Ann Thorac Surg 2004;77:1266-1271
© 2004 The Society of Thoracic Surgeons
Original article: cardiovascular
Role of transvalvular gradient in outcome from valve replacement for aortic stenosis
Arjuna Weerasinghe, FRCS, PhDa*,
Merangani Yusuf, FRCSa,
Thanos Athanasiou, PhDb,
Alan Wood, FRCSa,
Patrick Magee, FRCSa,
Rakesh Uppal, FRCSa
a Department of Cardiothoracic Surgery, St. Bartholomew Hospital and The London Chest Hospital, London, United Kingdom
b Department of Cardiothoracic Surgery, St. Mary's Hospital, London, United Kingdom
Accepted for publication October 2, 2003.
* Address reprint requests to Dr Weerasinghe, Department of Cardiothoracic Surgery, Hammersmith Hospital, Du Cane Rd, London W12 0NN, UK, United Kingdom
e-mail: a.weerasinghe{at}ic.ac.uk
 |
Abstract
|
|---|
BACKGROUND: Despite limitations the transvalvular gradient (TVG) still is commonly used in aortic stenosis when patients are referred for aortic valve replacement. We wished to ascertain if it had a role in predicting outcome from valve replacement rather than as an indicator of severity, specifically investigating if the TVG affected renal dysfunction, hospital stay, and medium-term survival after valve replacement.
METHODS: Six hundred and twenty-three consecutive patients who had aortic valve replacement were identified and 211 of these patients were isolated as first time replacement for aortic stenosis that formed the final study group, and were followed up for up to 48 months. Variables significant (p < 0.05) on univariate analysis were included in the logistic regression multivariate analysis (renal dysfunction, prolonged hospital stay) or Cox proportional hazard regression model (medium-term mortality).
RESULTS: A significant association was present between TVG and age (p = 0.001). Multivariate analysis demonstrated angina greater than or equal to grade 3 Canadian Cardiovascular Society (CCS 3; p = 0.014) and having nonelective surgery (p < 0.001) to be independent predictors of renal dysfunction. Angina greater than or equal to CCS 3 (p = 0.013) was the only independent predictor of prolonged hospital stay. Independent predictors of medium-term mortality on multivariate analysis were age (p = 0.043) and having a size 19 valve prosthesis (p = 0.015).
CONCLUSIONS: The TVG is inadequate as an independent predictor of the degree of aortic stenosis and outcome from aortic valve replacement in aortic stenosis. In contrast, the TVG may be a useful screening tool for detecting aortic stenosis in targeted elderly populations.
|
Introduction
|
|---|
Unlike coronary surgery, there is a scarcity of large-scale multicenter trials addressing the diagnosis and treatment of aortic valve disease, and the literature represents mostly the experiences reported by single institutions of relatively small numbers of patients; therefore management issues still remain controversial or uncertain.
Valve replacement for aortic stenosis in adults is influenced by the patients symptoms and measures of hemodynamic severity. The most widely used variable of hemodynamic severity is probably the transvalvular gradient (TVG). The inadequacy of the TVG as an indicator of the hemodynamic severity of aortic stenosis has been raised recently [1], and is also acknowledged in the report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines [2]. The pressure gradient across a stenotic valve is related to the valve orifice area and transvalvular flow. Thus, in the presence of a depressed cardiac output, relatively lower pressure gradients are frequently obtained in patients with more severe aortic stenosis. On the other hand, during exercise or other high-flow states, systolic gradients can be measured in valves that are only minimally stenotic. Patients with severe stenosis and low cardiac output often present with only modest transvalvular pressure gradients (ie, < 30 mm Hg). Such patients can be difficult to distinguish from those with low cardiac output and only mild to moderate stenosis.
In patients with low gradient stenosis and what appears to be moderate to severe stenosis, it may be useful to determine the transvalvular pressure gradient and calculate valve area and resistance during a baseline state and again during exercise or pharmacologic (dobutamine infusion) stress. Patients who do not have true, anatomically severe stenosis exhibit an increase in the valve area during an increase in cardiac output. In patients with severe stenosis, these changes may result in a calculated valve area that is higher than the baseline calculation but that remains in the severe range, whereas in patients without severe stenosis, the calculated valve area will fall outside the severe range with administration of dobutamine and indicate that severe aortic stenosis is not present.
