From the *Department of Anesthesia and Perioperative Medicine, Evidence-Based Perioperative Clinical Outcomes Research Group (EPiCOR), London Health Sciences Centre, University of Western Ontario, London, ON, Canada; †High Impact Technology Evaluation Centre, London Health Sciences Centre, London, ON, Canada; and ‡Division of Cardiac Surgery, London Health Sciences Centre, London, ON, Canada.
Accepted for publication January 5, 2010.
Address correspondence and reprint requests to Davy C. H. Cheng, MD, Department of Anesthesia and Perioperative Medicine, London Health Sciences Centre—UH, 339 Windermere Road, C3-172, London, ON, Canada N6A 5A5. E-mail: email@example.com.
The recent publication of the Veterans Affairs Randomized On/Off Bypass (ROOBY) trial1 has raised new questions about the role of off-pump coronary artery bypass surgery (OPCAB) versus conventional coronary artery bypass surgery (CCAB) using the cardiopulmonary bypass pump. In the ROOBY trial, 2203 patients scheduled to undergo emergency or elective coronary artery bypass grafting were randomized to OPCAB or CCAB, with follow-up for death or complications at 30 days and 1 year. At 30 days, no significant differences were found between OPCAB and CCAB for death (1.6% versus 1.2%, for OPCAB versus CCAB, respectively; P = 0.47) or for the composite of death, repeat revascularization, or myocardial infarction (7.0% versus 5.6%, respectively; P = 0.19). Similarly, no differences were found for other clinically relevant outcomes including stroke, cardiac arrest, renal failure requiring dialysis, need for reintervention, and need for new mechanical support. At 1 year, all-cause death was not significantly different between groups (4.1% versus 2.9% for OPCAB versus CCAB, respectively; P = 0.15). However, at 1 year, there was a significant increase in the primary composite outcome of death, repeat revascularization, or myocardial infarction for the OPCAB group (9.9% versus 7.4%; P = 0.04). Death from cardiac causes was significantly increased for OPCAB versus CCAB at 1 year (2.7% versus 1.3%; P = 0.01). Neuropsychologic outcomes were similar between groups at 1 year follow-up, but patency was generally reduced in the OPCAB group (although assessment of patency was incomplete because only one subgroup was available for imaging assessment during the study).
We welcome the ROOBY trial1 because it is the largest randomized trial of OPCAB versus CCAB published to date and because further evidence is immediately needed to define the rightful place of OPCAB in the management of patients requiring coronary artery revascularization. However, we are concerned that certain features of the ROOBY trial design may have biased the results against OPCAB, and this could prematurely truncate continued OPCAB research if ROOBY is interpreted without considering the potential limitations of the trial. In particular, a higher than usual conversion rate (12.4% in ROOBY versus 8% in meta-analyses)2,3 from OPCAB to on-pump CCAB may have caused higher rates of death and complications in the OPCAB arm of ROOBY trial because crossovers are (properly) analyzed by intention-to-treat in the OPCAB group (ie, analyzed in the group to which they were randomized, even though they did not receive OPCAB, and instead underwent CCAB by unplanned conversion to cardiopulmonary bypass pump). In numerous studies, conversion to CCAB surgery has been shown (randomized and nonrandomized) to increase the risk of death and serious complications for patients who were initially intended to undergo OPCAB.4–9 Given the number of patients who were converted from OPCAB to CCAB in ROOBY (137 of 1104 patients intended to undergo OPCAB), if each of these patients did poorly, they may have accounted for the majority of deaths and complications that were ascribed to the OPCAB arm. For example, the number of deaths that occurred in the OPCAB arm (including those who underwent OPCAB or who were converted immediately to CCAB) was 77 of 1104 patients, which is smaller than the total number of patients who were crossed over. Although it is correct to analyze the converted patients in the OPCAB group (because OPCAB was the intended procedure, and crossovers are sometimes the unintended consequence of undergoing OPCAB and should be accounted for therein), the observation of a high crossover rate in the ROOBY trial suggests that perhaps the potential for benefit of OPCAB over CCAB was overshadowed by unplanned crossovers, and unplanned crossovers may have been directly related to the lack of surgeon experience.
