Correlates and Causes of Death in Patients With Severe Symptomatic Aortic Stenosis Who Are Not Eligible to Participate in a Clinical Trial of Transcatheter Aortic Valve Implantation
Background—Transcatheter aortic valve implantation is currently being evaluated in patients with severe aortic stenosis who are considered high-risk surgical candidates. This study aimed to detect incidences, causes, and correlates of mortality in patients ineligible to participate in transcatheter aortic valve implantation studies.
Methods and Results—From April 2007 to July 2009, a cohort of 362 patients with severe aortic stenosis were screened and did not meet the inclusion/exclusion criteria necessary to participate in a transcatheter aortic valve implantation trial. These patients were classified into 2 groups: group 1 (medical): 274 (75.7%): 97 (35.4%) treated medically and 177 (64.6%) treated with balloon aortic valvuloplasty; and group 2 (surgical): 88 (24.3%). The medical/balloon aortic valvuloplasty group had significantly higher clinical risk compared with the surgical group, with significantly higher Society of Thoracic Surgeons score (12.8±7.0 versus 8.5±5.1; P<0.001) and logistic European System for Cardiac Operative Risk Evaluation (EuroSCORE) (42.4±22.8 versus 24.4±18.1; P<0.001). The medical/balloon aortic valvuloplasty group had a higher New York Heart Association functional class, incidence of renal failure, and lower ejection fraction. During median follow-up of 377.5 days, mortality in the medical/balloon aortic valvuloplasty group was 102 (37.2%), and during median follow-up of 386 days, mortality in the surgical group was 19 (21.5%). Multivariable adjustment analysis identified renal failure (hazard ratio [HR]: 5.60), New York Heart Association class IV (HR: 5.88), and aortic systolic pressure (HR: 0.99) as independent correlates for mortality in the medical group, whereas renal failure (HR: 7.45), Society of Thoracic Surgeons score (STS; HR: 1.09) and logistic EuroSCORE (HR: 1.45) were correlates of mortality in the in the surgical group.
Conclusion—Patients with severe symptomatic aortic stenosis not included in transcatheter aortic valve implantation trials do poorly and have extremely high mortality rates, especially in nonsurgical groups, and loss of quality of life in surgical groups.
Aortic stenosis (AS) is the most common valvular heart disease in the elderly; it affects 4.6% of adults older than 75 years.1 Aortic valve replacement (AVR) is recommended for symptomatic patients with severe AS or for those with left ventricular dysfunction, because their prognosis without surgical treatment is poor.2,3 Despite these facts, many symptomatic patients do not undergo operation. Data from the Euro Heart Survey on valvular heart disease revealed that ≤30% of patients with severe AS do not undergo AVR.4 Others reported even higher rates of ≤41%.5 In part, elderly patients with AS are reluctant to undergo surgery because of the potential of excess complications attributed to their advanced age and other comorbidities.
Transcatheter aortic valve implantation (TAVI) has been designed as a new option for this high-risk population. Current randomized TAVI trials for severe AS include high-risk patients for AVR defined as Society of Thoracic Surgeons (STS) score of >10 or a logistic European System for Cardiac Operative Risk Evaluation (EuroSCORE) of >20. In addition, all of the candidates for TAVI are evaluated by a study surgeon regarding their operative risk. Many patients with severe symptomatic AS who are screened for TAVI trials are not eligible based on strict inclusion/exclusion criteria. The aim of our study was to assess the prognosis of patients who are currently ineligible to participate in TAVI trials and to examine the clinical, echocardiographic, and hemodynamic factors associated with subsequent mortality.
Patient Population and Selection
The present was study was approved by the institutional review board of Washington Hospital Center. The study population included patients with symptomatic severe AS at high surgical risk, those considered non operable, and those who were referred for consideration for a randomized TAVI trial from April 2007 to July 2009. All were screened and consented for the study. This report details the outcomes of those patients who were not randomized because they failed to meet the inclusion/exclusion criteria of a TAVI study and were referred to an alternative treatment strategy.
