A MULTI-PROTEIN SIGNATURE OF POSTOPERATIVE DELIRIUM. (11th November 2018)

• Record Type: Journal Article
• Authors: Vasunilashorn, S; Zhou, W; Ngo, L; Dillon, S; Otu, H; Inouye, S; Libermann, T; Marcantonio, E
• Is Part Of: Innovation in aging. Volume 2(2018)Supplement 1
• Page Start: 571
• Page End: 572
• Publication Date: 2018-11-11
• Subjects: Aging -- Periodicals; Gerontology -- Periodicals; 612.67
• DOI: 10.1093/geroni/igy023.2115
• Languages: English
• ISSNs: 2399-5300

Abstract

Abstract: Delirium is common, morbid, and costly, yet its biology remains poorly understood. To advance understanding of delirium pathophysiology, we developed a multi-protein signature of delirium identifying proteins associated with delirium from unbiased quantitative proteomics using mass spectrometry (iTRAQ) and combining them with delirium biomarkers identified in our prior Luminex-based work (interleukin [IL]-6 and IL-2). Using 12 delirium cases and matched control pairs from the Successful Aging after Elective Surgery Study of adults age ≥70 undergoing major non-cardiac surgery (N=560; 24% delirium), iTRAQ was applied to identify a set of delirium-associated proteins. We then conducted enzyme-linked immunosorbent assays (ELISA) validation in 75 matched pairs. Combining validated proteomics proteins with IL-6 and IL-2, we considered all possible combinations of PREOP and POD2 proteins for predicting delirium using conditional logistic regression (1023 models). The best models were selected using the Akaike's Information Criteria (AIC) and area-under-the-curve (c-statistic). We identified and validated three proteins from proteomics: C-reactive protein (CRP), zinc alpha-2-glycoprotein (AZGP1) and alpha-1-antichymotrypsin. When combined with IL-6 and IL-2 for prediction of delirium, the best model included: PREOP: CRP and AZGP1, and POD2: IL-6, IL-2, and CRP (AIC: 73.64, c-statistic: 0.88). This model had better fit statistics (better predictive power for delirium) than models that included one protein at two timepoints (best single-protein model: CRP PREOP and POD2 [AIC: 82.81, c-statistic: 0.81]). Our multi-protein signature correctly classified delirium cases more often than any single-protein model. This multi-protein signature improves understanding of delirium pathophysiology and may help to identify patients at-risk for delirium.