|
|
||||||||
a Division of Cardiovascular Surgery, Mayo Clinic, Rochester, Minnesota
b Division of Cardiovascular Diseases, Mayo Clinic, Rochester, Minnesota
c Scott and White Clinic, Texas A & M Health Science Center, Temple, Texas
Accepted for publication April 30, 2009.
* Address correspondence to Dr Suri, Division of Cardiovascular Surgery, Mayo Clinic, 200 First St SW, Rochester, MN 55905 (Email: suri.rakesh{at}mayo.edu).
Presented at the Poster Session of the Forty-fifth Annual Meeting of The Society of Thoracic Surgeons, San Francisco, CA, Jan 26–28, 2009.
| Abstract |
|---|
|
|
|---|
Methods: One hundred fifty-two patients with aortic valve disease were randomly assigned to receive either the MM (n = 76) or an EP prosthesis. There were 89 men (59%), and the mean age was 76 years. Echocardiograms from preoperative, postoperative, predismissal, and 1-year time points were analyzed.
Results: Baseline characteristics and preoperative echocardiograms were similar between the two groups. The median implant size was 23 mm for both. There were no early deaths, and 10 patients (7%) died after dismissal. One hundred seven of 137 patients (78%) had a 1-year echocardiogram, and none required aortic valve reoperation. The mean aortic valve gradient at dismissal was 19.4 mm Hg (MM) versus13.5 mm Hg (EP; p < 0.0001), and at 1 year was 20.4 mm Hg versus 13.4 mm Hg (p < 0.0001). These differences were similar when the analysis was stratified by surgically measured annular size. The mean change in aortic valve gradient between predismissal and 1-year echocardiogram was +2.2 mm Hg (p = 0.02) for MM and –0.8 mm Hg (p = 0.33) for EP patients (p = 0.01 MM versus EP). The mean indexed aortic valve orifice area for MM and EP groups at dismissal and at 1 year was 0.9 cm2/m2 versus 1.1 cm2/m2, respectively (p < 0.01; p < 0.0001). During the first year after implantation, both groups demonstrated similar regression of LV mass index (MM, –32.4 g/m2 versus EP, –27.0 g/m2; p = 0.40). Greater preoperative LV mass index was the sole independent predictor of greater LV mass regression after surgery (p < 0.01).
Conclusions: Small differences in transprosthetic gradient and indexed aortic valve orifice area exist between porcine and bovine aortic valves. Despite this, both prostheses allow similar regression of LV mass during the first year after aortic valve replacement.
| Introduction |
|---|
|
|
|---|
Clinical outcomes after aortic valve replacement (AVR) generally have been attributed to successful minimization of transaortic pressure gradient, as determined by effective orifice area, transvalvular flow, body size, and cardiac output [1]. Evidence suggests that patients with clinically significant residual obstruction or patient-prosthetic mismatch after AVR have poorer functional recovery, lower exercise capacity, reduced regression of left ventricular (LV) hypertrophy, and an increased risk of both early and late death, particularly those with LV systolic dysfunction [2–6]. The variability in residual, flow-limiting gradients after insertion of different aortic valve prostheses may be clinically significant, and the surgical goal is to minimize intrinsic obstruction for a given annular size. The objective of the present study was to investigate the hemodynamic performance of two widely used, stented xenograft biologic prostheses. We also aimed to determine the effects of valve hemodynamics on regression of LV hypertrophy 1 year after surgery.
| Patients and Methods |
|---|
|
|
|---|
Patients
From January 2004 through September 2006, consecutive patients older than 50 years who were referred for isolated biologic AVR were considered for enrollment. Included were those undergoing concomitant procedures such as coronary artery bypass grafting or valve repair and those with a history of healed aortic valve endocarditis. Patients with prior nonaortic valve replacement, concomitant mitral, pulmonary, or tricuspid valve replacement, or active endocarditis were excluded. All patients provided informed consent and expressed willingness to follow requirements of the protocol.
Aortic Valve Replacement
Randomization numbers were generated by the statistician using the SAS program (SAS Institute Inc, Cary, NC), and printed on cards inside sealed envelopes. When a study patient entered the operating room, the envelope was opened and the surgeon was instructed regarding randomized valve type for implantation following annular measurement. Patients underwent standard cardiopulmonary bypass and cold-blood cardioplegic arrest. After aortotomy, excision of the native aortic valve cusps, and débridement of the aortic annulus, the native aortic annulus was measured by using a calibrated universal sizer (19, 21, 23, 25, 27, and 29 mm). Next, the annulus was measured twice further using aortic sizers supplied by the valve manufacturers, following their recommended techniques. Patients were prospectively randomized to receive an Edwards Perimount bovine pericardial valve (EP; Edwards Lifesciences, Irvine, CA) or Medtronic Mosaic porcine bioprosthesis (MM; Medtronic, Inc, Minneapolis, MN). Aortic valve prostheses were implanted using noneverting pledgeted stitches.
