Off-pump versus on-pump coronary artery bypass grafting—a surreal controversy?
- 27 Downloads
As per the findings of the SYNTAX Trial, it has been established that coronary artery bypass grafting (CABG) affords the best long-term results in patients with multivessel disease. However, it is still a point of contention whether ‘off-pump’ CABG (OPCAB) (on cardiopulmonary bypass) is better. Many studies appear to establish the superiority of one or the other. But there is still no well-planned multi-centre randomised controlled trial looking into all aspects of the issue that can, once and for all, put the controversy to rest. But as long as one is able to give the patient a good result, does it really matter whether the procedure is done off or on pump? This article looks at the available evidence with regard to various parameters comparing OPCAB and conventional CABG in an attempt to answer the question, if at all, it is a question!
KeywordsCABG Off-pump Conventional
Even after the SYNTAX Trial  indisputably established that CABG affords the best long-term results in patients with multivessel disease, a controversy that refuses to die down is that of OPCAB versus conventional CABG (CCAB) performed ‘on pump’. The last two decades have seen a see-saw battle rage between proponents of one or the other where hard evidence in favour of one is presented, only to be refuted a while later by the opposite camp, with a fresh set of scientific data. A retrospective analysis of the Society of Thoracic Surgeons (STS) Adult Cardiac Surgery Database in the USA revealed a decline in off-pump operations, currently being used in fewer than one in five patients . However, OPCAB continues to be a favoured modality in Asian countries  perhaps owing to reduction in early mortality besides the savings in costs incurred. The genesis of the debate lies in the basic need for a still, bloodless operating field so as to enable the surgeon create a coronary anastomosis that should ideally last the life of the patient. Interestingly, CABG started as an off-pump procedure, but with the introduction of cardiopulmonary bypass (CPB), it soon became a widespread operation, performed by surgeons of varying skill levels. However, the whole-body inflammatory reaction triggered by CPB led to a host of complications, apart from neurologic issues with microembolic showering from the aorta, which led to a revival of interest in OPCAB with a view to avoiding these problems.
The major trials
The initial enthusiasm for OPCAB soon gave way to reports of incomplete revascularisation, notably of the circumflex system . This led to the first few randomised controlled trials (RCT) comparing the two procedures. The randomised on/off bypass study (ROOBY) showed that at 1-year follow-up, patients undergoing OPCAB fared badly compared with their CCAB counterparts with respect to composite outcomes of death, myocardial infarction (MI), graft patency and repeat revascularisation . A rather thorny issue highlighted in this study was the surgeons’ inexperience, as patients underwent off-pump surgery by surgeons who had a minimum of 20 cases’ experience. Not entirely coincidentally, this study also revealed a 12.4% conversion rate from OPCAB to CCAB .
The Beating Heart Against Cardioplegic Arrest Studies (BHACAS 1 and 2) were early RCTs that provided a cohort of patients randomised to off-pump and on-pump surgery. Angelini et al. pooled data from these patients. The mean follow-up was 25.0 months for BHACAS 1 and 13.7 months for BHACAS 2. There was no difference in combined all-cause mortality between the two groups . The authors subsequently conducted a long-term follow-up study and found no difference in survival up to 6–8 years between on-pump and off-pump surgery . A Cochrane systematic review of RCTs of off-pump versus on-pump found an increased risk of death with off-pump in long-term follow-up studies . This finding came as a boost for the opponents of OPCAB who had been harping on the ROOBY study results, notwithstanding the sore point about the surgeons’ experience.
The Danish On-pump versus Off-pump Randomisation Study (DOORS) revealed a significantly higher rate of the primary composite outcome including all-cause mortality, repeat revascularization (RR), or nonfatal MI at 1 year and lower graft patency at 6 months following surgery in patients, who underwent off-pump CABG . However, the Coronary Artery Bypass Grafting Off- or On-Pump Revascularization Study (CORONARY) remains the largest RCT to date that recruited 4752 patients . At 1 year, the study reported no difference in primary composite of death, MI, stroke, or new renal failure requiring dialysis. The primary outcome event had occurred in 288 participants (12.1%) in the off-pump group and in 316 participants (13.3%) in the on-pump group. There was no significant difference between the rates of the individual primary outcome occurrences . At 5-year follow-up, there were no significant differences between the two groups in the rate of the composite outcome or in the rates of the components of the outcome, including repeat coronary revascularization, which was performed in 2.8% of the patients in the off-pump group and in 2.3% of the patients in the on-pump group . Two major factors providing a fillip to the off-pump camp from this landmark study were the experience of the participating surgeon (at least 2 years post-residency experience and 100 cases with either technique; contrast this with the experience laid out in the ROOBY trial), and the fact that no prior study had looked at long-term results.
