试验排除标准对真实世界类风湿关节炎靶向治疗比较效果的影响:FIRST登记研究
基于FIRST登记2003—2023年5462例启用b/tsDMARDs的类风湿关节炎患者,按III期试验常见排除标准判定不合格。39.7%至少符合一条排除标准,调整临床特征后两组26周缓解率无差异(24.0% vs 23.2%;OR 0.90;95%CI 0.78~1.04)。JAK抑制剂相对TNF抑制剂的缓解优势在试验合格人群中显著(OR 1.39;95%CI 1.09~1.76),在不合格人群中减弱且无统计学意义(OR 1.24;95%CI 0.81~1.91)。
要点
- 约40%真实世界患者不符合III期试验入选标准,但调整后缓解率与合格者相当。
- JAK抑制剂相对TNF抑制剂的优势在试验不合格人群中减弱且无统计学意义。
- 真实世界比较效果分析需考虑人群构成与日历时间。
核心结论
- 约40%常规诊疗患者不符合III期试验标准,调整后结局与合格者相当。
- 药物间疗效差异在试验不合格人群中小于合格人群。
研究设计与样本
- FIRST登记研究,分析2003—2023年启用b/tsDMARDs的5462例类风湿关节炎患者。
- 以III期试验常见排除标准(如肾/肝功能损害)定义RCT不合格。
- 结局为治疗持续性和26周CDAI缓解,采用倾向评分加权并按RCT合格性分层。
主要发现
- 39.7%(2167/5462)至少符合一条RCT排除标准,不合格者基线疾病活动度更高、功能状态更差。
- 倾向评分调整后两组缓解率无差异(24.0% vs 23.2%;OR 0.90;95%CI 0.78~1.04)。
- 不合格比例因药物不同:IL-6抑制剂47.4%、CTLA4-Ig 45.1%、JAK抑制剂31.0%。
- RCT合格人群中JAK抑制剂缓解优于TNF抑制剂(OR 1.39;95%CI 1.09~1.76)。
- RCT不合格人群中该优势减弱且无统计学意义(OR 1.24;95%CI 0.81~1.91)。
关键数值与推荐
- RCT不合格定义采用III期试验常见排除标准(如肾/肝功能损害)。
- 缓解定义为26周CDAI缓解。
- 比较效果采用倾向评分加权并按RCT合格性分层。
临床怎么用
- 解读b/tsDMARDs真实世界比较效果时需考虑试验排除标准造成的人群构成差异。
- 对不符合试验标准的患者,药物间疗效差异可能小于试验合格人群。
- 真实世界比较分析应结合日历时间解释治疗效应。
局限性
- 原文摘要未报告局限性相关内容。
以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。
原文
OBJECTIVES: Randomised controlled trials (RCTs) use restrictive eligibility criteria, raising concerns about external validity for heterogeneous routine-care populations. We aimed to characterise real-world patients with rheumatoid arthritis who would be ineligible for phase III RCTs, to evaluate their outcomes on biological/targeted synthetic disease-modifying anti-rheumatic drugs (b/tsDMARDs) in routine care and to assess how eligibility status influences comparative effectiveness estimates. METHODS: We analysed 5462 patients who initiated b/tsDMARDs between 2003 and 2023. RCT ineligibility was defined using common phase III trial exclusions (eg, renal/hepatic impairment). Outcomes were treatment persistence and Clinical Disease Activity Index remission at 26 weeks. Comparative effectiveness across drug classes was estimated using propensity score weighting, stratified by RCT eligibility. RESULTS: Overall, 39.7% (2167/5462) met at least one RCT exclusion criterion and ineligible patients had higher baseline disease activity and worse functional status. However, after propensity score adjustment for clinical characteristics, remission rates did not differ between groups (24.0% vs 23.2%; OR 0.90; 95% CI 0.78 to 1.04). The proportion ineligible differed by b/tsDMARDs, highest with interleukin 6 inhibitors (47.4%) and CTLA4-Ig (45.1%) and lowest with Janus kinase (JAK) inhibitors (31.0%). In the RCT-eligible cohort, remission with JAK inhibitors exceeded that with tumour necrosis factor inhibitors (OR 1.39; 95% CI 1.09 to 1.76), consistent with head-to-head RCTs. In the RCT-ineligible cohort, this advantage was attenuated and no longer statistically significant (OR 1.24; 95% CI 0.81 to 1.91). CONCLUSIONS: Approximately 40% of routine-care patients would be RCT ineligible. Although these patients achieved comparable adjusted outcomes, between-drug differences were smaller than in trial-eligible populations. Real-world comparative analysis should therefore consider composition and calendar time when interpreting treatment effects.
引用
- Kubo S, Tanaka Y, Miyazaki Y, et al. Impact of trial exclusion criteria on real-world comparative effectiveness of targeted therapies in rheumatoid arthritis: the FIRST Registry. RMD Open. 2026. PMID: 42150842. doi:10.1136/rmdopen-2026-006825
供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。