PROTECTED TAVR: No Less Stroke With Embolic Protection, but Some See Hope

There were fewer disabling strokes with protection and the procedure itself was safe, so it’s not the end of the story.

PROTECTED TAVR: No Less Stroke With Embolic Protection, but Some See Hope

BOSTON, MA—Those who were hoping that PROTECTED TAVR would provide a clear-cut result and bring closure to the question of whether cerebral embolic protection (CEP) reduces stroke in patients undergoing TAVI were disappointed Saturday, when the findings were presented here at TCT 2022.

The trial missed its primary endpoint, failing to demonstrate a significant difference in the rate of stroke within 72 hours of the procedure or before discharge—2.3% with use of the Sentinel cerebral protection system (Boston Scientific) versus 2.9% without (P = 0.30).

But that’s not the end of the story, according to Samir Kapadia, MD (Cleveland Clinic, OH), global principal investigator, because with embolic protection there was a lower rate of disabling stroke, a secondary endpoint for which the trial was not powered (0.5% vs 1.3%; P = 0.02).

He told TCTMD the trial also showed that use of embolic protection was safe, with no major complications, and that stroke was an unpredictable event—no specific factors that could be used to identify a patient group most likely to benefit from protection emerged.

“I wouldn’t say it’s negative,” Kapadia said about the trial, adding that it’s possible a larger study could have shown a significant advantage for protection on the primary endpoint. Moreover, he added, disabling stroke is a more meaningful outcome for patients. “This is, in my opinion, a positive trial, not a negative trial.”

He concluded that use of protection with the Sentinel device (Boston Scientific) should be considered for all patients undergoing TAVI as a way to reduce disabling strokes. The number needed to treat for this purpose was 125.

I don’t think we can completely discount and call this a negative trial, because disabling strokes are important and [are] associated with significant disability and morbidity and mortality. Sachin Goel

The results were published simultaneously online in the New England Journal of Medicine; along with an accompanying editorial by John Carroll, MD (University of Colorado School of Medicine, Anschutz Medical Campus, Aurora), and Jeffrey Saver, MD (David Geffen School of Medicine at UCLA, Los Angeles), that provides a more-sobering assessment of the trial.

“As a whole, this trial strikes a serious blow to the hope that this first CEP device would have a definitive effect in the prevention of TAVR-related strokes,” they write. “The ‘smoking gun’ of captured debris was not associated with high-quality evidence of stroke prevention. The safety of the device and its possible effect on disabling strokes provide a ray of hope, but the concern persists that CEP, approved and marketed on the basis of surrogate outcomes, may not actually provide a clear and incontestable clinical benefit.”

PROTECTED TAVR is one of the most eagerly awaited RCTs—and the largest—in structural interventions, Sachin Goel, MD (Houston Methodist, TX), commented to TCTMD.

Two important findings from the trial are that use of the Sentinel system was safe and feasible—with the proportion of patients in whom the filter couldn’t be placed similar to what’s seen in practice—and that disabling strokes were reduced, Goel said. He acknowledged that the primary endpoint was missed but said, “I don’t think we can completely discount and call this a negative trial, because disabling strokes are important and [are] associated with significant disability and morbidity and mortality.”

The safety information is particularly noteworthy, Goel said, because there had been concerns that the manipulations necessary to get the Sentinel filter in the proper position could actually increase the risk of stroke. “I think this is very reassuring that we haven’t seen an increase in the risk of stroke with this device,” he added.

Uptake of the Sentinel system has been spotty since the US Food and Drug Administration approved it in 2017 on the basis of the Sentinel IDE trial. That study showed that debris was captured by the filter in nearly all patients, although it also missed its primary endpoint by failing to show a reduction in the volume of new cerebral lesions on MRI.

Whether the PROTECTED TAVR results will actually change how cerebral embolic protection is used out in practice—where there are some strong believers in its benefits and others who are more cautious—and whether TAVI programs will take on the added cost of embolic protection without clear evidence of a clinical benefit remains to be seen.

