Distinct trajectory patterns of neutrophil-to-albumin ratio predict clinical outcomes after endovascular therapy in large vessel occlusion stroke

Abstract

Objective: To investigate the dynamic characteristics and prognostic value of neutrophil-to-albumin ratio (NAR) in patients with acute large vessel occlusion ischemic stroke (LVO-AIS) undergoing endovascular therapy (EVT).

Methods: In this retrospective cohort study, we consecutively enrolled 299 patients with anterior circulation LVO-AIS who underwent EVT between January 2018 and February 2024. NAR was measured at admission, day 1, and day 3 after EVT. The primary outcome was poor functional outcome at 90 days (modified Rankin Scale score 3-6). Secondary outcomes included symptomatic intracranial hemorrhage (sICH), malignant cerebral edema (MCE), and in-hospital mortality (IHM). Multivariable logistic regression and restricted cubic spline regression were used to analyze the association between NAR and functional outcomes. Latent class trajectory modeling (LCTM) was applied to identify NAR evolution patterns, and propensity score matching (PSM) was performed to balance baseline characteristics between different trajectory groups, followed by conditional logistic regression to assess their association with clinical outcomes.

Results: At 90-day follow-up, 197 patients (65.9%) had poor outcomes. The predictive value of NAR increased over time, with day 3 NAR showing the best predictive performance (poor outcome: AUC = 0.79; sICH: AUC = 0.70; MCE: AUC = 0.75; IHM: AUC = 0.81). Multivariable analysis showed that for each unit increase in day 3 NAR, the risk of 90-day poor outcome increased 2.81-fold (95% CI: 1.96-4.03, p < 0.001). LCTM analysis identified two distinct NAR evolution patterns: continuously increasing (31.1%) and peak-then-decline (68.7%). After PSM (63 patients per group), compared with the continuously increasing trajectory, the peak-then-decline trajectory was associated with significantly lower risks of poor functional outcome (OR = 0.38, 95% CI: 0.17-0.86, p = 0.020), sICH (OR = 0.38, 95% CI: 0.17-0.86, p = 0.020), MCE (OR = 0.25, 95% CI: 0.10-0.61, p = 0.002), and IHM (OR = 0.13, 95% CI: 0.04-0.42, p < 0.001).

Conclusion: NAR trajectory patterns independently predict clinical outcomes after EVT for LVO-AIS. Dynamic monitoring of NAR, particularly on day 3 post-procedure, may facilitate early risk stratification and development of targeted intervention strategies, providing a new biomarker tool for precision stroke management.

Keywords: acute ischemic stroke; clinical outcomes; endovascular therapy; large vessel occlusion stroke; latent class trajectory modeling; neutrophil-to-albumin ratio; propensity score matching.

Figure 1

Dynamic changes in neutrophil counts, serum albumin levels, and neutrophil-to-albumin ratio during the first 3 days after endovascular therapy. Box plots show the temporal evolution of neutrophil counts (×10⁹/L), serum albumin (g/L), and neutrophil-to-albumin ratio (NAR) at baseline, day 1, and day 3 after endovascular treatment. The boxes represent the interquartile range (IQR), with the horizontal line indicating the median. Whiskers extend to 1.5 times the IQR, and individual points represent outliers. Red lines connect the median values across time points. Statistical significance: *p < 0.05; **p < 0.01; ***p < 0.001; ns, not significant.

Figure 2

Restricted cubic spline analysis of the association between neutrophil-to-albumin ratio and risk of poor functional outcome after endovascular therapy. Restricted cubic spline regression analysis showing the nonlinear relationships between NAR measured at baseline, day 1, day 3, and the average of all time points (D) with the risk of 90-day poor functional outcome. The solid red line represents the adjusted odds ratio, and the pink shaded area represents the 95% confidence interval. The reference line (odds ratio = 1) is indicated by the dashed horizontal line. Blue histograms show the distribution of NAR values in the study population. P-overall indicates the significance of the overall association, while P-nonlinear indicates the significance of nonlinear components.

Figure 3

Neutrophil-to-albumin ratio trajectory patterns identified by latent class trajectory modeling in patients with anterior circulation large vessel occlusion stroke after endovascular therapy. Class-specific mean predicted trajectories of neutrophil-to-albumin ratio (NAR) evolution during the first 72 h after endovascular therapy. Two distinct trajectory patterns were identified through latent class trajectory modeling: Class 1 (black line, 31.1% of patients) represents a “continuously increasing” pattern. Class 2 (red line, 68.7% of patients) exhibits a “peak-then-decline” pattern.

Figure 4

Receiver operating characteristic curves for neutrophil-to-albumin ratio in predicting clinical outcomes after endovascular therapy. Receiver operating characteristic curves showing the predictive performance of NAR at different time points for (A) 90-day poor functional outcome (modified Rankin Scale score 3–6), (B) symptomatic intracranial hemorrhage, (C) malignant cerebral edema, and (D) in-hospital mortality.