Two more recent meta-analyses of RCTs on the subject focussed on short- and long-term outcomes. Deppe et al.  in a meta-analysis of RCTs on almost 16,900 patients found no difference in 30-day mortality, while Luo and colleagues  found no difference in patients after 6-month follow-up.
Completeness of revascularization and graft patency
Major concerns surrounding OPCAB relate to graft patency and completeness of revascularization. Early on, Puskas  and Magee  reported similar graft patency rates for OPCAB and on-pump CABG (ONCAB) at 30 days and 1 year. But contrary to this, the ROOBY trial demonstrated that the proportion of patients with fewer grafts than initially planned was higher in the off-pump arm than in the on-pump arm (17.8 vs. 11.1%, p < 0.01) . Similarly, the Off-Pump versus On-Pump Coronary Artery Bypass Grafting in Elderly Patients (GOPCABE) trial  revealed that fewer grafts were performed in the off-pump arm compared to on-pump. In the same vein, the CORONARY trial showed fewer grafts performed in the off-pump group and higher early revascularization rate at 30 days, but this difference petered out at 1 year [11, 12]. Palmer et al.  analysed the CABG arm of the Coronary Artery Revascularization (CARE) study, which included 1251 patients (654 on-pump CABG; 597 off-pump CABG). They found that off-pump patients received fewer grafts than on-pump patients (2.9 vs. 3.2). However, they demonstrated that the ratio of anastomoses performed to lesions was the same in both groups, indicating that the difference in the number of grafts had more to do with patient selection, i.e. patients needing fewer bypasses tended to undergo off-pump CABG.
Another, smaller sample-sized study from Japan actually found an increased number of distal anastomoses and greater patency rates (as assessed by coronary computed tomographic angiography) at 8–13 years in OPCAB patients compared with ONCAB; however, neither of these had statistical significance . The study also found that propensity-score-matched analysis revealed no difference in mid-term survival rate, Major Adverse Coronary and Cerebral Events (MACCE) free rate, graft patency rates, and number of distal anastomoses between ONCAB and OPCAB groups.
As mentioned earlier, stroke and other neurologic events are a major cause of morbidity and mortality, not to mention the increased costs. Avoidance of both aortic manipulation and use of CPB greatly reduce the risk of such complications. It stands to reason then, that OPCAB would score highly in this respect as manipulation of the aorta can easily be avoided by adopting an off-pump, anaortic approach. Surprisingly, however, the most prominent OPCAB versus CCAB trials to date failed to show a reduction in incidence of stroke with use of the former [5, 11, 18]. Deppe et al. in their subgroup meta-analysis of 13 trials found no difference in stroke rates between OPCAB and CCAB patients . However, Kowalewski et al. in an exhaustive meta-analysis that included 40 RCTs looking at the incidence of neurologic complications found OPCAB to be associated with a significant 28% reduction in odds of cerebral stroke . Kapetanakis et al.  reported that stroke rate following CCAB was 1.5 times (2.2 vs. 1.6%) that after OPCAB with partial aortic clamping and 3 times (2.2 vs. 0.8%) that after anaortic OPCAB (p = 0.01). Proximal aortic anastomotic devices such as heart string are also associated with a reduced stroke rate. A Swiss group  found that patients operated with a clampless heartstring device experienced significantly lower frequencies of stroke (0.7 vs. 2.3%; P = 0.04) and MACCE (6.7 vs. 10.8%; p = 0.001) than those undergoing partial aortic clamping, but similar to those who underwent total arterial anaortic grafting (stroke rate 0.8%; MACCE 7.9%). A recent meta-analysis (37,720 patients) found that anaortic OPCAB was associated with decreased risk of post-operative stroke, especially in patients at high stroke risk. In addition, the avoidance of CPB reduced the risk of short-term mortality, renal failure, atrial fibrillation and length of ICU stay. The operative techniques compared were anaortic OPCAB, off-pump CABG with heartstring device, OPCAB with partial clamp and CCAB . One of the present authors (OPY) has struck a note of caution with regard to jumping onto the bandwagon of anaortic OPCAB (despite the obvious reduction in stroke incidence), especially in patients with diffuse coronary lesions and other severe comorbidities .