Despite its limitations the TVG still is commonly used at the time patients are referred for aortic valve replacement, as a carry through of its historic usage and also due to its simplicity. We wished to ascertain if it had a role in predicting outcome from valve replacement rather than as an indicator of severity, thereby justifying its continued use. Specifically we set out to ascertain if the TVG affected renal dysfunction, caused a prolonged hospital stay, and affected medium-term survival.
|
Material and methods
|
|---|
Patient demographic data and perioperative data were gathered from the prospective cardiothoracic data base maintained at St. Bartholomews Hospital and The London Chest Hospital, both within the same health services trust in London, United Kingdom. Information on the valve gradients was added in at the time of analysis utilizing patient records. Survival data were obtained from the Department of National Statistics, United Kingdom.
Six hundred and twenty-three patients who had an aortic valve replacement between April 1, 1999 and October 31, 2001 were identified. From this group, patients who had concomitant coronary surgery, other valve surgery, redo-surgery, and patients having surgery for aortic regurgitation were excluded. A cohort of 211 patients having isolated first time aortic valve replacement for aortic stenosis formed the final study group. The patients were followed up for a period of 18 to 48 months.
Age and serum creatinine were used as continuous variables. An increase in serum creatinine of greater than or equal to 20% was defined as significant renal dysfunction based on a previous study by Weerasinghe and coworkers [3]. Left ventricular (LV) ejection fraction was scored as less than 30% = 1, 30% to 50% = 2, and more than 50% = 3. The TVG was used as a continuous variable when analyzed independently. The interaction between LV and TVG was analyzed by scoring TVG as less than 50 mm Hg = 1, 50 to 100 mm Hg = 2, more than 100 mm Hg = 3, and using the LV score/TVG score ratio in the statistical analysis.
Statistical methods
Data were expressed as mean ± standard deviation (SD). Numerical variables were compared using Student's t test or Mann-Whitney test where appropriate. For comparison of continuous variables t test was used. Categorical variables were compared using Fischer's exact test or 
2 test, where appropriate, and p values of equal to or less than 0.05 indicated a significant difference. All tests were two sided. The survival of a patient started at the time of aortic valve operation and ended at the event of death or at the time of the last follow-up (censoring). Kaplan-Meier statistical methods, used to calculate the probability of survival and differences between curves, were compared using the log rank test. Patients at risk were added at 12, 24, 36, and 42 months follow-up. The preoperative variables were compared by univariate analysis with regards to the effect on 30-day death and all cause medium-term mortality. Those variables proving to be significant (p < 0.05) were included in the logistic regression multivariate analysis (renal dysfunction and prolonged hospital stay) or Cox proportional hazard regression model (medium-term mortality). All statistical analysis was performed using the SPSS 11.0 for Windows software package (SPSS Inc., Chicago, IL).
|
Results
|
|---|
The mean age of the study population was 64.44 ± 14.36 years; 27 patients (12.8%) were 80 years old or older, whereas 184 patients (87.2%) were less than 80 years of age. One hundred and twenty-three patients (58.3%) were male and 88 patients (41.7%) were female. The mean TVG was 87.47 ± 27.85 mm Hg; 75 (35.5%) of 211 patients had impaired LV function (ejection fraction < 50%) and 125 patients (59.24%) had dyspnea of more than grade II New York Heart Association class. The mean preoperative creatinine was 119.97 ± 82.04 and the mean postoperative creatinine was 149.03 ± 109.28 µmol/L−1. Eighty of the operations (37.91%) were nonelective procedures. A biological prosthesis was used in 104 patients (49.29%) and a mechanical prosthesis was used in 107 patients (50.71%) in the group. The median length of stay following surgery was 8 days (interquartile range 6 to 12 days). Table 1
illustrates the complete demographic characteristics of the final study group.
A significant linear association was noted between age and the TVG (p = 0.001). The regression equation showed
This association between age and TVG is depicted in Figure 1
in the form of a scatter plot.
