The number of elderly patients is increasing at an

Arterial Grafting Results in Reduced Operative Mortality and Enhanced Long-Term Quality of Life in Octogenarians Paul A. Kurlansky, MD, Donald B. Williams, MD, Ernest A. Traad, MD, Roger G. Carrillo, MD, John S. Schor, MD, Melinda Zucker, BSN, Sam Singer, BA, and George Ebra, EdD Miami Heart Research Institute, Mount Sinai Medical Center, Miami Beach, Florida Background. Despite well-established benefits of arterial (ART) grafting, surgeons have been reluctant to use this conduit in octogenarians. This study explores the influence of arterial revascularization on operative and long-term outcomes of coronary artery bypass grafting surgery. Methods. A retrospective analysis was conducted of 987 consecutive patients 80 years of age or older who underwent isolated coronary artery bypass grafting between January 1989 and November 2000. Patients with saphenous vein graft only (SVG; n 574) were compared with those receiving arterial and saphenous vein grafts (ART SVG; n 413). Mean follow-up for SVG patients was 3.8 years (range, 4 months to 12.6 years) and 98.6% complete, and mean follow-up was 3.1 years for ART SVG patients (range, 2 months to 11.2 years) and 97.3% complete. Results. Patients with SVG had a significantly higher (p 0.009) operative mortality (11.1% versus 6.3%) and significantly longer postoperative length of stay (12.9 versus 10.7 days; p 0.002) than ART SVG recipients. More ART SVG than SVG patients were free of all postoperative complications (290 of 413; 70.2% versus 372 of 574; 64.8%; p 0.086). Multivariable analysis identified SVG as an independent predictor of operative mortality (p 0.014) and late mortality (p 0.040). When patients were matched by equivalent propensity scores to receive SVG only, operative mortality was higher for SVG patients in four of the five quintiles. At 10 years, 97.0% 1.2% of SVG and 92.9% 3.7% of ART SVG current survivors were free of all late major adverse cardiac events (p 0.565), and 95.5% of SVG patients and 97.5% of ART SVG patients were in Canadian class 1 or 2(p 0.162). On the SF-36 quality-of-life assessment, ART SVG patients scored significantly higher than both SVG patients and age-adjusted normal subjects. Physical health summary component scores were 36.8 11.0 for SVG and 41.0 10.3 for ART SVG (p 0.001). Mental health summary scores were comparable for the two groups. Conclusions. Arterial grafting confers an operative survival benefit, and an enhanced long-term quality of life in elderly patients. (Ann Thorac Surg 2003;76:418 27) 2003 by The Society of Thoracic Surgeons Presented at the Forty-ninth Annual Meeting of the Southern Thoracic Surgical Association, Miami Beach, FL, Nov 7 9, 2002. Address reprint requests to Dr Kurlansky, Miami Heart Research Institute, 801 Arthur Godfrey Rd, 5th Floor, Miami Beach, FL 33140; e-mail: doctorwu18@aol.com. The number of elderly patients is increasing at an unprecedented rate. It is projected that the number of citizens older than the age of 75 years will quadruple in the next 50 years [1]. As many as 40% of all octogenarians have symptomatic cardiovascular disease, which accounts for more than half of the mortality in this age group [2]. As individuals continue to live longer, a greater number will require coronary artery bypass grafting (CABG). Numerous advances in operative techniques and perioperative care have resulted in an increasing number of patients 80 years of age and older being referred for CABG [3]. It therefore becomes increasingly important to determine which operative techniques influence enhanced perioperative and long-term survival and quality of life. The left internal mammary artery (IMA), with its documented superior patency, has become the conduit of choice for surgical revascularization of the myocardium [4]. Initial enthusiasm for the use of this conduit was based on its demonstrated improvement in long-term survival [5]. Recent reports have shown that IMA grafting is associated with decreased operative mortality as well [6, 7]. Ferguson and colleagues [8], in an analysis of 99,942 patients 75 years of age and older, showed that IMA grafting provides an acute survival benefit, even when accounting for all associated risk factors in this patient cohort. However, these results also demonstrated a declining survival benefit from IMA grafting with increasing age. Because nearly three fourths of the patients in the study were younger than 80 years of age, the incremental benefits of IMA grafting in patients 80 years of age and older remains undefined. 2003 by The Society of Thoracic Surgeons 0003-4975/03/$30.00 Published by Elsevier Inc PII S0003-4975(03)00551-4

Ann Thorac Surg KURLANSKY ET AL 2003;76:418 27 ARTERIAL GRAFTING IN OCTOGENARIANS 419 Long-term data have suggested a survival benefit for IMA grafting in octogenarians [9, 10]. However, previous studies have not controlled for patient selection. From these findings, it is unclear as to whether survival benefits attributed to the IMA are a function of the choice of conduit or of specific patient characteristics. The focus of myocardial revascularization in octogenarians must be to prolong life, relieve distress, restore function, and prevent long-term disability and dependency. It should allow the individual to gain an enhanced functional status and a sense of well-being, and to assume a productive role in society. The influence of IMA grafting on long-term quality of life remains currently undefined. The purpose of this report is to examine the immediate and long-term benefits of arterial grafting in myocardial revascularization in patients 80 years of age and older. Furthermore, this study seeks to assess the patient s perceived quality of life in an effort to gain further insights into the health status and sense of well-being after CABG. Patients and Methods Patient Population The focus of this report involves 987 consecutive patients 80 years of age and older who underwent isolated CABG between January 1989 and November 2000. The sample (n 987) was divided into two groups; those who received only saphenous vein grafts (n 574; SVG) and those who received arterial grafts, namely, left IMA, right IMA, or radial arteries, and supplemental SVG (ART SVG; n 413). There were 318 men (55.4%) and 256 women (44.6%) in the SVG group and 281 men (68.0%) and 132 women (32.0%) in the ART SVG group (p 0.001). The mean was 83.5 3.0 years (range, 80 to 99 years) age in the SVG group and 82.5 2.5 years (range 80 to 92 years) in the ART SVG group (p 0.001). The coronary and perioperative risk factors and coronary angiography findings documented in the two patient groups