Abstract
For patients with acute myocardial infarction who undergo primary percutaneous transluminal coronary angioplasty (PTCA), it is important to promptly identify those in whom a significant delayed improvement of global left ventricular function is to be expected as a result of successful treatment. Methods: In 51 patients with acute myocardial infarction who underwent primary PTCA with a good angiographic result, the late outcome of the left ventricular ejection fraction (LVEF) was established after a 6-mo follow-up. In the early phase after infarction (within ∼1 wk of infarction), the contractile reserve of the infarct zone was evaluated using dobutamine echocardiography and compared with the assessment of infarct size, infarct severity, and mean tracer activity of the infarct zone obtained using resting 99mTc-sestamibi SPECT. Receiver-operating-curve (ROC) analysis was used to define the reliability of the various parameters for identifying the patients with a follow-up LVEF increase of ≥5 LVEF units. Results: Of the 48 patients without restenosis at follow-up, 34 showed significant LVEF improvement. The evaluation of the contractile reserve of the infarct zone achieved an ROC curve area of 0.75 ± 0.07 with 74% sensitivity, 71% specificity, and 73% overall accuracy. Of the 99mTc-sestamibi SPECT parameters, the extent of the infarct had no diagnostic value according to ROC analysis. The mean activity of the infarct zone had an ROC curve area of 0.64 ± 0.09 with 82% sensitivity, 50% specificity, and 73% overall accuracy. The infarct severity had an ROC area of 0.76 ± 0.08 (not significant vs. mean activity and vs. contractile reserve) with 77% sensitivity, 71% specificity, and 75% overall accuracy. Conclusion: Evaluation of the contractile reserve of the infarct zone using dobutamine echocardiography and assessment of the tracer activity of the infarct zone or infarct severity using 99mTc-sestamibi SPECT in the early phase after infarction are able to identify the patients in whom successful primary PTCA will be followed by significant late LVEF improvement.
- acute myocardial infarction
- echocardiography
- primary percutaneous transluminal coronary angioplasty
- 99mTc-sestamibi
The purpose of reperfusion therapy in patients with acute myocardial infarction is to achieve significant myocardial salvage and to limit the extent of irreversible tissue damage. Salvaged myocardium is defined precisely as the difference between the initial area at risk and the final infarct size. Perfusion SPECT with 99mTc-sestamibi before and after treatment has been shown to be highly accurate for this evaluation (1–4). Unfortunately, although the acquisition of SPECT after treatment (usually at the time of the patient’s discharge from the coronary care unit or from the hospital) is almost always feasible, performance of a study before reperfusion is often difficult because of technical and logistic problems (5). Perfusion images after treatment alone allow an accurate assessment of the infarct size but no direct visualization of salvaged myocardium (6–10). In clinical terms, the main features of salvaged myocardium are to be an area at high ischemic risk (11) and often to be dysfunctional because of postischemic stunning (12). The introduction of primary percutaneous transluminal coronary angioplasty (PTCA) has significantly improved the chance of restoring a normal coronary blood flow compared with thrombolysis, with a consequent decrease in the rate of ischemic events early and late after reperfusion (13). On the other hand, even a satisfactory angiographic result after primary PTCA with reestablished vessel patency does not always imply an effective tissue reperfusion with consequent actual myocardial salvage (14). Because the degree of left ventricular impairment, as expressed by the left ventricular ejection fraction (LVEF), remains a major prognostic factor even in patients who have undergone primary PTCA, there is still a need for methods that recognize viability within dysfunctioning segments and thus predict the final outcome of left ventricular function after infarction. The demonstration of contractile reserve within asynergic segments using dobutamine echocardiography is effective for the identification of stunned salvaged myocardium on a regional basis (15–20). Other data also suggest that the demonstration of tracer uptake over the viability threshold in 99mTc-sestamibi SPECT after reperfusion predicts reversible regional dysfunction (21). However, in both instances, scanty data are available on the relationship between regional viability and prediction of functional outcome defined in terms of global LVEF improvement. The aim of this study was to verify whether the evaluation of the contractile reserve with dobutamine echocardiography and perfusion imaging with 99mTc-sestamibi (both performed on the patient’s discharge from the coronary care unit) can identify the patients in whom a significant delayed increase in the LVEF after reperfusion therapy with primary PTCA is to be expected or to be ruled out and to compare the 2 methods in the same patient population.
