We read with great interest the study by Seo et al. [1] evaluating early diffusion-weighted imaging (DWI) lesions and clinical outcomes after carotid artery stenting (CAS) in symptomatic carotid stenosis. The authors studied 190 patients and found new DWI lesions within 24 hours in 52 patients (27.4%). Atrial fibrillation was independently associated with early DWI lesions, adding a patient-level signal to the imaging endpoint. They also reported low rates of periprocedural ischemic stroke with neurological deterioration (2.1%) and symptomatic intracranial hemorrhage (1.1%), with good 1-year functional outcome in 82.1% of patients. The finding that new DWI lesions were not independently associated with 1-year modified Rankin Scale (mRS) outcome, while baseline neurological severity was the primary determinant, is clinically important and plausible.

We suggest that this result should not be interpreted as evidence that post-CAS DWI lesions are clinically unimportant. Rather, it highlights a limitation of the binary endpoint “any new DWI lesion.” That category may combine a single punctate clinically silent lesion, multiple lesions, small non-eloquent lesions, eloquent-territory lesions, and ipsilateral, contralateral, or multi-territory patterns. These patterns are unlikely to carry the same procedural, mechanistic, or prognostic meaning. Therefore, the most transferable message of the study may be that post-CAS DWI should be reported as a procedural-safety phenotype, not reduced to a blunt binary endpoint.

Prior carotid revascularization literature supports this caution rather than contradicting Seo et al. [1]. Silent DWI infarcts have been considered candidate surrogate procedural-stroke outcomes [2], and post-CAS ischemic lesions have been related in selected studies to embolic signals, plaque characteristics, and subsequent cerebrovascular risk [3-6]. These observations make Seo et al. [1]’s result pattern especially instructive: DWI positivity may be frequent, symptomatic events may remain uncommon, and 1-year mRS may still be driven mainly by baseline neurological severity. The issue, therefore, is not whether DWI is meaningful, but whether binary DWI positivity is sufficiently specific for the clinical question being asked.

The atrial fibrillation signal in Seo et al. [1] is particularly relevant. Atrial fibrillation was independently associated with early DWI lesions, but this should be described as an association rather than a causal mechanism. It may identify a higher-risk embolic substrate or a more complex antithrombotic context. Future CAS studies could help distinguish, at least descriptively, procedure-related embolization from patient-level embolic vulnerability by reporting atrial fibrillation status, antithrombotic regimen, periprocedural medication changes, protection strategy, and lesion distribution.

A more granular DWI framework would ask not merely whether any lesion occurred, but what kind of ischemic injury occurred, when, in which vascular territory, and with what early neurological correlate. At minimum, future reports should distinguish lesion burden, vascular-territory pattern, eloquent or symptomatic involvement, protection strategy, and early neurological change. Such reporting would align the imaging endpoint with the actual safety question after CAS.

Finally, the outcome window matters. One-year mRS may be dominated by presenting neurological severity, recovery trajectory, comorbidity, and recurrent events, and may therefore be insensitive to some periprocedural ischemic injuries. Lack of an independent association with 1-year mRS in this retrospective cohort does not establish that DWI lesions are benign, particularly when lesions are multiple, eloquent, or symptomatic. Seo et al. [1] should be commended for connecting early imaging findings with longer-term clinical outcomes. Their results invite the field to refine DWI measurement after CAS so that imaging positivity becomes a clinically interpretable procedural-safety phenotype rather than a binary label.