Coronary atherosclerosis results from complex interactions of coronary endothelial cells (EC) and smooth muscle cells (SMC) with circulating inflammatory cells (monocytes and macrophages) and lipids, especially in the setting of certain hemodynamic conditions. Atherosclerotic lesions are characterized by accumulation and transformation of certain plasma lipoproteins (including low-density lipoproteins [LDL] and remnants of triglyceride-rich lipoproteins), complex carbohydrates, fibrous tissue, inflammatory cells, SMCs, and necrotic cell debris in the intimal space underneath a monolayer of ECs.

Atherosclerosis tends to occur in regions of arteries such as bifurcations, where shear stress caused by turbulent blood flow changes the cellular alignment of ECs and increases the permeability of the intima (through multiple pathways including BMP, TGF-beta, WNT, and Notch), with subsequent uptake of plasma LDL and TG-rich lipoproteins either by trans-endothelial transport or diffusion at the cell-cell junctions. The subsequent activation of the ECs occurs in response to oxidation of lipoprotein lipids and other inflammatory mediators, resulting in the expression of P-selectin, E-selectin, VACM1, and ICAM1 that promotes the adhesion of monocytes, recruitment of leukocytes, and chemotactic factors such as CCR2 and CCR5. Upon entering the intimal region, monocytes differentiate into macrophages and start a process of proliferation and uptake of oxidized LDL, giving rise to cholesterol-engorged macrophages. These “foam cells” frequently undergo apoptosis or necrosis, contributing to a growing “necrotic core” consisting of cholesterol esters, cholesterol crystals, and cell debris that increases the likelihood of lesion rupture. SMCs transform from a contractile to a proliferative state and migrate in the region underlying the ECs to form the “fibrous cap” that protects the lesion from rupture. SMC can also differentiate into osteochondrogenic descendants that deposit calcium phosphate mineral.

Fibrolipoid plaques may become thick enough to encroach on the lumen of the artery, producing a stenotic lesion. Probably episodically and sometimes over years, new material is deposited on the luminal side of the plaque, resulting in further narrowing and sometimes complete coronary occlusion. Small blood vessels form around and within the plaque. Gradual regression of plaque enlargement, seen clinically as regression of stenoses in a few patients, and development of collateral coronary blood flow can result in at least partial spontaneous restoration of antegrade regional myocardial blood flow.

Hemorrhage may occur suddenly within a plaque; this may suddenly increase the degree of coronary stenosis and precipitate acute myocardial infarction (MI) or unstable angina (UA) pectoris. Thrombosis occasionally complicates the coronary atherosclerotic process, generally when there is luminal narrowing. Sudden complete obstruction may result, and it is generally agreed that acute thrombotic occlusion is the genesis of acute MI in most patients. Rapid recanalization frequently follows this process. Platelet aggregation within the lumen of an already narrowed coronary artery may induce thrombosis or suddenly narrow the lumen and provoke an acute MI or UA, and it may play a role in development of the atherosclerotic plaque itself. Platelet aggregation releases thromboxane A ₂ , an extremely potent vasoconstrictor. Thus, interrelationships among atherosclerotic narrowing, platelet aggregation, and coronary spasm are important.

The atherosclerotic process usually affects multiple coronary arteries. In 1975, Gensini reported that 40% of patients with CAD sufficient to lead to cineangiographic study had important stenoses in all three major coronary arteries, and in an additional 30% of patients, two vessels were involved. Ninety-five percent of patients with complete occlusion of one artery had important stenoses in at least one of the other two arteries.

Atherosclerotic CAD usually involves the proximal portion of the larger coronary arteries, a biological reality that permits coronary artery bypass grafting to be effective. These stenoses occur particularly at or just beyond sites of bifurcation. Thus, stenoses in the main trunks of the LAD, circumflex (Cx) coronary artery, and RCA often involve the first of the secondary branches (first diagonal branch of LAD, obtuse marginal branch of Cx artery, and posterior descending branch of RCA). When CAD is more extensive in the main trunks, origins and first portions of secondary branches may be involved. Diffuse distal disease severe enough to render the patient unsuitable for CABG is uncommon, although it can occur, especially in diabetic patients. In 10% to 20% of patients with atherosclerotic CAD, the left main (LM) coronary artery is importantly stenotic (>50% narrowing). LM stenosis is often concomitant with multivessel disease; indeed, approximately two-thirds of patients with triple-vessel disease also have LM stenosis. In patients with multivessel disease, concomitant LM stenosis commonly involves the distal bifurcation and the ostia of the LAD and left Cx.

Occasionally, a major coronary artery may lie within the myocardium. This is most common in the middle third of the LAD, but sometimes, one or all the obtuse marginal branches of the left Cx artery or the ramus intermedius are buried in muscle throughout their course. These portions of artery are typically free of severe atherosclerotic changes.

When myocardial blood flow is sufficiently impaired in relation to myocardial oxygen demands, myocardial necrosis occurs. Diminished cellular glycogen, relaxed myofibrils, and sarcolemmal disruption are the first ultrastructural changes and are seen as early as 10 to 15 minutes after the onset of ischemia. Mitochondrial abnormalities are observed as early as 10 minutes after coronary occlusion by electron microscopy and are progressive. The resultant infarction may be subendocardial —that is, not involving the entire thickness of the ventricular wall, but only the inner third. In its most extreme form, subendocardial infarction may be diffuse and result from multiple-vessel disease. More often, however, subendocardial infarcts are regional and result primarily from a stenotic lesion in one or two coronary arteries. These infarcts are generally less extensive than so-called transmural infarcts but still have serious implications. A transmural MI involves the entire thickness of the ventricular wall. Transmural infarction usually results from a sudden increase in luminal narrowing or complete obstruction of the artery supplying that area or a sudden generalized increase in myocardial oxygen demand in the presence of a severely stenotic coronary artery. Although categorization of acute infarctions as subendocardial or transmural is convenient, most transmural MIs are not homogeneous but contain islands of viable muscle of varying numbers and size.

The process of infarction is complex. Animal studies indicate that some myocardial cells die after 20 minutes of complete coronary artery occlusion and that extensive myocardial cell death occurs after 60 minutes. Although these time frames may vary, some reperfusion generally occurs within the ischemic area of myocardium within minutes of onset of acute ischemia, particularly in the zone between ischemic and nonischemic myocardium (border zone). If this spontaneous reperfusion occurs within 3 to 4 hours, the amount of necrosis is limited, at times substantially, infarct size is reduced, and mortality is decreased. The process is complex because, in addition to these beneficial effects, spontaneous reperfusion can result in hemorrhage, edema, and ventricular electrical instability. ^,

Healing of the acute MI leaves a scarred area of myocardium. In most cases, this area is a mixture of fibrous tissue and viable myocardial cells in varying proportions. Such scarring is evident from (1) intraoperative inspection of areas of previous infarction at the time of CABG and (2) change from akinesia or dyskinesia to hypokinesia or normal wall motion in some LV wall segments when patients go from a symptomatic to an asymptomatic state after percutaneous coronary intervention (PCI) or CABG. When the scar is almost all fibrous tissue, it is usually large, and the affected region of the LV wall may remain akinetic or become aneurysmal.

In the aggregate, myocardial scarring leads to LV systolic and diastolic dysfunction and adverse ventricular remodeling, which may result in ischemic mitral regurgitation and the syndrome of chronic heart failure (HF) with elevated right atrial and jugular venous pressure, hepatomegaly, and fluid retention (see Chapter 10 ).

Several studies have emphasized the dynamic nature of coronary atherosclerotic plaque as a fundamental feature of CAD. ^, Fissuring or rupture of atherosclerotic plaques is probably the genesis of the acute coronary syndromes (ACS) termed unstable angina and acute MI . When this occurs, mural or occlusive coronary thrombi often coexist and contribute further to acute instability.

“Vulnerable plaques” are more a function of their composition than size. Coronary stenoses that produce less than 50% reduction in lumen diameter are often the site of the atherosclerotic plaque rupture that precipitates UA or acute MI. More severe stenoses also undergo plaque rupture, and total vessel occlusion may occur. However, an acute ischemic episode does not always develop in this scenario, possibly because severely stenotic lesions are long-standing and have stimulated development of a protective collateral circulation.

Certain atherosclerotic plaques appear to have a higher risk of rupture than others. These plaques are characterized by relative softness, a high concentration of cholesterol and cholesterol esters, and a lipid pool that tends to be situated eccentrically. Macrophages appear to contribute to plaque destabilization by amplifying inflammation and producing proteases that attack the fibrous cap. MI can also occur because of endothelial erosion. Neutrophils, although rare in lesions, appear to promote endothelial erosion via neutrophil traps and secretion of matrix metalloproteinases.

Rupture is through the cap of the plaque, and areas in which the cap lacks underlying collagen support, so-called thin cap fibroatheromas (TCFA), seem particularly vulnerable. ^, Improved spatial resolution of CT scanners and novel image reconstruction algorithms enable Computed Tomographic Coronary Angiography (CTCA) the quantification and characterization of atherosclerotic plaques. State-of-the-art CT imaging can, therefore, reliably assess the extent of CAD and differentiate between various plaque features, including plaque composition, plaque attenuation and pattern, calcifications, and remodeling. Features associated with high-risk plaques are also predictive of adverse clinical events.

CAD is usually first suspected with development of the symptom complex of angina pectoris or an acute MI, occasionally because of electrocardiographic (ECG) evidence of a silent acute MI, a positive ECG response to a graded exercise test, or sudden death with resuscitation. Occasionally, CAD is first suspected because of cardiomegaly and symptoms of chronic HF without any other obvious cause or prodrome.

The precise nature, location, duration, and severity of any chest pain are determined by carefully questioning the patient. Precipitating causes and maneuvers that relieve the pain are noted, as are any recent changes in pain pattern. Findings on physical examination are usually nonspecific.

Many noninvasive tests, beginning with a chest radiograph and ECG at rest and during exercise and then proceeding to more complex studies, are currently used to identify and quantify CAD and its sequelae. Such tests cannot yet define extent or distribution of anatomic coronary disease with great accuracy. From a surgical standpoint, therefore, properly performed coronary angiography remains the definitive diagnostic procedure. Although traditionally this could have been only performed via an invasive procedure, recent advances in contrast-enhanced CTCA make this less invasive technology an attractive substitute for conventional invasive coronary angiography in screening patients with suspected CAD. Increased spatial and temporal resolution, together with CTCA-derived physiologic assessment, a surrogate for fractional flow reserve (FFR) measured by invasive intracoronary pressure and velocity wires, render possible decision-making about revascularization solely based on computed tomography.

Although information from various invasive and noninvasive imaging studies is essential, proper assessment of LV function must integrate details from the patient’s history and physical exam. An ejection fraction (EF) of 35% has a different implication when accompanied by minimal LV enlargement seen on a chest radiograph than when enlargement is marked. An EF of 30% is much more ominous when accompanied by important elevation of jugular and right atrial pressure with hepatomegaly and fluid retention than when these pressures are normal. Exercise capacity may be variable in patients with similar EF and is prognostically important. Finally, in patients presenting with severely impaired LV function, assessment of myocardial viability is useful to differentiate scar from hibernating or stunned but viable myocardium.

Important associated conditions such as hyperlipidemia, arterial hypertension, and diabetes, and a history of MI, kidney disease, or smoking should be noted. Diabetic patients are at higher risk of adverse events after coronary artery bypass surgery. Glycosylated hemoglobin A1c (HbA1c) is commonly used to track control of diabetes, and it is a marker for the average blood glucose level over a three-to-four-month period prior to the measurement. Preoperative HbA1c testing may allow for more accurate risk stratification in patients undergoing both on-pump and off-pump coronary surgery. ^, Accurate glycemic control can also affect long-term survival after CABG, as evidenced by the importance of HbA1c in the SYNTAXES study. The American Heart Association (AHA) and the European Society of Cardiology (ESC) recommend a target of glycated HbA1c less than 7% as a class I recommendation. ^, Thus, diligent control of blood glucose is a key part of long-term medical management for CABG patients.

Because arteriosclerosis is the cause of CAD, its presence elsewhere in the circulatory system should be investigated. A history suggesting transient cerebral ischemic attacks or stroke, particularly when carotid bruits are present, must be carefully pursued. A history of intermittent claudication and presence of diminished femoral, popliteal, or pedal pulses are indicative of peripheral arterial occlusive disease. The thoracic and abdominal aorta are examined for possible aneurysm or occlusive disease and the presence of calcifications. Renal and pulmonary function should also be evaluated.

Invasive coronary angiography remains the gold-standard imaging study for CAD. During selective injection of radiocontrast medium into each of the coronary arteries, cineangiography images are obtained in multiple orthogonal views. Their quality must be sufficient to permit detailed assessment of both coronary ostia and all major and minor branches of the left and right coronary arterial systems. Multiple views are necessary to overcome the overlap and foreshortening due to compression of the 3D anatomy of the coronary vasculature onto a 2D film. Despite still being considered the “gold standard,” coronary angiography is an imperfect imaging methodology. Severity of a visualized stenotic lesion may be underestimated, and diameter of vessels distal to a stenosis is often underestimated. Moreover, there is considerable inter- and intraobserver variability. Finally, coronary angiography illustrates not the artery itself but rather the opacification of its lumen; accordingly, diagnosis and grading of coronary lesions derive from comparison to a “normal” reference segment, whose selection introduces bias and error.

Assessment of coronary arteries at operation by external palpation or probing of the open vessel cannot substitute for coronary angiography. When the arteries cannot be adequately filled by contrast media, however, or the available study is incomplete and cannot be repeated (this should be uncommon), intraoperative observations and epicardiac high-frequency ultrasound (HIFU or ECUS) can be useful to supplement angiographic findings.

Recording and reporting data from invasive coronary angiography.

Methods of recording and analyzing the data are crucial. A 75% cross-sectional area loss (50% diameter) is considered an important but moderate stenosis, and a 90% cross-sectional area loss (67% diameter) is considered severe ( Fig. 9.1 ). According to the 2021 ACC/AHA/SCAI Coronary Revascularization Guideline, a visually estimated diameter stenosis severity of ≥70% for non–left main disease and ≥50% for left main disease has been used to define significant stenosis and guide revascularization strategy. An angiographically intermediate coronary stenosis is defined as a diameter stenosis severity of 40% to 69% and generally warrants additional investigation to assess physiologic significance.

Diagrammatic representation of relationship between two methods of estimating severity of coronary artery stenosis.

(From Brandt PW, Partridge JB, Wattie WJ. Coronary arteriography: a method of presentation of the arteriogram report and a scoring system. Clin Radiol . 1977;28:361.)

Extent of important coronary artery stenoses has conventionally been summarized as “single-vessel,” “double-vessel,” or “triple-vessel” disease, usually with left main CAD considered as a separate category. This chapter uses the terms single-system, two-system, and three-system disease because each coronary system (LAD, Cx, and RCA) consists of several coronary branch vessels. Use of the term system is, therefore, more accurate than vessel .

These classifications have been criticized ^, because they do not indicate the amount of LV myocardium rendered ischemic by the lesions. For example, stenosis in the LAD system has a different significance when it lies at the origin of a large first diagonal artery than when it involves the middle third of the LAD beyond its major septal and diagonal branches. A single stenosis in the proximal portion of the Cx artery varies in significance depending on whether this artery is dominant. Single-system disease involving the proximal RCA has a different implication from that involving only the posterior descending branch of the RCA. Furthermore, the assessment of functional significance of the stenosis is usually lacking because invasive FFR measurements are performed in only a minority of diagnostic coronary angiograms.

The American College of Cardiology/American Heart Association Guidelines for Invasive Coronary Angiography adopted a classification system that divides the coronary arteries into a total of 27 segments ( Fig. 9.2 ). Recently, attention has been largely focused on the “complexity” of the coronary artery lesion, integrating anatomic, clinical, and functional characteristics for triaging between PCI and CABG. Angiographic features contributing to increasing complexity of CAD are summarized in Table 9.1 .

