- Email: [email protected]
Assessment of regional and transmural myocardial perfusion by means of intraoperative myocardial contrast echocardiography during coronary artery bypass grafting
Assessment of regional and transmural myocardial perfusion by means of intraoperative myocardial contrast echocardiography during coronary artery bypass grafting By using intraoperative myocardial contrast echocardiography, we assessed regional myocardial perfusion and transmural blood flow distribution immediately after myocardial revascularization. A total of 62 revascularized myocardial areas were studied in 31 patients undergoing coronary artery bypass grafting. The revascularized areas were divided into three different areas: S area, supplied by significantly stenosed coronary arteries (43 areas); C area, supplied by coronary collateral situation associated with totally occluded coronary arteries (12 areas); MI area, preexisting transmural myocardial infarction (7 areas). Myocardial contrast echocardiography was obtained by direct injection of 2 mI of sonicated 5 % human albumin into the saphenous vein grafts at rest and during atrial pacing. Each area was divided into two layers of endocardial and epicardial halves, and myocardial enhancement of peak intensity was measured for each half and endocardial/epicardial gray level ratio was calculated: (1) The peak intensity of myocardial enhancement in S area and C area was significantly higher than that in MI area at rest as well as during pacing after myocardial revascularization. There was no significant difference in the peak intensity between S area and C area both at rest and during pacing. In S area the peak intensity significantly increased during pacing (p < 0.01), whereas it did not change in C area and MI area. (2) S area demonstrated no significant change in endocardial/epicardial intensity ratio during pacing. In contrast, the ratio in C area significantly decreased during pacing. (3) In S area with preoperative percent increase of segmental wall thickening lower than 25 %, there was a significant correlation (r = 0.84, p < 0.001) between the peak intensity of myocardial enhancement and the postoperative changes of percent increase of segmental wall thickening in the revascularized areas. Thus, immediately after myocardial revascularization, intraoperative myocardial contrast echocardiography could provide a quantitative assessment of regional myocardial perfusion as well as blood flow distribution in the areas with myocardial infarction and with coronary collateral situation and in the areas supplied by stenosed coronary arteries. (J THORAC CARDIOVASC SURG 1992;104:1158-66)
Nobuaki Hirata, MD, Susumu Nakano, MD, Kazuhiro Taniguchi, MD, Mitsunori Kaneko, MD, Ryousuke Matsuwaka, MD, Toshiki Takahashi, MD, Kei Sakai, MD, Yasuhisa Shimazaki, MD, Hikaru Matsuda, MD, and Yasunaru Kawashima, MD, Osaka, Japan
During coronary artery bypass operations, bypass graft blood flow is often measured by electromagnetic From First Department of Surgery, Osaka University Medical School, Osaka, Japan. Received for publication Sept. 14, 1990. Accepted for publication Sept. 24,1991. Address for reprints: Hikaru Matsuda, MD, First Department of Surgery, Osaka University Medical School, I-I-50, Fukushima, Fukushima-ku, Osaka, Japan.
12/1/34009
I I 58
flowmeter to confirm establishment of new coronary blood flow. The measurement of the graft blood flow by the flowmeter, however, does not provide precise information about regional myocardial perfusion in the revascularized myocardium. Up to now no method has been available to directly evaluate the effectivenessof myocardial revascularization during operations. Recently myocardial contrast echocardiography has been developedto evaluate regional myocardial perfusion and transmural blood flow distribution.l' Goldman," Spotnitz.' Kabas,"
Volume 104 Number 4 October 1992
Intraoperative myocardial contrast echocardiography
I I 59
Fig. 1. Myocardial contrast echocardiogram (left ventricular short-axis view at the midpapillary muscle level) obtained by contrast injection into a left anterior descending coronary artery graft. Left, Control image obtained before contrast injection; right, peak contrast enhancement of anterolateral left ventricle after injection. Lower panel, Endocardial and epicardial peak intensity of enhancement per pixel.
and their coworkers have reported preliminary experiences with intraoperative use of myocardial contrast echocardiography in evaluating bypass grafting results. The present study was designed, using intraoperative myocardial contrast echocardiography, to assess regional myocardial perfusion and transmural distribution of new coronary blood flow in the revascularized myocardium, especially with coronary collateral flow and with preexisting transmural myocardial infarction.
Patients and methods A total of 62 revascularized myocardial areas in 31 patients undergoing coronary artery bypass grafting between February 1989 and February 1990 were studied by myocardial contrast echocardiographyduring the bypass operations. The 31 patients included 4 female and 27 male patients whose average age was 61 ± 13 years. Informed consent for the procedure was obtained from all patients, and preoperative coronary angiography was carried out on all of them. The operative techniques of coronary bypass grafting used included cardiopulmonary bypass under moderate hypothermia with cold potassium cardioplegic solution used for topical cooling for myocardial
protection. All patients underwent coronary bypass grafting with saphenous veins or internal mammary arteries, or both. Only saphenous veins were studied. The saphenous vein grafts were used to revascularize 14 anterior descending arteries, 12 diagonal branches, 19 left circumflex arteries, and 19 right coronary arteries. The graft blood flow was measured at rest by electromagnetic flowmeter (model MFV-3100, Nihon Kohden Corp., Tokyo, Japan) with 3 mm, 4 mm, or 5 mm flow probes, according to the respective diameters of the 48 individual saphenous vein grafts. Sixty-two revascularized areas studied by contrast echocardiography were divided into three different areas: 43 areas supplied by significantly (greater than 75%) stenosed coronary arteries (S area); 12 areas supplied by coronary collateral flow associated with totally occluded coronary arteries (C area); 7 areas with preexisting transmural myocardial infarction (anteroseptal, 2; inferior, 5) (MI area), which wereangiographically documented before the operation. A preexisting transmural myocardial infarction was defined by the existence of abnormal Q waveson the electrocardiogram and by clinical evidence. Intraoperative myocardial contrast echocardiography. When postoperative hemodynamics became stable after termination of cardiopulmonary bypass, two-dimensional echocardiographic images were obtained using a commercially avail-
1 1 6 0 Hirata et al.
