Chronic obstructive pulmonary disease (COPD) is the fourth leading cause of morbidity and mortality in developed countries. Pulmonary artery hypertension (PAH), frequently observed in patients with advanced COPD, is considered a predictor of worse outcome 12. PAH in COPD is mainly the result of hypoxic pulmonary vasoconstriction 3, however it has been recently shown also induced by pulmonary arterial walls remodelling, with aberrant intimal changes 45. Further evidences suggest that the initial event in the natural history of pulmonary hypertension due to COPD is the lesion of pulmonary endothelium by cigarette-smoke products 67, leading to critical changes in the expression of vascular mediators 78, which are responsible for impairment of endothelial function. Among a wide spectrum of vasoactive mediators, endothelin-1 (ET-1) plays a pivotal role in exerting vasoconstriction 9, bronchoconstriction 10, vascular 11 and airway 12 cells proliferation, via ET_A and ET_B receptors 1314. Increased expression of ET-1, it has been identified in the vessels of patients with IPAH 15. Plasma 1617, induced sputum 18, and urinary 17 levels of ET-1 are augmented in patients with COPD; however, there are still contrasting data about circulating ET-1 levels in COPD patients with PAH 192021. In addition, no data are yet available on ET-1 levels in exhaled breath condensate of COPD patients.

Nitric oxide (NO), produced by endothelial cells, is the central stimulus for releasing and dilating pulmonary arterial vasculature 22. NO production is oxygen dependent and lack of NO synthesis, under hypoxic conditions such as COPD, contributes to chronic hypoxic pulmonary vasoconstriction 23, leading to pulmonary artery hypertension 24. It is now appreciated that NO release is impaired in the pulmonary vasculature of COPD patients 25, while exhaled NO levels, increased in COPD 26, are significantly reduced in COPD associated with PAH 27. An imbalance between the excretion of vascular mediators has been shown important for developing idiopathic pulmonary hypertension 8 and might be also responsible in promoting secondary PAH 8. We hypothesized a pathological dysregulation of exhaled output between ET-1 and NO in COPD with PAH. Despite several issues about reproducibility, variability and sensitivity, measurement of different biological markers in COPD is currently a useful assay in assessing disease pathogenesis, predicting progression, and monitoring adequate therapies 28. The aims of the present study were to investigate exhaled breath condensate (EBC) and circulating levels of ET-1, as well as the exhaled NO (FENO) in COPD only, in COPD with pulmonary artery hypertension, and in a group of healthy subjects. We also examined whether concentrations of exhaled biomarkers, or circulating ET-1 levels were related to the severity of disease, as defined by lung function, or pulmonary artery pressure.

The Institutional Review and the Ethical Boards of the University of Bari approved protocols and all patients signed informed consent before participating in this study.

Demographic and clinical characteristics of 63 individuals recruited are presented in table 1. Endothelin-1 was detectable in the exhaled breath as well as in the blood of all subjects. The reproducibility of repeated ET-1 measurements was quite good, and was assessed by the Bland and Altman method 32 in six patients (mean difference 0.62 ± 1.214). NO was detectable in the exhaled of all individuals. All COPD patients were investigated with Echo-Doppler and then assigned to one of two groups according to the level of PaPs, using a cut-off of 35 mmHg. Those with a PaPs ≥ 35 mmHg (Group 1: N = 12; FEV₁% 39.35 ± 11.45; PaPs 51.08 ± 12.21 mmHg), and those with a PaPs < 35 mmHg (Group 2: N = 36; FEV₁% 48.13 ± 18.04; PaPs 26.05 ± 3.39 mmHg).