Screening included an interview, clinical examination, ECG, laboratory assessment, and 2D echocardiography including continuous wave Doppler examination of the aortic valve. The aortic valve flow gradient and valve area were calculated according to the recommendations of the American Society of Echocardiography.6 In addition, all had left and right cardiac catheterization and assessment of the peripheral vascular tree by angiography and computer tomography. STS score and the logistic EuroSCORE were calculated using the Web-based systems (http://22.214.171.124/STSWebRiskCalc261/de.aspx and http://www.euroscore.org, respectively). All of the patient data were prospectively entered into a dedicated database.
The study population was divided in 2 groups. Group 1 (medical group) included those unsuitable for surgical AVR or randomization in a TAVI trial. They were treated medically with or without balloon aortic valvuloplasty (BAV). Group 2 (surgical group) included those who underwent operative AVR after exclusion from a TAVI trial. The treatment decision was made by a multidisciplinary team (interventional cardiologist, cardiac surgeon, and primary cardiologist). For patients deemed inoperable, BAV was pursued if the patient had critical AS with deteriorating symptoms or for evaluation of the reversibility of pulmonary hypertension, left ventricular dysfunction, or mitral regurgitation. All of the patients were followed by telephone contact or office visit.
Statistical analysis was performed using SAS version 9.1 (SAS Institute, Inc). Continuous variables are presented as mean±SD and categorical variables as proportions and percentages. Follow-up days are presented as medians (25th, 75th percentiles). Differences among continuous variables were assessed by using Student t test. Categorical variables were compared using the χ2 test or Fisher exact test, as indicated. Significance was set at P<0.05. The correlates with clinical electrocardiographic, echocardiographic, and hemodynamic variables on mortality were studied with Cox proportional hazards regression analysis. To determine independent correlates of mortality, we initially performed a Cox proportional hazard regression univariable analysis using all of the variables recorded in the categories of clinical electrocardiographic, echocardiographic, and hemodynamic variables. In the medical/BAV group, all of the univariable correlates of morality with a P value ≤0.05 were then used in a stepwise multivariable Cox regression model. Cumulative survival curves were constructed using the Kaplan–Meier method, and the significance was assessed using the log-rank statistic. The hospital stay was calculated as the time from surgery to discharge or to death.
From April 2007 to July 2009, 469 patients with severe AS were referred for participation in a TAVI trial in our center. A total of 362 patients (77.1%) were screened but not randomized for the trial. The main exclusion criteria were low STS score <10% in 72 (20%), peripheral vascular or aorta disease in 58 (16%), aortic valve area >0.8 cm2 in 54 (15%), significant coronary artery disease requiring revascularization in 43 (12%), and renal failure in 25 (7%). Of these 362 patients, 75 (20.7%) had 2 exclusion criteria and 26 (7.1%) had 3 exclusion criteria causes (Figure 1). The 362 patients were divided into 2 groups according to treatment undertaken. Group 1 (medical group) included 274 medically treated patients (75.7%). Of those 274 patients, 177 (64.6%) had BAV after the initial evaluation. Group 2 (surgical group) included 88 patients (24.3%).
Baseline characteristics are displayed in Table 1. The surgical group had less comorbidities compared with the medical group, as represented by lower STS and logistical EuroSCOREs. A similar pattern for more favorable laboratory, echocardiographic, and hemodynamic values in the surgical group is shown in Table 2. Particularly significant are the greater degree of pulmonary hypertension in the medically managed group and the larger cardiac output in the surgically treated patients.
Figure 2 shows Kaplan–Meier survival curves for the 2 groups. The median follow-up was 378 days (range: 164 to 633) in the medically treated group and 386 days (range: 169 to 641) in the surgical group. The mortality rate in the surgical group was significantly lower compared with the medical/BAV group at 6 months, 1 year, and 2 years: 16.9% versus 31.8%, 22.9% versus 39.6%, and 28.1% versus 53.4%, respectively (P<0.001). Three patients (3.4%) died during surgery. The 30-day mortality rate in the surgical group was 12.5% (n=11), and in-hospital mortality was 17% (n=15; 4 patients died in hospital after 30 days of hospitalization). The median time from surgery to death was 12 days (range: 1 to 72 days). The median time from screening to death in the medical group was 65 days (range: 24 to 180 days).
