Low coronary flow in coronary artery bypass graft (CABG) represents a major problem for perioperative infarction and postoperative survival [1–4]. Thus, measurements of coronary bypass flow during CABG may be helpful to detect technical failures, coronary vasospasm or graft insufficiency for correction of the low flow situation while the chest is still open [5–8]. Non-invasive flow measurements in the early postoperative phase are difficult and associated with a low sensitivity and a high variability such as color-Doppler flow measurements and MR-phase contrast flow mapping [9–15]. Thus, the purpose of the present study was to determine intraoperative flow by the transit time flow technique and to compare these measurements to postoperative (days 5–7) flow measurements obtained by color-Doppler echocardiography and phase contrast MR-flow measurements. In 22 male patients, mean age 62±8.5 years, scheduled for elective CABG, intraoperative flow measurements were performed. All patients had a normal or slightly reduced ejection fraction (64±12%) and were non-diabetic. All received a left internal mammary artery (IMA) graft to the left anterior descending artery (LAD) using standard cardiopulmonary bypass techniques with mild hypothermia and cold blood cardioplegia. In all patients additional saphenous vein grafts were implanted. The total number of distal venous and arterial grafts was 2.8±0.4 per patient. All patients had an uneventful postoperative course. Informed consent was obtained from all patients. Transit time volume flow measurements (CardioMed 4008, Medi-Stim, Norway) were carried out 15 min after weaning from cardiopulmonary bypass (CPB) before chest closure [5,6]. No vasocative drugs were given before the measurements. The size of the flow probe was either 4 or 5 mm in diameter. Occasionally a 3 mm flow probe was used for measuring the IMA on its partly skeletonized segment. All IMA to LAD grafts and saphenous vein grafts were investigated (n=62). In addition the flow of the native right IMA was determined in three patients as a control measurement. Mean and phasic graft flow as well as arterial blood pressure were measured in all patients (Fig. 1) and systolic/diastolic flow ratio was calculated. (Patient R.H., 1920) Intra operative transit time flow measurement of the left IMA to LAD graft. The upper panel shows the phasic flow curve with a mean flow of 91 ml/min. The lower panels show curves of the arterial pressure, ECG and the calculated resistance. Color-Doppler flow measurements were performed during the early postoperative phase, usually on days 5–7. A 7 MHz Doppler flow probe was used on a Sequoia 256C Accuson System (Mountain View, CA, USA) in the second left or right intercostal space. After identification of the IMA, flow velocity and flow velocity time integral as well as systolic/diastolic flow ratio were calculated in all patients (Fig. 2) . For calculation of absolute flow, the color-Doppler velocity time integral was multiplied by IMA-cross-sectional area obtained from color-Doppler flow echocardiography. (Patient R.H., 1920) Postoperative color-Doppler examination of the left IMA to LAD graft. The curve shows a good signal with a peak velocity (v max) of 0.47 m/s and a mean velocity of 0.20 m/s. A 1.5 T whole body scanner was used with a circular body coil (Vision, Siemens Medical Systems, Erlangen, Germany). Localization of the graft position was performed with a single shot turbo spin echo (HASTE) sequence (5 mm slice thickness) in breath hold technique and ECG triggering (n=16). Then, MR-angiography was carried out using a 3D gradient echo fast low angle shot (FLASH) sequence with a section thickness of 1.5 mm after administration of gadolinium-DTPA contrast bolus injection (0.2 mmol/kg BW) (n=10 with contrast and n=6 without contrast). Postprocessing with use of the maximum intensity algorhythm (MIP) allows visualization of the graft in any plane. At the end of the examination graft flow measurements were preformed with a navigator echo 2D phase contrast (PC) sequence with about 30 time points (velocity measurements always perpendicular to the graft). The navigator echo tracking allowed to compensate for respiratory motion artifacts by tracking the variations of the dome of the right diaphragm (TR=65, TE=5.0 ms, section thickness 6 mm; VENC=75 cm/s) (n=16). A region of interest (ROI) was placed on the individual grafts and followed (repositioned) in each time point. Patency was assessed visually by a radiologist blinded to the patient' data. Identification of the distal anastomosis was performed from 3D FLASH with MIP reconstruction in all patients. Flow was measured in the midportion of the left IMA to LAD graft. The right IMA flow was measured in the midportion in situ of the right chest wall. The absolute flow was calculated by the multiplication of flow velocity (PC) by the vascular cross-sectional area (Fig. 3) . (Patient R.H., 1920) Postoperative MR phase contrast flow measurement of the left IMA to the LAD graft. The left panel shows a cross-section through the upper chest from a 2D phase contrast magnitude image after gadolinium DPTA bolus injection (0.2 mmol/kg BW). The left IMA shows a bright flow signal and is circled for area measurement (arrow 1). Note, on the opposite side of the sternum (arrow 2), the native right IMA artery (lateral) and vein (medial). The right panel shows the flow curve of the IMA with 30 time points. A biphasic systolic/diastolic flow curve with a higher diastolic component is shown with a mean flow of 65 ml/min. Data are the mean±SD. Statistical significance between the 3 groups was tested by a one way analysis of variance (ANOVA) for repeated measurements. If the analysis was significant, a multiple comparison test was applied. In addition a linear correlation comparison of intraoperative, color-Doppler and MR-flow measurements was performed. Intraoperative heart rate