Balloon aortic valvuloplasty (BAV) was developed as one of the first minimally invasive approaches to treat symptomatic severe aortic stenosis (AS). Alain Cribier first described the procedure in 1986 in a case series of three patients with severe calcific aortic stenosis. While the procedure was initially intended to be a minimally invasive alternative to surgical aortic valve replacement (SAVR), recognition of a high procedural complication rate, early restenosis, and lack of a mortality benefit eventually limited its overall utility.
Transcatheter aortic valve replacement (TAVR) or implantation (TAVI) has become a viable and durable therapy for patients with severe AS who have been deemed “inoperable” or “high-risk” for conventional SAVR. Since Alain Cribier first described the TAVR procedure in 2002, it is estimated that more than 125,000 procedures have been successfully performed worldwide in over 750 centers. During the past decade, procedural success, patient safety, and valve performance have improved dramatically due to technology enhancements, technique refinements, better patient selection, and a greater understanding of early and late clinical outcomes.
This chapter reviews the following: techniques for BAV; outcomes and complications with BAV; current indications for BAV; newer BAV technology; historical perspectives of TAVR; techniques for implantation of transcatheter valves; current and expanded clinical TAVR indications; updated TAVR clinical trial results; complications of TAVR; and an overview of the next generation of TAVR devices.
Balloon Aortic Valvuloplasty
BAV is a procedure in which one or more balloons are placed across a stenotic valve and inflated to fracture the calcified aortic valve leaflets. The procedure results in separation of the commissures and stretching of the aortic valve annulus. The immediate hemodynamic results include an increase in aortic valve area (although rarely greater than 1.0 cm 2 ) and a reduction of the transvalvular gradient. Despite only modest changes in the valve parameters, the procedure can lead to meaningful, albeit short-term, symptomatic improvement in patients.
The BAV procedure can be performed via a retrograde or an antegrade approach. The retrograde approach is more commonly utilized and involves access via the femoral artery with a 10 Fr to 14 Fr sheath. Anticoagulation is typically achieved with heparin dosing to achieve an activated clotting time (ACT) >250 but bivalirudin can be used in patients with heparin allergy. An extra-stiff guidewire (0.035 inch) is utilized for the procedure and is required for stabilization of the balloon during inflation and deflation. Care must be taken to position the wire with a gentle curve in the left ventricular (LV) apex to avoid the risk of ventricular perforation. Typically, a slightly undersized balloon, ~1 mm smaller than the annulus, is used to minimize risk of annular rupture while maximizing results. If an adequate result is not obtained with initial inflation, a larger balloon sized to the annulus may be used. The various balloons available (Zymed, Tyshack, Cristal) have different profiles and compliance curves and are typically inflated manually. Rapid ventricular pacing (160 to 180 bpm) via a temporary transvenous pacemaker can be utilized to stabilize the balloon during inflation by reducing forward cardiac output. Care must be taken to avoid prolonged pacing runs, which may cause ischemia and hemodynamic compromise. Vascular closure can be achieved by manual compression or the “preclose” technique with the Abbott Vascular Perclose device (Abbott Vascular Inc., California).
The antegrade transvenous approach has also been utilized for BAV and requires creation of a transcirculatory loop from the femoral vein to the ascending aorta via a transseptal puncture. An Inoue balloon (Toray, Tokyo, Japan) or a traditional valvuloplasty balloon can be used. One advantage of the Inoue balloon is the shape, which allows the waist of the balloon to fit the aortic valve annulus while the larger distal bulbous portion stretches the aortic leaflets more fully into the sinuses of Valsalva. Also, with one 26-mm balloon, multiple inflations can be performed at sizes ranging from 20 to 26 mm while assessing hemodynamic results in between. The potential benefits of the antegrade approach are reduced vascular complications, reduction of stroke, and greater increase in post-BAV area. However, it is more technically demanding given the need for a transseptal puncture and the subsequent looping of the guidewire in the LV apex.
The two largest registries that have evaluated BAV are the National Heart, Lung, and Blood Institute (NHLBI) and the Mansfield Scientific registries. The NHLBI registry evaluated 674 BAV patients immediately post procedure and up to 3 years later. High complication rates and in-hospital mortality were reported in this registry, with a 25% complication rate and a 3% mortality rate within the first 24 hours. The most common complication was the need for a transfusion in 20% due to vascular access issues. The cumulative cardiovascular mortality rate before discharge was 8%. Overall survival was 55% at 1-year, 35% at 2-year, and 23% at 3-year follow-up. Recurrent hospitalization (64%) and early restenosis were common. Echocardiography at 6 months demonstrated restenosis from the postprocedural valve area of 0.78 to 0.65 cm 2 .
The Mansfield Scientific Aortic Valvuloplasty Registry contained 492 patients and demonstrated comparable results. Approximately 20.5% of patients had complications after the procedure, with a 4.9% 24-hour mortality rate and a 7.5% mortality rate during the index hospital stay. Restenosis was also demonstrated to be nearly ubiquitous.
Although improved patient selection and technical improvements have led to a modest decrease in complication rates over the past 20 years, postprocedural morbidity remains high. In a contemporary series of 262 high-risk surgical or inoperable AS patients, the most common complications after BAV were intraprocedural death (1.6%); stroke (1.99%); coronary occlusion (0.66%); severe aortic regurgitation (1.3%); need for permanent pacemaker (0.99%); severe vascular complication (6.9%)—perforation (1.6%), ischemic leg (2.6%), pseudoaneurysm (1.99%), arterial-venous fistula (0.66%); acute kidney injury (11.3%); and new hemodialysis (0.99%). At approximately 6 months, there was a 50% mortality rate, and restenosis was evident as early as a few days postprocedure.
Given the frequency of these complications, their prompt identification and management are imperative. Despite smaller sheaths, vascular complications remain frequent and operators must have the skill set to manage them utilizing endovascular techniques with covered stents and prolonged balloon inflations. The most devastating complication for patients remains cerebrovascular injury, which occurs in 1% to 2% of patients. The etiology is typically atheroembolism from the ascending aorta or calcific embolism from the valve. The use of embolic protection devices in the future may further mitigate these complications.
The 2014 American College of Cardiology (ACC)/American Heart Association (AHA) guidelines on valvular heart disease specify that BAV might be reasonable (Class IIB recommendation) as a bridge to SAVR or TAVR in patients with severe AS. In the prior 2006 ACC/AHA valve guidelines, BAV was given a Class IIB recommendation for palliation in patients with co-morbidities that prevent aortic valve replacement, but this recommendation was removed in the 2014 guidelines owing to a lack of evidence. The updated 2012 European Society of Cardiology (ESC) and the European Association for Cardio-Thoracic Surgery (EACTS) valve guidelines also give a Class IIB recommendation for BAV as a bridge to SAVR or TAVR in hemodynamically unstable patients who are at high risk for surgery or in patients with symptomatic AS who require urgent major non-cardiac surgery. However, the ESC guidelines also state that BAV may be considered for palliation in patients unfit for SAVR or TAVR, but do not designate a formal class recommendation.
