A Carotid Revascularization Endarterectomy versus Stenting Trial (CREST) Sub-Study and a Discussion of the Pathophysiology of Carotid Artery Disease

 This CREST ad hoc sub-study was reported at the Society for Cardiovascular Angiography and Interventions (SCAI) Scientific Sessions (May 8-11, 2013).

In this sub-study, CREST trial patients who received angioplasty post carotid artery stent placement had a greater number of strokes versus the group who received angioplasty prior to stent placement only, but the increase in stroke was not statistically significant. 

Researchers also found that if angioplasty was performed post carotid stenting, the risk of restenosis was reduced by 64%.

Cath Lab Digest talks with Mahmoud B. Malas, MD, lead author of the sub-study and associate professor of surgery at Johns Hopkins University and chief of endovascular surgery at Johns Hopkins Bayview Medical Center in Baltimore, Maryland, about the implications of the study findings.

Dr. Malas, how did you arrive at the point where pre and/or post stent ballooning in carotid artery stenting became a focus? 

The whole idea for the trial started with the concept of reducing the complication rate of carotid stenting. Carotid stenting has been the most controversial procedure in recent time, and probably one of the biggest turf battles between different specialties. It has been done by interventional neuroradiologists, interventional neurologists, vascular surgeons, and interventional cardiologists. The initial studies on carotid stenting were disappointing, because the risk of stroke was as high as 10%, and many of the initial studies were terminated prematurely because of the prohibitive risk of stroke, given that the risk with carotid endarterectomy for asymptomatic patients can be as low as 2%; it can be as high as 5% for symptomatic patients, according to several randomized trials. We continue to learn how to improve the outcome of carotid stenting. 

I was trained at Montefiore Hospital at Albert Einstein in New York where they did some of the early work to identify how to prevent embolization associated with carotid intervention with distal protection and that led to the development of the first embolic protection filter. Filters used today during carotid stenting dramatically reduced the risk of stroke. We continue to improve our technique and patient selection criteria. This mirrors a steady decline in the risk of stroke from this procedure. Meanwhile, several studies were done in Europe: a German study, Stent-Protected Angioplasty versus Carotid Endarterectomy (SPACE); one in France, EV3S (the eV3’s Protégé RX), and the British study, the International Carotid Stenting Study (ICSS). All of these trials still showed that the risk of stroke with stenting is higher than endarterectomy. However, there was a dramatic decrease in that risk (4-7%) compared to the initial studies. Then the CREST trial came along, which took place in the U.S. and Canada. CREST is the largest study ever done comparing endarterectomy to stenting. It had 1600 patients in the lead-in-phase, and 2500 patients in the randomization phase. CREST was very strict in identifying investigators. To be a study interventionalist, you had to show at least 25 prior consecutive CAS procedures with an embolic protection filter with good outcome. CREST showed the lowest rates of stroke ever reported, for both stenting and endarterectomy (surgeons also had to be very well trained). Out of 110 investigators, there were few vascular surgeons on the study who did both procedures, probably less than five, including me. The beautiful thing about CREST, however, was that it included every possible specialty that could do one or both procedures, and the outcome was pretty darned good. CREST showed no difference between the two procedures in the long term, while immediately after the procedure, the risk of stroke with stenting was as low as 4%, the lowest ever reported. Likewise, the risk of stroke with endarterectomy was 2%, which is also the lowest that has ever been reported, for both symptomatic and asymptomatic patients. 

In addition to CREST, we were part of the randomized ACT-1 trial, and are enrolling in several carotid stent studies for high-risk patients for surgery. We are participating in the SAPPHIRE-W study, CAPTURE-2, CHOICE, and the flow reversal trial, FREEDOM. In our own data, for our last consecutive 100 carotid stents, we noticed that we had only one stroke. Our goal has been to reduce the risk of stroke to even less than 4%, and make the risk of stoke with stenting as good as it is with endarterectomy. At least in our own experience, we saw no difference in the risk of stroke between the two procedures, because we are very selective in our patient criteria, including avoiding difficult anatomy that can lead to stroke and using filters or flow reversal on all our patients. 

Early on, we used the most common technique: place the filter in the distal part of the internal carotid artery to protect the patient from a stroke, then balloon the lesion, very gently, with a small angioplasty balloon, and place the stent. After stent placement, we always went back and ballooned it again, because there was always residual stenosis. We might start with a 90% blockage, and after we balloon the first time and put in the stent, it would be perhaps a 30 or 40% residual stenosis. We wanted to make it perfect, so we would balloon it again. When performing the second ballooning, many patients would become hypotensive and bradycardic. It led us to ask ourselves, why are we doing this second ballooning? Our patient can become unstable hemodynamically, stays longer in the hospital, and we sometimes have to support them with medication drips after the procedure to maintain adequate blood pressure. 

That area of the carotid artery bifurcation has a sensor called the carotid body, with chemical and baro-receptors that sense the blood pressure changes in the body. When you have high blood pressure, the sensors become aware of it and try to reduce your blood pressure by lowering your heart rate and making the blood vessels dilate. When we put the stent near the carotid body, that stent pushes on the receptors, tricking the carotid body into believing that this patient has high blood pressure, and that it should signal to the brain to lower the heart rate and vasodilate. Thus, blood pressure starts dropping. When we balloon the stent, the trigger becomes even more exaggerated, and we see a lot more hypotension, which is the last thing you want for someone who already has a blockage in their carotid. 

I am also concerned that when I place a stent, it is pushing on that lesion, and when I balloon it, the struts of the stent might act almost like a knife and cut through the plaque, and little pieces can break off and go to the brain and cause a stroke. After all, this is really what causes stroke. While the pathophysiology of coronary disease is stenosis causing heart attack because of decreased blood supply to the heart muscle, and that of peripheral arterial disease is also decreased blood supply to the lower extremities, the pathology of carotid disease is embolic and not related to decreased blood flow from carotid stenosis. This is a very important concept to physicians who treat carotid disease, but often they seem to forget it and insist on making the vessel look better. I can remember how many times I heard my colleagues saying “It’s still a little narrow, let me balloon it more to make it look perfect.” Well, you are not there because you are trying to make the vessel perfect; the goal is to stabilize the lesion. The reason carotid stenting works, I believe, is because it stabilizes the plaque, and also it causes intimal hyperplasia in reaction to the stent, so it smoothes the surface of the plaque, rather than it being ratty and irregular, with little pieces of it breaking off. I have taken stents out before, surgically, and have seen how they have an amazing intimal hyperplasia layer on them, so that the plaque surface is no longer irregular. I believe that is the way stents work in the carotids. 

By ballooning the lesion, you are actually risking a shower of emboli, so I stopped ballooning altogether. In our 100-patient cohort, some patients were ballooned and some were not ballooned. My initial study showed that the risk of becoming hemodynamically unstable was 5-fold increased in the group that got post-stent ballooning. Since I have very little stroke, I didn’t have enough events to show a difference in the stroke rate between the two groups, because I only have one stroke and one transient ischemic attack (TIA). Both these complications were in the group that got ballooned after the stent. I had to have a larger data set, and so I turned to the CREST trial. In CREST, about 1100 patients received a stent. We focused on stroke and, of course, on the risk of restenosis, because we wondered whether we would have more restenosis by not ballooning the stent. We did find that the group that did not get ballooned had a higher risk of restenosis and that was statistically significant. But that was not a big problem overall. When most patients restenose, we don’t necessarily have to do anything about it, because of that smooth surface of intimal hyperplasia. Most people who have restenosis are asymptomatic. This was also true for the group in CREST that had restenosis. Most did not have symptoms and the majority did not end up having a procedure. We did see a three-fold increase in the risk of stroke in the group that got post-stent ballooning, however, once again, because even though CREST has 1100 patients, it had only 19 strokes in the stent group, and that was not enough, between the two separate groups, to show statistical significance. 

Every time you balloon, do you use an embolic protection device? What are your thoughts on the different embolic protection devices available today?

