|
|
||||||||
a Clinical Research Unit, Division of Cardiothoracic Surgery, Joseph B. Whitehead Department of Surgery, Atlanta, Georgia
b Department of Biostatistics, Rollins School of Public Health, Emory University School of Medicine, Atlanta, Georgia
Accepted for publication December 30, 2010.
* Address correspondence to Dr Thourani, Cardiothoracic Surgery, Crawford Long Hospital, 6th Floor, Medical Office Tower, 550 Peachtree St, Atlanta, GA 30308 (Email: vthoura{at}emory.edu).
Presented at the Poster Session of the Forty-sixth Annual Meeting of The Society of Thoracic Surgeons, Fort Lauderdale, FL, Jan 25–27, 2010.
| Abstract |
|---|
|
|
|---|
Methods: A retrospective review of patients with end-stage renal disease undergoing valve replacement from January 1996 through March 2008 at Emory Healthcare Hospitals was performed. Outcomes were compared using
2 tests and 2-sample t tests. Adjusted long-term survival up to 10 years was assessed with Kaplan-Meier plots and compared between biologic and mechanical replacements using the Cox proportional hazards model.
Results: A total of 202 patients underwent 211 valve replacement operations. Patient age was 20 to 83 years (mean age, 54.8 ± 14.0); 115 of 211 (54.5%) were male. Operations included the following: 100 of 211 (47.4%) isolated aortic; 49 of 211 (23.2%) isolated mitral; 4 of 211 (1.9%) isolated tricuspid; and 58 of 211 (27.5%) combined replacements. Thirteen (6.2%) patients underwent reoperative valve replacements. Most patients received bioprosthetic valves (143 of 211, 67.8%), while 68 of 211 (32.2%) received mechanical valves. Concomitant coronary artery bypass was performed in 53 of 211 (25.1%) patients. Thirty-day mortality was in 42 of 211 patients (19.9%) and was not different between bioprosthetic and mechanical replacements. Overall 10-year survival was 18.1% for all patients and was not influenced by valve type implanted.
Conclusions: For patients with end-stage renal disease treated with dialysis, valve replacement carries acceptable operative mortality. Long-term survival is similar among patients receiving bioprosthetic versus mechanical valve replacement. Careful risk assessment and choice of valve prosthesis should be performed prior to surgical intervention in this high-risk patient population.
| Introduction |
|---|
|
|
|---|
| Dr Thourani discloses that he has financial relationships with Edwards Lifesciences, Medtronic, Sorin, and St. Jude.
|
The prevalence of end-stage renal disease (ESRD) continues to increase in the United States. The 2009 annual data report from the USDRS [United States Renal Data System] revealed that as of 2007 there were 527,282 patients with ESRD in the United States [1]. Projections from this report estimate that the prevalence of ESRD patients in the US will approach 775,000 by the year 2020. This steady increase in ESRD patients parallels similar trends that have been seen around the world [2, 3]. Furthermore, the USDRS data report reveals that prevalence rates are increasing most rapidly among patients aged 65 and older.
Cardiovascular disease and its complications are the most common cause of death among ESRD patients, especially among the older age groups. The multiple shared risk factors that lead to the development of cardiovascular and renal disease cause the two disease processes to often exist and progress in concert with one another [4, 5]. Once patients reach ESRD and dialysis dependence, often they would have already developed significant symptoms related to coronary artery disease, heart failure, and valve-related cardiac abnormalities [6].
Although multiple studies have examined outcomes in patients with ESRD undergoing cardiac surgery, many of these studies have examined outcomes after all types of cardiac surgery and not focused solely on outcomes after valve-related procedures [3, 7–16]. Therefore, the purpose of the study is to investigate short-term and long-term outcomes in patients with preoperative dialysis undergoing cardiac valve surgery. Furthermore, we investigated the impact of biologic versus mechanical valve in those patients undergoing isolated aortic or mitral valve prosthesis.
| Material and Methods |
|---|
|
|
|---|
Cardiac catheterization was performed in all patients over the age of 40 or in younger patients with risk factors for coronary artery disease. The nephrology service was consulted prior to surgical intervention and managed the patients for renal replacement and electrolyte management in all patients. The infectious disease service was consulted prior to surgical intervention in those suspected or with known history of endocarditis. Standard cardiopulmonary bypass (CPB) techniques for valve operations were used in all patients. Per institutional protocol, epiaortic ultrasound was performed prior to central aortic cannulation and cross-clamp placement. When extensive aortic calcification was encountered, axillary or femoral artery cannulation was performed for arterial inflow. Surgical approach including valve insertion techniques and conduct of CPB and myocardial protection were left to the discretion of the 15 attending faculty cardiac surgeons active in the study. Typically, conventional CPB was performed utilizing roller head pumps, membrane oxygenators, cardiotomy suction, arterial filters, cold antegrade and retrograde blood cardioplegia, and moderate systemic hypothermia (32°C to 34°C). Near the discontinuation of CPB, a modified version of zero-balance ultrafiltration was initiated on all patients. The operative field was routinely flooded with carbon dioxide and removal of air maneuvers were performed in all cases prior to releasing the cross-clamp. Institution of postoperative CCRT [continuous renal replacement] and electrolyte management while on CRRT was at the discretion of the treating nephrologist. Postoperative extubation was generally performed after dialysis.
