Methods
Trial design2
- Subjects with IPF and DLco ≤35% predicted were randomized to receive nintedanib 150 mg bid plus sildenafil 20 mg tid or nintedanib 150 mg bid plus placebo for 24 weeks. Both subjects who were nintedanib-naïve and those who had previously taken nintedanib were eligible to participate.
RNA sequencing
- Analyses were based on total RNA extracted from whole blood samples taken at baseline and week 24.
- RNA quantity and quality were measured using a NanoDrop spectrophotometer.
- Total RNA sequencing, with approximately 50 million reads per sample, was performed using the TruSeq Stranded Total RNA Kit with Ribo-Zero Globin and a HiSeq 4000 (Illumina).
Analyses
- We analyzed changes in gene expression from baseline at week 24 in the two treatment groups:
- Data were log2 transformed prior to analysis.
- p-values were adjusted to control the false discovery rate at 5%.
- Changes in gene expression over 24 weeks were considered significant if adjusted p≤0.05 and|log2fold change|≥0.5 (i.e., there was a ≥1.4-fold difference between baseline and week 24).
- In a secondary investigation, we analyzed changes in the expression of nine genes that were downregulated after 12 weeks’ treatment with nintedanib in the INMARK trial in subjects with IPF and preserved lung function:3
- Changes were considered significant if unadjusted p≤0.05 and |log2fold change|≥0.5.
- Gene set variation analysis assessed the relative enrichment of this set of nine genes. Enrichment scores were tested using a simple linear model and moderated t-statistics.
- Pathways analyses were performed using EnrichR. The network was generated using Ingenuity Pathway Analysis (QIAGEN, Inc).