An 88-year-old man with a history of hypertension, insulin-dependent diabetes mellitus, moderate chronic obstructive pulmonary disease, and coronary artery disease with a remote inferior myocardial infarction was recently hospitalized for decompensated heart failure. He reported dyspnea with mild exertion while walking on level ground and associated substernal chest pressure consistent with New York Heart Association class III heart failure. He also reported 2-pillow orthopnea and progressive lower extremity edema. The physical exam demonstrated a grade III/VI late-peaking systolic crescendo-decrescendo murmur with obscuration of the second heart sound. The carotid upstrokes were blunted and delayed. The jugular venous pressure was estimated at 12 cm above the right atrium. A transthoracic echocardiogram demonstrated normal left ventricular function with an ejection fraction of 55% and a calcified trileaflet aortic valve with reduced leaflet mobility. Doppler interrogation of the aortic valve demonstrated a peak systolic velocity of 4.2 m/s and a mean systolic gradient of 45 mm Hg. The aortic valve area, calculated by the continuity equation, was found to be 0.8 cm². Based on the physical exam and echocardiography findings consistent with severe aortic stenosis, his predicted risk of mortality with surgical aortic valve replacement at 30 days was 11.5%.

Per the 2014 American Heart Association and American College of Cardiology (ACC) guidelines for the management of valvular heart disease, severe high-gradient aortic stenosis is defined as a peak aortic valve velocity greater than 4.0 m/s, a mean aortic valve gradient greater than 40 mm Hg, an aortic valve area of less than 1.0 cm², and an indexed aortic valve area of less than 0.6 cm²/m².¹

Calcific aortic stenosis is the most common form of acquired valvular heart disease in the United States that affects primarily older individuals, with a prevalence of 25% to 29% in patients older than age 65 years and 37% in those older than age 75 years.^2–4 Moderate to severe aortic stenosis develops in approximately 1.3% of those age 65 to 75 years and 2.8% of those older than age 75 years.⁵ Furthermore, the prevalence of aortic stenosis is likely to increase in the future as it is estimated that by 2020 there will be 20 million Americans older than age 80.⁶

Clinically, aortic stenosis is an insidious and progressive disease that develops over decades and has a long latency period prior to the development of symptoms.⁷ Compared to aged-matched asymptomatic controls, there is no difference in outcomes in patients with asymptomatic aortic stenosis.⁴ However, as described in 1968, once symptoms develop from aortic stenosis, the median survival is poor, at just 2 to 3 years.⁷ The classic triad of symptoms from aortic stenosis includes exertional angina, syncope, and dyspnea, with a median survival of 5 years when angina develops, 3 years when syncope develops, and 2 years when dyspnea develops.⁷ Even prior to the development of symptoms, there is a gradual progression in the degree of obstruction. A longitudinal study of 123 patients with at least mild aortic stenosis demonstrated an average increase in the peak aortic valve velocity of 0.3 m/s/yr, an average increase in the mean aortic valve gradient of 7 mm Hg/yr, and an average decrease in the aortic valve area by 0.1 cm²/yr.⁸

The pathophysiology of aortic stenosis is thought to be due to increased mechanical stress and reduced shear forces on the aortic valve cusps.⁹ This is illustrated well by the mechanics of congenital bicuspid aortic valve disease where unequal force is exerted on each of the valve leaflets, leading to premature degeneration of the valvular complex.⁹ Although patients with bicuspid aortic valves account for only 1% of the population with aortic stenosis, they account for approximately 50% of surgical aortic valve replacements.¹⁰ The injury to the valve complex is thought to occur from increased stress and subsequent invasion of the valvular endothelium by atherogenic lipoproteins, such as low-density lipoprotein and lipoprotein(a), with accompanying inflammation.⁹ The inflammatory component in the pathophysiology of aortic stenosis is suggested by an increased serum C-reactive protein, by an increased local temperature of the stenotic valve cusps, and by studies using fluorine-18-fluorodeoxyglucose positron emission tomography showing increased radiotracer uptake in the involved valve cusps.⁹ Furthermore, there is a transition to fibrosis through the differentiation of interstitial cells to myofibroblasts and then to osteoblasts that leads to valve calcification, reduced leaflet mobility, and valve obstruction, which then leads to pressure overload of the left ventricle, resulting in ventricular remodeling and hypertrophy (Figure 21-1).⁹

