Aortic stenosis (AS) is the most common acquired valvular disease in adults.¹ The incidence of AS has steadily increased along with population life expectancy and age, affecting almost 5% of patients over the age of 75 years.² The most common cause of valvular AS in adults is calcification of a normal trileaflet valve or a congenital bicuspid valve.³ Severe symptomatic AS has a poor prognosis when treated medically, with a mortality of almost 80% at 2 years.⁴ Surgical aortic valve replacement (SAVR) is currently the standard of care and accepted to alleviate symptoms and prolong survival; however, up to one-third of patients are denied SAVR because of prohibitive surgical risk (eg, advanced age, significant left ventricular dysfunction).⁵

In 2002, Cribier and colleagues⁶ successfully performed the first transcatheter aortic valve replacement (TAVR) in an elderly inoperable patient for the treatment of severe AS. This approach marked the feasibility of percutaneous valve implantation, and TAVR has emerged as a less invasive alternative to conventional SAVR.⁷ Results from the Placement of Aortic Transcatheter Valves (PARTNER) trial have allowed TAVR to become the standard of care for extremely high-risk or “inoperable” patients, and TAVR is a valid alternative to surgery for selected high-risk but “operable” patients with symptomatic AS.^8,9 TAVR has been performed for over 10 years with >100,000 implantations worldwide, and this innovative technology has allowed a paradigm shift in the treatment of AS.¹⁰

TAVR is recommended for patients with symptomatic, severe, calcific stenosis of a tricuspid aortic valve who have predicted survival >12 months and either prohibitive surgical risk (defined by >50% estimated risk of mortality or irreversible morbidity at 30 days, frailty, prior cobalt chest irradiation, porcelain aorta, coronary bypass graft adhesion to the posterior sternum preventing safe reentry into the chest, or severe pulmonary or hepatic disease) or high surgical risk (Society of Thoracic Surgeons [STS] mortality risk ≥8% with predicted actual mortality >15%).^11,12 Trials are ongoing to evaluate TAVR in a medium-risk population (STS mortality risk 3%-8%; PARTNER II trials).

Currently, the most data available for TAVR are based on the balloon-expandable SAPIEN valves (Edwards Lifesciences, Irvine, CA; Fig. 42-1). The Edwards SAPIEN transcatheter heart valve system has been used extensively worldwide and was used in the PARTNER I trial.⁷ Based on this trial, the SAPIEN valve was the first transcatheter aortic valve to be granted approval by the US Food and Drug Administration (FDA) in November 2011 for use in surgically inoperable patients and in 2012 for use in surgically high-risk patients in the United States.¹² The Edwards SAPIEN consists of a trileaflet, bovine pericardial valve mounted on a stainless steel stent with a polyethylene terephthalate skirt and seals at the aortic annulus with the stent frame deployed just beneath the coronary ostia. It was available in 2 sizes: 23 mm (for aortic annuli between 18 and 22 mm) and 26 mm (for aortic annuli between 21 and 25 mm) with a sheath and delivery catheter system of 22 Fr and 24 Fr, respectively. Dimensions and adequate requirements for the SAPIEN valve are provided in Table 42-1. More recently, the SAPIEN was replaced by the SAPIEN XT in Europe, Canada, and the United States (June 2014). Dimensions and adequate requirements for the SAPIEN XT valve are provided in Table 42-2. SAPIEN XT also consists of a bovine pericardial valve but is mounted to a thinner cobalt-chromium stent with fewer struts to allow for a lower profile delivery system. The SAPIEN XT is available in a wider range of sizes (23, 26, and 29 mm). Additionally, the Edwards eSheath (Edwards Lifesciences) combines a low profile with a dynamic expansion mechanism and is available in 16, 18, and 20 Fr (23-, 26-, and 29-mm valve, respectively).¹³ The eSheath allows for transient sheath expansion during delivery system passage and reduces the time the access vessel is expanded.

