Uncertainty about the benefit and potential complications of these procedures in very elderly patients has deterred physicians from referring their patients. Nonetheless, high success rates and acceptable complication rates for the use of percutaneous transluminal coronary angioplasty and coronary artery bypass surgery [6-13] have been reported in selected octogenarians. Thus, there is a need to define the effect of these procedures on the long-term outcome of very elderly patients.

We did a retrospective, observational study to evaluate the long-term outcome of patients 80 years or older who were hospitalized between 1985 and 1990 with an acute myocardial infarction and who had diagnostic cardiac catheterization. We tested the hypothesis that, after adjusting for clinical, demographic, hemodynamic, and anatomic variables, coronary revascularization is not significantly associated with improved survival in patients 80 years and older who are hospitalized for an acute myocardial infarction.

Methods

Top Methods Results Discussion Author & Article Info References

Study Sample

The ClinQuery system [14], a hospital-wide computer database, was used to identify patients hospitalized at Boston's Beth Israel Hospital from 1985 to 1990 who had a discharge diagnosis of acute myocardial infarction. Patients were stratified by age (<65 years, 65 to 79 years, 80 years) and compared in terms of their referral for cardiac catheterization, referral for coronary revascularization, and in-hospital mortality. From among the patients at least 80 years old, a sample was selected for detailed study that included all patients who met the above criteria, had cardiac catheterization before discharge, had at least one major coronary artery with greater than 70% stenosis, and did not have severe aortic stenosis (valve area <1.0 cm²) or severe mitral regurgitation. This sample was identified by ClinQuery and was confirmed by a review of hospital charts and procedure reports. The study sample was compared by age, in-hospital mortality, and peak creatine kinase level with the group of patients who had myocardial infarction, were at least 80 years old, and were not referred for cardiac catheterization.

Patient Characteristics and Hospital Course

Demographic, social, and clinical information about the study sample was obtained by chart review. Comorbidities were evaluated using the Charlson comorbidity score [15]. The indications for cardiac catheterization and data regarding hemodynamics, anatomy, left ventricular ejection fraction, procedural complications, and angioplasty results were obtained from chart review and procedure reports. Complications of cardiac catheterization were classified according to the methods described by Wyman and colleagues [16]. If left ventriculography was not done, an estimate of the left ventricular ejection fraction was obtained, if available, from either a radionuclide ventriculogram or an echocardiogram (in that preference order) done during the hospitalization. The study sample was further divided into three groups based on clinical revascularization strategy (percutaneous transluminal coronary angioplasty, coronary artery bypass grafting, or no revascularization) after cardiac catheterization and before hospital discharge. These groups were compared with respect to baseline demographic, clinical, anatomic, and hemodynamic variables. Information was also collected about procedural outcomes, length of stay, and mortality.

Clinical Follow-up

Information on the study group was obtained at least 1 year after hospital discharge by telephone contact with the patient, his or her referring physician, or a first-degree relative. Information was collected regarding mortality, recurrent myocardial infarction, functional status, and quality of life. Functional status was evaluated by inquiring if, after discharge, the patient was able to live independently and to perform activities of daily living without assistance. Quality of life was assessed by asking the patient or his or her surrogate to grade the patient's average quality of life after hospital discharge as either excellent, good, fair, or poor.

Statistical Analysis

Statistical comparisons between continuous variables were done using analysis of variance, and categorical variables were compared using the chi-square test. Kaplan-Meier survival curves were generated to show survival among patients alive at discharge who had angioplasty, bypass surgery, or medical therapy. These groups were compared by the log-rank test [17]. This analysis was repeated after stratifying the group by ejection fraction. The Cox proportional-hazards model [18] was used to determine if percutaneous transluminal coronary angioplasty and coronary bypass surgery (together and separately) were significantly associated with increased survival after adjusting for potential confounders. The model included the following variables: age, sex, ejection fraction, the presence of medication-dependent diabetes mellitus, history of chronic obstructive pulmonary disease, serum creatinine level greater than 220 µmol/L, the presence of three-vessel coronary disease, the presence of left main coronary artery stenosis greater than 50%, the cardiac index, and pulmonary capillary wedge pressure measured during the cardiac catheterization. The multivariate model of coronary revascularization included 83 patients for whom all the variables were available. The multivariate models of percutaneous transluminal coronary angioplasty and coronary artery bypass grafting included 67 patients and 49 patients, respectively. All statistical calculations were performed using STATA (Computing Resource Center, Los Angeles, California).

