Show Summary Details
Page of

Cardiac bypass and valve surgery 

Cardiac bypass and valve surgery
Cardiac bypass and valve surgery

Graham Cooper



New chapter including valve surgery and peroperative assessment of patients for cardiac surgery, includes content from retired Chapter 16.13.8.

Updated on 29 Oct 2015. The previous version of this content can be found here.
Page of

PRINTED FROM OXFORD MEDICINE ONLINE ( © Oxford University Press, 2015. All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a title in Oxford Medicine Online for personal use (for details see Privacy Policy).

Subscriber: null; date: 11 December 2017


Coronary artery bypass grafting (CABG)—the two main indications for are for relief of symptoms, usually angina and/or breathlessness, that persist even with optimal medical therapy (OMT), and/or prognosis. There is a prognostic benefit of CABG in patients with large volumes of ischaemia (i.e. affecting >12% of the ventricular mass), and the benefit of revascularization increases with increasing volumes of ischaemia. The overall mortality for elective CABG in the United Kingdom is around 1% and has continued to fall over the last decade despite an increasingly adverse risk profile of patients undergoing surgery. In randomized trials and large propensity-matched cohort registries CABG, in comparison to percutaneous coronary intervention (PCI) even with drug-eluting stents, has been shown to improve survival and to reduce the subsequent risk of myocardial infarction and recurrent angina. Approximately 80% of patients are alive a decade after surgery of whom around 70% are still free from angina.

Valve surgery—this is primarily performed for patients with severe valvular disease and symptoms. Indications also include deteriorating ventricular function and the requirement for coronary artery surgery in patients with coexistent valve disease. Mitral valve repair is a highly successful procedure in patients with nonrheumatic valvular regurgitation and is associated with an excellent long-term survival. Aortic valve disease is usually treated with aortic valve replacement. A range of biological and mechanical valves are available for valve surgery, with no difference in outcomes between mechanical and biological valves in respect of mortality, prosthetic valve endocarditis, or thromboembolism, but biological valves have a higher rate of reoperation and the haemodynamic profiles of the biological and newer mechanical valves are similar. Biological valves are particularly attractive for elderly patients in whom anticoagulation is deemed high risk, and now the commonest type of valves implanted worldwide. Patients with aortic stenosis may also be considered for transcatheter valve intervention (TAVI) when the risks of conventional surgery are high. The indications for TAVI for aortic stenosis are likely to expand significantly as the technique develops.


Valve surgery first developed in the 1920s for the treatment of congenital heart disease and mitral stenosis. The development of durable valve prostheses in the 1950s allowed surgery for a wider range of acquired valvular heart disease. Currently, degenerative disease causing aortic stenosis, aortic regurgitation, and mitral regurgitation is prevalent in North America and Europe; rheumatic heart disease remains a significant cause of valvular stenosis and/or regurgitation elsewhere. Every year, over 13 000 valve procedures are performed in the United Kingdom and almost 100 000 in the United States of America.

Coronary artery bypass grafting (CABG) has now been performed for almost half a century and it is estimated that approximately three-quarters of a million such operations are performed worldwide annually. Over the last decade the numbers of CABG operations have fallen in most developed countries because of improved optimal medical therapy and advances in percutaneous coronary intervention (PCI), while the numbers of CABG operations continue to increase in the developing world.

Attempts to improve the blood supply to the heart through indirect means were first attempted over a century ago. However, it was technological advances in the 1960s that allowed direct suturing of either the internal mammary artery (IMA) or saphenous grafts to the native coronary artery that led to dramatic improvements in the relief of angina and the explosive growth in CABG surgery. The publication of randomized trials comparing CABG to medical therapy in the 1970s demonstrated the superior efficacy of CABG in relieving angina and a subsequent meta-analysis of these trials also reported that CABG resulted in a survival benefit over a 10 year follow-up period. This led to further dramatic increases in the number of CABG operations in developed countries over the following two decades. Initially most CABG operations were performed using only saphenous vein grafts conduits but the demonstration of superior patency and clinical outcomes with an IMA graft eventually resulted in the vast majority of patients receiving an IMA graft to the anatomically and functionally most important coronary artery, the left anterior descending coronary artery. The superior angiographic patency of the IMA in comparison to vein grafts is largely explained by the tendency to development of intimal hyperplasia and atherosclerosis in vein grafts, a pathological process from which the IMA remains largely immune. Over the last decade there have been attempts to promote the use of more arterial grafts during CABG surgery, and particularly the use of both IMAs, as most patients need three bypass grafts. Over the last two decades there has also been considerable enthusiasm for the use of off-pump CABG to avoid the deleterious effects of cardiopulmonary bypass, but recent large trials have shown no difference in clinical outcome for the majority of patients whether CABG surgery is performed on or off pump.