Postoperative hospital stay was a median of 12 days (range: 8 to 19 days) in those who underwent operative treatment. Of the 73 patients discharged from the hospital, 13 (17.8%) were discharged to nursing care facilities because they were debilitated and dependent on a tracheotomy or a gastrostomy. Five patients (6.8%) had had a major cerebrovascular event.
Cause of Death
Table 3 details the causes of death in both groups. The cause of death was known to be cardiac related in 43.1% of the medically managed patients. In an additional 20.6%, the cause of death is unknown. Thus, cardiac death could account for nearly two thirds of the total. In the surgical group, 9 (47.3%) of the 19 deaths were cardiac.
Factors associated with mortality in the medically managed patients are presented in Table 4. After multivariable adjustment, the strongest correlates for mortality were renal failure (hazard ratio [HR]: 5.88; P<0.001), New York Heart Association class IV (HR: 4.16; P=0.001), and aortic systolic blood pressure (HR: 0.98; P=0.01). Age was not associated with mortality in the univariable and multivariable analyses. In the surgical group, renal failure (HR: 7.45 [95% CI: 2.28 to 24.4]; P=0.001), STS score (HR: 1.09 [95% CI: 1.03 to 1.16]; P=0.003), and logistic EuroSCORE (HR: 1.45 [95% CI: 1.15 to 1.81]; P=0.001) were associated with mortality in the univariable analysis (Table 5).
Our study results confirm an exceedingly poor prognosis for patients with symptomatic severe AS who are not candidates for surgery. Even the best in contemporary medical management with the addition of judiciously applied BAV, 37.2% had died by 1 year. Our findings agree with those of Gardin et al,7 who reported that, among symptomatic patients with moderate-to-severe AS treated medically, the mortality rates after onset of symptoms were ≈25% at 1 year and 50% at 2 years. Furthermore, Shareghi et al8 reported 1-, 2-, and 3-year mortality rates of 44%, 62%, and 71%, respectively, in inoperable AS patients after BAV.
In patients carefully selected for operative valve replacement, the 1-year mortality rate was 21.5%, which was lower than that in the medically managed patients. This difference is not surprising, because patients selected for “high-risk” surgery as defined by increased STS and EuroSCORE had fewer comorbidities than those who were medically managed.
Assessment of Operative Risk
TAVI is intended for use in symptomatic patients with severe calcific AS who have a prohibitive risk for open chest surgery because of comorbid conditions. Defining surgical outcomes is not straightforward, because different models have limitations in predicting risk. An STS risk score >10 and a logistic EuroSCORE >20 are most often used to define high risk. The EuroSCORE, based on a large patient database drawn across Europe, has been developed to predict hospital mortality after adult cardiac surgery.9 It is a well-established and validated tool to estimate the risk of in-hospital death of individual patients undergoing cardiac surgery.10 It has been shown to predict early11 and long-term mortality after heart valve surgery.12 Unfortunately, this algorithm has been shown to persistently overestimate mortality rate.
The STS score is derived from a voluntary registry of practice outcomes and estimates the risks of mortality, morbidity, renal failure, and length of stay after valvular and nonvalvular cardiac surgery.13 This score has been shown to underestimate the true morality rate after cardiac surgery, but it more closely reflects the operative- and 30-day mortalities for the highest-risk patients having AVR.14 The predicted mortality in the surgical AVR group was 8.5% by STS score and 24.4% by logistic EuroSCORE. The actual operative mortality rate (3.4%) and total in-hospital mortality rate (17.4%) are higher than the 30-day mortality rate predicted by STS score. Many of the high-risk features not included in the risk algorithms, such as porcelain aorta, frailty, cirrhosis, radiation chest wall, and chest wall deformities, can explain the higher 30-day mortality.
It is important to remember that STS score and EuroSCORE algorithms were constructed based on outcomes after surgery. These scores, therefore, have limited applicability to patients who are not surgical candidates. In our surgical group, both high STS and EuroSCOREs were associated with a greater operative mortality. Neither was, however, the strongest correlate of mortality in either population. Renal failure was the strongest correlate of outcome in the surgical population.
