Introduction

Chronic obstructive pulmonary disease (COPD), which is one of the leading causes of mortality and morbidity in the world, is often complicated by pulmonary hypertension (PH)1,2. The pulmonary hypertension in COPD is associated with an increase in mortality, the risk of repeated hospitalizations, and the increase in health care costs3,4. Therefore, early detection of PH is important for the management of patients with COPD.

Brain natriuretic peptide (BNP) and its N-terminal fragment (NT-proBNP) are the only biomarkers recommended by current guidelines for risk stratification of patients with PH5. Elevated blood levels of BNP or NT-proBNP in patients with COPD reflect the severity of PH and can be used as markers of PH in COPD, even at early disease stages6. NT-proBNP level is not only a reliable predictor of survival in PH, but also a good marker of treatment efficacy7.

C-type natriuretic peptide (CNP) is another promising biomarker for PH8. CNP is widely expressed in various tissues and the principal source of CNP is the vascular endothelium8,9. A significant 3-fold increase in CNP plasma levels was reported in COPD patients with cor pulmonale compared with age-matched controls10. Chronic arterial hypoxemia is believed to cause damage to the endothelium, resulting in more CNP leakage into the plasma11,12,13. Due to a short half-life, CNP plasma levels may not correspond to its concentration near the secretion site. At the same time, its N-terminal fragment (NT-proCNP) circulating in human blood in equimolar concentrations with CNP is considered a more reliable marker of CNP biosynthesis14.

Currently, NT-proCNP as a marker of pulmonary hypertension in COPD remains poorly evaluated. The objective of this study is to assess the diagnostic value of NT-proCNP in COPD patients with pulmonary hypertension.

Methods

The study was performed in accordance with the declaration of Helsinki and the subsequent revisions. The study was approved by the Medical Ethical Committee of the Pulmonology Research Institute, Moscow (protocol №3-02-17), and written informed consent was obtained from all patients.

Participants

Patients were included in the study if they were ≥ 40 years, had a confirmed diagnosis of COPD, ratio of the forced expiratory volume in one second (FEV1) to forced vital capacity (FVC) (FEV1/FVC) < 70% and an FEV1 ≤ 80% predicted.

Patients were excluded if they had left ventricular dysfunction, history of pulmonary embolism and chronic thromboembolic pulmonary hypertension, interstitial lung diseases, myocardial infarction, chronic renal impairment or an abnormal serum creatinine on admission (normal was defined as < 120 mmol/L), and body mass index (BMI) ≥ 35 m/kg2. We excluded patients if they were unable to perform spirometry. Subjects with sepsis, systemic inflammatory diseases, liver or neoplastic diseases or any other serious medical condition that would prevent their participation in the trial were also excluded.

All control subjects were staff of our Institute, non-smokers without known pathology of the respiratory and cardiovascular systems and were subject to the same exclusion criteria as COPD patients. Control subjects were approached individually and asked to participate. All control subjects gave written informed consent but did not undergo echocardiography. Control subjects (twelve males and six females) and COPD patients were well age-matched (controls: 62.7 ± 5.8 years).

Echocardiography

Echocardiography was performed using a digital echocardiographic equipment (En Visor CHD; Philips, Holland) equipped with a 2-MHz and 3-MHz probes by two experienced physicians (VVG and GSN). Tricuspid regurgitation flow was identified by color flow Doppler techniques, and the maximal jet velocity was recorded from the parasternal or apical window with a continuous wave Doppler echocardiographic probe. The trans-tricuspid pressure gradient (ΔP) was calculated from the maximum velocity of the tricuspid regurgitant jets (Vmax) using the simplified form of the Bernoulli equation: ΔP (mmHg) = V2 × 415. An estimation of systolic pulmonary artery pressure (SPAP) was obtained by adding an estimate of the right atrial pressure to ΔP16. Mean pulmonary artery pressure (MPAP) was calculated according a formula: MPAP = SPAP × 0.61 + 2 mmHg.

Pulmonary hypertension was defined as a calculated SPAP ≥ 40 mmHg and we also evaluated a subgroup of patients with severe PH defined as a calculated SPAP ≥ 55 mmHg.

N-Terminal pro C-Type natriuretic peptide

Venous blood samples were collected for NT-proCNP testing into tubes containing potassium EDTA. The NT-proCNP samples were analysed within 4 h or, in some cases, whole blood was centrifuged and the plasma stored at -800C until analysis. The NT-proCNP plasma concentrations were quantified using the Biomedica kit (Medizinprodukte GmbH and Co KG, A-1210, Austria). This NT-proCNP assay is a sandwich immunoassay consisting of a disposable device to which 250 mL of EDTA-anticoagulated whole blood or plasma is added.

Other measurements

Patients performed three acceptable forced expiratory measurements (Master Screen Body, Erich Jaeger, Germany), recorded sitting in an upright position and wearing a nose-clip. We recorded post-bronchodilation FEV1, FVC and their ratio (FEV1/FVC). The highest values were recorded, whether or not from the same attempt. The predicted values of parameters were calculated according to the ERS specification17.

The Modified Medical Research Council (mMRC) questionnaire and the COPD Assessment Test (CAT) were used to assess the severity of COPD symptoms18,19. Physical exercise tolerance was evaluated using the 6-minute walk test (6MWT)20.

