Dynamic serum lactate dehydrogenase monitoring during the acute phase and association with in-hospital all-cause mortality risk in large vessel occlusion acute ischemic stroke patients.

BACKGROUND

Lactate dehydrogenase (LDH), a key glycolytic enzyme released abundantly during cellular injury, has been established as a prognostic biomarker in ischemic stroke. However, the dynamic changes and predictive value of serum LDH during the acute phase in patients with large vessel occlusion acute ischemic stroke (LVO-AIS) remain insufficiently characterized.

METHODS

This retrospective cohort study consecutively enrolled 414 LVO-AIS patients who underwent endovascular treatment (EVT) at a comprehensive stroke center between January 2019 and November 2024. Serum LDH levels were measured at admission, on post-EVT day 1, and day 3. In-hospital all-cause mortality (IHM) served as the primary endpoint. Friedman tests assessed longitudinal trends in LDH levels, with Durbin-Conover post hoc pairwise comparisons. Progressively adjusted multivariable logistic regression models evaluated associations between LDH and IHM. Restricted cubic spline (RCS) regression explored potential non-linear relationships, while subgroup analyses and interaction testing assessed the consistency and robustness of associations.

RESULTS

Among 414 patients, 58 (14.0%) experienced IHM. Baseline LDH levels showed no significant difference between survivors and non-survivors (P = 0.108), whereas the non-survivor group demonstrated significantly elevated LDH levels on post-EVT days 1 and 3 (P < 0.001). In the fully adjusted model, each 10 U/L increase in day-3 LDH was associated with a 7% increased mortality risk (P < 0.001). The highest LDH quartile conferred a 10.75-fold increased mortality risk compared with the lowest quartile (P = 0.002). ROC analysis revealed good predictive performance for day-3 LDH levels (AUC = 0.74) and absolute change from baseline to day 3 (ΔLDH, AUC = 0.74). RCS analysis confirmed a significant linear dose-response relationship between LDH and IHM (both P-nonlinear > 0.05). No significant interaction effects were observed across subgroups (all P for interaction > 0.05).

CONCLUSIONS

Dynamic LDH monitoring, particularly day-3 post-EVT levels, provides valuable prognostic information for risk stratification in LVO-AIS patients. The temporal LDH pattern may reflect the evolution of cerebral tissue injury and reperfusion injury, facilitating early identification of high-risk patients requiring intensified monitoring and potential therapeutic interventions.