was significantly higher than that registered during postoperative MR and color-Doppler flow measurements (Table 1 ). Mean aortic pressure was slightly lower during bypass surgery although significantly. The rate–pressure product was not significantly different between the three groups in the ANOVA test for repeated measurements. Whereas it was increased significantly intraoperatively when compared to color-Doppler measurements, but not to MR-measurements in a multiple comparison test. Hemodynamics An original tracing of the intraoperative flow measurements is shown in a patient (Fig. 1) with an IMA to LAD graft. Mean flow of all patients was 33±17 ml/min and the systolic/diastolic flow ratio amounted to 0.44±0.12 (Table 2 ). The intraoperative flow of the right IMA (n=3) was 13±16 ml/min (significantly lower than the left IMA with a high systolic/diastolic flow ratio of 0.79±0.25). Flow and systolic/diastolic flow ratio in the left IMA grafted to the LAD compared to the right IMA in situ assessed by transit time, color Doppler and MR technique All left IMA to LAD grafts could be identified using the color-Doppler. The time required to visualize the left IMA graft was longer than the right in situ IMA. Good quality pulsed wave Doppler tracing could be obtained in all patients. Measurement of the vessel diameter was difficult. Mean velocity was 0.21±0.08 m/s and the total flow 66±54 ml/min. Systolic/diastolic flow ratio amounted to 0.67±1.0. The right in situ IMA flow amounted to 112±79 ml/min and the systolic/diastolic flow ratio to 0.75±0.05 (Table 2 and Fig. 5) Systolic/diastolic flow ratio in grafted (left columns) and ungrafted (right columns) IMA according to the measuring technique used: intraoperative transit time, postoperative color-Doppler and MR phase contrast. A higher systolic/diastolic flow ratio is measured with all three techniques in the ungrafted native right IMA artery as compared to the left IMA grafted to the LAD showing more diastolic flow due to coronary anastomosis. Color-Doppler yields a relatively high systolic flow pattern probably due to the proximal measurements (2nd intercostal space). All left IMA to LAD grafts could be identified. However, the image taken just after the gadolinium contrast bolus injection showed a very bright, easily distinguishable graft against the dark background of the chest cavity. The best representation was obtained with a MIP reconstruction and a 3D cinerotation loop in the sagittal/coronal axis after contrast injection allowing, in most cases, the left IMA graft to be traced to the distal LAD anastomosis even with contrast showing flow in the grafted coronary artery (Fig. 4) . MIP reconstruction of a 3D FLASH sequence after gadolinium injection showing clearly the left IMA to LAD graft, 2 saphenous vein grafts originating at the anterior border of the ascending aorta going to the circumflex coronary artery (left panel: sagital view). The right panel (coronal view) shows one SVG anastomosed to the right coronary artery and the right IMA in situ. All grafts were patent and the mean flow was 36±25 ml/min with a systolic/diastolic flow ratio of 0.43±0.17. The right IMA flow was 48±15 ml/min with a systolic/diastolic flow ratio of 0.52±0.14 (Table 2 and Fig. 5). There was a significant correlation between intraoperative direct flow measurements and postoperative MR-flow measurements (r=0.57; P<0.04). A similar but not significant correlation was observed between postoperative MR and color-Doppler flow measurements (r=0.44; P<0.02). However, no correlation existed between intraoperative and color-Doppler measurements. Systolic/diastolic flow ratios correlated between the postoperative color-Doppler and MR-flow measurements (r=0.88; P<0.008). There was a poor correlation between intraoperative and postoperative MR respectively color-Doppler systolic/diastolic flow ratio (Table 3 ). Linear correlations of flow and systolic/diastolic flow ratio between the 3 techniques used. TTF intraoperative transit time; CD color-Doppler; MR MR phase contrast The aim of the study was to assess postoperative patency and flow in grafted and ungrafted IMA by non-invasive methods, namely, color-Doppler and MR techniques and to compare them to intraoperative flow measurements. Intraoperative transit time volume flow measurements have been shown to be accurate and were easy to perform without complications [5,7]. Experimental and clinical studies have shown a close correlation to true blood flow measurements [5,16]. They represent the best quality assessment of the revascularization procedure since technical errors or severe vasospasm can easily be detected and treated [6]. The results of intraoperative flow measurements are not directly comparable with the postoperative results due to the fact that they are not taken simultaneously; however, there is a fair significant correlation of the IMA graft flow between the intraoperative transit time and the postoperative MR techniques. In the few right IMA arteries measured in situ a low flow and high resistance, probably due to vasospasm during preparation, was noted [5,6]. Assessment of flow velocity is well established by the color-Doppler method and used routinely for cardiac and vascular evaluations. In measuring coronary bypass flow using color-Doppler technique the main problem is signal penetration, insonation angle and spatial resolution [12,13,17,18]. Color-Doppler flow showed higher values than the other two techniques due to the difficulty in assessing the graft diameter (pixel blurring for vascular area) and the large variation between patients. There is a high systolic/diastolic flow ratio even in the grafted left IMA, probably due to the fact that the measurements were performed very proximally in the left IMA close to the aorta showing mainly systolic flow characteristics. This may explain the different findings of other authors with predominantly