Author + information
- Received June 1, 2018
- Revision received September 17, 2018
- Accepted September 19, 2018
- Published online February 4, 2019.
- Georg Goliasch, MD, PhDa,
- Andreas A. Kammerlander, MDa,
- Christian Nitsche, MDa,
- Carolina Donaa,
- Laurin Schachnera,
- Begüm Öztürka,
- Christina Binder, MDa,
- Franz Duca, MDa,
- Stefan Aschauer, MDa,
- Günther Laufer, MDb,
- Christian Hengstenberg, MDa,
- Diana Bonderman, MDa and
- Julia Mascherbauer, MDa,∗ ()
- aDepartment of Internal Medicine II, Medical University of Vienna, Vienna, Austria
- bDepartment of Cardiac Surgery, Medical University of Vienna, Vienna, Austria
- ↵∗Address for correspondence:
Dr. Julia Mascherbauer, Medical University of Vienna, Department of Internal Medicine II, Waehringer Guertel 18-20, A-1090 Vienna, Austria.
Objectives Decision making in severe aortic stenosis (AS) requires a comprehensive pre-operative evaluation of the risk-to-benefit ratio. The aim of this study was to assess whether certain pre-operative symptoms are associated with outcome after surgical aortic valve replacement (SAVR).
Background The cardinal symptoms of AS indicating a need for intervention are angina, symptoms of heart failure, and syncope. Nevertheless, it remains unknown whether the presence of these more advanced symptoms conveys an increased risk after SAVR and whether the detection of early symptoms in patients with asymptomatic AS should be emphasized more in routine clinical practice.
Methods A total of 625 patients with isolated severe AS undergoing elective SAVR were prospectively enrolled in this long-term observational study.
Results Patients experiencing syncope had significantly smaller left ventricular diameters (p = 0.02), left atrial diameters (p = 0.043), right ventricular diameters (p = 0.04), and right atrial diameters (p = 0.001), smaller aortic valve areas (p = 0.048), and lower indexed stroke volumes (p = 0.043) compared with patients without syncope. Syncope conveyed an increased risk for mortality after SAVR that persisted after multivariate adjustment for a bootstrap-selected confounder model, with an adjusted hazard ratio of 2.27 (95% confidence interval: 1.04 to 4.95; p = 0.04) for 1-year short-term mortality and an adjusted hazard ratio of 2.11 (95% confidence interval: 1.39 to 3.21; p < 0.001) for 10-year long-term mortality. In contrast, pre-operative dyspnea, angina, and reduced left ventricular function were not significantly associated with outcomes.
Conclusions This long-term observational study in a large contemporary cohort of patients with AS for the first time demonstrates that syncope represents an underestimated threat in aortic stenosis, associated with poor prognosis after SAVR. Importantly, other primary indications for SAVR (i.e., dyspnea, angina, and decreased left ventricular function) were associated with significantly better post-operative outcomes than syncope. Patients experiencing syncope displayed a specific pathophysiologic phenotype characterized by a smaller aortic valve area, smaller cardiac cavities, and lower stroke volumes.
Aortic stenosis (AS) represents the most common indication for valve replacement in Europe, with an ever-increasing disease prevalence due to the aging population (1). The 3 cardinal symptoms of AS indicating a need for intervention are angina, symptoms of heart failure, and syncope (2). Angina can be explained by an increased myocardial oxygen demand of the hypertrophied left ventricular (LV) myocardium and limited coronary blood supply. Heart failure symptoms have been shown to be caused by diastolic dysfunction, leading to increased filling pressures, post-capillary pulmonary hypertension, and subsequent right ventricular impairment (3). Syncope, finally, is considered the result of the inability of the heart to adequately increase stroke volume in the circumstances of increased peripheral demand (4).
LV remodeling response and the advent of symptoms are highly variable in AS (5). Symptoms of AS are typically deceptive in the beginning, with most patients initially experiencing decreased exercise tolerance, whereas the classic indicators angina, dyspnea, and syncope are considered late manifestations of the disease process (6) that are linked to specifically impaired hemodynamic patterns (i.e., markedly altered diastolic function and subsequently increased filling pressures for dyspnea in contrast to reduced stroke volume along with smaller cardiac cavities for syncope) (7). However, it is the common notion that after surgical aortic valve replacement (SAVR), symptoms diminish, quality of life improves, and long-term survival has been proclaimed to be similar to that expected for an age-matched population (8). Nevertheless, it remains unknown whether the presence of more advanced symptoms, synonymous for “advanced state of disease,” conveys an increased risk beyond SAVR and whether the detection of early symptoms in patients with asymptomatic AS should be emphasized more in routine clinical practice. Because decision making in severe AS requires a comprehensive pre-operative evaluation of the risk-to-benefit ratio, the aim of this study was to assess whether certain pre-operative symptoms might serve as surrogates of disease severity and are associated with outcomes after SAVR.
We prospectively enrolled consecutive adult patients with isolated severe AS undergoing elective SAVR at the Vienna General Hospital, a university-affiliated tertiary center in this observational study. Severe AS was defined in agreement with current guidelines as a peak aortic valve (AV) velocity >4 m/s, a mean valvular gradient ≥40 mm Hg, and a calculated AV area <1 cm2 (9). Patients undergoing elective SAVR presented with symptoms (exertional dyspnea, New York Heart Association [NYHA] functional class ≥II; exertional angina, Canadian Cardiovascular Society [CCS] class ≥II; syncope) or with reduced LV function (ejection fraction ≤50%). The study was approved by the ethics committee of the Medical University of Vienna.
Clinical measures and follow-up
Medical history, electrocardiogram, and a transthoracic echocardiogram were collected at baseline. Patient-prosthesis mismatch was defined as an effective orifice area indexed to body surface area ≤0.8 cm2/m2 (10). Traditional cardiovascular risk factors were recorded according to the respective guidelines (11). All-cause mortality was selected as the primary study endpoint and presented as 1-year short-term mortality and 10-year long-term mortality. Mortality was determined via retrieval query of the Austrian Death Registry, including screening for cause of death (according to the International Statistical Classification of Diseases and Related Health Problems-Tenth Revision).
Standard echocardiograms were obtained using commercially available equipment (Vivid 5 and Vivid 7, GE Healthcare, Little Chalfont, United Kingdom; and Acuson Sequoia, Siemens Healthcare, Erlangen, Germany). Cardiac morphology was assessed using diameters in standard 4- and 2-chamber views. LV ejection fraction was calculated using the biplane Simpson method and assessed by experienced readers (12). AS and other concomitant valvular regurgitations were quantified using an integrated approach and graded as none, mild, mild to moderate, moderate, moderate to severe, or severe according to the respective guidelines (9,13). Systolic pulmonary artery pressures were calculated by adding the peak tricuspid regurgitation systolic gradient to the estimated central venous pressure (12).
Discrete data are presented as counts and percentages and were analyzed by using a chi-square test. Continuous data are presented as medians and interquartile ranges (IQRs) and were compared using the Mann-Whitney U test. Cox proportional hazard regression analysis was applied to assess the impact of the respective primary indication for SAVR (i.e., exertional dyspnea, NYHA functional class II; exertional angina, CCS class ≥II; syncope; or reduced LV function) on post-operative short-term and long-term survival. To account for potential confounders, we used a stepwise bootstrap resampling procedure including all baseline variables to identify best-fitting variables for the final multivariate Cox regression model. Five hundred repeats with a p value of 0.05 for selection were performed for long-term mortality, and variables selected in 50% of all repeats were included in the final cofounder model (i.e., European System for Cardiac Operative Risk Evaluation score, diabetes, concomitant coronary artery bypass graft surgery, mean AV gradient). To test for interactions between primary indications for SAVR and all variables in the final model, we used Cox proportional hazard regression models with the respective indication for SAVR, a variable in question, and the interaction between both variables. The proportional hazards assumption was tested and satisfied in all cases using Schoenfeld residuals. Kaplan-Meier analysis was applied to evaluate the impact of the respective primary indication on survival and compared using the log-rank test. All statistical analyses were computed using SPSS version 24 (IBM, Armonk, New York) and Stata version 11 (StataCorp, College Station, Texas).