General considerations in assessing patients for cardiac surgery

The decision to proceed to cardiac surgery involves a careful assessment of the operative risk. In an ageing population with multiple comorbidities these considerations become increasingly important and significantly influence the decision to intervene with surgery and the choice between surgery and PCI. The presence of significant comorbidity exerts a greater influence when surgery is being performed for prognostic rather than symptomatic grounds. In some patients long-term prognosis is determined to a greater degree by their comorbidity than by their coronary or valve disease, and in those who have asymptomatic disease the benefits of intervention have to be carefully weighed against the risks.

All patients will have routine haematological and biochemical assessment, coronary angiography, and echocardiography. Patients undergoing valve surgery should have a dental assessment including a panoramic radiograph. Angiographic assessment can be refined by the use of pressure wire studies, particularly in those cases where the presence of a given coronary stenosis will determine the choice between PCI and surgery. In patients in whom coronary bypass surgery is being performed for prognostic benefit, (in particular those with significant left ventricular impairment) assessment with myocardial perfusion imaging or MRI will guide the decision to revascularize based on the extent of viable myocardium and reversible ischaemia. Right heart catheterization may be required in the assessment of mitral valve disease or where significant pulmonary hypertension has been identified on echocardiography.

Antiplatelet therapy with the exception of aspirin should be withdrawn in patients undergoing elective surgery (see Box

Adapted from Sousa-Uva M, Storey R, Huber K, et al., on behalf of ESC Working Group on Cardiovascular Surgery and ESC Working Group on Thrombosis (2013). Expert position paper on the management of antiplatelet therapy in patients undergoing coronary artery bypass graft surgery. Eur Heart J, 10, 1093.

A number of scoring systems have been developed to estimate the risks of cardiac surgery. In the EuroSCORE II a number of parameters have been identified on univariate analysis to influence the outcome of surgery, as shown in Table

Table Variables associated with mortality for cardiac surgery (EuroSCORE II)

Patient-related factors

Cardiac-related factors

Operation-related factors


1CCS angina class 4

Urgent operation


NYHA class (>II)

Emergency operation

Extracardiac arteriopathy

Left ventricular ejection fraction

Salvage operation

Neurological/musculoskeletal dysfunction

Recent myocardial infarction

Additional surgical procedures

Previous cardiac surgery

Pulmonary artery pressure

Postinfarct VSD

Serum creatinine

Acitve endocarditis

Thoracic surgery

Critical preoperative statea

Aortic arch surgery

Critical preoperative state is defined as ventricular tachycardia or fibrillation, aborted sudden death or cardiac massage, ventilation prior to surgery, inotropic support, ventricular assist device/balloon pump pre-operatively or acute renal failure (anuria or oliguria <10 ml/h).

Adapted from Nashef AM, et al. (2012). Euroscore II. Eur J Cardiothorac Surg, 41, 734-45.

The operative mortality in elderly patients has fallen substantially over the past 30 years and it is no longer unusual to consider surgery in patients over the age of 80 if their overall risk is acceptable. The risk of coronary artery bypass surgery in patients over the age of 85 is approximately 9% compared to less than 1% in those under 61; the corresponding figures for isolated valve surgery are 7% and 2.6% respectively. The risks are substantially affected by comorbidities such as chronic obstructive airways disease, cerebrovascular disease, and renal disease, which are more common in this age group. Frailty in elderly people, though increasingly important, is difficult to define and is probably best assessed by an experienced physician reviewing the patient, although attempts have been made to develop a frailty index to assist in decision-making.