Left ventricular dysfunction
Traditionally, this has been a major challenge for proponents of OPCAB, due to the fear of intraoperative hemodynamic instability; as a result, patients with severe left ventricular (LV) dysfunction have been under-represented in many trials. In the ROOBY trial , only 5.7% of the study population had an EF < 35% . Similarly, the DOORS trial comprised 5.3% of patients with EF < 30% . In a real-world scenario, surgeons often encounter such patients with the added burden of diffusely diseased vessels, unamenable to percutaneous coronary interventions (PCI). In fact, a study from the New York Statewide database revealed that, of 55,000 patients undergoing CABG, approximately 15% had an EF < 30% . A review of the STS National Database that included 25,667 patients undergoing elective or urgent primary CABG with an EF < 30% showed that off-pump CABG patients were older, had a higher proportion of women and had higher rates of comorbidities, including a lower estimated glomerular filtration rate (GFR), severe chronic lung disease, and preoperative arrhythmia than those undergoing on-pump surgery, resulting in a higher predicted mortality risk. Although the off-pump CABG patients in this study received fewer distal anastomoses (3.5 vs. 2.9, p < 0.001), risk adjusted outcomes such as in-hospital death, stroke, perioperative MI, MACCE, and prolonged ventilation significantly favoured off-pump CABG. Therefore, the predominant factors associated with lower in-hospital mortality after off-pump CABG in this report were the lower stroke and other neurologic event rates, transfusion requirements, and incidence of prolonged ventilation . A study from China showed that patients with preoperative EF of ≤ 35% had an increased risk of in-hospital death and incidences of post-operative low cardiac output and prolonged ventilation, but shared similar mid-term all-cause mortality and repeat revascularization after OPCAB with those having a preoperative EF of > 35% . As far back as 2011, Caputti et al. showed that CCAB actually led to a higher mortality, length of stay and complication rate, when compared with OPCAB in patients with severe LV dysfunction (EF < 20%) .
There have been no RCTs comparing outcomes of the two techniques of CABG in renal dysfunction. Studies have only included renal dysfunction as one of several comorbidities. An early Indian study showed that on-pump surgery was more deleterious to patients with non-dialysis-dependent renal insufficiency . A meta-analysis of ten retrospective studies including 14,072 patients with End-Stage Renal Disease (ESRD) on dialysis (11,310 on-pump CABG; 2762 off-pump CABG) revealed no differences in early mortality, re-exploration for bleeding, transfusions and atrial fibrillation (AF). However, the studies included in this meta-analysis were small and had a high degree of heterogeneity . However, the important issue is to look at the effects of either technique on renal parameters in those patients with underlying renal dysfunction who are not dialysis-dependent, a real-world practice scenario. A meta-analysis of nine observational studies and one RCT that included patients with non-dialysis-dependent renal dysfunction undergoing CABG (1850 CCAB; 1183 OPCAB) found that OPCAB was much more beneficial than CCAB at preventing development of acute renal failure. Early mortality was also lower following off-pump CABG . A study from South Korea found that patients with chronic kidney disease undergoing OPCAB had a lower risk of early death, prolonged ventilation and sternal wound infection. Again, the number of anastomoses was found to be lower in the OPCAB group when compared with those operated by CCAB, but this was not statistically significant. However, the use of CPB was found to be not associated with increased risk of mortality after 1 year of surgery .
Acute coronary syndrome
Patients undergoing CABG in an acute setting pose a higher surgical risk, as reported by one of the present authors , probably due to reperfusion injury or no reflow phenomenon following global ischemia due to aortic cross-clamping during CCAB. A group from Leipzig showed in 2006 that these potential complications could be avoided with CCAB on beating heart thereby avoiding the cross-clamp-associated ischemia . A study from Japan found no significant differences in mortality between techniques (OPCAB or CCAB) when applied in low, medium and high risk group patients undergoing CABG in a setting of acute coronary syndrome (ACS) . However, Davierwala, of the above-mentioned Leipzig group, averred that focused RCTs are necessary to prove the definitive benefit of OPCAB in ACS. Until then, surgeons should focus on achieving complete revascularization in such patients without compromising the purpose of performing the operation off-pump .
