Relationship between postoperative biomarkers of neuronal injury and postoperative cognitive dysfunction: A meta-analysis

• Published in: PloS one 2023-04, Vol.18 (4), p.e0284728-e0284728
• Main Authors: Wang, Xiaohua, Chen, Xinli, Wu, Fan, Liu, Yingchao, Yang, Yushen, Chen, Weican, Pan, Zhigang, Hu, Weipeng, Zheng, Feng, He, Hefan
• Format: Article
• Language: English
• Subjects: Alzheimer's disease; Alzheimers disease; Analysis; Bias; Biological markers; Biology and Life Sciences; Biomarkers; Biomarkers - metabolism; Clinical trials; Cognition disorders; Cognitive ability; Cognitive Dysfunction - diagnosis; Cognitive Dysfunction - etiology; Complications; Glial fibrillary acidic protein; Guillain-Barre syndrome; Humans; Injury analysis; Medicine and Health Sciences; Meta-analysis; Neuronal-glial interactions; Observational studies; Pathophysiology; Phosphopyruvate hydratase; Physical Sciences; Postoperative Cognitive Complications; Postoperative Complications - diagnosis; Postoperative Complications - etiology; Postoperative Complications - metabolism; Research and Analysis Methods; Risk factors; S100 Calcium Binding Protein beta Subunit; Sampling; Surgery; Tau protein
• ISSN: 1932-6203; 1932-6203
• DOI: 10.1371/journal.pone.0284728

Early biomarkers are needed to identify patients at risk of developing postoperative cognitive dysfunction (POCD). Our objective was to determine neuronal injury-related biomarkers with predictive values for this condition. Six biomarkers (S100β, neuron-specific enolase [NSE], amyloid beta [Aβ], tau, neurofilament light chain, and glial fibrillary acidic protein) were evaluated. According to the first postoperative sampling time, observational studies showed that S100β was significantly higher in patients with POCD than in those without POCD (standardized mean difference [SMD]: 6.92, 95% confidence interval [CI]: 4.44-9.41). The randomized controlled trial (RCT) showed that S100β (SMD: 37.31, 95% CI: 30.97-43.64) and NSE (SMD: 3.50, 95% CI: 2.71-4.28) in the POCD group were significantly higher than in the non-POCD group. The pooled data of observational studies by postoperative sampling time showed significantly higher levels of the following biomarkers in the POCD groups than in the control groups: S100β levels at 1 hour (SMD: 1.35, 95% CI: 0.07-2.64), 2 days (SMD: 27.97, 95% CI: 25.01-30.94), and 9 days (SMD: 6.41, 95% CI: 5.64-7.19); NSE levels at 1 hour (SMD: 0.92, 95% CI: 0.25-1.60), 6 hours (SMD: 0.79, 95% CI: 0.12-1.45), and 24 hours (SMD: 0.84, 95% CI: 0.38-1.29); and Aβ levels at 24 hours (SMD: 2.30, 95% CI: 1.54-3.06), 2 days (SMD: 2.30, 95% CI: 1.83-2.78), and 9 days (SMD: 2.76, 95% CI: 2.25-3.26). The pooled data of the RCT showed that the following biomarkers were significantly higher in POCD patients than in non-POCD patients: S100β levels at 2 days (SMD: 37.31, 95% CI: 30.97-43.64) and 9 days (SMD: 126.37, 95% CI: 104.97-147.76) and NSE levels at 2 days (SMD: 3.50, 95% CI: 2.71-4.28) and 9 days (SMD: 8.53, 95% CI: 7.00-10.06). High postoperative levels of S100β, NSE, and Aβ may predict POCD. The relationship between these biomarkers and POCD may be affected by sampling time.