|
|
||||||||
Ann Thorac Surg 2004;78:90-95
© 2004 The Society of Thoracic Surgeons
a Department of Cardiology, Cardiovascular Research Institute Maastricht (CARIM), University Hospital Maastricht, Maastricht, The Netherlands
b Department of Cardiothoracic Surgery, University Hospital Maastricht, Maastricht, The Netherlands
Accepted for publication February 6, 2004.
* Address reprint requests to Dr Maessen, Department of Cardiothoracic Surgery, 6202 AZ Maastricht, The Netherlands
e-mail: j.maessen{at}scpc.azm.nl
| Abstract |
|---|
|
|
|---|
METHODS: We performed MEDLINE and bibliographic searches and included 27 articles published between 1980 and 2003 about the outcome of valve replacement in 1546 aortic stenosis patients. To allow comparisons, we stratified the patients into early (06 months), intermediate (724 months), and late (25120 months) follow-up groups for the analysis of both LV mass regression and ejection fraction. We separately analyzed five articles that reported groups of patients with low preoperative ejection fraction.
RESULTS: Increase in ejection fraction after surgery is more pronounced in the patients that have low preoperative ejection fraction (28% ± 4.3%preop vs 40% ± 9.4%641 months follow-up). Patients with normal or high preoperative ejection fraction have variable outcomes. However, regression of LV mass is uniformly achieved regardless of age, sex, time of operation, or types of valve substitute. Furthermore, LV mass regresses predominantly within the first 6 months after surgery (g/m2, 181 ± 25.8preop vs 124 ± 276 months, 117 ± 1524 months, and 113 ± 14120 months follow-up).
CONCLUSIONS: This systematic review supports the concept that aortic stenosis patients with LV dysfunction show a clear functional improvement after aortic valve replacement. Ventricles regress rapidly and reach their approximate final size within the first 6 months of surgery.
| Introduction |
|---|
|
|
|---|
There are contrasting views about the timing and indications for surgical intervention for aortic stenosis. The results of AVR are also uncertain among aortic stenosis patients with reduced LV ejection fraction (LVEF) and low transvalvular mean gradient. Although these patients represent less than 5% of patients with aortic stenosis, it is in this group of patients that the effects of aortic valve surgery are least established. Increased perioperative risk and adverse late outcome have been reported in patients with reduced preoperative LVEF [4]. Moreover, published data are also inconsistent in regard to long-term benefits as far as the regression of LV hypertrophy and dilatation are concerned [5]. Medical literature pertaining to the effects and efficacy of aortic valve surgery is largely composed of numerous heterogeneous studies of relatively small population. The studies vary in terms of patient selection, outcomes evaluated, operative intervention, and timing for postoperative follow-up thus limiting the applicability of individual reports. Therefore, to discern the most consistently reported outcome variables, we systematically analyzed the current body of evidence and aimed to evaluate the early, intermediate, and late consequences of aortic valve replacement on LV size regression and LVEF.
To this end, we reviewed the existing literature pertaining to aortic valve stenosis patients from 19802003 using a comprehensive search strategy and predefined inclusion as well as exclusion criteria. We homogenized the available clinical data based on patient characteristics and preoperative and postoperative cardiac hemodynamic parameters. Depending on the size of the population, we allocated a defined weight to the individual studies before analyzing the categorical variables systematically.
| Material and methods |
|---|
|
|
|---|
Study selection criteria
Studies were considered eligible for this analysis if they evaluated the following interventions: isolated aortic valve stenosis in adults, documented preoperative and postoperative EF and LV mass index (LVMI) and at least 45 days follow-up postoperatively. Additionally, the included studies had to report sufficient data to calculate the differences in the EF and LVMI preoperatively versus postoperatively. Data were excluded if they did not report at least one of the clinical outcomes of interest.
Levels of evidence and grades of recommendation
We allocated the evidence level to individual studies according the guidelines formulated by Oxford Center for Evidence-based Medicine Levels of Evidence [6]. Unlike randomized controlled trials, aortic valve replacement is an outcome-based treatment without randomization and matched control cohorts. Because the studies are primarily based on existing normal parameters for the comparison of outcome, the selected studies customarily scored B grades of recommendation. Studies with inadequate follow-up or expert opinions without explicit critical appraisal were removed from the analysis.
Abstraction methods
Two authors independently abstracted study design and participant data onto pretested abstraction forms from each of these publications and reviewed bibliographies for additional potentially relevant studies. Abstraction discrepancies were resolved by repeated review and discussion. If two or more studies reported the same data from a single participant population, these data were included only once in the analysis. If a study presented data on two of the abovementioned clinical outcomes and if one did not meet our inclusion criteria (eg, studies that measured EF only in postoperative patients but measured LVMI both preoperatively and postoperatively), then data were abstracted only relevant to the outcome that met the inclusion criteria.
