The Take Home: ACC
The American College of Cardiology Scientific Session produced some groundbreaking findings with major implications for the practice of interventional cardiology. Most stunning was two trials that showed transcatheter aortic valve replacement bested surgery in low-risk patients with severe aortic stenosis; reaction to those data appears in the Cover Story.
Cardiology Today’s Intervention covered the meeting on-site and spoke to numerous prominent interventionalists, including Chandan Devireddy, MD, FACC, FSCAI, from Emory University School of Medicine; Cardiology Today’s Intervention Editorial Board Member Howard C. Herrmann, MD, from the Hospital of the University of Pennsylvania; Cardiology Today’s Intervention Associate Medical Editor Roxana Mehran, MD, from Icahn School of Medicine at Mount Sinai; Cardiology Today’s Intervention Editorial Board Member Sunil V. Rao, MD, FSCAI, FACC, from Duke University Medical Center and Durham VA Medical Center; Michael J. Reardon, MD, from Houston Methodist Hospital; Michael J. Rinaldi, MD, from Sanger Heart and Vascular Institute, Atrium Health; and Arnold H. Seto, MD, MPA, from Tibor Rubin VA Medical Center in Long Beach, California.
Seto: We know that 10% of patients who receive PCI also have atrial fibrillation, but have limited data on the safety of dual- and triple-therapy regimens. Too often physicians merely “add-on” additional antithrombotic therapies — either DAPT or anticoagulation — without sufficient thought to the safety of multiple therapies, and yet we know that triple therapy with aspirin, a P2Y12 antagonist and oral anticoagulant is associated with a high 4% to 16% (average 5%) major bleeding rate.
Part of the results of the AUGUSTUS trial, which studied 4,614 patients with AF who underwent PCI or had ACS, largely mirror the PIONEER-AF-PCI, RE-DUAL PCI and WOEST results, which suggested that dropping the aspirin and leaving patients on an anticoagulant and P2Y12 inhibitor was associated with reduced risk for bleeding. The other part of the trial, and what was most surprising, was the magnitude of the difference in bleeding between the vitamin K antagonist and apixaban (Eliquis, Bristol-Myers Squibb/Pfizer) groups. Patients who received apixaban had a 31% reduction in risk for major or clinically relevant nonmajor bleeding vs. those who received warfarin. Bleeding risk was reduced by 47% among patients who received placebo vs. aspirin. The highest rate of bleeding occurred in patients who were treated with clopidogrel, warfarin and aspirin (18.5%) and the lowest rate was in those treated with clopidogrel, apixaban and placebo (7.3%).
None of these trials are powered for efficacy, however, so there will always be a question in this regard.
The AUGUSTUS results reinforce the conclusions of the AF guidelines, which favor a direct oral anticoagulant over a vitamin K antagonist for anticoagulation. Due to the drugs still being newer and more costly than warfarin, pharmacy and insurance restrictions have slowed the uptake of direct oral anticoagulants, but the additional data and guidelines will continue to force change for the benefit of patients.
For patients with AF receiving PCI, AUGUSTUS reaffirms that triple therapy is unnecessary for many patients, and de-escalation of therapy can occur possibly as early as 14 days following PCI.
With the newer P2Y12 antagonists ticagrelor (Brilinta, AstraZeneca) and prasugrel (Effient, Daiichi Sankyo/Eli Lilly) showing superiority to clopidogrel in ACS, future trials would ideally compare each P2Y12 antagonist in the setting of a direct oral anticoagulant, to see if the risk/benefit equation of efficacy/risk would still favor ticagrelor and prasugrel. I believe it would, however, the patient’s bleeding risk would need to be taken in to account.
Mehran: This was the largest trial in patients who require an oral anticoagulant and present with ACS or have a need for PCI. Because of the 2x2 factorial design, the trial was able to show that apixaban was superior in reduction in bleeding events compared with vitamin K antagonists and aspirin withdrawal in these patients with AF undergoing PCI who require another antiplatelet regimen plus an oral anticoagulant will reduce bleeding by 90%. That was an incredible finding.
