Perioperative Ultrasound Flow Evaluation in Grafts and Native Vessel during CABG

Abstract

New blood in the vascular bed after coronary artery bypass grafting (CABG) may represent a turning point between ischemia and normal tissue nutrition. Quality control during CABG preoperatively is essential because errors lead to immediate consequences. With an understanding of hemodynamics, we can now search for less invasive tools for quantification of coronary blood flow changes over time after CABG. Ultrasound is becoming a key player for that purpose and we will present its application. Perioperatively, quality control in CABG should include target selection of native coronary arteries, graft selection, anastomose checking, and long-term flow follow-up. Because some grafts are unreachable for ultrasound evaluation, we should examine both arterial venous sides of coronary circulation. We will present the use of classic, epicardial ultrasonography and TTFM probes by looking for stenoses and competitive flow. We will present our research for quantification of new blood in coronary vessels after CABG. There we found constant increase in flow over the early postoperative period (20% per graft). By increasing graft number, coronary flow increases first linearly and then stepwise. Measured data and trends can be used in ambulatory monitoring and screening of ischemic complications after CABG.

1. Introduction

Ischemic heart disease (IHD) represents part of cardiovascular disease, which is connected to 30% of deaths worldwide. Especially in low- or moderately-developed countries, where 80% of cardiovascular disease is concentrated. IHD is a disease of coronary arteries and capillary bed that narrows vessel lumen, leading to reduced blood flow below heart needs. Flow reduction during rest is rare except in severely stenotic coronary arteries, hibernated myocardial segments, and after acute myocardial infarction. More often, there is flow reduction during maximal dilatation. The impact of atherosclerotic ischemic heart disease on quality of life and its connection to mortality is well-established.

Three overlapping treatments highly reduce the occurrence of adverse cardiovascular events and are used daily: medication, revascularization by percutaneous intervention, or surgical operation. Complications in macrocirculation during disease progression can be stopped or decelerated with adequate revascularization. ESC guidelines for myocardial revascularization promote the indications for the coronary artery bypass graft (CABG) over percutaneous procedure due to the higher beneficial effects of this procedure. The benefit is especially seen in patients with double vessel disease with proximal left anterior descending (LAD) stenosis, left main disease, triple vessel disease, and some other situations. In our practice and as we know in some other centers even single LAD stenosis is treated surgically if PCI procedure is impossible. When we are dealing with candidates for Re-Do or Re-Re-Do, the indications for surgical revascularization are generally the same as for first surgery. We take into account how big part of myocardium is at risk and the patency of LIMA-LAD (“no LIMA no Case”).

As millions of people are in danger to have CVD, 25% will have heart attacks in asymptomatic stage of the disease. All put together suggest that it will be important to develop more accurate method of screening of coronary artery disease and for quality control after revascularization procedures. All revascularization procedures should be subjected to quality control. They can be performed pre-OP, intra-OP, and post-OP. Given the possibility of harvesting of grafts from multiple locations, testing their suitability is necessary and falls under pre-OP quality control. Anatomical and functional tests of native coronary arteries, which are in the domain of the interventional cardiologist during coronary angiography, also fall under PreOP control. Considering the numerous cardiosurgical techniques of CABG OPCAB/ONCAB, graft configurations, and comorbidities, intra-OP quality control is very important and must be systematically carried out during the entire treatment. In the early and late Post-Op period, and due to the processes described in the further part of the text, Post-Op quality control should be carried out primarily to monitor bypass complications (anastomotic stenosis, graft occlusion, competition …), myocardium, but also possible progression of native coronary disease. As some grafts are inaccessible to flow measurements outside the OR, we also have to take into account their impact on coronary flow. Contemporary and reliable method will be flowmetric measurements on the venous side of the heart (coronary sinus) in addition to all mentioned above.

2. Quality control

CABG quality control is so important that every self-critical cardiac surgeon must always perform it. The reason is that it directly affects the early and late success of the surgical work. Quality control must include not only performed bypasses but also unperformed ones, given that incomplete revascularization is associated with increased mortality. Patency of bypasses is influenced by many variables (surgical technique, local findings, inflow-outflow, experience of the cardiac surgeon, etc.). The above reasons can lead to early anastomotic errors with future graft failure in 1–3%. The percentage is cumulative and increases with the number of anastomoses. Disease progression of native coronary arteries or progression of disease on grafts in the late post-OP period can lead to a decrease in total or regional blood volume in the myocardium. The aforementioned progression may require repeated diagnostic re-evaluation in terms of ischemic heart disease and consequent revascularization (CABG/PCI) and occurs in about 3% [1]. These conditions are much more complex and sometimes, unfortunately, cannot be solved by a cardiac surgeon. That is why early recognition of asymptomatic progression is very important, but it is often delayed, incidental, or discovered after the development of ischemic complications. Therefore, each phase of treatment of ischemic heart disease should be evaluated and documented as much as local personnel and technical possibilities allow. We believe that without going deep into the protocols of cardiac surgery centers, guideliness, experience of cardiac surgeons, the establishment of quality control, and strict adherence can distinguish the Center of Excellence from others. At the same time, it protects us from the law but also from non-benevolent colleagues.

In general, the quality control of CABG requires comprehensive judgment and individualized measures to ensure the safety and long-term outcome of patients. Although the importance of the volume of operations on the outcome and reduction of complications is described in the literature, systematic quality control can sometimes nullify this problem. Namely, Auerbach et al. who analyzed quality control showed that when quality control is maximized, CABG mortality at very low-volume hospitals is similar to very high-volume hospitals. If we transfer the point of this study (detailed and regular quality control of each step) to all other cardiac surgeries, we will probably have significantly better results in smaller centers as well. Of course, we do not want to ignore the connection between volume and experience and results, but we want to emphasize the unused potential of quality control. In this way, we encourage young cardiac surgeons in their professional development.

When to start quality control in surgical myocardial revascularization? The answer is yesterday. Unfortunately, even in centers that are equipped and staffed for detailed perioperative monitoring of bypass effects on coronary flow, these measurements rarely represent the protocol. In general, quality control should be started before the actual operation, and the measurement results should be documented. In this way, the effects of revascularization will be easily monitored during the entire perioperative period.

