Inflammation and the Vulnerable Brain: Understanding Postoperative Delirium and its Associated Long‐Term Cognitive Outcomes

Abstract Background Although growing evidence highlights inflammation as a common biological mechanism of delirium and Alzheimer’s disease (AD), not all individuals with high inflammation develop delirium, AD, and/or cognitive decline. Therefore other predisposing factors (e.g., brain vulnerability) likely influence the effect of inflammation on the brain. Method We used the Role of Inflammation after Surgery for Elders (RISE) study of adults aged ≥65 who underwent elective surgery under spinal anesthesia (n = 40), to investigate the interaction of brain vulnerability and inflammation on persistent neuronal injury. Brain vulnerability was examined from preoperative cerebrospinal fluid: 1) amyloid status, and 2) ATN classification (ATN+ vs. negative for all groups [ATN‐]). Plasma levels of C‐reactive protein (CRP), interleukin‐6, and chitinase 3‐like protein (CHI3L1/YKL‐40), previously associated with delirium, were assigned values of 1‐4 for each marker within sample‐based quartiles. For each patient, we summed these values for the three markers to create an inflammation score (range 0‐12). Persistent neuronal injury, measured using postoperative 1 month (PO1MO) plasma, was measured from neurofilament light chain (NfL) and glial fibrillary acidic protein (GFAP). Linear regression models examined the association between inflammation score and neuronal injury, stratified by brain vulnerability. Result On average, participants were age 75 years, 65% female, and 7% had Charlson scores ≥2. We identified a brain vulnerability x inflammation interaction on neuronal injury. Among amyloid+ patients, those in the highest inflammation score tertile [T3] vs T1 had 0.16 pg/ml higher GFAP (p<.05); however, among amyloid‐ patients, no inflammation and GFAP association was observed. No differences between inflammation and neuronal injury, by ATN classification, was observed. Conclusion The association between inflammation and persistent neuronal injury differs by amyloid pathology, suggesting the potential value of identifying older patients vulnerable to adverse postoperative cognitive outcomes who may benefit from interventions under conditions of high inflammation.