Plasma amino acid profiles show significant between-group differences during acute cerebrovascular events, and although individual amino acids have limited standalone diagnostic utility (AUC approximately 0.61-0.65), their integration with clinical variables increased the AUC from 0.734 to 0.803, providing complementary information for early diagnostic differentiation.
Key Findings
Results
Significant between-group differences in plasma amino acid concentrations were observed across acute cerebrovascular disease subtypes, particularly during the hyperacute phase at 3-4.5 hours after symptom onset.
Study included 297 participants: 122 patients with acute ischemic stroke (AIS), 50 with hemorrhagic stroke (HS), 50 with transient ischemic attack (TIA), and 75 control participants.
Six plasma amino acids were measured using high-performance liquid chromatography coupled with mass spectrometry (HPLC-MS).
Patients were sampled at two time points: 3-4.5 hours and 12 hours after symptom onset.
Control participants provided a baseline sample only, without repeated sampling.
Results
Individual amino acids demonstrated limited discriminative ability when assessed in isolation as diagnostic biomarkers.
AUC values for individual amino acids were approximately 0.61-0.65.
This limited performance was observed across the different acute cerebrovascular disease subtypes.
Statistical analyses included ROC curve analysis and multivariable logistic regression to evaluate discriminative ability.
The findings indicate that no single amino acid was sufficient as a standalone biomarker for differentiating acute cerebrovascular disease subtypes.
Results
Adding metabolomic variables to the clinical model significantly increased diagnostic accuracy for differentiating acute cerebrovascular disease subtypes.
The AUC increased from 0.734 (clinical model alone) to 0.803 (combined clinical and metabolomic model).
This improvement was statistically significant (DeLong p = 0.007).
The combined model approach suggests that amino acid profiles provide complementary rather than standalone diagnostic information.
Multivariable logistic regression was used to build and compare the clinical and combined models.
Results
Analysis of temporal changes in amino acid concentrations between the two sampling time points (3-4.5 hours and 12 hours) identified subtype-related metabolic patterns, particularly for differentiating AIS from TIA.
Dynamic changes in amino acid profiles between the hyperacute (3-4.5 h) and early acute (12 h) phases revealed metabolic patterns associated with stroke subtype.
The differentiation of acute ischemic stroke (AIS) from transient ischemic attack (TIA) showed particularly notable subtype-related metabolic patterns in temporal analysis.
This finding suggests that single time-point measurements may not capture the full diagnostic information available from amino acid profiling.
Distinguishing AIS from TIA is a recognized clinical challenge, especially in the hyperacute phase.
Discussion
Plasma amino acid profiles are associated with systemic metabolic alterations during acute cerebrovascular events, reflecting pathobiochemical changes across disease subtypes.
The study used a prospective observational clinical and laboratory design.
Six amino acids were selected for quantification via HPLC-MS, a high-sensitivity analytical method.
The study interpreted amino acid concentration differences as reflecting underlying pathobiochemical mechanisms rather than purely diagnostic signals.
Non-parametric tests and correlation analysis were used alongside ROC and regression analyses, appropriate for the distributional properties of metabolomic data.
What This Means
This research suggests that measuring specific amino acids in blood plasma may help doctors distinguish between different types of sudden brain events — including ischemic stroke (caused by a blood clot), hemorrhagic stroke (caused by bleeding), and transient ischemic attack (a temporary 'mini-stroke') — particularly in the critical first few hours after symptoms begin. The study measured six amino acids in blood samples taken from 297 people (including stroke patients and healthy controls) at approximately 3-4.5 hours and again at 12 hours after symptom onset, using a highly sensitive laboratory technique called HPLC-MS.
The findings indicate that while no single amino acid was accurate enough on its own to reliably tell these conditions apart (with diagnostic accuracy scores around 0.61-0.65 on a 0-1 scale), combining amino acid measurements with standard clinical information substantially improved diagnostic accuracy, raising the combined score from 0.734 to 0.803. Tracking how amino acid levels changed over time — rather than measuring them at just one moment — also revealed additional patterns that were useful for distinguishing ischemic stroke from transient ischemic attacks, which can be particularly difficult to tell apart in the early stages.
This research suggests that blood-based amino acid profiling could serve as a useful complement to existing clinical tools for the early diagnosis of acute brain events, rather than replacing them. Earlier and more accurate differentiation between stroke types matters because treatments differ significantly — for example, clot-busting therapy is used for ischemic stroke but would be harmful in hemorrhagic stroke. Further validation in larger studies would be needed before these findings could influence clinical practice.
Battakova S, Grigolashvili M, Klyuyev D, Kasatkin D, Shayakhmetova Y, Beisembayeva M, et al.. (2026). Disruption of amino acid metabolism in acute cerebrovascular diseases: diagnostic and pathobiochemical perspectives.. Frontiers in neurology. https://doi.org/10.3389/fneur.2026.1876777