The analysis should include the anatomy of the target and their functional tests in order to reveal the actual stenoses that will benefit the most from revascularization. Direct ultrasound evaluation of native coronary arteries (LM and RCA) is very difficult and unreliable, so it will not be the subject of this paper. Given that modern ultrasound machines are widely available today, measurement of the venous side of the coronary circulation—flow through the coronary sinus is possible. It gives us additional information about coronary flow, especially since it includes the return of blood from ultrasound-inaccessible grafts (venous bypasses). But, what is important is that the technique used should be simple for ambulatory monitoring. This can only be achieved with non-invasive ultrasound measurements. In addition to the analysis of the flow through the native coronary arteries and coronary sinus, it is necessary to record the anatomy preoperatively and, where possible and necessary, to functionally examine the grafts as well as the circulation that will be left behind after harvesting the grafts. We primarily mean LIMA/RIMA angiography and ultrasound tests of IMAs and Radial artery circulation and VSM/VSP veins. The most common preoperative technique for functional measurement of stenosis of native coronary arteries is—FFR. Functional assessment graft is possible either pre-OP or intra-OP. They include the measurement of flow volume, resistance index, etc. at rest and after drug administration (Adenosine, NTG, Dypiridamol, etc.) and sometimes intra-OP during free flow, which will be discussed later. A good example of a pre-OP analysis of grafts is the radial artery, which should be anatomically evaluated (diameters on ultrasound), and with an ultrasound-enhanced Allen’s test to examine its collateral circulation and its potency with an ischemic ultrasound test.

Pre-OP and Post-OP non-invasive assessment of IMA patency is challenging, but duplex ultrasound (DU) can be accepted as a reliable technique. The main goal of the examination is the detection of stenosis/occlusion or flow competition. DU may detect severe (>70%) stenosis by PW Doppler data (peak diastolic velocity/peak systolic velocity < 0.5 and diastolic fraction <50%). Considering that medium-severe stenoses (50–70%) at rest do not show changes in volumetric parameters except for slight accelerations of flow at the stenosis itself, it is possible that they may be missed during imaging. That’s why we present the below curve analysis with the use of some form of stress test (adenosine, NTG, Dypiridamol, etc.) in order to reveal borderline stenoses.

Intra-OP quality control includes the use of TTFM, epicardial and classic Doppler probes, and TTE/TEE. The goal is to examine the graft for stenosis, and competition by determining the total flow, flow in the systole/diastole phases, resistance indices, and other characteristics of the flow curve. And to quantify returning blood from coronary sinus, for screening of post-OP follow-up. A conclusion should be made about the potency of grafts, especially composite grafts, expected durability, etc. The best-known and most commonly used quality control tool in OR is the transit time flow measurement (TTFM). TTFM effectively detects the coronary graft failure but often needs additional knowledge to decide the cause of failure. If we add high-resolution epicardial ultrasound to TTFM we can improve the quality, safety, and effectiveness of CABG.

At the end of the day, there are several useful consequences of quality control. Pre-OP we will better choose native target and best graft/graft configuration. Intra-OP we will assess flow and if necessary re-anastomose immediately in OR before severe consequences occur. Post-OP we will detect failing grafts or native vessel stenosis earlier during the follow-up period before complications occur. This is after intra-OP judgment the most important data that we can extract from patient during follow-up. This early detection of failing graft is often challenging, because of technical limitations and because in busy outpatient departments examinations are often performed during rest where there is rest flow, taking into account only basic data from DU. But, even in that situation if we have data about pre-OP quantities of flow (with resistance indexes and other flow curve data analysis), intra-OP flow increase, coronary sinus flow, and expected early increase of flow in post-OP period at least we can individualize trend changes and make some conclusions before severe consequences develop. A typical mentioned situation will be when the graft occludes even without ECG changes, during the early ICU period, while the patient is at bed rest. This is the most important message in this chapter, possibility of noninvasive early detection of failing grafts/native vessels during follow-up. If we are suspicious of the mentioned situation or have a symptomatic patient (with other signs of ischemia) after CABG, there are plenty of diagnostic procedures for case evaluation. We can examine anatomical, functional, and physiological aspects of flow in post-CABG recurrent angina with assessment of left ventricular function, myocardial viability, etc. Furthermore, we can choose the best management strategies for this complex patient population. Who are often older with higher prevalence of comorbidities and complex atherosclerotic lesions. Diagnostic management can include besides re-coronarography other procedure: stress electro and stress echo examination, CT coronarography, MRI, myocardial scintigraphia, PET CT, etc. Some of them have limitations like expenses, radiation exposure, or insufficient sensitivity and specificity. But, in this chapter, we want to focus on pure noninvasive flow detection on the arterial and venous side of coronary circulation.

3. Coronary flow

Normal coronary artery flow is 5% of cardiac output, around 250 ml/min [2]. Resting flow through normal coronary arteries, according to invasive measurements, is 0.5–1.5 ml/gr/min (0.8–1.2 ml/gr/min). At the same time, maximal coronary artery flow is 3–4 ml/gr/min.

Resting flow through severely stenotic coronary arteries (= > 80%), hibernated myocardial segments or infarcted area is <0,5 ml/min/gr. At the same time, maximal coronary artery flow is <1 ml/min/gr.

Resting flow in intermediate stenosis (60–80%) is the same as in healthy coronary arteries but slightly near lower value, with differences in published literature [3]. Maximal flow through intermediate stenosis strongly correlates with degree of stenosis. Flow started to decrease with stenosis of 40% and become same as resting flow in stenosis = > 80%. Patients with microcirculatory disturbances (diabetes, hyperlipoproteinemia, arterial hypertension, nicotine abuse, etc.), even without angiographically evident stenosis also has decreased flow during rest or effort [4]. It is mainly because of vasodilatatory defects of endothelial cells.

After CABG resting net flow increases from 0,65 to 0,78 ml/min/gr and maximal flow increase from 0,85 to 1,0 ml/min/gr. To simplify: there is a 20% increase in blood flow per gram of myocardium for every graft.

CS flow is lower than sum of coronary artery flow. It is because there are: alternative drainage routes (Thebasian veins and tributaries not connected to CS) and because CS drains mainly left coronary artery territory [5]. Severe coronary artery stenosis can decrease the CS blood flow. Revascularization of severely stenotic vessels can increase coronary artery and consequently CS flow. Because of this coronary sinus flow can be used in post-OP Coronary Circulation Assessment by looking at trends over time.

In case we are averaging flow rate over a gram of myocardium (during CS, IMAs, or SVG volumetric measurements) we recommend ASE; Devereux formula for LV mass calculations. Formula is: LVgr = 0,8(1,04(LVDD + PWTD + IVSTD)3-(LVIDD)3)) + 0,6. Here, LVDD is LV diameter in diastole, IVSTD is diastolic diameter of septum, PWTD is diastolic diameter of posterior wall near apex of papillary muscle. With correction of whole myocardium mass by addition of right ventricle mass (according to certain conditions), what is outside of this paper.

