Author + information
- Received May 21, 1997
- Revision received May 27, 1998
- Accepted June 12, 1998
- Published online October 1, 1998.
- Tadamichi Sakuma, MDa,* (, )
- Yasuhiko Hayashi, MDa,
- Kotaro Sumii, MD∗,
- Michinori Imazu, MD∗ and
- Michio Yamakido, MD∗
- ↵*Address for correspondence: Dr. Tadamichi Sakuma, Division of Cardiology, Akane Foundation Tsuchiya General Hospital, 3-30 Nakajima, Naka-ku, Hiroshima, 730-8655 Japan
Objectives. This study sought to determine whether microvascular integrity in the risk area (RA) for myocardial infarction (MI) one day after recanalization predicts the outcome in patients with first acute MI.
Background. Immediately after recanalization, microcirculation in the RA is modified by both hyperemic response and microvascular impairment.
Methods. Fifty consecutive patients who underwent serial myocardial contrast echocardiography before and one day after recanalization (day 2) were studied. All patients had a completely occluded lesion in the left anterior descending coronary artery alone, and underwent successful reperfusion therapy. The relative size of the initial RA (RA ratio) and peak gray scale ratio (PGSR) within the RA on day 2 were determined. Patients were followed for a median of 22 months to evaluate clinical outcome.
Results. On day 2, PGSR was a median of 0.46. Study patients were subdivided into two groups, group A of 24 patients with acceptable opacification (PGSR > 0.46 on day 2) and group B of 26 patients without it. Major cardiac events (cardiac death, nonfatal MI and repeat admission for congestive heart failure) were more frequently observed in group B (28% vs. 4%, Cox hazard ratio = 8.5, p = 0.05, 95% confidence interval [CI] 1.03 to 69.9). The median value of the RA ratio was 0.45. Patients (n = 15) with RA ratio > 0.45 on day 1 and PGSR on day 2 ≤ 0.46 exhibited a 10.7-fold relative risk for major cardiac events (p = 0.005, 95% CI 2.06 to 55.8) and a 3.69-fold relative risk for composite cardiac events (major cardiac events and target lesion revascularizations) after the initial intervention (p = 0.004, 95% CI 1.51 to 9.04).
Conclusions. The assessment of both the size of the initial RA and microvascular integrity on day 2 enables precise determination of the efficacy of reperfusion therapy and prediction of the short- and intermediate-term prognoses of patients with recanalized MI.
During the last two decades, the efficacy of reperfusion therapy for acute myocardial infarction (MI) has been established (1), yet the degree of myocardial salvage after recanalizing culprit lesions varies from patient to patient (2). It is therefore necessary to evaluate myocardial viability during the acute stage, and to administer aggressive treatment for the anticipated left ventricular hemodynamic and structural changes based on individual patient cardiac condition. Several methods are available for evaluation of salvaged myocardial viability in the acute stage, including positron emission tomography (3)and myocardial scintigraphy (4). However, these methods may not always be available in the catheterization laboratory and are expensive. Myocardial contrast echocardiography (MCE) can easily be performed in the catheterization laboratory during the acute stage of MI and is reported to be one of the most promising methods for assessing microvascular integrity and consequent myocardial viability in the convalescent stage of MI.
The in-hospital prognoses of patients in whom angiographic no-reflow phenomena are observed immediately after reperfusion is unfavorable (5). Left ventricular function in patients with no-reflow phenomena detected using MCE is impaired in the chronic stage (6), in-hospital complications are frequent in such patients (7)and lack of reflow phenomena after initial reperfusion therapy is one of the predictors of left ventricular remodeling (7). However, microcirculation in the entire risk area (RA) changes rapidly immediately after reperfusion. In a previous study using serial MCEs (8), although we found no significant correlation between microvascular integrity immediately after recanalization and that in the convalescent stage of MI, we did find that microvascular integrity one day after recanalization was closely related to that in the convalescent stage. Thus, MCE one day after recanalization appears to enable the prediction of both the short- and intermediate-term prognoses of patients who have suffered acute MI. Nevertheless, the clinical prognoses of patients who have suffered a first acute MI and for whom serial MCE was performed during the acute phase has not been clearly determined. The present study was therefore designed to determine whether serial assessments of microvascular integrity in the RA for MI before and up to one day after recanalization of the culprit lesion are useful for predicting the short- and intermediate-term outcomes of patients who have experienced a first acute anterior or anteroseptal MI.
