy also could be considered a useful alternative in patients with a poor acoustic window and additive to echocardiography in patients with nondiagnostic echocardiographic studies (strategies C and D, Table 7⇑), especially in patients with contraindications for vasodilator stress. Theoretically, MIBI could be injected in these patients during the same stress test, conditional to on-line interpretation of the echocardiographic images and availability of the radiotracer and the gamma camera. This strategy seems reasonable and financially convenient in a laboratory with well-balanced experience in stress echocardiography and myocardial perfusion imaging. Addition of stress echocardiography to perfusion scintigraphy, however, provided little to no additional information and requires two separate tests because ischemia at MIBI can only be assessed off-line. Therefore, such a strategy does not seem cost-effective.
Limitations of the Study
Both animal31 and clinical studies21 have suggested that dobutamine is an appropriate stress agent to demonstrate abnormal wall motion caused by ischemia. However, vasodilators (adenosine or dipyridamole) might be more suitable to create blood flow heterogeneity detected by perfusion scintigraphy. Indeed, in the same animal model31 dipyridamole caused the most blood flow heterogeneity, making it particularly suited for myocardial perfusion studies. Published clinical data, however, conflict concerning the superiority of vasodilators to dobutamine for perfusion scintigraphy. Kumar et al32 found that dipyridamole thallium scintigraphy correlated better with coronary score. However, these results were based on a small group of patients and the dobutamine dose used was very low (20 μg/kg per minute). Marwick et al13 found in a larger series of 97 patients−using high-dose dobutamine−that the accuracy of dobutamine MIBI was comparable with adenosine MIBI (accuracies of 77% and 80%, respectively). Recently, these findings were confirmed by others.33 Although none of the aforementioned studies provides evidence for a superiority of vasodilator over dobutamine perfusion imaging for prognosis, the former stress modality is more routinely used in clinical cardiology. Future studies should provide information on the relative prognostic value of vasodilator versus dobutamine perfusion imaging and eventually versus dobutamine stress echocardiography.
The decision to perform coronary arteriography and subsequent coronary artery angioplasty or bypass graft surgery is influenced frequently by individual physicians’ biases and may be affected by the presence of abnormal findings on the stress study. Therefore, we excluded patients with early elective revascularizations, and “hard” cardiac events (nonfatal MI and cardiac-related death) were analyzed separately.
The scintigrams were scored with the use of a semiquantitative method. Quantitative methods may improve diagnostic accuracy34 ; however, such an improvement is usually marginal. In an in-depth prospective analysis of patients referred for 201Tl SPECT, Mahmarian et al34 reported that visual and quantitative methods were comparably sensitive for identifying patients with single, double, and triple coronary disease; however, quantitative tomography tended to be more specific. In contrast, in a prior study from our center, we reported similar sensitivity and a trend toward better specificity for semiquantitative analysis versus quantitative analysis of dipyridamole–exercise 201Tl scintigrams.35 Finally, the interobserver agreement on ischemia for the semiquantitative analysis in the present study was excellent (95%).
In a population unable to perform adequate exercise with suspected or known coronary artery disease and stable chest pain, probably reflecting the continuous spectrum of disease in the total population, dobutamine-atropine stress is a safe and feasible stress technique. Combined with either imaging modality−echocardiography or 99mTc sestamibi myocardial perfusion scintigraphy−it provides useful prognostic information additional to clinical data. For both imaging modalities, the single most important independent predictor for future nonfatal MI or cardiac death is any abnormal pattern, while an ischemic pattern provides additional, independent prognostic information. If a stress laboratory chooses to use echocardiography as the routine pharmacological stress test, the addition of perfusion scintigraphy could be useful but should be limited to patients with nondiagnostic echocardiographic studies. If a center prefers perfusion imaging as the first choice, the addition of stress echocardiography does not seem cost-effective.
Selected Abbreviations and Acronyms
|CHF||=||congestive heart failure|
|MIBI||=||99mTc sestamibi SPECT imaging|
|SPECT||=||single-photon emission computed tomography|
Dr Geleijnse is supported by the Dutch Heart Foundation (grant NHS 94.135) and Dr Elhendy by the Department of Cardiology, Cairo University Hospital, Cairo, Egypt. The authors would like to acknowledge Joyce Postma-Tjoa, nuclear technician, for her invaluable help in maintaining the database and Jeroen Vos, MD, for statistical advice.
Presented in part at the 67th Annual Scientific Sessions of the American Heart Association, Dallas, Tex, November 1994, and the Second International Conference of Nuclear Cardiology, Cannes, France, April 1995.
See Table 7⇑ legend for explanation of strategies. ECHO and MIBI strategies are reversed in this table.
- Received October 14, 1996.
- Revision received December 17, 1996.
- Accepted January 9, 1997.
