麻豆传媒

A redesigned embolic protection device failed to protect transcatheter aortic valve replacement (TAVR) patients against stroke, the REFLECT II study showed.

Efficacy outcomes were no better (numerically even trending worse) in the 112 patients who had the TriGuard 3 deflector placed during TAVR, compared with 119 pooled controls without embolic protection:

• All-cause mortality or any stroke at 30 days: 9.8% vs 6.7% (P=0.475)
• NIH Stroke Scale worsening predischarge: 14.1% vs 7.6% (P=0.176)
• Cerebral ischemic lesions: 85.0% vs 84.9% (P=1.000)
• Total cerebral lesion volume: median 215.39 mm³ vs 188.09 mm³ (P=0.405)

Nevertheless, safety of embolic protection was upheld, according to Jeffrey Moses, MD, of NewYork-Presbyterian/Columbia University Irving Medical Center in New York City, who presented the REFLECT II study at this year's virtual conference hosted by the Cardiovascular Research Foundation.

The 15.9% rate of 30-day major adverse cardiovascular events (MACE) in the TriGuard 3 group fell under the historical goal of 34.4% (P=0.0001 for non-inferiority), meeting the primary safety endpoint.

Against that, however, the MACE rate was numerically higher than the 7.0% seen in controls, although the difference did not achieve statistical significance (P=0.11).

In line with that difference in the MACE composite, most of its components were numerically more frequent with protected TAVR compared with control cases:

• Death: 2.5% vs 1.8% (P=1.00)
• Stroke: 8.3% vs 5.3% (P=0.057)
• Life-threatening or disabling bleeding: 5.7% vs 0% (P=0.12)
• Acute kidney injury (stage 2 or 3): 2.5% vs 0% (P=0.58)
• Coronary artery obstruction requiring intervention: 0.6% vs 0% (P=1.00)
• Major vascular complications: 7.0% vs 0% (P=0.04)
• Valve-related dysfunction requiring intervention: 0% vs 0%

Moses noted that the majority of strokes observed were non-disabling and occurred in-hospital.

Clinical Benefit of Embolic Protection Still Elusive

Neurologic events remain an Achilles heel of TAVR, occurring in 2%-6% of cases, Moses said in reviewing the literature.

TriGuard 3 was designed to sit below all the vessels going to the brain and deflect downstream any debris expressed during TAVR valve expansion. The self-positioning mesh deflector is inserted through the leg and thus easily fits into the existing workflow, he said.

During a press conference, discussant Dee Dee Wang, MD, of Henry Ford Health System in Detroit, alluded to operator learning curve as a potential reason for the disappointing results with the TriGuard.

"That is an important point we are emphasizing," Moses responded, noting that success was to some extent site-specific, perhaps due to operators confused about the "self-positioning" aspect of the device.

He emphasized that operators have to be careful inserting the TriGuard 3. A newer iteration of the system adds a loader to keep the device where it belongs during placement and has performed well in initial cases in Europe, he said.

An earlier-generation version of the deflector, the TriGuard HDH, was previously associated with reduced cerebral ischemic lesions on imaging but notably lacked definitive evidence of efficacy.

Thus, the combination of an early-stage device and relatively inexperienced doctors has made for a "frustrating" experience so far with the TriGuard, commented TCT director Martin Leon, MD, of NewYork-Presbyterian Hospital/Columbia University Medical Center in New York City, during a virtual panel session.

If operators are to have confidence in embolic protection, Leon said, larger studies that demonstrate clinical benefit are needed.

For now, the Sentinel filter is the only approved embolic protection device approved in the U.S.

investigators randomized patients to TAVR with the TriGuard 3 (n=157) or without protection (n=57). All were patients with severe native aortic valve stenosis undergoing planned transfemoral TAVR. MRI readers were blinded to assignments, though of course operators were not.

Age was approximately 80 years on average. Men accounted for 54.8% of the TriGuard 3 group and 61.4% of controls, the difference not being significant. Baseline characteristics were generally similar between groups, the exception being more prior stroke or transient ischemic attack in the TriGuard 3 group (17.2% vs 5.3%, P=0.026).

For the efficacy analysis, the investigators included the 112 TriGuard 3 patients who were enrolled after the early roll-out phase of the study and did not experience conversion to surgery or prolonged cardiac arrest prior to MRI. The comparison group consisted of 119 control patients pooled between REFLECT II and the REFLECT I study of the older TriGuard HDH device.

Safety was based on the entire 157-person TriGuard 3 cohort.

Notably, just 59.3% of TriGuard cases had full coverage across all three vessels throughout TAVR in REFLECT II. These patients had better efficacy results, namely smaller cerebral ischemic lesions.

A post hoc MRI analysis suggested that the device reduces the number of larger ischemic lesions.

"Improved device stability to achieve reliable, complete cerebral coverage might improve outcomes," Moses said.

Yet Registry Data Offer Hope

In a separate study, there was a potential decrease in strokes over 2 years of real-world embolic protection in TAVR.

In-hospital stroke rates in the STS/ACC TVT Registry numerically favored embolic protection in an instrumental variable analysis accounting for site-level preference (1.39% with embolic protection vs 1.54% without embolic protection, RR 0.90, 95% CI 0.68-1.13), according to David Cohen, MD, MSc, of Kansas City, Missouri.

Analyzed by propensity score weighting instead, in-hospital stroke handed embolic protection a significant advantage (1.30% vs 1.58%, RR 0.82, 95% CI 0.69-0.97), he said in a separate TCT Connect session.

The instrumental variable analysis may better account for measured and unmeasured confounding but yields wider confidence intervals.

Ultimately, both results "are consistent with a possible modest reduction in stroke" with protected TAVR, Cohen said. He estimated the relative risk reduction to be about 20%, with a number needed to treat of 300 to prevent one major stroke.

The modest stroke protection is "sobering" for the sole embolic protection device FDA approved for commercial use, the Sentinel, which is supposed to protect three out of the four great vessels going to the brain (i.e., the left and right carotid arteries and the right vertebral artery), the presenter said during a press conference.

Registry data thus support clinical equipoise and provide a strong rationale for ongoing large-scale randomized clinical trials testing whether embolic protection devices offer meaningful clinical benefit to patients undergoing TAVR, he concluded.

The registry study included transfemoral TAVR cases from January 2018 to December 2019, excluding emergent procedures, concurrent mitral procedures, and low-volume sites.

Cohen and colleagues found a gradual increase of hospitals adopting embolic protection in TAVR, reaching 28% by late 2019. The proportion of patients receiving protection reached 13%.

Hospitals varied widely in their use of embolic protection: 66% never used it at all during the study period, whereas 5% used it in more than half of their TAVR cases.

Embolic protection was not associated with increased vascular complications, major bleeding, or device failure, Cohen reported.

Comment