We enrolled 625 patients with isolated severe AS with a median age of 73 years (IQR: 64 to 78 years). Eighty-six percent of patients (n = 536) were in NYHA functional class II prior to surgery, 41% (n = 254) displayed exertional angina (CCS class ≥II), and 67 patients (11%) experienced syncope. Reduced LV function (ejection fraction ≤50%) was present in 47 patients (8%). The median additive European System for Cardiac Operative Risk Evaluation score was 6 (IQR: 5 to 8), 51% of patients were women, and 18% underwent additional coronary artery bypass graft surgery at the time of SAVR. Detailed baseline characteristics of the total study population are displayed in Table 1.
Syncope in AS associated clinical characteristics
Patients with syncope did not significantly differ in age (p = 0.38), sex (p = 0.15), pre-operative European System for Cardiac Operative Risk Evaluation score (p = 0.94), LV function (p = 0.72), and cardiovascular co-morbidities such as hypertension (p = 0.48), diabetes (p = 0.15), or concomitant coronary artery disease (p = 0.52) compared with patients without syncope. The prevalence of moderate or greater patient prosthesis mismatch did not significantly differ between patients with and without syncope (52% vs. 47%; p = 0.41). Fifty-nine patients with syncope had concomitant dyspnea (NYHA functional class ≥II). Patients experiencing syncope had significantly smaller end-diastolic LV diameters (45 mm [IQR: 42 to 51 mm] vs. 50 mm [IQR: 45 to 54 mm]; p = 0.02), left atrial diameters (55 mm [IQR: 48 to 60 mm] vs. 56 mm [IQR: 52 to 62 mm]; p = 0.043), right atrial diameters (49 mm [IQR: 46 to 54 mm] vs. 53 mm [IQR: 48 to 57 mm]; p = 0.001), and right ventricular end-diastolic diameters (30 mm [IQR: 27 to 33 mm] vs. 31 mm [IQR: 28 to 34 mm]; p = 0.04) compared with patients without syncope. Furthermore, AV area was significantly smaller in patients with syncope (0.55 cm2 [IQR: 0.50 to 0.80 cm2] vs. 0.60 cm2 [IQR: 0.55 to 0.79 cm2]; p = 0.048), and they were less likely to experience angina on exertion (43% vs. 24%; p = 0.006). Detailed baseline characteristics comparing patients with and without syncope are presented in Table 1.
Syncope and post-operative outcomes following AV replacement
During a follow-up period of 10 years, 33 patients (49%) experiencing pre-operative syncope died, compared with 189 patients (34%) without syncope (p = 0.029). Kaplan-Meier plots demonstrated that patients experiencing syncope had significantly worse long-term survival compared with those who experienced exertional dyspnea (NYHA functional class ≥II), exertional angina (CCS class ≥II), or reduced LV function (p = 0.004 for comparison between syncope and other primary indications for SAVR) (Figure 1).
Syncope was significantly associated with an impaired post-operative outcome, with a crude hazard ratio (HR) of 2.85 (95% confidence interval [CI]: 1.40 to 5.82; p = 0.004) (Table 2) for 1-year short-term mortality (Figure 2A) and a crude HR of 2.15 (95% CI: 1.45 to 3.19; p < 0.002) (Table 2) for 10-year long-term mortality (Figure 2B). This association persisted after multivariate adjustment for our bootstrap-selected confounder model, with an adjusted HR of 2.27 (95% CI: 1.04 to 4.95; p = 0.04) (Table 3) for 1-year short-term mortality and an adjusted HR of 2.11 (95% CI: 1.39 to 3.21; p < 0.001) (Table 3) for 10-year long-term mortality. In contrast, dyspnea (pre-operative NYHA functional class ≥II), exertional angina (CCS class ≥II), and reduced LV function were not significantly associated with short- or long-term outcomes in both univariate (Table 2) and multivariate (Table 3) analyses. Of note, the modes of death did not significantly differ between patients with and without syncope (cardiac death, 45% vs. 51%; malignancies, 25% vs. 17%; cerebrovascular death, 6% vs. 6%; respiratory death, 6% vs. 5%; other causes, 18% vs. 21%).
