The term minimally invasive coronary artery bypass grafting (CABG) is not well defined. According to one definition, avoidance of cardiopulmonary bypass (CPB) is considered essential in decreasing the morbidity associated with conventional CABG.¹ Other authors consider median sternotomy as a potential source for morbidity, due to increased risk of deep sternal wound infection and mediastinitis and delayed return to daily activities.² Accordingly, a number of surgical strategies have evolved to avoid the need for extracorporeal circulation and minimize surgical access. Furthermore, eluding aortic manipulation and complete arterial revascularization are operative strategies that focus on improving short and long-term results. Likewise, it was widely recognized that conventional harvesting techniques for bypass grafts are often associated with wound-healing problems, especially in diabetic patients. As a consequence, endoscopic harvesting techniques for both, venous and radial artery grafts have been developed.

Conventional CABG has been performed with CPB under cardioplegic arrest for decades. An empty, arrested heart, a bloodless surgical field, and an excellent exposure of all epicardial vessels have been considered to be the key factors for the success of this procedure. Excellent results and constantly declining mortality despite the ever-increasing risk profile of patients (Davierwala) have made standard CABG the “bread and butter” of our profession.³ Anecdotal reports on the deleterious effects of CPB and systematic reports that examined the pathophysiology of extracorporeal circulation began questioning the dogma, “the pump is your friend.” CPB is associated with (1) a systemic inflammatory response, (2) release of cytokines, (3) activation of the clotting cascade, (4) metabolic disturbances, and (5) microembolization among numerous other adverse effects. Although well tolerated by most patients, these effects alone or in combination may cause substantial morbidity, thus negatively affecting the results of the procedure. With an ever-aging population and increasing comorbidity, surgeons all over the world sought to further minimize the risk of CABG, and it seemed logical to question the role of CPB in CABG.

The evolution of off-pump coronary artery bypass (OPCAB) grafting is closely linked to the development of stabilizers that became available in the early 1990s. Initially, only pressure stabilizers were developed, but it soon became obvious that exposure of vessels on the posterolateral and inferior walls of the heart would require additional means of support. Hence, OPCAB gained greater popularity when vacuum-assisted stabilizers were introduced by the Utrecht group, which facilitated localized myocardial immobilization for all territories. Additionally, it was also recognized that OPCAB requires a team approach between the surgeon and anesthesiologist so as to prevent sudden hemodynamic changes during the procedure and to manage the same when they do occur.

Anesthesia Requirements

OPCAB is performed under general anesthesia in most centers. Incidental reports in literature demonstrate that OPCAB can also be performed under high epidural anesthesia in an awake patient breathing spontaneously.⁴ Standard monitoring is used. In addition, some centers prefer continuous cardiac output measurement using the PICCO or similar methods.⁵ A Swan-Ganz catheter is not always beneficial and contrarily can potentially cause arrhythmias when the heart is manipulated during the procedure. It is of utmost importance that the patient is kept warm at all times during the operation. Temperature management includes the use of a warming blanket and warm infusions, and maintaining higher the room temperatures. Volume management is essential because it is generally the preferred means to counterbalance hemodynamic changes. Exposure of the posterior wall results in severe hemodynamic compromise due to some degree of right ventricular compression under the right hemisternum, but can be treated adequately by increasing venous return by tilting the operating room table to the right in a Trendelenburg position. Excessive infusion of fluids should be avoided, especially in patients with end-stage renal disease (ESRD), as hemofiltration is not possible during surgery due to the absence of CPB. The use of inotropes should be preferably reserved only in patients with severe hemodynamic alterations, because they invariably increase the heart rate and make grafting more difficult. A review of human factors associated with manual control and tracking revealed that a human operator can, at his or her best, track a three-dimensional motion (such as the beating heart) only up to a frequency of 1 Hz, which corresponds to a heart rate of 60 beats per minute.⁶ Higher frequencies cannot be tracked; therefore the preferred heart rate should be maintained between 50 and 70 beats per minute to simplify anastomotic suturing. In case of atrial fibrillation (AF), pharmacologic reduction in the heart rate and temporary ventricular pacing using epicardial pacing wires may facilitate grafting. The use of a Cell Saver is recommended to minimize the risk of blood transfusion, which is rarely necessary. If less than 500 mL is accumulated, the blood is usually discarded in the Cell Saver.