Statistical analyses
For each study, a weighted mean was calculated from individual patient and outcome variables. If the study did not report individual data, then a weighted mean was calculated and this weighted mean was used to calculate the pooled mean. Similarly, a pooled variance was calculated from the individual variances. If a study did not report any measure of variance from an outcome, the overall mean pooled variance was used for that study according the study protocol published previously [7]. This analysis required setting a threshold to classify reduced versus normal EF. Because the literature has no clear consensus as to what constitutes a low or normal EF, 45% was chosen as the arbitrary threshold limit.
To check the eligibility of the data pooling, homogeneity of the variances was tested using Levene's statistic. Independent sample t test was used to compare the changes in EF or LVMI before and after surgery and test the efficacy of stentless valves over stented ones. In the patient groups with multiple follow-up, repeated-measures analysis of variance was used to compare the effect of AVR on LV mass regression. Categorical variables were compared by the
2 test. Similarly, multiple linear regression analysis tested the effect of age on the outcome variables. Analysis was performed using SPSS version 10.0 (SPSS Inc., Chicago, IL) and Microsoft Excel XP (Microsoft Corp., Redmond, WA).
| Results |
|---|
|
|
|---|
|
Change in EF after AVR
Because the time of follow-up varied between and within studies, we allocated three major subgroups to the published reports. The first (early follow-up) subgroup (N = 346) included the studies that reported the follow-up events within the first 6 months [1320]. The second (N = 471) and third (N = 454) subgroup included the reports that presented the followed data between 724 months (intermediate follow-up) [11, 13, 1828] and between 25120 months (late follow-up) [5, 9, 19, 22, 23, 28, 29], respectively (Table 1). Most of the studies with normal preoperative EF have shown an increase in EF after AVR (median increase, 6.8%). There is discernible increase in EF within the first 6 months of surgery and sustained improvement until 10 years (EF %, 56 ± 3.5preop vs 63 ± 3.106 months; 57.1 ± 4 preop vs 62.5 ± 5.1724 months and 57.5 ± 5.5preop vs 63 ± 4.425120 months). However, some of the studies with comparable preoperative EF have shown very little or no increase in EF after operation [5, 14, 20] (Fig 1).
As a summary effect, statistical analysis on weighted and pooled data showed no considerable change in EF at all three time periods chosen for comparison.
|
Regression and rate of regression of LV mass
A similar time stratification strategy was made to divide the study groups into early (N = 695) [10, 12, 13, 1820], intermediate (N = 506) [9, 11, 13, 1828], and late (N = 450) [9, 19, 22, 23, 28, 29, 33] follow-up groups (Table 1). In contrast to the unpredictable changes in EF, all the studies are in clear agreement for the regression of the LV mass after surgery (Fig 2).
The range of regression remained between 12% [18] to 43% [23] (median values, 20%). Comparisons between preoperative and postoperative follow-up yielded a consistent regression of LV mass (LVMI g/m2, 181 ± 34preop vs 136 ± 2106 months, 183 ± 19preop vs 128 ± 11724 months and 183 ± 19preop vs 120 ± 1425100 months; p < 0.05 at all instances). This indicated an invariable decrease in LV mass after surgery but to discern the rate of LV mass regression, we made a further analysis in seven studies [13, 18±20, 22, 23, 28] that reported the LVMI in more than one allocated time period after AVR. This showed a sharp fall in LVMI within first 6 months of surgery (LVMI g/m2, 181 ± 25.8preop vs 124 ± 276 months, 117 ± 1524 months and 113 ± 14120 months follow-up). Interestingly, the decrease in LVMI did not substantially change after 6 months (ppreop vs 06 months < 0.05; p06 months vs 724 months, NS; p724 months vs 25120 months, NS).
|
| Comment |
|---|
|
|
|---|
The development of concentric hypertrophy in patients with aortic stenosis is an adaptation of the left ventricle to the increase in intracavitary pressure [34], which allows it to maintain a normal relation between systolic wall stress and EF. Aortic valve replacement, therefore, decreases the ventricular afterload consequently leading to myocardial adaptation and regression of hypertrophy [35]. Most of the reports included in this review are in clear agreement on the improvement of EF after surgery. However, some studies have shown the opposite effects and thus denied the concept of AVR before the onset of symptoms [14]. Because aortic stenosis has the highest prevalence in the elderly population and is often associated with secondary cardiac diseases, it is conceivable that the rate of change of EF is influenced by multiple factors. The study by Connolly and associates showed 21% perioperative mortality in the patients with low preoperative EF and low systolic gradient. In that respect it is noteworthy that the studies that followed up low preoperative EF group of patients have included only the survivors for their analysis. Interestingly, the survivors demonstrated a substantial improvement in EF after surgery [32]. Studies have also argued the role of valve-prosthesis-patient mismatch for the outcome of AVR [36]. This review could not systematically analyze such an effect as most of the individual studies reported LV mass index without additional specifications on body surface areas of the individual patients. Furthermore, only a few studies compared the extent of LV mass regression depending on the prosthesis size.
