Article Text

Percutaneous aortic valve replacement: will we get there?
  1. Service de Cardiologie Pédiatrique
  2. Hôpital Necker Enfants Malades
  3. 149, rue de Sèvres
  4. 75743 Paris Cedex 15
  5. Paris, France
  1. Dr Philipp Bonhoeffer; philipp.bonhoeffer{at}

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Percutaneous implantation of a biological valve in the pulmonary position has recently enlarged the field of interventional cardiovascular catheterisation.1 Aortic valve disease is much more frequent than pulmonary valve disease and therefore a non-surgical approach for aortic valve replacement would be attractive. The percutaneous approach is obviously difficult because of the proximity of the aortic valve to the coronary orifices. We set out to develop a stent for valve implantation which avoids obstructing the coronary orifices. We describe here our initial experience using the new device for percutaneous valve replacement of the aortic valve in an animal study.

A newly designed stent with a deployment strategy in two steps was developed. The first step assures the orientation of the stent in the aortic orifice. The second acts as a support for the implantable valve. A naturally valved bovine jugular vein segment was mounted as previously described inside this stent (Numed, Osypka).2The venous wall was then reduced along the commissures in order to remove the wall in front of the coronary orifices (fig 1).

Figure 1

Left: venous valve before its suture in the supporting structure. Right: valve mounted in the platinum stent.  

The valved stent was first reduced on the outer balloon of the delivery system. The stent was then percutaneously inserted into a lamb through the right carotid onto a previously positioned guide wire. In the left ventricle, the valved stent was uncovered deploying the outer part of the stent. The system was thereafter pulled back in the area of the native aortic valve. The self centering mechanism allowed alignment of the commissures of the native valve with the commissures of the valve to be implanted. After successful centering, the balloons were successively inflated deploying the valved stent to the final diameter of 22 mm. The balloons were deflated and the delivery system was carefully retrieved, leaving the stent in position.

Transoesophageal echocardiography, and haemodynamic and angiographic evaluations were performed before and after the implantation and at the two week follow up. Data obtained confirmed the perfect function of the implanted valve and the absence of coronary artery obstruction (fig 2). There was no complication at clinical follow up after four weeks. During the entire study, the lamb was treated according to European regulations for animal experimentation.3

Figure 2

Left: aortic and ventricular view before angiography. Right: aortic angiography showing normal coronary arteries and competence of the implanted valve.

Initially, we wrongly thought that the venous wall was necessary for the valve to remain competent. Our in vitro and in vivo testing showed that its removal along the commissures did not alter the function of the valve. This new valved stent design theoretically allows an orthotopic valve replacement but its precise placement is delicate. The valved stent must be precisely placed in height. An implantation above the valve near to the coronary orifices is not suitable for their perfusion. An implantation below the coronary arteries, encroaching upon the left ventricle, can impair the function of the mitral valve and enhance the risk of paraprosthetic insufficiency. The high pressures in the left system complicate the situation by increasing the risk of embolisation in the ascending aorta during the inflation of the balloon catheter or in the left ventricle at the closure of the implanted valve. This emphasises the need to anchor the stent in the precise position. We resolved these problems with the new design of our stent.

The device as it is presently designed does not allow multiple attempts before definitive deployment. One can imagine what could happen if the device obstructed the coronary arteries. The hooks reduce the risk of malpositioning but their safety must be confirmed. In humans transoesophageal echocardiography could help to position the device during the procedure. The reproducibility of the technique needs to be proven before clinical application. Finally, the long term function of our venous valve is not yet established.

Both aortic regurgitation and aortic stenosis can potentially be treated using this technique. In the indication of aortic insufficiency, a possible limitation would be the size of the aortic root usually dilated. In aortic stenosis the calcifications and potential calcium emboli during dilatation are a major concern. The size of the delivery system might require balloon predilatation. The implantation of a stent thereafter is likely to improve the results of the balloon valvuloplasty in terms of valve area. Presumably this could improve the long term results.

Who should be the first patients to benefit from this new technique? Patients who require aortic valve replacement and who are at the highest risk of potential complications of mechanical valves are likely to primary candidates. These would include patients with high operative risks and patients who have contraindications to anticoagulation. For example, it could be a temporary solution for young women with aortic valve disease who wish to become pregnant.

The anatomy of the aortic area in lambs is very similar to humans, providing hope for the use of such a device in humans in the future. We have implanted a 22 mm venous valve which when fitted on the delivery system measured 20 French. Bigger valves are available and would not dramatically increase the size of the whole system, making percutaneous aortic valve replacement in the human possible.


This work has been supported by the grants of : “La Fondation de l'Avenir” and “La Federation Francaise de Cardiologie”.