Surgical Technique

Single or bilateral internal thoracic arteries (ITA) are harvested following a standard median sternotomy. The patient is heparinized (150-200 units/kg, ie, about half of CPB dosage). An activated clotting time (ACT) greater than 300 seconds is maintained. It may be necessary to open the right pleura in patients with severely dilated hearts, although some surgeons do it routinely to give the heart enough space during manipulation for exposure of the posterolateral wall. It avoids excessive compression of the right ventricle. Following pericardiotomy, it is advisable to hitch the left side of the pericardium onto the left sternal edge. Thereafter, the two edges of the sternum are spread out with the sternal retractor. This helps not only in rightward rotation of the heart, but also elevates the heart to a certain degree, which additionally facilitates exposure of vessels on the posterolateral wall. Subsequently, deep pericardial stay sutures are placed to further aid in exposure and manipulation of the heart. Numerous methods have been proposed for placing these sutures. Ideally, two sutures are placed deep into the pericardium (deeper than the level of the atrioventricular groove), one just medial to the inferior vena cava and the other just inferior to the left inferior pulmonary vein (Fig. 23-1). To avoid damage to the myocardium due to shearing, the sutures should be covered by plastic tubing; sponges may be used alternatively. Placement of the stay sutures should be done slowly because abrupt changes in positioning of the heart may cause undesired hemodynamic alterations or arrhythmias.

The sequence in which the grafting should be performed depends largely on the coronary anatomy and the grafting techniques used. The left anterior descending (LAD) artery is considered to be the most important target vessel and can be easily grafted without too much manipulation of the heart. Therefore, revascularization should start with an ITA graft to the LAD in most cases. A slight pull on the stay sutures easily brings the LAD into the operative field. The stabilizer is placed at the desired site of the anastomosis so as to ensure adequate distance between the pods of the stabilizer and the vessel for suturing. The anterior surface of the vessel is exposed by dissecting it free from the overlying tissue. If the vessel is covered by excessive fat or muscle, low-energy cautery and clipping of epicardial veins may help to minimize bleeding from the surrounding tissue. Mechanical compression of large diagonal branches should be avoided. Vacuum stabilizers should be locked only after vacuum is applied and the stabilizer firmly holds on to the surface. Only minimal pressure will then be required to immobilize the heart. The greater the pressure exerted by the stabilizer, the greater the force with which the heart beats against it, resulting in increasing wall motion and hence the difficulty in anastomosis. Temporary occlusion of the target vessel can be achieved in many ways. Complete or partially encircling 4-0 polypropylene sutures with or without pledgets or silastic tapes are commonly used widely. The occlusion tapes or sutures should be placed at a distance of at least 5 to 10 mm from the anastomosis to avoid compression and distortion of the target vessel at the level of the anastomosis, thus allowing for safe and comfortable suturing. The suture needs to be placed deep enough in the tissue to avoid damage of the target vessel. Injury to the accompanying coronary vein should be avoided to prevent bleeding at and around the anastomotic site. Distal coronary artery occlusion should be avoided and is rarely necessary even in occluded vessels with strong backflow. A shunt can be used in the latter case. Care must be taken to avoid occluding vessels in stented areas because the occlusion sutures/tapes may bend or kink an implanted stent. The incision is usually made before the occlusion suture is tightened to ensure that the vessel is full, as it will minimize the risk of back wall injury while opening the vessel. The suture then is tightened gently until the bleeding stops. This may not be necessary in totally occluded vessels with minimal coronary blood flow. A CO₂ blower-mister is used to maintain a clear field of vision during the anastomosis by blowing away the blood flowing retrograde from septal branches or the distal coronary artery. The CO₂ flow rate should not exceed 5 L/min. Excessive blowing can produce dissection or injury to the intima of both, the graft and the coronary artery, or cause air embolism. The use of coronary artery shunts is controversial because they also may cause endothelial damage. If shunts are used, care must be taken to ensure atraumatic placement. Another option is the use of a transparent reverse thermosensitive gel (LeGoo™, Pleuromed Inc., Woburn, MA) injected at the anastomotic site after incising the coronary artery, which temporarily blocks the blood flow allowing an almost bloodfree anastomotic site. The anastomosis can then be performed without a blower-mister, shunt, or occlusion snares. The gel completely dissolves with time (after approximately 15 minutes) or with the local application of a cold sponge or gauze.^7,8