View larger version (16K):
[in this window]
[in a new window]
|
Fig 1. Scatterplot illustrating the association between age and TVG. The regression equation demonstrated, TVG = 45.3 + 0.62 x Age (in years). (TVG = transvalvular gradient.)
|
|
Variables associated with a 20% or greater deterioration in renal dysfunction, a length of stay greater than 10 days, and medium-term mortality were investigated. Factors contributing on univariate analysis are depicted in Table 2.
Subsequently, multivariate analysis was performed and the predictors on multivariate analysis are indicated in Table 3.
The factors associated with renal dysfunction on univariate analysis were angina equal to or greater than grade 3 Canadian Cardiovascular Society (CCS 3), infective endocarditis, and having nonelective surgery. Multivariate analysis revealed angina greater than or equal to CCS 3 (p = 0.014, odds ratio [OR] = 2.43) and having nonelective surgery (p < 0.001, OR = 3.32) to be independent predictors of renal dysfunction.
Univariate predictors of a prolonged hospital stay were angina greater than or equal to CCS 3, infective endocarditis, and having nonelective surgery. Multivariate analysis demonstrated angina equal to or greater than CCS 3 (p = 0.013 and OR = 2.35) to be the only independent predictor of prolonged hospital stay.
The predictors of medium-term mortality on univariate analysis were, increasing age, diabetes, having nonelective surgery, and a valve size = 19. An ejection fraction more than 50% was seen to provide a protective role with improved survival on univariate analysis. Independent predictors of medium-term mortality on multivariate analysis were increasing age (p = 0.043, OR = 1.04 compared with a patient 1-year younger) and a valve size = 19 (p = 0.015, OR = 4.71).
Next we studied the Kaplan-Meier curves for medium-term survival, comparing patients with gradients more than 100 mm Hg with those with gradients equal to or less than 100 mm Hg (Fig 2).
The nondigression of the curves illustrates that the TVG clearly does not influence medium-term survival after aortic valve replacement for aortic stenosis. Survival time in the first group was 42.04 ± 1.5 months and in the second group it was 42.21 ± 1.23 months, giving a log rank value of 0.01 and p = 0.92 indicating no significant difference between the two groups.
View larger version (21K):
[in this window]
[in a new window]
|
Fig 2. Kaplan-Meier curves comparing medium-term outcome in patients with TVG more than 100 mm Hg and those with TVG less than or equal to 100 mm Hg. Survival time was 42.04 ± 1.5 months and 42.21 ± 1.23 months, respectively (log rank value of 0.01 and p = 0.92). (TVG = transvalvular gradient.)
|
|
View larger version (21K):
[in this window]
[in a new window]
|
Fig 3. Kaplan-Meier curves comparing medium-term outcome in patients receiving a valve size > 19 with those receiving a valve size = 19. Survival time was 42.6 ± 0.89 months and 36 ± 3.6 months, respectively (log rank value of 5.61 and p = 0.017).
|
|
Finally we studied the Kaplan-Meier curves for medium-term survival, comparing patients receiving a size more than 19 valve prosthesis with those receiving a size = 19 prosthesis. The two curves tracked closely for the first 24 months after valve replacement and then digressed markedly thereafter (Fig 3). Survival time when receiving a size more than 19 prosthesis was 42.6 ± 0.89 months and 36 ± 3.6 months if a size 19 prosthesis was used, giving a log rank value of 5.61 and p = 0.017.
|
Comment
|
|---|
Despite major advances having occurred in diagnostic techniques and surgical procedures for patients with valvular heart disease, the TVG is still widely used as a guide to aortic valve replacement in aortic stenosis. The limitations of the TVG as an indicator of severity of disease is due to its dependence on flow rate and has been previously reported [4–6]. In this study we investigated if the TVG had a continued role in aortic valve replacement for aortic stenosis as a predictor of outcome, rather than as an indicator for surgery.
Patients having nonelective surgery were seen to be more likely to suffer from renal dysfunction associated with surgery. In a comparable manner patients having nonelective coronary surgery are known to suffer greater renal dysfunction [3]. In a setting of emergency surgery, patients having aortic valve replacement are more likely to have had a period of preoperative low cardiac output and renal hypoperfusion due to either cardiogenic shock or septicemia. In addition, a preoperative angiogram just before surgery to assess concomitant coronary disease may have been contributory towards deterioration in renal function. The TVG indicates no direct influence in causing a significant alteration in renal function occurring due to aortic valve replacement. This is likely to be because it does not affect renal perfusion in the operative and postoperative period or because it has a minimal effect not obvious due to study size.
The TVG was not associated with a prolonged postoperative stay. The single factor that did affect postoperative stay was having more severe grades of angina. This may partly be explained by the possibility that these patients had a higher incidence of dysrrythmias, or required cardiovascular support resulting in a prolonged total postoperative stay. Furthermore, as observed above, more severe grades of angina were also seen to be associated with a greater deterioration of renal function, which may be explained by an increased likelihood of concomitant renal vasculopathy in these patients. Thus the higher likelihood of postoperative renal dysfunction may also contribute to the effect of angina on postoperative stay. The TVG is not directly related to any other factors such as rhythm disturbances, neurologic events or infections that would have caused a prolongation of the patients postoperative stay in hospital.
Of the study cohort 46 patients (21.8%) had angina greater than or equal to grade 3 CCS severity. Why then did these patients not have concomitant coronary artery bypass grafting? Previously it has been demonstrated that up to 57% of patients with aortic stenosis having angina may have angiographically normal coronary arteries [7]. These patients have been reported to have inadequate LV hypertrophy with lower LV muscle mass, increased LV peak systolic pressure, increased systolic and diastolic wall stress, and lower coronary flow reserve than those without angina [8]. Furthermore it has also been demonstrated (in the Euro-Heart Survey of acute coronary syndromes) that patients with valvular disease were less likely to have typical angina [9], thus making clinical assessment of myocardial ischemia in these patients even more complicated. In a recent editorial Gould and Carabello [10] suggest that the occurrence of angina in patients with aortic stenosis and angiographically normal coronary arteries heralds a critical decrease in diastolic myocardial perfusion. They further suggest the use of generating ischemia with progressive pacing tachycardia to achieve this critically short diastolic perfusion as a means of assessing severity. In this context it is conceivable that a lower threshold may need to be applied for surgical revascularisation in patients with aortic stenosis having angina of CCS equal to or greater than 3.
Because older patients are likely to have a decreased life expectancy irrespective of surgery, the influence of age on medium-term survival may be partly explained as a reflection of the natural life expectancy. A recent publication reports that the immediate operative mortality too is higher in patients more than 80 years of age [11]. The observation that a small aortic prosthesis may lead to decreased survival has been previously observed and is possibly related to patient-prosthesis mismatch [12, 13], resulting in impaired regression of LV hypertrophy. Regression of LV hypertrophy after aortic valve replacement has been revealed to be better in patients with a prosthesis size more than 21 mm (−21%) than in patients with a prosthesis size more than or equal to 21 mm (−8%) [14]. Because aortic valve replacement with a stentless bioprosthesis may be associated with a more complete regression of LV hypertrophy [15], stentless valves may need to be considered when a small prosthesis is used. The superior hemodynamic performance of stentless valves is due to their effective orifice area being generally larger than that of stented valves and also because a relatively larger prosthesis can be inserted than with a stented valve, into a small annulus. A further option to consider is the insertion of a larger prosthesis, achieved by enlargement of the aortic root or performing a supraannular implantation. Other surgical options that may have to be considered include the use of an aortic homograft or pulmonary autograft (Ross procedure).
Before this study it was known that the TVG is a poor indicator of severity of aortic stenosis in assessing the need for surgery. The results of this study clearly demonstrate that it is also inadequate even as a predictor of outcome from aortic valve replacement in aortic stenosis.
An incidental observation of the study was the linear association between TVG and age. Clustering is seen after 70 years of age, and measuring the TVG using a transthoracic echocardiogram may be a feasible mode of screening patients, despite its limitations as an isolated indicator of severity and outcome from valve replacement. This would allow early patient detection before the stenosis reaches severe proportions or decompensation of LV function occurs.
With advances in diagnostic methods, new methods more reflective of both the pathology and outcome need to be more widely utilized to replace the TVG. A recent publication [1] has reported that the LV stroke work loss expressed as a percentage is a better indicator of severity of aortic stenosis and as an indicator of the occurrence of a composite outcome of cardiac death or aortic valve replacement. The stroke work loss has yet to be studied as a potential indicator of outcome from aortic valve replacement for aortic stenosis.
