Debating ISAR-REACT 5: Ticagrelor vs Prasugrel in ACS

Robert F. Storey, BSc, BM, DM; Dirk Sibbing, MD, MHBA


September 12, 2019

This transcript has been edited for clarity.

Robert F. Storey, BSc, BM, DM: Hello. I'm Rob Storey, and I'm joined by Dirk Sibbing, who is one of the principal investigators in the ISAR-REACT 5 study.[1] Dirk, perhaps you could tell us a bit about the study design and its findings.


Dirk Sibbing, MD, MHBA: Thank you, Rob, for having me here. One open question about the potent antiplatelet agents prasugrel and ticagrelor: Which one is better? Debates have been going on for years. The ISAR-REACT 5 trial was an investigator-initiated multicenter trial that tried to address this question in 4018 patients randomized to receive ticagrelor (n = 2012) or prasugrel (n = 2006). It was an all-comers trial for patients with acute coronary syndrome (ACS).

All of the investigators did the trial with the assumption that ticagrelor might be the winner.

The trial enrolled both ST-segment elevation myocardial infarction (STEMI) and non-STEMI patients and, in comparison to other trials like PLATO,[2] had a planned invasive management for patients. For this reason, nearly all patients got a cath lab investigation and about 85% received a percutaneous coronary intervention (PCI). The trial was designed to show superiority for ticagrelor with respect to the endpoint, which was a combined incidence of death and minor stroke at 1 year after randomization. And, of course, when you are looking at potent agents, you have to look at bleeding. The key secondary endpoint was major bleeding events, defined according to the BARC (Bleeding Academic Research Consortium) criteria.

All of the investigators did the trial with the assumption that ticagrelor might be the winner. This was based on PLATO and on the strategy behind ticagrelor that implies pretreatment for non-STEMI and STEMI patients. The trial was not only a comparison of prasugrel and ticagrelor but also a strategy trial of pretreatment, because it included pretreatment for all patients and specifically for the non-STEMI patients. Of course, if you have the chance to pretreat, you have the chance to reduce periprocedural MI and other things. Taking all of this together, we hoped for superiority but what we found was the complete opposite.

As a matter of fact, we found superiority for prasugrel. We saw about a 9% event rate in the ticagrelor arm and 6.9% in the prasugrel arm. This was statistically significant. There was an absolute risk reduction when comparing the two groups of 2.4% corresponding to a number-needed-to-treat of 42. And we did not see a difference for bleeds. This difference in the primary endpoint was mainly driven by MI and stent thrombosis. I think we have to take the results as they are.

Results Are Discordant From Prior Studies -- Why?

Storey: I'm going to challenge you on a few points. This was an open-label study. Patients were coming to the cath lab with a clinician knowing which strategy had been used. And some of the patients randomized to prasugrel didn't actually receive it, I guess because they came to the cath lab and a decision was made not to use it.

This study showed results discordant from the much larger double-blinded studies. PLATO showed a superiority of ticagrelor over clopidogrel that was progressive over 12 months. TRITON-TIMI 38[3] also showed superiority of prasugrel over clopidogrel, although there are issues about loading regimens that might have disadvantaged clopidogrel. And also with prasugrel there was no advantage compared with clopidogrel in the TRILOGY ACS[4] study. It's really difficult to bring this all together and take the ISAR-REACT 5 results at face value on the basis of those large phase 3 double-blinded studies.

Sibbing: Your points are well taken and I will try to shed a little bit of light on this. When you have different agents (ISAR-REACT 5 is an example in this respect) and you try to extrapolate and compare data from trials like PLATO or TRITON-TIMI 38, there can be issues and it's always good to do a head-to-head comparison. A head-to-head comparison was done here. This would also be great for the novel oral anticoagulants if we could do it, but there are no data out on that respect.

Furthermore, the PLATO trial as it was designed was not a pure PCI trial. PLATO had a significant number of patients undergoing conservative treatment and a significant number of patients undergoing coronary artery bypass graft surgery. This is in contrast to TRITON-TIMI 38, which really was a PCI trial, as was the case for ISAR-REACT 5, which was designed to have PCI as the first-line treatment strategy. In looking at PLATO results, I agree that in the overall trial there was a mortality benefit; the results were really convincing and they still are. But if you dive a little bit deeper and look at the PCI-treated patients, you can see that most of the benefit, or at least a significant amount, was seen in the conservatively managed patients. We have to reflect on this when we compare all of these trials, apart from the fact that there are 15 years in between now. We have another treatment now. We have other options. ISAR-REACT 5 is very timely in this respect. And our guidelines really endorse PCI for revascularization.