diastolic flow in the IMA grafts [5,6,19]. MR angiography and flow measurements seem to be the noninvasive technique of choice, especially when using contrast agents since excellent visualization of all grafts and often the distal anastomosis was possible. Flow quantification can be performed at different and multiple levels which is an advantage compared to the color-Doppler technique. Systolic/diastolic flow ratio is comparable to measurements carried out during surgery and shows a higher diastolic than systolic flow, probably due to the fact that measurements were carried out in the mid or the distal portion of the left IMA graft. Measurements performed on native coronary arteries have shown limitations due to the small diameter, motion and tortuosity of native coronary arteries [20]. In contrast, when measuring flow in coronary artery bypass graft, especially saphenous vein grafts, several authors have shown good correlation to other techniques and are confident of such measurements due to the larger size of the vessels. Several authors have shown that such measurements are possible and that flow curves can be obtained reliably [11,14,21]. One major problem of the present study was the non-simultaneous assessment of coronary flow from intraoperative to early postoperative measurements. This fact is reflected by the differences in heart rate, mean aortic pressure and rate–pressure product (Table 1). However, simultaneous assessment of coronary flow by MR-imaging and color-Doppler flow techniques is not possible at the moment because intraoperative MR-possibilities are not available, metallic structures of the flow probes interfere with the MR-signal and because of the patient’s instrumentation (ventilation, multiple perfusors, etc.), monitoring and catheter access in the MR-unit. These limitations apply also for the Doppler flow wire techniques. Thus, a true simultaneous assessment of bypass flow is not possible in the patient and influences the conclusion of this study. MR-flow measurements are limited by their low temporal and spatial resolution. Motion artifact induced by cardiac contraction and respiration may have affected MR-flow measurements and, thus, its reliability. Despite these motion artifacts MR-measurements showed a significant correlation to the intraoperative flow measurements, whereas the color-Doppler did not (Table 3). Color-Doppler flow measurements may have been influenced by pixel blurring. As a result vascular area could not be measured exactly and might have led to overestimation of flow. Comparison of intra, and early postoperative flow measurements indicate the superiority of MR over color-Doppler techniques for assessing IMA graft flow. Color-Doppler flow measurements are associated with considerable over estimation of bypass flow although graft patency could be assessed in all patients. Intraoperative and early postoperative flow measurements may be helpful for assessing bypass patency and detecting early graft failures and its related morbidity and mortality. Current non-invasive technologies are, however, limited by several problems and may provide information on graft patency rather than quantitative flow measurements. The authors would like to thank the staff of cardiology and radiology for their expert help. This work was partly supported by a grant from the Swiss Heart Foundation. Dr Friedrich Mohr (Leipzig, Germany): Does the MR allow for an anastomotic judgment also? It does not look like it on the image, although the images are very nice. Dr Walpoth: We have even better images, but I just took a random image. Your question is a very important one and we looked at this. Actually, we used only one injection of contrast agent, trying to obtain all the necessary information in one single exam in order to reduce cost and time; in about half of the patients you can see the LIMA to LAD anastomosis. This needs some postprocessing because you have to rotate the heart to be multiplanar and otherwise you have the overlap of the contrast in the left ventricle. Dr Olaf Wendler (Homburg, Germany): I think you may have a systematic mistake because of correlating the flow measurements intraoperatively with postoperatively. It is known that low peripheral resistance or anemia may cause higher flow rates in the postoperative period. Dr Walpoth: Well, I agree with you. I mentioned that there is a limitation on trying to compare intra with postoperative flow measurements. I think you cannot compare it since the patients are in a different hemodynamic situation. Nevertheless, we recorded the pressure–rate product, and there was no significant difference between intra and postoperative measurements. However, I cannot make any comment on the peripheral resistance. Thus, there is no way of comparing these two measurements. However, our aim was mainly to see whether we could visualize the graft and whether we could measure flow: those are possible and it turned out that flow values were comparable between intraoperative and postoperative measurements. Dr Yves Louagie (Mont Yvoir, Belgium): Do you have contraindications to the use of the MRI and gadolinium injection in some patients? Particularly, do you consider that the large clips used to achieve hemostasis of LIMAs and gastroepiploic arteries contraindicate the use of MRI? Dr Walpoth: Small metallic clips on a IMA or vein grafts are no major problem. Mostly you will have a blackout where your clip is located but can visualize the graft and measure flow above or below that region. Regarding the gadolinium: there are a few contraindications, which are, for instance, renal insufficiency, but otherwise it is used as a single shot and is well tolerated by the patient.
No takes yet. Share an insight, caveat, or question.
Walpoth et al. (1999) studied this question.
Synapse has enriched 2 closely related papers on similar clinical questions. Consider them for comparative context: