Long-Term Outcomes Following Surgical Aortic Bioprosthesis Implantation
Original Investigation
Central Illustration
Abstract
Background:
Few data exist on long-term outcomes and structural valve degeneration (SVD) in consecutive unselected patients undergoing surgical aortic valve replacement (SAVR).
Objectives:
The goal of this study was to determine the long-term outcomes of a contemporary cohort of consecutive unselected SAVR recipients with a focus on evaluating clinical outcomes and SVD based on echocardiographic criteria.
Methods:
A total of 672 consecutive patients (mean age: 72 ± 8 years; 61.5% male) undergoing SAVR with a bioprosthesis between 2002 and 2004 were included. Baseline and follow-up data were prospectively collected in a dedicated database. Baseline post-operative echocardiographic data were obtained in the 624 patients alive at hospital discharge and in 209 patients at 10 years (87% of the patients at risk). SVD was defined as subclinical (increase >10 mm Hg in mean transvalvular gradient + decrease >0.3 cm2 in valve area and/or new-onset mild or moderate aortic regurgitation) and clinically relevant (increase >20 mm Hg in mean transvalvular gradient + decrease >0.6 cm2 in valve area and/or new-onset moderate-to-severe aortic regurgitation).
Results:
At a median follow-up of 10 years (interquartile range: 5 to 13 years), 432 patients (64.3%) had died. Older age, left ventricular dysfunction, atrial fibrillation, chronic obstructive pulmonary disease, greater body mass index, and diabetes mellitus were associated with an increased mortality risk (p < 0.05 for all). Clinically relevant SVD occurred in 6.6% of patients; 30.1% of patients had subclinical SVD. A greater body mass index and the use of a specific aortic bioprosthesis were independently associated with clinically relevant SVD (p < 0.05 for both), and 83% of these patients underwent aortic valve reintervention (valve-in-valve transcatheter aortic valve replacement in 44% of them).
Conclusions:
The 10-year mortality rate in elderly SAVR recipients of a bioprosthetic valve was considerable, chiefly determined by their older age and the presence of comorbidities. Clinically relevant SVD was infrequent, but close to one third of the population exhibited subclinical SVD. These results provide contemporary data on long-term clinical outcomes and SVD post-SAVR, and they should be taken into consideration when evaluating late clinical outcomes and valve durability after transcatheter aortic valve replacement.
Introduction
There has been a significant increase in the number of aortic valve replacement procedures in recent years, mainly in elderly patients, coinciding with the use of bioprosthetic valves (1). Currently, bioprostheses represent approximately 80% of the valves used in surgical aortic valve replacement (SAVR) (2). Increasing age, the quest to avoid chronic anticoagulation therapy, and novel utilization of less invasive therapies (valve-in-valve transcatheter aortic valve replacement [TAVR]) for treating bioprosthesis failure are potential reasons that have tipped the balance in favor of bioprosthetic valves. In fact, the 2017 American College of Cardiology/American Heart Association guidelines have already extended the indication for a bioprosthetic valve to younger patients (3). However, structural valve degeneration (SVD) limiting valve durability represents one of the major drawbacks of biological tissue.
SVD is a multifactorial process characterized by progressive leaflet degeneration that leads to valve dysfunction (stenosis and/or regurgitation). The reported durability of surgical bioprostheses at 10-year follow-up is >85% (4–7). Nonetheless, studies in SAVR patients often equate SVD with “reoperation” instead of valve performance to define valve durability, leading to a potential underestimation of the real incidence of SVD. Furthermore, to this point, there are few data on consecutive patients who underwent SAVR because most studies focused on a specific valve model instead of consecutive patients, without consensus on patients’ follow-up or SVD definition, hindering comparison between studies. This limitation is of particular importance in an era in which TAVR has emerged as an alternative to SAVR for treating intermediate-to-high risk patients, with multiple ongoing studies in lower-risk populations (8,9). In the near future, valve durability may become the most important criterion for selecting the type of valve replacement; thus, obtaining long-term contemporary data on consecutive unselected patients has a high clinical relevance.
The present study evaluated the long-term (≥10-year) outcomes of a contemporary cohort of consecutive unselected patients who underwent SAVR, with a focus on clinical outcomes (global and cardiovascular mortality) and SVD based on clinical and echocardiographic criteria.
Methods
Between January 2002 and December 2004, a total of 985 consecutive patients underwent SAVR in our center. Patients were considered eligible for the study if they received a bioprosthetic aortic valve either isolated or associated with other procedures such as coronary artery bypass graft (CABG), maze, aortoplasty, and/or replacement of the ascending aorta. Patients who received a mechanical valve (n = 236), homograft (n = 21), or those who had undergone replacement of multiple valves (n = 56) were excluded. This method led to a final study population of 672 patients.
The selection of a bioprosthetic (vs. mechanical) valve and valve model were determined by the surgeon responsible for the operation. Clinical, procedural, and echocardiographic data were prospectively gathered within a dedicated database. Clinical follow-up was undertaken during clinical visits or telephone contact. Clinical events were also prospectively recorded and linked to the public health, provincial vital statistics, and administrative hospital re-admission data (Régie de l’assurance maladie du Québec). The median follow-up was 10 years (interquartile range: 5 to 13 years), and no patient was lost to follow-up.
Echocardiographic assessment and SVD definitions
Comprehensive transthoracic echocardiograms (TTEs) were performed in all patients before hospital discharge (baseline post-SAVR TTE). At 10-year follow-up, echocardiographic evaluation was available in 209 patients (87% of the population at risk).
All TTE examinations were conducted according to American Society of Echocardiography guidelines (10). The mean transprosthetic gradient was calculated by using the modified Bernoulli formula. Absolute change in mean gradient was calculated as the gradient at follow-up minus the gradient at baseline post-operative echocardiogram. The effective orifice area (EOA) of the prosthesis was calculated by using the continuity equation. Color Doppler examination in multiple windows was used to distinguish between intraprosthetic versus paraprosthetic valve regurgitation. The severity of prosthetic valve regurgitation was assessed by using a multiparameter integrative approach (10,11). When necessary, a transesophageal echocardiography examination was performed to ensure an accurate evaluation of the localization and severity of prosthetic valve regurgitation.
Given that SVD is a progressive process with gradual changes in hemodynamic valve performance and severity of valve deterioration over time, we proposed to divide SVD into 2 stages: 1) subclinical SVD; and 2) clinically relevant SVD (12). The definition for subclinical SVD was based on the following echocardiographic criteria: 1) an increase in mean transvalvular gradient of >10 mm Hg with a concomitant decrease in EOA >0.3 cm2 (and/or decrease in Doppler velocity index >0.08), and/or new onset of at least mild intraprosthetic regurgitation or increase by at least 1 grade of pre-existent intraprosthetic valve regurgitation compared with the baseline post-operative echocardiogram, with the resulting regurgitation grade inferior or equal to moderate; and 2) changes in morphology (i.e., thickening, calcification, flail, pannus) and/or mobility (i.e., reduced, avulsed) of the bioprosthetic valve leaflets compared with the baseline post-operative echocardiogram. Clinically relevant SVD was defined as an increase in mean transvalvular gradient >20 mm Hg with a concomitant decrease in the EOA >0.6 cm2 (and/or decrease in Doppler velocity index >0.15), leading to severe aortic stenosis according to current guidelines (13,14), and/or new occurrence or increase of at least 1 grade of intraprosthetic regurgitation leading to moderate-to-severe aortic regurgitation.
Statistical analysis
Categorical variables are reported as number (percent) and continuous variables as mean ± SD or median (25th to 75th interquartile range), depending on variable distribution. Group comparisons were analyzed with the Student’s t-test or Wilcoxon rank sum test for continuous variables and the chi-square test or Fisher exact test for categorical variables. Univariable and multivariable Cox proportional hazard models were used to identify the factors associated with mortality, cardiac mortality, and SVD. The variables exhibiting a p value <0.05 in the univariable analysis were included in the multivariable model. Changes in mean transaortic gradient and aortic valve area measurements over time (discharge, 10-year) were evaluated with repeated measures analyses of variance. A p value <0.05 was considered statistically significant. Data were analyzed by using SAS version 9.3 (SAS Institute, Inc., Cary, North Carolina).
Results
The main baseline, procedural characteristics, and 30-day outcomes of the study population are shown in Table 1. The mean age of the study population was 72 ± 8 years, and 61.5% were male. The mean Society of Thoracic Surgeons Predicted Risk of Mortality score was 3.7 ± 3.5%. The 5 models of surgical aortic prostheses used were as follows: Medtronic Mosaic (Medtronic, Dublin, Ireland) (34.8%), CE-Pericardial Magna (Edwards Lifesciences, Irvine, California) (34.2%), CE-Perimount (Edwards Lifesciences) (16.1%), Medtronic Freestyle (Medtronic, Minneapolis, Minnesota) (7.7%), and Mitroflow (LinaNova, London, United Kingdom) (7.1%). Isolated SAVR was performed in 281 patients (41.8%), whereas 391 patients (58.2%) underwent concomitant procedures (mainly CABG) at the time of SAVR. The 30-day mortality rate was 7.1% (4.3% among those patients who underwent isolated aortic valve replacement).