The caveat is that I still believe one size does not fit all. The trial was not powered for the important rare ischemic events that may occur in this population. I don’t think we can say an oral anticoagulant plus a P2Y12 inhibitor without aspirin is the optimal therapy for 100% of these patients, but we can for a large majority of them, which is a giant leap forward.
Rinaldi: In two substudies of COAPT, one looked at quality of life metrics and the other looked at echocardiographic parameters and predictors.
In the echocardiographic substudy of 614 patients, mitral regurgitation improved more in the group assigned transcatheter mitral valve replacement (MitraClip, Abbott) at all times of follow-up compared with the group assigned medical therapy. In the quality of life substudy of 611 patients, Kansas City Cardiomyopathy Questionnaire score was better in the device group at 1 month and 24 months.
Patients who received clip therapy compared with patients who did not receive it saw their metrics in quality of life improve substantially. Those improvements were fairly immediate, and they were durable. The curves were fairly parallel, as the benefit did not drop off over time. The improvement in quality of life was across the board and not just in some small subgroup of patients, which is also important to note.
Previous studies and the COAPT trial showed that when you treat patients with the MitraClip, it mechanically improves the way their heart functions. Their ventricular volumes were remodeled in a positive way and everything improves from an echocardiographic standpoint. The heart pumps better, and that correlates completely with patients’ quality of life.
The question about placebo effect correlates with physiologic parameters, measuring how healthy the ventricle is with these ventricular volumes and ventricular measurements. It all fits tightly in a nice and positive way.
We know that patients with true mitral insufficiency benefit from MitraClip therapy, and the echocardiographic substudy tells us that it is not all grouped together. Whether patients had 3+ or 4+ mitral insufficiency, they similarly improved.
When researchers looked at all the subgroups, no matter what your left ventricular ejection fraction, you benefited. It did not matter how big the ventricle was, as all those markers of the health of the ventricle did not matter. Treating mitral insufficiency, if it was severe, helped patients.
There are other markers that suggest something sick about the heart that may make care futile. For instance, severe tricuspid insufficiency is associated with a poor prognosis. If you had no tricuspid insufficiency or 2+ to 3+ tricuspid insufficiency, you benefited equally from clip therapy as well. Even in patients with pulmonary hypertension, the clip group benefited regardless of severity.
These substudies reinforce the principles of the COAPT trial and showed a marked and clinically relevant benefit.
Devireddy: COAPT was a very dramatic result, and these two reports further demonstrate that mitral clipping has benefits in these carefully selected patients.
While patients were chosen based on these strict echocardiographic parameters and clinical features, they had a fairly dramatic regression in LV enlargement, which is a big problem for these patients.
There was some nuance that was raised and discretion about the fact that patients with pulmonary artery hypertension were excluded, but then as the trial went forward, the RV pressure did seem to be some marker of identification of patients at higher risk. That is going to need to be teased out a little bit about what that means in patient selection and in monitoring these patients overall. It is a fairly dramatic, sustained reduction in mitral regurgitation in these carefully selected patients.
With the recent FDA expanded indication for this device, all eyes are on the payers to provide some guidance of how they are going to navigate the pathway of patient selection.
Herrmann: It is important that we analyze this data set carefully to understand not only who the best candidates are for MitraClip in the functional mitral regurgitation space, but also to understand who did not benefit. We still have a high rate of mortality and HF hospitalization even in the treated group: close to half of patients at 2 years. It is hard to feel good about that despite the demonstrated improvement from the control group. We still have a lot to learn and analyze.
The echocardiographic substudy suggested that it may be difficult to select the best responders and those that won’t respond based on echocardiographic parameters alone. This analysis was more comprehensive than the initial subset analysis that was presented at TCT by Gregg Stone, MD, where a subset of people who had less mitral regurgitation and larger ventricles seemed to not get benefit, similar to the findings in MITRA-FR.
