Surgical bioprosthetic valves, when compared to mechanical valves, may prevent the need for long-term anticoagulation. However, progressive structural deterioration commonly develops 10 to 15 years after implantation, manifesting with stenosis, regurgitation, or a combination of both. Conventional treatment of patients with failed bioprosthetic valves is surgical reoperation. In these often high-risk patients, an attractive alternative to a “redo” surgical valve procedure is transcatheter aortic valve replacement (TAVR) of the failed bioprosthetic valve (ie, a valve-in-valve [VIV] procedure). In this article, we describe the potential benefits and challenges of this new and exciting interventional therapy.

BIOPROSTHETIC VALVES

Treatment of failed surgical valves requires familiarity with the structure of these valves and their fluoroscopic markers, in addition to a deep understanding of basic concepts (Figure 1). Surgical bioprostheses are most commonly xenografts; the vast majority of these are fashioned from either bovine pericardium or porcine aortic leaflets. These valves may be classified by their design: stented valves, which include a frame attached to a basal ring covered by a fabric sewing cuff, and stentless bioprostheses. These surgical valves may be positioned at the level of the annulus, or above it, in order to maximize the effective orifice area (supra-annular).

In some stented bioprostheses, such as Mitroflow (Sorin SpA, Milano, Italy) and Trifecta (St. Jude Medical, Inc., St. Paul, MN), valve leaflets are externally mounted to the stent. Bioprostheses are often characterized by label size, which is commonly the external diameter of the frame. A different measure is surgical valve internal diameter, a more relevant measure when planning a VIV procedure. These sizes may be provided by manufacturers' charts or websites.1,2 These characteristics and others have a profound clinical impact on VIV procedures.

Bioprosthetic valve degeneration is time related. At 10 years, structural failure is on the order of 10% to 30%, whereas at 15 years, it increases to 30% to 60%.3-5 The mode of failure is commonly stenosis (attributed to calcification, pannus, or thrombus formation), regurgitation (due to wear and tear, calcification, or infection), or a combination of both stenosis and regurgitation. Patients with degenerated surgical valves frequently present with recurrence of heart failure symptoms, deteriorating quality of life, and recurrent hospital admissions.

AORTIC VIV AND CLINICAL OUTCOMES

Until several years ago, there was no effective treatment for patients with failed bioprostheses who were considered inoperable. Some attempts of balloon valvuloplasty inside stenotic bioprostheses resulted in severe regurgitation or were followed by an emergent surgical procedure. The first VIV cases were performed in 2007 in Canada and Germany by implantation of a CoreValve device (Medtronic, Inc., Minneapolis, MN) inside a failed Mitroflow valve, as well as transfemoral and transapical implantations of Sapien devices (Edwards Lifesciences, Irvine, CA) in failed Carpentier- Edwards Perimount devices (Edwards Lifesciences). Currently, most aortic VIV experience has utilized either the self-expanding CoreValve or the balloon-expandable Sapien/Sapien XT valves.

The VIVID (VIV International Data) registry, also known as the “global VIV registry,” is an industry-independent collaboration that was introduced in 2010 to collect data from this widely distributed VIV experience. Currently, this registry includes more than 600 cases collected from over 60 centers worldwide.6 In the global registry, 61% of cases were performed with CoreValve and 39% with Sapien valves (Figure 2). Analysis of data from the global registry reveals that high-risk patients undergoing VIV procedures have 30-day death rates similar to TAVR for native aortic stenosis (9.6%).6 In addition, the procedure is clinically effective in the majority of patients, leading to marked improvement in symptoms and functional class (New York Heart Association class I/II in 84%) and a significant reduction in mean transvalvular gradients (16 ± 9 mm Hg), with no or trivial regurgitation in the majority of patients.6 The 1-year survival was 85.8%.