Our data show increased short- and long-term mortality after aortic valve replacement for patients with syncope. In contrast, other primary indications for SAVR (i.e., heart failure symptoms, angina, and decreased LV function) were not associated with adverse post-operative outcomes. Patients experiencing syncope displayed a specific pathophysiologic phenotype, characterized by smaller AV area and smaller cardiac cavities.
Pathophysiology of syncope in AS
This observational study in a large contemporary cohort of patients with AS undergoing surgical SAVR demonstrates for the first time the prognostic significance of pre-operative syncope, which is reflected by a more than 2-fold higher post-operative mortality compared with other primary indications for SAVR, although the distribution of classic cardiovascular risk factors was equal.
The exact pathophysiologic mechanism of syncope in AS remains unknown (14). The current concept assumes an exercise-induced sudden decrease in total peripheral resistance, which cannot be compensated, because cardiac output is restricted by the stenotic AV (15,16). Indeed, we found AV area to be closely related to the occurrence of syncope, supporting this hypothesis. In addition, our results extend previous findings by Park et al. (7), who described a certain pathophysiologic phenotype characterized by smaller left ventricles and left atria, along with reduced stroke volumes, that predispose for syncope. Our findings confirm that patients with AS with syncope on average present with smaller heart cavities, obviously resulting in lower stroke volumes. It is tempting to speculate that the combination of small, more hypertrophic and presumably stiffer left ventricles may represent a specific and more advanced disease entity among patients with AS. These changes may not regress after SAVR and cause persistent incapability to adequately increase cardiac output, finally contributing to the observed impairment of long-term outcomes. This adverse effect appears to be more pronounced within the early post-operative period, a period aggravated by transient myocardial stunning as a sequela of cardiopulmonary bypass.
Clinical implications for patient management
Although all cardinal symptoms of AS (i.e., angina, dyspnea, syncope, and reduced LV function) are considered late manifestations of the disease process, syncope appears to carry the highest risk for impaired survival after SAVR. Given the importance of symptom onset in clinical decision making, our results mandate that primary care physicians should explore “asymptomatic” patients with severe AS very carefully with regard to early symptoms, such as dizziness, fatigue, and a decrease in physical activity, because many patients unconsciously limit activities to avoid symptoms as valve obstruction slowly worsens (6,17). When symptomatic status remains unclear, or in high-risk patients with small AV area and cardiac cavities, standard treadmill exercise testing might be considered to detect provoked symptoms in order to potentially prevent the occurrence of syncope (18,19). Interestingly, other primary indications for SAVR (i.e., heart failure symptoms, angina, and decreased LV function) were not associated with adverse post-operative outcomes. However, only 8% of our study population presented with reduced LV function, which might limit the generalizability of our results in this regard, especially in light of recent studies that demonstrated worse clinical outcomes in patients with reduced LV function (20,21).
An emerging body of evidence, relying on retrospective and observational data, implies that early SAVR might lead to favorable outcomes compared with surgery after symptom onset (22–24). Furthermore, operative risk for isolated SAVR is low at established centers (25), and transcatheter aortic valve replacement has currently become an attractive, minimally invasive alternative to SAVR in patients at high and intermediate surgical risk (26,27). Therefore, there is currently growing interest in identifying asymptomatic or oligosymptomatic patients with severe AS who may benefit from early SAVR. Although it remains highly speculative, our data might support this rationale in oligosymptomatic patients before the occurrence of late disease manifestations such as syncope. Nevertheless, whether early SAVR can be translated into a clinically meaningful improvement of outcomes compared with “watchful waiting” is a matter of debate. Several randomized clinical trials (i.e., EARLY TAVR [NCT03042104], AVATAR [NCT02436655]) are currently under way to test this hypothesis.
