Meta-Analysis: A Look at Three Different GP IIb/IIIa Inhibitors for Anticoagulation During PCI

Can you explain the nature of your network meta-analysis and how the analysis was carried out?

There have been multiple randomized, controlled trials (RCTs) comparing different glycoprotein IIb/IIIa inhibitors (GPIs) with placebo in patients undergoing percutaneous coronary intervention (PCI) with anticoagulation and oral P2Y12 inhibitors. There have also been several randomized controlled trials comparing PCI with different GPIs. However, data is lacking to provide comparison between the different GPI regimens in combination with heparin vs heparin alone vs bivalirudin alone. Thus, we conducted a network meta-analysis¹ to enable comparison between high-dose bolus (HDB) tirofiban with heparin, eptifibatide with heparin, abciximab with heparin, heparin alone, and bivalirudin alone for patients undergoing PCI. The analysis was conducted by abstracting the data from 41 RCTs, and analyzed with specialized statistical software to enable direct and indirect comparison between the different groups.

Your analysis included a comparison of the three different GP IIb/IIIa inhibitors (GPIs): tirofiban, eptifibatide, and abciximab. What data currently exists for such a comparison, and how does your study complement these?

There have been previous meta-analyses comparing abciximab with small-molecule GPIs (tirofiban and eptifibatide), HDB tirofiban with abciximab, and the individual GPIs with heparin. This meta-analysis expands upon prior smaller meta-analyses to enable comparison of each individual GPI with any available anticoagulation strategy for PCI. This data may enable not only physicians but also pharmacies to compare outcomes between different GPI regimens to help educate formulary decisions.

Individual randomized trials evaluating clinical outcomes have not demonstrated a clear mortality benefit for GPI + heparin compared to heparin alone — did you find it surprising that tirofiban provided a significant reduction in all-cause mortality compared with heparin in your analysis? 

This finding is driven by inclusion of data from 16 RCTs with 2654 patients randomized to HDB tirofiban with comparison to a variety of other therapies. It is important to acknowledge that many of these studies included a small number of patients, which may impact the power of these findings. However, Sethi and colleagues² conducted a meta-analysis and found tirofiban with heparin led to a significant reduction in major adverse cardiac events (MACE) compared with heparin. Thus, further corroboration of these findings with larger randomized, controlled trials would be ideal. The critical point is that HDB tirofiban (25 mcg/kg) is superior to the previously utilized low-dose bolus regimen (10 mcg/kg).³  

GPIs are thought to have similar clinical profiles — did you observe any differences in the efficacy and safety outcomes between the three different GPIs? 

We found that bleeding was highest with eptifibatide. This may be driven by the prolonged infusion required in many of the clinical trials. Importantly, we showed that eptifibatide performed well in regards to MACE, but had increased all-cause mortality compared with HDB tirofiban. Therefore, we speculate that the increased all-cause mortality may be driven by higher bleeding risk with prolonged GPI infusion. As direct comparison between HDB tirofiban and eptifibatide in our analysis is limited to a single trial⁴, these differences are driven by indirect comparison and should be viewed as hypothesis generating.

With the advent of the newer oral P2Y12 inhibitors (i.e., prasugrel and ticagrelor), do you think there is still a role for the use of GPIs in contemporary PCI?

Recent studies have stressed that platelet inhibition through adenosine diphosphate or P2Y12 inhibition may have delayed onset in patients with acute coronary syndrome.⁵ This is further highlighted in a recent study that demonstrated that morphine significantly delays platelet inhibition with ticagrelor in patients with acute myocardial infarction (MI).⁶ Thus, adequate platelet inhibition early in the setting of PCI for acute coronary syndromes (ACS) remains a major concern⁷, as it requires balancing antiplatelet effects to avoid acute stent thrombosis without increasing the risk of adverse bleeding events. Thus, patients that present with ST-elevation myocardial infarction (STEMI) and high-risk NSTE-ACS, where there is large thrombus burden, may benefit from GPIs with a bolus-only strategy or utilization of a short infusion in an effort to mitigate bleeding risk. Further studies are required to assess the safety and efficacy of such a strategy.