Coronary artery segments used by the SYNTAX investigators. 1. RCA proximal: From the ostium to one-half the distance to the acute margin of the heart. 2. RCA mid: From the end of first segment to acute margin of heart. 3. RCA distal: From the acute margin of the heart to the origin of the posterior descending artery. 4. Posterior descending artery: Running in the posterior interventricular groove. 5. Left main: From the ostium Of the LCA through bifurcation into left anterior descending and left circumflex branches. 6. LAD proximal: Proximal to and including first major septal branch. 7. LAD mid: LAD immediately distal to origin of first septal branch and extending to the point where LAD forms an angle (RAO view). If this angle is not identifiable this segment ends at one-half the distance from the first septal to the apex of the heart. 8. LAD apical: Terminal portion of LAD, beginning at the end of previous segment and extending to or beyond the apex. 9. First diagonal: The first diagonal originating from segment 6 or 7. 9a. First diagonal a: Additional first diagonal originating from segment 6 or 7, before segment 8. 10. Second diagonal: Originating from segment 8 or the transition between segment 7 and 8. 10a. Second diagonal a: Additional second diagonal originating from segment 8. I I. Proximal circumflex artery: Main stem of circumflex from its origin of left main and including origin of first obtuse marginal branch. 12. Intermediate/anterolateral artery: Branch from trifurcating left main other than proximal LAD or LCX. It belongs to the circumflex territory. 12a. Obtuse marginal a: First side branch of circumflex running in genera/to the area of obtuse margin of the heart. 12b. Obtuse marginal b: Second additional branch of circumflex running in the same direction as 12. 13. Distal circumflex artery: The stem of the circumflex distal to the origin of the most distal obtuse marginal branch and running along the posterior left atrioventricular groove. 14. Left posterolateral: Running to the posterolateral surface of the left ventricle 14a. Left posterolateral a: Distal from 14 and running in the same direction. 14b. Left posterolateral b: Distal from 14 and 14a and running in the same direction. 15. Posterior descending: Most distal part of dominant left circumflex when present.

Table 9.1

Angiographic Features Contributing to Increasing Complexity of CAD

Multivessel disease Left main or proximal LAD artery lesion Chronic total occlusion Trifurcation lesion Complex bifurcation lesion Heavy calcification Severe tortuosity Aorto-ostial stenosis Diffusely diseased and narrowed segments distal to the lesion Thrombotic lesion Lesion length >20 mm

CAD , Coronary artery disease; LAD , left anterior descending.

In 2009, the Synergy between PCI with Taxus and Cardiac Surgery (SYNTAX) Score (SS) was introduced to aggregate angiographic data on the anatomic complexity of CAD (calcifications, bifurcation lesions, chronic total occlusion, etc.) into a single semiquantitative score. Important limitations of the SS include the cumbersome scoring system required for each lesion and the interobserver variability in its calculation. Accordingly, the 2021 AHA Guidelines state that calculation of the SS may be useful to guide revascularization (Class 2b recommendation). As CTCA imaging technology has substantially evolved in the last decade, the anatomic SS can also be reliably calculated with this noninvasive methodology.

The SS was subsequently revised in 2013 to incorporate the interaction of anatomic features with clinical characteristics and comorbidities to support more evidence-based decision-making by the heart team (SYNTAX Score-II). This score was redeveloped again in 2020 (SS-II 2020) to better predict the benefit of CABG versus PCI over 10 years. In 2011, after the FAME 1 trial (Fractional Flow Reserve Versus Angiography in Multivessel Evaluation) demonstrated that treatment based on FFR measurement in addition to angiography can lower rates of major adverse cardiac events (MACE) in patients with multivessel CAD undergoing PCI, an FFR-guided SS (Functional SYNTAX Score, FSS) was developed, essentially recalculating the original SS counting only ischemia producing lesions as assessed by FFR (FFR ≤0.8). FSS demonstrated better predictive accuracy for MACE following PCI than SS. Similarly, more recent studies have attempted to recalculate the SYNTAX Score II 2020, integrating the FFR results for intermediate lesions. Investigators found that by eliminating nonsignificant lesions (FFR ≤0.8), the functional SYNTAX Score II 2020 resulted in lower scores and downgraded the extent of functional CAD such that a significant portion of patients (40%) were reclassified to have an absolute risk difference (ARD) <4.5% (i.e., “equipoise”), comparing predicted outcomes after CABG versus PCI. Similarly, an interesting substudy of the FAME 3 trial showed that by calculating the FSS, up to 50% of all PCI patients of the trial had a low score (FSS≤22), compared to only 20% when functional information was excluded, and SS alone was calculated.

It is important to note, however, that whereas outcomes with PCI are improved using FFR in assessing coronary lesions, FFR has not been found to be useful in guiding CABG. This is likely because the consequences of stenting a nonflow-limiting lesion can include increased risk of MI and death in the event of stent thrombosis/restenosis, whereas the same is not true for CABG since graft closure due to native competitive flow through a nonflow-limiting lesion is usually a clinically silent event. Moreover, CABG protects against future plaque rupture in nonflow-limiting lesions and prevents future MI, whereas PCI does not. Thus, attempts to downgrade complexity of coronary disease with FFR may steer patients with multivessel disease inappropriately toward incomplete revascularization with PCI and away from CABG, with negative consequences. Despite the aforementioned, FFR information may be very useful to the Heart Team, helping to individualize the use of multiple arterial conduits.

Recently, application of machine learning algorithms to the SYNTAX study identified unsuspected but potentially relevant important prognostic factors predictive of 10-year mortality, such as C-reactive protein, patient-reported preprocedural physical and mental status, gamma-glutamyl transferase, and hemoglobin. Whatever the recording and reporting methods, data and images from coronary angiography should be critically reviewed by the operating surgeon before deciding for or against surgery, while deciding the surgical plan for the patient, and again immediately before/during the CABG operation.

Although conventional invasive CA (ICA) remains the gold standard to determine extent and severity of CAD and indications for CABG, CTCA has shown great technologic improvements over the last 2 decades ( Fig. 9.3 ). New generation CT scanners, like the GE Revolution scanner (Revolution CT; GE Healthcare, Milwaukee, WI, USA), have 160-mm coverage in the z-axis and 0.28s rotation speed, allowing for acquisition of the whole heart within a single beat. CT imaging, as opposed to ICA, generates a data set volume consisting of many 0.5-mm cross-sectional images in transaxial orientation, which can be reconstructed in any plane desired. Curved multiplanar reformation (MPR) allows for display of the entire coronary vessel in a single 2D image, with the ability to rotate the artery around a focal point. Maximum intensity projections can be helpful to reduce image noise and/or visualize a longer vessel segment. Volume-rendered 3D images allow for rapid interpretation of structural relationships, such as vessel course in relation to other cardiac or vascular structures. Improvement in the image acquisition process decreases the dose of radiation and allows images to be obtained even among patients in atrial fibrillation (AF) or with tachycardia. CTCA may, in the future, play a pivotal role as a “one-stop-shop” in screening, diagnosis, decision-making, and treatment planning with benefits for patients and the health economy ( Fig. 9.4 ).

(A-C) Multidetector computed tomographic volume rendering images show significant stenoses of major coronary arteries (arrows) , suggestive of three-system disease. These coronary lesions (arrows) were confirmed on conventional coronary arteriography (D–E), and patient underwent coronary artery bypass grafting. LAD, Left anterior descending coronary artery; LCx, left circumflex coronary artery; RCA, right coronary artery.

(From Lee et al. )

Pathway for evaluation of a patient with chest pain using cardiac CT. ECG, Electrocardiogram.

Modern CTCA could potentially provide the physician with noninvasive angiography, fractional assessment with CT-FFR by HeartFlow Inc. (Redwood City, CA, USA), combined with analysis of myocardial perfusion, plaque burden, presence of high-risk plaque, coronary vascular inflammation, regional LV wall motion, myocardial scar and fibrosis, percent myocardium at risk, and risk scores such as the Leaman score. Prospective multicenter studies have demonstrated the diagnostic accuracy of CTA, with a sensitivity between 85% and 99% and a specificity between 64% and 92%, in patients with suspected but unconfirmed CAD. ^, The PACIF study showed that in patients with stable CAD and adequate image quality, CT-FFR provides superior functional assessment of coronary stenosis compared with CTCA, SPECT, or PET imaging. Although most evidence on CTCA is derived from patients with chronic CAD, the European Society of Cardiology gives CTCA a Class IA recommendation in selected patients with non-ST elevation MI (NSTEMI) as an alternative to invasive coronary angiography. The SYNTAX III revolution trial suggested the feasibility of decision-making and planning based solely on noninvasive CCTA and clinical information. Additionally, the trial introduced the SYNTAX III score, which, with integration of CT-FFR and clinical characteristics, provides treatment recommendations based on the predicted 4-year mortality.

A pilot survey of the surgeons involved in the SYNTAX III revolution trial suggested the feasibility of planning surgical coronary revascularization solely based on CCTA for patients with left main or three-vessels disease (3VD). Finally, the FASTTRACK CABG study (NCT04142021) was designed to assess the feasibility of CTCA and FFR-CT to replace ICA as a sole surgical guidance method for planning and performing CABG in patients with complex CAD. The primary safety endpoint included a 30-day postoperative CTCA to evaluate the patency of bypass grafts and topographic adequacy of the revascularization procedure ( Fig. 9.5 ). This proof-of-concept trial demonstrated the feasibility and safety of performing CABG surgery based solely on noninvasive assessment of the coronary arteries ( Fig. 9.6 ), with an 82.9% agreement between the ICA and CTCA and 99.1% feasibility.

Postoperative CT angiogram demonstrating in panel A a sequential left internal mammary artery (LIMA) graft to the diagonal branch of the LAD (D1) and the LAD. Panel B demonstrates the CTA study of a radial artery (RA) graft from the LIMA to the obtuse marginal branch of the circumflex (OM) . In panel C, the right internal mammary artery (RIMA) is anastomosed to the posterior descending branch of the right coronary artery (PDA) and the 3rd obtuse marginal branch of the circumflex (OM3) .

Overview of the FAST TRACK CABG trial from Serruys et al. ICA , Invasive coronary angiography; CABG , coronary artery bypass grafting; CCTA , coronary computed tomography angiography; LAD , left anterior descending coronary artery; RA , radial artery; OM , obtuse marginal branch; LIMA , left internal mammary artery; D1 , diagonal branch.

Intravascular ultrasound (IVUS) uses a high-frequency miniaturized ultrasound transducer positioned on the distal tip of a coronary artery catheter to provide detailed cross-sectional images of the coronary vessel wall ( Fig. 9.7 A). Unlike coronary angiography, which details only luminal encroachment, IVUS provides images of the atherosclerotic plaque, characterizes its composition, and assesses severity of stenosis ( Fig. 9.7 B). When compared with formalin-fixed and fresh histologic specimens of coronary arteries, it correlates significantly ( P <.0001) with coronary artery cross-sectional area ( r = 0.94), residual lumen cross-sectional area ( r = 0.85), and percent cross-sectional area ( r = 0.84). The proximal and distal extents of the reference segment with the least plaque burden should be identified, and the minimal lumen area (MLA) should be localized to measure the luminal stenosis accurately. IVUS is the most sensitive and specific intravascular tool to assess calcium, an important piece of information in planning PCI. Fibrous plaques appear with an intermediate or high echogenicity depending on the density of the fibrous material. High lipid content (“soft plaques”) results in low echogenicity. Thrombus is mostly localized intraluminally or directly attached to the lumen and has an irregular appearance. IVUS is useful in the calculation of the plaque burden by dividing the plaque-media area by the external elastic lamina cross-sectional area. In the PROSPECT trial, plaque burden emerged as the most relevant independent correlate of MACE related to nonculprit lesions in patients with ACS.

Coronary intravascular ultrasound (IVUS) . (A) Schematic of IVUS catheter within blood vessel, cross-sectional imaging plane, and resultant image. Transducer types are shown. (B) Vessel wall thickness assessed by IVUS imaging. Mild (left, arrows) and moderate (right, arrows) amounts of intimal thickening are shown. (C) Angiographic underestimation of disease. Although angiogram (top) shows only mild luminal irregularities, two sites in left anterior descending coronary artery (arrows) show major arteriosclerosis by IVUS (below) .

(From Kimura and colleagues and Nissen and Yock. )

IVUS is useful in determining the need for CABG in situations when the severity of coronary artery stenosis cannot be precisely determined by angiography, particularly for intermediate-grade left main stenosis (LMS) and proximal LAD disease ( Fig. 9.7 C). The 2021 AHA/ACC revascularization guidelines give a Class 2a recommendation for the use of IVUS in indeterminate LMS; deferral of intervention in lesions with MLA≥6.0 to 7.5 mm ² (and 4.5–4.8 mm ² in Asian patients) is considered safe.

Following PCI, IVUS can detect the findings that are associated with an increased risk of restenosis and stent thrombosis, such as lesion under expansion, malposition of stent struts, coronary dissection, and remaining disease burden at the stent edge. The RENOVATE COMPLEX-PCI trial showed that among patients with complex CAD, IVUS–guided PCI led to a lower risk of a composite of death from cardiac causes, target vessel-related MI, or clinically driven target-vessel revascularization than angiography-guided PCI.

The concept of FFR was introduced by Nico H.J. Pijls and Bernard de Bruyne in 1996. FFR is a simple, reliable, and reproducible physiologic index of lesion severity in patients with intermediate stenosis and is another method to determine the need for revascularization, especially with PCI, in equivocal situations, particularly in the setting of stenosis of the left main coronary artery. The concept of FFR is illustrated in Fig. 9.8 . Pressure measured distal to the stenotic coronary lesion during maximum hyperemia (Pd) divided by mean aortic pressure (Pa) correlates with maximum myocardial blood flow in the presence of a stenosis Q ˙ S divided by the normal maximum myocardial blood flow Q ˙ N . The iFR, a resting physiologic index, is the instantaneous wave-free ratio (in diastole) of coronary pressure distal to the coronary lesion divided by the aortic pressure. The potential advantage of iFR is that it obviates the use of adenosine because it does not require a state of maximal hyperemia. The FAME 2 trial showed a significant benefit of PCI over medical therapy in patients with abnormal FFR with respect to death, MI, or urgent revascularization. The 2021 AHA/ACC revascularization guidelines give Class 1A recommendation for the use of FFR or iFR in patients with angina (or angina equivalent), undocumented ischemia, and angiographically intermediate stenosis to guide the decision to proceed to PCI. Deferral of PCI when the FFR is >0.8 or the iFR >0.89 is associated with low rates of long-term MACE, based on the results of the DEFER, DEFINE-FLAIR, and iFR-SWEDEHEART trials. However, according to the data from the VA CART Program, FFR is still underutilized in the United States (<20% of patients with intermediate lesions) because it is a time-consuming procedure that significantly adds to the cost of traditional invasive coronary angiography and can cause patients discomfort through the use of hyperemic agents. As a noninvasive alternative, postacquisition analysis of CTCA images by sophisticated software algorithms (such as HeartFlow Inc., Redwood City, CA, USA) can automatically perform lumen centerline assessment, luminal boundary determination, and myocardial mass assessment. CT-FFR is then calculated using a finite element mesh model and computational fluid dynamics methods. CT-FFR was shown to be accurate in the SYNTAX II trial and to have high specificity when combined with CTCA images in the detection of hemodynamically significant stenosis.

Concept of fractional flow reserve (FFR) . Pa , Mean aortic pressure; Pd , hyperemic distal coronary pressure; Q ˙ N , normal maximal myocardial blood flow; Q ˙ S , maximal myocardial blood flow in the presence of a stenosis.

(From Iwasaki and Kusachi. )

Although the role of FFR is well established in PCI decision-making, its capacity to guide surgical revascularization strategy is uncertain and controversial. On the one hand, it may be intuitive to select bypass targets according to the physiologic significance of the coronary lesion (for instance, to avoid issues related to competitive flow, graft occlusion, and return of angina). On the other hand, bypass surgery is intrinsically different from PCI: Surgery targets the distal, almost normal, portion of the vessel, and the procedural success of bypass is not influenced by target lesion complexity. Moreover, the use of FFR to inform surgical grafting carries the risk of jeopardizing the fundaments of a successful coronary operation: completeness of revascularization and protection of the coronary bed from disease progression. ^,

In addition, it must be remembered that the consequence of a failed stent in nonflow-limiting stenosis is MI, whereas the closure of a bypass graft distal to a nonflow-limiting lesion is almost always a clinically silent event. Finally, although competitive flow in native coronary arteries increases the likelihood of graft failure, this risk increases gradually as native flow increases; there is no sharp cutoff beyond which a bypass graft is doomed to failure. In a prospective study of 164 patients by Botman and colleagues, the patency rate at 1 year was influenced by the physiologic significance of the stenosis with an FFR cutoff value of 0.75 (91.2% vs. 79.8%; P >.0001); however, this finding had no clinical relevance as patients with patent or occluded bypass graft on nonsignificant lesions did not suffer from an excess of angina or coronary repeat interventions. In other words, occlusion of a graft to a vessel with a nonsignificant stenosis was a silent event.