The circumflex artery and its branches are usually grafted next, especially when a composite Y/T-grafting technique is used. Grafting the distal circumflex artery is considered to be the most challenging during OPCAB because exposure may be difficult. In patients with cardiomegaly and more importantly in those with right ventricular dilatation and dysfunction, it may be necessary to open the right pleura and divide the right-sided pericardium at the diaphragm up to the level of the inferior vena cava. This allows displacement of the heart under the right sternal edge into the right pleural cavity, thus preventing the compression of the right ventricle. Thereafter, the right coronary artery (RCA) and its branches are grafted. If the RCA is the dominant vessel and has a stenosis less than 80%, it may be necessary to use a shunt because loss of blood flow in the atrioventricular (AV) nodal artery during occlusion of the RCA can cause acute total AV block. It is therefore advisable to place and connect a temporary pacing wire before occluding such vessels.

To maximize both the short- and long-term benefits of an OPCAB procedure, composite arterial grafting is preferred. This will obviate the need for aortic clamping, which is another source for emboli and an independent predictor of stroke.⁹ In fact, some authors believe that avoiding aortic manipulation is the key factor in reducing the stroke risk following CABG below that after percutaneous coronary intervention (PCI).

If aortocoronary vein or radial grafts are used, the proximal or distal anastomosis can be performed first. However, in patients with critical ischemia or hemodynamic instability, it would be beneficial to perform the proximal anastomosis initially, as it can be performed without manipulating the heart and the myocardium is revascularized as soon as the distal anastomosis is completed. It is of utmost importance to reduce the systemic blood pressure during partial clamping of the aorta to minimize the risk of embolization and aortic dissection. This can be achieved pharmacologically or mechanically by briefly compressing the inferior vena cava. This maneuver should also be repeated prior to declamping. Intraoperative graft patency control using transient-time Doppler or other imaging techniques is recommended. If the operative field is dry, heparin usually is not completely antagonized. Postoperatively, aspirin is administered on the day of surgery, provided the patient does not bleed to muchion.

Special Situations

In patients with unstable angina, acute cardiogenic shock, or low ejection fraction (EF < 20%), the preoperative implantation of an intra-aortic balloon pump (IABP) may be useful. Alternatively, these patients can be operated on CPB without cardioplegic arrest (on-pump beating heart) to combine the advantages of preserved native coronary blood flow, decompression of the heart, and adequate organ perfusion.^13,14 Patients with AF have an irregular contraction pattern that may distract the surgeon while suturing (regular-motion patterns allow the development of coping strategies such as the “wait and see” strategy that are less effective when motion is unpredictable⁵). It may, therefore, be helpful to reduce the heart rate pharmacologically and temporarily pace the patient in a VVI mode. If AF is paroxysmal, epicardial ablation may be used for pulmonary vein isolation on the beating heart before grafting.

