Perioperative neurocognitive disorders (PND) include postoperative delirium within 7 days after surgery, delayed neurocognitive recovery up to 30 days, and postoperative neurocognitive disorder up to 12 months. These outcomes are related, but they are not the same. They arise from the interaction of baseline brain vulnerability and perioperative stress, including inflammation, vascular instability, blood–brain barrier injury, metabolic strain, and reduced neural reserve. Preoperative genetic profiling is useful because it can estimate latent susceptibility before surgery. Among current signals, APOE is the strongest and most biologically relevant locus. At the same time, Alzheimer’s disease polygenic risk scores (AD-PRS) can capture non- APOE common-variant burden across lipid transport, endosomal trafficking, innate immune signaling, complement activity, microglial regulation, mitochondrial stress, and neurovascular integrity. Recent perioperative cohort studies have begun to test preoperative APOE-based and polygenic neurocognitive risk in surgical patients. Large delirium genetics studies also show a strong signal at the APOE locus and support overlap between delirium risk and Alzheimer’s disease-related common-variant architecture. These findings support an APOE-aware framework in which APOE genotype is modeled separately from non- APOE AD-PRS. In clinical use, this genomic layer should be combined with baseline cognition, frailty, vascular comorbidity, surgery-related risk, and circulating biomarkers such as neurofilament light chain. This review summarizes the loci, molecular pathways, and translational model designs that can move preoperative genomic profiling from association to perioperative risk stratification.
Tan et al. (2026) studied this question.