4. Recording suites

Regardless of the time of measuring the hemodynamics of the grafts (pre-OP/post-OP), IMA grafts and coronary sinus are always available for measurements. Other grafts (aortocoronary: VSM/VSP or radial artery) are inaccessible to pre-OP/post-OP ultrasound flow measurement techniques. So, intraoperative measurements of all available grafts by DU or TTFM should always be recorded.

4.1 Arterial side of coronary circulation

Preoperative angiographic scan of IMAs is almost always possible and should be “must” in all CABG. Unfortunately, it is not performed routinely. Usually, because of fear of contrast overdose or in emergencies. The left internal mammary artery (LIMA) is the best first choice for performing CABG due to its location and structural characteristics. LIMA arises from the subclavian artery opposite to Thyreocervical trunk. IMAs are bigger in men and on right side [6]. LIMA descends behind the upper six ribs, 1.5 cm from the sternum. LIMA is elastic artery with intima, media, and adventitia layers. The composition of these layers varies throughout the route from the subclavian artery to its epigastric bifurcation. The media layer has two elastic lamina: external (contribution to elastic performance) and internal (contribution to muscular performance). All of these layers account for long patency of the IMAs in CABG surgery. Also, IMAs have fewer endothelial fenestrations and less intercellular junctions, and high intercellular communication. Because of this IMAs are highly resistant to manipulation during harvesting.

4.1.1 Equipment

Early use of Doppler Ultrasonography (DU) for LIMA hemodynamic assessment after CABG, during follow-up, date in late 1987 by Fusejima et al. [7]. It was generally used as an alternative to coronary angiography and it is the best method for the non-invasive-assessment of coronary graft stenosis. Up to date, DU has shown promising results in pre-OP/post-OP evaluation of IMAs function, regardless of the fact that the proximal portion is predominantly detected rather than distal. Improvements in technology and technique of DU nowadays help us in full-length IMAs detection. Numerous studies show that it is possible to perform ultrasonic flow measurements of IMAs. Measurement of IMAa flow by ultrasound (pre-OP/post-OP) can be performed using linear probes or transthoracic echocardiography (TTE) [8]. Imaging probe is put over the skin and use of commercial gels. The linear probe (suprascapular window) uses a main frequency of 7.5 MHz or more. PW (pulse Doppler) is the main tool in measurement with mandatory adjustment of Doppler steer, angle, and sample volume, as well as other parameters important for optimizing the image, which is in the domain of vascular ultrasound. When recording the IMAs flow by TTE (parasternal window), we use standard TTE probes with main frequency and Doppler frequency of 3.5 MHz or more [9]. The transducer should have high lateral and axial resolutions and depth is often set up to 4 cm.

4.1.2 Recording

The native IMAs vessel or as graft can be visualized either from the supraclavicular fossa or from the parasternal window. In the supraclavicular fossa, IMAs should be searched immediately after division from the subclavian artery, in the same location regardless of whether it was used for bypass or not. The patient is in a supine position in a warm room and the person examining him is behind the patient’s head. The hand rests on the bed so that micro-movements can be made in the area of the hand. The movements are from the palm, similar to computer-mouse movements, and the probe is in the left hand. The probe is placed in the left supraclavicular fossa, at an angle of 75–80 degrees. To obtain a proper scan of proximal IMAs we rotate the transducer clockwise and inclining it toward the anterior chest wall to become parallel to the IMA. We use slow micro movement to visualize IMAs at proper angle. The hemodynamics of LIMA is examined around 2.0 cm from the subclavian artery. During vessel search doppler scale should be set to around 20 cm/sec for optimal signal detection. In the parasternal window, the native vessel can be traced along several intercostal spaces. And if the IMA is grafted only in the upper intercostal spaces. Sometimes we cannot find IMAs proximally by this window, except during LAST procedures when we usually harvest the LIMA up to the II intercostal space. Sometimes it is easier to follow native IMAs from the distal end (below the bifurcation) to the prox because there are no ribs there. The best view is in the modified left parasternal window with probe oriented craniocaudally near left sternal border. The patient is in the left lateral decubitus position or supine. Vessel is located within the II–V intercostal space, using the TTE probe. IMA is identified as a tubular structure superficial to the heart and pericardium with color flow directed distally. Hemodynamics is recorded using the Pulsed-Wave Doppler.

Detection of the distal part of the grafted IMAs is also performed through the left parasternal window [9]. The patient is in the left lateral decubitus or supine position. In the long-axis view by using 2D and Color Doppler, the ventricle is first identified, then the space in front of the RVOT and the interventricular sulcus. LIMA is identified first by Doppler and then by PW as a tube with specific biphasic flow toward the apex. It is necessary to set the most favorable angle between the basic Doppler beam and the longitudinal axis of the graft, the SV (sample volume or gate) should be expanded in order to record the flow during the entire cycle or most of the cycle. The possibility of successful detection varies and according to this study, it is as high as 86%. The author states in the explanation that they used a lower-frequency probe (5 MHz).

4.1.3 Measurements

Scicchitano et al. in their research reveals that Doppler Ultrasonography (DU) can be used as a reliable non-invasive technique for follow-up IMA patency, even if it is challenging [8]. Authors show that it is possible to detect severe (>70%) stenosis of the IMA graft by DU during rest by flow curve analysis. In those situations, they found peak diastolic velocity/peak systolic velocity (PDV/PSV) < 0.5 and Diastolic Fraction (DF) < 50% as a reliable marker of stenosis.

Flow in normal coronary arteries predominantly occurs during diastole because myocardial contraction increases vascular resistance. Same is in normal LIMA grafts or moderate stenoses because LIMA grafts allow smooth flow into the recipient artery without turbulence. Also, there is the same diameter of LIMA grafts and native coronary arteries and arterial grafts are controlled by similar autoregulatory mechanisms that allow diameter changes in response to myocardial demands. Opposite to this is in ungrafted internal mammary arteries where flow occurs mainly during systole (as in typical in high resistance peripheral arteries). In severe stenoses LIMA grafts show same flow pattern as in ungrafted LIMA, there is loss of the diastolic flow.

