(Circulation. 1999;100:II-119.)
© 1999 American Heart Association, Inc.
Surgery for Coronary Artery Disease |
Risk Stratification for Coronary Bypass Surgery in Patients With Left Ventricular Dysfunction
Analysis of the Coronary Artery Bypass Grafting Patch Trial Database
From the Departments of Surgery (M.A., H.M.S., E.A.R.) and Medicine (W.W., J.T.B., M.P.), Columbia University College of Physicians and Surgeons, New York, NY.
Correspondence to Michael Argenziano, MD, c/o Eric A. Rose, MD, Division of Cardiothoracic Surgery, Milstein Hospital, Room 7-435, 177 Fort Washington Ave, New York, NY 10032. E-mail ma66{at}columbia.edu
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionReferences |
|---|
Background—Preoperative characteristics may influence morbidity and mortality in patients undergoing coronary artery bypass grafting (CABG). The CABG Patch Trial was designed to assess the impact of prophylactic insertion of an implantable cardioverter-defibrillator in patients undergoing high-risk CABG. This database was used to investigate the influence of symptomatic congestive heart failure (CHF) and angina on morbidity and mortality in CABG patients with ventricular dysfunction.
Methods and Results—Data were analyzed for 900 randomized
patients with an ejection fraction 
35% and an abnormal
signal-averaged ECG. Single-variable and stepwise multiple logistic
regression analyses were used for mortality and length-of-stay
(LOS) data. Severity of CHF and angina was graded by the New York Heart
Association (NYHA) and Canadian Cardiovascular Society
(CCS) classifications, respectively. Perioperative
mortality was 3.5% in 454 patients without clinical signs of heart
failure versus 7.7% in 443 patients with NYHA class I to IV heart
failure (P=0.018). By multiple logistic regression
analysis, mortality was significantly higher in patients with
preoperative symptomatic (NYHA class I to IV) heart failure
(odds ratio, 2.4; P=0.01) or reoperation (odds ratio,
3.8; P<0.0001). Mortality was not significantly
influenced by age, sex, the presence or severity of angina,
hypertension, left main coronary artery disease,
pulmonary disease, or severity of CHF (although LOS was
increased 0.7 days per NYHA class). Patients with a history of stroke
had a higher rate of perioperative stroke (16.4%
versus 3.6%, P=0.001) and an increased LOS (by 3.5
days).
Conclusions—Symptomatic heart failure and reoperation are predictors of increased operative mortality in patients with ventricular dysfunction and a positive signal-averaged ECG. Conversely, patients without heart failure symptoms may undergo CABG with relatively low mortality despite low ejection fraction. LOS is prolonged significantly by advanced age, history of stroke, and the presence and severity of heart failure.
Key Words: heart failure • angina • mortality • coronary artery bypass surgery
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Coronary artery bypass grafting (CABG) has been shown to prolong the life expectancy of several subgroups of patients with coronary artery disease,1 2 3 including those with left ventricular dysfunction.4 5 However, because surgical revascularization in these patients is associated with an increased risk of perioperative mortality,6 7 the long-term survival benefit of CABG surgery is dependent on low operative mortality rates. Risk factors for operative mortality after CABG have been defined by several investigators8 9 10 11 12 13 and include urgency of operation, left ventricular ejection fraction (LVEF), reoperation, age, sex, and presence of left main coronary artery disease. However, only a few studies have focused on clinical predictors of operative mortality after CABG in patients with left ventricular dysfunction and have reported inconsistent results. In one study of 466 patients with LVEF <0.40,7 operative mortality was related to ejection fraction, being increased in patients with lower LVEF. In contrast, in an analysis of 224 patients with LVEF <0.30,14 mortality was not influenced by decreasing ejection fraction but instead by clinical markers of heart failure, such as New York Heart Association (NYHA) class and left ventricular end-diastolic pressure (LVEDP). Finally, another study15 identified age and renal dysfunction as predictors of operative mortality but did not note an increased mortality in patients with heart failure.
