Author + information
- Received June 16, 2008
- Revision received October 9, 2008
- Accepted October 14, 2008
- Published online January 20, 2009.
- David X. Zhao, MD, FACC,
- Marzia Leacche, MD,
- Jorge M. Balaguer, MD,
- Konstantinos D. Boudoulas, MD,
- Julie A. Damp, MD,
- James P. Greelish, MD,
- John G. Byrne, MD, FACC⁎ (, )
- Writing Group on behalf of the Cardiac Surgery, Cardiac Anesthesiology, and Interventional Cardiology Groups at the Vanderbilt Heart and Vascular Institute
- ↵⁎Reprint requests and correspondence:
Dr. John G. Byrne, Vanderbilt Heart and Vascular Institute, Department of Cardiac Surgery, 1215 21st Avenue South, MCE–North Tower–Suite 5025, Nashville, Tennessee 37232-8802
Objectives This study sought to report our experience with a routine completion angiogram after coronary artery bypass surgery (CABG) and simultaneous (1-stop) percutaneous coronary intervention (PCI) at the time of CABG performed in the hybrid catheterization laboratory/operating room.
Background The value of a routine completion angiogram after CABG and 1-stop hybrid CABG/PCI remains unresolved.
Methods Between April 2005 and July 2007, 366 consecutive patients underwent CABG surgery, with (n = 112) or without (n = 254) concomitant 1-stop PCI (hybrid), all with completion angiography before chest closure. Among the 112 1-stop hybrid CABG/PCI patients, 67 (60%) underwent a planned hybrid procedure based on pre-operative assessment, whereas 45 (40%) underwent open-chest PCI (unplanned hybrid) based on intraoperative findings.
Results Among the 796 CABG grafts (345 left internal mammary artery, 12 right internal mammary artery/radial, and 439 veins), 97 (12%) angiographic defects were identified. Defects were repaired with either a minor adjustment of the graft (n = 22, 2.8%), with intraoperative open-chest PCI (unplanned hybrid, n = 48, 6%) or with traditional surgical revision (n = 27, 3.4%). Hybrid patients had clinical outcomes similar to standard CABG patients.
Conclusions Routine completion angiography detected 12% of grafts with important angiographic defects. One-stop hybrid coronary revascularization is reasonable, safe, and feasible. Combining the tools of the catheterization laboratory and operating room greatly enhances the options available to the surgeon and cardiologist for patients with complex coronary artery disease.
- hybrid coronary revascularization
- coronary angiography
- percutaneous coronary intervention
- coronary artery bypass surgery
With the exception of coronary artery bypass grafting (CABG) surgery, every major cardiovascular intervention—including percutaneous coronary intervention (PCI), heart valve surgery, and endovascular and traditional open vascular surgery—is accompanied by completion imaging to ensure procedural results. However, it has not been the standard of care to perform completion diagnostic imaging after CABG surgery, primarily because it has been previously either too cumbersome or too logistically difficult. Because modern CABG studies report early (12-month) graft failure rates ranging from 7.6% to 30% (1–3), and because technical errors are thought to be largely responsible, we now routinely perform open-chest completion coronary graft angiography in the hybrid suite with the goal of ensuring immediate graft quality control. We believe this may result in improved short-term and long-term graft patency.
In addition to routine completion angiography, the hybrid suite also allows the surgeon and interventional cardiologist, in a collaborative environment, to offer patients traditional surgery and/or PCI during the same operative session. Whereas in diabetic patients, CABG with the left internal mammary artery (LIMA) to the left anterior descending artery (LAD) provides outstanding long-term results and survival benefit over PCI or medical management (4–8), direct comparison of saphenous vein grafts (SVGs) with PCI has not been documented. Because recent studies report close to a 30% SVG occlusion rate within 12 to 18 months (1), and because the 12- to 18-month restenosis rate with drug-eluting stents (DES) seems superior to SVG (9–13), we hypothesized that a hybrid surgical approach with a LIMA graft to the LAD and PCI approach with DES to the right and/or the left circumflex coronary arteries is a valid alternative to standard CABG in selected patients.