Moderate to severe chronic obstructive airways disease (i.e. FEV1/FVC <0.7 and FEV1 <80% predicted) increases surgical mortality threefold and if combined with a DLCO of less than 50% the mortality increases tenfold. Many patients with chronic obstructive pulmonary disease (COPD) are wrongly classified prior to cardiac surgery and routine pulmonary function testing in patients with a smoking history or history of COPD is advised.

Carotid artery disease is associated with an increased risk of stroke during cardiac surgery; however, there is no evidence that routine screening of all patients is required. Screening of patients aged over 70 with an additional risk factor (carotid bruit, history of cerebrovascular disease, diabetes mellitus, or peripheral vascular disease) is probably justified. Intervention for carotid disease should be considered at or before surgery in patients with a history of cerebrovascular disease and a carotid stenosis (50-99% in men and 70-99% in women). The role of carotid surgery in asymptomatic patients is controversial but it should be considered in men with bilateral severe carotid stenosis or contralateral occlusion if the operative complication rate for carotid surgery is low and life expectancy is good.

The 30 day mortality of patients with acute renal failure in the postoperative period approaches 60% in some series. The risk is largely dependent on the baseline creatinine clearance (see Fig. Cardiac surgery in patients on dialysis carries a threefold greater mortality and patients are more likely to suffer a stroke pneumonia or sepsis in the postoperative period. There is some evidence that off-pump bypass surgery reduces the risks of surgery in this group of patients.

Fig. Risk of acute renal failure according to baseline creatinine clearance.

Risk of acute renal failure according to baseline creatinine clearance.

Adapted from Chertow GM, Lazarus JM, Christiansen CL, Cook EF, Hammermeister KE, Grover F, & Daley J (1997). Preoperative renal risk stratification. Circulation, 95(4), 878–884.

The decision to proceed to cardiac surgery involves a multidisciplinary team of cardiologists, surgeons, and physicians and detailed preoperative assessment is required for an informed decision to be made.


Indications for revascularization by either PCI or CABG are shown in Table The major indications for CABG are the relief of angina or breathlessness in patients who remain symptomatic despite optimal medical therapy and for prognosis in patients with substantial volumes of ischaemia (classified as involving >12% of the ventricle mass).

Table Indications for revascularization in stable angina or silent ischaemia

Subset of coronary disease by anatomy

Evidence class

For prognosis

Left main stem stenosis >50%


Any proximal LAD >50%


Two-vessel or three-vessel disease with impaired LV function


Proven large area of ischaemia (>10%)


Single remaining patent vessel >50% stenosis


Single-vessel disease without proximal LAD and without >10% ischaemia


For symptoms

Any stenosis >50% with limiting angina or angina equivalent unresponsive to optimal medical treatment


Dyspnoea/CHF and >10% LV ischaemia/viability supplied by >50% stenotic artery


No limiting symptoms with optimal medical therapy


Recent guidelines published in Europe and North America broadly agree that there is a prognostic advantage of CABG in patients with the most severe coronary artery disease and particularly in the presence of complex three-vessel disease and/or left main disease. Revascularization is also indicated in patients with impaired left ventricular function and severe coronary artery disease and especially with the demonstration of significant ischaemia and viable myocardium.

Non-ST-elevation myocardial infarction

Patients with non-ST-elevation myocardial infarction (NSTEMI) often require urgent revascularization by either PCI or CABG. For isolated one- or two-vessel disease, and particularly where the culprit lesions are not complex, PCI is an appropriate strategy. In contrast, for those patients with complex multivessel coronary artery disease CABG is still the preferred treatment option soon after medical stabilization of the patient using optimal medical therapy.

ST-elevation myocardial infarction

There is universal agreement that the primary treatment of ST-elevation myocardial infarction (STEMI) is immediate PCI, preferably within 90 min. There is a prohibitively high risk of CABG surgery in patients with acute myocardial infarction. CABG is therefore reserved for patients who exhibit persistent symptoms or evidence of ischaemia despite PCI or who become haemodynamically unstable, particularly with mechanical complications of myocardial infarction such as papillary muscle rupture or ventricular septal defect.

The CABG operation

The vast majority of CABG operations are performed through a median sternotomy which allows excellent access to all anatomical regions of the heart. In certain situations CABG can be performed through a minithoracotomy with or without the aid of robotic instruments.