Comparison of morbidity and mortality in women in OPCAB Vs CPB group
n = 379
n = 133
Blood requirement (units)
2.5 ± 1.2
4.3 ± 1.4
ICU stay (h)
29.3 ± 16.4
38.3 ± 17.3
Hospital stay (days)
6.81 ± 1.6
8.05 ± 2.1
Use of CPB has been cited as a source of myocardial damage due to cross-clamp induced ischemia, followed by declamping-associated reperfusion injury. In a randomised-prospective design by Rastan et al. , 40 elective patients with normal EF were assigned to either OPCAB or CCAB surgery. Before coronary occlusion and 1, 30, 60 and 90 min after reperfusion with the left internal mammary artery (LIMA) graft, coronary sinus (CS) blood was sampled to determine intraoperative myocardial ischemia (pH, lactate, pO2) and oxidative stress (malondialdehyde—MDA). Additionally, arterial blood was analysed 4, 12 and 24 h post-operatively to determine myocardial necrosis (CK-MB, cardiac troponin I), myocardial dysfunction (NT-proBNP) and inflammation (C-reactive protein). All the enzymes studied were higher post-operatively in the on-pump group. Therefore, OPCAB was associated with less myocardial injury than CCAB. In a similar study by Chowdhury et al., it was found that the total amount of heart-type fatty acid binding protein, cardiac troponin I, and high-sensitivity C-reactive protein released (measured in venous blood samples) was significantly higher in the on-pump group; and therefore OPCAB seemed to provide better myocardial protection than on-pump surgery . A non-randomised prospective study from Greece also found that a more severe inflammatory response associated with more complications and worse outcomes occurred in patients operated by CCAB than OPCAB. Biochemical markers assessed were various interleukins, tumour necrosis factor alpha, etc. . In a meta-analysis of observational studies reported by Attaran et al. , OPCAB versus CCAB revascularization was compared in female patients. Perioperative myocardial infarction rate was significantly lower with OPCAB; however, this did not translate into a reduction in 30-day mortality and off-pump surgery did not significantly improve other cardiovascular, renal or neurological outcomes.
The current evidence derived from RCTs demonstrates comparable early and late survival for both off-pump and conventional coronary artery bypass grafting. However, patients undergoing OPCAB appear to have fewer coronary anastomoses and increased repeat revascularisation rates, although these more often than not are statistically insignificant. There are some concerns about graft patency in OPCAB, but other studies have found no difference in long-term patency rates for both arterial and vein grafts. These studies have been confounded by controversies regarding surgeon experience. OPCAB grafting seems to reduce post-operative renal dysfunction, bleeding, transfusion requirements and respiratory complications while perioperative MI rates remain similar to CCAB grafting. Importantly, high-risk patient groups have been shown to benefit from off-pump coronary surgery. In conclusion, it can be best said that what ultimately matters is the safety of the patient and a good long-term result, regardless of the technique. As stated by the present authors, it is the surgeon and not the technique that should matter , provided a good long-term, reproducible result can be obtained.
- 9.Møller CH, Penninga L, Wetterslev J, Steinbrüchel DA, Gluud C. Off-pump versus onpump coronary artery bypass grafting for ischaemic heart disease. Cochrane Database Syst Rev. 2012;3:CD007224.Google Scholar
- 26.Topkara VK, Cheema FH, Kesavaramanujam S, et al. Coronary artery bypass grafting in patients with low ejection fraction. Circulation. 2005;112:I344–50.Google Scholar
- 41.Yadava OP, Prakash A, Kundu A, Yadava M. Coronary artery bypass grafting in women: is OPCAB mandatory? Indian Heart J. 2011;63:425–8.Google Scholar
- 45.Chowdhury UK, Malik V, Yadav R, et al. Myocardial injury in coronary artery bypass grafting: on-pump versus off pump comparison by measuring high-sensitivity Creactive protein, cardiac troponin I, heart-type fatty acid-binding protein, creatine kinase-MB, and myoglobin release. J Thorac Cardiovasc Surg. 2008;135:1110–9.CrossRefGoogle Scholar