2013 ESC Guidelines on Management of Stable CAD
2013 ESC Guidelines on Management of Stable CAD
Advances in techniques, equipment, stents and adjuvant therapy have established PCI as a routine and safe procedure in patients with SCAD and suitable coronary anatomy. The mortality risk associated with the procedure in SCAD is <0.5%. The efficacy of PCI in SCAD in comparison to medical therapy and CABG has been the subject of extensive evaluation.
8.1.1 Type of Stent and Dual Antiplatelet Therapy: Bare metal stents (BMS) are associated with a 20–30% rate of recurrence of angiographic stenosis within 6–9 months after implantation. Drug-eluting stents (DES) reduce the incidence of angiographic restenosis and ischaemia-driven repeat revascularization. For the first generation of DES, this benefit has been extensively demonstrated in spite of a slightly higher incidence of late and very late stent thrombosis, related to delayed endothelialization, which requires longer dual antiplatelet therapy (DAPT) to prevent stent thrombosis. First-generation sirolimus-eluting stents (SES) and paclitaxel-eluting stents (PES) have been extensively compared in head-to-head randomized controlled trials. Angiographic results were better with SES and translated into significant differences in terms of repeat revascularization. The most recent or 'second- generation' DES (with thinner struts and biodegradable or more biocompatible polymers) showed superior clinical outcomes for both efficacy and safety when compared with first-generation DES. Second-generation DES—preferring those tested in large all-comers trials and compared with other DES with proven outcome—are therefore the recommended option in SCAD patients with no contra-indication to DAPT (see Table 30).
A recent meta-analysis confirmed that clopidogrel pretreatment in stable patients undergoing elective PCI does not reduce mortality or major adverse cardiac events (MACE), as compared with clopidogrel administration in the catheterization laboratory. On the basis of several randomized trials and this meta-analysis—and contrary to a diffuse common practice—SCAD patients who undergo diagnostic coronary angiography with the possibility of undergoing ad-hoc PCI (revascularization within the same procedure) should not be treated with clopidogrel before the coronary anatomy is known. The bleeding risk of routine DAPT, administered before catheterization in patients who do not require stenting (no significant CAD or CAD requiring CABG surgery), is not balanced by a detectable benefit in terms of ischaemic events in those undergoing PCI. Despite the overwhelming advantages shown in ACS patients—and especially diabetic patients—in the absence of randomized clinical trials, the use of prasugrel or ticagrelor cannot be recommended in SCAD patients undergoing elective PCI. An off-label use of these drugs is, however, common practice in some high-risk patients, especially in cases of documented stent thrombosis. After stenting, premature discontinuation of antiplatelet therapy is a major risk factor for stent thrombosis and should be avoided.
Current guidelines recommended 6–12 months of DAPT after first-generation stents. New-generation DES have been associated with lower rates of stent thrombosis, and recent data from registries and randomized controlled trials suggested that a shorter duration of DAPT might be sufficient in stable coronary patients. Considering the risk–benefit ratio of DAPT beyond 6 months—and while waiting for more information from the ongoing studies exploring various durations of treatment including more than a year—we endorsed the current ESC recommendation of 6–12 months of DAPT in SCAD patients undergoing PCI revascularization with a latest-generation DES (see section 9.5 for more details and for recommendations). Shorter durations (1–3 months) are reasonable in patients with high bleeding risk or undergoing undeferrable surgery or on concomitant anticoagulant treatment for which the use of clopidogrel only has shown significant advantages in a single small-scale trial (What is the Optimal antiplatElet and anticoagulant therapy in patients with oral anticoagulation and coronary StenTing (WOEST)).
8.1.2 Intracoronary Assessment of Stenosis Severity (Fractional Flow Reserve, Intravascular Ultrasound and Optical Coherence Tomography) (See Web Addenda). When non-invasive stress imaging is contra-indicated, non-diagnostic, or unavailable, the measurement of FFR during adenosine infusion is particularly helpful in identify haemodynamically or functionally significant stenosis, inducing ischaemia, justifying revascularization (see Table 31). In patients with FFR >0.80, studies in the BMS era have demonstrated that medical treatment provides better outcomes than immediate revascularization. Accordingly, a patient with a stenosis and an FFR >0.80 (two measurements or during adenosine infusion) should not be revascularized. The recent Fractional Flow Reserve vs. Angiography for Multivessel Evaluation (FAME-2) study confirmed that SCAD patients with stenoses having FFR ≤0.80 gain a benefit from PCI revascularization in addition to OMT, a benefit driven only by the reduced need for urgent revascularization (the study was stopped prematurely by the DSMB for that reason). Patients without ischaemia have excellent outcomes on medical therapy alone. Although the study suffers from significant limitations (the open nature of the trial may have affected the decision on 'urgent' revascularization; low-risk population), FFR can guide PCI in a clinically-effective way.
Fractional flow reserve, although in general not useful in very high grade lesions (angiographically >90%), which practically always have an FFR ≤0.8, may help the decision on when to revascularize in many uncertain clinical conditions. One such condition is 'multivessel disease', which occurs in a very heterogeneous population. In these patients, FFR measurement may change the strategy of revascularization (PCI vs. CABG) and the extent of revascularization according to the functional assessment of stenoses in critical coronary locations. Another critical situation for revascularization is LM stenosis, a lesion site that is of major prognostic importance and often determines the type of treatment. A recent single-centre study showed that angiography is not always reliable in the determination of the severity of LM stenoses and that deferred revascularization, if FFR >0.80, may be a safe approach.
Another situation relevant to these Guidelines, where FFR may be useful, is in 'post-ACS' patients. Once the culprit lesion has been treated, the patient can be considered as a stable or stabilized CAD patient. Non-invasive stress testing/imaging immediately after the acute phase may be impossible, contra-indicated or hazardous. Non-culprit stenoses in patients with recent ACS can be evaluated by FFR, either during the index procedure or in a staged procedure.
The use of intravascular ultrasound (IVUS) has been broadly investigated in SCAD with many different subsets of lesions (see Table 31). Unlike FFR, IVUS is an imaging diagnostic tool and does not provide assessment of the functional severity of a stenosis. Previously accepted cut-off limits of 3.5 or 4.0 mm for major epicardial artery stenosis and 6.0 mm for left main stenosis have been shown to be unreliable and poorly correlated with FFR, with somewhat better results when the absolute IVUS measurements are corrected for the reference vessel size. Once the indication to treatment is established, when more information is needed, IVUS is far superior to FFR because it provides an anatomical characterization of the lesion in terms of vessel size and plaque composition and can control stent expansion and strut apposition. More recently, optical coherence tomography (OCT) has been developed as a new intracoronary imaging tool with superior resolution (<10 μm) able to offer detailed assessment of superficial components including measurements of the thickness of the fibrous cap of lipidic plaques. The usefulness of OCT in SCAD patients with vulnerable plaques has not been well established, and certainly treatment of non-functionally critical lesions based only on the presence of elements of instability is not recommended. The facilitated technique of image acquisition allows optimization of stent expansion and apposition, and long-term assessment of stent healing.
8.2.1 Arterial vs. Venous Grafts. For the last 25 years the principle technique underpinning CABG has been the use of an internal mammary artery (IMA) to the LAD coronary artery with supplemental vein grafts as required. This followed a seminal publication from the Cleveland Clinic in 1986, showing that an IMA to the LAD improved survival and reduced the subsequent incidence of MI, recurrent angina and the need for repeat revascularization.
Since then, several angiographic studies have confirmed the superior patency of both IMA grafts in comparison to vein grafts both early and late after CABG. Most importantly, this superior graft patency appears to translate into a survival benefit. In 2001 a systematic review comparing single (SIMA) and bilateral (BIMA) IMA grafting reported a significant survival benefit with BIMA grafts with a hazard ratio for death of 0.81. Recent studies have reported that a survival benefit of BIMA grafts extends to the second and third decades of follow-up, and especially in patients with diabetes.
Previous concerns that the use of BIMA grafting may increase early postoperative mortality and/or morbidity have been dispelled by the Arterial Revascularization Trial (ART) which, in one of the largest trials ever conducted in cardiac surgery, randomized 3102 patients to SIMA or BIMA, with supplemental grafts as necessary. Whilst the primary outcome of this trial is 10-year survival, an interim analysis of safety at 1 year showed similar mortality of around 2% in both groups, with no difference in the incidence of MI, death or stroke but a slight increase in the incidence of sternal wound reconstruction in the BIMA group (1.9 vs. 0.6%). The data are currently being analysed to determine the key patient and operating factors that predispose to sternal dehiscence.