On the contrary, retrospective studies have suggested a benefit of FFR-driven CABG in both graft patency and survival at 6 years. Two underpowered RCTs (GRAFFITI and FARGO) ^, and a meta-analysis of these trials suggest that FFR-guided revascularization led to fewer anastomosis and wider utilization of off-pump CABG (OPCAB), without difference in clinical outcomes compared to traditionally angiographic guided CABG. Finally, one area in which the application of FFR in CABG could be beneficial is the choice of conduits to be used to bypass a specific vessel. The IMPAG trial examined the impact of preoperative invasive FFR and multiarterial graft patency at 6 months, and results demonstrated a correlation between FFR and patency with a suggested soft cutoff of FFR 0.78 for arterial grafts to the left system and of 0.71 for arterial grafts to the right system. These new data may inform the safe, systematic expansion of multiple arterial grafting in coronary surgery.

In general, both severity and distribution of CAD tend to increase with time, ^, although the rate of increase is highly variable and difficult to predict, and regression also occurs in a small minority of lesions. ^, In general, over 2 years, in patients with already important stenoses, 20% of the stenoses increase in severity, and about half of patients develop important new lesions. ^, The mechanism of increase in severity is variable, but atherosclerotic plaque rupture and thrombosis play important roles. The closest approximation to a true “natural history” study of subclinical coronary atherosclerosis was recently reported by Fuchs and colleagues, who studied 9533 asymptomatic persons aged 40 years or older without known ischemic heart disease with a CTCA that was blinded to treatment and outcomes. CAD was defined as “obstructive” if causing stenosis less than 50% and “extensive” if it involved a third or more of the coronary tree. Obstructive disease was found in 10% of these asymptomatic subjects. At a median follow-up of 3.5 years, the risk of MI in asymptomatic subjects with subclinical atherosclerosis was increased at least eightfold in the presence of obstructive CAD (for obstructive-extensive adjusted relative risk [RR] 12.48 [confidence interval {CI} 5.50–28.12]; for obstructive-nonextensive adjusted relative risk, 8.28 [CI, 3.75–18.32]). Both noncontrast coronary artery calcium scoring and high-risk plaque features defined by visual coronary CTA analysis were each independently associated with elevated risk for acute MI.

The SCOT-HEART (Scottish Computed Tomography of the HEART) trial demonstrated a significantly lower rate of the combined endpoint of cardiovascular death or nonfatal MI (2.3% vs. 3.9% during 5-year follow-up) in patients in whom CTCA was performed in addition to routine testing (mostly exercise ECG). The DANE-HEART study (ClinicalTrials.gov: NCT05677386) is an ongoing RCT aimed at assessing if preventive treatment guided by CTCA reduces the risk of MI or death as compared with current primary cardiovascular prevention practice.

Not all of the usually accepted risk factors for presence of stenotic CAD have been helpful in predicting its rate of progression. ^{, ,} Aggressiveness of the atherosclerotic process seems to be a risk factor for progression. Surrogates include young age at presentation with symptomatic CAD, peripheral arterial disease, diabetes, hyperlipidemia, cardiometabolic syndrome, and HIV. ^{, , ,} Mohammad and colleagues studied the progression of coronary stenosis <50% in patients with a first-time coronary angiogram between 1989 and 2017 (n = 2,661,245 coronary artery segments in 248,736 patients) over 15 years using the Swedish Coronary Angiography and Angioplasty Registry. The stenosis progression and incidence rate were 2.6% and 1.45 (95% CI 1.43–1.46) per 1000 segment-years, respectively, in patients with ≥4 risk factors, especially in younger patients and women. The greatest progression was found in the proximal and middle segments of the LAD, especially in male diabetic patients (twofold increased risk). In the absence of risk factors, the progression and incidence rate was 0.6% and 0.36 (95% CI 0.34–0.39), respectively, increasing to 8.1% and 4.01 (95% CI 3.89–4.14) per 1000 segment-years, respectively. The specific risk factors with the highest impact on plaque progression were (1) obstructive (single vessel) CAD on index angiography, (2) diabetes, (3) active smoking, (4) hyperlipidemia, and (5) hypertension ( Fig. 9.9 ).

Impact of risk factors on progression of coronary atherosclerosis. Cumulative probability of stenosis progression from nonobstructive (<50% diameter stenosis) to obstructive (≥50% diameter stenosis) disease over 15 years in patients with vs. without diabetes (A), hypertension (B), hyperlipidemia (C), smoking (D). HR indicates hazard ratio and IR incidence rate.

(From Mohammad MA, Stone GW, Koul S, et al. On the natural history of coronary artery disease: a longitudinal nationwide serial angiography study. J Am Heart Assoc . 2022;11[21]:e026396.)

Nature of the atherosclerotic plaque is a risk factor because plaque rupture is frequently the inciting event leading to progression in severity of coronary artery stenosis. Eccentric positioning of the lipid pool within the plaque appears to predispose to rupture and thus progression in severity of stenosis. Rheologic factors play a role, and the more severe the stenosis, the more rapid the progression toward total occlusion. ^{, ,}

The PROSPECT study studied 697 patients with acute ACS who underwent three-vessel ICA and IVUS after PCI for the primary event. Subsequent MACE events were adjudicated to be related to either the originally treated culprit lesion or nonculprit lesions. At a median follow-up of 3.4 years, the investigators found that most nonculprit lesions responsible for subsequent events were angiographically mild at baseline. However, those lesions were characterized by a plaque burden of 70% or greater, a small luminal area of 4.0 mm ² or less, and a TCFA.

Chronic coronary syndrome (CCS) has been defined to better characterize patients with CAD. This definition has replaced the terms stable angina or stable ischemia heart disease. Conversely, ACS, including UA and acute MI, defines the unstable phases due to atherothrombotic events. The 2019 ESC Guidelines suggest a stepwise strategy for the initial diagnostic management of patients with angina and suspected CAD, and crucial to this approach is the assessment of event risk in every patient with suspected or newly diagnosed CAD to decide the most appropriate therapy. The process of risk stratification serves to identify patients at high risk for adverse cardiovascular events who will benefit from revascularization. High event risk is defined as a cardiac mortality rate >3% per year, and low event risk as a cardiac mortality rate <1% per year. Event risk stratification uses clinical evaluation and noninvasive and invasive testing for myocardial ischemia and coronary anatomy. High event risk features on different testing modalities include an area of ischemia ≥10% of the LV myocardium on PET or SPECT, ≥3/16 segments with stress-induced hypokinesia or akinesia on stress-echo, ≥2/16 segments with perfusion defects or >3 dobutamine-induced dysfunctional segments on CMR, three-vessel disease with proximal stenoses, LM disease, or proximal LAD disease on CTCA or ICA with FFR ≤0.8, iwFR ≤0.89 on invasive functional testing. ^,

Multivariable analysis is used to generate incremental risk factors for various unfavorable events after CABG, PCI, or medical treatment. However, those identified are often surrogates for more basic risk factors, and at times, several surrogates for the same basic risk factors appear. Box 9.1 presents the basic risk factors as currently perceived.

Reduced regional coronary flow reserve.

The risk of adverse events (death, MI, or revascularization) associated with a given coronary lesion is important in the decision-making between medical therapy and revascularization. As discussed later, both the anatomic distribution of CAD and physiologic analysis of coronary stenoses are helpful in estimating the possibility of complications. Moreover, the use of CTCA to measure CAD distribution, severity, and plaque analysis provides additional information for risk stratification and prognostication. ^,

Reduced regional flow reserve results from severity of the coronary arterial stenoses and number of coronary arterial systems with important stenoses with the left main coronary artery considered an additional “system.”

Coronary flow plays an important role in maintaining myocardial function. ^, CFR, the available vasodilator reserve capacity of the coronary circulation, evaluates all aspects of the coronary circulation as it assesses both epicardial blood vessel flow and microcirculation. It has shown consistently strong prognostic value for adverse cardiac events, especially when used in combination with FFR. CFR is the ratio of coronary blood flow during maximal vasodilation divided by coronary blood flow during resting conditions. It can be measured by either a temperature-sensitive guide wire applying the coronary thermodilution technique or a Doppler sensor-equipped guide wire measuring Doppler flow velocity. ^, The diagnostic accuracy of CFR for inducible myocardial ischemia on noninvasive stress testing is 81%, at an optimal cutoff value of 1.9, the same as the accuracy for FFR. ^, Several factors, however, have hampered the adoption of CFR assessment in clinical practice. CFR assessment may be associated with an increase in procedural time, and CFR is intrinsically sensitive to changes in hemodynamic conditions. Assessing CFR and FFR in combination provides more details into the pathophysiologic substrate of CAD. Although with a cutoff of 0.8 for FFR and 2.0 for CFR, the diagnostic accuracy for myocardial ischemia is equivalent for both tests, disagreement between the two occurs in up to 40% of cases. The occurrence of abnormal CFR with normal FFR (when the flow is hampered without a drop in pressure) might characterize two pathophysiologic substrates: dominant microvascular disease or dominant diffuse epicardial disease. The latter is at high risk for adverse events, and its optimal management remains to be elucidated. Results from the DEFINE-FLOW study suggest that in lesions that are FFR positive (≤0.8) and CFR negative (≥2.0), revascularization should not be deferred as the rate of adverse events is not noninferior compared to stenosis that is FFR and CFR negative. Finally, the concept of coronary flow capacity (CFC) has been introduced to stratify coronary lesions on the basis of the combination of both maximal coronary flow and CFR values. This is based on the assumption that myocardial ischemia originates when both maximal coronary flow and the reserve capacity of the coronary circulation are below ischemic thresholds, and as such, myocardial ischemia is unlikely once CFR or maximal flow is among normal values. Additional risk factors relating to reduction in regional CFR include (1) specific vessel(s) diseased, (2) location of stenosis within the vessel, and (3) severity of stenosis.

Studies from the 1970s and 1980s suggest that time-related survival for a heterogeneous group of patients with single-system stenosis is high, approximately 90% to 95% at 5 years ( Fig. 9.10 A). At 15 years, survival is about 50%. Single-system disease with stenosis in the RCA appears to confer better survival than can be expected with LAD disease, at least for 5 years (RCA 96% vs. LAD 92%). When single-system stenosis is in the LAD, a very proximal location (proximal to the large septal branch) imposes less favorable survival than more distal lesions (proximal 90% at 5 years vs. distal 98%). Although not conclusively demonstrated, more severe stenoses (>90%), especially those proximal in the artery, probably impose higher time-related mortality than less severe stenoses.

Survival of medically treated men with coronary artery disease, stable angina of at least 6 months duration, and less than severe left ventricular dysfunction enrolled in the US Veterans Administration (VA) Cooperative Study (solid squares) . For comparison, survival is shown of a population matched for age and gender from the 1976 US life tables (solid line) . Data for other groups of medically treated patients published earlier by Burggraf and colleagues, Oberman and colleagues, and Webster and colleagues are also shown. Lower survival in the last three groups may have been the result of less restrictive selection of patients than for the VA group and better medical treatment in the more recent VA group. Data from the Coronary Artery Surgery Study (CASS) , in which important stenosis meant a 70% diameter reduction, are also presented. These data include patients treated medically in the current era with all types of ventricular function. Left main coronary artery data from CASS refer to left main coronary artery plus triple-system disease. (A) Single-system disease. (B) Two-system disease. (C) Three-system disease. (D) Left main coronary artery disease.

(From Kirklin and colleagues. )

Patients with two-system disease as a heterogeneous group have lower survival than those with single-system stenoses, with 5-year survival of about 88% ( Fig. 9.10 B). At 15 years, survival is about 56%. When the LAD is one of the two systems, the same effects of location and severity as mentioned for single-system disease pertain. Differences in outcome between single- and two-system disease are not as great as those between two- and three-system disease.

As a heterogeneous group, patients with three-system disease have a 5-year survival without interventional treatment of about 70% ( Fig. 9.10 C) and a 10- and 15-year survival of about 60% and 40%, respectively. Factors affecting survival in patients with important single-system disease involving the LAD also affect survival in patients with three-system disease. Also, the greater the number of systems with important proximal stenoses, the lower the time-related survival: at 5 years, survival with no, one, two, and three systems with proximal stenoses is 71%, 64%, 51%, and 45%, respectively.

Important left main coronary artery disease imposes an even lower survival: 40% to 60% at 5 years ( Fig. 9.10 D). Survival falls to about 10% to 26% by 15 years ( Fig. 9.11 ). To better characterize CAD severity, the group from Duke University introduced the Coronary Artery Disease Prognostic Index, where possible combinations of coronary lesions were ranked and weighted. The index considers both the number of diseased vessels and any significant involvement of the LAD, particularly the proximal LAD, especially if severe. In their original study, Mark and colleagues demonstrated a significant benefit of coronary revascularization for patients with higher CAD index.

Nomograms of specific solutions of multivariable risk factor equations illustrating effect of number of coronary artery systems with important stenoses on time-related freedom from cardiac death in patients randomly assigned to initial medical treatment in Veterans Administration randomized trial of chronic stable angina. For this depiction, patients were censored if they crossed over to coronary artery bypass grafting. Values for each risk factor in the specific solutions of the multivariable equation represented by these nomograms are provided in the American College of Cardiology/American Heart Association Joint Task Force Subcommittee on Coronary Artery Bypass Graft Surgery. (A) Patients with normal left ventricular (LV) function. (B) Patients with importantly impaired LV function. ACC, American College of Cardiology; AHA, American Heart Association; CABG, coronary artery bypass grafting; LM, left main coronary artery disease; SD, systems diseased.

More recently, the modified Duke CAD index was employed to examine the association of all-cause death with CTCA-defined extent and severity of CAD. In the study by Min and colleagues, patients with higher modified Duke CAD index were significantly older and had significantly higher risk factors (male gender, DM, dyslipidemia, arterial hypertension). The Duke prognostic CAD index was a significant predictor of all-cause mortality in patients treated medically ( Fig. 9.12 ). Similar results were found by Reynolds et al looking at a subgroup of 2,415 participants of the ISCHEMIA trial with CTCA interpretable for Duke score; the study found a graded association between the severity of CAD and all-cause mortality, risk of MI (both spontaneous and periprocedural) and the primary composite endpoint of cardiovascular death, MI, or hospitalization for UA, HF, or resuscitated cardiac arrest.

Cumulative survival of patients with severe plaque or moderate/severe left main plaque by coronary computed tomographic angiography. Risk-adjusted P <.001 (controlling for age, family history, and dyslipidemia). LAD, Left anterior descending.

(From Min and colleagues )

Severity of angina is a surrogate for the basic risk factor of severity of reduction in coronary blood flow reserve ( Fig. 9.13 ). Also, graded exercise testing (GXT) is a surrogate for the basic risk factor of severity of reduced coronary blood flow reserve and is related to outcome events in CAD patients. For example, in the heterogeneous group of patients randomly assigned to initial medical treatment in the European Coronary Surgery Study Group randomized trial, 1-, 5-, and 10-year survival was 94%, 83%, and 71% in patients with a mildly positive GXT but 92%, 77%, and 62% in those with a strongly positive GXT. Similarly, in a study from the CASS Registry, survival at 12 years after medical treatment was substantially lower among patients with a strongly positive GXT (55% for men, 62% for women) than among those with a mildly positive test (75% for men, 82% for women). On the other hand, more recent data on patients with stable CAD enrolled in the ISCHEMIA trial showed that increasing ischemia severity was not associated with an increase in the 4-year event rates for all-cause mortality or the 5-component primary endpoint.

Nomograms illustrating effect of severity of angina (expressed as Canadian class) on survival in patients randomly assigned to initial medical treatment in the Veterans Administration randomized trial of chronic stable angina (same equation as in Fig. 9.11 ). (A) Survival, leaving patients in follow-up evaluation after crossover to coronary artery bypass grafting (CABG) (“intent to treat” analysis). (B) Freedom from cardiac death, censoring patients at time of crossover to CABG.

Progression of coronary arteriosclerosis.

