BEST EVIDENCE TOPIC – ADULT CARDIAC
Interactive CardioVascular and Thoracic Surgery 19 (2014) 149–159 doi:10.1093/icvts/ivu075 Advance Access publication 21 March 2014
Does off-pump coronary artery bypass graft surgery have a beneficial effect on long-term mortality and morbidity compared with on-pump coronary artery bypass graft surgery? Umar A.R. Chaudhry, Christopher Rao, Leanne Harling and Thanos Athanasiou* Department of Surgery and Cancer, St Mary’s Hospital, Imperial College London, London, UK * Corresponding author. Department of Surgery and Cancer, Imperial College London, 10th Floor, QEQM Building, St Mary’s Hospital Campus, South Wharf Road, London W2 1NY, UK. Tel: +44-20-33127630; fax: +44-20-33126309; e-mail: [email protected] (T. Athanasiou). Received 20 December 2013; received in revised form 16 February 2014; accepted 27 February 2014
Abstract A best evidence topic in cardiac surgery was written according to a structured protocol. The question addressed was whether off-pump coronary artery bypass grafting (CABG) surgery offered superior long-term outcomes compared with on-pump CABG surgery. Best evidence papers were considered to be those that had a follow-up period of ≥5 years, had >50 patients in either cohort, did not utilize concomitant interventions nor comprised low-risk, high-risk or sub-population groups. Where potential duplicate data sets from the same institution were likely, the more credible and recently published study was included. Two hundred and fifty-six papers were found as a result of the reported search, of which 16 represented the best evidence to answer the clinical question. The authors, journal, date and country of publication, patient group studied, study type, relevant outcomes and results of these papers are tabulated. The 16 studies comprised 4 prospective randomized controlled trials (RCTs), with the remaining 12 retrospective, of which 8 were propensity-score matched. All 4 RCTs contained fewer than 450 participants. Two studies concluded with a survival advantage towards on-pump CABG: one, a large registry-based study, the Veterans Affairs, with >25 000 patients, and another, a propensity-matched retrospective study involving almost 8000 patients. The remaining 14 studies all provided evidence to suggest comparable long-term survival. In addition, all other long-term outcomes mentioned within these studies including angina recurrence, myocardial infarction heart failure, need for revascularization, stroke, graft patency, cognitive and quality of life showed similar results between the two groups. We conclude that off-pump CABG surgery may have similar or slightly reduced long-term survival compared with on-pump CABG surgery. Other long-term indicators such as cardiovascular or cerebrovascular events or neuro-psychological outcomes were similar between the two groups. Despite these conclusions, the evidence is limited by substantial variability in patient selection and study methods. The CORONARY (coronary artery bypass surgery off- or on-pump revascularization study) trial recently presented results, which showed no significant differences in composite outcomes at 1 year; it will be interesting to observe whether these comparable outcomes are maintained for a much longer time frame.
INTRODUCTION A best evidence topic was constructed according to a structured protocol. This is fully described in the ICVTS [1].
different studies present early outcomes but little data comparing the two techniques are provided regarding long-term outcomes, particularly survival. A colleague suggests that off-pump CABG surgery may be more beneficial in the long-term as it negates the effects of cardiopulmonary bypass (CPB). You are unconvinced and resolve to check the literature yourself.
Three-part question In [ patients undergoing coronary artery bypass grafting (CABG) surgery], is (the off-pump or on-pump technique) superior in terms of [long-term survival and other late outcomes]?
Clinical scenario You are at an international conference discussing the benefits of off-pump compared with on-pump CABG surgery. Several
Search strategy Medline from 1950 to December 2013 was searched using PubMed interface using the following terms: (‘coronary artery bypass’*) AND (‘coronary artery bypass, off-pump’*, ‘off-pump’, ‘OPCAB’, ‘beating heart’) AND (‘cardiopulmonary bypass’*, ‘conventional coronary artery bypass’, ‘on-pump’, ‘ONCAB’) AND (‘long-term’ OR ‘late’) AND (‘survival’, ‘mortality’, ‘death’, ‘follow-up’, ‘outcomes’).
© The Author 2014. Published by Oxford University Press on behalf of the European Association for Cardio-Thoracic Surgery. All rights reserved.
BEST EVIDENCE TOPIC
Keywords: Coronary artery bypass • Off-pump • Cardiopulmonary bypass • Long-term • Outcomes
U.A.R. Chaudhry et al. / Interactive CardioVascular and Thoracic Surgery
150
Table 1: Best evidence papers Author, date, journal and country Study type (level of evidence)
Patient group
Outcomes
Key results
Comments
Angelini et al. (2009), J Thorac Cardiovasc Surg, UK [2]
401 participants from two randomized trials (BHACAS 1 and 2) at the Bristol Heart Institute, UK, were recruited overall between March 1997 and November 1999: 200 off-pump 201 on-pump
Survival
No difference between off-pump vs on-pump CABG groups: [HR = 1.24; 95% CI: 0.72–2.15, P = 0.44]. Overall, 29 deaths in off-pump vs 23 deaths in on-pump category
Conclusions:
No difference between off-pump vs on-pump CABG groups: [HR = 0.77; 95% CI: 0.51–1.18, P = 0.24]. Overall, 39 MACEs in off-pump (3 revascularization, 8 MI, 28 angina) vs 49 MACEs in on-pump (3 revascularization, 8 MI, 38 angina) category
Limitations:
Prospective, randomized, controlled trial (level 2 evidence)
Survival-free from MACEs Mean follow-up period for off-pump and on-pump survival was 6.3 years and 6.4 years, respectively, and survival-free from MACEs or death was 5.4 and 5.2 years, respectively BHACAS 2 trial (September 1998 to November 1999) held later excluded patients with: LVEF 0.99]
QoL
No difference between off-pump vs on-pump CABG groups across a range of domains
Survival (matched patients)
Off-pump vs on-pump all-cause mortality:
Survival and MACEs were compared using Cox regression analysis Patient follow-up was through a national database, and by annual questionnaires or via family practitioners for MACEs
No difference between off-pump vs on-pump CABG groups: [HR = 0.84; 95% CI: 0.58–1.24, P = 0.39]
Long-term survival and MACE outcomes with off-pump CABG are similar to those with on-pump CABG
At 5 years: 14.5 vs 13.5%; (RR = 1.08; 95% CI: 1.02– 1.15) At 10 years: 25.2 vs 23.6%; (RR = 1.07; 95% CI: 1.03– 1.12). Overall HR = 1.06 (95% CI: 1.00–1.13; P = 0.036)
Conclusions: On-pump CABG may be associated with increased long-term survival Limitations: Retrospective data VA population—patients were more likely to be male and younger Conversion to on-pump not recorded. Analysis was not on an intention-to-treat basis Long period of recruitment with unknown levels of surgical experience at each centre No details as to patient selection to either cohort were provided
Demographics: after adjustment, no significant differences in preoperative risk factors were found, but off-pump
Continued
U.A.R. Chaudhry et al. / Interactive CardioVascular and Thoracic Surgery
151
Table 1: (Continued) Author, date, journal and country Study type (level of evidence)
Patient group
Outcomes
Key results
Comments
Survival and 10-year all-cause mortality (crude and multivariate model for unmatched cohorts; propensity-matched cohort)
Unmatched: Off-pump CABG associated with improved crude all-cause mortality but not following adjustment. Crude: HR = 0.92 (95% CI: 0.85–0.98) Multivariable: HR = 0.99 (95% CI: 0.92–1.06)
Conclusions:
patients had fewer distal anastomoses (2.66 vs 3.18, P < 0.0001) Cox proportional hazard models were used to estimate an HR for survival
Retrospective cohort study (level 3 evidence)
50 676 patients who underwent CABG in Sweden between 1998 and 2008 using the SWEDEHEART registry: 3337 off-pump 47 339 on-pump 1:1 propensity-score-matched cohort using logistic regression
Overall survival at 10 years: off-pump 73% and on-pump 71% (P = 0.017)
The mean follow-up period for the entire cohort was 7.1 years Excluded patients comprised those who had previous cardiac surgery, concomitant procedures or an emergency surgery Demographics: unmatched patients had significant differences in all baseline and perioperative characteristics except for levels for COPD and heart failure and use of internal mammary use
Matched: Similar overall long-term survival [HR = 1.02 (95% CI: 0.91–1.16)] and survival at 10 years (73 vs 72%; P = 0.56) All-cause mortality or rehospitalization for MI, heart failure or stroke (composite outcome)
Cox proportional hazards regression was used to estimate the HR. Multivariate adjustments were made to crude associations between off-pump and on-pump CABG
Unmatched: Off-pump CABG associated with improved composite outcome but not following adjustment. Crude: HR = 0.94 (95% CI: 0.88–0.99) Multivariable: HR = 0.99 (95% CI: 0.94–1.05) Freedom from composite outcome at 10 years: Off-pump 49% and on-pump 44% (P = 0.021)
Patient follow-up was determined by February 2011 using personal identity numbers and national registers
Off-pump CABG had similar long-term outcomes compared with on-pump CABG Limitations: Retrospective observational study, with elements of selection bias Small proportion of CABG procedures undertaken as off-pump (6.6%), with a continuous decline over the recruitment period Unknown propensity-matched details: which baseline variables were matched; differences between matched baseline and perioperative characteristics and the number of patients matched Unknown conversion rates to on-pump CABG Missing or incomplete data for certain variables
BEST EVIDENCE TOPIC
Dalén et al. (2013), Ann Thorac Surg, Sweden [4]
Matched: Similar all-cause mortality or rehospitalization [HR = 0.99 (95% CI: 0.90–1.10)] and freedom from composite outcome at 10 years (49 vs 43%; P = 0.22) Di Mauro et al. (2007), Ann Thorac Surg, Italy [5] Retrospective cohort study (level 3 evidence)
2833 patients in total with multivessel coronary disease between November 1994 and December 2001 at the Ferarrotto Hospital, Catania, Italy. Two analyses comparing off-pump and on-pump CABG were created with respect to preoperative normal and abnormal renal function. Patients were then consequently matched using propensity scores for
Late outcome
Ten-year mortality on-pump (7.0%) vs off-pump (5.3%); HR = 1.3 (95% CI: 0.91–1.9; P = 0.141)
Conclusions: Off-pump CABG provides better late outcomes in patients with normal creatinine preoperatively, but similar outcomes for those with abnormal creatinine levels. Surgical strategy does not influence late mortality in the overall population
