Embolic Protection During Transcatheter Aortic Valve Replacement

Can you tell us about your experience with transcatheter aortic valve replacement (TAVR)?

We started our TAVR program in 2012 with FDA-clearance, and since that time, have fully adopted TAVR into clinical practice. We have grown over the years perform approaching 300 TAVR procedures each year. We support TAVR for intermediate risk patients and close to 65% of our total aortic stenosis volume is now done by TAVR. We also believe in an efficient program and work using a minimalist approach with conscious sedation and a totally percutaneous access for an overwhelming majority of patients. We can routinely complete 5 cases in a day in our hybrid room or cath lab. We have a team of 2 interventionalists and 2 surgeons (although that is growing) to do the procedures. Our dedicated cardiac computed tomography (CT) program, directed by Dr. Marcus Scherer, performs all pre-procedural imaging and analysis. TAVR has been a very successful program for us.

Is it the use of the minimalist approach that allows you to do as many as 5 cases a day?

Most of the efficiency that allows us to reliably complete 5 cases in one room in one day is related to the minimalist approach. The minimalist approach is, first and most important, good for patients, as it improves recovery time and reduces the risk of complications, but it is also good for the providers and hospital system, because we can work efficiently, reducing turnover times between cases.

What is the risk of stroke during a TAVR procedure?

Stroke is a persistent problem in TAVR, and while many of the early problems with TAVR, such as access site complications and perivalvular leak, have been dramatically improved by advances in devices and technique, TAVR stroke rates have remained fairly constant. Even looking at the more contemporary trials with moderate risk TAVR patients, for instance, PARTNER IIA, with the Sapien XT valve (Edwards Lifesciences), PARTNER II S3i, with the S3 valve, and SURTAVI, with the CoreValve (Medtronic), the risk of stroke in each of those trials remains in the 3-5% range and the risk of major, disabling stroke is between 1-3%. That is a persistent issue. If you compare it to percutaneous coronary intervention (PCI) stroke data, it is a thousand-fold higher. In TAVR, 1 to 3 in a hundred patients are going to have a major, disabling stroke, and this is an important issue we need to address. It is not that surgery is any better. The surgery arm of contemporary TAVR trials has shown consistently higher stroke rates than TAVR, so in terms of stroke, TAVR is still better. If you look at stroke rates in the Society of Thoracic Surgeons/American College of Cardiology (STS/ACC) Transcatheter Valve Therapy (TVT) Registry, median rates are about 1 to 1.5%. The trial data is reflected in real life. Of course, real-life registries will under-report smaller strokes, and those strokes are important, too, because they can affect cognitive function and vision. Patients’ greatest fear isn’t necessarily death, but a major, disabling stroke. The risk of stroke still needs to be improved upon and I think we now have the tools to make an impact.

How are you trying to avert the risk of stroke?

We use femoral access and avoid predilating before valve deployment. If we use a self-expanding prosthesis, we minimize repositioning as much as possible. There is no way to really impact stroke rates significantly, just based on technique, other than that. What we have adopted is embolic protection, and the currently available FDA-cleared device is the Sentinel Cerebral Protection System (CPS) (Claret Medical, now part of Boston Scientific). Certainly, there are other devices that are investigational, but those are not available in the U.S. outside of clinical trials. Once the Sentinel device became available, we felt it was important to add it to our armamentarium. Some programs have tried to add embolic protection selectively for what they perceived as higher-risk cases only, but an analysis of the data suggests there isn’t a subset of patients that necessarily have a higher stroke rate. It is not that more calcified valves are more likely to have stroke — that doesn’t pan out. Our feeling was that if we were going to adopt embolic protection that it should be adopted uniformly on all patients where its use was anatomically favorable. The Sentinel CPS has a filter that sits in the brachiocephalic artery and a second filter that sits in the left carotid artery. It is deployed from a right radial or brachial access. It requires adequate right upper extremity access, and anatomy that is without too much tortuosity or significant disease in the vessels. We have found that greater than 80 to 90% of patients do have favorable anatomy, so it is being applied in a significant portion of our TAVR population. In our procedural evaluation pathway, in addition to evaluating lower extremity vascular access, we now also evaluate upper extremity access with computed tomography (CT). We are looking to make sure the arteries are not overly tortuous and that we can safely and efficiently deploy the Sentinel device. Concern has been voiced that manipulating the aortic arch during deployment could cause iatrogenic emboli.  While unproven, this is a reasonable concern, so we adopted a minimal manipulation protocol to our workflow. We hold ourselves to a “10-minute or less deployment time” limit to reduce the risk of iatrogenic emboli from arch manipulation in complex anatomy. An appropriate procedural analogy would be to carotid stenting. It has been said that in carotid stenting, if you spend more than 10 minutes trying to get a sheath up the carotid artery, you should stop, as the risk of causing emboli from arch manipulation becomes too great. We think the same applies with embolic protection. If it takes more than 10 minutes to deploy a device, this reflects unfavorable anatomy and potentially more risk. While we have adopted this approach in all our TAVR cases, this “10-minute rule” applies to a very small number of cases and with experience, can largely be predicted from the pre-procedural CT angiography (CTA). This approach also adds to procedural efficiency. We were already routinely completing 5 TAVRs a day, typically finishing before 5pm. By implementing this 10-minute attempt dictum, we are still able to get those 5 cases completed. In reviewing our data, we found embolic protection use added less than 10 minutes per case and this was early in our experience, as we were still learning the device. The SENTINEL US IDE showed a median 4-minute device deployment time. With anatomy prescreening, device prep done in parallel, and standard patient prep process, we believe this average 4-minute device time is realistic in clinical practice. Once we became efficient with the device and added it to our workflow, it really didn’t change procedure times very much at all.

Why remove emboli rather than deflect it elsewhere?

Capturing emboli makes sense. We know from the SENTINEL US IDE that virtually all patients have emboli, as Sentinel CPS was able to capture and remove embolic debris in 99% of TAVR patients, regardless of valve type and patient risk profile. All TAVR devices, including ones that are not yet available in the U.S., cause significant emboli, and this correlates with post procedural imaging studies that nearly uniformly show cerebral emboli hits. Procedural transcranial Doppler evaluation shows most emboli occur during TAVR device positioning and deployment, and balloon valvuloplasty procedures. There is a compelling reason to capture the emboli and keep them from going to the brain. In our experience, it is not infrequent that we remove the Sentinel device and find material captured in the baskets. These aren’t just microemboli, but sometimes visible pieces of calcium or tails of thrombus large enough to cause major stroke if they had reached the brain.

What does the term “protected TAVR” mean to you?

Protected TAVR/protected TAVI means using a device (and in the U.S., that device is the Sentinel CPS) to keep emboli from traveling to the brain. From a clinical outcomes standpoint, the brain is the most sensitive to emboli. So when we say “protected TAVR”, we mean the use of a device to keep emboli from traveling to the brain, as the consequences can be devastating. 

What have you seen as a result of implementing embolic protection during TAVR?

Our “protected TAVR” program is in the early stages, but in the first 40 patients that received full protection with the Sentinel device, none had strokes. Prior to this, we had about a 1.5% stroke rate. Those numbers aren’t large enough for true statistical significance, but numerically, it is certainly in a favorable direction.

At what point in your patient evaluation process do you decide whether to use embolic protection?

It is mostly done when we pre-screen the anatomy by CTA. As we review anatomy for femoral access, we now routinely assess anatomy for embolic protection. We assess subclavian tortuosity, the angle of the left subclavian origin, and assess for vessel size and disease. It has become part of our routine process. The patient worksheet is mailed to every member of the team, including the OR and cath lab staff. If we are using embolic protection, they will know to have the arm prepped for transradial access, and they also have the Sentinel device available for quick prep. Radial access is obtained while we are getting access everywhere else. Then we move forward with the procedure.

How are you discussing the risk of stroke and use of embolic protection with patients?