The flow characteristics of IMAs proximally, distally before and after CABG are different, as stated in the study by Goto and et al. [8]. Measurement of the proximal segment was performed through the supraclavicular window with linear probes and the distal segment through the parasternal window with TTE probes. Ultrasound scan of IMAs was possible in 80% of cases for all vessel lengths and 90% for only prox segment [10]. There was a difference in the flow curve of grafted and ungrafted IMAs. In general, the ungrafted IMA has a typical triphasic flow curve like typical muscular circulation, while the grafted one has a low-resistance flow curve like typical parenchymatous circulation. Native IMAs belong to peripheral arteries and have a characteristic highly resistant triphasic (rarely biphasic) flow with a high systolic and low diastolic phase [8]. It is a typical flow of peripheral arteries in which the vascular bed is perfused dominantly during the systole. The first component of the flow curve corresponds to the anterograde flow during systole and is measured by the peak velocity, i.e., peak systolic velocity (PSV). The second component of the flow curve corresponds to the reverse or rejected wave during early diastole and is measured at the maximum speed at the beginning of diastole, i.e., end-diastolic velocity (EDV). Although it is called EDV, it means early diastole. The third component of the flow curve is the later anterograde flow in the end diastole, i.e., Late Diastole Velocity. After IMAs anastomosis with coronary low-resistance circulation, deformation of the flow curve in IMAs occurs. Perfusion is performed dominantly (2/3 of the total blood) in diastole. IMAs then show two peaks in flow: systolic (now smaller) and diastolic (much bigger than before) [8]. The flow amplitude during systole decreases at the expense of an increase in amplitude during diastole. The increase in diastolic flow is the result of blood inflow into the low-resistance coronary vascular bed of the parenchymatous organ - the heart. There are also differences in the flow curve of grafted IMAs proximally and distally. In the proximal part of the graft, the flow is biphasic with a systolic-diastolic component (like the flow in the arterial end of an AV fistula). Other authors such as Federici A. also stated these characteristics in the flow curve along the IMAs graft [11]. At distal end, dominant systolic flow becomes the dominant diastolic flow. In general, this study showed that DU is safe, reliable, and reproducible when it comes to the evaluation of coronary flow through the IMAs graft. Although there are no guidelines that define the reference value for native and grafted IMAs and the distinction between diseased of normal. Several well conducted studies can be used for developing local vascular diagnostic criteria. For example, mentioned authors above quantify flow, resistance index, and diameters of IMAs. In grafted IMAs diameters do not increase significantly (2.21 ± 0.18 vs. 2.27 ± 0.22 mm). There was a significant increase of flow volume in IMAs in the early postoperative period (39.77 ± 21.59 vs. 25.96 ± 13.17 mL/min) and resistance index RI/RI decrease (1.43 ± 0.46 vs. 4.20 ± 1.49). They also compare DU and TTFM: MGF (TTFM) has a moderate correlation with the flow volume on DU. Probably because TTFM was recorded intra-OP and DU post-OP. The same was for PI¹⁰.

What we are interested in when analyzing the flow is the presence of stenoses/occlusions or competition. The authors described the phenomena and quantified pathologies so that may have a practical place [8]. Evaluation near the anastomosis is the best location to evaluate competitive flow. It is challenging due to the existence of two flows (native and in the graft) and due to myocardial contraction. Evaluation for degree of graft stenosis by DU is possible, but it is unreliable in moderately severe diameter stenoses (50–70%). Because, with these stenoses at rest, the flow velocities do not increase and the basic flow curve value mostly retains the physiological shape. In the case of severe stenoses, there is a deformation of the flow curve. Authors Hata M et al. monitored the correlation of stenoses on angiography and DU and monitored the diameters of IMAs and the characteristics of the curve in stenosis [12]. Doppler waveform along LIMA shows a different pattern from proximal part to the distal one. Trend over time in post-OP period shows LITA diameter increase in the first-month post-CABG (initial diameter 1.71 ± 0.72 vs. diameter after 30 days 1.99 ± 0.31 mm). The diastolic peak velocity and diastolic/systolic velocity ratio (D/S) in the postoperative early phase also increased by 0.26 ± 0.08 m/sec. and 1.54 ± 0.04, respectively. Also, there is a decrease in diastolic peak velocity and D/S ratio of less than 1.0 (D/S ratio > 1.0 was associated with a good angiographic finding).

Also, there is a difference in the flow curve in severe stenoses in the proximal and distal part of the IMA graft. Proximal stenosis (imaging distal to the stenosis) results in a monophasic flow curve with a preserved systolic component (loss of diastolic). This is a typical pathophysiological feature of peripheral artery disease progression. Namely, in the periphery, as the disease progresses, the three-phase curve first disappears and becomes biphasic and then monophasic, and finally the flow stops. On the other hand, the distal stenosis (imaging proximal to the stenosis) retains the biphasic shape due to the elasticity of the IMS, but the diastolic component is reduced. In stenoses of IMAs >75%, there are additional characteristics of the flow curve at PW. First of all, there is an increase in PSV at rest if we are near a stenosis (as a double/triple increase in flow). If, on the other hand, we record an increase in the peak of diastolic velocity (PDV), the risk of graft stenosis is small. There is also the PDV/PSV ratio as a useful marker of the presence of significant stenosis (and it is not dependent on the imaging angle) In case of occlusions, the flow is interrupted (with minimal flow into the occluded graft until it thromboses). Occlusion is present when there is no flow or when we have two sharp curves “to and from” the anastomosis with minimal amplitude. According to the works of Bach et al. a normal IMA graft has an increase in diastolic phase from proximal to distal [13]. The proximal PDV/PSV ratio is 0.6 and the distal ratio is 1.4. The cutoff is >1 as a sign of patent graft. Absence of this increase or reduction of PDV/PSV ratio < 0.5/0.6 is a predictor of severe graft stenosis [14]. These values correlate with angiographic regular findings [12]. The next important parameter of the flow curve of IMAs grafts is Diastolic Fraction (DF). It is the proportion of diastolic flow to the entire flow, i.e., velocity time integral - VTI (i.e., the percentage of the area under the curve that belongs only to diastole). The formula is DF = DVTI/(DVTI + SVTI). Where DVTI is diastolic VTI, DSTI is systolic VTI. The use of DF is more accurate than the absolute value of DVTI, and has less inter and intra-observer variability. A cutoff for hemodynamically significant stenosis (>70%) at rest is DF < 50% [15]. DF significantly positively correlates with invasive measurements [16].

Other authors also described the use of TTE in imaging LIMA grafts [17]. They detect flow in 80% of cases. Flow was biphasic (systolic and diastolic) in normal grafts or if the stenosis is <70%. The biphasic pattern is predominantly diastolic. In stenoses of IMAs >70%, the flow pattern is predominantly systolic. Diastolic fraction (DF) of VTI in normal IMAs and stenosis <70% was 0.81 ± 0.11 and in stenosis >70% it was 0.25 ± 0.06. That is, DF < 0.5 predict severe graft stenosis with a sensitivity and specificity of 100%. Also Ratio PSV/EDV in normal grafts and stenoses <70% was 0.61 ± 0.31 and in stenoses >70% it was 3.21 ± 0.49. This ratio predicted severe stenosis with a sensitivity and specificity of 100 and 90%, respectively.