The Journal of Thoracic and Cardiovascular Surgery
Fig. 2. Myocardial contrast echocardiogram obtained by contrast injection into a circumflex posterolateral graft.
Left, Before contrast injection; right, peak contrast enhancement of posterior left ventricle. Lower panel, As explained in Fig. I. able phased-array system (model SSH-270A ultrasound system, Toshiba Corp., Tokyo, Japan) with a 3.75 MHz transducer. Myocardial contrast echocardiography was performed by imaging an epicardial short-axis view at the midpapillary muscle levelof the left ventricle and by a bolus injection of 2 ml of 5% sonicated human albumin directly into the individual saphenous vein grafts through a 24-gauge catheter. The 24-gauge catheter was inserted from the branch or a puncture near the branch of the saphenous vein grafts. Myocardial contrast echocardiography was obtained at rest and during atrial pacing. Pacing was performed in 32 revascularized myocardial areas (S area, 20; C area, 7; MI area, 5) of 15 patients and set at 1.3 times the heart rate before pacing. Sonication was performed with a commercially available sonicating system (Sonifier model 250, Branson Corporation, Danbury, Conn.). The mean size of microbubbles prepared with sonicated 5% human albumin was 4.6 ± 2.6 (1.8 to 9.0) JLm. The size was observed directly through polarized light microscopy (model Biophot VBD-FT-I, Nikon Corp., Tokyo, Japan). Echocardiographic images were recorded by a videotape recorder (model BR 8610, Victor Corp., Tokyo, Japan) from approximately 10 seconds before injection of the contrast agent until the contrast enhancement was no longer evident. Gain settings were adjust-
ed at the beginning of each recording and were not changed during the remainder of the study, so that all images were recorded at the same gain setting. Analysis of myocardial contrast echocardiography. We analyzed the contrast echocardiographic images according to the method of Lim and colleagues.' A commercially available microprocessor-based offline echocardiographic view system was used that consisted of a personal computer (model PC-9801, NEC Corp., Tokyo, Japan) and a high-speed image processor capable of digitizing the echocardiographic field (models 68322 and 64000, Nexus Corp., Tokyo, Japan). This system was used to convert two-dimensional echocardiographic images on the videotape to a 512 X 512 pixel matrix image with 256 gray levels per pixel and to quantify the intensity of echocardiographic signals in the regions of interest. Enddiastolic short-axis images at the midpapillary muscle level of the left ventricle were used for analysis. End-diastole was defined as the point of the peak of the electrocardiographic R wave. The region of interest was defined as the area where the intensity was enhanced by injection of the contrast agent into the individual saphenous vein grafts. So that regional myocardial blood flow distribution could be assessed, each area was divided into two layers of epicardial and endocardial halves, and the
Volume 104 Number 4 October 1992
Intraoperative myocardial contrast echocardiography
I I6I
Fig. 3. Myocardial contrast echocardiogram obtained by contrast injection into a right coronary artery graft to an area with preexisting transmural myocardial infarction. Lower panel, As explained in Fig. I. intensitieswere measured for each half of each area. Regions of interest were traced by hand. Each value of the intensity of myocardial enhancement was used as the value subtracted from the value before injection of the contrast agent, and peak intensity was determined from time-intensity curve. The ratio of endocardial to epicardial intensity, endocardial/epicardial intensity ratio, was calculated. Figs. I to 3 present representative intraoperative two-dimensional images of myocardial enhancement obtained by injection of contrast medium into the individual saphenous vein graft in the revascularized areas with and without preexisting myocardial infarction before myocardial revascularization. Analysis of regional wall motion. So that regional wall motion in the grafted areas could be assessed quantitatively, the percent increase in segmental wall thickening during systolewas used in the two-dimensional echocardiographic images of the short-axis midpapillary muscle level of the left ventricle, in which myocardial contrast echocardiography was performed. After tracing of endocardium and epicardium, wall thickness was measured at the center of the segment of myocardial enhancement on the end-systolic and end-diastolic images, and the percent increase in segmental wall thickening during systole was calculated. The regional wall motion was assessed, with no administration of catecholamine for the maintenance of hemodynamics after myocardial revascularization, in 44 revascularized areas(S area, 27; C area, 9; MI area, 6)of22 patients. This examination was performed 30 minutes before the beginning of cardiopulmonary bypass and again simultaneously with the myocardial contrast echocardiography after termination of
cardiopulmonary bypass. The examined areas were defined as the same grafted areas assessed by myocardial contrast echocardiography. Reproducibility of results. Intraobserver and interobserver variability was determined by measuring the peak intensity. The intraobserver correlation was y = 1.04x - 0.87; r = 0.98, and the interobserver correlation was y = 0.98x - 0.56, r = 0.98. The reproducibility of our measurement of the peak intensity was assessed in 20 patients by repeated injection. There was no difference between the values (35 ± 13 versus 37 ± 8),andthe absolute difference was 3 ± 2. Statistical analysis. All data were expressed as mean ± standard deviation. Wilcoxon rank-sum test or Wilcoxon signed-rank test was used to compare the values. Results During myocardial contrast echocardiography, no electrocardiographic or hemodynamic changes were observed in any of the patients studied. Relationship between the graft blood flow and peak intensity of myocardial enhancement by injection of contrast agents into individual saphenous vein grafts in revascularized areas. There was a significant correlation (r = 0.46; P < 0.005) between the graft blood flow and the peak intensity of myocardial enhancement (Fig. 4).
The Journal of Thoracic and Cardiovascular Surgery
1 16 2 Hirata et af.
•
110
•
100
r = 0.42
80 >, .~
70
Q) +-'
60
~
50
til C
c
CIl
Q)
o,
• • •
• ... ...• ... ... • ...
20 10
o
10
20
30
0
•
•
•
...
30
• 0
0
0
40
p
• • ••
90
••
• 0 0
•
• • ••
• S
oc ... MI
40
50
60
70
80
90
100 110 120 (mi/min)
Graft Blood Flow Fig. 4. Relationship between the graft bloodflow measured by electromagnetic flowmeter and the peak intensity of myocardial enhancement in the revascularized areas obtained from myocardialcontrast echocardiography. Table I. Peak intensity after myocardial revascularization, preoperative and postoperative percent increase of segmental wall thickening, and graft blood flow in the preoperative ischemic dysfunction regions as shown by percent increase of segmental wall thickening less than 25% in the S area Graft location 1. 2. 3. 4. 5.
