Author + information
- Received December 30, 2001
- Revision received May 9, 2002
- Accepted August 20, 2002
- Published online March 5, 2003.
- Tomasz Kukulski, MD*,
- Fadi Jamal, MD*,
- Lieven Herbots, MD*,
- Jan D’hooge, PhD*,
- Bart Bijnens, PhD*,
- Liv Hatle, FESC*,
- Ivan De Scheerder, MD, PhD* and
- George R Sutherland, FESC*,* ()
- ↵*Reprint requests and correspondence:
Dr. George R. Sutherland, University Hospital Gasthuisberg, Department of Cardiology, Herestraat 49, B-3000 Leuven, Belgium.
Objectives The goal of this study was to investigate whether the changes in myocardial deformation measured with ultrasonic strain could accurately identify acutely ischemic myocardium during coronary angioplasty.
Background Early identification of acute myocardial ischemia has important clinical implications. The accuracy of ultrasonic strain for the detection of acute myocardial ischemia has been validated in animal experiments but has not been investigated in the clinical setting.
Methods In 73 patients (64 ± 12 years), either radial or longitudinal strain values were monitored in the “at-risk” segments before, during, and early after right, circumflex, and left anterior descending coronary angioplasty. Based on the visual wall motion assessed before the angioplasty, segments were divided into normokinetic (group I) and hypo/akinetic (group II). Strain data in the “at-risk ” segments were compared with values derived from the adjacent nonischemic segments and normal values in 20 controls.
Results Coronary occlusion induced a marked reduction in the systolic strain both in the radial (from 49 ± 6.9% to 23 ± 4.6% in group I and from 21.9 ± 11% to 11.3 ± 8.4% in group II, p < 0.001) and longitudinal directions. Concomitantly, postsystolic strain increased (from 3.8 ± 3.1% to 14.6 ± 9.5% in group I, and from 4.4 ± 3.7% to 11.3 ± 7.8% in group II in radial direction, p < 0.001). Upon reperfusion, all deformation parameters returned to near preocclusion values. In comparison with control, baseline, and reperfusion data, the systolic and postsystolic strain parameters measured during total coronary occlusion identified acutely ischemic myocardium with a sensitivity of 86% to 95% and a specificity of 83% to 89%.
Conclusions In this model of acute ischemia, ultrasonic strain indexes differentiate acutely ischemic segments from both normal and dysfunctional myocardium. This should be a promising new approach to the bedside monitoring of acute ischemic changes in regional myocardial function.
The early identification of acutely ischemic myocardium has both important therapeutic and prognostic implications (1). The routine diagnosis of acute myocardial ischemia is currently based on the combination of clinical, electrocardiographic (ECG), and biochemical markers (2). However, typical chest pain and ECG changes occur only in a limited number of patients with established acute myocardial ischemia (3). Furthermore, the specific rise in the blood-sampled cardiac enzymes is delayed in time in relation with the onset of the ischemic insult and, when present, already reflects myocardial cellular damage (2). In the temporal sequence of the ischemic cascade, regional wall motion abnormalities appear early after flow reduction (4). Despite several attempts to noninvasively quantify ischemia-induced changes in myocardial contractility (5–10), the routine clinical evaluation of regional function is still based on the visual assessment of wall motion. This qualitative approach has well-documented limitations (11–13)and may fail to identify subtle ischemia-induced changes in regional mechanics.
With the introduction of color Doppler myocardial imaging (CDMI), it is currently possible to noninvasively quantify in-plane myocardial deformation (i.e., regional strain) (14,15). The accuracy of such ultrasonic strain measurements has been recently validated in a comparison with sonomicrometry in an experimental ischemia model (16), and ranges of normal values have been lately established in healthy volunteers (17). Furthermore, deformation indexes appear to be less affected by global cardiac motion and segmental tethering than myocardial velocities (16,18). We postulated that ultrasonic strain could quantify the changes in regional myocardial function during acute ischemia in humans. Thus, the aim of this study was to investigate the value of ultrasonic deformation indexes in the identification of acutely ischemic myocardium using routine coronary angioplasty as a clinical model of acute coronary occlusion.