While BAV can be utilized as a bridge to TAVR in patients who are at extreme risk, it has also been utilized as a selection strategy for TAVR and SAVR in patients with severe AS but with other potential causes of their symptoms such as severe lung disease. Temporary improvement in symptoms after BAV would support aortic stenosis as the cause, and replacement of the aortic valve would be warranted. And last, BAV has been used to temporize patients with acute hemodynamic failure while formulating a decision between SAVR, TAVR, and medical therapy.
Next Generation BAV Devices
With the advent of TAVR, there has been renewed interest in BAV. Several new devices have now been developed with the hope of improving both the safety and efficacy of BAV as a stand-alone procedure, as well as improving preparation (predilatation) for subsequent TAVR. Four such devices are the InterValve V8 (InterValve, Plymouth, Minnesota), Bard TRUE™ dilatation balloon catheter (Bard Medical, Burlingame, California), the CardioSculpt scoring balloon (AngioScore, Fremont, California), and the Pi-Cardia LeafLex system (Pi-Cardia, Beit Oved, Israel) ( Figure 29-1 ).
The InterValve V8 ( Figure 29-1A ) and Bard TRUE™ balloon ( Figure 29-1B ) are Food and Drug Administration (FDA) approved and available for use. Both of these devices attempt to address limitations of the current devices. The InterValve V8 has a dumbbell shape, which allows it to lock into the valve anatomy and limit balloon movement. The waist of the balloon is 5 to 7 mm less than the proximal and distal bulbous segments of the balloon, and this shape is maintained throughout inflation. The proximal bulb allows for hyperextension of the leaflets into the sinus to enhance valve opening, and the smaller waist reduces the risk of annular dissection. Furthermore, a rapid balloon inflation and deflation time minimizes ischemic time and hypotension. The balloon comes in 22-, 24-, 26-, and 28-mm sizes. Bard TRUE™ dilatation balloon catheter is made of Kevlar composite balloon material with a precisely reproduced size and shape. It has also been designed for fast inflation and deflation, rewrapping, and puncture resistance. The balloon comes in sizes 20, 22, 24, and 26 mm × 4.5 cm length.
The CardioSculpt BAV ( Figure 29-1C ) and the Pi-Cardia LeafLex system ( Figure 29-1D ) are two devices undergoing investigation. The CardioSculpt device is a scoring balloon, which consists of a balloon encased in a nitinol scoring element. In theory, it also allows for better seating and stability of the device. Also the balloon has rapid deflation times and excellent rewrap, reducing deflated device profile. The Pi-Cardia LeafLex system is not a balloon but a catheter that delivers mechanical shock waves to fracture calcium within the aortic valve. This allows for increased leaflet compliance with an increase in aortic valve area. Clinical results from both these devices may demonstrate their efficacy and potential role in patients.
Transcatheter Aortic Valve Replacement (Implantation)—General Concepts
Historical Perspectives and Unmet Clinical Need
During the past 50 years, the standard of care for symptomatic AS has been SAVR, which in most patients is associated with prolonged survival, improved symptoms, and few procedural complications. However, both the risks and the recovery after SAVR are less favorable in elderly AS patients, especially those with multiple co-morbidities, including prior cardiac surgery, chronic lung disease, peripheral vascular disease, prior stroke, renal failure, coronary artery disease, and frailty. In addition, there are anatomic factors, such as porcelain aorta and chest wall deformities, that increase the risk of conventional SAVR. For these reasons, it is estimated that at least one-third of patients with symptomatic severe AS are either not candidates or are denied surgical therapy. This has prompted investigation into alternative less invasive catheter-based approaches such as BAV and TAVR. From the mid-1980s to the mid-1990s, BAV was selectively used in high-risk AS patients, but as previously discussed, the recognition of procedural complications, prohibitive early restenosis, and lack of mortality benefit relegated the use of BAV to a small clinical niche: either palliative therapy or as a bridge to definitive valve replacement.
The first catheter-based aortic valve replacement was performed in 1965 by Davies in a canine model for the temporary relief of aortic insufficiency. Andersen performed the first contemporary transcatheter aortic valve replacement procedure using a stent-based porcine bioprosthesis in pigs in 1992. The first implantation of a transcatheter valve in a human was performed by Bonhoeffer in 2000 involving the percutaneous replacement of a pulmonary valve in a right-ventricle to pulmonary-artery prosthetic conduit, using a bovine jugular valve to treat a 12-year-old boy with severe stenosis and regurgitation of the valved prosthesis. Cribier’s first-in-human landmark TAVR procedure in 2002 was undertaken as a last resort in a patient with cardiogenic shock, failed BAV, and multiple co-mordibities. Since that time, TAVR has become the standard of care in patients who are “inoperable” and is an important alternative in patients who are high risk for surgery, with more than a dozen different device variations either commercially available or under active investigation.
In the current ESC/EACTS and ACC/AHA guidelines for the management of valvular heart disease, it is recommended that the following patients be considered for TAVR procedures: patients with severe, symptomatic, calcific stenosis of a trileaflet aortic valve who have aortic and vascular anatomy suitable for TAVR, an expected survival >12 months, and surgical risk assessment by a multidisciplinary heart team indicating the following:
A prohibitive surgical risk as defined by an estimated 50% or greater risk of mortality or irreversible morbidity at 30 days or other factors such as frailty, prior chest wall radiation therapy, porcelain aorta, severe hepatic or pulmonary disease
A high surgical risk with an expected 30-day mortality at least 15%, as an alternative to SAVR
Heart Team Model and Risk Assessment
Given the complexity regarding the management of elderly patients with severe AS, a collaborative heart team model is essential for appropriate patient selection and subsequent care. This multidisciplinary team consists of experienced cardiac surgeons, interventional cardiologists, imaging specialists, heart failure specialists, cardiac anesthesiologists, intensivists, neurologists, geriatricians, nurses, and social workers. The coordinated approach of the heart team results in more comprehensive patient evaluations, facilitated gathering of essential data, improved communication with patients and families, superior decision making, and ultimately, better clinical outcomes. The importance of the heart team model is emphasized in both the European and the US TAVR guidelines.
