Dayong Shen, Siyan Chen, Yuting Zhang, Aparna Sharma Chivukula, Shasha Zhu, Chao Zhao, Zeheng Li, Jie Sun, Wei Zhang, Lu Yu
Admission inflammatory-metabolic biomarkers showed complementary, outcome-specific associations after EVT. CTI was most consistently associated with END and unfavorable 90-day functional outcome, whereas cell-derived inflammatory indices were more strongly associated with HT. Combining CTI and NLR may refine early risk stratification after EVT, although prospective external validation is required before clinical application.
BACKGROUND: Clinical outcomes after endovascular therapy (EVT) for acute ischemic stroke (AIS) remain heterogeneous despite advances in reperfusion care. Whether admission inflammatory-metabolic biomarker profiles provide complementary and outcome-specific prognostic information beyond established clinical and imaging variables remains uncertain.
METHODS: This multicenter retrospective cohort included consecutive adults with anterior-circulation large-vessel occlusion AIS treated with EVT at three centers between 2020 and 2025. The primary outcome was unfavorable 90-day functional outcome (modified Rankin Scale score 3-6); secondary outcomes were early neurological deterioration (END) and hemorrhagic transformation (HT). Multivariable logistic regression, restricted cubic splines, interaction analyses, and internally validated prediction models were used.
RESULTS: Among 740 patients, 424 (57.3%) had an unfavorable 90-day functional outcome. Per 1-SD increase, the C-reactive protein-triglyceride glucose index (CTI) was independently associated with unfavorable outcome (adjusted odds ratio [aOR] 2.07, 95% CI 1.65-2.61) and END (aOR 1.74, 95% CI 1.46-2.08), but not with HT. Several cell-derived inflammatory indices were independently associated with HT, including the neutrophil-to-lymphocyte ratio (NLR; aOR 1.47, 95% CI 1.25-1.74) and systemic immune-inflammation index (SII; aOR 1.35, 95% CI 1.15-1.58). The association between CTI and unfavorable outcome was stronger in patients with poor collateral circulation (P for interaction = 0.011). Adding CTI and NLR increased the baseline model AUC from 0.792 to 0.836 (ΔAUC 0.044, 95% CI 0.025-0.065; cross-validated AUC 0.818).
CONCLUSION: Admission inflammatory-metabolic biomarkers showed complementary, outcome-specific associations after EVT. CTI was most consistently associated with END and unfavorable 90-day functional outcome, whereas cell-derived inflammatory indices were more strongly associated with HT. Combining CTI and NLR may refine early risk stratification after EVT, although prospective external validation is required before clinical application.