James J et al. showed that TTE probes can image the IMAs graft in 81% of cases. Such high percentages generally encourage doctors to use non-invasive tools when patients have borderline pathologies or when recurrence of anginal symptoms after CABG is present [18]. Thus, they noticed, like other authors, two flow patterns through the grafts, both of which were biphasic (systolic and diastolic). In a normal graft or moderate (<70%) stenosis, blood flow velocity was maximal during diastole. In patients with native IMA or in severe (>70%) graft stenosis, blood velocity was maximal during systole, and low velocities were recorded during diastole. The diastolic fraction of the velocity time integrals in normal grafts was 0.77 ± 0.07 and 0.27 ± 0.01 in severe stenosis of grafts. 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 in normal grafts was 0.54 ± 0.26 and 3.45 ± 0.74 in diseased ones. A systolic-to-diastolic peak velocity ratio > 1 predicted severe stenosis with a sensitivity of 100% and specificity of 85%. Authors conclude that TTE allows the identification of the LIMA grafts and measurement of blood flow. Compared with patent grafts or those with moderate lesions, severe stenoses demonstrated different Doppler velocity patterns. Velocity patterns of severely stenosed grafts were similar to those in the ungrafted LIMA control group.

As we can see ultrasound is a totally non-invasive and non-infectious method, unlike coronary angiography. It can also be used to record the functionality of the grafts and the state of the coronary circulation. And simple curve characteristics such as flow rate can be used in the analysis of graft functionality. An example is the study of Hirata et al. which showed blood flow velocity at the anastomotic sites of 83 ± 228 cm/s in stenotic cases and 59 ± 28 cm/s in normal anastomotic cases [17].

4.1.4 Flow reserve

As noted above, flow measurement of IMAs grafts can be performed at rest and during stress testing. This is important because gafted IMAs flow velocities far from stenosis may not be changed in patients with significant stenosis if performed during rest (basal flow). That is why the assessment of the grafted IMAs stenosis is more reliable if we calculate so-called flow reserve. Knowledge from invasive studies transferred to DU can be used in such situations. So, in maximal coronary dilatation, the flow increases much more in normal blood vessels than in stenosed ones (both in ml/min and cm/sec). In this way, moderately severe stenoses can be revealed. Generally, during the stress test, there is an increase in the demand for oxygen, and an auto-regulatory increase in the flow rate, which is a confirmation of the patent graft. Coronary vasodilators such as dipyridamole (which acts on resistance vessels) or nitroglycerin (NTG, which acts on capacitance vessels) can be used for this purpose. With the use of both drugs, the diastolic component of the PDV flow curve increases without an increase in PSV if the findings are normal. Dipyridamole also increases the heart rate and diameter of the LIMA-graft, with increase in vessel capacity. NTG increases PDV in LIMA, without effects on diameter. Some non-pharmacological measures can also affect the flow curve like hyperventilation and Valsalva maneuver (but only in native IMAs). Thus, during hyperventilation, blood velocity increased while during the expiratory effort of the Valsalva maneuver, there is a decrease in mean flow [19]. These changes in flow are also present in TEE measurements of native coronary artery stenoses [20]. Here authors caused a stress test with dipyridamole infusion, and the test showed that there was no increase in flow in any severe stenosis. But, that increase occurs in patent-normal grafts. Dipyridamole baseline PDV and DV ratios were in normal patients on LAD 3.22 ± 0.96 and 3.04 ± 0.88 and in stenosed patients 1.46 ± 0.45 and 1.48 ± 0.49.

Other authors also showed that the DU measurement of IMAs grafts and the stress test represent a simple protocol and have promising results [21]. Constantly present criticisms of minimally invasive direct coronary artery bypass surgery (MIDCAB) in terms of slightly worse graft patency are a good reason to present the results of flow reserve and its connection with graft stenoses. Advances in ultrasound have made it easier to record the flow of IMAs grafts. Monitoring of patency and flow reserve should be part of regular ambulatory monitoring. In this study, flow reserve was measured after adenosine infusion. Patent IMAs grafts had predominantly diastolic flow at rest with PDV 48 ± 20 cm/s, with diastolic TVI 21 ± 10, with PDV/PSV ratio 1.3 ± 0.6, with DF (TVI) of 70% ± 11. Native IMAs (contralateral, high resistance circulation) had a resting flow of PDV/PSV ratio of 0.2 ± 0.1, with a DF (TVI) of 30% ± 10. Occluded IMAs grafts at rest had no flow or a minimal systolic peak was present. After adenosine during the stress test, there is an increase in flow in normal IMAs grafts: aPDV 105 ± 28 cm/s, diastolic TVI 37 ± 19 cm, adenosine/baseline flow ratios of 2.4 ± 0.9 and 2.0 ± 0.7, the diastolic flow velocity reserve was inversely related to baseline diastolic flow.

In the end, we have an interesting statement in the work of Dubey et al. comparing TTE and angiography in measuring severe stenoses of IMAs graft [22]: “It is suggested that angiography may be reserved for cases in which Doppler echocardiography fails to visualize the internal mammary artery or reveals an abnormal flow pattern.” Even we may not fully agree with this statement, because the use or exclusion of additional diagnostic tools requires an additional clinical picture. In this paper, the authors compared angiographic data with Doppler echocardiography findings and scintigraphy. They visualized via TTE IMAs in 98% of cases. In 9% there were abnormal angiograms. Thallium scanning in patients with suboptimal angiographic results but non-severe stenosis (4%) showed no evidence of myocardial ischemia in the left anterior descending artery territory.

4.1.5 Limitations

Limitations of the use of DU (pre-OP/post-OP) in the detection of stenosis/occlusion or competitive flow with or without anginal signs are related for several reasons. First one is the staff and second is guidelines/protocols. An expert is needed to monitor the patient pre-OP and post-OP. Even he can miss borderline stenoses if changes in PSV, EDV, and various ratios are not detected at the right spot. Also, error is possible if recordings were not compared with previous or baseline velocities and flow ratios. Due to these limitations, flow monitoring in the grafts should be supplemented by flow monitoring in the venous arm of the coronary circulation (coronary sinus). This volumetric calculation can attenuate hardly detectable marginal stenoses measured at rest and unreachable venous grafts (outside OR). When dealing with guidelines and protocols there are not enough well-conducted studies that compare DU and CT angiography in graft stenosis. However, above mentioned non-invasive measurements can serve as base for local protocol development and for screening of failing grafts or native vessel progression.

