Advisory Panel Votes to Recommend FDA Approval of Ticagrelor

Reed Miller

July 28, 2010

July 28, 2010 (Updated July 30, 2010) (Adelphi, Maryland) — The FDA Cardiovascular and Renal Drugs Advisory Committee voted to recommend approval of antiplatelet drug ticagrelor (Brilinta, AstraZeneca) for preventing thrombotic events in patients with acute coronary syndromes (ACS).

Even though the pivotal trial did not show a benefit for ticagrelor in US patients, the panel was sufficiently impressed with the results from the totality of clinical data on the drug and voted seven to one in favor of approving the drug for patients with either ST-elevation-MI or unstable angina/non-ST-elevation MI who will be treated with PCI and then voted seven to one again in favor of STEMI/NSTEMI patients who will be treated with medical management.

PLATO Provided Phase 3 Data

Ticagrelor was tested in the phase 3 Platelet Inhibition and Patient Outcomes (PLATO) trial, which enrolled 18 624 ACS patients in 43 countries and randomized them to ticagrelor (180-mg loading dose, 90 mg twice daily thereafter) or clopidogrel (Plavix, Bristol Myers Squibb/Sanofi-Aventis) (300- to 600-mg loading dose, 75 mg thereafter). Patients also received aspirin, at a dose of 75 mg to 100 mg per day, unless they could not tolerate it. After one year, 9.8% of the patients on ticagrelor and 11.7% of the clopidogrel patients reached the primary composite end point of cardiovascular-related death, MI, or stroke (hazard ratio 0.84; p<0.001). The components of the primary end point were predefined secondary end points, and MI and cardiovascular death were reduced in the ticagrelor group but stroke was the same. Bleeding events were about the same in both groups.

If approved, ticagrelor would become the first reversibly binding oral ADP-receptor antagonist on the US market. Ticagrelor is a direct-acting P2Y12-receptor antagonist in the class of cyclopentyl-triazolo-pyrimidines (CPTPs). The potential advantage of ticagrelor is that, unlike clopidogrel and prasugrel (Effient, Lilly/Daiichi), it is not a thienopyridine and can be "turned on and off" much faster than those agents; it therefore can prevent clots day-to-day without putting patients at higher risk for bleeding should they need surgery.

In the overall PLATO population, both medically managed and planned-invasive subgroups had better results with ticagrelor than clopidogrel (HR=0.85 and 0.84, respectively), with similar results in no-PCI and PCI subgroups. The comparative benefit was also preserved across the index event types of STEMI and NSTEMI, but with some attenuation of benefit in unstable-angina patients (HR=0.96; 95% CI 0.75–1.22).

What's The Matter With America?

As expected, a major focus of the meeting was getting a grip on why PLATO showed a statistically insignificant trend toward worse outcomes with ticagrelor vs clopidogrel among the North American patients in the study, who made up 1800 of the total 18 000 patients. AstraZeneca's statistics consultant Dr Gary Koch (University of North Carolina, Chapel Hill) explained that the difference in aspirin maintenance doses was the only variable that correlated to different outcomes between the North American and non-North American patients.

Patients in both the US and Canada had primary outcomes unfavorable toward ticagrelor (US: HR=1.27, 95% CI 0.92–1.75; Canada: HR=1.17, 95% CI 0.59–2.31). There were 10 other countries with a total of 3222 patients (17% of all patients) with hazard ratios above 1.0, but individually no one country showed a statistically significant outcome. The US contributed the second-highest number of subjects of the 43 countries representing PLATO, behind Poland with 2666 patients.

However, nobody can say for certain why North America was an outlier in the study. At the meeting, FDA reviewer Dr Robert Fiorentino concluded, "As of now, there is no clear biological or pathophysiological explanation for a ticagrelor-effect modifier that has been identified. There has been speculation, but I think we'd like to see confirmatory data to support those claims."