are summarized in Table 1. The patients preoperative anginal symptoms were defined by the Canadian Cardiovascular Classification System (CCS). Unstable angina patients in each group were those with class 3 or 4 symptoms. In the SVG group, 17 patients (3.0%) were in class 2, 257 patients (44.8%) in class 3, and 300 patients (52.2%) in class 4. In the ART SVG group, 27 patients (6.5%) were in class 2, 210 patients (50.8%) in class 3, and 176 patients (42.6%) in class 4. A significantly greater number of patients in the SVG group had class 4 symptoms (52.3% versus 42.6%; p 0.003). Surgical Technique Details of the surgical technique applied in the current series have been discussed previously [11]. The majority of all operations (n 967; 98.0%) were performed with the assistance of cardiopulmonary bypass, the remaining procedures (n 20; 2.0%) were performed off-pump. Table 1. Comparison of Preoperative Clinical Variables and Risk Factors by Patient Group Variables SVG a ART SVG a p Value No. of patients 574 (100.0) 413 (100.0) Sex Male 318 (55.4) 281 (68.0) 0.001 Female 256 (44.6) 132 (32.0) Age (y) Mean 83.5 3.0 82.5 2.5 0.001 Range 80 to 99 80 to 92 Age groups (y) 80 82 241 (42.0) 249 (60.3) 83 85 208 (36.2) 112 (27.1) 86 88 93 (16.2) 44 (10.7) 89 91 22 (3.8) 7 (1.7) 92 10 (1.7) 1 (0.2) Coronary risk factors Family history of CAD 118 (20.6) 85 (20.6) 0.993 Hypertension 377 (65.7) 272 (64.9) 0.953 Hyperlipidemia 166 (28.9) 129 (31.2) 0.433 Smoking history 127 (22.1) 97 (23.5) 0.614 Diabetes mellitus 151 (26.3) 96 (23.2) 0.273 Perioperative risk factors Renal dysfunction 90 (15.7) 44 (10.7) 0.023 Cerebral vascular disease 87 (15.2) 56 (13.6) 0.482 Peripheral vascular disease 126 (22.0) 66 (16.0) 0.019 Prior myocardial infarction 353 (61.5) 216 (52.3) 0.004 History of congestive heart 167 (29.1) 92 (22.3) 0.016 failure Unstable angina 495 (86.2) 326 (78.9) 0.002 Coronary angiography Three-vessel disease 508 (88.5) 372 (90.1) 0.613 Two-vessel disease 52 (9.1) 37 (9.0) Single-vessel disease 14 (2.4) 4 (0.2) Left main disease ( 0.50) 174 (30.3) 126 (30.5) 0.948 Ejection fraction 0.50 208 (36.2) 200 (48.4) 0.001 0.30 to 0.50 264 (46.0) 165 (40.0) 0.30 90 (15.7) 43 (10.4) Not done 12 (2.1) 5 (1.2) a Numbers in parentheses are percentages. ART SVG arterial and saphenous vein graft; artery disease; SVG saphenous vein graft. CAD coronary Operative Data A total of 2,068 coronary artery grafts were performed (mean, 3.6 per patient; range, 1 to 6) in the SVG group. There were 539 (26.1%) SVGs to the left anterior descending coronary artery, 292 (14.1%) to the diagonal arteries, 744 (36.0%) to the circumflex or the obtuse marginal arteries, and 493 (23.8%) to the right coronary artery or the posterior descending coronary artery. In the ART SVG group, there were 1,507 coronary artery grafts performed (mean, 3.7 per patient; range, 1 to 6). The left IMA was used to the left anterior descending coronary artery in 392 (26.0%) grafts, to the diagonal arteries in 64 (4.2%) grafts, to the circumflex or the obtuse marginal arteries in 9 (0.6%) grafts, and to the right coronary artery

420 KURLANSKY ET AL Ann Thorac Surg ARTERIAL GRAFTING IN OCTOGENARIANS 2003;76:418 27 Fig 1. Increasing use of the left internal mammary artery during years of operation. or the posterior descending artery in 3 (0.2%) grafts. The right IMA was used to the left anterior descending coronary artery in 5 (0.3%) grafts, to the circumflex or the obtuse marginal arteries in 5 (0.3%) grafts, and to the right coronary artery or the posterior descending artery in 12 (0.8%) grafts. Radial arteries were used to the left anterior descending coronary artery in 3 (0.2%) grafts, to the diagonal artery in 1 (0.1%) graft, and to the right coronary artery or the posterior descending artery in 3 (0.2%) grafts. Saphenous vein grafts were used to the left anterior descending coronary artery in 13 (0.9%) grafts, to the diagonal arteries in 129 (8.6%) grafts, to the circumflex or the obtuse marginal arteries in 500 (33.2%) grafts, and to the right coronary artery or the posterior descending artery in 368 (24.4%) grafts. In summary, 94.2% (468 of 497) of arterial grafts were performed with a left IMA, and 94.9% (392 of 413) of the SVG ART patients received a left IMA to the left anterior descending coronary artery. Figure 1 demonstrates the increasing use of the left IMA in this cohort of patients during the course of the study. The mean cardiopulmonary bypass time was 57.1 18.1 minutes (range, 10 to 190 minutes) in the SVG group and 66.0 23.9 minutes (range, 29 to 240 minutes) in the ART SVG group (p 0.001). The mean aortic crossclamping time was 26.5 11.1 minutes (range, 4 to 83 minutes) in the SVG group and 37.5 17.2 minutes (range, 12 to 115 minutes) in the ART SVG group (p 0.001). Operative Urgency The operation was performed in the SVG group electively in 455 patients (79.3%), urgently in 75 patients (13.1%), and emergently in 44 patients (7.7%). In the ART SVG group, the operation was performed electively in 327 patients (79.2%), urgently in 66 patients (16.0%), and emergently in 20 patients (4.8%). An urgent operation was defined as being required within 48 hours in an effort to prevent further clinical deterioration. An emergency operation was defined as one performed when a patient was suffering from intractable angina or failure that did not respond to ordinary clinical measures. It included patients receiving respiratory assistance who were transferred from other institutions or those who suffered decompensation in the cardiac catheterization laboratory requiring intervention such as defibrillation, external cardiac massage, balloon counterpulsation, or inotropic support. All other patients in the series were considered elective. Data Sources Perioperative data were obtained by prospective review of the patient s hospital record, catheterization reports, cineangiograms, and echocardiography. Follow-up information was obtained through comprehensive questionnaires and by telephone interview with surviving patients, family members, or the patient s personal physician. Follow-up data included activity level, current symptoms, diagnostic tests, occurrence of late cardiac events, medications being taken, and an assessment of quality of life (QOL). Patients were asked to describe their functional capacity and were ranked according to the CCS classification system. A patient registration form and a patient follow-up form were completed for each patient in the study. These data collection instruments provided standardized reporting of each patient s clinical status before and after the operation. Quality-of-life assessment was conducted with the