Invasive coronary angiography is the gold standard investigation for graft quality assessment during surgery [7,27]. All newer techniques should be compared with ICA for their validity [28]. A study by Hol et al. [8] performed intraoperative angiography for 186 patients with 427 grafts. The CABG was performed on-pump, off-pump through sternotomy, and off-pump MIDCAB. Out of 427 grafts, 18 grafts (4.2%) were revised based on angiography results. The revision rate was 1.1%, 6.4%, and 6.5% for on-pump, off-pump through sternotomy, and MIDCAB groups, respectively. Twelve out of eighteen grafts had a problem in conduits while six out of eighteen had a problem in the distal anastomosis area. The authors believed intraoperative angiography could save many grafts which otherwise could have been occluded, and encouraged the use of intraoperative angiography [8]. Another study by the same group showed a 5% graft revision rate based on on-table angiography [7]. At the same time, they believed that sometimes it was hard to interpret the intraoperative angiography results because not all negative findings affected future follow-up patency. The sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) of intraoperative angiography, compared with follow-up angiography in their study, were 42%, 82%, 38%, and 84%, respectively [7]. Izzat et al. reported an 8% graft revision rate based on intraoperative angiography results [10]. The unavailability of an angiography machine in the operating room (OR) in most hospitals is one of the challenges for its routine use, and the availability of more hybrid operating rooms in hospitals will address this issue. A hybrid OR serves as a complete OR as well as a complete catheterization laboratory. It allows surgeons to perform angiography on completion of surgery and detect abnormal grafts. It also offers the opportunity to solve the problem with either PCI or surgical Intervention before the patient leaves the OR [27].

Intraoperative fluorescence imaging is an intraoperative invasive angiography-like modality but uses fluorescent indocyanine green (ICG) dye. This is less invasive compared with catheter-based angiography. The dye is injected through the central vein and images of grafts, anastomosis, and native coronary vessels are achieved. ICG dye has been in clinical use for many decades but its use in quality assessment in CABG surgery started at the beginning of the 21st century [29,30,31]. Desai et al. performed a study of 120 patients with 348 grafts who performed ICG angiography intraoperatively. They found that 5 out of 120 patients (4.2%) had major graft problems which needed revision of anastomosis or construction of new grafts. Six patients with twenty-two grafts underwent X-ray angiography as a pilot study for a future randomized clinical trial. The sensitivity, specificity, PPV, and NPV of ICG compared with conventional angiography for more than 50% graft stenosis were 100% each. ICG detected 3 grafts to be dysfunctional and 19 grafts patent, and conventional angiography gave the same results. The authors concluded that ICG angiography could detect patients who needed graft revision and would otherwise have gone unnoticed [24]. Another study by Waseda et al. compared intraoperative transient time flow measurement (TTFM) with the IFI system. They analyzed 137 patients with 507 grafts with ICG angiography and found 21 grafts with unsatisfactory TTFM results to be acceptable by ICG angiography. At the same time, six grafts with acceptable TTFM results were considered graft failure by ICG angiography which needed immediate revision. The authors believed the IFI system enabled on-site assessment of the grafts and anastomosis and provided functional and morphological information [32]. Another OPCAB study by Oliver et al. included 38 patients with 124 grafts. Out of 124 grafts, 107 were analyzed. Out of 107 grafts, 4 grafts needed revision. Three grafts had anastomotic stenosis and one graft had conduit dissection. The authors believed that ICG images were equivalent to angiography without catheter insertion. Additionally, ICG images could help to show the course of coronaries that would be difficult to find in obese patients. They believed ICG-based imaging technology was feasible and easy to use for the assessment of graft quality and patency [33]. Desai et al. in a randomized control trial, compared ICG with conventional angiography in 46 patients with 139 grafts. There were two false-negative and no false-positive results on the ICG angiogram. Compared with a conventional angiogram, an ICG angiogram had 100% specificity and 83% sensitivity. They encouraged the use of ICG angiography during CABG surgery [23].

