To identify serum biomarkers predictive of death and near-death (ECMO) in children with life-threatening influenza infection.
Objective:From January to August 2009, 22 sites in the PALISI Network enrolled 151 children admitted to the ICU with community Methods: acquired influenza infection. We collected extensive clinical data, performed multiplex viral PCR testing, collected serum and followed children to hospital discharge. Sera from 97 subjects were analyzed for a multiplex panel of 30 cytokines and chemokines using the Luminex platform as well as ELISA for IFNbeta. Results were interpolated from 5-parameter fits of standard curves using the corresponding recombinant proteins. All samples were analyzed in duplicate (coefficient of variation < 10%). We used linear and logistic regression to assess whether serum biomarkers on presentation were associated with severity of illness (PRISM III APS), death and death and/or ECMO.In the 97 children, 9 died and 3 additional children were ECMO survivors. Half of the sample (49%) had 2009 Pandemic H1N1 Results: (2009P-H1N1) and the other half had seasonal flu (70% influenza A, 30% influenza B). All analyses were adjusted for age, gender, underlying health conditions and immune compromise. Of the inflammatory mediators evaluated, serum MIP-1alpha and IL10 were strongly associated with illness severity (P=0.001, P=0.007). Serum levels of IL-6, IL-8, IL-10, GCSF MCP-1 and IP-10 were found to be strong risk factors for fatality (death or ECMO) (P=0.01 to 0.0008). IL-6, IL-8, IL-10, MCP-1 and GCSF were still significant (P=0.01 to 0.04) after additional adjustment for bacterial coinfection and PRISM APS. Addition of these serum biomarkers and clinical factors in a multivariate model predicting fatality (death or ECMO) increased the area under the ROC from 0.906 for clinical factors alone to 0.939 with addition of IL-6 and to 0.925 with addition of IL-8. After adjusting for age, gender and immune compromise, serum IL-2 was significantly higher in children with 2009P-H1N1 compared to those with seasonal flu (OR 2.79, 95% CI 1.58, 4.92, P = 0.0004).
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.