6. 7.
8. 9. 10. 11. 12. 13. 14.
Diag. RCA LCX LCX LCX RCA LCX RCA LAD RCA RCA RCA Diag. LCX
Peak intensity
Preop. % increase (1)
Postop. % increase (2)
(2)-(1)
Graft bloodflow
99
14.5 10.0 10.0 25.0 25.0
49.1 43.2 58.0 37.0 40.0 28.2 35.2
34.6 33.2 48.0 12.0 15.0 7.1
41.7
26.9 23.9 22.5 -3.1 13.1 3.7
20 50 83 65 65 50 50 65 80 70 30 50
79 86 34 47 39 36 80 64 31 13 58
22 54
21.1 15.3 14.8
19.6 20.0 22.9
16.0 24.1 24.8
43.5 42.5
19.8 29.1 27.8 35.3
19.9
10.5
52
Diag., Diagonal branch; RCA, right coronary artery; LCX; left circumflex artery; LAD, left anterior descending artery; S area, areas supplied by significantly stenosed coronary arteries.
Regional myocardial perfusion at rest and during rapid pacing after myocardial revascularization. The peak intensities of myocardial enhancement in S area, 43 areas supplied by significantly stenosed coronary
arteries(at rest, 48 ± 25; during pacing, 58 ± 20), and C area, 12 areas supplied by coronary collateral flow associated with totally occluded coronary arteries (at rest, 51 ± 26; during pacing, 52 ± 23), were signifi-
Volume 104 Number 4 October 1992
Intraoperative myocardial contrast echocardiography
p
80
I I63
< 0.01
II
70
.....>,
60
'Vi cQ)
50
c
40
.....
o _
at rest
during pacing
~
ell
Q)
Q.
30 20 10 S-area
C-area
MI-area
Fig. 5. Peak intensity of myocardial enhancement in the revascularized areas. S-area, Areas supplied by significantly stenosed coronary arteries; C-area, areas supplied by coronary collateral flow associated with totally occluded coronary arteries; MI-area, areas with preexisting transmural myocardial infarction.
cantly higher than the peak intensity of MI area, 7 areas with preexisting transmural myocardial infarction (at rest, 14 ± 4; during pacing, 19 ± 8) at rest as well as during pacing (p < O.OI;p < 0.01). There was no significant difference in the peak intensity between S area and C area at rest. During pacing, however, the peak intensity in S area significantly increased (p < 0.01), whereas that in C area and MI area did not change (Fig. 5). Myocardial blood flow distribution at rest and during rapid pacing after myocardial revascularization. S area demonstrated no significant change in endocardial/ epicardial intensity ratio during pacing (at rest, 0.90 ± 0.13; during pacing, 0.87 ± 0.14). On the other hand, the ratio in C area (at rest, 0.83 ± 0.11; during pacing, 0.70 ± 0.09) significantly decreased during pacing (Fig. 6). Relationship between regional myocardial perfusion and wall motion after myocardial revascularization. Percent increase of segmental wall thickening in S area was from 29% ± 12% (before) to 40% ± 13% after myocardial revascularization. On the other hand, in C area and MI area the values did not change (C area, 22% ± 6% and 25% ± 8%; MI area, 21% ± 6% and 21% ± 6% before and after myocardial revascularization, respectively) (Fig. 7). In the 14 S areas with preoperative percent increase of segmental wall thickening lower than 25% (Table I), there was a significant correlation (r = 0.84;p < 0.001) between the peak intensity of
myocardial enhancement and the postoperative changes of percent increase of segmental wall thickening in the revascularized areas (Fig. 8). Discussion For evaluating the effectiveness of myocardial revascularization, it is mandatory to assess regional myocardial perfusion and transmural distribution of new coronary blood flow in the revascularized myocardium. The measurement of graft flow alone does not provide precise information about regional myocardial perfusion in the revascularized myocardium. Experimental and clinical studies have shown that myocardial contrast echocardiography can visualize regional myocardial perfusion and transmural distribution, and these studies have indicated that the intensity of myocardial echocardiographic contrast enhancement may also be related to the quantity of regional myocardial perfusion." 7-10 Furthermore, Keller and colleagues 11 have demonstrated that sonicated albumin microbubbles (sonicated agents) during myocardial contrast echocardiography behave like intravascular tracers of red cell flow. Therefore we thought that intraoperative use of myocardial contrast echocardiography would enable the assessment of regional myocardial perfusion by measuring the peak intensity of myocardial enhancement by direct injection of sonicated contrast medium into individual saphenous vein grafts. Surgeons could thus evaluate the effectiveness of graft function immediately after
The Journal of Thoracic and Cardiovascular Surgery
1 1 6 4 Hirata et al.
1.1 1.0 0 +-'
ro
~
Q..
w
<,
0 "'0
0.9 0.8 0.7
c w 0.6
0.5 0.4 at rest
during pacing at rest during pacing at rest
S-area
C-area
during pacing
MI-area
Fig. 6. Endocardial/epicardial (Endo/Epi) intensity ratios at rest and during rapid atrial pacing. S-area, Areas supplied by significantly stenosed coronary arteries; C-area, areas supplied by coronary collateral flowassociated with totally occluded coronary arteries; MI-area, areas with preexisting transmural myocardial infarction.
'0.0
c c
80
Q) ~
o :..c 70 m 60
I
P<0.01
+-'
3:
~ +-'
c
Q)
E
'0.0 Q)
(J)
c
50 40 30
en
20
.... o
10
Q)
~
Q)
c
~
? I I f~I I~i ..---before after S-area
before after C-area
before after MI-area
Fig. 7. Regional wall motion before and after myocardial revascularization. S-area, Areas supplied by significantly stenosed coronary arteries; C-area, areas supplied by coronary collateral flow associated with totally occluded coronary arteries; MI-area, areas with preexisting transmural myocardial infarction.
Volume 104 Number 4 October 1992
Intraoperative myocardial contrast echocardiography
100
•
90
80
•
70 60 ~
50
~
40
Q)
Q..
10
•
• •
30
20
I I65
r =0.84
•
p
•
• o
10
20
30
40
50
change in regional wall motion (postop-preop
% increase in segmental wall thickening)
Fig. 8. Relationship between the peakintensity of myocardial enhancement and the percentincrease of segmental wallthickening (preoperatively, in the areas with the percent increase of wall thickening lower than 25% in the S area [areas supplied by significantly stenosed coronary arteriesD.
myocardial revascularization. Quantitatively, regional myocardial perfusion would then be analyzed retrospectively from videotape. There has been no information about transmural perfusion of new coronary blood flow in the revascularized areas, especially in the coronary collateral situation and with preexisting transmural myocardial infarction. This information would be interesting in determining the indication of myocardial revascularization for such areas. The present study provided valuable information regarding the regional and transmural myocardial perfusion after myocardial revascularization. In the present study, regional myocardial perfusion in the revascularized areas supplied by the stenosed coronary arteries (S area) increased during atrial pacing compared with that at rest. Myocardial blood flow distribution as shown by the endocardial/epicardial intensity ratio did not change during atrial pacing. Atrial pacing induces an increase in myocardial oxygen demand, and blood flow is preferentially delivered to the more viable myocardium. The response of myocardial perfusion during atrial pacing in the S area seems favorable, proving
that myocardial revascularization to the S area is effective. On the other hand, regional myocardial perfusion in the revascularized areas supplied by collateral coronary arteries (C area), while being favorably compared with that in the S area at rest, did not increase during atrial pacing. Moreover, the myocardial perfusion in the endocardial halves of these areas substantially decreased during atrial pacing. Such unfavorable responses of myocardial perfusion during atrial pacing indicate underperfusion of subendocardial blood flow in the C area, even after myocardial revascularization. The number of viable myocytes in the C area, especially in the endocardial halves of these areas, is likely fewer than the number in the S area. We speculate subendocardial infarction in the greater portion of the C area, although the intraoperative visual appearance of myocardium both in the S area and in the C area showed no myocardial fibrosis. In the areas of preexisting transmural myocardial infarction (MI area), postoperative myocardial perfusion was significantly lower than that in noninfarcted areas. There was also no significant change in endocardial/epi-
I I66
The Journal of Thoracic and Cardiovascular Surgery
Hirata et al.
cardiaI intensity ratio at rest and during atrial pacing. These findings indicate poor myocardial perfusion in the infarcted areas despite myocardial revascularization. We revascularized the MI area because preoperative examinations documented the likelihood of myocardial viability in these areas. During the operation the areas appeared as islands of fibrosis rather than as scar. The effectiveness of myocardial revascularization on the infarcted areas depends on myocardial viability. We think that myocardial viability may be demonstrated by contrast enhancement, and we further believe that myocardial contrast echocardiography during preoperative coronary angiography would enable the assessment of myocardial viability in the infarcted areas and thus would provide some indication for surgical revascularization to these infarcted areas. Myocardial contrast echocardiography provided simultaneous information on regional myocardial perfusion and regional wall motion of the left ventricle immediately after myocardial revascularization. Regional wall motion improved postoperatively in the S area but not in the C area. This fact was attributable to the difference of myocardial perfusion during atrial pacing between these two groups. Regional wall motion showed little improvement or no change after operation in the areas that had almost normal wall motion at the preoperative study. However, we found that in the poorly contracting areas of the left ventricle with preoperative ischemic dysfunction, there was a good correlation between postoperative improvement in regional wall motion and postoperative myocardial perfusion. Significantly improved regional wall motion was associated with good regional myocardial perfusion after myocardial revascularization. In addition to the demonstrated potential of intraoperative myocardial contrast echocardiography, the greatest ischemic areas to be revascularized may be defined by injection of sonicated agents with cardioplegic solution before grafting. If differential rate and degree of whiting out of sonication could be discerned in detail, we also believe that this method in which cardioplegic solution is used can determine the most valuable vessel to bypass first in multivessel diseases. This determination may be difficult by coronary angiographic findings alone. In conclusion, intraoperative myocardial contrast echocardiography would enable surgeons to assess regional myocardial perfusion in the revascularized myocardium and thus to evaluate the effectiveness of coronary bypass grafting by direct injection of sonicated contrast medium
into individual saphenous vein grafts. The present study suggested that the majority of the areas supplied by coronary collateral flow associated with totally occluded coronary arteries had subendocardial infarction. REFERENCES 1. DeMaria AN, Bommer WJ, RiggsK, DajeeA, Keown 0 L,
2.
3.
4.
5.
6.
7.
8.
9.
10.
II.
Mason DT. Echocardiographicvisualization of myocardial perfusionby left heart and intracoronary injections of echo contrast agents [Abstract]. Circulation 1980;62(pt 2): III 43. Armstrong WF, Mueller TM, Kinney EL, Tickner EG, DillonJC, Feigenbaum H. Assessmentof myocardialperfusion abnormalities with contrast-enhanced two-dimensional echocardiography. Circulation 1982;66: 166-73. Lim YJ, Nanto S, Masuyama T, et al. Visualization of subendocardial myocardial ischemia with myocardialcontrast echocardiography in humans. Circulation 1989;79: 233-44. Goldman ME, Mindich B. Intraoperative cardioplegic contrast echocardiography for assessing myocardialperfusion during open heart surgery. J Am ColI Cardiol 1988; II :59-65. Spotnitz WD, Kaul S. Intraoperative assessmentof myocardial perfusion using contrast echocardiography. Echocardiogr J Cardiovasc Ultrasound Allied Technol 1990;7:209-28. Kabas JS, Kisslo J, FlickCL, et al. Intraoperativeperfusion contrast echocardiography. Initial experienceduring coronary artery bypassgrafting. J THORAC CARDIOVASC SURG 1990;99:536-42. Maurer G, Ong K, Haendchen R, et al. Myocardial contrast two-dimensional echocardiography: comparison of contrast disappearance rates in normal and underperfused myocardium. Circulation 1984;69:418-29. Taylor A, Collins SM, Skorton OJ, Kieso RA, Melton J, Kerber RE. Artifactual regional gray level variability in contrast enhanced two dimensional echocardiographic images: effect on measurement of coronary perfusion bed. J Am Coli Cardiol 1985;6:831-8. Kaul MW, KellyP, Oliner JO, Glasheen WP, KellerMW, Watson DO. Assessmentof regional myocardialbloodflow in-vivo with myocardial contrast two-dimensional echocardiography. J Am Coli Cardiol 1989;13:468-82. Cheirie J, Zoghbi WA, Minor ST, et al. Assessment of myocardial perfusion in humans by contrast echocardiography. I. Evaluation of regional coronary reserve by peak contrast intensity. J Am ColI Cardiol 1988;11 :735-43. Keller MW, Segel SS, Kaul S, During B. The behaviorof sonicated albumin microbubbles within the microcirculation: a basis for their use during myocardial contrast echocardiography. Circ Res 1989;65:458-67.
