On superficial review, CREST2 appears rigorous and its conclusions undeniable, especially since it is published in the highly reputable New England Journal of Medecine. However, a detailed analysis reveals that the study is poorly designed and has severe limitations. As a result, some of its conclusions are invalid and others are misleading.
Serious design flaws of CREST2
The duration of the study is far too short
The authors of CREST2 have overlooked the most serious limitation of their trials: They compare two by two, three treatments. Two of which, carotid surgery and angioplasty, have a large part of their complications occurring in the first month of the study, while the third treatment, purely medical, which leaves the atherosclerotic plaque in place, has its complications occurring over time. Beyond the post-operative period, all patients have a risk of non-carotid stroke, mainly cardioembolic. But those only receiving medical treatment also have the risk of rupture of the fibrous cap of their atherosclerotic plaque or even of its evolution towards occlusion. The stroke rate of asymptomatic carotid stenoses is low, which considerably increases the time required for interventional treatments to neutralize the rate of postoperative complications.
Therefore, the shorter the study, the more it is biased against interventional procedures. Ideally, such research should track patients for at least as long as half the group survives, roughly 10 years.
Medical treatment alone, even at 4 years, has a much higher stroke rate than angioplasty or surgery. In addition, this approach does not account for the portion of only medically treated patients who may live beyond four years and up to 30 years but still carry the high-risk atheromatous plaque.
It is not possible to really compare surgery and angioplasty
Crest2 is made up of two separate trials.
At 4 years, angioplasty proved its superiority over medical treatment alone.
And surgery does not proved its superiority over medical treatment alone, likely due to an insufficient follow-up period.
Within 44 days post-procedure, the complication rates are similar. There were 9/617 post-surgery strokes and 8/616 post-angioplasty strokes.
The difference occurs later, related to the rate of stroke, which is partly cardioembolic.
Concluding that carotid endarterectomy does not demonstrate its superiority over medical treatment at four years sends a misleading message: surgery is not effective on asymptomatic stenoses, but carotid angioplasty is effective, although there is no real comparison between surgery and angioplasty.
A trial must be stopped if one treatment is superior to the other.
Is it ethical, given that in the angioplasty-medical treatment alone trial, angioplasty has demonstrated its superiority over medical treatment-alone, to continue this trial which will expose patients in the medical treatment-alone group to an increased risk of stroke?
Commentary on the limitations of CREST2.
First limitation
“First, although stroke adjudicators were unaware of the treatment assignments, patients and treating physicians were not.”
This limitation is correct and not open to criticism.
Second limitation
“Second, various changes in medical-therapy practices occurred during the trials that could lower stroke rates and negate any additional benefit of revascularization…”
The authors only discuss the potential progress of medical treatment, focusing their discussion on improving the care of diabetic patients. They do not discuss the potential advances in surgery and angioplasty for diabetic patients.
Furthermore, the authors overlook a limitation of medical treatment alone. In everyday life, patients medically treated for asymptomatic severe carotid stenosis will not have, like those in CREST2, such regular, frequent, and expert follow-up, with Rankin and NIHSS assessments at each visit and risk factor management directly supervised by a commercial company. And if they request it, free medications, including alurocumab, offered by the pharmaceutical company that produces it.
Despite this superabundance of attention, the results of medical treatment alone have not really improved since the publication of the ACAS trial 30 years ago (2). This can be explained by the fact that the rupture of the fibrous cap, responsible for 70 to 80% of strokes of carotid origin, does not appear to be delayed by this very high level of care.
Third limitation
“Third, revascularization was performed only by well-trained and certified high-volume operators…”
Although surgeon complication rates are low (~1.5%), further reductions could be achieved through optimized indications and perioperative protocols.
For example, regarding the invasive management of diabetic patients. In CREST2, there is no benefit of interventional treatments compared to medical treatment alone in diabetic patients (Figure 3). This increased risk had already been observed in the ACAS (3) and NASCET (4) trials. This is probably due to the increased risk of hemodynamic stroke in diabetic patients, aggravating the risk of ischemic and hemorrhagic strokes (5).
This increased risk of decrease in vascular reserve and cerebral vasomotor activity in diabetics patients can be detected pre-operatively (6) or during the operation (7). It can be prevented, in the event of a decrease in vascular reserve, regardless of the technique, surgical (8) or endovascular (9).
