Abstract
The study of the human response to injury has been hampered by the inherent heterogeneity in the models and methods used. By studying a standard injury longitudinally, using individual patient-level analysis, we endeavoured to better describe its dynamics. We analysed clinical variables, clinical laboratory and plasma cytokines from 20 patients at five time points. Clustering analysis showed two prototype patterns of cytokine behaviour: a concordant type, where cytokines behave the same way for all patients (notably IL-0 and TNFα), and a variable type, where different patterns of expression are seen for different patients (notably IL-8, IL-6 and IL-1RA). Analysis of the cytokines at the individual patient-level showed a strong four-way correlation between IL-1RA, GCSF, MIP-1β and MCP-1. As it holds for most patients and not just on average, this suggests that they form a network which may play a central role in the response to gastro-intestinal injuries in humans. In conclusion, the longitudinal analysis of cytokines in a standard model allowed the identification of their underlying patterns of expression. We propose that the two prototype patterns shown may reflect the mechanism that separates the common and individual aspects of the injury response.