Several different surgical risk algorithms are utilized by heart teams for the selection of patients for TAVR. The two most common risk assessment tools are the Society of Thoracic Surgeons (STS) score and the logistic EuroSCORE. The STS score is derived from outcomes data of 24 covariates in 67,292 patients undergoing isolated SAVR in the United States, while the logistic EuroSCORE is derived from 12 covariates from 14,799 patients undergoing all forms of cardiac surgery in Europe. While both have been shown to be accurate for estimating risk (i.e., 30-day surgical mortality) in low-risk patients with AS, their accuracy is far less precise in higher risk patients. The two scores differ predominantly in the covariates utilized in the respective models and in the populations studied. It is generally acknowledged that the STS score is more accurate at estimating SAVR mortality in higher risk populations. There is reasonable consensus that an STS score ≥8 is deemed to be high risk for SAVR and that these patients should also be considered for TAVR. While both the STS and logistic EuroSCORE can aid in the selection of patients for TAVR, they serve as only one aspect of the selection process and should be utilized in the context of the entire clinical picture.
Several important concepts relevant to risk assessment for TAVR require further consideration. In the unique patient population currently screened for possible TAVR (elderly patients with co-morbidities or anatomic limitations), many risk factors are not represented in the standard risk scores, including frailty, dementia, hepatic disease, and anatomic factors (e.g., porcelain aorta or “hostile” chest). These ignored or under-represented co-morbidities must be considered by the heart team during risk assessment. At the extreme end of the risk spectrum are the so-called futile AS patients, wherein there is little hope of meaningful quality of life and/or limited life expectancy (e.g., untreatable malignancy or severe dementia), despite successful TAVR therapy. Although this may be a difficult societal conundrum, it is the responsibility of the heart team to thoughtfully identify these patients, such that TAVR may be sensitively withheld as a treatment option. Importantly, surgical risk is a continuum and the categorization of risk status into discrete groups is somewhat arbitrary and depends on definitions that are changing over time and may be different in the rarified confines of a clinical trial versus real-world community standards. Finally, since the predictors of early and late outcomes after TAVR are different compared with SAVR, specific risk assessment models for TAVR would be clinically useful and are being actively evaluated.
Anatomic Screening and Need for Multimodality Imaging
The information gathered during multimodality anatomic screening should be utilized to make an informed judgment on the candidacy of a patient for TAVR and in the overall management of patients with AS. The salient imaging data needed for comprehensive TAVR screening include (1) confirmation of the diagnosis of tri-leaflet, calcific, and severe valvular AS; (2) determination of left ventricular size and function; (3) coronary artery anatomy; (4) peripheral vasculature of sufficient size and suitability for catheter access and prosthesis delivery; and (5) geometry, measurement, and calcium patterns of the left ventricular outflow tract, proximal aorta, and the aortic annulus for appropriate device selection. Imaging for anatomic screening consists of a combination of echocardiography, angiography, and multi-slice computed tomography (MSCT).
Echocardiography is clearly the gold standard for assessing the etiology and severity of AS. Other important anatomic findings best determined by echocardiography are left ventricular mass, size, and function; right ventricular size and function; and other valvular lesions (especially mitral and tricuspid regurgitation). Usually, transthoracic echocardiography is sufficient, but in some patients with difficult imaging planes transesophageal echocardiography is preferred. Coronary angiography is crucial in every patient to determine the need for concomitant revascularization, given the frequent co-existence of coronary artery disease and AS. Peripheral angiography is also recommended to assess tortuosity, size, and calcification of the distal aorta, iliac, and femoral vessels. However, MSCT with contrast is the best imaging study to quantitatively measure the lumen dimensions of peripheral arteries and their suitability for a given TAVR system. MSCT is also the recommended imaging study to determine the optimal transcatheter valve size, which may differ depending on the specific transcatheter valve type. These three-dimensional (3D) reproducible measurements of the annulus region using validated algorithms derived from high-quality contrast MSCT have become the global standard modality in selecting the correct valve size. Intraprocedural 3D echocardiography can also be used to confirm the annulus measurements and to assist in valve sizing. MSCT is also helpful in measuring the location and height of the coronary arteries; patterns of calcification in the aortic valve, aorta, and left ventricular outflow tract; and the shape, angulation, and size of the proximal aorta. Much of the success of TAVR and the recent improvements in clinical outcomes have been directly linked to meticulous preprocedural planning using the aforementioned multimodality imaging studies.
TAVR is always performed in a sterile environment, either a catheterization laboratory or an operating room, with fluoroscopic and angiographic digital imaging capabilities. Most recently, there has been a growing interest in using a “hybrid” catheterization laboratory–operating room suite for TAVR. These hybrid procedure rooms combine the advantages of a high-resolution angiographic catheterization lab with the concomitant availability of a sterile environment for surgical management of complications and to facilitate nonpercutaneous alternative access routes.
The presence of cardiac anesthesiology to supervise sedation and analgesia control and to assist with hemodynamic monitoring and management has been an important requirement for TAVR to provide optimal care of these high-risk AS patients. There is growing controversy whether general anesthesia versus monitored anesthesia control (conscious sedation) is necessary or preferred in all or most patients during TAVR procedures. Similarly, the requirement of intraprocedural transesophageal echocardiography in every case has been highly debated. The more traditional approach incorporates general anesthesia with transesophageal echocardiography to help guide the procedure, including confirmation of valve sizing and positioning, assessment of paravalvular regurgitation, and rapid recognition of complications. Nevertheless, an increasing number of TAVR operators prefer a more “minimalist” approach, without general anesthesia and employing only transthoracic echocardiography, as needed. The reasons for this less invasive TAVR strategy are reduced resource consumption, fewer anesthesia-related complications, more rapid patient ambulation, and shorter durations of hospital stay. Thus far, experienced operators adopting this simplified approach have had equivalent procedural outcomes. Perhaps a stratified patient-specific approach is most reasonable, wherein lower risk patients or those with anticipated intubation morbidities (e.g., severe chronic obstructive pulmonary disease [COPD]), can be triaged to the minimalist strategy and the higher risk patients can be managed using a more intense strategy with general anesthesia and transesophageal echocardiography guidance. As sheath sizes decrease and operator experience increases, likely the majority of TAVR worldwide will be performed in catheterization laboratories using conscious sedation.