Results

Top Methods Results Discussion Author & Article Info References

Study Sample

A total of 4694 patients were discharged from Boston's Beth Israel Hospital from 1985 to 1990 with a diagnosis of acute myocardial infarction, including 1743 patients who were younger than 65 years, 1736 patients who were 65 to 79 years old, and 1215 patients who were 80 years and older. The in-hospital mortality rate of these patients increased significantly with age, from 4% among the patients younger than 65 years, to 14% among the patients 65 to 79 years old, to 22% among the patients 80 years and older (P = 0.001). The percentage of patients who were referred for cardiac catheterization decreased significantly with age, from 60% among the patients younger than 65 years, to 37% among the patients 65 to 79 years old, to 11% among the patients 80 years and older (P = 0.001). Among the patients who had cardiac catheterization, referral for coronary revascularization before hospital discharge also decreased significantly, from 92% among the patients younger than 65 years and 95% among the patients 65 to 79 years old to 64% among the patients 80 years and older (P = 0.001). Among all the patients with a discharge diagnosis of acute myocardial infarction who had percutaneous transluminal coronary angioplasty, the in-hospital mortality rate increased significantly with age, from 2% among the patients younger than 65 years, to 5% among the patients 65 to 79 years old, to 12% among the patients 80 years and older (P = 0.001). Among all the patients with a discharge diagnosis of acute myocardial infarction who had coronary artery bypass grafting alone (without valve replacement), the in-hospital mortality rate ranged from 3% among the patients younger than 65 years, to 7% among the patients 65 to 79 years old, to 3% among the patients 80 years and older (P = 0.01).

Of the 1215 patients who were 80 years and older and had a discharge diagnosis of acute myocardial infarction, 142 (11%) had diagnostic cardiac catheterization before hospital discharge. On the basis of chart review, patients were excluded who were hospitalized for reasons other than the myocardial infarction (36 patients), had severe aortic stenosis or mitral regurgitation (8 patients), did not have significant coronary artery disease (3 patients), or died before a definitive decision about coronary revascularization could be made (2 patients). The final study group comprised 93 patients. Compared with the patients with myocardial infarction who were 80 years and older and did not have cardiac catheterization Table 1, the 93 study patients were younger (83.1 years compared with 85.9 years, P = 0.001) and had a lower in-hospital mortality rate (15.1% compared with 22.6%, P = 0.001) despite having a higher peak creatine kinase level (11.2 mmol/L compared with 6.6 mmol/L, P = 0.001).

Study Sample Characteristics

The 93 study patients were characterized by their independence and general good health. Almost all of the patients were living at home (89 patients) and caring for themselves (91 patients) before they were hospitalized. They had few comorbidities, with an average Charlson score of 2.0. Among the study group, 7 (9%) had a history of chronic obstructive pulmonary disease, 25 (30%) were taking medication for diabetes mellitus, and 11 (13%) had renal insufficiency (serum creatinine level >220 µmol/L). Seventeen (18%) patients smoked cigarettes.

The indications for cardiac catheterization were postinfarction angina (80%), postinfarction congestive heart failure (11%), cardiogenic shock (4%), positive results from an exercise test (4%), and the evaluation of coronary anatomy after an uncomplicated non-Q wave infarction (1%). Major complications of the cardiac catheterization included one femoral artery injury that required surgical repair and one severe but nonfatal allergic reaction to the contrast dye.

