1. David J McCormack MFSTEd FRCSEd (CTh) Advanced Clinical Fellow
Cardiothoracic Journal Club ROOBY Trial (Veterans Affairs Randomised On/Off Bypass)
David J McCormack MFSTEd FRCSEd (CTh)
Advanced Clinical Fellow

2. What proportion of Waikato patients is ROOBY applicable to?5%
20%
50%
90%

8. Off-Pump Proposed Advantages Off-Pump Proposed Challenges
No aortic cannulation
No atrial cannulation
No cross-clamping
No blood-plastic interface
No cardioplegia
K+ load
fluid load
coronary air embolism
No risk of bypass machine failure -> air embolism
Less cost
Less equipment
Less staff
Needs skilled staff
Technically more difficult
Not all coronary arteries easily reached by technique
Potential for MI without cardioplegia
Incomplete revascularisation more frequent.
Worse long-term graft patency
Proposed neuropsychological benefit not shown in trial
Difficult in diffuse disease or small artery disease

9. Hypothesis or objective
‘We hypothesised there would be no difference between the on-pump and off-pump procedures for the two primary outcomes’
Primary Outcomes
Major morbidity and mortality at 1 month
Major morbidity and mortality at 1 year
Secondary Outcomes
Completeness of revascularisation
1-year graft patency
Neuropsychological test scores

10. Methods – Study Design
The ROOBY trial was a controlled, single-blind, randomized trial conducted from February 2002 through May 2008 at 18 VA medical centres.
Participating surgeons were required to document that they had performed at least 20 off-pump CABG surgeries, including some in which complete revascularization was performed for all vascular territories of the heart.
The pre-study off-pump experience of the surgeons averaged 120 cases (median, 50).
Sixteen sites had training programs in which cardiothoracic trainees (postgraduate year 6 to 10) were designated before randomization as the primary surgeon or first assistant surgeon.
Postoperatively, enrolled patients were followed every 2 months for up to 1 year.

11. Methods – Inclusion Criteria
Patients undergoing isolated CABG surgery
Scheduled or elective procedures

12. Methods – Exclusion Criteria
Any clinically significant valve disease (i.e., moderate, moderate-to-severe, or severe valve disease).
Requiring immediate surgery.
Small target vessels (<1.1 mm in internal diameter) or diffuse coronary disease.
Clinical reservations of the surgical team regarding patients with risk-factor profiles that predisposed them to an extremely high risk of an adverse event.
The inability or unwillingness of the patient to provide consent.

13. Methods – Randomisation
While in the preoperative holding area, patients underwent randomization.
Automated central telephone system in a blocked randomization scheme.
Assignments of the patients were balanced for each attending surgeon.

14. Methods – Surgical Details
A standard median sternotomy was performed in all patients.
Mandatory use of stabilisation device for off-pump surgery.
Conversion from the assigned procedure to the other procedure was performed when clinically necessary

15. Methods – Primary Endpoints
Short-term composite end point (<30d / discharge)
death
reoperation
new mechanical support
cardiac arrest,
coma,
Stroke
renal failure requiring dialysis
Long-term composite end point (30d – 1y)
death from any cause
nonfatal myocardial infarction
repeat revascularization
“Composite end points were selected for feasibility and fiscal reasons”

16. Methods – Secondary Endpoints
Completeness of revascularization (determined by the number of graft s performed as compared with the number planned)
Graft patency at 1 year
Scores on a battery of neuropsychological tests which were performed preoperatively and 1 year post- operatively.

17. Methods – Statistical Analysis
Power calculation indicated a sample size of patients
Two-tailed continuity-corrected chi-square test.
P value of 0.05 to indicate statistical significance for primary end points,
Power of 0.80, a 10% rate of loss to follow-up, and the ability to detect a reduction of 40% in the rate of the primary 1-year composite end point in the off-pump group as compared with the expected 8% rate in the on-pump group.
This sample size also allowed the detection of a 30% reduction in the primary short-term composite endpoint from the expected 14% rate in the on-pump group.

18. Methods – Statistical Analysis
The two primary end points – continuity corrected chi square tests.
Study-centre effects – Mantel–Haenszel chi-square tests.
Baseline characteristic and secondary outcomes were compared with the use of chi-square techniques and t-tests as appropriate.
Log-rank tests with On-Pump versus Off-Pump Coronary-Artery Bypass Surgery Kaplan–Meier curves were used to report time to death.

19. Results Consort Diagram

22. Results – Baseline Characteristics

23. Results – Baseline Characteristics

24. Results – Baseline Characteristics

25. Results – Primary Endpoints

26. Results – Primary Endpoints

27. Results – Primary Endpoints

28. Results – Secondary Enpoints

29. Results – Secondary Endpoints

30. Results – Secondary Endpoints
Less grafts performed in off-pump group.
More patients ‘undergrafted’ in off-pump group.

31. Conclusion
Our trial did not show any overall advantage to the use of the off-pump as compared with the on-pump cardiac surgical approach for coronary bypass.
Instead, there was a consistent trend toward better outcomes in patients undergoing the conventional on- pump CABG technique.
This including better 1-year composite outcomes and 1-year patency rates.
No significant differences between the off-pump and the on-pump techniques were identified in neuropsychological outcomes or the use of major resources.

32. Strengths of the study
Prospective, controlled, randomised and blinded trial.
Large multicentre trial.
Achieved sufficient patients to power study.
Impressive angiographic follow up
Good clinical follow up rate (XX %)

33. Limitations of the study
Surgical intervention not performed by experts.
Minimum of 20 cases with some cases grafting all coronary territories.
Self documentation of experience (expertise).
Residents cases included.
Large cohort of patients excluded
Limited description of 3282 excluded patients (44% of eligible patients)
Limited applicability to our local cohort.
No standard operation in either limb.

34. Limitations of the study
No description of reason for cross over
Off-pump -> on-pump (12.4%)
On-pump -> off-pump (3.6%)
Historic study
CORONARY Trial
On/Off Trial
Cologne Metanlysis
GOPCABE

35. General Discussion Points
Landmark paper – largely negative for off-pump.
Less grafts per patient in off-pump surgery (2.9 vs 3.0).
What level of experience should be required?
Should resident cases be performed?
Intention to treat analysis not used.

36. What proportion of Waikato Patients is ROOBY applicable to?5%
20%
50%
90%