| Clinical characteristics | |
| Age, yrs | 72 ± 8 |
| BMI, kg/m2 | 28 ± 5 |
| Male | 413 (61.5) |
| Diabetes mellitus | 161 (23.9) |
| Hypertension | 404 (60.1) |
| Dyslipidemia | 495 (73.6) |
| COPD | 116 (17.2) |
| NYHA functional class | |
| I–II | 308 (45.8) |
| III–IV | 363 (54.0) |
| eGFR <60 ml/min | 279 (41.5) |
| Coronary artery disease | 362 (53.8) |
| Previous valve surgery | 36 (5.4) |
| Previous stroke | 49 (7.3) |
| STS PROM score, % | 3.7 ± 3.5 |
| Baseline echocardiography | |
| LVEF, % | 57 ± 14 |
| LVEF <50 | 107 (15.9) |
| Mean gradient, mm Hg | 39.8 ± 16.8 |
| Aortic valve area, cm2 | 0.80 ± 0.29 |
| Aortic regurgitation | |
| None/trace | 297 (58.9) |
| Mild | 138 (27.4) |
| Moderate/severe | 69 (13.7) |
| Procedural variables | |
| Prosthesis type | |
| CE-Perimount | 108 (16.1) |
| CE-Pericardial Magna | 230 (34.2) |
| Medtronic Freestyle | 52 (7.7) |
| Medtronic Mosaic | 234 (34.8) |
| Mitroflow | 48 (7.1) |
| Valve size | |
| 19 mm | 48 (7.1) |
| 21 mm | 160 (23.8) |
| 23 mm | 204 (30.3) |
| 25 mm | 181 (26.9) |
| 27 mm | 65 (9.7) |
| 29 mm | 14 (2.1) |
| Concomitant procedures | 391 (58.2) |
| SAVR + CABG | 298 (44.4) |
| SAVR + CABG + other∗ | 49 (7.2) |
| SAVR + other∗ | 44 (6.6) |
| In-hospital outcomes | |
| Death | 48 (7.1) |
| Stroke | 30 (4.5) |
| Sepsis | 12 (1.8) |
| Atrial fibrillation | 357 (53.1) |
| Renal failure | 48 (7.1) |
| Dialysis | 25 (3.7) |
| Pacemaker | 17 (2.5) |
| Discharge echocardiography | |
| LVEF, % | 55 ± 13 |
| Mean gradient, mm Hg | 14.5 ± 6.1 |
| Aortic valve area, cm2 | 1.4 ± 0.4 |
| Aortic regurgitation | |
| None/trace | 516 (93.7) |
| Mild | 31 (5.6) |
| Moderate/severe | 4 (0.7) |
| Moderate prosthesis–patient mismatch | 230 (42.1) |
| Severe prosthesis–patient mismatch | 127 (23.3) |
Valve performance was evaluated through a TTE in all patients alive at hospital discharge (n = 624). Post-operative baseline echocardiographic data revealed a mean gradient of 14.5 ± 6.1 mm Hg and a mean EOA of 1.4 ± 0.4 cm2. Severe prosthesis–patient mismatch was observed in 127 patients (23.3%). Aortic regurgitation was none or trace in 516 patients (93.7%), whereas mild and moderate/severe aortic regurgitation were observed in 31 patients (5.6%) and 4 patients (0.7%), respectively.
Late clinical outcomes
Late clinical outcomes of the study population are summarized in Table 2. At a median follow-up of 10 years (interquartile range: 5 to 13 years), all-cause (n = 432) and cardiovascular (n = 151) mortality rates were 64.3% and 22.5%, respectively. The rates of late (>30 days’ post-SAVR) death and cardiac death were 57.1% and 15.6%, respectively. Reoperation was performed in 69 patients (10.3%) during the follow-up period, and 49 of them underwent redo aortic valve replacement. Reasons for redo aortic valve replacement were as follows: early nonstructural valve dysfunction (n = 3), SVD (n = 36), endocarditis (n = 7), valve thrombosis (n = 2), and aortic dissection (n = 1). The mean time to reoperation was 8 ± 5 years. The Kaplan-Meier curves at 10-year follow-up for global mortality, cardiovascular mortality, reintervention, and mortality + reintervention are shown in Figure 1.
| Death | 432 (64.3) |
| Cardiac death | 151 (22.5) |
| Late death | 384 (57.1) |
| Late cardiac death | 105 (15.6) |
| Reintervention | 69 (10.3) |
| Aortic valve replacement | 55 (79.7) |
| Redo SAVR | 33 (47.8) |
| Redo TAVR | 16 (23.2) |
| Redo MVR | 5 (7.2) |
| Redo TVR | 1 (1.4) |
| CABG | 5 (7.3) |
| Repl Asc Ao | 6 (8.7) |
| Other | 3 (4.3) |
| Rehospitalization | 444 (66.1) |
| Rehospitalization for cardiac reasons | 235 (34.9) |
Kaplan-Meier Estimates for Clinical Events at 10-Year Follow-up
(A) Freedom from death. (B) Freedom from cardiac death. (C) Freedom from redo aortic valve replacement. (D) Freedom from death + aortic valve replacement.
The factors associated with cumulative global and cardiovascular death are shown in Table 3. Older age (hazard ratio [HR]: 1.08 for each increase of 1 year; 95% confidence interval [CI]: 1.06 to 1.09; p < 0.001), diabetes mellitus (HR: 1.72; 95% CI: 1.35 to 2.18; p < 0.001), chronic obstructive pulmonary disease (COPD) (HR: 2.17; 95% CI: 1.67 to 2.81; p < 0.001), BMI (HR: 1.02 for each increase of 1 kg/m2; 95% CI: 1.00 to 1.05; p = 0.037), atrial fibrillation (HR: 1.44; 95% CI: 1.06 to 1.96; p = 0.028), and left ventricular dysfunction at baseline (HR: 1.45 for each decrease of 5% in left ventricular ejection fraction; 95% CI: 1.16 to 1.81; p < 0.001) were independently associated with an increased risk of all-cause death. Older age (HR: 1.06 for each increase of 1 year; 95% CI: 1.03 to 1.09; p < 0.001), COPD (HR: 2.74; 95% CI: 1.78 to 4.20; p < 0.001), previous CABG (HR: 2.50; 95% CI: 1.42 to 4.41; p < 0.001), diabetes mellitus (HR: 1.56; 95% CI: 1.03 to 2.38; p = 0.036), previous stroke (HR: 1.91; 95% CI: 1.02 to 3.58; p = 0.039), and larger body mass index (BMI) (HR: 1.05 for each increase of 1 kg/m2; 95% CI: 1.02 to 1.09; p = 0.004) were associated with an increased risk of cardiovascular mortality.
