Author + information
- Received September 10, 2012
- Revision received November 20, 2012
- Accepted November 27, 2012
- Published online April 9, 2013.
- Robert Zilberszac, MD⁎,
- Harald Gabriel, MD⁎,
- Michael Schemper, PhD†,
- David Zahler, MD⁎,
- Martin Czerny, MD‡,
- Gerald Maurer, MD⁎ and
- Raphael Rosenhek, MD⁎,⁎ ()
- ↵⁎Reprint requests and correspondence:
Dr. Raphael Rosenhek, Department of Cardiology, Vienna General Hospital, Medical University of Vienna, Waehringer Guertel 18-20, 1090 Vienna, Austria
Objectives This study sought to describe the natural history of combined stenotic and regurgitant aortic valve disease.
Background Data on outcome and prognostic factors in combined aortic valve disease are scarce.
Methods This study prospectively followed 71 consecutive asymptomatic patients (21 women, age 52 ± 17 years) with at least moderate aortic stenosis in combination with at least moderate aortic regurgitation and preserved left ventricular function (ejection fraction ≥55%).
Results During a median potential follow-up of 8.9 years, 50 patients developed an indication for aortic valve replacement and no cardiac deaths were observed. Overall event rates were high with an event-free survival for the entire patient population of 82 ± 5%, 62 ± 6%, 49 ± 6%, 33 ± 6%, and 19 ± 5% at 1, 2, 3, 4, and 6 years, respectively. There was 1 operative and no post-operative deaths. Peak aortic jet velocity (AV-Vel) independently predicted event-free survival. Patients with an AV-Vel between 3 and 3.9 m/s had an event-free survival of 94 ± 4%, 88 ± 6%, 65 ± 9%, and 51 ± 9% after 1, 2, 4, and 6 years, respectively, compared with 92 ± 4%, 67 ± 7%, 38 ± 8%, and 12 ± 6% for patients with an AV-Vel between 4 and 4.9 m/s and 67 ± 8%, 39 ± 10%, 17 ± 9%, and 0% for patients with an AV-Vel ≥5 m/s (p < 0.0001).
Conclusions Asymptomatic patients with combined aortic valve disease can be safely followed until surgical criteria defined for aortic stenosis, aortic regurgitation, or the aorta are reached. However, high event rates can be expected even in younger patients and those with only moderate disease. AV-Vel, which reflects both stenosis and regurgitant severity, provides an objective and easily assessable predictive parameter.
Aortic valve disease is frequently encountered in clinical practice with a prevalence of 2% in patients between 65 and 75 years and of 6% in patients older than 75 years in population-based studies (1). Aortic stenosis (AS) is the most common single-valve disease accounting for one-third of patients referred to treatment, followed by aortic regurgitation (AR), which accounts for 10% of those referred to surgery (2). Data on combined stenotic and regurgitant aortic valve disease are scarce, both with regard to its epidemiological prevalence and even more importantly with regard to its natural history (3). Knowledge of this information would be needed for an adequate clinical decision-making process (4). The management guidelines conclude that data on mixed valve disease are lacking and thus making evidence-based recommendations is difficult (5,6). In clinical practice, management of patients with combined aortic disease follows the recommendations concerning the predominant lesion and in the case of balanced stenosis and regurgitation on how well the lesion is tolerated by the patient (5). However, the combination of pressure and volume load represents a distinct form of myocardial stress. Symptoms warrant surgery, both in AS and in AR (7,8). Furthermore, surgery is indicated in asymptomatic patients with AS or AR when left ventricular function is impaired (5,9,10). In AR, marked left ventricular dilation is also considered as an indication for surgery (9,11). Early elective surgery may be considered in asymptomatic AS patients with rapid hemodynamic progression (increase in peak aortic jet velocity [AV-Vel] of at least 0.3 m/s/year) in the presence of a calcified aortic valve (5,6,12).