The radial artery has also been proposed as a second arterial graft, rather than a second IMA graft. In two randomized trials, the radial artery patency at 1 year was variously reported to be 'superior' and 'equivalent' to that of vein grafts. In an additional small, randomized trial the 5-year patency of the radial artery was significantly superior to that of vein grafts when placed to the circumflex coronary system.
Nevertheless, despite angiographic and clinical evidence of the potential superiority of arterial grafts, the reality is that the vast majority of bypass grafts—with the exception of the IMA to the LAD—are performed with saphenous vein grafts. Best current evidence suggests that the patency rate of saphenous vein grafts is slightly lower in off-pump surgery and when harvested using endoscopic, rather than open techniques.
8.2.2 On-pump vs. Off-pump Surgery (See Web Addenda). Off-pump surgery was initially proposed almost three decades ago. Numerous randomized trials and meta-analyses have shown no significant beneficial effect on mortality, but there have been reductions in stroke, transfusion, re-operation for peri-operative bleeding and postoperative complications, possibly at the cost of an excess of repeat revascularization with off-pump CABG. The two largest randomized trials, the Veterans Affairs (VA) Randomized On/Off Bypass (ROOBY) (n = 2203) and The CABG Off or On Pump Revascularization Study (CORONARY) (n = 4752) both reported no difference in the primary composite endpoint at 30 days. ROOBY reported a poorer outcome (death or complication) in the off-pump composite endpoint at one year (9.9 vs. 7.4%) while CORONARY has still to report at the time of writing. In contrast to the randomized trials, several large propensity-matched registries, which generally include higher-risk patients, have reported a reduction in mortality in patients undergoing off-pump CABG, although off-pump surgery is still performed in a minority of centres.
8.3.1 General Rules for Revascularization (See Web Addenda). The decision to revascularize a patient should be based on the presence of significant obstructive coronary artery stenosis, the amount of related ischaemia and the expected benefit to prognosis and/or symptoms (Figure 5). There are many clinical, anatomical, technical and environmental factors that may be discussed before the benefit of revascularization can be anticipated (Table 32, Figure 5). The vast number of possible combinations makes absolute recommendations difficult to mandate in every situation. In this regard, for a given patient in a given hospital, clinical judgement with consensual rather than individual decision-making, with a Heart Team discussion, should prevail, although this has to be individualized since, in many patients, the preferred approach is often quite clear-cut.
(Enlarge Image)
Figure 5.
Global strategy of intervention in stable coronary artery disease (SCAD) patients with demonstrated ischaemia. CABG = coronary artery bypass graft; CAD= coronary artery disease; LAD = left anterior descending; LV = left ventricular; OMT = optimal medical treatment; PCI = percutaneous coronary intervention.
Indication of revascularization for prognosis or symptoms (see Table 32).
Not suitable for revascularization due to anatomy or clinical conditions.
See section 9.
When technically feasible, with an acceptable level of risk and a good life expectancy, revascularization is indicated in chronic angina refractory to OMT. It can also be considered as first-line treatment in the situations discussed below.
8.3.1.1 Post-myocardial Infarction: The Swiss Interventional Study on Silent Ischemia Type II (SWISSI II) trial, involving 201 patients with a recent ST-segment elevation MI or non-ST segment elevation MI, investigated whether revascularization with PCI was better than drug therapy in stable patients with silent myocardial ischaemia (see description below).
During a lengthy 10-year follow-up period, the primary endpoint, which was survival free of cardiac death, non-fatal MI or revascularization, was significantly better in the PCI group.
PCI also significantly reduced the rates of cardiac death and all-cause mortality or MI. In addition, objective evidence of ischaemia was reduced in the revascularization group.
The Danish trial in Acute Myocardial Infarction (DANAMI) compared a deferred invasive strategy of PCI or CABG with a conservative strategy in 503 patients with inducible myocardial ischaemia, who had received thrombolysis for a first MI. Stress testing was performed at discharge and in patients randomized to the invasive strategy; angiography was performed within two weeks of stress testing. Patients with unstable angina were excluded.
Angina plus ischaemia was present in 25%, angina alone in 16% and 57% had silent ischaemia on stress testing. At 2.5 years follow-up, the invasive strategy was associated with a reduction in the incidence of re-infarction and less frequent angina. This was noted in patients with both symptomatic and asymptomatic ischaemia.
In contrast, the Occluded Artery Trial (OAT), following a strategy of routine PCI 3 to 28 days after acute MI, found no discernible benefit in terms of death, re-infarction, or heart failure at 4-year follow-up among asymptomatic or minimally symptomatic patients with occlusion of the infarct-related artery. The findings of the OAT study, however, should not be interpreted as applying to all patients experiencing ST-elevation MI, but just to those with a late occluded artery and no- or minimal angina. Two smaller studies [The Open Artery Trial (TOAT) and Desobstruction Coronaire en Post-Infarctus (DECOPI)] dealt with similar situations of stable patients after a Q-wave MI, without residual ischaemia and a persistent total occlusion of the infarct-related artery, and these studies did not show any clinical benefit of stenting over medical therapy.
Post-lytic revascularization studies. Old trials (not discussed here) comparing an invasive- with a conservative approach after fibrinolytic therapy did not show any differences in patient outcomes, but these studies antedated the use of stenting and modern antiplatelet therapies. In contrast, more recent randomized studies, comparing systematic early PCI with a conservative ischaemia guided strategy, have demonstrated favourable trends with early PCI and a significant reduction of death or MI in a meta-analysis.
8.3.1.2 Left Ventricular Dysfunction: In general, revascularization improves survival in 'sicker' patients, especially in the presence of LV dysfunction. From the early days of coronary angiography, it has been well recognized that LV dysfunction is one of the most powerful indicators of an adverse prognosis. As techniques of revascularization improved, LV dysfunction has become a prime target—as opposed to a contra-indication—for coronary revascularization. Several older studies, including a meta-analysis, suggested that survival was improved by CABG over medical therapy in patients with mild-to-moderate systolic dysfunction. The CASS randomized trial of bypass surgery vs. medical therapy demonstrated no overall differences in survival, except in the subset of patients with an ejection fraction (EF) of 0.35–0.49, in association with triple vessel disease. The more contemporary Surgical Treatment for Ischemic Heart Failure (STICH) trial of subjects with more severe impairment of LV function (EF <0.35) demonstrated no survival difference at 5 years between CABG and OMT, although CV mortality was reduced, as were hospitalization rates for major CV causes, in the CABG group. Moreover, if the data are analysed by treatment received and as per protocol, due to the large number of crossovers to both therapies, the differences in all-cause mortality reached statistical significance in favour of the CABG group; in this respect the trial can be considered to have demonstrated a modestly positive result in favour of surgery, with potentially important clinical implications. The subset analysed by viability testing is inconclusive.
8.3.1.3 Multivessel Disease and/or Large Ischaemic Territory: Observational studies from the CASS registry and the meta-analysis of seven randomized trials—comprising a total of 2649 patients—of CABG vs. medical therapy suggested a survival advantage of surgery in patients with three-vessel disease (or LM disease), but no difference in patients with one- or two-vessel disease, except in patients with involvement of the proximal LAD plus one other major coronary artery. In addition, these studies demonstrated a greater efficacy from CABG over medical therapy in respect of symptom relief, bearing in mind the caveat that, in these trials, the methods of medical therapy and secondary prevention were obsolete by today's standards. In the more contemporary Medical, Angioplasty, or Surgery Study (MASS II) trial of CABG, PCI and medical therapy, patients treated with CABG enjoyed a better survival and lower rates of subsequent MI and need for additional revascularization procedures over a follow-up period of 10 years. The importance of the severity of symptoms was emphasized by two studies from the CASS registry, which demonstrated that, in patients with mild angina pectoris and triple vessel disease, survival advantage was confined to those with mild-to-moderate LV dysfunction. On the other hand, among patients with severe angina, survival was improved irrespective of LV function. In addition, the greater the number of proximal stenosis, the greater the surgical benefit. Observational studies also support a survival advantage for CABG in patients with double-vessel disease in the presence of severe or extensive ischaemia or severe angina. The concept of a revascularization benefit in patients with extended ischaemia is currently being tested in the ongoing International Study of Comparative Health Effectiveness with Medical and Invasive Approaches (ISCHEMIA).