We prospectively selected 90 consecutive patients with stable angina in whom elective coronary angioplasty was to be performed. Patients with either a prior history or ECG evidence of transmural myocardial infarction, previous coronary bypass grafting, atrial or ventricular arrhythmia, pacemaker, bundle branch block, or left ventricular ejection fraction <40% were not to be admitted in the study. After inclusion in the study, five patients were found to have a poor acoustic window, and four others developed ischemia-induced arrhythmias. These nine patients were not included in the subsequent data analysis.
The catheterization was performed using standard techniques. Coronary collateral flow was assessed from angiograms by an expert investigator blinded to the echocardiographic data using the Rentrop classification (19). Due to the potentially protective effect of collateral circulation on myocardial function (8,19), eight patients with a coronary collateral Rentrop grade of 2 or 3 were analyzed as a separate group and were not entered into the quantitative data analysis of the noncollateral patients. In the remaining 73 patients (36 to 88 years; mean, 64 ± 12 years; 17 females), the site of angioplasty was either the right (n = 32), circumflex (n = 25), or left anterior descending coronary artery (n = 16). The coronary stenosis severity in the vessel to be monitored during angioplasty was assessed visually and was defined to be >90% in all patients. No patient had evidence of an occluded coronary artery before the angioplasty procedure. Single balloon inflation was carried out in four patients. In the remainder, multiple inflations (two to six) were performed. The duration of each inflation ranged from 30 to 90 s. Intracoronary stents were implanted in 53/73 patients (72%). Immediately after the procedure, the angiographic control demonstrated a residual stenosis of <30% in the dilated artery in all patients. The heart rate averaged 70 ± 11 min−1before coronary angioplasty and increased nonsignificantly during balloon inflation (74 ± 14, p = NS).
Twenty subjects (41 to 76 years; mean, 57 ± 11 years; five females) served as a control group. All had normal resting ECGs, echocardiograms, and ECG treadmill tests. Both patients and normal subjects gave informed consent before participation in the study, which had been approved by the institutional ethics committee.
Echocardiographic data acquisition
Echocardiographic studies were performed with the patients laying in a supine position. Data was acquired using a Vingmed System 5 (GE Vingmed, Horten, Norway) and a 2.5 MHz probe. B-mode CDMI myocardial velocity data was acquired at a frame rate of 147 s−1using an imaging sector angle of 45°.
A parasternal short-axis or apical two-, four-chamber view was used for the assessment of either radial or longitudinal function, respectively. The posterobasal segment, imaged in the short-axis view, was considered “at-risk” during right or circumflex coronary angioplasty (20). The inferior mid- and basal segments imaged in the apical two-chamber view were considered “at-risk” segments during right coronary artery percutaneous transluminal coronary angioplasty (PTCA). Both the septal mid- and apical segments, imaged in an apical four-chamber view, were considered “at-risk” during left anterior descending coronary angioplasty. Baseline data on “at-risk” segment function were acquired immediately before the angioplasty procedure. Occlusive data were collected 30 to 60 s after the onset of the first balloon inflation. Reperfusion data were recorded 120 s after balloon deflation.
Echocardiographic data analysis
Strain data were postprocessed from color myocardial velocity loops with dedicated software (TVI 6.0, GE Vingmed) (14,21). All steps of analysis were described in detail in the previous papers (17,18). For strain computation, end-diastole was chosen as the reference time-point. This was defined to occur at the R-top on the ECG trace. The timing of end-systole (aortic valve closure [AVC]) was derived from the anteroseptal myocardial velocity profile (Fig. 1). This mechanical event correlates very well with timing of peak negative dP/dt (first temporal derivative of left ventricular pressure) (22). End-systolic strain (ϵSYS) was defined to be the magnitude of deformation between the end-diastolic reference time-point to the end of systole (i.e., AVC). Peak strain (ϵPEAK) was defined to be the highest strain value obtained for the radial direction and the lowest strain value for the longitudinal direction over a mean RR interval. Postsystolic strain (ϵPS) was calculated as the difference between ϵPEAKand ϵSYS(ϵPS= ϵPEAK− ϵSYS). The time delay from AVC to peak strain (Tϵ) was also calculated. As prior animal studies had shown a combination of decrease in systolic deformation and an increase in postsystolic (early diastolic) deformation to be induced by acute ischemia (23), an additional parameter combining systolic and early diastolic deformation values was calculated (postsystolic strain index [PSI] = (ϵPEAK− ϵSYS)/ϵPEAK) (Speqle, KU Leuven, Belgium) (Fig. 2). This PSI represents the relative amount of ischemia-related segmental thickening/shortening, which was found to occur after AVC. The intra- and interobserver variability (in percent of mean values) ranged from 8% to 12% for all strain parameters.
