Predictive value of systemic immune-inflammatory markers for lung cancer-associated cerebral infarction: a
Dewei Zhu1, Haoliang Wang1, Zhixin Yan1
1Department of Neurology, First Affiliated Hospital of Henan Medical University (Henan Medical Key Laboratory of Neurology) (Henan Joint International Laboratory of Neurorestoratology for Senile Dementia) (Henan Key Laboratory of Neurorestoratology and Protein Modification) Xinxiang 453100, Henan, China.
Abstract:
To clarify the prognostic role of immune inflammatory biomarkers in lung cancer-related cerebral infarction (CI). This retrospective cohort study enrolled 152 lung cancer patients, and they were divided into a lung cancer with CI group (n=48) and a lung cancer without CI group (n=104); an additional 52 patients without lung cancer were included as the simple CI group. Observed endpoints included neutrophil (NEUT), lymphocyte (LYMPH), platelet (PLT), red cell distribution width (RDW), C-reactive protein (CRP), procalcitonin (PCT), systemic inflammatory index (SII), neutrophil/lymphocyte ratio (NLR), platelet-to-lymphocyte ratio (PLR), and Essen stroke risk score (ESRS). Cox regression model was used to screen independent risk factors for lung cancer-associated CI, and receiver operating characteristic (ROC) curves with area under the curve (AUC) were used to assess the predictive value of biomarkers. The incidence of CI in 152 lung cancer patients was 31.6%. Compared with lung cancer patients without CI, those with CI were older, had a higher proportion of stage IV disease, and showed elevated levels of ESRS, NEUT, PLT, RDW, CRP, PCT, SII, NLR, and PLR, while LYMPH levels were decreased (all P<0.001). Age ≥65 years, stage IV disease, SII≥1250×109/L, NLR≥7, and CRP≥10 mg/L were independent risk factors (all P<0.001). The SII-NLR-CRP combined model had the highest AUC (0.918), with 89.4% sensitivity and 90.4% specificity in predicting lung cancer-related CI. Systemic immune inflammatory markers have significant predictive value for lung cancer-related CI. The SII-NLR-CRP combined model exhibits excellent predictive efficacy for this condition, which is superior to individual biomarkers and provides a reliable reference for clinical evaluation.