The US patient group tended to be heavier, with more diabetes and more prior MIs, PCIs, and bypass surgery than non-US patients. The US population had more NSTEMIs but fewer unstable-angina cases than the non-US patients. The proportion of patients who started the antiplatelet regiment 12 or more hours after the index event was higher in the US than outside the US, and the US group included more patients referred to invasive management despite having a higher proportion of NSTEMI vs STEMI patients. Drug-eluting stents were used more commonly in the US, and US compliance rates with the drug regimen were lower.

It does seem unlikely that the unfavorable US outcome can be explained entirely by a single factor alone, including aspirin.

"But of all of these factors pointed out here, the most striking is the difference in aspirin dose," Fiorentino said. The median dose in the US was 325 mg vs 100 mg outside the US, and the mean dose in the US was 217 mg vs 100 mg outside the US. "But it was not clear if effect modifiers related to higher-dose aspirin could explain these differences independently, [and] no specific factor was correlated with higher aspirin dose in the US; overall, there was no other single factor identified that appeared to be acting as some kind of surrogate for higher-dose aspirin that was the true causal factor for the regional interaction."

The panelists generally agreed that differences in aspirin doses are probably related to why the US patients did worse with ticagrelor, but it is not certain if that's because of some drug interaction or because the differences in aspirin maintenance dose are a surrogate for some other unknown differences between the patients. The consensus of the panel was expressed by chair Dr Sanjay Kaul (University of California, Los Angeles). "There is great uncertainty on this issue. Certain questions are irresolvable," he said, but the panel felt this issue should not derail the new drug application for ticagrelor, given the strong overall efficacy results in PLATO.

There is great uncertainty on this issue. Certain questions are irresolvable.

The discrepancy in outcomes between the non-North American and North American patients--especially the US patients--was cited as the reason for the single "no" vote from Dr Mori J Krantz (University of Colorado, Denver). "My only trepidation was the issue of the US population. It was fairly large . . . and I think that it went in the opposite direction was a little bit discomforting," Krantz said, adding that he believes the company should conduct another study to resolve this issue before the FDA approves ticagrelor.

Panelist and biostatistician Dr James Neaton (University of Minnesota, Minneapolis) voted in favor of recommending approval and said the best way to handle this mystery is for ticagrelor's labeling to include some statement about the possible impact of low-dose aspirin on the effectiveness of ticagrelor. "It's not the ideal solution, but I don't see anything else in the data that offers a clue. At least, the public should be aware of the post hoc findings that the sponsor did . . . and make a decision for themselves."

Getting to the Bottom of the Aspirin Question

Several of the panelists recommended that that FDA require AstraZeneca, as a condition of ticagrelor's approval, to agree to conduct a postmarket study that will clarify the how aspirin interacts with ticagrelor or why US patients had worse outcomes than non-US patients in PLATO.

In its briefing materials for the meeting, FDA reviewers recommended that the agency require a new clinical trial in patients with non-ST- or ST-segment-elevation ACS with at least 50% of the population from the US region. The proposed trial need not be a repetition of the PLATO study, the agency says.

The company is proposing to conduct a large multinational, randomized double-blind, "double-dummy" study, with over half the patients in the US, comparing ticagrelor with placebo in patients with stable coronary disease who are at least 12 months beyond the last ACS event. Patients in the study would be on a "background" aspirin dose of 70 to 150 mg a day. End points of this trial would be cardiovascular death, MI, and stroke.

But Kaul suggested that this study alone "will not help us address the issues [that are] driving this disparity, [which] have more to do with a practice of care differences rather than doses of aspirin." He would like the company to conduct another study that is similar to PLATO but smaller and more focused on the US. He also suggested that the company further study ticagrelor in different patient subtypes, especially patients with unstable angina and those with positive biomarkers, "to get a better idea of where the risk/benefit profile is going to be most optimal in the different cohorts."

Dr Henry Black (New York University) argued, "There are very important unanswered questions about aspirin, and unless there is a mandate to investigate this issue going forward, it will not happen when the drug is free-living in the marketplace." He lamented that there has yet to be study comparing aspirin plus clopidogrel vs clopidogrel alone in patients undergoing intervention or patients with acute coronary syndromes in order to understand how aspirin interacts with antiplatelet drugs. "It's an important question to be answered, and I'd like to see not only a look at aspirin dose, but even the possibility that patients might benefit from the elimination of aspirin from this regimen when other potent platelet inhibitors are used in patients with acute coronary disease."

