Comparative Efficacy and Safety of 7 Antiplatelet Drugs and different combinations for the Treatment of Adults With ischemic stroke or transient ischemic attack : A Bayesian Network Meta-Analysis

Wenkui Li, Chuyue Wu, Qingyuan Wu, Lina Zhang, Yong Liu, Shengli Chen2023 · DOI

This document outlines a research protocol for a systematic review and meta-analysis. As it is a protocol (a pre-registered plan for research), it describes the intended methodology rather than the final results of the study.

TL;DR

Researchers are launching a comprehensive Bayesian network meta-analysis to rank seven major antiplatelet drugs and their combinations for treating non-cardioembolic ischemic stroke. By synthesizing data from existing randomized controlled trials, the study aims to identify which treatments offer the best protection against recurrent strokes while minimizing the dangerous risk of bleeding.

What Problem Does This Paper Address?

Ischemic stroke and transient ischemic attacks (TIAs) are leading causes of disability and death worldwide. To prevent recurrence, clinicians typically prescribe antiplatelet therapy. However, the "best" choice is far from settled. There are at least seven common drugs—including household names like aspirin and newer agents like ticagrelor—used either alone or in various combinations.

The challenge is twofold:

  1. The Efficacy-Safety Trade-off: A drug that is highly effective at preventing clots (efficacy) often carries a higher risk of causing internal bleeding (safety).
  2. Lack of Head-to-Head Data: While many trials compare Drug A to a placebo or Drug A to Drug B, we rarely have trials that compare all seven drugs simultaneously. This makes it difficult for doctors to know which specific regimen is truly superior.

Methods

The authors plan to use a Bayesian Network Meta-Analysis (NMA). Unlike a traditional meta-analysis that only compares two treatments, an NMA allows researchers to compare multiple treatments across a "network" of trials.

Key methodological steps include: * Systematic Review: A rigorous search of major medical databases (PubMed, EMBASE, etc.) to find all relevant Randomized Controlled Trials (RCTs). * Bayesian Framework: Using Bayesian statistics allows the researchers to calculate the probability of each drug being the "best" for a specific outcome. This provides a more intuitive ranking than frequentist p-values. * Software Stack: The analysis will be performed using ADDIS (Automated Data Extraction and Design of Information Systems) and R software, which are standard for complex network modeling. * Quality Control: The team will use the Cochrane Risk of Bias Tool to ensure that low-quality or "cherry-picked" trial data doesn't skew the results.

Datasets

As this is a meta-analysis, the "dataset" consists of the collective results from previously published Randomized Controlled Trials. The researchers will pull data from:

  • EMBASE
  • PubMed
  • Web of Science
  • The Cochrane Library

The study focuses specifically on adults with non-cardioembolic ischemic stroke or TIA, excluding strokes caused by heart issues (like atrial fibrillation), which require different types of medication (anticoagulants).

Key Insights and Ideas

  • Comprehensive Scope: The study will evaluate seven specific drugs: aspirin, clopidogrel, cilostazol, dipyridamole, ticagrelor, ticlopidine, and prasugrel.
  • Multi-Dimensional Outcomes: Success isn't just measured by preventing another stroke. The study will look at a "quad-threat" of outcomes: recurrent stroke/TIA, risk of bleeding, risk of death, and risk of myocardial infarction (heart attack).
  • Indirect Comparison: The "aha moment" of network meta-analysis is its ability to infer how Drug A compares to Drug C, even if no trial ever compared them directly, as long as both have been compared to Drug B.

Key Results

Not discussed in the paper. As this is a protocol, the data collection and analysis phases are currently in progress. The "Review Stage" at the time of this publication was listed as "Preliminary searches."

Limitations and Future Work

While the protocol is robust, the eventual study will face a few inherent limitations common to meta-analyses:

  • Trial Heterogeneity: Different trials may have followed patients for different lengths of time or used slightly different dosages, which can introduce "noise" into the network.
  • Data Availability: The analysis is only as good as the published RCTs available. If certain drug combinations haven't been studied in a randomized format, they cannot be included.
  • Future Directions: Once the ranking is established, the authors imply this could lead to updated clinical guidelines, helping neurologists move toward more personalized antiplatelet prescriptions based on a patient's specific risk profile for bleeding vs. clotting.

Summary updated

Abstract

Review question / Objective: Patients: patients with non-cardioembolic ischemic stroke or transient ischemic attack.Intervention and comparition: antiplatelet drugs such as aspirin
BibTeX
@misc{li2023comparative,
  author = {Wenkui Li and Chuyue Wu and Qingyuan Wu and Lina Zhang and Yong Liu and Shengli Chen},
  title = {{Comparative Efficacy and Safety of 7 Antiplatelet Drugs and different combinations for the Treatment of Adults With ischemic stroke or transient ischemic attack : A Bayesian Network Meta-Analysis}},
  year = {2023},
  doi = {10.37766/inplasy2023.4.0039},
}
Figure