After subtracting the color myocardial velocity data from the CDMI clips, the underlying digital gray-scale B-mode data could be displayed in cineloop format. Two independent experts carried out the visual assessment of wall motion before the PTCA and myocardial segments were subsequently subdivided into segments with baseline normal (normokinetic, n = 55) and abnormal function (hypo/akinetic, n = 43).
Gray scale and CDMI echocardiogram follow-up data were collected from 51 of the 73 patients 12 months after the PTCA procedure. Regional strain parameters were analyzed for both radial and longitudinal function in both left anterior descending and right coronary artery/circumflex “at-risk” segments. After right circumflex artery PTCA, patients scanned for radial segmental function were subdivided into group 1, with normal pre-PTCA baseline function; group 2, with abnormal baseline function; and group 3, patients in whom ischemic events occurred during the 12-month follow-up period. In the last group, there were two patients who underwent coronary artery bypass grafting, two who had recurrent angina, one who underwent a further PTCA procedure in the same vessel, and one who had a transmural infarction in the “at-risk” territory.
Data are expressed as mean ± SD. For multiple comparisons an analysis of variance along with a post-hoc Scheffé test were performed. A Student ttest was used for paired comparisons. A Mann-Whitney nonparametric test was used for the comparison between collateralized and noncollateralized segments. Statistical significance was inferred when p < 0.05. In order to test the diagnostic accuracy of strain parameters, receiver operating characteristic (ROC) curve analysis was performed. A cutoff value with optimal sensitivity and specificity was determined for each parameter.
For each cardiac cycle, the normal radial systolic thickening of the posterobasal myocardium was characterized by positive strain values (Fig. 1). During diastole, the myocardium thinned, and strain values decreased, reaching the zero value at the end-diastolic reference point. Conversely, in the longitudinal direction, the myocardium shortened during systole. This resulted in negative strain values (Fig. 1) for the longitudinal direction. During diastolic myocardial lengthening, longitudinal strain values returned to zero at end-diastole. The radial ϵSYS(56 ± 13%) differed consistently from longitudinal ϵSYSvalues (−23 ± 5%, p < 0.001 and −19 ± 4%, p < 0.001). The strain indexes derived ⇓for normal controls are presented in Tables 1 to 3. ⇓
In the segments, which were defined to have normal thickening on visual inspection (group I), both radial and longitudinal strain profiles were similar to those observed in controls (Fig. 1 and 3a). ⇓
Conversely, in group II segments with abnormal visual wall motion (Fig. 3b), the strain profile was consistently different from normal segments (Fig. 1). In both the radial and the longitudinal directions, systolic deformation was diminished. In addition, after AVC, the early diastolic deformation was characterized by ongoing postsystolic thickening or shortening. The quantitative deformation parameters are listed in Tables 1 to 3. The ϵSYSwas only slightly decreased in group I in comparison with controls (49.6 ± 6.9% vs. 56.3 ± 11.7% in the radial, p = 0.02, and −21.2 ± 4.5% vs. −23.2 ± 4.7% in the longitudinal direction, p = NS) and markedly reduced in the inferior segments (longitudinal ϵSYS−13.6 ± 3.1% vs. −18.9 ± 3.7% in normal control). In group II segments, ϵSYSwas markedly reduced (21.9 ± 11% in the radial and −5.2 ± 4.5% in the longitudinal direction, p <0.001 vs. normal). The ϵPEAKwas also significantly decreased. In addition, the relative postsystolic deformation (PSI) in group II was significantly higher than in group I and control segments in both the radial and longitudinal directions.