The study population consisted of 50 consecutive patients (32 men and 18 women) ranging from 44 to 88 years of age (mean, 66 years) who were admitted to the Coronary Care Unit at Tsuchiya General Hospital with the diagnosis of a first acute anterior or anteroseptal MI, and who underwent serial MCE examination before and immediately after successful recanalization and one day thereafter. They were prospectively selected for this study if they met the following criteria between February 1993 and April 1995: cardiopulmonary condition on arrival corresponded to Killip classification I, II or III; the culprit lesion for acute MI was located in the sixth or seventh segment as defined by the American Heart Association Committee Report (9); single-vessel coronary artery disease was present; Thrombolysis In Myocardial Infarction trial (TIMI) grade 0 or I flow was confirmed on multidirectional coronary angiography after selective infusion of 5 mg isosorbide dinitrate; collateral circulation feeding the infarct-related coronary artery (IRA) on the initial angiogram corresponded to Rentrop collateral filling grade 0, I or II (10); recanalization achieved TIMI grade II or grade III flow without ≥70% diameter residual stenosis on multidirectional observation by quantitative coronary angiography within 24 h after the onset of symptoms; technically adequate two-dimensional echocardiographic images were obtained for MCE; and informed consent could be obtained from them or from their families.
Definitive diagnosis of acute MI was made based on the following criteria: chest pain lasting for more than 30 min; ST-segment elevation in at least two consecutive anterior leads on electrocardiographic examination; regional anterior or anteroseptal wall motion abnormalities observed in the territory of the left anterior descending coronary artery (LAD) on two-dimensional echocardiograms; and confirmation of the diagnosis by a typical pattern of increase and decrease over time in total creatine phosphokinase.
During the period for enrollment of eligible patients, 149 consecutive patients with acute anterior or anteroseptal MI underwent emergent coronary angiography. Following exclusions, 66 patients remained. However, no informed written consent could be obtained from five of these patients. Of the remaining 61 patients, no repeat coronary angiograms or MCEs were performed for 4 on the second hospital day, in 1 case due to pneumonia, in 2 cases to cerebral vascular accident, and in 1 case to acute renal insufficiency. For seven other patients, inadequate image quality was obtained for analysis with our off-line computer system. Thus, a total of 50 patients were finally enrolled for this cohort study.
Angiographic study and MCE
On the first hospital day, emergent coronary angiography was performed following admission. Selective coronary angiography was performed using the Seldinger technique after intravenous injection of 50 U/kg heparin, and multiple projections including at least two orthogonal angiographic views with minimal vessel overlaps were recorded at 30 frames/s. Provided that patients did not have severe congestive heart failure, that is, pulmonary capillary wedge pressure less than 18 mm Hg as measured using a Swan–Ganz catheter inserted via the femoral vein, left ventriculograms in the 30° right anterior oblique plane were obtained prior to recanalization. After confirming occlusion of the proximal portion of the LAD, MCEs were performed prior to recanalization, as follows. MCE was performed by selective intracoronary injection of 2 ml sonicated ioxaglic acid (Hexabrix-320, Tanabe, Osaka, Japan) generated using an ultrasonic homogenizer (Model 250, Sonifier-II, Branson Ultrasonics Corporation, Danbury, Connecticut) at a rate of 1 ml/s with a 5F diagnostic catheter. Sonicated microbubbles were generated by sonicating the contrast agent at 50 watts for 30 s (11).