4.1.6 Intra-OP

Intraoperatively, the most useful and surgeon-friendly equipment for flow measurement is TTFM. Other vascular Doppler devices can give the same information but use is limited to additional knowledge about vascular ultrasound and flow calculation. TTFM is a device that uses ultrasound media delay for calculation of flow. It does not use Doppler effects (reflection from erythrocytes) for generating flow curves and is widely used outside medicine. The principle of ultrasonic velocity measurement is simple. Probe send signal, through vessel, on opposite side to reflect and return to calculate time delay, giving flow velocity. By using probe with specific diameter device calculate flow in ml/min. The basics of using TTFM are easy to learn [23]. Values that suggest a patent graft are flow >15 c/min, PI <5.0–3.0, and Diastolic Filling 60–80% (for left-sided vessels) and 45–55% (for right-sided vessels). For maximum data extraction multiple measurements are needed. Primarily because early identification of suboptimal grafts and early revision will minimally impact operative quality. Often during OPCAB we perform TTFM after anastomosis with/without proximal snares, with/without stabilizer, and after protamine. In addition to the above, in ONCAB, we measure flow during arrest and on/off pump. But why do we have suboptimal grafts instantly? There are several reasons for instant IMAs graft low/no flow: graft/anastomosis stenosis and competitive flow. The main reason for stenosis is intra-OP errors, which can be repaired by reanastomosing. The main reason for competitive flow is the presence of moderate native vessel stenosis and patent anastomosis per se. Competitive flow reduces the blood flow within the graft and leads to the string phenomenon. Our unpublished experiences show that competition results in occlusion of either a native LAD or LIMA graft within 30 days, and therefore special attention should be paid to it. Clinical importance of those conditions is in protocol consequences. If there is low flow, the surgeon snares the native coronary artery proximally and distally to the anastomosis and record TTFM. If flow increases significantly, it means that competitive flow is present. But, this is often at risk because vessel spasm or plaque mobilization can occur. If flow does not increase then stenosis or low runoff is present. Often surgeon in this situation reanastomose vessel which can improve the blood flow, except if present competitive flow or severe distal vascular bed disease. Differentiation between stenosis and competitive flow using TTFM is difficult because common values that surgeons use are mean flow and PI [24].

Besides them, other data can be extracted from the flow curve for this differentiation. That’s: DF, and graft flow FFT (Fast Fourier Transformation) ratio results in F0/H1 and F0/H2 (FO fundamental frequency, H1 first harmonic). The author found that when the cause is competitive flow, there is systolic backflow, and F0/H2 > 14.89. When the cause is anastomotic stenosis, the waveform maintains a bimodal state and F0/H2 is in a normal state - 1.17. So, to avoid unnecessary re-anastomosis it is important to look for additional value during TTFM. In general, as mentioned surgeons look for basic TTFM parameters: mean flow, PI, and DF. Low flow, high PI, and low DF mean graft problem, but, it does not identify cause. For differentiation between competitive flow and stenosis, other data from waveform shape and FFT ratio (especially F0/H2) can be used. They are not available in standard TTFM machines, but new versions will probably overcome that.

4.1.7 Epicardial probes

In addition to flow measurements through the supraclavicular and parasternal windows, DU probe measurements can be used. The authors in this study compared epicardial probe Doppler data with angiography [25]. They revealed that the percentages of systolic and diastolic reverse flow (%sRF, %dRF), as well as PI, were predictors of early postoperative graft failure. They show that mean flow velocity < 12.5 cm/s, %sRF > 9.3%, %dRF > 4.1%, and PI > 4.4 were predictors of early graft failure. Nowadays it is easy to substitute an epicardial probe with a conventional linear ultrasound transducer (under sterile covering). Some manufacturers deliver epicardial probes with TTFM probes. Such an upgrade, in addition to the advantages described below, is very useful when trying to save the LIMA graft at all costs. The most common fear of surgeons is the use of a dissected LIMA versus the fear of discarding a patent LIMA because of discrete hematoma. These two conditions can be differentiated with epicardial probes [26].

Epicardial probes can be used for flow quality control of IMAs or venous bypasses like any other DU probe. But, because of accessibility, we can record the morphology-geometry of the anastomosis with these probes. Rune et al used TTFM and epicardial ultrasound in the analysis of anastomotic geometry. They measured the length of the anastomosis (DA), the diameter of the LIMA (DM), the diameter of the LAD at the toe of the anastomosis (D1) and 5 mm distal to the anastomosis (D2), as well as the ratios between these variables [27]. Flow was analyzed by Doppler and TTFM. They show that it was possible to accurately confirm the patency of anastomoses and that mean ratios of D1/D2, DA/D2, and DM/D2 were 0.89 ± 0.13, 3.01 ± 1.04, and 1.32 ± 0.32, respectively. Those data are useful because they show that anastomosis geometry (orientationally, the length of the anastomosis should be >2 times the diameter of the recipient vessel) correlates with patency.

4.1.8 Graft configuration

Han et al. analyzed different bypass configurations and their effect on coronary flow [28]. Total flows (by TTFM) in IMAs are best when BIMA grafts are used ipsilaterally, i.e., to the corresponding native vessels (LIMA left and RIMA right). If we use BIMA grafts for the left basin, then it does not matter which IMAs feed the LAD (the flow is the same MF, PI, or DF). If we compare the flow in native RIMA with free RIMA, the flow in native RIMA is better. On the other hand, there is no difference in flow and resistance indices (PI, MGF, and DF) through free RIMA (e.g., if it was short or injured) whether it was sutured proximally to the aorta or the LIMA. Olaf et al. analyzed the flow reserve of prox IMAs in composite configurations and showed that it is adequate for second arterial grafts [29].

