Newer data may justify update to CMS reimbursement policy for carotid artery stenting
Click Here to Manage Email Alerts
In the U.S., the use of carotid artery stenting has never been as widespread as that of carotid endarterectomy, which is partly a function of the CMS coverage decision. That may be about to change.
The CMS coverage decision has not been updated since 2008, when CAS technology was much less proven than it is now. In a state-of-the-art review published in July in the Journal of the American College of Cardiology, Christopher J. White, MD, MSCAI, FACC, FAHA, FESC, FACP, system chairman for cardiovascular disease and director of the John Ochsner Heart & Vascular Institute at Ochsner Medical Center, New Orleans, and professor of medicine at the University of Queensland-Ochsner Clinical School, and colleagues outline the case for approving reimbursement for CAS for stroke prevention in symptomatic patients with carotid artery stenosis of at least 50% and in asymptomatic patients with CAS of at least 70%. Except for patients enrolled in approved trials, CMS currently limits reimbursement for CAS to procedures in patients with symptomatic carotid artery stenosis for whom surgery is too risky.
“The CMS national coverage decision is outdated and is not consistent with the current evidence supporting equipoise for CAS and carotid endarterectomy for stroke prevention,” White and colleagues wrote. “This places Medicare patients at a disadvantage by restricting their options for less invasive treatment. The current evidence strongly supports an indication for CAS performed by qualified operators in appropriately selected patients.”
Healio spoke to White about the CMS policy, the advances in CAS since 2008, the importance of appropriate patient selection, the evidence for equipoise between stenting and surgery and the importance of operator and hospital volume.
Healio: Why has the CMS coverage decision on CAS not been updated for so long?
White: It is multifactorial. The CMS workload and queue is more work than CMS can handle, frankly. I am empathetic to that. I have worked closely with CMS on other things, and they have a good group. They set priorities each year, but can only handle so many of them. They review things in great detail and have to be very accurate; CMS is not allowed to make mistakes.
What has happened is that there has not been enough demand clinically for CMS to act on this, even though the data we quoted in the paper were there more than 10 years ago with the CREST trial. It’s a matter of prioritization for them.
Why now? Over the last few years, there has been a new carotid procedure introduced, transcarotid artery revascularization (TCAR). It is basically a surgeon’s procedure, a cutdown on the carotid artery with protective stenting. It has caught fire among the vascular surgery community. The uptick in volume has been largely on that side, and the surgeons are wanting to expand access. Indeed, there were a number of surgeons who were authors on this paper. That would have been difficult to do in 2007.
What has happened now is that carotid stenting is no longer the enemy of carotid endarterectomy. It has become the partner of carotid endarterectomy.
Healio: What are the most important things about carotid stenting that have changed since the last CMS update in 2008?
White: We started with carotid stenting as a solution for patients at high surgical risk. People that for whatever reason — anatomic or medical comorbidities — performing an open operation was going to put them at increased risk. And therefore, we had an alternative.
What we did not have, because carotid stenting was a fairly immature procedure, was a lot of data about who the optimal carotid stenting candidates were, and how to better select carotid stenting patients for lower risk for periprocedural outcomes.
That has now happened. A major part of the paper is aimed at outlining some of the ways we now have to better select good candidates for carotid stenting. So, if we agree there is equipoise in patients who are equally good candidates for either surgery or stenting, then the next discussion is which patients are better for surgery and which patients are better for carotid stenting. We can now identify a candidate pool that is much more ideal for carotid stenting because of lower risk, and we also can identify patients at increased risk of stenting in which to avoid carotid stenting.
Healio: Which patients with carotid stenosis are appropriate to treat with carotid stenting?
White: There is a lot of variation in the papers. Some say carotid stenosis greater than 50%, others greater than 69%. What we chose to do, since we are making the argument for equipoise, is to simply match the indications for endarterectomy to carotid stenting. Whatever would be the surgical indication would match the carotid stenting indication, and that’s for both symptomatic and asymptomatic patients. We were careful to make those numbers align, which makes it much easier for CMS.
The data suggest that is true, but there is variation in the literature. We simply said, for clarity, let’s just harmonize the indications with endarterectomy.
Healio: What is the evidence that there is equipoise between the two procedures?
White: There is equipoise on several levels. The most important one is that the NHLBI is currently sponsoring a trial in patients with asymptomatic carotid stenosis called CREST-2. In that trial, the design is a parallel-arm trial of stenting or surgery combined with best medical therapy compared with best medical therapy alone. There is no comparison between stenting and surgery in the first arm. What that means is, when I see a patient in the clinic who I can enroll in this trial, if they get assigned to the stenting/surgery arm, I can pick whether they get stenting or surgery, because the NHLBI is telling me there is equipoise. The government is already telling us that when CREST-2 was designed, they agreed that there is equipoise. That is also apparent to CMS.
The second part is the data. We now have three large randomized trials — CREST, ACT I and ACST-2 — in asymptomatic patients that show very good equipoise in stenting vs. surgery, with low rates of perioperative stroke, death and MI. In the early 2000s, the guideline for endarterectomy was if you wanted to work on symptomatic patients, your rate of perioperative stroke and death had to be less than 6%, and if you wanted to work on asymptomatic patients, your rate had to be less than 3%. We are achieving those numbers in both categories now. We can find data to justify just about anything, but the level I evidence is all very confirmatory that we are achieving equipoise. That is what led the NHLBI to approve the CREST-2 design.
Healio: To what extent does volume play a role in procedural success?
White: When the CREST trial was ongoing, we had a roll-in where potential investigators had to perform at least 20 cases at an acceptable complication rate before they could officially enroll patients. L. Nelson “Nick” Hopkins, MD, FACS, and colleagues published those roll-in data that demonstrated the vascular surgeons had significantly more problems with carotid stenting than other specialists. In the early 2000s, they just did not have a lot of volume. They did not have a lot of experience with catheter-based procedures and they were early on the learning curve. Once they developed that experience during their roll-in procedures, they did fine. If we had included that roll-in period in CREST, we would have had seriously contaminated data, which is what we saw in some of the European trials that did not have a roll-in phase, leading to much worse outcomes for carotid stenting. People who were not experts at the procedure were encountering more problems.
So, not only are patient selection criteria important, but so is operator experience. There is no substitute for being good at this procedure with relatively high volume. It does not matter what your specialty is, but you need to be able to do this procedure at a high rate and a high volume.
What that means is when CMS updates its coverage decision, someone who implants their first carotid stent is not likely to get the same result as someone who has implanted 5,000. There is no way to factor that in to reimbursement. So, patients and referring physicians need to understand that there is clearly a relationship between volume — both operator volume and hospital or team volume — and outcomes. A knowledgeable patient or referring physician would choose the higher-volume operator in the higher-volume hospital. Patients should ask how many of these procedures are done at the institution. It’s a common discussion with my patients, and I am happy to have it. Referring doctors should understand they should be favoring places where there is a high rate of procedures in order to get the best outcomes.
Healio: What can the medical community do to encourage CMS to update its coverage decision?
White: The ball is rolling. Again, it’s just a matter of the CMS prioritization list. We are looking forward to an updated coverage decision soon.
Reference:
For more information:
Christopher J. White, MD, MSCAI, FACC, FAHA, FESC, FACP, can be reached at cwhite@ochsner.org; Twitter: @stent4u.
Collapse
Healio Interviews
Click Here to Manage Email Alerts