What Can Ticagrelor Do?

Although they acknowledged disappointment that about 15% of patients in the study did not complete the prespecified cardiovascular follow-up, the panelists were generally satisfied with the adequacy of the PLATO design. The FDA asked the panel to consider whether the agency ought to break out different patient subsets from the overall ACS indication--PCI vs medical management and STEMI vs NSTEMI. The panel declined the option to consider different subsets separately, noting the inherent hazards of subgroup analysis.

Neaton said, "The indication is the population that was defined at randomization, which just happens to include all of these groups, and so breaking them apart doesn't make much sense."

Black added, "When you first see these people, you don't know where they're going to end up, and since there's no group that did particularly worse [in PLATO], I don't think it's necessary to worry too much about [further specifying the indications within ACS]."

Overall, PLATO showed a 22% relative reduction in all-cause mortality with ticagrelor vs clopidogrel, and about 85% of the deaths were cardiovascular. The study was powered to look at cardiovascular mortality, but not all-cause mortality, so there was some discussion during Wednesday's meeting about whether ticagrelor's labeling should include a claim that it can reduce all-cause mortality or just cardiovascular mortality. Dr Robert Temple, the FDA's Center for Drug Evaluation and Research's director of the Office of Medical Policy, said, "We have something of a bias against [claims of reducing] all-cause mortality unless the things other than CV mortality are also whipping along, and that's not too likely to be true here. This is going to be driven by the cardiovascular mortality, so it's really just a cardiovascular-mortality claim, but still impressive."

Biostatistician Neaton agreed that the claim should just be for cardiovascular mortality, although he said it wouldn't make much difference if the label claimed reduction in all-cause mortality. The other statistician on the panel, nonvoting advisor Dr Ralph B D'Agostino (Boston University, MA), noted, "Saying [the mortality difference] is driven by cardiovascular mortality is an appropriate way of handling the actual results."

Unlike clopidogrel, ticagrelor requires two doses a day, so the patient representative to the panel, Debra Madden (National Breast Cancer Coalition, Newtown, CT), pointed out that the labeling needs to stress the importance of patient compliance.

One of the PLATO investigators who presented on behalf of the sponsor at the meeting, Dr Robert Harrington (Duke Clinical Research Institute, Durham, NC), told heartwire that even though ticagrelor outperformed clopidogrel in PLATO, "I don't think it's the end of clopidogrel at all. In acute coronary syndromes, you'll have a couple of choices--ticagrelor, clopidogrel, and prasugrel. I think that ticagrelor will be a good choice in those situations [where it was tested in PLATO]. That will give clinicians choices to make in certain situations. [For example,] when clopidogrel becomes generic, there will be cost issues to consider."

Harrington pointed out that there are a number of indications in which ticagrelor hasn't been tested but clopidogrel has shown to be effective--for example, in elective angioplasty, STEMI with fibrinolysis, or peripheral arterial disease. "So there's no way that there will be a wholesale switch. In the acute coronary syndrome realm, you can make a pretty compelling case that the best data--the broadest category of patients treated, the broadest strategies of treatment--indicate ticagrelor. So maybe in that area it will become the dominant medication over a period of time."

Commenting on the advisory panel's decision, Dr Timothy Gardner (Christiana Care Health System, Newark, DE) told heartwire that "ticagrelor will provide a good option for patients with STEMI or unstable angina who may need urgent surgical intervention." Ticagrelor provides similar antiplatelet protection to ticagrelor but with much shorter duration of drug activity when stopped, which should lead to a reduction in excess bleeding in surgical patients who require antiplatelet treatment, he explained.

"The PLATO trial was a well-designed and well-conducted study, and the safety and efficacy of ticagrelor were confirmed," Gardner said. He acknowledges the issues with the North American subset of patients, but "this uncertainty should not hold up approval of the drug because of its obvious benefits for many patients."

Comments

3090D553-9492-4563-8681-AD288FA52ACE

processing....