Given the association of aortic stenosis with coronary atherosclerosis and the involvement of atherogenic lipid particles, 5-hydroxy-3-methylglutaryl–coenzyme A (HMG-CoA) reductase inhibitors, or “statins,” have been studied to prevent progression of aortic stenosis. The Simvastatin and Ezetimibe in Aortic Stenosis (SEAS) study randomized 688 patients with mild to moderate aortic stenosis to simvastatin plus ezetimibe or placebo and showed no difference in major adverse cardiovascular events or disease progression requiring aortic valve replacement at a mean follow-up of 52 months. Analysis of secondary end points showed no difference in the rates of patients being referred for surgical aortic valve replacement (28.3% in the treatment arm and 29.9% in the control arm; hazard ratio [HR], 1.0; 95% confidence interval [CI], 0.84-1.18; P = .97).¹¹ Another randomized trial looked at the use of atorvastatin in 155 patients with at least mild aortic stenosis and showed that there was no difference in progression of the peak aortic valve velocity or aortic valve area between the 2 groups after 25 months of follow-up.¹² A randomized trial of the effect of rosuvastatin on echocardiographic parameters of aortic stenosis in 269 patients showed no difference in peak aortic valve velocities at a median follow-up of 3.5 years.¹² Based on these trials, lipid-lowering therapy with statins is not indicated for the secondary prevention of aortic stenosis progression.

Another potentially appealing therapeutic target for those with aortic stenosis is inhibition of the renin-angiotensin system with angiotensin-converting enzyme inhibitors (ACEIs) or angiotensin receptor blockers (ARBs) due to their antifibrotic effects. However, the vast majority of studies with ACEIs or ARBs for this application have failed to demonstrate a clear benefit.⁹ One observational study of 2117 patients in Tayside, Scotland demonstrated a reduction in all-cause mortality in patients with aortic stenosis (HR, 0.77; 95% CI, 0.65-92; P <.0001) in favor of those on ACEs or ARBs; however, the study is limited by its observational nature and other confounding factors.¹³

More recently, there has been interest in the osteoprotegerin–receptor activator of nuclear factor-κB (RANK)–RANK ligand (RANKL) pathway. There appears to be an association between aortic stenosis and altered bone metabolism as well as evidence of increased RANKL in stenotic valve tissue.¹⁴ Furthermore, there is evidence to suggest reduced calcification of bioprosthetic aortic valves in patients on bisphosphonate therapy.¹⁵ However, no randomized clinical trials with bisphosphonates or anti-RANKL monoclonal antibodies have been conducted in patients with aortic stenosis. Apart from routine echocardiographic surveillance to monitor for progression of aortic stenosis, there is no effective medical or preventative therapy for asymptomatic aortic stenosis.

As discussed earlier, medical therapy for aortic stenosis is limited, and once symptoms develop, survival is poor without more definitive therapy with aortic valve replacement.⁷ Therefore, the indications for aortic valve replacement for severe aortic stenosis are linked tightly to the presence of symptoms.¹ Valve replacement be achieved via surgical aortic valve replacement (SAVR) or transcatheter aortic valve replacement (TAVR) with discussion of the merits of each to follow.

SAVR

The first aortic valve replacement was performed by Drs. Hufnagel and Harvey in 1952.¹⁶ Since that time, numerous aortic valve replacements have been done with favorable outcomes.¹⁷ In a large surgical database of patients with aortic valve replacement, the average survival at 5, 10, and 15 years was 94.6%, 84.7%, and 74.9%, respectively.¹⁷ There are no prospective randomized controlled trials comparing SAVR with medical therapy; however, a Veterans Affairs hospital registry of patients who refused aortic valve replacement demonstrated a significant survival benefit with aortic valve replacement.¹⁷ Survival in the SAVR arm at 1, 3, and 5 years was 92%, 85%, and 73%, respectively, whereas survival in the medical therapy arm was 65%, 29%, and 16%, respectively (P <.0001 for each comparison).¹⁷ Therefore, the benefit of aortic valve replacement in patients with severe symptomatic aortic stenosis is clear; however, cardiac surgery is invasive and aortic valve replacement can be associated with a 5% to 10% in-hospital mortality that increases with age^18,19 and a 1-year mortality rate as high as 20% in some series.²⁰ As such, many otherwise eligible patients with severe aortic stenosis are turned down for SAVR due to their prohibitively high surgical risk, with one retrospective review suggesting that this occurs at a frequency of more than 30%.²¹ This indicates the large unmet need that exists for those who are not eligible or are at increased risk for surgery.

