Since the introduction of the 18-Fr CoreValve self-expanding transcatheter aortic valve replacement (TAVR), over 70,000 implants have been performed worldwide. The use of this self-expanding bioprosthesis has provided an alternative to surgery in patients who are suboptimal for conventional surgical aortic valve replacement and has resulted in improved survival and quality of life for thousands of patients worldwide. There are a number of potential advantages of a self-expanding bioprosthesis over alternative transcatheter designs, including the progressive self-expansion of the inflow frame reducing the degree of paravalvular regurgitation (PVR) over time; supra-annular location of the porcine pericardial valve, which improves hemodynamics and potentially improves long-term durability; and large cell diameter, which provides access to coronary arteries after implantation.

The purposes of this chapter are to review the CoreValve self-expanding frame design and newer iterations of the self-expanding prosthesis, discuss the clinical evidence for use of the self-expanding devices in an expanding population, and outline the risks and benefits of this device in patients with aortic stenosis.

The CoreValve System consists of 3 components: a transcatheter bioprosthesis, the delivery catheter system, and the compression loading system. The transcatheter bioprosthesis is composed of a self-expanding nitinol frame that supports a trileaflet porcine pericardial valve available in 23-, 26-, 29-, and 31-mm diameters that treat an annulus range from 18 to 29 mm¹ (Fig. 44-1). The inflow portion of the frame is designed to conform to the annulus and to stabilize the frame at the annular location.¹ The lowest 12 mm of the frame contains a porcine pericardial skirt to seal the annulus.¹ The valve is located in a supra-annular position at the waist (constrained portion) of the valve frame.¹ The outflow portion of the valve frame is constructed to support the valve commissures and orient the frame to facilitate laminar flow. All valve sizes are delivered using an 18-Fr catheter delivery system. The valve is deployed without rapid pacing and is partially repositionable until annular contact with the transcatheter heart valve (THV) is made. The CoreValve bioprosthesis has now been replaced commercially with the Evolut R transcatheter system (see below), which allows repositioning of the valve if the initial deployment is suboptimal.

Early European registries between 2007 and 2010 demonstrated the value of self-expanding transcatheter replacement in patients who were not optimal candidates for surgical valve replacement,^2-4 but transcatheter replacement was limited by 2-dimensional echocardiographic-based valve sizing, the availability of only 26- and 29-mm sizes, and delivery systems that had high friction, resulting in forward movement into the ventricle of the device with deployment. In the initial CE Mark series, 126 patients (mean age, 82 years; 42.9% male; mean logistic European System for Cardiac Operative Risk Evaluation score [EuroSCORE], 23.4%) with severe aortic valve stenosis underwent treatment with the CoreValve self-expanding bioprosthesis.³ The overall technical success rate was 83.1%, and the 30-day mortality rate was 15.2%, without significant differences in the subgroups.³ All-cause mortality at 2 years was 38.1%, and the mean aortic valve gradients remained unchanged from 30 days (8.5 ± 2.5 mm Hg) to 2 years (9.0 ± 3.4 mm Hg).³ There was no incidence of structural valve deterioration in this early series.³ The major limitations of these early series were postprocedural PVR (10%-15%)⁵ and high rates of conduction system disorders resulting in relatively high permanent pacemaker implantation rates (20%-40%).^2-4

The European Advance Registry included 1015 patients at high risk for surgical aortic valve replacement with an age of 81 ± 6 years and mean logistic EuroSCORE of 19.4% ± 12.3%.⁶ Self-expanding TAVR resulted in a 30-day all-cause mortality of 4.5%, rate of stroke of 3.0%, and life-threatening or disabling bleeding rate of 4.0%.⁶ The 12-month rates of all-cause mortality were 11.1%, 16.5%, and 23.6% among patients with a logistic EuroSCORE ≤ 10%, EuroSCORE of 10% to 20%, and EuroSCORE > 20% (P <.05), respectively.⁶ At 3 years, the Society of Thoracic Surgeons Predicted Risk of Mortality (STS PROM) ≤7% group had a lower rate of all-cause mortality (28.6% compared with 45.9% in patients with an STS >7%; P <.01).⁷ In patients with STS PROM ≤7%, mortality at 3 years was higher in those with moderate or severe aortic regurgitation at discharge than in those with mild or less aortic regurgitation (39.9% vs 22.9%, respectively; P <.01).⁷