Correlates of Mortality in Nonoperative Patients
Because neither STS nor EuroSCORE is readily applicable to the medically treated group, we undertook a detailed analysis of their clinical features at baseline and their relation to prognosis (Table 4). Functional status and hemodynamics are well-known determinants of prognosis in patients with severe AS who are treated medically,15,16 and renal failure is reported to be a significant correlate of worse outcome in high-risk patients undergoing AVR surgery.12,17,18 Our results are congruent with the following observations: renal failure was the correlate for mortality in both operative and nonoperative groups, and in the medically managed group, New York Heart Association class and systolic aortic pressure were additional independent correlates of mortality. In the surgical group, previous bypass, STS score, and EuroSCORE were associated with mortality.
Age was not associated with higher mortality in either the medical/BAV group or in the surgical group. Nearly all of the patients were older, with an average age near 80 years. Thus, it is likely that the lack of an age spread and the influence of comorbid conditions obscured any effect of age. The notion that comorbid processes are more important than age alone in the prognosis of these patients is supported by the relatively good outcomes reported recently for surgical AVR in very elderly patients (>80 years).19–21
Quality of Life Issues
In an elderly population, mortality risk is not the only consideration. It is clear from our data that quality of life considerations are important. One fifth of the patients who were discharged alive after operation experienced a reduced quality of life, and some lost their independence. Moreover, even successful valve replacement does not prevent mortality attributable to the patient’s comorbid conditions. Mortality was attributed to a noncardiac cause in 36.2% of the unoperated group and 52.6% of the surgical group. In the surgical group the noncardiac deaths were mostly related to the surgery and its complications.
Using the Edward-Sapien valve, Webb et al22 reported their results in 168 patients. Of those 168 patients, 113 with transfemoral approach had a 1-year mortality rate of 26%. In the Canadian multicenter TAVI program,23 the mortality at 1 year with the transfemoral approach (n=162) was 25% and with the transapical approach (n=177) was 22%. The pooled analysis of Registry of EndoVascular Implantation of Valves in Europe (REVIVE); tRans-catheter EndoVascular Implantation of VALves (REVIVAL); and Placement of AoRTic TraNscathetER Valves Trial Europe (PARTNER EU)24 showed similar results with the transfemoral approach (n=222) with 1-year mortality at 25%. The 1-year mortality rate with the transapical approach (n=281) was 42%, but the risk profile of the patients included in the transapical group was higher compared with the transfemoral group. In our cohort, the 1-year mortality rate in the surgical group was comparable to other TAVI trials; however, those patients were at much lower risk compared with the medical or TAVI patients and still had high rates of complications. The potential benefits of TAVI in complex patients are supported by encouraging preliminary data; its future will ultimately be defined in the ongoing largest randomized multicenter trial, Placement of Aortic Transcatheter Valves, which is enrolling in North America and Canada. In cohort A, patients are randomized between TAVI and surgical AVR, and in cohort B patients who are not suitable surgical candidates are randomized to medical therapy versus TAVI.
This study confirms that patients with severe symptomatic AS who are not included in the TAVI trials have exceptionally poor 1-year prognoses. These patients are exposed to high mortality rates irrespective of the treatment modality, medical or surgical. Importantly, quality of life was often compromised by the morbidity associated with the surgical procedure. Clinical factors such as renal failure, New York Heart Association class, and decreased blood pressure are more predictive of mortality than is the risk score model in medically managed patients. Renal dysfunction was the strongest correlate of outcome in the surgically treated group. It remains to be seen whether this patient population may benefit from TAVI with respect to overall mortality rates and quality of life.
Presented at the 2009 American Heart Association meeting in Orlando, Fla, November 14–18, 2009.
Bonow RO, Carabello BA, Chatterjee K, de Leon AC Jr, Faxon DP, Freed MD, Gaasch WH, Lytle BW, Nishimura RA, O'Gara PT, O'Rourke RA, Otto CM, Shah PM, Shanewise JS, for the American College of Cardiology/American Heart Association Task Force on Practice Guidelines. 2008 focused update incorporated into the ACC/AHA 2006 guidelines for the management of patients with valvular heart disease: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (writing committee to revise the 1998 guidelines for the management of patients with valvular heart disease). Endorsed by the Society of Cardiovascular Anesthesiologists, Society for Cardiovascular Angiography and Interventions, and Society of Thoracic Surgeons. J Am Coll Cardiol. 2008; 52: e1–e142.