Rate of progression of coronary arteriosclerosis, which could also be termed aggressiveness of the arteriosclerotic process , was the subject of a large study by Mohammad and colleagues, who analyzed the Swedish Coronary Angiography and Angioplasty Registry to characterize the natural history of CAD. All coronary artery segments with less than 50% luminal stenosis in patients with a first-time coronary angiogram between 1989 and 2017 were identified (2,661 245 coronary artery segments in 248,736 patients) and followed until a clinically indicated angiography within 15 years was performed or until death or end of follow-up. Progression to obstructive CAD occurred in 26,644 (2.6%) coronary artery segments in 13,853 (12.9%) patients, with an estimated rate of progression of 1.45 per 1000 segment-years. Atherosclerosis progressed more rapidly in the left coronary artery compared with the RCA, in proximal segments compared with distal segments, and in the middle and proximal portions of the LAD compared with other coronary artery tree segments ( Fig. 9.14 ). The specific risk factors with the highest impact on plaque progression were (1) obstructive (single-vessel) CAD on index angiography, (2) diabetes, (3) active smoking, (4) hyperlipidemia, and (5) hypertension. A linear increase in plaque progression rates from 0.6% to 8.1% was observed with increased risk-factor burden from 0 to ≥4 risk factors. The relative importance of multiple risk factors was greater in younger patients; in particular, younger women exhibited a significantly higher risk.

Distribution of progressive coronary lesions from Mohammad and colleagues. The anatomic distribution of coronary artery segments progressing to atherosclerotic obstructive coronary artery disease is shown in (A). The distribution of coronary artery segments that progressed and were revascularized with either percutaneous coronary intervention or coronary artery bypass grafting is shown in (B). The middle portion of the left anterior descending artery (LAD) was the most common site of progression to obstructive disease, whereas proximal LAD was the most common site when lesions were revascularized, reflecting a treatment bias. Cx , Left circumflex artery; LMCA , left main coronary artery; PDA/RPD/LPD , posterior descending artery/right posterior descending artery/left posterior descending artery; RCA , right coronary artery.

(Mohammad MA, Stone GW, Koul S, et al. On the Natural History of Coronary Artery Disease: A Longitudinal Nationwide Serial Angiography Study. J Am Heart Assoc . 2022;11(21):e026396. Figures reproduced according to CC BY-NC-ND 4.0 license: https://creativecommons.org/licenses/by-nc-nd/4.0/ ).

An important advance has been the demonstration that the progression of arteriosclerotic CAD can be slowed and that regression of some lesions in some circumstances can be initiated by intensive lipid-lowering therapy. ^, Plaque regression may occur as a result of a reduction in plaque lipid content, macrophage content, and inflammatory state. Plaque regression can be defined as (i) an increase in luminal diameter measured by coronary angiography (a surrogate measure for reduced plaque size), (ii) a reduction in plaque volume, and (iii) a change in plaque composition (i.e., necrotic core volume and fibrous cap thickness). Treatment strategies for plaque regression include dietary and lifestyle modifications as well as pharmacologic interventions. Avoiding excessive caloric intake and obesity, together with regular exercise, smoking cessation, and reducing alcohol intake, is associated with plaque regression. Aggressive pharmacologic lipid-lowering therapy is the mainstay of plaque interventions; high-intensity statins therapy can induce plaque regression, reducing total atheroma volume (TAV) somewhere in the vicinity of 0%–20%, compared with progression of around 10% among control subjects. ^, Furthermore, the use of ezetimibe in combination with OMT has been associated with plaque regression in multiple IVUS studies. Recently introduced inhibitors of proprotein convertase subtilisin/kexin type 9 (PCSK9) have shown promise in promoting plaque regression by increasing hepatic clearance of serum LDL-C. In the GLAGOV trial, greater reductions in TAV were seen in the evolocumab group compared with OMT alone. Omega-3 fatty acids like eicosapentaenoic acid (EPA) act as precursors for important inflammatory mediators that stabilize endothelial dysfunction. In a study by Welty and colleagues, the use of EPA and docosahexaenoic acid showed a triglyceride and non–high-density lipoprotein cholesterol reduction, which, if associated with strict blood pressure control (SBP < 125 mmHg), lead to coronary plaque regression and reduced cardiac events. Finally, the positive effects on plaque regression of antihypertensive agents (amlodipine and losartan), colchicine, and anticoagulants (NOACs and warfarin) have been the focus of smaller studies.

Coexisting conditions

Older age.

Older age at presentation is a risk factor for death in patients with ischemic heart disease and probably acts as a coexisting condition rather than directly affecting CAD.

Diabetes.

Diabetes is a strong risk factor for death in CAD patients because of its effect as a coexisting condition and its accelerating effect on the arteriosclerotic process. Fig. 9.15 illustrates the powerful effect of diabetes in elderly patients who have undergone PCI. Importantly, the FREEDOM trial and numerous other studies have demonstrated that diabetic patients with triple vessel CAD are best treated with CABG rather than PCI, as freedom from death or the combination of death, MI, and CVA are improved with CABG compared to PCI among diabetic patients.

Nomogram illustrating effect of diabetes on survival after percutaneous coronary intervention (PCI) in elderly patients (same equation as in Fig. 9.11 ).

(Equation from Garrahy and colleagues. )

Hypertension.

The strong effect of hypertension as a risk factor for death in CAD patients is related to kidney damage, intracranial complications, LV hypertrophy, and acceleration of the arteriosclerotic process ( Fig. 9.16 ).

Nomogram illustrating effect of hypertension on time-related probability of freedom from cardiac death in patients randomly assigned to initial medical treatment in the Veterans Administration randomized trial of chronic stable angina (same equation as in Fig. 9.11 ). CABG, Coronary artery bypass grafting.

Gender.

Although overall mortality is lower in women with angina than in men, for patients older than 65, relative risks are similar. ^,

Other comorbidity.

Any serious coexisting disease adversely affects survival in patients with CAD. Of particular importance because of their prevalence in this group of patients are chronic obstructive pulmonary disease , chronic renal disease, and liver disease, in particular, if associated with liver cirrhosis. Ongoing smoking can also be considered an important coexisting condition.

Many patients come to CABG taking β-adrenergic receptor or calcium channel blocking agents, angiotensin-converting enzyme (ACE) inhibitors, angiotensin receptor blockers (ARB), digitalis preparations, antiarrhythmic agents, metformin, and platelet antiaggregating drugs. Recently, the use of sacubitril/valsartan, sodium-glucose cotransporter 2 (SGLT-2) inhibitors, and glucagon-like peptide-1 (GLT-1) agonists has increased significantly in patients affected by CAD. Some patients who present with ACS are referred for CABG while receiving intravenous heparin and/or nitroglycerin.

It is advisable in most circumstances to continue β-adrenergic receptor up to the time of surgery. Several studies have shown a tendency toward development of acute MI in patients in whom β-adrenergic receptor agents are acutely discontinued. Boudoulas and colleagues demonstrated a significant increase in adrenergic tone in most patients the day before the operation that could be reduced by propranolol. Propranolol has also been shown to lessen prevalence of intraoperative ventricular arrhythmias without compromising LV function in low or moderate doses. Patients receiving preoperative β-adrenergic receptor agents, amiodarone, or sotalol are less likely to develop AF postoperatively.

Digitalis preparations, although less frequently part of contemporary medical therapy, can be safely discontinued preoperatively unless AF is present. They can be administered intraoperatively and postoperatively for control of heart rate if AF or other atrial arrhythmias are present, typically when a contraindication to amiodarone or sotalol exists or when these first-line drugs are ineffective.

Drugs that inhibit platelet aggregation, such as abciximab, eptifibatide, tirofiban, clopidogrel, prasugrel, and ticagrelor, bind to the glycoprotein IIb/IIIa platelet receptors. When feasible, these drugs should be discontinued at the appropriate time interval before operation (each has a different half-life) if the patient has a stable coronary syndrome because their use is associated with increased perioperative bleeding and increased transfusion of blood products. Low-dose acetylsalicylic acid (ASA) 100 mg should be continued up to the day of surgery to reduce the risk of perioperative MI (Class IIa recommendation 2017 ESC Guidelines).

Patients who have received plasminogen-activating (fibrinolytic) agents such as streptokinase, alteplase, and reteplase preoperatively require careful intraoperative and postoperative management to mitigate against excessive bleeding (see Chapter 4 , section 3).

Management of perioperative medications by class is summarized as follows:

• Antihypertensive agents:

  • •

    ACE-inhibitors (e.g., ramipril, lisinopril, enalapril): last dose 24 hours before surgery

  • •

    ARB (e.g., losartan, valsartan): last dose 24 hours before surgery

• Antiplatelet agents:

  • •

    Clopidogrel: last dose 5 days before surgery

  • •

    Ticagrelor: last dose 3 to 5 days before surgery

  • •

    Prasuguel: last dose 7 days before surgery ^,

• Anticoagulants:

  • •

    Warfarin: last dose 5 days before surgery

  • •

    Apixiban: last dose 3 days before surgery ^,

  • •

    Rivaroxaban: last dose 3 days before surgery ^,

  • •

    Dabigatran: last dose 3 to 5 days before surgery ^,

• Diabetes control meds

  • •

    Metformin: last dose 2 days before surgery

  • •

    SGLT-2 inhibitors (e.g., canagliflozin, dapagliflozin, and empagliflozin): last dose 3 days before surgery

  • •

    GLP-1 agonist: last dose 14 days before surgery

The prime objective of CABG is to obtain complete revascularization by bypassing all severe stenoses (at least 50% diameter reduction) in all coronary arterial trunks and branches having a diameter of about 1.25 mm or more. ^, Because five or more individual conduits cannot be conveniently used in most patients, at least some of the grafts may require sequential (side-to-side) anastomoses. To increase the likelihood that the entire graft will remain patent, the distal end-to-side anastomosis of a sequential graft should be made, whenever possible, to a relatively large artery with a substantial proximal stenosis and good runoff. Although it is not clearly established whether grafts with more than one distal anastomosis have the same, higher, or lower patency rates than those with only a single distal anastomosis, several studies suggest that sequential grafts are associated with higher mean flows and improved graft patency. As a general principle, conservation of conduit by employing sequential grafting is prudent because of the possibility of subsequent CABG or peripheral arterial procedures that may require use of SV grafts (SVGs).

A widely used strategy involves routine use of the left ITA to the LAD and segments of SV to the remaining coronary arteries requiring revascularization. The right ITA, one or both radial arteries, and the right gastroepiploic artery can also be used in combination with the left ITA. Sequential anastomoses with the ITA and RA can be performed with satisfactory results. Fig. 9.17 and Fig. 9.18 show the most widely used combinations and configurations of bypass grafts. Details of graft placement are often individualized according to location and severity of arteriosclerotic disease, surgeon preference, availability of suitable conduit, and knowledge of the long-term function of various conduits. A detailed discussion on conduits performance and choice will follow later in the chapter.

Combinations and configurations of saphenous vein bypass grafts. (A) Vein graft is anastomosed side to side to a diagonal branch of the left anterior descending coronary artery (LAD) and end to side to LAD. (B) In circumflex system, vein graft is anastomosed side to side to one or more proximal marginal branches and end to side to most distal marginal branch. (C) Sequential grafts to circumflex system (Cx) can be extended to include branches of right coronary artery (RCA). (D) In RCA system, vein graft can be anastomosed side to side to posterior descending coronary artery and end to side to one or more left ventricular branches of RCA. (E) Sequential grafts to RCA system can be extended to include branches of Cx artery. Direction of a sequential graft to RCA and Cx artery systems (configuration C or E ) is chosen so that the largest coronary artery branch is placed at end of sequence.

Combinations and configurations of internal thoracic artery (ITA) bypass grafts. (A) Left ITA is most often used to bypass the left anterior descending coronary artery (LAD). (B) Sequential grafting using left ITA may include a diagonal branch of LAD. (C) Right ITA can be used to bypass right coronary artery (RCA) alone or in combination with an ITA graft to LAD system. (D, Alternatively, right ITA can be passed through transverse sinus and anastomosed to one or more marginal branches of left circumflex coronary artery. (E) Right ITA can be brought across midline and used to bypass LAD, and if indicated, left ITA can be anastomosed to one or more marginal branches of left circumflex coronary artery. (F) When extensive revascularization of posterior wall of left ventricle (LV) is required, a posteriorly positioned sequential vein graft (or radial artery) in combination with a left ITA graft to the LAD is typically used. (G) Radial artery graft may be used as a sequential graft to bypass arteries on lateral and posterior surfaces of LV. Radial artery can be anastomosed proximally to left ITA, which is used to bypass the LAD. Alternatively, radial artery can be anastomosed directly to ascending aorta. (H) Right gastroepiploic artery or splenic artery may be used to bypass branches of right and circumflex coronary arteries in combination with ITA or other grafts to LAD circulation.

The surgeon must decide which coronary artery branches should be bypassed, and pursuant to the primary objective of CABG, namely to achieve complete revascularization, they are usually well advised to err on the side of grafting borderline lesions in significant coronary branches because bypass grafts to borderline coronary artery targets may remain patent and failure of a graft due to native competitive flow is usually a clinically silent event.

A median sternotomy is made through a skin incision that overlies the lower two-thirds of the sternum, and at the same time, the left RA (or a segment of greater SV or other conduit) is harvested, typically by a minimally invasive endoscopic technique. Before the pericardium is opened, the left ITA (and the right unless contraindicated) is/are harvested, preferably by skeletonized technique using an ultrasonic scalpel to avoid thermal injury to the conduit and to the chest wall. Heparin is administered as dissection of the ITA(s) is completed, and the ITA(s) is/are then divided distally. The harvested ITA(s) may be pharmacologically dilated with papaverine, nitroglycerin, or other medications, applied externally, or carefully instilled into the distal vessel. A soft bulldog clamp or clip is placed on the artery at the distal end, and the remaining segment of the artery on the chest wall is ligated or clipped. The pericardium is opened, and pericardial stay sutures are placed. Because arteriosclerosis is frequently present in the ascending aorta and proximal aortic arch in patients with CAD, particularly elderly patients, epiaortic ultrasonographic scanning of the aorta may be used before aortic cannulation to obtain information that can guide safe positioning of cannulae and aortic clamps. Purse-string sutures are placed in the chosen site of the ascending aorta and the right atrial wall for insertion of cannulae for CPB as well as for catheters to permit delivery of cardioplegia into the aortic root and the coronary sinus.

After full heparinization (400 U/kg) aiming to an ACT >480 sec, CPB is established, typically using an angled aortic cannula to direct arterial inflow axially within the ascending aorta and a single dual-stage venous cannula to return venous blood to the perfusion reservoir; normally the patient is cooled to mild hypothermia (34°C bladder temperature). Catheters for administering cardioplegic solution are placed into the ascending aorta and coronary sinus through the previously placed purse-string sutures and secured with tourniquets. The aorta is clamped with a padded cross-clamp, and cardioplegic solution is infused. If cold saline or ice slush is used for topical cooling during administration of cold cardioplegia, a phrenic nerve pad is recommended to protect the left phrenic nerve and to thermally isolate the heart from the abdominal organs (e.g., liver), which are always warmer than the desired myocardial temperature.

With the heart retracted out of the pericardial cavity and toward the head of the patient by an assistant standing to the surgeon’s left aided by deep pericardial sutures, the first anastomosis is made to the distal RCA or to the posterior descending artery (PDA) (see Fig. 9.17 D). (See “ Distal Anastomosis ” later in this chapter.) Sequential anastomoses of the chosen right-sided conduit can be performed to more distal branches of the RCA or to the posterior marginal branches of the LCx coronary artery (see Fig. 9.17 E).

Although the suture pattern varies according to the target vessel, the same uncompromisable general principles are applied to every coronary anastomosis: (i) the 7- or 8-0 needle passes are always from the inside to the outside of the coronary target; (ii) intima to intima opposition is essential with every stitch, and (iii) construction of a tall and symmetric anastomosis with perfect visualization of every single stitch is an inflexible expectation. Strict adherence to these principles is the sine qua non of precise and reproducible coronary bypass surgery.

If the graft is to have inflow from the aorta, it is then distended gently with cardioplegic solution, positioned along the right atrium up to the right side of the ascending aorta, and transected at the point that will permit a smooth course of the conduit back to the aorta without kinking or tension. The proximal end of the graft is then cannulated with an adaptor and reconnected to the cardioplegia inflow via a multiply bifurcated manifold. With each subsequent dose of retrograde blood cardioplegia administered after completion of each graft, cardioplegia is also given through the completed free (vein or RA) grafts. (Of course, if the right graft has inflow from the RITA, it will not be used for administration of cardioplegia and will be left occluded with a soft bulldog clamp until just prior to cross-clamp removal.)