Results

Bakaeen et al reported recently that the number of OPCABs performed in the United States peaked in 2002 (23%) and 2008 (21%), followed by a steady decline to 17% in 2012. The last 5 years has witnessed a reduction in OPCAB rates not only in intermediate-volume centers performing between 50 and 200 cases a year, but also in high-volume centers performing >200 cases a year. They further reported that 84% of centers performed fewer than 50 off-pump cases per year, 34% of surgeons performed no off-pump operations, and 86% of surgeons performed fewer than 20 off-pump cases per year.¹⁷ However, in some countries, especially, those with limited medical and financial resources, OPCAB still accounts for almost 80% of CABG procedures. Some units almost exclusively perform OPCAB with no patient selection. One of the major reasons for this decline is the results of several randomized controlled trials that have been unfavorable for OPCAB. The ROOBY trial reported a worse composite outcome of death and complications (9.9 vs 7.4%, p = .04), higher rate of incomplete revascularization (17.9 vs 11.1%; p < .0001), and a lower graft patency rate (82.6 vs 87.8%, p < .01) for OPCAB than on-pump CABG at 1 year. The Danish off-pump versus on-pump randomization study (DOORS) which included 900 patients > 70 years of age found no difference in the compound clinical endpoints between the two techniques at 30 days and 6 months.¹⁶ However, the DOORS study group recently published a subanalysis, which revealed significantly inferior graft patency rates after OPCAB than after on-pump CABG at 6 months (79 vs 86%; p = .01).¹⁷ Nevertheless, both these trials were not sufficiently powered to determine clinically relevant differences between the two techniques of surgical revascularization with respect to death, myocardial infarction, stroke, or renal failure. Sergeant and colleagues as well as Puskas and coworkers have pointed out that large cohorts of patients would be required to reveal a significant statistical difference in these outcomes between the two operative techniques.^18,19 Hence, the CORONARY trial was conducted, which randomized 4752 patients to OPCAB (2375) and on-pump CABG (2377). No differences in composite outcomes were found between groups at 30 days, 30 days to 1 year, and at 1 year after surgery.²⁰ Similarly, the German Off-Pump Coronary Artery Bypass Grafting in Elderly Patients (GOPCABE) study, which involved 2539 patients > 75 years of age randomized to OPCAB or on-pump CABG, showed no difference in clinical composite or individual components of death, stroke, MI, or renal replacement therapy at 30 days and 12 months between the two groups.²¹ However, patients undergoing OPCAB required significantly more repeat revascularization at 30 days, which nonetheless evened out at 12 months. This finding was keeping in lieu with those of Khan et al, who reported reduced graft patency in OPCAB patients three months postoperatively.²² Contrary to this, other studies like the SMART and Prague IV study that provide angiographic data on graft patency reveal equal patency rates for on- and off-pump bypass surgery.^23-25 This controversy concerning OPCAB has remained unresolved ever since the first reports on OPCAB regarding higher rates of incomplete revascularization^26-29 and lower graft patency²¹ were published.

OPCAB does play an important role in stroke prevention, especially in patients at higher risk. A meta-analysis that involved all major randomized controlled trials comparing OPCAB with on-pump CABG between 2011 and 2010 revealed a significant 30% reduction in the occurrence of postoperative stroke with OPCAB [risk ratio (RR): 0.70]. There was no significant difference in mortality (RR: 0.90) or myocardial infarction (pooled RR: 0.89).³⁰ There is growing evidence that neurocognitive outcome is better and stroke rate is lower after OPCAB.^31-33 Several studies have proven CPB to be an independent predictor of adverse neurologic outcomes.³⁴ The embolic burden that is measured by transit-time Doppler of the medial cerebral artery is reduced significantly during OPCAB.^35-37 Neurological complications can be further reduced during OPCAB by performing the operation without aortic manipulation. In a recent retrospective analysis of 12,079 CABG patients, Moss and colleagues reported that the ratio of observed to expected stroke rate increased as the degree of aortic manipulation increased, from 0.48 in the no-touch group, to 0.61 in the clampless facilitating device group, and to 0.95 in the clamp group. Aortic clamping was independently associated with an increase in postoperative stroke compared with a no-touch technique (adjusted odds ratio (OR), 2.50; p < .01).³⁸ OPCAB has also been associated with a reduced risk of acute renal failure,^39-41 especially in high-risk patients with preoperative renal insufficiency.^42,43 A meta-analysis of 22 randomized studies involving 4819 patients affirmed these findings. OPCAB was associated with a 40% lower odds of postoperative acute kidney injury (OR 0.60; 95% confidence interval (CI) 0.43 to 0.84; p = .003) and a nonsignificant 33% lower odds for dialysis requirement (OR 0.67; 95% CI 0.40 to 1.12; p = .12) than on-pump CABG.⁴⁴ The kidney function substudy of the CORONARY trial, which enrolled 2932 patients, revealed that OPCAB reduced the risk of acute kidney injury when compared to on-pump CABG (17.5 vs 20.8%, p = .01); however, there was no significant difference between the two groups in the loss of kidney function at 1 year (17.1 vs 15.3%, p = .23).⁴⁵