Short-Form (SF) 36 developed by Ware and associates [12]. The SF-36 is a standardized instrument composed of 36 items designed to measure eight dimensions of overall health. These include physical functioning, social functioning, role limitations attributed to emotional problems, mental health, vitality (energy/fatigue), bodily pain, and general health perception. For each dimension, item scores are computed, totaled, and converted into a scale, which ranges from 0 for worst health to 100 representing best health. Two summary components a physical and mental health score are also computed [13]. A 98.8% follow-up was obtained in the SVG group with 7 patients lost to follow-up. In the ART SVG group, a 97.3% follow-up was achieved with 11 patients lost to follow-up. Statistical Analysis Data are presented as frequency distributions and simple percentages. Values of continuous variables are expressed as mean standard deviation. Univariate analysis of selected preoperative and postoperative discrete variables was accomplished by 2, the continuityadjusted 2 analysis, or a two-tailed Fisher s exact test with the appropriate degrees of freedom to test for the equality of proportions in the case of categorical variables. Two-sample Student s t test (two-tailed) were used to test for the equality of the means of continuous variables. To identify predictors of hospital mortality, a multivariate analysis by forward stepwise logistic regression of 26 preoperative and intraoperative variables was performed. Propensity score technology applying logistic regression was used to calculate a probability score for each patient selected for SVG only on the basis of 25 preoperative and intraoperative risk factors [14, 15]. A Cox proportional hazards regression model [16] was used to discern the influence of multiple clinical variables on

Ann Thorac Surg KURLANSKY ET AL 2003;76:418 27 ARTERIAL GRAFTING IN OCTOGENARIANS 421 late survival. Regression coefficients and odds ratio using the 95% confidence interval were calculated to determine the relative influence of each covariate on the survivor function. Coefficients were computed by the method of maximum likelihood. Patient survival for those discharged from the hospital was expressed by actuarial analysis according to the method of Kaplan and Meier [17] using time zero as the date of operation and late death as the end point (with variability expressed as the standard error of the mean). The equality of survival distribution for the two patient groups was tested with the log-rank algorithm [18]. Data collected were subjected to both quantitative and qualitative analysis using the biostatistical capabilities of the Patient Analysis and Tracking Systems (PATS; Axis Clinical Software, Inc, Portland, OR) and the Number Cruncher Statistical Systems (NCSS; Kaysville, UT). A significant difference between measurements was defined as p less than or equal to 0.050. Results Hospital Morbidity Rate The overall incidence of postoperative morbidity for the two groups was low, with a majority of patients in the SVG group (64.8%; n 372) and in the ART SVG group (70.2%; n 290) experiencing no hospital complications (p 0.086). Patients with deep sternal infection included those with instability of the sternum with positive wound cultures necessitating an additional surgical procedure, such as incision and drainage, debridement, or secondary closure. Respiratory insufficiency included patients who required intubation for more than 48 hours or tracheostomy (or both). Cerebrovascular accident referred to a neurologic deficit that remained unresolved and present for more than 24 hours, and renal dysfunction was defined as a creatinine level greater than or equal to 2.0 mg/dl. Myocardial infarction was defined as a new onset of Q waves with or without elevation of myocardial enzymes, or a substantial elevation of myocardial enzymes alone. Low cardiac output referred to clinical evidence of hypotension, oliguria, and peripheral vascular constriction with normal or supranormal left ventricular filling pressure or a measured cardiac index of less than 2 L min 1 m 2, necessitating the administration of inotropic agents or use of intraaortic balloon pump (IABP), or both. A between-group comparison of each of the hospital complications revealed no significant difference between the groups except for renal insufficiency (p 0.007). Respiratory insufficiency approached statistical significance (p 0.058). The hospital complication rates for the two groups are presented in Table 2. Placement of the IABP was required in 47 patients (8.2%) in the SVG group. Thirty patients (5.2%) had an IABP placed preoperatively, 9 (1.6%) intraoperatively, and 8 (1.4%) postoperatively. In the ART SVG group, 34 Table 2. Comparison of Hospital Complications by Patient Group Complications SVG a ART SVG a p Value No. of patients 574 (100.0) 413 (100.0) Reoperation for 14 (2.4) 15 (3.6) 0.274 bleeding Respiratory insufficiency 113 (19.7) 62 (15.0) 0.058 Cerebral vascular 15 (2.6) 19 (4.6) 0.091 accident Perioperative MI 7 (1.2) 8 (1.9) 0.363 Renal dysfunction 62 (12.5) 30 (7.3) 0.007 Low cardiac output 103 (17.9) 60 (14.5) 0.154 Cardiac arrest 53 (9.2) 27 (6.5) 0.126 Deep sternal infection 10 (1.7) 4 (1.0) 0.311 a Numbers in parentheses are percentages. ART SVG arterial and saphenous vein graft; infarction; SVG saphenous vein graft. MI myocardial patients (8.2%) required placement of an IABP. Nineteen patients (4.6%) had an IABP placed preoperatively, 4 (1.0%) intraoperatively, and 11 (2.7%) postoperatively. There was no significant difference in the use of the IABP between the two groups. None of the patients requiring the use of the IABP in either group experienced a major vascular complication. The average postoperative length of stay for SVG patients was 12.9 10.3 days and 10.7 7.4 days for ART SVG patients (p 0.002). Hospital Mortality Rate Hospital mortality was defined as death occurring during the operation or the hospitalization during which the procedure was performed or death occurring after discharge from the hospital but within 30 days of the surgical procedure, unless the cause was unrelated to the operation. The hospital mortality rate was 11.1% (64 of 574) for SVG patients and 6.3% (26 of 413) for ART SVG patients (p 0.009). The overall hospital mortality rate for the series was 9.1% (90 of 987). The elective mortality rate was 10.1% (46 of 455) for SVG patients and 4.9% (16 of 327) for ART SVG; the urgent mortality rate was 10.7% (8 of 75) for SVG patients and 9.1% (6 of 66) for ART SVG; and the emergent mortality rate was 22.7% (10 of 44) for SVG patients and 20.0% (4 of 16) for ART SVG patients. A between-group comparison of the mortality rates for various urgency categories revealed a significant difference (p 0.011) for elective urgency but not for urgent or emergent (or salvage). The mortality rate for first operation was 11.3% (57 of 505) for SVG patients and 6.7% (25 of 375) for ART SVG patients. The mortality rate for reoperation was 10.1% (7 of 69) in SVG patients and 2.6% (1 of 38) in ART SVG patients. A comparison of the influence of surgical history on hospital mortality rates revealed a significant difference between the groups on first operation (p 0.027) but not for reoperation. Twenty-six preoperative and intraoperative variables were entered into a forward stepwise logistic regression

