Work

Gut microbiota-derived tryptophan metabolites and inflammatory factors enhance prediction of early neurological deterioration after intravenous thrombolysis in stroke patients: a comparative model study

Year: 2026

Type: article

Abstract: Background and purpose Early neurological deterioration (END) after stroke thrombolysis predicts poor outcomes. This study developed predictive nomograms that incorporate gut-derived tryptophan metabolites and inflammatory markers to predict END. Methods We evaluated demographic, clinical, and laboratory variables, including gut-derived tryptophan metabolites and inflammatory markers. A multivariable logistic regression model was developed to predict END and presented as a nomogram. Model performance was assessed using variance inflation factors, ROC analysis, calibration curves, and decision curve analysis. Net Reclassification Improvement and Integrated Discrimination Improvement were used to compare models. Results In this study, END occurred in 11.26% of stroke patients and was independently associated with systolic blood pressure (SBP), fasting blood glucose, and serum levels of indole-3-propionic acid (IPA), indole-3-lactic acid (ILA), and interferon γ (IFN γ). Two nomogram models were developed to predict END: one based on conventional risk factors (SBP and glucose) and another incorporating emerging biomarkers (SBP, glucose, ILA, IPA, and IFN γ). The models achieved AUC scores of 0.719 (95% CI 0.658–0.780) and 0.909 (95% CI 0.879–0.939), respectively. Comparing the models showed a categorical Net Reclassification Improvement of 0.601 ( p = 0.000) and an Integrated Discrimination Improvement of 0.197 ( p = 0.000). Conclusion Incorporating gut microbiota-derived tryptophan metabolites and inflammatory factors improves prediction of END among thrombolyzed stroke patients.

Source: Frontiers in Neurology

DOI: 10.3389/fneur.2026.1934189