Endovascular thrombectomy after large-vessel ischemic
stroke: a meta-analysis of individual patient data fromfive randomised trials
HERMES CollaboratorsHighly Effective Reperfusion evaluated in Multiple Endovascular Stroke trials (HERMES)
18/02/2016 HERMES Collaboration 1
Disclosure slide
• Medtronic provided an unrestricted grant to the University of Calgary for the HERMES collaboration
18/02/2016 HERMES Collaboration 2
HERMES collaboration
• 5 trials published in early 2015
• All supported endovascular thrombectomy as a definitive treatment for anterior circulation, large vessel occlusive ischemic stroke
• Pooled analyses of individual patient data will allow:
– Greater precision
– Analysis of subgroups
18/02/2016 HERMES Collaboration 3
Methods
• A statistical analysis plan was formulated and finalized prior to data pooling
• Two statisticians collated and pooled the data – both are independent of the investigators and the relevant sponsors
• A steering committee oversaw the analysis
• All interpretation and analyses were independent of the study sponsors
18/02/2016 HERMES Collaboration 4
Methods
• A literature review confirmed that 5 major trials examining modern endovascular stroke treatment were published at the time of analysis
• A mixed effects model, with study as a random variable was used to assess the treatment effect across the 5 studies
• Sub-groups of interest were age, sex, occlusion location, ASPECTS score, treatment with alteplase and time from randomization
18/02/2016 HERMES Collaboration 5
Baseline Characteristics(N=1287)
18/02/2016 HERMES Collaboration 6
Baseline Characteristics
18/02/2016 HERMES Collaboration 7
Overall Treatment EffectNNT = 2.6
18/02/2016 HERMES Collaboration 8
Overall Treatment Effect
18/02/2016 HERMES Collaboration 9
Treatment effect is the same in men and women (pint=0.36)
18/02/2016 HERMES Collaboration 10
Treatment effect by agemRS 0-2 at 90 days
18/02/2016 HERMES Collaboration 11
Treatment effect is strong across occlusion sites (pint=0.35)
18/02/2016 HERMES Collaboration 12
Effect size by NIHSS, time to randomization, and tPA use
18/02/2016 HERMES Collaboration 13
Treatment effect is present for all ASPECTS category (pint=0.49)
18/02/2016 HERMES Collaboration 14
NOTE:42% = 5;30% = 4;28% < 4
18/02/2016 HERMES Collaboration 15
Constant Effect Size by Age
18/02/2016 HERMES Collaboration 16
Constant effect size by NIHSS
18/02/2016 HERMES Collaboration 17
Conclusions
• Endovascular treatment is a highly effective therapy across all subgroups
• These data may provide additional support for endovascular treatment in subgroups not addressed in individual trials
• Caution is still warranted in interpreting results for groups with small numbers
• The full paper and supplementary data are published on-line today in the Lancet
18/02/2016 HERMES Collaboration 18
18/02/2016 HERMES Collaboration 19
Highly Effective Reperfusion evaluated in Multiple Endovascular Stroke trials
(HERMES) collaboration
18/02/2016 HERMES Collaboration 20
Thank You!