The Italian CoreValve Registry reported the 5-year outcomes after self-expanding TAVR. The all-cause mortality rate was 55.0% at 5 years, and the overall neurologic event rate was 7.5%.⁸ Rehospitalization occurred in 46% patients, most of which (42.7%) were due to acute heart failure. Importantly, mean transaortic gradients remained low 5 years later (12.8 ± 10.9 mm Hg).⁸ Late prosthesis failure occurred in 5 cases (1.4%); among these, redo TAVR was successfully carried out in 2 patients (0.6%).⁸ The remaining 3 cases of prosthesis failure did not undergo additional invasive interventions.⁸ Ten patients (2.8%) showed late mild stenosis with a mean transaortic gradient ranging from 20 to 40 mm Hg.⁸ No other cases of structural or nonstructural valvular deterioration were observed.⁸

The US CoreValve Extreme Risk and High Risk Pivotal Trial and Registries were performed to evaluate the safety and efficacy of the CoreValve bioprosthesis in patients with severe aortic stenosis. This portfolio of studies evaluated patients who were deemed extreme risk or high risk for surgery by a local heart team and included both pivotal trials and continued access and expanded use registries.

CoreValve US Extreme Risk Study

Given the lack of clinical equipoise for subsequent randomized study in patients without surgical options, the CoreValve US Pivotal Study was a prospective, multicenter, nonrandomized investigation that included 489 patients with severe aortic stenosis who were deemed at extreme risk for surgery, defined as a ≥50% predicted 30-day mortality or irreversible morbidity at 30 days.¹ The primary end point was a composite of all-cause mortality or major stroke at 12 months and was compared with a prespecified objective performance goal.¹ The rate of all-cause mortality or major stroke at 12 months was 26.0%, compared with a 43.0% performance goal (P <.0001).¹ Individual 30-day and 12-month events included all-cause mortality (8.4% and 24.3%, respectively) and major stroke (2.3% and 4.3%, respectively).¹ The frequency of moderate or severe PVR was lower 12 months after self-expanding TAVR (4.2%) than at discharge (10.7%; P = .004 for paired analysis).¹

The benefits of self-expanding TAVR in extreme-risk patients were sustained at 2 years; the rate of all-cause mortality or major stroke was 38.0% (all-cause mortality, 36.5%; major stroke, 5.1%).⁹ Multivariable predictors of all-cause mortality at 2 years included the presence of coronary artery disease and admission from an assisted living center (Fig. 44-2).⁹ An STS PROM >15% was also predictive of 2-year all-cause mortality. The frequency of moderate or severe PVR (4.3% at 1 year; 4.4% at 2 years) was unchanged between the first and second years.⁹ These findings further reinforce that 2-year clinical outcomes are determined by comorbid conditions rather than valve performance in these complex patients.⁹

Clinical outcomes after self-expanding TAVR using an alternative access approach were also evaluated in the CoreValve US Extreme Risk Study. Reardon et al¹⁰ evaluated 150 patients with prohibitive iliofemoral anatomy who were treated with the CoreValve THV delivered by way of the subclavian artery (n = 70) or a direct aortic approach (n = 80). The preoperative aortic valve area was 0.72 ± 0.27 cm², and mean aortic valve gradient was 49.5 ± 17.0 mm Hg.¹⁰ After the TAVR, the effective aortic valve area was 1.82 ± 0.64 cm² at 1 month and 1.85 ± 0.51 cm² at 12 months.¹⁰ The mean aortic valve gradient was 9.7 ± 5.8 mm Hg at 30 days and 9.5 ± 5.7 mm Hg at 12 months.¹⁰ The primary end point was all-cause mortality or major stroke at 12 months and occurred in 15.3% of patients at 30 days and 39.4% at 12 months.¹⁰ The individual rates of all-cause mortality and major stroke were 11.3% and 7.5% at 30 days and 36.0% and 9.1% at 12 months, respectively.¹⁰ These data show that the CoreValve THV delivered by an alternative access provides a suitable alternative for treatment of extreme-risk patients with symptomatic severe aortic stenosis who have prohibitive iliofemoral anatomy and no surgical options. Three-dimensional access planning and real-time image guidance for direct aortic access TAVR may be facilitated using co-registration of multidetector computed tomography (CT) and noncontrast DynaCT image co-registration.¹¹

TAVR with a self-expanding bioprosthesis resulted in substantial improvements in both disease-specific and generic health-related quality of life in extreme-risk patients, but there remained a large minority of patients who died or had very poor quality of life despite TAVR.¹² In a detail health status analysis that evaluated 436 patient who were extreme risk for surgery, there was substantial improvement in both disease-specific and generic health status measures, with an increase in the Kansas City Cardiomyopathy Questionnaire Overall Summary Score (KCCQ-OS) of 23.9 points at 1 month, 27.4 points at 6 months, and 27.4 points at 12 months (P <.003 compared with baseline).¹² A large minority of patients (39%) had a poor outcome after TAVR that was predicted by the presence of being wheelchair dependent or having a lower mean aortic valve gradient, prior coronary artery bypass grafting, oxygen dependency, or very high predicted mortality with surgical aortic valve replacement, and low serum albumin.¹²

CoreValve US High-Risk Pivotal Trial

The CoreValve High-Risk Pivotal Trial enrolled patients deemed to be at elevated surgical risk.¹³ Two cardiac surgeons and 1 interventional cardiologist at each site needs to confirm the estimated 30 days surgical mortality or irreversible morbidity to be >15%. Because of the limitations of the standard STS PROM, additional non-STS factors, such as severe comorbidity, frailty, and disability, were also considered in the assessment of surgery risk.^14,15 A total of 795 patients were randomly assigned in a 1:1 fashion to self-expanding TAVR or surgical aortic valve replacement. The primary end point, the rate of death from any cause at 1 year in the as-treated population, was significantly lower in TAVR patients than in surgery patient (14.2% vs 19.1%, respectively), with an absolute reduction in risk of 4.9 percentage points (upper boundary of the 95% confidence interval [CI], –0.4; P <.001 for noninferiority; P = .04 for superiority).

Hemodynamic differences between self-expanding transcatheter and surgical bioprostheses were found in the CoreValve US High-Risk Clinical Study.¹⁶ Compared with surgery patients, TAVR patients had a lower mean aortic valve gradient, larger valve area, and less patient-prosthesis mismatch (all P <.001), but more PVR at discharge, which decreased at 1 year.¹⁶ Surgery patients experienced a significant right ventricular systolic dysfunction at discharge and 1 month, with normal right ventricular function at 1 year.¹⁶ Preimplantation aortic regurgitation rated as mild or worse was associated with reduced mortality hazard for both the TAVR patients (hazard ratio [HR], 0.48; 95% CI, 0.27-0.85; P = .01) and the surgery patients. Aortic regurgitation rated as mild or worse after TAVR was associated with increased risk for all-cause mortality (HR, 1.95; 95% CI, 1.08-3.53; P = .03).¹⁶ This benefit was consistent along a broad array of patient subsets (Fig. 44-3).

Kaul¹⁷ provided an overview of the benefit-risk analysis of the CoreValve US Pivotal Trial suggesting an overall net benefit of self-expanding TAVR in these patients (Fig. 44-4).

