|
|
||||||||
Department of Cardiac and Thoracic Surgery, Flinders Medical Centre, Flinders Private Hospital, Ashford Hospital, and the Flinders University of South Australia, Adelaide, South Australia, Australia
Accepted for publication October 6, 2009.
* Address correspondence to Dr Baker, Cardiac Surgery Research, Level 6, Flinders Private Hospital, Adelaide, South Australia, 5042, Australia (Email: rob.baker{at}flinders.edu.au).
| Abstract |
|---|
|
|
|---|
Methods: Octogenarians undergoing coronary artery bypass grafting (CABG) ± concomitant valvular procedure between 1992 and 2005 from three institutions were included in this study. The survival data of 606 octogenarians (414 isolated CABG, 192 concomitant valve procedures) were analyzed with multivariable proportional hazard models.
Results: There were 271 deaths and 2,675 person years of survival for analysis, and median follow-up was 7.15 years (95% confidence interval 6.47 to 7.82 years). Five-year survival for isolated CABG and concomitant valve procedures was 66.5% and 61.5%, respectively. An increase in mortality risk was attributable to older age, hypercholesterolemia, severely impaired left ventricular function, tobacco smoking history and high creatinine (
0.15 mmol/L). Time spent on cardiopulmonary bypass was the only intraoperative risk factor associated with an increase in mortality risk (hazard ratio 1.01, 95% confidence interval: 1.00 to 1.02; p < 0.001).
Conclusions: This study showed that from the intraoperative parameters examined only time spent on cardiopulmonary bypass was associated with long-term survival. Surgeons may be assisted in patient selection by identifying the factors that influence long-term survival among octogenarians and development of a preoperative risk model specific for this age group.
| Introduction |
|---|
|
|
|---|
Compounding this information gap it cannot be assumed that the intraoperative risk factors for mortality identified in younger cardiac surgery cohorts will necessarily translate to octogenarians who may undergo substantially different perfusion protocols [3]. For example, Willcox and van Uden [3] commented in their qualitative review that little is known about the role of perfusion parameters on elderly patient survival and only selected studies have reported multiple intraoperative risk factors such as cardiopulmonary bypass (CPB) time and transfusion of red blood cells [4] or revascularization with the left internal mammary artery (LIMA) [5, 6]. Thus, describing the preoperative risk factors for survival among octogenarian cardiac surgery patients in conjunction with potential intraoperative and perfusion risk factors may assist cardiothoracic surgical teams in deciding both suitable octogenarian candidates for surgery and optimal surgical strategies. The aim of this analysis was therefore to determine the influence of preoperative and intraoperative factors on long-term survival among octogenarian cardiac surgery candidates in a series of consecutive patients from three centers.
| Material and Methods |
|---|
|
|
|---|
Anesthetic, Surgical, and Cardiopulmonary Bypass Technique
Anesthetic technique consisted of midazolam, pancuronium, and fentanyl for induction and maintenance with isoflurane or sevoflurane, nitrous oxide, and (or) propofol as required. Before aortic cannulation, heparin was given at a dose of 300 IU/kg to achieve a target activated clotting time of 400 seconds or greater before commencement of CPB.
After median sternotomy, and harvesting of arterial and (or) venous conduit, CPB was instituted using an ascending aortic and either a two-stage right atrial or bicaval cannulation. Cardiopulmonary bypass was performed utilizing roller pumps and the circuit included a hard shell membrane oxygenator, PVC or biopassive tubing (SMARxT; Cobe Cardiovascular, Arvada, CO), and a 40 micrometer arterial line filter. Routine CPB protocol included nonpulsatile flow (rate 1.8 to 2.4 Lpm/m2), alpha-stat pH management, gravity venous drainage, and systemic temperature management (28°C to 36°C). Myocardial protection was achieved by using intermittent antegrade hyperkalemic cardioplegia. The initial or induction dose of cardioplegia was given for 2 minutes (250 mL/minute), and then the maintenance dose was given approximately every 20 minutes as required through the procedure. Seventy-three cases were performed with electric fibrillation of the heart. Attempts were made at all procedures to revascularize all vessels deemed operable by the respective surgeons. The heart was arrested and the target coronary artery was opened, and distal anastomoses between the bypass graft and native coronary artery were performed under aortic cross-clamping followed by valve surgery if indicated. Proximal anastomoses were performed with partial aortic clamping. Gradual weaning from bypass started after completion of the proximal anastomoses. Cardiotomy suction was utilized only in valve procedures. Patients were rewarmed at rates not in excess of 1°C per minute and separated from CPB when nasopharyngeal temperatures exceeded 36.5°C. On completion of all anastomoses and weaning off CPB, protamine was given to return the activating clotting time to preoperative levels. At the end of surgery patients were transferred to the intensive care unit and managed according to unit protocol.
