Diffusion Tensor Imaging Along the Perivascular Space (DTI-ALPS) in Ischemic Stroke: A Systematic Review of Diagnostic and Prognostic Performance for Post-Stroke Cognitive Impairment.

Background/

Objectives

Post-stroke cognitive impairment (PSCI) affects ~40% of survivors. Diffusion Tensor Imaging Analysis Along the Perivascular Space (DTI-ALPS) is a fast, contrast-free surrogate of perivascular (glymphatic-aligned) diffusivity that may stratify PSCI risk. We systematically synthesized evidence on the diagnostic and prognostic performance of ALPS in ischemic stroke.

Methods

Following PRISMA 2020, we searched PubMed/MEDLINE, Scopus, and Web of Science from inception to August 2025 for human ischemic stroke studies reporting ALPS and cognitive or functional outcomes. Eligible designs were cohort or case-control. Outcomes included group differences, associations with cognition (Montreal Cognitive Assessment [MoCA]/Mini-Mental State Examination [MMSE]), prognostic accuracy for PSCI/functional outcome, and longitudinal change. Risk of bias was appraised with QUADAS-2 (diagnostic) and QUIPS (prognostic). The protocol was registered on OSF. Heterogeneity among studies precluded a meta-analysis.

Results

Five single-center cohorts (n per cohort 29-120) from Asia, Europe, and the USA used 3T DTI with ventricular-level ALPS ROIs. Across studies, ALPS was lower after stroke, with early ipsilesional depression and partial recovery over weeks to months. ALPS correlated with cognition (MoCA r ≈ 0.43-0.56) and discriminated early cognitive impairment (AUC 0.868; sensitivity 96%, specificity 66%). Follow-up ALPS predicted poor 6-month outcome (AUC 0.786). In lacunar stroke with small-vessel disease, higher baseline ALPS related to better cognitive trajectories and lower incident dementia risk (HR ≈ 0.33), though associations attenuated after adjustment for diffusion microstructural covariates (PSMD/MD). Reporting of acquisition parameters and ROI methods varied; overall risk of bias was moderate.

Conclusions

DTI-ALPS shows consistent post-stroke reductions, recovery-sensitive trajectories, and promising-though context-dependent-prognostic value for PSCI and longer-term outcomes. Clinical translation will require standardized acquisition/analysis, multimodal adjustment, prespecified cut-offs, and prospective multicenter validation.