8.3.1.4 Left Main Coronary Artery Disease: The survival advantages for bypass surgery in patients with 50% or greater stenosis of the LM coronary artery was established by the striking differences noted in the Veterans Administration Cooperative Study in a subgroup of 113 patients, and confirmed in a subsequent meta-analysis, and in studies from the CASS registry. The data now need to be re-interpreted in the light of more recent data evaluating the functional severity of LM stenoses and the possibility that revascularization can be safely deferred if FFR >0.80.
Irrespective, LM CAD (stenosis 50% or greater) continues to be a Class 1 indication for revascularization. No further randomized, controlled trials of bypass surgery or PCI vs. medical therapy are likely to be undertaken in patients with LM CAD.
8.3.2 Revascularization in Lower-risk Populations. 8.3.2.1 The Randomized Studies (See Web Addenda): The older randomized studies that investigated revascularization vs. OMT are selectively reviewed in the web addenda. The three most recent studies are also the largest and most informative studies for this comparison of revascularization with OMT.
The Clinical Outcomes Utilizing Revascularization and Aggressive Drug Evaluation (COURAGE) trial (n = 2287) compared PCI + OMT with OMT only, in patients with SCAD or ischaemia and coronary lesions suitable for PCI. The target study population for the COURAGE trial were patients with chronic angina pectoris Canadian Cardiovascular Society (CCS) Class I–III), stable post-MI patients and asymptomatic patients with objective evidence of myocardial ischaemia. All patients had angiographically defined CAD, with at least one vessel meeting AHA/American College of Cardiology (ACC) Class I or II indications for PCI. Patients with a prior CABG were accepted. Patients with stenosis >80% in one or more vessels, subtending a large area of myocardium, could be enrolled even in the absence of objective ischaemia. The primary endpoint of all-cause death or non-fatal MI did not differ between the two groups during a mean follow-up of 4.6 years. However, in patients who were invasively treated, freedom from angina was significantly better up to 3 years of follow-up. In a sub-study, patients with >10% ischaemia on stress myocardial perfusion scintigraphy had a higher rate of death or MI. More PCI + OMT patients exhibited significant ischaemia reduction (33 vs. 19%; P = 0.0004). Patients with ischaemia reduction had lower unadjusted risk for death or MI, particularly if baseline ischaemia was moderate to severe.
The Clinical Outcomes Utilizing Revascularization and Aggressive Drug Evaluation (BARI-2D) trial (n = 2368) evaluated whether PCI or CABG (choice left to the discretion of the treating physician), combined with OMT, would be better than OMT alone in patients with SCAD and type 2 diabetes mellitus. The target study population were patients with a diagnosis of type 2 diabetes and angiographically documented CAD for which revascularization was not required for prompt control of severe or unstable angina. Patients with stenosis >70% presenting with angina symptoms were eligible for randomization, even without documented ischaemia. In contrast, approximately 30% of patients were asymptomatic with a positive stress test. The primary endpoint of all-cause mortality at 5 years follow-up did not differ between the two treatment strategies, nor did rates of MI or stroke. The patients with most severe disease were selected for CABG rather than PCI and were a higher-risk group that drew a greater benefit from early revascularization (reduction of MI compared with OMT).
In the Fractional Flow Reserve vs. Angiography for Multivessel Evaluation (FAME-2) trial, 888 SCAD patients with functionally significant stenosis (FFR ≤0.80) were randomly assigned to FFR-guided PCI plus OMT, or to OMT alone. The target study population were patients who had at least one functionally significant stenosis and, on average, large areas of ischaemic myocardium (mean FFR value of 0.68) while the low-risk patients with non-ischaemic FFR values were not randomized but followed in a separate registry. The study was stopped prematurely by the Data Safety Monitoring Board, due to a highly significant reduction in hospital re-admission and urgent revascularization in the FFR ≤0.80-PCI group, compared with the FFR ≤0.80-OMT group. There was no difference in rates of death or MI between the two strategies. In patients without ischaemia (registry), the outcome appeared to be favourable with OMT only.
Altogether, seven major (n of 200 or more) randomized trials of revascularization vs. medical therapy in chronic SCAD have been published over the past 10 years (Table 33). Typically, populations of these studies were selected after an angiogram, had demonstrated at least one significant stenosis of an epicardial coronary artery in patients with typical or suspected angina—with or without documented myocardial ischaemia—with, in general, good LV function, no comorbidities and excluding patients at high angiographic risk, patients with LM coronary disease, CABG, multivessel disease, or lesions deemed to be treated with revascularization without further discussion for OMT only.
The results of these studies comparing myocardial revascularization with OMT have been somewhat consistent in confirming that, except for better symptom relief and lesser frequency of urgent revascularization, there is no advantage of revascularization over OMT alone to reduce mortality in angiographically selected patients presenting with SCAD, acknowledging the possibility of crossover from medical therapy to intervention during follow-up. Although interventional and surgical techniques have improved in the past two decades, medical therapy has also improved over the same period.
As a result, OMT may substantially improve long-term outcomes of patients treated conservatively but also of patients undergoing revascularization, reducing the impact of revascularization itself on survival in non-ACS patients.
8.3.2.2 Limitations of the Randomized Studies (See Web addenda): All of these studies have their limitations, which may limit their general applicability. These limitations are discussed in the additional web addenda. In brief:
8.3.2.3 Overall Interpretation: In low-risk, stable CAD patients, after an ischaemic documentation and a careful clinical and angiographic selection, the strategy of initial OMT is safe and should be the default approach. When a period of OMT has not been adequately conducted, cardiologists and surgeons should be more conservative when making a decision over revascularization, especially in case of high-risk comorbidities, difficult anatomies, mildly symptomatic patients or in patients without extensive provocable ischaemia. The trials have shown that, despite frequent crossovers to revascularization, the majority of patients remain on OMT alone for the duration of the trial.
When initial OMT has failed and patients remain symptomatic, or when the ischaemic risk appears important, the various options need to be discussed (OMT reinforcement or revascularization). The advantages, limitations and advice from the Heart Team must be fully presented in the discussion with the patient.
The early hazards of revascularization are well known: early periprocedural MI, stent thrombosis or late restenosis (much reduced now by second-generation DES) after PCI, peri-operative MI, stroke, cognitive dysfunction, surgical wound infection, prolonged hospital stay and rehabilitation after CABG. Potential advantages of an initial revascularization strategy (PCI or CABG) include better relief of symptoms and no significant excess of mortality, fewer drugs, fewer hospital visits and less revascularization within the first year with globally improved QoL. The advantage of revascularization over OMT on symptom relief is, however, blunted over time. OMT is safer in the short term, and as safe as revascularization for mortality up to 5 years in patients meeting the low-risk inclusion criteria of these trials. However, OMT requires larger doses and numbers of medications that may have a direct impact on adherence to treatment, side-effects, drug interactions, QoL and long-term cost to the patient and third party payers.
8.3.2.4 Ongoing Studies for Management of Stable Coronary Artery Disease Patients With Demonstrated Ischaemia: Several studies have suggested that patients with more extensive ischaemia benefit from revascularization—and that this benefit could translate into a long-term survival benefit if ischaemia is severe and reduction of ischaemia is significant. This hypothesis has been poorly investigated prospectively, although the positive randomized trials 'Asymptomatic Cardiac Ischaemia Pilot' (ACIP) and SWISSI II—with subset analyses of the CABG population in BARI 2D, plus the results of PCI in FAME 2—strongly suggest that ischaemia plays a key role in the benefit afforded by revascularization.
The hypothesis of deciding upon an invasive approach prior to angiography and not after (as in COURAGE and BARI 2D)—on the basis of documented clinically meaningful ischaemia during stress testing or haemodynamic assessment of stenoses identified at the time of angiography—certainly needs re-evaluation. This hypothesis is currently being evaluated in randomized trials: in the ongoing ISCHEMIA trial, patients are randomized—before coronary angiography—to a conservative OMT strategy or to an invasive strategy when they have documented myocardial ischaemia, the primary endpoint being death or MI.
8.4.1 Recent Data and Recommendations. The relative indications for PCI and CABG in SCAD patients have been clearly defined by the recent recommendations. There has been an increasing recognition of the value of the Heart Team in reaching consensus over if, when, and how to revascularize patients. Figures 6 and 7 show suggested algorithms to help simplify the decision-making process and to possibly avoid the need for systematic discussion of every patient with locally agreed protocols (refer to specific ESC Guidelines on myocardial revascularization for class and LOE concerning the respective indications of PCI and CABG). The Guidelines emphasize the importance of OMT in all patients and for both procedures, and the pivotal role of the heart team in most decisions over revascularization in patient with multivessel or left main disease. This is particularly true for patients with three-vessel disease when they present with a syntax score ≥22 or when complete revascularization is not achievable by one technique of revascularization, or when they have diabetes. For these patients, CABG should most often be the preferred option.
