Could Revascularization of Carotid Stenosis Reverse Cognitive Deficits?

NEW YORK, NY—Whether intervening on asymptomatic, high-grade carotid artery stenosis will provide cognitive benefits remains an unanswered question, one the CREST-2 investigators hope to provide some insight into with the add-on CREST-H study.

Brajesh Lal, MD (University of Maryland School of Medicine, Baltimore), a co-principal investigator (PI) of CREST-2, discussed the issue here at the VEITHsymposium 2019 last week.

Prior cohort studies, including ACCOF-1 and InCHIANTI, have shown that asymptomatic carotid stenosis is associated with cognitive deficits, as well as mobility impairments and increased fall risk. In addition, an analysis of the first 200 patients enrolled in CREST-2, showed evidence of cognitive impairment compared with 30,000 participants in the REGARDS study, a national cohort study exploring stroke risk factors in adults 45 years and older. CREST-2, meanwhile, consists of two parallel trials comparing carotid stenting and endarterectomy separately with intensive medical therapy in patients with asymptomatic, high-grade carotid stenosis.

“Regardless of whether we looked at a composite cognitive function [endpoint] or looked at individual cognitive function domains, stenosis patients performed far worse than those without stenosis,” Lal said.

The possibility of reversing cognitive and mobility impairments in patients with asymptomatic carotid stenosis ≥ 70% has been tested in preliminary studies. In a small, nonrandomized, prospective study of 46 patients, which was published in 2011, cognitive function improved 6 months after revascularization with either stenting or carotid endarterectomy. That finding was replicated in 14 subsequent studies, Lal said. And in an interim analysis of 42 patients in the ongoing single-center, nonrandomized, prospective ACCOF-2 study, there were trends toward improvement in several measures of mobility a year after revascularization, with statistical significance achieved for dynamic gait index, gait speed, and the MiniBEST score (an assessment of balance).

Additional evidence to inform this question may be forthcoming from CREST-2, which has cognitive function as a secondary endpoint. Moreover, the investigators are conducting an add-on study, CREST-H, to see whether cognitive impairment related to MRI-detected hemodynamic hypoperfusion is reversible with carotid revascularization. The plan is to enroll about 350 patients, Lal said.

“Brain hypoperfusion results in cognitive and mobility impairment. Carotid stenosis, we now know, is associated with brain hypoperfusion. And there are increasingly stable results demonstrating that stenosis in the carotid artery is associated with cognitive and mobility impairment,” Lal explained, suggesting that how “asymptomatic” is defined in the context of carotid stenosis should be reconsidered based on the evidence of cognitive impairment in this setting. “There is a strong rationale to include these endpoints in the [CREST-2] trial.”

No Impact on Long-term Dementia Risk

In a separate presentation, Alison Halliday, MD (University of Oxford, England), reported 15-year data on dementia risk from a subset of patients participating in the ACST-1 trial, which randomized patients with asymptomatic carotid stenosis to immediate endarterectomy or indefinite deferral of any carotid procedure. Previously reported 10-year results showed that successful surgery reduced stroke risk over that span.

Carotid stenosis is associated with cognitive aging in addition to an elevated stroke risk, Halliday explained, “so might prophylactic interventions such as [carotid endarterectomy] prevent cognitive aging and hence dementia?”

To address that question, her team looked at data from 1,600 trial participants enrolled in the United Kingdom and Sweden because those countries have “good” dementia information. At 15 years, 13.3% had developed some form of dementia. That rate was highest in patients who were at least 75 years old at study entry (44.5%), intermediate in those ages 65 to 74 (21.4%), and lowest in those younger than 65 (5.4%).

There was no clear impact of immediate versus deferred endarterectomy on overall dementia risk (18.7% vs 19.9%; RR 0.94; 95% CI 0.72-1.24), although subgroup analyses suggested that surgery may have been successful at lowering risk in patients with diabetes and in those without a prior brain infarct.

“These interventions from midlife onwards, they might prevent several years of cognitive aging for some of the patients, so this lack of overall effect does not preclude protection from carotid intervention,” Halliday said.

Some Skepticism

During a panel discussion, Lal was asked about the discrepancy between the promising preliminary results he reported and two meta-analyses showing that carotid intervention does not prevent cognitive decline.

“The answer lies in the fact that we’ve improved our ability to define and quantify over the years cognitive impairment using in-person testing. So that’s one aspect of it,” Lal responded.

“The second aspect of it is that all asymptomatic patients are not created equal and that’s why [there is] the focus on brain-perfusion assessment as a potential tool to separate out those patients that have flow-related cognitive impairment,” he continued. “Those are the ones that are potentially going to be benefitted, not those that have dementia associated with [multiple] infarcts or potentially with intracranial atherosclerosis, just as two examples, or frankly Alzheimer’s disease. It’s a mixed bag. So when we look at all patients with impairment to try to improve them, that signal is going to get lost as to which patients will benefit.”

The idea that cognitive impairment associated with carotid stenosis can be mitigated by revascularization was met with some skepticism, however.

Panelist J. David Spence, MD (Robarts Research Institute, London, Canada), noted that a minority of patients with asymptomatic stenosis have reduced cerebrovascular blood flow. “So likening asymptomatic stenosis to reduced flow is nonsense. In the average patient with asymptomatic stenosis, it is very clear that intervention will increase the risk of dementia because the highest predictor of cognitive decline is stroke, and stenting and endarterectomy cause more strokes than intensive medical therapy.”

He said that there’s a subgroup of patients with reduced blood flow who might derive a cognitive benefit from revascularization, but that they already have a guideline-recommended indication for a procedure. “Furthermore, learning effects and a placebo effect have such huge effects on cognitive testing that the only way to clearly show this is going to be a . . . trial with a sham control.”

Thomas Brott, MD (Mayo Clinic, Jacksonville, FL), one of the PIs of the CREST-2 trial, replied: “My eminent colleague has very persuasive arguments, but we want evidence-based medicine, not eminence-based medicine. The CREST-2 trial has a secondary endpoint, cognitive function, and we have randomized them. Learning effects are important and we’re doing our best. We’re not going to be able to have a sham trial in this setting, but with the CREST results we should be able to answer . . .  your questions.”

Later in the day, Anne Abbott, MBBS, PhD (Monash University, Melbourne, Australia), discussed the merits of CREST-H, pointing out that once the patients are divided based on whether they have reduced blood flow and were randomized to an intervention, the comparisons will involve small patient numbers.

“So it’s likely that the H aspect of CREST-2 will not show a procedural indication with respect to cognition because the study seems underpowered; stenting and endarterectomy are being combined as one intervention, which is inappropriate; the validity of MRI perfusion for hemodynamic impairment is not explained; and the clinical significance of 0.6 for standard-deviation change in a battery of cognitive tests [the primary outcome] is unclear.”