Access Differences, Stent Thrombosis

Storey: There was a high use of femoral access that might have disadvantaged the pretreatment approach compared with the more contemporary radial access studies in terms of bleeding rates and implications for cessation of therapy.

Sibbing: The trial started in 2013 and enrolled patients up to 2018. Practice has changed over time. We all know how the guidelines have changed, and if we started the trial today, the rates of radial and femoral would be different. On the other hand, I think the key point to be taken from ISAR-REACT 5 is that efficacy was not affected by the access site. Bleeding was affected, I completely agree, and we could see different results if it was a radial-only trial. But efficacy results have to be reflected upon and they do not necessarily depend on access site.

Storey: Quite a bit of the benefit of prasugrel versus ticagrelor in the study was related to stent thrombosis rates, which were higher in the ticagrelor group. Again, that does not quite resonate with me because we've published our center evidence looking at ticagrelor pretreatment and found a huge reduction in the stent thrombosis rates with ticagrelor moving on from clopidogrel.[5] Stent thrombosis is a very rare event apart from mainly morphine-treated STEMI patients, but we have recognized this delayed absorption issue which we can now adapt our treatment to. Were you surprised by the difference in stent thrombosis given what we know? Of course, you have done lots of work on platelet function and platelet reactivity and the risk for stent thrombosis. Would you have expected similar rates?

Sibbing: I would have expected similar rates both for stent thrombosis and for MI. If you look at the MIs, it's also STEMIs that are significantly higher in the ticagrelor group. With the results as they are, we have to ask ourselves how this could happen.

On one side we have an irreversible agent like prasugrel, and on the other side a reversible agent like ticagrelor; it may well be that this reversible mode of action can be an issue for some patients who don't take the medication for 1,2, or 3 days, for whatever reason. We all know in clinical practice that this is going to happen. People may forget the tablet in the morning and maybe even forget it the next morning. Maybe we need to investigate further on this. This kind of temporary noncompliance can cause events in some patients.


Storey: Patients had to pay for their own drugs because drugs were not provided as part of the study, is that correct?

Sibbing: That is correct, yes.

Storey: Could there have been any cost issues that led patients to try and save in terms of dosing?

Sibbing: Yes. I'm lucky to say that in Germany, cost is not so much of an issue if the drug is prescribed. Patients [are able to obtain] ticagrelor and prasugrel in the pharmacy. To add to this important comment, on the other side, the discontinuation rate in ISAR-REACT 5 was comparatively low. If you look at THEMIS,[6] for example, you have about a 20% discontinuation rate. If you look at GLOBAL LEADERS,[7] for ticagrelor there was a discontinuation rate of about 20%. In ISAR-REACT 5 [the discontinuation rate] is slightly above 10%. The investigators did a great job of keeping patients on their tablets.

Storey: There is experience in terms of using ticagrelor. Certainly in my center, we've gained a lot of experience from using it first-line over the years, so we know how to handle the transient and often mild dyspnea related to ticagrelor. There is a learning curve in terms of using it and keeping patients on it or learning how to switch if you need to.

Sibbing: I fully agree. Dyspnea is an issue on ticagrelor, but a discontinuation rate of around 10% shows that it's just fine. Concerning the performance of the intake in the trial and comparing it with other trials, I honestly do not see a real issue here. I think we did our job very well in that respect.

Can We Trust These Results?

Storey: I'm going to be a little provocative and make a counterargument. Because of some of the design issues and the open-label nature of the study, this is hypothesis-generating in terms of contemporary practice and would need perhaps a larger and preferably blinded study to really compare the agents' efficacy.

Sibbing: But in this study of more than 4000 patients, we lost very few patients to follow-up, compliance to treatment was excellent, and combined endpoint of death and minor stroke resulted in a relative increase in risk above 30%, so I think we can't take that for granted. We have to reflect it in our practice. I am not saying that all ACS patients should now be on prasugrel and none should be prescribed ticagrelor. But specifically for the PCI cohort, if you have good reason to [prescribe] prasugrel, I think the ISAR-REACT 5 trial results support your strategy.

Storey: Great. Thanks very much, Dirk, for that great conversation. I'm sure this debate is going to go on and on and that the audience will take their own position. Thanks very much.

Sibbing: Thank you.

Follow Medscape on Facebook, Twitter, Instagram, and YouTube


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.