Initial attempts at percutaneous aortic valve replacement (AVR) were compromised due to the difficulties associated with vascular passage of a relatively bulky prosthetic valve compressed onto a standard valvuloplasty catheter (1- 2). Subsequent technical and procedural improvements allowed the development of a more predictable therapy (3). Although outcomes continue to improve, limitations (aortic injury, atheroembolism, difficulty crossing the native valve) remain apparent (4). We describe a new delivery system and next-generation balloon-expandable valve in a case series of 25 high-risk patients undergoing transarterial AVR.

Percutaneous AVR was approved for clinical use in patients with symptomatic aortic stenosis in whom the risk associated with open heart surgery was considered prohibitive by a team of cardiologists and cardiac surgeons. Written informed consent was obtained. Patients were excluded if the diameter of the aortic annulus was <18 or >26 mm as assessed by transesophageal echocardiography, if there was severe iliofemoral arterial disease, or if a reasonable quality or duration of life was considered unlikely.

Procedures were performed in a catheterization laboratory under general anesthesia. Percutaneous femoral artery puncture was utilized followed by percutaneous suture “pre-closure” (Prostar, Abbott, Boston, Massachusetts) or surgical cutdown. The basic procedure of transarterial AVR has been previously described (3,4).

The RetroFlex 2 delivery system (Edwards Lifesciences, Irvine, California) incorporates a coaxial nose cone catheter, a balloon catheter, and a deflectable pusher catheter (Figure 1). The catheter is advanced through the sheath and into the aorta. An articulation control knob allows steering around the aortic arch and through the central orifice of the stenotic native valve. The deflectable pusher catheter is retracted, and the nose cone is advanced so as to expose the deployment balloon (Figure 2). The prosthesis is positioned within the native valve. The deployment balloon is inflated and deflated while right ventricular burst pacing is used to temporarily reduce transvalvular flow. The balloon is utilized to guide the withdrawal of the nose cone through the newly implanted valve. The nose cone and pusher catheter are approximated and removed through the delivery sheath.

Initial reports of balloon-expandable valve replacement utilized a valve constructed from a stainless steel tubular frame with equine pericardial leaflets and a fabric sealing cuff (Cribier-Edwards, Edwards Lifesciences). This basic valve design was subsequently modified utilizing bovine pericardial leaflets and lengthening the sealing cuff without modification of the frame (SAPIEN, Edwards Lifesciences).

A next-generation transcatheter heart valve (SAPIEN XT, Edwards Lifesciences) was utilized in 3 patients. A cobalt-chromium frame permits thinner struts without loss of structural integrity. Thinner struts and a more open design allow for a lower crimped profile (Figure 3), while maintaining the valve's radial stiffness. Bovine pericardial leaflets are matched for thickness and elasticity and incorporate Thermafix anticalcification treatment. The scalloped geometry and attachment method of the leaflets have been modified to achieve a naturally closed design and enhance valve durability (Figure 4).

The femoral arterial sheath required is determined by the crimped diameter of the prosthetic valve and its delivery system. A 23-mm SAPIEN valve crimped on the RetroFlex 2 delivery catheter requires a 22-F internal diameter sheath, whereas a 26-mm SAPIEN valve requires a 24-F sheath. When crimped on the current RetroFlex 2 catheter, the 26-mm next-generation low-profile valve allows use of a 22-F sheath.

Continuous variables are presented as mean ± SD, medians, and interquartile ranges (IQRs), as appropriate. Categorical variables are stated as frequencies and percentages. Normality was tested with the Kolmogorov-Smirnov goodness-of-fit test. For comparison of continuous variables before and after transarterial AVR, the Student t test was used, and for comparison of groups, a 1-way analysis of variance F test was utilized. Categorical variables were compared by the Fisher exact test. A 2-sided value of p < 0.05 was considered statistically significant.

The RetroFlex 2 system was utilized in 25 patients. Mean age was 85 years (IQR: 79 to 88 years). Patients had multiple comorbidities, including prior thoracotomy (28%), renal insufficiency (40%), porcelain aorta (12%), grade 3 or 4 mitral regurgitation (12%), severe lung disease (24%), and severe frailty (40%). Baseline characteristics are presented in (Table 1).

Procedural success, defined as implantation of a prosthesis at the intended site, was achieved in all patients. Tortuosity of the thoracic aorta, horizontal aortic roots, and severely stenotic, heavily calcified valves, although problematic with earlier delivery systems, were relatively easily managed with the new delivery system. Percutaneous needle puncture was utilized for femoral arterial access in all patients and percutaneous suture closure in the majority. Outcomes are presented in (Table 2). Society of Thoracic Surgeons and logistic EuroSCORE estimates of operative 30-day mortality were 8.9% and 21.0%, respectively. All patients remained alive at 30 days.

The SAPIEN valve was utilized in 22 patients and the next generation low profile cobalt-chromium SAPIEN XT valve in 3. At discharge, echocardiographic aortic mean gradient had fallen from 49.3 ± 17.9 mm Hg to 10.6 ± 2.9 mm Hg and effective orifice area had risen from 0.59 ± 0.15 cm² to 1.6 ± 0.27 cm² (Table 3). No patient had more than mild valvular regurgitation, and only 1 patient had more than mild paravalvular regurgitation (graded moderate). All patients had normal prosthetic valve function at 1-month echocardiographic follow-up.

Transcatheter valve replacement continues to evolve rapidly. In the initial experience with transarterial AVR, procedural success was achieved in 78% of patients with a 30-day mortality of 11% (3). The current success rate of 100% and 30-day mortality of 0% in 25 high-risk patients is encouraging. Technological and procedural improvements as well as increasing experience likely share responsibility for these improving outcomes.

Although mortality was not observed, morbidity was. The major source of morbidity was related to vascular access. Pre-procedural screening and improved vascular management appear to be reducing vascular morbidity and mortality risk (4). Consequently, a major thrust of current development is to reduce the diameter of the valve and delivery system. The development of a new, thinner cobalt-chromium SAPIEN XT valve frame with a more open design allows for tighter crimping of the valve and an improved delivery profile. Further modification of the delivery system to accommodate this next-generation valve is expected to reduce the delivery profile to 18- or 19-F (Figure 5). Such a reduction in profile will likely increase the population of patients amenable to this therapy and improve outcomes further. The thinner struts of the next-generation cobalt-chromium low-profile device did not appear to result in a reduction in orifice area (Figure 6). We have not observed structural valve failure in our larger experience with almost 200 valve implants. Comparable or improved durability with the next-generation valve is anticipated by in vitro testing but remains to be clinically demonstrated.

Advances in transcatheter valves, delivery systems, and procedural technique may result in improved outcomes and suggest the potential for a broader role of this therapeutic modality.