The CABG Patch Trial16 was a prospective, randomized trial designed to test the hypothesis that prophylactic insertion of an implantable cardioverter-defibrillator (ICD) would reduce mortality rates in CABG patients predicted to be at high risk for late arrhythmic death. Results of that trial have been previously reported.17 18 19 The purpose of the present study was to use the CABG Patch Trial database to identify predictors of perioperative mortality in a large group of CABG patients with preoperative left ventricular dysfunction.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Study Patients
Randomization and data collection in the full-scale CABG Patch Trial were conducted between March 12, 1993, and February 5, 1996, at 37 centers in the United States and Europe. Informed consent was solicited from each enrolled patient. Inclusion criteria were a clinical indication for elective CABG, age <80 years, LVEF <0.36, and an abnormal signal-averaged ECG. The last criterion was selected to reduce the number of patients with a low risk of arrhythmic events after CABG surgery.20 Exclusionary criteria included emergency CABG, history of sustained ventricular tachycardia, ventricular fibrillation or cardiac arrest with inducible ventricular tachycardia, previous or concomitant mitral or aortic valve surgery, or concomitant surgery for arrhythmia or ventricular aneurysm. A complete list of proposed and actual reasons for exclusion from the trial has been published previously.21
If the surgeon agreed intraoperatively that randomization was appropriate, patients were randomized to undergo routine conclusion of surgery or ICD insertion with epicardial rate-sensing electrodes and patch leads. Patients were followed at 3-month intervals for 42 months, corresponding to the anticipated useful battery life of the ICDs. During the recruitment period, 71 855 patients were screened, 1422 eligible patients were identified, 1055 were enrolled and signed consent forms, and 900 were randomized. Details of the randomization procedure have been described previously.16 21
Statistical Analysis
Severity of clinical heart failure and of angina pectoris was
graded by the NYHA and Canadian Cardiovascular Society
(CCS) classifications, respectively. For the present
analysis, randomized patients were classified into 3 subgroups
according to severity of angina: those without symptoms of angina for
the past 3 months, those with CCS class I/II symptoms of angina, and
those with CCS class III/IV symptoms of angina. Patients were also
classified into subgroups according to severity of heart failure
symptoms: those without symptoms, those with NYHA class I/II symptoms,
and those with NYHA class III/IV symptoms.
Perioperative mortality and hospital length of stay
were analyzed for patients in each heart failure and angina
symptom grouping. Perioperative mortality was defined
in the present study by death within 30 days of CABG or within 90
days of CABG in continually hospitalized patients. Length of stay was
defined by the number of days from the date of coronary bypass
surgery to hospital discharge. Fisher’s exact test or Mantel-Haenszel
2 tests were used for mortality data, and
2-sided unpaired t tests were applied to length-of-stay
data.
Single-variable tests of significance were performed on
perioperative mortality and postoperative
length-of-stay data by use of a set of candidate variables selected
by literature review, including angina and heart failure symptom group,
age, sex, LVEF, LVEDP, and history of chronic obstructive
pulmonary disease (COPD), hypertension, diabetes, and stroke.
Logistic and linear regression models were used to analyze
perioperative mortality and length of stay,
respectively. Next, stepwise multiple regression analyses were
performed with the use of those variables that were statistically
significant in the single-variable analyses. All
statistical tests were performed to reject the null hypothesis at a
level of 
=0.05. To prevent confounding due to shortened lengths of
stay for the patients who died perioperatively, patients in this group
were not included in linear regression models of length of
stay.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Patient Demographics and Preoperative Characteristics
A summary of the demographic profile and preoperative medication regimens of the 900 randomized patients is shown in Table 1
. Data regarding congestive heart
failure (CHF) class were missing for 3 patients, and data regarding
angina class were missing for 57 patients, yielding 840 patients with
complete symptom data. Mean age was 63.5 years, and the majority of
patients were male, smoked cigarettes, or had a history of 
1
myocardial infarction. Significant proportions of the study cohort were
hypertensive, hypercholesterolemic, or diabetic, and
almost 11% had had previous coronary artery bypass surgery.
All patients had low preoperative ejection fraction (<0.36), and 16%
had severely depressed ventricular function (LVEF <0.20).
Fifty percent of patients had clinical heart failure, whereas 75% had
angina pectoris. Table 2 lists the
breakdown of patients according to CHF and angina class.
|
|
Perioperative Mortality
Overall perioperative mortality was 5.6% in the
840 randomized patients for whom complete data on symptoms were
available (Table 3).
Perioperative mortality rates were 3.3% among patients
without heart failure, 8.2% among patients with class I/II heart
failure, and 7.7% among patients with class III/IV heart failure.
Perioperative mortality rates were 4.3% among patients
without angina, 5.9% among patients with class I/II angina, and 6.1%
among patients with class III/IV angina. In single-variable
analyses (Table 4), there was a
statistically significant association between
perioperative mortality and heart failure symptom group
(P=0.018) but not between perioperative
mortality and angina symptom group (P=0.932). The only other
variable that was significantly associated with
perioperative mortality in the single-variable
analyses was reoperation, ie, a history of CABG before the
index surgery. In stepwise multiple logistic regression
analyses with the log odds of perioperative
mortality as the dependent variable, class I/II and class III/IV
heart failure symptoms were each associated with a 2.4-fold increase in
the odds of perioperative mortality compared with
absence of heart failure (Table 5).
Compared with primary CABG, reoperation was associated with a 3.8-fold
increase in the odds of perioperative mortality.
|
|
|
|
Hospital Length of Stay
Among the randomized patients for whom complete data on symptoms
were available, average length of stay was 10.3±10.0 (median, 8)
days(Table 6). Unadjusted lengths of stay
were 9.2±6.9 (median, 7) days among patients without heart failure,
11.9±14.8 (median, 8) days among patients with class I/II heart
failure symptoms, and 10.9±9.5 (median, 8) days among patients with
class III/IV heart failure symptoms. Unadjusted lengths of stay were
10.2±10.1 (median, 8) days among patients without angina, 10.3±10.3
(median, 7) days among patients with class I/II angina symptoms, and
10.3±9.8 (median, 8) days among patients with class III/IV angina
symptoms. In single-variable analyses (Table 7), there was a statistically significant
association between length of stay and heart failure symptom group
(P=0.008) but not between length of stay and angina symptom
group (P=0.247). Age in decades, history of stroke,
defibrillator group status, and LVEF were also significantly associated
with length of stay in single-variable analyses. In
stepwise multiple linear regression analyses with length of
stay as the dependent variable, class I/II heart failure symptoms
were associated with 2.3 additional days in the hospital compared with
absence of heart failure (P=0.006), and class III/IV
symptoms were associated with 1.4 additional days in the hospital
(P=0.08). Age, history of stroke, and defibrillator group
status remained statistically significant in multiple regression
analyses (Table 8).
|
|
|
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Since the advent of CABG for the treatment of ischemic heart disease, several investigators have attempted to define preoperative characteristics predictive of postoperative success and survival. In an early report, Kennedy et al10 performed discriminant analyses of data from the Coronary Artery Surgery Study (CASS) to determine those variables that would predict operative mortality. In 6176 patients who underwent isolated CABG from 1975 through 1978, overall operative mortality was 2.3% and was increased by age, high-grade left main coronary artery occlusion, female sex, poor left ventricular wall motion, increased LVEDP, and the presence of pulmonary rales on physical examination. Myers et al11 performed a further analysis of 8991 patients in the CASS study who underwent isolated CABG from 1974 through 1979 and identified CCS angina classification and CHF score as additional determinants of mortality. Notably, the CASS study excluded patients with LVEF <0.35. More recently, Christakis et al9 studied 7334 patients who underwent isolated CABG at 3 teaching hospitals from 1982 to 1986. Overall operative mortality for CABG was 3.7% and was significantly higher with emergency surgery, depressed LVEF, increasing age, and previous CABG surgery. In a study investigating the influence of clinical heart failure on operative mortality, Wechsler and Junod13 examined nearly 6000 patients who underwent CABG at 2 centers. In one center, operative mortality was 0.75% in patients without CHF and 6.8% in patients with class IV CHF; in the other, mortality was 3.0% without CHF and 14.9% with class IV CHF. Additionally, these authors found that patients with class IV angina fared worse than did those with milder degrees of angina.
In contrast to the above reports, which represent only a small
fraction of the published literature on risk factors for mortality
after CABG, relatively few studies have focused on patients selected on
the basis of ventricular dysfunction. Before the
present analysis, the largest group of patients with
ventricular dysfunction to be studied was reported by
Hochberg et al.7 In this 1983 analysis of data
from 466 patients with ejection fractions <0.40, overall operative
mortality after CABG increased progressively as ejection fraction
decreased, from 11% in patients with ejection fractions between 0.20
and 0.39 to 37% in patients with ejection fractions <0.20. A decade
later, Hausmann et al14 reported a prospective study of
224 patients with ejection fractions of 0.10 to 0.30 who underwent
CABG. Overall operative mortality was 8.9%, and in single-variable
analyses, LVEDP >24 mm Hg, number of prior myocardial
infarctions, and class III or IV heart failure were significantly
associated with operative mortality. In that analysis, symptoms
of angina and LVEF were not correlated with mortality. Subsequently,
Kaul et al15 reported a study of 210 patients with
ejection fractions of 0.20 who underwent CABG, with an operative
mortality of 10%. Early mortality was significantly correlated with
age but was not influenced by preoperative heart failure, previous
CABG, or the sex of the patient.
In the present study, perioperative mortality was significantly increased by the presence of heart failure symptoms as well as previous CABG surgery, and hospital length of stay was increased by advanced age, ICD implantation, and the presence and severity of heart failure symptoms. Notably, mortality was not influenced by age, sex, presence or severity of angina, hypertension, LVEDP, left main coronary artery disease, or severity of CHF. Although mortality increased progressively from 3.5% in patients without CHF to 5.7%, 8.1%, and 9.1% in patients with NYHA class I, II, and III CHF, respectively, a linear relationship between CHF class and mortality could not be established because patients with class IV CHF had a mortality of only 4.7%. Although this decrement in mortality in patients with the most advanced degrees of heart failure might be explained by proportionately greater improvements in myocardial function after revascularization, it is also likely due to the fact that patients with class IV CHF who undergo CABG are highly selected, with optimal revascularization targets and minimal comorbid conditions.22 Finally, the low overall operative mortality encountered in our cohort was especially notable in light of the fact that all patients had an abnormal preoperative signal-averaged ECG, which has been shown to predict poor recovery of myocardial function (and poor prognosis) after CABG.23
A potential limitation of the present study was the sample size of 900 patients that constituted the CABG Patch Trial, which is relatively small compared with other studies of risk factors among unselected CABG patients. With the low mortality rates encountered in our cohort, power to detect differences in mortality according to certain variables may have been limited. Nonetheless, compared with other published studies focusing on patients with ejection fractions <0.40, our analysis examined approximately twice the number of patients as the next largest study.7
In conclusion, the present study demonstrates that despite traditional concerns about operative risks in patients with ventricular dysfunction, patients without heart failure symptoms may undergo CABG with low mortality despite low ejection fraction, high LVEDP, advanced age, an abnormal signal-averaged ECG, or comorbid conditions such as diabetes, hypertension, or chronic obstructive pulmonary disease. Furthermore, although the presence of clinical heart failure symptoms confers a >2-fold increase in the risk of death, patients with low ejection fraction and advanced heart failure may still expect a perioperative mortality rate of <10%. This increased mortality risk may not be prohibitively high in patients who stand to benefit significantly from surgical revascularization but certainly underscores the importance of proper selection of patients with surgically correctable ischemia. Finally, length of hospital stay, an increasingly important economic outcome variable, is significantly increased by advanced age, a history of stroke, and the presence and severity of clinical heart failure.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
1. Herd JA, Wood AJ, Blumenthal J, Dougherty JE. Medical therapy in the elderly. J Am Coll Cardiol. 1987;10(suppl A):18A–21A.
2.
Baker DW, Jones R, Hodges J, Massie BM, Konstam MA,
Rose EA. Management of heart failure, III: the role of
revascularization in the treatment of patients with
moderate or severe left ventricular systolic
dysfunction. JAMA. 1994;272:1528–1534.
3.
CASS Principal Investigators and Associates.
Coronary Artery Surgery Study (CASS): a randomized trial of
coronary artery bypass surgery: survival data.
Circulation. 1983;68:939–950.
4.
Alderman EL, Fisher LD, Litwin P, Kaiser GC, Myers WO,
Maynard C, Levine F, Schloss M. Results of coronary artery
surgery in patients with poor left ventricular function
(CASS). Circulation. 1983;68:785–795.
5. Bounous EP, Mark DB, Pollock BG, Hlatky MA, Harrell FE Jr, Lee KL, Wechsler AS, Pryor DB, Califf RM. Surgical survival benefits for coronary disease patients with left ventricular dysfunction. Circulation. 1988;78(suppl III, pt 2):III-1151–III-1157.
6. Christakis GT, Weisel RD, Fremes SE, Ivanov J, David TE, Goldman BS, Salerno TA. Coronary artery bypass grafting in patients with poor ventricular function. J Thorac Cardiovasc Surg. 1992;103:1083–1092.[Abstract]
7. Hochberg MS, Parsonnet V, Gielchinsky I, Hussain SM. Coronary artery bypass grafting in patients with ejection fractions below forty percent. J Thorac Cardiovasc Surg. 1983;86:519–527.[Abstract]
8. Zubiate P, Kay JH, Mendez AM. Myocardial revascularization for the patient with drastic impairment of function of the left ventricle. J Thorac Cardiovasc Surg. 1977;73:84–86.[Abstract]
9. Christakis GT, Ivanov J, Weisel RD, Birnbaum PL, David TE, Salerno TA. The changing pattern of coronary artery bypass surgery. Circulation. 1989;80(suppl I):I-151–I-161.
10. Kennedy JW, Kaiser GC, Fisher LD, Maynard C, Fritz JK, Myers W, Mudd JG, Ryan TJ, Coggin J. Multivariate discriminant analysis of the clinical and angiographic predictors of operative mortality from the Collaborative Study in Coronary Artery Surgery (CASS). J Thorac Cardiovasc Surg. 1980;80:876–877.[Abstract]
11. Myers WO, Davis K, Foster ED, Maynard C, Kaiser GC. Surgical survival in the Coronary Artery Surgery Study (CASS) Registry. Ann Thorac Surg. 1985;40:245–260.[Abstract]
12.
Higgins TL, Estafanous FG, Loop FD, Beck GJ, Blum JM,
Paranandi L. Stratification of morbidity and mortality outcome by
preoperative risk factors in coronary artery bypass patients: a
clinical severity score. JAMA. 1992;267:2344–2348.
13. Wechsler AS, Junod FL. Coronary bypass grafting in patents with chronic congestive heart failure. Circulation. 1989;79(suppl I):I-92–I-96.
14. Hausmann H, Warnecke H, Ennker J, Topp H, Schiesser A, Hempel B, Friedel N, Schmitt D, Hetzer R. Survival predictors in patients with a left ventricular ejection fraction of 10–30% receiving a coronary bypass: analysis of preoperative variables. Cardiovascular Surgery. 1993;2:558–562.
15.
Kaul EK, Agnihotri AK, Fields BL, Riggins LS, Wyatt DA,
Jones CR. Coronary artery bypass grafting in patients with an
ejection fraction of twenty percent or less. J Thorac
Cardiovasc Surg. 1996;111:1001–1012.
16. The CABG Patch Trial Investigators and Coordinators. The Coronary Artery Bypass Graft (CABG) Patch Trial. Prog Cardiovasc Dis. 1993;36:97–114.[Medline] [Order article via Infotrieve]
17. Spotnitz HM, Herre JM, Raza ST, Hammon JW Jr, Baker LD Jr, Fitzgerald DM, Kron IL, Bigger JT. Effect of implantable cardioverter-defibrillator implantation on surgical morbidity in the CABG Patch Trial. Circulation. 1998;98(suppl II):II-770–II-80.
18. Spotnitz HM, Herre JM, Baker LD, Fitzgerald DM, Kron IL, Bigger JT. Surgical aspects of a randomized trial of defibrillator implantation during coronary artery bypass surgery: the CABG Patch Trial. Circulation. 1996;94(suppl II):II-248–II-253.
19.
Bigger JT Jr, for The CABG Patch Trial Investigators.
Prophylactic use of implanted cardiac defibrillators in
patients at high risk for ventricular arrhythmias
after coronary artery bypass graft surgery. N Engl
J Med. 1997;337:1569–1575.
20. Kron IL, Cope JT, Baker LD Jr, Spotnitz HM. The risks of reoperative coronary artery bypass in chronic ischemic cardiomyopathy: results of the CABG Patch Trial. Circulation. 1997;96(suppl II):II-21–II-25.
21. Curtis AB, Cannom DS, Bigger JT Jr, DiMarco JP, Estes M III, Steinman RC, Parides MK. Baseline characteristics of patients in the Coronary Artery Bypass Graft (CABG) Patch Trial. Am Heart J. 1997;133:787–798.
22. Hung J, Kelley DT, Baird DK, Hendel PN, Leckie BD, Grant AF, Uren RF. Aorto-coronary artery bypass grafting in patients with severe left ventricular dysfunction. J Thorac Cardiovasc Surg. 1980;79:718–723.[Medline] [Order article via Infotrieve]
23. Cook JR, Flack JE, Gregory CA, Deaton DW, Rousou JA, Engelman RM. Influence of the preoperative signal-averaged electrocardiogram on left ventricular function after coronary artery bypass surgery in patients with left ventricular dysfunction. Am J Cardiol. 1998;82:285–289.[Medline] [Order article via Infotrieve]
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||