Mortality Assessment
Long-term survival was ascertained from the National Death Index provided by the Australian Institute of Health and Welfare for use in epidemiologic studies and medical research. National Death Index data provided all cause mortality up to December 31, 2006 and this date was taken as the censor date for patient survival, enabling a minimum of 12-month follow-up.
Statistical Analysis
Statistical analyses were performed using SPSS 15.0 (SPSS Inc, Chicago, IL). The early mortality rates 30 days or less across the study period were evaluated for linear trend. Potential preoperative risk factors for long-term mortality were evaluated with the Mantel-Cox log-rank test for categoric variables and univariable Cox proportional hazard models for continuous variables to ascertain the association with survival. The risk factors examined were principally derived from previous survival research among octogenarian and nonoctogenarian series [2, 4–11], determined by prevalence in the sample and according to available data across the study period, and included the following: age (continuous variable), concomitant valvular procedure (versus isolated CABG), urgency of operation (elective versus emergency, urgent), congestive heart failure, renal disease, female sex, cerebrovascular disease, myocardial infarction less than 30 days preoperatively, reoperative procedure, hypercholesterolemia, diabetes mellitus (type I or type II vs none), peripheral vascular disease, history of tobacco smoking, left ventricular ejection fraction ([LVEF] normal [0.60] category versus impaired [0.31 to 0.59], severe [<0.30]), body mass index (normal vs underweight, overweight, obese, morbidly obese), minimum preoperative hemoglobin (Hb), and preoperative creatinine prior to institution of CPB (analyzed categorically as quintiles). The hospital institution was evaluated by an interaction term with year of operation. The prevalence of intraaortic balloon pump in the sample was low (n = 7, 1.2% of total) and thus this variable was not considered.
Due to regression model constraints, only preoperative covariates p less than or equal to 0.20 were retained for further analysis to establish a baseline hazard model of risk factors and to avoid overfitting. Utilizing the mean of covariates method these covariates were entered into a proportional hazard model in block fashion at the first step (model 1). Given the particular interest in intraoperative covariates, these were forced into the regression model at the second step (model 2) regardless of univariable association and included the following: LIMA for revascularization, defibrillation, total time on CPB (minutes), maximum arterial outlet temperature (°C), minimum hematocrit value, minimum nasopharyngeal temperature (°C), and use of blood products intraoperatively (transfusion of red blood cells, frozen plasma), and total number of grafts.
Multicollinearity statistics were deemed acceptable from examination of squared multiple correlations less than 0.90 and inspection of correlations between regression coefficients. Proportionality of hazards was inspected by covariate interactions with time, and examination of the baseline hazards function plot and log-minus-log plot of survival function.
| Results |
|---|
|
|
|---|
|
|
2 (13) = 8.45, p = 0.81, indicating that early survival rates did not significantly change across the observation period.
Long-Term Survival: Preoperative Risk Factors
The Kaplan-Meier plot of actuarial survival is shown in Figure 1
for isolated CABG and concomitant valve procedures. Survival for isolated CABG patients at 1, 3, and 5 years was 91%, 77.8%, and 66.5%, respectively, and for valve-CABG patients at 1, 3, and 5 years was 86.2%, 74.4%, and 61.5%, respectively. By comparison, the expected survival for Australian males aged 80 to 84 is 7.9 years, while for females of the same age it is 9.7 years, and for males and females aged 85 years or greater the life expectancy is 5.6 and 6.9, respectively [12].
|
Long-Term Survival: Intraoperative Risk Factors
Inspection of the intraoperative risk factors presented in Table 1 showed that time on CPB, LIMA use, and minimum nasopharyngeal temperature were associated with mortality in univariable analysis. In the multivariable analysis depicted in Table 2 (model 2), however, adjustment for other risk factors showed that only CPB time remained associated with greater mortality risk (hazard ratio 1.01, 95% CI: 1.00 to 1.02; p < 0.001). With regard to bypass time, a 1% increased mortality risk was evident for every 1 minute increase in bypass time, and this translated to a 10.4% increase in mortality risk for an increase in bypass time of 10 minutes. The mortality risk attributable to preoperative risk factors evident in model 1 was not attenuated for age, hypercholesterolemia, severely impaired LVEF, smoking, and the highest creatinine quintile as observed in model 2.
| Comment |
|---|
|
|
|---|
Internal mammary artery grafts were used for revascularization in 37.5% of patients supporting the claim that uptake in the elderly has been disproportionate with LIMA's more commonly employed in younger patients [7]. The results with respect to the LIMA did not confirm favorable survival among octogenarians but suggested a trend toward lower mortality risk. Given that the number of grafts was also not associated with survival a potential explanation is that the variance attributable to total graft number was confounded by the significant effect found for time spent on CPB, which is a general marker of operative complexity.