Rosenhek R, Binder T, Porenta G, Lang I, Christ G, Schemper M, Maurer G, Baumgartner H. Predictors of outcome in severe, asymptomatic aortic stenosis. N Engl J Med. 2000; 31: 343: 611–617.
Iung B, Baron G, Butchart EG, Delahaye F, Gohlke-Bärwolf C, Levang OW, Tornos P, Vanoverschelde JL, Vermeer F, Boersma E, Ravaud P, Vahanian AA. prospective survey of patients with valvular heart disease in Europe: the Euro Heart Survey on Valvular Heart Disease. Eur Heart J. 2003; 24: 1231–1243.
Bouma BJ, van Den Brink RB, van Der Meulen JH, Verheul HA, Cheriex EC, Hamer HP, Dekker E, Lie KI, Tijssen JG. To operate or not on elderly patients with aortic stenosis: the decision and its consequences. Heart. 1999; 82: 143–148.
Schiller NB, Shah PM, Crawford M, DeMaria A, Devereux R, Feigenbaum H, Gutgesell H, Reichek N, Sahn D, Schnittger I. Recommendations for quantitation of the left ventricle by two-dimensional echocardiography. American Society of Echocardiography Committee on Standards, Subcommittee on Quantitation of Two-Dimensional Echocardiograms. J Am Soc Echocardiogr. 1989; 2: 358–367.
Roques F, Nashef SA, Michel P, Gauducheau E, de Vincentiis C, Baudet E, Cortina J, David M, Faichney A, Gabrielle F, Gams E, Harjula A, Jones MT, Pintor PP, Salamon R, Thulin L. Risk factors and outcome in European cardiac surgery: analysis of the EuroSCORE multinational database of 19030 patients. Eur J Cardiothorac Surg. 1999; 15: 816–822.
Nashef SA, Roques F, Hammill BG, Peterson ED, Michel P, Grover FL, Wyse RK, Ferguson TB, for the EuroSCORE Project Group. Validation of European System for Cardiac Operative Risk Evaluation (EuroSCORE) in North American cardiac surgery. Eur J Cardiothorac Surg. 2002; 22: 101–105.
Pinna-Pintor P, Bobbio M, Colangelo S, Veglia F, Giammaria M, Cuni D, Maisano F, Alfieri O. Inaccuracy of four coronary surgery risk-adjusted models to predict mortality in individual patients. Eur J Cardiothorac Surg. 2002; 21: 199–204.
Otto CM, Burwash IG, Legget ME, Munt BI, Fujioka M, Healy NL, Kraft CD, Miyake-Hull CY, Schwaegler RG. Prospective study of asymptomatic valvular aortic stenosis: clinical, echocardiographic, and exercise predictors of outcome. Circulation. 1997; 95: 2262–2270.
Monin JL, Lancellotti P, Monchi M, Lim P, Weiss E, Piérard L, Guéret P. Risk score for predicting outcome in patients with asymptomatic aortic stenosis. Circulation. 2009; 120: 69–75.
Kolh P, Kerzmann A, Honore C, Comte L, Limet R. Aortic valve surgery in octogenarians: predictive factors for operative and long-term results. Eur J Cardiothorac Surg. 2007; 31: 600–606.
Webb JG, Altwegg L, Boone RH, Cheung A, Ye J, Lichtenstein S, Lee M, Masson JB, Thompson C, Moss R, Carere R, Munt B, Nietlispach F, Humphries K. Transcatheter aortic valve implantation: impact on clinical and valve-related outcomes. Circulation. 2009; 119: 3009–3016.
Rodés-Cabau J, for the Canadian Transcatheter Aortic Valve Implantation Program. Acute and late follow-up from a multicenter consecutive series of patients with inoperable or high-surgical risk aortic stenosis. Presented at the 21st Annual Scientific Sessions Transcatheter Cardiovascular Therapeutics, September 21–25, 2009; San Francisco, CA.
Leon MB. A meta-analysis of all available Edwards Sapien TAVI clinical results: focus on risk stratification, learning curve issues, and echo results. Presented at the 21st Annual Scientific Sessions Transcatheter Cardiovascular Therapeutics, September 21–25, 2009; San Francisco, CA.