When we talk to patients about aortic stenosis and risk, we find that one of the things they fear most is stroke. That said, most of these patients are highly symptomatic and understand that aortic stenosis is a progressive disease. The majority are eager to move forward despite the stroke risk. We haven’t talked to our patients specifically about embolic protection, because not every patient is necessarily going to receive it, based on anatomic constraints. At this point, our approach is more that the Sentinel device has become part of our protocol and it is simply how we take care of our patients. On the other hand, however, I do think embolic protection should be discussed more with our referring physicians. Since many physicians have a choice as to where to send their patients, I believe that given that choice, most physicians would preferentially refer to centers using embolic protection. As evidence for this, on multiple occasions, I have attended conferences where physicians were asked, ”If you or a family member was having a TAVR, would you want embolic protection?” Each time, everyone picks embolic protection. Yet when asked, “Do you use embolic protection?”, only a minority of hands go up. There are a number of potential reasons why people don’t use embolic protection, but one of the greatest barriers has been cost. Beyond the potential for marketing advantage, we as a system decided to commit to the additional expense associated with the device, because we believe it is the right thing to do. On August 2nd, the Centers for Medicare & Medicaid (CMS) approved the Sentinel Cerebral Protection System for a new technology add-on payment (NTAP) as part of the FY 2019 Inpatient Prospective Payment System (IPPS) final rule which will kick in from October 1, 2018.  With some of the significant financial barriers to use of the Sentinel device mitigated, there may be a substantial increase in the adoption of this technology.

It sounds like adopting embolic protection for TAVR is the best approached from a program level.

TAVR is truly a system process; there isn’t just one provider that does TAVR. This procedure was the beginning of the first real heart team, with interventional cardiologists, surgeons, and advanced imaging specialists coming together with program coordinators, nurse coordinators, and advanced care practitioners, to provide a service that is truly integrated, where multiple members of the team are working together. That’s something in medicine we talk about, but doesn’t always happen. TAVR is the one place where it really happens. We, as a team, decided that use of embolic protection as a default strategy was the right thing to do. With team buy-in, implementation was easy. The unreimbursed cost of the device was an important consideration, but our cardiovascular service line administrator has been an integral part of our team and felt the message was compelling. Now with NTAP granted, cost becomes even less of a consideration.

How did your team approach it?

We didn’t face the same pressures that some other institutions have faced. We have a very good relationship with our main cardiovascular service line administrator. We sat down, as we often do, with him as part of our team, showed the evidence, and said, we think this is compelling, important, and the right thing to do for patient safety. We have been responsible in our approach to TAVR and have tried to save the system money through efficiency where it clinically didn’t affect patient care. With this level of cooperation and mutual respect, we were easily able to obtain system support. Not all systems have administrators that are as engaged. We have been lucky in that regard.

As we look ahead, what do you see happening with embolic protection?

Embolic protection will be adopted rapidly once a reimbursement code is in place, to the point that it becomes the standard of care. Instead of should we do embolic protection, we will ask, in which patients should we not do embolic protection? Who is anatomically not feasible? And how hard should we push the boundaries of anatomical feasibility and who should we not protect? The use of embolic protection is going to change rapidly. I think there will be a complete 180-degree turn within the next 12 months. 

Any final thoughts?

There has been some pushback against the use of this technology, because the SENTINEL trial just missed the primary endpoint, but I think that conclusion is an overly simplistic way of looking at the data. In the SENTINEL trial, stroke rates were numerically lower in the group that received protection; it just didn’t meet statistical significance, because the trial was underpowered. However, there were definitely favorable trends. If, instead of the 30-day stroke rates, we look at only the procedural stroke rates (within 3 days), there was a statistically significant reduction in neurologist-adjudicated clinical stroke. 

From a purist’s standpoint, randomized trials are the closest thing to truth in medicine that we can get, and to some extent, cardiologists pray at the altar of randomized trials. However, the problem with randomized trials is that their results reflect the truth only in the population studied and may not reflect the truth in every population. Registries are much more reflective of the real world, but we can’t measure all the confounders and can’t take out all the bias. Randomized trials eliminate the confounding and bias, but only apply to a small subset of patients. No data set, be it a randomized trial or registry, is perfect. We must take the data, both randomized and registry, and put it through a filter of clinical experience and common sense. The Sentinel device makes mechanistic sense, and there is a reasonable amount of compelling post-market data that supports its use. 

Disclosure: Dr. Michael Rinaldi reports the following disclosures: Abbott Vascular- Advisory Board, Teaching Courses, Consultant, Speaker; Boston Scientific – Advisory Board, Consultant, Speaker; Edwards – Teaching Courses, Consultant; Cordis – Advisory Board, Consultant, Speaker; 4C – Advisory Board.

Dr. Michael Rinaldi may be contacted at michael.rinaldi@atriumhealth.org.