The heart is then retracted to the right by the assistant. To better expose the lateral wall, the apex should grasped with a dry sponge and twisted anticlockwise. The right pericardial suture should be loose and if necessary, the right pleural space may be opened to accommodate the heart. A separate conduit is anastomosed to one or more of the marginal branches of the LCx artery (see Fig. 9.17 B). The graft is properly oriented to avoid twisting, and the heart is repositioned in the pericardial cavity. The graft is distended gently with cardioplegic solution, cut to the appropriate length, and reconnected to the cardioplegia manifold. This graft is usually positioned anterior to the pulmonary artery. A third segment of conduit can be anastomosed to one or more diagonal branches of the LAD and treated similarly.

The ITA is cut to the appropriate length, and a bulldog clamp is placed on the proximal portion. If the operation has been performed using hypothermia, rewarming is begun at this time. The pericardium is incised widely to permit proper alignment of the ITA with the LAD and its diagonal branches, taking care to avoid injury to the left phrenic nerve. A pad is placed beneath the LV, and the LAD is isolated and incised. The distal end of the ITA is spatulated and sutured to the LAD and sequentially to a diagonal branch of the LAD if indicated and anatomically feasible. (see Fig. 9.18 A and B). The angle of takeoff of the diagonal branch from the LAD needs to be acute enough to allow a perfect lie of the ITA graft, avoiding kinking.

Before removing the aortic cross-clamp, two or three openings are made with a punch in the isolated ascending aortic segment. The grafts are sutured to the aorta so that they are free of kinking, tension, and twisting. Cardioplegia may be administered during this time via the retrograde cardioplegia catheter and the remaining free grafts connected to the cardioplegia manifold. There is a large body of evidence implicating the arteriosclerotic ascending aorta as an important source of emboli to the brain and other organs during cardiac surgery. Thus, this strategy, called the single cross-clamp technique, is associated with improved neurologic outcomes compared to construction of the proximal conduit-to-aorta anastomosis using a partially occluding (side-biting) clamp. If all grafts share inflow from ITA(s), there are no proximal anastomoses on the aorta. Cardioplegia is initially administered antegrade and then, in repeated doses, retrograde. The soft occluding clamps on the ITA(s) are removed, and coronary inflow restored prior to cross-clamp removal.

Air is manually evacuated from the ascending aorta through the last, most anterior (untied) proximal anastomosis, and then the aortic cross-clamp is removed. Gentle suction is maintained on the aortic cardioplegia catheter to evacuate any more air that may infrequently enter the aorta from the left cardiac chambers after cross-clamp removal. This risk is further reduced by routine insufflation of CO ₂ gas into the mediastinum throughout the case. Once rewarming has been completed and the heart is beating well, the patient is weaned off CPB, heparin is reversed with protamine, and cannulae are removed. All cannulation sites are secured by oversewing with pledgeted monofilament sutures. Temporary atrial and ventricular pacing wires and chest drainage tubes are placed, and the operation is completed in the standard manner.

Perfection of techniques and equipment for stabilizing the beating heart has resulted in an increased number of CABG procedures performed without use of CPB (OPCAB). Multivessel off-pump coronary revascularization is generally performed through a median sternotomy. With development of stabilization devices that permit adequate and safe exposure of all surfaces of the LV, OPCAB can be safely performed in patients with three-system disease and with left main CAD.

Methodology for OPCAB continues to evolve toward the use of multiple arterial inflows and avoiding aortic manipulation altogether to minimize the risk of perioperative stroke. Many of the past issues relating to preoperative selection and preparation of patients, their anesthetic management, and conduct of the operation are now resolved, and OPCAB can be safely performed in the large majority of patients by appropriately trained and experienced surgeons. Perioperative factors favoring OPCAB must be weighed carefully against relative or absolute contraindications to the procedure. Because OPCAB is not utilized routinely in most centers, the following description provides detailed steps in the operation. The discussions of conduit harvest and preparation also apply to coronary artery bypass grafting performed utilizing CPB.

All patients undergoing OPCAB require invasive monitoring. At a minimum, an arterial line, central venous line, and urinary catheter are required. Pulmonary artery catheters are used routinely in patients with left ventricular or biventricular impairment. Monitoring of pulmonary artery pressures can be particularly useful during retraction of the heart for the construction of distal anastomoses. Transesophageal echocardiogram (TEE) is utilized routinely to identify areas of hypokinesis or significant mitral regurgitation during periods of regional myocardial ischemia and cardiac displacement. Intraoperative TEE is also used to assess the aortic arch and descending thoracic aorta for atherosclerosis and to identify other pathology, especially valvular heart disease.

Close communication between the surgeon and anesthesiologist is essential during OPCAB. The surgeon must inform the anesthesiologist when and how the heart is being displaced, when a coronary artery is occluded, and when a shunt is used. The anesthesiologist must keep the surgeon informed of ischemic changes detected by the ECG, occurrence of arrhythmias, and the patient’s overall hemodynamic status, paying close attention to the pulmonary arterial pressure waveform, which may serve as an early warning of worsening regional or global ischemia. When the heart is displaced, amplitude of the ECG signal may be greatly decreased, and thus, ST-segment changes may underestimate ischemia. In contrast to conventional CABG procedures, maintenance of normothermia is critically important throughout the case, as the ability to actively rewarm the patient by CPB is forfeited. Normothermia can be maintained by routine use of warm intravenous and irrigating fluids, application of a heated mattress or blanket, humidification of the airway, and maintaining a warm ambient temperature in the operating room.

A median sternotomy is made through a skin incision that overlies the lower two-thirds of the sternum, and the linea alba is incised further caudad to allow wide opening of the lower portion of the sternotomy, a feature that facilitates rotating the (full and beating) heart during OPCAB. At the same time, the left RA and/or a segment of the greater SV are harvested using a minimally invasive endoscopic technique. Before the pericardium is opened, the left ITA and right ITA (if indicated) are completely mobilized, preferably in an atraumatic, skeletonized fashion. Heparin is administered as dissection of the last ITA is being completed. For OPCAB, 150 IU/kg is administered to achieve a target activated clotting time (ACT) of >350 seconds. A heparin infusion is subsequently started (rate 6000 IU/hr). ACT is checked every 30 minutes, and heparin infusion is adjusted or supplemented to maintain an ACT >350 seconds.

The ITA(s) is/are then divided. A bulldog clamp is placed on the artery near the open end, and the distal segment of the artery on the chest wall is ligated or clipped. A wide inverted “T”-shaped pericardiotomy is performed, dividing the pericardium along the diaphragm. The phrenic nerves should be identified and protected during pericardiotomy. It is important to divide the pericardium laterally to the level of the LV apex to facilitate cardiac displacement. The left and right pleural spaces are opened widely. Care is taken during the dissection to thoroughly cauterize or clip any large vessels encountered and to avoid the phrenic nerves. It is also important to divide the diaphragmatic muscle slips, which insert on the right side of the xiphoid to allow elevation of the right sternal border, creating space for rightward cardiac displacement. Epiaortic ultrasonographic scanning of the ascending aorta to detect severe arteriosclerosis that may affect the conduct of the operation is performed in every patient.

Hemodynamic stability during manipulation of the heart can be preserved by specific techniques, including elevating the caudad portion of the operating table or placing the patient in the head-down (Trendelenburg) position to increase preload by redistributing blood volume from the legs, rotating the operating table, and opening the right pleural space by incising the pericardium from the diaphragm. The latter two maneuvers minimize the manipulation required to create optimal exposure of the lateral wall of the LV. Another effective first-line treatment of hypotension is the administration of intravenous fluids. An assessment of the patient’s intravascular volume status is made (by direct visual inspection of the right atrium, by measurement of central venous pressure, and by TEE examination) prior to manipulation of the heart, and preload is subsequently optimized. Placing and using temporary atrial (and ventricular in case of complete heart block during proximal RCA occlusion) epicardial pacing wires will prevent bradycardia and improve cardiac output by maintaining an optimal heart rate. Preoperative insertion of an IABP in high-risk patients may increase tolerance of cardiac manipulation.

Exposure of the LV surfaces containing the arteries to be bypassed is achieved by different combinations of elevation (apex of heart toward the ceiling), lateral displacement (apex of the heart to right or left), and rotation (lateral wall rotated anteriorly by use of the suction-based positioning devices). These maneuvers are facilitated by placement of a deep posterior pericardial traction suture approximately two-thirds of the way between the inferior vena cava and left pulmonary vein at the point where the pericardium reflects under the posterior left atrium ( Fig. 9.19 ), the most dependent portion of the pericardial cavity. The suture should not be placed too deeply because this may result in injury to deeper structures of the posterior mediastinum, including the descending thoracic aorta, esophagus, left lung, and pulmonary veins. When this suture is retracted toward the patient’s left hip, it elevates the base of the heart toward the ceiling and points the apex vertically with remarkably little change in hemodynamics. When the deep pericardial traction suture is retracted toward the left shoulder, the heart rotates from left to right. Lesser degrees of elevation and exposure can occasionally be obtained without the deep stitch by placing a warm laparotomy pad underneath the heart. We now routinely use a 2″ × 36″ sterile ribbon gauze wedged between a rubber catheter and the deep pericardial stitch to facilitate cardiac positioning. The two ends of this ribbon gauze can be pulled laterally, cephalad or caudad, to elevate, rotate, and position the heart to expose various coronary artery targets, as first described by Sergeant and colleagues Compression of the heart against the sternal edge or pericardium should be avoided, as this will inevitably result in hemodynamic compromise. The routine maneuver of wedging a rolled towel(s) under the right limb of the sternal retractor will elevate the right sternal edge and help avoid such compression.

Maneuvers to facilitate exposure of surfaces of left ventricle. (A) Heavy suture (size 0 silk or polyester) is placed in posterior pericardium opposite oblique sinus and midway between right and left inferior pulmonary veins. (B) Suture is placed through a wide strip of cloth tape. (C) Snare is placed over both ends of suture and tightened.

The heart may be allowed to roll with gravity into the left or right chest, facilitated by table rotation or tension on traction sutures and gauze ribbons. With proper exposure and retraction, the heart should never be compressed against the sternum or pericardium. Right pericardial traction sutures are always released when exposing the left side of the heart, and similarly, left pericardial traction sutures are released when exposing the RCA.

When grafting of the lateral wall or inferior wall is planned, a suction-based cardiac positioning device (Starfish EVO; Medtronic, Inc.; or ACROBAT and EXPOSE, Maquet, Inc.) is also used. These devices use suction to capture the epicardial surface of the heart and elevate and displace the heart to provide exposure of coronary targets with little hemodynamic compromise. When employed appropriately, they rotate the heart along the axis of the vena cavae while elevating it out of the pericardial well, thereby avoiding compression or kinking of the atria, vena cavae, pulmonary artery, and pulmonary veins. The main drawback of the suction positioning devices is the potential creation of an epicardial hematoma that, especially in frail elderly patients, can evolve into a bleeding hematoma with the consequent risk of epicardial rupture and bleeding. This complication can be reliably avoided by utilizing the deep pericardial traction suture(s) and gauze sling to accomplish most of the cardiac displacement so that the suction-based cardiac positioning device is not obliged to exert excessive force on the epicardium.

Next, the artery to be grafted must be stabilized. The current generation of coronary stabilization devices also rely on suction rather than compression to maintain epicardial tissue capture ( Fig. 9.20 ). This characteristic allows the device to achieve coronary stabilization at the mechanical median of the cardiac cycle rather than compressing the cardiac chambers excessively. Thus, stabilization is maintained while mechanical interference with ventricular function is minimized. Once the device is applied, a few seconds of adjustment may be needed to ensure hemodynamic stability. If hemodynamics are compromised, the degree of compression should be reduced.

Stabilization devices to facilitate exposure of segment of artery to be grafted. (A) Generic device that depresses myocardium on both sides of artery. (B) Generic device that elevates myocardium on both sides of artery, using suction.

In OPCAB surgery, choosing the optimal sequence of grafting to maintain hemodynamic stability and avoid critical ischemia is one of the most important factors in the success of the operation. As a general rule, collateralized vessels are grafted first and then reperfused by releasing flow through the ITA or performing the proximal anastomosis. The last coronary target grafted is the collateralizing vessel(s). This strategy avoids interrupting vital flow from the collateralizing vessel to the collateralized territory until after the collateralized vessel has been grafted. The use of an intracoronary shunt during grafting of a collateralizing vessel is a routine safeguard.

At times, the proximal anastomoses may be performed early in the operative sequence to aid in early reperfusion of a collateralized vessel. If the LITA to LAD artery graft must be performed first, it is necessary to leave a long mammary graft to avoid tension on the LITA-LAD anastomosis during subsequent displacement of the heart to expose other target vessels. This is especially true in the case of severe LMD when the LAD territory needs to be revascularized first. The common practice of performing the left ITA to LAD anastomosis first is based on the principle of restoring flow to the anterior wall and septum of the LV before substantial manipulation of the heart is performed to expose the LCx arterial branches. This approach may be valid in many patients, but it may be problematic if the LAD provides substantial collateral flow to the remainder of the LV or if the LITA graft is short. In the former situation, an intracoronary shunt is essential; in either situation, first grafting a lateral wall coronary target(s) may be helpful.

The anterior wall vessels (LAD and diagonals) may be exposed with very little manipulation of the heart, and an apical positioner device is often not necessary ( Fig. 9.21 A). The deep traction stitch and/or ribbon gauze is secured to the drapes on the patient’s left side, and the coronary stabilizer is usually brought over the anterior wall from the caudal rack portion of the sternal retractor.

Exposure of posterior and anterior descending coronary arteries. (A) Posterior descending coronary artery. Heart is elevated with minimal lateral displacement by upward traction toward the head on ends of tape and downward traction on the snare. Stabilizing device is then applied. (B) Anterior descending coronary artery. Ends of tape are pulled upward and slightly to patient’s left, and downward traction in opposite direction is exerted on snare. Stabilization device is then applied.

By comparison, the lateral wall vessels are more difficult to expose and are best approached by rolling the apex of the heart toward/under the right sternal border. The traction sutures on the right pericardium are released, and the right pleural cavity can be opened widely ( Fig. 9.21 B). The left-sided traction sutures are pulled up taut on the sternal retractor, and the table is rotated sharply to the right to assist in rolling the heart toward/under the right sternal border, which should be elevated on rolled towels wedged under the right limb of the sternal retractor. The deep stitch and/or ribbon gauze is pulled toward the patient’s left and secured to the sternal retractor. The coronary stabilizer is mounted on the right side of the sternal retractor adjacent to the cardiac positioning device, and its arm reaches across the heart, aiding in both the presentation and stabilization of the obtuse marginal coronary arteries. The cardiac positioner may be applied to the left lateral wall of the heart rather than the apex, as needed.

Exposure of the posterior descending branch of the RCA is achieved by marked elevation of the apex of the heart with minimal lateral displacement. The deep traction stitch or ribbon gauze is pulled toward the patient’s left hip or directly caudad and clamped to the drapes. The coronary stabilizer is attached to the left limb of the sternal retractor opposite the cardiac positioning device, which is attached to the right limb of the sternal retractor. The patient is placed in Trendelenburg position with the bed tilted to the patient’s right. The base of the heart is elevated, and the apex is oriented vertically toward the ceiling. The cardiac positioner may be applied to the apex of the heart and helps to elevate and elongate the heart. Care should be taken to avoid folding or kinking the right ventricular outflow tract by excessive cephalad tilting of the heart; this will grossly impair hemodynamic stability.

Exposure of the distal RCA is obtained by traction on the right pericardial traction sutures, release of the left pericardial traction sutures/gauze sling, rotation of the table to the left, use of Trendelenburg position, and application of the suction-based cardiac positioning device to the RV free wall or acute margin of the heart.

After exposing the appropriate wall of the LV, the artery to be grafted is encircled at least 1 cm proximal to the chosen site of the arteriotomy with an elastic vessel loop. Encircling the artery distal to the arteriotomy site is not recommended. After the stabilization device is applied, the artery is incised, and anastomosis to the appropriate conduit is performed using techniques described later in this chapter. Performing the anastomosis can be facilitated by use of a misted blowing device that sprays humidified carbon dioxide over the anastomotic site to disperse blood and optimize visibility. This device should be directed at the anastomotic site only during the actual placement of sutures to minimize injury to the endothelium of the coronary artery and conduit. An intracoronary shunt may also facilitate performing the anastomosis by maintaining a bloodless field and preventing inadvertent suturing of the back wall of the coronary artery. Gentle insertion of the shunt is important to avoid injury to the endothelium ( Fig. 9.22 ). When the anastomosis is performed on a distal RCA that is not critically narrowed or occluded proximally, a shunt may be particularly valuable because bradycardia and hypotension may occur when the RCA is temporarily occluded. A shunt may also provide sufficient distal flow to an arterial segment that may be an important source of collateral blood flow to other segments of the LV that have not been bypassed, thus preventing hypotension and cardiac decompensation.