We hope the results of this study will further catalyze the quest for a unifying indicator of severity that is also a reliable predictor of outcome.
|
References
|
|---|
- Bermejo J., Odreman R., Feijoo J., Moreno M.M., Gomez-Moreno P., Garcia-Fernandez M.A. Clinical efficacy of Doppler-echocardiographic indices of aortic valve stenosis: a comparative test-based analysis of outcome. J Am Coll Cardiol 2003;41:142-151.[Abstract/Free Full Text]
- Bonow R.O., Carabello B., de Leon A.C., Jr, Edmunds L.H., Jr, et al. Guidelines for the management of patients with valvular heart disease: executive summary. A report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (Committee on Management of Patients with Valvular Heart Disease). Circulation 1998;98:1949-1984.[Free Full Text]
- Weerasinghe A., Hornick P., Smith P., Taylor K., Ratnatunga C. Coronary artery bypass grafting in non-dialysis-dependent mild-to-moderate renal dysfunction. J Thorac Cardiovasc Surg 2000;121:1083-1089.
- Burwash I.G., Pearlman A.S., Kraft C.D., Miyake-Hull C., Healy N.L., Otto C.M. Flow dependence of measures of aortic stenosis severity during exercise. J Am Coll Cardiol 1994;24:1342-1350.[Abstract]
- Bermejo J., Garcia-Fernandez M.A., Torrecilla E.G., et al. Effects of dobutamine on Doppler echocardiographic indexes of aortic stenosis. J Am Coll Cardiol 1996;28:1206-1213.[Abstract]
- Burwash I.G., Thomas D.D., Sadahiro M., et al. Dependence of Gorlin formula and continuity equation valve areas on transvalvular volume flow rate in valvular aortic stenosis. Circulation 1994;89:827-835.[Abstract/Free Full Text]
- Rapp A.H., Hillis L.D., Lange R.A., Cigarroa J.E. Prevalence of coronary artery disease in patients with aortic stenosis with and without angina pectoris. Am J Cardiol 2001;87:1216-1217.[Medline]
- Julius B.K., Spillmann M., Vassalli G., Villari B., Eberli F.R., Hess O.M. Angina pectoris in patients with aortic stenosis and normal coronary arteries: mechanisms and pathophysiological concepts. Circulation 1997;95:892-898.[Abstract/Free Full Text]
- Hasdai D., Lev E.I., Behar S., et al. Acute coronary syndromes in patients with pre-existing moderate to severe valvular disease of the heart: lessons from the Euro-Heart Survey of acute coronary syndromes. Eur Heart J 2003;24:623-629.[Abstract/Free Full Text]
- Gould K.L., Carabello B.A. Why angina in aortic stenosis with normal coronary arteriograms?. Circulation 2003;107:3121-3123.[Free Full Text]
- Florath I., Rosendahl U.P., Mortasawi A., et al. Current determinants of operative mortality in 1400 patients requiring aortic valve replacement. Ann Thorac Surg 2003;76:75-83.[Abstract/Free Full Text]
- Blais C., Dumesnil J.G., Baillot R., Simard S., Doyle D., Pibarot P. Impact of valve prosthesis-patient mismatch on short-term mortality after aortic valve replacement. Circulation 2003;108:983-988.[Abstract/Free Full Text]
- Pibarot P., Dumesnil J.G. Haemodynamic and clinical impact of prosthesis–patient mismatch in the aortic valve position and its prevention. J Am Coll Cardiol 2000;36:1131-1141.[Abstract/Free Full Text]
- Barner H.B., Labovitz A.J., Fiore A.C. Prosthetic valves for the small aortic root. J Cardiac Surg 1994;9(Suppl):S154-157.
- Pibarot P., Dumesnil J.G., Jobin J., Cartier P., Honos G., Durand L.G. Haemodynamic and physical performance during maximal exercise in patients with an aortic bioprosthetic valve: comparison of stentless versus stented bioprostheses. J Am Coll Cardiol 1999;34:1609-1617.[Abstract/Free Full Text]
Top
Abstract
Introduction