That analysis was done with a dichotomous cut point on LV end-diastolic volume and the primary endpoint. This analysis was done using LV and end-diastolic volume as a continuous variable and a slightly different endpoint. The present study used mortality and first HF hospitalization rather than mortality and all HF hospitalizations. I am not sure whether this explains the difference.
This study suggests that we still do not know exactly how to pick the patients who are going to get the most benefit from MitraClip beyond the COAPT trial general inclusions and exclusions. I would like to understand better why the patients who did not get benefit did not get it, so that we can avoid unnecessary therapy and potentially offer other therapies to these patients.
The quality of life analysis was impressive. The researchers demonstrated real and sustained benefits and answered a lot of potential statistical criticisms with careful analyses.
Additional analyses that I would like to see include those based on residual mitral regurgitation and stratified by ejection fraction to try to understand how much of the observed benefit is due to mitral regurgitation reduction. We presume it is, but does it matter how much mitral regurgitation reduction? Is there a role for transcatheter mitral valve placement because one gets better mitral regurgitation reduction, even though it is a more complex procedure with its own issues of safety and efficacy?
With the FDA approval for secondary mitral regurgitation, these questions have become even more important and relevant. We need to carefully understand not only who benefits, but also who does not benefit so that we can apply these technologies rationally and to our patients’ best benefit.
TAVR in Bicuspid Valves
Reardon: Early-generation TAVR valves did not perform very well in patients with bicuspid valves, but indications are the newer generations are better. What the researchers did was look at patients with bicuspid valves treated with Sapien 3, a later generation of the balloon-expandable valve (Edwards Lifesciences). After propensity matching of 2,691 patients in each group, they found that patients with bicuspid valves did just as well as patients with tricuspid valves in terms of death at 30 days (P = .82) and stroke or death at 1 year (HR = 0.97; 95% CI, 0.81-1,16).
Jeffrey J. Popma, MD, and colleagues previously used the same registry to show the same thing for the Evolut (Medtronic), a later-generation self-expanding valve.
The thing to bear in mind is that all these patients have been selected by their implanting physicians, who have determined that this is a bicuspid valve on which they are comfortable performing TAVR. We don’t know how many have been screened out by the implanting teams, although I think it’s a relatively small number.
We are already treating patients with bicuspid valves with TAVR if they are at intermediate risk or high risk for surgery. We feel very comfortable treating bicuspid valves in those populations. The patients who we decide are better off with surgery tend to have aneurysms that are getting borderline large or other anatomic characteristics.
I am study chair of a nonrandomized trial that is underway looking at TAVR in low-risk patients with bicuspid valves. It would be great to do a randomized trial, but I think at this point it’s not pragmatically possible, now that data from the large randomized trials showing the benefit of TAVR in low-risk patients are available. We expect to produce the early results of that study by the end of the year. We want to see how endpoints such as paravalvular leak, stroke and mortality are at 30 days. The truth is, if you are doing OK at 30 days, you will probably do OK long-term. If the valve works anatomically, it will continue to work.
In the intermediate- and high-risk bicuspid populations, we are already saying that if a patient can have either TAVR or surgery, they get TAVR. If they are not anatomically suitable for TAVR, they get surgery. If you see a patient who is a candidate for a bioprosthetic surgical valve, you have to talk to them about TAVR as a potential option or you are not giving them complete informed consent. That does not mean they have to have TAVR, but it needs to be part of the conversation.
This is a seismic shift because 80% of this population has bicuspid valves. Not all of them are going to be candidates for TAVR, but at least 50% of them, if not more, would be bioprosthetic candidates.
Devireddy: The RADIANCE-HTN SOLO trial produced fascinating data. It is a small (69 patients) randomized controlled trial of a renal denervation system (Paradise, Recor Medical) against a sham procedure, a level of evidence that we rarely see in CV trials.