Continued
U.A.R. Chaudhry et al. / Interactive CardioVascular and Thoracic Surgery
152
Table 1: (Continued) Author, date, journal and country Study type (level of evidence)
Patient group
Outcomes
Key results
Comments
Limitations:
preoperative and perioperative characteristics: 942 off-pump 942 on-pump
Retrospective analysis Single centre
Mean follow-up was 7.5 years Small number of patients with abnormal creatinine levels (80/942)
Other patients were excluded on the basis of incomplete data, intraoperative death and preoperative chronic dialysis Demographics: propensity-matched patients had similar preoperative characteristics for both analyses Patient selection: patients were allocated to off-pump surgery depending on vessel size, level of calcification, mechanical instability and surgical experience. Patients converted to on-pump were deemed intention-to-treat Patients were followed up annually at the outpatient clinic or by contacting the patient or cardiology. Follow-up ended in June 2006 Filardo et al. (2011), Ann Thorac Surg, USA [6] Retrospective cohort study (level 3 evidence)
8081 consecutive patients enrolled between January 1997 and December 2008 at the Baylor University Medical Centre, TX, USA: 732 off-pump 7349 on-pump
Survival
Total follow-up was 48 165 years
Unadjusted off-pump vs on-pump: HR = 1.21 (95% CI: 1.04–1.41; P = 0.012). At 5 years: 77.4 vs 80.8% At 10 years: 54.7 vs 62.3%
Conclusions:
Adjusted off-pump vs on-pump: HR = 1.18 (95% CI: 1.02–1.38; P = 0.012)
Limitations:
Patients were excluded if they had had previous valvular procedures, preoperative endocarditis, a ventricular assist device or there were missing values for MI timing
Long-term survival following on-pump CABG is significantly greater
Observational study at a single centre Three surgeons performed a significant proportion of off-pump procedures Insufficient information regarding patient selection for either CABG technique
Demographics: propensity-score analysis was used for adjustment using a logistic regression model and recognized STS risk factors. Off-pump patients were significantly older, had renal impairment, greater levels of previous PCI, greater EF and were more likely to be elective procedure cases
Rates of conversion not reported
Survival estimates and effects were calculated using the Kaplan–Meier method or Cox proportional hazard models Survival was measured until June 2009 using a national register García Fuster et al. (2013), Eur J Cardiothorac Surg, Spain [7]
1752 consecutive patients were included between January 1995 and Jun 2011 at the University General Hospital of Valencia, Spain, with complete eGFR data:
Long-term mortality (three different types: unmatched; multivariable model adjusted with propensity scores; propensity-matched group)
72 off-pump vs 418 on-pump deaths during follow-up periods
Conclusions: No advantage in long-term mortality for off-pump CABG patients. There were no
Continued
U.A.R. Chaudhry et al. / Interactive CardioVascular and Thoracic Surgery
153
Table 1: (Continued) Author, date, journal and country Study type (level of evidence)
Patient group
Retrospective cohort study (level 3 evidence)
350 off-pump 1402 on-pump
Outcomes
Mean follow-up periods were 6.4 and 10.5 years, respectively, for off-pump and on-pump CABG Excluded patients had undergone concomitant cardiac, reoperation or urgent/emergency procedures or had end-stage renal disease. 250 off-pump were propensity-score matched to 250 on-pump CABG patients
Key results
Comments
Off-pump vs on-pump: Unmatched HR = 0.96 (95% CI: 0.70–1.31; P = 0.80) Multivariable model HR = 0.87 (95% CI: 0.60–1.27; P = 0.49) Matched HR = 0.97 (95% CI: 0.60–1.55; P = 0.90)
protective effects of off-pump CABG in patients with impaired renal function Limitations: Single-centre, retrospective, non-randomized study with a long enrolment period Limited availability of follow-up methods Shorter follow-up periods for off-pump CABG patients
Demographics: patients were classified in terms of different eGFR stages. Unmatched groups had significant differences in nearly all patient characteristics except for gender, diabetes and LVEF. Matched patients were different in terms of their age and serum creatinine levels
Exclusion of cross-over data
Patient selection for CABG technique was at the discretion of the surgeon Cox regression was used as predictors for long-term mortality Long-term mortality was identified from a national database
Retrospective cohort study (level 3 evidence)
219 patients required revascularization between June 1989 and July 1990 at the Loma Linda University Medical Centre, CA, USA: 107 off-pump 112 on-pump
Survival (off-pump vs on-pump)
80 vs 79% (P = 0.8)
Number of reinterventions
22 vs 8
Percentage recatheterizations (off-pump vs on-pump)
30 vs 16%
Follow-up was at 7 years in January 1997 from patient records or via contact with patient or their practitioner
Conclusions: Similar survival rates for the two groups. Possible increased reintervention rates following off-pump Limitations: Single-centre, retrospective, non-randomized study with no cross-over data
Revascularization choice was made by the three surgeons who performed the procedures
No information regarding patient selection or preoperative characteristics are mentioned
Comparison between the two groups was made by χ 2 test
All CABG surgeries were completed by three surgeons Performed more than 20 years ago Hueb et al. (2010), Circulation, USA [9]
308 undergoing CABG surgery were randomly assigned between March 2001 and March 2006 at the Heart Institute of the University of São Paulo, Brazil:
Five-year outcomes (on-pump vs off-pump)
Combined incidence of overall mortality, stroke, MI or revascularization: HR = 0.71 (95% CI: 0.41–1.22; P = 0.21)
Conclusions: No differences in composite end-points between the two strategies at 5 years
Continued
BEST EVIDENCE TOPIC
Gundry et al. (1998), J Thorac Cardiovasc Surg, USA [8]
U.A.R. Chaudhry et al. / Interactive CardioVascular and Thoracic Surgery
154
Table 1: (Continued) Author, date, journal and country Study type (level of evidence)
Patient group
Outcomes
Key results
Prospective randomized, controlled trial (level 2 evidence)
155 off-pump 153 on-pump
Off-pump vs on-pump: Death MI Revascularization Angina Stroke Positive Treadmill test
8.4 vs 5.2% (P = 0.18) 6.5 vs 1.9% (P = 0.05) 6.5 vs 5.9% (P = 0.84) 11.8 vs 6.7% (P = 0.09) 1.9 vs 3.2% (P = 0.50) 10.1 vs 13.7% (P = 0.18)
Follow-up was at 5 years with 6-monthly visits Inclusion criteria: >70% proximal multivessel coronary stenosis, stable angina; preserved LV function; technically feasible; no previous CABG. Exclusion criteria: emergency or concomitant surgery; no written informed consent. Patients were analysed using an intention-to-treat principle
Comments
Limitations: Limited sample size Single-centre experience No differences were available for baseline demographic characteristics
Off-pump CABG patients tended to have fewer distal anastomoses and grafts, spent less time in ICU, took less time to extubate and then to discharge Data were treated as either dichotomous or continuous and analysed appropriately. Event-free survival was analysed graphically using Kapan–Meier plots and Cox regression Järvinen et al. (2013), Coron Artery Dis, Finland [10] Retrospective cohort study (level 3 evidence)
508 patients at the Heart Centre of Tampere University Hospital, Tampere, Finland, were assessed during May 1999 and November 2000 in terms of health-related QoL during 12 years following CABG: 56 off-pump 452 on-pump
10-year survival rates
Off-pump: 85.7% On-pump: 78.1% (P = 0.187)
Mean QoL at 12 years
General tendency in a decrease of scores. Onpump had a significant improvement after 12 years in 8 categories from baseline. Off-pump had improvement in 7/8 categories
Mean follow-up period was 11.8 years. All patients provided written informed consent
Conclusions: CPB had no effect on patient’s health-related QoL or survival Limitations: Preoperative pathology of coronary arteries dictated patient selection for off-pump CABG Small number of patients within off-pump category
Off-pump CABG patients were younger, had a greater female proportion, had a lower EuroSCORE, fewer multivessel disease and critical left main stem disease
Single-centre, retrospective, non-randomized study with no cross-over data
Patients were selected for CPB based on the surgeon’s preference and by the patient’s status
No evidence of matching or adjustment for baseline variables
Follow-up was by means of a questionnaire χ 2 test was used to compare off-pump and on-pump groups Puskas et al. (2011), Ann Thorac Surg, USA [11]
297 patients with multivessel disease during March 2000 and August 2001 were enrolled in the SMART trial at the Emory University Hospital, Atlanta, GA, USA: 98 off-pump 99 on-pump
Survival
At the end of follow-up, there were 26 deaths off-pump vs 31 deaths on-pump Off-pump vs on-pump: 5-year survival:
Conclusions: Similar long-term survival, late graft patency, incidence of recurrent or residual myocardial ischaemia and need for reintervention
Continued
U.A.R. Chaudhry et al. / Interactive CardioVascular and Thoracic Surgery
155
Table 1: (Continued) Author, date, journal and country Study type (level of evidence)
Patient group
Prospective, randomized, controlled trial (level 2 evidence)
Mean follow-up period was 7.5 years
Outcomes
No exclusions were in place. Patients were withdrawn if found to have mitral valve disease. Surgeries were performed by a single surgeon
Key results
Comments
92.9 vs 81.8% (P = 0.02) 7-year survival: 83.7 vs 73.7% (P = 0.09)
Limitations:
Off-pump vs on-pump: Recurrent angina
25.6 vs 11.4% (P = 0.09)
Percutaneous reintervention
2.3 vs 2.3% (P = 1.0)
Patient characteristics were generally similar; although on-pump patients had an increased incidence of previous CVA. 1 off-patient was crossed over to on-pump CABG, whereas 3 on-patients were crossed over to off-pump CABG
Graft patency (computed tomographic angiography)
76 vs 83.5% (P = 0.44)
Survival information was gained via national databases, with a study cut-off date of March 2009
Graft patency (ischaemia on positron emission tomography scanning)
35.3 vs 41.0% (P = 0.62)
Survival
Off-pump (3.6% deaths) vs on-pump (4.8% deaths): HR = 0.91 (95% CI: 0.70–1.12; P = 0.87)
Small sample size, with a fewer proportion available for follow-up graft patency and residual ischaemia Single-surgeon, single-centre trial
Off-pump vs on-pump:
Long-term survival was predicted by Kaplan–Meier estimates or Cox proportional hazard models
Retrospective cohort study (level 3 evidence)
704 consecutive patients with multivessel disease were compared having undergone surgery from January 2002 to December 2002 at Harefield Hospital, London, UK: 307 off-pump 397 on-pump The 307 off-pump patients were propensity score matched to 307 on-pump patients using 26 variables
Readmission to hospital for a cardiac cause
Need for repeat reintervention
Follow-up at 10 years was 100% All surgeons had previously performed >100 off-pump CABG procedures, and choice was determined by a group of specialists
Off-pump (3.3%) vs on-pump (3.8%): HR = 0.96 (95% CI: 0.78– 1.10) Off-pump (0.7% repeat revascularization; 0.3% redo-CABG) vs on-pump (0.9% repeat revascularization; 0.3% redo-CABG): HR = 0.93 (95% CI: 0.87– 1.05)
Conclusions: Off-pump CABG does not adversely affect on survival, readmission or reintervention Limitations: Non-randomized retrospective, single-centre study Specific exclusion criteria not provided Lack of follow-up data methods
Unmatched patient characteristics revealed that off-pump CABG patients were more likely to be male, diabetic, have high cholesterol, PVD, two-vessel disease, higher serum creatinine, more urgent. Off-pump surgeries utilized more right internal mammary arteries, fewer saphenous vein grafts and fewer grafts/patient. 3 off-pump patients were converted to CPB Long-term outcomes of interest were analysed using survival curves or Cox proportional hazards Robertson et al. (2012), J Thorac Cardiovasc Surg, Canada [13]
1285 patients were selected between January 1997 and June 2003 at the Maritime Heart Center, Nova Scotia, Canada. Following propensity-score matching according to baseline variables: 308 off-pump 308 on-pump
Long-term survival
Freedom from cardiac readmissions (MI, CHF or revascularization)
Off-pump vs on-pump: 5-year survival: 90.1 vs 90.8% 8-year survival: 80.8 vs 78.7% Off-pump vs on-pump: 5-year freedom: 77.2 vs 73.8%
Conclusions: Both techniques produced comparable results Limitations:
Continued
BEST EVIDENCE TOPIC
Raja et al. (2013), Biomed Res Int, UK [12]
U.A.R. Chaudhry et al. / Interactive CardioVascular and Thoracic Surgery
156
Table 1: (Continued) Author, date, journal and country Study type (level of evidence)
Patient group
Retrospective cohort study (level 3 evidence)
Median follow-up time was 5.9 years
Outcomes
Key results
Comments
8-year freedom: 68.3 vs 68.1%
Differences in operative techniques, surgical experience and completeness in revascularization
Patients included involved non-redo, isolated CABG and those being a resident of Nova Scotia
Non-randomized retrospective, single-centre study
Following propensity-score matching, all preoperative characteristics were similar. Operatively, off-pump CABG involved lower number of distal anastomoses, fewer saphenous vein grafts, greater level of total arterial grafts and lower proportion of complete circumflex revascularization. 37 patients were converted to CPB, but were treated as intention-to-treat A single surgeon assessed the eligibility for surgery Kaplan–Meier analysis compared late outcomes Follow-up was available from longitudinal data through population records Selnes et al. (2009), Ann Thorac Surg, USA [14] Prospective cohort study (level 3 evidence)
395 patients were grouped into four categories having been admitted at the John Hopkins University School of Medicine or nearby hospitals between September 1997 and October 2003. Two of those categories: Off-pump: 75 (between March 1998 and October 2003) On-pump: 152 (between September 1997 and March 1999) The other two groups were non-surgical cardiac comparison and healthy heart comparison Mean duration of follow-up was 5.8 and 6.9 years for off-pump and on-pump groups, respectively. Follow-up was at 6 years
Deaths at 72 months (off-pump vs on-pump) Cardiac outcomes at 72 months (off-pump vs on-pump): Hospitalized Chest pain reported MI PCI New CABG AF New diagnosis of hypertension New diagnosis of diabetes Stroke TIA Cognitive outcomes
The inclusion criteria focused on the ability of patients to undergo neuro-psychological testing, which was the primary focus of the study. Other inclusion criteria involved being able to give written informed consent, sit upright, English-speaking and not mechanically ventilated
24 vs 17%
47 vs 58% 21 vs 22% 4 vs 6% 10 vs 10% 0 vs 0% 21 vs 28% 7 vs 13% 6 vs 9% 4 vs 6% 3 vs 7% No consistent differences in cognitive outcomes. Both groups had a lower baseline performance and a greater degree of cognitive decline than the healthy heart control group. Compared with baseline, neither group was dramatically worse
Conclusions: Vascular disease may have an impact of cognitive outcomes but management strategy is not a determinant for long-term outcomes Limitations: Non-randomized Bias involvement after selective drop-out of poorly performing subjects Different time frames for off-pump and on-pump for patient enrolment No direct comparison to detect significance was available for baseline characteristics and outcomes Off-pump patients from neighbouring hospitals were included
Direct comparison and statistical analysis were not available between the two groups in question for baseline variables
Continued
U.A.R. Chaudhry et al. / Interactive CardioVascular and Thoracic Surgery
157
Table 1: (Continued) Author, date, journal and country Study type (level of evidence)
Patient group
Outcomes
Key results
Comments
Sigurjónsson et al. (2012), Laeknabladid, Iceland [15]
720 consecutive patients who underwent CABG surgery at Landspítali-The National University Hospital, Reykjavik, Iceland, between January 2002 and December 2006: 207 off-pump 513 on-pump
5-year survival
Off-pump: 93% On-pump: 92% (P = 0.87)
Conclusions:
Retrospective cohort study (level 3 evidence)
Myocardial revascularization in Iceland yields similar long-term survival in Iceland Limitations: No details on collection of long-term data and method of determining choice of surgery
Mean follow-up was 5.0 years Concomitant surgical patients were excluded. A greater proportion of off-pump patients were male and had a higher NYHA class. 23 conversions to CPB were made, but treated as intention-to-treat
Non-randomized retrospective, single-centre study
Kaplan–Meier methods were used to estimate survival
Prospective, randomized, controlled trial (level 2 evidence)
281 CABG patients who had an Octopus stabilizer device utilized across three centres in Netherlands were enrolled between 1998 and 2000: 142 off-pump 139 on-pump Mean follow-up was at 5.2 years Patients referred for CABG surgery for the first time and deemed technically feasible were included. Emergency or concomitant surgical patients or those with a Q-wave MI or unable to provide informed consent or undergo neuro-psychological testing were excluded from the trial 10 off-pump patients underwent on-pump surgery, whereas 5 on-pump patients underwent off-pump surgery
All deaths (off-pump vs on-pump) at 5 years
Retrospective cohort study
CPB did not affect cognitive and cardiac outcomes at 5 years 21.1 vs 18.0% (P = 0.55) 4.9 vs 6.5% (P = 0.62) 17.7 vs 12.3% (P = 0.23) 1.4 vs 3.6% (P = 0.28) 7.7 vs 6.0% (P = 0.47)
Cognitive outcomes (off-pump vs on-pump) at 5 years
Standard definition of cognitive decline: 50.4 vs 50.4% (P > 0.99)
May not reflect CABG surgery population: younger population and lower preoperative risk
More conservative definition of cognitive decline: 33.3 vs 35.0% (P = 0.79)
Only reflected 11% of all CABG procedures during the study enrollment period
Limitations: 16 patients did not receive assigned treatment
Small number of participants QoL
The patient’s general practitioner initially provided follow-up data A total of 8580 patients underwent CABG surgery between July and December 2000 using two major New York State’s CSRS Database: 2640 off-pump 5940 on-pump
Conclusions:
Cardiovascular events (off-pump vs on-pump) at 5 years: Total MI Recurrent angina Stroke Repeat revascularization
The proportions of patients suffering an event during follow-up were compared using absolute risk difference, Fisher’s exact test and event curves if appropriate
Wu et al. (2012), Circ Cardiovasc Qual Outcomes, USA [17]
8.5 vs 6.5% (P = 0.65)
Survival
No differences in overall quality of life and in seven of the eight domains between the two groups, and both showed improvement following baseline levels. The domain of limitations due to physical health problems fared better following on-pump CABG Among 2631 matched pairs, 782 off-pump patients vs 725 on-pump patients died during follow-up: HR = 1.10 (95% CI: 0.99–1.21; P = 0.07) In a sensitivity analysis, which involved patients
Differences in baseline characteristics were not reported
Conclusions: No statistically significant differences in long-term mortality between on-pump and off-pump CABG Limitations:
Continued
BEST EVIDENCE TOPIC
van Dijk et al. (2007), JAMA, Netherlands [16]
U.A.R. Chaudhry et al. / Interactive CardioVascular and Thoracic Surgery
158
Table 1: (Continued) Author, date, journal and country Study type (level of evidence)
(level 3 evidence)
Patient group
Outcomes
Propensity score matching at a 1:1 ratio was conducted to create 2631 matched patients. Differences in baseline risk factors were no longer present following matching
Key results
Comments
treated by high-volume surgeons who performed CABG in >50% of their cases: HR = 1.12 (95% CI: 0.95–1.32; P = 0.17)
Retrospective, non-randomized CABG surgery performed more than 10 years ago
Median follow-up period was 7.2 years No exclusions were reported. 43 patients from the original cohort were converted to on-pump surgery Mortality data were available through databases Differences in survival were examined through Kaplan–Meier analysis and hazard ratio for death was obtained via Cox proportional hazards regression model AF: atrial fibrillation; BHACAS: beating heart against cardioplegic arrest studies; CABG: coronary artery bypass surgery; CHF: congestive heart failure; CI: confidence interval; COPD: chronic obstructive pulmonary disease; CPB: cardiopulmonary bypass; CSRS: cardiac surgery reporting system; CVA: cerebrovascular accident; EF: ejection fraction; eGFR: estimated glomerular filtration rate; HR: hazard ratio; ICU: intensive care unit; LV: left ventricular; LVEF: left ventricular ejection fraction; MACEs: major adverse cardiovascular events; MI: myocardial infarction; NYHA: New York Heart Association; OR: odds ratio; PCI: percutaneous coronary intervention; PVD: peripheral vascular disease; QoL: quality of life; RCT: randomized controlled trial; SMART: surgical management of arterial revascularization therapies; STS: Society of Thoracic Surgery; SVT: supraventricular arrhythmia; SWEDEHEART: Swedish Web-system for Enhancement and Development of Evidence-based care in Heart Disease Evaluated According to Recommended Therapies; TIA: transient ischaemic attack; VA: veterans affairs.
