Before you have your stroke ask your doctor what subpopulation you should be in so you are treated properly. It is your responsibility to have the type of stroke your doctor can treat, NOT your doctor's responsibility to be able to treat all stroke patients. Remember that and plan accordingly.
Amrou Sarrajhttps://orcid.org/0000-0001-5726-44781, Ameer E Hassanhttps://orcid.org/0000-0002-7148-76162, Michael Abraham3, Marc Ribo4, Spiros Blackburn5, Michael Chen6, Muhammad Shazam Hussain7, Vitor Mendes Pereira8, Santiago Ortega-Gutierrez9, Clark Sitton10, Phillip W Lavori11, Chunyan Cai12, Mohammed Rahbar12, Deep Pujarahttps://orcid.org/0000-0001-6187-894X1, Faris Shakerhttps://orcid.org/0000-0003-4462-52321, Maarten G Lansberghttps://orcid.org/0000-0002-3545-692713, Bruce Campbellhttps://orcid.org/0000-0003-3632-943314, James C Grotta15, Gregory W Albers13, and on behalf of the SELECT2 Investigators
RationaleRandomized evidence for endovascular thrombectomy safety and efficacy in patients with large core strokes is lacking.
Aims
To demonstrate endovascular thrombectomy efficacy and safety in patients with large core on non-contrast CT or perfusion imaging (CT/MR) and determine if there is heterogeneity of treatment effect in large cores based on the imaging modality.
Design
SELECT2 is a prospective, randomized, multi-center, assessor-blinded controlled trial with adaptive enrichment design, enrolling up to 560 patients.
Procedure
Patients who meet the clinical criteria and have anterior circulation large vessel occlusions with large core on either NCCT (ASPECTS 3–5) or perfusion imaging (CTP [rCBF < 30%] and/or MRI [ADC < 620] ≥ 50 cc) will be randomized in a 1:1 ratio to undergo endovascular thrombectomy or medical management (MM) only up to 24 h of last known well.
Study outcomesThe distribution of 90-day mRS scores is the primary outcome. Functional independence (mRS = 0–2) rate is a secondary outcome. Other secondary outcomes include safety (symptomatic ICH, neurological worsening, mortality) and imaging outcomes.
Analysis
A normal approximation of the Wilcoxon-Mann-Whitney test (the generalized likelihood ratio test) to assess the primary outcome. Functional independence rates, safety and imaging outcomes will also be compared.
Discussion
The SELECT2 trial will evaluate endovascular thrombectomy safety and efficacy in large cores on either CT or perfusion imaging and may provide randomized evidence to extend endovascular thrombectomy eligibility to larger population.
Registration: ClinicalTrials.gov–NCT03876457
Keywords
Acute stroke therapy, thrombectomy, clinical trial, large vessel occlusion, ischemic stroke, protocols
1Department of Neurology, UT McGovern Medical School, Houston, USA
2Department of Neurology, Valley Baptist Medical Center, Harlingen, USA
3Department of Neurology, University of Kansas Medical Center, Kansas City, USA
4Department of Neurology, Vall d'Hebron Hospital Universitari, Barcelona, Spain
5Department of Neurosurgery, UT McGovern Medical School, Houston, USA
6Department of Neurosurgery, Rush University, Chicago, USA
7Department of Neurology, Cleveland Clinic, Cleveland, USA
8Department of Neurology, Toronto Western Hospital, Toronto, Canada
9Department of Neurology, University of Iowa Hospitals, Iowa City, USA
10Department of Neuroradiology, UT McGovern Medical School, Houston, USA
11Biomedical Data Science, Stanford University, Stanford, USA
12CCTS, UT McGovern Medical School, Houston, USA
13Department of Neurology, Stanford University, Stanford, USA
14Department of Neurology, Royal Melbourne Hospital, Parkville, Australia
15Department of Neurology, Memorial Hermann–Texas Medical Center, Houston, USA
*Investigators listed in eTable 3.
Corresponding author(s):
Amrou Sarraj, UT McGovern Medical School, 6431 Fannin St #7.044; Houston, TX 77030, USA. Email: Amrou.Sarraj@uth.tmc.edu
Introduction and rationale
The randomized trials1–3 that established endovascular thrombectomy (EVT) efficacy and safety focused on patients with favorable imaging profiles on Non-contrast Computed Tomography (NCCT) (Alberta Stroke Program Early CT Score (ASPECTS ≥ 6)) or perfusion imaging (presence of mismatch defined as small ischemic core with a large penumbra). As a result, these trials largely excluded patients with large ischemic core (ASPECTS < 6 or an ischemic core volume >50–70 cc). Thus, the role of EVT in these patients is not established.
Patients with larger baseline ischemic core are less likely to achieve functional independence given irreversible brain damage has already occurred. Moreover, reperfusion of large infarcts may increase the risk of hemorrhagic transformation. Therefore, EVT risk–benefit ratio is not well-defined in this subpopulation.
Large core definition also differs between NCCT and perfusion (Figure 1). While NCCT-ASPECTS detects hypodense tissue, perfusion identifies regions of very low blood flow or volume and magnetic resonance detects tissue with restricted diffusion of water molecules. When these images are obtained concurrently, there is often disparity between them.4 How these disagreements may affect EVT efficacy and safety is unclear.
Figure 1. Illustration of large core on CT and perfusion imaging – illustrates imaging eligibility based on (a) NCCT, (b) CT or (c) MR-perfusion.
Large core patients represent an important sub-population that physicians encounter often.4,5 Of 445 enrolled in SELECT, 105 (24%) had large core on either Computed Tomography (CT) or CT Perfusion (CTP). EVT resulted in higher functional independence rates, a shift towards better outcomes, smaller final infarct volumes and less infarct growth. There was a nominally higher rate of symptomatic Intracerebral Hemorrhage (ICT) with EVT. A Highly Effective Reperfusion evaluated in Multiple Endovascular Stroke Trials (HERMES) post-hoc analysis of only 50 patients with CTP infarct volume ≥70 cc showed EVT was associated with potential benefit up to 150 cc.6 A recent meta-analysis of data from cohort studies and RCTs subgroups showed EVT to be associated with improved functional independence and lower mortality without significant increase in symptomatic ICH across the various large core definitions on the different imaging modalities, core size thresholds, and time windows.7 This gap in evidence results in inconsistent EVT treatment and calls for further randomized evidence to establish EVT efficacy and safety in this population and assess if there was a difference in treatment effect on different imaging modalities, size cutoffs or treatment time windows.
Methods
Study objectives
•
Primary aim: Evaluate if EVT as compared to MM will be efficacious and safe in patients with large core on NCCT (ASPECTS 3–5) or ≥50 cc using CT perfusion (relative cerebral blood flow, rCBF < 30%) or diffusion MRI (apparent diffusion co-efficient, ADC < 620), treated within 0–24 h from last known well (LKW).
•
Secondary aim: Assess if there is heterogeneity of EVT treatment effect in large core based on different imaging modalities.
SELECT2 is an investigator-initiated, multicenter, phase-III, prospective, randomized, open-label, blinded outcome (PROBE) clinical trial. Local Institutional Review Boards approvals are obtained prior to inclusion of the study sites. Patients (up to 560) with LVO in the intracranial or cervical ICA and MCA/M1 segment (the horizontal segment proximal to the Sylvian fissure) who have large core on NCCT (ASPECTS 3–5), CT perfusion (rCBF < 30% ≥ 50 cc) or MR diffusion (ADC < 620 ≥ 50 cc) are enrolled at 30 sites in the United States, Canada, Australia, New Zealand and Europe (Figure 1). ASPECTS determination is made by site investigators. Eligible patients or their legally authorized representatives provide informed consent prior to enrollment. Patients are randomized through a secure web-based system to receive EVT versus MM (Figure 2). Eligibility criteria are provided in eTable 1. Details regarding randomization and treatment allocation are provided in the supplemental materials.
Figure 2. Study design: randomization algorithm.
Study assessments
Clinical assessments are performed at baseline, 24 h after randomization, hospital discharge, 30 and 90 days and include modified Rankin Scale (mRS), National Institutes of Health Stroke Scale (NIHSS), and quality of life measure (Neuro-QoL). The Neuroimaging Core lab at the UTHealth consisting of experienced neuro-radiologists blindly adjudicates study images. eTable 2 describes the schedule of events.
Study outcomes
The primary outcome is the distribution of mRS scores at 90 days, performed by certified raters blinded to patient characteristics and treatment allocation, in-person or by telephone. Functional independence (90-day mRS = 0–2) is a secondary outcome. Additional details regarding secondary and imaging outcomes are provided in the supplemental materials.
Safety monitoring
SELECT2 has an independent Data and Safety Monitoring Board (DSMB) (comprised of vascular neurologists, neurointerventionalists, and a biostatistician) and an independent medical safety monitor. The description of their duties is provided in the supplemental materials.
Statistical analysis
Two interim and a final analysis will be conducted using adaptive enrichment design. The primary analysis will be under intention-to-treat principle. The primary outcome will be evaluated using a normal approximation of the Wilcoxon-Mann-Whitney test. A statistical analysis protocol (SAP) will be finalized prior to the data lock at the first interim analysis. The core salient features of the SAP as well as details regarding adaptive enrichment design and sample size justification are provided in the supplemental materials. Projected-90-day mRS distribution used for power-analysis is shown in Figure 3.
Figure 3. Projected 90-day mRS distribution – demonstrates projections based on SELECT prospective cohort data and simulations used for study power and sample size calculations.
Study organization
A steering committee, of vascular neurologists and neurointerventionalists with experience in conducting clinical trials, advises the Global PI on study management on quarterly basis.
Discussion
Patients presenting with large core infarcts on imaging are an important subpopulation without high-level evidence thus a trial to assess EVT efficacy and safety has been recognized among the highest priorities.8
Large core definition may differ between imaging modalities which would probably influence EVT outcomes in this population.5 In SELECT2, we will assess if there is heterogeneity in treatment effect in those with large core on CT as compared with those on perfusion images. Other RCTs evaluate EVT potential benefit in patients with large core on CT or MRI without including perfusion images. SELECT2 will provide high-level evidence on the possible additional benefit of perfusion images in identifying those who may benefit and those who may have a higher risk of reperfusion injury.
Furthermore, SELECT2 has a unique design that will examine subgroups with paucity of data that precludes conclusive EVT efficacy and safety. One such subgroup is patients with ASPECTS 6–10 and large ischemic core (≥50 mL) on perfusion imaging presenting within 6 h of LKW. Enrollment of these patients was limited in early window trials (eFigure 1). Furthermore, in SELECT, patients with ASPECTS 6-10 and ischemic core ≥50 mL demonstrated significantly worse functional independence and safety when compared to those with ASPECTS 6–10 and ischemic core <50 mL.9 Additionally, in the late window; only 18 patients with ASPECTS 6–10 and ischemic core of 50–70 mL were enrolled in DEFUSE3 trial3 while DAWN enrollment cutoff was 50 mL.2 Thus, patients with ASPECTS6-10 but ischemic core ≥50 mL lack definitive evidence of EVT benefit beyond 6 h of LKW. SELECT2 have pre-specified analyses evaluating EVT efficacy and safety in these subgroups.
Several variables can affect EVT outcomes in large core strokes. Infarct size and reperfusion time are the two most important.5 SELECT2 has an adaptive enrichment design to assess those.
Whether thrombectomy safety and efficacy can be extended to large core patients remains to be seen. If so, to what stroke size, on what imaging modality and how late are also unknowns. All of these are unanswered questions of important significance that we attempt to address in the SELECT2 trial.
Conclusion
The SELECT2 trial will evaluate EVT safety and efficacy in patients with large core on different imaging modalities up to 24 h.
