Evaluation of the systemic immune-inflammation index in the diagnosis of Trousseau syndrome initially presenting with acute ischemic stroke

Abstract Background Diagnosing Trousseau syndrome (TS) in its early stages presents significant difficulties owing to the lack of distinctive symptoms. Emerging research suggests that immune-mediated inflammatory processes play a crucial role in TS development. This investigation explores the evaluation of the utility of the systemic immune-inflammation index (SII) in the early diagnosis of TS with acute ischemic stroke (AIS) as its initial manifestation. Methods This study employed a retrospective design to examine 30 patients treated at Affiliated Hospital of Jiangsu University from 2018 to 2025. Clinical records were reviewed to obtain demographic information, National Institutes of Health Stroke Scale (NIHSS), hematological parameters: complete blood count, C-reactive protein (CRP), coagulation profiles, and oncological biomarkers. The SII was derived through computational analysis using the formula: (neutrophil count × platelet count)/lymphocyte count. Participants were stratified into two distinct cohorts based on the temporal relationship between AIS onset and cancer diagnosis: the AIS-TS group (n = 14) comprised patients whose AIS represented the initial manifestation of TS, while the NAIS-TS group (n = 16) included individuals who developed AIS subsequent to their cancer diagnosis. Logistic regression analysis was implemented to investigate potential associations between SII values and AIS-TS classification. Additionally, the relationship between SII and D-dimer index (DDI) concentrations was examined through bivariate correlation analysis. Correlation analyses employed Pearson’s or Spearman’s correlation coefficient, selected based on the normality of variable distributions. To assess the diagnostic accuracy of AIS-TS, receiver operating characteristic (ROC) curve analysis was employed as the evaluation approach. Results In the logistic regression model, a total of 5 variables with P