Search outcome Two hundred and fifty-six papers were found as a result of the reported search. From these, 16 papers were identified that provided the best evidence to answer the question. These are presented in Table 1.
RESULTS Angelini et al. [2] studied 401 patients from two separate prospective randomized clinical trials (RCT) with varying exclusion criteria, for a period of up to 6.4 years. There were no differences between off-pump and on-pump CABG surgery in: survival; survival-free from major adverse cardiac events (MACEs) (including myocardial infarction (MI), angina recurrence and need for revascularization); survival-free from death or MACEs; graft patency and quality of life (QoL). Bakaeen et al. [3] retrospectively utilized the large Veterans Affairs (VA) registry, to propensity score 8911 off-pump patients in a 1:3 manner. After a median follow-up of 6.7 years, overall survival following on-pump CABG surgery was significantly improved. Dalén et al. [4] retrospectively observed patients undergoing CABG operations in Sweden and found similar long-term survival
and other composite outcomes (mortality, rehospitalization for MI, heart failure or stroke) following 1:1 propensity-score matching and after a mean period of 7.1 years. Di Mauro et al. [5] were primarily concerned with the effect of preoperative renal function and its impact on postoperative outcomes. A single centre, retrospective analysis of the combined and matched cohort, including 1884 patients, showed no favourability to either technique after 7.5 years. Off-pump CABG did provide better late outcomes in patients with normal creatinine preoperatively. Filardo et al. [6] retrospectively analysed 8081 consecutive patients at a single institution. Despite adjustment of risk factors, patients following on-pump CABG had a significantly improved long-term survival. García Fuster et al. [7] retrospectively studied 1752 consecutive patients with varying degrees of preoperative renal function. Off-pump patients were followed up for a shorter period but did not have an advantage for late mortality. Gundry et al. [8] retrospectively enrolled 219 patients during 1989–90 for 7 years demonstrating similar survival but potentially increased reintervention rates following off-pump CABG surgery. Hueb et al. [9] prospectively compared 155 off-pump patients with 153 on-pump patients who met their inclusion/exclusion criteria for their RCT. Follow-up at 5 years revealed no differences in
U.A.R. Chaudhry et al. / Interactive CardioVascular and Thoracic Surgery
Clinical bottom line We sought to identify whether off-pump CABG conferred mortality or morbidity benefit over on-pump CABG in the long-term. Two of the 16 studies showed improved survival following on-pump surgery, whereas the remaining 14 showed no difference in late mortality (≥5 years). No differences were observed for other morbidity outcomes such as MACEs (including MI, recurrence of angina and heart failure or revascularization), stroke, graft patency, cognitive and QoL. A few limitations must be considered. Firstly, only four of the studies were prospective RCTs and sample sizes were small. Secondly, there is substantial variability between the studies in terms of patient cohorts, selection criteria, preoperative and operative risk factors, study methods and reporting of results, which may have an immeasurable effect on long-term outcomes. In summary, these results suggest that off-pump has long-term morbidity outcomes comparable with those of on-pump following CABG surgery. Long-term survival may also be similar to or slightly favourable towards on-pump CABG surgery. Procedural differences such as graft patency and completeness of revascularization as well as other factors need to be considered. Large, prospective RCTs
with adequate preoperative matching are now required to definitively measure and compare long-term outcomes. Conflict of interest: none declared.
REFERENCES [1] Dunning J, Prendergast B, Mackway-Jones K. Towards evidence-based medicine in cardiothoracic surgery: best BETS. Interact CardioVasc Thorac Surg 2003;2:405–9. [2] Angelini GD, Culliford L, Smith DK, Hamilton MC, Murphy GJ, Ascione R et al. Effects of on- and off-pump coronary artery surgery on graft patency, survival, and health-related quality of life: long-term follow-up of 2 randomized controlled trials. J Thorac Cardiovasc Surg 2009;137:295–303. [3] Bakaeen FG, Chu D, Kelly RF, Ward HB, Jessen ME, Chen GJ et al. Performing coronary artery bypass grafting off-pump may compromise long-term survival in a veteran population. Ann Thorac Surg 2013;95: 1952–8; discussion 1959–60. [4] Dalen M, Ivert T, Holzmann MJ, Sartipy U. Long-term survival after off-pump coronary artery bypass surgery: a Swedish nationwide cohort study. Ann Thorac Surg 2013;96:2054–60. [5] Di Mauro M, Gagliardi M, Iaco AL, Contini M, Bivona A, Bosco P et al. Does off-pump coronary surgery reduce postoperative acute renal failure? The importance of preoperative renal function. Ann Thorac Surg 2007;84: 1496–502. [6] Filardo G, Grayburn PA, Hamilton C, Hebeler RF Jr, Cooksey WB, Hamman B. Comparing long-term survival between patients undergoing off-pump and on-pump coronary artery bypass graft operations. Ann Thorac Surg 2011;92:571–7; discussion 577–8. [7] Garcia Fuster R, Paredes F, Garcia Pelaez A, Martin E, Canovas S, Gil O et al. Impact of increasing degrees of renal impairment on outcomes of coronary artery bypass grafting: the off-pump advantage. Eur J Cardiothorac Surg 2013;44:732–42. [8] Gundry SR, Romano MA, Shattuck OH, Razzouk AJ, Bailey LL. Seven-year follow-up of coronary artery bypasses performed with and without cardiopulmonary bypass. J Thorac Cardiovasc Surg 1998;115:1273–7; discussion 1277–8. [9] Hueb W, Lopes NH, Pereira AC, Hueb AC, Soares PR, Favarato D et al. Five-year follow-up of a randomized comparison between off-pump and on-pump stable multivessel coronary artery bypass grafting. The MASS III Trial. Circulation 2010;122:S48–52. [10] Jarvinen O, Hokkanen M, Huhtala H. Quality of life 12 years after on-pump and off-pump coronary artery bypass grafting. Coron Artery Dis 2013;24:663–8. [11] Puskas JD, Williams WH, O’Donnell R, Patterson RE, Sigman SR, Smith AS et al. Off-pump and on-pump coronary artery bypass grafting are associated with similar graft patency, myocardial ischemia, and freedom from reintervention: long-term follow-up of a randomized trial. Ann Thorac Surg 2011;91:1836–42; discussion 1842–3. [12] Raja SG, Husain M, Popescu FL, Chudasama D, Daley S, Amrani M. Does off-pump coronary artery bypass grafting negatively impact long-term survival and freedom from reintervention? Biomed Res Int 2013;2013: 602871. [13] Robertson MW, Buth KJ, Stewart KM, Wood JR, Sullivan JA, Hirsch GM et al. Complete revascularization is compromised in off-pump coronary artery bypass grafting. J Thorac Cardiovasc Surg 2013;145:992–8. [14] Selnes OA, Grega MA, Bailey MM, Pham LD, Zeger SL, Baumgartner WA et al. Do management strategies for coronary artery disease influence 6-year cognitive outcomes? Ann Thorac Surg 2009;88:445–54. [15] Sigurjonsson H, Helgadottir S, Oddsson SJ, Sigurdsson MI, Geirsson A, Arnorsson T et al. [Outcome of myocardial revascularisation in Iceland]. Laeknabladid 2012;98:451–6. [16] van Dijk D, Spoor M, Hijman R, Nathoe HM, Borst C, Jansen EW et al. Cognitive and cardiac outcomes 5 years after off-pump vs on-pump coronary artery bypass graft surgery. J AmMed Assoc 2007;297:701–8. [17] Wu C, Camacho FT, Culliford AT, Gold JP, Wechsler AS, Higgins RS et al. A comparison of long-term mortality for off-pump and on-pump coronary artery bypass graft surgery. Circ Cardiovasc Qual Outcomes 2012;5:76–84.
BEST EVIDENCE TOPIC
composite end-points including death, MI, need for revascularization, angina recurrence or stroke. Järvinen et al. [10] retrospectively analysed 56 off-pump patients vs 452 on-pump patients after 11.8 years to conclude no overall effect in survival or health-related QoL. Puskas et al. [11] prospectively reviewed 98 off-pump and 99 on-pump patients in their RCT who underwent CABG surgery by a single surgeon. There were similar incidences in survival, recurrent angina, percutaneous reintervention and graft patency after 7.5 years. Raja et al. [12] retrospectively propensity-score matched 307 patients in each group and showed that off-pump did not adversely impact on survival, readmission for a cardiac causes or need for reintervention after 10 years. Robertson et al. [13] demonstrated comparable results in longterm survival and freedom from cardiac readmissions for MI, heart failure or revascularization after propensity-matching 308 off-pump and on-pump patients. Selnes et al. [14] observed generally similar long-term cognitive and cardiac outcomes at 6 years having compared four different cohorts, two of which included 75 off-pump and 152 on-pump patients. Sigurjónsson et al. [15] retrospectively reviewed 720 consecutive patients at 5 years and showed similar long-term survival. van Dijk et al. [16] conducted an RCT involving 281 patients using an Octopus stabilizer device and at 5 years concluded that CPB had no effect on survival, MACEs (MI, angina recurrence, stroke, repeat revascularization), cognitive and QoL outcomes. Wu et al. [17] utilized New York’s Cardiac Surgery Reporting System database to 1:1 propensity-score match 2631 patients and follow-up for a period of 7.2 years. There was no significant difference in long-term mortality either overall or in their subgroup analysis of patients treated by high-volume surgeons who performed CABG in >50% of their cases.
159
Interactive CardioVascular and Thoracic Surgery 19 (2014) 149–159 doi:10.1093/icvts/ivu075 Advance Access publication 21 March 2014
Does off-pump coronary artery bypass graft surgery have a beneficial effect on long-term mortality and morbidity compared with on-pump coronary artery bypass graft surgery? Umar A.R. Chaudhry, Christopher Rao, Leanne Harling and Thanos Athanasiou* Department of Surgery and Cancer, St Mary’s Hospital, Imperial College London, London, UK * Corresponding author. Department of Surgery and Cancer, Imperial College London, 10th Floor, QEQM Building, St Mary’s Hospital Campus, South Wharf Road, London W2 1NY, UK. Tel: +44-20-33127630; fax: +44-20-33126309; e-mail: [email protected] (T. Athanasiou). Received 20 December 2013; received in revised form 16 February 2014; accepted 27 February 2014
Abstract A best evidence topic in cardiac surgery was written according to a structured protocol. The question addressed was whether off-pump coronary artery bypass grafting (CABG) surgery offered superior long-term outcomes compared with on-pump CABG surgery. Best evidence papers were considered to be those that had a follow-up period of ≥5 years, had >50 patients in either cohort, did not utilize concomitant interventions nor comprised low-risk, high-risk or sub-population groups. Where potential duplicate data sets from the same institution were likely, the more credible and recently published study was included. Two hundred and fifty-six papers were found as a result of the reported search, of which 16 represented the best evidence to answer the clinical question. The authors, journal, date and country of publication, patient group studied, study type, relevant outcomes and results of these papers are tabulated. The 16 studies comprised 4 prospective randomized controlled trials (RCTs), with the remaining 12 retrospective, of which 8 were propensity-score matched. All 4 RCTs contained fewer than 450 participants. Two studies concluded with a survival advantage towards on-pump CABG: one, a large registry-based study, the Veterans Affairs, with >25 000 patients, and another, a propensity-matched retrospective study involving almost 8000 patients. The remaining 14 studies all provided evidence to suggest comparable long-term survival. In addition, all other long-term outcomes mentioned within these studies including angina recurrence, myocardial infarction heart failure, need for revascularization, stroke, graft patency, cognitive and quality of life showed similar results between the two groups. We conclude that off-pump CABG surgery may have similar or slightly reduced long-term survival compared with on-pump CABG surgery. Other long-term indicators such as cardiovascular or cerebrovascular events or neuro-psychological outcomes were similar between the two groups. Despite these conclusions, the evidence is limited by substantial variability in patient selection and study methods. The CORONARY (coronary artery bypass surgery off- or on-pump revascularization study) trial recently presented results, which showed no significant differences in composite outcomes at 1 year; it will be interesting to observe whether these comparable outcomes are maintained for a much longer time frame.
INTRODUCTION A best evidence topic was constructed according to a structured protocol. This is fully described in the ICVTS [1].
different studies present early outcomes but little data comparing the two techniques are provided regarding long-term outcomes, particularly survival. A colleague suggests that off-pump CABG surgery may be more beneficial in the long-term as it negates the effects of cardiopulmonary bypass (CPB). You are unconvinced and resolve to check the literature yourself.
Three-part question In [ patients undergoing coronary artery bypass grafting (CABG) surgery], is (the off-pump or on-pump technique) superior in terms of [long-term survival and other late outcomes]?
Clinical scenario You are at an international conference discussing the benefits of off-pump compared with on-pump CABG surgery. Several
Search strategy Medline from 1950 to December 2013 was searched using PubMed interface using the following terms: (‘coronary artery bypass’*) AND (‘coronary artery bypass, off-pump’*, ‘off-pump’, ‘OPCAB’, ‘beating heart’) AND (‘cardiopulmonary bypass’*, ‘conventional coronary artery bypass’, ‘on-pump’, ‘ONCAB’) AND (‘long-term’ OR ‘late’) AND (‘survival’, ‘mortality’, ‘death’, ‘follow-up’, ‘outcomes’).
© The Author 2014. Published by Oxford University Press on behalf of the European Association for Cardio-Thoracic Surgery. All rights reserved.
BEST EVIDENCE TOPIC
Keywords: Coronary artery bypass • Off-pump • Cardiopulmonary bypass • Long-term • Outcomes
U.A.R. Chaudhry et al. / Interactive CardioVascular and Thoracic Surgery
150
Table 1: Best evidence papers Author, date, journal and country Study type (level of evidence)
Patient group
Outcomes
Key results
Comments
Angelini et al. (2009), J Thorac Cardiovasc Surg, UK [2]
401 participants from two randomized trials (BHACAS 1 and 2) at the Bristol Heart Institute, UK, were recruited overall between March 1997 and November 1999: 200 off-pump 201 on-pump
Survival
No difference between off-pump vs on-pump CABG groups: [HR = 1.24; 95% CI: 0.72–2.15, P = 0.44]. Overall, 29 deaths in off-pump vs 23 deaths in on-pump category
Conclusions:
No difference between off-pump vs on-pump CABG groups: [HR = 0.77; 95% CI: 0.51–1.18, P = 0.24]. Overall, 39 MACEs in off-pump (3 revascularization, 8 MI, 28 angina) vs 49 MACEs in on-pump (3 revascularization, 8 MI, 38 angina) category
Limitations:
Prospective, randomized, controlled trial (level 2 evidence)
Survival-free from MACEs Mean follow-up period for off-pump and on-pump survival was 6.3 years and 6.4 years, respectively, and survival-free from MACEs or death was 5.4 and 5.2 years, respectively BHACAS 2 trial (September 1998 to November 1999) held later excluded patients with: LVEF 0.99]
QoL
No difference between off-pump vs on-pump CABG groups across a range of domains
Survival (matched patients)
Off-pump vs on-pump all-cause mortality:
Survival and MACEs were compared using Cox regression analysis Patient follow-up was through a national database, and by annual questionnaires or via family practitioners for MACEs
No difference between off-pump vs on-pump CABG groups: [HR = 0.84; 95% CI: 0.58–1.24, P = 0.39]
Long-term survival and MACE outcomes with off-pump CABG are similar to those with on-pump CABG
At 5 years: 14.5 vs 13.5%; (RR = 1.08; 95% CI: 1.02– 1.15) At 10 years: 25.2 vs 23.6%; (RR = 1.07; 95% CI: 1.03– 1.12). Overall HR = 1.06 (95% CI: 1.00–1.13; P = 0.036)
Conclusions: On-pump CABG may be associated with increased long-term survival Limitations: Retrospective data VA population—patients were more likely to be male and younger Conversion to on-pump not recorded. Analysis was not on an intention-to-treat basis Long period of recruitment with unknown levels of surgical experience at each centre No details as to patient selection to either cohort were provided
Demographics: after adjustment, no significant differences in preoperative risk factors were found, but off-pump
Continued
U.A.R. Chaudhry et al. / Interactive CardioVascular and Thoracic Surgery
151
Table 1: (Continued) Author, date, journal and country Study type (level of evidence)
Patient group
Outcomes
Key results
Comments
Survival and 10-year all-cause mortality (crude and multivariate model for unmatched cohorts; propensity-matched cohort)
Unmatched: Off-pump CABG associated with improved crude all-cause mortality but not following adjustment. Crude: HR = 0.92 (95% CI: 0.85–0.98) Multivariable: HR = 0.99 (95% CI: 0.92–1.06)
Conclusions:
patients had fewer distal anastomoses (2.66 vs 3.18, P < 0.0001) Cox proportional hazard models were used to estimate an HR for survival
Retrospective cohort study (level 3 evidence)
50 676 patients who underwent CABG in Sweden between 1998 and 2008 using the SWEDEHEART registry: 3337 off-pump 47 339 on-pump 1:1 propensity-score-matched cohort using logistic regression
Overall survival at 10 years: off-pump 73% and on-pump 71% (P = 0.017)
The mean follow-up period for the entire cohort was 7.1 years Excluded patients comprised those who had previous cardiac surgery, concomitant procedures or an emergency surgery Demographics: unmatched patients had significant differences in all baseline and perioperative characteristics except for levels for COPD and heart failure and use of internal mammary use
Matched: Similar overall long-term survival [HR = 1.02 (95% CI: 0.91–1.16)] and survival at 10 years (73 vs 72%; P = 0.56) All-cause mortality or rehospitalization for MI, heart failure or stroke (composite outcome)
Cox proportional hazards regression was used to estimate the HR. Multivariate adjustments were made to crude associations between off-pump and on-pump CABG
Unmatched: Off-pump CABG associated with improved composite outcome but not following adjustment. Crude: HR = 0.94 (95% CI: 0.88–0.99) Multivariable: HR = 0.99 (95% CI: 0.94–1.05) Freedom from composite outcome at 10 years: Off-pump 49% and on-pump 44% (P = 0.021)
Patient follow-up was determined by February 2011 using personal identity numbers and national registers
Off-pump CABG had similar long-term outcomes compared with on-pump CABG Limitations: Retrospective observational study, with elements of selection bias Small proportion of CABG procedures undertaken as off-pump (6.6%), with a continuous decline over the recruitment period Unknown propensity-matched details: which baseline variables were matched; differences between matched baseline and perioperative characteristics and the number of patients matched Unknown conversion rates to on-pump CABG Missing or incomplete data for certain variables
BEST EVIDENCE TOPIC
Dalén et al. (2013), Ann Thorac Surg, Sweden [4]
Matched: Similar all-cause mortality or rehospitalization [HR = 0.99 (95% CI: 0.90–1.10)] and freedom from composite outcome at 10 years (49 vs 43%; P = 0.22) Di Mauro et al. (2007), Ann Thorac Surg, Italy [5] Retrospective cohort study (level 3 evidence)
2833 patients in total with multivessel coronary disease between November 1994 and December 2001 at the Ferarrotto Hospital, Catania, Italy. Two analyses comparing off-pump and on-pump CABG were created with respect to preoperative normal and abnormal renal function. Patients were then consequently matched using propensity scores for
Late outcome
Ten-year mortality on-pump (7.0%) vs off-pump (5.3%); HR = 1.3 (95% CI: 0.91–1.9; P = 0.141)
Conclusions: Off-pump CABG provides better late outcomes in patients with normal creatinine preoperatively, but similar outcomes for those with abnormal creatinine levels. Surgical strategy does not influence late mortality in the overall population
Continued
U.A.R. Chaudhry et al. / Interactive CardioVascular and Thoracic Surgery
152
Table 1: (Continued) Author, date, journal and country Study type (level of evidence)
Patient group
Outcomes
Key results
Comments
Limitations:
preoperative and perioperative characteristics: 942 off-pump 942 on-pump
Retrospective analysis Single centre
Mean follow-up was 7.5 years Small number of patients with abnormal creatinine levels (80/942)
Other patients were excluded on the basis of incomplete data, intraoperative death and preoperative chronic dialysis Demographics: propensity-matched patients had similar preoperative characteristics for both analyses Patient selection: patients were allocated to off-pump surgery depending on vessel size, level of calcification, mechanical instability and surgical experience. Patients converted to on-pump were deemed intention-to-treat Patients were followed up annually at the outpatient clinic or by contacting the patient or cardiology. Follow-up ended in June 2006 Filardo et al. (2011), Ann Thorac Surg, USA [6] Retrospective cohort study (level 3 evidence)
8081 consecutive patients enrolled between January 1997 and December 2008 at the Baylor University Medical Centre, TX, USA: 732 off-pump 7349 on-pump
Survival
Total follow-up was 48 165 years
Unadjusted off-pump vs on-pump: HR = 1.21 (95% CI: 1.04–1.41; P = 0.012). At 5 years: 77.4 vs 80.8% At 10 years: 54.7 vs 62.3%
Conclusions:
Adjusted off-pump vs on-pump: HR = 1.18 (95% CI: 1.02–1.38; P = 0.012)
Limitations:
Patients were excluded if they had had previous valvular procedures, preoperative endocarditis, a ventricular assist device or there were missing values for MI timing
Long-term survival following on-pump CABG is significantly greater
Observational study at a single centre Three surgeons performed a significant proportion of off-pump procedures Insufficient information regarding patient selection for either CABG technique
Demographics: propensity-score analysis was used for adjustment using a logistic regression model and recognized STS risk factors. Off-pump patients were significantly older, had renal impairment, greater levels of previous PCI, greater EF and were more likely to be elective procedure cases
Rates of conversion not reported
Survival estimates and effects were calculated using the Kaplan–Meier method or Cox proportional hazard models Survival was measured until June 2009 using a national register García Fuster et al. (2013), Eur J Cardiothorac Surg, Spain [7]
1752 consecutive patients were included between January 1995 and Jun 2011 at the University General Hospital of Valencia, Spain, with complete eGFR data:
Long-term mortality (three different types: unmatched; multivariable model adjusted with propensity scores; propensity-matched group)
72 off-pump vs 418 on-pump deaths during follow-up periods
Conclusions: No advantage in long-term mortality for off-pump CABG patients. There were no
Continued
U.A.R. Chaudhry et al. / Interactive CardioVascular and Thoracic Surgery
153
Table 1: (Continued) Author, date, journal and country Study type (level of evidence)
Patient group
Retrospective cohort study (level 3 evidence)
350 off-pump 1402 on-pump
Outcomes
Mean follow-up periods were 6.4 and 10.5 years, respectively, for off-pump and on-pump CABG Excluded patients had undergone concomitant cardiac, reoperation or urgent/emergency procedures or had end-stage renal disease. 250 off-pump were propensity-score matched to 250 on-pump CABG patients
Key results
Comments
Off-pump vs on-pump: Unmatched HR = 0.96 (95% CI: 0.70–1.31; P = 0.80) Multivariable model HR = 0.87 (95% CI: 0.60–1.27; P = 0.49) Matched HR = 0.97 (95% CI: 0.60–1.55; P = 0.90)
protective effects of off-pump CABG in patients with impaired renal function Limitations: Single-centre, retrospective, non-randomized study with a long enrolment period Limited availability of follow-up methods Shorter follow-up periods for off-pump CABG patients
Demographics: patients were classified in terms of different eGFR stages. Unmatched groups had significant differences in nearly all patient characteristics except for gender, diabetes and LVEF. Matched patients were different in terms of their age and serum creatinine levels
Exclusion of cross-over data
Patient selection for CABG technique was at the discretion of the surgeon Cox regression was used as predictors for long-term mortality Long-term mortality was identified from a national database
Retrospective cohort study (level 3 evidence)
219 patients required revascularization between June 1989 and July 1990 at the Loma Linda University Medical Centre, CA, USA: 107 off-pump 112 on-pump
Survival (off-pump vs on-pump)
80 vs 79% (P = 0.8)
Number of reinterventions
22 vs 8
Percentage recatheterizations (off-pump vs on-pump)
30 vs 16%
Follow-up was at 7 years in January 1997 from patient records or via contact with patient or their practitioner
Conclusions: Similar survival rates for the two groups. Possible increased reintervention rates following off-pump Limitations: Single-centre, retrospective, non-randomized study with no cross-over data
Revascularization choice was made by the three surgeons who performed the procedures
No information regarding patient selection or preoperative characteristics are mentioned
Comparison between the two groups was made by χ 2 test
All CABG surgeries were completed by three surgeons Performed more than 20 years ago Hueb et al. (2010), Circulation, USA [9]
308 undergoing CABG surgery were randomly assigned between March 2001 and March 2006 at the Heart Institute of the University of São Paulo, Brazil:
Five-year outcomes (on-pump vs off-pump)
Combined incidence of overall mortality, stroke, MI or revascularization: HR = 0.71 (95% CI: 0.41–1.22; P = 0.21)
Conclusions: No differences in composite end-points between the two strategies at 5 years
Continued
BEST EVIDENCE TOPIC
Gundry et al. (1998), J Thorac Cardiovasc Surg, USA [8]
U.A.R. Chaudhry et al. / Interactive CardioVascular and Thoracic Surgery
154
Table 1: (Continued) Author, date, journal and country Study type (level of evidence)
Patient group
Outcomes
Key results
Prospective randomized, controlled trial (level 2 evidence)
155 off-pump 153 on-pump
Off-pump vs on-pump: Death MI Revascularization Angina Stroke Positive Treadmill test
8.4 vs 5.2% (P = 0.18) 6.5 vs 1.9% (P = 0.05) 6.5 vs 5.9% (P = 0.84) 11.8 vs 6.7% (P = 0.09) 1.9 vs 3.2% (P = 0.50) 10.1 vs 13.7% (P = 0.18)
Follow-up was at 5 years with 6-monthly visits Inclusion criteria: >70% proximal multivessel coronary stenosis, stable angina; preserved LV function; technically feasible; no previous CABG. Exclusion criteria: emergency or concomitant surgery; no written informed consent. Patients were analysed using an intention-to-treat principle
Comments
Limitations: Limited sample size Single-centre experience No differences were available for baseline demographic characteristics
Off-pump CABG patients tended to have fewer distal anastomoses and grafts, spent less time in ICU, took less time to extubate and then to discharge Data were treated as either dichotomous or continuous and analysed appropriately. Event-free survival was analysed graphically using Kapan–Meier plots and Cox regression Järvinen et al. (2013), Coron Artery Dis, Finland [10] Retrospective cohort study (level 3 evidence)
508 patients at the Heart Centre of Tampere University Hospital, Tampere, Finland, were assessed during May 1999 and November 2000 in terms of health-related QoL during 12 years following CABG: 56 off-pump 452 on-pump
10-year survival rates
Off-pump: 85.7% On-pump: 78.1% (P = 0.187)
Mean QoL at 12 years
General tendency in a decrease of scores. Onpump had a significant improvement after 12 years in 8 categories from baseline. Off-pump had improvement in 7/8 categories
Mean follow-up period was 11.8 years. All patients provided written informed consent
Conclusions: CPB had no effect on patient’s health-related QoL or survival Limitations: Preoperative pathology of coronary arteries dictated patient selection for off-pump CABG Small number of patients within off-pump category
Off-pump CABG patients were younger, had a greater female proportion, had a lower EuroSCORE, fewer multivessel disease and critical left main stem disease
Single-centre, retrospective, non-randomized study with no cross-over data
Patients were selected for CPB based on the surgeon’s preference and by the patient’s status
No evidence of matching or adjustment for baseline variables
Follow-up was by means of a questionnaire χ 2 test was used to compare off-pump and on-pump groups Puskas et al. (2011), Ann Thorac Surg, USA [11]
297 patients with multivessel disease during March 2000 and August 2001 were enrolled in the SMART trial at the Emory University Hospital, Atlanta, GA, USA: 98 off-pump 99 on-pump
Survival
At the end of follow-up, there were 26 deaths off-pump vs 31 deaths on-pump Off-pump vs on-pump: 5-year survival:
Conclusions: Similar long-term survival, late graft patency, incidence of recurrent or residual myocardial ischaemia and need for reintervention
Continued
U.A.R. Chaudhry et al. / Interactive CardioVascular and Thoracic Surgery
155
Table 1: (Continued) Author, date, journal and country Study type (level of evidence)
Patient group
Prospective, randomized, controlled trial (level 2 evidence)
Mean follow-up period was 7.5 years
Outcomes
No exclusions were in place. Patients were withdrawn if found to have mitral valve disease. Surgeries were performed by a single surgeon
Key results
Comments
92.9 vs 81.8% (P = 0.02) 7-year survival: 83.7 vs 73.7% (P = 0.09)
Limitations:
Off-pump vs on-pump: Recurrent angina
25.6 vs 11.4% (P = 0.09)
Percutaneous reintervention
2.3 vs 2.3% (P = 1.0)
Patient characteristics were generally similar; although on-pump patients had an increased incidence of previous CVA. 1 off-patient was crossed over to on-pump CABG, whereas 3 on-patients were crossed over to off-pump CABG
Graft patency (computed tomographic angiography)
76 vs 83.5% (P = 0.44)
Survival information was gained via national databases, with a study cut-off date of March 2009
Graft patency (ischaemia on positron emission tomography scanning)
35.3 vs 41.0% (P = 0.62)
Survival
Off-pump (3.6% deaths) vs on-pump (4.8% deaths): HR = 0.91 (95% CI: 0.70–1.12; P = 0.87)
Small sample size, with a fewer proportion available for follow-up graft patency and residual ischaemia Single-surgeon, single-centre trial
Off-pump vs on-pump:
Long-term survival was predicted by Kaplan–Meier estimates or Cox proportional hazard models
Retrospective cohort study (level 3 evidence)
704 consecutive patients with multivessel disease were compared having undergone surgery from January 2002 to December 2002 at Harefield Hospital, London, UK: 307 off-pump 397 on-pump The 307 off-pump patients were propensity score matched to 307 on-pump patients using 26 variables
Readmission to hospital for a cardiac cause
Need for repeat reintervention
Follow-up at 10 years was 100% All surgeons had previously performed >100 off-pump CABG procedures, and choice was determined by a group of specialists
Off-pump (3.3%) vs on-pump (3.8%): HR = 0.96 (95% CI: 0.78– 1.10) Off-pump (0.7% repeat revascularization; 0.3% redo-CABG) vs on-pump (0.9% repeat revascularization; 0.3% redo-CABG): HR = 0.93 (95% CI: 0.87– 1.05)
Conclusions: Off-pump CABG does not adversely affect on survival, readmission or reintervention Limitations: Non-randomized retrospective, single-centre study Specific exclusion criteria not provided Lack of follow-up data methods
Unmatched patient characteristics revealed that off-pump CABG patients were more likely to be male, diabetic, have high cholesterol, PVD, two-vessel disease, higher serum creatinine, more urgent. Off-pump surgeries utilized more right internal mammary arteries, fewer saphenous vein grafts and fewer grafts/patient. 3 off-pump patients were converted to CPB Long-term outcomes of interest were analysed using survival curves or Cox proportional hazards Robertson et al. (2012), J Thorac Cardiovasc Surg, Canada [13]
1285 patients were selected between January 1997 and June 2003 at the Maritime Heart Center, Nova Scotia, Canada. Following propensity-score matching according to baseline variables: 308 off-pump 308 on-pump
Long-term survival
Freedom from cardiac readmissions (MI, CHF or revascularization)
Off-pump vs on-pump: 5-year survival: 90.1 vs 90.8% 8-year survival: 80.8 vs 78.7% Off-pump vs on-pump: 5-year freedom: 77.2 vs 73.8%
Conclusions: Both techniques produced comparable results Limitations:
Continued
BEST EVIDENCE TOPIC
Raja et al. (2013), Biomed Res Int, UK [12]
U.A.R. Chaudhry et al. / Interactive CardioVascular and Thoracic Surgery
156
Table 1: (Continued) Author, date, journal and country Study type (level of evidence)
Patient group
Retrospective cohort study (level 3 evidence)
Median follow-up time was 5.9 years
Outcomes
Key results
Comments
8-year freedom: 68.3 vs 68.1%
Differences in operative techniques, surgical experience and completeness in revascularization
Patients included involved non-redo, isolated CABG and those being a resident of Nova Scotia
Non-randomized retrospective, single-centre study
Following propensity-score matching, all preoperative characteristics were similar. Operatively, off-pump CABG involved lower number of distal anastomoses, fewer saphenous vein grafts, greater level of total arterial grafts and lower proportion of complete circumflex revascularization. 37 patients were converted to CPB, but were treated as intention-to-treat A single surgeon assessed the eligibility for surgery Kaplan–Meier analysis compared late outcomes Follow-up was available from longitudinal data through population records Selnes et al. (2009), Ann Thorac Surg, USA [14] Prospective cohort study (level 3 evidence)
395 patients were grouped into four categories having been admitted at the John Hopkins University School of Medicine or nearby hospitals between September 1997 and October 2003. Two of those categories: Off-pump: 75 (between March 1998 and October 2003) On-pump: 152 (between September 1997 and March 1999) The other two groups were non-surgical cardiac comparison and healthy heart comparison Mean duration of follow-up was 5.8 and 6.9 years for off-pump and on-pump groups, respectively. Follow-up was at 6 years
Deaths at 72 months (off-pump vs on-pump) Cardiac outcomes at 72 months (off-pump vs on-pump): Hospitalized Chest pain reported MI PCI New CABG AF New diagnosis of hypertension New diagnosis of diabetes Stroke TIA Cognitive outcomes
The inclusion criteria focused on the ability of patients to undergo neuro-psychological testing, which was the primary focus of the study. Other inclusion criteria involved being able to give written informed consent, sit upright, English-speaking and not mechanically ventilated
24 vs 17%
47 vs 58% 21 vs 22% 4 vs 6% 10 vs 10% 0 vs 0% 21 vs 28% 7 vs 13% 6 vs 9% 4 vs 6% 3 vs 7% No consistent differences in cognitive outcomes. Both groups had a lower baseline performance and a greater degree of cognitive decline than the healthy heart control group. Compared with baseline, neither group was dramatically worse
Conclusions: Vascular disease may have an impact of cognitive outcomes but management strategy is not a determinant for long-term outcomes Limitations: Non-randomized Bias involvement after selective drop-out of poorly performing subjects Different time frames for off-pump and on-pump for patient enrolment No direct comparison to detect significance was available for baseline characteristics and outcomes Off-pump patients from neighbouring hospitals were included
Direct comparison and statistical analysis were not available between the two groups in question for baseline variables
Continued
U.A.R. Chaudhry et al. / Interactive CardioVascular and Thoracic Surgery
157
Table 1: (Continued) Author, date, journal and country Study type (level of evidence)
Patient group
Outcomes
Key results
Comments
Sigurjónsson et al. (2012), Laeknabladid, Iceland [15]
720 consecutive patients who underwent CABG surgery at Landspítali-The National University Hospital, Reykjavik, Iceland, between January 2002 and December 2006: 207 off-pump 513 on-pump
5-year survival
Off-pump: 93% On-pump: 92% (P = 0.87)
Conclusions:
Retrospective cohort study (level 3 evidence)
Myocardial revascularization in Iceland yields similar long-term survival in Iceland Limitations: No details on collection of long-term data and method of determining choice of surgery
Mean follow-up was 5.0 years Concomitant surgical patients were excluded. A greater proportion of off-pump patients were male and had a higher NYHA class. 23 conversions to CPB were made, but treated as intention-to-treat
Non-randomized retrospective, single-centre study
Kaplan–Meier methods were used to estimate survival
Prospective, randomized, controlled trial (level 2 evidence)
281 CABG patients who had an Octopus stabilizer device utilized across three centres in Netherlands were enrolled between 1998 and 2000: 142 off-pump 139 on-pump Mean follow-up was at 5.2 years Patients referred for CABG surgery for the first time and deemed technically feasible were included. Emergency or concomitant surgical patients or those with a Q-wave MI or unable to provide informed consent or undergo neuro-psychological testing were excluded from the trial 10 off-pump patients underwent on-pump surgery, whereas 5 on-pump patients underwent off-pump surgery
All deaths (off-pump vs on-pump) at 5 years
Retrospective cohort study
CPB did not affect cognitive and cardiac outcomes at 5 years 21.1 vs 18.0% (P = 0.55) 4.9 vs 6.5% (P = 0.62) 17.7 vs 12.3% (P = 0.23) 1.4 vs 3.6% (P = 0.28) 7.7 vs 6.0% (P = 0.47)
Cognitive outcomes (off-pump vs on-pump) at 5 years
Standard definition of cognitive decline: 50.4 vs 50.4% (P > 0.99)
May not reflect CABG surgery population: younger population and lower preoperative risk
More conservative definition of cognitive decline: 33.3 vs 35.0% (P = 0.79)
Only reflected 11% of all CABG procedures during the study enrollment period
Limitations: 16 patients did not receive assigned treatment
Small number of participants QoL
The patient’s general practitioner initially provided follow-up data A total of 8580 patients underwent CABG surgery between July and December 2000 using two major New York State’s CSRS Database: 2640 off-pump 5940 on-pump
Conclusions:
Cardiovascular events (off-pump vs on-pump) at 5 years: Total MI Recurrent angina Stroke Repeat revascularization
The proportions of patients suffering an event during follow-up were compared using absolute risk difference, Fisher’s exact test and event curves if appropriate
Wu et al. (2012), Circ Cardiovasc Qual Outcomes, USA [17]
8.5 vs 6.5% (P = 0.65)
Survival
No differences in overall quality of life and in seven of the eight domains between the two groups, and both showed improvement following baseline levels. The domain of limitations due to physical health problems fared better following on-pump CABG Among 2631 matched pairs, 782 off-pump patients vs 725 on-pump patients died during follow-up: HR = 1.10 (95% CI: 0.99–1.21; P = 0.07) In a sensitivity analysis, which involved patients
Differences in baseline characteristics were not reported
Conclusions: No statistically significant differences in long-term mortality between on-pump and off-pump CABG Limitations:
Continued
BEST EVIDENCE TOPIC
van Dijk et al. (2007), JAMA, Netherlands [16]
U.A.R. Chaudhry et al. / Interactive CardioVascular and Thoracic Surgery
158
Table 1: (Continued) Author, date, journal and country Study type (level of evidence)
(level 3 evidence)
Patient group
Outcomes
Propensity score matching at a 1:1 ratio was conducted to create 2631 matched patients. Differences in baseline risk factors were no longer present following matching
Key results
Comments
treated by high-volume surgeons who performed CABG in >50% of their cases: HR = 1.12 (95% CI: 0.95–1.32; P = 0.17)
Retrospective, non-randomized CABG surgery performed more than 10 years ago
Median follow-up period was 7.2 years No exclusions were reported. 43 patients from the original cohort were converted to on-pump surgery Mortality data were available through databases Differences in survival were examined through Kaplan–Meier analysis and hazard ratio for death was obtained via Cox proportional hazards regression model AF: atrial fibrillation; BHACAS: beating heart against cardioplegic arrest studies; CABG: coronary artery bypass surgery; CHF: congestive heart failure; CI: confidence interval; COPD: chronic obstructive pulmonary disease; CPB: cardiopulmonary bypass; CSRS: cardiac surgery reporting system; CVA: cerebrovascular accident; EF: ejection fraction; eGFR: estimated glomerular filtration rate; HR: hazard ratio; ICU: intensive care unit; LV: left ventricular; LVEF: left ventricular ejection fraction; MACEs: major adverse cardiovascular events; MI: myocardial infarction; NYHA: New York Heart Association; OR: odds ratio; PCI: percutaneous coronary intervention; PVD: peripheral vascular disease; QoL: quality of life; RCT: randomized controlled trial; SMART: surgical management of arterial revascularization therapies; STS: Society of Thoracic Surgery; SVT: supraventricular arrhythmia; SWEDEHEART: Swedish Web-system for Enhancement and Development of Evidence-based care in Heart Disease Evaluated According to Recommended Therapies; TIA: transient ischaemic attack; VA: veterans affairs.
