Abstract
CREST-2 findings are unchanged using an aggregated control group for the CAS and CEA trial
Jagiellonian University Department of Cardiac and Vascular Diseases, Krakow, Poland
St. John Paul II Hospital, Krakow, Poland
Department of Bioinformatics and Telemedicine, Jagiellonian University Medical College, Krakow, Poland
Center for Digital Medicine and Robotics, Jagiellonian University Medical College, Krakow, Poland
Jagiellonian University Department of Interventional Cardiology, Krakow, Poland
CardioVascular Center, Frankfurt, Germany
Image-Guided Therapy Research Facility, University of Dundee, Dundee, UK
University Department of Neurology with Stroke Unit, Holy Spirit Multispecialty Regional Hospital, Sandomierz, Poland
MVZ-Department Structural Heart Disease, Asklepios Clinic/Klinik St Georg, Hamburg, Germany
Department of Radiology, Aretaieion University Hospital, National and Kapodistrian University of Athens, Athens, Greece
Klinik für Diagnostische und Interventionelle Neuroradiologie, Klinikum Bremen Mitte Bremen, Bremen, Germany
Department of Radiology, Ninewells Hospital, Dundee, UK
Chair of Neuroradiology, University of Dundee, Dundee, UK
Adv Interv Cardiol 2026; 22, 2 (84): 178–189
Introduction
CREST-2 comprised two parallel observer-blinded randomised controlled trials evaluating carotid revascularisation (carotid artery stenting – CAS, or carotid endarterectomy – CEA) plus intensive medical management (IMM; supervised pharmacologic therapy and risk factor control and coached lifestyle modification) versus IMM alone in patients with asymptomatic ≥ 70% carotid stenosis. Each trial carried an independent IMM control arm. The primary endpoint (peri-procedural stroke/death or ipsilateral ischaemic stroke thereafter by 4 years) occurred, in the CAS trial, in 2.8% vs. 6.0% (IMM + CAS vs. IMM; p = 0.02). The effect of CEA did not reach significance; 3.7% vs. 5.3% (IMM + CEA vs. IMM; p = 0.24).
Aim
To test the hypothesis that the divergent control-arm event rates – rather than true differential efficacy of the interventional treatments – could underlie the CAS efficacy and CEA failure in CREST-2, we used a single combined control group as a balanced reference for both interventional treatment arms.
Material and methods
An Aggregated Control Group of IMM-only (n = 1,252) was formed by merging the CREST-2 IMM control arms patient data and outcomes. Kaplan-Meier analysis was performed for the CAS and CEA treatment vs. the Aggregated Control Group, consistent with the trial statistical methodology (intent-to-treat).
Results
In the Aggregated Control Group, the primary endpoint occurred in 5.65% of patients (95% CI: 4.47–7.12). Absolute risk reduction with CAS + IMM (n = 616) was 2.85% (relative risk reduction 50.4%; number-needed-to-treat 35, p = 0.0089). Absolute risk reduction with CEA + IMM (n = 617) was 1.95% (p = 0.0871, a maintained lack of statistical significance).
Conclusions
Aggregated Control Group analysis of CREST-2, minimising control arms biases, confirmed CAS efficacy (50.4% relative risk reduction) and maintained failure of CEA in primary prevention of ipsilateral stroke in patients with asymptomatic carotid stenosis under intensive medical management. This refutes control-arm disparity as an explanation for the divergent outcomes with CAS vs. CEA.
Keywords
asymptomatic carotid artery stenosis, carotid atherosclerosis, stroke, carotid-related stroke, medical management, interventional therapy, carotid artery stenting, carotid endarterectomy, CREST-2
Indexed in
Integrated with