Our study reflects the experience of a single tertiary care center with a long-standing heart valve clinic program. Almost 75% of patients are referred to our institution when already symptomatic. Delays in referral and symptom reporting in patients with AS are known to be common, therefore our patient population might represent a more advanced spectrum of patients with AS. Thus, a center-specific bias cannot be excluded. However, major advantages of single-center studies are 1) inclusion of a homogenous patient population; 2) adherence to a constant clinical routine; 3) standardized intra- and perioperative treatment; 4) constant quality of echocardiographic examinations; and 5) constant follow-up. Syncope may have been misinterpreted as a symptom of AS, whereas it might have been caused by a different underlying disease. However, because cardiovascular mortality was not significantly different between patients with and without syncope, this concern seems to appear less relevant to us. Furthermore, the number of patients who experienced syncope are limited and concomitant assessment of LV mass, diastolic function, and invasive hemodynamic data would have been desirable. Nevertheless, to the best of our knowledge, the present data contain the most comprehensive information on the impact of pre-operative symptoms on prognosis following SAVR.
This long-term observational study in a large contemporary cohort of patients with AS for the first time demonstrates that syncope represents an underestimated threat in AS, associated with poor outcome, even beyond SAVR. Importantly, patients with AS presenting with heart failure symptoms or angina had significantly better post-operative outcomes than those with syncope. Patients experiencing syncope displayed a specific pathophysiologic phenotype characterized by a smaller AV area, smaller cardiac cavities, and lower stroke volume.
COMPETENCY IN PATIENT CARE AND PROCEDURAL SKILLS: Although all cardinal symptoms of AS (i.e., angina, dyspnea, and syncope) and reduced LV function are considered late manifestations of the disease process, syncope appears to be associated with the worst outcomes, with significantly higher mortality rates both operatively and in the long run.
TRANSLATIONAL OUTLOOK: Given the prognostic importance of symptom onset in AS, our results mandate a careful assessment of asymptomatic patients with severe AS with regard to early symptoms. Although it remains highly speculative, our data might support the rationale of early AVR in oligosymptomatic patients before the occurrence of late disease manifestations such as syncope.
The authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- aortic stenosis
- aortic valve
- Canadian Cardiovascular Society
- confidence interval
- hazard ratio
- interquartile range
- left atrium
- left ventricular
- New York Heart Association
- right atrium
- right ventricle
- surgical aortic valve replacement
- Received June 1, 2018.
- Revision received September 17, 2018.
- Accepted September 19, 2018.
- 2019 American College of Cardiology Foundation
- Kammerlander A.A.,
- Marzluf B.A.,
- Graf A.,
- et al.
- Singh A.,
- Chan D.C.S.,
- Greenwood J.P.,
- et al.
- Park S.J.,
- Enriquez-Sarano M.,
- Chang S.A.,
- et al.
- Kvidal P.,
- Bergstrom R.,
- Horte L.G.,
- Stahle E.
- Nishimura R.A.,
- Otto C.M.,
- Bonow R.O.,
- et al.
- Piepoli M.F.,
- Hoes A.W.,
- Agewall S.,
- et al.
- Lang R.M.,
- Badano L.P.,
- Mor-Avi V.,
- et al.
- Baumgartner H.,
- Hung J.,
- Bermejo J.,
- et al.
- Rajani R.,
- Rimington H.,
- Chambers J.B.
- Taniguchi T.,
- Morimoto T.,
- Shiomi H.,
- et al.
- Ito S.,
- Miranda W.R.,
- Nkomo V.T.,
- et al.
- Kang D.H.,
- Park S.J.,
- Rim J.H.,
- et al.
- Taniguchi T.,
- Morimoto T.,
- Shiomi H.,
- et al.
- Reardon M.J.,
- Van Mieghem N.M.,
- Popma J.J.,
- et al.