After median sternotomy one or both IMAs are harvested while vein graft from the leg and/or the radial artery may also be simultaneously be harvested as additional conduits. The left IMA remains attached proximally to the subclavian artery and the right IMA can either be left in situ or anastomosed as a composite graft to the left IMA.

Around 80% of all CABG operations are done with cardiopulmonary bypass by draining venous blood from the right atrium to the extracorporeal perfusion circuit, where it is oxygenated and cooled, and then returned to the ascending aorta so that the heart and lungs are effectively bypassed. A large clamp is then placed across the ascending aorta and a cardioplegia solution—usually either crystalloid or blood containing high concentrations of potassium—is used to arrest the heart to provide the surgeon with a motionless, bloodless operating field. After performance of the distal anastomosis the aortic clamp is removed so that the heart is reperfused and then the top end of the radial artery or vein graft is sewn to the ascending aorta after isolating part of the ascending aorta with a side-biting clamp.

If the operation is performed off pump (without the use of cardiopulmonary bypass) a stabilizing device is used to immobilize a small area of the heart to allow the anastomosis to be performed to the coronary artery.


The 10 year survival for a standard CABG operation using an IMA and saphenous vein grafts is expected to be in the region of 80%. Half of late deaths are due to vein graft failure and this has been a driving force for increasing the use of two IMAs. At 10 years the patency of the IMA is around 95% in comparison to 25 to 50% for vein grafts. Recent studies have shown that the patency of the IMA remains at over 90% two decades after follow-up.

In younger patients there is general agreement to try to maximize the use of mammary arteries and radial arteries because of their improved patency over the longer term. There is evidence that two IMAs improves survival and freedom from the need for further interventions in comparison to a single IMA.

Recurrent angina

As for survival, there is increasing evidence that the more frequent use of arterial grafts also reduces recurrent rates of myocardial infarction and recurrent angina.

Secondary prevention

The use of secondary prevention is mandatory in patients who have undergone any revascularization whether by PCI or CABG. Minimum therapy should be at least one antiplatelet medication, β‎-blockers, statins, and angiotensin converting enzyme (ACE) inhibitors in the presence of impaired left ventricular function.

The choice between CABG and PCI

There is strong evidence from randomized trials such as SYNTAX and FREEDOM (in diabetic patients), and from several large-scale propensity-matched registries with tens of thousands of patients, of a persistent survival advantage of CABG by around 5%, 3 to 5 years after intervention (see also Chapter 16.13.5). In patients with the most severe disease the difference in survival in favour of CABG is around 10%. These survival curves continue to diverge with further duration of follow-up, suggesting that over the longer term the benefits of CABG may be even greater. This difference between CABG and PCI has persisted despite advances in PCI technology from balloon angioplasty to bare metal stents to drug-eluting stents and to the newer generation of drug-eluting stents. The likely reason for the persistent survival advantage of CABG is that placing bypass grafts to the mid-coronary vessels makes the complexity of proximal coronary artery disease irrelevant and protects against the development of new proximal disease, which is still common despite optimal medical therapy. In contrast, PCI can only deal with localized suitable proximal culprit lesions and has no prophylactic benefit against the development of new disease.

Heart valve surgery


The indications for valve surgery are covered in more detail elsewhere (see Chapter 16.6). In brief, surgery is indicated for symptomatic (breathlessness, angina, syncope) severe valve disease or for asymptomatic severe valve disease with evidence of pathophysiological changes, e.g. abnormal exercise test for asymptomatic severe aortic stenosis, left ventricular dysfunction, pulmonary hypertension, or atrial fibrillation for asymptomatic severe mitral regurgitation.

Repair or replacement

The suitability and success of valve repair rather than replacement depends on valve pathology, the pathophysiological consequences, and surgical expertise. The advantages of valve repair are the avoidance of anticoagulation, prosthetic valve dysfunction, and paravalvular leak, with lower procedural risks and better long-term outcome.

