Longitudinal Blood Transcriptomic Changes Predict Lung Function Decline in Idiopathic Pulmonary Fibrosis

The Rationale: Molecular markers of disease activity that are predictive of forced vital capacity (FVC) progression in idiopathic pulmonary fibrosis are needed. Objectives: Develop a predictor using longitudinal within-patient gene expression differences (ΔGE) in peripheral blood mononuclear cells (PBMC) to predict of FVC progression. Methods: Patients in the training cohort (n=74) experiencing ≥10% relative reduction in FVC% of predicted over 12 months were categorized as progressors in contrast to the remaining stable patients. Baseline to 4-month within-patient ΔGE were correlated with FVC status. FVC-predictor genes were prioritized by two-group comparison with FDR<5%, logistic LASSO regression with p<0.05, and 10-Fold Cross-Validation with ≥50% support. Receiver operating characteristic with area under the curve (AUC) analyses were conducted in training subsets and independent validation cohorts from UChicago (n=27), UPMC (n=35), and Imperial (n=24) where different transcriptome assay platforms and varying transcriptome sampling times were used to derive ΔGE. Results: Intra-subject Compared to cross-sectional analysis of baseline GE, our longitudinal ΔGE variation approach demonstratedlargely reduced within-group sample variation and increased statistic power fin progressor and stable groups for prediction model development. A 25-gene FVC-predictor separated “progressors” from “stable” by Principal Component Analysis in the training and subsets of the training cohort. The resulting FVC-predictor consistently demonstrated high discriminatory performance independent of transcriptome assay platforms and varying sampling times in validation cohorts (AUC= 0.77-0.80). TGF beta was the highest-ranking canonical pathway by Gene Set Enrichment Analysis. Conclusions: Our novel short-term longitudinal within-patient ΔGE approach identified a FVC-predictor which may reflect disease activity and prove to be a reliable biomarker predictive of future FVC decline. Subjects included in this analysis were participants in COMET-IPF (Correlating Outcomes with biochemical Markers to Estimate Time-progression in Idiopathic Pulmonary Fibrosis), a prospective, observational study correlating biomarkers with disease progression (NCT01071707) (E1). This multicenter investigation recruited subjects at nine clinical centers in the US. Inclusion criteria required diagnosis of IPF was confirmed using a multidisciplinary diagnostic approach per international guidelines using expertise from clinicians, radiologists, and pathologists at the local, enrolling clinical center and age 35-80 years. Subjects were excluded if the diagnosis of IPF was >4 years prior to screening or if there was a diagnosis of collagen-vascular disorder, FEV1/FVC <0.60, evidence of active infection at screening, or comorbid conditions other than IPF likely to result in death within one year. Subjects underwent protocol-directed visits every 4 months after the baseline (0 visit) for a minimum of 1 year, establishing four transcriptome sampling timepoints. Registry patients with PBMC gene expression (GE) sampling and pulmonary function tests (PFT) over at least two time points were included in each cohort. Informed consent was obtained from all participants. The study protocol was reviewed by the institutional review board of each participating center.