Furthermore, the incidence of postoperative AF^43,46 and blood levels of biochemical markers for myocardial injury (eg, creatinine kinase and troponin) are also reduced after OPCAB.^47-49 Blood loss is less, and transfusion rate is reduced.¹⁹ Overall, OPCAB reduces hospital costs by 15 to 35%,^50,51 possibly owing to a decrease in the length of hospital stay and resource utilization.⁵⁰

Most studies favor OPCAB with respect to operative and short-term mortality. A retrospective analysis of 17,969 OPCAB patients (8.8% of total) in the Society of Thoracic Surgeons database for the years 1999 and 2000 identified a significant survival advantage for OPCAB compared to on-pump CABG, as was demonstrated by risk-adjusted multivariate logistic regression analysis (OR 0.76, 95% CI 0.68 to 0.84) and conditional logistic regression of propensity-matched groups (OR 0.83, 95% CI 0.73 to 0.96).⁵² Similar results have been reported from another multicenter analysis by Mack and colleagues, which included 7283 patients. Following propensity score matching and multivariate regression analysis, the use of CPB was identified as an independent predictor of mortality (OR 2.08, 95% CI 1.52 to 2.83, p < .001).⁵³ OPCAB seems to offer a survival benefit especially in high-risk populations (eg, the elderly, those with EF < 30%, and obese patients).^19,50,54,55 The International Society for Minimally Invasive Cardiac Surgery Consensus Group, in a thorough meta-analysis, published that OPCAB reduces mortality and length of stay and the incidence of postoperative myocardial infarction (MI), renal failure, AF, and transfusion rate in mixed-risk and high-risk patients¹⁹ (Fig. 23-2). Few reports on mid-term survival showed similar outcomes for OPCAB and on-pump CABG at 2 and 4 years.^26,56

There are limited reports comparing the long-term outcomes of OPCAB and on-pump CABG. Three major publications in the last 5 years have produced different results. The long-term follow-up of patients enrolled in the Beating Heart Against Cardioplegic Arrest Study (BHACAS 1 and 2) showed a similar likelihood of graft occlusion between OPCAB (10.6%) and on-pump CABG (11.0%) groups (OR: 1.00; p > .99). Similarly, no differences were noted in the hazard of death (hazard ratio (HR): 1.24) or major adverse cardiac-related events or death (HR: 0.84) between the two groups.⁵⁷ Contrarily, a large propensity-matched analysis of 5203 patients showed a similar risk of death at 30 days (OR: 0.70; p = .31) and up to 1 year (HR: 1.11; p = .62) for OPCAB and on-pump CABG, but a higher risk of death for the former at a median follow-up duration of 6.4 years (HR: 1.43; p < .0001).⁵⁸ A recent meta-analysis of 32 studies comparing the long-term outcomes of OPCAB and on-pump CABG concluded that OPCAB has similar mid-term mortality and morbidity to on-pump CABG in a low-risk population. When observational studies were excluded a comparable long-term mortality was observed between the two surgical revascularization techniques.⁵⁹ To summarize, OPCAB is a technically challenging operation, which requires a prolonged learning curve compared to conventional bypass graft surgery. Some surgeons perform OPCAB almost exclusively, whereas others never use it at all. OPCAB may not be the best option for every patient or for all cardiac surgeons. It is, however, an important alternative and must be mastered with the same technical precision as conventional CABG.⁶⁰ True comparisons between OPCAB and on-pump CABG are extremely difficult even by randomized controlled trials as there are many factors such as quality of the coronary vessels anastomosed, type of conduits used, grafting techniques (sequential, composite total arterial, venous grafts, etc), and competency and experience of the surgeons to perform OPCAB that cannot be accounted for. All the same, we have to wait for the results of long-term follow-up of large-scale, multi-institutional RCTs.