(Enlarge Image)
Figure 6.
Percutaneous coronary intervention (PCI) or coronary artery bypass graft surgery (CABG) in stable coronary artery disease without left main coronary artery involvement. CABG = coronary artery bypass graft; LAD = left anterior descending; PCI = percutaneous coronary intervention.
>50% stenosis and proof of ischaemia, >90% stenosis in two angiographic views, or FFR = 0.80.
CABG is the preferred option in most patients unless patients co-morbidities or specificities deserve discussion by the heart team. According to local practice (time constraints, workload) direct transfer to CABG may be allowed in these low risk patients, when formal discussion in a multidisciplinary team is not required (adapted from ESC/EACTS Guidelines on Myocardial Revascularization 2010).
(Enlarge Image)
Figure 7.
Percutaneous coronary intervention (PCI) or coronary artery bypass graft surgery (CABG) in stable coronary artery disease with left main coronary artery involvement. CABG = coronary artery bypass graft; PCI = percutaneous coronary intervention.
>50% stenosis and proof of ischaemia, >70% stenosis in two angiographic views, or fractional fow reserve = 0.80.
Preferred option in general. According to local practice (time constraints, workload) direct decision may be taken without formal multidisciplinary discussion, but preferably with locally agreed protocols (adapted from ESC/EACTS Guidelines on Myocardial Revascularization 2010).
In SYNergy between percutaneous coronary intervention with TAXus and cardiac surgery (SYNTAX), 1800 patients with three-vessel or left main coronary artery disease were randomized to undergo CABG or PCI. Publication of the 5-year results of the SYNTAX trial have confirmed the initial findings, with a higher rate of major adverse cardiac or cerebrovascular events at 12 months in the PCI group, in large part because of an increased rate of repeat revascularization. At 5 years, all-cause death was 13.9% with PCI, against 11.4% with CABG (P = 0.10) and cardiac death was 9.0 vs. 5.3% (P = 0.003) in favour of CABG. MACE was significantly reduced with CABG also.
Interestingly, this benefit was driven by the upper two tertiles of the SYNTAX score; although PCI and CABG performed as well on all endpoints for SYNTAX scores of 22 or less, there was a clear benefit with CABG at 5 years, especially in patients with scores of 33 or more. In patients with intermediate or high SYNTAX scores, MACE was significantly increased with PCI (intermediate score, 25.8% of the CABG group vs. 36.0% of the PCI group; P = 0.008; high score, 26.8 vs. 44.0%; P < 0.0001).
These findings are consistent with the survival benefit of CABG reported in several large propensity-matched registries comparing outcome of PCI and CABG. Indeed, in a recently published study of 7235 pairs of patients, matched for numerous baseline characteristics, the overall 8-year survival rates were 78.0% for CABG and 71.2% for stenting (HR 0.68; 95% CI 0.64–0.74; P < 0.001). For anatomic groups, the HRs ranged from 0.53 (P < 0.001) for patients with three-vessel disease involving the proximal LAD to 0.78 (P = 0.05) for patients with two-vessel disease but no disease in the LAD artery. A lower risk of death after CABG was observed in all subgroups stratified by a number of baseline risk factors. Most recently, the Asymptomatic atrial fibrillation and Stroke Evaluation in pacemaker patients and the atrial fibrillation Reduction atrial pacing Trial (ASSERT) reported survival in 86 244 CABG and 103 549 PCI propensity-matched patients with two- or three-vessel CAD. At 4-year follow-up there was increased mortality with PCI, compared with CABG. Despite statistical adjustment, this huge registry cannot eliminate confounding variables and the fact that sicker patients may have been assigned to PCI.
In SYNTAX, the results for 705 patients with LMS disease differ from the remaining patients with three-vessel CAD. For these patients, there was no overall difference between CABG and PCI in terms of death (8.4% CABG vs. 7.3% PCI; P = 0.64) or MI (4.1 vs. 6.9%; P = 0.14) but a higher incidence of stroke with CABG (4 vs. 1.2%; P = 0.02). The advantage of CABG was reduced repeat revascularization at 12% vs. 20% for stents (P = 0.004).
'Premier of Randomized Comparison of Bypass Surgery vs. Angioplasty Using Sirolimus-Eluting Stent in Patients with Left Main Coronary Artery Disease' (PRECOMBAT), another randomized trial of 600 patients with left main stem (LMS) disease, reported a composite endpoint of death, cerebrovascular accident and MI as 4.7% for CABG and 4.4% for PCI. Furthermore, the incidence of stroke was substantially lower than in SYNTAX and similar for PCI (0.4%) and CABG (0.7%). It should be acknowledged that Left Main SYNTAX was a subgroup analysis and PRECOMBAT was not powered to detect a difference in hard clinical endpoint. Accordingly, further large randomized controlled trials are needed to establish the optimal mode of LM revascularization with this degree of complexity [e.g. the Evaluation of XIENCE PRIME or XIENCE V vs. Coronary Artery Bypass Surgery for Effectiveness of Left Main Revascularization (EXCEL)].
Meanwhile, angiographic characteristics of the LM disease are key in selection between PCI and CABG (calcifications, ostial/mid/distal, LM size, distal lesions, etc.) and, for at least lower severity of LMS disease, PCI produces at least equivalent—if not superior—outcomes to CABG.
The Design of the Future Revascularization Evaluation in patients with Diabetes mellitus: Optimal management of Multivessel disease (FREEDOM) trial (see chapter 9.2 on diabetes for details)—which demonstrated a significant reduction on the primary ischaemic outcome at 5 years in patients treated with CABG vs. PCI—taking into consideration previous trials, suggests that there is a significant mortality benefit from bypass surgery vs. PCI in diabetic patients with multivessel disease when both options of revascularization are technically feasible, albeit at the price of an increased risk of non-fatal stroke.
The role of hybrid coronary revascularization (Figure 5) combining Left Internal Mammary Artery (LIMA)-to-LAD artery grafting and PCI of at least one non-LAD artery is evolving and is an option in patients with multivessel disease and technical issues over one of the two techniques of revascularization, comorbidities, prior history of revascularization with limitations in graft conduits or access for PCI (e.g. occlusion). It allows complete revascularization with the advantages of both modes of revascularization. No randomized study, and only small series of patients have so far been published, precluding any firm recommendation.
8.4.2 Target Populations of the Randomized Studies (See Web Addenda). Over the last two decades there have been approximately 20 trials of PCI vs. CABG, which have consistently reported no overall difference in survival between the two interventional techniques, possibly related to the low risk of the populations studied.
In contrast, several propensity-matched registries have consistently demonstrated a survival benefit of CABG after intervention, accompanied by a marked reduction in the need for repeat intervention, although it may still be susceptible to confounding factors.
8.5.1 Scores (See Web Addenda). Interventional and surgical scores have been developed to evaluate the risk of the different revascularization strategies. Although we lack prospective validation of these scores in randomized studies comparing CABG to PCI, the recommendations—and now the practice—are heavily based on these scores as tools for decision-making in individual patients.
8.5.2 Appropriate Utilization of Revascularization (See Web Addenda). Appropriateness criteria are based upon expert consensus as to when a procedure is appropriate. This is, however, an important and complex area of concern as the cost of imaging and revascularization comes under increasing but appropriate scrutiny.
8. Revascularization
8.1 Percutaneous Coronary Intervention
Advances in techniques, equipment, stents and adjuvant therapy have established PCI as a routine and safe procedure in patients with SCAD and suitable coronary anatomy. The mortality risk associated with the procedure in SCAD is <0.5%. The efficacy of PCI in SCAD in comparison to medical therapy and CABG has been the subject of extensive evaluation.