These results suggested that other than total CPB time the remaining CPB-related and intraoperative risk factors did not significantly influence the risk of all-cause mortality in the long term after adjustment. Such intraoperative risk factors may reflect acute intraoperative risk and are perhaps associated with early morbidity. For example, intraoperative red blood cell transfusion is associated with low-output heart failure [13], while CPB temperature management may be related to stroke [14]. Octogenarians have been reported to present for CABG with lower hematocrit than nonoctogenarians, and increased age is associated with greater need for transfusion and packed red blood cells [8]. Hematocrit levels and requirement for intraoperative blood products bore no significant influence on long-term survival in this sample, though were recently implicated in long-term survival after cardiac surgery in adults [15, 16].
The findings with respect to traditional risk factors supported work elsewhere suggesting that many risk factors for late death among octogenarians are noncardiac comorbidities [5–7, 17, 18]. Specifically, age was a risk factor for mortality supporting other studies elsewhere that reported short-term mortality [8, 19] and late mortality [17, 20]. However, this result is not universal as, for example, Kamiya and colleagues [21] found that increasing age affected neither prognosis nor quality of life among octogenarians receiving CABG or percutaneous coronary intervention. Of the indicators of renal insufficiency, low creatinine was associated with a reduced mortality risk whereas the highest quintile of creatinine was associated with an increased mortality risk. This parallels the findings of de Vincentiis and colleagues [22] who showed that preoperative renal insufficiency (creatinine > 2.00 mg/dL) was a predictor of hospital mortality in octogenarians undergoing aortic valve replacement and further supports the impact of poor renal function as a predictor of short-term [23] and long-term [24] mortality. The results with regard to LVEF support previous series on operative mortality [25] and late mortality [26, 27] as it is well documented that a severely impaired left ventricle portends greater mortality risk after cardiac surgery as does congestive cardiac failure [7].
Limitations
The results are presented with several limitations, including the retrospective design, examination of concomitant valve procedures, and also the sample size that was, however, comparable with other studies [7, 8, 18]. There was a potential unknown selection bias due to the lack of information on octogenarian patients considered for surgery but not operated on. The univariable selection of preoperative risk factors for multivariable hazard models may limit the conclusions drawn about potential risk factors. For example, the recent work by Billah and colleagues [28] suggested that along with the risk factors explored here, both New York Heart Association class and use of inotropes were important variables in determining preoperative mortality risk in cardiac surgery candidates. The final limitation of this report relates to the scope of perfusion-related data as continuous variables such as flow, pressure, and other parameters could not be analyzed with respect to mortality. Multicenter and international partnerships such as the Perfusion Downunder Collaboration and the International Consortium for Evidence Based Perfusion [29], through prospective multicenter registries, will be able to provide access to a broader range of perfusion and intraoperative data in the future.
Traditional risk factors increased the risk of mortality among octogenarians after cardiac surgery, and perfusion and intraoperative risk factors were not found to significantly influence mortality risk other than the total CPB time. In general there is limited information from large series documenting the role of intraoperative risk factors combined with preoperative variables in determining long-term survival in octogenarian cardiac surgery patients. As an increasing number of octogenarians are referred for cardiac surgery, closer examination of these factors will aid surgeons in deciding which patients are suitable for cardiac surgery and further inform suitable intraoperative management. Identification of a range of preoperative and intraoperative risk factors may allow optimization of patient characteristics before surgery that may translate to a survival benefit. Octogenarians can also be informed about the individual risks imposed by surgery and provide a clearer scope for long-term recovery.
| Acknowledgments |
|---|
|
|
|---|
| References |
|---|
|
|
|---|
Related Article
This article has been cited by other articles:
![]() |
A. Saxena, D. T. Dinh, C.-H. Yap, C. M. Reid, B. Billah, J. A. Smith, G. C. Shardey, and A. E. Newcomb Critical Analysis of Early and Late Outcomes After Isolated Coronary Artery Bypass Surgery in Elderly Patients Ann. Thorac. Surg., November 1, 2011; 92(5): 1703 - 1711. [Abstract] [Full Text] [PDF] |
||||
![]() |
W. A. Ahmed, P. J. Tully, J. L. Knight, and R. A. Baker Female Sex as an Independent Predictor of Morbidity and Survival After Isolated Coronary Artery Bypass Grafting Ann. Thorac. Surg., July 1, 2011; 92(1): 59 - 67. [Abstract] [Full Text] [PDF] |
||||
![]() |
D. J. LaPar, C. M. Bhamidipati, T. B. Reece, J. C. Cleveland, I. L. Kron, and G. Ailawadi Is off-pump coronary artery bypass grafting superior to conventional bypass in octogenarians? J. Thorac. Cardiovasc. Surg., January 1, 2011; 141(1): 81 - 90. [Abstract] [Full Text] [PDF] |
||||
![]() |
C. W. Akins Invited Commentary Ann. Thorac. Surg., January 1, 2010; 89(1): 111 - 111. [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |