CREATE Centre, Division of Infection and Immunity, Cardiff University School of Medicine, Wales, UK.
Rheumazentrum Ruhrgebiet, Ruhr-University Bochum, Herne, Germany.
Arthritis Res Ther. 2019 Jan 22;21(1):32. doi: 10.1186/s13075-019-1812-3.
Spondyloarthritis comprises a group of inflammatory diseases, characterised by inflammation within axial joints and/or peripheral arthritis, enthesitis and dactylitis. An increasing number of biologic treatments, including biosimilars, are available for the treatment of spondyloarthritis. Although there are a growing number of randomised controlled trials assessing treatments in spondyloarthritis, there is a paucity of data from head-to-head studies. Comparative data are required so that clinicians and payers have the level of evidence required to inform clinical decision-making and health economic assessments. In the absence of head-to-head studies, statistical methods such as network meta-analyses and matching-adjusted indirect comparisons (MAICs) are used for assessing comparative effectiveness.Network meta-analysis can be used to compare treatments for trials using a common comparator (e.g. placebo); however, for those without a common comparator or where considerable heterogeneity exists between the study populations, a MAIC that controls for differences in study design and baseline patient characteristics may be used. MAICs, unlike network meta-analyses, are of value for longer-term comparisons beyond the placebo-controlled phase of clinical trials, which is important for chronic diseases requiring long-term treatment, like spondyloarthritis. At present, there are a number of limitations that restrict the effectiveness of MAIC, such as the poor availability of individual patient-level data from trials, which results in patient-level data from one trial being compared with published whole-population data from another. Despite these limitations, drug reimbursement agencies are increasingly accepting MAIC as a means of comparative effectiveness and greater methodological guidance is needed.This report highlights a number of challenges that are specific to conducting comparative studies like MAIC in spondyloarthritis, including disease heterogeneity, the paucity of biomarkers and the duration of studies required for radiographic endpoints in this slow-progressing disease.
脊柱关节炎包括一组炎症性疾病，其特征为中轴关节内炎症和/或外周关节炎、附着点炎和指（趾）炎。目前有越来越多的生物制剂治疗药物，包括生物类似药，可用于治疗脊柱关节炎。尽管有越来越多的随机对照试验评估脊柱关节炎的治疗方法，但缺乏头对头研究的数据。需要比较数据，以便临床医生和支付者拥有所需的证据级别，以告知临床决策和卫生经济评估。由于缺乏头对头研究，统计方法如网络荟萃分析和匹配调整间接比较（MAIC）用于评估比较疗效。网络荟萃分析可用于比较使用共同对照剂（如安慰剂）的试验中的治疗方法；然而，对于没有共同对照剂或研究人群之间存在相当大的异质性的情况，可使用 MAIC 来控制研究设计和基线患者特征的差异。与网络荟萃分析不同，MAIC 可用于临床试验安慰剂对照阶段之外的长期比较，这对于需要长期治疗的慢性疾病（如脊柱关节炎）很重要。目前，有许多限制因素限制了 MAIC 的有效性，例如试验中个体患者水平数据的可用性较差，这导致一个试验的患者水平数据与另一个试验的全人群数据进行比较。尽管存在这些限制，但药物报销机构越来越接受 MAIC 作为比较疗效的一种手段，需要更多的方法学指导。本报告强调了在脊柱关节炎中进行 MAIC 等比较研究的一些具体挑战，包括疾病异质性、生物标志物的缺乏以及这种进展缓慢的疾病中需要进行放射学终点研究的时间。