Reardon et al¹⁸ reported the 2-year results of the CoreValve US High-Risk Pivotal Trial and found that the rate of 2-year all-cause mortality was significantly lower in the TAVR group (22.2% vs 28.6% in the surgery group; log-rank P <.05), with an absolute reduction in risk of 6.5 percentage points. Importantly, the rate of death or major stroke at 2 years was significantly lower in the TAVR patients (24.2% vs 32.5% in surgery patients; log-rank P = .01).¹⁸ The hemodynamic (Fig. 44-5) and clinical beneficial effect of TAVR over surgery was maintained at 3 years after the procedure.¹⁹

Overall stroke rates are lower in patients undergoing TAVR than patients undergoing surgery.²⁰ The 30-day, 1-year, and 2-year stroke rates were 4.9%, 8.7%, and 10.9%, respectively, for TAVR and 6.2%, 12.5%, and 16.6%, respectively, for surgical aortic valve replacement (P = .46, .11, and .05, respectively).²⁰ The impact of stroke was profound, with an all-cause mortality in TAVR patients with a major stroke of 83.3% and in surgery patients of 54.5% (P = .29).²⁰ Late major stroke was disproportionately higher (23.8% at 2 years) among patients with poor iliofemoral access randomized to surgery.²⁰ Predictors of early stroke included peripheral vascular disease and falls within 6 months, and late strokes were predicted by severe aortic calcification and high Charlson scores.²⁰ Lack of dual antiplatelet therapy use during and after TAVR was associated with early stroke.²⁰

Prosthesis-Patient Mismatch

Another potential reason for the benefit of TAVR over surgery is superior hemodynamics in the TAVR group, understanding that the frequency of patient-prosthesis mismatch (PPM) is higher in patients undergoing surgery than TAVR.²¹ In an analysis of patients enrolled in the US High-Risk Pivotal Trial randomized to TAVR or surgery, patients were characterized as having severe PPM, defined as an effective orifice area index (EOAi) ≤0.65 cm²/m², or no PPM, defined as an EOAi >0.65 cm²/m².²¹ The incidence of severe PPM in the surgery group at 1 year was 25.7% compared with 6.2% in the TAVR group (P <.0001). At 1 year, the rates of all-cause mortality and acute kidney injury were significantly greater in patients with severe PPM compared with no severe PPM (20.6% vs 12.0% [P = .0145] for death; Fig. 44-6; and 19.2% vs 8.5% [P = .0008] for acute kidney injury, respectively).²¹ These findings demonstrate that severe PPM is more common in patients treated with surgery than those treated with TAVR and that patients with severe PPM are a greater risk for death and acute kidney injury than patients without severe PPM.²¹

Cost-Effectiveness Analysis

A health status evaluation in patients enrolled in the CoreValve US Pivotal Trial found that health status improved substantially in surviving patients who were treated with either self-expanding TAVR or surgery.²² TAVR via the iliofemoral route was associated with better early health status compared with surgery.²² In this analysis, among surviving patients eligible for iliofemoral access, there was a clinically relevant early benefit with self-expanding TAVR across all disease-specific and generic health status measures.²² Among the alternative access cohort, however, most health status measures were similar for TAVR and surgery.²² There were no consistent differences in health status between TAVR and surgery at the later time points.²²

Reynolds et al²³ performed a formal economic analysis in patients enrolled in the CoreValve US High-Risk Pivotal Trial using empirical data regarding survival and quality of life and medical resource use and hospital costs through 12 months. Relative to surgery, TAVR reduced initial length of stay an average of 4.4 days, decreased the need for rehabilitation services at discharge, and resulted in superior 1-month quality of life.²³ In contrast, index admission and projected lifetime costs were higher with TAVR than with surgery (Δ $11,260 and $17,849 per patient, respectively).²³ TAVR provided a lifetime gain of 0.32 quality-adjusted life-years. Lifetime incremental cost-effectiveness ratios were $55,090 per quality-adjusted life-year gained and $43,114 per life-year gained.²³ Sensitivity analyses indicated that a reduction in the initial cost of TAVR by approximately $1650 would lead to an incremental cost-effectiveness ratio <$50,000 per quality-adjusted life-year gained.²³

The benefits of self-expanding TAVR were examined in a number of patient subsets.

Female Sex

Despite a higher cardiovascular risk profile in men, studies have not shown differences in outcomes after self-expanding TAVR between men and women.²⁴ Forrest et al²⁵ compared the clinical outcomes in women and men in 3687 patients (1708 women and 1979 men) enrolled in the CoreValve US Pivotal Trials and Registries. Women tended to be slightly older and to have more frailty, but fewer cardiac comorbidities, higher left ventricular systolic function, less coronary artery disease, and fewer previous strokes.²⁵ All-cause mortality was 5.9% for women and 5.8% for men at 30 days (P = .87) and 24.1% and 21.3%, respectively, at 1 year (P = .08).²⁵ Similarly, the incidence of stroke was 5.7% in women and 4.0% in men at 30 days (P = .02) and 9.3% and 7.7%, respectively, at 1 year (P = .05).²⁵ Women had a higher incidence of bleeding, including more life-threatening bleeds, and a greater incidence of major vascular complications than men at 30 days.²⁵

In women at increased risk for surgery, women treated with self-expanding TAVR had better outcomes than women treated with surgery. Baseline characteristics and predicted risk were comparable in the 2 groups, although the frequency of diabetes mellitus was lower in patients undergoing TAVR (33.3% vs 45.3% in the surgery group; P = .02) in the US High-Risk Pivotal Trial.²⁶ One-year mortality was lower in TAVR patients (12.7% vs 21.8% in surgery patients; P = .03).²⁶ The composite all-cause mortality or major stroke rate was also lower in TAVR patients (14.9% vs 24.2% in surgery patients; P = .04; Fig. 44-7).²⁶

Prior Coronary Artery Bypass Graft

Patient with prior coronary artery bypass graft may benefit from self-expanding TAVR compared with repeat surgery. In an analysis of 226 patients with prior coronary artery bypass graft surgery enrolled in the CoreValve US High-Risk Trial, 1-year all-cause mortality was 9.6% for TAVR versus 18.1% for surgical aortic valve replacement (P = .06); cardiovascular mortality was 7.0% for TAVR versus 13.8% for surgical aortic valve replacement (P = .09).²⁷ A combination of STS PROM risk score >7 and age >80 years was a significant predictor of mortality, with TAVR demonstrating a survival advantage (P = .03).²⁷ No differences were seen for stroke.²⁷ The surgical aortic valve replacement group had longer intensive care unit and hospital stays and increased incidence of acute kidney injury, life-threatening or disabling bleeding, and major adverse cardiac and cerebrovascular events (P <.05).²⁷

Reduced Left Ventricular Function

It is estimated that one-third of patients with symptomatic aortic stenosis have reduced left ventricular ejection fraction (LVEF) before TAVR. In an analysis by Dauerman et al,²⁸ 156 patients from the CoreValve US Extreme and High-Risk trials with LVEF ≤40% were evaluated. Early LVEF recovery was defined as an absolute increase of ≥10% in ejection fraction at 30 days.²⁸ One-year outcomes were compared between patients with and without early recovery.²⁸ Early LVEF recovery occurred in 62% of patients, generally before discharge. By 30 days, LVEF increased >17% compared with baseline in the early recovery group with minimal increase in the no early recovery group (48.9% ± 8.8% vs 31.5% ± 6.9%, respectively; P <.001).²⁸ One-year all-cause mortality was numerically (but not statistically) higher in the no early recovery group (24% vs 12%; P = .07). Absence of previous myocardial infarction and baseline mean gradient ≥40 mm Hg were identified as predictors of early LVEF recovery.²⁸