In addition, neurological adverse events are not as prevalent in real-world experience. Fortunately, the stroke risk has been lower than anticipated, with a 2% rate of major stroke in the global registry. It seems that several specific adverse events (ie, annulus rupture, aortic dissection, cardiac tamponade, significant regurgitation, and cardiac conduction defects) are less common after VIV procedures compared to native aortic valve TAVR (see the Complications of Aortic Valve Implantation sidebar). It is possible that the sewing ring of a surgical bioprosthesis protects the surrounding structures from injury. Nevertheless, VIV procedures have unique challenges and technical considerations that need to be examined carefully.

CHALLENGES AND LIMITATIONS OF THE VIV APPROACH

Malposition

As opposed to native aortic stenosis, in which bulky calcification provides adequate fluoroscopic markers to guide transcatheter heart valve (THV) positioning, degenerated surgical bioprostheses often have only mild leaflet calcification and, in addition, may also have limited fluoroscopic markers. Especially problematic are stentless surgical valves, the vast majority of which have no fluoroscopic markers. The risk of malposition, either more aortic or more ventricular, is a concern, as it was described in approximately 15% of patients in the global VIV registry and resulted in the need for attempted CoreValve device retrieval (8.9%) and implantation of a second THV device (8.4%).6

This issue can be addressed with use of additional imaging and, in particular, with real-time transesophageal echocardiography. Other well-known techniques that may facilitate correct positioning include locating a pigtail catheter within the noncoronary sinus, as well as multiple contrast injections during deployment. It is crucial to study images from previous VIV implantations in a similar setting. Operator inexperience and an incomplete understanding of the procedure might contribute to many of these events. The development of newer THV devices that may be resheathable and repositionable could offer an advantage in dealing with this complication.

Coronary Obstruction

Perhaps the most dreaded complication during VIV procedures is ostial coronary occlusion. This has been described in 3.5% of the VIV registry, mostly involving the left coronary artery and manifesting with immediate hemodynamic collapse.6 However, delayed presentation was described as well. The etiology may include displacement of surgical valve leaflets in a tubular fashion, covering the coronary ostium or the sinotubular junction over the ostia. Risk factors include stentless surgical valves with supra-annular designs, low-lying coronary ostia, narrow aortic root and sinotubular junction, severely stenosed bioprostheses with bulky leaflets (rather than regurgitant), previous root reconstruction, reimplanted coronaries, and stented valves with externally mounted leaflets. The technical approach in high-risk cases includes positioning of an angioplasty guidewire within the coronary at risk before valve deployment, with a balloon or stent ready to be deployed, if needed.

Prosthesis Underexpansion

Ideal THV function requires optimal prosthesis expansion. When deployed within a surgical bioprosthesis, the THV device is often constrained by the rigid annulus. Thus, although average mean gradients after native TAVR range between 5 to 15 mm Hg, these tend to be higher with VIV, on the order of 10 to 25 mm Hg. Data from the global VIV registry reveal that postprocedure gradient depends on the mechanism of failure of the surgical valve, the size of the surgical valve, and the type of THV prosthesis.6 There were higher postprocedural gradients in stenotic bioprostheses and in cases in which a Sapien device was implanted inside small surgical valves. The elevated postprocedural gradient (mean ≥ 20 mm Hg) after VIV implantation in a small surgical aortic valve (inner diameter < 20 mm) was seen in 58% of Sapien VIV procedures versus 20% of CoreValve VIVs. This dissimilarity is attributed to the fact that, usually, the CoreValve functional part is located above the annulus (supra-annular). It is anticipated that with the introduction of the 20-mm Sapien XT valve, as well as with the development of more specific THV prostheses dedicated for VIV procedures and improved patient screening, the incidence of high postprocedure gradients may decline. The THV underexpansion seen in many VIV cases causes prosthetic-patient mismatch. Suboptimal THV hemodynamic performance, impaired leaflet coaptation, and abnormal leaflet-frame contact my all lead to lower THV valve durability.

EVALUATING CANDIDATES FOR VIV

The risk for undergoing conventional surgery should be assessed, and disease severity should be confirmed using echocardiography, coronary evaluation, and peripheral vascular assessment. Although screening is mostly similar to that in patients with native aortic stenosis for conventional TAVR, there are specific issues to address for VIV procedures.7 It is crucial that the operators be familiar with the particular characteristics of the surgical bioprosthesis. Frequently, obtaining a detailed operative report is helpful. Several publications can assist in understanding the fluoroscopic appearance of various bioprosthetic valves and the implications for THV positioning.1,2

Results from previous echocardiographic examinations should be obtained, and isolated patient-prosthetic mismatch of the surgical valve should be excluded. Patients with small surgical valves and high, but stable, gradients since the original surgical procedure should probably not be referred for VIV. Additionally, patients with regurgitant bioprostheses secondary to active endocarditis and paravalvular leakage should be excluded, mainly by using transesophageal echocardiography. An essential step in screening patients for a VIV procedure is evaluating the risk for coronary obstruction. Imaging modalities such as echocardiography, aortography, and CT are useful. Aortic root injection in a left anterior oblique/cranial projection is also very helpful in that it may show the proximity of the mobile bioprosthetic leaflets to the left coronary ostium.

SUMMARY AND FUTURE PERSPECTIVES

The majority of surgical heart valves currently implanted are bioprosthetic tissue valves. Because these valves have limited durability, and a redo surgical procedure is considered a high-risk operation, we can speculate that VIV, as a less-invasive approach, will be increasingly performed. The efficacy of VIV implantations is similar to standard TAVR procedures. However, the procedure is technically more demanding and requires additional careful planning and anticipation of possible complications; thus, it should be reserved for experienced centers/physicians. Attention should be given to specific adverse events that are more common after VIV, mainly THV device malposition, ostial coronary obstruction, and elevated postprocedural gradients.

We can anticipate that the rate of device malposition and coronary obstruction will decrease with improved patient screening and operator experience. However, the issue of elevated postprocedural gradients may have no resolution until more specific devices are designed for this clinical setting. The use of small THV devices, such as the 23-mm CoreValve Evolut and the 20-mm Sapien XT, will be examined. It should be noted that residual, moderate stenosis after VIV procedures may be a reasonable outcome in patients who are considered inoperable. The major limitation of the VIV approach may be THV durability. Thus, VIV seems promising, but it may take several years until we can fully appreciate the efficacy and limitations of this approach for patients with failed bioprostheses.

Konstantinos Spargias, MD, is with Hygeia Hospital in Athens, Greece. He has stated that he has no financial interest related to this article.

Danny Dvir, MD, is from the Catheterization Laboratory, St. Paul's Hospital in Vancouver, Canada. He has stated that he has no financial interest related to this article. Dr. Dvir may be reached at danny.dvir@gmail.com.

  1. Gurvitch R, Cheung A, Ye J, et al. Transcatheter valve-in-valve implantation for failed surgical bioprosthetic valves. J Am Coll Cardiol. 2011;58:2196-2209.
  2. Bapat V, Mydin I, Chadalavada S, et al. A guide to fluoroscopic identification and design of bioprosthetic valves: a reference for valve-in-valve procedure. Catheter Cardiovasc Interv. 2013;81:853-861.
  3. Jamieson WR, Munro AI, Miyagishima RT, et al. Carpentier-Edwards standard porcine prosthesis: clinical performance to seventeen years. Ann Thorac Surg. 1995;1995:999-1006.
  4. Jamieson WR, Miyagishima RT, Burr LH, et al. Carpentier-Edwards porcine prosthesis: clinical performance assessed by actual analysis. J Heart Valve Dis. 2000;9:530-535.
  5. Ruel M, Kulik A, Rubens FD, et al. Late incidence and determinants of reoperation in patients with prosthetic heart valves. Eur J Cardiothorac Surg. 2004;25:364-370.
  6. Dvir D, Webb J, Brecker S, et al. Transcatheter aortic valve replacement for degenerative bioprosthetic surgical valves: results from the global valve-in-valve registry. Circulation. 2012;126:2335-2344.
  7. Webb JG, Dvir D. Transcatheter aortic valve replacement for bioprosthetic aortic valve failure: the valve-in-valve procedure. Circulation. 2013;127:2542-2550.