Whereas no predictors of outcome have been defined for combined aortic valve disease, AV-Vel is of interest in this regard because it is affected by the severity of both AS and AR. It may thus be a reliable parameter reflecting the overall hemodynamic load on the left ventricle. AV-Vel has been previously identified as an important predictor of outcome in asymptomatic patients with AS without coexisting AR (12–15).
Therefore, we prospectively studied a cohort of consecutive asymptomatic patients with combined aortic valve disease having at least moderate stenosis and moderate regurgitation in order to assess the natural history of these patients and to assess the predictive value of AV-Vel in these patients.
All patients who were studied in our outpatient clinic for valvular heart disease between 1995 and 2009 and who were found to have a combination of at least moderate AS with at least moderate AR were prospectively included into the study when they had preserved left ventricular function (ejection fraction ≥55%). Exclusion criteria were additional hemodynamically significant valve lesions (moderate or severe) or the presence of symptoms.
According to these criteria, 71 consecutive patients (age 52 ± 17 years; 21 women) were identified. At entry, 22 patients had moderate AS and moderate AR; 33 patients had moderate AR and severe AS; 7 patients had moderate AS and severe AR; and 9 patients had severe AR and severe AS. Thirty-five patients had bicuspid aortic valves, and 36 patients had tricuspid valves. All but 24 patients had moderate-to-severe valve calcification, and 18 of those 24 patients had bicuspid aortic valves. Color Doppler showed additional mild mitral regurgitation in 32 patients and mild-to-moderate mitral regurgitation in 4. Sixteen patients had mild and 6 had mild-to-moderate tricuspid regurgitation.
Clinical data were recorded at study entry as follows: age, sex, history of hypercholesterolemia (total cholesterol >200 mg/dl or patient under lipid-lowering therapy), diabetes mellitus, arterial hypertension (on the basis of the average of repeated measurements: blood pressure >140/90 mm Hg for patients with nonsevere AR and systolic blood pressure >140 mm Hg for patients with severe AR), and coronary artery disease (history of myocardial infarction, angioplasty, coronary artery bypass grafting, or angiographically documented coronary artery stenosis).
Echocardiography was performed using commercially available ultrasound systems. All patients underwent a comprehensive examination including M-mode, 2-dimensional echocardiography, conventional and color Doppler by an experienced echocardiographer. Multiple transducer positions were used to record AV-Vel and aortic valve area was calculated using the continuity equation (16).
An integrated approach was used for the quantification of AS and AR as proposed by current recommendations (17–19). Severe AS was defined by an aortic valve area ≤1.0 cm2 and moderate AS by a valve area of >1.0 to 1.5 cm2. A vena contracta of more than 6 mm (20) and a prominent holodiastolic flow reversal in the descending aorta thoracalis were used to define severe AR and moderate AR was defined by a vena contracta of 3 to 6 mm. When assessing the severity of AR, coexisting AS that is associated with left ventricular hypertrophy and impaired relaxation may lead to a prolonged pressure half-time. In the grading of lesion severity, potential influences of the combined lesion on the respective measures of severity were thus considered (3). The presence of concomitant AR may lead to higher transaortic flow rates and consequently to higher transvalvular gradients.
The degree of aortic valve calcification was scored according to previously described criteria (12).
Patients were followed prospectively after the initial examination at the outpatient clinic for valvular heart disease. Patients were scheduled to undergo follow-up exams every 6 months based on a watchful waiting approach and were referred to surgery once they had reached an indication for surgery of either AR or AS (3,4). Before undergoing surgery, coronary angiography was performed in all patients. The follow-up information was obtained from interviews with the patients, their relatives, and their physicians. Information regarding the development of cardiac symptoms, eventual aortic valve replacement, and death was obtained. Exercise testing was performed in selected patients when doubt about whether they were truly asymptomatic existed. The decision to perform an exercise test was made on an individual basis according to clinical judgment. An exercise test was considered as an indication for surgery when limiting dyspnea or angina occurred at subnormal workload. Rapidly reversible breathlessness and fatigue at workloads close to the age-sex predicted maximum workloads were not considered as an indication for surgery.
