| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
J Thorac Cardiovasc Surg 2006;131:90-98
© 2006 The American Association for Thoracic Surgery
Surgery for Acquired Cardiovascular Disease |
a Department of Thoracic and Cardiovascular Surgery, The Cleveland Clinic Foundation, Cleveland, Ohio.
b Department of Quantitative Health Sciences, The Cleveland Clinic Foundation, Cleveland, Ohio.
Read at the Eighty-fifth Annual Meeting of The American Association for Thoracic Surgery, San Francisco, Calif, April 10-13, 2005.
Received for publication March 31, 2005; accepted for publication May 9, 2005. * Address for reprints: Joseph F. Sabik III, MD, Department of Thoracic and Cardiovascular Surgery, The Cleveland Clinic Foundation, 9500 Euclid Ave, Desk F24, Cleveland, OH 44195. (Email: sabikj{at}ccf.org).
 |
Abstract |
|---|
 Top Abstract IntroductionPatients and MethodsResultsDiscussionConclusionAppendixDiscussionReferences |
|---|
METHODS: From 1971 to 1998, 48,758 patients underwent primary isolated coronary artery bypass grafting, and 1000 per year were followed every 5 years (n = 26,927). A multivariable time-related analysis was performed to model freedom from coronary reoperation and to identify patient and operative variables associated with occurrence of coronary reoperation.
RESULTS: Freedoms from reoperative coronary artery bypass grafting were 99.6%, 98.4%, 93%, 82%, 72%, and 65% at 1, 5, 10, 15, 20, and 25 years, respectively. Risk of reoperation (hazard function) demonstrated a short, rapidly declining early phase, followed by a long, slow-rising late phase. Patient variables that increased the likelihood of coronary reoperation included younger age (P < .0001), higher total cholesterol (P = .0004) and triglyceride levels (P = .0005), lower high-density lipoprotein (P = .0002) level, diabetes mellitus (P < .0001), and more extensive coronary artery disease (P = .01). Increasing extent of arterial grafting performed at primary coronary artery bypass grafting decreased occurrence of coronary reoperation (P < .0001).
CONCLUSION: Patient factors associated with arteriosclerosis progression and type of bypass conduit influence the need for or bias toward repeat coronary artery bypass grafting. Aggressive patient risk-factor reduction and extensive arterial coronary revascularization at primary coronary artery bypass grafting should result in fewer coronary reoperations.
|
Introduction |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionConclusionAppendixDiscussionReferences |
|---|
|
Coronary artery bypass grafting (CABG) is a temporary treatment for a chronic disease. Arteriosclerosis progresses in native coronary arteries and bypass grafts after revascularization, resulting in recurrent ischemic events and need for further intervention. Risk-factor reduction and arterial grafting have been advocated to improve long-term results of myocardial revascularization.
1-5
For a better understanding of how patient characteristics and operative techniques influence the likelihood of coronary reoperation, we identified factors predictive of the need for or bias toward reoperative CABG.
|
Patients and Methods |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionConclusionAppendixDiscussionReferences |
|---|
Statistical Methods
Analysis
Freedom from reoperation was estimated nonparametrically using the Kaplan–Meier method
6
and parametrically using a multiphase hazard method.
7
(For additional details, see http://www.clevelandclinic.org/heartcenter/hazard). The parametric method involved determining the number of hazard phases, the equation form for each, and parameters characterizing the distribution of times to reoperation.
To identify preoperative patient characteristics and operative techniques (Appendix) associated with reoperation, multivariable analyses were performed in the hazard-function domain. A directed stepwise entry of variables into the model, including appropriate transformations of continuous and ordinal variables, was performed.
8
A P value of .05 or less was considered significant for retaining variables in the model. Interactions among significant variables were sought. Bootstrap aggregation (bagging) was used for variable and interaction validation.