Intracoronary shunt can be used to facilitate performing distal anastomosis.

Proximal anastomoses of radial arteries or vein grafts to the ascending aorta are performed after identifying a segment of disease-free ascending aorta by epiaortic ultrasound (EUS). Aortic proximal anastomoses may be performed before or after the distal anastomoses. The systolic blood pressure is lowered to <95 mmHg, and a clampless proximal anastomosis system (Heart String, Getinge, Wayne, NJ, USA) is deployed. The vein grafts are occluded with a soft bulldog clamp until they are de-aired with a 25-gauge needle. Arterial grafts are not punctured but are allowed to back-bleed prior to removal of the proximal seal system. Proximal anastomoses are not performed in the presence of diffuse grade III or any grade IV or V atherosclerotic involvement of the ascending aorta. The partial occluding clamp, aka “side-biting clamp,” although still part of the typical surgical tray, is not applied on the ascending aorta except in the most unusual emergency circumstances.

After completion of every graft, Transit Time Flow Measurement with MiraQ Cardiac (Medistim, Norway) is used as a routine quality control. Acceptable minimum mean graft flow and maximal pulsatility index (PI) depend on many factors, including mean arterial pressure, conduit size, run-off quality, configuration of composite grafts, quality of the anastomosis, degree of upstream stenosis and presence of native or graft competitive flow. Although a detailed analysis of all those variables is beyond the scope of this chapter, a mean graft flow ≥20 ml/min with a PI ≤3 is generally acceptable. Lower MGF and higher PI (especially if in combination) warrant a functional test for competitive flow (i.e., snaring the proximal native coronary) and further interrogation of the conduit and distal anastomosis with the EUS imaging probe and may ultimately mandate revision of the graft or anastomosis. After completing all grafts and documenting the patency of each graft, protamine is given to fully reverse the heparin, returning the ACT to the preoperative baseline. All grafts are retested with the Medistim flow probe prior to chest closure.

Performing off-pump coronary revascularization provides the unique opportunity to significantly lower the risk of intraoperative stroke. For this reason, a preferred routine approach is an “aortic no-touch” (or “anaortic”) technique, with dual inflow from both ITAs and liberal use of composite extension grafts (I-grafts) and T/Y-grafts, thus obviating any aortic manipulation. A common configuration is a RITA-radial I-graft for the inferior wall, a LITA graft to the anterior wall, and a LITA-radial T-graft to the lateral wall ( Fig. 9.23 ). Alternative arrangements to provide total arterial revascularization (TAR) are numerous and include RITA-radial extension through the transverse sinus sequentially to the lateral and inferior walls and LITA to the anterior wall ( Fig. 9.24 A), LITA to the anterior wall and LITA-radial T-graft to the inferior and lateral wall ( Fig. 9.24 B), in situ RITA to the anterior wall and LITA to the lateral wall ( Fig. 9.24 C), and finally RITA to the lateral wall via the transverse sinus (may require RA I-graft extension), LITA to the LAD and RA T-graft off the LITA to the diagonal artery ( Fig. 9.24 D).

Use of a radial artery graft in conjunction with bilateral internal internal thoracic artery conduits. The radial artery may be anastomosed into side to the internal thoracic artery to increase conduit length, or it may be anastomosed end-to-side from the left internal thoracic artery (LITA) to use as a conduit to the circumflex system. Composite arterial grafts are especially useful when aortic atherosclerosis makes aortic manipulation hazardous as illustrated here. RITA , Right internal mammary (thoracic) artery; RA , radial artery; PDA , posterior descending coronary artery; Cx marginal, circumflex marginal coronary artery; LAD , left anterior descending coronary artery.

Alternative configurations of arterial grafts to provide total arterial revascularization. (A) RITA-radial extension through the transverse sinus sequentially to the lateral and inferior walls and LITA to the anterior wall; (B) LITA to the anterior wall and LITA-radial T-graft to the inferior and lateral wall; (C) in situ RITA to the anterior wall and LITA to the lateral wall; (D) RITA to the lateral wall via the transverse sinus (may require radial artery extension), LITA to the LAD and radial artery T-graft to the diagonal artery. RIMA , Right internal mammary artery; PDA , posterior descending coronary artery; Cx marginal, circumflex marginal coronary artery; PDA , posterior descending coronary artery; LITA , left internal thoracic artery; RITA right internal thoracic artery; CxM1 , first obtuse marginal branch of the circumflex coronary artery; LAD , left anterior descending coronary artery.

After preoperative examination of both legs with the patient standing, supplemented by ultrasonic imaging when indicated, the right or left greater SV may be chosen for harvesting. In the rare instance that a patient is not a candidate for bilateral ITA harvest or RA use and a suitable segment(s) of vein cannot be found in the legs by any method, use of alternative conduits becomes necessary. These include the right gastroepiploic artery, inferior epigastric artery (IEA), and, in rare circumstances, the splenic or ulnar artery. The cephalic vein can be taken from wrist to shoulder, but its walls are usually thinner than those of leg veins, and its late patency is worse.

For removal of the greater SV, the leg is abducted, and the knee is flexed about 30 degrees and supported ( Fig. 9.25 A). The greater SV is most commonly harvested through a series of vertical incicions ( Fig. 9.25 B) or endoscopically. Using small transverse or vertical incisions, a lighted dissector is introduced into the wounds ( Fig. 9.26 A). A plane of dissection anterior to the vein is established with a balloon-tipped dilator or other device ( Fig. 9.26 B), and the dissector is used to isolate the vein and its branches ( Fig. 9.26 C). The branches are clipped and divided with a cautery, and after ligating its proximal and distal ends, the vein is removed. The time required to remove veins with this technique is somewhat longer than with the conventional method, but a significantly lower incidence of leg wound complications and leg discomfort has been observed.

Removing greater saphenous vein. (A) Location of greater saphenous vein and line of incision. (B) Multiple small incisions over saphenous vein.

Endoscopic removal of greater saphenous vein. (A) Insertion of lighted dissector into incision over saphenous vein. (B) Plane of dissection immediately above saphenous vein is established with a balloon-tipped dilator. (C) Isolation of saphenous vein and branches with lighted dissector. Inset, Endoscopic view.

If the traditional open technique is employed, preoperative ultrasound mapping of the vein is of value to precisely direct the incision and avoid the creation of flaps. If the vein from the lower leg is to be used, the initial skin incision is made just anterior to the medial malleolus. If the upper portion of the vein is used, the initial skin incision is made in the groin. The desired plane is accessed by blunt dissection with scissors down to the level of the vein. Skin and subcutaneous fat are undermined with the scissors, staying just superficial to the SV and spreading the tips of the scissors over the vein. A continuous incision or multiple small incisions over the length of the vein may be used ( Fig. 9.25 B). Creation of flaps is avoided, and care is taken to preserve the saphenous nerve.

Whenever possible, a single long segment (usually 50–65 cm) of the greater SV is removed. About 12 to 15 cm may be needed for diagonal branches of the LAD, about 20 to 24 cm for marginal Cx branches, and about 18 to 22 cm for the RCA and its branches. When the usable vein has been exposed and its length measured, the proximal (femoral) end is isolated and divided between ligatures. Branches may be ligated with fine sutures and divided or divided between hemostatic clips ( Fig. 9.27 A). When clips are used to occlude branches, they should be applied parallel to the longitudinal axis of the SV rather than transversely to avoid narrowing or kinking the vein graft. ( Fig. 9.27 A [upper inset]). After division of all branches and removal from its bed, the vein is divided between ligatures at its peripheral end ( proximal end of the graft) and removed. Evidence supports a “no-touch” SVG harvest technique to improve long-term patency rates. With this technique, a pedicle of fat tissue around the vein is left intact, thus decreasing the risk of direct vein damage and spasm.

Division of side branches of saphenous vein using fine ligatures or hemostatic clips. (A) Venous branches should be secured just flush with saphenous vein to avoid narrowing (upper inset) or creating diverticula (lower inset) , which can be a nidus for thrombus formation. (B) Avulsed branches are secured with a double-loop 7-0 polypropylene suture.

The lesser saphenous vein can be removed with the patient in the prone or supine position with the leg either straightened (prone position) and elevated or flexed at the knee and rotated medially (supine position). If the prone position is used, the patient is subsequently repositioned and redraped in the supine position. An initial skin incision is made posterior to the lateral malleolus and extended superiorly toward the popliteal fossa. The vein is divided at the level where it penetrates the deep fascia to join the popliteal vein. The sural nerve lies parallel to the vein and is preserved. Branches of the vein are secured and divided as described for the greater SV.

Leg incisions are closed with continuous absorbable sutures in the subcutaneous layers. A small drainage catheter may be placed beneath these layers in the thigh, and the skin is closed with a continuous subcuticular suture or with metal clips. This can be done immediately after the vein is removed or deferred until after CPB has been discontinued and protamine administered.

The vein is removed to a preparation table, and a small adaptor is inserted into the open peripheral end and secured with a ligature. The clamp previously placed on the central end of the graft is removed, and the vein is flushed with a room-temperature, heparinized, balanced salt solution (500 mL) to which a small amount of heparinized blood (30 mL) has been added. Vasodilating drugs (nitroprusside, papaverine) can be added to mitigate vasospasm. Branches that have been avulsed are secured with fine polypropylene sutures (see Fig. 9.27 B). The vein is placed in the heparinized solution at room temperature until used.

It is generally agreed that details of removing and storing SVs until their insertion are important in minimizing damage, which may include intimal disruption, deposition of platelets and leukocytes on the intimal surface, damage to smooth muscle, and disruption of the extracellular matrix. Despite this general agreement, opinions differ as to which maneuvers or interventions result in the least injury. Overdistention of the vein ^, and venous spasm are surely disadvantageous. No consensus exists as to the optimal temperature to maintain the graft or the optimal solution to flush and distend the graft before its insertion.

The epicardium is incised over the area of the coronary artery that has been selected for anastomosis using a fine scalpel blade with a rounded end. The anterior surface of the artery can be cleared by gentle brushing with the scalpel blade. Even when crystalloid cardioplegic solution has been infused, careful inspection of the artery usually reveals a thin central line that is red or translucent, indicating the location of the lumen. The target vessel is opened with a specialized coronary knife (22.5° Straight BVI Beaver Optimum™ Blade, Visitec International, MA), and the arteriotomy is extended with fine coronary scissors (Scanlan Inc., Minneapolis, MN). It is crucial to open the coronary artery exactly in the midline of the anterior wall and to stay in the anterior midline while extending the coronary arteriotomy ( Fig. 9.32 ). The incision in the coronary artery is of variable length, depending upon local conditions of the artery, presence of coronary calcification, and size of conduit. In general, the coronary arteriotomy is at least twice as long as the diameter of the vessel, erring on the side of a longer and more capacious, rather than a shorter or more restrictive, anastomosis. The anastomotic site is kept free of blood by dispersing retrograde bleeding from the distal end of the arteriotomy with a humidified CO ₂ blower (Medtronic Inc., Minneapolis, MN). It is important that the surgeon’s assistant blow on the target only when the surgeon is placing the needle through the tissue of the conduit or target vessel so as to minimize potential trauma to the intima of the vessels that can occur with excessive use of the blower. Excellent visualization is critical for a precise anastomosis. Optical magnification with 3.5× loupes, a headlight, and specialized fine coronary needle drivers and forceps (Scanlan Inc., Minneapolis) are used for all anastomoses. Fine 7-0 or 8-0 polypropylene suture with an 8-mm 3/8 circle taper needle (Deklene Max, Teleflex Medical, Morrisville, NC) is used for all distal anastomoses to optimize precision unless severe calcifications or thickened conduit mandates use of a heavier needle (Prolene Everpoint cardiovascular needle, Ethicon Inc., Bridgewater, NJ). Although the suture pattern varies according to the target vessel, the same uncompromising principles apply to every coronary anastomosis: (i) the needle passes are always from the inside to the outside of the coronary target; (ii) intima to intima opposition is essential with every stitch; and (iii) construction of a tall and symmetric anastomosis with perfect visualization of every single stitch. These principles are the sine-qua-non of precise and reproducible surgical coronary revascularization.

Distal anastomosis in coronary artery bypass grafting. (A) Anterior wall of coronary artery is opened with a scalpel. (B) and (C) Incision is enlarged to the appropriate length with angled scissors.

The distal end of the conduit (arterial or venous) is prepared for the most distal anastomosis by beveling it so that the circumference of the opening is slightly larger than the opening of the artery. The incision is made about 10%–20% longer than that in the artery, and sutures in the conduit are placed slightly farther apart than those in the coronary to create the desired “tall” anastomosis. Especially if the conduit is small, a larger vertical incision provides a larger hood to suture over the distal anastomosis, mitigating the threat of size mismatch between the conduit and coronary arteries.

The technique of anastomosis uses one double-armed 8-0 or 7-0 polypropylene suture placed as a continuous stitch ( Fig. 9.33 A). Stitches in the coronary artery are always placed from intima to adventitia (inside to outside), optimizing suture placement and avoiding the potential danger of pushing intimal calcium away from the coronary media/adventitia, which can occur when a needle is passed outside to inside, resulting in coronary artery dissection. Each stitch pierces the intima near the vessel edge but often emerges through periarterial tissue 0.5 to 1.0 mm away from the edge. Stitches in the conduit are passed from outside to inside. The stitches are generally placed separately through the coronary artery and the coronary conduit unless it is convenient to place them with one pass of the needle holder. Even then, the conduit and coronary artery should be held apart so that the needle can be visualized after it has pierced the conduit and before it pierces the coronary intima (no “blind” stitches are taken) ( Fig. 9.33 A). This disciplined maneuver ensures that every stitch is placed accurately in the intima layer of the coronary artery and of the conduit and that no extraneous tissue has been incorporated.

Technique of anastomosis for left coronary system grafts. (A) Suture is passed through artery and continued to heel of anastomosis. Inset, Direction of suture placement. (B) Vein is approximated to artery. (C) Suture line is continued around toe of anastomosis and ends of suture tied.

The suturing pattern of the anastomosis varies depending on which target vessel is being grafted. For anastomoses to the left coronary system , two classic patterns are employed, one for more horizontal targets (LAD and diagonals) and another for more steeply vertical targets (ramus intermedius and marginal branches of the LCx). In both patterns, the suture is passed first through the conduit (outside to inside) two stitches from the heel. For LAD/diagonal anastomosis, suture placement begins on the surgeon’s right of the heel, and for lateral wall targets, suture placement starts from the surgeon’s left of the heel. The suture is then passed through the coronary artery (inside to outside) and continued to the heel of the anastomosis with the third pass of the suture. The conduit is suspended by the first assistant approximately 1.0 to 2.0 cm above the coronary artery target so that loops of suture are short. After five sutures, the conduit is approximated to the coronary artery by gentle in-line tension on both ends of the suture ( Fig. 9.33 B). The suturing is then continued around the lateral side and the toe of the anastomosis, and the two ends of the suture are tied ( Fig. 9.33 C).

For anastomoses to the right coronary system , the suture is passed first through the conduit (outside to inside), two sutures away from the heel on the surgeon’s right-hand side of the heel ( Fig. 9.34 A). It is then passed through the artery (inside to outside) and continued to the heel of the anastomosis and around for another two stitches. The conduit is approximated to the coronary artery, and the suture line is completed around the lateral side and the toe of the anastomosis and remaining near wall of the artery ( Fig. 9.34 B). The two ends of the suture are then tied ( Fig. 9.34 C). The suture techniques described earlier describe patterns for a right-handed surgeon performed from the patient’s right side of the operating table; left-handed surgeons may alter some aspects of these patterns.

Technique of anastomosis for right coronary system grafts. (A) Suture line begins near midpoint of arterial wall and is continued to toe of anastomosis. Inset, Direction of suture placement. (B) Vein is approximated to artery, and suture line is continued around heel of anastomosis. (C) Suture ends are tied.