The MM valve is a third-generation porcine bioprosthesis implanted in the supraannular aortic position. The valve is fixed in glutaraldehyde with the use of zero pressure "Physiologic Fixation," mounted on a flexible polymer stent, and treated with alpha-amino oleic acid to mitigate calcification. The Perimount bioprosthesis is a second-generation trileaflet valve implanted in an intrasupraannular position. It is composed of bovine pericardium that is preserved in glutaraldehyde and mounted on an Elgiloy frame. The valve is treated with the XenologiX anticalcification process.
Echocardiography
Study patients underwent transthoracic Doppler echocardiography preoperatively, intraoperatively, before hospital dismissal, and at 1 year postoperatively [7–10]. Left ventricular dimensions were measured according to the recommendations of the American Society of Echocardiography. Left ventricular mass was calculated with the corrected American Society of Echocardiography formula [10] using two-dimensional, M-mode or two-dimensional linear LV surface echocardiographic measurements. These methods were validated in both animal models and human autopsy studies [11], and have subsequently been used in large population-based series examining changes in LV mass in hypertensive patients [12, 13]. The modified Bernoulli equation was used to calculate peak and mean pressure gradients across the prosthetic valve, and the effective orifice area was determined by the continuity equation. Multiplane transesophageal echocardiography was used intraoperatively or whenever additional information was not available with transthoracic echocardiography.
Statistical Analysis
The SAS statistical software package was used for data analysis. Continuous data are reported as mean (± standard deviation) or median as appropriate, and categorical data are reported as number and percent of total. Categoric variables were compared using the
2 test, and Student's t tests were used to compare continuous parameters. The Kaplan-Meier technique was used to estimate LV mass as a function of time. Cox proportional hazards regression analysis identified predictors of LV mass regression. Variables included in the multivariate analysis were those significant in the univariate analysis, and stepwise selection was used to create the final multivariate model. Probability values less than 0.05 were considered statistically significant.
| Results |
|---|
|
|
|---|
|
|
|
During the first year after implantation, both groups had similar regression in LV mass index (MM, –32.4 g/m2; EP, –27.0 g/m2; p = 0.40). The percentage change in LV mass was also similar between groups (MM, –19%; EP, –21%; p = 0.59). Whereas increased preoperative mean transaortic gradient, LV mass, and LV mass index were all univariate predictors of improved LV mass regression, larger preoperative LV mass index was the sole independent predictor of greater LV mass regression 1 year after surgery (p < 0.01) using multivariate modeling (Table 4).
|
| Comment |
|---|
|
|
|---|
Most studies have consistently demonstrated a small hemodynamic advantage of the EP compared with the MM in the aortic position during the early period after implantation. Seitelberger and associates [16] randomized 86 patients undergoing valve replacement and reported that mean gradients were lower at dismissal for the EP group (20.3 versus 17.4 mm Hg) receiving 21-mm valves. However, no significant hemodynamic differences were observed when the echocardiographic results at dismissal or at the 6-month follow-up were stratified on the basis of universal annular size. Eichinger and colleagues [17] from Munich randomized 139 patients to receive MM or EP and reported that the mean systolic gradient was significantly lower for the EP group (10.9 versus 14.2 mm Hg) but that AVA was not different. Walther and coworkers [18] randomized 100 patients to receive MM or EP and found that transprosthetic gradients were lower at dismissal for patients with the 25-mm EP and also lower after a mean follow-up of 438 days for those receiving either a 23- or 25-mm EP bioprosthesis.
Studies examining LV mass regression after AVR as a metric of favorable LV reverse remodeling have found results similar to ours. Seitelberger and colleagues [16] reported LV mass regression of approximately 50 to 70 g 6 months after valve replacement, and they also observed no differences between EP and MM groups. Eichinger and associates [19] described similar results from a series of patients undergoing stress echocardiography 10 months after either EP or MM implantation. Despite lower transprosthetic gradients with the EP valve, LV mass regression was indistinguishable between the prosthetic groups. Similar to the findings of our current study, their results also showed that greater preoperative LV hypertrophy was predictive of more complete LV mass regression at follow-up. A slightly discrepant result was reported by Walther and colleagues [18], who observed considerable regression of LV mass for all patients after implantation of either an EP or MM prosthesis but also discovered that LV mass index was slightly lower for patients who received the 23-mm EP compared with those having the 23-mm MM (156 versus 187 g/m2). Unusual was the fact that both the preoperative and follow-up LV mass values were high in that series. Finally, others have confirmed the hypothesis that LV mass regression may have little to do with small hemodynamic differences [20], even in those with patient-prosthetic mismatch [21]. Rather, preoperative LV mass appears to be a robust independent predictor of reverse remodeling after AVR [22].
If important differences in LV mass regression cannot be elucidated from randomized series of patients receiving EP and MM prostheses, what about "newer-generation" devices designed to have more favorable hemodynamics? Dalmau and coworkers [23] compared outcomes after implantation of the Magna bovine pericardial prosthesis (Edwards Lifesciences) and the MM in a randomized study of 86 patients. Although patients receiving the Magna had a significantly lower mean transaortic gradient (10.2 versus 17.1 mm Hg) and larger AVA (1.99 versus 1.69 cm2) after 1 year, no significant difference in LV mass regression was observed. They similarly concluded that small variations in prosthetic hemodynamics may not be important for LV mass regression in patients older than 70 years.