9,10
Presentation
To simplify graphs, actuarial estimates are presented at yearly intervals. To depict reoperation stratified by patient characteristics or operative technique, actuarial estimates were made of each stratum, and parametric predicted freedom from reoperation, calculated for each patient in each stratum from the risk-factor model, was superimposed. Seventy percent confidence limits equivalent to 1 SE are given for actuarial and parametric estimates.
Variables and Definitions
Values of preoperative patient characteristics and operative techniques used in the multivariable analyses were obtained at primary CABG. Left ventricular function was graded as normal (ejection fraction [EF] 
60%), mild dysfunction (EF 40%-59%), moderate dysfunction (EF 25%-39%), or severe dysfunction (EF <25%).
A coronary artery system was considered importantly stenotic if it contained an obstruction at least 50% of the diameter. Incomplete revascularization was defined as failure to graft any system containing at least 50% stenosis, or left anterior descending coronary artery (LAD) and circumflex systems for at least 50% left main trunk stenosis.
|
Results |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionConclusionAppendixDiscussionReferences |
|---|
|
|
|
|
|
Operative procedure
More extensive ITA grafting lowered risk of reoperation (P < .0001; Figure 4). However, ITA grafts to the RCA were not associated with lower risk until the latter half of the series (P = .002; Figure 5). Intra-aortic balloon pump insertion at primary operation increased risk of early reoperation (P < .0001). Patients operated on later in the series were less likely to undergo reoperation (P < .0001).
|
|
|
Discussion |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionConclusionAppendixDiscussionReferences |
|---|
Etiologies include incomplete revascularization, graft failure from technical problems, and arteriosclerosis progression in native coronary arteries or bypass grafts. Treatment includes medical management to control symptoms and surgical or percutaneous reintervention to improve myocardial blood flow and decrease ischemia. The decision about appropriate treatment must be made for each patient by weighing the risks of treatment against the benefits. Comorbidities and amount of ischemic myocardium are important considerations. Healthy patients with large amounts of myocardium at risk are more likely to undergo reoperation, whereas those with multiple comorbidities, patent arterial grafts, and small areas of ischemic myocardium are more likely to receive medical therapy or percutaneous intervention. To understand what drives the need for or bias toward for coronary reoperation, we identified factors associated with the likelihood of undergoing reoperation.
Principal Findings
Patient characteristics and operative procedures during primary operation each influenced the likelihood of undergoing reoperation. Some patient factors increased risk (eg, risk factors for arteriosclerosis), whereas others decreased it. Operative factors that influenced risk included conduit choice and success of the primary revascularization.
Risk factors for arteriosclerosis
Diabetes mellitus and an abnormal lipid profile were associated with increased risk of undergoing reoperation. These are known risk factors for arteriosclerosis and probably increased the likelihood of reoperation by promoting arteriosclerosis progression in native coronary arteries and arteriosclerosis development in bypass grafts. In contrast, elevated high-density lipoprotein protected against need for reoperation, most probably by lessening the risk of arteriosclerosis.
Comorbidity and symptoms
Smoking, older age, renal insufficiency, and worse left ventricular function decreased the likelihood of reoperation. All these factors elevate mortality and morbidity associated with reoperation and thus may have biased the treatment decision against reoperation and in favor of medical therapy or percutaneous intervention.
Patients who had more severe symptoms at the primary operation were more likely to undergo reoperation. Such patients may have a greater sensitivity to ischemia. Thus, if ischemia recurs, they may be more likely to have worse angina and to require reoperation to relieve symptoms.
ITA grafts
More extensive ITA grafting lowered the risk of reoperation. ITA grafts have better long-term patency than saphenous vein grafts because of their freedom from arteriosclerosis.
12
It is logical to assume that if grafts that are more likely to remain patent are used at primary operation, important ischemia is less likely to recur, and therefore the need for reoperation will be lower. Incremental benefits of one and then two ITA grafts in terms of both survival and freedom from reoperation have been previously demonstrated.
1-3
Sergeant and colleagues
13
found that the likelihood of reintervention (reoperation and percutaneous therapy) was strongly reduced by increasing the number of ITA grafts at primary operation. However, they noted no decrease in recurrence of angina with arterial grafting.
11,13
They hypothesized that physicians and surgeons were biased against reintervening in patients with multiple arterial grafts—perhaps appropriately so! The survival benefit from reoperating on patients with patent arterial grafts, particularly to the anterior wall of the left ventricle, is small; thus, reintervention may not be in their best interest because the amount of myocardium at risk may not be large enough to justify reoperation.
ITA grafting to the RCA did not decrease occurrence of reoperation until the latter half of this study. A possible explanation is that percutaneous therapy was not available during the early years, so reoperation was the only way to improve myocardial blood flow. After percutaneous therapy was introduced, there may have been a treatment bias against reoperation in these cases. A second explanation is that with greater surgical experience, it became known that ITA graft patency was lower when used to bypass the RCA than the left-sided coronary arteries.
12
This is due to native blood vessel competitive flow.
14
Use of ITAs to graft RCAs likely became more selective with greater surgical experience, and thus they were more likely to remain patent when used for this purpose later in the study.
Success of primary operation
Insertion of an intra-aortic balloon pump at the primary operation was associated with increased risk of undergoing reoperation within 18 months. Need to place a pump at operation suggests failure of revascularization, possibly from early graft failure or incomplete revascularization. Unsuccessful primary operation would result in failure to relieve symptoms of ischemia, and patients would likely have symptoms early after surgery and require reintervention.
Incomplete revascularization was found by univariable analysis, but not by multivariable analysis, to be associated with reoperation. This may be due to incomplete revascularization being highly correlated with another variable found to be significant in the multivariable analysis (eg, intra-aortic balloon pump); in that case, incomplete revascularization would no longer appear important.
Limitations
Although many patient and operative variables were included in the analysis, important variables may not have been recorded. This is perhaps why some factors identified are difficult to explain. For example, it is hard to understand how incomplete revascularization to the RCA and elevated triglyceride level both lowered the early risk of reoperation. It is possible that these significant variables are correlated with other factors we did not identify or investigate. They also may have been found to be significant by chance alone.
The study period includes three decades of coronary surgery. Both surgical and medical therapy changed during this period, and many treatments available today were not available early in this study. However, using all three decades of data allowed us to evaluate the influence of surgical treatments that were common early in the series but not today, such as saphenous vein grafting to the LAD.
Values of variables used in the analyses were obtained at the primary operation. Thus, we were unable to evaluate how changes in these affected occurrence of reoperation. For example, elevated lipid levels at first revascularization increased the likelihood of undergoing reoperation; however, we do not know whether lowering them after revascularization altered risk.
We evaluated only reoperation; however, during the time of this study, percutaneous intervention became an important treatment for myocardial ischemia, and it is likely that many patients receiving percutaneous therapy today would have undergone reoperation in the past. This may explain why patients later in the series were less likely to undergo reoperation.
Parametric Analysis
To identify time-related risk factors for reoperation, a parametric analysis was used. We believe that this type of analysis has several advantages relative to a Cox-proportional hazards analysis, which assumes the influence of a risk factor is present and of equal importance throughout the period of risk. This may not be true. Some factors may be associated with early risk; others may not exert an effect until much later. Parametric analysis solves this weakness by identifying (1) different phases of risk, (2) which risk factors are associated with each phase, and (3) the strength of risk factors in each phase. Thus, in this study, many risk factors associated with late reoperation had no effect during the early phase. Rather, they were related to either arteriosclerosis progression or arterial grafts, which would be expected to influence the need for late but not early reoperation.
