Author + information
- Received July 22, 2004
- Revision received September 18, 2004
- Accepted September 21, 2004
- Published online January 4, 2005.
- John G. Byrne, MD, FACC*,* (, )
- Marzia Leacche, MD*,
- Daniel Unic, MD*,
- James D. Rawn, MD*,
- Daniel I. Simon, MD, FACC†,
- Campbell D. Rogers, MD, FACC† and
- Lawrence H. Cohn, MD, FACC*
- ↵*Reprint requests and correspondence:
Dr. John G. Byrne, Department of Cardiac Surgery, Vanderbilt University Medical Center, Nashville, Tennessee 37232-8815
Presented at the Annual Scientific Session of the American College of Cardiology, March 7 to 10, 2004, New Orleans, Louisiana.
Objectives The goal of this study was to determine if a “hybrid” approach to the treatment of complex combined coronary and valve disease is superior to the results predicted by a Society of Thoracic Surgeons' (STS) algorithm with conventional coronary artery bypass graft (CABG)/valve surgery in high-risk patients.
Background With advancements in percutaneous coronary interventions (PCIs), some patients requiring coronary revascularization and valve surgery may benefit from a hybrid approach involving initial planned PCI followed by valve surgery, rather than conventional CABG/valve surgery.
Methods We retrospectively analyzed 26 consecutive patients with coronary artery and valve disease who underwent planned initial PCI followed by valve surgery during the same hospital stay between September 1997 and August 2003. We calculated the predicted mortality at the time of PCIand compared it with the observed mortality.
Results There were 12 male and 14 female patients with a median age of 72 years (range 53 to 91 years). Balloon angioplasty was performed in all patients, followed by stenting in 22 (85%) patients. Within a median of 5 days (range 0 to 14 days), 15 patients (58%) underwent primary and 11 patients (42%) underwent re-operative valve surgery. Operative mortality was 1 of 26 patients (3.8%), dramatically lower than the STS-predicted mortality of 22%. Median blood loss was 900 ml, and 22 patients (85%) required blood transfusions. Survival at 1, 3, and 5 years was 78%, 56%, and 44%, respectively.
Conclusions Hybrid initial PCI followed by staged valve surgery represents an excellent alternative to conventional CABG/valve surgery in some high-risk patients, particularly those who present in shock after myocardial infarction. Lower mortality rates come at the cost of more bleeding and transfusion requirements.
Emergency coronary artery bypass grafting (CABG) combined with valve surgery as well as some re-operative valve/CABG operations are high-risk procedures with a modern operative mortality for each approaching 20% (1–4). The optimal treatment for these patients is not firmly established (5).
With our recently developed techniques for minimally invasive primary (6) and re-operative (7) valve surgery, the “hybrid” approach makes particular sense. We have previously shown that the minimally invasive approach for re-operative valve surgery minimizes the possibility of graft injury and avoids unnecessary mediastinal dissection, thereby decreasing bleeding complications (7–9). In the presence of verified coronary or graft disease, an initial staged percutaneous coronary intervention (PCI) might help avoid excessive surgery, enabling patients to benefit fully from minimally invasive valve surgery.
We hypothesized that the hybrid approach—consisting of planned initial PCI of the culprit native coronary artery or CABG followed by valvular surgery after patient stabilization—would have better outcomes than those predicted by the Society of Thoracic Surgeons (STS) algorithm for combined CABG/valve surgery after acute coronary syndromes (ACS) and patients undergoing complex valve reoperations.
After obtaining the institutional review board approval, we retrospectively reviewed the medical records of 26 consecutive patients who underwent PCI followed by valve surgery from September 1997 to August 2003. The decision for the hybrid approach was made jointly by the referring and interventional cardiologist and the cardiac surgeon. In all patients, coronary and valvular lesions were documented by diagnostic catheterization and echocardiography, respectively, and hemodynamic status was established in the catheterization laboratory. After a treatment plan was established, the interventional cardiologist proceeded with PCI (balloon angioplasty with or without stenting) of the culprit lesion (native vessels or saphenous vein grafts). After a period of hemodynamic stabilization during which patients received antiaggregation treatment, valve surgery was performed, typically within a week of PCI. Patients requiring re-operative valvular procedures in which concomitant CABG was not needed were approached through a minimally invasive incision when appropriate (7,8).
Definitions of pre-PCI patient characteristics
Acute myocardial infarction (AMI): creatine-kinase values >2 times the upper limit of the normal, or troponin T or I maximal level >1.0 ng/ml on one occasion within 24 h of clinical event and changes from the baseline or of the ST-T in serial electrocardiograms; unstable angina: angina at rest or new onset of exertional angina of at least Canadian Cardiovascular Society class III; low cardiac output syndrome: cardiac index ≤2.0 l/min/m2requiring inotropic support to maintain a systolic pressure >90 mm Hg for at least 30 min, and/or placement of an intra-aortic balloon pump; cardiogenic shock: systolic blood pressure <80 mm Hg or cardiac index <1.8 despite maximal treatment; renal insufficiency: baseline creatinine level ≥1.5 mg/dl.