For sequential anastomoses in which the conduit will lie perpendicular to the arterial branches, the conduit can be opened perpendicular to its long axis for the side-to-side anastomosis if it is of sufficient diameter ( Fig. 9.35 A). The incision should not exceed one-third of the circumference of the conduit. The anastomosis is constructed exactly as for an end graft, with sutures passing outside to inside on the conduit and inside to outside on the coronary artery. The first stitch is passed outside to inside through the conduit, two stitches from the heel of the incision made in the conduit, and then passed inside to outside through the coronary artery two stitches away from the heel of the coronary arteriotomy. Care is taken to ensure optimal orientation of the conduit ( Fig. 9.35 B). If the vein conduit is small or if a RA or ITA is used, the conduit is incised parallel to its long axis Fig. 9.35 C), and a diamond anastomosis is created . The suture is passed outside to inside through the conduit two stitches to the right end of the midpoint of the longitudinal incision ( Fig. 9.35 D) and then passed inside to outside through the coronary artery two stitches to the right of the heel. The suture line is continued leftward across the heel of the anastomosis and completed as described earlier. In short, the suture patterns described are effective for all types of anastomoses; orientation of the conduit to the coronary artery is determined by the careful placement of the first stitch for each anastomosis.

Technique of sequential grafting. (A) Vein graft is opened perpendicular to its long axis. (B) Anastomosis is begun at midpoint of arteriotomy on its right side. (C) Alternatively, vein is incised parallel to its long axis. (D) In alternative approach (diamond anastomosis with longitudinal incision in graft), suture is passed through graft close to right end of incision and continued across heel of anastomosis. (E) End-to-side anastomosis. Inset, Direction of suture placement for side-to-side and end-to-side anastomoses. (F) and (G) Side-to-side anastomosis. Technique is identical to that in Fig. 9.25 , G and H . A, Coronary artery; G, graft.

If the conduit and artery are parallel rather than perpendicular, a side-to-side anastomosis is performed with the same suture pattern as for end-to-side anastomoses, shown in Fig. 9.35 F and Fig. 9.35 G. After each sequential anastomosis is completed, the conduit is gently distended by injection of cardioplegia solution or a balanced salt solution (if the operation is performed on an arrested heart) or by releasing soft bulldog clamps and filling the new bypass graft with inflow (if the operation is performed off-pump) to assess optimal graft geometry and avoid excessive length or tension between sequential anastomoses.

Endarterectomy is most often performed on the LAD and less frequently on the distal RCA and LCx. If the coronary arteries are diffusely diseased or occluded, endarterectomy, usually combined with coronary bypass grafting, may be necessary to achieve revascularization. CEA can usually be avoided by utilizing other available options, which include creating multiple jump grafts to different segments of a diffusely diseased coronary artery and thoroughly grafting adjacent coronary targets of better quality, which may provide collateral flow. Nonetheless, it is occasionally necessary and should be performed when indicated because successful CEA leads to satisfactory long-term survival and good relief of angina. CEA is most appropriate when a large LAD is chronically and very severely diseased or occluded over most or all of its length and no suitable bypass site exists. In this scenario, the patient may suffer relentless angina from the large region of myocardial ischemia.

An on-pump, arrested heart approach is employed, and an “open” endarterectomy technique is preferred over the older “closed traction” technique. The LAD is incised in the anterior midline with a fine scalpel blade down through epicardium and adventitia to the media layer over the entire length of the diseased segment; this will typically include the middle third and proximal part of the distal third of the LAD with a total length of 5 to 10 cm. The distal extent of the anterior midline LAD arteriotomy should not be so distal as to extend beyond the reach of the in situ LITA conduit that is used to reconstruct the LAD after endarterectomy. The “core” of thickened media and intima is identified and gently dissected circumferentially with a fine curved CEA spatula (Scanlan Inc., Minneapolis, MN). The endarterectomy spatula may be used to bluntly dissect the core beyond the distal LAD incision to facilitate a combination of closed traction distally to remove apical obstruction and open endarterectomy of the mid and proximal LAD ( Fig. 9.36 ).

Endarterectomy of right coronary artery (RCA). (A) RCA is incised just proximal to origin of posterior descending coronary artery (PDA) . (B) Endarterectomy plane is developed with a dissector. (C) Atheromatous core is divided over a clamp. (D) Core is freed distally from arterial wall. (E) Core is teased from posterior descending coronary artery and distal RCA. (F) Core is removed from proximal RCA. (G) Vein graft is sewn to edges of artery (see Fig.9.34 ).

Care must be taken to preserve the adventitia of the LAD, including its diagonal and septal branches. Proximally, the occlusive core typically breaks free cleanly with gentle traction. The remaining LAD adventitia is carefully cleaned of debris. The skeletonized in situ LITA is generously spatulated and used to reconstruct the anterior two-thirds of the LAD using two fine polypropylene sutures. These sutures are placed normally on the ITA and deeply through the adventitia of the LAD with the intent of excluding all of the LAD that does not contain the ostia of the diagonals and septal perforating branches of the LAD. This technique results in the maximum removal of obstruction and the minimal residual adventitia exposed to blood flow. It is anticipated that ECs from the LITA may repopulate the LAD adventitia over time, contributing to longer-term graft patency.

When an attempt at routine coronary bypass encounters a branch of the RCA or LCx that cannot be bypassed due to severe diffuse atherosclerosis, a similar endarterectomy can be performed in a limited localized fashion to facilitate the performance of a coronary bypass graft to that region of the heart. A somewhat longer-than-usual anastomosis may be combined with local closed traction endarterectomy to accomplish a satisfactory solution, applying the same principles presented earlier.

When CABG is performed with CPB, the proximal anastomoses on the aorta are typically constructed after all distal anastomoses are completed. This allows the surgeon to optimize graft lengths while the heart is orthotopic in the pericardial well. Although a partially occluding clamp on the ascending aorta for construction of proximal anastomoses is occasionally used, performing both distal and proximal anastomosis under a single padded cross-clamp yields better neurologic outcomes and is considered the preferred routine technique. ^,

With the aid of a punch (in the case of an on-pump operation) or a proximal anastomotic device (in the case of an off-pump operation), one or more aortotomy is made in the ascending aorta. Conduits are positioned so that they will be free of kinking or tension when the heart is filled with blood. Each conduit is cut obliquely and incised to create a circumference that is 20% to 30% larger than that of the circular opening in the aorta, resulting in a “tall” anastomosis. A double-armed 5-0 or 6-0 polypropylene suture line is begun two stitches to the right of the heel of the conduit, passing the suture from outside to inside on the conduit and then inside to outside on the aorta ( Fig. 9.37 A). This suture line is continued leftward across the heel of the anastomosis. The conduit, which is initially suspended by the first assistant 1.0–2.0 cm above the aorta, is now approximated to the aorta by exerting gentle in-line tension on both ends of the suture. The suture line is continued across the toe of the anastomosis ( Fig. 9.37 B), and the ends of the suture are tied on the right side of the anastomosis ( Fig. 9.37 C).

Proximal anastomosis in coronary artery bypass grafting. (A) After incising conduit on its undersurface and creating an opening in aorta, suture line is passed from outside to inside on conduit and then inside to outside on aorta. (B) Conduit is approximated to aorta. (C) Suture line is continued across toe of anastomosis and completed.

Patients referred for CABG who have preoperative AF should have the AF addressed surgically during coronary revascularization because this does not increase perioperative risk and has been shown to prolong life and reduce complications over longer-term follow-up. If patient’s comorbidities do not preclude on-pump CABG, then a comprehensive left and right atrial Cox-Maze IV lesion set with exclusion of the left atrial appendage (LAA) should be added to the planned coronary bypass procedure. If atherosclerosis of the ascending aorta precludes aortic cannulation for CPB, bilateral pulmonary vein isolation may be achieved without CPB using bipolar radiofrequency clamps and combined with clip occlusion of the LAA (Atricure Inc., Cincinnati, OH, USA). Finally, in the patient for whom the Maze procedure is likely to be unsuccessful or for whom an extended operation poses serious risk, amputation or occlusion of the LAA should be considered the minimum appropriate intervention for any CABG patient with AF.

If the patient is hemodynamically unstable at the time of operation, an IABP is inserted if one is not already in place. Appropriate monitoring devices are inserted, the sternum is quickly opened, and CPB expeditiously established. The left ITA can be mobilized after CPB is established, with the heart decompressed but still beating while maintaining suitable coronary and cerebral perfusion pressures. RA and SV conduits are harvested as needed. Optimal myocardial resuscitation and preservation are essential under these circumstances. Use of warm antegrade cardioplegic induction, followed by cold antegrade cardioplegia and repeated doses of cold retrograde cardioplegia, and concluded with warm retrograde “hot shot” resuscitation combined with controlled aortic reperfusion may be advantageous.

The choice of conduits during emergency coronary bypass surgery in the setting of cardiogenic shock should be carefully considered. Although it is almost always appropriate to harvest the left ITA (even after initiation of CPB if necessary), the use of multiple arterial conduits in this setting is less clearly indicated and depends on local expertise and individual patient factors, including reversibility of shock, likelihood of high-dose postoperative alpha-adrenergic pressor agents, body habitus, and patient age. The remainder of the operation is completed as described under “Surgical Strategy” earlier in this chapter. If the IABP is insufficient to permit safe discontinuation of CPB, use of extracorporeal membrane oxygenation (ECMO), Impella (Abiomed, Inc., Danvers, MA), or a temporary ventricular assist device may be necessary. The recently introduced Impella 5.5 (Abiomed, Inc., Danvers, MA) device may be the optimal support for the stunned left ventricle that cannot wean from CPB and may be inserted either centrally via the ascending aorta or via the right axillary artery where it may be tunneled and later removed without reopening the sternum.

Low-dose oral β-adrenergic receptor blocking agents may be given beginning 4 to 6 hours postoperatively if heart rate allows and inotropes have been weaned off. This regimen is continued until discharge as prophylaxis against supraventricular tachyarrhythmias. ^, In most patients, beta-blockers are up-titrated before discharge and continued long-term. Prophylaxis is especially important in elderly patients, who are more susceptible to postoperative AF. ^, Biatrial pacing may reduce the prevalence of postoperative AF. A protocol for prevention of postoperative AF includes continuing beta-blocking agents to the morning of surgery. In patients with a CHA2DS2-VASC score ≥2, 150 mg of amiodarone is given intraoperatively (if not bradycardic), followed by 200 mg orally twice a day for a total of 15 doses. Perioperative administration of amiodarone to prevent postoperative AF is supported by the 2021 AHA/ACC guidelines. Intraoperatively, a posterior pericardiotomy may be useful. Careful attention is paid to maintaining a K ≥4.4 mmol/L and Mg ≥2.2 mmol/L. Anticoagulation is commenced only after 48 hours of persistent AF with the choice between intravenous unfractionated heparin, warfarin, and a NOAC, depending on the clinical scenario and surgeon preference.

Heparin (5000–7500 units every 8–12 hours) can be administered subcutaneously as prophylaxis against deep venous thrombosis and pulmonary embolism in the first 48 to 72 hours postoperatively. This may be especially important in obese patients and those who have had off-pump CABG.

Aspirin (81 mg/day) should be administered almost immediately after operation and continued indefinitely postoperatively. Efficacy of this regimen has been well established, particularly when anastomoses have been made to smaller coronary arteries ( Table 9.5 ). Ticlopidine, clopidogrel, or prasugrel can be used in patients who are allergic to or intolerant of aspirin. The role of DAPT to prevent graft occlusion has been the focus of multiple trials and systematic reviews. In the DACAB trial that included 500 patients with 1460 SVGs, the SVG patency rate 1 year after CABG was 88.7% in patients treated with ticagrelor + aspirin, 82.8% in patients treated with ticagrelor alone, and 76.5% in patients treated with aspirin alone. The difference between ticagrelor + aspirin and aspirin alone was statistically significant, whereas that between ticagrelor alone and aspirin alone was not statistically significant. A systematic review and meta-analysis by Deo and colleagues showed that DAPT reduces the risk of early SVG occlusion at the expense of a higher risk of major bleeding in the early postoperative period. Greater benefit was seen in patients undergoing OPCAB, probably due to the temporary state of relative hypercoagulation that follows OPCAB.

A meta-analysis by Nocerino and colleagues showed that during a treatment duration of up to 1 year, single antiplatelet therapy led to an absolute risk of venous graft occlusion that was 4.2% higher compared to DAPT, a statistically significant difference. In contrast, arterial grafts did not benefit from this intensified therapeutic approach. The network meta-analysis by Solo and colleagues found that dual antiplatelet therapy with either aspirin plus ticagrelor or aspirin plus clopidogrel is more efficacious than aspirin monotherapy in preventing SVG failure after CABG. Furthermore, the issue of clopidogrel resistance and optimal combination with aspirin has been studied by Kim and colleagues who found that aspirin in combination with ticagrelor significantly reduced the rate of MACE without an increase in the rate of overall major bleeding after OPCAB in patients with clopidogrel resistance.

Oral vasodilators play an important role in promoting RA patency. The RADIAL study showed that after controlling for known confounding, calcium channel blocker (CCB) therapy was associated with a significantly lower risk of MACE (multivariate Cox hazard ratio [HR] 0.52, 95% CI 0.31–0.89, P <.02) and RA graft occlusion (multivariate Cox HR 0.20, 95% CI 0.08–0.49, P <.001). The optimal duration of CCB therapy is not certain, but 6–12 months seems to be beneficial. As discussed later, control of cardiovascular risk factors is important to minimize risk of atherosclerosis in vein grafts.

A well-developed body of knowledge exists concerning risk factors for arteriosclerosis. The consensus is that this knowledge should be focused on patients who have established CAD and undergo CABG. Smoking cessation must be emphasized to the patient before operation. An appropriate body weight should be maintained, even if special dieting is required. Numerous large trials have demonstrated that semaglutide, an antidiabetic medication that mimics glucagon-like peptide-1 (GLP-1), is effective in achieving impressive weight loss in obese patients. This weight loss has been associated with marked reduction in adverse cardiovascular events. The role of such new medications for secondary prevention in patients with CAD who have undergone CABG remains to be explored. Hypertension and saturated fats in the diet must be controlled, and serum lipids should be kept within guidelines-directed levels through dietary measures and administration of statin medications if required. ^{, , ,}

The benefit of long-term beta-blocker therapy for secondary prevention following revascularization in patients with previous MI or impaired left ventricular function is well established ; on the other hand, the routine use of beta-blockers in patients with stable CAD and normal LV is not beneficial to reduce the risk of cardiovascular events after complete revascularization. In patients after CABG, beta-blockers are recommended and should be started as soon as possible to reduce the incidence or clinical sequelae of postoperative AF.

In nearly all patients, cardiac stress testing following revascularization should be driven by the presence of symptoms rather than routinely performed. The goal of cardiac stress testing after revascularization is not only to evaluate for restenosis but also to determine the functional status and patient symptoms. Routine stress testing of asymptomatic patients post-PCI or post-CABG is rarely recommended. ^, Noninvasive cardiac stress testing plays a valuable role in symptomatic patients following CABG to assess the potential ischemic etiology of chest pain due to incomplete revascularization, the development of thrombosis or stenosis in bypass grafts, and the progression of native coronary heart disease.

Risk of early (hospital) death has been extensively studied. Risk stratification models have been created to provide accurate prediction of operative risk for groups of patients undergoing CABG. Large databases have been established in single institutions and multicenter studies, and analyses of these data have established the predictive power of certain preoperative variables, the most important of those are:

• •

  Older age ( Fig. 9.38 ) ^,

  

  Operative mortality for coronary artery bypass grafting in various age cohorts in New York State Cardiac Surgery Reporting System for 1991–1992 ( n = 30,972).

  (From Eagle and colleagues. )

• •

  Female gender

• •

  Previous CABG

• •

  Urgency of operation

• •

  Increasing LV dysfunction

• •

  Left main disease

• •

  Increasing extent of CAD

• •

  Renal dysfunction

The relative mortality risks of these variables are shown in Table 9.6 for six of the data sets. Older age has consistently predicted operative risk after CABG ^, (see Fig. 9.38 ). Female gender is also independently associated with increased operative risk. Previous CABG and emergency operations are associated with substantial increases in relative risk (2.0 or greater). Other variables associated with increased early mortality are recent (<30 days) STEMI and several common comorbidities (e.g., diabetes, metabolic syndrome, end-stage renal disease, valvar heart disease, pulmonary hypertension, chronic obstructive pulmonary disease, severe peripheral arterial disease, anemia, AF). ^, Importantly, failure to use an ITA graft is also associated with increased early (and late) mortality. An important risk factor for mortality that is not considered in the STS score or EuroScore II is the presence of liver disease. Patients with advanced liver failure with a Child-Pugh score of B and a MELD score above 12 are at increased risk of mortality. ^,

Patient outcomes following OPCAB have been studied in several prospective randomized trials. A consistent finding among most of these studies has been that OPCAB in experienced hands is a safe technique for coronary revascularization; however, the magnitude of the overall clinical benefit to the patient from avoiding CPB remains a point of contention.

