Peripheral Blood Mononuclear Cell Gene Expression Profiles May Predict Poor Outcome in Idiopathic Pulmonary Fibrosis [Agilent]

Background: In this study we aimed to identify peripheral blood mononuclear cell (PBMC) gene expression profiles predictive of poor outcomes in idiopathic pulmonary fibrosis (IPF) Methods: Microarray analyses of PBMC were performed in 120 patients from discovery (n=45) and replication cohorts (n=75). Genes and pathways associated with transplant-free survival (TFS) were identified and confirmed by qRT-PCR. Findings: 52 genes were predictive of TFS in a discovery cohort (FDR<5%, Cox score above 2.5 or below -2.5). Clustering the replication cohort samples using these genes distinguished two patient groups with significantly different TFS (hazard ratio 1.96, 95%CI 1.01-3.8, P=0.018). Decreased expression of “The co-stimulatory signaling during T cell activation” Biocarta pathway and in particular CD28, ICOS, LCK and ITK was associated with shorter TFS times in each cohort (FDR<5%). qRT-PCR expression of CD28, ICOS, LCK and ITK correlated with the microarray results in the discovery cohort (P<0.05) and their decreased expression was predictive of shorter TFS in the replication cohort (P<0.05). A genomic and clinical model demonstrated an area under the ROC curve of 78.5% at 2.4 months for death and lung transplant prediction. Interpretation: Our results suggest that CD28, ICOS, LCK and ITK are outcome biomarkers in IPF. PBMC from 75 patients with the diagnosis of IPF were obtained within 30 minutes from blood draw. Total RNA was extracted, labeled and hybridized to Agilent Whole Human Genome Oligo Microarray, 4 x 44K. Patients were followed from blood draw until death, transplant or last follow up. Hierarchical clustering and gene-set analysis with censored outcome data were used to study the association of gene expression and outcome in this cohort (replication cohort)