8.1.1 Type of Stent and Dual Antiplatelet Therapy: Bare metal stents (BMS) are associated with a 20–30% rate of recurrence of angiographic stenosis within 6–9 months after implantation. Drug-eluting stents (DES) reduce the incidence of angiographic restenosis and ischaemia-driven repeat revascularization. For the first generation of DES, this benefit has been extensively demonstrated in spite of a slightly higher incidence of late and very late stent thrombosis, related to delayed endothelialization, which requires longer dual antiplatelet therapy (DAPT) to prevent stent thrombosis. First-generation sirolimus-eluting stents (SES) and paclitaxel-eluting stents (PES) have been extensively compared in head-to-head randomized controlled trials. Angiographic results were better with SES and translated into significant differences in terms of repeat revascularization. The most recent or 'second- generation' DES (with thinner struts and biodegradable or more biocompatible polymers) showed superior clinical outcomes for both efficacy and safety when compared with first-generation DES. Second-generation DES—preferring those tested in large all-comers trials and compared with other DES with proven outcome—are therefore the recommended option in SCAD patients with no contra-indication to DAPT (see Table 30).
A recent meta-analysis confirmed that clopidogrel pretreatment in stable patients undergoing elective PCI does not reduce mortality or major adverse cardiac events (MACE), as compared with clopidogrel administration in the catheterization laboratory. On the basis of several randomized trials and this meta-analysis—and contrary to a diffuse common practice—SCAD patients who undergo diagnostic coronary angiography with the possibility of undergoing ad-hoc PCI (revascularization within the same procedure) should not be treated with clopidogrel before the coronary anatomy is known. The bleeding risk of routine DAPT, administered before catheterization in patients who do not require stenting (no significant CAD or CAD requiring CABG surgery), is not balanced by a detectable benefit in terms of ischaemic events in those undergoing PCI. Despite the overwhelming advantages shown in ACS patients—and especially diabetic patients—in the absence of randomized clinical trials, the use of prasugrel or ticagrelor cannot be recommended in SCAD patients undergoing elective PCI. An off-label use of these drugs is, however, common practice in some high-risk patients, especially in cases of documented stent thrombosis. After stenting, premature discontinuation of antiplatelet therapy is a major risk factor for stent thrombosis and should be avoided.
Current guidelines recommended 6–12 months of DAPT after first-generation stents. New-generation DES have been associated with lower rates of stent thrombosis, and recent data from registries and randomized controlled trials suggested that a shorter duration of DAPT might be sufficient in stable coronary patients. Considering the risk–benefit ratio of DAPT beyond 6 months—and while waiting for more information from the ongoing studies exploring various durations of treatment including more than a year—we endorsed the current ESC recommendation of 6–12 months of DAPT in SCAD patients undergoing PCI revascularization with a latest-generation DES (see section 9.5 for more details and for recommendations). Shorter durations (1–3 months) are reasonable in patients with high bleeding risk or undergoing undeferrable surgery or on concomitant anticoagulant treatment for which the use of clopidogrel only has shown significant advantages in a single small-scale trial (What is the Optimal antiplatElet and anticoagulant therapy in patients with oral anticoagulation and coronary StenTing (WOEST)).
8.1.2 Intracoronary Assessment of Stenosis Severity (Fractional Flow Reserve, Intravascular Ultrasound and Optical Coherence Tomography) (See Web Addenda). When non-invasive stress imaging is contra-indicated, non-diagnostic, or unavailable, the measurement of FFR during adenosine infusion is particularly helpful in identify haemodynamically or functionally significant stenosis, inducing ischaemia, justifying revascularization (see Table 31). In patients with FFR >0.80, studies in the BMS era have demonstrated that medical treatment provides better outcomes than immediate revascularization. Accordingly, a patient with a stenosis and an FFR >0.80 (two measurements or during adenosine infusion) should not be revascularized. The recent Fractional Flow Reserve vs. Angiography for Multivessel Evaluation (FAME-2) study confirmed that SCAD patients with stenoses having FFR ≤0.80 gain a benefit from PCI revascularization in addition to OMT, a benefit driven only by the reduced need for urgent revascularization (the study was stopped prematurely by the DSMB for that reason). Patients without ischaemia have excellent outcomes on medical therapy alone. Although the study suffers from significant limitations (the open nature of the trial may have affected the decision on 'urgent' revascularization; low-risk population), FFR can guide PCI in a clinically-effective way.
Fractional flow reserve, although in general not useful in very high grade lesions (angiographically >90%), which practically always have an FFR ≤0.8, may help the decision on when to revascularize in many uncertain clinical conditions. One such condition is 'multivessel disease', which occurs in a very heterogeneous population. In these patients, FFR measurement may change the strategy of revascularization (PCI vs. CABG) and the extent of revascularization according to the functional assessment of stenoses in critical coronary locations. Another critical situation for revascularization is LM stenosis, a lesion site that is of major prognostic importance and often determines the type of treatment. A recent single-centre study showed that angiography is not always reliable in the determination of the severity of LM stenoses and that deferred revascularization, if FFR >0.80, may be a safe approach.
Another situation relevant to these Guidelines, where FFR may be useful, is in 'post-ACS' patients. Once the culprit lesion has been treated, the patient can be considered as a stable or stabilized CAD patient. Non-invasive stress testing/imaging immediately after the acute phase may be impossible, contra-indicated or hazardous. Non-culprit stenoses in patients with recent ACS can be evaluated by FFR, either during the index procedure or in a staged procedure.
The use of intravascular ultrasound (IVUS) has been broadly investigated in SCAD with many different subsets of lesions (see Table 31). Unlike FFR, IVUS is an imaging diagnostic tool and does not provide assessment of the functional severity of a stenosis. Previously accepted cut-off limits of 3.5 or 4.0 mm for major epicardial artery stenosis and 6.0 mm for left main stenosis have been shown to be unreliable and poorly correlated with FFR, with somewhat better results when the absolute IVUS measurements are corrected for the reference vessel size. Once the indication to treatment is established, when more information is needed, IVUS is far superior to FFR because it provides an anatomical characterization of the lesion in terms of vessel size and plaque composition and can control stent expansion and strut apposition. More recently, optical coherence tomography (OCT) has been developed as a new intracoronary imaging tool with superior resolution (<10 μm) able to offer detailed assessment of superficial components including measurements of the thickness of the fibrous cap of lipidic plaques. The usefulness of OCT in SCAD patients with vulnerable plaques has not been well established, and certainly treatment of non-functionally critical lesions based only on the presence of elements of instability is not recommended. The facilitated technique of image acquisition allows optimization of stent expansion and apposition, and long-term assessment of stent healing.
8.2 Coronary Artery Bypass Surgery
8.2.1 Arterial vs. Venous Grafts. For the last 25 years the principle technique underpinning CABG has been the use of an internal mammary artery (IMA) to the LAD coronary artery with supplemental vein grafts as required. This followed a seminal publication from the Cleveland Clinic in 1986, showing that an IMA to the LAD improved survival and reduced the subsequent incidence of MI, recurrent angina and the need for repeat revascularization.
Since then, several angiographic studies have confirmed the superior patency of both IMA grafts in comparison to vein grafts both early and late after CABG. Most importantly, this superior graft patency appears to translate into a survival benefit. In 2001 a systematic review comparing single (SIMA) and bilateral (BIMA) IMA grafting reported a significant survival benefit with BIMA grafts with a hazard ratio for death of 0.81. Recent studies have reported that a survival benefit of BIMA grafts extends to the second and third decades of follow-up, and especially in patients with diabetes.
Previous concerns that the use of BIMA grafting may increase early postoperative mortality and/or morbidity have been dispelled by the Arterial Revascularization Trial (ART) which, in one of the largest trials ever conducted in cardiac surgery, randomized 3102 patients to SIMA or BIMA, with supplemental grafts as necessary. Whilst the primary outcome of this trial is 10-year survival, an interim analysis of safety at 1 year showed similar mortality of around 2% in both groups, with no difference in the incidence of MI, death or stroke but a slight increase in the incidence of sternal wound reconstruction in the BIMA group (1.9 vs. 0.6%). The data are currently being analysed to determine the key patient and operating factors that predispose to sternal dehiscence.
The radial artery has also been proposed as a second arterial graft, rather than a second IMA graft. In two randomized trials, the radial artery patency at 1 year was variously reported to be 'superior' and 'equivalent' to that of vein grafts. In an additional small, randomized trial the 5-year patency of the radial artery was significantly superior to that of vein grafts when placed to the circumflex coronary system.
Nevertheless, despite angiographic and clinical evidence of the potential superiority of arterial grafts, the reality is that the vast majority of bypass grafts—with the exception of the IMA to the LAD—are performed with saphenous vein grafts. Best current evidence suggests that the patency rate of saphenous vein grafts is slightly lower in off-pump surgery and when harvested using endoscopic, rather than open techniques.
8.2.2 On-pump vs. Off-pump Surgery (See Web Addenda). Off-pump surgery was initially proposed almost three decades ago. Numerous randomized trials and meta-analyses have shown no significant beneficial effect on mortality, but there have been reductions in stroke, transfusion, re-operation for peri-operative bleeding and postoperative complications, possibly at the cost of an excess of repeat revascularization with off-pump CABG. The two largest randomized trials, the Veterans Affairs (VA) Randomized On/Off Bypass (ROOBY) (n = 2203) and The CABG Off or On Pump Revascularization Study (CORONARY) (n = 4752) both reported no difference in the primary composite endpoint at 30 days. ROOBY reported a poorer outcome (death or complication) in the off-pump composite endpoint at one year (9.9 vs. 7.4%) while CORONARY has still to report at the time of writing. In contrast to the randomized trials, several large propensity-matched registries, which generally include higher-risk patients, have reported a reduction in mortality in patients undergoing off-pump CABG, although off-pump surgery is still performed in a minority of centres.
8.3 Revascularization vs. Medical Therapy
8.3.1 General Rules for Revascularization (See Web Addenda). The decision to revascularize a patient should be based on the presence of significant obstructive coronary artery stenosis, the amount of related ischaemia and the expected benefit to prognosis and/or symptoms (Figure 5). There are many clinical, anatomical, technical and environmental factors that may be discussed before the benefit of revascularization can be anticipated (Table 32, Figure 5). The vast number of possible combinations makes absolute recommendations difficult to mandate in every situation. In this regard, for a given patient in a given hospital, clinical judgement with consensual rather than individual decision-making, with a Heart Team discussion, should prevail, although this has to be individualized since, in many patients, the preferred approach is often quite clear-cut.
(Enlarge Image)
Figure 5.
Global strategy of intervention in stable coronary artery disease (SCAD) patients with demonstrated ischaemia. CABG = coronary artery bypass graft; CAD= coronary artery disease; LAD = left anterior descending; LV = left ventricular; OMT = optimal medical treatment; PCI = percutaneous coronary intervention.
Indication of revascularization for prognosis or symptoms (see Table 32).
Not suitable for revascularization due to anatomy or clinical conditions.
See section 9.
When technically feasible, with an acceptable level of risk and a good life expectancy, revascularization is indicated in chronic angina refractory to OMT. It can also be considered as first-line treatment in the situations discussed below.
8.3.1.1 Post-myocardial Infarction: The Swiss Interventional Study on Silent Ischemia Type II (SWISSI II) trial, involving 201 patients with a recent ST-segment elevation MI or non-ST segment elevation MI, investigated whether revascularization with PCI was better than drug therapy in stable patients with silent myocardial ischaemia (see description below).
During a lengthy 10-year follow-up period, the primary endpoint, which was survival free of cardiac death, non-fatal MI or revascularization, was significantly better in the PCI group.
PCI also significantly reduced the rates of cardiac death and all-cause mortality or MI. In addition, objective evidence of ischaemia was reduced in the revascularization group.
The Danish trial in Acute Myocardial Infarction (DANAMI) compared a deferred invasive strategy of PCI or CABG with a conservative strategy in 503 patients with inducible myocardial ischaemia, who had received thrombolysis for a first MI. Stress testing was performed at discharge and in patients randomized to the invasive strategy; angiography was performed within two weeks of stress testing. Patients with unstable angina were excluded.
Angina plus ischaemia was present in 25%, angina alone in 16% and 57% had silent ischaemia on stress testing. At 2.5 years follow-up, the invasive strategy was associated with a reduction in the incidence of re-infarction and less frequent angina. This was noted in patients with both symptomatic and asymptomatic ischaemia.
In contrast, the Occluded Artery Trial (OAT), following a strategy of routine PCI 3 to 28 days after acute MI, found no discernible benefit in terms of death, re-infarction, or heart failure at 4-year follow-up among asymptomatic or minimally symptomatic patients with occlusion of the infarct-related artery. The findings of the OAT study, however, should not be interpreted as applying to all patients experiencing ST-elevation MI, but just to those with a late occluded artery and no- or minimal angina. Two smaller studies [The Open Artery Trial (TOAT) and Desobstruction Coronaire en Post-Infarctus (DECOPI)] dealt with similar situations of stable patients after a Q-wave MI, without residual ischaemia and a persistent total occlusion of the infarct-related artery, and these studies did not show any clinical benefit of stenting over medical therapy.
Post-lytic revascularization studies. Old trials (not discussed here) comparing an invasive- with a conservative approach after fibrinolytic therapy did not show any differences in patient outcomes, but these studies antedated the use of stenting and modern antiplatelet therapies. In contrast, more recent randomized studies, comparing systematic early PCI with a conservative ischaemia guided strategy, have demonstrated favourable trends with early PCI and a significant reduction of death or MI in a meta-analysis.
8.3.1.2 Left Ventricular Dysfunction: In general, revascularization improves survival in 'sicker' patients, especially in the presence of LV dysfunction. From the early days of coronary angiography, it has been well recognized that LV dysfunction is one of the most powerful indicators of an adverse prognosis. As techniques of revascularization improved, LV dysfunction has become a prime target—as opposed to a contra-indication—for coronary revascularization. Several older studies, including a meta-analysis, suggested that survival was improved by CABG over medical therapy in patients with mild-to-moderate systolic dysfunction. The CASS randomized trial of bypass surgery vs. medical therapy demonstrated no overall differences in survival, except in the subset of patients with an ejection fraction (EF) of 0.35–0.49, in association with triple vessel disease. The more contemporary Surgical Treatment for Ischemic Heart Failure (STICH) trial of subjects with more severe impairment of LV function (EF <0.35) demonstrated no survival difference at 5 years between CABG and OMT, although CV mortality was reduced, as were hospitalization rates for major CV causes, in the CABG group. Moreover, if the data are analysed by treatment received and as per protocol, due to the large number of crossovers to both therapies, the differences in all-cause mortality reached statistical significance in favour of the CABG group; in this respect the trial can be considered to have demonstrated a modestly positive result in favour of surgery, with potentially important clinical implications. The subset analysed by viability testing is inconclusive.
8.3.1.3 Multivessel Disease and/or Large Ischaemic Territory: Observational studies from the CASS registry and the meta-analysis of seven randomized trials—comprising a total of 2649 patients—of CABG vs. medical therapy suggested a survival advantage of surgery in patients with three-vessel disease (or LM disease), but no difference in patients with one- or two-vessel disease, except in patients with involvement of the proximal LAD plus one other major coronary artery. In addition, these studies demonstrated a greater efficacy from CABG over medical therapy in respect of symptom relief, bearing in mind the caveat that, in these trials, the methods of medical therapy and secondary prevention were obsolete by today's standards. In the more contemporary Medical, Angioplasty, or Surgery Study (MASS II) trial of CABG, PCI and medical therapy, patients treated with CABG enjoyed a better survival and lower rates of subsequent MI and need for additional revascularization procedures over a follow-up period of 10 years. The importance of the severity of symptoms was emphasized by two studies from the CASS registry, which demonstrated that, in patients with mild angina pectoris and triple vessel disease, survival advantage was confined to those with mild-to-moderate LV dysfunction. On the other hand, among patients with severe angina, survival was improved irrespective of LV function. In addition, the greater the number of proximal stenosis, the greater the surgical benefit. Observational studies also support a survival advantage for CABG in patients with double-vessel disease in the presence of severe or extensive ischaemia or severe angina. The concept of a revascularization benefit in patients with extended ischaemia is currently being tested in the ongoing International Study of Comparative Health Effectiveness with Medical and Invasive Approaches (ISCHEMIA).
8.3.1.4 Left Main Coronary Artery Disease: The survival advantages for bypass surgery in patients with 50% or greater stenosis of the LM coronary artery was established by the striking differences noted in the Veterans Administration Cooperative Study in a subgroup of 113 patients, and confirmed in a subsequent meta-analysis, and in studies from the CASS registry. The data now need to be re-interpreted in the light of more recent data evaluating the functional severity of LM stenoses and the possibility that revascularization can be safely deferred if FFR >0.80.
Irrespective, LM CAD (stenosis 50% or greater) continues to be a Class 1 indication for revascularization. No further randomized, controlled trials of bypass surgery or PCI vs. medical therapy are likely to be undertaken in patients with LM CAD.
8.3.2 Revascularization in Lower-risk Populations. 8.3.2.1 The Randomized Studies (See Web Addenda): The older randomized studies that investigated revascularization vs. OMT are selectively reviewed in the web addenda. The three most recent studies are also the largest and most informative studies for this comparison of revascularization with OMT.
The Clinical Outcomes Utilizing Revascularization and Aggressive Drug Evaluation (COURAGE) trial (n = 2287) compared PCI + OMT with OMT only, in patients with SCAD or ischaemia and coronary lesions suitable for PCI. The target study population for the COURAGE trial were patients with chronic angina pectoris Canadian Cardiovascular Society (CCS) Class I–III), stable post-MI patients and asymptomatic patients with objective evidence of myocardial ischaemia. All patients had angiographically defined CAD, with at least one vessel meeting AHA/American College of Cardiology (ACC) Class I or II indications for PCI. Patients with a prior CABG were accepted. Patients with stenosis >80% in one or more vessels, subtending a large area of myocardium, could be enrolled even in the absence of objective ischaemia. The primary endpoint of all-cause death or non-fatal MI did not differ between the two groups during a mean follow-up of 4.6 years. However, in patients who were invasively treated, freedom from angina was significantly better up to 3 years of follow-up. In a sub-study, patients with >10% ischaemia on stress myocardial perfusion scintigraphy had a higher rate of death or MI. More PCI + OMT patients exhibited significant ischaemia reduction (33 vs. 19%; P = 0.0004). Patients with ischaemia reduction had lower unadjusted risk for death or MI, particularly if baseline ischaemia was moderate to severe.
The Clinical Outcomes Utilizing Revascularization and Aggressive Drug Evaluation (BARI-2D) trial (n = 2368) evaluated whether PCI or CABG (choice left to the discretion of the treating physician), combined with OMT, would be better than OMT alone in patients with SCAD and type 2 diabetes mellitus. The target study population were patients with a diagnosis of type 2 diabetes and angiographically documented CAD for which revascularization was not required for prompt control of severe or unstable angina. Patients with stenosis >70% presenting with angina symptoms were eligible for randomization, even without documented ischaemia. In contrast, approximately 30% of patients were asymptomatic with a positive stress test. The primary endpoint of all-cause mortality at 5 years follow-up did not differ between the two treatment strategies, nor did rates of MI or stroke. The patients with most severe disease were selected for CABG rather than PCI and were a higher-risk group that drew a greater benefit from early revascularization (reduction of MI compared with OMT).
In the Fractional Flow Reserve vs. Angiography for Multivessel Evaluation (FAME-2) trial, 888 SCAD patients with functionally significant stenosis (FFR ≤0.80) were randomly assigned to FFR-guided PCI plus OMT, or to OMT alone. The target study population were patients who had at least one functionally significant stenosis and, on average, large areas of ischaemic myocardium (mean FFR value of 0.68) while the low-risk patients with non-ischaemic FFR values were not randomized but followed in a separate registry. The study was stopped prematurely by the Data Safety Monitoring Board, due to a highly significant reduction in hospital re-admission and urgent revascularization in the FFR ≤0.80-PCI group, compared with the FFR ≤0.80-OMT group. There was no difference in rates of death or MI between the two strategies. In patients without ischaemia (registry), the outcome appeared to be favourable with OMT only.
Altogether, seven major (n of 200 or more) randomized trials of revascularization vs. medical therapy in chronic SCAD have been published over the past 10 years (Table 33). Typically, populations of these studies were selected after an angiogram, had demonstrated at least one significant stenosis of an epicardial coronary artery in patients with typical or suspected angina—with or without documented myocardial ischaemia—with, in general, good LV function, no comorbidities and excluding patients at high angiographic risk, patients with LM coronary disease, CABG, multivessel disease, or lesions deemed to be treated with revascularization without further discussion for OMT only.
The results of these studies comparing myocardial revascularization with OMT have been somewhat consistent in confirming that, except for better symptom relief and lesser frequency of urgent revascularization, there is no advantage of revascularization over OMT alone to reduce mortality in angiographically selected patients presenting with SCAD, acknowledging the possibility of crossover from medical therapy to intervention during follow-up. Although interventional and surgical techniques have improved in the past two decades, medical therapy has also improved over the same period.
As a result, OMT may substantially improve long-term outcomes of patients treated conservatively but also of patients undergoing revascularization, reducing the impact of revascularization itself on survival in non-ACS patients.
8.3.2.2 Limitations of the Randomized Studies (See Web addenda): All of these studies have their limitations, which may limit their general applicability. These limitations are discussed in the additional web addenda. In brief:
Some patient subsets that are commonly encountered in clinical practice were under-represented and the amount of evidence may appear insufficient, or even contradictory to other studies.
Patients were considered for inclusion only after coronary angiography was performed: therefore conclusions from these trials cannot be extended to patients with unknown coronary anatomy.
Crossover rates from OMT to revascularization were high and much higher than initially expected, suggesting that revascularization was merely deferred in 33–42% of patients randomized to a conservative approach.
Documented ischaemia was not mandatory for enrolment in COURAGE or in BARI 2D. Many patients with severe ischaemia—and, as such, considered to be at higher risk—were not randomized in the study.
The rapid evolution of revascularization techniques (e.g. DES for PCI and arterial grafts for CABG) and antiplatelet, anticoagulant, lipid-lowering and anti-ischaemic drugs render many of the studies obsolete by contemporary standards or difficult to interpret (e.g. stents used were mostly BMS).
OMT was particularly well carried out in these trials (not reflecting current practice), which emphasizes the need to educate physicians in clinical practice about the necessity and the scope of OMT.
COURAGE and BARI 2D failed to meet the statistical endpoint of superiority and, as such, were neutral trials demonstrating that an initial approach of intervention was neutral upon death or MI.
Several meta-analyses of randomized studies have shown divergent results on hard outcomes, as have registries with propensity analyses.
8.3.2.3 Overall Interpretation: In low-risk, stable CAD patients, after an ischaemic documentation and a careful clinical and angiographic selection, the strategy of initial OMT is safe and should be the default approach. When a period of OMT has not been adequately conducted, cardiologists and surgeons should be more conservative when making a decision over revascularization, especially in case of high-risk comorbidities, difficult anatomies, mildly symptomatic patients or in patients without extensive provocable ischaemia. The trials have shown that, despite frequent crossovers to revascularization, the majority of patients remain on OMT alone for the duration of the trial.
When initial OMT has failed and patients remain symptomatic, or when the ischaemic risk appears important, the various options need to be discussed (OMT reinforcement or revascularization). The advantages, limitations and advice from the Heart Team must be fully presented in the discussion with the patient.
The early hazards of revascularization are well known: early periprocedural MI, stent thrombosis or late restenosis (much reduced now by second-generation DES) after PCI, peri-operative MI, stroke, cognitive dysfunction, surgical wound infection, prolonged hospital stay and rehabilitation after CABG. Potential advantages of an initial revascularization strategy (PCI or CABG) include better relief of symptoms and no significant excess of mortality, fewer drugs, fewer hospital visits and less revascularization within the first year with globally improved QoL. The advantage of revascularization over OMT on symptom relief is, however, blunted over time. OMT is safer in the short term, and as safe as revascularization for mortality up to 5 years in patients meeting the low-risk inclusion criteria of these trials. However, OMT requires larger doses and numbers of medications that may have a direct impact on adherence to treatment, side-effects, drug interactions, QoL and long-term cost to the patient and third party payers.
8.3.2.4 Ongoing Studies for Management of Stable Coronary Artery Disease Patients With Demonstrated Ischaemia: Several studies have suggested that patients with more extensive ischaemia benefit from revascularization—and that this benefit could translate into a long-term survival benefit if ischaemia is severe and reduction of ischaemia is significant. This hypothesis has been poorly investigated prospectively, although the positive randomized trials 'Asymptomatic Cardiac Ischaemia Pilot' (ACIP) and SWISSI II—with subset analyses of the CABG population in BARI 2D, plus the results of PCI in FAME 2—strongly suggest that ischaemia plays a key role in the benefit afforded by revascularization.
The hypothesis of deciding upon an invasive approach prior to angiography and not after (as in COURAGE and BARI 2D)—on the basis of documented clinically meaningful ischaemia during stress testing or haemodynamic assessment of stenoses identified at the time of angiography—certainly needs re-evaluation. This hypothesis is currently being evaluated in randomized trials: in the ongoing ISCHEMIA trial, patients are randomized—before coronary angiography—to a conservative OMT strategy or to an invasive strategy when they have documented myocardial ischaemia, the primary endpoint being death or MI.
8.4 Percutaneous Coronary Intervention vs. Coronary Artery Bypass Graft (See Web Addenda)
8.4.1 Recent Data and Recommendations. The relative indications for PCI and CABG in SCAD patients have been clearly defined by the recent recommendations. There has been an increasing recognition of the value of the Heart Team in reaching consensus over if, when, and how to revascularize patients. Figures 6 and 7 show suggested algorithms to help simplify the decision-making process and to possibly avoid the need for systematic discussion of every patient with locally agreed protocols (refer to specific ESC Guidelines on myocardial revascularization for class and LOE concerning the respective indications of PCI and CABG). The Guidelines emphasize the importance of OMT in all patients and for both procedures, and the pivotal role of the heart team in most decisions over revascularization in patient with multivessel or left main disease. This is particularly true for patients with three-vessel disease when they present with a syntax score ≥22 or when complete revascularization is not achievable by one technique of revascularization, or when they have diabetes. For these patients, CABG should most often be the preferred option.
(Enlarge Image)
Figure 6.
Percutaneous coronary intervention (PCI) or coronary artery bypass graft surgery (CABG) in stable coronary artery disease without left main coronary artery involvement. CABG = coronary artery bypass graft; LAD = left anterior descending; PCI = percutaneous coronary intervention.
>50% stenosis and proof of ischaemia, >90% stenosis in two angiographic views, or FFR = 0.80.
CABG is the preferred option in most patients unless patients co-morbidities or specificities deserve discussion by the heart team. According to local practice (time constraints, workload) direct transfer to CABG may be allowed in these low risk patients, when formal discussion in a multidisciplinary team is not required (adapted from ESC/EACTS Guidelines on Myocardial Revascularization 2010).
(Enlarge Image)
Figure 7.
Percutaneous coronary intervention (PCI) or coronary artery bypass graft surgery (CABG) in stable coronary artery disease with left main coronary artery involvement. CABG = coronary artery bypass graft; PCI = percutaneous coronary intervention.
>50% stenosis and proof of ischaemia, >70% stenosis in two angiographic views, or fractional fow reserve = 0.80.
Preferred option in general. According to local practice (time constraints, workload) direct decision may be taken without formal multidisciplinary discussion, but preferably with locally agreed protocols (adapted from ESC/EACTS Guidelines on Myocardial Revascularization 2010).
In SYNergy between percutaneous coronary intervention with TAXus and cardiac surgery (SYNTAX), 1800 patients with three-vessel or left main coronary artery disease were randomized to undergo CABG or PCI. Publication of the 5-year results of the SYNTAX trial have confirmed the initial findings, with a higher rate of major adverse cardiac or cerebrovascular events at 12 months in the PCI group, in large part because of an increased rate of repeat revascularization. At 5 years, all-cause death was 13.9% with PCI, against 11.4% with CABG (P = 0.10) and cardiac death was 9.0 vs. 5.3% (P = 0.003) in favour of CABG. MACE was significantly reduced with CABG also.
Interestingly, this benefit was driven by the upper two tertiles of the SYNTAX score; although PCI and CABG performed as well on all endpoints for SYNTAX scores of 22 or less, there was a clear benefit with CABG at 5 years, especially in patients with scores of 33 or more. In patients with intermediate or high SYNTAX scores, MACE was significantly increased with PCI (intermediate score, 25.8% of the CABG group vs. 36.0% of the PCI group; P = 0.008; high score, 26.8 vs. 44.0%; P < 0.0001).
These findings are consistent with the survival benefit of CABG reported in several large propensity-matched registries comparing outcome of PCI and CABG. Indeed, in a recently published study of 7235 pairs of patients, matched for numerous baseline characteristics, the overall 8-year survival rates were 78.0% for CABG and 71.2% for stenting (HR 0.68; 95% CI 0.64–0.74; P < 0.001). For anatomic groups, the HRs ranged from 0.53 (P < 0.001) for patients with three-vessel disease involving the proximal LAD to 0.78 (P = 0.05) for patients with two-vessel disease but no disease in the LAD artery. A lower risk of death after CABG was observed in all subgroups stratified by a number of baseline risk factors. Most recently, the Asymptomatic atrial fibrillation and Stroke Evaluation in pacemaker patients and the atrial fibrillation Reduction atrial pacing Trial (ASSERT) reported survival in 86 244 CABG and 103 549 PCI propensity-matched patients with two- or three-vessel CAD. At 4-year follow-up there was increased mortality with PCI, compared with CABG. Despite statistical adjustment, this huge registry cannot eliminate confounding variables and the fact that sicker patients may have been assigned to PCI.
In SYNTAX, the results for 705 patients with LMS disease differ from the remaining patients with three-vessel CAD. For these patients, there was no overall difference between CABG and PCI in terms of death (8.4% CABG vs. 7.3% PCI; P = 0.64) or MI (4.1 vs. 6.9%; P = 0.14) but a higher incidence of stroke with CABG (4 vs. 1.2%; P = 0.02). The advantage of CABG was reduced repeat revascularization at 12% vs. 20% for stents (P = 0.004).
'Premier of Randomized Comparison of Bypass Surgery vs. Angioplasty Using Sirolimus-Eluting Stent in Patients with Left Main Coronary Artery Disease' (PRECOMBAT), another randomized trial of 600 patients with left main stem (LMS) disease, reported a composite endpoint of death, cerebrovascular accident and MI as 4.7% for CABG and 4.4% for PCI. Furthermore, the incidence of stroke was substantially lower than in SYNTAX and similar for PCI (0.4%) and CABG (0.7%). It should be acknowledged that Left Main SYNTAX was a subgroup analysis and PRECOMBAT was not powered to detect a difference in hard clinical endpoint. Accordingly, further large randomized controlled trials are needed to establish the optimal mode of LM revascularization with this degree of complexity [e.g. the Evaluation of XIENCE PRIME or XIENCE V vs. Coronary Artery Bypass Surgery for Effectiveness of Left Main Revascularization (EXCEL)].
Meanwhile, angiographic characteristics of the LM disease are key in selection between PCI and CABG (calcifications, ostial/mid/distal, LM size, distal lesions, etc.) and, for at least lower severity of LMS disease, PCI produces at least equivalent—if not superior—outcomes to CABG.
The Design of the Future Revascularization Evaluation in patients with Diabetes mellitus: Optimal management of Multivessel disease (FREEDOM) trial (see chapter 9.2 on diabetes for details)—which demonstrated a significant reduction on the primary ischaemic outcome at 5 years in patients treated with CABG vs. PCI—taking into consideration previous trials, suggests that there is a significant mortality benefit from bypass surgery vs. PCI in diabetic patients with multivessel disease when both options of revascularization are technically feasible, albeit at the price of an increased risk of non-fatal stroke.
The role of hybrid coronary revascularization (Figure 5) combining Left Internal Mammary Artery (LIMA)-to-LAD artery grafting and PCI of at least one non-LAD artery is evolving and is an option in patients with multivessel disease and technical issues over one of the two techniques of revascularization, comorbidities, prior history of revascularization with limitations in graft conduits or access for PCI (e.g. occlusion). It allows complete revascularization with the advantages of both modes of revascularization. No randomized study, and only small series of patients have so far been published, precluding any firm recommendation.
8.4.2 Target Populations of the Randomized Studies (See Web Addenda). Over the last two decades there have been approximately 20 trials of PCI vs. CABG, which have consistently reported no overall difference in survival between the two interventional techniques, possibly related to the low risk of the populations studied.
In contrast, several propensity-matched registries have consistently demonstrated a survival benefit of CABG after intervention, accompanied by a marked reduction in the need for repeat intervention, although it may still be susceptible to confounding factors.
8.5 Scores and Decisions (See Web Addenda)
8.5.1 Scores (See Web Addenda). Interventional and surgical scores have been developed to evaluate the risk of the different revascularization strategies. Although we lack prospective validation of these scores in randomized studies comparing CABG to PCI, the recommendations—and now the practice—are heavily based on these scores as tools for decision-making in individual patients.
8.5.2 Appropriate Utilization of Revascularization (See Web Addenda). Appropriateness criteria are based upon expert consensus as to when a procedure is appropriate. This is, however, an important and complex area of concern as the cost of imaging and revascularization comes under increasing but appropriate scrutiny.
Source...