In one single-center randomized study of 200 patients, there were no differences in early and late patency rates in any coronary distribution for patients undergoing OPCAB or conventional CABG. In contrast, the multicenter ROOBY trial of 2203 veterans showed decreased graft patency rates for patients who underwent OPCAB at 1-year follow-up and decreased survival in the OPCAB cohort at 5-year follow-up. Individual operator variability may explain the differences between these studies.

A large meta-analysis of 41 RCTs up to 2006 by Sedrakyan and colleagues suggested that OPCAB was associated with a significant reduction in the rate of perioperative stroke, AF, and wound infection. The CORONARY trial of 4752 patients from 19 countries showed no difference at 5 years follow-up in the composite outcome of death, stroke, MI, renal failure, and repeat revascularization. The GOPCABE trial studied on- or off-pump CABG outcomes in elderly patients (≥75 years old). At 5 years follow-up, the study showed no difference in the composite outcome of death, MI, and repeat revascularization. Incomplete revascularization was associated with lower survival rates, irrespective of the type of surgery.

The question of absolute mortality or stroke benefit with OPCAB versus conventional on-pump surgery is difficult to address with an adequately powered RCT. Indeed, a study sample size of 19,506 would be required to provide a power of 80% to detect a 30% relative risk reduction in the rate of in-hospital mortality or stroke (∼2% event rate).

A single-center risk-adjusted analysis of approximately 15,000 primary isolated CABG patients found that patients with higher predicted risk of mortality had a significant 30-day mortality benefit with OPCAB ( Fig. 9.39 ). In another large single-center review of over 10,000 isolated CABG cases, propensity-matched analysis showed that OPCAB was associated with improved long-term survival at 10 and 15 years. Similarly, a risk-adjusted comparison of outcomes after on-pump coronary artery bypass (ONCAB) versus off-pump coronary artery bypass (OPCAB) within more than 876,081 primary isolated CABG patients in the STS National Cardiac Database reported that OPCAB was associated with reduced 30-day risk of death, stroke, acute renal failure, morbidity or mortality and prolonged length of stay. Patients with the highest predicted risk of mortality had the highest apparent benefit from OPCAB.

Regression curve comparison of observed mortality rates for off-pump coronary artery bypass grafting (OPCAB) and coronary artery bypass grafting (CABG) on cardiopulmonary bypass (CPB) across all levels of predicted risk. STS , The Society of Thoracic Surgeons.

In contrast, a propensity-matched analysis of the New Jersey Department of Health mandatory registry of isolated CABG performed by surgeons experienced with on-pump and off-pump techniques revealed increased mortality and increased need for repeated revascularization at 10 years in the OPCAB group. Further, in a large national registry study reported by Kim and colleagues, patients undergoing off-pump CABG had a significantly higher risk of late death (HR 1.43, 95% CI 1.19–1.71, P <.0001) compared with those undergoing on-pump CABG ( Fig. 9.40 ), and in subgroup analyses, on-pump CABG conferred survival benefits in most demographic, clinical, and anatomic subgroups compared with off-pump CABG.

Long-term survival of 5203 patients from Korea who underwent elective isolated coronary artery bypass grafting (off-pump, n = 2333; on-pump, n = 2870) from 1989 through 2012. As seen in panel (A) during the first postoperative year, there is no significant survival difference between the on- and off-pump groups. However, the difference becomes significant later in favor of on-pump surgery compared with off-pump surgery (B).

(Kim JB, Yun SC, Lim JW, et al. Long-term survival following coronary artery bypass grafting: off-pump versus on-pump strategies. J Am Coll Cardiol . 2014;63[21]:2280-2288.)

Unplanned conversion from off-pump to un-pump is a well-known risk factor for mortality (10.5% for OPCAB converted vs. 1.6% for OPCAB not-converted in a report from Kirmani and colleagues ). Data from the European Coronary Artery Bypass Grafting (E-CABG) registry show the important role of surgeon experience and center volume in improving early outcomes of OPCAB. In a recent analysis of the registry, centers performing more than 50% of coronary revascularization off-pump with a minimum of 20 OPCAB per surgeon per year had lower procedure time, lower rate of conversion to on-pump, lower rate of early postprocedural PCIs, and lower 30-day mortality compared to lower volume practices.

Similar results were obtained from a recent analysis of the National Inpatient Sample. For high-volume off-pump coronary artery bypass centers (≥164 cases per year) and surgeons (≥48 cases per year), off-pump coronary artery bypass reduced mortality compared with on-pump coronary artery bypass in cases requiring a single graft (odds ratio [OR] 0.66, 95% CI 0.49–0.89 and OR 0.33; 95% CI 0.22–0.47, respectively) or two or more grafts (OR 0.82, 95% CI 0.66–0.99 and OR 0.63, 95% CI 0.49–0.81, respectively).

The largest meta-analysis of over 100 RCTs comparing OPCAB with conventional CABG showed no statistically significant difference in 30-day and 1-year mortality and MI and found a lower rate of 30-day stroke, renal and respiratory failure, postoperative AF, RBC transfusion, and wound infection with OPCAB. On the other hand, patients undergoing OPCAB had fewer grafts performed, a higher incidence of graft occlusion at 30 days and 1 year, higher incidence of coronary reinterventions, and perhaps lower 5-year survival compared to conventional CABG, highlighting once again the importance of surgeon and center experience and volume in achieving complete and durable off-pump coronary revascularization.

Finally, similar results were reported by Comancini and colleagues in a recent systematic review and meta-analysis of 22 studies; there was no significant difference in long-term all-cause mortality between OPCAB and ONCAB (HR 1.000, 95% CI 0.92–1.08, P =.95). Meta-regression identified older age as a significant factor favoring OPCAB.

In general, after isolated CABG, approximately 98% of heterogeneous groups of patients survive at least 1 month, and 97%, 92%, 81%, and 66% survive 1, 5, 10, and 15 years or more, respectively ( Fig. 9.41 A). The hazard function for death has an early and rapidly declining phase that merges with a constant hazard phrase at about 6 months, and this gives way to a gradually rising phase of hazard at about 1 year ( Fig. 9.41 B). This gradually rising phase probably results from closure of grafts, progression of native arterial disease, and noncardiac comorbidities. This phase of hazard is favorably affected by use of ITA grafts and multiple arterial grafts (MAG). ^,

Survival after coronary artery bypass grafting (CABG). (A) Time-related survival in a heterogeneous group of patients. Circles, Individual deaths, positioned along horizontal axis at time of death and along vertical axis according to Kaplan-Meier estimator; vertical bars, 70% confidence limits; (parentheses), number of patients still being traced; solid line (not visible in some areas because of density of circles), nomogram of separately determined parametric survival; dashed lines enclose 70% confidence limits; dot-dash-dot line, survival of age-gender-ethnicity–matched general population. (B) Hazard function for death in same group of patients. Note (1) early, rapidly declining phase; (2) constant phase, which elevates entire hazard function above baseline; and (3) slowly rising third phase.

(Modified from Sergeant and colleagues. )

The SYNTAX Extended Survival (SYNTAXES) study was an investigator-driven extension of follow-up of the SYNTAX trial. Overall mortality at 10 years after CABG was 24% (21% in patients with three-vessel disease without left main disease and 28% in patients with left main disease), with 95.7% of patients receiving an arterial graft to the LAD, 27.6% receiving bilateral ITAs, and 18.9% undergoing TAR. In a post-hoc analysis of the SYTAXES, Thuijs and colleagues found that at the longest follow-up of 12.6 years, all-cause death occurred in 23.6% of patients receiving MAG compared to 40% receiving single arterial graft (SAG, P =.038).

The FAME 3 trial randomized patients with three-vessel disease to FFR-guided PCI or angiographically-guided CABG ^, ; the trial reported an all-cause mortality at 1 and 3 years after CABG of 0.9% and 3.4%, respectively (mortality for cardiac causes 0.5% and 1.2%, respectively). In this trial, 97% of patients received a LITA graft, and 24.5% received MAG. Importantly, the primary endpoint of MACE favored CABG over PCI at 1 and 3 years follow-up.

In contrast with the widely held belief that long-term survival benefit after CABG is dependent primarily on the use of the LITA to revascularize the LAD territory, multiple recent studies support the concept that the use of MAG or TAR further enhances long-term survival. ^, A recent systematic review and meta-analysis by Gaudino and colleagues showed that the use of the RA as a second arterial graft is associated with better survival at 10 years follow-up compared to the use of ITA plus SVGs alone (HR 0.73; 95% CI 0.57–0.93; P =.01). A meta-analysis of 9 observational studies by Yi and colleagues showed that the use of bilateral ITAs is associated with a significant reduction in mortality at 10 years (HR 0.79; 95% CI 0.75–0.84). Similarly, a propensity-matched study by Puskas and colleagues showed that BITA grafting conferred a 35% reduction (95% CI 12%–52%, P =.006) in the long-term risk of mortality. Furthermore, Goldstone and colleagues, using a State-maintained clinical registry including all 126 nonfederal hospitals in California, showed that a second arterial conduit is associated with significantly lower mortality (10.6% vs. 13.1% at 7 years; HR 0.79, 95% CI 0.72–0.87), and lower risk of MI (HR 0.78, 95% CI 0.70–0.87) and repeat revascularization (HR 0.82, 95% CI 0.76–0.88) compared to LITA plus SVGs.

Likewise, an analysis of the mandatory cardiac surgery registry of the state of New Jersey by Chikwe and colleagues demonstrated that, after adjusting for baseline characteristics, multiarterial CABG is associated with lower 10-year mortality compared to single arterial CABG in 3,588 propensity-matched pairs (HR 0.80; CI 0.72–0.90; P =.006). Ren and associates, using data from the Australian and New Zealand Cardiac Surgery Database, compared mortality among patients receiving a SAG, MAG, or TAR. At a median follow-up of 5 years, the mortality rate was significantly lower in MAG versus SAG patients (HR 0.79, 95% CI 0.76–0.83, P <.01) and TAR versus MAG with ≥1 supplemental vein graft (HR 0.85, 95% CI 0.80–0.91, P <.01).

Finally, a comprehensive risk-adjusted analysis of more than one million primary CABG patients in the STS National Cardiac Database demonstrated that MAG was associated with superior long-term survival compared to SAG among virtually all patient subgroups, including stable coronary disease, ACS, and acute infarction. Survival with MAG was equivalent to that with SAG for patients aged >80 years and those with severe CHF, renal failure, peripheral vascular disease, or obesity. Only patients with BMI ≥40 kg/m ² had superior survival with SAG. The authors conclude that MAG is associated with superior survival and should be the surgical multivessel revascularization strategy of choice for patients with a BMI <40 kg/m ² ( Fig. 9.42 ).

Multi arterial grafting is associated with a 12-year survival benefit in most demographic, comorbidity, coronary artery disease, and surgical subgroups, and the benefit appears greater in young patients. Importantly, the survival advantage of multiarterial coronary artery bypass was comparable in urgent and elective procedures but was less pronounced in off-pump compared with on-pump coronary artery bypass surgery. CHF , Congestive heart failure; CVD , cardiovascular disease; Dis , disease; DM , diabetes mellitus; EF , ejection fraction; eGFR , estimated glomerular filtration rate; IPW , inverse probability weighting; MI , myocardial infarction; NYHA , New York Heart Association; OW , overweight; PCI , percutaneous coronary intervention; PVD , peripheral vascular disease; UW , underweight.

(From Sabik et al. )

Most deaths early and late after CABG are from cardiac failure, which could be termed cardiac death ( Table 9.7 ). A smaller proportion of deaths (about 15%) are sudden, considerably less than in the natural history of patients with CAD. Although it is difficult to be certain, about 25% of all deaths early and late after CABG are unrelated to ischemic heart disease or the operation. As mentioned earlier, the recent FAME 3 trial showed cardiac mortality to account for approximately half of deaths at one year and for about one-third of deaths at 3 years. ^,

Many patient-specific risk factors for death in patients with arteriosclerotic heart disease pertain to patients who have undergone CABG (see Box 9.1 ). In addition, procedural and institutional incremental risk factors affect outcomes ( Table 9.8 ).

The Califf-Harrell nomogram and its conceptual bases are fundamental to understanding risk factors and outcomes after interventional therapy ( Fig. 9.43 ). Only patient risk factors categorized as “reversible ischemia” (see Box 9.1 ) can be neutralized by CABG (or PCI), and when these are neutralized, clinical outcomes are determined in large part by other risk factors. “Vessel stenoses neutralized” (see Fig. 9.43 ) depicts potential for improvement (“comparative benefit”) by interventional therapy because this possibility is determined by number of systems (vessels) stenosed. Actual improvement, or “comparative benefit,” is determined by the actual number of systems with important stenoses that are, in fact, neutralized by interventional therapy.

Nomogram of an equation describing comparative benefit of coronary artery bypass grafting (CABG) or percutaneous coronary intervention (PCI) over medical treatment. Comparative benefit is difference in percent freedom from cardiovascular deaths after two treatments at some time . Patient’s position on horizontal axis conceptually is determined by sum of all patient risk factors, such that the higher the percent freedom from cardiovascular death, the fewer the risk factors in the particular patient. Comparative benefit, along vertical axis, is determined by interrelationship between patient’s position along horizontal axis and number of coronary arteries with stenoses completely neutralized by interventional therapy (isobars) . Values of isobars are in fact hazard ratios that represent effect on comparative benefit of number of arteries with important stenosis in patients undergoing CABG, assuming all stenoses are neutralized by interventional therapy. Nomogram pertains when time (t) is 2 to 7 years (time of approximately proportional hazards) after CABG (or PCI) and when early risks of intervention (CABG or PCI) are negligible. Abstract equation is as follows: Comparativebenefit = S CABG ( t ) − S M ( t ) S M ( t ) h = SCABG ( t ) h, Hazard ratio; LM, left main coronary artery; S CABG, survival after CABG; 5 D, vessels diseased (stenosis >50%); S _M , survival with medical treatment; SD, standard deviation; t, time.

(Modified from Califf and colleagues, and from Califf RM, Harrell FE Jr: personal communication; 1990.)

The greater the number and severity of patient-specific risk factors, the greater the comparative survival benefit of interventional therapy. At the same time, 5- or 10-year survival is reduced when patient risk factors are numerous. Conversely, the favorable effect of interventional therapy is difficult to demonstrate, even though present on conceptual grounds, in most patients with good LV function and few unfavorable risk factors.

A classic parsimoniously derived multivariable risk factor equation for death after CABG has been derived by Blackstone, working with Sergeant and colleagues ^{, ,} ( Table 9.9 ). Strength and shape (time of maximal effect) of risk factors are best determined by nomograms that also have the advantage of presenting risk-adjusted depictions. Number of systems with important stenoses is a weak risk factor for death after CABG in the first 5 years ( Fig. 9.44 A) when revascularization is complete and the ITA is used for the LAD graft. Also, the number of diseased vessels and even left main coronary artery stenosis, as well as the number of distal anastomoses, are not risk factors for hospital death when an adequate operation has been performed. This reality that outcomes after CABG are not impacted by preoperative complexity of CAD is a critically important difference between CABG and PCI, where outcomes after PCI are powerfully impacted by complexity of CAD at baseline.

Nomograms of specific solutions of KU Leuven multivariable risk factor equation for death after coronary artery bypass grafting (CABG), illustrating strength and shape of certain risk factors for death. In all figures, dashed lines represent 70% confidence limits. (A) Risk-adjusted effect on survival after CABG of number of stenosed systems. (B) Risk-adjusted effect of moderate and severe left ventricular dysfunction. (C) Risk-adjusted effect of left ventricular ejection fraction (EF) , continuously represented along horizontal axis, on survival to 5, 10, and 20 years after operation. (D) Risk-adjusted effect of pre-CABG unstable angina, both after “cooling” and when refractory to treatment, compared with moderately severe but stable angina.

Only gold members can continue reading. Log In or Register to continue

Share this:

• X
• Facebook

Related

Related posts:

1. Cardiac reoperations: Considerations and techniques
2. Acute and chronic complications of myocardial infarction
3. Cardiac rhythm disturbances
4. Generating knowledge from information, data, and analyses

Stay updated, free articles. Join our Telegram channel

Join