| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Ann Thorac Surg 2007;83:2036-2043
© 2007 The Society of Thoracic Surgeons
a Department of Cardiothoracic Surgery, University Hospital, Linköping, Sweden
b Department of Cardiothoracic Anesthesia, University Hospital, Linköping, Sweden
Accepted for publication January 22, 2007.
* Address correspondence to Dr Svedjeholm, Department of Cardiothoracic Surgery, University Hospital, Linköping, SE-581 85, Sweden (Email: rolf.svedjeholm{at}lio.se).
 |
Abstract |
|---|
 Top Abstract IntroductionPatients and MethodsResultsCommentAcknowledgmentsReferences |
|---|
Methods: All patients undergoing aortic valve replacement (AVR) for aortic stenosis from 1995 through 2000 in the southeast region of Sweden (n = 398) were compared with a cohort, matched for age and sex, undergoing coronary artery bypass grafting (CABG [n = 398]). Risk factors for 5-year mortality were analyzed.
Results: Forty-five AVR and 47 CABG patients required treatment for postoperative heart failure. Thirty-day, 1-year, and 5-year mortality in patients with and without postoperative heart failure after AVR were 6.7% versus 1.4% (p = 0.05), 8.9% versus 4.0% (p = 0.13), and 42.2% versus 14.2% (p < 0.0001) respectively. Corresponding results in the CABG group were 21.3% versus 1.1% (p < 0.0001), 25.5% versus 3.1% (p < 0.0001), and 36.2% versus 11.1% (p = 0.0015). Postoperative heart failure, preoperative renal dysfunction, procedure-associated stroke, body mass index less than 19 kg/m2, older age, preoperative atrial fibrillation, and preoperative anemia turned out as independent risk factors for 5-year mortality after AVR. In the CABG group, postoperative heart failure, diabetes mellitus, older age, and procedure-associated stroke emerged as independent risk factor for 5-year mortality.
Conclusions: Postoperative heart failure was associated with high early mortality after CABG whereas the grave consequences of postoperative heart failure after AVR for aortic stenosis became evident only with time.
|
Introduction |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentAcknowledgmentsReferences |
|---|
|
Patients and Methods |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentAcknowledgmentsReferences |
|---|
Forty-five patients undergoing AVR (11.3%) and 47 patients undergoing CABG (11.8%) fulfilled criteria for PHF. Demographic and periprocedural data were registered prospectively in a computerized institutional database (Summit Vista for Windows; version 1.98.1; Summit Medical Systems Inc, Minneapolis, MN). All fields were defined in a data dictionary. Data on late mortality and cause of death was retrieved from the Swedish Civil Registry and the Swedish Cause of Death Registry respectively, and included follow-up to January 2006. Average follow-up time was 7.2 ± 1.7 years (range, 5.2 to 11.2). One AVR patient was lost to follow-up. Data fields used in the univariate and multivariate analyses were 99.4% and 99.9% complete, respectively.
The study was performed according to the Helsinki Declaration of Human Rights and was approved by the Ethics Committee for Medical Research at the University Hospital of Linköping. Owing to the nature of the study, the Ethics Committee waived the need for patient consent.
Definitions
Postoperative heart failure was defined as a hemodynamic state secondary to pump failure unable to meet systemic demands without treatment other than correction of volume or vascular resistance. A low cardiac output can be sufficient to supply the body demands in an anesthetized or sedated patient and, hence, reliance on markers for adequate circulation, in particular mixed venous oxygen saturation (SvO2), and echocardiographic evaluation rather than fixed hemodynamic criteria were employed to diagnose PHF [6–8]. The relationship between SvO2 and systolic arterial pressure (SAP) [SvO2 < 50% and SAP < 130 mm Hg; SvO2 < 55% and SAP < 110 mm Hg; SvO2 < 60% and SAP < 90 mm Hg; SvO2 < 65% and SAP < 70 mm Hg] after correction of shivering and hypovolemia provide our guidelines to recognize inadequate circulation. In the majority of patients, PHF was evident at weaning from cardiopulmonary bypass, with inability to wean from cardiopulmonary bypass or deteriorating circulation and increasing filling pressures after weaning from cardiopulmonary bypass. In the remaining patients, echocardiographic evidence of left ventricular or right ventricular dysfunction, or both, associated with the signs of inadequate circulation were used to diagnose PHF. Treatment consisted of continuous infusion of inotropes longer than 30 minutes with or without intra-aortic balloon pump and usually also adjunctive metabolic support with glucose-insulin-potassium or intravenous glutamate, or both [6].
Emergency operation was defined as a procedure that could not be postponed to the following day and was therefore usually performed immediately but not later than 24 hours from acceptance. Urgent operations were defined as scheduled procedures on patients unable to leave the hospital because of clinical condition. The effective orifice areas for the different prostheses used in this study are based on in vivo Doppler echocardiographic measurements reported in the literature [5].