Balloon Expandable Valves
Technology Overview and Early Access Approaches
All TAVR systems are composed of three integrated components: a support frame (usually metallic), a bioprosthetic tri-leaflet valve, and a delivery catheter. The support frame is crimped onto the delivery catheter immediately prior to valve implantation and is expanded by either retracting a sheath or inflating an underlying balloon. Balloon-expandable TAVR systems (Edwards Lifesciences, Irvine, California) were the earliest used in patients and have undergone several generations of evolution. However, many technology features have remained constant over time, including the tubular-slotted metallic frame geometry, pericardial bioprosthetic valve leaflets sewn to the frame, a fabric “skirt” covering the bottom of the frame, and out-of-body circumferential crimping of the valve and frame assembly onto the delivery catheter. The Cribier-Edwards valve became available in 2004 and was used in many of the early feasibility cases in Europe and the United States. This TAVR system had both 23- and 26-mm valve sizes and a stainless steel frame with an attached equine pericardial trileaflet valve that was directly crimped onto a commercial balloon valvuloplasty catheter. Cribier’s first case and many of the earliest cases were performed using an antegrade transfemoral vein approach, wherein after right femoral vein access, a transseptal puncture provided entry to the left heart, followed by positioning a stiff guidewire across both the mitral and aortic valves. Navigating these first generation devices and large-sized catheters with roughened distal edges across the interatrial septum and the tortuous intracardiac anatomy was challenging. The initial antegrade transfemoral vein transseptal access procedures required very experienced operators with advanced skills and resulted in many intraprocedural complications. Specifically, the generation of a large guidewire loop inside the left ventricle, which was required to avoid traction on the anterior mitral valve leaflet, was extremely difficult to maintain throughout the procedure and often resulted in severe mitral regurgitation with hemodynamic collapse.
Difficulties with the unpredictability of the antegrade transseptal approach resulted in modifications of both the access approach and the delivery catheter. A simpler and more familiar access site, typically used with BAV procedures, was the femoral artery with retrograde transaortic entry into the left ventricle. This can be accomplished with direct percutaneous access or through an open surgical exposure of the common femoral artery. Webb and colleagues reviewed their initial experience with the Cribier-Edwards valve via a retrograde transfemoral approach in a case series of 50 patients. For this purpose, a steerable delivery catheter with a deflectable tip was also developed to safely advance the TAVR system within the vasculature and to better align the valve assembly with the central valve orifice. In 2007, the next generation Edwards-SAPIEN (Edwards Lifescience, Irvine, California) transcatheter valve was introduced ( Figure 29-2A ). The major differences included a change from equine to bovine pericardium valve leaflet material, which enabled surgical valve-like consistency in tissue processing (decalcification, thickness, flexibility, and tensile strength), as well as further improvements in the delivery catheter. The third generation SAPIEN XT valve (20-, 23-, 26-, and 29-mm sizes) began clinical evaluations in 2010 and represented a more radical design change of all system components, with a major goal to importantly reduce the overall profile ( Figure 29-2B ). The support frame had less metal and was changed from stainless steel to a thinner cobalt alloy, the valve geometry was modified to allow partial closing in the open position, and the delivery catheter was reduced in diameter by 33% for all valve sizes, with improved transitions to facilitate advancement and crossing. The marked reduction in system profile was in part due to an endovascular docking maneuver, such that the valve was crimped onto the catheter shaft for arterial entry, and in the descending aorta, the balloon was pulled back underneath the valve for subsequent deployment. The SAPIEN XT is the current commercially available balloon expandable TAVR system in the United States.
Most recently, the fourth generation SAPIEN 3 device ( Figure 29-3A ) has completed enrollment in clinical trials in the United States (PARTNER II registry) and has received CE approval in Europe. The overall system profile has been further reduced, with most valve sizes introduced through a 14 Fr expandable sheath. The frame geometry has been modified with larger cells distally, and in addition to the internal skirt, an external skirt has been added to fill gaps and to prevent paravalvular regurgitation.
Procedural Details for Sapien or Sapien XT Implantation
The typical approach for balloon expandable valve implantation is via the transfemoral approach if arterial access permits. For the current generation Sapien XT, a 16 Fr, 18 Fr, or 20 Fr e-sheath is required for the 23-, 26-, and 29-mm valves, respectively. As noted earlier, procedures are typically performed in a hybrid OR under either general anesthesia or conscious sedation. In addition to fluoroscopy, echocardiography (transthoracic echocardiogram [TTE] or transesophageal echocardiogram [TEE]) should be available for procedural guidance and/or postdeployment assessment. A temporary pacemaker is required for the procedure and is placed at the beginning via the femoral or internal jugular vein. Aortography is performed to identify a coplanar view for valve deployment. It is essential that the nadirs of all three cusps be in the same plane to ensure proper deployment of the valve ( ). The view can be identified by the preoperative CT scan or though fluoroscopy and aortography. Arterial access for the valve delivery sheath can be obtained either percutaneously or via a surgical cutdown; however, as sheath sizes continue to decrease, the majority of procedures will be done utilizing percutaneous access and closure with the “preclose” technique. In the near future, there may be dedicated closure devices for large vessel access.
The remaining steps of the transfemoral retrograde TAVR procedure are as follows:
Placement of the valve delivery sheath after proper dilation of the artery.
Guidewire crossing of the stenosed native aortic valve and positioning of a tightly curved Amplatz extra-stiff guidewire in the left ventricular apex.
Although some operators are avoiding predilatation, valvuloplasty is usually performed with an under-sized balloon using transient rapid right ventricular pacing at heart rates of 180 beats to 200 beats per minute to minimize pulsatile transaortic flow.
Advancement of the steerable delivery catheter and crimped valve assembly through the vasculature to a coaxial position above the native valve.
Crossing the native valve, retraction of the sheath, and final positioning in a transvalvar location with approximately 60% to 70% of the prosthesis above the annulus confirmed using coplanar fluoroscopic views +/− transesophageal echo (prosthesis will shorten ~3 mm from the ventricular side during deployment) ( ).
Deployment of the bioprosthetic valve by slow balloon inflation during rapid right ventricular pacing to insure a stable platform ( ).
Assessment of paravalvular regurgitation (by hemodynamics, angiography, and echocardiography).
If significant paravalvular aortic regurgitation is present, postdilatation may be performed with addition of 1 to 2 cc of volume to the balloon delivery catheter (postdilatation should not be performed in scenarios where there is increased risk for annular or root injury such as aggressive valve oversizing [>20%] or severe left ventricular outflow tract [LVOT] calcification).
Removal of the catheters and suture-based closure of the percutaneous arteriotomy site or surgical repair of the access site.
Alternative Access Approaches
In patients with severe peripheral artery disease and/or marked vessel tortuosity, or concerning pathological anatomy of the ascending aorta, there was a pressing need for an alternative to retrograde transfemoral arterial access. The first important nontransfemoral access site for balloon expandable transcatheter valves was the antegrade transapical approach, wherein the left ventricular apex was exposed via a small left anterolateral intercostal incision (fifth or sixth intercostal space) to expose the left ventricular apex. After purse string or mattress sutures are placed to secure the apex, direct needle puncture allows introduction of a hemostatic sheath into the left ventricle. The valve prosthesis is crimped in the antegrade direction onto the delivery catheter and is introduced through the sheath to an optimal transannular location, followed by deployment with balloon expansion during rapid right ventricular pacing. Bleeding from the apical entry site is always a concern, and careful surgical closure is required to avoid complications.
