Association Between Serum Lactate Dehydrogenase Level and Hematoma Expansion in Patients with Primary Intracerebral Hemorrhage: A Propensity-Matched Analysis

Abstract

The present study aimed to explore whether a higher serum lactate dehydrogenase (sLDH) level on admission is associated with hematoma expansion (HE) in patients with primary intracerebral hemorrhage (ICH). This single-center prospective observational study of patients with primary ICH aged 19 years or older was conducted at the Dehua County Hospital from January 2018 to May 2021. Clinical data and demographic information and outcomes were collected and analyzed. The association between increased sLDH levels and HE was assessed in univariate and multivariate analyses. Propensity-score matching (PSM) analysis was implemented to reduce baseline differences between the groups. Of 609 patients with ICH screened, 360 who metall eligibility criteria were enrolled in the study (mean age, 59.83 ± 12.64 years; 60.28% female patients), of whom 69 (19.17%) developed early HE. sLDH levels were statistically higher in the HE group compared with the non-HE group (236.0 [222.30-275.50] U/L vs. 209.6 [179.30-253.8] U/L; P < 0.001). Multivariable analysis showed that higher sLDH levels were still statistically associated with HE (odds ratio [OR], 0.08; 95% confidence interval, 0.03-0.210; P < 0.001). After PSM, the matched HE group had a significantly higher sLDH level than did the matched non-HE group (236.0 [222.0-279.10] vs. 216.30 [173.0-278.7] U/L; P= 0.003). The area under the curve of 0.704 (95% confidence interval, 0.654-0.751; P < 0.0001) (sensitivity, 92.75%; specificity, 52.58%), and the optimal cutoff value for sLDH level as a predictor for HE in patients with primary ICH was determined as 211.0U/L. The area under the curve of the logistic regression model based on these predictors (the TsL (time from onset to initial computed tomography,sLDH) modelbased on these predictors: sLDH, time from onset to initial computed tomography) was 0.817, with a sensitivity of 84.06% and specificity of 72.51% for HE. The TsL model produced the best ability to predict HE compared with single sLDH. sLDH levels were statistically correlated with poor outcome. The current PSM analysis study shows that increased serum LDH level is statistically associated with HE. Our findings indicate that the TsL model constructed by sLDH and time from onset to initial computed tomography markedly enhances the prediction of HE after ICH.