A second benefit of parametric analysis is that risk prediction can be made for an individual patient. With the risk factors identified in this analysis, an individual's likelihood of undergoing reoperation can be determined. This information may be useful in advising patients about the risks and benefits of revascularization.
|
Conclusion |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionConclusionAppendixDiscussionReferences |
|---|
|
Appendix |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionConclusionAppendixDiscussionReferences |
|---|
Demographic
Gender, age (years), height (cm), weight (kg), body surface area (m2), body mass index (kg · m–2)
Symptoms
New York Heart Association functional class (I, II, III, IV)
Left Ventricular Function
Left ventricular function (normal and mild, moderate, and severe dysfunction), previous myocardial infarction, left ventricular segmental wall motion abnormalities (septal, anterior, inferior, lateral, apical, basilar, none)
Cardiac Comorbidity
Family history of coronary artery disease, atrial fibrillation, complete heart block
Noncardiac comorbidity
History of cigarette smoking, peripheral vascular disease, carotid stenosis, hypertension, diabetes mellitus (diet controlled, orally treated, insulin treated), renal insufficiency
Preoperative Laboratory Values
Total cholesterol level, high-density lipoprotein level, low-density lipoprotein level, triglyceride level, creatinine level, blood urea nitrogen, hematocrit
Coronary Artery Anatomy and Stenosis
Dominance (left, right, codominant), number of coronary artery systems with at least 50% stenosis (one, two, three), left main trunk stenosis (any, 50% stenosis, 70% stenosis), LAD stenosis (any, 50% stenosis, 70% stenosis), circumflex stenosis (any, 50% stenosis, 70% stenosis), RCA stenosis (any, 50% stenosis, 70% stenosis)
Procedure
Complete revascularization; incomplete revascularization of LAD, circumflex, or RCA system; any ITA grafting; ITA graft to LAD, circumflex, or RCA; any saphenous vein grafting; saphenous vein graft to LAD, circumflex, or RCA
Experience
Date of operation
Postoperative Management
Postoperative intra-aortic balloon pump
|
Discussion |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionConclusionAppendixDiscussionReferences |
|---|
Today, Sabik and coauthors have narrowed their focus to the risk of reoperation and extended the follow-up to 25 years, at which point 65% of their patients have avoided coronary reoperation. Just as important, they have demonstrated that increased use of arterial conduits very significantly increased freedom from reoperation.
I think there are at least three "freedoms" that might be relevant to our understanding of the long-term benefits of CABG. The most comprehensive is freedom from all ischemic events, even those not requiring any kind of intervention. Next in line, at least with respect to frequency, is freedom from any reintervention, followed by its smaller subset, freedom from reoperation, which was the focus of this discussion.
Many of you will remember a series from Belgium that was published in 1991 by Sergeant and Blackstone, in which freedoms from all ischemic events were 54% at 10 years and a mere 22% at 15 years. This is in contrast to the data that you have just heard, in which freedoms from reoperation were 93% at 10 years and 82% at 15 years. So if we assume, for the sake of argument, that the 1991 figures still apply, this leaves a 50% to 60% gap between ischemic events and reoperation.
I think it is interesting to speculate, as some have, that surgical treatment might actually increase this gap if it produces a stable ischemic syndrome or increases reluctance to reoperate, and there are those who believe that bilateral ITA bypasses have both those effects. Be that as it may, we should certainly hope that improvements in surgery have narrowed the gap by reducing the frequency of nonsurgical ischemic events. With this in mind, Dr Sabik, and understanding that this was not the purpose of your study, do you have any information or opinion regarding the freedom from all ischemic events in this group or the frequency of nonsurgical reinterventions?
According to one of your tables, revascularization was judged to be incomplete in 80% of the group not undergoing reoperation. This seemed improbably high to me and was significantly higher than what you observed in the reoperative group (P < .0001). If I understand your analysis, a 50% stenosis is indistinguishable from a 99% stenosis in the tabulation. Is it possible that the group not undergoing reoperation had a higher frequency of debatable 50% to 60% lesions of no surgical significance and that the reoperative group had a higher frequency of lesions more severe than 75%?
I also noticed that bilateral ITA conduits were used in 8% of all patients, and one or two ITAs were used in 71% of all patients. Both figures seem low. I assume this is because ITA use has increased with time. How frequently do you use bilateral ITAs today?
I would also like to know your group's view on the advisability of sequential ITA bypasses. And should the desire to use bilateral ITAs influence the choice of off-pump versus on-pump CABG?
Dr Sabik. Thank you for your insightful comments and questions. We have not determined the freedom from all ischemic events after coronary surgery. However, we have begun to analyze our data to determine freedom from all coronary reinterventions, both reoperation and percutaneous intervention, after CABG. Freedoms from all coronary reinterventions are about 80% at 10 years and about 60% at 20 years. Risk factors for reintervention are similar to those we presented today for reoperation. Risk factors for arteriosclerosis increase the likelihood of reintervention, and arterial grafting lowers the risk. An interesting finding in this analysis is that freedom from percutaneous intervention after coronary surgery is not decreased by ITA grafting at the first operation. We believe that this suggests a bias in treatment of patients with recurrent ischemia and patent arterial grafts. This bias, however, may be appropriate. We are unlikely to reoperate on a patient with recurrent angina from stenosis of a diagonal who has patent ITA grafts to his or her LAD and circumflex. The benefit of reoperation does not outweigh the risk. However, percutaneous intervention to the diagonal to relieve symptoms may be appropriate.
Completeness of revascularization was 80% in the no reoperation group and 74% in the reoperation group. Incomplete revascularization was found by univariable analysis, but not by multivariable analysis, to be associated with reoperation. This may have been due to incomplete revascularization being highly correlated with another variable found to be significant in the multivariable analysis. We defined incomplete revascularization as failure to graft any coronary system containing a stenosis of at least 50% or failure to graft the LAD and circumflex in patients with at least 50% stenosis of the left main trunk. We did not specifically determine whether the grade of coronary lesion was less in the incompletely revascularized no reoperation group than in the reoperation group. However, we did find the degree of coronary stenosis to be directly related to the likelihood of reoperation.
This study covers three decades of surgical revascularization, and our use of ITA during this period changed. In the 1970s, many patients did not receive even single ITA grafting, and few underwent bilateral ITA grafting. This explains the overall low bilateral ITA use. Our bilateral ITA use increased in the latter two decades of the study. However, inclusion in this study of patients operated on early in the series was extremely valuable, because it allowed us to determine how different revascularization strategies affected the likelihood of coronary reoperation.
We use ITAs to perform both sequential grafts and composite T or Y grafts. The technique used depends on the quality of the ITA, the coronary anatomy, and the preference of the surgeon. If the ITA is small, we use it as a single graft. If it is of good quality, we may use it as a sequential graft if the coronary anatomy allows sequential grafting without kinking of the ITA.
We do not believe that the quality of surgical revascularization should be compromised to avoid cardiopulmonary bypass, unless bypass poses an important risk to the patient. If the same quality of revascularization can be performed off pump as on pump, and the morbidity of the operation can be lowered by performing it off pump, then off-pump revascularization may be the best way to proceed.
Dr Joan Ivanov (Toronto, Ontario, Canada). This was a typically excellent presentation from your group, Dr Sabik. Reoperation, to me, seems to be a very difficult outcome at which to look, because it is probably the one outcome that is most subject to any kind of selection bias. So can you discuss your predictors, such as age and gender, in that domain?
Second, we have noticed quite a significant reduction, a temporal trend, in the numbers of patients coming back for reoperative CABG. Did you use any kind of temporal marker in your models to try and adjust for time?
Dr Sabik. I agree that the decision to reoperate is highly dependent on selection bias, and that is why in our presentation we stated that our objective was to determine how patient characteristics and the primary operation influence the need for or bias toward coronary reoperation.
How is younger age a risk factor for the likelihood of having a reoperation? We believe it is a surrogate for a patient characteristic, such as a gene that predisposes one toward aggressive arteriosclerosis. Or it may be a surrogate for level of activity. Younger people are more likely to be active than older patients, and they are therefore more likely to have recurrent angina and to be referred for reoperation. An older patient may have had similar arteriosclerosis progression but be less active and without symptoms and therefore not be referred for surgery. We believe that male gender is similarly a surrogate for a patient characteristic not included in our analysis.