The transapical approach was often preferred by surgeon TAVR operators for the following reasons:
More precise control of transcatheter valve positioning, due to the close proximity of the entry and deployment sites
Less need for predilatation with valvuloplasty balloons, as the ventricular surfaces of the native aortic valve were less resistant to crossing
Avoidance of “hostile” proximal ascending aorta pathology
An early perception of reduced periprocedural strokes
Disadvantages associated with the transapical approach include the possibility of early and late access site bleeding from the left ventricular apex, hemodynamic instability due to placement of a large intraventricular sheath, especially in patients with small hypercontractile ventricles or severe left ventricular dysfunction, the requirement to use general anesthesia in all patients, and the sequelae of a left thoracotomy procedure resulting in increased patient pain and delayed recovery. With the reduced catheter profiles of modern TAVR systems, the need for alternative transapical access has significantly diminished.
In addition to transapical access, other interesting alternative access concepts have been developed for patients with unsuitable anatomy for the standard percutaneous transfemoral technique. For some TAVR systems, there is a preference for the subclavian/axillary artery approach, via surgical cutdown, with either direct access or use of a prosthetic graft to the artery. Recently, the direct aortic approach has become more popular, due to the familiarity with standard surgical procedures involving exposure of the ascending aorta and aortic root cannulation. The direct aortic approach requires an upper hemisternotomy or an upper right parasternal intercostal incision to expose a disease-free portion of the ascending aorta just below the origin of the innominate artery. This retrograde access site is used to insert a short sheath followed by the TAVR delivery system. Other less commonly used access alternatives include direct iliac and distal descending aorta exposure via a retroperitoneal incision for placement of an iliac conduit, direct exposure of a carotid artery, and transcaval access with placement of a sheath from the inferior vena cava to the abdominal aorta followed by closure of the aortotomy hole using implantable occluder devices.
Early Feasibility Trials
After the initial first-in-human experiences with TAVR, four different nonrandomized feasibility studies for balloon expandable Edwards transcatheter valves (Cribier-Edwards or Edwards SAPIEN) were conducted: REVIVE II, REVIVAL II, PARTNER EU, and TRAVERCE. These feasibility registries demonstrated that TAVR could successfully be performed in a safe and efficacious manner in high-risk AS patients, and affirmed the short- to intermediate-term durability of the first generation Edwards transcatheter valves.
The earliest feasibility trial was the multicenter REVIVE II registry, which consisted of 106 patients in Canada and Europe who underwent retrograde transfemoral-TAVR with the Cribier-Edwards valve. The nearly concurrent US transfemoral REVIVAL II registry comprised another 55 similar high-risk patients. In a pooled analysis of the two trials, TAVR was attempted in 161 patients with successful valve deployment in 142 (88.2%). Thirty-day major adverse events were 18.6%, with 18 (11.2%) deaths, 5 (3.1%) myocardial infarctions, and 7 (4.3%) cerebrovascular events. Adverse vascular events occurred in 15.5% of patients, and 4.9% needed permanent pacemakers. The 1-year survival was 73.8% and multivariate analysis demonstrated that prior coronary artery bypass surgery, baseline NYHA class, and procedural vascular complications were strong predictors of 1-year mortality. A subsequent transapical REVIVAL II feasibility study with the Edwards SAPIEN valve was initiated with 40 patients in the United States. There was higher than anticipated valve migration or embolization (12.5%); mortality and strokes were 17.5% and 5.5% at 30 days and 36% and 9% at 6 months. This small early study demonstrated that while transapical TAVR was feasible, it was associated with significant morbidity and mortality, perhaps in part due to the increased co-morbidities in the patients ineligible for transfemoral vascular access.
The European-based PARTNER EU feasibility study with the Edwards SAPIEN valve included concurrent transapical and transfemoral patient cohorts. This registry consisted of 130 patients, of whom 61 patients underwent transapical and 69 underwent transfemoral TAVR. Overall, successful valve deployment was 95.4% in the transapical patients and 96.4% in the transfemoral patients. In transapical patients, mortality was 18.8% at 1 month and 50.7% at 1 year, compared with 8.2% at 1 month and 21.3% at 1 year in the transfemoral patients. At 1 year, improvement in NYHA class was observed in 78.1% of surviving transapical patients and in 84.8% of transfemoral patients. In addition, quality of life, measured using the Kansas City Cardiomyopathy Questionnaire, improved in 73.9% in the transapical group and in 72.7% in the transfemoral group. No evidence of structural valve deterioration was observed during 1-year follow-up. The single arm TRAVERCE trial included 168 European patients who underwent transapical TAVR with either the Cribier-Edwards or the Edwards SAPIEN valve. For the entire group, 95.8% had successful valve implantation; the remainder had valve migration, embolization, or severe valve regurgitation. Overall, mortality at 30 days, 6 months, and 1 year was 15%, 30%, and 37%, respectively. Other outcomes included conversion to conventional surgery in 5.4%, early stroke in 1.2%, and new permanent pacemakers in 6%. Again, there appeared to be higher early and late mortality associated with the transapical approach, but it was difficult to determine if differences in baseline patient characteristics (co-morbidities) or the transapical access route were responsible.
SOURCE and Other Registries
The SAPIEN Aortic Bioprosthesis European Outcome (SOURCE) registry was created to gather clinical outcome data on the Edwards SAPIEN 23- and 26-mm transcatheter valves, for both the transfemoral and transapical approaches, during the early phase of commercialization in Europe ( Table 29-1 ). At the time, it was the largest single registry of TAVR patients and it was meant to reflect a consecutive case “real-world” experience. The SOURCE registry included 2307 consecutive patients in 37 different centers from 14 different countries and was divided into two cohorts, based on the time of enrollment; cohort 1 were patients enrolled from November 2007 to January 2009, and cohort 2 were patients enrolled from February 2009 to December 2009. Due to the large size of the Edwards SAPIEN TAVR system (outer sheath diameters 8 to 9 mm), the majority of patients (62.7%) were treated using transapical access. The mean logistic EuroSCORE was 27.6% in the transapical group and 23.9% in the transfemoral group, indicating different risk profiles. The mean age of all patients was 81.6 years and 57.8% were women. All data were site reported, there were no core laboratories or formal event adjudication committees, and currently, 2-year follow-up results have been reported.
