ORIGINAL RESEARCH article
Front. Neurol.
Sec. Stroke
Volume 16 - 2025 | doi: 10.3389/fneur.2025.1622586
This article is part of the Research TopicFutile recanalization after successful thrombectomy for acute ischemic stroke, including incomplete microvascular reperfusionView all 3 articles
Modified Small Vessel Disease Score as the Top Predictor of Stroke Outcome After Thrombectomy: A CT-Based Machine Learning Study
Provisionally accepted
- University of São Paulo, Ribeirão Preto, Ribeirão Preto, Brazil
Select one of your emails
You have multiple emails registered with Frontiers:
Notify me on publication
Please enter your email address:
If you already have an account, please login
You don't have a Frontiers account ? You can register here
Background: Mechanical thrombectomy (MT) improves outcomes in ischemic stroke (IS) due to large vessel occlusion (LVO), but ~50% of patients fail to achieve functional independence. Objectives: We investigated whether cerebral small vessel disease (cSVD), assessed by the modified Small Vessel Disease (mSVD) score and Brain Frailty Score (BFS), outperforms individual CT markers in predicting 90-day outcomes after MT. Design: Prospective cohort with retrospective analysis Methods: We included 351 patients with anterior circulation LVO treated with MT. Admission CT was used to score cSVD markers (leukoaraiosis, atrophy, lacunes) and compute mSVD and BFS. Eight logistic regression models and a Random Forest algorithm were used to predict poor outcome (modified Rankin Scale [mRS] 3–6). Model performance was evaluated using AUC-ROC and compared via DeLong tests. Results: Poor outcomes were associated with older age, higher NIHSS, systolic blood pressure, glycemia, and more severe leukoaraiosis and atrophy. Severe mSVD (score=3) independently predicted poor outcomes (OR=3.267;CI:1.731–6.168;p=0.009). mSVD outperformed BFS and individual CT markers (AUC=0.904 vs. 0.889/0.898; DeLong p<0.05) and ranked as the top predictor in Random Forest (importance=42.05). Treatment efficacy declined with increasing mSVD: NNT=6.44 and NNH=1.18 for mSVD=3 vs. NNT=1.12 and NNH=9.29 for mSVD=0. A secondary model incorporating 24h NIHSS and hemorrhagic transformation improved discrimination (AUC=0.954), but mSVD remained a key independent predictor. Conclusions: In this prospective study in a middle-income country, mSVD score was the strongest predictor of post-thrombectomy outcome, outperforming BFS and isolated imaging markers. While cSVD does not contraindicate MT, it reflects reduced cerebrovascular resilience. Integrating mSVD into baseline CT evaluation may enhance risk stratification and treatment guidance.
Keywords: Stroke, Ischemic stroke (IS), Thrombectomy, reperfusion therapies, Low and Middle Income Countries, machine learning, artificial intelligence, Cerebral small vessel disease
Received: 04 May 2025; Accepted: 13 Aug 2025.
Copyright: © 2025 Goulart, Martins-Filho, Camilo, Abud and Pontes-Neto. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
* Correspondence: Thiago Goulart, University of São Paulo, Ribeirão Preto, Ribeirão Preto, Brazil
Disclaimer: All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher.