Background Cardiac catheterization is the only practical method of assessing internal mammary artery graft patency. A noninvasive method would be useful in patients with recurrence of anginal symptoms after coronary artery bypass graft surgery. We hypothesized that transthoracic echocardiography could provide information on blood velocity and anatomy and therefore has the potential to allow measurement of blood flow.
Methods and Results High-frequency (5 MHz) transthoracic echocardiography was performed on 41 consecutive patients (mean age, 67±6 years) who had had left internal mammary artery grafts to the left anterior descending coronary artery (LAD) and were undergoing coronary angiography because of recurrence of anginal symptoms. The results were compared with those from 19 patients (mean age, 58±11 years) in whom an ungrafted left internal mammary artery was assessed and with those from 15 patients (mean age, 61±12 years) who had angiographically normal coronary arteries in whom the LAD was studied. Doppler velocity profiles of the left internal mammary graft were obtained in 35 of the 41 study patients (81%). In all cases, a biphasic pattern of blood flow was recorded that corresponded to systole and diastole. Two different flow patterns were observed. In 25 patients with a normal graft or moderate (<70%) stenosis (group A), blood flow velocity was maximal during diastole. This pattern was also seen in the LAD control group. In 10 patients with severe (>70%) graft stenosis (group B), blood velocity was maximal during systole, and low velocities were recorded during diastole. This pattern was also seen in the ungrafted internal mammary artery control group. The diastolic fraction of the velocity time integrals for group A was 0.77±0.07 and for group B was 0.27±0.01 (P<.05). A diastolic velocity time integral fraction <0.5 predicted severe stenosis with a sensitivity and specificity of 100%. The ratio of systolic-to-diastolic peak velocities for group A was 0.54±0.26 and for group B was 3.45±0.74 (P<.05). A systolic-to-diastolic peak velocity ratio >1 predicted severe stenosis with a sensitivity of 100% and specificity of 85%. Mean graft blood flow was 63±21 mL/min. There was no significant difference in mean blood flow between any of the patient groups studied.
Conclusions High-frequency transthoracic echocardiography allows identification of the left internal mammary grafts and measurement of blood flow. Compared with patent grafts or those with moderate lesions, severe stenoses demonstrated different Doppler velocity patterns. Use of this technique may allow noninvasive detection of significant stenoses of the left internal mammary artery graft.
In recent years, the left internal mammary artery (LIMA) has become the conduit of choice for the surgical bypass of stenosis of the left anterior descending artery (LAD) because of the greater longevity of this vessel compared with saphenous vein grafts. Assessment of LIMA graft patency currently requires invasive investigation. A noninvasive method of assessment would be useful in patients with recurrence of anginal symptoms after coronary artery bypass graft surgery and would also allow measurement of blood flow after therapeutic interventions. We previously described an ultrasound method for examination of blood flow in the ungrafted LAD.1 2 The objective of this report was to use ultrasound to assess LIMA flow in patients with bypass grafts.
Forty-one consecutive patients (mean age, 67±6 years) undergoing coronary angiography because of recurrence of anginal symptoms were studied. All had had coronary artery bypass graft surgery using the LIMA to the LAD 1 to 11 years previously (mean, 8±2 years). In each patient a careful attempt was made to visualize the LIMA and to obtain a pulsed-wave Doppler measurement of LIMA blood flow velocity.
Two control groups were also studied. Patients were selected only if high-quality images of the vessel and a Doppler velocity profile could be obtained. In 19 patients (mean age, 58±11 years) with angina but no previous surgery, the ungrafted LIMA was assessed. In 15 patients (mean age, 61±12 years) who had angiographically normal coronary arteries, the LAD was assessed. Informed written consent was obtained from all participants.
Echocardiography was performed using a Vingmed CFM 750 ultrasound unit and a 6.3-MHz mechanical sector transducer. The transducer has a focal length of 40 mm, with lateral and axial resolutions of 0.6 and >0.3 mm, respectively, operating at its nominal frequency of 5 MHz. It is a broad bandwidth transducer that allows the transmission and reception of ultrasound frequencies over a wide range for both imaging and Doppler. In this study a Doppler frequency of 4 MHz was used.