Definitions of perioperative events
Hospital mortality: death for any reason occurring within 30 days after the surgery or after 30 days occurring during the same hospitalization; low cardiac output syndrome: cardiac index ≤2.0 l/min/m2, requiring inotropic support to maintain a systolic pressure >90 mm Hg for at least 30 min, and/or placement of an intra-aortic balloon pump and/or ventricular assist device; bleeding: necessity of re-exploration of the thorax for suspected bleeding during the postoperative period; stroke: evidence in the postoperative period of a new central neurological deficit persisting for >72 h, whereas, if the neurological deficit resolved in 72 h, it is considered a transient ischemic attack.
STS risk algorithm for predicted surgical mortality at the time of PCI
Predicted mortality was calculated utilizing algorithms for the pre-operative calculation of risk for AVR + CABG and MVR + CABG provided by the STS. These algorithms are based on data collected between 1994 and 1997 from approximately 100,000 of these combined CABG/valve procedures. Algorithms for outcomes since 1997 are not yet available. However, unadjusted and risk-adjusted mortalities for these operations have not significantly changed.
Data are expressed as the median value or percentage. Long-term survival rates were calculated using the Kaplan-Meier method. Statistical software package STATA 7.0 for Windows (STATA, College Station, Texas) was used to calculate our results.
From September 1997 to August 2003, 26 patients underwent staged initial PCI followed by valve surgery; pre-PCI data documenting the high-risk profile of this patient cohort are summarized in Table 1.Indications for PCI are summarized in Table 2.The median number of vessels undergoing PCI was 2 (range 1 to 3), with balloon angioplasty performed in 26 patients (100%) and stenting in 22 of 26 patients (85%). Drug-eluting stents were used in 3 of 26 patients (12%). Distribution of vessels undergoing PCI was left anterior descending coronary in 30%, saphenous vein grafts in 11%, LIMA-left anterior descending coronary artery graft in 4%, obtuse marginal coronary in 19%, left circumflex coronary in 23%, and right coronary artery in 19%. Surgical procedures are summarized in Table 3.Concomitant CABG to non-culprit vessels was performed in 10 of 26 patients (38%): 9 primary mitral and 1 primary aortic. No re-operative patients underwent concomitant CABG. Median cardiopulmonary bypass time was 141 min (range 60 to 249 min) with a median aortic clamp time of 91 min (range 10 to 142 min). Twenty-six patients (100%) were receiving aspirin at the time of surgery, and 18 of 26 patients (69%) were receiving clopidogrel. Postoperative bleeding and other complications were common (Table 4).
One patient (3.8%) died during the postoperative course. This 75-year-old patient developed acute papillary muscle rupture and cardiogenic shock after AMI. The coronary lesion was successfully managed by PCI and stenting. Immediate mitral valve replacement was required because of severe mitral regurgitation. Massive gastrointestinal bleeding on postoperative day 28 was the immediate cause of death.
The median predicted mortality for conventional valve/CABG surgery at the time of PCIfor this cohort of patients was 22% (range 3.5% to 63.5%) according to the STS risk stratification algorithms (Fig. 1).Survival at 1, 3, and 5 years was 78%, 56%, and 44%, respectively (Fig. 2).There were no cases of subacute stent thrombosis or postoperative Q-wave myocardial infarction.
The principal finding in this report is that “hybrid” initial PCI followed by staged valve surgery represents an excellent alternative to conventional high-risk CABG/valve surgery in patients who present after an ACS, and in some patients who require complex re-operative valve surgery. A lower operative mortality rate (3.8% vs. 22%) compared with that predicted by the STS algorithm was achieved at the cost of increased rates of reoperation for bleeding, greater blood loss, and higher transfusion requirements.
Aspirin and platelet inhibitors are universal adjuvant therapies to PCI. All of our patients received aspirin, and 69% received additional clopidogrel in the period between PCI and surgery. The median blood loss after surgery for the entire cohort was almost 1 l. A total of 85% received some type of blood transfusion, and our re-operation rate for bleeding was 8%. The only death in the series was related, in part, to a bleeding complication. Although our study documented a high incidence of bleeding complications, this was not accompanied by high mortality. Despite potent antiplatelet agents, a sensible algorithm for dosing and timing may help minimize bleeding risk. Subacute stent thrombosis, as reported after noncardiac surgery where antiplatelet agents are stopped (10), was not observed in this study, likely the result of the antiplatelet effects of cardiopulmonary bypass as well as the routine use of postoperative clopidogrel.
In patients with increased risk for bleeding, the choice of surgical approach gains greater importance. We have previously documented benefits in terms of decreased blood loss and transfusion requirements in patients undergoing both primary and re-operative minimally invasive valvular surgery (6–9). The minimally invasive approach is particularly useful in re-operative surgery and in some stabilized patients after an ACS. Thus, a tailored approach to meet specific patient needs is required.
Because most patients in this study (92%) presented with an ACS, the hybrid approach would seem of obvious benefit to most clinicians, with the goal of first stabilizing the coronary lesions and then, during the same hospital admission, addressing the valve lesion. Although performing re-operative valve surgery in the presence of aspirin and clopidogrel might seem anathema to most surgeons, in two patients we judged that the risk posed by dissecting the heart for concomitant CABG—in the absenceof clopidogrel—was greater than the risk posed by bleeding in the setting of powerful antiplatelet medications. Ideally, these patients would be best managed by either a longer staging duration so that the clopidogrel can be stopped (three to six months with drug-eluting stents), or by a very short staging duration (under 6 h), so that clopidogrel's actions are just beginning to take effect once the surgery has been completed. One way to accomplish the shorter staging approach is to initially perform the PCI in the catherization laboratory and then directly transfer the patient to the operating room for re-operative valve surgery. However, this approach would require not only holding an operating room open, but also two hospital fees for the use of both the catheterization laboratory and the operating room, as well as additional hand-offs among care providers, making it an inefficient and cost-ineffective approach. Another solution may be to develop hybrid operating rooms/catheterization suites where PCI could be done first followed immediately by re-operative valve surgery. Such hybrid suites could also be used for elective minimally invasive primary valve surgery with accompanying PCI. Recently, we have been performing elective cases involving both procedures in a staged fashion more often, but, without hybrid suites, timing and coordination are critical.
None of the patients with previous CABG in this study undergoing re-operative valve surgery received additional grafts, but rather they had previous vein grafts or native coronaries stented as needed. We believe that a significant left anterior descending coronary artery lesion should be treated with a LIMA graft when possible, and non-left anterior descending coronary artery native vessels or vein grafts treated with PCI. Recent improvements in PCI techniques and stent design may offer safer intervention for diseased grafts and prolong reestablished patency (11,12).
Our operative mortality rate is far below that predicted by the STS algorithm. We used pre-PCI data in the STS algorithm to calculate the predicted mortality for emergency valve + CABG at the time of PCIin order to document the benefits of PCI to stabilize patients before valve surgery. Statistical analysis to compare the predicted versus the observed mortality rates was not possible because the STS algorithm predicts a percent likelihood of death for each patient, thereby generating a range of percentages (3.5% to 63.5%). Observed mortality, on the other hand, is a discrete variable (alive or dead). However, even if statistical significance is not achieved, the difference in mortality (3.8% vs. 22%) using the hybrid approach certainly seems clinically significant. Our five-year survival of 44% correlates with previously reported results (4) and underscores the poor patients substrate represented by this cohort.
This was a single-center retrospective study of a heterogeneous group of patients with mixed valve and coronary lesions requiring both primary and re-do cardiac procedures, in whom a variety of PCI procedures were performed. In addition, patients were selected for PCI on the basis of favorable anatomy for this procedure. A larger study will be needed to make more definitive recommendations.
Our data demonstrate that, for selected patients with mixed coronary and valvular heart disease, a team approach whereby coronary lesions are treated in the catheterization laboratory and then valvular disease is treated in the operating room provides early outcomes far superior to those predicted for valve/CABG surgery. The “hybrid” approach will likely become more common, particularly for complex patients who require a tailored strategy to help minimize risk.
Supported by the BWH Heart Research Fund of Brigham & Women's Hospital.
- Abbreviations and acronyms
- acute coronary syndrome
- acute myocardial infarction
- coronary artery bypass graft/grafting
- percutaneous coronary intervention
- Society of Thoracic Surgeons
- Received July 22, 2004.
- Revision received September 18, 2004.
- Accepted September 21, 2004.
- American College of Cardiology Foundation
- Edwards F.H.,
- Peterson E.D.,
- Coombs L.P.,
- et al.
- Herlitz J.,
- Brandrup-Wognsen G.,
- Caidahl K.,
- et al.
- Thourani V.H.,
- Weintraub W.S.,
- Guyton R.A.,
- et al.
- Hasdai D.,
- Lev E.I.,
- Behar S.,
- et al.
- Byrne J.G.,
- Karavas A.N.,
- Adams D.H.,
- et al.
- Wilson S.H.,
- Fasseas P.,
- Orford J.L.,
- et al.
- Saia F.,
- Lemos P.A.,
- Lee C.H.,
- et al.