Although there is an obvious LV mass regression after AVR, the role of age and sex for the outcome of the surgery cannot be ignored. Age is an important and independent factor for LVH and this could explain the observation of incomplete LV mass regression in most of the study groups. Although there were some discrepancies in the patient age group in our analysis, the differences noticed on the extent of LV mass regression compared with age were statistically insignificant and, therefore, did not justify the age-dependent difference in EF and LV mass regression. Sex differences in LV adaptation to aortic stenosis have been described recently [4, 37, 38]. However, the several descriptors of LV geometry and function can be largely eliminated after normalizing for body surface area and there are no sex differences in surgical mortality or long-term outcome of the patients [37]. With the emergence of more descriptive and bigger surgical trials, future reviews should also address additional parameters of clinical outcome (eg, transaortic jet velocity, valve calcification index, mean systolic gradient, change in myocardial stiffness, and exercise tolerance testing).
A few comments concerning this review have to be made because they have implications in the interpretation of the analyzed results. These are, in fact, limitations that include modest magnitude of the inclusion parameters (only EF and LVMI as these were the only consistently reported parameters in most of the studies), the variability in the medical management of the patients, duration of the disease with its unclear starting point, and heterogeneity of the published reports. Even though the published articles we analyzed addressed the same question regarding the change in EF and LV mass after aortic valve replacement, there were several intraarticle and interarticle discrepancies about the sample size, study groups, reporting strategies, patient follow-up, and techniques pertaining to the acquiring of the hemodynamics.
This review did not completely fulfill the requirements of a meta-analysis as the studies were performed without the inclusion of case-matched control groups. All the patients included in the study underwent AVR, thus the clinical dilemma differentiating LV dysfunction due to aortic stenosis or other subtle forms of cardiomyopathies could not be differentiated. As the review embraces the studies published within last 23 years, the clinical strategies regarding patient management, selection of appropriate aortic valves, and other technological advances rendered the study groups less comparable. Because of these confounding factors, it is important to suggest that individual studies still need to be evaluated on their own stance, although a systematic review provides a gross picture of the existing practice and its impact on patient mortality and morbidity.
Although aortic stenosis patients with preoperative low EF and secondary cardiac diseases constitute a small subset and seem to have relatively higher surgical mortality, these patients should not be denied aortic valve replacement only on the grounds of low EF. The LV mass regression is an independent (of age, sex and types of valve substitutes) and most consistent effect of valve replacement. The clinical follow-up of these patients should specifically focus on the first 6 months postoperatively as the ventricles revert to their final size within this short but crucial period of time.
| Acknowledgments |
|---|
|
|
|---|
| References |
|---|
|
|
|---|
62 years with unoperated severe valvular aortic stenosis. Am J Cardiol 1993;72:846-848.[Medline]
This article has been cited by other articles:
![]() |
A. Germing and A. Mugge What the cardiac surgeon needs to know prior to aortic valve surgery: impact of echocardiography Eur. J. Cardiothorac. Surg., June 1, 2009; 35(6): 960 - 964. [Abstract] [Full Text] [PDF] |
||||
![]() |
V. S. Lad Valve Size Does Matter in the Young: Letter 1 Ann. Thorac. Surg., January 1, 2009; 87(1): 353 - 354. [Full Text] [PDF] |
||||
![]() |
M. J. Silka and Y. Bar-Cohen Should patients with congenital heart disease and a systemic ventricular ejection fraction less than 30% undergo prophylactic implantation of an ICD?: Patients With Congenital Heart Disease and a Systemic Ventricular Ejection Fraction Less Than 30% Should Undergo Prophylactic Implantation of an Implantable Cardioverter Defibrillator Circ Arrhythmia Electrophysiol, October 1, 2008; 1(4): 298 - 306. [Full Text] [PDF] |
||||
![]() |
Y.-Y. Lam, M. G Kaya, W. Li, V. S Mahadevan, A. A Khan, M. Y Henein, and M. Mullen Effect of endovascular stenting of aortic coarctation on biventricular function in adults Heart, November 1, 2007; 93(11): 1441 - 1447. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. A. Borger, S. M. Carson, J. Ivanov, V. Rao, H. E. Scully, C. M. Feindel, and T. E. David Stentless Aortic Valves are Hemodynamically Superior to Stented Valves During Mid-Term Follow-Up: A Large Retrospective Study Ann. Thorac. Surg., December 1, 2005; 80(6): 2180 - 2185. [Abstract] [Full Text] [PDF] |
||||
![]() |
K. Imanaka, O. Kohmoto, S. Nishimura, Y. Yokote, and S. Kyo Impact of postoperative blood pressure control on regression of left ventricular mass following valve replacement for aortic stenosis Eur. J. Cardiothorac. Surg., June 1, 2005; 27(6): 994 - 999. [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 |