Conventional bypass surgery is performed in an operating room (OR), whereas PCI is performed in a cardiac catheterization laboratory; the physical separation of these 2 procedural suites poses logistical limitations and potential risk for 2-staged hybrid revascularization. At Vanderbilt, we have constructed a hybrid operating room that combines the tools of a cardiac catheterization laboratory and an operating room into 1 procedural suite.
The hybrid procedural suite
In 2005, the hybrid room was constructed in the cardiac catheterization laboratory suites; the hybrid catheterization laboratory/OR is a fully functional catheterization laboratory as well as a fully functional OR (Fig. 1). In this 1 procedural suite, every cardiovascular procedure can be performed, from aortic arch reconstruction to routine catheter-based intervention.
After obtaining institutional review board approval, we retrospectively collected data on 366 consecutive patients who underwent CABG surgery with (n = 112) or without (n = 254) concomitant 1-stop PCI from April 2005 to June 2007. All patients enrolled in the study were referred for CABG surgery based on clinical indications. Consent was obtained before the procedure for intraoperative completion angiography. Before chest closure, an interventional cardiologist, or a surgeon under the supervision of a cardiologist, performed coronary graft angiography. If angiographic defects were identified, revision was carried out surgically or percutaneously based on the angiographic findings. Repeat angiography was then performed to ensure correction of these defects before chest closure. For analysis, 2 independent cardiologists who were not involved in the care of the patient reviewed each coronary angiogram. Angiographic defects were divided into 3 categories based on the location within the graft or target vessel: 1) conduit defects; 2) anastomotic defects; and 3) target vessel errors. Defect repair methods were subdivided into 3 categories based on repair technique: 1) a minor but important adjustment of the graft without additional cardiopulmonary bypass or aortic cross-clamping; 2) intraoperative open-chest PCI; or 3) traditional surgical revision requiring additional cardiopulmonary bypass or aortic cross-clamping or off-pump revision of surgical anastomosis. Sixty-seven patients (18%) underwent planned hybrid revascularization. For these patients, both the cardiac surgeon and the interventional cardiologist reviewed the coronary angiogram in advance. If technically feasible for both surgical and percutaneous approaches, the hybrid approach was presented to the patient and a separate consent was obtained. Patients who underwent planned a hybrid procedure received 300 mg of clopidogrel orally immediately before entering the OR. Unplanned hybrid procedure patients, those in whom the decision to perform PCI was based on intraoperative findings, received 300 mg of clopidogrel via nasogastric tube at the time the decision for open-chest PCI was made. All hybrid procedure patients received clopidogrel 75 mg/day for >6 months. All patients received daily 325-mg aspirin after the surgery. Before heparinization, a femoral sheath, for a later catheter-based procedure, was placed in the left or right femoral artery using the Seldinger technique. For on-pump CABG surgery, 300-U/kg heparin was given to obtain an activated clotting time >400 s. For off-pump CABG, the dose of heparin was 125 U/kg. At the end of the surgery, the sheath was typically removed after the heparin was reversed with protamine, and a Syvek patch (Marine Polymer Technologies, Danvers, Massachusetts) was applied with 12 min of manual compression of the groin.
Data are expressed as mean and standard deviation or percentage. For continuous variables, a Mann-Whitney U test was used and data were expressed as median. The 2-tailed Fisher exact test was used for categorical variables. Data analysis and statistics were performed with STATA 9.0 for Windows (Stata Corp., College Station, Texas).
Completion angiography after CABG
The entire cohort of patients underwent completion angiography regardless of whether or not concomitant PCI (hybrid) was performed. Pre-operative characteristics are listed in Table 1, a summary of grafts performed in Table 2, and operative characteristics in Table 3. There were 97 of 796 (12%) angiographic defects in 89 patients. There were 54 conduit defects (6.8%), 30 anastomotic defects (3.7%), and 13 target vessel errors (1.6%). A minor adjustment was required in 22 (2.8%) grafts, intraoperative open-chest PCI in 48 (6%), and surgical revision in 27 (3.4%). Twenty-five of 345 (7%) LIMAs had a major defect, of which 10 were located in the conduit and 15 at the distal anastomosis. Thirty-seven of 445 (8%) SVGs had a major defect, of which 25 were located in the conduit. Angiographic findings versus type of revision method are summarized in Table 4.
Figures 2 and 3⇓⇓ show examples of defects found on completion angiography: a chest tube obstructing an SVG before (Fig. 2A) and after its removal (Fig. 2B); the effect of an SVG kink of a (reversed) vein valve, before (Fig. 3A) and after PCI (Fig. 3B). Intraoperative angiography added an average 20 ± 12 min to the surgery and 108 ± 86 ml contrast (Visipaque, GE Healthcare, Chalfont St. Giles, United Kingdom). Table 5 summarizes post-operative characteristics. In the earlier series of patients (the first 15 months of the study) the rate of angiographic defects was 15%, whereas in the last 12 months it was 11% (p = 0.3).
There were 67 planned and 45 unplanned hybrid procedure patients. DES were used in 95 of 112 (84%) and bare-metal stents (BMS) in 9 of 112 (8%). A combination of DES and BMS were used in 8 of 112 (7%). Tables 1, 3, and 5 summarize the pre-operative, intraoperative, and post-operative characteristics of hybrid (n = 112) versus standard CABG (n = 254) patients. There was no significant difference in operative mortality between hybrid (n = 3, 2.6%) versus standard CABG (n = 4, 1.5%) (p = 0.33, 95% confidence interval: −0.060 to −0.019). The indications for planned hybrid procedures (n = 67) included an attempt to minimize surgical risk (n = 32), poor conduits (n = 3), ungraftable vessels (n = 29), or stenting of the left subclavian artery (n = 3) (Table 6). In unplanned hybrid procedures (n = 45), the decision for open-chest PCI was based on intraoperative findings. The indications for unplanned PCI were graft defects (n = 43) identified by routine completion angiogram, poor conduits (n = 1), or ungraftable vessels (n = 2). One patient underwent both (planned left subclavian stenting and unplanned PCI for graft defect). These 43 PCI-treated defects are a subset of the 97 total defects identified by intraoperative angiography. The remaining 54 defects were treated surgically. The timing and sequence of the hybrid procedures are summarized in Table 6.
Two patients (0.5%) sustained catheter-related complications: 1 unplanned hybrid procedure patient developed a small perforation of the posterior descending artery, managed percutaneously, and 1 planned hybrid procedure patient sustained an injury of the femoral artery requiring open surgical repair because of an enlarging hematoma. One of 112 hybrid procedure patients (1%) sustained acute stent thrombosis and died. Another patient who underwent balloon angioplasty of a small posterior descending artery had a reocclusion after the procedure and underwent reopening of the vessel with successful stenting.
Intraoperative completion angiography
The clinical impact of SVG failure was documented in a randomized study of 3,041 patients, wherein the common end point of death and new myocardial infarction was more common if SVG failure was present (14% vs. 1%) (1). The impact of SVG failure was even more pronounced if the need for revascularization was added to death and new myocardial infarction (26% vs. 2%) (1). In the same randomized trial, 1-year angiography documented that the SVG graft failure rate was 30%, whereas the LIMA failure rate was 8% (1). Technical errors are believed to be responsible for early graft failure, which is present in 5% to 20% of patients at discharge (14). In the current study, we identified that 12% of all grafts had some form of angiographic defect, which may have otherwise been missed intraoperatively. Completion angiography changed surgeon behavior in the operating room by providing immediate feedback on issues that were difficult to appreciate with the naked eye. We became more selective regarding the significance of apparently minor (to the naked eye) kinks in vein grafts, the effect of (reversed) vein valves on antegrade flow, and the effect of conduit-target mismatch, in which the SVG diameter is significantly greater than the target vessel. These considerations were perhaps only minor considerations before our use of routine completion angiography. Although the technique for CABG surgery has largely remained the same over the last 35+ years, the technical difficulties of performing CABG have increased over time, because of the presence of more diffuse coronary artery disease and smaller coronary targets, especially in diabetic patients, the very population in whom CABG affords the most benefit.
We also observed a 7% (25 of 345) incidence of LIMA angiographic findings, remarkably similar to the findings in the PREVENT IV (Project of Ex-Vivo Vein Graft Engineering via Transfection) trial (1), in which 8% of LIMA grafts had a significant defect at 1 year. We believe the 8% defect rate in the PREVENT IV trial likely represents findings that could have been corrected at the time of surgery. Our data, together with the PREVENT IV data, further support our argument that intraoperative imaging should perhaps eventually become the standard of care for CABG surgery. Our defect rate decreased from 15% to 11% during the study period, and we anticipate that we will continue to improve as we continue to measure our performance. Long-term (12-month) graft computed tomography angiographic follow-up for these patients is in progress, and we hypothesize that correction of these defects at the time of surgery may help reduce the rate of graft failure.
The LIMA-LAD bypass has been shown to be the most durable and proven treatment to significantly reduce the subsequent risk of death, myocardial infarction, recurrent angina, and the need of further revascularization (15,16). However, the benefit of surgical revascularization after LIMA-LAD remains less clear (8). This is reflected in the lack of survival benefit in reoperative CABG surgery when the LIMA is already patent (17). Thus, as PCI continues to improve, the future of CABG may possibly be limited to the LIMA-LAD performed through a small anterior thoracotomy with PCI to non-LAD vessels (18).
Whether non-LAD vessels should be grafted, stented, or left alone will need to be determined based on conduit and target vessel quality. These considerations suggest that selected patients might benefit from a combined planned hybrid approach. In our series this planned hybrid approach was used in 67 (18%) patients whom the cardiologist and cardiac surgeon believed would be better served by a combined approach rather than by conventional CABG surgery.
For planned hybrid procedure patients, in centers without a hybrid suite, a 2-stage approach, using 2 suites (a traditional OR and traditional catheterization laboratory), is certainly reasonable. In a 2-stage hybrid procedure, either CABG or PCI is performed first followed days later by the other procedure. The CABG first staging is probably preferable because it avoids clopidogrel-related bleeding complications during CABG, has the advantage of a protected environment with a (presumably perfectly patent) LIMA-LAD graft, and the LIMA graft patency can be verified at the time of PCI (although any defects identified days after CABG could likely only be addressed with PCI). The only disadvantage of the CABG first staging is that, in the event of PCI complication/failure, a second and much higher-risk operation would need to be performed. Whether CABG or PCI is performed first, in both scenarios, 2 procedures need to be performed with 2 operative costs and 2 operative teams. A 1-stop procedure, as described in the current report, has the advantage of avoiding the potential problems related to 2 separate procedures and hands-off, but also any complications can be resolved in 1 setting, and graft patency can be confirmed immediately. Also, many patients just do not want to undergo 2 procedures.
Forty-five (40%) hybrid procedure patients had an unplanned hybrid procedure that was performed based on intraoperative findings either because of graft defects, a poor conduit, or an ungraftable coronary vessel. Had these operations been performed in a standard operating room, these patients may have left the OR with a suboptimal result. Once a defect was discovered on completion angiogram, the decision for unplanned PCI versus surgical revision was made by consultation between a cardiac surgeon and an interventional cardiologist. PCI was the preferred method for revising angiographic defects when they were located in the conduit or target vessel, whereas traditional surgical revision was more commonly used for anastomotic revisions. In a traditional setting, without PCI as an option—provided a defect could be identified—the only choice would have been to harvest another conduit and/or repeat the anastomosis with the potential for another conduit of poor quality. Consistent with prior reports, we used DES in the revision of an SVG conduit rather than BMS, which results in lower restenosis rates (19).
Our inability to document a lower operative risk in hybrid CABG/PCI patients, in contrast to our prior study on hybrid valve/PCI patients in which we showed a reduction in the observed mortality (20), is probably attributable to the fact that CABG is, in general, a lower-risk endeavor, and therefore, it is more difficult to prove a benefit.
Bleeding complications versus stent thrombosis
A major concern with the hybrid approach is the need for antiplatelet agents. In a prior study on hybrid valve/PCI patients, we documented an increased amount of chest tube output, reoperation for bleeding, and transfusions requirements (20). In the present series, 19% of the patients underwent their procedures while on clopidogrel, and 31% received clopidogrel as part of their hybrid procedure. We found similar chest tube output drainage, and the rate of reoperation for bleeding was not different between groups.
Acute stent thrombosis is an important complication and a potential major limitation of the hybrid procedure. In our series, 1 of 112 (1%) patients developed acute in-stent thrombosis and died. This was an elderly man with severe ascending aortic disease in whom a tight left main lesion was present. We performed LIMA-LAD off-pump followed by left main PCI. The patient developed acute stent thrombosis after heparin reversal and died. Although this rate of stent thrombosis was within the reported rate of acute stent thrombosis (21), we believe that antiplatelet strategies need particular attention in hybrid procedure patients because of the need to reverse heparin with protamine. Because the antiplatelet effect of clopidogrel is both time and dose dependent with a maximal inhibition of platelet aggregation 4 to 24 h after a dose of 300 to 600 mg (22,23), we currently administer the dose of clopidogrel immediately before surgery in the holding area for a planned hybrid procedure or, for an unplanned hybrid procedure, during the surgery at the time the decision for PCI is made. We also use a 300-mg loading dose rather than 600 mg, as given for isolated PCI, as a compromise to balance the risk of significant bleeding against stent thrombosis. The antiplatelet effects of cardiopulmonary bypass may be protective against the risk of stent thrombosis (24). Thus, antiplatelet strategies may need to be more intense with off-pump hybrid CABG/PCI patients. Also, patient response to clopidogrel may differ and pre-operative platelet studies may help to identify patients with potential clopidogrel resistance. An alternative to heparin in selected hybrid procedure patients is bivalirudin (25). Currently, however, there is no consensus on the best monitoring strategy for bivalirudin anticoagulation on cardiopulmonary bypass.
Cardiac catheterization–related complications
We placed the femoral arterial sheath before heparinization to avoid potential access site hematomas, and we report 1 patient (0.5%) with arterial access complication. We also avoided completion angiography in patients with advanced vascular disease (grade 4 to 5 atheroma in the descending thoracic aorta on transesophageal echocardiogram). Although we report a 1% incidence of new stroke, it is difficult to determine whether stroke was catheter or surgery related. Our data support that angiography should be performed in the vast majority of patients undergoing CABG in the hybrid suite. We also acknowledge that there are important differences between diagnostic angiography and completion angiography. In the latter, there are fresh proximal and distal anastomoses and the hemodynamic changes associated with the post-operative condition. However, with careful technique, our interventional cardiologists were able to perform completion angiography with a very low rate of catheter-related complications.
We were exceedingly meticulous about sterile technique, and we report a sternal wound infection rate of 1%. Because the angiograms are performed with an open chest, and the camera is moved into place over the open chest, we have adopted the practice of quintuple draping (3 drapes on the open wound and 2 on the camera). Because the hybrid suite contains many more moving parts, and more personnel are involved, infection control is exceedingly important. Team discipline, control of traffic, and policing of behavior are crucial for a successful hybrid program.