The most recent generation of the SAPIEN series of valves is the SAPIEN 3 (S3). The S3 consists of a bovine pericardial valve mounted on a cobalt-chromium stent with an outer skirt (polyethylene terephthalate) designed to enhance sealing and minimize paravalvular leak. The S3 is a taller valve (decreases the likelihood of native leaflet tissue prolapse) with 4 rows and 4 columns of cells between each commissure for high radial force. The S3 is available in 20, 23, 26, and 29 mm and can be delivered through a reduced profile 14- and 16-Fr (for the 29-mm valve) eSheath. Dimensions and adequate requirements for the S3 valve are provided in Table 42-3. In addition, the latest iteration of the delivery system (Commander) has several characteristics intended to facilitate accurate positioning, including increased flexion capability to engage the native valve coaxially and a fine alignment mechanism to make precise adjustments prior to deployment. The low-profile S3 valve and delivery system may facilitate fully percutaneous implantation in a broader range of patients with more accurate positioning and less paravalvular regurgitation. Improvements in valve technology will make it feasible for less experienced centers to make the transition to a complete percutaneous approach using local anesthesia and fluoroscopic guidance in the catheterization laboratory.¹⁴

The assessment of a patient’s surgical risk, clinical status, and feasibility for TAVR is best accomplished by a multidisciplinary heart team consisting of clinical cardiologists, cardiac imaging specialists, interventional cardiologists, and cardiothoracic surgeons. We begin to establish surgical risk using the STS Predicted Risk of Mortality (PROM) score, which evaluates factors such as age, renal function, type of surgery, and pulmonary function. An STS-PROM score (mortality rate) of >8% signifies high risk (as defined by the inclusion criteria of the PARTNER trial cohort A group), and a score of 4% to 8% signifies moderate surgical risk (as defined by the inclusion criteria of the PARTNER II trial cohort A group).⁹

The STS-PROM scoring system provides a formal means of risk stratification; however, it fails to include several important patient characteristics, including frailty and disability, which are particularly relevant to the elderly AS cohort. Several standardized models of frailty exist, and all seem to indicate that frailty has a significant impact on outcomes after cardiac surgery. Elderly patients with a low level of independent functioning (Katz Index of independent activities of daily living) were found to have twice the surgical mortality and 6 times the risk of prolonged institutional care following cardiac surgery.¹⁵ Even crude measures of disability, such as gait speed, have been found to increase the risk of mortality or major morbidity by 2 to 3 times for any given STS-PROM score after cardiac surgery.¹⁶ A high frailty score (which includes assessment of gait speed, grip strength, serum albumin, and activities of daily living) has been independently associated with increased 1-year mortality following TAVR.¹⁷ These findings serve to remind clinicians that although risk stratification begins with validated models, a comprehensive assessment must consider other factors that may not be so easily measured.

A small but significant number of patients with AS are unable to have surgery due to a myriad of anatomic and medical conditions. Reasons for technical inoperability include porcelain aorta, previous mediastinal radiation, chest wall deformity, and potential for injury to previous bypass graft on sternal reentry.¹⁸ Cirrhosis, lung disease, and frailty are other medical conditions that when present are often thought to imply excessive surgical risk. Patients undergoing TAVR secondary to anatomic and advanced medical comorbidities have significant survival benefit when compared to standard therapy.¹⁸ Patients with bicuspid aortic valves have been excluded from most TAVR registries and trials. However, there are data to suggest that TAVR can be performed with good clinical outcomes in patients with bicuspid aortic valves.¹⁹ Patients with moderate dementia who would not fare well after cardiopulmonary bypass can be considered for TAVR. Patients who cannot afford to have a prolonged recovery after SAVR due to ongoing chemotherapy, need for additional operations, and upcoming transplantation can be considered for TAVR. Consideration of such factors outside of formal risk prediction models makes it crucial to allow assessment of anticipated risk for SAVR or TAVR by an experienced heart team.

Perhaps the 2 most significant challenges to TAVR are the larger size of the delivery sheaths and the reliance on indirect visualization of the heart. To overcome these obstacles and perform safe TAVR, meticulous preprocedure planning and intraprocedure imaging are needed. Evaluation prior to TAVR should seek to determine several factors, including (1) iliofemoral vessel size, calcification, and tortuosity; (2) anatomic details of the aortic valve leaflets; (3) annulus, sinotubular, and sinus of Valsalva dimensions; and (4) distance from the aortic annulus to the coronary ostia.

TAVR was initially conceived as a fully percutaneous procedure performed through the iliofemoral vessels without the need for a vascular cutdown. The transfemoral (TF) approach is the preferred access route for TAVR because it is less invasive than alternative routes. For patients without adequate iliofemoral vessels, operators have investigated more novel routes for delivery of the SAPIEN bioprosthesis. The SAPIEN device can be deployed via a transapical (implantation through the cardiac apex after a lateral thoracotomy), transaortic (direct cannulation of the ascending aorta with a mini-sternotomy), transcarotid (following a cutdown with Javid shunt), transcaval, or transseptal approach.^6,20 In earlier Edwards SAPIEN registries, 30% to 50% of patients required an alternative approach due to peripheral artery disease or vessels unable to accommodate the larger delivery sheaths available at the time. In more recent reports using the 16/18/20 Fr SAPIEN XT and 14/16 Fr S3 delivery sheaths, an alternative access site has been used in 7% to 10% of cases.^21,22

To determine which patients are suitable for a true percutaneous approach and which would be best served by an alternative access approach, we propose multimodality imaging assessment as an important step in planning TAVR. Imaging guides both access route and device sizing.

