The systemic immune-inflammation index as a superior predictor of functional outcome following mechanical thrombectomy for acute ischemic stroke: a retrospective cohort study

TL;DR

This study shows that the systemic immune-inflammation index (SII) better predicts recovery after stroke treatment with mechanical thrombectomy than other blood markers.

Contribution

The study introduces and validates a new SII-based model for predicting stroke outcomes after mechanical thrombectomy, outperforming traditional biomarkers.

Findings

SII alone had higher predictive accuracy (AUC: 0.834) than PLR or NLR.

The optimal model (baseline + SII) achieved an AUC of 0.863, significantly better than the baseline model.

SHAP analysis confirmed SII as the most influential variable in outcome prediction.

Abstract

Despite high recanalization rates with mechanical thrombectomy (MT) for acute ischemic stroke (AIS), functional outcomes remain variable. Systemic inflammation is a key driver of secondary brain injury post-reperfusion. The systemic immune-inflammation index (SII), calculated as (platelet count × neutrophil count)/lymphocyte count, integrates multiple inflammatory pathways and has shown prognostic value in cardiovascular diseases and stroke treated with intravenous thrombolysis. However, its role in predicting outcomes specifically for AIS patients undergoing MT remains underexplored. This study aimed to develop and validate an SII-based model for predicting 90-day functional outcomes after MT and to compare its performance with traditional inflammatory biomarkers, namely neutrophil-to-lymphocyte ratio (NLR) and platelet-to-lymphocyte ratio (PLR).

We retrospectively analyzed data from 387 AIS patients treated with MT. The cohort had a median age of 68 years [interquartile range (IQR): 59–75], 67.2% were male, and the median time from stroke onset to thrombectomy was 340 min (IQR: 242.5–465.5). Inflammatory markers were measured at admission, such as SII, platelet lymphocyte ratio (PLR), neutrophil lymphocyte ratio (NLR) and 90-day modified Rankin Scale (mRS) scores. Patients were divided into good (90-day mRS ≤ 2; n = 151) and poor (mRS > 2; n = 236) outcome groups. We constructed and compared four logistic regression models: clinical baseline, baseline + SII, baseline + PLR, and baseline + NLR. Model performance was assessed using the area under the receiver operating characteristic curve (AUC), net reclassification improvement (NRI), integrated discrimination improvement (IDI), calibration, and decision curve analysis (DCA).

SII alone showed higher predictive accuracy (AUC: 0.834) than PLR or NLR. The optimal model (baseline + SII) achieved an AUC of 0.863, significantly improving outcome prediction over the baseline model (AUC: 0.655). Shapley Additive exPlanations (SHAP) analysis confirmed SII as the most influential variable (74.2% contribution). The model demonstrated good calibration and clinical utility across a range of probability thresholds.

A model incorporating the SII provides superior accuracy for predicting 90-day functional outcome after MT compared to models using NLR or PLR. As an easily obtainable composite biomarker, SII enhances risk stratification and could aid early clinical decision-making for AIS patients undergoing endovascular therapy.

Figures