TAVR

TAVR was first attempted in animals in 1965 with the Davies valve and then again with the Anderson valve in 1992.^16,22 The first human transcatheter valve was implanted by Dr. Bonhoeffer who placed a transcatheter pulmonic valve in 2000.²³ The first human transcatheter aortic valve replacement was performed by Cribier in 2002.²⁴ This initial TAVR implant required a 24-Fr femoral vein sheath and transseptal puncture. The initial use of TAVR was focused on patients who were at too high risk or not candidates for conventional SAVR.²⁵ Since the first TAVR procedure was performed in 2002, there have been significant advances in the technology and technique that have translated into improved results.²⁶

Randomized Controlled Trials of TAVR

PARTNER Trials

The first randomized trial to assess clinical outcomes with TAVR was the PARTNER trial with the Edwards Sapien balloon-expandable transcatheter heart valve (Edwards, Irvine, CA). There were 2 different risk populations studied in the trial.^27,28 In cohort A, 699 high-risk surgically eligible patients with severe symptomatic aortic stenosis with New York Heart Association (NYHA) class II or greater symptoms were randomized to SAVR or TAVR. High risk for the trial was defined by a Society of Thoracic Surgeons (STS) predicted risk of mortality (PROM) of >10% or surgeon assessment of the risk of mortality with SAVR of >15%. This resulted in a patient population with a mean STS PROM of 11.8%. The primary end point of mortality was not statistically different at 1 year, with a mortality rate of 24.2% in the TAVR arm and 26.8% in the SAVR arm. Follow-up studies have shown durability of these outcomes, with 3-year data demonstrating comparable mortality.²⁰ In terms of secondary end points, major vascular complications at 30 days were more common in the TAVR group (11% vs. 3.2%, P <.001), whereas new-onset atrial fibrillation (8.6% vs. 16%, P = .006) and major bleeding (9.3% vs. 19.5%, P <.001) were more common in the SAVR group.

In cohort B, 358 inoperable patients with severe symptomatic aortic stenosis and NYHA class II or greater heart failure symptoms were randomized to TAVR with the Edwards Sapien balloon-expandable transcatheter heart valve or medical therapy.²⁸ The mean STS PROM in this inoperable cohort B population was 11.6%, which was lower than that noted in the high-risk cohort A population, in whom the mean STS PROM was 11.8%. This has been attributed to the inclusion of patients with features that affect overall surgical outcomes but are not captured by the STS score, such as patients with porcelain aorta (12.1%), chest wall deformity or radiation (13.1%), oxygen dependency (23.5%), and frailty (23.1%). Despite the similar STS score, however, the overall outcome in the TAVR arm in regard to mortality was worse in cohort B as compared to cohort A, confirming a higher risk population in cohort B. The primary outcome of all-cause death had occurred in 30.7% of the TAVR group as compared with 50.7% in the medical care arm at 1 year (HR, 0.55; 95% CI, 0.40-0.74; P <.001). This difference in mortality occurred despite 83.8% of patients in the medical care arm having balloon aortic valvuloplasty (BAV). Similar survival advantages were seen at 3 years in cohort B, with a survival of 46% in the TAVR arm and 19% in the medical therapy arm.²⁹ However, the trial did show an increased risk of stroke (5% vs. 1.1%, P = .06) and vascular complications (16.2% vs. 1.1%, P <.001) in the TAVR arm compared with standard of care.²⁸ A subgroup analysis of patients with an STS PROM of >15% in cohort B demonstrated no difference in mortality at 1 year, although the overall PARTNER trial demonstrated a clear mortality benefit of TAVR over medical therapy in inoperable patients with severe symptomatic aortic stenosis.²⁸

US PIVOTAL CoreValve Trials

The US PIVOTAL CoreValve trial randomized 795 high-risk patients with severe symptomatic aortic stenosis to TAVR with the Medtronic (Minneapolis, MN) self-expanding CoreValve heart valve or SAVR, with all-cause mortality as the primary outcome. In both an as-treated and an intent-to-treat analysis, TAVR fared better then SAVR, with a risk of all-cause mortality at 1 year of 14.2% versus 19.1% (P <.001 for noninferiority and P = .04 for superiority).³⁰ Unlike the PARTNER trial, there was no statistically significant difference in risk of stroke; however, major vascular complications (5.9% vs. 1.7%, P = .003) and need for pacemaker implantation (19.8% vs. 7.1%, P <.001) were more common in the TAVR group. Despite these adverse events, the high-risk group of the PIVOTAL trial was the first to demonstrate a potential survival advantage to TAVR in a high-risk patient population. It is notable that unlike the PARTNER trial, risk was not driven by STS PROM score but rather by surgeon consensus. This resulted in a lower mean STS PROM score in this high-risk cohort (7.4%) as compared to the mean STS PROM score of 11.8% in cohort A of the PARTNER trial.