CREST-2 Reaches A Surgical Milestone

The Carotid Revascularization and Medical Management for Asymptomatic Carotid Stenosis Trial (CREST-2) includes two concurrently running two-arm trials assessing intensive medical management (IMM) with or without carotid revascularization.¹ One trial tests endarterectomy versus IMM; the other tests stenting versus IMM. Each trial had an original recruitment target of 1,240 patients with ≥70% asymptomatic carotid stenosis, randomized 1:1 in each arm. CREST-2 reached an important milestone on September 29, 2023, when the endarterectomy trial achieved its prespecified sample-size. This goal was achieved despite the COVID-19 pandemic, which at one point suspended new enrollments.² The National Institute of Neurological Disorders and Stroke (NINDS) continues to provide CREST-2 with direct support and guidance as well as indirect support through access to the central Institutional Review Board of the NINDS StrokeNet clinical trial network. Most surgical trials never achieve their recruitment goals. In 2,542 surgical randomized controlled trials (RCTs) initiated between 2010 and 2014, only 45.9% achieved target-enrollment, and only 14.6% achieved target-enrollment on time.³ The Stent-Supported Percutaneous-Angioplasty of the Carotid Artery Versus Endarterectomy-2 (SPACE-2) was the first sizable RCT with a medical-management group to conclude in >2 decades,⁴ but it closed prematurely due to low enrollment, thereby compromising power. SPACE-2 investigators attributed their recruitment challenges to a financial environment where all hospitals (including academic centers) are “expected to maintain annual performance targets,” “a lack of incentives” such as instances where only participation in a trial attracts reimbursement for a procedure, and “a lack of academic self-awareness.”⁵ The stakes are high in RCTs, since a positive finding may be missed for failure to recruit an adequate number of patients. CREST-2 surgeon investigators demonstrated exceptional dedication by completing enrollment in the endarterectomy trial and deserve credit (Supplementary Material).

Since the original trials for asymptomatic disease,^6–8 perioperative monitoring, periprocedural pharmacotherapy, and surgical technique have improved significantly.⁹ The 5-year stroke or death rate was 2.5% in the SPACE-2 trial, a substantial reduction from the 11.8% rate reported in the Asymptomatic Carotid Surgery Trial (ACST) almost two decades ago.⁷ Concurrent with improvements in perioperative management, there have been improvements in the intensity, effectiveness, and quality of vascular risk factor management. A recent retrospective cohort study of 3,737 patients managed non-operatively in the Kaiser Permanente Northern California health system reported a 5-year ipsilateral stroke risk of 4.7%.¹⁰ Evidence of improved risk factor management can also be seen within randomized trials. In ACST, there was a steady increase in the rate of lipid lowering drug use by calendar year. In the same trial, there was an increase in the use of anti-hypertensive drugs by calendar year, with a corresponding decrease in diastolic blood pressure. In the Carotid Revascularization, Endarterectomy versus Stent Trial (CREST), achievement of the low-density lipoprotein (LDL) cholesterol goal improved from 59.1% at baseline to 73.6% after 48 months in the trial (P<0.001), achievement of the systolic blood pressure goal improved from 51.6% to 65.1% (P<0.001), achievement of the glucose goal improved from 74.9% to 80.7% (P=0.0101), and nonsmoking status improved from 74.4% to 80.9% (P<0.0001).¹¹ Outside of clinical trials, a declining base rate of stroke likely reflects better medical management. At baseline, 74% of patients had a blood pressure <140/90 mm Hg and the median LDL level was 87 mg/dL in the Kaiser Permanente cohort study.¹⁰ For investigators, randomization in CREST-2 is therefore ethically supported by the uncertainty principle: there are data suggesting that either approach – medical therapy with or without revascularization - may be similarly effective.¹² It also appears that many Vascular Surgeons have clinical equipoise regarding the performance of endarterectomy for asymptomatic disease. Among Medicare-beneficiaries, the number of endarterectomies performed has dropped from 81,306 in 1999 to 36,325 in 2014.¹³

Evolving guideline recommendations also reflect growing uncertainty about endarterectomy for asymptomatic disease. In 2006, the American Heart Association guideline regarded endarterectomy for asymptomatic disease as a Class 1A recommendation;¹⁴ by 2014, it had been downgraded to Class 2A.¹⁵ The 2022 Society for Vascular Surgery (SVS) guideline assigned endarterectomy a Grade 1B recommendation, i.e., with moderate quality of evidence.¹⁶ We anticipate that CREST-2 will help fill this gap in evidence soon. In 2023, the European SVS considered IMM a Class 1A recommendation unless patients had ≥5 years of life-expectancy, favorable anatomy, and additional radiologic high-risk features like intraplaque hemorrhage (IPH) or microembolization.¹⁷ While radiologic high-risk features are associated with symptomatic plaques,¹⁸ plaque biomarker-guided revascularization in asymptomatic patients lacks support from randomized trials. Some may argue that it would be unethical not to treat a patient with biomarker evidence of an unstable plaque. Clearly such individuals are at increased risk of ipsilateral stroke. In addition, the Atherothrombosis Intervention in Metabolic Syndrome with Low High-Density Lipoprotein/High Triglycerides magnetic resonance imaging (AIM-HIGH MRI) carotid sub-study found that carotid plaques with IPH can progress over two years despite intensive lipid lowering therapy.¹⁹ The requirement, however, for randomized trial evidence of efficacy for revascularization in the asymptomatic unstable plaque population is reasonable because of the potential for increased periprocedural complications by virtue of plaque instability. The type of revascularization performed (endarterectomy or stenting) may also impact the relative benefit. A meta-analysis of nearly 500 stented patients found that the presence of IPH doubled the risk of post-procedural infarction detected on diffusion-weighted MRI.²⁰ It will be harder to find enough patients with IPH to randomize than it was to find patients to randomize in CREST-2, but the rate of endpoints, and, hence, sample size requirements would be expected to be favorable in such a study.

With recruitment complete, the CREST-2 endarterectomy trial now continues in follow-up. Patients will require continued titration of medical management to achieve the two primary trial goals of IMM: maintain systolic blood pressure <130 mm Hg and low-density lipoprotein <70 mg/dL. Patients will be followed for up to 4 years after enrollment in CREST-2. However, benefits from revascularization may accrue beyond that time point since the “price” for endarterectomy is paid early, while the “price” for IMM accrues over a longer period. Therefore CREST-2 sites have been invited to participate in a CREST-2 Long-term Observational Extension study to collect even longer-term outcomes data. This study is designed for targeted efficiency. Patients are being followed for stroke endpoints through 6-monthly telephone interviews, and strokes are confirmed by a review of medical-records by a central adjudication committee.

Randomized procedural trials are onerous and require physicians to courageously accept that there are no substitutes for reliably defining standard-of-care. While there is growing interest in non-randomized real-world evidence (RWE) studies, often-times incorporating propensity-score matching to emulate RCT conditions, there remain important differences in the results generated between RWE studies and their randomized counterparts.²¹ Randomized procedural trials require dedication to advancing the field and a high level of trust between patients and their surgeons. We are grateful to our surgeons, coordinators, staff, patients, and their family members, and look forward to completing recruitment in the stenting trial soon.