The mean diameter of the sonicated ioxaglic acid was measured using computed planimetry. The size of microbubbles was microscopically compared with that of polybead blue-dyed microspheres 10 μm in diameter (Polysciences, Inc., Warrington, Pennsylvania). The mean diameter was 10.3 ± 3.9 μm (mean ± SD). Two repeat two-dimensional echocardiographic images in the long apical view were recorded on super-VHS video cassette tape using a high-fidelity video recorder (Model SVO-9500MD, Sony, Tokyo, Japan) and a phased-array system (Model SSD 870 Ultrasound Systems, Aloka, Tokyo, Japan) with a 3.5-MHz transducer. We attempted to visualize images formed by the plane defined by three points, the junction of the mitral valve cusp, that of the aortic valve cusp and the cardiac apex. The gain settings and sensitivity time control were held constant throughout the observations after adjusting for optimal identification of the endocardium and epicardium. Once we had determined the optimal settings for each patient, the same settings were used on all occasions. We defined the RA for MI as a contrast deficit in the end-diastolic phase in the territory of the LAD observed after injection of the sonicated contrast media before opening the IRA (12). After determination of the RA, thrombolytic therapy (intravenous or intracoronary infusion of tissue-type plasminogen activator, or intracoronary infusion of urokinase) or primary percutaneous transluminal coronary angioplasty (PTCA) was performed. For patients in whom TIMI grade II or grade III flow could not be obtained with thrombolytic therapy alone, rescue PTCA was performed. When PTCA was performed, bolus injection of heparin up to a total amount of 10,000 U was added. Twelve-lead electrocardiograms (ECGs) were recorded before recanalization and serially up to the end of the catheterization procedures. At the time of recanalization of the culprit lesion, we carefully observed whether reperfusion arrhythmias and/or ST-segment re-elevation appeared. Reperfusion arrhythmia was defined as the transient appearance of premature ventricular contractions, accelerated idioventricular rhythm, ventricular tachycardia or ventricular fibrillation immediately after recanalization. Significant transient ST-segment re-elevation at 60 ms after the J point was defined as an increase in ΣST (sum of ST-segment elevations in V1through V6of >5 mm immediately after recanalization compared with that before re-establishment of anterograde flow. Myocardial contrast echocardiographies were performed immediately after successful recanalization with a 5F diagnostic catheter. We attempted carefully to depict the same outline as that in the initial examination of prereflow MCE. Since remaining regions of pronounced narrowing decrease coronary flow reserve and maximum opacification in the RA (13), we excluded patients with high-grade residual stenosis from this study from the first hospital day onward. Each patient was in principle administered intravenous nitroglycerin, intravenous heparin and oral aspirin after being transferred to the Coronary Care Unit. On the second hospital day, coronary angiography was performed and MCE was repeated using the same method as described above. In both convalescent (1 month after recanalization) and remote stages (6 months and/or 1 year after recanalization), whenever feasible, coronary angiograms and left ventriculograms were obtained again in a similar manner.
The study protocol was approved by the Institutional Review Board of Akane Foundation Tsuchiya General Hospital. The registered patients and/or their family received written information concerning the study and gave informed consent for participation.
Analysis of catheterization and MCE data
The coronary angiograms were analyzed by an experienced operator using a commercially available quantitative coronary angiographic software program (POLY C2 LAR C2 D.C.I., Coronary Quantification Software Package, Philips Medical Systems, The Netherlands BV). Using the projection revealing the severest stenosis, the minimal lumen diameter, reference coronary artery dimensions and percent diameter stenosis were measured.
To evaluate the size and microvascular integrity of the RA, we employed a commercially available off-line computer (Macintosh Quadra 800, Apple Computer Japan, Inc., Tokyo, Japan) equipped with a video capturing card (Quick Capture Card, Data Translation Corporation, Marlboro, Massachusetts). The images at end-diastolic phase on electrocardiographic monitoring were accumulated in the hard disk as sequential memory media approximately 15 s prior to injection of the sonicated medium and up to 1 min after injection. Computed planimetry was used to measure the size of the initial RA. A mean gray scale in both the RA and a region of the posterior wall (as a control site) was computed using public domain image analyzing software (i.e., NIH Image ver. 1.56), excluding the endocardial and epicardial borders. After converting composite signals into 640 × 480 ×8-bit format, we analyzed the peak gray scale ratio (PSGR) within the RA normalized to the peak gray scale of the left circumflex artery area with subtraction of the baseline using the following formulas (14): PGSR = A/X, A = BSPGS in RA, and X = BSPGS in LCX, where PGSR is the peak gray scale ratio, BSPGS the baseline-subtracted peak gray scale value after opening the IRA and LCX the region of interest in the territory of the left circumflex coronary artery. For serial measurements of PGSR, image analysis was performed after the same outline as prereflow examination in the long apical view was carefully delineated in each stage using the off-line computer system. We defined the ratio of relative size of the initial RA (RA ratio) as follows: RA ratio = RA in the territory of LAD/total longitudinal cross-sectional area in the apical long axial view of the left ventricle. The off-line measurements of areas in regions of interest and gray scale levels of intensity were determined as the means of values of two measurements obtained by each of two independent observers who were unaware of the other’s findings.
