Lorenzi Elizabeth, Crawford Amy M, Anderson Craig S, Menon Bijoy, Chen Xiaoying, Mistry Eva, Khatri Pooja, Elm Jordan J, Beall Jonathan, Saville Benjamin R, Berry Scott M, Lewis Roger J
Berry Consultants LLC, Austin, TX (E.L., A.M.C., S.M.B., R.J.L.).
George Institute for Global Health, University of New South Wales, Sydney, Australia (C.S.A., X.C.).
Stroke. 2025 Jan;56(1):198-208. doi: 10.1161/STROKEAHA.124.045754. Epub 2024 Dec 20.
Clinical trials of treatments for stroke have generally utilized 2-arm, randomized designs to evaluate a single intervention against a control. Running separate clinical trials, with each addressing a single therapeutic question, is resource intensive and slows evidence generation, especially in a field with rapidly expanding treatment options and evolving practices. Platform trials-randomized clinical trials designed to evaluate multiple interventions that may enter and exit the ongoing platform based on a master protocol-accelerate the investigation of multiple therapeutic options within a single infrastructure. This in turn has the potential to accelerate access to new interventions for patients with stroke that can save lives and improve outcomes. In the context of acute ischemic stroke, 2 new platform trials have been established, the STEP trial (StrokeNet Thrombectomy Endovascular Platform) and ACT-GLOBAL (A Multi-Factorial, Multi-Arm, Multi-Stage, Randomised, Global Adaptive Platform Trial for Stroke), to address multiple therapeutic questions simultaneously using a multifactorial design including Bayesian modeling and other adaptive features. These trials are designed to maximize the information obtained from each participant, to align clinical research more closely with the complexities of clinical care, and to accelerate the identification of effective therapies. This article explores conceptual, practical, and statistical considerations in the design and implementation of adaptive platform trials and highlights their potential to accelerate the identification of new therapies, management, and rehabilitation in stroke.
中风治疗的临床试验通常采用双臂随机设计，以评估单一干预措施与对照措施的效果。开展单独的临床试验，每个试验只解决一个治疗问题，这需要大量资源，而且会减缓证据生成的速度，尤其是在一个治疗选择迅速增加且实践不断演变的领域。平台试验——旨在评估多种干预措施的随机临床试验，这些干预措施可根据主方案进入和退出正在进行的平台——可在单一基础设施内加速对多种治疗选择的研究。这反过来有可能加快中风患者获得新干预措施的速度，从而挽救生命并改善治疗结果。在急性缺血性中风的背景下，已经开展了两项新的平台试验，即STEP试验（中风网血栓切除术血管内平台试验）和ACT-GLOBAL试验（一项针对中风的多因素、多臂、多阶段、随机、全球适应性平台试验），以采用包括贝叶斯建模和其他适应性特征在内的多因素设计同时解决多个治疗问题。这些试验旨在最大限度地从每个参与者身上获取信息，使临床研究更紧密地与临床护理的复杂性相匹配，并加速有效治疗方法的识别。本文探讨了适应性平台试验设计与实施中的概念、实践和统计方面的考虑因素，并强调了它们在加速中风新疗法、管理和康复识别方面的潜力。