Shahzad G et al. analyzed the flows in the arms of the composite graft [30]. Generally, the flow reserve of the proximal IMA is sufficient for composite grafts. Namely, in the radial arm of the composite graft, flows of 161 ± 81 mL/min were measured, and in the IMAs arm 137 ± 57 mL/min (together it is 298 ± 101 mL/min). In free flow through both grafts simultaneously, the proximal flow is 226 ± 84 mL/min. It is less by 24 ± 14% compared to the mentioned total flow. The first sum of flow can be considered as maximum and the second sum of flow proximal to the arms as current. After creating the anastomosis, the flow in the radial branch alone was 88 ± 49 mL/min and through the IMAs branch 60 ± 45 mL/min while they had flowed at the same time (measurement before branching, i.e., prox IMAs 140 ± 70 mL/min). Using these numbers, we can determine the surrogate of coronary reserve, i.e., graft flow reserve. Thus, flow reserve (prox IMAs before and after anastomosis with both open branches) is 160%, i.e., 1.6. This flow reserve value is consistent with earlier studies that measured a value of 1.8 during the first week and first month after CABG [31]. In addition to the configuration of the graft, the flow can be affected by subclavian disease or spasm of the IMAs, so it is necessary to carefully handle the graft during harvesting and during the recording of the flow and exceptionally use dilatation instruments for that purpose (probe 1.5 mm).

In addition to assessing the adequate configuration of the grafts in terms of potency for the expected increase in flow, the increased risk for the development of competitive flow should also be taken into account if we use grafts in a composite configuration. Namely, the longer the graft, the lower the perfusion pressure at the end, the longer the delay of the pulse wave, and the greater the risk of developing competition. The consequences of competition are not negligible because competition disturbs the anterograde flow in diastole and the delay of the pulse wave in IMAs promotes retroflow in early systole. The consequences of this oscillating flow around the anastomosis are seen on the endothelium, where NO and prostacyclins are released, causing the string sign, i.e., physiological vasoconstriction of that part of the graft. It usually develops in medium-severe stenoses of native coronary arteries with patent anastomosis and leads to graft or native vessel failure. This phenomenon is even more pronounced with composite grafts because, in addition to the aforementioned anastomoses, they deepen the phase delay of the pulse wave. So that the prevention of competition is related not only to the selection of the target (moderately severe vs. severe stenoses) but also to the configuration of the grafts (avoid composite grafts in moderately stenotic coronary arteries, particularly in the RCA territory).

4.1.9 CAD and PAD

Patients with combined coronary artery disease and peripheral artery disease (CAD and PAD), especially severe aortoiliac occlusive pathology, require special caution when selecting grafts [32]. In these conditions, the IMAs represent part of the collateral flow to the lower extremities (e.g., Sy Lerish). So, careful avoidance of IMAs grafts is a wise approach. In all these situations, angiographic visualization of the internal mammary artery is crucial to confirm whether it is involved in collateral circulation or not. The situation is similar for subclavian disease, regardless of whether LIMA-LAD is planned or performed earlier [33]. In order to avoid the development of LAD steal syndrome, it is necessary to systematically examine the IMAs inflow artery.

4.1.10 CAD and PCI

In addition to the intraoperative quality control of the bypass, we have the opportunity to evaluate the flow through the stents. A typical example is RCA stenting in complex and long-term operations. Then when we suspect that the stent has become blocked/bent (during OPCAB manipulations or during valvular surgery). Then, after installing the stabilizer foot, we prepare the native RCA on the most accessible part and measure the flow through TTFM. If everything is normal, the flow is typical diastolic. In our center, the flow measurement of native RCA is performed only by indication.

In addition to the configuration of the grafts and PAD, the measured flow can be affected by the use of CPB. The variations are small but should be kept in mind. In general, the flow is higher and the resistance index is lower with ONCAB (probably on the basis of ischemia and consequent vasodilatation) [34]. Here, the authors quantified OPCAB/ONCAB flow differences. MGF (mean graft flow) was higher in ONCAB versus OPCAB (for all grafts 32 vs 28 mL/min; 30 vs 27 mL/min for arterial grafts and 35 vs 31 mL/min for venous grafts). PI was lower in the ONCAB group (2.1 vs 2.3, for all grafts). Diastolic fraction (DF) was slightly lower in the ONCAB group (65 vs 67.5%). The backflow was also lower in the ONCAB group (0.6 vs 1.3) with trends similar to MGF and PI for venous and arterial grafts. We must not forget that cardiac surgeons sometimes refrain from revising possible stenoses. Because of the above, flow objectification should be daily, especially when we have borderline stenoses.

Other situations that can affect flow in the LIMA graft are therapeutic radiation (in carcinomas). Although we have papers suggesting that the flow is equal, the durability of such grafts is no longer the same as native unirradiated ones which must be kept in mind.

4.2 Venous side of coronary circulation

In order to complete flow evaluation of whole coronary circulation before and after CABG, it is also necessary to quantify changes in coronary sinus flow using noninvasive method (TTE). Of course, we can identify all three coronary arteries and their segments but because of accuracy and reliability, it is easy to measure flow on venous side of coronary circulation—CS blood flow [35]. Another reason is that imaging at the venous side incorporates all new blood from unreachable to DU grafts Post-OP. Currently, there are no TTE-based diagnostic criteria for Ischemic Heart Disease of follow-up. Nor pre or nor post, but there are publications that describe TTE as a screening method for Coronary Artery Disease in isolated populations [36, 37, 38]. Quantifications of CS blood flow changes after CABG can be performed using Echocardiography [39, 40, 41]. Generally, CS flow increase after CABG for 20% per gram of myocardium per graft [42]. Anny sudden decrease during follow-up period can suggest a new native or bypass flow problem occurrence.

TTE for coronary sinus recording is performed in lateral decubitus position. With B mode, 4 chamber view is used to find CS after dorsal probe angulation. CS can then be found in atrioventricular grove and traced to distal segment. Then probe is rotated until minimal angle of insonation is recorded (about 30 degree). In this position, we have to record VTI with Doppler angle of 60% and in B/M mode diameter of CS under same systemic hemodynamic parameters [43]. After few cardiac cycles, VTI is traced and average value is memorized together with CS diameter and systemic hemodynamic data. For recordings in short axis view the patient is in same position. Distal CS segment is recorded behind mitral valve with minimal probe angulation. Using this view Doppler angle can be even more decreased. Banjanovic B. et al. used following formula for flow calculations [44]: Q = (VTI × CSA) × HR. Where Q is flow, (VTI × CSA) is stroke volume, HR is heart rate, CSA = Cross Section Area of CS (πr², r = radius if vessel is round), VTI is average velocities calculated automatically after tracing heart cycle velocities. They adopted above formula as CSA = CSA × 0,39. Reason for this is ellipsoid shape of cross-section area of CS with a ratio 2:1. As they had recorded a longer cross-section diameter in four-chamber view calculation of cross section area was πab (“a” was half of longer CS diameter, “b” was half of shorter CS diameter that mean “b” was 1/4 of measured diameter).