Comparison of Revascularization Strategies

After cardiac catheterization, 41 patients underwent percutaneous transluminal coronary angioplasty, 18 patients underwent coronary artery bypass surgery, and 34 patients did not undergo coronary revascularization. None of the patients who were selected for medical therapy had an absolute contraindication to coronary revascularization. These patients either refused the offer of bypass surgery (4 patients), or their physicians felt that the risks of surgery outweighed its potential benefits (30 patients). There were few crossovers in treatment after initial selection of a management strategy. One patient had coronary bypass surgery 2 days after percutaneous transluminal coronary angioplasty because of persistent angina and hypotension. In addition, 1 patient selected for angioplasty and 2 patients selected for medical therapy had bypass surgery after discharge.

The admission characteristics of the three groups are shown in Table 2. The group that was referred for surgery was slightly younger than the other two groups. There were, however, no substantial absolute differences in potentially important prognostic factors, nor was there a combination of risk factors that was unbalanced in a particular direction.

Procedural Results

Percutaneous transluminal coronary angioplasty was technically successful, defined as achieving less than 50% stenosis, in 36 of 41 patients (88%). Major complications of percutaneous transluminal coronary angioplasty included one instance of injury to the femoral artery that required surgical repair. There were no cases of abrupt closure or coronary artery damage resulting in cardiogenic shock, death within 24 hours, or transfer for emergent coronary bypass surgery. Major complications of bypass surgery included one case of postoperative necrotic bowel (with patient survival) and one case of stroke (with a minor residual neurologic deficit). Overall, five in-hospital deaths occurred after a revascularization procedure.

Clinical Follow-up

Data were obtained for at least 1 year after hospital discharge. Recurrent myocardial infarctions occurred in 9 patients (6% of the angioplasty patients, 0% of the bypass surgery patients, and 28% of the medical therapy patients). Among the patients who were alive at discharge, those who had revascularization were more likely to be able to take care of themselves and live independently after hospital discharge (89% [32 of 36] angioplasty patients, 89% (16 of 18) surgery patients, and 52% [13 of 25] medical therapy patients). The average quality of life was also more likely to be considered good or excellent (86% in the angioplasty [31 of 36]) and surgical patients (89% [16 of 18]) compared with the medical therapy group (44% [11 of 25]).

Univariable Survival Analysis

The 14 patients who died during hospitalization included 5 of 41 patients (12%; 95% CI, 4% to 26%) in the angioplasty group, none of 18 patients (0%; CI, 0% to 19%) in the bypass surgery group, and 9 of 34 patients (26%; CI, 13% to 44%) in the medical therapy group (P = 0.001). At 1 year, the mortality rate was 24% (CI, 15% to 47%) for patients who had angioplasty, 6% (CI, 0% to 27%) for patients who had bypass surgery, and 44% (CI, 27% to 62%) for those who received medical therapy (P = 0.01). Survival curves were plotted for each of the revascularization groups (Figure 1). These curves showed a significant difference in survival among the three groups, with the surgery group having the best survival and the medical therapy group, the worst. When the groups were further stratified based on ejection fraction (Figures 2 and 3), patients with an ejection fraction of 0.55 or greater (mean, 0.64 ± 0.9) showed no statistical differences according to therapy (P = 0.3). Among patients with an ejection fraction less than 0.55 (mean, 0.39 ± 0.12), however, the groups differed significantly, with the medical therapy group having the worst survival (P = 0.005).

View larger version (13K): [in this window] [in a new window]   Figure 1. Kaplan-Meier plot of the survival for the 79 patients who survived the hospitalization. Patients were stratified by revascularization strategy. The number of patients at risk at each time point is indicated. CABG = coronary artery bypass graft; PTCA = percutaneous transluminal coronary angioplasty; Rx = therapy.

View larger version (13K): [in this window] [in a new window]   Figure 2. Kaplan-Meier plot of the survival for the 38 patients who survived the hospitalization and had an ejection fraction of 0.55 or greater. Patients were stratified by revascularization strategy. The number of patients at risk at each time point is indicated. CABG = coronary artery bypass graft; PTCA = percutaneous transluminal coronary angioplasty; Rx = therapy.

View larger version (12K): [in this window] [in a new window]   Figure 3. Kaplan-Meier plot of the survival for the 39 patients who survived the hospitalization and had an ejection fraction of less than 0.55. Patients were stratified by revascularization strategy. The number of patients at risk at each time point is indicated. CABG = coronary artery bypass graft; PTCA = percutaneous transluminal coronary angioplasty; Rx = therapy.

The association of coronary revascularization and age, sex, ejection fraction, diabetes mellitus, three-vessel coronary disease, and pulmonary capillary wedge pressure with survival was tested in a bivariate analysis (Table 4). Higher left ventricular ejection fraction (P = 0.03), the absence of three-vessel coronary artery disease (P = 0.03), and the performance of coronary revascularization (P = 0.003) were each significantly associated with survival.

Multivariable Survival Analysis

In the multivariable analysis, the performance of coronary revascularization continued to show a significant association with survival in a model that adjusted for age, sex, ejection fraction, comorbidity score, presence of medication-dependent diabetes mellitus, history of chronic obstructive pulmonary disease, creatinine level greater than 220 µmol/L, number of major coronary arteries with greater than 70% stenosis, presence of left main coronary artery stenosis greater than 50%, past history of myocardial infarction, past history of congestive heart failure, and the cardiac output and pulmonary capillary wedge pressure measured during the cardiac catheterization (Table 4). When the revascularization procedures were considered separately, coronary artery bypass surgery remained independently associated with survival in a model that was adjusted for possible confounders, whereas percutaneous transluminal coronary angioplasty did not.

Discussion

Top Methods Results Discussion Author & Article Info References

In our study sample, only 8% of the patients 80 years and older who were hospitalized with an acute myocardial infarction were referred for cardiac catheterization before hospital discharge. Our principal finding is that outcomes among the group who were selected by their physicians for coronary revascularization were excellent. These patients had good survival, had a low rate of recurrent myocardial infarction, and commonly achieved a good or excellent quality of life after discharge. Further, almost all the patients who had coronary revascularization recovered sufficiently to care for themselves after discharge. Therefore, the hypothesis that patients 80 years and older who are hospitalized with a myocardial infarction have such a poor prognosis after coronary revascularization that invasive procedures should be considered futile must be rejected.

Our results are consistent with studies in other settings that suggest selected older patients can do well after invasive cardiovascular procedures [6-13, 19-23]. For instance, Jeroudi and colleagues [20] showed an 80% survival rate at 3 years for 54 octogenarians who had angioplasty from 1980 to 1988, done primarily for treatment of unstable angina. Cumulative freedom from major cardiac events among these patients was 78% at 3 years. These findings are similar to our results with angioplasty, although only 5 of their patients had a myocardial infarction.

The results of bypass surgery in our series were excellent compared with other series of very elderly patients, and careful patient selection surely contributed to this favorable outcome. In particular, our results are better than those reported by Edmunds and colleagues [6]. In that series, 101 consecutive patients underwent open-heart surgery for various indications and had a 90-day mortality rate of 29% and a 3-year mortality rate of 41%. Fifty-eight of 100 patients, however, had clinically significant aortic or mitral valvular disease, and 1 patient had an aortic dissection repair. These patients would have been excluded from our study sample because we restricted our study to patients considered for coronary revascularization only. Also, Edmunds and colleagues' cohort had significant comorbidity (10% had a history of stroke) and poor functional status (one half were classified as New York Heart Association functional class IV). Even with these adverse preoperative conditions, the authors concluded that "operation may be an effective therapeutic option" for very elderly patients with unmanageable cardiac symptoms.

Our results are consistent with those of Ko and colleagues [9], who evaluated the experience of 65 octogenarians who had cardiac catheterization at their institution for various indications. Only one in-hospital death occurred among the 36 patients who had surgery, and compared with medical therapy, bypass surgery was significantly associated with survival in a model that adjusted for ejection fraction, the presence of valvular disease, and New York Heart Association class. The 3-year probability of survival in the surgical patients was 77% and in the medical patients was 55%.

It is important to emphasize that the study group was carefully selected and represented a small proportion of all the patients 80 years and older who were hospitalized with myocardial infarction. Our results cannot be generalized to the large number of patients who were not referred for cardiac catheterization. The patients in our study generally were highly functional before admission, had low comorbidity, had preserved left ventricular ejection fraction after the myocardial infarction, and had evidence of clinically significant postinfarction ischemia. Although our findings are relevant for these patients, our results do not address whether invasive procedures are underused in the remaining group of very elderly patients.

Another important issue regarding our study is that because of its design and relatively small number of patients, it is not able to prove that coronary revascularization is beneficial in older patients. The interpretation of our findings must be made with the understanding that the decision to refer a patient for coronary revascularization includes selection factors that we could neither identify nor quantify for this analysis. Within the study group, important differences existed among the patients selected for angioplasty, bypass surgery, and medical therapy. Medical therapy was reserved for the patients with extensive coronary artery disease and the worst hemodynamic profile and the worst left ventricular function. Patients selected for bypass surgery also had extensive coronary artery disease but a slightly better hemodynamic profile and left ventricular ejection fraction than the patients selected for medical therapy. The patients selected for angioplasty had the least extensive coronary disease and the best hemodynamic profile and left ventricular ejection fraction. Given these differences among the groups, it is not surprising that significant differences existed in the in-hospital mortality and the long-term survival of hospital survivors in the three groups.

A randomized trial would be desirable to determine more definitively the effect of revascularization for elderly patients with acute myocardial infarction, but these observational data are the only information currently available to assist physicians and their patients. The inclusion criteria were designed to produce a fairly homogeneous group of patients who could have been included in a randomized trial and limited the study to patients who had no contraindications to coronary revascularization.

Although the study was observational in design, it has the strength of representing the entire experience of a large tertiary-care hospital with this patient group. Therefore, it does show that these patients have a good outcome after their procedure. We can only infer that the procedure may have contributed to that outcome. The speculation is supported by statistical comparisons between patients with different clinical strategies. Patients who had coronary revascularization had a better survival rate than patients who received medical therapy.

After stratifying by ejection fraction, these differences by revascularization group remained significant in the patients with an ejection fraction that was less than 0.55 (mean, 0.39 ± 0.12). This finding is consistent with studies of younger patients in other settings that have suggested that coronary revascularization selectively benefits patients with abnormal ejection fractions [24]. Finally, a Cox proportional-hazards model was developed to test the association of revascularization with increased survival after adjusting for important differences between the groups. These results suggest that revascularization may have benefited these patients, an issue that requires further study.

The applicability of our results to other clinical settings is not known. The study represents the experience of a single hospital, a community-based teaching hospital with a large elderly population. The success of angioplasty and surgery may be related, in addition to patient selection, to operator experience and expertise at the institution. Nonetheless, these results should be relevant to other busy medical centers that routinely perform coronary revascularization.

We believe that our study has important implications for the care of elderly patients who are hospitalized with an acute myocardial infarction. Physicians should choose a clinical strategy with the knowledge that many older patients can tolerate the procedures, avoid serious complications, survive the hospitalization, return to independent living, and do well for years. Further, coronary revascularization may provide a survival benefit in certain situations. Further study is required to document the risks, costs, and benefits of coronary revascularization for the treatment of older patients after myocardial infarction.

Presented at the Samuel Levine Young Investigator Award Session of the 1992 Meeting of the American Heart Association.