Echocardiography

Transthoracic echocardiography (TTE) is the initial imaging modality of choice due to its ease and safety. TTE is important to establish the diagnosis of AS and valve morphology (ie, bicuspid vs trileaflet valve), assess aortic annulus size, and evaluate left ventricular function and other concomitant valvular disorders, right ventricular dysfunction, and pulmonary hypertension. TTE is useful in identifying the location of any large, calcific nodules, their relationship to the coronary ostia, and calcium that tracks into the left ventricular outflow tract (Fig. 42-2). Precise sizing of the aortic annulus is crucial for proper valve prosthesis selection and in order to avoid complications of inappropriate sizing (see section on computed tomography [CT]). TTE tends to underestimate true open surgical sizing and transesophageal echocardiography (TEE) may be helpful to provide better assessment of the annulus (Fig. 42-3).^23,24 Balloon sizing of the annulus during balloon aortic valvuloplasty may be particularly helpful when the annulus is questionably too large for a given valve; supravalvular aortography is performed during balloon aortic valvuloplasty with rapid pacing, and any leakage of contrast into the left ventricle suggests that a larger valve will be needed (Fig. 42-4).²⁵ Invasive techniques to assess aortic annular sizing may not be necessary with the advent of cross-sectional 3-dimensional TEE and CT imaging.²⁶

Computed Tomography

CT provides a wealth of cardiac, thoracic, and peripheral information in assessing a patient’s feasibility for TAVR and procedural planning.²⁶ A CT scan will show in detail the aortic root anatomy (including annular, sinotubular, and sinus of Valsalva dimensions as well as coronary heights above the annulus), aortobifemoral anatomy, and details of arterial diameters and calcification. A gated, contrast CT with less than 1-mm slices will allow detailed 3-dimensional reconstruction, which can be invaluable for evaluation of the aortic annulus and peripheral arteries. Given the ellipsoid shape of the aortic annulus, 2-dimensional imaging techniques, such as TTE, may not accurately represent the size of the annulus (Fig. 42-5). Three-dimensional gated CT imaging of the aortic root may provide even more accurate assessment than surgical sizing.²⁷

An appreciation of the complex, 3-dimensional, and nearly uniformly ovoid shape of the aortic annulus is important for appropriate prosthesis sizing.²⁸ Inappropriate prosthesis size selection is associated with paravalvular regurgitation, aortic annular rupture, coronary occlusion, and device embolization. CT annular area measurements are usually performed in systole at 25% or 35% of the RR interval when the annulus is largest.²⁹ Nominal external areas for the SAPIEN 23- and 26-mm valves are 415 and 531 mm², respectively, and for the SAPIEN XT 23-, 26-, and 29-mm valves, they are 415, 531, and 661 mm², respectively. Area oversizing of 8% to 10% is considered optimal; however, most patients will not meet this stringent target due to the large increments between prostheses sizes, and therefore, a range extending up to 20% oversizing is acceptable in proposed sizing algorithms.³⁰ For borderline aortic annular dimensions, intentional underexpansion of balloon-expandable valves by reducing the balloon volume by 10%, which results in an intermediate inflow diameter between the fully expanded valve and the next smaller transcatheter valve, has been proposed.³⁰ The S3 valve incorporates a paravalvular sealing system, and the target annular oversizing range is 5% to 15% (see Table 42-3 for nominal area for the 20-, 23-, 26-, and 29-mm valves). If oversizing by 20% to 30% is anticipated, underfilling the deployment balloon by 1 to 3 mL or undersizing the prosthesis by up to 5% should be considered.³¹

Another application of cardiac CT is to establish a fluoroscopic projection that defines the precise aortic valve plane that is perpendicular to the axis of implantation and allows optimal deployment of the valve prosthesis. By using CT to determine the “deployment angle,” operators can avoid performing multiple aortic root angiograms. Additionally, CT imaging provides information such as leaflet length, leaflet calcification, and height from the annulus to the coronary ostia. These factors may predict the complication of coronary occlusion during TAVR. Generally, a distance from annulus to ostium of 11 to 14 mm is considered reasonable, and less than 7 mm is possibly prohibitive for TAVR, particularly with narrow sinuses of Valsalva (see Table 42-1). Adverse root features, including more than minimal left ventricular outflow tract calcification, increase the risk for aortic rupture and paravalvular regurgitation.³²

CT assessment of the minimum diameter of the iliofemoral vasculature is important in assessing the feasibility of TF TAVR. The Edwards SAPIEN bioprosthesis requires a minimum diameter of 8 mm throughout the iliofemoral system for the 26-mm valve and of 7.3 mm for the 23-mm valve (see Table 42-1). The outer diameter of the 22- and 24-Fr Edwards delivery sheaths are 8.4 and 9.2 mm, respectively (see Table 42-1). The SAPIEN XT requires a minimum of 6.0 mm for the 23-mm valve, 6.5 mm for the 26-mm valve, and 7.0 mm for the 29-mm valve (see Table 42-2). The S3 requires a minimum of 5.5 mm for the 23- and 26-mm valves and 6.0 mm for the 29-mm valve (see Table 42-3). It is also important to assess the degree of tortuosity and calcification of the pelvic vasculature. Extensive calcification in relatively straight common iliac arteries maybe acceptable as long as the dimension is larger than the outer diameter of the sheath. Moderate calcification in tortuous external iliac arteries (which frequently dive deep into the pelvis) may increase the likelihood of major vascular complications. A noncalcified artery will often stretch 1 mm without major complication using a nonexpandable sheath, although dissection can still occur postprocedure. With the e-Sheath, we have seen noncalcified arteries temporarily stretch >2 mm without complication. Contrast-enhanced CT is preferred in the assessment of the peripheral vasculature, but in patients with significant renal insufficiency, a noncontrast study may be adequate.

Angiography

Invasive coronary and bypass graft angiography is also needed to evaluate anatomy suggestive of a large burden of ischemic myocardium; however, complete revascularization may not be necessary prior to TAVR.³³ An invasive, digitally subtracted distal abdominal aortobifemoral angiogram using a marked pigtail catheter can be complimentary to CT and very helpful in defining the size and course of the iliofemoral system. In patients with borderline diameter of the pelvic vessels, intravascular ultrasound may be helpful to determine candidacy for the TF approach.³⁴

Carotid Ultrasound

Carotid duplex ultrasound may identify severe, flow-limiting carotid disease that may increase the risk of stroke associated with significant cardiogenic shock during TAVR, although this is exceedingly rare with TF TAVR. Percutaneous or surgical carotid revascularization may be considered prior to TAVR; however, procedure-related hypotension may be devastating and, thus, this procedure is not recommended by our center. We also measure the diameter of the common carotid artery in addition to the disease burden of the carotids to determine candidacy for a transcarotid approach if no other access is feasible or safe.

TF TAVR is commonly performed under general anesthesia but may be safely performed in experienced centers using only local anesthesia and minimal or no sedation.¹⁴ The advantage of general anesthesia is that operators early in their experience can perform TAVR without being rushed or concerned about airway management if a complication occurs. TAVR performed with conscious sedation allows for quicker recovery and hospital discharge of the patient. It also decreases the utilization of resources for TAVR, which may have important economic implications.

The following text describes the practical aspects of TF TAVR for patients with native valve AS using the Edwards SAPIEN valve.

Antibiotics

Routine antibiotic prophylaxis should be administered immediately prior to the procedure. Often, the same antibiotics that are used perioperatively for open surgical aortic replacement are used (eg, cefuroxime). Drugs effective against gram-positive skin flora such as cefazolin and vancomycin (in cases of penicillin allergy) are appropriate when a fully percutaneous approach is used. Antibiotic prophylaxis for the implanted bioprosthesis is indicated with subsequent dental procedures according to intersociety guidelines.³⁵

Patient Preparation

Patients are prepared and draped in the supine position with exposure from the upper thigh through the subxiphoid area, which should be available if emergent pericardiocentesis is required. An adhesive, iodophor-impregnated incise drape may be helpful in preventing access site infection.

Intraoperative Echocardiography

Intraoperative echocardiography may be transesophageal or transthoracic and is helpful in positioning the new valve prior to deployment as well as in evaluating postdeployment valve function.³⁶ In the setting of hypotension after valve deployment, echocardiography is critical in assessing for valvular regurgitation, pericardial effusion, left ventricular dysfunction, and injury to the aortic root. Aortic regurgitation, if present, should be identified as transvalvular or paravalvular, quantified, and treated appropriately; management of aortic regurgitation will be discussed later in this chapter.

