On the Approval and Use of Vascular Devices: What's in a Label?

In 2005, more than 90% of non-coronary endovascular stent use (by physicians performing vascular intervention) was off-label. It dropped to a “low” of approximately 60% in 2006. Astonishingly, the on-label indication for the vast majority of such devices was (and is)… temporary relief of malignant bile-duct obstruction in patients with a less than 6-month life expectancy! See anything wrong with this picture? Let us first formulate and answer the three critical questions surrounding these issues. 1) What is “label”? 2) What, if anything, is wrong with off-label use? and; 3) Why does this happen?

• Label is the indication or specific use or purpose for which the FDA approves a drug or device. In truth, label is what “FDA approval” is all about! Its importance and profound implications are huge and cannot be overemphasized. Off-label refers to the use of a medical device (or drug) for a non-approved indication, or application in a non-approved anatomical location or site. Often, the instructions for use (IFU) and label go together and are essentially one and the same — but not always. For some devices (i.e., aortic stent-grafts and several others), the IFU may include detailed instructions on technical and procedural aspects that go well beyond label.

• What’s wrong with off-label use? Why would it be bad to use a stent in a non-indicated location, i.e., a biliary-label device in the superficial femoral artery (SFA)? The answer to this question should be rather obvious: by and large, these devices have not been tested or validated (safety and efficacy) for anything other than the labeled indication. Examples of serious adverse events abound: air embolism with carotid use, stent dislodgement from the delivery catheter during vascular deployment, frequent stent fractures in the SFA, and many more. Having said this, it is also important to note that, in general, there is no direct or consistent link between off-label use and medical risk. The risk depends largely on two things: how different the off-label use is from the on-label application of a given approved device, and what evidence (or the lack thereof) exists to support off-label use. There are situations where off-label use has been studied extensively and has in fact become the accepted standard treatment. In such scenario, not offering off-label use may be improper.

• Why does this happen? Why would manufacturers pursue a non-vascular label (indication) for a device that is clearly ‘intended’ for vascular use? Again, this is a “no-brainer”. To gain agency approval and release for commercialization of a class 3 (implantable) vascular-label device, manufacturers are required to go through the so-called Pre-Market Approval (PMA) pathway that implies the need for extensive documentation and demanding engineering and clinical testing. The process tends to be quite prolonged and very costly. Contrary to this, FDA approval of non-vascular devices (i.e., biliary stents) is far less demanding through the Pre-Market Notification 510(k) pathway. Documentation and required testing are simple and straightforward – without need for extensive fatigue or clinical testing. Essentially, all that is necessary is to demonstrate substantial equivalence to a predicate similar device that is already available on the US market. Needless to say, this state of affairs is totally unacceptable, and one might argue that it has reached a point where almost “everyone” agrees something must be done to correct it. The appropriate branches of the FDA and all manufacturers have entered into a dialogue that will hopefully lead to a workable solution in the near future. Industry will be encouraged to pursue on-label vascular device submissions as the PMA pathway becomes streamlined and more “user-friendly”. Another important focus will be the development of improved engineering testing standards and methodologies, including computer simulation.

Today, there is absolutely no doubt that minimizing off-label use is a most important and worthy goal. After all, it must be remembered that it carries significant risks, such as the potential for widespread clinical application of a device lacking valid safety and effectiveness data, and the inability to fully inform patients of possible adverse events. Furthermore, off-label practices imply that the physician and the hospital assume all risk and liability in the absence of FDA evaluation. According with a recent presentation by an FDA officer*, we have at present in the USA the following vascular-indication approved stents: 18 coronary, 5 carotid, 4 iliac, 2 renal, 2 SFA, 1 TIPS, and 1 venous. The number of approved non-vascular stents is in the 100s (mostly biliary-indication devices). These numbers will likely shift significantly within the next few years as the PMA process becomes less burdensome. In the meantime, physicians must continue to do what’s best for their patients — off-label device use included. However, I would urge everyone to give consideration to all of the issues discussed in this editorial and to try to determine in each case whether the potential “costs” can be justified by the expected benefit to the patient.¹

* Data presented by Jennifer Goode at ISET 2007.