November 16, 2009

November 16, 2009 (Orlando, Florida) — The fallout from the CHAMPION trials of the new IV antiplatelet agent cangrelor (the Medicines Company) in PCI has been one of the main talking points at the American Heart Association (AHA) 2009 Scientific Sessions so far. The first IV P2Y12 antagonist did inhibit platelet activity more effectively than clopidogrel, so why didn't this translate into a reduction of the primary end point in the two trials?

Several theories have been suggested, the main one focusing on the issue of periprocedural MIs as an end point and whether these results were confounded by an unusually short time from admission to PCI in the CHAMPION trials.

Most experts polled by heartwire said there was a need for an IV agent for acute use in the cath lab, and they would like to see more trials conducted with the drug.

The Trials' Results

The CHAMPION-PCI and CHAMPION-PLATFORM trials both compared a two-hour infusion of cangrelor with a clopidogrel 600-mg loading dose in predominantly ACS patients for whom a PCI strategy had already been decided. The main difference between the two trials was that in the PCI trial, the clopidogrel was given at the same time as the cangrelor infusion was started (before the PCI procedure), whereas in the PLATFORM trial, cangrelor was started before the PCI procedure and clopidogrel was given immediately after the PCI procedure on the way out of the cath lab.

As previously reported by heartwire , neither trial showed any significant difference between the two treatments in terms of the primary end point--death/MI/ischemic-driven revascularization at 48 hours. But some subgroup and exploratory analyses, particularly in the PLATFORM study, have suggested some benefits of cangrelor. The two studies were also published online on November 15, 2009 in the New England Journal of Medicine [1,2]

The Issue of Periprocedural MIs

On the issue of periprocedural MIs, one of the designated discussants of the trials, Dr David Faxon (Brigham and Women's Hospital, Boston, MA), who was on the CHAMPION data and safety monitoring board, commented: "There was a remarkably short time from admission to PCI in these studies (an average of six to eight hours). In most prior trials, this time has been much longer. This very short time to PCI in CHAMPION will have made it difficult to discern a periprocedural MI from an MI that was already under way before randomization." He added: "There is no uniform definition of periprocedural MI, and it can be particularly difficult to discern when patients present with positive biomarkers that continue to rise during the procedure. There was a higher rate of periprocedural MI in the CHAMPION studies (7%) compared with what we would expect (about 5%). In addition, the primary end point was significantly reduced in troponin-negative patients, in whom the definition of periprocedural MI is much easier to discern. I think this is a legitimate argument."

Because of all these caveats, and the benefits seen on death and stent thrombosis in the PLATFORM trial, I think it is wrong to interpret these trials as completely negative.

Lead investigator of CHAMPION-PCI, Dr Robert Harrington (Duke Clinical Research Institute, Durham), elaborated to heartwire : "We used a relatively liberal definition for periprocedural MI of three times upper limit of normal for CKMB rises. That meant we picked up a lot of periprocedural MIs, but when the bar is set relatively low like that it is easy to pick up events that were already under way, particularly when we had such a short time to PCI. In hindsight, it would probably have been better to use a stricter definition like five times the upper limit of normal or even eight times." He added that further analyses of the data were now under way based on this stricter definition.

CHAMPION-PLATFORM lead investigator, Dr Deepak Bhatt (Brigham and Women's Hospital), reiterated that the secondary end points of death and stent thrombosis, which were not dependent on enzyme rises, were significantly reduced in the CHAMPION PLATFORM trial. And exploratory analysis from both studies has shown impressive reductions in the composite end point of death/Q-wave MI/ischemia-driven revascularization. He also noted that when the two trials were analyzed together as a meta-analysis, which has not yet been presented, this exploratory analysis became stronger.

In the TRITON trial, periprocedural MI was used to carry the analysis, but in this trial it is being used to display information for obscuring benefit.

But cochair of the late-breaking clinical-trials session at which the trials were presented, Dr Eric Bates (University of Michigan Medical Center, Ann Arbor), asked whether the end points of death and stent thrombosis seen in the CHAMPION PLATFORM study were still significantly reduced in the combined analysis of the two studies, to which Bhatt replied that they were not.

Bates said he was confused about the role of periprocedural MI in composite end points. "In the TRITON trial, periprocedural MI was used to carry the analysis, but in this trial it is being used to display information for obscuring benefit," he pointed out.

