The Take Home: TCT

Issue: January/February 2017

The Transcatheter Cardiovascular Therapeutics Scientific Symposium was held from Oct. 29 to Nov. 2, 2016, in Washington, D.C. Cardiology Today’s Intervention was onsite and spoke to some of the most prominent experts in interventional cardiology, including Cardiology Today’s Intervention Chief Medical Editor Deepak L. Bhatt, MD, MPH, with Brigham and Women’s Hospital and Harvard Medical School; Samir Kapadia, MD, with Cleveland Clinic; Patrick T. O’Gara, MD, with Shapiro Cardiovascular Center at Brigham and Women’s Hospital and past president of the American College of Cardiology; and Vinod H. Thourani, MD, with Emory University, for their opinions on the most important presentations.

EXCEL and NOBLE

Bhatt: EXCEL and NOBLE were probably the most anticipated presentations at TCT. These trials examined patients with left main disease undergoing PCI vs. CABG. The results were very interesting because EXCEL showed noninferiority of PCI vs. CABG, but PCI did not meet noninferiority criteria vs. CABG in NOBLE.

The results looked discordant on the surface. In EXCEL, the primary endpoint of death, MI or stroke was similar between the groups at 3 years (PCI group, 15.4%; CABG group, 14.7%; HR = 1; 95% CI, 0.79-1.26; P for noninferiority = .02). In NOBLE, the primary endpoint of death, stroke, nonprocedural MI or repeat revascularization at 5 years occurred in 29% of the PCI group vs. 19% of the CABG group (HR = 1.48; 95% CI, 1.11-1.96; P for superiority of CABG = .0066). I am sure there will be much discussion in the interventional community and in the cardiac surgery community as to which trial was better, which was more representative as a whole, and which was more representative of U.S. sites and U.S. practice.

However, there were some common themes. There were signals in both trials that in longer duration of follow-up with respect to the endpoint of MI, CABG was better than PCI. In EXCEL, and going back to earlier trials such as SYNTAX, an area where PCI generally beats CABG is for early periprocedural stroke. For target vessel revascularization, current and older data support the superiority of CABG.

I think that what it all comes down to for complex revascularization such as for patients with left main disease is decision-making by a heart team, much like we are doing with transcatheter aortic valve replacement these days. For patients with ostial left main disease, PCI with a current-generation DES is probably the way to go. For patients with, for example, more complex bifurcation left main disease, a high SYNTAX score and chronic total occlusions or additional bifurcation lesions, CABG is probably preferred. Even there, heart team input is important because if a patient is at very high risk for stroke or surgical complications, maybe they should undergo PCI, but similar patients with low surgical risk should probably undergo surgery.

PCI and CABG are two good options for left main disease, but exactly which one to use depends on the patient’s risk factors, not only clinical risk factors such as stroke but also angiographic risk. Integrating all those data into a heart team approach will provide the best outcomes to our patients.

Kapadia: EXCEL and NOBLE appear different but are similar in many respects. In the short term, results for PCI vs. CABG were similar at 1 year. It’s a good message that both can be done safely.

Later, the curves separated, which was apparent after 5 years in NOBLE. But even in EXCEL, the mortality curves had started to separate at 3 years. The patient population was quite similar in both, though we don’t know how many lesions were revascularized and how complicated revascularization was for the patients in NOBLE.

PAGE BREAK

Revascularization rates of left main arteries with stents were similar and low in both trials. The difference came from other areas of revascularization, not the left main itself. It is a good message that if the anatomy is not very complicated, and you are not going to revascularize many areas other than the left main, then outcomes are outstanding.

I’m not sure that the differences in the stents used (EXCEL: Xience, Abbott Vascular; NOBLE: BioMatrix Flex, Biosensors) explain everything. We do not even have the stent they used in NOBLE.

WATCHMAN US POST-APPROVAL STUDY

O’Gara: One of the most interesting late-breaking clinical trials at TCT concerned the real-world experience with use of a left atrial appendage closure device (Watchman, Boston Scientific) for stroke prevention in patients with nonvalvular atrial fibrillation who are poor candidates for anticoagulation.

There has been concern in the community that this device and its implantation can be a complicated procedure with significant complications and morbidity. The investigators of this study reviewed 3,822 implantations, the majority of which were done by low-volume and relatively inexperienced operators who underwent a proctoring experience.

The implant success rate was 95.6% and median procedure time was 50 minutes. The overall complication rate was 1.63%, the rate of pericardial effusion was 0.29%, the rate of procedure-related stroke was 0.078%, the rate of device embolization was 0.24% and the rate of death was 0.078%.

It’s gratifying and relieving to see that the rate of major complications with the device are relatively low, especially the risk for pericardial tamponade, device migration or procedure-related death. This is a very important addition to our understanding of where these devices might fit in in the spectrum of interventions we can apply to patients who have AF and are at risk for stroke or systemic embolism.

