Principal component analysis of the cytokine and chemokine response to human traumatic brain injury

• Published in: PloS one Vol. 7; no. 6; p. e39677
• Main Authors: Helmy, Adel, Antoniades, Chrystalina A, Guilfoyle, Mathew R, Carpenter, Keri L H, Hutchinson, Peter J
• Format: Journal Article
• Language: English
• Published: United States Public Library of Science 22.06.2012; Public Library of Science (PLoS)
• Subjects: Biochemistry; Bioindicators; Biology; Biomarkers; Brain; Brain injuries; Brain Injuries - metabolism; Brain research; Catheters; Chemokines; Chemokines - metabolism; Cytokines; Cytokines - metabolism; Data processing; Discriminant analysis; Ethics; Fluid; Fluids; Growth factors; Head injuries; Human behavior; Humans; Inflammation; Inflammatory response; Injury analysis; Intensive care; Mathematics; Medical imaging; Medicine; Microdialysis; Neurosciences; Neurosurgery; Peripheral blood; Principal Component Analysis - methods; Principal components analysis; Statistical analysis; Statistical methods; Statistics; Studies; Trauma; Traumatic brain injury; United Kingdom > UK
• ISSN: 1932-6203; 1932-6203
• DOI: 10.1371/journal.pone.0039677

There is a growing realisation that neuro-inflammation plays a fundamental role in the pathology of Traumatic Brain Injury (TBI). This has led to the search for biomarkers that reflect these underlying inflammatory processes using techniques such as cerebral microdialysis. The interpretation of such biomarker data has been limited by the statistical methods used. When analysing data of this sort the multiple putative interactions between mediators need to be considered as well as the timing of production and high degree of statistical co-variance in levels of these mediators. Here we present a cytokine and chemokine dataset from human brain following human traumatic brain injury and use principal component analysis and partial least squares discriminant analysis to demonstrate the pattern of production following TBI, distinct phases of the humoral inflammatory response and the differing patterns of response in brain and in peripheral blood. This technique has the added advantage of making no assumptions about the Relative Recovery (RR) of microdialysis derived parameters. Taken together these techniques can be used in complex microdialysis datasets to summarise the data succinctly and generate hypotheses for future study.