Detection of the LIMA
Patients were examined in the left lateral position by using a modified left parasternal window. Long-axis images of the left ventricle were obtained, and then the area anterior to the right ventricular outflow tract and the anterior interventricular sulcus was carefully examined by using combined imaging and color flow mapping. The LIMA graft was identified as a tubular structure with color flow directed from base to apex and containing characteristic Doppler flow signals. Once the position of the LIMA was identified, intraluminal flow signals were obtained using the pulsed Doppler method. The long-axis sections were carefully adjusted to minimize the angle between the Doppler beam and the long axis of the artery and also to ensure that the sampling volume was located within the vessel lumen for as much of the cardiac cycle as possible. The Doppler signal and two-dimensional (2D) echocardiogram were then recorded.
Detection of the LAD
With the same procedure described above, the transducer was placed in the left parasternal area, and the left ventricle was imaged in its long axis. The ultrasound beam was angled laterally and superiorly to identify the anterior interventricular sulcus. By using combined imaging and color flow mapping, the distal LAD was identified as a tubular structure located within the anterior interventricular sulcus containing characteristic Doppler flow signals. The long-axis sections were carefully adjusted to minimize the angle between the Doppler beam and the long axis of the artery and also to ensure that the sampling volume was located within the vessel lumen for as much of the cardiac cycle as possible. The Doppler signal and 2D echocardiogram were then recorded.
All images were recorded on super VHS videotape and analyzed later by one observer who was blind to the angiographic result.
2D and Doppler Echocardiographic Measurements
Measurements of vessel diameters were performed using internal electronic calipers on frozen frame images from the 2D recordings. A leading edge–to–leading edge technique was used.
Velocity measurements were performed using the internal analysis package on the ultrasound unit. Measurements were calculated taking into consideration the angle between the Doppler beam and the longitudinal axis of the blood vessel as determined by the 2D echocardiogram. Six parameters were measured: (1) peak systolic velocity; (2) peak diastolic velocity; (3) mean systolic velocity; (4) mean diastolic velocity; (5) systolic velocity time integral; and (6) diastolic velocity time integral. Values for each parameter were obtained by averaging measurements from 5 to 7 consecutive cardiac cycles.
Blood flow was obtained from the product of the mean blood velocity multiplied by the cross-sectional area of the related artery. In addition, comparison of blood velocity patterns was made by calculating the systolic-to-diastolic peak velocity ratios and the diastolic fraction of the velocity time integral (ie, the diastolic velocity time integral divided by the diastolic plus systolic velocity time integral).
Coronary angiography was performed on the day after echocardiography by standard Judkins technique with a single-plane imaging system. Images were recorded onto 35-mm cine-film at a frame rate of 25 frames per minute and were analyzed independently by an expert observer. The LIMA grafts were examined by using multiple projections and classified according to visually determined percent narrowing as severe (>70%), moderate (40% to 70%), or normal (<40%).
Results are presented as mean±SD. Data were analyzed using ANOVA. A Fisher protected least-significant-difference test was performed if the ANOVA showed significant differences. A value of P<.05 was considered significant.
At angiography, 12 patients had an occluded or severely stenosed LIMA graft. Moderate stenosis was observed in 3 patients. The LIMA graft was normal in 26 patients. In all cases, the stenosis occurred at a point distal to the site of echocardiographic assessment.
Echocardiographic Detection of LIMA Grafts
2D images and Doppler velocity patterns of the LIMA graft were obtained in 35 of the 41 patients (81%). Of the 6 patients in whom it was not possible to obtain adequate images for analysis, the graft was occluded in 2 and was normal in the other 4. Apart from these 6, the study patients were assigned to two groups: group A, 25 patients with a normal LIMA graft or moderate (<70%) stenosis; group B, 10 patients with severe (>70%) graft stenosis. Individual patient characteristics and hemodynamic results for the entire study group are given in Table 1⇓.
Flow Patterns in the LIMA Grafts
In all cases, there was a biphasic pattern of blood flow corresponding to systole and diastole. Two different patterns were observed. In group A patients, who had a normal graft or moderate stenosis, flow was dominant during diastole (Fig 1⇓). In group B patients, who had an occluded or severely stenosed LIMA graft, flow was dominant during systole, and low velocity profiles were recorded during diastole (Fig 2⇓).
Flow Patterns in Ungrafted LIMA and Normal LAD
Two characteristic flow patterns were seen in the control groups. In the ungrafted LIMA controls, flow occurred mainly during systole with low velocities during diastole. In the LAD controls, flow was dominant during diastole (Table 2⇓).
Velocity patterns in group A were similar to those in the LAD control group. Velocity patterns in group B were similar to those in the ungrafted LIMA control group.
Measurement of Blood Flow
Measurement of LIMA diameter was possible in 33 of 35 patients, allowing calculation of blood flow. Mean graft diameter was 0.23±0.06 cm. Mean blood flow for the entire group was 63±21 mL/min. There was no significant difference in mean blood flow between any of the patient groups studied (Table 3⇓).
Prediction of Severe LIMA Graft Stenosis
The diastolic fraction of the velocity time integrals for normal LIMA grafts or those with moderate stenosis (group A) was 0.77±0.07 and for severely stenosed LIMA grafts (group B) was 0.27±0.01 (P<.05). A diastolic velocity time integral fraction <0.5 predicted severe stenosis with a sensitivity and specificity of 100%. The ratio of systolic-to-diastolic peak velocities for normal LIMA grafts was 0.54±0.26 and for severely stenosed LIMA grafts was 3.45±0.74 (P<.05; Fig 3⇓). A systolic-to-diastolic peak velocity ratio of >1 predicted severe stenosis with a sensitivity of 100% and specificity of 85%.
We report a technique for the assessment of LIMA patency using transthoracic echocardiography. The LIMA was identified by using combined 2D and color flow echocardiography that allowed detection of blood velocity patterns by Doppler. Measurement of blood flow was possible because both velocity and vessel diameter were obtained. Using this transthoracic approach, velocity and diameter data were obtained in 81% of patients. Two distinct patterns of flow velocity were observed. In patients with normal grafts or moderate stenosis, the blood velocity pattern was similar to that seen in normal coronary arteries. With significant stenosis (>70%), the blood velocity pattern reverted to that observed in ungrafted internal mammary arteries. Blood flow and flow velocities were measured in the distal segment of the LIMA graft. No stenosis was detected at or proximal to this part of the vessel during angiography, so that differences in blood velocity profiles were not attributable to stenosis at the site of measurement.
Noninvasive visualization of LIMA grafts has been attempted in previous studies.3 4 However, imaging of the proximal and mid-LIMA using parasternal and supraclavicular approaches has a low detection rate (55%). These studies used higher frequency (7.5 MHz) transducers than those used in this study, which may have greater tissue attenuation. Other possible reasons for the absence of a Doppler signal include variations in the shape of the chest wall or position of the grafted LIMA.
High-frequency ultrasound and Doppler techniques provide accurate and reliable morphological and physiological information in carotid and femoral arteries.5 6 Recently these techniques have been extended to the study of coronary arteries and grafted vessels.1 7 8 Transthoracic echocardiography has allowed imaging of short segments of the coronary arteries and also assessment of blood velocity profiles in the distal segment of the LAD.2 9 However, imaging of the coronary arteries has proved difficult, partly because of unfavorable chest-wall configurations, coexistent chronic obstructive airways disease, and the small size of these vessels. The coronary arteries are tortuous and mobile so that it is difficult to obtain accurate velocity information by Doppler throughout the entire cardiac cycle. The LIMA is less mobile than the coronary arteries and is close to the surface of the chest, making it accessible to imaging by high-frequency echocardiography.
In normal coronary arteries, most blood flow occurs during diastole because myocardial compression during systole increases distal vascular resistance. Coronary blood flow in the LAD control group demonstrated this characteristic pattern. In our study, the blood flow patterns and velocities in patients with normal LIMA grafts or moderate stenoses were similar to those in the normal LAD group, suggesting that LIMA grafts allow smooth flow into the recipient artery without the development of turbulence. This is consistent with other known characteristics of these vessels: The diameter of LIMA grafts is approximately the same as that of recipient coronary arteries and arterial grafts are controlled by similar autoregulatory responses that allow changes in diameter in response to changing myocardial blood flow demands.10 11 Blood flow in ungrafted internal mammary arteries occurs mainly during systole similar to flow in peripheral arteries. An occluded LIMA graft acts as a blind-ended tube similar to an ungrafted LIMA, resulting in loss of the diastolic component of coronary flow.