Megan Coylewright, MD (Erlanger Health System, Chattanooga, TN), editor of the American College of Cardiology’s CardioSmart, said, “We were hoping there would be a very clear message that stroke was reduced with cerebral protection, and the results show that there is no difference in stroke between control and cerebral protection. And typically in studies we stop there, and we are curious about the subgroup analysis but we take the finding that the study was powered for as a way to guide clinical practice.” She noted that the trial was not powered to detect a difference in disabling stroke.

For his part, Goel thought there would be some impact. “I’m not sure that all nonbelievers will start believing in this and start using this, but I think there will be more believers now than nonbelievers perhaps, and the shift will be towards using more cerebral embolic protection,” he predicted.

Discussion during a press conference reflected uncertainty about the clinical impact of the trial.

Michael Young, MD (Dartmouth-Hitchcock Medical Center, Hanover, NH), said that his group does not universally use embolic protection during TAVI. This trial supports the safety and feasibility of using the Sentinel, as well as a reduction in disabling stroke, he said.

He added, however, that the trial “muddies the water somewhat for programs that are low-to-medium volume,” pointing to the number needed to treat to reduce disabling stroke. “I think it’s going to be a tough sell for some of those programs to ubiquitously use that for all their patients.”

Still, patient preference may affect use, he suggested. “Patients are savvy and they do their research, and I’ve had a number of patients asking, ‘Do you use cerebral embolic protection?’” he said. “And I think this may potentially change the equation and how people think about it.”

And it’s not just patient preference that could influence adoption of embolic protection. At the main presentation of the results, several physician panelists responded to a question about whether they would opt for embolic protection if a loved one were undergoing TAVI. Of the seven, only one responded no, noting that at her center in Montreal, Canada, protection is not currently used in any patients.


Stroke has been a feared complication of TAVI from the beginning, and remains so despite substantial reductions in risk over time. One study using data from the STS/ACC TVT Registry showed that early postprocedural stroke was associated with a substantially higher risk of 30-day mortality (16.7% vs 3.7%) and a reduced likelihood of being discharged home (36.1% vs 78.9%).

The Sentinel cerebral protection system was developed as a way to capture debris created by the TAVI procedure to prevent it from embolizing to the brain. Though the Sentinel IDE trial missed its primary endpoint, a post hoc analysis suggested that there was a lower rate of stroke in the 72 hours after the procedure in the protection group (3.0% vs 8.2%), providing the impetus to perform the PROTECTED TAVR trial.

PROTECTED TAVR, conducted across 51 centers in Australia, Europe, North America, enrolled 3,000 patients (mean age 79 years; 40% women) who were undergoing TAVI, and randomized them to use of the Sentinel protection system or not. Patients at all levels of surgical risk were included, with an average STS score of 3.3% in the protection arm and 3.4% in the control arm. Neurology professionals examined all patients before and after TAVI.

Overall, the Sentinel system was successfully deployed in 94.4% of patients in whom an attempt was made.

The primary endpoint of stroke shortly after the procedure was based on NeuroARC definitions. Most of the strokes recorded were ischemic, both in the disabling and nondisabling categories. There was a lower rate of disabling strokes in the protection arm, but that difference was not seen for nondisabling strokes (1.7% vs 1.5%; P = 0.67).

Of the eight disabling strokes in the embolic protection arm, two were hemorrhagic and would not have been prevented by the Sentinel device. Only one, an ischemic stroke in the left MCA territory, occurred in what should have been a protected vessel.

There were no differences between the Sentinel and control groups in terms of several other outcomes, including all-cause mortality/stroke (2.7% vs 3.0%), mortality (0.5% vs 0.3%), stroke/TIA/delirium (3.1% vs 3.7%) or acute kidney injury (0.5% in each arm). Only one patient had a vascular complication at the access site for the protection system.

Subgroup analyses showed consistency in the lack of difference between trial arms in terms of all strokes and in the lower risk of disabling stroke with cerebral embolic protection.

What Will Change?

PROTECTED TAVR is likely to fuel a debate about how extensively cerebral embolic protection should be deployed in practice.

