New PLATO Analysis: Aspirin Dose May Explain Discordant North American Outcomes

Although the regional variations in the effect of ticagrelor vs. clopidogrel found in the PLATO trial may have arisen from chance alone, the poor outcomes for North American patients seen with the new, more potent antiplatelet drug could be due to the fact that they tended to receive higher aspirin doses compared with the rest of the world. The findings, published online June 27, 2011, ahead of print in Circulation, confirm early suspicions that aspirin dose could influence ticagrelor’s efficacy.

Published in the New England Journal of Medicine in September 2009, the PLATO (PLATelet inhibition and patient Outcomes) trial randomized more than 18,000 ACS patients in 43 countries worldwide to the PSY12 antagonist ticagrelor (AstraZeneca, Wilmington, DE; 180-mg loading dose, 90 mg twice daily thereafter) or clopidogrel (300- to 600-mg loading dose, 75 mg daily thereafter). All patients also received aspirin.

At 12 months, the primary endpoint (CV death, MI, or stroke) was observed in 9.8% of ticagrelor patients compared with 11.7% of clopidogrel patients (HR 0.84; 95% CI 0.77-0.92; P < 0.001). But a prespecified subanalysis also revealed an interaction between treatment and region (P = 0.045), with ticagrelor having less effect among North American patients. Concerns about this disparity have been credited with thus far stalling US Food and Drug Administration approval of ticagrelor.

What’s Happening in North America?

For the current paper, researchers led by Kenneth W. Mahaffey, MD, of the Duke Clinical Research Institute (Durham, NC), explored the geographic interaction in PLATO.

Two independent statistical groups participating in the analysis found no differences between how the trial was conducted in the United States compared with the rest of the world. Although US patients (n = 1,413) had worse adherence to their randomized treatment drug (defined as > 80% compliance at each visit) than did patients elsewhere (n = 17,211) at 62.0% vs. 84.7%. Discontinuation also was more common in the United States than in the rest of the world (31.4% vs. 22.0%). Both adherence and discontinuation rates, however, were consistent between the ticagrelor and clopidogrel groups within the United States.

Importantly, a higher proportion of patients in the United States took concomitant high-dose aspirin (median ≥ 300 mg/day) than those located elsewhere (53.6% vs. 1.7%).

The new analysis confirmed that US patients exhibited a trend toward higher rates of the primary outcome with ticagrelor (12.6% vs. 10.1% with clopidogrel; HR 1.27, 95% CI 0.92-1.75; P = 0.1459). While in the rest of the world clopidogrel lowered the risk of CV death, MI, and all-cause mortality, US patients obtained similar outcomes for each endpoint regardless of which drug they received.

Although Dr. Mahaffey and colleagues were unable to rule out that chance alone was responsible for the difference in ticagrelor’s effect, aspirin was also identified as a possible culprit. Of 37 baseline and post-randomization factors evaluated, only aspirin dose substantially influenced the regional interaction. Results for the primary endpoint varied by aspirin dose in both the United States and the rest of the world (table 1).

Table 1. Effect of Aspirin Dose on Primary Endpoint

 

Landmark analyses at time points ranging from randomization to 180 days indicated that, in patients recruited outside of the United States who were receiving low-dose aspirin, ticagrelor was consistently associated with better results than clopidogrel. But in US-based patients receiving low-dose aspirin, outcomes were equivalent between the 2 drugs. In patients receiving high-dose aspirin, clopidogrel produced superior outcomes among both US and non-US patients. Due to small sample sizes, particularly in the United States, the relationships often failed to reach statistical significance.

“Number one, the overall trial was not powered to look at these subgroups,” cautioned Dr. Mahaffey in an American Heart Association-sponsored media briefing. “Number two, . . . subgroup analyses that are defined by post-randomization events, such as the dose of aspirin, are very complicated and potentially hazardous even when we use very rigorous statistical methodologies.”

Low-Dose Aspirin Preferred

Therefore, “the treatment-by-region interaction is still apparent [and] the aspirin hypothesis is intriguing,” said Dr. Mahaffey, “but even the simulations and analyses we did show that the play of chance, just random happenstance in a large trial of 18,000 patients in 43 countries, [could be responsible]. I think the important message, though, is that—with current guidelines recommending low-dose aspirin in patients with acute coronary syndromes for long-term therapy—in a setting of potent P2Y12 inhibition, the most favorable outcomes are most likely to be realized [with the lower dose].”

Elliott Antman, MD, of Harvard Medical School (Boston, MA), who comoderated the discussion, agreed that “we are on our most secure footing” when looking at the main trial results rather than subgroup analyses.

“Speaking now as a clinician reading the medical literature and thinking about aspirin, I don’t really see anything that points to a benefit from high-dose aspirin over low-dose aspirin. In fact, I am concerned about the risk of gastrointestinal bleeding being higher when we use high-dose aspirin in the maintenance phase,” he commented. “So personally, I’ve begun to switch very soon after PCI to a lower maintenance dose of aspirin across the board [no matter which] antiplatelet therapy I’m giving, be that clopidogrel or prasugrel, or in the future ticagrelor.”

 

 

---

 

Source:Mahaffey KW, Wojdyla DM, Carroll K, et al. Ticagrelor compared with clopidogrel by geographic region in the PLATelet inhibition and patient Outcomes (PLATO) trial. Circulation. 2011;Epub ahead of print.

 

 

 

 

Related Stories:

• FDA Ticagrelor Surprise: Further PLATO Analyses Required Before Approval
• Optimal Aspirin Dosage with Ticagrelor May Be Lower Than Expected
• Questions Regarding PLATO Cover Old Ground