Contrast nephropathy is another major concern of intraoperative angiography and a hybrid approach. We document a low incidence of new renal failure (4%). Our experience was consistent with previous reports that angiography after cardiac surgery is safe without increased risk of contrast nephropathy (26). Moreover, in the hybrid group, we did not observe an increased rate of acute renal failure or an increase of serum creatinine, despite greater contrast load compared with the completion angiography group. We currently do not perform a routine completion angiogram if the pre-operative creatinine value is >2.0 mg/dl.
Added time and cost
As expected, there was a learning curve and much of the added time early in the series was related to our caution with all of the moving parts, all of the different personnel involved, as well as our goal of training surgeons in catheter-based techniques. However, as surgeons become more adept at completion angiography, this added time will likely be reduced significantly.
Building hybrid suites is certainly more expensive than building either a standard operating room or a standard catheterization laboratory. However, this all-purpose hybrid suite can be used for every cardiovascular procedure, including traditional cardiac surgery, traditional vascular surgery, endovascular aortic procedures, electrophysiology procedures, or interventional coronary procedures. There is also a financial benefit of synergy of teams, cross-training personnel, with overlapping of the OR team with the interventional team.
This was a retrospective study with inherent selection bias. The decision to perform the hybrid procedure was made by the surgeon and the interventional cardiologist on an individual and highly selective basis.
Implications and the hybrid paradigm
Our data suggest that routine completion graft imaging should eventually become the standard of care in CABG surgery. Furthermore, with the increased complexity of patients referred for cardiac surgery, a team approach, combining traditional cardiac surgical techniques and PCI, may be beneficial, especially in high-risk populations.
The technology used in a hybrid OR is well established, but the ability to combine them into 1 operative suite can be challenging. This requires a high level of collaboration between interventional cardiologists and cardiac surgeons. The willingness and ability to form this collaborative working environment seems to be the greatest barrier (27).
The authors would like to thank the following contributing authors:
Rashid M. Ahmad, MD
Stephen K. Ball, MD
John H. Cleator, MD, FACC
Robert J. Deegan, MD, PhD
Susan S. Eagle, MD
Pete P. Fong, MD, FACC
Joseph L. Fredi, MD, FACC
Steven J. Hoff, MD
Henry S. Jennings, III, MD, FACC
John A. McPherson, MD, FACC
Robert N. Piana, MD, FACC
Mias Pretorius, MD
Mark A. Robbins, MD, FACC
David A. Slosky, MD, FACC
Annemarie Thompson, MD
For a list of definitions, please see the online version of this article.
Funding and technical support for this project were provided by the Vanderbilt Heart and Vascular Institute and the Cardiac Surgery Research Fund of Vanderbilt University Medical Center, Nashville, Tennessee.
- Abbreviations and Acronyms
- bare-metal stent(s)
- coronary artery bypass grafting
- drug-eluting stent(s)
- left anterior descending artery
- left internal mammary artery
- operating room
- percutaneous coronary intervention
- saphenous vein graft
- Received June 16, 2008.
- Revision received October 9, 2008.
- Accepted October 14, 2008.
- American College of Cardiology Foundation
- Hoffman S.N.,
- TenBrook J.A.,
- Wolf M.P.,
- Pauker S.G.,
- Salem D.N.,
- Wong J.B.
- (2007) The final 10-year follow-up results from the BARI randomized trial. J Am Coll Cardiol 49:1600–1606.
- Serruys P.W.,
- Ong A.T.,
- van Herwerden L.A.,
- et al.
- Grube E.,
- Silber S.,
- Hauptmann K.E.,
- et al.
- Colombo A.,
- Chieffo A.
- Byrne J.G.,
- Leacche M.,
- Unic D.,
- et al.
- Daemen J.,
- Serruys P.W.
- Patrono C.,
- Bachmann F.,
- Baigent C.,
- et al.
- Gurbel P.A.,
- Bliden K.P.,
- Hiatt B.L.,
- O'Connor C.M.
- Schachner T.,
- Bonatti J.,
- Bonaros N.,
- et al.
- Kotter J.P.