MATERIALS AND METHODS
Patient Population and Study Protocol
The patient population included 51 consecutive patients (43 men; mean age, 59 ± 12 y) with acute myocardial infarction who were admitted to our coronary care unit within 6 h of the onset of symptoms (mean time, 138 ± 90 min; range, 30–360 min) and who were referred to the catheterization laboratory for emergency primary PTCA. The diagnosis of acute myocardial infarction required the presence of typical chest pain lasting >30 min together with a >0.1-mV ST segment elevation in at least 2 contiguous electrocardiographic leads. Other inclusion criteria were total occlusion of the infarct-related artery; successful primary PTCA (defined as Thrombolysis in Myocardial Infarction trial flow grade 3 (22) and residual stenosis of <30% of lumen diameter); and willingness to participate in the study. Patients were excluded in the case of clinical instability or cardiogenic shock and if the echocardiographic window was inadequate for satisfactory visualization of the entire left ventricular wall. Of 62 otherwise eligible patients, 7 subjects (11%) were excluded on the basis of their clinical status and 4 (7%) were excluded because of poor quality of the echocardiographic images. The study protocol included the acquisition of 2-dimensional (2D) echocardiography to evaluate regional wall motion and global left ventricular function on admission of the patient. Dobutamine echocardiography was performed at a mean of 3 ± 1.1 d (range, 2–5 d), and resting 99mTc-sestamibi SPECT was performed at a mean of 6.5 ± 1.2 d (range, 4–9 d) after index infarction. Follow-up 2D echocardiography and coronary angiography were repeated 6 mo later. The ethics committee of our institution approved the study protocol.
Coronary Angiography
All angiograms were analyzed in a random sequence by 2 experienced observers who were unaware of other results. Discrepancies were resolved by consensus. In all patients, the infarct-related artery was analyzed before and after primary PTCA to assess residual stenosis. The same views of coronary arteries were used on follow-up to assess the arterial patency and restenosis rate.
Low-Dose Dobutamine Protocol
Dobutamine (5 μg/kg of body weight per minute) was infused under continuous electrocardiographic and blood pressure monitoring. Infusion was continued for 5 min and then increased to 10 μg/kg/min for an additional 5 min. The criteria for early interruption of dobutamine infusion included hypotension, angina, and significant ventricular arrhythmia.
2D Echocardiography
Patients were studied at rest and under low-dose dobutamine infusion in the left lateral decubitus using a commercially available scanner (Aloka SSD 870, 2.5- and 3.5-MHz transducers; Aloka Co., Tokyo, Japan). All standard views were acquired and recorded on videotape for off-line evaluation of regional wall motion. For calculation of the LVEF, 2D echocardiographic images were transferred to the hard disk of a TomTec P90 (TomTec Imaging Systems, Munich, Germany) medical off-line computer analysis system and digitized. LVEF was calculated under baseline resting conditions on admission and follow-up examination.
99mTc-Sestamibi SPECT
The tracer dose was 740–925 MBq (20–25 mCi) and was injected to the resting patient lying supine. Images were collected 60 min later using a single-head, large-field-of-view tomographic gamma camera equipped with an ultra-high-resolution collimator and with a 20% window centered on the 140-keV photopeak of 99mTc. Sixty projections of 20 s each were acquired. Image reconstruction was performed using filtered backprojection. No attenuation or scatter correction was used. The transaxial slices were realigned along the heart axis. For quantitative evaluation of SPECT images, the short-axis slices, from the first with apical activity to the last with activity at the base, were used. Their count profiles were generated by computer software and plotted onto a 2D volume-weighted polar map.
