The Take Home: ACC Scientific Sessions

Issue: May/June 2013

In March, attendees from all over the world gathered in San Francisco for the 62nd annual American College of Cardiology Scientific Sessions. In total, 19,627 attendees from 109 countries took part in the 3-day meeting, which featured 1,845 poster presentations and 466 sessions.

Cardiology Today’s Intervention was onsite during the meeting and spoke with several experts on some of the most anticipated and talked-about trials presented. Among them were Cardiology Today’s Intervention Chief Medical Editor Deepak L. Bhatt, MD, MPH, with Harvard Medical School; Editorial Board members Paul A. Gurbel, MD, with the Sinai Center for Thrombosis Research at the Sinai Hospital of Baltimore, Dean J. Kereiakes, MD, with the Christ Hospital Heart and Vascular Center at The Christ Hospital, Cincinnati, and Matthew Price, MD, with Scripps Clinic, San Diego; and Society for Cardiovascular Angiography and Interventions representative Stephen R. Ramee, MD, with the Ochsner Medical Center in New Orleans.

CHAMPION PHOENIX

Paul A. Gurbel

Gurbel: The CHAMPION PHOENIX trial supports the platelet hypothesis: Therapy with the more potent platelet inhibitor strategy (ie, cangrelor, The Medicines Company) in the patient at high risk for thrombosis will result in less thrombotic event occurrence. In CHAMPION PHOENIX, the potent thrombogenic stimulus was PCI. Similar observations were made many years ago in the area of PCI with parenteral glycoprotein IIb/IIIa inhibitor therapy. In the ESPRIT trial, patients undergoing non-urgent PCI who were treated with eptifibatide, a reversible agent, had less subsequent ischemic events and the benefit was seen early, within 48 hours, most notably in periprocedural MI. The ESPRIT trial was stopped early because of superior efficacy with eptifibatide vs. placebo. In the CHAMPION PHOENIX trial, use of a reversible parenteral P2Y12 inhibitor, cangrelor, in patients not sufficiently pretreated with clopidogrel (Plavix, Sanofi-Aventis) was associated with a similar early benefit, in this case a lower rate of a composite of death, MI, ischemia-driven revascularization or stent thrombosis 48 hours after randomization compared with clopidogrel (4.7% vs. 5.9%; adjusted OR=0.78; P=.005). Cangrelor produces a much more rapid and potent level of P2Y12 inhibition than clopidogrel and its pharmacodynamic superiority vs. clopidogrel directly translated into a superior antithrombotic effect. We also learned from the TRITON and PLATO trials that more potent P2Y12 blockade with the new oral P2Y12 inhibitors, prasugrel (Effient, Daiichi Sankyo/Eli Lilly) and ticagrelor (Brilinta, AstraZeneca), also translated into superior antithrombotic effects.

In summary, the clinical response observed in large-scale trials has followed the pharmacodynamic effect and this is exactly what happened in the CHAMPION PHOENIX trial. For me, this trial further solidifies the important role that platelet physiology plays in the genesis of thrombotic events early after PCI. It is a very important study and the investigators should be congratulated for their contribution. In the United States, many coronary interventions are performed ad hoc in the absence of adequate clopidogrel pretreatment. Cangrelor provides another important option for the interventional cardiologist to achieve immediate, potent and reversible platelet inhibition.

PREVAIL

Matthew Price

Price: The objective of the PREVAIL trial was to acquire more data regarding procedural safety with the Watchman device (Boston Scientific) and to confirm the efficacy that was observed in PROTECT AF. The data show that there has clearly been an improvement in procedural safety over time, with the success rates for implanting the device significantly higher (95.1% vs. 90.9%; P=.04) and vascular complications lower (4.6% vs. 8.7%; P=.004) than in PROTECT AF. What is really impressive is the low risk for adverse procedural events with operators who are new to the procedure, which shows that with newer techniques, operators are cognizant of potential problems and it appeared to be a much safer procedure than previously. The important efficacy outcome with regard to AF was the rate of ischemic stroke and systemic embolism, which is really what we are trying to prevent when we give anticoagulation. The results show that the noninferiority criteria for this endpoint was met. This builds upon the recent data looking at long-term outcomes in PROTECT AF, which by landmark analysis showed that there was a superior outcome with left atrial appendage occlusion compared with warfarin (Coumadin, Bristol-Myers Squibb) with respect to ischemic stroke, which has not been demonstrated with the oral factor Xa inhibitors.