But Bhatt argued back that TRITON had a much stricter definition of periprocedural MI, needing two postprocedural elevated values rather than just one in CHAMPION. "If we had used that definition in CHAMPION, we would have seen a stronger benefit for cangrelor," he added.

Other possible factors that may have contributed to the negative result, mentioned by Faxon in his talk, were a lower event rate than anticipated in the placebo group--8% vs the expected 11%--and that the duration of the cangrelor infusion (two hours) may not have been long enough. Faxon added to heartwire : "Because of all these caveats, and the benefits seen on death and stent thrombosis in the PLATFORM trial, I think it is wrong to interpret these trials as completely negative."

Rationalization After the Fact?

But clinical-trials guru Dr Sanjay Kaul (Cedars Sinai Medical Center, Los Angeles) disagreed. Commenting on the CHAMPION trials for heartwire , he said: "The CHAMPION investigators have overestimated the relevance of some of their observations. This is rationalization after the fact. They are being overenthusiastic about the death and stent-thrombosis results in CHAMPION-PLATFORM study. The most likely explanation for these findings is the play of chance. The counterintuitive benefit in troponin-negative patients could also just be statistical fluke."

The most likely explanation for these findings is the play of chance.

Kaul says that any claims of noninferiority of cangrelor to clopidogrel must also be questioned. "The noninferiority assessment was not prespecified, and the study was not powered to address it," he pointed out.

Designated discussant of the CHAMPION PCI trial, Dr Alan Michelson (Center for Platelet Research Studies, Boston, MA) added some more possible reasons for the negative results in CHAMPION trials, including the fact it had a tougher control arm--clopidogrel 600 mg--than previous trials of new agents, which have tended to use clopidogrel 300 mg as the control. "We now know that 600-mg is better than 300-mg clopidogrel, so perhaps there is no benefit over this higher clopidogrel dose," he said.

Interaction With Clopidogrel?

Another reason could be problems transitioning from the IV cangrelor to oral clopidogrel in the trials, Michelson suggested. Noting that a previous mechanistic study by Steinhubl et al [3] has shown a partial recovery of platelet function between stopping cangrelor and starting clopidogrel, he commented: "It may be that cangrelor is blocking the P2Y12 receptor, not allowing clopidogrel to act when it is first given."

But Bhatt said a detailed analysis from the platelet-function substudy in CHAMPION has not shown any interaction between cangrelor and clopidogrel. However, Harrington commented to heartwire that it may be better to transition to a different oral molecule after using cangrelor. "Ticagrelor [AstraZeneca] is the obvious choice, as it is more similar to cangrelor than clopidogrel is." He also pointed out that there was another compound in phase 2 development--elinogrel (Portola/Novartis), which has both IV and oral formulations, which would alleviate any concerns about possible interactions between different agents when transitioning from the IV to oral products.

Do We Need an IV Agent?

At an AHA press conference on the CHAMPION trials, it was asked whether there was a need for an IV agent when ticagrelor, which has a quicker on-off effect than clopidogrel or prasugrel (Effient, Lilly/Daiichi), has shown such good results and should be available soon.

Most experts said they thought an IV agent would still have a role. Faxon commented: "There is a great interest in cangrelor, as it is the first IV P2Y12 antagonist, which we would welcome for acute use in the cath lab. There are a large number of patients who present to the cath lab for angiography and are then given PCI straight away. There is no time for pretreatment, so an IV agent that we can just switch on and switch off would be very attractive. It would also be particularly useful for the very ill patients who might suddenly need to go to urgent CABG."

What Now for Cangrelor?

The CHAMPION investigators are keen for more trials of cangrelor to be conducted. Harrington commented: "I think our results are provocative. I think there is something there that needs to be investigated further. If we did another trial, we would have discussions about many things--the definition of periprocedural MI or whether to use biomarker-diagnosed MI at all as an end point; whether to use a longer infusion, and which oral drug to transition to. It would be great to do a trial with just Q-wave MI, but then we would need a much larger trial size to get enough events."

The Medicines Company issued a statement saying, "Based on the CHAMPION data and the analysis and encouragement by experts at the AHA, we are confident in the pharmacology, potential beneficial clinical effects, and suitable safety of cangrelor. We will continue every effort to make the compound available to patients and cardiologists worldwide. More studies are almost always helpful in defining the clinical value of a potentially leading product. Cangrelor has the potential to save lives, and we will seek approval for it, guided by the needs of regulators."