Do your findings have financial implications for healthcare providers?

Given the safety and efficacy demonstrated for HDB tirofiban in our network meta-analysis, I think it is reasonable to consider utilization of tirofiban. Tirofiban appears to be less expensive than alternative GPIs, which may improve the cost to the hospital.

In your findings, GPIs have generally demonstrated an ischemic benefit, but with an increased bleeding risk compared to heparin. Is there a strategy to maximize the ischemic benefit while minimizing the bleeding risk of GPIs?
As mentioned above, patients that present with STEMI and high-risk NSTE-ACS, where there is large thrombus burden, may benefit from GPIs with a bolus-only strategy or utilization of a short infusion in an effort to mitigate bleeding risk. Further studies are required to assess the safety and efficacy of such a strategy. However, I believe utilization of GPIs will always increase the risk of bleeding when compared with bivalirudin or heparin. The question is whether this bleeding risk can be minimized.

What is the most important takeaway from your analysis?

I think this analysis provides a concise comparison between different anticoagulation/antiplatelet strategies for PCI. Furthermore, the analysis confirms that all GPI strategies reduce MACE at the cost of increased bleeding in the setting of PCI when compared with a bivalirudin-alone or heparin-alone strategy.

References

1. Lipinski MJ, Lee RC, Gaglia MA Jr, Torguson R, Garcia-Garcia HM, Pichard AD, Satler LF, Waksman R. Comparison of heparin, bivalirudin, and different glycoprotein IIb/IIIa inhibitor regimens for anticoagulation during percutaneous coronary intervention: A network meta-analysis. Cardiovasc Revasc Med. 2016 Dec; 17(8): 535-545. doi: 10.1016/j.carrev.2016.09.011. 1.    
2. Sethi A, Bahekar A, Doshi H, Bhuriya R, Bedi U, Singh S, Khosla S. Tirofiban use with clopidogrel and aspirin decreases adverse cardiovascular events after percutaneous coronary intervention for ST-elevation myocardial infarction: a meta-analysis of randomized trials. Can J Cardiol. 2011; 27: 548-554.
3. Valgimigli M, Biondi-Zoccai G, Tebaldi M, van’t Hof AW, Campo G, Hamm C, et al. Tirofiban as adjunctive therapy for acute coronary syndromes and percutaneous coronary intervention: a meta-analysis of randomized trials. Eur Heart J. 2010; 31: 35-49.
4. Mardikar HM, Hiremath MS, Moliterno DJ, Mathew R, Arora R, Deo D, et al. Optimal platelet inhibition in patients undergoing PCI: data from the Multicenter Registry of High-Risk Percutaneous Coronary Intervention and Adequate Platelet Inhibition (MR PCI) study. Am Heart J. 2007; 154: 344.e1-e5.
5. Parodi G, Valenti R, Bellandi B, Migliorini A, Marcucci R, Comito V, et al. Comparison of prasugrel and ticagrelor loading doses in ST-segment elevation myocardial infarction patients: RAPID (Rapid Activity of Platelet Inhibitor Drugs) primary PCI study. J Am Coll Cardiol. 2013; 61: 1601-1606.
6. Kubica J, Adamski P, Ostrowska M, Sikora J, Kubica JM, Sroka WD, et al. Morphine delays and attenuates ticagrelor exposure and action in patients with myocardial infarction: the randomized, double-blind, placebo-controlled IMPRESSION trial. Eur Heart J. 2016; 37: 245-252.
7. Alexopoulos D, Bhatt DL, Hamm CW, Steg PG, Stone GW. Early P2Y12 inhibition in ST-segment elevation myocardial infarction: Bridging the gap. Am Heart J. 2015; 170: 3-12.

Disclosure: Dr. Michael Lipinski reports he has received research grant support from Medicure Inc.

He can be contacted at michael.j.lipinski@medstar.net.