Patients were classified according to their valve procedure and the type of implant received. During this period the medical records of 202 qualified patients, representing 211 valve operations, were analyzed. Of these, 100 (47.4%) were isolated aortic valve replacement (AVR) cases, 49 (23.2%) were isolated mitral surgery, and the remaining 62 (29.4%) were either isolated tricuspid or pulmonic valve operations or double valve operations. Four double valve patients received both a mechanical and a biologic valve but were classified for comparison purposes as biologic valves. The Emory University Institutional Review Board approved this study and waived consent.
Measurements
Extracted STS records included demographic data, preexisting conditions, intraoperative variables, and clinical outcomes. Prior to analysis, 24 preoperative and operative risk factors were harvested from the STS database. Standard STS definitions for each risk factor and outcome were used. Race was dichotomized as either Caucasian or non-Caucasian. Chronic lung disease was ordinally measured in some latter years and dichotomously measured in earlier years; in this study it was dichotomized. New York Hospital Association (NYHA) heart failure classification was dichotomized as class III-IV or I-II. The STS predicted risk of mortality was gathered where available; it was missing for 79 patients (37.4%) undergoing operations for which the predicted risk formula has not been validated. Operative variables included CPB time, cross-clamp time, and an indicator for whether an intraaortic balloon pump was inserted intraoperatively.
The primary outcomes examined in this study were in-hospital mortality and major adverse cardiac events ([MACE], a composite of death, permanent stroke, and myocardial infarction) as well as long-term survival; however, other complications were collected and summarized. The SSDI [Social Security Death Index] was incorporated into our STS database to identify death dates for deceased patients as of December 31, 2007. Cause of death was not available for any of the patients; those patients still alive on this date were considered censored.
Data were 100% complete for each valve surgery type, valve implant type, and all major postoperative hospital outcomes. Data were missing for the following preoperative variables: Caucasian race (n = 14, 6.6%); ejection fraction (n = 38, 0.180); New York Heart Association classification (n = 42, 19.9%); last creatinine level (n = 27, 12.8%); and STS predicted risk of mortality (n = 79, 37.4%). A multiple imputation algorithm was employed where necessary to impute missing values so that the whole sample could be analyzed in a multivariable fashion (Schaffer [17] and Molenberghs and Kenward [18]).
Statistical Analysis
Kaplan-Meier curves were generated that provided unadjusted survival estimates at postoperative points in time. Survival differences in surgical strata were determined by log-rank tests.
To statistically evaluate the effects of valve implant type (mechanical or biologic) on short-term and long-term outcomes, multivariable regression models were constructed separately for all patients and then for subsets undergoing isolated AVR (n = 100) or isolated mitral valve surgery (n = 49). For in-hospital death and MACE, logistic regression models were constructed for each subset. For long-term survival endpoints, Cox proportional hazards regression models were fit for each subset. Each logistic and Cox model were adjusted for the same preoperative covariates to control for potential selection bias: age, chronic lung disease, cerebrovascular accident, ejection fraction, infectious endocarditis, gender, heart failure, concomitant CABG, body mass index, and diabetes mellitus. Adjusted odds ratios and hazard ratios, along with 95% confidence intervals, were computed for the logistic and Cox models, respectively, to determine the relative strength of the associations of the variables in the model to the outcomes. The records of 9 patients, all of whom survived hospitalization, were not used in the survival analyses due to the fact that their surgery happened after December 31, 2007, the cutoff SSDI date. Additionally, 9 more patient operations were withheld from the survival analysis because they were redo operations in the same patient.
The data were managed and analyzed using SAS Version 9.1 (SAS Institute, Cary, NC). Unadjusted comparisons between groups were performed with
2 tests (categoric variables), 2-sample t tests (numeric variables), and Mann-Whitney tests when significant outliers were present for numeric variables. All statistical tests were 2-sided using a p = 0.05 level of significance.
| Results |
|---|
|
|
|---|
|
|
|
|
Long-Term Survival
Kaplan-Meier survival estimates were calculated for the entire cohort and separately for biologic and mechanical patients (Fig 1). Median survival for the cohort was 1.2 years and did not differ by implant type (p = 0.87). The similarity in survival by valve implant type persisted for isolated AVR patient (p = 0.83, Fig 2) and isolated mitral replacements (p = 0.79, Fig 3).
|
|
|
| Comment |
|---|
|
|
|---|
The uremic milieu of ESRD patients and the associated derangements in calcium and phosphate metabolism have a well-recognized effect on calcification in the body and particularly of cardiac valves [20]. The progression of aortic stenosis in ESRD patients is accelerated compared with patients with normal renal function [21–23]. One small series [22] documented an annual decrease in aortic valve area among ESRD patients with aortic stenosis of 0.23 cm2/year compared with 0.05 cm2 to 0.1cm2/year among nonuremic patients. The reality of this fundamental physiologic difference in the ESRD patient, coupled with two early case reports from the 1970s [24, 25], generated an early concern for accelerated calcification of bioprosthetic valves. These concerns persisted for decades and were manifested as recently as 2008 in the American College of Cardiology/American Heart Association practice guidelines [26], which considered the use of bioprosthetic valves in ESRD patients potentially harmful.