|REGISTRY||VALVE(s) USED||PATIENT * POPULATION||AVERAGE † RISK SCORE||ACCESS||MAJOR ENDPOINTS|
|SOURCE (Cohort 1 and 2)||ESV 100%||N = 2307 |
|EuroSCORE 26.0%||TF: 60.2% |
|30 days: |
All-cause mortality 9.6%
Major vascular complication 5%
Major bleeding 3.6%
New PPM 7%
|1 year: |
TF: All-cause mortality 7.5%
New PPM 6.7%
TA: All-cause mortality 10.9%
New PPM 7.1%
|Canadian Multicenter||ESV/SXT 100%||N = 339 |
|STS 9.8%||TF: 48.6% |
|30 days: |
Procedural success rate 93.3%
All-cause mortality 10.4% (TF 9.5%, TA 11.3%)
Stroke 0.6% (TF 0.6%, TA 0.6%)
Major vascular complication 13% (TF 13%, TA 13%)
New PPM 4.9% (TF 3.6%, TA 6.2%)
|1 year: |
TF: All-cause mortality 25%, TA: All-cause mortality 22%
|2 year: |
TF: All-cause mortality 35%, TA: All-cause mortality 36%
|SOURCE XT||SXT 100%||N = 2688 |
|EuroSCORE 20.5%||TF: 62.7% |
|30 days : |
All-cause mortality 6.3%
Major vascular complication 14.5%
Major bleeding 8%
New PPM 9.5%
|1 year: |
All-cause mortality 19.5%
All rehospitalizations 29.4%
Moderate or severe PVR 6.2%
|PREVAIL (TA and TF)||SXT 100%||TA: N = 212 |
|TA: EuroSCORE 24.1||TF: 60% TA: 40%||TA 30 days: |
All-cause mortality 7.5%
Major vascular complications 0.9%
New PPM 12%
|TA 1 year: |
All-cause mortality 17%
New PPM 13.1%
|TF: N = 141 |
|TF: EuroSCORE 22.4||TF 30 days: |
All-cause mortality 8.5%
Major vascular complication 11.4%
New PPM 8.7%
|TF 1 year: |
All-cause mortality 17%
New PPM 13.5%
|Piazza et al.||MCV 100%||N = 646 |
|EuroSCORE 23%||TF: 100%||30 day: |
Procedural success 97%
All-cause mortality 8%
Composite of death, stroke, or MI 9.3%
|Italian Registry||MCV 100%||N = 659 |
|EuroSCORE 23%||TF: 90% |
|30 days: |
All-cause mortality 5.4%
|1 year: |
Composite of death, major stroke, MI, and life-threatening bleeding 30.4%
All-cause mortality 23.6%
|2 year: |
Composite of death, major stroke, MI, and life-threatening bleeding 36.5%
All-cause mortality 30.3%
|3 year: |
Composite of death, major stroke, MI, and life-threatening bleeding 40.3%
All-cause mortality 34.8%
|Australia/NZ Registry||MCV 100%||N = 441 |
|EuroSCORE 17.3%||TF: 88.9% |
|1 year: |
All-cause mortality 12%
|2 year: |
All-cause mortality 22.1%
|ADVANCE||MCV 100%||N = 1015 |
|EuroSCORE 19.4%||N/A||30 day: |
All-cause mortality 4.5%
|1 year: |
All-cause mortality 17.9%
|UK TAVI||ESV 48%, MCV 42%||N = 870 |
|EuroSCORE 19%||TF: 69% |
|30 days : |
Procedural success 97%
All-cause mortality 7.1%
Major vascular complication 6.3%
New PPM 16%
Moderate/Severe AR 14%
|1 year: |
All-cause mortality 21%
|2 year: |
All-cause mortality 26%
|Comparison of TF vs. other access: |
TF had lower mortality and surgical conversion, but more AR and vascular complications
|ESV vs. MCV: |
ESV had less AR, need for second valve, and PPM, but greater surgical conversion
|FRANCE 2||ESV 67%, MCV 33%||N = 3195 |
|EuroSCORE 22%||TA: 18% |
|30 days: |
Procedural success 96.9%
All-cause mortality 9.7%
|1 year: |
All-cause mortality 24%
Incidence of all PVR 64.5%
|GARY||ESV 53%, MCV 42%, Other 5%||N = 3875 |
|EuroSCORE 25%||TA: 30%||TA 30 days: |
All-cause mortality 9.0%
|TA 1 year: |
All-cause mortality 28.0%
New PPM 14.1%
|Non-TA: 70%||Non-TA 30 days: |
All-cause mortality 5.6%
|Non-TA 1 year: |
All-cause mortality 20.7%
New PPM 26.2%
|PRAGMATIC||ESV 43%, MCV 57%||N = 793 |
|EuroSCORE 21%||TF: 100%||MCV vs. ESV at 30 days: |
Procedural success 94% vs. 96%
All-cause mortality 7.5% vs. 5.0%
Major stroke 3.5% vs. 1.5%
Major vascular complication 9.1% vs. 15%
Life-threatening bleed 12% vs. 14%
PPM 23% vs. 5.9%
|MCV vs. ESV at 1 year: |
All-cause mortality 17% vs. 14%
The 30-day, 1-year, and 2-year all-cause mortality for transapical patients was 11%, 26%, and 34.5%, respectively, and for the transfemoral patients it was 7.6%, 19.9%, and 26.9%, respectively. Most of the late deaths were from noncardiac causes, as the 2-year cardiac mortality was only 12.8% after transapical TAVR and 9.6% after transfemoral TAVR. Increased major bleeding was more frequent in transapical patients (3.9% vs. 2.3%), whereas greater vascular access-related complications were observed in the transfemoral patients (major—11.3% vs. 2.0%; minor—10.4% vs. 1.0%). At 2 years, the rates of all strokes were 5.9% in the transapical group and 5.8% in the transfemoral group, and new pacemakers were needed in 8.7% of transapical and 9.3% of transfemoral patients. Additionally, 1.9% of transapical patients needed a re-intervention of the bioprosthetic valve, compared with 0.5% of transfemoral patients. At 2 years, there was a sustained, similar symptom improvement (reduction in NYHA class in survivors) for both access approaches.
The Multicenter Canadian Study is another real-world registry that sought to capture the early experiences with balloon expandable valves in the Canadian population. The study included 339 patients judged to be nonoperable or at very high surgical risk who underwent either transfemoral or transapical TAVR from January 2005 to June 2009 in six Canadian centers with the Cribier-Edwards, Edwards SAPIEN, or the SAPIEN XT balloon expandable valve. The average age was 81 years, mean Society of Thoracic Surgeons (STS) score was 9.8%, and there was a nearly equal distribution of transapical and transfemoral cases (52% vs. 48%). Long-term outcomes including nearly 4 years of systematic follow-up results have been published.
