Background: 

Delirium is an acute state of confusion associated with adverse postoperative outcomes. Delirium is diagnosed clinically using screening tools; most cases go undetected. Identifying a delirium biomarker would allow for accurate diagnosis, application of therapies, and insight into causal pathways. To agnostically discover novel biomarkers of delirium, we conducted a case-control sub-study using the VISION-Cardiac Surgery biobank. Our objective was to identify candidate biomarkers to investigate in future studies.

Methods: 

We obtained a convenience sample of 30 patients with delirium on postoperative day 1 matched to 30 matched controls by age, sex, ethnicity, center and cardiopulmonary bypass time. The Olink Explore 3K platform was used to identify blood protein alterations on postoperative day 3. Protein concentrations were expressed as normalized protein expression (NPx) units (log₂ fold scale). We compared protein expression between cases and controls using a paired t-test and reported significantly different biomarkers based on a False Discovery Rate (FDR)-adjusted p-value<0.05.

Results: 

Of 2,865 unique serum proteins, 26 (0.9%) were significantly associated with delirium status; all were elevated in cases versus controls at an FDR<0.05. Pathway analysis identified “calcium-release channel activity” (P_adj=0.02) and “Guanosine 5’ triphosphate (GTP)-binding” (P_adj=0.005) functions as characteristic of proteins associated with delirium. The top three differentially expressed biomarkers were FKBP1B (P_adj=0.003), C2CD2L (P_adj=0.004), and RAB6B (P_adj=0.004). The inflammatory biomarker IL-8 (CXCL8) (mean difference = 2.36; P=3.6×10^-4) was also associated with delirium.

Discussion: 

We identified 26 biomarkers significantly associated with delirium; all are novel except for IL-8. We did not identify an association between delirium and recognized neuro-inflammatory proteins and markers of brain injury, which supports using biomarkers to differentiate between delirium and other neurological conditions. While exploratory, our findings support using biomarkers to diagnose postoperative delirium and validate using agnostic screens to identify potential delirium biomarkers.