Even though coronary angiography is the gold standard and most reliable method for intraoperative graft patency, when considering the cost, availability, additional surgery time, and risk, it may not be a very practical and feasible way of assessing graft patency. Assessing the patency of the graft through a transient time ultrasound is a noninvasive, simple, and reliable method of flow measurement of the graft while the patient is still in the OR [34]. TTFM was first used for quality control in CABG in 1998 by Walpoth et al. and has been constantly in use since [35]. The 2018 ESC/EACTS guidelines on myocardial revascularization recommend the use of TTFM for intraoperative quality assessment [36]. TTFM can detect technical errors and provide the opportunity to correct the problem during surgery. The three main parameters measured during TTFM and its recommended values are mean graft flow (MGF) of >15 mL/min, pulsatility index (PI) of less than 5, and diastolic filling (DF) of greater than 50% [22,37,38]. A revision is recommended if two of the above three criteria are not met. A study by Walker et al. [25] compared TTFM with intra or postoperative angiography in 160 LITA-LAD grafts. The proportion of FitzGibbon type A grafts was 152/160 (95%), and 8 grafts were defective with 3 being type B and 5 being type O grafts. Based on the above given parameters of TTFM, no graft would have been identified as defective. According to the FitzGibbon grading, a graft with normal flow or less than 50% reduction in diameter is type A, a graft with greater than 50% stenosis or reduction in diameter is type B, and a graft without flow that is considered to be occluded is type O [39]. In the study by Walker et al. out of three parameters measured by TTFM, only MGF was significantly different between patent and defective grafts with a mean flow of 34.3 ± 16.8 mL/min and 23.9 ± 12.5 mL/min, respectively. Considering the above values as a standard, their study would have predicted six false positives based on MGF, one false positive based on PI, and two false positives based on DF. They believed that high flow and lower PI were not a guarantee against graft dysfunction [25]. A study by D’Ancona and colleagues for 161 patients with 323 distal anastomoses who underwent OPCAB where all patients’ bypass grafts were evaluated using TTFM intraoperatively, found that 32 grafts (9.9%) needed revision based on the results obtained by TTFM. The decision to either accept the graft or revise the graft was based on the flow curves and PI or both. All the revised grafts were found to have technical errors such as thrombus, kinking of the graft, intimal flap, or dissection. They found the TTFM very helpful and strongly recommended the use of TTFM in both off-pump and on-pump CABG surgeries [11]. Normally the mean flow in the sequential grafts is higher than the individual graft [40]. A study by Yang et al. introduced a new TTFM method which they named flow reduction TTFM for sequential grafting [41]. In sequential grafting, because of the high flow in the graft, less than critical anastomotic defects might not significantly decrease the flow and, therefore, might be missed by conventional TTFM. They compared the conventional and new (flow reduction) TTFM methods for all sequential anastomoses. In the conventional method, the probe would be placed 2 cm proximal to the target anastomosis with normal graft flow, while in the new method, a bulldog clamp was applied a few centimeters distal to target anastomosis to reduce the flow, and then the flow was measured the same way as the conventional method. Two distal anastomoses in the middle of the sequential grafts were found defective on flow reduction TTFM, which were missed by conventional TTFM and were revised subsequently. They believed that the temporary flow reduction method increased the sensitivity of TTFM for less than critical anastomotic defects of sequential grafting [41]. A randomized trial (GRIIP) compared the imaging group (revision based on TTFM and IFI) and control group (revision based on surgeon judgment and conventional approach) with 78 patients in each group and the same number of grafts in both groups. The major adverse cardiovascular events (MACE) (MI, repeat revascularization, and death) were similar between the groups (7.7%). After one year, angiography was performed for 55 patients with 160 grafts, and 52 patients with 152 grafts, in imaging and control groups, respectively. Single or multiple graft occlusion was comparable between the groups (30.9% imaging group) and (28.9% control group). They believed the use of TTFM and IFI was safe but did not lead to a reduction in graft occlusion at 1-year follow-up [22]. A study by Leviner et al. suggested different cut-off values of TTFM for on-pump vs. off-pump CABG [42]. An RCT comparing TTFM to angiography showed nine grafts’ TTFM values to be normal but had greater than 50% stenosis on the angiogram. Meanwhile, two grafts had abnormal TTFM values which were normal on the angiogram [23]. The limitation of TTFM is its lower sensitivity to detect greater than 50% occlusion of the graft [23,25,43]. One limitation of the TTFM was its inability to detect technical errors distal to the anastomosis. Normal TTFM values were shown in the case of distal LIMA-LAD anastomosis occlusion because of a technical error with the preserved retrograde flow in LAD and very limited or no antegrade flow in LAD distal to the anastomosis [23].