Left ventricular regional wall motion was assessed in each case by the centerline method described by Sheehan et al. (15), in which the mean value of the standard deviation of chord length (i.e., SD/chord) was calculated using the normalized chord length in the LAD region (from 10 to 66 in chords out of 100 chords) in the end-diastolic perimeter of the left ventricle.
Reproducibility of results of MCE analysis
To determine the interinjection reproducibility of both peak contrast intensity and the RA ratio, we compared the results obtained from the analysis of 11 randomized paired injections performed at 4-min intervals. To assess interobserver and intraobserver reproducibility, a subset of 22 contrast injections, selected at random, was quantified by an independent observer and later by the initial observer. The interinjection, intraobserver and interobserver reproducibilities for peak contrast intensity in RA were tested for absolute error (i.e., absolute value of the peak gray scale on the second injection subtracted from that on the first injection) and percent error (i.e., absolute error divided by the peak gray scale on the first injection) (16). The interinjection, intraobserver and interobserver absolute errors in peak contrast intensity were 3.7 ± 4.7, 3.3 ± 2.8 and 4.0 ± 6.9 (mean ± SD), respectively. The percent errors in peak contrast intensity were 6.4 ± 4.9%, 6.6 ± 5.3% and 8.7 ± 9.5%, respectively. Similarly, the interinjection, intraobserver and interobserver absolute errors in the RA ratio were 0.04 ± 0.03, 0.03 ± 0.02 and 0.05 ± 0.03, respectively. The percent errors were 6.5 ± 5.8%, 5.4 ± 4.8% and 7.8 ± 6.5%, respectively.
The incidences of clinical events were determined at 6 and 12 months after hospital discharge and then once a year thereafter. Follow-up data were obtained from review of the patients’ hospital record, personal communication with the patients’ attending physicians and telephone interviews. Follow-up data were obtained for all patients.
The clinical events included in patient outcome determination were cardiac death, nonfatal myocardial reinfarction, repeat hospitalization for congestive heart failure (NYHA class III or class IV) and target lesion revascularization. The major cardiac events included cardiac death, nonfatal myocardial reinfarction and repeat hospitalization for congestive heart failure. The composite cardiac events included in determination of cardiac event-free survival were cardiac death, nonfatal myocardial reinfarction, repeat hospitalization for congestive heart failure and revascularization of the target lesion. Cardiac death was defined as death due to coronary artery disease. Myocardial reinfarction was confirmed by a typical pattern of increase and decrease over time in total creatine phosphokinase in association with the appearance of new, pathologic Q waves. The revascularization procedures for the target lesions were considered to include all PTCAs which included routine balloon coronary angioplasty, directional coronary atherectomy and metallic coronary artery stenting to the initial culprit lesion in the IRA, and coronary artery bypass surgery (CABG) involving the initial culprit lesions. If repeat interventions were performed to eliminate the cause of myocardial reinfarction or congestive heart failure, myocardial reinfarction or repeat hospitalization for congestive heart failure was recorded as the cardiac event. Follow-up angiography was performed in principle at 6 months (actual range, 3 to 8 months) and 12 months (actual range, 9 to 14 months) after the onset of symptoms of the initial MI. Restenosis was defined as reduction greater than 50% of the initial gain derived from PTCAs and/or thrombolytic therapy performed on the first hospital day, with greater than 50% diameter stenosis in the initial culprit lesion. Target lesion revascularizations at follow-up angiography were as a general rule attempted for patients who exhibited recurrence of severe narrowing with greater than 70% diameter stenosis on quantitative coronary angiography. For patients with stenosis between 50% to 70%, those were performed in accordance with the decision of the physicians performing angioplasty blind to MCE data or on the basis of ischemic change on treadmill tests. Clinical follow-up events were studied in accordance with the intention-to-treat principle. In-hospital events were included in the analysis of follow-up events.