For the assessment of outcome, endpoints were defined as cardiac death or indication for aortic valve replacement according to the accepted indications for AR and AS in agreement with current guidelines (5,6).
Deaths were classified as noncardiac or cardiac-based. Cardiac deaths were classified as directly related to aortic valve disease (sudden death, congestive heart failure) or to other cardiac pathology. Patients who had a noncardiac cause of death were censored at the time of death.
Quantitative baseline patient characteristics are expressed as mean ± SD. Event-free probabilities and corresponding standard errors have been obtained by the Kaplan-Meier method (Figs. 1 to 3⇓⇓⇓). Follow-up was quantified by means of the reverse Kaplan-Meier estimator (21). The effect of different predictors (age, sex, hypercholesterolemia, diabetes mellitus, arterial hypertension, coronary artery disease, aortic valve area, and aortic valve jet velocity) was studied by simple and multiple Cox regression models. The strength of the effect was quantified by the unadjusted and adjusted hazard ratio estimates. The assumption of proportional hazards was assessed by adding interactions of all prognostic factors with the log of time. A p value <0.05 was considered to indicate statistical significance.
For the graphical representation of the time-dependent predictor (AV-Vel), the origin (time 0) for each of the Kaplan-Meier curves (refers to Fig. 2A) signifies either classification at study entry or at the time of switching to a higher AV-Vel group. Individuals changing to a higher group during follow-up are not censored in the initial group. This permits an adequate graphical representation of the magnitude of the effect of the time-dependent prognostic factor (AV-Vel), which was analyzed as a time-dependent covariate in a standard Cox regression. Figure 2B represents the outcome of patients stratified according to their last documented AV-Vel: patients were censored in the lower AV-Vel group when their AV-Vel progressed and they switched to a higher group.
Follow-up information was complete for 70 patients (98.6%). The baseline patient characteristics are summarized in Table 1.
Pressure half-time was significantly shorter in patients with severe AR than in those with moderate AR (293 ± 25 ms vs. 388 ± 15 ms; p = 0.0021). Furthermore, left ventricular end-diastolic diameters were significantly larger in patients with severe AR as compared to patients with moderate AR (61.4 ± 1.7 mm vs. 53.1 ± 0.8 mm; p < 0.001). At baseline, 36 patients presented with an AV-Vel of 3.0 to 3.9 m/s, 28 patients with an AV-Vel of 4.0 to 4.9 m/s, and 7 patients with an AV-Vel of ≥5.0 m/s. During follow-up, 26 patients had a progression of their AV-Vel from 3.0 to 3.9 m/s to 4.0 to 4.9 m/s, 23 patients from 4.0 to 4.9 m/s to ≥5m/s (11 of these had already progressed from 3.0 to 3.9 m/s to 4.0 to 4.9 m/s) and 3 patients progressed from 3.0 to 3.9 m/s to ≥5.0 m/s.
During a median potential follow-up of 8.9 years, 50 patients developed criteria warranting aortic valve replacement. No cardiac deaths occurred in asymptomatic patients without an indication for surgery. Event-free survival was 82 ± 5%, 62 ± 6%, 49 ± 6%, 33 ± 6%, and 19 ± 5% at 1, 2, 3, 4, and 6 years, respectively (Fig. 1).
One noncardiac death occurred in a patient who did not develop criteria for surgery. This patient died of late sequelae after a major stroke with hemiparesis and severe disability. Three patients had an indication for surgery due to progressive symptoms but refused aortic valve replacement surgery. Of these, 1 patient died of congestive heart failure, 1 of a stroke, and the cause of death remained unknown in 1 patient. No sudden deaths occurred during follow-up, and 1 patient died 10 days after aortic valve replacement.
