Validation of the Riboleukogram to Detect Ventilator-Associated Pneumonia after Severe Injury
*J. Perren Cobb, MD 1 , Ernest E. Moore, MD2 , *Doug L. Hayden, MA3 , *Joseph P. Minei, MD4 , *Joseph Cuschieri, MD5 , *Jingyun Yang, PhD3 , *Bernard H. Brownstein, PhD1, *Laura Hennessy, RN6 , *Philip H. Mason, PhD3, *William S. Schierding, MS1, *David J. Dixon, PhD1, Ronald G. Tompkins, MD, ScD3 , *H. Shaw Warren, MD3, *David A. Schoenfeld, PhD 3 , Ronald V. Maier, M 5
1 Washington University in St. Louis, St. Louis, MO; 2 University of Colorado, Denver, CO; 3 Massachusetts General Hospital, Boston, MA; 4 University of Texas Southwestern, Dallas, TX; 5 University of Washington, Seattle, WA; 6 Harborview Medical Center, Seattle, WA
OBJECTIVE(S): The most common complication following injury is ventilator-associated pneumonia (VAP). We hypothesized that circulating leukocyte RNA profiles (“riboleukograms”) can detect VAP. A pilot microarray study of 11 VAP patients suggested previously that 85 leukocyte genes can be used to diagnose VAP. Validation of this gene set was tested herein using data from an existing, independent patient cohort.
METHODS: 158 intubated, severely injured patients were enrolled at 5 centers, where 57 (36%) developed VAP (ATS definition; 72% with BAL quantitative cultures). Patient age was 34.2 +/- 11.1 years; 65% were male. Circulating leukocyte GeneChip U133 2.0 expression values were measured at time 12 h and 1, 4, 7, 14, 21, and 28 d after injury. To determine the association of gene expression with initial onset of VAP, DChip normalized leukocyte expression data were analyzed using repeated measures logistic regression and generalized estimating equations.
RESULTS: Using these expression values measured on each study day, 27 (32%) of the 85 genes were associated with the diagnosis of VAP 1-4 days prior to diagnosis (FDR<0.05). Of these 27 genes, the majority are associated with early neutrophil activation in response to (and for the subsequent clearance of) bacterial infections.
CONCLUSIONS: Using the largest patient cohort yet reported, our results validate those described in an earlier (pilot) study, confirming that riboleukograms can detect the development of VAP days prior to attending diagnosis. The clinical implications of these findings are substantial (earlier diagnosis, better outcomes), and motivate development and testing of a predictive model for VAP.