Predictive Value of Composite Inflammatory Markers for Stroke Prognosis: A Prospective Cohort Study

BACKGROUND: Novel composite inflammatory markers' role in stroke prognosis is understudied, and the best predictor is unclear, requiring further exploration. OBJECTIVES: This study aimed to systematically evaluate the associations of 6 novel composite inflammatory markers on stroke prognosis. METHODS: Based on the Third China National Stroke Registry (CNSR-III), we analyzed 14,476 participants aged ≥ 18 years with ischemic stroke or transient ischemic attack. Novel composite inflammatory markers were calculated, including neutrophil-to-lymphocyte ratio (NLR), platelet-to-lymphocyte ratio (PLR), monocyte-to-lymphocyte ratio (MLR), systemic immune-inflammation index (SII), systemic inflammatory response index (SIRI), and pan-immune inflammation value (PIV). Outcomes at 1 year included stroke recurrence (primary outcome) and ischemic stroke, composite vascular events, mortality, and poor functional outcome (mRS 2-6) as secondary outcomes. Multivariable Cox or logistic regression analyses were performed to estimate the association of composite inflammatory markers with these outcomes. XGBoost, IDI, and NRI were utilized to evaluate predictive performance. Additionally, subgroup analyses and sensitivity tests were conducted to assess the robustness of the findings. RESULTS: There were 1403 (9.7%), 1300 (9.0%), 1484 (10.3), 483 (3.3%) and 2865 (21.7%) patients with recurrent stroke, ischemic stroke, composite vascular events, mortality, and poor functional outcome within 1 year. SII, PIV, PLR, NLR, SIRI, and MLR were positively associated with stroke recurrence, with hazard ratios (95% CIs) of highest-quartile vs. lowest-quartile 1.39 (1.20, 1.61), 1.40 (1.21, 1.62), 1.16 (1.00, 1.34), 1.42 (1.22, 1.66), 1.38 (1.18, 1.61) and 1.18 (1.02, 1.37), respectively. A graded increase in stroke risk across quartiles was observed (Log-rank p < 0.05). All six markers were positively associated with ischemic stroke, composite vascular events, mortality, and poor functional outcome. SIRI demonstrated better predictive performance [IDI: 0.004 (0.002, 0.007); NRI: 0.082 (0.052, 0.107)] than the other five composite inflammatory markers for stroke recurrence. Among eight XGBoost models, the AUC increased from 0.598 (base) to 0.610 (adding NIHSS/mRS), and further to 0.624 with the addition of SIRI, which was the highest. CONCLUSIONS: Novel composite inflammatory markers were significantly associated with an increased risk of adverse outcomes in stroke patients. When accounting for the influence of inflammatory markers, SIRI demonstrated stronger predictive ability for stroke recurrence than the other markers.