Indications for aortic valve replacement surgery
During follow-up, surgery was indicated in 50 patients. When reaching an indication for surgery, AS was the predominant lesion in 28 patients; AR was the predominant lesion in 8 patients; and both regurgitation and stenosis were severe in 14 patients. Indications for surgery were as follows: symptoms (n = 33); symptoms unmasked by an exercise test (n = 3); rapid hemodynamic progression of AS, defined by an increase in AV-Vel of at least 0.3 m/s within 1 year in the presence of a calcified aortic valve (n = 5); aortic aneurysm (n = 3; of which 2 had bicuspid aortic valves); aortic dissection in a patient with a tricuspid aortic valve (n = 1); before major noncardiac surgery (n = 2); endocarditis (n = 2); and left ventricular dysfunction (n = 1). No patient required surgery because of criteria of left ventricular dilation.
Forty-three of these patients underwent aortic valve replacement, whereas 6 patients refused surgery. One patient was denied surgery because of metastatic colorectal cancer from which he died after 1 year. Fifteen patients received a biological valve prosthesis, 17 a mechanical valve prosthesis, and 10 underwent a Ross procedure. One patient received concomitant aortocoronary bypass surgery.
One death occurred in a 67-year-old man 10 days after aortic valve replacement and single coronary artery bypass surgery. This patient had presented with mild exertional dyspnea and a pre-operative AV-Vel of 5.7 m/s. No post-operative deaths were observed during follow-up.
Potential prognostic factors
Among patients with combined aortic valve disease, patients with moderate and those with severe AS at baseline did not have significantly different event-free survival rates: patients with an aortic valve area of <1.0 cm2 had an outcome that was comparable to those with a valve area between 1.0 and 1.5 cm2 (p = 0.57). Also the degree of AR did not have prognostic value and patients with moderate AR and those with severe AR had a similar outcome (p = 0.81).
Patients with moderate AS and moderate AR had event-free survival rates of 100 ± 0%, 75 ± 10%, 24 ± 10%, and 18 ± 9% at 1, 2, 4, and 6 years, respectively, as compared to 77 ± 8%, 53 ± 9%, 42 ± 9%, and 19 ± 8% for patients with severe AS and moderate AR; 71 ± 17%, 54 ± 20%, 36 ± 20%, and 18 ± 16% for patients with moderate AS and severe AR and 76 ± 15%, 63 ± 17%, 42 ± 21%, and 21 ± 18% for patients with severe AS and severe AR (p = 0.9) (Fig. 3). However, AV-Vel was a significant predictor of outcome allowing further risk stratification among patients with combined stenotic and regurgitant aortic valve disease. The likelihood of having an event with an AV-Vel between 3.0 and 3.9 m/s is low (Fig. 2B), however rapid progression to a higher velocity group with the subsequent occurrence of events is common in these patients (Fig. 2A).
Event-free survival rates for patients with an AV-Vel between 3.0 and 3.9 m/s were 94 ± 4%, 88 ± 6%, 65 ± 9%, and 51 ± 9% after 1, 2, 4, and 6 years, respectively. Event-free survival rates were significantly worse for patients with an AV-Vel between 4.0 and 4.9 m/s: 92 ± 4% at 1 year; 67 ± 7% at 2 years; 38 ± 8% at 4 years; and 12 ± 6% at 6 years. Event-free survival rates were even worse for patients with an AV-Vel ≥5.0 m/s: 67 ± 8% at 1 year; 39 ± 10% at 2 years; 17 ± 9% at 4 years; and 0% at 6 years (p < 0.001) (Fig. 2A).
Event-free survival rates for patients with coronary artery disease were 63 ± 17%, 25 ± 15%, and 0 ± 0% after 1, 2, 4, and 6 years, respectively, as compared with 85 ± 5%, 69 ± 6%, 36 ± 7%, and 22 ± 6% after 1, 2, 4, and 6 years, respectively, for patients where no coronary artery disease was documented upon cardiac catheterization (p = 0.0019). In univariate analysis, age and arterial hypertension were significant predictors of outcome (p = 0.005 and p = 0.0003, respectively).