The initial endpoint that had been reported was 2-month difference in BP and safety. The prespecified endpoints were met.
What happened from 2 months to 6 months is that patients were then allowed to be titrated on medications per the discretion of the treating physician, but with their treatment assignment still blinded. If the patient was hypertensive, he or she was started on medications per a prespecified pathway of titration.
What the study showed was that patients who had received renal denervation had a better ability to reach treatment guideline goals compared with patients in the sham procedure group. The mean systolic BP reduction at 6 months was 18.1 mm Hg with renal denervation vs. 15.6 mm Hg with sham control (difference adjusted for baseline BP and number of medications, 4.3 mm Hg; 95% CI, –7.9 to –0.6; P = .024).
There were some interesting bar graphs from the study. Comparing the renal denervation vs. the sham procedure group, in the patients who reached goal in both groups, there was a difference in how much medication they needed to get to that goal BP. It seems to be — and this has been suggested from these trials over the last year — that renal denervation is providing an effect that may be equivalent to adding another BP medicine but without having to take that medicine every day, and all the issues that go along with compliance.
Next will be 12-month data, and some of the larger trials are hoping to finish enrollment this year. If we are lucky, we can see another late-breaking trial in the next 12 months.
Rao: SAFARI-STEMI is a bold study and the investigators should be congratulated for doing this challenging study. It is the largest dedicated STEMI transradial vs. transfemoral trial performed, even though, at 2,292 patients, it stopped well short of its planned sample size.
The rates of 30-day mortality were similar in both groups (transradial, 1.5%; transfemoral, 1.3%; RR = 1.15; 95% CI, 0.58-2.3), and did not vary by age, sex, administration of bivalirudin, preloading with ticagrelor (Brilinta, AstraZeneca), BMI, creatinine clearance or diabetes status. There were also no differences between the groups in reinfarction (RR = 1.07; 95% CI, 0.57-2), stroke (RR = 2.24; 95% CI, 0.78-6.42), death/reinfarction/stroke (RR = 1.17; 95% CI, 0.77-1.79), stent thrombosis (RR = 1.07; 95% CI, 0.57-2) or any metric of bleeding.
This study showed how challenging doing trials in the radial and STEMI spaces can be. It did reiterate that TIMI bleeding is probably not the correct bleeding metric to use for vascular access site strategy trials (RIVAL also showed no difference between transradial and transfemoral with respect to TIMI bleeding). It also showed that when bleeding is measured, transfemoral performance improves significantly.
In the STEMI setting, time is still of the essence, so until operators become proficient with transradial access, they should use whatever access site they are comfortable with. However, since the total body of data still supports transradial access, interventional cardiologists and cath lab staff should become proficient with the transradial approach for non-urgent cases and eventually adopt a “radial first” approach for primary PCI for STEMI as well.
One important thing from the presentation was that even if you include SAFARI-STEMI, the pooled data still support trans-radial access for STEMI PCI.
Disclosures: Devireddy reports he is on the data safety monitoring board for Medtronic and works at a participating center for the PARTNER 3 study with the Edwards Lifesciences TAVR device. Herrmann reports he was an investigator in the COAPT trial, consults for companies including Edwards Lifesciences and Medtronic and received institutional research funding for several mitral valve therapy trials. Mehran reports she has financial ties with multiple pharmaceutical and device companies, including AstraZeneca, Bayer, Bristol-Myers Squibb, Daiichi Sankyo and Janssen, and served on the executive committee of the AUGUSTUS trial. Rao and Reardon report no relevant financial disclosures. Rinaldi reports he was a primary investigator for the COAPT trial; teaches courses for Abbott and Edwards Lifesciences; is a speaker for Abbott, Boston Scientific and Edwards Lifesciences, and is on the advisory board for Abbott and Boston Scientific. Seto reports he is on the speaker’s bureau for Janssen.