The sensitivity and specificity for detection of severe LIMA graft stenosis was 100% and 85%, respectively, in successfully imaged patients. Three patients had moderate stenosis, but it was not possible to differentiate them from patients with normal LIMA graft function on the basis of resting blood velocity profiles and blood flow. There was no significant difference in blood flow between any of the patient groups studied. However, flow measurements were made at rest, and at rest differences in basal flow may not occur even in patients with significant stenosis.4 12 13 The functional impairment of a vessel is probably better assessed by measuring flow reserve. Invasive studies have shown that when maximal coronary dilatation is induced, blood flow or blood flow velocity increases more in normal vessels than in those affected by significant stenosis.14 15 16 Measurement of blood flow after the use of coronary vasodilators would allow noninvasive assessment of flow reserve in the LIMA and may be useful in the assessment of graft stenoses of moderate severity. This technique has been reported in the assessment of LAD stenoses using transesophageal echocardiography.17
The LIMA is a small vessel, and accurate measurement of blood flow is particularly dependent on accurate measurement of vessel diameter. In a pathological study, Kenny et al18 found close agreement between measurements of coronary artery luminal diameter by epicardial echocardiography and histology. Therefore, high-quality images of the LIMA should allow reliable diameter measurements to be made. Assessment of absolute blood velocity may be limited in some patients by the large incident angle between the Doppler beam and blood flow. We measured only those Doppler recordings that provided high-quality narrow spectral traces. These are known to lead to reproducible velocity profiles in other vessels.5 6 Calculation of the systolic-to-diastolic peak velocity ratio and the diastolic fraction of the velocity time integral allows assessment of flow profiles without the need for absolute values.
Blood velocity in the LIMA graft is the result of instantaneous pressure gradients along the vessel throughout the cardiac cycle. Parameters that affect this gradient will lead to alterations in the velocity profile. Invasive studies have shown that residual flow in the recipient artery may compete with flow in the patent LIMA and reduce flow in the graft. The LIMA graft may remain patent and may function normally at a later stage when native coronary flow ceases.19 Severe left ventricular dysfunction may affect flow in the LIMA graft and LAD by causing a reduction in cardiac output and also by reducing the myocardial compressive force during systole. This may alter velocity profiles in the LIMA graft in an unpredictable manner. Assessment of the contribution of these factors was not made in this study.
High-frequency transthoracic echocardiography allows identification of LIMA grafts and measurement of blood flow. We have described a technique for the noninvasive measurement of LIMA blood velocity profiles and flow using high-frequency transthoracic echocardiography. Compared with patent grafts or those with moderate lesions, severe stenoses demonstrated different Doppler velocity patterns. Use of this technique may allow noninvasive detection of significant stenoses of the LIMA graft and also assessment of blood flow after therapeutic and pharmacological interventions.
Presented in part at the 67th Scientific Sessions of the American Heart Association, Dallas, Tex, November 14-17, 1994.
- Copyright © 1995 by American Heart Association
Kenny A, Wisbey CR, Shapiro LM. Profiles of coronary blood flow velocity in patients with aortic stenosis and the effect of valve replacement: a transthoracic echocardiographic study. Br Heart J. 1994;71:57-62.
Kyo S, Matsumura N, Yokote Y, Takamoto S, Omoto R. Evaluation of patency of internal mammary artery grafts: a comparison of two-dimensional Doppler echocardiography and coronary angiography. J Cardiol. 1990;20:606-616.
Kohler TR, Nance DR, Cramer MM, Vandenburghe N, Strandness DE. Duplex scanning for diagnosis of aortoiliac and femoropopliteal disease: a prospective study. Circulation. 1987;76:1074-1080.
Fusejima K, Takahara Y, Sudo Y, Murayama H, Masuda Y, Inagaki Y. Comparison of coronary hemodynamics in patients with internal mammary artery and saphenous vein coronary artery bypass grafts: a noninvasive approach using combined two-dimensional and Doppler echocardiography. J Am Coll Cardiol. 1990;15:131-139.
Mehta J, Pepine CJ. Effect of sublingual nitroglycerin on regional flow in patients with and without coronary disease. Circulation. 1978;58:803-807.
Feldman RL, Pepine CL, Conti CR. Magnitude of dilation of large and small coronary arteries by nitroglycerin. Circulation. 1981;64:324-333.
Hoffman JIE. Maximal coronary flow and the concept of coronary vascular reserve. Circulation. 1984;70:153-159.
Wilson RJ, White CW. Intracoronary papaverine: an ideal coronary vasodilator for studies of coronary circulation in conscious humans. Circulation. 1985;73:444-451.
Marcus M, Wright C, Doty D, Eastham C, Laughlin D, Krumm P, Fastenow C, Brody M. Measurements of coronary velocity and reactive hyperemia in the coronary circulation of humans. Circ Res. 1981;49:877-891.
Iliceto S, Marangelli V, Memmola C, Rizzon P. Transesophageal Doppler echocardiography evaluation of coronary blood flow velocity in baseline conditions and during dipyridamole-induced coronary vasodilation. Circulation. 1991;83:61-69.
Kenny A, Fuller CA, Cary NRB, Shapiro LM. Histopathological validation of high frequency epicardial echocardiography of the coronary arteries in vitro. Br Heart J. 1991;65:326-331.