| Univariate | Multivariate | |||
|---|---|---|---|---|
| HR (95% CI) | p Value | HR (95% CI) | p Value | |
| Global mortality | ||||
| Age, yrs | 1.06 (1.05–1.08) | <0.001 | 1.08 (1.06–1.09) | <0.001 |
| BMI, kg/m2 | 1.03 (1.01–1.05) | 0.005 | 1.02 (1.00–1.05) | 0.037 |
| Diabetes mellitus | 1.57 (1.27–1.93) | <0.001 | 1.72 (1.35–2.18) | <0.001 |
| Hypertension | 1.31 (1.07–1.59) | 0.008 | ||
| Dyslipidemia | 0.79 (0.64–0.97) | 0.026 | ||
| COPD | 2.43 (1.94–3.04) | <0.001 | 2.17 (1.67–2.81) | <0.001 |
| NYHA functional class III–IV | 1.27 (1.05–1.54) | 0.015 | ||
| eGFR <60 ml/min | 1.48 (1.23–1.79) | <0.001 | ||
| Coronary artery disease | 1.41 (1.16–1.70) | <0.001 | ||
| Previous CABG | 1.67 (1.18–2.37) | 0.004 | ||
| Atrial fibrillation | 1.85 (1.43–2.39) | <0.001 | 1.44 (1.06–1.96) | 0.028 |
| Previous stroke | 1.51 (1.08–2.12) | 0.016 | ||
| STS PROM score, % | 1.10 (1.04–1.17) | <0.001 | ||
| LVEF <50% | 1.45 (1.20–1.76) | <0.001 | 1.45 (1.16–1.81) | 0.001 |
| Valve size ≤23 mm | 1.37 (1.12–1.67) | 0.002 | ||
| Moderate/severe prosthesis–patient mismatch∗ | 1.43 (1.13–1.81) | 0.003 | ||
| Cardiovascular mortality | ||||
| Age, yrs | 1.06 (1.03–1.08) | <0.001 | 1.06 (1.03–1.09) | <0.001 |
| BMI, kg/m2 | 1.05 (1.03–1.08) | <0.001 | 1.05 (1.02–1.09) | 0.004 |
| Diabetes mellitus | 1.52 (1.07–2.16) | 0.028 | 1.56 (1.03–2.38) | 0.036 |
| Hypertension | 1.60 (1.13–2.26) | 0.008 | ||
| COPD | 2.38 (1.65–3.44) | <0.001 | 2.74 (1.78–4.20) | <0.001 |
| NYHA functional class III–IV | 1.55 (1.12–2.17) | 0.010 | ||
| eGFR <60 ml/min | 1.96 (1.42–2.70) | <0.001 | ||
| Coronary artery disease | 1.49 (1.08–2.07) | 0.017 | ||
| Previous CABG | 3.14 (1.99–4.94) | <0.001 | 2.50 (1.42–4.41) | <0.001 |
| Atrial fibrillation | 2.12 (1.43–3.18) | <0.001 | ||
| Previous stroke | 1.92 (1.16–3.19) | 0.011 | 1.91 (1.02–3.58) | 0.039 |
| STS PROM score, % | 1.16 (1.08–1.26) | <0.001 | ||
| LVEF <50 | 1.38 (1.00–1.92) | 0.047 | ||
| Mean gradient, mm Hg | 0.99 (0.98–1.00) | 0.009 | ||
| Valve size ≤23 mm | 1.56 (1.11–2.21) | 0.011 | ||
Structural valve degeneration
TTE examinations were performed at a median follow-up of 10 years (range: 10 to 15 years) in 209 patients (87% of the population at risk). The changes in mean transprosthetic gradient and EOA over time are shown in Figure 2. Overall, there was a mild but significant increase in mean transprosthetic gradient (20.8 ± 14.3 mm Hg, mean increase of 6.6 ± 1.0 mm Hg vs. baseline post-SAVR TTE [p < 0.001]) and a decrease in EOA (1.1 ± 0.4 cm2, mean decrease of 0.3 ± 0.04 cm2 vs. baseline post-SAVR TTE [p < 0.001]), respectively, at 10-year follow-up.
Hemodynamic Bioprotheses Performance at 10-Year Follow-up
Overall changes in mean gradient (solid line) and effective orifice area (dashed line) between discharge and last follow-up.
Throughout the study period, 41 patients developed clinically relevant SVD (6.6% among survivors at hospital discharge), with a mean transvalvular gradient increase of 26 ± 3 mm Hg and EOA mean decrease of 0.7 ± 0.1 cm2 (Figure 3). Stenosis was the leading cause of clinically relevant SVD in 24 patients (mean transvalvular gradient increase and EOA decrease of 36 ± 4 mm Hg and 1.1 ± 0.1 cm2, respectively). Fifteen patients developed severe intraprosthetic aortic regurgitation (Figure 4), and 2 patients had mixed valve dysfunction with severe aortic stenosis and regurgitation. Thirty-four of the 41 patients with clinically relevant SVD (82.9%) underwent reintervention; 19 patients had redo SAVR and 15 patients received a valve-in-valve TAVR procedure. Seven patients with clinically relevant SVD were deemed inoperable. The mean time to clinically relevant SVD was 9.8 ± 3 years. The Kaplan-Meier curve at 10-year follow-up for clinically relevant SVD is shown in Figure 5. A Kaplan-Meier survival curve comparing the study population and an age- and sex-matched general population is shown in Online Figure 1.
Hemodynamic Changes in Patients With SVD
(A) Changes in mean gradient (solid line) and effective orifice area (dashed line) between discharge and follow-up in patients with subclinical structural valve degeneration (SVD). (B) Changes in mean gradient (solid line) and effective orifice area (dashed line) between discharge and follow-up in patients with clinically relevant SVD.
