Non-invasive biomarkers and pulmonary function in smokers

Abstract

Limited information exists regarding measurement, reproducibility and interrelationships of non-invasive biomarkers in smokers. We compared exhaled breath condensate (EBC) leukotriene B4 (LTB4) and 8-isoprostane, exhaled nitric oxide, induced sputum, spirometry, plethysmography, impulse oscillometry and methacholine reactivity in 18 smokers and 10 non-smokers. We assessed the relationships between these measurements and within-subject reproducibility of EBC biomarkers in smokers. Compared to non-smokers, smokers had significantly lower MMEF % predicted (mean 64.1 vs 77.7, p = 0.003), FEV1/FVC (mean 76.2 vs 79.8 p = 0.05), specific conductance (geometric mean 1.2 vs 1.6, p = 0.02), higher resonant frequency (mean 15.5 vs 9.9, p = 0.01) and higher EBC 8-isoprostane (geometric mean 49.9 vs 8.9 pg/ml p = 0.001). Median EBC pH values were similar, but a subgroup of smokers had airway acidification (pH < 7.2) not observed in non-smokers. Smokers had predominant sputum neutrophilia (mean 68.5%). Repeated EBC measurements showed no significant differences between group means, but Bland Altman analysis showed large individual variability. EBC 8-isoprostane correlated with EBC LTB4 (r = 0.78, p = 0.0001). Sputum supernatant IL-8 correlated with total neutrophil count per gram of sputum (r = 0.52, p = 0.04) and with EBC pH (r = -0.59, p = 0.02). In conclusion, smokers had evidence of small airway dysfunction, increased airway resistance, reduced lung compliance, airway neutrophilia and oxidative stress.

Figure 1

(A) Natural log (Ln) sGaw in non-smokers and smokers. Non-smokers (n = 10) and smokers (n = 18). Bars represent geometric means, t test used for comparison. (B) Resonant frequency (RF) in non-smokers and smokers. Non-smokers (n = 10) and smokers (n = 18). Bars represent means, t test used for comparison.

Figure 2

Natural log (Ln) methacholine challenge PD20 in non-smokers and smokers. Non-smokers (n = 10) and smokers (n = 15). Bars represent geometric means, t test used for comparison.

Figure 3

(A) Natural log (Ln) exhaled breath condensate 8-isoprostane concentration in non-smokers and smokers. Non-smokers (n = 10) and smokers (n = 18). Bars indicate geometric mean values, t test used for comparison. (B) Exhaled breath condensate pH in non-smokers and smokers. Non-smokers (n = 10) and smokers (n = 17). Bars indicate median values, Mann-Whitney U test used for comparison.

Figure 4

(A) Bland Altman plot of EBC pH reproducibility between visit 1 and visit 2 (n = 14). (B) Bland Altman plot of EBC 8-isoprostane reproducibility between visit 1 and visit 2 (n = 18). (C) Bland Altman plot of EBC LTB₄ reproducibility between visit 1 and visit 2 (n = 18).

Figure 5

(A) Pearson’s correlation in smokers between natural log (Ln) exhaled breath condensate 8-isoprostane and LTB4 (n = 18). (B) Spearman’s correlation in smokers between sputum supernatant IL-8 (pg/ml) and EBC pH (n = 15). (C) Pearson’s correlation in smokers between sputum supernatant IL-8 (pg/ml) and natural log (Ln) sputum total neutrophil count (n = 15).