Techniques for mitral valve repair for degenerative technique are well established with excellent long-term outcomes with respect to reoperation. More than 90% of degenerative mitral valves are suitable for repair using a combination of techniques: resection or plication of prolapsing or redundant leaflet tissue; chordal replacement with Gore-Tex neochords; or annuloplasty, usually with implantation of a prosthetic ring or band to support the repair and prevent further annular dilatation. The cumulative reoperation rate is less than 1%/year, better for isolated posterior leaflet repair (0.5%), and worse for bileaflet (0.9%) or anterior leaflet (1.6%) repairs. Current guidelines support early mitral valve repair for asymptomatic severe mitral regurgitation, with a high expectation of successful durable repair and low procedural mortality.

Surgical repair for rheumatic mitral valve disease is more limited, depending on the extent and chronicity of rheumatic changes: closed and open commissurotomy may be performed to palliate mitral stenosis. Several techniques for aortic valve repair for aortic regurgitation in bicuspid and trileaflet valves have been described to treat cusp, commissural, and annular pathology in selected cases, but long-term outcomes are unknown.

Surgical approaches

The majority of valve procedures are performed through a median sternotomy on cardiopulmonary bypass, as described for CABG. Several minimal-access approaches have been described that allow better cosmesis compared with median sternotomy. Aortic valve replacement may be undertaken through a partial upper sternotomy with a J-shaped or inverted T sternal incision through the third or fourth intercostal space, or through a right anterior thoracotomy. The mitral valve may be approached through a lower partial sternotomy, right thoracotomy, or a port access approach through the right chest using a thoracoscopic camera for guidance and specialized instruments; robotic mitral valve surgical techniques have also been developed, but these are limited to specialized centres owing to the high costs of a surgical robot. Depending on the exposure, these minimal-access approaches may require peripheral cannulation for cardiopulmonary bypass, with specialized surgical equipment for venting and arresting the heart, and clamping the aorta. There is a recognized learning curve for these newer surgical approaches, and, although a shorter in-hospital stay and faster early recovery have been reported, the medium-term outcomes remain equivalent to standard open approaches.

Transcatheter valve intervention

Percutaneous valve intervention techniques have been developed that have replaced surgery in cases with prohibitive surgical risk. Transcatheter aortic valve implantation (TAVI) for aortic stenosis uses standard pericardial bioprosthetic valves mounted in balloon-expandable or self-expanding alloy frames, implanted through the femoral or subclavian artery, ascending aorta, or left ventricular apex, depending on the type of device, presence of vascular disease, and institutional expertise. The procedural success rate is 95% with a 90% or less 30-day mortality and less than 2% stroke rate. TAVI is recommended for inoperable patients (logistic EuroSCORE ≥ 20, STS PROM ≥ 8) following the PARTNER B study that found a significant reduction in 2 year all-cause mortality with TAVI compared with optimal medical therapy in inoperable severe aortic stenosis (43.3% vs 68%). The PARTNER A study found TAVI to be noninferior to surgical aortic valve replacement with respect to 2 year all-cause mortality (33.9% vs 35%) in a high-risk surgical cohort (STS predicted mortality ≥10). TAVI devices for aortic regurgitation have not yet been widely introduced. Further improved devices are under development to facilitate intraprocedural positioning and to reduce the risks of acute coronary ostial occlusion and paravalvular leak.

Types of valve prosthesis

Biological valves

Biological or bioprosthetic valves may be xenografts, homografts (allografts), or autografts. Xenograft valves are made from glutaraldehyde-fixed animal leaflet tissue with a proprietary anticalcification treatment, most commonly bovine pericardium or porcine aortic valve mounted in an alloy frame for a stented valve, or a whole porcine aortic root for stentless valve. The advantages of stented xenograft valves are the ease of implantation, the avoidance of long-term anticoagulation, and the ease of reoperation; the development of transcatheter valve-in-valve implantation offers an additional less invasive option. Porcine stentless valves became popular in the 1990s because of their excellent haemodynamics and avoidance of long-term anticoagulation; however, these valves are more challenging to implant reliably, either as a subcoronary implant or as a mini-root replacement, and the rate of structural valve deterioration is higher than for stented valves.

Homografts (allografts) are antibiotic-treated cryopreserved cadaveric grafts including the aortic root and valve. Homografts are resistant to infection and are used for aortic root replacement, particularly for aortic valve endocarditis, in younger patients to avoid the need for anticoagulation, and where there is extensive periannular infection and tissue destruction to allow left ventricular outflow tract reconstruction. However, the durability at 10 years is similar to pericardial bioprosthetic valves, the reoperation rate for structural valve deterioration at 15 years is as high as 20% in patients aged 41 to 60 years, and reoperation is challenging owing to homograft calcification.