There are published data about immediate flow changes after CABG recorded using TTE and TEE [2, 42]. Immediate coronary flow increase after triple CABG is up to 50% of native coronary flow. Increase is much higher if there are no microvascular disturbances. Over time, flow increase due to graft and capillary bed vasodilatation. The practical importance of repeated imaging of flow through the CS with TTE is in the trend. Because trend is constant and positive during the early post-OP period. So, any sudden stop in increase or even decrease could lead to suspicion of graft/native vessel failure. That’s why Banjanovic et al. performed TTE flow imaging before and two times after CABG (1 and 6 postoperative day [44]. They measure CS diameter, Velocity Time Integral (VTI), and systemic hemodynamic data. Data needed for LV mass calculation were recorded once. The results shown below were obtained at rest due to the careful selection of patients (all patients had at least one severe stenosis of native coronary arteries). They show preoperatively very low CS flow (mean 181 ± 72 ml/min) as others. In published data, we found that normal CS flow is higher (mean 327 ± 125 ml/min). During follow up according to their results there was a constant increase in CS flow from first postoperative day (mean 276 ± 79 ml/min) to last measurement on six postoperative days (mean 355 ± 99 ml/min), in Figure 1.

Figure 1.

Flow through CS during repeated measurements (ml/min), after CABG.

Other published data show even higher immediate postoperative/post-PCI CS flow (mean 451 ± 102 ml/min) [45]. Our understanding of this flow discrepancy is in flow calculation. We take into account ellipsoid shape of terminal CS (2:1) during calculation, so our flow data were less for 0,39 times. Also using angle of 60 degrees during the Doppler VTI calculation brings us with error of 10%. CS flow discrepancy could be because of recording position wary in terminal CS segment from patient to patient. But this was probably nullified by repeated measurements, as it will be during future follow-up in outpatient departments. According to the literature, net flow through three grafts after CABG is about 110–130 ml/min. As net native coronary artery flow is about 250 ml/min during rest, volume of new blood in coronary artery bed after CABG is 50% of net native flow during rest [2, 46]. For even more individualization of results, they performed CS flow calculation over LV mass. They found preoperatively low CS flow per gram of LV (mean 0,68 ± 0,30 ml/gram/min). Which corresponds to invasive measurements. And there was also CS flow increase over time per gram of LV, from day 1 mean flow 1,13 ± 0,35 ml/gram/min to day 6 mean flow 1,30 ± 0,46 ml/gram/min. All data above support correct indication for CABG because coronary vessel bed was empty and constantly filled with new blood in following 6 days. Other variables recorded were: CS diameter was preoperatively mean 7,5 ± 1,1 mm, first postoperative day mean 8,2 ± 1,4 mm, and 6 postoperative days 9,3 ± 1,3 mm. As flow is proportional to the forth square of diameter, those changes can have huge impact on flow calculations. In this paper, calculation error was decreased three times by fact that we record elliptic segment of CS. Heart rate also has a direct positive impact in flow calculations. In our results, heart rate increase from mean 68 (±8,9) preoperatively to 87 (±11) first and same at the sixth postoperative day. As heart rate was constant postoperatively, the impact on error in flow calculation was decreased in Table 1.

	Preoperative	1 postoperative day	6 postoperative day
Flow (VTI syst + diast) ml/min	181 (±72)	276 (±79)	355 (±99)
Flow (VTI syst + diast) ml/min/per I graft	120 (±10)	240 (±69)	232 (±55)
Flow (VTI syst + diast) ml/min/per II graft	175 (±65)	281 (±94)	378 (±89)
Flow (VTI syst + diast) ml/min/per III graft	186 (±78)	269 (±52)	352 (±100)
Flow (VTI syst + diast) ml/min/per IV graft	234 (±43)	379 (±165)	448 (±8)
Flow (VTI syst + diast) ml/gram LV	0,68 (±0,30)	1,13 (±0,35)	1,30 (±0,46)
CS VTI (syst + diast)	12,0 cm (±3,4)	11,9 cm (±3,9)	12,2 cm (±3,3)
CS VTI (max)	35 cm (±7,7)	41 cm (±14,5)	34 cm (±5,9)
CS diameter	7,5 mm (±1,1)	8,2 mm (±1,4)	9,3 mm (±1,3)
Heart rate	68 (±8,9)	87 (±11)	87 (±11)

Table 1.

Flow through CS during repeated measurements after CABG.

CS-Coronary Sinus LV-left ventricle, RV-right ventricle.

Authors show that it is possible to detect effect of graft number on CS flow. That there is evidence of maximal amount of new blood in the coronary bed after four bypasses and that same amount of new blood was whatever the patient gets three or two grafts, in Figure 2. Meaning that the increase in flow by graft number is linear at the beginning and after that is stepwise.

Figure 2.

CS flow comparison during repeated measurements, effect of graft number.

5. Conclusion

Functional evaluation of each graft is crucial during the follow-up of patients who underwent CABG intervention. The gold standard investigation for the assessment of conduit stenosis remains CA, but this is an invasive technique with many risks. Different non-invasive methods had been recently developed for coronary graft evaluation.

DU seems to be the best non-invasive tool to assess graft patency and coronary sinus flow. In the early post-OP after CABG there is a constant increase in coronary flow. The increase is significant overall after classifying per gram of LV and graft number. That’s why measurements via DU should cover IMAs grafts, coronary sinus and include intra-OP TTFM measurements of all other grafts (GSV/SSV or radial artery). This means that intra-OP (TTFM) graft flow measurements must be available to outpatient departments.

What data should be recorded for stenosis or competition evaluation? For practical use depending on used equipment that’s are as show below. For DU—IMA native/grafts diameter: PSV, EDC, PSV/EDV, EDV/PSV, DF, ml/min, PI/RI, VTIsistolic, VTI diastolic, heart rate. For DU—coronary sinus: VTI, diameter, ml/min. For TTFM—all grafts: ml/min, PI/RI diastolic filling, retro flow pik. For epicardial probe: all grafts: same as for DU and %sRF, % dRF, D1-anastomotic length, D0-vessel diameter, ratio D1/D0.

DU has good reproducibility, ease of use, and significant correlation with angiographic results which make DU the best technique to be adopted in clinical ambulatory practice in order to follow-up with patients who underwent CABG. Thus, the periodicity of flow measurement should be regularly increased from weeks to every 6 months. If during follow-up period flow decrease or a flow curve deform that is indication of the disease progression. In the situation we should offer a re-evaluation to all patients depending on clinical picture (e.g. ERGO, scintigraphy or coronary re-angiography, etc.).