Results are given as mean values ± SD or as proportions. Continuous variables were compared using Student’s t-test. Statistical analysis of discrete variables was performed with the chi-square test, and the Fisher exact test was used when appropriate. Correlations between two different groups were expressed by Pearson’s correlation coefficient. The significance of each correlation was determined, since the normalized correlation coefficient depended on the tdistribution. The individual effect of certain variables on cardiac event-free survival was evaluated with the Cox proportional hazard model. Using a stepwise selection process, variables were entered, or removed, from the regression equation on the basis of computed significance probability. When evaluating cardiac events, only the first cardiac event (in patients with >1 event) was considered. Findings of p < 0.05 (two-sided) were considered significant.
Serial MCEs were performed 31 min (range: 25 to 36) before opening of the IRA and 29 min (range: 13 to 39) after recanalization, and 23 h (range: 18 to 28) after opening of the IRA, on average. No clinical adverse effects of MCE procedures were observed except for transient hypotension in two patients. Peak gray scale ratio on the first hospital day was a median of 0.33 (0.37 ± 0.21), while PGSR on the second hospital day was a median of 0.46 (0.48 ± 0.28). The 50 patients enrolled were divided into two groups: 24 patients in group A with second-day PGSR greater than 0.46 (median 0.64, range 0.50 to 1.25), and 26 patients in group B with second-day PGSR less than or equal to 0.46 (median 0.27, range 0.05 to 0.46).
Table 1shows baseline characteristics of the two groups. No significant differences were found between the two groups in any of these characteristics. Table 2summarizes clinical characteristics and angiographic findings for each group. Between the two groups there were no significant differences in the site of culprit lesion, degree of collateral circulation immediately before recanalization, RA ratio, cardiac index on day 2, time from onset of symptoms to recanalization, frequency of coronary angioplasty, peak value of creatine phosphokinase, appearance of reperfusion arrhythmia, grade of TIMI flow on day 1, percent diameter stenosis on day 1 or percent diameter stenosis on day 2. There were, however, significant differences between them in hemodynamic variables and ECG findings. There was no significant difference in administered medications between the two groups (Table 2). The enrolled patients were followed up for a median of 21.8 months (range 3 days to 39 months). Table 3shows the clinical outcomes of the study patients. During the follow-up period, of the 24 patients in group A, 1 died due to cerebrovascular accident immediately after elective CABG. On the other hand, of the 26 patients in group B, 1 died due to low cardiac output syndrome on the third hospital day, 1 died suddenly due to ventricular fibrillation on the 15th hospital day and 1 died due to cardiac rupture of oozing type on the fifth hospital day. No myocardial reinfarctions or repeat admissions for congestive heart failure occurred in group A, while two patients in group B suffered myocardial reinfarction and two were readmitted for congestive heart failure. Two patients in group A and three patients in group B died of noncardiac causes. Seven patients in group A underwent repeat interventions during the follow-up period. Of the 26 group B patients, 10 underwent repeat interventions. Thirty-five (83%) of 42 patients, with 3 patients who suffered cardiac death and 5 who suffered noncardiac death excluded, underwent repeat diagnostic catheterization during the remote stage. Left ventriculograms could be obtained for 30 of the patients who were alive at 12 months after recanalization. Left ventricular ejection fraction and end-systolic ventricular volume were worse in group B than those in group A (Table 2) (17).
As determined using the Cox proportional hazard model, for major cardiac events the relative risk for the 26 group B patients was 8.5-fold that of the 24 group A patients (chi-square = 3.96, p = 0.05, 95% confidence interval [%CI] 1.03 to 69.9). For target lesion revascularization, the relative risk for group B patients was 1.67-fold that of group A patients, but the difference between groups in risk was not significant (chi-square = 1.08, p = 0.29, 95% CI 0.63 to 4.41). For composite cardiac events, the relative risk for group B patients was 2.13-fold that of group A patients, but again the difference between groups in risk was not significant (chi-square = 2.57, p = 0.10, 95% CI 0.85 to 5.35). One patient in group A and seven in group B had major cardiac events over their clinical course. If the cut-off value for prediction of major cardiac events was set at a PGSR of 0.46, sensitivity, specificity, positive predictive accuracy and negative predictive accuracy were 87.5%, 54.8%, 26.9% and 95.8%, respectively. Both specificity and positive predictive accuracy were too low for prediction of cardiac events. In a stepwise variable selection process for predicting major cardiac events, only the RA size (chi-square = 6.94, p = 0.008) and the PGSR on day 2 (chi-square = 3.94, p = 0.046) were entered. Therefore, with the vertical axis as PGSR and the horizontal axis as RA ratio, patients who suffered major cardiac events were concentrated among those for whom the RA ratio was high, and insufficient recovery from microvascular impairment was found even one day after reopening of the culprit lesion (Fig. 1,A). For the group of all 50 examined patients, the median value of the RA ratio was 0.45 (0.45 ± 0.10). Patients (n = 15) in whom the RA ratio was >0.45 and for whom PGSR one day after recanalization was ≤0.46 had a 10.7-fold relative risk for major cardiac events after initial intervention (chi-square = 7.95, p = 0.005, 95% CI 2.06 to 55.8). They also had a 3.44-fold relative risk for target lesion revascularization (chi-square = 6.17, p = 0.013, 95% CI 1.29 to 9.11) and a 3.69-fold relative risk for composite cardiac events (chi-square = 8.17, p = 0.004, 95% CI 1.51 to 9.04) (Fig. 1).
MCE one day after recanalization
Performance of serial MCEs appears to be clinically useful since it enables not only determination of both the location and size of the area at risk for necrosis prior to recanalization but also observation of the serial changes of restoration from microvascular impairment within the initial RA (8,18).
Infarct size in experimental animal models tends to be underestimated due to hyperemic response immediately after recanalization (19). Immediately after recanalization, coronary blood flow and volume can change rapidly. Microvascular conditions might be significantly altered by hyperemic response, microvascular spasm, microvascular embolism with microthrombi and/or neutrophilic infiltration, interstitial edema and/or hemorrhagic infarction in myocytes or other factors (20). Recently, Kates et al. (21)showed that the absence of perfusion abnormalities after reperfusion did not rule out the presence of necrosis in dogs subjected to prolonged ischemia. Conversely, they showed that the presence of a perfusion defect after reperfusion did not indicate the presence of necrosis but rather abnormal microvascular reserve. The presence of hyperemia along with microvascular impairment prevents exact determination of the infarct size immediately after recanalization. We believe that determination of when and how microvascular integrity should be estimated for accurate prediction of that in the convalescent stage is a matter of greatest importance. We have already reported that the greater part of rapid change in the injured microvasculature occurred by one day after opening of the IRA, and that microvascular integrity in the RA one day after recanalization was closely related to that in the convalescent stage (8). Microvascular integrity in the RA one day after recanalization is relatively stable compared with that immediately after recanalization, and therefore its determination using MCEs on the second hospital day is thought to be useful for prediction of the clinical outcomes of patients who have suffered acute MI.
Usefulness of MCE for predicting cardiac events
No significant differences were found between our two groups in age, Killip class, systolic blood pressure on arrival, grade of TIMI flow on day 1, degree of collateral circulation immediately before recanalization or residual stenosis of the culprit lesion on day 2. The 26 group B patients, however, had significantly higher pulmonary wedge pressure levels before recanalization, frequency of Q-wave infarction and rate of occurrence of transient ST-segment re-elevation immediately after recanalization than did the 24 group A patients. Furthermore, the patients in group B exhibited more serious impairment of left ventricular ejection fraction and regional wall motion abnormality on day 1 than did the patients in group A. In the univariate analysis, the relationships between clinical parameters and major cardiac events were as follows: pulmonary wedge pressure levels prior to recanalization (chi-square = 0.77, p = 0.38), frequency of Q-wave infarction (chi-square = 0.37, p = 0.54), rate of occurrence of transient ST-segment re-elevation (chi-square = 0.14, p = 0.71), left ventricular ejection fraction on day 1 (chi-square = 3.01, p = 0.08), PGSR on day 1 (chi-square = 2.21, p = 0.14), PGSR on day 2 (chi-square = 3.06, p = 0.08) and RA size (chi-square = 3.90, p = 0.05). However, in the stepwise variable selection process using the Cox hazard regression analysis model, only the RA size (chi-square = 6.94, p = 0.008) and the PGSR on day 2 (chi-square = 3.94, p = 0.046) among these variables were entered from the regression equation on the basis of significant computed probability in predicting major cardiac events. The size of the RA is poorly reflected by measured values of hemodynamic variables (12)and could not be assessed with precision by the site of coronary occlusion within a vessel during the acute stage of infarction (22). Furthermore, prolonged myocardial ischemia causes myocardial stunning in the acute stage as well as in the convalescent stage (20). Hence, we were unable to predict with precision the viability of myocardium in the RA using hemodynamic variables alone in the acute stage. Our findings demonstrated that patients who exhibit a high degree of microvascular damage within a relatively large initial RA are at high risk for in-hospital cardiac death due to low output syndromes, cardiac rupture, myocardial reinfarction and repeat admission for congestive heart failure, but that other patients with acceptable recovery from microvascular impairment suffered no such events, with the exception of one patient who underwent CABG in the remote stage. Although the reason why a large extent of microvascular impairment in the RA was associated with myocardial reinfarction could not be determined from the findings of our study alone, any of several mechanisms, such as abrupt thrombus formation in the epicardial main coronary artery with abolition of vascular remodeling (23)after large MI, decrease in anterograde flow due to angiographically undetectable flap formation and/or focal dissection in the culprit lesion of the epicardial main coronary artery and impaired coronary flow reserve due to sparse microvasculature in the RA, might possibly have participated in reocclusion of the initial IRA.
Our results agree with the finding that the size of the initial RA as assessed by Technetium-99m sestamibi imaging is a significant determinant of the likelihood of subsequent cardiac death (4). Furthermore, our findings are consistent with reports that patients who suffered a large MI affecting greater than 40% of total left ventricular mass often suffered cardiogenic shock or cardiac death due to low output syndrome (24)and had very poor outcomes even if reperfusion therapy was performed (4).
Comparison with previous studies
Few findings are available concerning the relationship between microvascular integrity within the initial RA and subsequent clinical outcomes. Recently Ito et al. (7)reported the in-hospital outcomes of patients in whom significant no-reflow phenomenon within the RA was detected by MCE. However, the difference in accuracy of prediction of prognosis between MCE performed on day 1 and that performed on day 2 was not clear. Figure 2shows receiver operator characteristic curves for clinical outcomes using PGSR values obtained by MCE on days 1 and 2. The PGSR obtained by MCE one day after recanalization was more useful for predicting major cardiac events than was the PGSR obtained by MCE immediately after recanalization.
Further, in previous studies, intermediate-term prognoses of patients with reperfused acute myocardial infarction were not determined using both initial RA and microvascular integrity after recanalization. In our study, the median value of PGSR on day 1 was 0.33. The factor “RA ratio >0.45 and PGSR on day 1 ≤0.33” was not an independent predictor of major cardiac events (chi-square = 1.46, p = 0.23) but was an independent predictor of composite cardiac events (chi-square = 4.31, p = 0.038). On the other hand, the factor “RA ratio >0.45 and PGSR on day 2 ≤0.46” was an independent predictor of both major cardiac events (chi-square = 8.23, p = 0.004) and composite cardiac events (chi-square = 10.72, p = 0.001). Serial observation using MCE thus enables precise prediction of the short- and intermediate-term prognoses of patients with recanalized MI.