Nobuaki Hirata, MD, Susumu Nakano, MD, Kazuhiro Taniguchi, MD, Mitsunori Kaneko, MD, Ryousuke Matsuwaka, MD, Toshiki Takahashi, MD, Kei Sakai, MD, Yasuhisa Shimazaki, MD, Hikaru Matsuda, MD, and Yasunaru Kawashima, MD, Osaka, Japan
During coronary artery bypass operations, bypass graft blood flow is often measured by electromagnetic From First Department of Surgery, Osaka University Medical School, Osaka, Japan. Received for publication Sept. 14, 1990. Accepted for publication Sept. 24,1991. Address for reprints: Hikaru Matsuda, MD, First Department of Surgery, Osaka University Medical School, I-I-50, Fukushima, Fukushima-ku, Osaka, Japan.
12/1/34009
I I 58
flowmeter to confirm establishment of new coronary blood flow. The measurement of the graft blood flow by the flowmeter, however, does not provide precise information about regional myocardial perfusion in the revascularized myocardium. Up to now no method has been available to directly evaluate the effectivenessof myocardial revascularization during operations. Recently myocardial contrast echocardiography has been developedto evaluate regional myocardial perfusion and transmural blood flow distribution.l' Goldman," Spotnitz.' Kabas,"
Volume 104 Number 4 October 1992
Intraoperative myocardial contrast echocardiography
I I 59
Fig. 1. Myocardial contrast echocardiogram (left ventricular short-axis view at the midpapillary muscle level) obtained by contrast injection into a left anterior descending coronary artery graft. Left, Control image obtained before contrast injection; right, peak contrast enhancement of anterolateral left ventricle after injection. Lower panel, Endocardial and epicardial peak intensity of enhancement per pixel.
and their coworkers have reported preliminary experiences with intraoperative use of myocardial contrast echocardiography in evaluating bypass grafting results. The present study was designed, using intraoperative myocardial contrast echocardiography, to assess regional myocardial perfusion and transmural distribution of new coronary blood flow in the revascularized myocardium, especially with coronary collateral flow and with preexisting transmural myocardial infarction.
Patients and methods A total of 62 revascularized myocardial areas in 31 patients undergoing coronary artery bypass grafting between February 1989 and February 1990 were studied by myocardial contrast echocardiographyduring the bypass operations. The 31 patients included 4 female and 27 male patients whose average age was 61 ± 13 years. Informed consent for the procedure was obtained from all patients, and preoperative coronary angiography was carried out on all of them. The operative techniques of coronary bypass grafting used included cardiopulmonary bypass under moderate hypothermia with cold potassium cardioplegic solution used for topical cooling for myocardial
protection. All patients underwent coronary bypass grafting with saphenous veins or internal mammary arteries, or both. Only saphenous veins were studied. The saphenous vein grafts were used to revascularize 14 anterior descending arteries, 12 diagonal branches, 19 left circumflex arteries, and 19 right coronary arteries. The graft blood flow was measured at rest by electromagnetic flowmeter (model MFV-3100, Nihon Kohden Corp., Tokyo, Japan) with 3 mm, 4 mm, or 5 mm flow probes, according to the respective diameters of the 48 individual saphenous vein grafts. Sixty-two revascularized areas studied by contrast echocardiography were divided into three different areas: 43 areas supplied by significantly (greater than 75%) stenosed coronary arteries (S area); 12 areas supplied by coronary collateral flow associated with totally occluded coronary arteries (C area); 7 areas with preexisting transmural myocardial infarction (anteroseptal, 2; inferior, 5) (MI area), which wereangiographically documented before the operation. A preexisting transmural myocardial infarction was defined by the existence of abnormal Q waveson the electrocardiogram and by clinical evidence. Intraoperative myocardial contrast echocardiography. When postoperative hemodynamics became stable after termination of cardiopulmonary bypass, two-dimensional echocardiographic images were obtained using a commercially avail-
1 1 6 0 Hirata et al.
The Journal of Thoracic and Cardiovascular Surgery
Fig. 2. Myocardial contrast echocardiogram obtained by contrast injection into a circumflex posterolateral graft.