High-frequency epicardial echocardiography (HEE) was developed for the assessment of morphological and functional information of the graft and anastomosis quality. Several clinical reports have established its efficacy intraoperatively [26,44,45,46,47]. Anastomosis assessment is performed by using a high-frequency linear probe, with a frequency ranging from 10 to 13 MHz. A study by Suematsu et al. used HEE and power Doppler imaging for the assessment of coronary arteries and graft anastomoses during CABG. The maximal luminal diameter of the graft at the site of anastomosis was obtained intraoperatively by using power Doppler imaging and was compared with a postoperative coronary angiogram. Excellent correlation was observed between these modalities and they concluded that the patency of anastomosed graft could quickly be evaluated by epicardial echocardiography [44]. Similarly, a study by Budde et al. evaluated three different types of intentionally created coronary anastomosis construction errors (suture cross-overs, purse-string, or deep toe stitch) by using HEE with a linear probe of 13-MHz frequency. HEE enabled the detection of construction errors with high specificity and sensitivity [45].

On the contrary, Hol et al. conducted a study where they compared the measurement of graft diameters and graft quality assessment using epicardial ultrasonography with those measured using intraoperative angiography. They found poor correlations between the two methods for graft diameter. However, epicardial ultrasonography detected 13% of the abnormal grafts while ICA found 23% of the abnormal grafts. The specificity, sensitivity, PPV, and NPV for HEE compared with ICA were 90%, 22%, 40%, and 79%, respectively. The authors concluded that although epicardial ultrasonography was useful in assessing graft quality intraoperatively, angiography was superior in identifying grafts that require revision. Therefore, epicardial ultrasonography has the potential of assessing graft morphology but its ability to predict graft that needs revision should be further evaluated in comparative studies [26].

Transesophageal echocardiography (TEE) was one of the accepted modalities used for the assessment of grafts intraoperatively. Several studies by Hiroshima University Hospital performed an intraoperative assessment of grafts by using TEE [48,49,50]. In 90.4% of cases, the left internal thoracic artery (LITA) was successfully visualized using TEE. The clamp-and-decamp test, when combined with TEE, allows for the evaluation of the LITA’s patency, stenosis, or the existence of a remnant branch [48]. Additionally, another study by Orihashi et al. evaluated the quality of saphenous vein or gastroepiploic artery grafts to the posterior descending artery (PDA) [49]. The grafts were successfully visualized in 95.2% of cases. Postoperative ICA results were well correlated to the intraoperative TEE results. The flow signals were easily detected in 17/20 grafts, but were hard to detect in the other 3 patients, which were determined to be occluded on postoperative angiography [49].

In a retrospective study of 51 patients by Kuroda et al. who underwent LITA-LAD grafting evaluated by TEE intraoperatively and were examined with ICA postoperatively [50], the researchers measured the flow velocity intraoperatively by TEE after anastomosis of the LITA graft. The LITA was detected in 88% of patients intraoperatively with TEE. Peak and mean velocities and velocity time integral ratios were measured and the critical values for them were 0.60, 0.73, and 1.06, respectively. The specificity for peak velocity, mean velocity, and velocity time integral ratio, was 92%, 94%, and 89%, respectively, while the sensitivity was found to be 100% for each [50]. The authors concluded that the assessment of LITA quality by TEE intraoperatively was a useful and powerful tool during CABG surgery.