Finally, the Ross procedure described in 1962 uses a pulmonary autograft for aortic root replacement with the pulmonary outflow tract replaced by a homograft. The pulmonary autograft is viable tissue and is able to grow in young patients, has excellent haemodynamics with a low thromboembolic risk, and is resistant to infection. The complexity of the Ross procedure limits its use to specialist centres for selected cases, e.g. women of childbearing age keen to avoid anticoagulation. The Ross procedure is complicated by homograft stenosis in 10 to 20% and aneurysmal dilatation of the autograft causing aortic regurgitation; the 10 year structural valve deterioration rate is up to 30%.

’Sutureless’ or rapid-deployment valves are bioprosthetic aortic valves incorporating many features of transcatheter valves but allowing faster implantation in the debrided aortic annulus after open surgical resection of the diseased valve. Cardiopulmonary bypass and cardioplegic arrest are still required, but these valves facilitate minimally invasive approaches and allow shorter procedure times, although the longer-term benefits have yet to be confirmed.

Mechanical valves

Mechanical valves offer the advantages of excellent durability but the disadvantages of long-term anticoagulation and the risks of bleeding; modern low-profile valves have better haemodynamic properties and lower thromboembolic risk than earlier generations. The PROACT study is comparing standard anticoagulation against lower intensity anticoagulation for high thromboembolic risk cases and dual antiplatelet therapy for low-risk cases with the On-X bileaflet valve: early results are encouraging, with a 0.6%/year thromboembolic event rate and 0.4%/year significant bleeding rate.

Meta-analyses of the randomized studies comparing mechanical with biological valves have found no difference in outcomes between mechanical and biological valves with respect to mortality, prosthetic valve endocarditis, or thromboembolism; biological valves have a higher rate of reoperation, mechanical valves a higher risk of significant bleeding complications. The Veterans Administration study found a better 15 year survival for mechanical valves, but the Edinburgh Heart Valve trial found no difference in survival at 20 years. The choice of valve prosthesis for an individual patient depends on several factors including, most importantly, the wishes of the patient, age and life expectancy, metabolic factors predisposing to calcification and early structural valve deterioration (e.g. chronic kidney disease), any contraindication to anticoagulation, expectation of pregnancy, previous infection, and risk of reoperation. There has been a steady increase in the proportion of biological valves implanted over the last decade with these valves now making up more than 80% of valves implanted.


Anticoagulation for prosthetic valves

Anticoagulation is required for all currently available mechanical valves. The intensity of anticoagulation depends on valve characteristics and its position, and patient factors such as a history of thromboembolism, atrial fibrillation, left atrial enlargement, and left ventricular dysfunction. Current recommendations for anticoagulation are summarized in Box

Management of anticoagulation for noncardiac surgery

Anticoagulation is usually stopped for noncardiac surgery depending on the prosthesis type and bleeding risk of surgery. Patients with modern bileaflet or tilting disk mechanical aortic valves at low risk of thromboembolism and with no risk factors such as atrial fibrillation, history of thromboembolism or hypercoagulability, or left ventricular dysfunction, may stop warfarin 3 to 5 days before surgery, with no need for bridging therapy with low molecular weight or unfractionated heparin. In all other cases, bridging therapy is indicated before and after surgery for an INR of 2.0 or less; heparin should be resumed after surgery as soon as the immediate risk of bleeding has passed.

Excessive anticoagulation

Anticoagulation may need to be reversed because of an excessive INR, for bleeding, or for emergency surgery. Prothrombin complex concentrate is recommended for rapid reversal for bleeding. A mildly elevated INR with no signs of bleeding may be managed by the omission and/or adjustment of warfarin doses. Oral vitamin K and omission of warfarin are recommended for the correction of a higher INR with no bleeding.

Complications of cardiac surgery

Operative mortality

The overall mortality for all CABG in the United Kingdom is around 1.8%, being just under 1% for elective CABG and approximately 2% for all urgent CABG. Overall mortality has remained low despite an increasing risk profile in patients who are ever more elderly with significant comorbidities. Valve surgery caries a slightly higher risk: the mortality rates for uncomplicated mitral valve repair and aortic valve replacement are approximately 2%. A consistently low mortality almost certainly reflects improvements in medical management of patients as well improvements in anaesthetic, surgical. and perfusion techniques.