The procedural success was 93.3%, 30-day all-cause mortality was 10.4% (transfemoral—9.5%, transapical—11.3%), and 30-day strokes were 2.3% (transfemoral—3%, transapical—1.7%). Patients with either porcelain aorta (18%) or frailty (25%) exhibited 30-day outcomes similar to the rest of the study population, and porcelain aorta patients tended to have a better survival at 1 year. At a mean follow-up of 42 months, 55.5% of patients had died, and the causes of late death were noncardiac in 59.2%, cardiac in 23.0%, and unknown in 17.8%. Predictors of late mortality were chronic obstructive pulmonary disease, chronic kidney disease, chronic atrial fibrillation, and frailty. A mild nonclinically significant decrease in valve area was seen at 2-year follow-up, but no further reduction in valve area was observed during the later follow-up evaluations (mean 3.5 years). No late changes in paravalvular regurgitation and no cases of structural valve failure were observed during the extended follow-up period.
The lower profile SAPIEN XT TAVR system was studied in three registries—PREVAIL TA, PREVAIL TF, and SOURCE XT. The PREVAIL TA registry included 212 patients undergoing TAVR via transapical access and the PREVAIL TF registry included 141 patients with transfemoral access. All-cause mortality was 7.5% and 17% at 30 days and 1 year in the transapical cohort and 8.5% and 17% at 30 days and 1 year in the transfemoral cohort. Other outcomes including stroke, myocardial infarction (MI), and acute kidney injury were also similar between the two approaches, although major vascular complications were higher in the transfemoral patients.
The SOURCE XT registry enrolled 2688 patients in 93 centers from 17 European countries between July 2010 and October 2011. Vascular access in the trial was not limited to transfemoral and transapical, but also included a small number of direct aortic access cases (3.7%). Unlike the original SOURCE registry, SOURCE XT included standardized valve academic research consortium (VARC) endpoint definitions, core laboratories, and an independent clinical events committee. Compared with SOURCE, patients in the SOURCE XT registry had a lower logistic EuroSCORE (overall, 20.4%), a significant minority (28%) did not have general anesthesia during the procedure, and due to the smaller delivery systems, patients were more likely to be treated using the transfemoral access route (62.7%).
All-cause mortality was lower in SOURCE XT than in the previous SOURCE registries: 6.3% at 30 days and 19.5% at 1 year (cardiac mortality 10.8% at 1 year). Importantly, vascular complications were no longer associated with increased mortality, due to improved management strategies and fewer major events. Periprocedural (within 48 hours) and 1-month strokes were 2.2% and 6.3% and the need for periprocedural pacemakers was 5.7%. The frequency of moderate or severe paravalvular regurgitation was only 5.5% at 1 month and 6.2% at 1 year. There was dramatic symptom improvement in survivors at 1 year and all-cause mortality at 1 year was better in women (p = 0.008) and in patients with a lower logistic EuroSCORE (<15% vs. ≥15%; p = 0.003). The 1-year mortality was almost double in the transapical versus the transfemoral patients (27.2% vs. 15.0%) and multivariate analysis clearly indicated that transapical access was a strong predictor of 1-year mortality (hazard ratio [HR] 1.64; 95% confidence interval [CI], 1.28, 2.09; p < 0.0001).
The PARTNER Trial
The Placement of Aortic Transcatheter Valves (PARTNER) trial is the seminal trial that demonstrated the safety and efficacy of TAVR for high-risk and inoperable AS patients and set the benchmark for all subsequent TAVR device approval trials ( Figure 29-4A ) ( Table 29-2 ). PARTNER was the first multi-center, randomized controlled trial of TAVR compared with accepted standard therapies in carefully defined patient populations. The results of this trial ultimately led to the approval of the Edwards SAPIEN valve by the U.S. Food and Drug Administration (FDA) and has informed the worldwide cardiology community of the benefits and concerns of TAVR as an important new therapy for patients with AS.
|TRIAL||PRIMARY VALVE STUDIED||TREATMENT ARMS||POPULATION STUDIED||MAJOR STUDY RESULTS|
|PARTNER 1A 1-Year Outcomes||Edwards SAPIEN balloon-expandable valve||TAVR vs. SAVR||High-risk surgical candidates (STS ≥ 10%) N = 699||30 day: |
All-cause mortality 3.4% vs. 6.5% (p = 0.07)
Stroke 4.6% vs. 2.4% (p = 0.12)
Major vascular complications 11% vs. 3.2% (p < 0.001)
Major bleeding 9.3% vs. 19.5% (p < 0.001)
Atrial fibrillation 8.6% vs. 16% (p = 0.006)
|1 year : |
All-cause mortality 23% vs. 26.8% (p = NS)
Stroke 6.0% vs. 3.2% (p < 0.08)
|PARTNER 1A 2-Year Outcomes||Edwards SAPIEN balloon-expandable valve||TAVR vs. SAVR||High-risk surgical candidates (STS ≥ 10%) N = 699||All-cause mortality 33.9% vs. 35% (p = NS)|
|Stroke 7.7% vs. 4.9% (p = NS)|
|PARTNER 1A 3-Year Outcomes||Edwards SAPIEN balloon-expandable valve||TAVR vs. SAVR||High-risk surgical candidates (STS ≥ 10%) N = 699||All-cause mortality 44.2% vs. 44.8% (p = NS)|
|Stroke 8.2% vs. 9.3% (p = NS)|
|PARTNER 1B 1-Year Outcomes||Edwards SAPIEN balloon-expandable valve||TAVR vs. standard therapy||Inoperable patients N = 358||30 day: |
All-cause mortality 5.0% vs. 2.9% (p = NS)
Stroke 5.0% vs. 1.1% (p = 0.06)
Vascular complications 16% vs. 1.1% (p < 0.001)
|1 year: |
All-cause mortality 31% vs. 51% (p < 0.001)
Death or hospitalization 43% vs. 72% (p < 0.001)
NYHA ≥ III 25% vs. 58% (p < 0.001)
|PARTNER 1B 2-Year Outcomes||Edwards SAPIEN balloon-expandable valve||TAVR vs. standard therapy||Inoperable patients N = 358||All-cause mortality 43% vs. 68% (p < 0.001)|
|Cardiac mortality 31% vs. 62% (p < 0.001)|
|Stroke 13.8% vs. 5.5% (p = 0.01)|
|Rehospitalization 35% vs. 73% (p < 0.001)|
|PARTNER 1B 3-Year Outcomes||Edwards SAPIEN balloon-expandable valve||TAVR vs. Standard therapy||Inoperable patients N = 358||All-cause mortality 80.9% vs. 54.1% (p < 0.001)|
|Cardiac mortality 74.5% vs. 41.4% (p < 0.001)|
|Rehospitalization 75.5% vs. 42.3% (p < 0.001)|
|PARTNER 2A||Edwards SAPIEN XT balloon-expandable valve||TAVR vs. SAVR||Intermediate-risk surgical candidates (STS ≥ 4%) N = 2000||Enrollment completed; currently in the follow-up period|
|PARTNER 2B||Edwards SAPIEN XT balloon-expandable valve||TAVR with SAPIEN XT vs. SAPIEN||Inoperable patients N = 358||30 day: All-cause mortality, disabling stroke, or rehospitalization 17% vs. 15.3% (p = NS) |
