Association and Risk-Stratification Value of the Positive Remodeling Index of Middle Cerebral Artery Atherosclerotic...

Atherosclerosis of the middle cerebral artery (MCA) is a major contributor to ischemic stroke, and infarcts confined to the perforator territory can produce considerable neurological dysfunction even when the affected volume is comparatively small. As the severity of luminal narrowing measured by conventional methods does not, on its own, account for the variety of observed infarction patterns, additional contributions from plaque remodeling and plaque vulnerability are likely. High-resolution magnetic resonance imaging (HR-MRI) permits fine-grained characterization of intracranial plaques. The present work was undertaken to examine how the positive remodeling index (PRI) of MCA atherosclerotic plaques relates to perforator-territory infarction and to provide a preliminary appraisal of the value of PRI for stratifying the risk of this infarct subtype.

Patients admitted to our institution between January 2021 and December 2024 with acute ischemic stroke attributable to MCA atherosclerosis were reviewed retrospectively, yielding 283 cases. Each patient completed both conventional MRI and HR-MRI vessel wall imaging. Based on where the infarct was located, patients were separated into a perforator-territory infarction group (n = 127) and a non-perforator-territory infarction group (n = 156). The cross-sectional vessel area at the culprit lesion and at the reference segment were quantified from HR-MRI, from which the PRI was derived. Clinical and imaging variables showing an independent relationship with perforator-territory infarction were identified through univariate followed by multivariable logistic regression. The apparent discrimination of PRI was summarized using receiver operating characteristic (ROC) analysis, and the stability of the derived cutoff was examined by bootstrap resampling together with stratified 10-fold cross-validation. The added value of the PRI was judged using net reclassification improvement (NRI), integrated discrimination improvement (IDI), and decision curve analysis (DCA).

PRI values were greater among patients with perforator-territory infarction than among those without. In the multivariable model, perforator-territory infarction was independently linked to the PRI (odds ratio [OR] = 8.67, 95% confidence interval [CI]: 3.42-21.98, p < 0.001), diabetes mellitus (OR = 2.18, 95% CI: 1.21-3.92, p = 0.009), homocysteine (OR = 1.03, 95% CI: 1.00-1.05, p = 0.032), severe stenosis (OR = 2.45, 95% CI: 1.38-4.35, p = 0.002), marked enhancement (OR = 2.32, 95% CI: 1.28-4.21, p = 0.006), and intraplaque hemorrhage (OR = 1.95, 95% CI: 1.09-3.49, p = 0.025). Used alone, the PRI achieved moderate discrimination, with an area under the curve (AUC) of 0.728 (95% CI: 0.670-0.786). After internal validation, the bootstrap-corrected AUC was 0.716 (95% CI: 0.654-0.775) and the stratified 10-fold cross-validated AUC was 0.714 (95% CI: 0.650-0.777). A Youden-based cutoff of PRI = 1.12 was identified post hoc from the ROC curve, and bootstrap resampling located this threshold predominantly within the 1.10-1.14 interval. Incorporating the PRI into the base model raised the AUC from 0.794 to 0.832 (p = 0.018), accompanied by a continuous NRI of 0.318 (95% CI: 0.154-0.482) and an IDI of 0.061 (95% CI: 0.028-0.094) (both p < 0.001).

Among patients with MCA atherosclerotic stroke, a higher PRI was independently related to perforator-territory infarction. Within this single-center retrospective cohort, the PRI demonstrated moderate discriminative capacity and may represent a candidate HR-MRI vessel wall imaging marker for risk stratification. Confirmation of its threshold and its place in clinical practice will require multicenter prospective investigation and validation in independent samples.