Neurological injury

Significant neurological injury is arguably the most feared complication of cardiac surgery and occurs with an incidence of around 1 to 2% during surgery or in the perioperative period. Of patients with neurological injury approximately one-third will die, one-third will remain severely disabled, and one-third will make a good recovery. The incidence of stroke is statistically higher in patients with left main disease than those with isolated three-vessel disease and this may reflect a concomitant higher burden of carotid artery disease in patients with left main disease. The major risk factors for stroke are advanced age, significant disease of the ascending aorta, carotid artery disease, previous neurological injury, and the development of postoperative atrial fibrillation. There is strong evidence that CABG performed off pump using a no-touch aortic technique is the strongest surgical methodology for reducing incidence of stroke.

Local and cardiac complications

Sternal wound dehiscence is another particularly troublesome complication of median sternotomy. The overall incidence is around 0.6% and the major risk factors are insulin-dependent diabetes and especially in combination with obesity. In such patients the use of two IMAs leads to a small but significant increase in this risk of sternal dehiscence and is therefore generally avoided. The treatment of sternal dehiscence is prolonged and complex and usually requires a period of suction dressings followed by reconstruction with muscle flaps.

Pleural effusion

Pleural effusions are usually small and self-limiting and easily treated by chest drainage. They can also occur after patient discharge as a late event.

Pericardial effusion

All patients develop pericardial effusions after cardiac surgery and in the vast majority they are self-limiting and require no specific therapy. A small percentage of patients may develop significant pericardial effusions which can usually be drained by a small incision under the xiphisternum or by using a thoracoscope through the pleural cavity and the pericardium. Pericardial effusions can also appear after patient discharge and can usually be drained without having to reopen the full sternotomy.

Atrial fibrillation

Atrial fibrillation occurs temporarily in around 30% of patients after CABG and the incidence can be reduced by peri- and postoperative β‎-blockade. It is now standard practice to also anticoagulate these patients as well as treat with amiodarone for 6 weeks. If the patient remains in atrial fibrillation after this period then cardioversion is indicated.

Conduction defects

Cardiac conduction defects are common after valve surgery, particularly aortic valve replacement owing to the proximity of the atrioventricular node and bundle of His to the right coronary-noncoronary commissure: conduction pathways may be damaged during valve debridement, by direct injury from a suture, or by postoperative oedema. First-degree or higher degrees of heart block are common after aortic valve surgery and most surgeons routinely place epicardial atrial and ventricular pacing wires for temporary postoperative pacing. Complete heart block requiring implantation of a permanent pacemaker is needed in 3 to 8% of aortic valve replacement cases, being more common in the elderly, pre-existing conduction defects, and redo valve surgery.

Structural valve deterioration

Acute primary valve failure is rare in current mechanical or biological valves, but emergent or urgent reoperation is indicated. Structural valve deterioration is a complication of biological valves owing to a leaflet fibrosis, and calcification causing progressive valvular stenosis, and perforation and leaflet tearing leading to regurgitation. Structural valve deterioration develops at a predictable rate related to younger patient age, valve position, mitral more affected than aortic, altered calcium metabolism (e.g. chronic kidney disease), and pregnancy. Pericardial valves deteriorate more slowly than porcine bioprostheses. The indications for reoperation for structural valve deterioration are the same as for native valve disease, based on symptoms, ventricular size and function, and pulmonary hypertension.


The incidence of clinical thromboembolic events is up to 2.3 cases per 100 patient-years. The risk is similar for biological and anticoagulated mechanical valves. Risk factors for thromboembolism include the prosthesis type and position, a history of thromboembolism or hypercoaguability, atrial fibrillation and left atrial size, and left ventricular dysfunction. Thromboembolism with a mechanical valve is managed by ensuring that the INR is in the therapeutic range, or if the INR is already therapeutic, by increasing the target INR or adding low-dose aspirin.