Left, Before contrast injection; right, peak contrast enhancement of posterior left ventricle. Lower panel, As explained in Fig. I. able phased-array system (model SSH-270A ultrasound system, Toshiba Corp., Tokyo, Japan) with a 3.75 MHz transducer. Myocardial contrast echocardiography was performed by imaging an epicardial short-axis view at the midpapillary muscle levelof the left ventricle and by a bolus injection of 2 ml of 5% sonicated human albumin directly into the individual saphenous vein grafts through a 24-gauge catheter. The 24-gauge catheter was inserted from the branch or a puncture near the branch of the saphenous vein grafts. Myocardial contrast echocardiography was obtained at rest and during atrial pacing. Pacing was performed in 32 revascularized myocardial areas (S area, 20; C area, 7; MI area, 5) of 15 patients and set at 1.3 times the heart rate before pacing. Sonication was performed with a commercially available sonicating system (Sonifier model 250, Branson Corporation, Danbury, Conn.). The mean size of microbubbles prepared with sonicated 5% human albumin was 4.6 ± 2.6 (1.8 to 9.0) JLm. The size was observed directly through polarized light microscopy (model Biophot VBD-FT-I, Nikon Corp., Tokyo, Japan). Echocardiographic images were recorded by a videotape recorder (model BR 8610, Victor Corp., Tokyo, Japan) from approximately 10 seconds before injection of the contrast agent until the contrast enhancement was no longer evident. Gain settings were adjust-
ed at the beginning of each recording and were not changed during the remainder of the study, so that all images were recorded at the same gain setting. Analysis of myocardial contrast echocardiography. We analyzed the contrast echocardiographic images according to the method of Lim and colleagues.' A commercially available microprocessor-based offline echocardiographic view system was used that consisted of a personal computer (model PC-9801, NEC Corp., Tokyo, Japan) and a high-speed image processor capable of digitizing the echocardiographic field (models 68322 and 64000, Nexus Corp., Tokyo, Japan). This system was used to convert two-dimensional echocardiographic images on the videotape to a 512 X 512 pixel matrix image with 256 gray levels per pixel and to quantify the intensity of echocardiographic signals in the regions of interest. Enddiastolic short-axis images at the midpapillary muscle level of the left ventricle were used for analysis. End-diastole was defined as the point of the peak of the electrocardiographic R wave. The region of interest was defined as the area where the intensity was enhanced by injection of the contrast agent into the individual saphenous vein grafts. So that regional myocardial blood flow distribution could be assessed, each area was divided into two layers of epicardial and endocardial halves, and the
Volume 104 Number 4 October 1992
Intraoperative myocardial contrast echocardiography
I I6I
Fig. 3. Myocardial contrast echocardiogram obtained by contrast injection into a right coronary artery graft to an area with preexisting transmural myocardial infarction. Lower panel, As explained in Fig. I. intensitieswere measured for each half of each area. Regions of interest were traced by hand. Each value of the intensity of myocardial enhancement was used as the value subtracted from the value before injection of the contrast agent, and peak intensity was determined from time-intensity curve. The ratio of endocardial to epicardial intensity, endocardial/epicardial intensity ratio, was calculated. Figs. I to 3 present representative intraoperative two-dimensional images of myocardial enhancement obtained by injection of contrast medium into the individual saphenous vein graft in the revascularized areas with and without preexisting myocardial infarction before myocardial revascularization. Analysis of regional wall motion. So that regional wall motion in the grafted areas could be assessed quantitatively, the percent increase in segmental wall thickening during systolewas used in the two-dimensional echocardiographic images of the short-axis midpapillary muscle level of the left ventricle, in which myocardial contrast echocardiography was performed. After tracing of endocardium and epicardium, wall thickness was measured at the center of the segment of myocardial enhancement on the end-systolic and end-diastolic images, and the percent increase in segmental wall thickening during systole was calculated. The regional wall motion was assessed, with no administration of catecholamine for the maintenance of hemodynamics after myocardial revascularization, in 44 revascularized areas(S area, 27; C area, 9; MI area, 6)of22 patients. This examination was performed 30 minutes before the beginning of cardiopulmonary bypass and again simultaneously with the myocardial contrast echocardiography after termination of
cardiopulmonary bypass. The examined areas were defined as the same grafted areas assessed by myocardial contrast echocardiography. Reproducibility of results. Intraobserver and interobserver variability was determined by measuring the peak intensity. The intraobserver correlation was y = 1.04x - 0.87; r = 0.98, and the interobserver correlation was y = 0.98x - 0.56, r = 0.98. The reproducibility of our measurement of the peak intensity was assessed in 20 patients by repeated injection. There was no difference between the values (35 ± 13 versus 37 ± 8),andthe absolute difference was 3 ± 2. Statistical analysis. All data were expressed as mean ± standard deviation. Wilcoxon rank-sum test or Wilcoxon signed-rank test was used to compare the values. Results During myocardial contrast echocardiography, no electrocardiographic or hemodynamic changes were observed in any of the patients studied. Relationship between the graft blood flow and peak intensity of myocardial enhancement by injection of contrast agents into individual saphenous vein grafts in revascularized areas. There was a significant correlation (r = 0.46; P < 0.005) between the graft blood flow and the peak intensity of myocardial enhancement (Fig. 4).
The Journal of Thoracic and Cardiovascular Surgery
1 16 2 Hirata et af.
•
110
•
100
r = 0.42
80 >, .~
70
Q) +-'
60
~
50
til C
c
CIl
Q)
o,
• • •
• ... ...• ... ... • ...
20 10
o
10
20
30
0
•
•
•
...
30
• 0
0
0
40
p
• • ••
90
••
• 0 0
•
• • ••
• S
oc ... MI
40
50
60
70
80
90
100 110 120 (mi/min)
Graft Blood Flow Fig. 4. Relationship between the graft bloodflow measured by electromagnetic flowmeter and the peak intensity of myocardial enhancement in the revascularized areas obtained from myocardialcontrast echocardiography. Table I. Peak intensity after myocardial revascularization, preoperative and postoperative percent increase of segmental wall thickening, and graft blood flow in the preoperative ischemic dysfunction regions as shown by percent increase of segmental wall thickening less than 25% in the S area Graft location 1. 2. 3. 4. 5.
6. 7.
8. 9. 10. 11. 12. 13. 14.
Diag. RCA LCX LCX LCX RCA LCX RCA LAD RCA RCA RCA Diag. LCX
Peak intensity
Preop. % increase (1)
Postop. % increase (2)
(2)-(1)
Graft bloodflow
99
14.5 10.0 10.0 25.0 25.0
49.1 43.2 58.0 37.0 40.0 28.2 35.2
34.6 33.2 48.0 12.0 15.0 7.1
41.7
26.9 23.9 22.5 -3.1 13.1 3.7
20 50 83 65 65 50 50 65 80 70 30 50
79 86 34 47 39 36 80 64 31 13 58
22 54
21.1 15.3 14.8
19.6 20.0 22.9
16.0 24.1 24.8
43.5 42.5
19.8 29.1 27.8 35.3
19.9
10.5
52
Diag., Diagonal branch; RCA, right coronary artery; LCX; left circumflex artery; LAD, left anterior descending artery; S area, areas supplied by significantly stenosed coronary arteries.
Regional myocardial perfusion at rest and during rapid pacing after myocardial revascularization. The peak intensities of myocardial enhancement in S area, 43 areas supplied by significantly stenosed coronary
arteries(at rest, 48 ± 25; during pacing, 58 ± 20), and C area, 12 areas supplied by coronary collateral flow associated with totally occluded coronary arteries (at rest, 51 ± 26; during pacing, 52 ± 23), were signifi-
Volume 104 Number 4 October 1992
Intraoperative myocardial contrast echocardiography
p
80
I I63
< 0.01
II
70
.....>,
60
'Vi cQ)
50
c
40
.....
o _
at rest
during pacing
~
ell
Q)
Q.
30 20 10 S-area
C-area
MI-area
Fig. 5. Peak intensity of myocardial enhancement in the revascularized areas. S-area, Areas supplied by significantly stenosed coronary arteries; C-area, areas supplied by coronary collateral flow associated with totally occluded coronary arteries; MI-area, areas with preexisting transmural myocardial infarction.
cantly higher than the peak intensity of MI area, 7 areas with preexisting transmural myocardial infarction (at rest, 14 ± 4; during pacing, 19 ± 8) at rest as well as during pacing (p < O.OI;p < 0.01). There was no significant difference in the peak intensity between S area and C area at rest. During pacing, however, the peak intensity in S area significantly increased (p < 0.01), whereas that in C area and MI area did not change (Fig. 5). Myocardial blood flow distribution at rest and during rapid pacing after myocardial revascularization. S area demonstrated no significant change in endocardial/ epicardial intensity ratio during pacing (at rest, 0.90 ± 0.13; during pacing, 0.87 ± 0.14). On the other hand, the ratio in C area (at rest, 0.83 ± 0.11; during pacing, 0.70 ± 0.09) significantly decreased during pacing (Fig. 6). Relationship between regional myocardial perfusion and wall motion after myocardial revascularization. Percent increase of segmental wall thickening in S area was from 29% ± 12% (before) to 40% ± 13% after myocardial revascularization. On the other hand, in C area and MI area the values did not change (C area, 22% ± 6% and 25% ± 8%; MI area, 21% ± 6% and 21% ± 6% before and after myocardial revascularization, respectively) (Fig. 7). In the 14 S areas with preoperative percent increase of segmental wall thickening lower than 25% (Table I), there was a significant correlation (r = 0.84;p < 0.001) between the peak intensity of
myocardial enhancement and the postoperative changes of percent increase of segmental wall thickening in the revascularized areas (Fig. 8). Discussion For evaluating the effectiveness of myocardial revascularization, it is mandatory to assess regional myocardial perfusion and transmural distribution of new coronary blood flow in the revascularized myocardium. The measurement of graft flow alone does not provide precise information about regional myocardial perfusion in the revascularized myocardium. Experimental and clinical studies have shown that myocardial contrast echocardiography can visualize regional myocardial perfusion and transmural distribution, and these studies have indicated that the intensity of myocardial echocardiographic contrast enhancement may also be related to the quantity of regional myocardial perfusion." 7-10 Furthermore, Keller and colleagues 11 have demonstrated that sonicated albumin microbubbles (sonicated agents) during myocardial contrast echocardiography behave like intravascular tracers of red cell flow. Therefore we thought that intraoperative use of myocardial contrast echocardiography would enable the assessment of regional myocardial perfusion by measuring the peak intensity of myocardial enhancement by direct injection of sonicated contrast medium into individual saphenous vein grafts. Surgeons could thus evaluate the effectiveness of graft function immediately after
The Journal of Thoracic and Cardiovascular Surgery
1 1 6 4 Hirata et al.
1.1 1.0 0 +-'
ro
~
Q..
w
<,
0 "'0
0.9 0.8 0.7
c w 0.6
0.5 0.4 at rest
during pacing at rest during pacing at rest
S-area
C-area
during pacing
MI-area
Fig. 6. Endocardial/epicardial (Endo/Epi) intensity ratios at rest and during rapid atrial pacing. S-area, Areas supplied by significantly stenosed coronary arteries; C-area, areas supplied by coronary collateral flowassociated with totally occluded coronary arteries; MI-area, areas with preexisting transmural myocardial infarction.
'0.0
c c
80
Q) ~
o :..c 70 m 60
I
P<0.01
+-'
3:
~ +-'
c
Q)
E
'0.0 Q)
(J)
c
50 40 30
en
20
.... o
10
Q)
~
Q)
c
~
? I I f~I I~i ..---before after S-area
before after C-area
before after MI-area
Fig. 7. Regional wall motion before and after myocardial revascularization. S-area, Areas supplied by significantly stenosed coronary arteries; C-area, areas supplied by coronary collateral flow associated with totally occluded coronary arteries; MI-area, areas with preexisting transmural myocardial infarction.
Volume 104 Number 4 October 1992
Intraoperative myocardial contrast echocardiography
100
•
90
80
•
70 60 ~
50
~
40
Q)
Q..
10
•
• •
30
20
I I65
r =0.84
•
p
•
• o
10
20
30
40
50
change in regional wall motion (postop-preop
% increase in segmental wall thickening)
Fig. 8. Relationship between the peakintensity of myocardial enhancement and the percentincrease of segmental wallthickening (preoperatively, in the areas with the percent increase of wall thickening lower than 25% in the S area [areas supplied by significantly stenosed coronary arteriesD.
myocardial revascularization. Quantitatively, regional myocardial perfusion would then be analyzed retrospectively from videotape. There has been no information about transmural perfusion of new coronary blood flow in the revascularized areas, especially in the coronary collateral situation and with preexisting transmural myocardial infarction. This information would be interesting in determining the indication of myocardial revascularization for such areas. The present study provided valuable information regarding the regional and transmural myocardial perfusion after myocardial revascularization. In the present study, regional myocardial perfusion in the revascularized areas supplied by the stenosed coronary arteries (S area) increased during atrial pacing compared with that at rest. Myocardial blood flow distribution as shown by the endocardial/epicardial intensity ratio did not change during atrial pacing. Atrial pacing induces an increase in myocardial oxygen demand, and blood flow is preferentially delivered to the more viable myocardium. The response of myocardial perfusion during atrial pacing in the S area seems favorable, proving
that myocardial revascularization to the S area is effective. On the other hand, regional myocardial perfusion in the revascularized areas supplied by collateral coronary arteries (C area), while being favorably compared with that in the S area at rest, did not increase during atrial pacing. Moreover, the myocardial perfusion in the endocardial halves of these areas substantially decreased during atrial pacing. Such unfavorable responses of myocardial perfusion during atrial pacing indicate underperfusion of subendocardial blood flow in the C area, even after myocardial revascularization. The number of viable myocytes in the C area, especially in the endocardial halves of these areas, is likely fewer than the number in the S area. We speculate subendocardial infarction in the greater portion of the C area, although the intraoperative visual appearance of myocardium both in the S area and in the C area showed no myocardial fibrosis. In the areas of preexisting transmural myocardial infarction (MI area), postoperative myocardial perfusion was significantly lower than that in noninfarcted areas. There was also no significant change in endocardial/epi-
I I66
The Journal of Thoracic and Cardiovascular Surgery
Hirata et al.
cardiaI intensity ratio at rest and during atrial pacing. These findings indicate poor myocardial perfusion in the infarcted areas despite myocardial revascularization. We revascularized the MI area because preoperative examinations documented the likelihood of myocardial viability in these areas. During the operation the areas appeared as islands of fibrosis rather than as scar. The effectiveness of myocardial revascularization on the infarcted areas depends on myocardial viability. We think that myocardial viability may be demonstrated by contrast enhancement, and we further believe that myocardial contrast echocardiography during preoperative coronary angiography would enable the assessment of myocardial viability in the infarcted areas and thus would provide some indication for surgical revascularization to these infarcted areas. Myocardial contrast echocardiography provided simultaneous information on regional myocardial perfusion and regional wall motion of the left ventricle immediately after myocardial revascularization. Regional wall motion improved postoperatively in the S area but not in the C area. This fact was attributable to the difference of myocardial perfusion during atrial pacing between these two groups. Regional wall motion showed little improvement or no change after operation in the areas that had almost normal wall motion at the preoperative study. However, we found that in the poorly contracting areas of the left ventricle with preoperative ischemic dysfunction, there was a good correlation between postoperative improvement in regional wall motion and postoperative myocardial perfusion. Significantly improved regional wall motion was associated with good regional myocardial perfusion after myocardial revascularization. In addition to the demonstrated potential of intraoperative myocardial contrast echocardiography, the greatest ischemic areas to be revascularized may be defined by injection of sonicated agents with cardioplegic solution before grafting. If differential rate and degree of whiting out of sonication could be discerned in detail, we also believe that this method in which cardioplegic solution is used can determine the most valuable vessel to bypass first in multivessel diseases. This determination may be difficult by coronary angiographic findings alone. In conclusion, intraoperative myocardial contrast echocardiography would enable surgeons to assess regional myocardial perfusion in the revascularized myocardium and thus to evaluate the effectiveness of coronary bypass grafting by direct injection of sonicated contrast medium
into individual saphenous vein grafts. The present study suggested that the majority of the areas supplied by coronary collateral flow associated with totally occluded coronary arteries had subendocardial infarction. REFERENCES 1. DeMaria AN, Bommer WJ, RiggsK, DajeeA, Keown 0 L,
2.
3.
4.
5.
6.
7.
8.
9.
10.
II.
Mason DT. Echocardiographicvisualization of myocardial perfusionby left heart and intracoronary injections of echo contrast agents [Abstract]. Circulation 1980;62(pt 2): III 43. Armstrong WF, Mueller TM, Kinney EL, Tickner EG, DillonJC, Feigenbaum H. Assessmentof myocardialperfusion abnormalities with contrast-enhanced two-dimensional echocardiography. Circulation 1982;66: 166-73. Lim YJ, Nanto S, Masuyama T, et al. Visualization of subendocardial myocardial ischemia with myocardialcontrast echocardiography in humans. Circulation 1989;79: 233-44. Goldman ME, Mindich B. Intraoperative cardioplegic contrast echocardiography for assessing myocardialperfusion during open heart surgery. J Am ColI Cardiol 1988; II :59-65. Spotnitz WD, Kaul S. Intraoperative assessmentof myocardial perfusion using contrast echocardiography. Echocardiogr J Cardiovasc Ultrasound Allied Technol 1990;7:209-28. Kabas JS, Kisslo J, FlickCL, et al. Intraoperativeperfusion contrast echocardiography. Initial experienceduring coronary artery bypassgrafting. J THORAC CARDIOVASC SURG 1990;99:536-42. Maurer G, Ong K, Haendchen R, et al. Myocardial contrast two-dimensional echocardiography: comparison of contrast disappearance rates in normal and underperfused myocardium. Circulation 1984;69:418-29. Taylor A, Collins SM, Skorton OJ, Kieso RA, Melton J, Kerber RE. Artifactual regional gray level variability in contrast enhanced two dimensional echocardiographic images: effect on measurement of coronary perfusion bed. J Am Coli Cardiol 1985;6:831-8. Kaul MW, KellyP, Oliner JO, Glasheen WP, KellerMW, Watson DO. Assessmentof regional myocardialbloodflow in-vivo with myocardial contrast two-dimensional echocardiography. J Am Coli Cardiol 1989;13:468-82. Cheirie J, Zoghbi WA, Minor ST, et al. Assessment of myocardial perfusion in humans by contrast echocardiography. I. Evaluation of regional coronary reserve by peak contrast intensity. J Am ColI Cardiol 1988;11 :735-43. Keller MW, Segel SS, Kaul S, During B. The behaviorof sonicated albumin microbubbles within the microcirculation: a basis for their use during myocardial contrast echocardiography. Circ Res 1989;65:458-67.