422 KURLANSKY ET AL Ann Thorac Surg ARTERIAL GRAFTING IN OCTOGENARIANS 2003;76:418 27 Table 3. Multivariate Analysis of Preoperative and Intraoperative Variables Associated With Hospital Mortality in Octogenarian Patients Undergoing Isolated Coronary Artery Bypass Predictor Beta Estimate Standard Odds Error 2 p Value a Ratio Confidence Limits (95%) Preoperative Age 0.0384 0.0028 390.84 0.001 1.0 1.0 1.0 Congestive heart failure 0.5418 0.2422 4.84 0.028 1.7 1.1 2.8 Renal insufficiency 0.8839 0.2720 9.68 0.002 2.4 1.4 4.1 (creatinine 2.0 mg/dl) IABP inserted preoperatively 1.3382 0.3647 11.21 0.001 3.8 1.9 7.8 Intraoperative Saphenous vein conduit 0.6033 0.2512 6.10 0.014 1.8 1.1 3.0 IABP inserted intraoperatively 2.0803 0.6054 9.46 0.002 8.0 2.4 26.2 a Only significant variables (p 0.050) are listed. IABP intraaortic balloon pump. model to identify independent correlates of hospital mortality. Of the variables entered into the multivariate model, four preoperative (age [p 0.001], congestive heart failure [p 0.028], renal insufficiency [p 0.002], and IABP inserted preoperatively [p 0.001]) and two intraoperative variables (saphenous vein conduit only [p 0.014] and IABP inserted intraoperatively [p 0.002]) were found to be predictors of increased hospital mortality (Table 3). Subgroup multivariate analyses of patients age 80 to 83 years (n 611) and 84 years and older (n 356) revealed that choice of conduit was a significant predictor of hospital mortality in the patients age 80 to 83 years (p 0.005) but was not for patients 84 years and older (p 0.765). Although the age distribution was somewhat arbitrary and was selected to achieve parity in the two groups, the underlying trend toward diminishing the impact of arterial grafting on hospital mortality with increasing age is clearly demonstrated. The information in Table 4 provides a comparison of the hospital mortality rates for the two patient cohorts after grouping of the total study population into five quintiles using propensity score analysis. The distribution of cases into quintiles was based on the calculated propensity of receiving an SVG only. Patients who received an SVG only were then compared with those who had an ART SVG within each quintile. Although the Table 4. Comparison of Hospital Mortality by Quintiles for Patients 80 Years of Age and Older Undergoing Isolated Coronary Artery Bypass Grafting With Saphenous Vein Grafts and Arterial and Saphenous Vein Grafts Quintile SVG Group a ART SVG Group a p Value 1 28/160 (17.5) 3/37 (8.1) 0.119 2 15/129 (11.6) 4/68 (5.9) 0.147 3 7/122 (5.7) 7/75 (9.3) 0.340 4 10/92 (10.9) 6/106 (5.7) 0.180 5 4/71 (5.6) 6/127 (4.7) 0.511 a Numbers in parentheses are percentages. ART SVG arterial and saphenous vein graft; infarction; SVG saphenous vein graft. MI myocardial hospital mortality was higher for SVG than ART SVG patients, in four of the five quintiles, these differences did not achieve statistical significance. In the lowest propensity score group (quintile 1), 81.2% of the patients received an arterial graft whereas in the highest propensity score group (quintile 5) only 35.9% of the patients received an arterial graft. Long-Term Follow-up Follow-up data were collected for 510 SVG patients (98.6%) and 387 ART SVG patients (97.3%) discharged from the hospital. The follow-up for SVG ranged from 4.0 months to 12.6 years (mean, 3.8 years) and for ART SVG from 2 months to 11.2 years (mean, 3.1 years). The cumulative follow-up for SVG patients was 1,940.1 patient-years and for ART SVG 1,214.8 patient-years. To identify independent predictors of late death, a Cox proportional hazards regression model was created to measure the effects of various prognostic factors on time-to-response (operation to late death). Cox regression analysis demonstrates the independence of 11 covariates: age (p 0.003), male sex (p 0.001), date of surgery (p 0.001), peripheral vascular disease (p 0.001), preoperative renal insufficiency (p 0.002), congestive heart failure (p 0.001), three-vessel disease (p 0.002), saphenous vein conduit only (p 0.040), postoperative renal insufficiency (p 0.013), pulmonary insufficiency (p 0.014), and cerebral vascular accident (p 0.005) on late death (Table 5). At the completion of the follow-up, 247 (48.4%) of the 510 SVG hospital survivors and 247 (63.8%) of the 387 ART SVG hospital survivors were alive. There were 254 (49.8%) late deaths in the SVG group; 219 (42.9%) were cardiac related and 35 (6.9%) were noncardiac related. In the ART SVG group there were 129 (33.3%) late deaths; 109 (28.2%) were cardiac-related and 20 (5.2%) were noncardiac related. The actuarial survival data for 210 SVG patients and 387 ART SVG patients discharged from the hospital are shown in Figure 2. At 5 years, survival for SVG patients ( standard error of the mean) was 61.5% 2.5% and 64.5% 3.2% for ART SVG patients. At 10 years it was 15.2% 3.4% for SVG patients

Ann Thorac Surg KURLANSKY ET AL 2003;76:418 27 ARTERIAL GRAFTING IN OCTOGENARIANS 423 Table 5. Preoperative, Intraoperative, and Postoperative Variables Influencing Late Mortality as Evidenced by Cox Regression Analysis in Octogenarian Patients Undergoing Coronary Artery Bypass Surgery Predictor Regression Coefficient SEM Relative Hazard p Value a Preoperative Age 0.0531 0.0171 1.0545 0.003 Male sex 0.3828 0.1129 1.4664 0.001 Date of surgery 0.3120 0.0687 1.3661 0.001 Peripheral vascular 0.4368 0.1181 1.5478 0.001 disease Renal insufficiency 0.4621 0.1394 1.5874 0.002 (creatinine 2.0 mg/dl) Congestive heart failure 0.4587 0.1161 1.5820 0.001 Three-vessel disease 0.5769 0.2034 1.7806 0.002 Intraoperative Saphenous vein conduit 0.2375 0.1167 1.2681 0.040 Postoperative Renal insufficiency 0.4573 0.1750 1.5797 0.013 (creatinine 2.0 mg/dl) Pulmonary insufficiency 0.3838 0.1515 1.4679 0.014 Cerebral vascular accident 0.9713 0.3079 2.6414 0.005 a Only significant variables (p 0.050) are listed. and 30.2% 6.6% for ART SVG patients. The