A total of 2054 aSAH patients in our institution between 2010 and 2018 were collected, and 874 patients treated by microsurgical clipping were enrolled based on the upper criteria (figure 1). Incidence of poor outcome following aSAH was 13.8%( 121/874). Basic clinical characteristics of patients with aneurysmsal subarachnoid hemorrhage were shown on table 1 and 2. The median serum LDH level(U/L) in the good outcome group (180.096±50.237) was obviously lower than that in the poor outcome group (227.554±83.002)(p=0.000). The receiver operating characteristic (ROC) curve of serum LDH level for poor outcome of aSAH patients at the 3-month follow-up is shown in Figure 2. The area under the ROC curve (AUC) was 0.702(95% confidence interval [CI], 0.650 - 0.754; p=0.000). The optimal cutoff value for serum LDH level as a predictor for 3-month poor outcome (mRS>2) was determined as 201.5U/L in the ROC curve. On this level, the sensitivity was 59.5%, the specificity was 76.4%.
In order to analyze the predictors for poor outcome of aSAH patients, the variables significance level at p < 0.10 including age,hypertension,diabetes,Hunt-Hess,Fisher,anterior communicating artery aneurysm, basilar artery aneurysm, delay ischemic neurological deficit(DIND), hydrocephalus, pneumonia, serum LDH(>201.5U/L) were included in the univariate analysis. The results revealed that age, hypertension, Hunt-Hess grade,Fisher grade, delay ischemic neurological deficit, hydrocephalus, pneumonia, serum LDH(>201.5U/L) were associated with 3-month poor outcome(table 2, p < 0.05). In multivariate analysis, however, Hunt-Hess grade,Fisher grade, delay ischemic neurological deficit, pneumonia, serum LDH(>201.5U/L) were still significantly associated with outcome, whereas, age,hypertension, diabetes, anterior communicating artery aneurysm, basilar artery aneurysm and hydrocephalus were not. Patients with Hunt-Hess grade IV-V had a 1.6-fold increased risk of developing poor outcome (OR 1.637, 95% CI 1.266-2.118, p=0.000). Fisher grade 4 had a 1.5-fold increased risk of developing poor outcome(OR 1.517, 95% CI 1.182-1.946, p=0.001). DIND had a 4.2-fold increased risk of developing poor outcome(OR 4.234, 95% CI 2.412-7.432, p=0.000). Pneumonia had a 3.8-fold increased risk of developing poor outcome(OR 3.848, 95% CI 2.386-6.206, p=0.000). Serum LDH level greater than >201.5U/L was associated with a 2.7-fold increase risk of developing poor outcome(OR 2.702, 95% CI 1.645-4.440, p=0.000) (table 2). After PSM, there were no significant differences in Hunt-Hess grade,Fisher grade, delay ischemic neurological deficit, pneumonia between good outcome and poor groups (table 1 and 2). In the logistic regression model (table 2), serum LDH(>201.5U/L) was still considered as an independent risk factor of poor outcome (OR 2.426, 95% CI 1.378-4.271, p = 0.002).
Interestingly, we found that serum LDH level was associated with Hunt-Hess and Fisher grade, it was revealed that serum LDH level(U/L) was 163.880±35.571 in Hunt-Hess grade I group, which was lower than that in II (174.981±49.616), III (188.306±50.702), IV (225.609±69.509), and V (252.851±93.302). There were statistically significant differences between groups (p=0.000), and there was obvious trend that serum LDH level will increase concomitantly with increasing grade of Hunt-Hess (shown in figure 3). Serum LDH level was 169.492±41.621 in Fisher grade 1 group, which was lower than that in grade 2 (177.097±42.621), grade 3(198.709±72.553) and grade 4 (210.811±68.962). There were statistically significant differences between grade 4 vs 3, 4 vs 2, 4 vs 1, 3 vs 2, 3 vs 1(p=0.000). There was obvious trend that serum LDH level will increase concomitantly with increasing Fisher grade(show in figure 4).
It was also shown that serum LDH level correlated with neurological functional outcome. Serum LDH level was 179.247±46.761 in no symptoms (mRS 0) group, which was lower than that in no significant to slight disability (mRS 1-2) (193.977±69.399), moderate to serious disability (mRS 3-4) (205.918±59.203) and severe disability to death (mRS 5-6) (234.188±108.336). There were statistically significant differences (p=0.000). There was obvious trend that serum LDH level would increase with deterioration of neurological function (show in figure 5).