Major vascular events 9.6% vs. 15.5% (p = 0.04)
Procedural: Less anesthesia time, aborted procedures, multiple valve implants, and need for hemodynamic support with SAPIEN XT
|1 year: All-cause mortality, disabling stroke, or rehospitalization 33.9% vs. 34.7% (p = NS) |
All-cause mortality 22.5% vs. 23.7% (p = NS)
Disabling stroke 4.5% vs. 4.6% (p = NS)
Rehospitalization 19% vs. 17.4% (p = NS)
|CoreValve Pivotal Trial Extreme Risk||Medtronic self-expanding CoreValve||Iliofemoral TAVR with CoreValve (Standard therapy comparison arm removed after results of PARTNER 1B)||Inoperable patients N = 471||30 day: All-cause mortality or major stroke 9.3% |
All-cause mortality 7.9%
Major stroke 2.4%
Moderate to severe PVR 11.5%
|1 year: All-cause mortality or major stroke 25.5% |
All-cause mortality 24.0%
Major stroke 4.1%
Moderate to severe PVR 4.1% (no severe PVR)
|CoreValve Pivotal Trial High Risk||Medtronic self-expanding CoreValve||SAVR vs. TAVR||High-risk surgical candidates (STS ≥ 10%) N = 699||30 day: All-cause mortality 4.5% vs. 3.3% (p = NS) |
MACCE 10.4% vs. 7.7% (p = NS)
Major stroke 3.1% vs. 3.9% (p = NS)
|1 year: All-cause mortality 19.1% vs. 14.2% (p = NS) |
MACCE 27.3% vs. 20.4% (p = 0.03)
Major stroke 7.0% vs. 5.8% (p = NS)
|SURTAVI||Medtronic self-expanding CoreValve||SAVR vs. TAVR||Intermediate-risk surgical candidates (STS ≥ 4% ≤ 10%) N = 2000||Completing enrollment and entering follow-up period|
|CHOICE||Edward SAPIEN XT/ Medtronic self-expanding CoreValve||Balloon-expandable valves vs. self-expanding valves||High-risk surgical candidates (STS ≥ 10%) N = 241||Procedural: All-cause mortality 0% vs. 0% (p = NS) |
Implant ≥2 valves 0.8% vs. 5.8% (p = 0.03)
Device success 95.8% vs. 77.5% (p < 0.001)
|30 day: |
All-cause mortality 4.1% vs. 5.1% (p = NS)
Combined safety endpoint 18.2% vs. 23.1% (p = NS)
Stroke 5.8% vs. 2.6% (p = NS)
NPP 17.3% vs. 37.6% (p = 0.001)
During the course of PARTNER development and enrollment, it became clear that more robust standardized clinical trial processes were necessary to assess study outcomes in patients with AS. The Valve Academic Research Consortium (VARC) was convened in 2009, including representatives from surgical and cardiology societies in Europe and the United States, prominent Academic Research Organizations, the FDA, and several expert consultants. Standardized endpoint definitions for all important clinical outcomes were carefully developed and published in the first consensus document in 2011. Thereafter, upon testing these definitions in clinical trial settings, several additions and revisions were published as the VARC-2 consensus document in 2012. This dynamic process of creating optimal and standardized endpoint definitions has strengthened the TAVR evidence-based medicine effort and VARC definitions were immediately incorporated into the PARTNER trials.
The purpose of the PARTNER trial was to study the Edwards-SAPIEN TAVR system in discrete high-surgical-risk patients using rigorous clinical trial methodologies (randomization vs. control therapies, core laboratories, adjudicated and carefully defined clinical events, etc.). PARTNER began enrollment in 2007 and included 1057 patients (of 3105 screened patients) with severe AS (defined as an aortic valve area <0.8 cm 2 or aortic valve area index <0.5 cm 2 /m 2 , and a mean aortic valve gradient >40 mm Hg, or a peak aortic jet velocity >4.0 m/sec), and cardiac symptoms (NYHA Class II or greater) in two parallel randomized trials in whom conventional SAVR was associated with high or prohibitive risk. Patients were divided into two cohorts :
Cohort A: those who were not considered to be suitable candidates for surgery because they had co-existing conditions that would be associated with a predicted probability of 50% or more of either death by 30 days after surgery or a serious irreversible condition.
Cohort B: those who were considered to be candidates for surgery despite being at high surgical risk, defined by an STS risk score of 10% or higher or the presence of co-existing conditions that would be associated with a predicted risk of death by 30 days after surgery of 15% or higher.
In the inoperable cohort, 358 patients were randomly assigned (1 : 1) to either transfemoral TAVR or to standard therapy (medical therapy with or without adjunctive BAV). In the high-risk cohort, 699 patients were randomly assigned (1 : 1) to either transfemoral TAVR or SAVR (244 vs. 248 patients) or, if the peripheral vascular anatomy was not suitable to accommodate the large sheaths, transapical TAVR or SAVR (104 vs. 103 patients). The primary endpoint for both cohorts was all-cause mortality, with cohort B powered for superiority versus standard therapy over the course of the trial and cohort A powered for noninferiority versus SAVR at 1 year. In both studies, the follow-up was at least 1 year in all patients before assessing the primary endpoint results and other outcomes. Thus far, findings from PARTNER have been presented with 5-year follow-up for the inoperable cohort and with 3-year follow-up for the high-risk cohort.
Patients in the inoperable cohort averaged 83 years old, more than half were female, mean STS score was 11.7%, multiple co-morbidities were prevalent (including frailty and COPD), >90% had NYHA functional Class III or Class IV symptoms, and ~80% received BAV (at least once) in the standard therapy arm. The primary endpoint analysis at 1 year indicated a reduction in all-cause mortality from 50.8% with standard therapy to 30.7% after transfemoral TAVR (p < 0.0001; Figure 29-5A ); the number needed to treat was only 5 patients. This dramatic reduction in mortality was sustained over time with landmark analyses indicating incremental TAVR mortality benefits up to 3 years ( Figure 29-5B ). Importantly, in standard therapy patients who did not receive crossover to TAVR (allowed after 1 year) or out-of-protocol valve replacement, there was only one survivor at 5 years, which recapitulates the dire prognosis of “untreated” severe AS. Improved mortality after TAVR was demonstrated in all age groups, although limited in patients with the highest baseline STS scores (>15%). Other benefits associated with TAVR therapy in these inoperable patients included significant reductions in rehospitalizations and improved cardiac symptoms with significant quality-of-life enhancement. Notable complications associated with TAVR therapy were:
Major vascular complications and bleeding associated with the large delivery catheters
Increased strokes (6.7% vs. 1.7%)