I found these results provocative in light of this long-term, landmark data from PROTECT AF because the landmark analysis showed superiority with ischemic stroke and systemic embolism, but this was counterbalanced by concerns with the early upfront procedural hazard. It’s clear with PREVAIL that this hazard has been reduced with increasing knowledge and experience with the procedure. Therefore, the long-term PROTECT AF data combined with the safety data from PREVAIL really lay important groundwork for this technology moving forward in clinical practice. We will of course need to carefully assess the composite clinical endpoint as more follow-up accrues, and it will be informative to combine the efficacy outcomes of both PROTECT AF and PREVAIL, given the relatively small sample size enrolled in the latter study.

However, interpreting the results of PREVAIL may prove challenging for clinicians because it uses a novel approach called Bayesian noninferiority. Bayesian analyses have several advantages compared with classic randomized clinical trials in that you incorporate your pretest and post-test probabilities as time goes on; so as you get more information, you can understand more of the effect of your intervention. As more trials will likely be using such a design in the future, it will be very important for us to understand how to appropriately interpret the results of Bayesian methods so that we can assess such data critically and accurately.

PAGE BREAK

PARTNER II cohort B

Stephen R. Ramee

Ramee: The PARNER II cohort B trial was a very important trial that was a follow-up to the PARTNER I cohort B trial, which looked at inoperable patients receiving transcatheter aortic valve replacement. The definition of inoperable in this situation was patients in whom two surgeons consider are too high risk to undergo aortic valve surgery with open surgical repair. These tend to be patients who without TAVR would die. The mortality risk with surgery is estimated to be more than 50%. The importance of PARTNER II cohort B was it compared the new and commercially unavailable valve, Sapien XT (Edwards Lifesciences), with the commercially available Sapien valve. The new valve is lower in profile (18F vs. 24F), can be put in patients with smaller iliac arteries and has a lower risk of rupture of those arteries and vascular complications.

Fortunately, all major endpoints were the same between the old and new valve in terms of durability, safety and freedom from cardiac events with a similar survival benefit. The one important difference was that vascular complications were nearly cut in half with the new lower-profile valve (9.6% vs. 15.5%; P=.04). The importance of that is multiple, but the main thing is we will be able to treat inoperable patients with smaller iliac arteries once the FDA acts on these data and approves this device, which will greatly expand the population of patients we can treat. In my own patient population, for instance, I have 40 patients waiting for the FDA to approve the Sapien XT valve because their iliacs are too small for the currently approved device and they are too sick to undergo surgery. So I am really looking forward to the FDA acting quickly on these data and the company being able to release the valve. It is going to save a lot of lives.

STREAM

Deepak L. Bhatt

Bhatt: The STREAM trial was an interesting study of nearly 2,000 patients with STEMI who presented within 3 hours of onset of chest pain and were unable to get to a cath lab for primary PCI for at least an hour. Patients, who were all outside the United States, were randomly assigned to one of two treatment arms: either treatment with primary PCI as soon as possible (although not within the first hour) or to receive fibrinolytic therapy and then go on to catheterization within the next 6 to 24 hours, although rescue PCI was allowed. The overall trial did not find any significant differences in the primary endpoint, which was a composite of various ischemic endpoints between those two arms of the study.

As a result, the authors concluded that either approach could be used and in fact the lytic-based approach may be preferable because it allows for the possibility of deferring urgent PCI until later, for example the next morning instead of tonight in the middle of the night. That is an appealing sort of interpretation. I happened to be the editorialist for the study (Bhatt DL. N Engl J Med. 2013;368:1446-1447) and had a little bit of a different interpretation of the data. Clearly this was an important and well-done study, but my interpretation was that timely PCI wins because the lytic-based strategy wasn’t superior, but as it turns out there was more intracranial hemorrhage in that arm. The absolute rates were low, but still there was a fivefold difference — 0.2% vs. 1% going against the lytic arm. It’s true that in the course of the trial the investigators changed the trial so that the elderly patients, those aged 75 years or older, received a half-dose lytic instead of a full dose and that intracranial hemorrhage signal seemed to largely be attenuated; but nevertheless in the overall trial it was still a significant finding prior to that trial amendment.

Therefore, in the current time, primary PCI remains the treatment of choice and, in my opinion, this is probably going to be the last facilitated PCI trial to look at the issue of whether giving lytics or something else upfront is a better approach; from these results, it appears that is not the case — timely primary PCI is still the way to go.

PAGE BREAK

MASS COMM

Dean J. Kereiakes

Kereiakes: Investigators of the MASS COMM trial found that in select Massachusetts-based hospitals, MACE rates at 30 days and 1 year for patients treated with nonemergency PCI were nearly identical between hospitals without (no SOS) vs. with (SOS) cardiac surgical facilities on site (P for noninferiority for both <.001). However, many questions remain with respect to the conduct, analysis and interpretation of the trial that impact the generalizability of findings.