Balloon coronary occlusion induced considerable alteration in the segmental deformation of the “at-risk” myocardium (Fig. 3). Interestingly, systolic strain was reduced, and postsystolic strain increased not only in group I but also in group II segments. However, myocardial bulging characterized by paradoxical mid/late systolic lengthening in the longitudinal direction (positive systolic strain) was observed only in 6 of 16 patients undergoing left anterior descending coronary angioplasty. In the radial direction, none of the patients showed paradoxical systolic thinning during right or circumflex coronary artery angioplasty. In both groups I and II, radial and longitudinal ϵSYSwas decreased by near 50% compared with the baseline values. At the same time, postsystolic deformation increased in absolute values (ϵPS) and even more strikingly in relative values (PSI). Conversely, ϵPEAKwas less affected by the ischemic insult, especially in group II segments. Furthermore, ϵPEAKwas significantly delayed in both groups resulting in a prolonged Tϵ during coronary occlusion in comparison with baseline data. All these changes were consistently observed in the radial and longitudinal directions for all “at-risk” segments (Tables 1 to 3).
Regional function of the adjacent and remote segments
Adjacent to the “at-risk” segments, inferior apical (two-chamber view) and basal septal (four-chamber view) segments were considered as segments not affected by acute ischemia during PTCA balloon occlusion of the right coronary artery and left anterior descending artery, respectively. In these segments ϵSYSdid not change for the duration of the artery occlusion (−16.7 ± 10.2% vs. −15.8 ± 7.5% at baseline and −19.1 ± 7.2% vs. −18.2 ± 8.1% at baseline, respectively, p = NS). Basal-anteroseptal segments (long-axis parasternal view) were considered as remote segments not affected by acute posterior wall ischemia. Also in these segments, ϵSYSremained unchanged during right coronary artery/circumflex occlusion (11.8 ± 7.3% vs. 13.3 ± 7.9% at baseline, p = NS).
Collaterals (grades 1 to 3 according to the Rentrop scale) supplying the “at-risk” area to undergo the angioplasty procedure could be visually identified on the angiograms in 16/57 patients who underwent right coronary artery/circumflex angioplasty and in 7/16 patients who underwent left anterior descending coronary artery angioplasty. Adequate collateral flow (grades 2 to 3) to the right coronary artery/circumflex territory was present in 7/57 patients and to the left anterior descending coronary artery territory in 1/16 patients. During occlusion, right coronary artery/circumflex “at-risk” segments with adequate collateral flow (group A) demonstrated less postsystolic strain (6.3% ± 3.0% vs. 14.1% ± 9.5%, p = 0.01) and lower PSI (0.23 ± 0.16 vs. 0.44 ± 0.22, p = 0.01) compared with noncollateralized segments (group C) (Table 4). The Tϵ appeared to be the strongest discriminator between group A and group C (74 ± 10 ms vs. 106 ± 18 ms, p = 0.0002). These data support previous observations suggesting the protective role of collateral circulation on myocardial function (8,19). The individual response of collateralized and noncollateralized segments during acute coronary occlusion is shown in Figure 4. Note that postsystolic deformation did not change in collateralized segments (group A) during acute coronary occlusion, while, in noncollateralized segments (group C), postsystolic strain increased significantly (3.9 ± 3.5% vs. 14.1 ± 9.5%, p < 0.001).
ROC curve analysis
Figure 5shows the results of ROC curve analysis for strain parameters and the corresponding cutoff values. The PSI appeared as the most accurate parameter in differentiating acutely ischemic segments during coronary occlusion from baseline normal and abnormal segments with a sensitivity of 92% to 95% and specificity of 89%. On the contrary, ϵPEAKdemonstrated a lower diagnostic accuracy in comparison with ϵSYSand PSI.
Follow-up results are summarized in Table 5. In group 1 patients, regional function in the “at-risk” segment did not change at 12 months after PTCA as measured by strain (ϵSYS) and wall motion score (WMS). Group 2 patients had improved their function significantly at one year after PTCA as measured by ϵSYS, although visual assessment did not show a significant WMS improvement. Group 3 patients showed reduction in ϵSYS, while WMS remained unchanged (see Fig. 6for details). Analyzing the individual segmental response to revascularization, 19 of 27 dysfunctional segments (70%) showed an improvement as measured with ϵSYSwhile, using eyeball assessment, only 15/27(55%) demonstrated functional improvement. An increase of more than one standard deviation in ϵSYSwas used in order to classify functional improvement in the “at-risk” segment.