equality of survival distribution for the two groups of patients was tested, and no significant difference (p 0.122) was noted. On completion of the follow-up, many current survivors in both groups were clinically and functionally improved and conducting their activities of daily living. In the SVG group 95.5% and in the ART SVG group 97.5% of the patients were in CCS class 1 or 2 at follow-up. The between-group comparison did not achieve statistical significance (p 0.162). The freedom from major adverse late cardiac events (nonfatal myocardial infarction, percutaneous coronary intervention, absence of reoperation or CABG) in SVG and ART SVG current survivors is shown in Figure 3. The freedom from major adverse late cardiac events at 10 years for SVG patients was 97.0% 1.2% and 92.9% 3.7% for ART SVG patients. A comparison of the freedom from major adverse cardiac events for the two cohorts of patients revealed no significant difference (p 0.565). Quality-of-Life Assessment At follow-up QOL assessment was conducted for all survivors in the SVG group (241 of 247; 97.5%) and in the ART SVG group (244 of 247; 98.8%) willing and able to complete the SF-36. Table 6 gives the means and standard deviations for each of the eight health scale scores as well as the physical and mental health summary component scores for the two patient groups. Between-group comparisons on the eight health scale Fig 2. Actuarial survival of patients 80 years of age and older who underwent isolated coronary artery bypass grafting with saphenous vein grafts (SVG; dark line) and arterial and saphenous vein grafts (ART SVG; light line). Number of patients at risk is in parentheses; results are mean standard error of the mean. Fig 3. Comparison of freedom from nonfatal myocardial infarction, percutaneous coronary intervention, and reoperation in current survivors of isolated coronary artery bypass grafting with saphenous vein grafts (SVG; dark line) and arterial and saphenous vein grafts (ART SVG; light line). Number of patients at risk is in parentheses; results are mean standard error of the mean.

424 KURLANSKY ET AL Ann Thorac Surg ARTERIAL GRAFTING IN OCTOGENARIANS 2003;76:418 27 Table 6. SF-36 Quality of Life Health Status Scale Scores for Patients 80 Years of Age and Older Undergoing Isolated Coronary Artery Bypass Grafting With Saphenous Vein Grafts and Arterial and Saphenous Vein Grafts Scale Score scores and the two summary component scores revealed that the ART SVG group performed significantly better on all scales except role-emotional, mental health, and the mental health summary component. A similar comparison of the physical and mental health summary scores for ART SVG patients with age-adjusted normal values revealed that ART SVG patients achieved significantly higher scores than age-adjusted normal values in the physical health summary (p 0.001). However, in the mental health summary, ART SVG patients achieved a higher score than age-adjusted normal values, but the difference did not achieve statistical significance (p 0.295). These findings demonstrate that ART SVG patients have an enhanced QOL when compared with patients receiving SVG only and with age-adjusted normal values. Moreover, SVG patients scored lower than controls in the physical and mental health summary scales. When responses to the SF-36 were analyzed by quintiles, patients in the ART SVG group achieved higher scores than SVG patients in the physical health summary component within every quintile. In quintile 2, this difference achieved statistical significance (p 0.006). In the mental health summary component the ART SVG group achieved higher scores in quintiles 2, 3, and 5, with quintile 4 being statistically different (p 0.033). The SVG patients achieved higher scores in the mental health component in quintiles 1 and 2. However, these differences did not achieve statistical significance. Comment SVG Group (n 241) ART SVG Group (n 244) p Value Physical functioning 39.2 28.9 49.1 30.6 0.001 Role physical 58.4 45.8 67.5 43.5 0.024 Bodily pain 66.3 29.4 75.2 25.2 0.001 General health 53.9 22.2 59.8 19.7 0.002 Vitality 48.8 21.5 53.5 20.5 0.014 Social functioning 71.8 31.0 79.7 26.9 0.003 Role emotional 70.0 42.2 76.0 39.6 0.108 Mental health 70.3 19.2 72.2 19.8 0.279 Physical health summary 36.8 11.0 41.0 10.3 0.001 Mental health summary 50.3 10.9 51.4 9.8 0.260 ART SVG arterial and saphenous vein graft; vein graft. SVG saphenous As the population ages and surgical techniques and perioperative care improves, octogenarians comprise an ever-increasing proportion of cardiac surgical practice [19]. Despite the higher mortality and morbidity in this elderly population, operative results have continued to improve [20, 21]. Moreover, despite the higher risks, studies have demonstrated a survival advantage both in the short and the long term for surgical therapy in these patients [22, 23]. Therefore, determination of the optimal surgical approaches for the care of octogenarians has far-reaching implications for both health care and public policy [24]. The IMA graft is widely accepted as the conduit of choice in CABG surgery, primarily because of its superior patency and enhanced long-term survival benefits [5]. Recent studies have demonstrated an acute survival benefit for IMA grafting [6, 7]. Careful evaluation of this phenomenon in elderly patients also supports these findings [8]. Nevertheless, concerns regarding increased operative and pump times, sternal fragility, adequacy of acute postoperative graft flow in acutely ischemic patients, and lack of potential benefits in patients with a shortened life expectancy have generally discouraged the use of arterial grafting in octogenarians. Our data for the past two decades demonstrate no increase in sternal infection, perioperative myocardial infarction, or low cardiac output syndrome, nor any evidence of increased organ system injury in patients undergoing arterial grafting. The only demonstrable differences in morbidity in the present series were an increase in postoperative renal insufficiency and a borderline increase in postoperative respiratory insufficiency in SVG patients. It should be noted that in the present study the two groups were not matched for comorbidities or other measures of operative risk. Nonetheless, the outcomes in this study do not suggest that there is any apparent increase in morbidity rates attributable to the use of the IMA. Is there a demonstrable benefit in the use of the IMA? Early survival benefits from arterial grafting have been shown to diminish with increasing patient age [7, 8]. This corroborates our finding that even though the use of the SVG was found on multivariable analysis to be an independent predictor of hospital mortality, subgroup analysis of patients 84 years of age or older failed to demonstrate any survival advantage for arterial grafting. Even though the findings of Ferguson and associates [8] from the Society of Thoracic Surgeons National Database are compelling, nearly three fourths of their patients were younger than 80 years of age, and no subgroup analysis of octogenarians was conducted. To better define the relationship between arterial grafting and perioperative mortality rates in this cohort of patients, propensity analysis was performed post hoc. The hospital mortality was higher for the SVG than ART SVG patients in four of the five quintiles; however, mortality differences failed to achieve statistical significance in any of the five groupings. This finding suggests one of two possibilities: either there simply is no difference in hospital mortality between SVG and ART SVG patients when the groups are appropriately matched, or the number of patients in each of these subgroups was too small to reveal such a difference. Although the former explanation is certainly plausible, it raises the question of why the use of the SVG only without arterial conduits proved to be a significant predictor of hospital mortality on multivariate analysis. Whereas it is possible that SVG

Ann Thorac Surg KURLANSKY ET AL 2003;76:418 27 ARTERIAL GRAFTING IN OCTOGENARIANS 425 may be a surrogate for some other undefined variable, the number, extent, and clinical relevance of the variables considered in the analysis make this a less attractive explanation. If, on the other hand, there is a relevant clinical difference in perioperative mortality based on the use or nonuse of the LIMA graft but this difference is too small to detect when the study cohort is divided into fifths, then it must be determined why arterial grafting leads to a lower operative mortality. It must further be understood why this survival benefit seems to diminish with increasing age. Early graft failure can be an important cause of inhospital mortality. It has been established that early SVG occlusion occurs in approximately 15% of grafts, whereas IMA grafts acutely occlude only 1% to 2% of the time [25, 26]. Discrepancies in diameter between conduit and native vessel, arterial runoff, and hemostatic or endothelial factors have all been implicated in the greater tendency of SVGs to occlude [27 29]. Because elderly patients tend to have dilated or ectatic veins with impaired flow and more diffuse coronary artery disease with resulting impaired runoff, it should not be surprising that this group of patients may be more vulnerable to early SVG closure and increased hospital mortality [21, 30]. However, as clearly demonstrated in this study, age itself is a significant predictor of hospital mortality. It is therefore not surprising that age is a competing risk factor for mortality. As age increases, the impact of arterial grafting on mortality may diminish and ultimately be overwhelmed by the negative influence of age. As previously noted, for patients 70 years of age and older (mean age, 73 years), the use of arterial grafting has been found to be an independent predictor of improved long-term survival [30]. Although the actuarial curves in the present study begin to diverge at 7 years after discharge from the hospital, the equality of survival distribution for the two groups approached statistical significance. This finding may be influenced by the relatively short period of follow-up (3.8 years for SVG patients and 3.3 years for ART SVG patients), and the consequent small number of patients available for late follow-up. On the other hand late survival may be more heavily influenced by medical therapy or other clinical factors rather than the type of conduit used in these elderly patients. Most long-term studies demonstrate an increasing late survival benefit for IMA grafting with time [9, 10, 31]. In an effort to compare similarly matched groups of patients, propensity analysis technology was applied to late mortality analysis but did not reveal any significant difference between groups of hospital survivors. However, it should be noted that the number of patients in each of the 10 groups (SVG and ART SVG for each of the five quintiles) might have been too small to demonstrate a significant difference with the reduced length of follow-up. Inasmuch as a greater percentage of the arterial grafts were performed in recent years, during a period in which improved surgical techniques, enhanced perioperative care, and decreased mortality rates in octogenarians was occurring, care was taken to include the date of operation in both the multivariable analyses and in the computation of the propensity scores. However, the date of operation was not found to be a significant predictor of early mortality in the present series. The choice of conduit was a significant independent predictor of both early and late mortality. It may be likely that the increasing use of arterial conduits is one of the most compelling reasons for the decreased operative mortality rate observed in the current study. The application of structured questionnaires to measure QOL is essential in an effort to obtain an objective assessment of the patient s personal sense of well-being and functional health status. Quality of life is a complex abstract and multidimensional construct, which encompasses several components including the individual s physical, psychological, and social domains of health. The ultimate goal of CABG surgery in octogenarians is restoration of the patient s physical, social, and psychological functioning. Enhancing the individual s functional status facilitates the ability to perform normal activities of daily living to meet most basic needs, perform usual roles, and maintain a sense of well-being. Patient self-reported health status has been found to be predictive of clinical outcome in patients with symptomatic coronary artery disease [32]. The SF-36 is increasingly being recognized as an excellent gauge of clinical status after CABG surgery [33]. In the present study the SF-36 demonstrated a clear advantage in physical functioning for ART SVG patients compared with SVG patients (p 0.001). In the mental health summary component, ART SVG patients scored higher than SVG patients. Moreover, when ART SVG patients were compared with age-adjusted normal values they scored significantly higher in the physical health summary component (p 0.005) but comparable in the mental health summary component. These QOL outcomes clearly demonstrate that CABG surgery in octogenarians is successful