Dr Ivanov. And a temporal trend?
Dr Sabik. We used date of surgery to determine the influence of time on the need for or bias toward reoperation. We found that patients operated on in the 1970s and 1980s had a much higher likelihood of undergoing reoperation than those operated on in the 1990s, and we believe this was due to the introduction of percutaneous coronary therapy in the late 1980s.
Dr Harold L. Lazar (Boston, Mass). In your large database, did you have any evidence to suggest that the use of statins or angiotensin-converting enzyme inhibitors in these patients might have prolonged the time to revascularization?
Dr Sabik. We do not have that information in our database.
Dr Brian Buxton (Heidelberg, Austria). Thank you for a lovely article and for drawing our attention to the 20-year results. I would like to follow up Dr Smith's comments on the right ITA by asking you a few more details about its use during the period of the study and what your current recommendations are. Specifically, do you use left or right coronary targets in situ or free grafting?
Dr Sabik. We believe that ITAs should be used preferentially to graft left-sided coronary arteries. We know that their patency—and therefore effectiveness—is greater when they are used to bypass left-sided coronary arteries. We use the left ITA routinely to bypass the LAD and the right ITA to bypass the next most important left-sided coronary artery. We use a radial artery to graft any remaining left-sided coronary artery and a saphenous vein graft to bypass the RCA.
Dr Buxton. What about the proximal attachment to the right ITA? Currently, do you leave it in situ, do a Y graft, or do an aortic anastomosis?
Dr Sabik. We use all three techniques, depending on the quality of the ITAs and the coronary anatomy.
Dr John D. Puskas (Atlanta, Ga). Just to reiterate what Dr Smith asked, the overall series had 8% bilateral ITA use. What is your current use?
Dr Sabik. Overall, we use bilateral ITAs in about 25% of our patients undergoing primary coronary revascularization. This overall percentage is influenced by the characteristics of our patient population. We are unlikely to use bilateral ITAs in obese, insulin- treated patients with diabetes and are very likely to use them in non–insulin-treated patients younger than 70 years.
| Earn CME credits at http://cme.ctsnetjournals.org/cgi/hierarchy/ctsnetcme_node;JTCS.
|
|
References |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionConclusionAppendixDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  J. F. Sabik III, A. Stockins, E. R. Nowicki, E. H. Blackstone, P. L. Houghtaling, B. W. Lytle, and F. D. Loop Does Location of the Second Internal Thoracic Artery Graft Influence Outcome of Coronary Artery Bypass Grafting? Circulation, September 30, 2008; 118(14_suppl_1): S210 - S215. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Azmoun, R. Ramadan, N. Al-Attar, C. Kortas, S. Ghostine, C. Caussin, M.-L. Bourachot, B. Lancelin, M. Slama, and R. Nottin Exclusive Internal Thoracic Artery Grafting in Triple-Vessel-Disease Patients: Angiographic Control Ann. Thorac. Surg., June 1, 2007; 83(6): 2098 - 2102. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. R. R. Jaganath, S. Theodore, and K. M. Cherian Internal thoracic artery grafts to right coronary system. J. Thorac. Cardiovasc. Surg., August 1, 2006; 132(2): 442 - 442. [Full Text] [PDF]
|
|
|
|
|
|
J. F. Sabik III, E. H. Blackstone, A. M. Gillinov, N. G. Smedira, and B. W. Lytle Occurrence and Risk Factors for Reintervention After Coronary Artery Bypass Grafting Circulation, July 4, 2006; 114(1_suppl): I-454 - I-460. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
O. Ashraf Redo coronary bypass grafting: Role of arterial grafts and time interval J. Thorac. Cardiovasc. Surg., July 1, 2006; 132(1): 209 - 210. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |
