Rick De Graaf, Department of Radiology, Maastricht University Medical Centre, Maastricht, The Netherlands, presented the results of the first comparative study of four dedicated stents for the venous system at the VEITHsymposium (17–20 November 2015, New York, USA). De Graaf spoke to Vascular News about the study, his views on current devices, and the characteristics of the ideal venous stent.
Why is it important to compare the differences between the venous stents currently available on the market?
It is important because of all the different characteristics of the venous anatomy in relation to the arterial anatomy. The stent manufacturers produce venous stents mainly based on stents that they already have for the arterial anatomy, and without working on a new design based on anatomical demands, such as constant angles and the movements inside the pelvis during ambulation. It is important to see how these stents behave over time. Sometimes the stent has a tendency to straighten and because the movement of the vein is very subtle, the tendency of the stent to become straight and pull the vein with it is potentially harmful. The idea with the study involving 3D analysis was to see how the stents performed over time—if they would straighten, kink, fracture or be significantly compressed. At the moment nobody knows how these stents behave in the venous system over an extended period of time.
How was the study conducted?
We used stents that have manufacturers’ recommendation to be placed in the venous anatomy—Vici (Veniti), Zilver Vena (Cook Medical), sinus-Venous (Optimed) and sinus-XL Flex (Optimed). Ten consecutive stents of each manufacturer implanted for post-thrombotic disease and May-Thurner disease were analysed, according to a similar protocol: pre-dilatation with a 16mm non-compliant balloon, up to the external iliac vein, and then with a 14mm balloon down to the healthy vein segment, usually at the level of the common femoral vein. Directly post-stenting, a 3D cone-beam computed tomography scan was performed by which stent angles and surface area could be measured, respectively, at high flex points and high pressure points. These measurements were repeated at six and 12 months with duplex X-rays in four directions. Stent integrity, ie. stent fractures, was also assessed.
With these measurements we could compare the angles over time, and assess if there was any compression at the hard pressure points or if there was kinking or straightening at the more severe flex points. We excluded post-thrombotic patients with disease below the level of the femoral confluence and included only ipsilateral obstructions.
What have the results shown?
The goal was clearly not to try to show differences in patency rates or to suggest that any stent was better or worse than others, but was rather to distillate specific features and thereby help manufactures develop improved designs. All the stents have specific pros and cons. We looked at a number of factors—ease of deployment, positioning, whether the device straightens, how much it straightens, how strong is it, how much radial force it exerts, and whether the integrity of the stent might suffer.
Looking at the Vici stent, we could see it is a very strong stent, with a sufficient surface area, performing well in May-Thurner syndrome patients, with good stent integrity. When you put pressure on the vein you do not see indentation of the stent. This differs from the Zilver Vena device, which is weaker as it does not have a closed cell design. However, the Zilver Vena stent is very flexible, does not foreshorten and stent deployment is reliable. Ease of deployment was also very positive with Vici as the device comes out and you can still reposition it slightly. With the sinus-Venous device, for example, you have to be very accurate with the initial deployment because there is no change of repositioning. You have to deploy it deliberately and stack the segments near to each other to get the optimal stent configuration and characteristics. It is an extremely flexible stent and with the correct deployment technique it also has a high crush resistance. The sinus-XL Flex stent is easier to deploy, though it is less flexible and it has a tendency to kink at the flex points.
The Vici device is also a rigid stent and shows similar characteristics. If you have it in your hand you can force it into a curve, however, I do not call this flexibility, because when you let it go it straightens forcefully. Stents should follow the venous anatomy and not the contrary; the stent should not dictate how the vein is going to behave. If you change the venous anatomy at one location due to stent rigidity, it might backfire on you at another and cause early or late stent-related patency loss.
What would the characteristics of the ideal venous stent be?
The ideal stent should be strong at the high pressure points but flexible at the flex points, and there are more flex points in the venous anatomy than you would imagine, because the venous anatomy differs from its arterial counterpart. In the pelvis—in front of L5, in the promontory area, there is a high flex point of 120 degrees; deeper in the pelvis, it is about 130–140 degrees; and at the level of the inguinal ligament, we again have 130–140 degrees, creating an S shape. Thus the ideal venous stent should be able to form that S shape effortlessly. That is also what I tell physicians and manufacturers when they ask me what the ideal stent should look like: just try to make an S shape and observe if there are any stress signs visible, like small kinks.
The most important aspect is that we treat relatively young patients, so venous stents will be in the body for 50 years and have to perform well over that entire time. We are not used to that with coronary or peripheral artery stents, which are usually in patients for 5–20 years. Standard fatigue tests have been developed accordingly performing only a limited number of cycles. Principally, we should perform tests representing 50 years of specific stress movements related to the venous anatomy. Venous stenting is a different ball game and without proper tests we need to extrapolate what the stent is going to do after 20 years. With 3D imaging and multiplanar follow-up we have shown that straightening and kinking already happens during follow-up of only one year. It is very likely that these drawbacks continue to develop and increasingly create problems throughout following years. It is important to emphasise these shortcomings and eliminate them so we can move closer to the optimal stent configuration perfectly designed for the venous vasculature.
Based on the study results, what is your take-home message?
There is not a perfect venous stent right now. However, it is too early to exclude any stent because we still have very limited data. The idea with the study was not to put a specific stent forward for use but to try to provide more information to the industry so they can create a better stent. Our unanimous goal should be not to accept any stent-related patency loss in the future.