in improving the patient s clinical and functional status. Clinical assessment of functional capacity using the CCS classification system likewise demonstrated excellent outcomes with 95.1% of SVG only and 97.6% of ART SVG in classes 1 or 2. Although these data suggest a strong relationship between clinical assessment and the patient s perceived functional status, the SF-36 instrument clearly yields a more complete evaluation of the patient s health status. These findings support the use of arterial revascularization and, more specifically, IMA grafting in octogenarians. Moreover, the long-term benefits of QOL are apparent in all patients as documented by objective standardized measures. Although compelling, these results are not definitive and must be viewed with caution owing to a series of important study limitations. This series is one of the largest experiences of isolated CABG surgery in octogenarians reported from a single institution. Even so, cohort size clearly limited subgroup analysis, particularly within quintiles. Although the findings of a community-based clinical practice may have broad

426 KURLANSKY ET AL Ann Thorac Surg ARTERIAL GRAFTING IN OCTOGENARIANS 2003;76:418 27 applicability, caution must be exercised, as there may be factors that may not have been directly accounted for which can limit generalizing from these findings. Although efforts were made to account for patient selection criteria through propensity score analysis, this was a retrospective study. Confirming these results with similar patient groups would assist in solidifying the present findings. Despite the long-term follow-up in a number of patients, the mean length of follow-up in the present series was less than half the current life expectancy of an 80-year-old individual in this country. Continued follow-up of the current cohort will enable us to answer many more critical questions regarding the long-term impact of arterial grafting in this cohort of patients. In the present study, arterial grafting in octogenarians has demonstrated reduced operative mortality rates, enhanced QOL, and reduced late cardiac events. Further studies are needed to more clearly delineate the longterm survival benefits of octogenarians undergoing revascularization of the myocardium. The authors thank Debra D. Guest, EdD, for technical assistance in the preparation of this report. References 1. Spencer G. U.S. Bureau of the Census: projections of the population of the United States, by age, sex and race: 1988 2080. Washington, DC: US Government Printing Office; 1989. Current Population Reports Series P-25, No 1018. 2. National Center for Health Statistics. National Health Interview Survey 1983 1985. Hyattsville, Maryland: 1986. 3. Peterson ED, Jollis JG, Bebchuk JD, et al. Changing mortality following myocardial revascularization in the elderly: the National Medicare Experience. Ann Intern Med 1994;121: 919 27. 4. Ferguson TB Jr, Dziuban SW, Edwards FH, et al. The STS National Database. Current changes and challenges for the new millennium. Ann Thorac Surg 2000;69:680 91. 5. Loop FD, Lytle BD, Cosgrove DM, et al. Influence of the internal mammary artery graft on 10-year survival and other cardiac events. N Engl J Med 1986;314:1 6. 6. Edwards FH, Clark RE, Schwarz M. Impact of internal mammary artery conduits on operative mortality in coronary revascularization. Ann Thorac Surg 1994;57:27 32. 7. Leavitt BJ, OConnor GT, Olmstead EM, et al. Use of the internal mammary artery graft and in-hospital mortality and other adverse outcomes associated with coronary artery bypass surgery. Circulation 2001;103:507 12. 8. Ferguson TB Jr, Coombs LP, Peterson ED. Internal thoracic artery grafting in the elderly patient undergoing coronary artery bypass grafting: room for process improvement? J Thorac Cardiovasc Surg 2002;123:869 80. 9. Akins CW, Daggett WM, Vlahakes GJ, et al. Cardiac operations in patients 80 years old and older. Ann Thorac Surg 1997;64:606 15. 10. Craver JM, Puskas JD, Weintraub WW, et al. 601 Octogenarians undergoing cardiac surgery: outcome and comparison with younger age groups. Ann Thorac Surg 1999;67:1104 10. 11. Galbut DL, Traad EA, Dorman MJ, et al. Seventeen-year experience with bilateral internal mammary artery grafts. Ann Thorac Surg 1990;49:195 201. 12. Ware JE, Snow KK, Kosinski M, Gandek B. SF-36 Health survey manual and interpretation guide. Boston: The Health Institute, New England Medical Center, 1993. 13. Ware J, Kosinski M, Keller S. SF-36 Physical and mental health summary scales: a user s manual. 2nd ed. Boston: The Health Institute, New England Medical Center, 1994. 14. Rosenbaum PR, Rubin DB. Reducing bias in observational studies using subclassification on the propensity score. J Am Stat Assoc 1984;79:516 24. 15. Rubin DB. Estimating causal effects from large data sets using propensity scores. Ann Intern Med 1997;127:757 63. 16. Cox DR. Regression models and life-tables. J R Stat Soc [B] 1972;34:187 220. 17. Kaplan EL, Meier P. Nonparametric estimation from incomplete observations. J Am Stat Assoc 1958;53:457 81. 18. Mantel N. Evaluation of survival data and two new rank order statistics arising in its consideration. Cancer Chemother Rep 1966;50:163 70. 19. Alexander KP, Peterson ED. Coronary artery bypass grafting in the elderly. Am Heart J 1997;134:856 64. 20. Peterson ED, Cowper PA, Jollis JG, et al. Outcomes of coronary artery bypass graft surgery in 24,461 patients aged 80 years or older. Circulation 1995;92(Suppl 2):II-85 91. 21. Alexander KP, Anstrom KJ, Muhlbaier LH, et al. Outcomes of cardiac surgery in patients age 80 years: results from the National Cardiovascular Network. J Am Coll Cardiol 2000; 35:731 8. 22. The TIME Investigators. Trial of invasive versus medical therapy in elderly patients with chronic symptomatic coronary-artery disease (TIME): a randomised trial. Lancet 2001; 358:951 7. 23. Graham MM, Ghali WA, Faris LP, et al. Survival after coronary revascularization in the elderly. Circulation 2002; 105:2378 84. 24. Sollano JA, Rose EA, Williams DL, et al. Cost-effectiveness of coronary artery bypass surgery in octogenarians. Ann Surg 1998;228:297 306. 25. Motwani JG, Topol EJ. Aortocoronary saphenous vein graft disease: pathogenesis, predisposition and prevention. Circulation 1998;97:916 31. 26. Russo P, Orszulak TA, Schaff HV, et al. Use of internal mammary artery grafts for multiple coronary artery bypasses. Circulation 1986;74(Suppl 3):III-48 52. 27. O Connor NJ, Morton JR, Birkmeyer JD, et al. Effect of coronary artery diameter in patients undergoing coronary bypass surgery: the Northern New England Cardiovascular Disease Study Group. Circulation 1996;93:652 5. 