Presence and Severity of AR After SAVR
(A) Overall (discharge and 10-year follow-up). (B) Subclinical SVD (discharge and 10-year follow-up). (C) Clinically relevant SVD (discharge and 10 year follow-up). AR = aortic regurgitation; SAVR = surgical aortic valve replacement; SVD = structural valve degeneration.
Kaplan-Meier Curves for Clinically Relevant SVD at 10-Year Follow-up
Freedom from clinically relevant structural valve degeneration (SVD) at 10-year follow-up.
A total of 63 patients (30.1% among patients with echocardiographic follow-up at 10 years) had echocardiographic criteria of subclinical SVD (Figure 3). Fourteen patients showed stenosis as the main reason for subclinical SVD, with a mean transvalvular gradient increase of 14.8 ± 3.8 mm Hg along with a mean decrease in EOA of 0.5 ± 0.2 cm2 compared with baseline post-SAVR TTE (p < 0.001). Forty patients showed aortic regurgitation as the leading cause of SVD (Figure 4), and 9 patients developed a mixed disease (stenosis and regurgitation).
The factors associated with post-SAVR SVD are shown in Table 4. In a multivariable analysis, increasing age (HR: 1.09 for each increase of 1 year; 95% CI: 1.05 to 1.12; p < 0.001) is associated with subclinical SVD; and increasing BMI (HR: 1.06 for each increase of 1 kg/m2; 95% CI: 1.01 to 1.13; p = 0.03), and the implantation of a Mitroflow valve (HR: 3.13; 95% CI: 1.34 to 7.36; p = 0.009) were the factors associated with clinically relevant SVD.
| Univariate | Multivariate | |||
|---|---|---|---|---|
| HR (95% CI) | p Value | HR (95% CI) | p Value | |
| Subclinical SVD | ||||
| Age, yrs | 1.09 (1.05–1.13)∗ | <0.001 | 1.09 (1.05–1.12)∗ | <0.001 |
| COPD | 2.30 (1.17–4.54) | 0.016 | ||
| eGFR <60 ml/min | 1.82 (1.10–3.01) | 0.019 | ||
| Stentless prosthesis | 4.60 (1.12–18.9) | 0.034 | ||
| Valve size ≤23 mm | 1.58 (0.95–2.66) | 0.081 | ||
| Clinically relevant SVD | ||||
| BMI, kg/m2 | 1.09 (1.04–1.16)† | 0.001 | 1.06 (1.01–1.13)† | 0.03 |
| Diabetes mellitus | 2.06 (1.07–3.96) | 0.030 | ||
| Mitroflow valve | 4.38 (2.03–9.46) | <0.001 | 3.13 (1.34–7.36) | 0.009 |
| Valve size ≤23 mm | 1.85 (0.96–3.53) | 0.063 | ||
The rates of subclinical and clinically relevant SVD according to age, valve model, and valve size are shown in Figure 6. Clinically relevant SVD was more frequent in patients <65 years of age compared with patients between 65 and 75 years of age and patients >75 years of age (p < 0.01 for both age ranges). Conversely, subclinical SVD was more prevalent in patients >75 years of age compared with patients <65 years of age (p = 0.005). There were differences regarding valve model: the rate of clinically relevant SVD was significantly higher with Mitroflow compared with other valve models (p = 0.009). There were no differences in the rate of SVD according to valve size.
Rate of Clinically Relevant and Subclinical SVD
Rate of clinically relevant/subclinical structural valve degeneration (SVD) according to: (A) age; (B) valve model; and (C) valve size.
Discussion
This contemporary patient series reporting on the 10-year outcomes of 672 consecutive unselected patients undergoing SAVR with a bioprosthetic valve in the pre-TAVR era showed that close to two-thirds of valve recipients had died at the 10-year follow-up, with increasing age and cardiac (low left ventricular ejection fraction and atrial fibrillation) and noncardiac (larger BMI, diabetes, and COPD) comorbidities driving this heightened mortality risk (Central Illustration). Most deaths (65%) were not cardiovascular related. Clinically relevant SVD occurred in 6.6% of patients during the study period, whereas close to one-third of patients had subclinical hemodynamic changes at the 10-year follow-up echocardiogram consistent with subclinical SVD. A greater BMI and the use of the Mitroflow valve model were associated with an increased risk for SVD, and most (>80%) patients with clinically relevant SVD had a reintervention (TAVR in about one-half of them).
Late Outcomes After SAVR With a Bioprosthesis
Event-free survival curves at 10-year follow-up (upper panels). Rates of subclinical and clinically relevant structural valve degeneration (SVD) at 10-year follow-up. Changes in valve hemodynamics over time, overall, and in patients with subclinical and clinically relevant SVD (lower panels). AR = aortic regurgitation; SAVR = surgical aortic valve replacement.
Clinical outcomes
Classically, most surgical valve studies have reported outcomes based on a specific valve model instead of consecutive unselected patients, and this method may have led to some bias regarding clinical outcomes. It may have also partially prevented an appropriate comparison between studies due to several reasons: 1) population baseline characteristics are different in each study; 2) large variability exists in the time period of surgery, with many surgical series including patients with >10 years of difference regarding the timing of surgery (5,15), which may indeed entail differences in technique and experience of surgeons’ team; and 3) differences in the definition of outcomes (clinical and echocardiographic) between studies.
