Neuroimaging characteristics associated with early neurological deterioration in single subcortical infarction patients

Abstract

Background and purpose Single subcortical infarctions (SSIs) are frequently complicated by early neurological deterioration (END), which adversely impacts long-term functional outcomes. However, the neuroimaging predictors of END in SSI patients remain insufficiently characterized. This study aimed to comprehensively evaluate the predictive value of multiple neuroimaging biomarkers for END in SSI patients. Methods We retrospectively enrolled consecutive patients diagnosed with acute SSIs in the middle cerebral artery (MCA) perforating territory. Neuroimaging assessments included apparent diffusion coefficient (ADC) and relative ADC (rADC) values, axial infarct location, MCA plaque status evaluated by whole-brain vessel wall imaging, and cerebral small vessel disease (CSVD) burden. END was defined as an increase of ≥2 points in the total NIHSS score or ≥1 point in the motor subscore within 72 h of admission. Results A total of 129 SSI patients were enrolled, of whom 32 (24.8%) developed END. Lower ADC and rADC values, larger maximum infarct diameter, more proximal axial infarct location, and the presence of MCA plaques were significantly associated with END on univariate analysis. rADC demonstrated superior predictive performance over ADC, with an optimal cut-off of 0.61 (sensitivity 84%, specificity 50%). On multivariable logistic regression, rADC < 0.61 (OR = 4.23, 95% CI: 1.58–11.61, P = 0.004) and MCA plaque presence (OR = 2.80, 95% CI: 1.14–6.89, P = 0.025) emerged as independent predictors of END. Neither CSVD burden nor any individual CSVD imaging marker was significantly associated with END. Conclusions rADC and MCA plaque are independent neuroimaging predictors of END in SSI patients, while CSVD burden showed no significant association. These findings support early magnetic resonance imaging-based risk stratification and individualized therapeutic decision-making in clinical practice.