Fourth limitation
“Fourth, transcarotid-artery revascularization came into frequent use after approximately one half the patients had undergone randomization…”
Transcarotid-artery revascularization (TCARE) potentially elevates hemodynamic stroke risk, especially in diabetic patients with diminished cerebrovascular reserve, and warrants caution based on preoperative assessment.
Other shortcomings of this randomized multicenter study
Lack of publication regarding the causes of post-operative complications
The main purpose of randomized, multicenter study is to compare treatments, not to improve them. However, this is regrettable because the publication of the summary of complications analyzed in morbidity and mortality staff meetings would allow to understand certain errors and to improve treatments.
Lack of analysis of the causes of post-procedural strokes
Furthermore, this study lacks analysis of the cause of strokes occurring after the operative period. Differentiating between carotid strokes and cardioembolic strokes would have made this trial more precise and convincing.
Conclusion
CREST2 trials should never have been published in their current form, and even worse, with these conclusions.
The main interest of CREST2 is to show the lack of short-term effectiveness of medical treatment alone and the remarkable progress of angioplasty, but it does not mention the cost of treatments, which is important when two treatments are equivalent.
Stroke rate of interventional treatments could be reduced by a better understanding of the mechanism of perioperative complications.
Continued monitoring of patients, which is planned in CREST2, will allow for more refined results and likely to be more in line with reality.
Pending longer-term results, it is always necessary to combine medical treatment with an interventional technique, depending on the indications and contraindications of each of these two complementary techniques.
References
- Medical Management and Revascularization for Asymptomatic Carotid Stenosis. Brott TG, Howard G, Lal BK, et al; CREST-2 Investigators. N Engl J Med. 2025 Nov 21;394(3):219-231. doi: 10.1056/NEJMoa2508800. Epub 2025 Nov 21. PMID: 41269206
- Executive Committee for the Asymptomatic Carotid Atherosclerosis Study. Endarterectomy for asymptomatic carotid artery stenosis. JAMA. 1995 May 10;273(18):1421-8.
- Young B, Moore WS, Robertson JT, Toole JF, Ernst CB, Cohen SN, Broderick JP, Dempsey RJ, Hosking JD. An analysis of perioperative surgical mortality and morbidity in the Asymptomatic Carotid Atherosclerosis Study. ACAS Investigators. Asymptomatic Carotid Atherosclerosis Study. Stroke. 1996;27:2216–2224. doi: 10.1161/01.str.27.12.2216
- Barnett H, Taylor W, Eliaswiz M, Fox AJ, Ferguson GG, Haynes RB, Rankin RN, Clagett GP, Hachinski VC, Sackett DL, et al, for the North American Symptomatic Carotid Endarterectomy Trial Collaborators. Benefit of carotid endarterectomy in patients with symptomatic moderate or severe stenosis. N Engl J Med. 1998;339:1415–1425. doi: 10.1056/NEJM199811123392002
- Diabetes, stroke severity and hemodynamic cerebral ischemia. Petitjean C, Labreuche J. Cardiovasc Diabetol Endocrinol Rep. 2025 Dec 21;11(1):48. doi: 10.1186/s40842-025-00264-2. PMID: 41422063
- Hemodynamic Stroke: Emerging Concepts, Risk Estimation, and Treatment. Wegener S, Baron JC, Derdeyn CP, et al. Stroke. 2024 Jun;55(7):1940-1950. doi: 10.1161/STROKEAHA.123.044386. Epub 2024 Jun 12. PMID: 38864227
- Samra SK, Dy EA, Welch K, et al. Evaluation of a cerebral oximeter as a monitor of cerebral ischemia during carotid endarterectomy. Anesthesiology. 2000 ;93:964–70
- Teng L, Fang J, Zhang Y, Liu X, Qu C, Shen C. Perioperative baseline β-blockers: an independent protective factor for post-carotid endarterectomy hypertension.Vascular. 2021;29:270–279. doi: 10.1177/1708538120946538
- Abou-Chebl A, Reginelli J, Bajzer CT, Yadav JS. Intensive treatment of hypertension decreases the risk of hyperperfusion and intracerebral hemorrhage following carotid artery stenting. Catheter Cardiovasc Interv. 2007;69:690–696. doi: 10.1002/ccd.20693
