STEMI: Best P2Y12 inhibitor according to network meta-analysis

Original Title: Optimal P2Y12 inhibitor in patients with ST-segment elevation myocardial infarction undergoing primary percutaneous coronary intervention: a network meta-analysis.

Reference: Rafique AM et al. J Am Coll Cardiol Intv 2016;9:1036–46.

Courtesy of Dr. Alejandro Lakowsky.

Researchers drew up a network meta-analysis incorporating 37 trials with more than 88,000 patients undergoing ST elevation myocardial infarction (STEMI) to compare the safety and efficacy of clopidogrel, prasugrel, ticagrelor and cangrelor.

Some results were expected, such as:

• Lower antithrombotic action of clopidogrel clinically compared to prasugrel and ticagrelor.
• Little data on cangrelor to draw strong conclusions.

But others were provoking and surprising.

• To start with, major bleeding rates were similar between clopidogrel, prasugrel and ticagrelor. The available large randomized studies comparing prasugrel and ticagrelor vs. clopidogrel (Triton-TIMI 38 y Plato), had found both drugs produce more major bleeding events unrelated to surgery than clopidogrel. Researchers justify these findings with the fact that the STEMI population is slightly younger (3 years in average) and has less comorbidities than ACS or stent thrombosis populations.

• Even more striking is the fact that prasugrel was associated to significantly lower mortality and MACE incidence than ticagrelor, both at 30 days and one year after the event. No randomized studies compare these drugs clinically, head to head. The authors have based their conclusions on only 4 small randomized studies (approximately 200 patients in all) with surrogate end points (mainly platelet aggregometry) that had been designed to study pharmacokinetics and pharmacodynamics, not clinical efficacy, which had been added non-randomized data of post-hoc analyzes of other studies that had not been designed to compare the clinical benefit of P2Y₁₂ inhibitors. Most available data are from the in-hospital phase, no longer than 30 days post MI, conclusions at 12 months being drawn indirectly by meta-analysis.

Network meta-analyses are complex indirect studies introduced the last decade, designed as an alternative tool to conventional pair-wise meta-analyzes.^2-4 Systematic reviews have traditionally used pair-wise meta-analyzes that assess studies head to head to compare A vs B. When no such studies exist, when the number of patients is limited, or when more than one treatment needs to be assessed, we can use these indirect comparisons of a network of clinical trials, where treatments are related through a third study (A and B in relation to C).

Finally, sometimes network meta-analyses can include some head to head trials in addition to indirect evidence, which has been called mixed comparisons. Depending on the condition to be researched and the number of trials to be included, these network meta-analyses can become quite complex. A reliable network meta-analysis is based on a series of assumptions:

1. All relevant studies have been included.
2. Individual studies are not biased.
3. Trials are homogeneous as regards design and population.
4. There are no changes as regards effect, and when there are, the analysis is adjusted accordingly.
5. In mixed comparisons, trials are consistent, i.e. indirect and direct evidence match.

Despite this potential complexity, network analyses does not escape certain considerations and requisites characteristic of all meta-analyses, which Rafique et all have not entirely complied with in this publication. The studies included by them are based on different populations with different designs and definitions.

For example, Triton and Plato ─the most representative clinical trials on prasugrel and ticagrelor─ are rather different and hard to compare. Triton randomized patients only after knowing their coronary anatomy. After agreeing on an intervention strategy, Triton researchers did not allow the use of loading doses higher than 300 mg clopidogrel, as did Plato researchers, who enrolled patients prior catheterization and allowed any kind of strategy: intervention, surgery or medical treatment. Cardiovascular mortality in the control group was 2.4% in the Triton and 5.1% in the Plato.^5,6 From these completely different universes, Rafique et all extracted data from STEMI subgroups to include in the meta-analysis.

Methodologically, this meta-analysis also shows other serious limitations:

• There is no information on study heterogeneity, or publication bias. Most patients came from observational studies, not randomized.
• There is no mention of the quality of studies.
• There are no reports on sensitivity studies. These could have given information as to how the model behaves with different adjustments and if results are consistent based on direct or indirect evidence.
• Finally, the authors quote in their references different opinion studies by Serebruany and DiNicolantonio, who have been accused of bias and lack of objectivity.^7,8

At present, an ongoing randomized study is comparing prasugrel vs. ticagrelor head to head in patients with ACS treated with an invasive strategy with clinical end points: the ISAR-REACT 5, conducted by A. Kastrati et all in Germany, who plan to include more than 4 thousand patients. ⁹ This study will provide direct information on clinical outcomes with both drugs in one homogeneous population, as well as its advantages and limitations in different subgroups.

Meanwhile, comparing pears with apples does not seem the best way to choose what P2Y₁₂ inhibitors to use in STEMI patients.

Bibliography:

• Rafique AM, Nayyar P, Wang TY, et al. Optimal P2Y12 inhibitor in patients with ST-segment elevation myocardial infarction undergoing primary percutaneous coronary intervention: a network meta-analysis. J Am Coll Cardiol Intv 2016;9:1036–
• Jansen JP, Fleurence R, Devine B, Itzler R, Barrett A, Hawkins N, Lee K, Boersma C, Annemans L, Cappelleri JC: Interpreting indirect treatment comparisons and network meta-analysis for health-care decision making: report of the ISPOR Task Force on Indirect Treatment Comparisons Good Research Practices: Part 1. Value Health 2011, 14: 417-428.
• Thompson SG: Systematic Review: Why sources of heterogeneity in meta-analysis should be investigated. BMJ 1994, 309: 1351-1355.
• Jansen JP, Naci H. Is network meta-analysis as valid as standard pairwise meta-analysis? It all depends on the distribution of effect modifiers. BMC Medicine 2013; 11:159-166.
• Wiviott SD, Braunwald E, McCabe CH, et al. Prasugrel versus clopidogrel in patients with acute coronary syndromes. N Engl J Med 2007;357:2001-15.
• Wallentin L, Becker RC, Budaj A, et al. Ticagrelor versus clopidogrel in patients with acute coronary syndromes. N Engl J Med 2009;361:1045-57.
• James SK, Pieper KS, Cannon CP et al. Ticagrelor in Patients With Acute Coronary Syndromes and Stroke. Interpretation of Subgroups in Clinical Trials. Stroke 2013; 44:1477-1479.
• Montalescot G; van’t Hof AW. The hazards of approximations and extrapolations: Insights from the ATLANTIC study. Thromb Haemost 2015;114:11-13.
• Schulz S, Angiolillo DJ, Antoniucci D, et al. Randomized Comparison of Ticagrelor versus Prasugrel in Patients with Acute Coronary Syndrome and Planned Invasive Strategy—Design and Rationale of the Intracoronary Stenting and Antithrombotic Regimen: Rapid Early Action for Coronary Treatment (ISAR-REACT) 5 Trial. J of Cardiovasc Trans Res 2014; 7:91–

Courtesy of Dr. Alejandro Lakowsky, MTSAC.

Conflict of interest: AstraZeneca.

Your opinion matters to us. You are most welcome to leave your comment, reflection, question or any ideas down below.