The 30-day mortality rate observed in our study (7.1%) seems to be slightly higher than that reported in SAVR studies including specific bioprosthetic valve models (7,16–21) but is similar to that reported by Goldstone et al. (22) and Barreto-Filho et al. (1) including 82,755,924 consecutive Medicare fee-for-service beneficiaries undergoing SAVR in the United States between 1999 and 2011. In their studies, the mortality rates of SAVR recipients during a time period similar to our study (2001 and 2002) were 7.3% and 7.1%, respectively. The fact that our study relates to the pre-TAVR era may have also influenced early and late outcomes. In fact, it is well known that the emergence of TAVR translated into a significant improvement in the results of SAVR (23), probably because the highest risk group of patients were increasingly referred for the less invasive TAVR approach. Also, the majority of patients had a concomitant procedure (mainly CABG), with a mortality rate <5% among those patients who had isolated SAVR.
Surgical studies have also shown a large variability in survival rates at 10-year follow-up, ranging from 37.4% (24) to 64% (25). Our study reported a high rate of late mortality post-SAVR (57.1%), which may be explained by the inclusion of an elderly group of patients with a significant comorbidity burden. In fact, long-term survival after SAVR is highly dependent on patients’ age and comorbid conditions (26). The 10-year survival rate of the present cohort (48%) is lower than that estimated for the general population of the province of Quebec for individuals of the same age and sex (62%). Similar to previous studies (4,5,24,27), advanced age and comorbidities such as diabetes mellitus, left ventricular dysfunction, COPD, and previous CABG were associated with an increased mortality risk, and they should be taken into consideration when evaluating the global risk-benefit of a SAVR procedure in a specific patient.
Structural valve degeneration
Most previous studies in the surgical valve field have equated SVD with the need for reoperation without any specific criteria to define SVD and/or the indication for reoperation (12). However, reoperation does not necessarily imply SVD and vice versa; many patients with SVD may be deemed inoperable because of high or prohibitive surgical risk, leading to an underestimation of the real incidence of SVD. Also, surgical studies evaluating valve durability have mainly focused on a specific valve model instead of consecutive unselected patients. It is well known that each model of aortic bioprosthesis may have a particular degenerative pattern (16–21), and this factor might result in a large variability in the rate of SVD between studies. Previous SAVR studies including a smaller cohort of patients and using echocardiographic criteria have reported rates of SVD at 10-year follow-up of approximately 20% (27,28).
Despite a mild increase in transvalvular gradients over time, our study offers encouraging global data regarding the performance of aortic bioprostheses at 10-year follow-up, with a mean transvalvular gradient of about 20 mm Hg (delta mean gradient compared with baseline post-operative echocardiogram <7 mm Hg). Indeed, clinically relevant SVD was uncommon (6.6%), with most patients exhibiting SVD presenting mild subclinical changes in valve performance. Subclinical changes on echocardiography should probably alert clinicians about possible structural changes within the bioprosthesis warranting several considerations such as: 1) additional imaging tests to confirm/dismiss a more relevant SVD; 2) closer follow-up considering that the time period between subclinical and clinically relevant SVD remains unknown; and 3) a more aggressive therapeutic approach to address modifiable risk factors that promote atherosclerotic disease, which in turn may play a role in the progression of SVD (29–31). In our study, advanced age, greater BMI, and the use of a Mitroflow valve were risk factors associated with bioprosthesis deterioration. Classically, younger age has been one of the factors determining bioprosthesis deterioration in surgical studies (12,32). However, as previously mentioned, most of these studies associated reoperation with SVD, and younger (vs. older) patients are more likely to be reoperated, which may partially explain this apparent inverse relationship. Interestingly, clinically relevant SVD was more frequent in patients <75 years of age, but age was not associated with an increased rate of clinically relevant SVD in the univariable and multivariable analyses.
Greater BMI and the associated metabolic risk factors such as diabetes and dyslipidemia have been associated with biological tissue degeneration in native (33) and biological aortic surgical (34) valves. SVD is an active process modulated by numerous mechanisms, and several studies support the implication of lipoproteins as one of the mechanisms that trigger inflammatory responses in the cusps of bioprosthetic valves, finally leading to valve calcification and deterioration (33,35,36). In addition, other molecules such as cytokines, found in abdominal adipose depot in obese patients, also possess proinflammatory and proatherogenic properties that exacerbate the inflammatory environment and induce low-density lipoprotein oxidation (33). A potential lipid-mediated mechanism as a pathway for SVD highlights the importance of both identifying the modifiable risk factors and increasing the therapeutic effort to delay valve deterioration. Previous studies have already shown a higher incidence of SVD associated with the Mitroflow prosthesis (16,37). This outcome has been attributed to the lack of anticalcification treatment during tissue preparation and fixation. The latest model of this valve with phospholipid reduction treatment may be associated with lower rates of SVD.
When SVD occurs, the standard treatment for dysfunctional surgical valves has been redo SAVR. The mortality associated with redo aortic valve surgery has decreased significantly over the last 2 decades. The rate of in-hospital mortality ranges from 2% to 7% in most studies (38,39) and long-term outcomes are encouraging, especially in young patients, New York Heart Association functional class I/II cases, and nonendocarditic or elective cases (40). However, several factors have been associated with an increased risk of morbidity and mortality in redo surgery cases, especially in the elderly, and many patients are refused for reoperation (41,42). Recently, valve-in-valve procedures have emerged as a less invasive alternative to redo surgery in patients at high risk (3,38). In the present study, the majority of patients underwent redo SAVR. As a reflection of the current clinical environment, an important percentage of patients (44%) underwent a TAVR valve-in-valve procedure. Although the early outcomes of TAVR valve-in-valve procedures have been well described, future studies are needed to determine valve durability after this intervention. Considering the durability issues associated with bioprosthetic valves, the use of newer mechanical valves with improved thrombogenicity properties allowing less aggressive anticoagulation regimens (On-X valve model; CryoLife Inc., Kennesaw, Georgia) or even no anticoagulation (Lapeyre-Triflo Furtiva aortic valve; Novostia SA, Neuchatel, Switzerland) (43) should be taken into account, particularly in younger patients with longer life expectancy.
Study limitations
This study was a retrospective analysis of prospectively collected data, and there was no independent echocardiographic core laboratory analysis. Baseline post-operative echocardiography was performed at hospital discharge instead of 1 to 3 months after the index procedure as recommended by guidelines. Data on stroke volume, heart rate, and hemoglobin levels at the time of echocardiograms were not available, as these factors may have influenced transvalvular gradient values. Echocardiographic follow-up at 10 years was incomplete (echocardiographic data not available in 31 patients). However, the 87% rate of echocardiographic data available at 10-year follow-up is one of the highest ever reported in the SAVR literature (16,27).
Conclusions
The late outcomes within a large contemporary series of consecutive and unselected patients undergoing SAVR with a bioprosthetic valve revealed a relatively high (>50%) mortality rate at 10-year follow-up, with older age and comorbidities (cardiac and noncardiac) determining poorer outcomes. This finding highlights the importance of both judicious patient evaluation/selection and a systematic follow-up of these patients, particularly those at increased risk due to the presence of comorbidities. Overall, surgical aortic bioprostheses exhibited a satisfactory hemodynamic performance over time, with clinically relevant SVD being infrequent. However, there was a significant increase in mean transprosthetic gradients coupled with a decrease in EOA at 10-year follow-up, with 30% of the population exhibiting subclinical SVD. The involvement of modifiable factors such as larger BMI and valve model in the process of SVD should stimulate further therapeutic efforts to improve valve durability. Future studies on the factors involved in and the evolution of subclinical SVD are warranted. Meanwhile, the results of this study should be taken into consideration when evaluating late clinical outcomes and valve durability issues after TAVR.
COMPETENCY IN PATIENT CARE AND PROCEDURAL SKILLS: In patients undergoing aortic valve replacement with biological prostheses, modifiable risk factors and prosthesis model predicted SVD. Although the durability of the valve prostheses was satisfactory overall, by 10 years, subclinical SVD developed in nearly one-third of patients.
TRANSLATIONAL OUTLOOK: Further studies are needed to assess the impact of risk factor modification, such as weight reduction, on the evolution of SVD after bioprosthetic aortic valve replacement. Also, future studies on the factors involved in and the evolution of subclinical SVD are warranted.
Appendix
1. : "Trends in aortic valve replacement for elderly patients in the United States, 1999-2011". JAMA 2013; 310: 2078.
2. : "Isolated aortic valve replacement in North America comprising 108,687 patients in 10 years: changes in risks, valve types, and outcomes in the Society of Thoracic Surgeons National Database". J Thorac Cardiovasc Surg 2009; 137: 82.
3. : "2017 AHA/ACC focused update of the 2014 AHA/ACC guideline 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". J Am Coll Cardiol 2017; 70: 252.
4. : "The St. Jude Toronto stentless bioprosthesis: up to 20 years follow-up in younger patients". Heart Surg Forum 2015; 18: E129.
5. : "Aortic valve replacement with Toronto SPV bioprosthesis: optimal patient survival but suboptimal valve durability". J Thorac Cardiovasc Surg 2008; 135: 19.
6. : "Long-term clinical outcomes 15 years after aortic valve replacement with the Freestyle stentless aortic bioprosthesis". Ann Thorac Surg 2014; 97: 544.
7. : "Pericardial stentless valve for aortic valve replacement: long-term results". Ann Thorac Surg 2016; 102: 1956.
8. : "Transcatheter aortic valve implantation: current and future approaches". Nat Rev Cardiol 2011; 9: 15.
9. : "Future of transcatheter aortic valve implantation—evolving clinical indications". Nat Rev Cardiol 2018; 15: 57.
10. : "Recommendations for evaluation of prosthetic valves with echocardiography and Doppler ultrasound: a report from the American Society of Echocardiography's Guidelines and Standards Committee and the Task Force on Prosthetic Valves". J Am Soc Echocardiogr 2009; 22: 975.
11. : "Recommendations for the imaging assessment of prosthetic heart valves: a report from the European Association of Cardiovascular Imaging endorsed by the Chinese Society of Echocardiography, the Interamerican Society of Echocardiography and the Brazilian Department of Cardiovascular Imaging". Eur Heart J Cardiovasc Imaging 2016; 17: 589.
12. : "Aortic bioprosthetic valve durability: incidence, mechanisms, predictors, and management of surgical and transcatheter valve degeneration". J Am Coll Cardiol 2017; 70: 1013.
13. : "2014 AHA/ACC guideline 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". J Am Coll Cardiol 2014; 63: e57.
14. : "Guidelines on the management of valvular heart disease (version 2012): the Joint Task Force on the Management of Valvular Heart Disease of the European Society of Cardiology (ESC) and the European Association for Cardio-Thoracic Surgery (EACTS)". Eur J Cardiothorac Surg 2012; 42: S1.
15. : "Aortic valve replacement with the Mitroflow pericardial bioprosthesis: durability results up to 21 years". J Thorac Cardiovas Surg 2008; 136: 688.
16. : "Early structural valve deterioration of Mitroflow aortic bioprosthesis: mode, incidence, and impact on outcome in a large cohort of patients". Circulation 2014; 130: 2012.
17. : "20-year experience of 1712 patients with the Biocor porcine bioprosthesis". J Thorac Cardiovasc Surg 2009; 137: 76.
18. : "Twenty-year durability of the aortic Hancock II bioprosthesis in young patients: is it durable enough?". Eur J Cardiothorac Surg 2014; 46: 825.
19. : "Carpentier-Edwards pericardial valve in the aortic position: 25-years experience". Ann Thorac Surg 2013; 96: 486.
20. : "Medtronic Mosaic porcine bioprosthesis: assessment of 12-year performance". J Thorac Cardiovasc Surg 2011; 142: 302.
21. : "Long-term clinical and echocardiographic follow-up of the Freestyle stentless aortic bioprosthesis". Circulation 2012; 126: s198.
22. : "Mechanical or biologic prostheses for aortic-valve and mitral-valve replacement". N Engl J Med 2017; 377: 1847.
23. : "Surgical aortic valve replacement outcomes in the transcatheter era". J Thorac Cardiovasc Surg 2015; 150: 1582.
24. : "Hancock II bioprosthesis for aortic valve replacement: the gold standard of bioprosthetic valves durability?". Ann Thorac Surg 2010; 90: 775.
25. : "Carpentier-Edwards supra-annular aortic porcine bioprosthesis: clinical performance over 20 years". J Thorac Cardiovasc Surg 2005; 130: 994.
26. : "Late results of heart valve replacement with the Hancock II bioprosthesis". J Thorac Cardiovasc Surg 2001; 121: 268.
27. : "Very long-term outcomes of the Perimount aortic valve in patients aged 60 or younger". Ann Thorac Surg 2015; 100: 853.
28. : "ApoB/ApoA-I ratio is associated with increased risk of bioprosthetic valve degeneration". J Am Coll Cardiol 2013; 61: 752.
29. : "Cardiovascular risk factors as predictors of early and late survival after bioprosthetic valve replacement for aortic stenosis". J Heart Valve Dis 2007; 16: 483.
30. : "Age and valve size effect on the long-term durability of the Carpentier-Edwards aortic pericardial bioprosthesis". Ann Thorac Surg 2001; 72: 753.
31. : "Hypercholesterolemia is a risk factor for bioprosthetic valve calcification and explantation". J Thorac Cardiovasc Surg 2003; 126: 969.
32. : "Late incidence and determinants of reoperation in patients with prosthetic heart valves". Eur J Cardiothorac Surg 2004; 25: 364.
33. : "Metabolic syndrome negatively influences disease progression and prognosis in aortic stenosis". J Am Coll Cardiol 2006; 47: 2229.
34. : "Metabolic syndrome is associated with faster degeneration of bioprosthetic valves". Circulation 2006; 114: I512.
35. : "Incidence, risk factors, clinical impact, and management of bioprosthesis structural valve degeneration". Curr Opin Cardiol 2017; 32: 123.
36. : "Spectrum of calcific aortic valve disease: pathogenesis, disease progression, and treatment strategies". Circulation 2005; 111: 3316.
37. : "Long-term results of the Mitroflow aortic pericardial bioprosthesis in over 800 patients: limited durability and mechanisms of dysfunction". Eur J Cardiothorac Surg 2017; 52: 264.
38. : "Transcatheter valve-in-valve and valve-in-ring for treating aortic and mitral surgical prosthetic dysfunction". J Am Coll Cardiol 2015; 66: 2019.
39. : "Re-operation for bioprosthetic aortic structural failure—risk assessment". Eur J Cardiothorac Surg 2003; 24: 873.
40. : "In which patients is transcatheter aortic valve replacement potentially better indicated than surgery for redo aortic valve disease? Long-term results of a 10-year surgical experience". J Thorac Cardiovasc Surg 2014; 148: 500.
41. : "Redo aortic valve surgery: early and late outcomes". Ann Thorac Surg 2011; 91: 1120.
42. : "Repeat heart valve surgery: risk factors for operative mortality". J Thorac Cardiovasc Surg 2001; 122: 913.
43. : "Mechanical versus biological valve replacement strategies". Expert Rev Cardiovasc Ther 2016; 14: 423.
Abbreviations and Acronyms
| BMI | body mass index |
| CABG | coronary artery bypass graft |
| CI | confidence interval |
| COPD | chronic obstructive pulmonary disease |
| EOA | effective orifice area |
| HR | hazard ratio |
| SAVR | surgical aortic valve replacement |
| SVD | structural valve degeneration |
| TAVR | transcatheter aortic valve replacement |
| TTE | transthoracic echocardiogram |
Footnotes
Drs. Rodriguez-Gabella and Asmarats were supported by a grant from the Fundacion Alfonso Martin Escudero, Madrid, Spain. Dr. Pibarot has received research contracts with Edwards Lifesciences and Medtronic. Dr. Josep Rodés-Cabau holds the Canadian Research Chair “Fondation Famille Jacques Larivière” for the Development of Structural Heart Disease Interventions. All other authors have reported that they have no other relationships relevant to the contents of this paper to disclose. Subodh Verma, MD, served as Guest Editor for this paper.
Listen to this manuscript's audio summary by JACC Editor-in-Chief Dr. Valentin Fuster.