Search outcome Two hundred and fifty-six papers were found as a result of the reported search. From these, 16 papers were identified that provided the best evidence to answer the question. These are presented in Table 1.
RESULTS Angelini et al. [2] studied 401 patients from two separate prospective randomized clinical trials (RCT) with varying exclusion criteria, for a period of up to 6.4 years. There were no differences between off-pump and on-pump CABG surgery in: survival; survival-free from major adverse cardiac events (MACEs) (including myocardial infarction (MI), angina recurrence and need for revascularization); survival-free from death or MACEs; graft patency and quality of life (QoL). Bakaeen et al. [3] retrospectively utilized the large Veterans Affairs (VA) registry, to propensity score 8911 off-pump patients in a 1:3 manner. After a median follow-up of 6.7 years, overall survival following on-pump CABG surgery was significantly improved. Dalén et al. [4] retrospectively observed patients undergoing CABG operations in Sweden and found similar long-term survival
and other composite outcomes (mortality, rehospitalization for MI, heart failure or stroke) following 1:1 propensity-score matching and after a mean period of 7.1 years. Di Mauro et al. [5] were primarily concerned with the effect of preoperative renal function and its impact on postoperative outcomes. A single centre, retrospective analysis of the combined and matched cohort, including 1884 patients, showed no favourability to either technique after 7.5 years. Off-pump CABG did provide better late outcomes in patients with normal creatinine preoperatively. Filardo et al. [6] retrospectively analysed 8081 consecutive patients at a single institution. Despite adjustment of risk factors, patients following on-pump CABG had a significantly improved long-term survival. García Fuster et al. [7] retrospectively studied 1752 consecutive patients with varying degrees of preoperative renal function. Off-pump patients were followed up for a shorter period but did not have an advantage for late mortality. Gundry et al. [8] retrospectively enrolled 219 patients during 1989–90 for 7 years demonstrating similar survival but potentially increased reintervention rates following off-pump CABG surgery. Hueb et al. [9] prospectively compared 155 off-pump patients with 153 on-pump patients who met their inclusion/exclusion criteria for their RCT. Follow-up at 5 years revealed no differences in
U.A.R. Chaudhry et al. / Interactive CardioVascular and Thoracic Surgery
Clinical bottom line We sought to identify whether off-pump CABG conferred mortality or morbidity benefit over on-pump CABG in the long-term. Two of the 16 studies showed improved survival following on-pump surgery, whereas the remaining 14 showed no difference in late mortality (≥5 years). No differences were observed for other morbidity outcomes such as MACEs (including MI, recurrence of angina and heart failure or revascularization), stroke, graft patency, cognitive and QoL. A few limitations must be considered. Firstly, only four of the studies were prospective RCTs and sample sizes were small. Secondly, there is substantial variability between the studies in terms of patient cohorts, selection criteria, preoperative and operative risk factors, study methods and reporting of results, which may have an immeasurable effect on long-term outcomes. In summary, these results suggest that off-pump has long-term morbidity outcomes comparable with those of on-pump following CABG surgery. Long-term survival may also be similar to or slightly favourable towards on-pump CABG surgery. Procedural differences such as graft patency and completeness of revascularization as well as other factors need to be considered. Large, prospective RCTs
with adequate preoperative matching are now required to definitively measure and compare long-term outcomes. Conflict of interest: none declared.
REFERENCES [1] Dunning J, Prendergast B, Mackway-Jones K. Towards evidence-based medicine in cardiothoracic surgery: best BETS. Interact CardioVasc Thorac Surg 2003;2:405–9. [2] Angelini GD, Culliford L, Smith DK, Hamilton MC, Murphy GJ, Ascione R et al. Effects of on- and off-pump coronary artery surgery on graft patency, survival, and health-related quality of life: long-term follow-up of 2 randomized controlled trials. J Thorac Cardiovasc Surg 2009;137:295–303. [3] Bakaeen FG, Chu D, Kelly RF, Ward HB, Jessen ME, Chen GJ et al. Performing coronary artery bypass grafting off-pump may compromise long-term survival in a veteran population. Ann Thorac Surg 2013;95: 1952–8; discussion 1959–60. [4] Dalen M, Ivert T, Holzmann MJ, Sartipy U. Long-term survival after off-pump coronary artery bypass surgery: a Swedish nationwide cohort study. Ann Thorac Surg 2013;96:2054–60. [5] Di Mauro M, Gagliardi M, Iaco AL, Contini M, Bivona A, Bosco P et al. Does off-pump coronary surgery reduce postoperative acute renal failure? The importance of preoperative renal function. Ann Thorac Surg 2007;84: 1496–502. [6] Filardo G, Grayburn PA, Hamilton C, Hebeler RF Jr, Cooksey WB, Hamman B. Comparing long-term survival between patients undergoing off-pump and on-pump coronary artery bypass graft operations. Ann Thorac Surg 2011;92:571–7; discussion 577–8. [7] Garcia Fuster R, Paredes F, Garcia Pelaez A, Martin E, Canovas S, Gil O et al. Impact of increasing degrees of renal impairment on outcomes of coronary artery bypass grafting: the off-pump advantage. Eur J Cardiothorac Surg 2013;44:732–42. [8] Gundry SR, Romano MA, Shattuck OH, Razzouk AJ, Bailey LL. Seven-year follow-up of coronary artery bypasses performed with and without cardiopulmonary bypass. J Thorac Cardiovasc Surg 1998;115:1273–7; discussion 1277–8. [9] Hueb W, Lopes NH, Pereira AC, Hueb AC, Soares PR, Favarato D et al. Five-year follow-up of a randomized comparison between off-pump and on-pump stable multivessel coronary artery bypass grafting. The MASS III Trial. Circulation 2010;122:S48–52. [10] Jarvinen O, Hokkanen M, Huhtala H. Quality of life 12 years after on-pump and off-pump coronary artery bypass grafting. Coron Artery Dis 2013;24:663–8. [11] Puskas JD, Williams WH, O’Donnell R, Patterson RE, Sigman SR, Smith AS et al. Off-pump and on-pump coronary artery bypass grafting are associated with similar graft patency, myocardial ischemia, and freedom from reintervention: long-term follow-up of a randomized trial. Ann Thorac Surg 2011;91:1836–42; discussion 1842–3. [12] Raja SG, Husain M, Popescu FL, Chudasama D, Daley S, Amrani M. Does off-pump coronary artery bypass grafting negatively impact long-term survival and freedom from reintervention? Biomed Res Int 2013;2013: 602871. [13] Robertson MW, Buth KJ, Stewart KM, Wood JR, Sullivan JA, Hirsch GM et al. Complete revascularization is compromised in off-pump coronary artery bypass grafting. J Thorac Cardiovasc Surg 2013;145:992–8. [14] Selnes OA, Grega MA, Bailey MM, Pham LD, Zeger SL, Baumgartner WA et al. Do management strategies for coronary artery disease influence 6-year cognitive outcomes? Ann Thorac Surg 2009;88:445–54. [15] Sigurjonsson H, Helgadottir S, Oddsson SJ, Sigurdsson MI, Geirsson A, Arnorsson T et al. [Outcome of myocardial revascularisation in Iceland]. Laeknabladid 2012;98:451–6. [16] van Dijk D, Spoor M, Hijman R, Nathoe HM, Borst C, Jansen EW et al. Cognitive and cardiac outcomes 5 years after off-pump vs on-pump coronary artery bypass graft surgery. J AmMed Assoc 2007;297:701–8. [17] Wu C, Camacho FT, Culliford AT, Gold JP, Wechsler AS, Higgins RS et al. A comparison of long-term mortality for off-pump and on-pump coronary artery bypass graft surgery. Circ Cardiovasc Qual Outcomes 2012;5:76–84.
BEST EVIDENCE TOPIC
composite end-points including death, MI, need for revascularization, angina recurrence or stroke. Järvinen et al. [10] retrospectively analysed 56 off-pump patients vs 452 on-pump patients after 11.8 years to conclude no overall effect in survival or health-related QoL. Puskas et al. [11] prospectively reviewed 98 off-pump and 99 on-pump patients in their RCT who underwent CABG surgery by a single surgeon. There were similar incidences in survival, recurrent angina, percutaneous reintervention and graft patency after 7.5 years. Raja et al. [12] retrospectively propensity-score matched 307 patients in each group and showed that off-pump did not adversely impact on survival, readmission for a cardiac causes or need for reintervention after 10 years. Robertson et al. [13] demonstrated comparable results in longterm survival and freedom from cardiac readmissions for MI, heart failure or revascularization after propensity-matching 308 off-pump and on-pump patients. Selnes et al. [14] observed generally similar long-term cognitive and cardiac outcomes at 6 years having compared four different cohorts, two of which included 75 off-pump and 152 on-pump patients. Sigurjónsson et al. [15] retrospectively reviewed 720 consecutive patients at 5 years and showed similar long-term survival. van Dijk et al. [16] conducted an RCT involving 281 patients using an Octopus stabilizer device and at 5 years concluded that CPB had no effect on survival, MACEs (MI, angina recurrence, stroke, repeat revascularization), cognitive and QoL outcomes. Wu et al. [17] utilized New York’s Cardiac Surgery Reporting System database to 1:1 propensity-score match 2631 patients and follow-up for a period of 7.2 years. There was no significant difference in long-term mortality either overall or in their subgroup analysis of patients treated by high-volume surgeons who performed CABG in >50% of their cases.
159