One of the goals of minimally invasive cardiac surgery has been to avoid sternotomy to reduce the amount of surgical trauma and avoid wound complications. Therefore, coronary artery bypass techniques on the beating heart without sternotomy were developed. The operation was termed minimally invasive direct coronary artery bypass (MIDCAB), and since its introduction in the mid-1990s,^61-65 it has found a widespread application. In some centers, MIDCAB is the preferred method of surgical revascularization for isolated LAD disease. In addition, MIDCAB is a valuable alternative to standard CABG or OPCAB in selected high-risk patients with multivessel disease and extensive comorbidity, who are at a prohibitively high risk for sternotomy and CPB.

Anesthesia

Standard monitoring techniques are applied, and temperature management is as described previously for OPCAB procedures. Single-lung ventilation is achieved by using a double-lumen tube or bronchus blocker to provide selective right lung ventilation. Short-acting anesthetics are used to allow early extubation.

Surgical Technique

Standard MIDCAB is usually performed through a left anterolateral minithoracotomy with the patient in a 10 to 30° right lateral position. Following a 5 to 6 cm skin incision at the level of the fifth intercostal space or the inframammary fold in females, the pectoralis muscle is displaced bluntly with minimal division of its muscle fibers (muscle-sparing approach). This decreases the likelihood of developing a lung hernia that has been sometimes reported with this approach. The chest is then usually entered one intercostal space higher than the actual incision. Excessive rib spreading must be avoided at all times to prevent dislocation or fracture. Excision of a rib is almost never necessary. LITA harvest is usually performed under direct vision, but endoscopic harvest utilizing a harmonic scalpel or telemanipulation systems has also been reported.^65-67

The intrathoracic fascia is divided to facilitate LITA harvest, which can be performed in a pedicled or skeletonized manner, from the fifth intercostal space to the origin of the subclavian artery. Using a sceletonized technique offers the advantage of gaining more graft length, but is, of course, more time-consuming. Additional length can also be gained by dividing the mammary vein at its junction with the subclavian vein. Side branches are clipped or cauterized based on the preference of the surgeon. Heparin is administered prior to distal transsection of the graft. As in a sternotomy OPCAB, the ACT is maintained at a level greater than 300 seconds. The pericardium should be opened at a point approximately corresponding to the course of the LAD and extended to the groove between the aorta and the pulmonary artery. This will facilitate location of the target vessel in the presence of excessive epicardial fat or an intramuscular course. The target vessel is identified. To enhance exposure, one or two pericardial stay sutures may be used to position the heart. It is of utmost importance to have the anastomotic site in direct vision. Standard reusable pressure stabilizers are then used to immobilize the target vessel (Fig. 23-3). Vacuum stabilizers in general are too bulky and not required for a single anastomosis to the LAD. Proximal LAD occlusion is performed using a 4-0 pledgeted suture or vessel loops. Ischemic preconditioning is not helpful, but some surgeons use it to feel more comfortable knowing that ischemia is well tolerated. The use of shunts is rarely necessary, but they may be used based on surgeon preference, presence of significant retrograde coronary flow or evidence of ischemia or hemodynamic instability. Distal occlusion should be avoided whenever possible (99% of patients). A blower-mister is used in all cases to achieve a bloodless field. The anastomosis then is performed in a standard manner using a 7-0 or 8-0 polypropylene running suture. Finally the LITA pedicle can be fixed to the epicardial tissue. Graft patency assessment is performed routinely using transit-time Doppler flow measurements. A single chest tube is inserted into the left pleural space, and intercostal nerve blockade is applied using local anesthetics. Before closing the chest with one or two strong rib sutures, the single-lung ventilation is stopped and the left lung is inflated under direct vision of the surgeon to prevent it from causing an avulsion of the LITA and also confirm that the LITA lies perfectly without undue tension. This is facilitated by holding the LITA close to the mediastinum using the suction tip or long forceps. Extubation can be performed in the operating room or a few hours postoperatively. Antiplatelet therapy is usually administered on the day of surgery.

Results

Many papers have reported excellent results with the use of this approach. Immediate angiographic patency rates range between 94 and 98% and are thus similar to those following conventional CABG.^64,68 At 6 months, 94% patency has been reported.⁶⁹ The mid-term results published recently by Reser et al demonstrated an overall survival of 92.4 ± 0.2% and MACCE-free survival of 96.1 ± 1.7% at 24 months.⁷⁰ In our own series of 1768 patients, who underwent MIDCAB from 1996 to 2009, in-hospital mortality was 0.8% (predicted mortality by Euroscore 3.8%), and stroke rate was 0.4%. Conversion to sternotomy was necessary in 1.7%. A total of 712 patients underwent routine postoperative angiogram demonstrating a 95.5% early patency rate. Short-term target-vessel reintervention was required in 3.3% of patients. At 6-month follow-up, graft patency was 95.2% (n = 423). The 5- and 10-year survival was 88.3% (95% CI 86.6 to 89.9%) and 76.6% (95% CI 73.5 to 78.7%) (Fig. 23-4A). The corresponding freedom from MACCE and angina was 85.3% (95% CI 83.5 to 87.1%) and 70.9% (95% CI 68.1 to 73.7%) (see Fig. 23-4B).⁷¹ These results are comparable to those of other groups.^69,72,73

In-hospital mortality after MIDCAB is <1% and compares favorably with 1.4% mortality following off-pump single-vessel bypass and 3.6% mortality after on-pump single-vessel bypass reported in the registry of the German Society for Thoracic and Cardiovascular Surgery.⁷⁴ It is also lower than the 2.4% mortality reported from the STS database.⁷⁵ A propensity score adjusted comparison of MIDCAB versus full sternotomy revealed similar operative mortality, late survival, and need for repeat revascularization at a mean follow-up of 6.2 years, but a lower rate of surgical site infection (2.8 vs 0.7%; p = .04).⁷⁶ Another smaller study from the UK also found MIDCAB comparable to OPCAB with regard to operative mortality, MI, stroke, reinterventions, and ICU stay. However, hospital stay was significantly lower for MIDCAB patients.⁷⁷ In general, rates of perioperative major complications such as MI and the need for target-vessel reintervention are low and comparable with standard CABG. However, due to anaortic surgery perioperative stroke is rare and considerably lower compared to conventional techniques. A few randomized trials comparing the LIMA-LAD with the MIDCAB approach to bare metal stenting for isolated proximal LAD disease have demonstrated better early patency and superior freedom from target-vessel reintervention and angina up to 5 years of follow-up in patients undergoing MIDCAB.^78-83 The 10-year results of the randomized trial conducted by our group comparing MIDCAB with bare metal stents for proximal LAD disease showed no differences in the primary composite endpoint of death, MI, and reintervention (47 vs 36%; p = .12) and hard endpoints (death and infarction) between the two revascularization modalities. However, a higher target vessel revascularization rate in the PCI group (34 vs 11%; p < .01) was observed.⁸⁴ Similar results were also reported when drug-eluting stents were compared to MIDCAB.⁸⁵ These results have been further confirmed not only by meta-analyses^86,87 but also propensity-matched data from large registries.⁸⁸

Some studies have pointed out that the lateral approach is associated with more pain than sternotomy chiefly due to excessive rib spreading required for visualizing the LITA during harvest.⁶⁵ Rib dislocation or fracture has been reported infrequently with this approach. Direct harvesting of the LITA is regarded as being technically challenging and has been one of the main arguments used by surgeons to disregard MIDCAB as the primary operation for patients who require surgical revascularization of the LAD. Limited working space and incomplete vision are blamed for insufficient graft length, incomplete mobilization, and the occasional reports of LITA or subclavian vein injury.