Prosthetic valve thrombosis

Thrombosis of a mechanical valve may be a life-threatening complication. The diagnosis is suggested by heart failure, signs of a low cardiac output, or thromboembolism with reduced or absent prosthetic valve sounds, new murmurs, or documented inadequate anticoagulation. Mitral and tricuspid valves are more commonly involved. Echocardiography or fluoroscopy usually confirm reduced leaflet or disk motion caused by an occluding thrombus. Emergency reoperation is recommended for left-sided valve thrombosis with shock or New York Heart Association (NYHA) III or IV symptoms or cases with large thrombi (>0.8 cm2 on transoesophageal echocardiography (TOE)) but the operative mortality is up to 30%. Fibrinolysis with tPA or streptokinase may be used for left-sided valves with less severe symptoms (NYHA I and II) or smaller thrombus burdens and for patients unsuitable for reoperation; fibrinolysis is recommended for right-sided valve thrombosis. Fibrinolysis for left-sided valve thrombosis is associated with a 15 to 20% risk of systemic embolism or death.

Prosthetic valve endocarditis

Prosthetic valve endocarditis (PVE) is more common early after surgery, with an incidence up to 3% at 1 year. Mechanical valves are more commonly involved over the first year, but the incidence for mechanical and biological valves is similar thereafter. Early PVE (within 1 year) in most commonly due to nosocomial coagulase-negative staphylococci; late PVE (after 1 year) is caused by a similar range of organisms as native valve endocarditis. PVE follows a more aggressive course than native valve endocarditis with early perivalvular tissue destruction and abscess formation. TOE is important to establish the diagnosis and identify complications indicating early surgery. Medical therapy is usually ineffective in PVE. Early surgery is recommended for heart failure, abscess formation, valve dehiscence or other dysfunction, or infection with a resistant organism; surgery is also indicated for a persistent bacteraemia despite adequate antibiotic therapy or recurrent embolism from vegetations. The operative mortality for early surgery for PVE is up to 35%.

Paravalvular leak

A paravalvular leak may develop because of poor surgical technique, suture dehiscence, poor native tissue strength, and infection: PVE must always be excluded in the setting of new paravalvular leak. A small leak may cause a haemolytic anaemia due to mechanical red cell damage. Iron and folic acid supplements may be beneficial. Reoperation is indicated for heart failure, a persistent need for transfusion, or an impaired quality of life. Large leaks, particularly mitral, may cause volume overload: the development of intractable heart failure is an indication for reoperation. Catheter-based approaches may be helpful to avoid redo surgery.

Further reading

Iqbal J, et al. (2013). Optimal revascularization for complex coronary artery disease. Nat Rev Cardiol, 10, 635–47.Find this resource:

    Mohr FW, et al. (2013). Coronary artery bypass graft surgery versus percutaneous coronary intervention in patients with three-vessel disease and left main coronary disease: 5-year follow-up of the randomised, clinical SYNTAX trial. Lancet, 381(9867), 629–38.Find this resource:

      Partridge JS, et al. (2012). Frailty in the older surgical patient: a review. Age Ageing, 41, 142–7.Find this resource:

        Taggart DP (2013). Current status of arterial grafts for coronary artery bypass grafting. Ann Cardiothorac Surg, 2, 427–30.Find this resource:

          Taggart DP, et al. (2001). Effect of arterial revascularisation on survival: a systematic review of studies comparing bilateral and single internal mammary arteries. Lancet, 358(9285), 870–5.Find this resource:

            Vahanian A, et al. (2012). Guidelines on the management of valvular heart disease (version 2012) The Joint Task Force on the Management of Valvular Heart Disease of the European Society of Cardiology (ESC) and the European Association for Cardio-Thoracic Surgery (EACTS). Eur Heart J, 33, 2451–96.Find this resource:

              Vohra HA, et al. (2013). Outcomes following cardiac surgery in patients with preoperative renal dialysis. Interact Cardiovasc Thorac Surg, 18, 103–11.Find this resource:

                Wijns W, et al. (2010). Task Force on Myocardial Revascularization of the European Society of Cardiology (ESC) and the European Association for Cardio-Thoracic Surgery (EACTS); European Association for Percutaneous Cardiovascular Interventions (EAPCI), Guidelines on myocardial revascularization. Eur Heart J, 31, 2501–55.Find this resource: