| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Dartmouth-Hitchcock Medical Center, Lebanon, New Hampshire
b Maine Medical Center, Portland, Maine
c Catholic Medical Center, Manchester, New Hampshire
d Fletcher Allen Health Care, Burlington, Vermont
e Concord Hospital, Concord, New Hampshire
f Eastern Maine Medical Center, Bangor, Maine
g Portsmouth Regional Hospital, Portsmouth, New Hampshire
Accepted for publication July 9, 2007.
* Address correspondence to Dr Dacey, Department of Cardiothoracic Surgery, Dartmouth-Hitchcock Medical Center, Lebanon, 03756, New Hampshire (Email: lawrence.j.dacey{at}hitchcock.org).
Presented at the Forty-second Annual Meeting of The Society of Thoracic Surgeons, Chicago, IL, Jan 30–Feb 1, 2006.
 |
Abstract |
|---|
 Top Abstract IntroductionPatients and MethodsResultsCommentDiscussionReferences |
|---|
Methods: We identified 1693 patients, aged 80 to 89, with two-vessel disease (57.6%) or three-vessel disease (42.4%), without left main disease, undergoing a first, nonemergency revascularization from 1992 to 2001. Adjusted hazard ratios (HR) were calculated for CABG versus PCI. Because survival curves for these procedures crossed midway through year 1, results were analyzed separately for the first 6 months and 6 months to 8 years.
Results: PCI was performed in 54.6% of patients with two-vessel disease and 23.7% of those with three-vessel disease. More CABG patients were men (54.7% versus 43.3%). The CABG patients had more peripheral vascular disease (23.1% versus 15.2%) and congestive heart failure (24.5% versus 13.1%) but less renal failure (4.6% versus 9.1%) and fewer prior myocardial infarctions (48.7% versus 53.6%). In-hospital mortality was 3.0% for PCI and 5.9% for CABG (p = 0.005). CABG was associated with poorer survival than PCI during the first 6 months (HR, 1.32; p = 0.135). Survival from 6 months to 8 years was significantly better with CABG for the group as a whole (HR, 0.72; p = 0.005) and for patients with two-vessel disease (HR, 0.68; p = 0.016), and there was a nonsignificant trend for those with three-vessel disease (HR, 0.75; p = 0.177).
Conclusions: Patients aged 80 years or older with multivessel disease must consider the trade-off between the increased early risks of CABG in return for improved long-term survival.
|
Introduction |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentDiscussionReferences |
|---|
The randomized trials comparing PCI with CABG have shown that survival was essentially equivalent for most patient populations [3, 4]; however, these trials did not include patients older than age 80. In addition, the generalizability of those studies has recently been called into question. Reports from two large registries [5, 6] have shown better survival for populations similar to those in the Bypass Angioplasty Revascularization Investigation (BARI) [7] undergoing CABG versus PCI.
Given the absence of relevant information on revascularization in octogenarians from randomized trials and the question of their generalizability, we sought to use our large regional registry of consecutive coronary revascularizations to assess long-term survival after PCI and CABG for patients aged 80 years and older who may have been candidates for either procedure.
|
Patients and Methods |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentDiscussionReferences |
|---|
Data Collection
Data were obtained from the registries of the Northern New England Cardiovascular Disease Study Group (NNECDSG), a voluntary research consortium composed of clinicians, research scientists, and hospital administrators. The goal of the group is to foster the continuous improvement in the quality of care, safety, and effectiveness of cardiovascular interventions through the analysis of process and outcomes data and the timely feedback of these data to the clinicians involved in providing these services [8–10]. Data are prospectively collected on all PCIs and CABGs in the region and are periodically validated using administrative data sources.
The data collected for this study included patient demographics, comorbid conditions, cardiac history, cardiac anatomy, cardiac function, procedural indication and priority, procedural details, and outcomes. The priority of a patient’s intervention was defined as urgent or elective. Urgent indicated that medical factors required the patient to stay in the hospital for the procedure before discharge. Elective was defined by medical factors that indicated the need for the procedure but allowed the patient to be discharged from the hospital with readmission at a later date. Complete definitions of these variables have been previously published [11].
Institutional review board (IRB) approval was obtained at each participating medical center. The IRBs of seven of our eight member centers have designated the NNECDGS as a Quality Improvement Registry, and therefore, patient consent was not required. Written patient consent was obtained for the one remaining center.
Patient Follow-Up
The outcome measure for this study was all-cause mortality. Mortality through December 31, 2001, was determined by a probabilistic match of the regional registry to the National Death Index (NDI; United States Department of Health and Human Services) using a combination of patient name, Social Security number, date of birth, gender, date last known alive, and last known state of residence. The sensitivity of the NDI to identify deaths is 92% to 99%, depending on which identifiers are available [12, 13].
Statistical Analysis
Characteristics of the PCI and CABG populations were summarized by percentages for categoric variables and means for continuous variables and compared using 
2 tests and t tests, respectively. Crude survival curves were created according to the nonparametric Kaplan-Meier method and compared using a log-rank test. A Cox proportional hazard regression model was used to compare PCI versus CABG patients while adjusting for baseline characteristics. Hazard ratios (HRs) and 95% confidence intervals (95% CIs) were calculated using PCI as the referent group. A HR of less than 1 indicated better survival for the CABG population; a HR exceeding 1 indicated better survival for the PCI patients. We adjusted for the patient characteristics of age, sex, peripheral vascular disease, diabetes, chronic obstructive pulmonary disease (COPD), congestive heart failure (CHF), renal insufficiency (creatinine 
2 mg/dL) or renal failure, cancer, left ventricular ejection fraction, history of myocardial infarction (MI), and priority at intervention. Analyses were performed using Stata 8.0 software (Stata Corp, College Station, TX).
|
Results |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentDiscussionReferences |
|---|
Table 1 summarizes the characteristics of the PCI and CABG populations. The CABG patients were younger (mean age, 82.8 years versus 83.1 years; p = 0.026) and were more often male (54.7% versus 43.3%; p < 0.001). They had more peripheral vascular disease (23.1 versus 15.2%, p < 0.001) and CHF (24.5% versus 13.1%, p < 0.001) but less renal dysfunction (4.6% versus 9.1%, p < 0.001) and fewer prior MIs (48.7% versus 53.6%, p = 0.009). The groups were not significantly different with regard to diabetes, COPD, cancer, ejection fraction of less than 0.40, or priority at intervention. Patients undergoing CABG were more likely to be fully revascularized. In-hospital mortality was 3.0% for PCI and 5.9% for CABG (p = 0.005).
|
The CABG patients had significantly more strokes than the PCI patients (2.84% versus 0.57%, p = 0.001). Most of the CABG procedures were performed with cardiopulmonary bypass (87% versus 13% off-pump). There was no significant difference in the stroke rate between CABG patients done on or off pump. Mediastinitis occurred in 1.42% of CABG patients.
Crude Kaplan-Meier survival curves for PCI versus CABG are shown for the group overall (Fig 1) and stratified by the number of diseased coronary vessels (Figs 2 and 3).
After 6 months, survival after CABG was better than after PCI (log-rank p = 0.016), with respective survival at 1, 5, and 8 years of 88.1%, 68.0%, and 49.3% for CABG and 87.8%, 60.6%, and 45.8% for PCI (Table 2). For patients with 2VD, median survival was 7.7 years for those undergoing PCI and 8.0 years for those undergoing CABG. Among patients with 3VD, median survival was 6.0 years for PCI and 7.7 years for CABG.
|
|
|
|
2 mg/dL, cancer, diabetes, ejection fraction, prior MI, and priority at surgery. Crude and adjusted HRs are presented in Table 3.
|
|
Comment |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentDiscussionReferences |
|---|
Previous reports have also shown lower in-hospital mortality for patients undergoing PCI versus CABG in this age group. In the largest collaborative study of coronary revascularization in very aged patients, Peterson and colleagues [2] examined the survivorship experience of patients aged 75 years or older undergoing PCI or CABG. This project involved six PCI registries (n = 48,439) and eight CABG registries (n = 180,709). They found that after adjustment for potential confounding factors, in-hospital mortality favored PCI with a risk difference of 2.9% (PCI, 3%; CABG, 5.9%). Our findings of crude in-hospital mortality are similar, with an absolute risk difference of 2.9% (risk difference of 4% among patients with 2VD, and 1.8% among patients with 3VD).
There are scant data directly comparing the long-term survival of PCI versus CABG in very elderly individuals. Individual studies of survival focusing on either PCI or CABG have demonstrated favorable results for both procedures in this age group [14, 15], but there is little information to aid in the decision of which revascularization strategy to pursue in the very elderly patient.
Randomized trials in the general population, which have not included the elderly, have reported equivalency of survival in patients assigned to PCI or CABG [3, 4]. In the short term, this appears to be true. However, a meta-analysis of nine of these randomized trials by Hoffman and colleagues [16] showed a survival benefit for CABG versus PCI at 4 years among diabetic patients and at 5 to 8 years among nondiabetic patients.
Recent studies of the general population from large registries have also found a long-term survival benefit of CABG over PCI among patients with multivessel coronary disease. Malenka and colleagues [6] found the adjusted long-term survival for patients with 3VD was better after CABG than PCI (HR, 0.60; p < 0.01) but not for patients with 2VD (HR, 0.98; p = 0.77). Hannan and colleagues [5] reported that in patients with two or more diseased coronary arteries, CABG was associated with higher rates of long-term survival. The adjusted HRs for the long-term risk of death of CABG relative to PCI ranged from 0.76 for patients with 2VD with involvement of the nonproximal left anterior descending artery to 0.64 for 3VD with proximal left anterior descending artery involvement. Brener and colleagues [17] showed that among 6033 patients at The Cleveland Clinic presenting with multivessel disease and comorbid conditions, CABG was associated with better survival than PCI, even after adjusting for known cofounders.
In a study from Alberta, Canada, that specifically looked at the elderly, Graham and colleagues [18] reported long-term survivorship among 21,573 patients undergoing catheterization, of which 15,392 were younger than 70 years, 5198 were 70 to 79, and 983 patients were older than 80. Among the octogenarians, 133 underwent CABG, 289 underwent PCI, and 561 had medical management. Long-term survivorship in all age groups was significantly better with CABG versus PCI, and both forms of revascularization showed significant improvement compared with medical management alone.
A number of reasons may explain why we observed a survival benefit among patients undergoing CABG versus PCI. One possibility may be related to the completeness of revascularization. Previous reports have shown improved survival with more complete revascularization [19]. Our data registries do not have a systematic assessment of myocardium at risk or territories revascularized. Nonetheless, in our series among patients undergoing CABG, three or more vessels were bypassed in 84% of patients with 2VD and in 96% of patients with 3VD. For patients with 2VD and 3VD undergoing PCI, 64% had only one vessel attempted. This suggests that either the PCI strategy is more focused on a culprit lesion angioplasty or that surgical strategies allow clinicians to bypass diseased segments of the coronary tree that may not be possible by PCI.
Another possibility involves the specific area on the artery where the intervention is occurring. Interventional cardiologists at present can only image the lumen of the vessel, rather than the entire vessel itself, as is done in the operating room. PCI involves stenting on the most diseased segment of a vessel. Atherosclerotic segments in the vessel wall proximal or distal to the luminal irregularities cannot be identified and may progress to blockages more quickly than normal areas. CABG allows bypassing to the least diseased segment of the artery. This may afford patients with longer vessel patency and subsequent improved survival.
The favorable median survivorship (7.7 years) of this cohort of octogenarians with multivessel coronary disease is a potentially surprising finding. However, studies have previously shown that octogenarians undergoing invasive cardiac surgery have a mean survival of more than 5 years [14, 20, 21]. Fruitman and colleagues [22] studied 126 octogenarians in Nova Scotia, Canada, and found that patients undergoing cardiac procedures had an actuarial survival of 83% at 1 year and 80% at 2 years. The study population achieved a nearly identical survivorship as the population matched for age and sex.
From a patient standpoint, mortality alone may not be the most apt metric for assessing a successful revascularization strategy. Although some individuals undergo PCI or CABG for strictly for longevity, many in this age group seek treatment to achieve an improved quality of life. Previous studies have demonstrated the effectiveness of PCI and CABG in enhancing the quality of life in elderly patients by providing both improved functional status and relief from angina [23, 24]. Studies comparing quality of life after CABG or PCI have shown either equivalent improvement [25] or a greater improvement with CABG compared with PCI [26–28]. As more octogenarians are being offered aggressive treatment for their coronary disease, it will be increasingly important to supply them with survival and quality-of-life data to aid them in making their decisions. Thus, it would be helpful for registries to begin recording quality-of-life assessment after interventions.
We acknowledge several limitations to the current study. First, although we did adjust for important known patient and disease characteristics before intervention, there may remain unmeasured differences in the patient groups that could have confounded our results. However, we investigated and adjusted for those patient and disease factors that have been previously reported to be associated with in-hospital mortality in these procedures.
Second, additional factors outside of the index admission may also influence a patient’s survivorship, such as mental health, socioeconomic status, support systems, and functional status. We do not have data on these factors.
Third, this was not a randomized trial, and the procedure selected represents the preferences of the patients and providers involved. We have no information on whether the patients for one procedure were truly candidates for the other. Patients undergoing PCI may have been turned down for CABG; and likewise, patients having CABG may have been felt to have no viable catheter-based options.
Fourth, the PCI and CABG procedures both continue to evolve. Although this study was done in the era of intracoronary stenting during PCI, it precedes the era of drug-eluting stents [29]. Survival of patients who have PCI with drug-eluting stents may change and will be an area of intense interest. Likewise in CABG, techniques involving off-pump surgical revascularization are being applied more commonly in the elderly population than was done in this study and may or may not decrease the morbidity and mortality of these surgical patients [30, 31]. Nonetheless, our findings reveal the long-term effect of the practice of revascularization strategies among octogenarians in northern New England from 1992 to 2001.
Fifth, the reported sensitivity of the NDI for correctly identifying vital status is good (0.92 to 0.99) but not perfect, and we could have overestimated survival. Most important, we were able to provide comparable patient identifiers for the PCI and CABG cohorts such that there should have been no bias introduced into the analysis by differential matching of patients to the NDI.
In conclusion, this study demonstrates a favorable survivorship for octogenarians undergoing either CABG or PCI for treatment of multivessel coronary artery disease. CABG surgery is associated with higher in-hospital mortality than PCI, but for patients surviving this acute postoperative phase, CABG affords a more favorable survivorship up to 8 years after the procedure. Age alone should not be a deterrent for aggressive treatment of coronary disease.
|
Discussion |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentDiscussionReferences |
|---|
We owe them thanks for undertaking such a large review to expand our knowledge base in octogenarians. This is likely the largest paper of its kind to address this group. Also, we should thank Dr Dacey for a masterful presentation of his data today.
Obviously, limitations in this type of study are significant. As he mentioned, essentially no drug-eluting stents were involved, and CABG was predominantly on-pump, and off-pump was used sparingly, and I don’t believe it is broken down further as to independent risk factors for risk in those groups.
Probably the most important thing in this paper is a call for us to change how we evaluate our outcomes, in my opinion. I would challenge us to look carefully at our STS [Society of Thoracic Surgeons] database and its end points for utility in today’s marketplace. In other words, we must not congratulate ourselves any further on simple mortality rates and morbidity in-hospital, but work to create new metrics for measuring benefits to patients, metrics whose terms would contrast measurable quality-of-life items and, with improved imaging technique, demonstrate better our efforts at revascularization. We must also better evaluate other end-organ morbidity of CABG such as neurologic injury or pulmonary dysfunction as regards to other therapies such as simply medical or interventional approaches to coronary disease. Having said these things, which are limits of our current common ways of investigating patients, there are questions that arise for Dr Dacey.
One, do you have any morbidity or length-of-stay data for this series? Two, could you break out any differences in those who had off-pump CABG versus the whole or those who had coated stents versus simple ones? And finally, do you have any cost data or return-to-function data on these two cohorts?
Thank you once again to the STS for the chance to share these comments and to Dr Dacey and his colleagues on this fine work.
DR DACEY: Thank you very much, Dr Calhoon. The only hard data as far as morbidity we have following this procedure is stroke. That is obviously the biggest concern to this age group. In my own experience, you really can’t scare an 86-year-old saying that, "you might die." They know they might die. What they want is quality of life, but what they seem to be "deathly" afraid of, if you will pardon the expression, is having a stroke afterwards. We did look back at that. About 2.8% of the patients having CABG had a stroke as opposed to 0.6% of the patients undergoing PCI. Stroke, we found previously, is a horrible predictor for death: about 25% of patients that have a stroke will die in hospital. Trying to do efforts to avoid stroke, whether that could be off-pump surgery—and there is some indication that off-pump surgery may reduce stroke particularly in this age group—I think would be important.
With regards to on- and off-pump, we don’t have the outcomes based on that. About 13% of these patients were done off-pump, the vast majority within the last few years. About 25% of the patients done in 2000 and 2001 were done off-pump. I would agree that to have a meaningful discussion with these patients, with octogenarians, we really do need quality of life, not just survival data.
DR PAUL KURLANSKY (Miami, FL): I rise to congratulate Dr Dacey and the Northern New England Group, which has always provided us with superb follow-up and very insightful research. I have one comment and one question. Regarding quality of life, our work we reported to the American College of Cardiology a couple of years ago on over a thousand octogenarians who underwent coronary bypass surgery demonstrated that the SF-36 [Medical Outcomes Study Short Form 36] patient self-assessment of quality of life was excellent in long-term survivors and in fact was equal to or better than, depending upon which parameter was examined, a comparable control group of the general population.
My question for you is whether or not you subjected your data to any propensity scale or other statistical techniques that might study specifically or address specifically the impact of the decision to choose coronary bypass or PCI on the outcomes in these patients as a way to help further elucidate which patients might benefit more from CABG and which might benefit more from PCI in the long-term? Thank you.
DR DACEY: Thank you. That is an excellent question. To answer it directly, no, we had not done propensity scores on the various variables. But as you suggest and allude to, there are a number of published reports in the data suggesting that of survivors, somewhere between 83% and 93% rate their health as excellent and improved compared to what it was prior to the operation, and indeed, the vast majority—over 80%—would agree to go through the procedure again. There is a very high percentage of these people that actually get back to living independently at home. So the outlook is quite good if you do indeed survive.
DR DOUGLAS E. WOOD (Seattle, WA): Dr Dacey, I have one additional question. How do you justify censoring the deaths in the first 6 months for your longer-term analysis? This does not seem to be justifiable in comparing the long-term results and I think weakens the results overall.
DR DACEY: I think the reason to do that is because the curves cross, as I say, midway through year 1 with survival analysis, and I am told that by the curves crossing, if you try to analyze just that portion with the cross and the curves, that you assign different proportional hazards. So statistically, as I say, dividing them into two seems to make statistical sense.
DR WOOD: No, I would disagree. It appears that this skews the analysis to favor the CABG patients because it ignores the deaths in the first 6 months, which of course is what one is interested in is the overall deaths in the whole population over time. I would just encourage you to reanalyze similar to your first slide where you had the whole curves instead of censoring the 6-month data, which lessens its overall credibility.
DR DACEY: Thank you.
|
References |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentDiscussionReferences |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 |