28. Lesperance J, Bourassa MG, Biron P, et al. Aorta to coronary artery saphenous vein grafts: preoperative angiographic criteria for successful surgery. Am J Cardiol 1972;30:459 65. 29. Kurlansky PA. Is there hypercoaguable state following offpump coronary artery bypass surgery? What do we know and what can we do? J Thorac Card Surg, in press. 30. Gardner TJ, Greene PS, Rykiel MF, et al. Routine use of the left internal mammary artery graft in the elderly. Ann Thorac Surg 1990;49:188 94. 31. Morris RJ, Strong MD, Grunewald ML, et al. Internal thoracic artery for coronary artery grafting in octogenarians. Ann Thorac Surg 1996;62:16 22. 32. Spertus JA, Jones P, McDonell M, et al. Health status predicts long-term outcome in outpatients with coronary disease. Circulation 2002;106:43 9. 33. Kiebzak GM, Pierson LM, Campbell M, Cook JW. Use of the SF36 General Health Status Survey to document healthrelated quality of life in patients with coronary artery disease: effect of disease and response to coronary artery bypass graft surgery. Heart Lung 2002;31:207 13.

Ann Thorac Surg KURLANSKY ET AL 2003;76:418 27 ARTERIAL GRAFTING IN OCTOGENARIANS 427 DISCUSSION DR JAMES W. JONES (Columbia, MO): That was a very well presented paper, and it probably is one of the largest series, if not the largest series, of octogenarians having coronary artery bypass graft operations. This experience is from an institution that pioneered the use of the internal thoracic artery. I think that this is a little different than most of the papers in the past. What we have done in the past is try and show that there was not a higher operative morbidity and mortality associated with the use of this graft in the elderly, and I think you have done very well to show that. There always are methodological problems with a 20-year sample size, because we are not doing things, I hope, the same way we were 20 years ago, and therefore I wonder whether you have corrected adequately for surgeon s preference in determining graft choice? I prefer to use arterial grafts, and I think you have done us all a favor by showing the long-term effects of the benefits in this patient group that has as much right to a superior operation as do the younger patients. I remain doubtful that the difference in early patency rates and not surgeon s preference is the reason for the benefit in operative mortality you have shown. Surgeon s preference likely introduces factors that did not go into your multivariate analysis. Please explain. DR KURLANSKY: Thank you for your comments. You raise two very interesting points. One is the longevity of the study, and that makes it a very difficult database to work with, because even if you wanted to use a risk model, there is no risk model. The risk model from 20 years ago is not the same as 10 years ago, is not the same as currently. One thing that we did find a little bit helpful is in our multivariable analysis we included the date of surgery in the analysis, because we were very concerned about the fact that surgical practices have changed, the mortality is going down from surgery in these patients, and the higher percentage of patients who had the internal mammary artery graft are in the more recent years. So, therefore, maybe it has nothing to do with the fact that they got arterial grafts. It is just a function of the fact that they were operated on in a time span when we are doing better operating on these patients, and in fact it is corroborated by the fact that the mortality is going down and the use of the internal mammary artery is going up. That is why we specifically included both in the multivariable analysis and in the multivariable analysis that went into the propensity scales the date of surgery, and, in fact, date of surgery was not statistically a predictor of mortality or survival either in the perioperative or in the late survival and choice of conduit was. What that tells me is that choice of conduit may be one of the biggest factors that has gone into dropping the operative mortality. The other issue really gets to the core issue, why is this true? If it is true, why is it true? The Society of Thoracic Surgeons came out with this finding, then the Northern New England Group came out with this finding, and more recently, Dr Ferguson, looking at patients 75 years of age and greater, using the Society of Thoracic Surgeons database, also found this advantage, even using propensity scale analysis. His patients are not exactly the same as ours because both he and we found, and others have found, that the incremental benefit for internal mammary artery grafting decreases with increasing patient age, and 75% of his patients were younger than 80 years old. So his cohort is not exactly the same as ours. But, in any event, in all three studies you find that the use of the internal mammary artery graft has a survival benefit, an acute survival benefit. I am not sure I agree with you that the graft patency is the same when they leave the hospital. I cannot prove it to you, I do not have angiographic studies, and I also, unfortunately in this database, cannot tell you the exact cause of death in the patients who died, which would be extremely helpful. Certainly we know from historical studies that have documented graft patency that early graft closure with saphenous vein grafts varies anywhere from 10% to 25% but tends to run around 15%, whereas with internal mammary artery grafts it tends to run 3% or less. So therefore I can not help but speculate that graft patency certainly may have a role in improving even the acute operative mortality in these patients. DR FREDERICK L. GROVER (Denver, CO): I thought your data on the quality of life as measured by the SF-36 were quite interesting. I would like to know, however, whether you measured quality of life preoperatively (ie, baseline) so that you could evaluate change from baseline postoperatively. This is important because there might have been a difference between your two groups as the vein group appears to be a higher risk group of patients. DR KURLANSKY: It is an excellent point, and we do not. The only thing we tried to deal with that a little bit was to use age-matched controls, and even the age-matched controls are not exact because the age-matched controls that come from the SF-36 data are age 75 and greater, and so many of those patients are not even 80 years old. The few people who have studied using a baseline study preoperatively and postoperatively in coronary artery bypass graft surgery have shown a dramatic improvement, but whether or not we started out with a lower quality-of-life group of patients is undetermined and an excellent point.