First, the validity of the comparison made — outcomes following nonemergent PCI performed at no-SOS vs. SOS hospitals — is dependent, at least in part, on the quality of PCI performed. The researchers acknowledge this as a “between hospital variation” treatment effect, which accounted for a difference in MACE across hospitals by 14 absolute percentage points at 30 days and 17 percentage points at 12 months. This difference between participating centers is striking and appears most marked among no-SOS centers. One might question the poolability of these widely discrepant outcomes data.

The minimum annual operator PCI volumes at both SOS and no-SOS centers is also shocking (annual minimums of 1, 5 and 11). How did the study investigators selectively choose those operators who participated in the study? If the annual operator PCI volume required to participate was 75 per year as stated in the manuscript, how were minimum volumes of one, five and 11 PCIs recorded? Even more concerning are the per patient and per lesion PCI success rates, which are presented as if they reflect the results of the participating SOS and no-SOS centers. Indeed, the per patient and per lesion PCI procedural success rates shown in Table 4 of the manuscript (Jacobs AK. N Engl J Med. 2013;368:1498-1508) appear remarkably low and fall well below the 10th percentile of a contemporary ACC/NCDR report. It would be interesting to know how these rates correspond to similar PCI procedural success rates reported to ACC/NCDR by these participating centers.

Given the degree of heterogeneity among hospitals within treatment groups with respect to the primary endpoints as observed in MASS COMM, it would require multiple years of continuous observation before these apparently discrepant outcomes could be proven to be statistically significantly different. Thus, the wide confidence intervals that surround low incidence occurrence adverse outcomes, such as death, in low procedural volume institutions, obscure the timely and accurate analysis of quality.

Finally, the MASS COMM researchers have not reported a cost-economic analysis as previously done by the C-PORT E investigators. In C-PORT E, for example, the cost of PCI at no SOS facilities was significantly greater than at SOS centers. This discrepancy was most marked for low volume (<200 PCI/year) no-SOS facilities, which applies to all but one of the MASS COMM no-SOS centers.

ADAPT-DES Subanalysis

Price: An ADAPT-DES subanalysis presented by Ajay Kirtane, MD, SM, demonstrated that with lower levels of platelet reactivity, there is more bleeding (6.7% vs. 5.6%; P=.04) and with increasing platelet reactivity by quintiles there is a greater risk of stent thrombosis (1.3% vs. 0.5; P=.0001) in clopidogrel-treated patients undergoing PCI. So, this analysis showed that by using HRs across quintiles it seems that there is a therapeutic window of platelet function within which you can maximize your ischemic benefit but minimize your bleeding risk. There have been other smaller studies demonstrating a therapeutic window, but this is the most robust demonstration in a very large group of patients — more than 8,000 patients in the United States — that point-of-care platelet function testing can be used to identify a therapeutic window for P2Y12 inhibition. This is important in the context of many clinical scenarios, including the decision of what drug to use and how long to use dual antiplatelet therapy. One of the downsides of prolonged DAPT is cost and bleeding. So if you can find the therapeutic window where you get the largest net clinical benefit or the window where your bleeding risk is reduced but you maintain your clinical efficacy, there may be less of a drawback of prolonged DAPT.

Disclosure: Bhatt reports receiving research grants from Amarin, AstraZeneca, Bristol-Myers Squibb, Eisai, Ethicon, Medtronic, Sanofi-Aventis and The Medicines Company and reports involvement in unfunded research for FlowCo, PLx Pharma and Takeda. Gurbel serves as a consultant for Accumetrics, AstraZeneca, Bayer, Boehringer Ingelheim, CSL, Daiichi Sankyo/Eli Lilly, Haemonetics, Iverson Genetics, Medtronic, Merck, Nanosphere, Novartis, Pozen and Sanofi-Aventis; receives grants or grants pending from AstraZeneca, CSL, Daiichi Sankyo/Eli Lilly, Duke Clinical Research Institute, Haemoscope, Harvard Clinical Research Institute, Medtronic, the National Institutes of Health, Pozen and Sanofi-Aventis; receives payment for lectures, including service on speakers’ bureaus, from Daiichi Sankyo/Eli Lilly, Iverson Genetics, Merck, Nanosphere and Sanofi-Aventis; and receives payment for development of educational presentations from the Discovery Channel and Schering-Plough. Kereiakes reports no relevant financial disclosures; he is the chairman for the Mission: Lifeline initiative for Ohio. Price reports receiving consultant fees/honoraria/being on the speakers’ bureau for Accumetrics, AstraZeneca, Boston Scientific, Daiichi-Sankyo/Eli Lilly, Janssen Pharmaceuticals, Medicure, The Medicines Company, Merck/Schering Plough, Terumo and W. L. Gore. Ramee has been involved as in investigator in all PARTNER trials.