Data Analysis
For the echocardiographic evaluation, the left ventricular wall was divided in 16 segments (23), and 2 infarct zones (anterior and inferior, the latter including also the lateral wall) were defined as theoretic maximal risk areas (17). Because the apical lateral and the apical inferior segments can be perfused by anterior (left anterior descending) and posterior (right coronary or left circumflex) arteries, they were considered as overlap segments; thus, each infarct zone included 9 segments (17). Two experienced observers, who were unaware of patient data and image sequence, examined independently the studies and scored the segmental wall motion as follows: 1 = normal, 2 = hypokinesis, 3 = akinesis, and 4 = dyskinesis (17). In evaluating wall motion abnormalities, attention was also paid to systolic thickening in the central portion of the segment. Discrepancies were resolved by consensus. The appropriate infarct zone was identified in each patient and its wall motion score index, which is the sum of the segment scores divided by 9, was derived for baseline and dobutamine echocardiograms. The contractile reserve was defined by the difference between baseline and the dobutamine wall motion score index (17). The LVEF was measured with the modified Simpson’s rule algorithm (24). The mean values of 3 measurements of the technically best cardiac cycles were taken from each examination. Improvement of global LVEF after revascularization was arbitrarily defined as an increase of ≥5 LVEF units in the follow-up control compared with the baseline value (25).
For SPECT evaluation, the polar map displays were divided into 16 segments, matching with the echocardiographic ones. Using an automated procedure, segment tracer activity was calculated as the total of the normalized counts of the pixels included within the segment divided by the number of pixels. The segment with maximal activity was then normalized to 100, and the activity of the other segments was expressed as a percentage of the peak activity segment (26). Segmental 99mTc-sestamibi activity was classified as mild to moderately abnormal if it was <75% and ≥50% of the peak value and as severely abnormal if it was <50% of the peak value. The same infarct zones defined for echocardiography were considered. For each infarct zone, the following parameters were determined: mean 99mTc-sestamibi activity value, calculated as the sum of the activity in the related segments divided by 9; extent of infarct, expressed as the total number of segments with abnormal uptake; and severity index of the infarct, calculated as the number of segments with severely decreased 99mTc-sestamibi uptake divided by the total number of segments with abnormal uptake.
Statistical Analysis
Data are expressed as the mean ± SD. Comparisons of continuous variables within and between groups were performed using the Student t test for paired or unpaired samples, with the Bonferroni correction for repeated measures, as appropriate. The diagnostic power of the echocardiographic and scintigraphic variables for predicting a significant improvement in LVEF was assessed using receiver-operating-characteristic (ROC) curves. Each area was expressed as ±SE and compared using Wilcoxon statistics. P < 0.05 was considered statistically significant.
RESULTS
Patient Characteristics
The location of the infarct, according to the electrocardiographic changes and the wall motion abnormalities, involved the anterior infarct zone in 26 patients and the inferior infarct zone in 25 patients. The infarct-related artery was the left anterior descending artery in 26 patients, the left circumflex artery in 2 patients, and the right coronary artery in 23 patients. Thirty-three patients were affected by single-vessel coronary artery disease, 12 by 2-vessel disease, and 6 by 3-vessel disease. The LVEF on admission was 44% ± 9%. Primary PTCA was successful in all patients and was performed without complications. Stenting of the dilated infarct-related artery was performed in 26 patients. The peak creatine phosphokinase release reached 2,922 ± 2,453 IU/L.
Follow-Up
On follow-up at 6 mo, 3 patients had severe (>70%) restenosis of the infarct-related vessel and, therefore, were excluded from analysis. In the remaining 48 patients, the baseline LVEF (43.4% ± 8.6%) increased to 51.9% ± 11.4% at 6 mo (P < 0.00001). However, only 34 patients showed a significant (≥5 LVEF units) improvement. The 6-mo LVEF of these patients was 55.6% ± 8.9%, significantly higher than on admission (43.3% ± 8.6%; P < 0.00001). The LVEF was unchanged in 14 patients (42.8% ± 12% at 6 mo vs. 43.6%±9.2% on admission; not significant). Table 1 compares the main features of the 2 groups.
Dobutamine Echocardiography
All 48 patients still had an abnormal infarct zone wall motion at the time of dobutamine echocardiography. The mean baseline wall motion score index was 2.1 ± 0.47 (range, 1.22–2.77). The dobutamine infusion protocol was concluded in all patients without complications. According to the comparison between baseline and dobutamine wall motion score index, 19 patients showed no contractile reserve, and the remaining 29 had a decrease in wall motion score index ranging from 0.11 to 0.89 (mean, 0.35 ± 0.19). A significant difference in the dobutamine-induced change in wall motion score index was registered between the 34 patients with significant LVEF improvement (0.26 ± 0.23) and the 14 without significant improvement (0.08 ± 0.15; P < 0.05). The area under the ROC curve for contractile reserve was 0.75 ± 0.07. The best threshold of the decrease in wall motion score index to detect patients with significant global LVEF improvement was set at ≥0.11. With this cutoff, sensitivity was 74% and specificity was 71%, with 86% positive predictive value, 53% negative predictive value, and 73% overall accuracy.
99mTc-Sestamibi SPECT
Of the 432 segments included within the infarct zones of the 48 patients in whom functional follow-up analysis was completed, 290 showed an abnormal 99mTc-sestamibi uptake, with a mean of 6 ± 2.4 segments per patient (range, 0–9). The mean activity of the infarct zone was 65.9% ± 12.7% (range, 39.8%–86.2%). A severely depressed 99mTc-sestamibi uptake was registered in 93 segments, with a mean of 1.9 ± 2.2 (range, 0–7) segments per patient. The severity index of the infarct ranged from 0 to 0.78, with a mean of 0.27 ± 0.26. Table 1 compares the 99mTc-sestamibi SPECT results in patients with LVEF improvement versus those without LVEF improvement. The single variable that was significantly different between the 2 groups was the severity index of the infarct.
Figure 1 and Table 2 compare the diagnostic reliability of the various SPECT variables for recognizing patients with significant LVEF improvement on follow-up. The area under the ROC curve for the extent of infarct was just above the limit of no discrimination. The area under the ROC curve for the mean 99mTc-sestamibi activity of the infarct zone was significantly larger (P < 0.05). With a cutoff of ≥60% of peak activity, sensitivity was 82%, specificity was 50%, and overall accuracy was 73%. The best ROC curve was obtained using the severity index of the infarct, with an area significantly larger than that of the extent of the infarct (P < 0.01) but not larger than that of 99mTc-sestamibi activity (P = 0.09). Using a cutoff of ≤0.33, 77% sensitivity, 71% specificity, and 75% overall accuracy were achieved. No significant difference in the area under the ROC curve was registered between the contractile reserve using dobutamine echocardiography and either the mean 99mTc-sestamibi activity or the mean severity index of the infarct in perfusion SPECT. Accordingly, the diagnostic reliability (sensitivity, specificity, predictive values, and overall accuracy) of the 3 parameters (contractile reserve, infarct severity, and mean 99mTc-sestamibi activity) was closely comparable.
DISCUSSION
The main finding of this study is that 99mTc-sestamibi perfusion SPECT and dobutamine echocardiography are able to recognize with satisfactory accuracy the patients in whom the LVEF will spontaneously improve after successful primary PTCA for acute myocardial infarction. Therefore, independent of the extent of the initial area at risk and of the final infarct size, the presence of myocardial stunning caused by the presence of salvaged myocardium as a result of effective reperfusion can be inferred.
Reperfusion therapy is nowadays the standard treatment for patients with acute myocardial infarction who are admitted to the hospital within a reasonable time interval after the onset of symptoms (27). Thrombolytic therapy and primary PTCA have been shown to be effective in limiting the extent of myocardial necrosis, thereby improving symptoms and decreasing the rate of complications and the early and late mortality (13,28). However, the effectiveness of therapy in the individual patient is influenced by several factors and is more difficult to establish. This is especially true for systemic thrombolysis, because even the recanalization of the infarct-related vessel remains uncertain. However, in the case of direct PTCA, a satisfactory angiographic result with reestablished vessel patency does not always imply an effective tissue reperfusion with consequent actual myocardial salvage (14). Therefore, imaging techniques able to identify the risk area involved by the myocardial infarction, the final damage after reperfusion therapy, and, by difference, the extent of myocardial salvage would be desirable. Myocardial perfusion SPECT with 99mTc-sestamibi before and after treatment fulfills all of the requirements of an ideal imaging modality for this particular issue (1–4), but various practical problems limit the wide application of imaging before reperfusion to all patients with acute myocardial infarction (5). Myocardial contrast echocardiography performed immediately before reperfusion and repeated thereafter has been used, but published data indicate that this method has a low specificity (19,21,29). Thus, major efforts have been undertaken to identify imaging modalities that are able to give at least part of the needed information with a single examination performed after the reperfusion therapy has been completed but before the patient’s discharge.
Dobutamine echocardiography is highly reliable in identifying the asynergic segments that will show spontaneous function recovery after acute myocardial infarction (15–20). These segments, which include viable stunned myocardium, can be differentiated because of the presence of contractile reserve during inotropic stimulation from the irreversibly damaged regions. So far, limited data on the relationship between the extent of contractile reserve and the presence of significant late increase of global LVEF have been available (20). According to our results, the detection of any contractile reserve in the infarct zone is able to recognize with good accuracy the patients in whom the LVEF will increase on late follow-up after myocardial infarction. Even a minimal wall motion improvement during dobutamine infusion appears predictive of late significant improvement of the LVEF, so that a lower threshold of the wall motion score change of the infarct zone was selected compared with that in previous reports (17,20). On the other hand, a remarkable proportion of patients without a contractile reserve showed a significant LVEF improvement on follow-up, and thus the low negative predictive value appears to be the most important limitation of dobutamine echocardiography in our series.
The assessment of infarct size in a 99mTc-sestamibi SPECT study acquired at the time of hospital discharge is a very useful parameter for stratifying patients’ risk and a possible endpoint for establishing the effectiveness of different therapeutic regimens (5,9). A close relationship exists between the infarct size assessed with 99mTc-sestamibi SPECT performed at the time of hospital discharge and the LVEF late after acute myocardial infarction (6,7). In addition, this parameter has an important prognostic implication (8). However, myocardial salvage may be present independently from the degree of final impairment of left ventricular function. A discrepancy between infarct size and the LVEF at hospital discharge has been shown to be an indirect sign of the presence of myocardial stunning, but the diagnostic reliability of this approach is not defined (10). Other data suggest that the simple recognition of 99mTc-sestamibi activity over the viability threshold in asynergic segments early after myocardial infarction is predictive of spontaneous recovery at 1 mo (21). Recently, Hambÿe et al. (30) reported that a specific pattern of decreased 99mTc-sestamibi uptake and even more depressed β-methyl-p-iodophenylpentadecanoic acid activity is predictive of significant recovery of the LVEF after revascularization in infarct patients. Starting with these premises, we attempted to establish whether any parameters that can be obtained from 99mTc-sestamibi SPECT after reperfusion are predictive of spontaneous LVEF increase. Because the initial extent of the area at risk was unknown, we chose to evaluate the SPECT data on the basis of standard infarct zones, as in previous studies using dobutamine echocardiography (17,19,20). Among the examined parameters, we considered the extent of infarct, the mean tracer activity of the infarct zone, and, finally, the severity of the residual damage, expressed by the incidence of segments with severely reduced activity over the segments with abnormal uptake. On the basis of our results, the first parameter did not differentiate the patients with late functional improvement from the others. This was shown by the lack of a significant difference in the mean extent of infarct between the 2 groups of patients. In addition, the results of ROC analysis showed an area under the curve just above the limit of diagnostic uncertainty. Conversely, the areas under the ROC curves constructed using the mean 99mTc-sestamibi activity and, above all, the severity index were significantly larger and were not significantly different than that of the contractile reserve assessed by dobutamine echocardiography. Accordingly, the sensitivity, specificity, and overall accuracy of those parameters were very close to the related values achieved using dobutamine echocardiography. Interestingly, considering either the mean activity of the infarct zone or the severity of the infarct, instead of the extent of the infarct, the reliability of 99mTc-sestamibi SPECT alone approached the diagnostic reliability of the combined evaluation of perfusion and metabolism proposed by Hambÿe et al. (30), with the advantage of a simpler and widely available imaging protocol. With regard to the possible physiopathologic explanation of our results, it is reasonable that patients in whom the residual damage after revascularization includes a prevalence of severely necrotic segments (with reduced mean tracer activity) do not to show a significant recovery on follow-up, because the proportion of residual viable stunned myocardium within the infarct area is probably limited and irreversible myocardial damage is already established.
The limitations of this study must be considered. The choice of analyzing the data on the basis of theoretic maximal risk areas (standard infarct zones) was justified by the lack of an objective criterion to define the true initial risk area. However, this is an important limitation because the possibility of infarcts extending over boundaries of the infarct zones cannot be ruled out. In our series we did not observe asynergic segments outside the infarct zone in any patient, but this remains theoretically possible. The use of a low-dose dobutamine protocol precluded the possibility of examining the presence of ischemia or biphasic responses. Similarly, the execution of stress imaging could have increased the value of perfusion scintigraphy. However, the study protocol required as inclusion criterion the optimal result of primary PTCA and, hence, a low likelihood of inducible ischemia early after the procedure had to be expected. Although all patients were in apparently stable clinical conditions, the time interval between the execution of dobutamine echocardiography and 99mTc-sestamibi SPECT is another potential limitation for a comparison of the 2 methods. With regard to the SPECT parameters that we evaluated, none of them is optimal and devoid of potential criticisms. The mean activity within the infarct zone attempts to approximate the threshold method that is widely adopted for the evaluation of myocardial viability in patients with chronic coronary artery disease and left ventricular dysfunction (26,31). It is noteworthy that the best threshold identified for this parameter by ROC analysis was set at ≥60% of peak activity, in agreement with previous studies (21,26,31). The estimate of the extent of the infarct in terms of the number of infarct zone segments with abnormal tracer uptake is very rough. Also, our proposal of a severity index could be criticized for the same reason. Most previous studies of infarct size used a fixed activity threshold over the entire left ventricular wall and did not consider a segment model (5–10), so this approach was difficult to reconcile with the concept of infarct zone and to compare with echocardiography. Furthermore, severity was not considered. The assessment of the extent and severity of the infarct using the comparison with a normalcy database could be a possible alternative (32), but, again, it would be difficult to adapt the data to the use of a standard infarct zone.
CONCLUSION
The results of this study confirm the value of dobutamine echocardiography for the detection of stunned salvaged myocardium early after acute myocardial infarction and define its diagnostic reliability for the prediction of LVEF outcome. In addition, the perfusion pattern evaluated by means of a single 99mTc-sestamibi SPECT study after reperfusion therapy is able to identify those patients in whom myocardial salvage will lead to a delayed increase in the LVEF. In particular, the presence of functional recovery is related to the infarct severity and mean tracer uptake. These parameters, together with the infarct size, could be useful in assessing the effectiveness of acute reperfusion therapy in patients with acute myocardial infarction and in guiding the subsequent management of the patient.
Footnotes
Received Mar. 21, 2001; revision accepted Aug. 29, 2001.
For correspondence or reprints contact: Roberto Sciagrà, MD, Nuclear Medicine Unit, Department of Clinical Physiopathology, University of Florence, Viale Morgagni 85, 50134 Florence, Italy.
E-mail: r.sciagra{at}dfc.unifi.it