However, beginning in the late 1990s, a series of reports detailing valve replacement in ESRD patients [27–29] including one from our own institution [7] questioned the concern for bioprosthetic implantation in the ESRD patient. Notably, the work of Herzog and colleagues [29] utilized the United States Renal Data System database to identify two decades (1978 to 1998) of dialysis patients who underwent heart valve replacement surgery. Their cohort of 5,858 patients demonstrated no significant survival difference related to the type of valve implanted with two-year survival rates of 39.7 ± 3.5% for biologic valves versus 39.7 ± 1.4% for mechanical. As interest in this question has persisted, additional single-center studies have demonstrated the same equivalence of survival among valve prosthesis choice [30, 31]. More recent practice guidelines from the National Kidney Foundation's National Disease Outcomes Quality Initiative do not discourage the implantation of bioprosthetic valves in ESRD patients.
Our data echo the results of the previously published series with regard to in-hospital and short-term mortality. Our overall in-hospital mortality of 19.9% closely approximates the results of previous publications [7, 29–31]. Additionally, it also reaffirms the relatively poor survival of these patients at 5 years post-replacement, 19% in our series compared with 14.8% in the series from Herzog and colleagues [29]. Most importantly, it confirms previous findings [7, 28–32] regarding the lack of effect that prostheses choice has on midterm survival. Furthermore, our data illustrate that even at long-term follow-up to 10 years the choice of a bioprosthetic versus a mechanical valve replacement does not significantly affect patient survival. These consistencies suggest that the long-term outcomes of these patients will be dictated more by their chronic ESRD rather than their underlying prosthetic choice.
This study is limited by its observational nature and the inherent limitations of a retrospective database study. Despite using logistic regression analysis to control for confounding variables, it is likely that all factors influencing selection bias were not accounted for in this analysis. Also, the study has a small sample size (although it is one of the largest reports to date) and the statistical tests used for comparison purposes are probably underpowered. A multiinstitutional study to evaluate this high-risk population is warranted. The long-term mortality dates are not accompanied with cause of death information, so an assumption is made that noncardiac-related deaths are evenly divided among comparison groups. Further, we were unable to incorporate surgeon identity into the multivariable model though our best evidence suggests it is not influential in this study. Decision-making regarding whom to operate on, and which types of prostheses to implant, were left to the discretion of the attending surgeon. Additionally, this analysis did not have access to the dialysis vintage (antecedent time on dialysis) for each patient. Dialysis vintage has an established inverse relationship with life expectancy and its effect on outcomes after cardiac surgery is a current research focus of our group. Furthermore, the heterogeneity of the study population and myriad comorbid conditions present may make it difficult to draw broad conclusions based on these data.
Ultimately, this large, retrospective review represents one of the largest single-center evaluations of prosthesis type and long-term survival after valve replacement in patients with ESRD. It confirms the conclusions of other recent studies regarding the clinically equivalent performance of mechanical and bioprosthetic valves in ESRD patients. Finally, it suggests that long-term survival in these patients will most likely be impacted by the continued improvement in dialysis techniques and aggressive risk factor modification rather than by improvements in cardiac surgery.
| Acknowledgments |
|---|
|
|
|---|
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
G. Bianchi, M. Solinas, S. Bevilacqua, and M. Glauber Are bioprostheses associated with better outcome than mechanical valves in patients with chronic kidney disease requiring dialysis who undergo valve surgery? Interact CardioVasc Thorac Surg, September 1, 2012; 15(3): 473 - 483. [Abstract] [Full Text] [PDF] |
||||
![]() |
V. B. Pai, C. K. Tai, K. Bhakri, and S. Kolvekar Should we use mechanical valves in patients with end-stage renal disease? Interact CardioVasc Thorac Surg, August 1, 2012; 15(2): 240 - 243. [Abstract] [Full Text] [PDF] |
||||
![]() |
V. H. Thourani, E. L. Sarin, P. D. Kilgo, O. M. Lattouf, J. D. Puskas, E. P. Chen, and R. A. Guyton Short- and long-term outcomes in patients undergoing valve surgery with end-stage renal failure receiving chronic hemodialysis J. Thorac. Cardiovasc. Surg., July 1, 2012; 144(1): 117 - 123. [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 |