DEFLECT III: 30-Day Results Show Successful, Safe TAVR With Novel Neuroprotection Device

PARIS, France—Cerebral protection during TAVR with a novel neuroprotection device results in greater freedom from ischemic brain lesions, fewer neurological deficits, and improved cognitive function at 30 days, according to results of an exploratory study presented May 19, 2015, at EuroPCR and published simultaneously online in the European Heart Journal.

Next Step:  DEFLECT III: 30-Day Results Show Successful, Safe TAVR With Novel Neuroprotection Device

The TriGuard device (Keystone Heart; Caesarea, Israel) consists of a nitinol single-wire frame and a mesh filter (pore size 130 μm). It is delivered femorally with a 9-Fr sheath, positioned across all cerebral vessels, and maintained by a stabilizer.

Preliminary intention-to-treat data from DEFLECT III showed promise for the device, with 87% technical success—coverage of all 3 vessels—and equivalent in-hospital safety outcomes with and without TriGuard.

Andreas Baumbach, MD, of Spire Bristol Hospital (Bristol, England), presented the final 30-day results of 85 TAVR patients (average age 82.4 years; 45.9% men) randomized to the procedure with (n = 46) or without embolic protection (n = 39) at 13 centers in Europe and Israel between February 2014 and March 2015.

Technical success in this case was 88.9%, and there were no instances of interference with the TAVR valve or device failure. Rates of the primary endpoint (MACCE: all-cause mortality, all stroke, life-threatening bleeding, stage 2/3 acute kidney injury, and major vascular complications) and its individual endpoints were similar between the study and control groups both in the hospital and at 30 days (table 1).

Table 1. Safety Outcomes

There were 2 in-hospital strokes in each group, 1 death from pneumonia in a stroke patient in the TriGuard arm, and 2 acute procedural deaths due to aortic ring rupture in the control arm. No additional strokes or deaths occurred in either group at 30 days. 

Complete freedom from ischemic brain lesions on diffusion-weight MRI was a relative 46% higher in the TriGuard group compared with controls in the intention-to-treat population, and a relative 57% higher than controls in the per-protocol population. TriGuard was also associated with fewer new neurologic deficits than seen in controls as detected by a worsening in NIH Stroke Scale score (3.1% vs 15.4%), improved Montreal Cognitive Assessment (MoCA) scores, better performance on a delayed memory task at discharge, and a more than 2-fold increase in recovery of normal cognitive function (MoCA score > 26) at 30 days.

Mean total fluoroscopy time was longer in the TriGuard group (28.4 vs. 18.8 min; P < .001). Two-thirds of patients were placed on dual antiplatelet therapy, while the remainder received either aspirin or clopidogrel monotherapy.

‘Lesions Cannot Be Ignored’

“Neuroprotection with the TriGuard device in DEFLECT III was safe,” Dr. Baumbach concluded. “It confirms improved diffusion-weighted MRI surrogate measures [and] it provides new evidence of clinical benefit.”

In a population with ongoing embolic risk, he continued, neuroprotection is important. The upcoming REFLECT randomized trial “should confirm our results,” Dr. Baumbach reported.

Session moderator Stephan Windecker, MD, of Bern University Hospital (Bern, Switzerland), noted the discrepancy between clinical and imaging event rates. “If you look at clinical event rates in contemporary series, it’s about 3-4% of stroke, yet if you perform imaging, it’s nearly every patient who has lesions. Are we too insensitive with clinical observations? Or are we oversensitive with the imaging?” he asked.

Dr. Baumbach explained that “emboli and the things we see have a large spectrum of sequelae. So of course it matters what kind of volume or lesions you have in the brain, but it also matters where the location of the infarct is for the clinical picture. Volume alone doesn’t give us information about the severity of those lesions.”

But after results from this study and CLEAN-TAVI, “we cannot ignore these lesions,” he stressed. “They may not produce overt stroke that we as cardiologists are able to detect, but they certainly produce neurological deficits that a neurologist can see.”

Panelist Neil E. Moat, MBBS, MS, of the Royal Brompton and Harefield National Health Service Foundation Trust (London, England), asked “why the magnitude of the change you saw by the imaging lesions is less than you hoped.”

“The absolute [change] doesn’t show you the full story,” Dr. Baumbach replied, adding that the location of the lesion and the patient’s neurological outcome are “most important” and will be heavily weighted in the REFLECT trial. “We do see a previously unseen rate of patients without any new embolic lesion and that should be the target that we want to achieve in these patients.”

Updated VARC Criteria to Elucidate Findings Further

In an accompanying editorial, A. Pieter Kappetein, MD, of Erasmus University Medical Center (Rotterdam, the Netherlands), writes, “The analysis and the importance of embolic debris and especially comparisons between devices is not so straightforward.”

Although the researchers refer to the fact that only 2% of patients did not have ischemic brain lesions using the Montage filter (Claret) in CLEAN-TAVI, Dr. Kappetein says, “the discrepancy in endpoint and stroke definition, the number of neurological tests performed, and the way strokes are assessed makes it difficult to compare differences in incidence across studies…. There is no doubt that the more neurological tests are performed the more neurological deficits will be found.”

Improvements in technology, better patient selection, and increased operator experience decreased the stroke incidence by up to 4%, he observes, “emphasizing the need for surrogate endpoints to be used in order to be able to compare outcomes across studies.”

DEFLECT III “is a major step forward as they introduced a set of neuropsychological tests able to show a > 2-fold increase in the proportion of patients with restoration of normal cognitive function using the protection device,” he continues. “The study also shows the limitation of these tests as in > 25% of the patients the 30-day results were not obtainable.”

Lastly, the study was based on VARC-2 criteria, which “does not provide guidance on cerebral imaging endpoints, timing of imaging, or integrating a neurocognitive test battery,” Dr. Kappetein writes, adding that a VARC-3 update is planned for 2015. “Hopefully, with these new criteria, trials can be conducted to test novel devices and new therapeutic regimens (anticoagulation) while they will also enable comparison across studies.”


1. Lansky AJ, Schofer J, Tchetche D, et al. A prospective randomized evaluation of the TriGuard HDH embolic deflection device during transcatheter aortic valve implantation: results from the DEFLECT III trial. Eur Heart J. 2015;Epub ahead of print.
2. Kappetein AP. Reflect on DEFLECT. Eur Heart J. 2015;Epub ahead of print.


  • The study was supported by Keystone Heart and the National Institute of Health Research Bristol Cardiovascular Biomedical Research Unit.
  • Dr. Baumbach reports serving on the scientific advisory board and receiving research grants from Keystone Heart.
  • Dr. Kappetein reports no relevant conflicts of interest.

Related Stories