CHOICE-2, FASTEST and CREST 2 results revealed during Opening Main Event
Researchers presented their findings on three Late-Breaking Science trials at the Opening Main Event of #ISC26 on Wednesday.
They found that:
- CHOICE-2 strongly supported intra-arterial thrombolysis following thrombectomy for acute ischemic stroke.
- FASTEST signals rFVIIa benefit in ICH.
- CREST 2 found that revascularization after asymptomatic carotid artery stenosis did not improve cognition.
Thrombolysis after thrombectomy improves outcomes for acute ischemic stroke
Thrombolysis following successful thrombectomy significantly improved outcomes for acute large vessel ischemic strokes. The addition of adjunctive intra-arterial alteplase after successful thrombectomy improved functional outcomes by an adjusted risk difference of 15% compared to thrombectomy alone. The CHOICE-2 trial confirmed results of the CHOICE trial showing improved outcomes with thrombectomy + thrombolysis vs. thrombectomy alone.
Ángel Chamorro, MD, PhD
The original CHOICE trial, originally presented at #ISC22, showed an 18% improvement in outcomes for combination therapy vs. thrombectomy alone, but the trial was terminated early due to expiration of the placebo agent. CHOICE-2 was designed to replicate and confirm the initial findings in a larger, adequately powered study.
A total of 433 patients were randomized to adjunctive alteplase (214) or thrombectomy alone (219) following large vessel occlusive stroke within 24 hours of symptom onset. The median age of patients was 76 years, and 51% were female. CHOICE-2 was conducted across 14 stroke centers in Spain.
The primary outcome was modified Rankin Scale (mRS) score of 0 or 1 at 90 days. Secondary outcomes included hypoperfusion on follow-up CT perfusion. Safety outcomes were symptomatic intracranial hemorrhage and all-cause mortality.
At 90 days, 57.5% of patients in the alteplase group had mRS of 0/1 vs. 42.5% in the thrombectomy-only arm (adjusted risk difference [ARD]= 15%, 95% CI 5.7-24.3, p=0.002) and a number need to treat of 7. There was statistically significant reduction in hypoperfusion at follow-up, 28.6% vs. 50.5% in the alteplase group (ARD=−22.0%, 95% CI −31.5% to −12.4%; p < 0.001). Symptomatic intracranial hemorrhage occurred rarely in both groups, although mortality at 90 days was 12.1% in the alteplase group and 6.4% in the thrombectomy-alone group (ARD = 5.9%; 95% CI 0.5%-11.3%; p=0.04).
The investigators noted that this difference warrants further study.
“The efficacy results were highly anticipated based on the CHOICE trial, which had already shown very promising signals,” Chamorro said. What surprised us in CHOICE-2 was the mortality difference, which was mainly driven by the unexpectedly low 6% death rate in the control arm mortality in the alteplase group.
CHOICE and similar trials are changing the paradigm of ischemic stroke treatment, Chamorro added. Angiography can only assess occlusion in macroscopic vessels. Micro-vessels account for 90% to 95% of total cerebral blood flow and are invisible on angiography. Thrombolysis appears to clear at least some of these micro-vessel blockages that cannot be imaged with angiography.
“Some might suggest that patients with less good angiographic results might also benefit from a CHOICE approach,” Chamorro said. “I assume this is something that will be further tested. If people are convinced that the combination therapy is safe, a 15% absolute improvement in the rate of cure for large vessel occlusion will have a quite positive cost-benefit. In a few months or years, we will just give the combination treatment.”
Subgroup analysis suggests potential benefit for rFVIIa in patients with ICH
Subgroup analysis of the largest imaging-based hyperacute stroke trial suggested that early use of recombinant Factor VIIa (rFVIIa) in acute spontaneous intracerebral hemorrhage (ICH) may reduce bleeding and improve outcomes in subgroups of patients with highest risk of ongoing bleeding. Overall findings of the FASTEST trial were neutral, and the trial was halted early for futility. Further testing of rFVIIa in patients at highest risk of continued bleeding is continuing.
Joseph Broderick, MD, FAHA
FASTEST compared rFVIIa vs. placebo for patients with ICH treated within two hours of onset or last known well, said Kazunari Toyoda, MD, PhD, FAHA, deputy director general of the National Cerebral and Cardiovascular Center in Osaka, Japan. A total of 328 patients were randomized to rFVIIa and 298 to placebo across 91 sites in Canada, Germany, Japan, Spain, U.K., and the U.S. All patients received both CT imaging of the head at baseline and 24 hours and baseline CT angiography (CTA) when obtained as standard of care.
The primary outcome was the modified Rankin Score (mRS) of 0-2, 3, and 4-6 at 180 days. Secondary outcomes included differences in bleeding between baseline and 24-hour head CT vs. placebo. Preplanned interim analyses were conducted after 200 and 430 participants had completed follow-up; the trial met prespecified criteria for futility at the second interim analysis.
There was no evidence of benefit for clinical outcome in the overall study population, Broderick said. Life-threatening thromboembolic complications occurred in 4.6% of the rFVIIa arm vs. 1.3% in the placebo arm, relative risk 3.41 (95% CI 1.14 – 10.15, p=0.02). rFVIIa was associated with decreased growth of ICH (-3.7 ml) and ICH and intraventricular hemorrhage (IVH) (−5.2 mL) between baseline and 24-hour CT scans vs. placebo.
Prespecified subgroup analyses at 180 days favored rFVIIa in patients with at least one spot sign (OR 1.86, 95% CI 0.94-3.68) and those treated within 90 minutes (OR 1.82, 95% CI 0.0.98 –3.4) although it was not statistically significant. In tertiary and exploratory analyses of benefit at 90 days for patients who had a spot sign (OR 2.52, 95% CI 1.2, 5.28) and in those treated within 90 minutes (OR 2.66, 95% CI 1.43, 4.96).
Patients with a spot sign also showed a substantial decrease in mean ICH growth (-9.4 ml) and ICH + IVH growth (-14.1 ml) compared to much smaller decreases in those lacking a spot sign. Decrease in moderate (≥- 6.1 ml) or severe (≥-12.5 ml) growth of ICH was substantially greater in both the spot sign and ≤90 minutes-to-treatment subgroups.
This reduction in growth of ICH is the largest observed with any medical therapy for spontaneous ICH, Broderick said. He added that larger volumes of hematoma expansion have been linked to increased risk for poor outcomes.
“A spot sign is a very good marker of continued hemorrhagic expansion and is even a better marker the closer you get to time zero,” Broderick said. “And the closer to time zero you can treat, the less bleeding you are likely to have with rFVIIa treatment. It makes biological and clinical sense, which is why we’re doing FASTEST Part 2 in these subgroups.”
Carotid artery stenosis revascularization did not affect cognition
A secondary analysis of the CREST-2 trial found that neither medical nor surgical interventions (carotid artery stenting or carotid endarterectomy) to treat asymptomatic carotid artery stenosis had any effect on cognition. Researchers followed more than 2,000 participants in the parent trial for up to four years after intervention and found no difference between revascularization plus intensive medical therapy vs. intensive medical therapy alone. A sub-analysis of patients with the lowest baseline cognition scores likewise found no change after revascularization.
Ronald M. Lazar, PhD, FAHA, FAAN
The parallel CREST-2 randomized clinical trials published in the NEJM in November 2025 found that stenting, but not endarterectomy, reduced stroke and death risk within four years for asymptomatic carotid stenosis compared to intensive medical therapy alone. The trials enrolled 2,485 patients with high-grade (≥70%) asymptomatic carotid stenosis in 155 centers across five countries. Patients were considered as treatment candidates for either transfemoral carotid artery stenting (1,245) or carotid endarterectomy (1,240) and then randomized either to the procedure or to medical therapy alone. The mean age was about 70, and roughly 37% were female. Both cohorts were compared with patients receiving intensive medical therapy alone.
The primary results showed that stents, but not endarterectomy, reduced stroke risk for asymptomatic carotid stenosis compared to medical therapy. A secondary analysis focused on changes in cognition from pre-treatment baseline up to four years. Cognition was assessed by telephone at baseline and annually after treatment using validated instruments for attention, learning, memory and executive function.
Lazar noted that those domains are particularly sensitive to perfusion failure seen in carotid disease.
An ad hoc sub-analysis explored cognition changes in patients with the poorest baseline cognition based on Z scores. Results confirmed the main finding with no difference between revascularization and medical therapy in cognition scores, even in a subgroup more likely to show improvement in cognition from improved perfusion.
Clinicians have long recognized that a decline in cognition in someone with severe carotid artery stenosis (≥70% stenosis) typically means that blood flow has declined to the brain. For patients with ≥70% stenosis, a change in cognition is recognized as one sign to reassess carotid stenosis progression.
“We have to look at other mechanisms of cognitive decline in those with asymptomatic carotid stenosis,” Lazar said. “It also means that our assumptions that restoring blood flow is sufficient to reverse neurologic injury need to be re-examined. Maybe the injury from this chronic disease is permanent, and reversibility may not be as achievable as we originally thought. Clinicians are going to conclude that we’re not going to get people back to their cognitive baseline by recanalization, so the only reason to do it is to reduce a risk for stroke or death.”