- Copyright © 1997 by American Heart Association
Severi S, Picano E, Michelassi C, Lattanzi F, Landi P, Distante A, L’Abbate A. Diagnostic and prognostic value of dipyridamole echocardiography in patients with suspected coronary artery disease: comparison with exercise electrocardiography. Circulation. 1994;89:1160-1173.
Ryan T, Armstrong WF, O’Donnel JA, Feigenbaum H. Risk stratification after acute myocardial infarction by means of exercise two-dimensional echocardiography. Circulation. 1991;83(suppl III):III-32-III-37.
Picano E, Severi S, Michelassi C, Lattanzi F, Masini M, Orsini E, Distante A, L’Abbate A. Prognostic importance of dipyridamole-echocardiography test in coronary artery disease. Circulation. 1989;80:450-457.
Stratmann HG, Williams GA, Wittry MD, Chaitman BR, Miller DD. Exercise technetium-99m sestamibi tomography for cardiac risk stratification of patients with stable chest pain. Circulation. 1994;89:615-622.
Picano E. Stress echocardiography: from pathophysiological toy to diagnostic tool. Circulation. 1992;85:1604-1612.
Marwick T, Willemart B, D’Hondt AM, Willemart B, Wijns W, Detry JM, Melin J. Selection of the optimal nonexercise stress for the evaluation of ischemic regional myocardial dysfunction and malperfusion: comparison of dobutamine and adenosine using echocardiography and 99m Tc-MIBI single photon emission computed tomography. Circulation. 1993;87:345-354.
Günalp B, Dokumaci B, Uyan C, Vardareli E, Isik E, Bayhan H, Özgüven M, Öztürk E. Value of dobutamine technetium- 99m-sestamibi SPECT and echocardiography in the detection of coronary artery disease compared with coronary angiography. J Nucl Med. 1993;34:889-894.
Schiller NB, Shah PM, Crawford M, DeMaria A, Devereaux R, Feigenbaum H, Gutgesell H, Reichek N, Sahn D, Schnittger I, Silverman NH, Tajik AJ. Recommendations for quantification of the left ventricle by two-dimensional echocardiography: American Society of Echocardiography Committee on Standards, Subcommittee on Quantitation of Two-Dimensional Echocardiograms. J Am Soc Echocardiogr. 1989;2:358-367.
Forster T, McNeill AJ, Salustri A, Reijs AEM, El-Said EM, Roelandt JRTC, Fioretti PM. Simultaneous dobutamine stress echocardiography and 99-m technetium isonitrile single photon emission computed tomography in patients with suspected coronary artery disease. J Am Coll Cardiol. 1993;21:1591-1596.
Geleijnse ML, Marwick TH, Boersma E, Deckers JW, Melin JA, Fioretti PM. Optimal pharmacological stress testing for the diagnosis of coronary artery disease: a probabilistic approach. Eur Heart J. 1995;16(suppl M):3-10.
Marwick TH, D’Hondt AM, Mairesse GH, Baudhuin T, Wijns W, Detry JM, Melin JA. Comparative ability of dobutamine and exercise stress in inducing myocardial ischemia in active patients. Br Heart J. 1994;72:31-38.
Ruffolo RR. The pharmacology of dobutamine. Am J Med. 1987;294:244-248.
Mertes H, Sawada SG, Ryan T, Segar DS, Kovacs R, Foltz J, Feigenbaum H. Symptoms, adverse effects, and complications associated with dobutamine stress echocardiography: experience in 1118 patients. Circulation. 1993;88:15-19.
Poldermans D, Fioretti PM, Boersma E, Forster T, van Urk H, Cornel JH, Arnese M, Roelandt JRTC. Safety of dobutamine-atropine stress echocardiography in patients with suspected or proven coronary artery disease: experience in 650 consecutive examinations. Am J Cardiol. 1994;73:456-459.
Nesto RW, Kowalchuck GJ. The ischemic cascade: temporal sequence of hemodynamic, electrocardiographic and symptomatic expressions of ischemia. Am J Cardiol. 1987;57:23C-27C.
Fung AY, Gallagher KP, Buda AJ. The physiologic basis of dobutamine as compared with dipyridamole stress interventions in the assessment of critical coronary stenosis. Circulation. 1987;76:943-951.
Kumar EB, Steel SA, Howey S, Caplin JL, Aber CP. Dipyridamole is superior to dobutamine for thallium stress imaging: a randomised crossover study. Br Heart J. 1994;71:129-134.
Laarman GJ, Bruschke AVG, Verzijlbergen JF, Go TL, Bal ET, Van der Wal EE, Ascoop CAPL. Thallium-201 scintigraphy after dipyridamole infusion with low exercise: quantitative analysis vs visual analysis. Eur Heart J. 1990;11:162-172.