哮喘联盟

   摘要
   研究背景：哮喘严重急性加重(SevEx)属罕见事件，因此，传统的急性加重试验需要大样本和长期的研究才有足够的说服力。
   研究目标：建立一种包含临床相关恶化及严重急性加重信息的日记变量的复合终点，以便在早期临床发展阶段时做出决策，且可以使临床试验所需纳入的样本更少，持续时间更短。
   实验方法：研究纳入8项6～12个月使用布地奈德/福莫特罗(Symbicort SMART®)或福莫特罗(Oxis®)进行临床试验的数据(患者15524例)，对其中标准化收集的哮喘急性加重及日记卡变量数据进行分析。
   从获得的日记变量，如PEF、使用SABA缓解的次数、哮喘症状、夜间醒来的次数，我们证明了基于之前定义的恶化阈值及斜率的一系列日记事件来评估趋势。并通过Random Forest法确定最具预测性的变量，复合恶化（CompEX)被定义为第一个日记事件或严重急性加重（SevEx）。通过与 SevEx对比，评估作为分险指标的效果。
   研究结果：跟踪3个月时，相对8%有SevEx，20%的患者有经历CompEx事件，且不同研究在进行CompEx对比时区别不大。至少有50%的SevEx不能通过日记事件获得。通过PEF及单独使用SABA缓解定义的CompEx与更好的集体效能和统计效能相关。治疗3个月后CompEx的评估效能与SevEX的比较，发现随着时间延长及纳入的样本量足够大时，CompEx与SevEx评估效能相近。
   研究结论：与SevEx相比，CompEx使哮喘急性发作风险评估减少到3个月且试验所需纳入的样本更少。


 

（南方医科大学南方医院 赵文驱 赵海金 ）
2015,ERS Meeting

CompEx – A novel composite asthma exacerbation endpoint

Anne Fuhlbrigge, Thomas Bengtsson, Stefan Peterson, Alexandra Jauhiainen, Malin Fageras
2015,ERS Meeting
OA3258, http://abstract.ersnet.org/my-abstract-book-2015/

Abstract
Background: Severe exacerbations (SevEx) are rare events; hence, conventional exacerbation trials are large and lengthy to allow sufficient power.
Aims & Objectives: Establish a composite endpoint of diary variables capturing clinically relevant deteriorations and SevEx that can guide decision making during early clinical development and allow trials to contain fewer patients and be of shorter duration.
Methods: Data from eight 6–12 month trials (15524 patients) with budesonide/formoterol (Symbicort SMART®) or formoterol (Oxis®) using standardised collection of asthma exacerbations and diary card variables were investigated.
From available diary variables, ie PEF, rescue SABA use, asthma symptoms, night-time awakenings, we identified a series of diary events based on predefined threshold values of deteriorations from baseline and slopes to assess trends. Random Forest methodology identified the most predictive variables. Composite Exacerbation (CompEx) was defined as first event of either diary event or SevEx. Performance was assessed by comparing treatment effects as hazard ratios relative to SevEx.
Results: When censored at 3 months, 20% of patients experienced CompEx events vs 8% for SevEx with small variations between studies and CompEx algorithms. At least 50% of all SevEx were not captured by a diary event. CompEx defined by PEF and rescue SABA use alone was associated with best overall performance and statistical power. Treatment effects on CompEx at 3 months tracked the effect on SevEx. CompEx showed a similar effect profile as SevEx over time and in enriched populations.
Conclusion: CompEx allows evaluation of exacerbation risk reduction in 3-month trials involving fewer patients compared with SevEx.