Figures
- FIGURE 1
Differential gene expression of sputum RNA-sequencing data. a) Volcano-plot of differentially expressed genes concerning time to exacerbation after inhaled corticosteroid (ICS) withdrawal (false discovery rate <0.1). Blue and red dots represent differentially expressed genes with negative and positive log2 fold change, respectively. b) Heat map of differentially expressed genes. Subjects were clustered according to time to exacerbation (days; log2 transformed). c–e) Spearman correlation testing comparing the enrichment score of PRISE #1 with the c) time to exacerbation, d) sputum eosinophil counts (log10), and e) macrophage cell counts (log10).
- FIGURE 2
Cumulative hazard plots of risk of exacerbations. PRISE #1 enrichment scores are stratified by tertiles. The first tertile served as the reference for second and third tertiles. HR: hazard ratio. *: p<0.05.
- FIGURE 3
Bayesian network model predicting exacerbation phenotype. The binary exacerbation phenotype is based on “time to exacerbation” (dichotomised (> or < mean) into early (n=20) versus late/nonexacerbators (n=23)). Colour coding represents genes that were part of PRISE #1 and #2.
Tables
- TABLE 1
Baseline participant clinical and demographic characteristics
Number of patients 43 Current/former smokers 23/20 Number of exacerbations 38 Males/females 36/7 Number of exacerbations in the past 12 months) 0 19 1 17 ≥2 7 Age (years) 64.5 (59.4–70.8) Smoking history (pack-years) 38.0 (25.5–53.0) Body mass index (kg·m−2) 25.9 (23.6–28.0) Daily dose of ICS (µg)# 800 (500–1000) Pre-study treatment regimen ICS¶ 8 ICS+LABA¶ 30 ICS+LAMA¶ 2 ICS+LAMA+LABA¶ 3 Post-bronchodilator lung function FEV1 (% pred) 64.1 (52.0–76.1) FEV1/FVC 0.52 (0.44–0.56) TLC (% pred) 112.1 (105.9–122.0) RV (% pred) 135.5 (124.9–158.5) RV/TLC (% pred) 114.9 (103.4–127.4) Baseline inflammatory characteristics Sputum cells Eosinophils (%) 1.8 (0.7–4.0) Neutrophils (%) 72.8 (67.0–80.3) Macrophages (%) 18.0 (13.4–25.3) Lymphocytes (%) 0.7 (0.1–1.2) Bronchial epithelial cells (%) 2.0 (0.65–3.55) Sputum supernatant proteins ECP (µg·L−1) 194.0 (58.6–718.0) LTB4 (ng·mL−1) 0.49 (0.28–1.21) Data are presented as n or median (interquartile range). ICS: inhaled corticosteroid; LABA: long-acting β-agonist; LAMA: long-acting muscarinic antagonist; FEV1: forced expiratory volume in 1 s; FVC: forced vital capacity; TLC: total lung capacity; RV: residual volume; ECP: eosinophilic cationic protein; LTB4: leukotriene-B4. #: daily dose of ICS calculated for indexed budesonide equivalent; ¶: additional short-acting β-agonist and/or short-acting muscarinic antagonist as needed.
- TABLE 2
Monovariate Cox regression hazard ratios
Hazard ratio (95% CI) p-value Generalised R2 History of exacerbations 1.5 (1–2.1) 0.046* 0.08 Season of ICS withdrawal (not in November, December or January#) 0.3 (0.13–0.68) 0.004* 0.152 Sex (male) 0.31 (0.13–0.78) 0.012* 0.12 Smoking status (current smoker) 0.68 (0.36–1.30) 0.24 0.03 Pack-years smoking ≥38¶ 0.49 (0.25–0.96) 0.038* 0.095 Sputum eosinophils ≥3%+ 2.3 (1.1–4.6) 0.021* 0.11 % Sputum eosinophils (stratified by tertiles) Second tertile (N=15) 1.4 (0.62–3.2) 0.42 0.124 Third tertile (N=15) 2.8 (1.17–6.5) 0.02* PRISE #1 (ES stratified by tertiles) Second tertile§ (N=13) 2.3 (0.99–5.2) 0.052 0.202 Third tertile§ (N=15) 3.7 (1.57–8.5) 0.003* ICS: inhaled corticosteroid; ES: enrichment score. #: dichotomised as “outside” versus “in” the period; ¶: dichotomised using the median value at baseline; +: dichotomised ≥3% (N=15), <3% (N=28). §: compared to the first tertile. *: p<0.05.
- TABLE 3
Multiple linear Cox regression including either sputum eosinophils or PRISE #1
Hazard ratio (95% CI) p-value Generalised R2 Sputum eosinophils History of exacerbations 1.40 (0.97–2.03) 0.071 0.377 Sex (male) 0.33 (0.13–0.85) 0.022* Season of ICS withdrawal (not in November, December, or January#) 0.36 (0.15–0.88) 0.025* Pack-years smoking ≥38¶ 0.70 (0.3301.46) 0.337 Sputum eosinophils ≥3%+ 2.24 (1.09–4.58) 0.027* PRISE #1 History of exacerbations 1.53 (1.03–2.28) 0.035 0.495 Sex (male) 0.47 (0.17–1.28) 0.14 Season of ICS withdrawal (not in November, December, or January#) 0.27 (0.10–0.69) 0.006* Pack-years smoking ≥38¶ 0.50 (0.22–1.15) 0.105 PRISE #1 (ES stratified by tertiles) Second tertile (N=13) 2.59 (1.11–6.05) 0.027* Third tertile (N=15) 5.35 (2.12–13.47) <0.001* ICS: inhaled corticosteroid; ES: enrichment score. #: dichotomised as “outside” versus “in” the period; ¶: dichotomised using the median value at baseline; +: dichotomised as ≥3% (N=15) or <3% (N=28). *: p<0.05.
- TABLE 4
ICS sensitivity of PRISE #1 and #2 genes in SYMBEXCO and independent RNA-sequencing dataset
Gene SYMBEXCO: ICS treatment → no ICS (8 weeks) Independent RNA-seq dataset: no ICS → ICS treatment (8 weeks) Log2 fold change Nominal p-value Log2 fold change Nominal p-value CLC (PRISE #1) 0.76 <0.01* −2.88 0.01* LGALS12 (PRISE #1) 0.52 0.01* −1.52 0.01* CD24 (PRISE #1) 0.30 0.04* NA NA ALOX15 (PRISE #1) 0.42 0.04* −1.17 <0.01* EMR4P (PRISE #1) 0.22 0.13 −0.63 0.16 IL1RL1 (PRISE #1) 0.15 0.36 −1.32 0.03* ALG3 (PRISE #2) −0.02 0.85 −0.49 0.02* CCDC152 (PRISE #2) −0.02 0.91 0.27 0.16 SEPP1 (PRISE #2) 0.01 0.96 −0.22 0.08 ICS: inhaled corticosteroid. *: p<0.05.
Supplementary Material
Please note: supplementary material is not edited by the Editorial Office, and is uploaded as it has been supplied by the author.
Supplementary material 00097-2021.SUPPLEMENT