This study has confirmed the ability of ultrasonic strain indexes to differentiate acutely ischemic myocardium, not only from normal but also from dysfunctional myocardium. To our knowledge, for the first time this could be achieved noninvasively using a bedside imaging technique.
Changes in regional myocardial function during acute ischemia
An acute reduction in regional myocardial blood flow induces, within a few seconds, a local contractile dysfunction, which results in an alteration in the regional deformation pattern (24–26). During systole, the radial thickening and circumferential/longitudinal shortening of the ischemic segment are decreased. In addition, the segmental relaxation is considerably impaired during the ischemic insult, and, as a consequence, the physiologic early diastolic thinning and lengthening are replaced by ongoing postsystolic thickening and shortening (23,25). Such consistent changes in early diastolic deformation have been proposed as an early marker of regional ischemia (27,28). In our study, it was the combination of systolic and diastolic strain indexes (which reflect the acute alterations in both local contraction and relaxation) that appeared as the most accurate method for identifying severely ischemic segments.
We excluded from the study patients with transmural myocardial infarction and recent episodes of unstable angina. Nevertheless, wall motion was already abnormal in 43 of the 98 analyzed “at-risk” segments. Deformation indexes in these segments were further impaired during coronary occlusion. Although we could not establish whether the impaired function of these segments was related to myocardial hibernation, stunning, nontransmural infarction, or ischemia at rest (because we did not correlate our baseline findings with either perfusion imaging [contrast echocardiography or single-photon emission tomography using Tc-99m sestamibi or a low-dose dobutamine provocation test]), we demonstrated functional improvement in 19 of 27 (70%) “at-risk” dyssynergic segments evaluated 12 months after PTCA, thus confirming that these segments were viable. The potential differences in the response to an acute ischemic insult of each of these pathophysiologic entities require additional investigation and were not the aim of our study.
Other quantitative imaging techniques used for the characterization of acute ischemia
The attempt to quantify the ischemia-induced changes in regional myocardial function using noninvasive imaging is not new (5–10,29). In an experimental model of acute coronary occlusion, Kerber et al. (30)have shown the potential value of gray-scale M-mode for the assessment of segmental asynergy. Myocardial deformation data could be derived from gray-scale M-mode using a manual tracing of endocardial and epicardial borders (8), but only for radial deformation. Myocardial Doppler radial velocities are consistently altered during PTCA (7), and reflect changes in myocardial motion but not in true regional deformation. For the longitudinal function, Henein et al. (31)have used the mitral ring displacement assessed with gray-scale M-mode to investigate the changes in myocardial wall longitudinal shortening during coronary angioplasty. In contrast with our results, these authors did not detect any increase in the longitudinal postsystolic shortening during balloon inflation in patients with stable angina. This difference could be explained by the fact that mitral ring displacement does not reflect the changes in the individual deformation of each myocardial segment as assessed with CDMI-derived regional strain (15).
Coronary angioplasty was used as a clinical model of acute coronary occlusion to simulate the occurrence of acutely ischemic events. The natural history of acute myocardial infarction and unstable angina involves subtotal, total, intermittent coronary thrombosis or spasm, and potentially longer periods of ischemia than those investigated here. Therefore, our results should be transposed with caution to the clinical situations where acute ischemia is potentially involved, and further investigation is required. The angle dependency of ultrasonic strain has already been established (16). Furthermore, the CDMI technique allows only one-dimensional in-plane strain to be quantified. Currently, all three-dimensional components of myocardial deformation cannot be measured using ultrasonic techniques. However, despite these potential limitations, our results showed that one-dimensional high-temporal resolution strain could already give valuable information on regional myocardial mechanics and function. Another important fact is the need to precisely identify AVC because, in the absence of this, the quantitation of postsystolic events is limited.
The identification and quantitation of myocardial deformation abnormalities could serve as a valuable adjunct to the conventional diagnostic approach to acute coronary syndromes. Furthermore, ultrasonic strain data acquisition is totally noninvasive and safe and is suitable for the repetitive monitoring of myocardial function at short intervals after reperfusion therapy or coronary revascularization. This could lead to an early diagnosis of acute regional ischemia and, therefore, may help in limiting the extent of myocardial necrosis. For the long-term observation, with new ultrasonic strain parameters, we could, better than using visual assessment, distinguish between segments that benefited and did not benefit from PTCA.
Using a new quantitative imaging tool, ultrasonic strain, we have defined a specific pattern of acutely ischemic changes within the “at-risk” segment in either left anterior descending or right coronary artery/circumflex territory for both radial and longitudinal direction. We observed significant changes in regional deformation (reduction of ϵSYS) during acute ischemia even in segments that did not deteriorate visually. Quantitative analysis of myocardial deformation appeared to be superior over conventional visual analysis of contractility for the evaluation of late beneficial effects of percutaneous revascularization (Appendix).
The division of segments into groups is summarized below:
Group I = segments with normal pre-PTCA function assessed visually;
Group II = segments with abnormal pre-PTCA function assessed visually;
Group A = collateralized segments graded as 2 to 3 according to the Rentrop scale;
Group B = collateralized segments graded as 1 according to the Rentrop scale;
Group C = noncollateralized segments;
Group 1 = follow-up patients with normal segmental function pre-PTCA;
Group 2 = follow-up patients with abnormal segmental function pre-PTCA;
Group 3 = follow-up patients with ischemic events occurring through one-year follow-up.
- aortic valve closure
- color Doppler myocardial imaging
- peak strain
- postsystolic strain
- end-systolic strain
- postsystolic strain index
- percutaneous transluminal coronary angioplasty
- receiver operating characteristic
- time delay from aortic valve closure to peak strain
- wall motion score
- Received December 30, 2001.
- Revision received May 9, 2002.
- Accepted August 20, 2002.
- American College of Cardiology Foundation
- Rawles J.M.
- Braunwald E.,
- Antman E.,
- Beasley J.,
- et al.
- Hauser A.M.,
- Gangadharan V.,
- Ramos R.G.,
- Gordon S.,
- Timmis G.C.
- Dupouy P.,
- Geschwind H.,
- Pelle G.,
- et al.
- Koh T.W.,
- Carr-White G.S.,
- Delouse A.C.,
- Ferdinand F.D.,
- Pepper J.R.,
- Gibson D.G.
- Koch R.,
- Lang R.M.,
- Garcia M.J.,
- et al.
- Hoffmann R.,
- Lethen H.,
- Marwick T.,
- et al.
- Kvitting J.P.,
- Wigstrom L.,
- Strotmann J.M.,
- Sutherland G.R.
- Urheim S.,
- Edvardsen T.,
- Torp H.,
- Angelsen B.,
- Smiseth O.
- Cohen M.,
- Rentrop K.P.
- Stamm R.B.,
- Gibson R.S.,
- Bishop H.L.,
- Carabello B.A.,
- Beller G.A.,
- Martin R.P.
- D’hooge J.,
- Heimdal A.,
- Jamal F.,
- et al.
- Takayama M.,
- Norris R.M.,
- Brown M.A.,
- Armiger L.C.,
- Rivers J.T.,
- White H.D.
- Tennant R.,
- Wiggers C.J.
- Theroux P.,
- Franklin D.,
- Ross J. Jr..,
- Kemper W.S.
- Villarreal F.J.,
- Lew W.Y.,
- Waldman L.K.,
- Covell J.W.
- Safwat A.,
- Leone B.J.,
- Norris R.M.,
- Foëx P.
- Derumeaux G.,
- Ovize M.,
- Loufoua J.,
- Pontier G.,
- André Fouet X.,
- Cribier A.
- Kerber R.E.,
- Marcus M.L.,
- Wilson R.,
- Ehrhardt J.,
- Abboud F.M.
- Henein M.Y.,
- O’Sullivan C.,
- Davies S.W.,
- Sigwart U.,
- Gibson D.G.