In the multivariate analysis (Table 2), AV-Vel and coronary artery disease were the only significant predictors of outcome. The hazard ratio for an increase of AV-Vel from 3.0 to 3.9m/s to 4.0 to 4.9 m/s or from 4.0 to 4.9 m/s to ≥5.0 m/s was 3.3 (95% confidence interval [CI]: 2.1 to 5.3). Age, sex, hypercholesterolemia, diabetes mellitus, and arterial hypertension were not found to be significant predictors of outcome in the multivariate analysis. A time-dependent effect of aortic valve area was detected as an interaction of aortic valve area with the log of time. However, this effect was not further explored by weighted Cox regression due to imprecision in the determination of aortic valve area related to technical limitations of the continuity equation and the lack of clinical relevance and plausibility of such a potential effect in daily clinical practice. The effect of all other considered factors appeared constant in time.
The main disease etiologies were bicuspid and degenerative calcific aortic valve disease, accounting for approximately one-half of the patients, each. Obviously it cannot be excluded that some of the extensively calcified valves have an underlying masked bicuspid etiology (22). In an analysis of surgical findings in patients who had undergone aortic valve surgery for combined aortic valve disease in the 1960s and 1970s, post-inflammatory disease accounted for 69% of the cases of combined aortic valve disease (23). The fact that no patient in this series had rheumatic disease can be explained by its decreasing incidence and the fact that these patients frequently present with concomitant mitral disease, which was an exclusion criterion in the present study. Six percent (n = 4) of the patients required surgery because of aortic aneurysm or dissection, of which one-half had a bicuspid aortic valve. The ascending aorta should thus be routinely assessed during the follow-up visits.
Natural history of combined aortic valve disease
This is the first study to specifically assess the natural history and the outcome of a relatively large group of patients with combined stenotic and regurgitant aortic valve disease. Most available evidence on the natural history of aortic valve disease is related to either AS or AR. Due to a strict lack of data for combined aortic valve disease, no specific recommendations have been issued so far for combined stenotic and regurgitant aortic valve disease. Although a substantial number of the patients included in the present study had only moderate combined aortic valve disease and although they were relatively young, with a mean age of 51 years, the overall event rate was high: 38% of the patients required surgery within 2 years and 67% within 4 years. Thus, close clinical and echocardiographic follow-up is recommended in this patient group, even when a patient presents with combined aortic valve disease of only moderate severity. Patients should also be educated about the likelihood of a rapid onset of symptoms and that aortic valve surgery may be required in the near future.
Indications for surgery
The current indications for surgery as well as recommendations on the scheduling of follow-up intervals have been defined for single-valve disease and are “extrapolated” to patients with combined valve disease.
Although many clinicians follow these patients conservatively until surgical criteria defined for AS, AR, or the aorta are reached, the present data confirm for the first time that such a strategy is safe and results in good surgical and post-operative outcomes. One cardiac death occurred 10 days after the procedure in the only patient in our series that had undergone concomitant aortocoronary bypass surgery. At the same time, patients can expect to have a good outcome when surgery is performed at that stage. Coronary artery disease, which was relatively rare in this younger patient population with a prevalence of 11%, was a significant predictor of outcome in multivariate analysis. Combined aortic valve disease seems to be less well tolerated in the presence of coronary artery disease, which leads to a relatively early symptom onset.
Peak aortic jet velocity as a measure of severity of combined aortic valve disease
The precise quantification of the severity of each component of combined aortic valve disease is complex because the measures of severity may be influenced by the coexisting lesion (3). As has already been mentioned, AR may lead to higher transvalvular gradients related to an increased transaortic flow rate, and pressure half-time may be prolonged when left ventricular hypertrophy with impaired relaxation is present due to AS (24). AV-Vel has the potential advantage of being a simple, reproducible, and quantitative measure that comprehensively assesses the overall hemodynamic load of combined aortic valve disease. In contrast, event-free survival was comparable for the 4 patient subgroups that were stratified according to semiquantitative severity of AS and AR. Even patients having moderate stenosis and moderate regurgitation had a high event rate. The present study demonstrates that AV-Vel is a solid parameter that allows the prediction of event-free survival in patients with combined aortic valve disease. Its value is thus also of importance, when scheduling control exams in these patients and should be integrated into the discussion of the timing of surgery.