Special Instruments

• 
  

  “Micropuncture” access kits that use a 21-gauge needle and 0.018-inch wire for arterial entry facilitate fluoroscopy-guided access.

  
  
• 
  

  Uninterrupted capture during rapid ventricular pacing is critical to stable valve deployment, and for this reason, a Mullins (transseptal) sheath may be placed across the tricuspid valve to support the temporary pacing wire.

  
  
• 
  

  Suture-based, percutaneous arteriotomy closure devices such as the Perclose ProGlide and ProStar XL (Abbott Vascular, Abbott Park, IL) are needed for fully percutaneous TF TAVR procedures.

  
  
• 
  

  When TF TAVR is performed without an anesthesiologist, prepared 10-mL syringes of norepinephrine (16 μg/mL) and epinephrine (0.1 mg/mL) are helpful when immediately available to operators for as-needed treatment of hypotension.

Antithrombotic Therapy

Unfractionated heparin should be used after arterial access is obtained to maintain an activated clotting time (ACT) of 250 to 300 seconds. Heparin administration based on baseline ACT levels results in a significantly lower occurrence of major bleeding with TAVR.³⁷ Direct thrombin inhibitors have been used for those with heparin allergies. We routinely administer a loading dose of clopidogrel prior to TF TAVR based on anecdotal evidence. Dual antiplatelet therapy with aspirin and clopidogrel is indicated after TAVR to prevent thromboembolic complications, but there is no consensus on the indicated duration of therapy.¹¹ Six months of therapy with aspirin and clopidogrel after TF TAVR are reasonable in most patients, and aspirin is indicated indefinitely thereafter. Clopidogrel can probably be omitted for patients who will be therapeutically anticoagulated with warfarin.

Arterial and Venous Access

In addition to femoral arterial access for the valve’s delivery sheath, a second (typically 6 Fr) arterial access is needed for insertion of a pigtail catheter into the aortic root. Venous access is needed for insertion of the temporary pacing wire.

Device Preparation

The family of SAPIEN valves uses dedicated delivery systems, and the balloon inflation lumen is connected to an Atrion inflation device (Atrion Medical, Arab, AL). The balloon ends in a nose cone, which facilitates crossing of the stenotic native valve. A crimping tool is a compression device that symmetrically collapses the valve over the delivery balloon catheter. The crimper is composed of a crimp aperture, which allows compression by means of a rotary knob; a balloon gauge, which verifies the delivery balloon diameter at full inflation; and a crimp gauge to verify the size of the balloon/valve assembly and to ensure that the delivery system has been suitably crimped. The newer generation SAPIEN valves (XT and S3) do not require measurement of the delivery balloon prior to crimping as a predetermined volume in the balloon is adequate. The entire system needs to be de-aired carefully. It is important to always check to see if the valve is mounted correctly on the balloon (fabric skirt toward the nose cone for the TF approach). The SAPIEN valves can be fully crimped for 15 minutes before insertion without compromising subsequent valve performance. Preparation of the delivery system and valve should be timed appropriately with the flow of the case.

Step by Step Procedure

After appropriate anesthesia, arterial and venous accesses are obtained on the side contralateral to where the delivery sheath will be inserted. A temporary pacing wire is advanced to a stable position in the right ventricular apex, tested, and secured. The wire may be placed through a Mullins sheath (positioned across the tricuspid valve) for stabilization. The most stable position for the pacer lead is at the right ventricular apex. Placement of the lead in the base of the right ventricle free wall can lead to loss of capture or perforation (Fig. 42-6).

For fully percutaneous procedures, a 6-Fr Judkins right number 4 or similar catheter is then advanced to the aortic-iliac bifurcation, where it is used to selectively engage the contralateral iliac artery and create a digital “roadmap” of the femoral artery where the delivery sheath will be inserted. Using this image, a 21-gauge needle is placed coaxially into the center of the common femoral artery under real-time fluoroscopic guidance. Using the CT scan for preprocedure planning, the location of the puncture can be selected to avoid the superficial femoral artery/profunda bifurcation and avoid a high arteriotomy through the inguinal ligament. Proper location of this puncture is critical to facilitate subsequent arteriotomy closure. Areas of heavy calcification are purposefully avoided. A 6- or 7-Fr sheath is inserted. Surgical cutdown to the planned femoral access site or external iliac conduit may alternatively be performed (Fig. 42-7).