Stentless aortic valves are designed to minimize internal stent-related transprosthetic flow limitation and initially were heralded as a more "physiologic surrogate" for the native aortic valve. A recent study comparing two bovine pericardial valves, the stentless Sorin Freedom and the stented Sorin More, showed a lower postoperative gradient (17 versus 31 mm Hg) and a larger indexed AVA (1.1 versus 0.8 cm2/m2) in the stentless group [24]. Although LV reverse remodeling at 6 months was better for patients with the stentless prosthesis, LV mass regression continued in the EP group past 6 months, and results were equivalent after 1 year. A similar message was reported in a recent meta-analysis [25] of 10 studies that compared 919 patients who received either a stentless or a stented aortic valve prosthesis. Although the mean aortic valve gradient and LV mass index were lower in the stentless group at 6-month follow-up, no significant differences in LV mass regression were observed at 1 year. For current-era biologic aortic valve prostheses it appears, therefore, that the ability to link hemodynamic performance with the kinetics and degree of LV mass regression after AVR is tenuous at best. Small differences in valve hemodynamics likely have more to do with propagating marketing strategy than predicting reversal of LV hypertrophy [14, 26, 27].
What then is the best way to compare the hemodynamic efficiency of various aortic valve bioprostheses? Several authors have made the point that after surgical débridement of the aortic annulus, the degree of correlation between true annular dimension (measured with a universal sizer) and labeled valve size is unpredictable [23, 28, 29]. We were less concerned with this discrepancy and instead compared differences among patients after stratifying them by prosthesis types into three groups based on universal annular dimension. We hypothesized that each of the sizes within the groups would behave in a hemodynamically similar manner and might thus be a better method to compare differences between prosthetic groups. The results of the stratified analysis were, in fact, similar to the overall findings. Others have suggested that surgeons should compare the internal diameter of different valve prostheses at the time of implantation to determine which device might best optimize postoperative AVA [16]. Eichinger and associates [17] reported that the AVA to annulus area index (the quotient obtained after dividing the AVA by the aortic annulus area) could be used to compare prostheses and identify the device with the maximal orifice area available for LV ejection. However, although these authors found lower mean gradients for EP compared with MM (sizes 21 and 23 mm), they were unable to show a difference in the AVA to annulus area index. The problem with linking hemodynamic performance to "available orifice area" is that it does not account for heterogeneity in flow-limiting internal prosthetic design. On the basis of our current data, we suggest that implantation of an appropriately sized MM or EP aortic bioprosthesis is likely to offer satisfactory hemodynamic performance and comparable regression of LV mass, regardless of small hemodynamic differences.
Limitations
This prospective study was designed to emulate current surgical practice, and therefore decisions regarding implantable prosthetic size were made before randomization to device. Operations were performed in a group practice with multiple surgeons, all of whom used a similar surgical implantation technique. Although we cannot be certain that small differences in sizing tendencies did not exist, clinical monitoring during the roll-in phase of the study decreased the likelihood of this becoming a significant confounding factor. One-year echocardiographic data were available for 107 patients (78% of those eligible), and we have no reason to believe that patients were lost to follow-up in a nonrandom fashion. Residual postoperative gradients were relatively low in this study, and the differences between prosthetic groups were small. These data do not allow us to speculate on the extent to which higher levels of residual prosthetic obstruction might compromise patient survival or LV mass regression. Finally, it is logical to maintain that severe residual obstruction would be deleterious in those with poor LV function and should be avoided.
Conclusions
Prospectively randomized patients who underwent implantation of an aortic MM or EP bioprosthesis had similar regression in LV mass, despite small differences in prosthetic hemodynamics 1 year after surgery. These data suggest that the choice of biologic substitute has less influence on patient outcome than previously suspected.
| Acknowledgments |
|---|
|
|
|---|
Mayo Clinic does not endorse the products mentioned in this article.
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
S. Martens, J. Sadowski, F. S. Eckstein, K. Bartus, B. Kapelak, H.-H. Sievers, C. Schlensak, and T. Carrel Clinical experience with the ATS 3f Enable(R) Sutureless Bioprosthesis Eur J Cardiothorac Surg, September 1, 2011; 40(3): 749 - 755. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. J. Dalmau, J. M. Gonzalez-Santos, J. A. Blazquez, J. A. Sastre, J. Lopez-Rodriguez, M. Bueno, M. Castano, and A. Arribas Hemodynamic performance of the Medtronic Mosaic and Perimount Magna aortic bioprostheses: five-year results of a prospectively randomized study Eur J Cardiothorac Surg, June 1, 2011; 39(6): 844 - 852. [Abstract] [Full Text] [PDF] |
||||
![]() |
K. K. Baskin and H. Taegtmeyer Taking pressure off the heart: the ins and outs of atrophic remodelling Cardiovasc Res, May 1, 2011; 90(2): 243 - 250. [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 |