Complications presented refer to in-hospital events occurring at our institution. Intraoperative myocardial infarction was diagnosed by biochemical markers of myocardial injury or by findings at autopsy [5]. Severe systolic left ventricular dysfunction corresponds to an ejection fraction of 0.30 or less.
Preoperative unstable hemodynamic state was defined as a circulatory state requiring either inotropic treatment or mechanical circulatory assist immediately before surgery. Anemia was defined according to reference values given by the Linkoping University Hospital laboratory (blood hemoglobin less than 134 g/L for men and less than 117 g/L for women).
Early and late mortality was defined as mortality occurring within and later than 30 days from surgery, respectively.
Statistical Analysis
To analyze risk factors for mortality, Cox regression was considered. However, the proportionality assumption was not met, and therefore univariate and multivariate logistic regression was chosen to evaluate impact of risk factors on 5-year survival. Variables were tested one at a time, and those with a p value less than 0.25 and those that previously had been shown to influence long-term survival after cardiac surgery were then tested with forward stepwise multiple logistic regression. Results are given as odds ratios (OR) with 95% confidence intervals (CI). Hosmer-Lemeshow goodness-of-fit statistics were calculated for the final models. Cumulative long-term survival was assessed with Kaplan-Meier analysis. Fisher’s exact test was used for comparison of dichotomous variables, and the Mann-Whitney U test was used for comparison of continuous variables. Statistical significance was defined as p less than 0.05. Statistical analyses were performed with Statistica 6.0 (StatSoft, Tulsa, Oklahoma), and SPSS 14.0 (SPSS, Chicago, Illinois).
|
Results |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentAcknowledgmentsReferences |
|---|
|
Mortality
Crude mortality for AVR and CABG patients, with and without PHF, respectively, are shown in Table 2, and cumulative survival is demonstrated in Figure 1. Causes of death are given in Table 3. The average time to death within the 5-year follow-up for PHF patients in the AVR and CABG group was 2.5 ± 1.5 and 0.9 ± 1.5 years, respectively (p = 0.007).
|
|
|
Risk Factors for 5-Year Mortality After AVR for Aortic Stenosis
Postoperative heart failure emerged as the independent risk factor, with the lowest p value for both overall 5-year mortality and for late mortality (between 30 days and 5 years). Preoperative renal dysfunction, periprocedural stroke, body mass index less than 19 kg/m2, age, preoperative atrial fibrillation, and preoperative anemia also turned out as independent risk factors for overall 5-year mortality and late mortality (between 30 days and 5 years). Univariate and multivariate risk factors for mortality with Hosmer-Lemeshow goodness of fit for the models are presented in Tables 1 and 4.
|
|
|
Comment |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentAcknowledgmentsReferences |
|---|
The valve patients studied represented the largest homogenous cohort of valve procedures in our clinical practice and included all patients with aortic stenosis without clinically significant regurgitation or associated coronary artery disease within an area of 1 million inhabitants operated on during a 5-year period. Therefore, no referral selection bias should be present. The matching procedure resulted in a CABG cohort with a higher proportion of women and higher average age than in our general CABG population, which should be considered when comparing the results for CABG patients with other studies.
Although PHF has been recognized as a major cause of in-hospital mortality after cardiac surgery, its long-term consequences have not been described previously. The lack of a universally accepted definition of PHF may partly explain the shortage of studies addressing this issue. Therefore, our definition of PHF also may be debated. Nevertheless, the patients studied represented those 10% of our patients who had the poorest hemodynamic state due to pump failure at weaning from cardiopulmonary bypass or during the early postoperative phase.
In CABG patients, PHF has previously been reported to be the major cause of early postoperative mortality [1, 2]. Our results are consistent with these findings. The survival curves and the fact that PHF disappeared as a significant risk factor for mortality occurring after 30 days corroborates that PHF was a risk factor mainly for early mortality after CABG, although some influence on late mortality cannot be excluded. Procedure-associated stroke, diabetes, and age were the only significant independent risk factors for mortality occurring between 30 days and 5 years after surgery. These risk factors are not controversial and agree with previous experience, although our CABG cohort was matched with regard to sex and age for patients operated on for aortic stenosis [9–12].
Our study demonstrates differences in temporal patterns of mortality after PHF in CABG and AVR patients. In a recent study, we found that PHF after CABG was more closely related to ischemic events during early stages of surgery and intraoperative myocardial infarction, which could explain the less favorable short-term outcome [4]. In contrast, PHF after AVR for aortic stenosis appears as a comparatively benign condition looking at short-term mortality. The true importance of this complication becomes evident only with time as demonstrated by the Kaplan-Meier analysis (Fig 1) with equalized mortality in patients with PHF in both groups after approximately 3 to 4 years.
The role of PHF for long-term mortality was evaluated in relation to other risk factors and periprocedural events. In the AVR group, PHF emerged as the risk factor with highest statistical significance both for overall and late mortality within the 5-year follow-up. Several investigators have reported risk factors for operative and in-hospital mortality after AVR [10, 13–19]. Although most studies on long-term follow-up after AVR have focused on valve performance or valve-related complications, a few also reported on risk factors for long-term mortality [13, 20, 21]. Important information about the predictive value regarding risk factors known at admission has been gained. In contrast to most previous studies, our study is characterized by its homogeneous cohort of aortic stenosis patients, and by an analysis that takes off from variables known at discharge from hospital with the intention to evaluate the impact of PHF and other periprocedural events on long-term outcome. Addressing postoperative mortality from this angle, the role of periprocedural stroke is not surprising [9]. Furthermore, age, preoperative renal dysfunction, preoperative atrial fibrillation, low body mass index, and anemia have been reported as risk factors for mortality after cardiac surgery [10, 13–20, 22–25]. Choice of prosthesis and patient-prosthesis mismatch have been shown to play a role in postoperative hemodynamics and regression of left ventricular mass, whereas data regarding its impact on clinical outcome and survival are equivocal [26–29]. In our study, there was a trend toward significant association between small effective orifice area and overall mortality in the univariate analysis only. Evidently, other variables emerged as factors of greater importance for long-term outcome. These findings are consistent with the views of the Ad Hoc Liaison Committee for Standardizing Definitions of Prosthetic Heart Valve Morbidity, which has acknowledged that patient variables may be more responsible for outcome than prosthesis-related factors [30].
It can be argued that the event of PHF is only a substitute for other risk factors and underlying causes of PHF. In a previous study, we identified five preoperative variables (hypertension, history of congestive heart failure, severe systolic left ventricular dysfunction, pulmonary hypertension, and preoperative hemodynamic instability) and two intraoperative variables (aortic cross-clamp time, intraoperative myocardial infarction) as independent risk factors for PHF after AVR for aortic stenosis [5]. To account for these risk factors, they were also tested in the multivariate model; however, none turned out as an independent risk factor for 5-year mortality. On the other hand, the event of PHF per se may not necessarily be the cause of the poor outcome, but it possibly identifies patients with reduced life expectancy within different risk groups with higher accuracy than the respective risk factors by themselves. A recent study demonstrated a similar sustained negative influence of preoperative left ventricular diastolic dysfunction on long-term survival [27]. It was speculated that diastolic dysfunction represented nonreversible structural myocardial abnormalities, although the authors acknowledged the lack of direct histopathologic evidence. Regardless of the causal relationships, we emphasize the importance of PHF as a prognostic indicator for long-term survival after surgery for aortic stenosis.
There is an ongoing discussion regarding the optimal timing of surgery and whether asymptomatic patients with severe aortic stenosis should be operated on. Particularly in elderly patients, the transition from asymptomatic to symptomatic stage may be difficult to detect [31]. We suggest that the recognition of the prognostic implications of PHF warrants more meticulous surveillance of asymptomatic patients with regard to previously identified risk factors for PHF [5]. Furthermore, in symptomatic patients with aortic stenosis, surgery should be performed without undue delay to optimize the benefit of valve surgery.
The serious consequences of PHF after surgery for aortic stenosis also call for periprocedural efforts to avoid PHF and to evaluate treatment of PHF. The latter evidently requires long-term follow-up for appropriate assessment. Efforts to reach consensus about diagnostic criteria for PHF that are easily applicable in routine clinical setting are desirable to facilitate future research on this issue. The event of PHF deserves to be emphasized in the discharge records to alert the physician responsible for postoperative follow-up. To conclude, we suggest that the recognition of PHF as an important long-term prognostic factor after surgery for aortic stenosis should have implications not only for preoperative surveillance and timing of surgery but also for perioperative management and postoperative follow-up.
|
Acknowledgments |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentAcknowledgmentsReferences |
|---|
|
References |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentAcknowledgmentsReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  J. Holm, E. Hakanson, F. Vanky, and R. Svedjeholm Mixed venous oxygen saturation predicts short- and long-term outcome after coronary artery bypass grafting surgery: a retrospective cohort analysis Br. J. Anaesth., September 1, 2011; 107(3): 344 - 350. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Gada, P. A. Scuffham, B. Griffin, and T. H. Marwick Quality-of-Life Implications of Immediate Surgery and Watchful Waiting in Asymptomatic Aortic Stenosis: A Decision-Analytic Model Circ Cardiovasc Qual Outcomes, September 1, 2011; 4(5): 541 - 548. [Abstract] [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 |
