Reply to: After Carotid Artery Stenting, Diffusion Lesions Need a Lesion-Burden Estimand

Article information

Neurointervention. 2026;.neuroint.2026.00864

Publication date (electronic) : 2026 August 14

doi : https://doi.org/10.5469/neuroint.2026.00864

Kwon-Duk Seo, MD, PhD¹, Taemin Kim, MD², Minsoo Sung, MD³, Sang Hyun Suh, MD, PhD^,1

¹Department of Radiology, Gangnam Severance Hospital, Yonsei University College of Medicine, Seoul, Korea

²Department of Neurology, Danwon Hospital, Ansan, Korea

³Department of Neurology, Gangnam Severance Hospital, Yonsei University College of Medicine, Seoul, Korea

Correspondence to: Sang Hyun Suh, MD, PhD Department of Radiology, Gangnam Severance Hospital, Yonsei University College of Medicine, 211 Eonju-ro, Gangnam-gu, Seoul 06273, Korea E-mail: suhsh11@yuhs.ac

Received 2026 July 21; Accepted 2026 August 8.

Dear Editor,

We thank the correspondent for the thoughtful and constructive commentary on our study of early diffusion-weighted imaging (DWI) lesions after carotid artery stenting (CAS) in symptomatic patients [1]. We share the central premise of the Letter: dichotomizing periprocedural ischemic injury into the presence or absence of “any new DWI lesion” may not capture its full procedural or prognostic significance. We would like to clarify the scope of our analysis and the methodological requirements for a more granular assessment.

Our study addressed a deliberately narrow question: whether the presence of any new DWI lesion within 24 hours after CAS was associated with functional outcome at 1 year. Binary DWI positivity, the primary imaging endpoint of our study, was not independently associated with 1-year functional outcome. However, the confidence interval was wide and compatible with a clinically meaningful effect in either direction. Moreover, the 1-year modified Rankin Scale (mRS) was chosen to reflect sustained functional status after the acute recovery and rehabilitation phases, but it is strongly influenced by presenting neurological severity, recovery trajectory, and intercurrent events and may be relatively insensitive to limited periprocedural ischemic injury. Our findings should therefore be interpreted as the absence of a detectable association for a binary endpoint in this cohort, not as evidence that all post-CAS DWI lesions are benign or that periprocedural embolization has no other clinically relevant consequences. The association between atrial fibrillation and early DWI lesions should likewise be interpreted cautiously. Only 10 patients had atrial fibrillation, resulting in a highly imprecise estimate; this finding should therefore be regarded as hypothesis-generating rather than causal. Larger studies with detailed antithrombotic and lesion-distribution data will be required to clarify this association.

Some of the procedural and clinical elements proposed in the Letter were available in our cohort. Distal embolic protection was used uniformly, whereas proximal balloon occlusion was not used, precluding comparison across protection strategies. Early neurological changes were also documented: 4 patients experienced mild neurological deterioration of 1–2 points on the National Institutes of Health Stroke Scale, and all events were attributable to lesions within the vascular territory of the treated artery. Preprocedural antiplatelet status, intravenous thrombolysis, and balloon angioplasty were likewise recorded and were not significantly associated with early DWI lesions, although these estimates were imprecise because of the limited subgroup sizes.

We agree that lesion count, volume, vascular-territory distribution, and eloquent involvement may provide information beyond binary positivity. However, a reliable post hoc granular analysis was not feasible in this dataset. As acknowledged in the original report, preprocedural DWI was not uniformly repeated immediately before CAS during the study period from 2005 to 2023. Consequently, some lesions identified after the procedure may have arisen in the interval between the index cerebrovascular event and revascularization [1]. Acquisition parameters were not harmonized for quantitative lesion measurement, and DWI was interpreted by a single blinded reader without prespecified segmentation criteria or a second reader for reliability assessment. Retrospective quantification under these conditions would have been vulnerable to both measurement and attribution bias, and quantitative lesion-burden estimates derived from heterogeneous imaging could give a false impression of precision.

A further challenge is that the optimal lesion-burden estimand has not yet been established. Candidate measures include lesion count, total lesion volume, maximum lesion size, territorial distribution, eloquent involvement, and composite phenotypes, and different measures may classify the same patients differently. Previous studies have linked post- CAS DWI lesions to subsequent cerebrovascular risk, embolic signals, and plaque features, but have used heterogeneous definitions and imaging intervals [2-5]. Future studies should therefore prospectively prespecify a primary lesion-burden measure within a standardized imaging protocol and validate it against early neurological change, longer-term functional status, and recurrent events. Cognitive outcomes may be particularly informative, since the mRS captures global disability and may not detect subtle consequences of small, multiple, or strategically located ischemic lesions. Existing studies of post-CAS DWI lesions and cognition are small and have yielded mixed findings, and few have incorporated both pre- and postprocedural DWI with cognitive testing [2,6,7]. Cognition therefore remains an uncertain, though complementary, outcome to the mRS rather than an alternative to it.

In conclusion, our findings support the narrower interpretation that binary early DWI positivity was not independently associated with 1-year functional outcome in this cohort. We agree that prospective studies using standardized imaging and prespecified lesion-burden measures are needed to determine whether specific patterns of post-CAS ischemic injury have distinct clinical implications.