I always use a protection device. It is important to understand that even though the filter works and reduces the risk of embolization, it is not 100% protection, because the filter still allows the blood to go through, so it will allow some microemboli to pass as well. 

Frank Veith and Takao Ohki did the initial work at Montefiore that showed how important it is to have a filter. They demonstrated how the filter worked in vitro. After endarterectomy surgeries, they took the plaque and put it in tubes, and then ballooned it while a filter was placed at the end of the tube. They showed how the filter, every single time, captured debris. Even if it was microscopic, there is always debris in the filter. 

Different filters on the market have different pore diameters. Some are a lot smaller and capture a lot more emboli, but then the drawback is that when the filters have very small pores, they can clot during the procedure. There is also certain debris that goes around the filter, even if it is well placed. The filter should be placed in a segment of the artery that is straight, not tortuous, and must be sized properly to be well apposed to the vessel wall. Still, filters are not perfect. That is why we are still having strokes in these procedures. We risk a stroke just passing the lesion with the filter initially, before it is deployed. That is why we are now testing flow reversal, which does not use a filter, but uses a reversal flow in the brain to protect the brain. That’s not the scope of this study, however. 

Several studies of transcranial Doppler during carotid stenting showed how debris is visible in the middle cerebral artery during stenting and during every step of the procedure. You see it when you first select the vessel from the arch, when you pass the lesion with the filter and during initial balloon angioplasty and when you place the stent. Yet the highest load of emboli is always during the post stent ballooning. This is the part of the procedure when we are pushing the stent against the plaque and risking more debris breaking off. 

It happens during every procedure we do, but the brain has such amazing collateral vessels, that it is forgiven, in a way. If it is a large piece of debris that blocks a major artery, you have a stroke, but when it is little tiny debris, the other blood vessels can compensate. People have done MRI studies on patients after carotid stenting that show little infarcts in the brain, even though the patient is completely asymptomatic. There is obviously some controversy about the importance of those infarcts in asymptomatic patients. 

So what do we take from this? We found a much larger number of strokes in the group that got ballooned post stent. However, the only thing that was statistically significant is restenosis. Where do you go from here? It’s a judgment call. At the end, there is no one single study, no matter how well it is done, that can change practice. In the end, the treating team just has to use their common sense and judgment, and tailor their knowledge to every patient. Every patient is different: different anatomy, different problems, and different risk factors. So I think this study is a good start, because it is the first and the largest study to show post-stent ballooning does increase the risk of stroke, although it is not statistically significant, again. Post-stent ballooning does reduce the risk of restenosis. Restenosis, most likely, is not going to be symptomatic or cause any problems. Is the operator willing to take the risk of stroke? Obviously, the risk of stroke is much serious than restenosis. Especially if restenosis is going to be most likely asymptomatic.

We have long-term follow-up on over 80% of our patients at Hopkins. Only three or four patients have moderate restenosis, and none of them required an intervention, because they were asymptomatic. You don’t really have to make the lumen look perfect. You don’t have to make it 100% open. It is acceptable to leave 30-40% residual stenosis. That is just my own theory and it is yet to be proven. More studies need to be done with larger numbers, and studies need to be designed specifically to prove that point. 

In CREST, it was just operator’s choice as to what they chose to do with ballooning.

Right, and there were three different groups in CREST. The group that had pre-and post-stent ballooning is the largest group, 696 patients. The second group was smaller, 344 patients, where operators did primary stenting, skipping pre-stent ballooning. Then, after they placed the stent, they ballooned. The idea behind primary stenting is preventing shower emboli. Others believe the opposite, because you are advancing the stent through a very narrow vessel without dilating the lumen to allow the stent to fit through, thus increasing the risk of shower emboli. Finally, the third group, a small group of 69 patients, ballooned before stenting, made room for the stent, placed the stent, deployed it, and then left it alone and did not balloon after the stent. This is the group that had the lowest number of stroke out of the three groups, but did have a higher risk of restenosis.

Dr. Malas can be contacted at mbmalas@jhmi.edu.