Mechanistic insights into postoperative delirium using untargeted high-throughput proteomics in elderly patients - a case-control study

Postoperative delirium (POD) in elderly patients is a serious complication after surgery with an unclear molecular pathogenesis. Perioperative untargeted high-throughput proteomic profiling may provide insights into underlying mechanistic molecular patterns and help identify patients at high risk, guiding prevention and therapy. This study is a monocentric substudy of the European BioCog project, a prospective multicentre observational study involving elderly patients aged ≥ 65 undergoing elective surgery with a Mini-Mental State Examination score of ≥ 24. POD was assessed twice daily for up to 7 days using the Nursing Delirium Screening Scale (Nu-DESC) and the Confusion Assessment Method for the Intensive Care Unit (CAM-ICU). We conducted high-throughput liquid chromatography mass spectrometry on sequential pre- and postoperative plasma samples. Data are available via ProteomeXchange with the identifier PXD071265. Data were analysed using a matched case-control design, employing both cross-sectional and longitudinal approaches, along with pathway enrichment analysis. A total of 226 highly abundant proteins were investigated in 168 patients (50% POD incidence). Multiple pathways, particularly those involved in the activation of the innate immune response and the complement system, were associated with POD. Butyrylcholinesterase showed the most robust regulation, with preoperative downregulation and postoperative upregulation in patients with POD, whereas it was further downregulated postoperatively in those without POD. Catalytic enzyme activity showed significant decrease in both groups post-surgery. A combined set of 8 preoperative proteins classified patients with POD with 86% sensitivity and 79% specificity. Untargeted high-throughput proteomics is a feasible approach to characterise pathways involved in POD pathogenesis. This case-control study identified a protein signature associated with POD, emphasising the need for larger cohorts to confirm these observations and improve the mechanistic understanding of POD.