SENTINEL

Thourani: One of the biggest concerns for those of us who do transcatheter therapies is postprocedural stroke. It is sometimes more devastating than dying from a procedure. It is very important for us to find ways to decrease stroke after transcatheter valve therapies.

I was one of the authors of the SENTINEL study, the first large randomized trial evaluating a neuroprotection device (Sentinel, Claret Medical). We implanted the device in 363 patients at 19 U.S. centers. Patients were stratified into three groups: a safety arm, a device imaging arm and a control imaging arm. All patients underwent transcatheter therapies with a balloon-expandable or self-expanding valve. All had an MRI performed after the procedure within a week before discharge. This was a remarkable and well-designed trial.

What was found was that according to an analysis from an adjudicated core lab, 99% of patients had debris within the device. That’s quite remarkable.

However, most likely because the patient population was small, there were no changes in neurocognitive function found to be within the device or control arms. The rate of MACCE in the device groups was 7.3%, which was noninferior to the historical performance goal of 18.3% (P for noninferiority < .001) and was not statistically different from the control group (9.9%; P = .41).

Also, MRI results showed new lesion volume was greater in the control imaging group than in the device group (178 mm3 vs. 102 mm3; reduction, 42%). However, this was not statistically significant (P = .33) despite the large reduction in infarct sizes within the brain on MRI. Although we saw a large amount of capture, we did not see statistical differences between the control arm and the device arm, most likely because the power was on the small side.

PAGE BREAK

We’re seeing for the first time an adjunctive device being used in transcatheter valve procedures to decrease some of the morbidity. That should help us design and implement a variety of adjunctive therapies to decrease stroke, one of the most common and devastating morbidities associated with TAVR.

O’Gara: One trial that caught my attention has to do with the use of embolic protection during the performance of TAVR. This is of course to reduce the risk for stroke, which is the Achilles’ heel of both transcatheter and surgical AVR. Any randomized trial designed to answer whether we can reduce stroke incidence is of very high importance to the cardiology and cardiac surgical community.

The SENTINEL trial was possibly not powered to show a significant difference with the use of the embolic protection device using an endpoint of MRI-determined volume of a new diffusion-weighted imaging abnormality assessed 2 to 7 days after the procedure.

Although SENTINEL might be advertised as a negative trial, it adds to the body of information and validates the need for us to continue to push the envelope of investigation so we can help these patients.

I have to admit I’m a bit disappointed. This may not be different, unfortunately, from what we’ve seen in surgical trials. We may need to go back to the drawing board and see what other kinds of interventions might be available to reduce stroke.

Kapadia: When we started doing TAVR in the United States in 2005, there was evidence that TAVR was better than medical therapy for people with severe aortic stenosis who are not candidates for surgery, but a 5% stroke rate was identified as the most important limitation of the technology. The stroke rate was higher for TAVR than surgery in the first PARTNER trial. There was no neurological evaluation in the PARTNER trial because at that time we didn’t know better, and identification of stroke was a bit of an issue. Since then, TAVR has shown a lower stroke rate than surgery, but it is still 2.5% to 3%.

Then we began to ask how we could protect against stroke during TAVR. The Sentinel device is one of two under investigation in the U.S. When we designed the trial, on which I was a principal investigator, we used a surrogate endpoint of new lesions on MRI. One vertebral artery is not covered by this device, and there was debate about whether to check for lesions in protected territories or all territories; we decided to check in protected territories.

The most important question was, if you manipulate all these devices in the carotid artery, will that itself cause more strokes? We found the device was extremely safe. There was no increased risk for stroke just because of the device. The stroke rate in the non-device group was 9.1%; the number is that high because neurologists are evaluating the patients. The device group’s stroke rate was 5.6%, a 44% reduction associated with protection by the device. This was clinically significant but statistically not significant. What does that mean?

In addition, new lesion volume was reduced in the device group but was also statistically not significant. Approximately 25% of patients were lost to follow-up because they required a pacemaker, and we could not perform MRI on them. We knew something like that would happen.

This was an exciting trial for the field because it was the first of its kind and the technology is new. There are three other trials with this device and one with another (TriGuard HDH, Keystone Heart) that all point in the same direction. Whether there will be an FDA panel to investigate these devices and who will use them are questions which will come up.

In our clinical practice, a lot of people ask us about these devices, because they know stroke is the most important fear for patients undergoing TAVR.

PARTNER I FIVE-YEAR ECHO

Kapadia: This was a core lab-adjudicated trial designed in response to concerns that TAVR valves are not durable. The results were amazingly parallel. Whether you measured the gradient, the valve area or anything else, the valves were identical at baseline, 3 months and 5 years and the curves were superimposable.

This gives us a lot of confidence to say that, at least at 5 years, we are not seeing a difference, and concerns about durability may not be warranted.

Disclosure: Bhatt, Kapadia, O’Gara and Thourani report no relevant financial disclosures.