Several limitations affect interpretation of the results of this study. First, the range of the 95th percentile CI for cardiac events was too wide to permit prediction of the clinical outcomes of patients with recanalized MI. Further study may be needed with registration of a larger number of patients. Second, we excluded Killip class 4 patients, those with multivessel coronary artery disease and those with a time to recanalization of greater than 24 h from the onset of symptoms. Thus, some subjects who might have suffered severe left ventricular dysfunction were excluded from the study. Third, although MCE performed during pharmacologically induced coronary hyperemia might permit accurate evaluation of infarct size in experimental animal models, this adjunctive procedure by itself inevitably affects not only the natural course of microvascular impairment but also the clinical outcomes of study patients. We therefore conducted the present study without adjunctive pharmacologic hyperemia. Fourth, we could not infuse the sonicated contrast media via the right coronary artery in several patients. Therefore, although the patients with Rentrop collateral grade III were excluded, the size of initial RA might be overestimated in such patients. Lastly, our MCE procedures required invasive techniques. Further, sonicated ioxaglic acid is not a physiologic tracer like sonicated albumin. It more likely acted as a deposit tracer. Grayburn et al. (25)showed that MCE using intravenous dodecafluoropentane accurately defined the myocardial area at risk and infarct size, and had the potential for reliable, noninvasive assessment of reperfusion following therapeutic interventions. In the near future, serial MCEs using such contrast media will enable interindividual variability in topography in the region of ongoing microvascular impairment to be determined, and will permit noninvasive determination of the efficacy of reperfusion treatment.
This study was performed to determine whether microvascular integrity in the RA for MI predicts the clinical outcome of patients who have suffered their first acute anterior or anteroseptal MI. Our findings demonstrated the importance of determining the actual size of the RA just before recanalization. Furthermore, both short- and intermediate-term cardiac events tended to occur in those patients with both large initial RA and inferior potential for recovery from microvascular impairment even at one day after recanalization. However, our procedure is invasive. In the near future, use of intravenous echocardiographic contrast agents may permit noninvasive evaluation of serial changes in the RA, and prediction of the risk of cardiac events in patients who have suffered acute MI.
Presented in part at the 68th Scientific Sessions of the American Heart Association, Anaheim, California, November 1995.
- coronary artery bypass surgery
- infarct-related coronary artery
- left anterior descending coronary artery
- myocardial contrast echocardiography
- myocardial infarction
- risk area
- peak gray scale ratio
- percutaneous transluminal coronary angioplasty
- Thrombolysis In Myocardial Infarction Trial
- Received May 21, 1997.
- Revision received May 27, 1998.
- Accepted June 12, 1998.
- American College of Cardiology
- Clements I.P,
- Christian T.F,
- Higano S.T,
- Gibbons R.J,
- Gersh B.J
- Maes A,
- Van D.W.F,
- Nuyts J,
- Bormans G,
- Desmet W,
- Mortelmans L
- Miller T.D,
- Christian T.F,
- Hopfenspirger M.R,
- Hodge D.O,
- Gersh B.J,
- Gibbons R.J
- Ito H,
- Tomooka T,
- Sakai N,
- et al.
- Ito H,
- Maruyama A,
- Iwakura K,
- et al.
- Austen W.G,
- Edwards J.E,
- Frye R.L,
- et al.
- Cohen M,
- Rentrop K.P
- Reisner S.A,
- Ong L.S,
- Lichtenberg G.S,
- et al.
- Kaul S
- Ismail S,
- Jayaweera A.R,
- Goodman N.C,
- Camarano G.P,
- Skyba D.M,
- Kaul S
- Sheehan F.H,
- Bolson E.L,
- Dodge H.T,
- et al.
- Shapiro J.R,
- Reisner S.A,
- Amico A.F,
- Kelly P.F,
- Meltzer R.S
- Pfeffer M.A,
- Braunwald E
- Ito H,
- Iwakura K,
- Oh H,
- et al.
- Villanueva F.S,
- Camarano G,
- Ismail S,
- Goodman N.C,
- Sklenar J,
- Kaul S
- Kloner R.A
- Haronian H.L,
- Remetz M.S,
- Sinusas A.J,
- et al.
- McCallister B.D,
- Christian T.F,
- Gersh B.J,
- Gibbons R.J
- Grayburn P.A,
- Erickson J.M,
- Escobar J,
- Womack L,
- Velasco C.E