Inexperience is particularly concerning for OPCAB, which has a characteristically steep learning curve. In the ROOBY trial, the primary surgeon or first assistant was a trainee in 16 of 18 centers postgraduate year 6 to 10, which is likely an important departure from the more advanced experience of surgeons in a number of other published trials. Previous studies were generally single center; often single surgeon trials involving experienced surgeons show lower crossover rates.5 Because 55.4% of OPCAB was performed by residents in ROOBY trial, this may have caused the high rate of crossover and death. Perhaps the ROOBY trial addresses the issue of lack of generalizability of the superior results of single center studies. Perhaps ROOBY also provides a timely indicator that OPCAB is not ready for prime time in low-risk patients as a routine procedure delivered largely by inexperienced surgeons. The fact that there was more incomplete revascularization in OPCAB group than in the CCAB group (17.8% versus 11.1%) also provides corroboration that lesser experience may have overshadowed the true potential of OPCAB in the hands of an experienced surgeon.
The ROOBY trial showed 12.4% conversion of OPCAB to CCAB, and a pivotal question in interpreting ROOBY would be to determine how many of these 137 (12.4% OPCAB to CCAB) contributed to the 30 days and 1 year composite outcome difference? Because ROOBY does not report the outcome of converted patients separately, we performed sensitivity analysis assuming that crossovers in ROOBY contributed to mortality in the OPCAB group according to the range of crossover death rates reported in previous studies. Exploration of the plausible range of values for increased risks of adverse outcomes in the converted patients allows us to project whether the converted patients may have contributed inordinately to adverse effects within the OPCAB group and, also, allows us to hypothesize whether these excess events may be remediable if the they could be prevented by reducing the conversions resulting from inexperience. As shown in Table 1, the conclusions in ROOBY are not robust and generalizable if these conversions could have been prevented by restricting OPCAB to more experienced surgeons. Mortality might even be significantly lower for OPCAB if 26% of the 137 converted patients in ROOBY died (contributing inordinately to the OPCAB death rate) and if these could have been prevented by more experienced surgeons (Table 1). ROOBY results should be interpreted as applicable primarily to settings with low experience and higher likelihood of conversions to on-pump.
Another important consideration when interpreting the ROOBY trial in the context of the totality of available evidence relates to the fact that ROOBY trial included only male patients and, also, excluded high-risk patients. Exploration of the baseline demographics of ROOBY suggests that the results apply to younger, lower risk patients with normal left ventricular ejection fraction (>82%). The ROOBY trial showed an actual mortality of 1.6% versus 1.2% for OPCAB versus CCAB, both of which are better than the expected 30-day risk of mortality of 1.8% to 1.9% in this population. Furthermore, meta-analysis of randomized trials (mostly of low-risk patients) showed that there was no difference in death and most clinical complications for OPCAB versus CCAB.2 On the other hand, a meta-analysis of 24,989 high-risk patients (mostly from observational studies) suggests that OPCAB may improve survival compared with CCAB (odds ratio = 0.58, 95% confidence interval = 0.49–0.68).3
Accordingly, the jury is still out on OPCAB. Further randomized controlled trial evidence in high-risk patients who are most likely to benefit from OPCAB surgery (with experienced surgeons and low risk of crossover) is eagerly awaited.3,10,11
1.Shroyer AL, Grover FL, Hattler B, et al; for the Veterans Affairs Randomized On/Off Bypass (ROOBY) Study Group. On-pump versus off-pump coronary-artery bypass surgery. N Engl J Med
2.Cheng DC, Bainbridge D, Martin JE, Novick RJ. Does off-pump coronary artery bypass reduce mortality, morbidity and resource utilization when compared to conventional coronary artery bypass? A meta-analysis of randomized trials. Anesthesiology
4.Légaré JF, Buth KJ, Hirsch GM. Conversion to on pump from OPCAB is associated with increased mortality: results from a randomized controlled trial. Eur J Cardiothorac Surg
5.Reeves RC, Ascione R, Caputo M, Angelini GD. Morbidity and mortality following acute conversion from off-pump to on-pump coronary surgery. Eur J Cardiothorac Surg
6.Landoni G, Pappalardo F, Crescenzi G, et al. The outcome of patients requiring emergency conversion from off-pump to on-pump coronary artery bypass grafting. Eur J Anaesthesiol
7.Patel NC, Patel NU, Loulmet DF, et al. Emergency conversion to cardiopulmonary bypass during attempted off-pump revascularization results in increased morbidity and mortality. J Thorac Cardiovasc Surg
8.Jin R, Hiratzka LF, Grunkemeier GL, et al. Aborted off-pump coronary artery bypass patients have much worse outcomes than on-pump or successful off-pump patients. Circulation
. 2005;112(9 suppl):I332–I337.
9.Edgerton JR, Dewey TM, Magee MJ, et al. Conversion in off-pump coronary artery bypass grafting: an analysis of predictors and outcomes. Ann Thorac Surg