Prasugrel has a proposed indication for the prevention of atherothrombotic events in patients with acute coronary syndrome (ACS) managed with percutaneous coronary intervention (PCI). This sub-group analysis suggests that prasugrel has a greater net treatment benefit in patients with diabetes mellitus compared to those without. However, results from such analyses should be viewed with caution, and prasugrel’s overall increased risk of bleeding, as demonstrated in the full TRITON-TIMI 38 paper, remains a concern.
Area Prescribing Committees should engage with local clinicians and agree a protocol for use should prasugrel be licensed and launched to the market. Its possible impact may depend not only on acquisition cost, but also on identification of those patients for whom prasugrel may be appropriate, bearing in mind the possibility of increased efficacy but at the expense of an increased bleeding risk.
What is the background to this?
Prasugrel is an antiplatelet agent with a similar mode of action to clopidogrel. The manufacturer (Daiichi-Sankyo/Lilly) submitted an application for licensing to the European Medicines Agency in February 2008 for the proposed indication of prevention of atherothrombotic events in patients with acute coronary syndrome (ACS) managed with percutaneous coronary intervention (PCI). Daiichi Sankyo/Lilly have recently announced that the US Food and Drug Administration (FDA) did not complete its review of prasugrel for the goal date of September 26, 2008. The companies suggest the submission is complex so delays are not surprising and they remain optimistic. NICE are reviewing prasugrel plus aspirin for the treatment of ACS with PCI as part of their 17th wave of technology appraisals, with an expected publication date of August 2009. More information on the possible place in therapy of prasugrel is available from the NPC’s recently published On the Horizon Future Medicines Bulletin (NHSnet connection required). More information on ACS can be found on the cardiovascular floors of NPC.
A phase III randomised controlled trial (RCT) of prasugrel plus aspirin vs. clopidogrel plus aspirin in patients with moderate to high risk ACS undergoing scheduled PCI was published in November 2007. TRITON-TIMI 38 demonstrated improved efficacy for prasugrel over clopidogrel in reducing atherothrombotic events, but at the expense of an increased major, and in some cases, fatal bleeding risk. Pre-specified sub-group analysis in TRITON-TIMI 38 indicated that the benefit of prasugrel over clopidogrel was greater in patients with diabetes mellitus than in those without. This has been explored more fully in this recently published paper.
What does this study claim?
This pre-specified sub-group analysis of TRITON-TIMI 38 claims that prasugrel was particularly efficacious in patients with diabetes mellitus (no distinction made between type 1 and type 2 diabetes). In patients without diabetes, the primary endpoint occurred in 9.2% of the prasugrel group vs. 10.6% of the clopidogrel group (hazard ratio [HR] 0.86; 95% confidence interval [CI] 0.76 to 0.98; P=0.02; number needed to treat [NNT]=71). In patients with diabetes, these figures were 12.2% vs. 17.0%, respectively (HR 0.70; 95% CI 0.58 to 0.85; P<0.001; NNT=21). As in the full RCT, the reduction in the primary endpoint in both the diabetes sub-group and the non-diabetes subgroup was mainly due to a reduction in non-fatal MI.
With regard to safety, in patients without diabetes, major bleeding occurred in 2.4% of the prasugrel group vs. 1.6% of the clopidogrel group (HR 1.43; 95% CI 1.07 to 1.91; P=0.02; number needed to harm [NNH]=125). In patients with diabetes, there was no significant difference in major bleeding between the prasugrel (2.5%) and clopidogrel (2.6%) group.
Post-hoc analyses of net clinical benefit (defined as the composite of all-cause mortality, non-fatal MI, non-fatal stroke or non-fatal major bleeding not related to CABG) suggested the net clinical benefit of prasugrel was greater for patients with diabetes than for those without diabetes. This was because patients with diabetes tended to have a greater reduction in ischaemic events without an observed increase in major bleeding.
This paper reports results from a sub-group analysis which, while pre-specified, still needs to be viewed with caution, as such analyses can be misleading. Randomisation was not stratified by diabetic status and the possibility of an imbalance between treatment groups exists. The authors suggest prasugrel, compared with clopidogrel, was particularly beneficial in patients with diabetes, with a more favourable net clinical benefit (weighing up atherothrombotic benefits vs. bleeding risks). However, TRITON-TIMI 38 studied a carefully defined patient group and, if prasugrel comes to the market, balancing its risks and benefits for individual patients will be vital, despite their diabetic status. Based on the full TRITON-TIMI 38 trial, for every 1,000 patients treated with prasugrel plus aspirin rather than clopidogrel plus aspirin for 14.5 months, 23 will have a primary CV endpoint prevented but 6 will have a major bleed. Prasugrel may be suitable for patients who are at high risk of ischaemic events but at low risk of bleeding. Clopidogrel may be a safer choice in patients at a lower risk of ischaemic events but higher risk of bleeding. Differentiating between these groups in clinical practice may be difficult.
Design: Pre-specified subgroup analysis of patients with diabetes mellitus from the randomised controlled trial, TRITON-TIMI 38.
Patients: TRITON-TIMI 38 included 13,608 patients with moderate-to-high risk ACS with scheduled PCI. 3,146 (23%) had diabetes mellitus.
Intervention and comparison: 99% of patients were given aspirin 75–162mg daily. All patients were randomised to either prasugrel (60mg loading dose then 10mg daily) or clopidogrel (300mg loading dose then 75mg daily) for 6 to 15 months (median 14.5 months).
Outcomes: The primary efficacy endpoint was a composite of cardiovascular death, non-fatal MI, or non-fatal stroke. The key safety endpoint was TIMI major bleeding not related to coronary-artery bypass grafting (CABG).
Results: Rates of thrombotic events were higher among subjects with diabetes than subjects without diabetes, with highest rates seen in diabetic patients on insulin. In patients without diabetes, the primary endpoint occurred in 9.2% of the prasugrel group vs. 10.6% of the clopidogrel group (HR 0.86; 95% CI 0.76 to 0.98; P=0.02; NNT=71). In patients with diabetes, the primary endpoint occurred in 12.2% of the prasugrel group vs. 17.0% of the clopidogrel group (HR 0.70; 95% CI 0.58 to 0.85; P<0.001; NNT=21). As in the full RCT, the reduction in the primary endpoint in both the diabetes sub-group and the non-diabetes subgroup was mainly due to a reduction in non-fatal MI. Bleeding rates were similar in patients with and without diabetes, regardless of diabetes treatment. In patients without diabetes, major bleeding occurred in 2.4% of the prasugrel group vs. 1.6% of the clopidogrel group (HR 1.43; 95% CI 1.07 to 1.91; P=0.02; NNH=125). In patients with diabetes, there was no significant difference in major bleeding between the prasugrel (2.5%) and clopidogrel (2.6%) group.
Sponsorship: The TRITON-TIMI 38 trial was supported by Daiichi Sankyo Co Ltd and Eli Lilly and Co.