Metabolomic analysis of arterial blood near intracranial occlusions reveals biochemical signatures of acute ischemic stroke.

BACKGROUND

Acute ischemic stroke is a major cause of death and disability, and timely assessment of symptom onset is essential for appropriate treatment. Conventional imaging often lacks accuracy in estimating the timing of ischemic injury, especially in patients with unclear or unwitnessed onset. Metabolomic profiling may provide biochemical information reflecting tissue damage, collateral circulation, and clinical progression.

METHODS

We performed an untargeted metabolomic study of arterial blood collected during mechanical thrombectomy from one hundred sixty-one patients with acute ischemic stroke and known onset time. Blood was drawn near the intracranial occlusion and analyzed with gas chromatography and mass spectrometry. Two hundred eleven consistent metabolic signals were identified. Multivariate models based on partial least squares discriminant analysis and least absolute shrinkage and selection operator regression were used to examine associations between metabolic patterns and clinical, radiological, and procedural features, including onset time, neurological worsening within forty-eight hours, and collateral flow.

RESULTS

Distinct metabolic signatures were associated with occlusion site, onset time, and clinical evolution. Elevated glucose and lactate were linked to early neurological worsening. Increased three-hydroxybutyric acid and niacinamide were associated with larger clot burden. Gamma-aminobutyric acid and branched-chain amino acids correlated with higher clinical severity and unfavorable outcome. Pathway analysis showed alterations in carbohydrate, amino acid, and fatty acid metabolism.

CONCLUSIONS

Arterial blood metabolomics reveals metabolic reprogramming associated with acute ischemic injury and compensatory responses. These findings support the potential use of metabolomic profiling for patient stratification and therapeutic decision-making, which requires validation in independent cohorts.