I think it’s going to be a tough sell for some of those programs to ubiquitously use [the Sentinel] for all their patients. Michael Young

The problem with stroke after TAVI, Goel said, is that there aren’t any good predictors of who’s going to be affected. At his center, the Sentinel device has been used in a minority of cases, reserved for those with prior stroke or a bicuspid aortic valve as well as those undergoing a valve-in-valve procedure.

Goel noted that there is another similar trial—BHF PROTECT-TAVI, with a projected enrollment of 7,000 patients—still ongoing; a meta-analysis combining the results of that trial with those from PROTECTED TAVR is already planned. In addition, there are other devices in the pipeline that may provide broader protection against stroke during TAVI.

Based on the results, Goel said he will consider using embolic protection more frequently during TAVI, and will discuss these data with patients when making a decision about TAVI.

Coylewright said the focus on disabling stroke by the PROTECTED TAVR investigators reminded her of the story around left atrial appendage occlusion, where trial evidence indicated a reduction in disabling strokes versus oral anticoagulation. “We want to lean into the outcome of disabling stroke because it shows a difference and because we believe that it is important to our patients,” she said. “The study was not powered for that, though. Thus it makes it difficult to interpret the result showing less disabling stroke as a means to change our clinical practice.”

Coylewright highlighted the role patients themselves will play in determining how extensively embolic protection is used moving forward: “I think one of the things that is needed is further engagement with patients undergoing TAVR to help us understand how critical this all-stroke outcome is to them versus disabling stroke in order to further understand how to deploy the therapy.”

Though there didn’t seem to be a signal of risk associated with use of the Sentinel system, Coylewright said, “there are many other things that heart teams will be considering, which includes the additional access point during the procedure, time spent during the procedure, and cost to the patient and the hospital and the health system.”

As for how the PROTECTED TAVR results would shift use of embolic protection out in practice, Coylewright couldn’t predict, but she said once a study is published, the decision to change practice on the basis of the results is based partly on the science and partly on the discussions that are sparked locally and nationally. In this case, each heart team will have to review the results, have a detailed discussion, and make a decision about whether to change practice, she said.

Commenting during a panel discussion at the press conference, Robert Cubeddu, MD (NCH Heart Institute, Naples, FL), said “a lot of this just boils down to cost, and we need to be realistic about that.” He pointed out that many programs across the country are operating on tight margins: “Cost-effectiveness is something we always look at in everything we do in medicine, so I’d like to see how perhaps [the Centers for Medicare & Medicaid Services] will stand on this.”

Cubeddu indicated the embolic protection will continue to be used, however. “It takes 5 minutes to do the procedure. We’ve seen that it’s safe and technically feasible in over 94%, 95% of patients. And if you’re able to prevent one disabling stroke out of 125, and you’re that one patient, I think we would all love to benefit from this,” he said.

Todd Neale is the Associate News Editor for TCTMD and a Senior Medical Journalist. He got his start in journalism at …

Read Full Bio
  • Kapadia SR, Makkar R, Leon M, et al. Cerebral embolic protection during transcatheter aortic-valve replacement. N Engl J Med. 2022;Epub ahead of print.

  • Carroll JD, Saver JL. Does capturing debris during TAVR prevent strokes? N Engl J Med. 2022;Epub ahead of print.

  • PROTECTED TAVR was funded by Boston Scientific.
  • Kapadia reports receiving grant/research support from Boston Scientific and being a stock shareholder of or having equity/options from Admedus, Anteris, and Navigate.
  • Carroll reports receiving lecture fees from Edwards Lifesciences and serving as local principal investigator for trials run by Edwards Lifesciences and Medtronic. He is chair of the steering committee of the STS/ACC TVT Registry, which has provided data on the topic of cerebral protection using the Sentinel device during TAVI.
  • Saver reports personal fees from Abbott, Amgen, Bayer, Filterlex, Johnson & Johnson, Medtronic, MIVI, Novo Nordisk, Rapid Medical, Roche, as well as grants from Edwards Lifesciences, outside the submitted work.
  • Goel reports no relevant conflicts of interest.