In the natural history of COPD, right heart failure leading to increased pulmonary artery pressure is a common finding 3536. PAH, in this scenario, is likely due to a primary hypoxic vasoconstriction associated to the direct toxic effect of tobacco smoke into the intrapulmonary vessels, with abnormal production of substances that control vasoconstriction, vasodilatation, and vascular cell proliferation, ultimately leading to extensive pulmonary artery remodelling 237. Early detection of an imbalanced output between essential mediators, such as endothelin-1 and NO, might be a novel approach to rapidly investigate and eventually treat this critical disease. The present study was designed to investigate endothelin-1 and NO concentration in the exhaled breath, as well as circulating levels of ET-1 in COPD patients with or without pulmonary hypertension. Our findings show that endothelin-1 levels in EBC are increased in all COPD patients. In addition, we observed ET-1 levels in EBC significantly higher in COPD associated with PAH, compared to COPD only. Several previous studies detected elevated levels of ET-1 in EBC in interstitial pulmonary disease 38, in non small cell lung cancer patients (NSCLC) 39, and in unstable bronchial asthma 40, similar to those observed in our COPD patients, while in the present, we found significantly higher EBC levels of ET-1 in COPD with PAH. Carpagnano and colleagues hypothesized that ET-1, in pathological conditions (interstitial pulmonary disease and NSCLC), could be in part produced by airway epithelium itself 3839, while the reason of an exaggerated ET-1 release in EBC in COPD with PAH is still unknown. However, it appears that ET-1, primary produced by the pulmonary endothelium, could be in part generated by the epithelium, and, in addition, could be partially the result of an uptake into the airway epithelium. Furthermore, we found elevated circulating levels of ET-1 in arterial and venous samples in COPD, and significantly higher levels in COPD with PAH. Increased circulating ET-1 levels have been recently shown in exacerbation of COPD 41, whereas there are still controversial data about ET-1 levels in the blood of patients with PAH associated with COPD or emphysema 192021. While a study disclosed increased levels of ET-1 only in pulmonary arterial sample 19, two reports showed persistent increased circulating ET-1 levels, which correlated to the severity of pulmonary hypertension 2021. In our group of COPD patients with PAH, ET-1 levels in arterial sample positively correlated with PaPs, however we provide new evidence that also ET-1 in EBC is related to PaPs. Interestingly, in the same group, we observed a negative correlation between ET-1 levels in arterial blood and FEV₁%, but not between ET-1 levels in EBC and FEV₁%. This finding sustains the previous hypothesis that pulmonary endothelium represents the mayor source of ET-1, which could be only in part produced by the airway epithelium, where it may contribute to induce partial bronchoconstriction, although the latter observation needs to be validated by further histological evidences. As previously demonstrated 42, an inverse correlation between PaPs and PaO₂ was observed in this group, confirming that pulmonary artery pressure and PaO₂ are closely negatively related in PAH induced by COPD. Regarding to exhaled NO, we found reduced levels of FENO in COPD with PAH compared to the group of COPD. Clini and co-workers previously showed reduced levels of FENO in PAH induced by COPD compared to COPD only, which negatively correlated with PaPs 27. In our experience, FENO levels were also significantly lower in these patients, but we failed to find correlation with Echo-Doppler assessment. In summary, we observed a pathological dysregulation output in the breath between ET-1 and NO in COPD with PAH. Whether the imbalance in the release of these two mediators is an early cause or result of pulmonary hypertension due to COPD is unknown, however the easy and quick detection by a non-invasive and useful assay might encourage further investigations.

This study has different limitations, likely due to methodological assays. First, the control group consisted by healthy non smokers, and not by ex-smokers with no airflow obstruction, that would have been represented a better control group. However our COPD patients were all non smokers from at least 1 year, we consider that smoking habit might not have significantly influenced exhaled breath results. Second, in COPD patients, the pulmonary arterial pressure was evaluated by echocardiography analysis, which does not correspond to the gold standard; however in 4 patients was previously performed a right heart catheterization and the results were comparable with the echo Doppler assessment. Moreover, although no COPD patient was under steroid treatment, we could not exclude that other treatments, such as antihypertensive, nitro-derivate, or other drugs, might have influenced the results. Finally, no significant difference in the treatment between COPD and COPD with pulmonary hypertension was evident.

In conclusion, we found increased expression of ET-1 in EBC and in the blood, in COPD patients with PAH compared to COPD only. ET-1 levels, both in EBC and in the arterial samples, positively correlated to pulmonary-artery systolic pressure, while arterial concentration of ET-1 negatively correlated with FEV₁%, indicating that enhanced ET-1 levels in biological sample might early occur in PAH secondary to COPD, and actively participate in stimulating vascular and airway remodelling, through an endogenous increased production by the endothelium, and by a partial release and uptake into the airway epithelium. Finally, we showed an imbalance output in the breath between ET-1 and NO in this group of patients, suggesting that suppression of NO, in pulmonary hypertension, might have been caused in part by ET-1 43. Although the novel treatment for pulmonary artery hypertension has been used successfully for IPAH, it remains to be investigated with randomized multicentric studies whether this new therapy could provide beneficial clinical effects in PAH secondary to COPD.

ATS: American Thoracic Society; COPD: Chronic Obstructive Pulmonary Disease; D_LCO: Carbon monoxide diffusion capacity; EBC: Exhaled Breath Condensate; ERS: European Respiratory Society; ET-1: Endothelin-1; FENO: Exhaled Nitric Oxide; FEV₁: Forced Expiratory Volume in the first second; FVC: Forced Vital Capacity; HS: Healthy Subjects; NO: Nitric Oxide; PAH: Pulmonary Artery Hypertension; IPAH: Idiopathic Pulmonary Artery Hypertension; PaPs: Pulmonary-artery systolic pressure; ppb: parts per billion; RV: Residual Volume; 6MWD: 6-minutes walking distance.

The authors declare that they have no competing interests.

PC designed the study, carried out the laboratory research and the patients' characterisation for the classification of the different patient groups, and wrote the manuscript. CS designed the study, recruited the patients, coordinated the study, and assisted in performing the statistical analysis. MM performed the statistical analysis, participated in interpretation of results, in writing the manuscript, and critically reviewed the manuscript. TSM assisted in the patients' characterisation, performed laboratory analysis and interpretation of results. GDG recruited the patients and helped with study design. FG assisted in the patients' characterisation, in statistical analysis, and interpretation of results. RAC performed the statistical analysis and participated in reviewing the manuscript. SD assisted in the patients' characterisation and in statistical analysis. PP recruited the patients and helped with study design. AS participated in interpretation of results and critically reviewed the manuscript. OR conceived and supervised the study as head of the lung research group, participated in its design and coordination and revised the manuscript. All authors read and approved the final manuscript.

The authors would like to thank Gaia Iacoviello MD, Fabio Cardinale MD, and G. Elisiana Carpagnano MD, for technical support