Association of postoperative elevation of biomarkers of neural injury in blood with postoperative neurocognitive disorders in adults receiving general anaesthesia: a systematic review.

Abstract

Perioperative neural injury, evidenced by postoperative elevation in neurofilament light (NfL) and glial fibrillary acidic protein (GFAP), is a potential biological basis of perioperative neurocognitive disorders. This systematic review evaluated the association between perioperative biomarker changes and postoperative cognitive disorder (POCD) and postoperative delirium (POD) in patients older than 60 yr undergoing surgery under general anaesthesia. The protocol was registered (CRD42023479136). We searched MEDLINE, EMBASE, Cochrane Library, Scopus, Web of Science and CINAHL between 1990 and May 2025. We included studies reporting the association between biomarker changes and POCD or POD and used the Quality in Prognosis Studies tool for quality assessment. Results were summarised using effect direction plots and and GRADE approach. We identified 12 studies (1212 participants). Most studies of NfL reported a postoperative increase. Two studies reported the association between changes in NfL and POCD and three reported the association between changes in GFAP and POCD. All exhibited a high risk of bias and yielded inconsistent findings. A postoperative increase in NfL was associated with POD in four of six studies, with substantial variability in patient populations, adjustment for confounders, and statistical methods that limit interpretation. There was no consistent association between GFAP increase and POD. The significance of postoperative increases in plasma NfL and GFAP remains uncertain. Low-certainty evidence supports an association between an increase in NfL and POD, with very low certainty of evidence for an association between an increase in GFAP and POD, and either biomarker and POCD. Further high-quality studies are needed.