Moreover, no additional prognostic information is provided by aortic valve area. This may not be surprising, because aortic valve area reflects the severity of AS only, without accounting for AR severity.
We have previously shown that AV-Vel predicts outcome in isolated AS across the entire spectrum from mild to very severe disease (14,15). Patients with combined aortic disease with an AV-Vel of 3.0 to 3.9 m/s have lower event-free survival rates (65 ± 9% at 2 years and 51 ± 9% at 4 years) than do patients with isolated moderate AS (AV-Vel 3.0 to 3.9 m/s) with an event-free survival of 79 ± 4% and 61 ± 5% at 2 and 4 years, respectively (14).
For an AV-Vel of 4.0 to 4.9 m/s and an AV-Vel ≥5.0 m/s, event rates are comparable for patients with combined aortic disease and those with isolated AS. For patients with an AV-Vel of 4.0 to 4.9 m/s, event-free survival rates at 2 and 4 years are 67 ± 7% and 38 ± 8% when they have combined aortic disease as compared to 70 ± 5% and 39 ± 16% for patients with isolated AS, respectively. For patients with an AV-Vel ≥5.0 m/s, event-free survival rates at 2 and 4 years are 39 ± 10% and 17 ± 9% when they have combined aortic disease as compared to 36 ± 5% and 12 ± 4% for patients with isolated AS, respectively (15).
Regurgitant orifice area and regurgitant volume measures were not systematically used in the present study; however, a careful integrated approach for the assessment of AR was performed as recommended (17,18). The study was not designed to assess and compare surgical risks and is limited in this regard by the number of patients undergoing surgery.
Asymptomatic patients with combined aortic valve disease can be safely followed until surgical criteria defined for AS, AR, or the aorta are reached. Good surgical and post-operative outcomes are achieved when following such a strategy. However, high event rates can be expected despite a relatively young age of the population and even in patients with moderate disease severity. AV-Vel, which reflects both stenosis and regurgitant severity, provides an objective and easily assessable parameter to risk stratify these patients.
The authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- aortic regurgitation
- aortic stenosis
- peak aortic jet velocity
- confidence interval
- Received September 10, 2012.
- Revision received November 20, 2012.
- Accepted November 27, 2012.
- American College of Cardiology Foundation
- Iung B.,
- Baron G.,
- Butchart E.G.,
- et al.
- Unger P.,
- Rosenhek R.,
- Dedobbeleer C.,
- Berrebi A.,
- Lancellotti P.
- Rosenhek R.,
- Iung B.,
- Tornos P.,
- et al.
- Vahanian A.,
- Baumgartner H.,
- Bax J.,
- et al.,
- ESC Committee for Practice Guidelines
- Bonow R.O.,
- Carabello B.A.,
- Chatterjee K.,
- et al.
- Dujardin K.S.,
- Enriquez-Sarano M.,
- Schaff H.V.,
- Bailey K.R.,
- Seward J.B.,
- Tajik A.J.
- Klodas E.,
- Enriquez-Sarano M.,
- Tajik A.J.,
- Mullany C.J.,
- Bailey K.R.,
- Seward J.B.
- Otto C.M.,
- Burwash I.G.,
- Legget M.E.,
- et al.
- Rosenhek R.,
- Klaar U.,
- Schemper M.,
- et al.
- Rosenhek R.,
- Zilberszac R.,
- Schemper M.,
- et al.
- Otto C.M.,
- Pearlman A.S.,
- Comess K.A.,
- Reamer R.P.,
- Janko C.L.,
- Huntsman L.L.
- Tribouilloy C.M.,
- Enriquez-Sarano M.,
- Bailey K.R.,
- Seward J.B.,
- Tajik A.J.
- Roberts W.C.,
- Ko J.M.
- de Marchi S.F.,
- Windecker S.,
- Aeschbacher B.C.,
- Seiler C.