ABSTRACT NUMBER: ESOC2026A962 ASSOCIATION OF DTI-ALPS-DERIVED GLYMPHATIC PROXIES WITH COGNITIVE AND CLINICAL OUTCOMES AFTER STROKE: A META-ANALYSIS

Abstract Background and aims Alterations in glymphatic-related processes have been associated with post-stroke pathophysiological changes, and may be related to subsequent cognitive impairment. Diffusion tensor imaging analysis along the perivascular space (DTI-ALPS) provides a noninvasive proxy of glymphatic-related processes. This meta-analysis aimed to quantify post-stroke changes in ALPS indices and to examine their associations with cognition and clinical severity. Methods PubMed, EMBASE, Web of Science, and CENTRAL were systematically searched for diffusion MRI studies assessing the ALPS index in patients with stroke. Studies reporting group comparisons with healthy controls or correlations between ALPS and outcomes were included. Random-effects models were used to pool standardized mean differences and correlation coefficients. Results Eleven studies including 592 patients with stroke and 648 healthy controls were analyzed. Compared with healthy controls, patients with stroke showed significantly lower ALPS indices (SMD = -1.41, 95% CI -1.83 to -0.98, p < 0.001). Ipsilateral regions of interests (ROIs) exhibited greater ALPS reductions than contralateral ROIs. Higher ALPS indices were associated with better global cognitive performance (r = 0.42, p < 0.001). Across stroke subtypes, higher ALPS indices were associated with reduced edema and hematoma volumes in intracerebral hemorrhage and lower National Institutes of Health Stroke Scale scores in ischemic stroke (all p < 0.05). Conclusions DTI-ALPS-derived indices are reduced after stroke and are associated with cognitive performance, lesion burden, and neurological severity. These findings support ALPS as a noninvasive imaging surrogate of glymphatic-related processes in stroke and suggest its potential relevance for phenotyping post-stroke brain injury and cognitive vulnerability. Conflict of interest Shuchang Zhong, Jingjing Lou, Jie Zhang: nothing to disclose