Statistics

Statistical analyses were performed using SPSS version 22.0 (SPSS Inc., Chicago, IL, USA) and STATISTICA version 7.0 (Statsoft Inc., USA). All data were stated as mean ± standard deviation (SD) or median (25th to 75th percentiles). Normality of variables was evaluated with the Shapiro-Wilk test. Data conforming to a normal distribution were analyzed using an independent-samples t-test, while non-normally distributed data were analyzed with the Mann-Whitney U test. The correlations between variables were concluded by Pearson’s correlation test. Receiver-operating characteristic (ROC) curve analysis was performed for plasma NT-proCNP levels. Differences were judged to be significant at a probability level of P < 0.05.

Results

A total of 94 patients (84 men and 10 women) with stage II-IV COPD according to GOLD classification participated in the study. The patients had an age of 63.5 ± 7.5 years, a smoking history of 25.2 ± 6.9 pack-years, a disease duration of 13.2 ± 6.1 years, mean FEV1 of 33.7 ± 10.2% predicted, and a CAT score of 28.0 ± 6.2 points (Table 1).

Table 1 Clinical parameters of COPD patients participating in the study.
Full size table

The studied patients with and without PH did not differ in age, duration of the disease and frequency of exacerbations. However, the patients with PH had more severe impairment in FEV1 (p = 0.001), mMRC score (p = 0.002), CAT score (p < 0.001), less distance walked in 6-MWT (p = 0.005) and higher levels of С-reactive protein (CRP) and fibrinogen (p < 0.001) (Table 1).

The mean estimated SPAP level was 42.4 ± 13.9 mmHg in the whole group of patients with COPD, 29.4 ± 4.7 mmHg in patients without PH, and 52.9 ± 8.9 mmHg in patients with PH (p < 0.001).

Plasma NT-proCNP levels in the patients with COPD (2.65 ± 0.33 pg/mL) were significantly (p < 0.001) higher than the reference values in control group (0.24 ± 0.22 pg/mL). Statistically significant differences in plasma NT-proCNP levels were found in patients with and without PH (1.42 ± 0.35 pg/mL vs. 3.63 ± 0.95 pg/mL, p < 0.001) and also in patients with non-severe and severe PH (3.31 ± 0,65 pg/mL vs. 4.50 ± 1.15 pg/mL, p = 0.009) (Fig. 1).

Fig. 1
figure 1

Plasma NT-proCNP levels in controls and patients with COPD according to their systolic pulmonary artery pressure.

Full size image

Statistically significant correlations were shown between plasma NT-proCNP levels and a number of selected clinical, laboratory and functional parameters, including SPAP (r = 0.77, p < 0.001), mMRC score (r = 0.42, p = 0.001), CAT score (r = 0.60, p < 0.001), SpO2 (r = − 0.47, p = 0.001), CRP (r = 0.46, p = 0.001) and fibrinogen (r = 0.45, p = 0.001) (Table 2).

Table 2 Correlations between selected clinical, laboratory and functional data and NT-proCNP levels in COPD patients.
Full size table

A receiver-operating characteristic curve (ROC) for plasma NTproCNP was performed for diagnosis of PH in COPD patients. The area under the curve (AUC) of plasma NT-proCNP for the prediction of PH was 0.82 (95% confidence interval [CI] 0.74–0.91, p = 0.001) (Fig. 2). The optimal threshold value when maximizing specificity and positive predictive value was estimated as 2.16 pg/mL, sensitivity was 84.6% and specificity was 81.0%.

Fig. 2
figure 2

A receiver operator characteristic curve of plasma NT-proCNP as a predictor of PH in COPD patients.

Full size image

Discussion

Our study demonstrated that patients with COPD and pulmonary hypertension had higher plasma NT-proCNP levels compared to those without pulmonary hypertension. In patients with COPD, plasma NT-proCNP levels correlated significantly with systolic pulmonary artery pressure. Significant correlations were also shown between plasma NT-proCNP levels and a number of key clinical and physiological parameters, including mMRC score and CAT score, SpO2, C-reactive protein and fibrinogen levels. ROC curve also demonstrated the high diagnostic value of NT-proCNP level for the diagnosis of pulmonary hypertension in COPD patients.

While BNP and NT-proBNP have been examined extensively in several types of pulmonary hypertension, only limited data exists on the role of CNP and NT-proCNP as diagnostic biomarkers for PH10,14.

Cargill et al. demonstrated that in 16 patients with COPD and cor pulmonale the plasma CNP concentration was significantly increased compared with young or elderly controls and with patients with congestive heart failure10. The COPD in patients included in this study was quite severe (mean FEV1 30.2% (range 15–42%), mean PaO2 on air 6.09 (range 3.95–7.49) kPa), but no hemodynamic data of patients were presented.

In another study Kaiser et al. examined plasma levels of NT-proCNP and other subtypes of natriuretic peptides in 62 patients with precapillary PH14. Patients with PH had higher values of circulating NT-proCNP, but the difference did not reach statistical significance, and in the ROC-analysis AUC for NT-proCNP was 0.70. The patients with PH involved into the study of Kaiser et al. were quite different from our patients as they had heterogeneous types of PH and some patients also had renal impairment.

The mechanism underlying high plasma NT-proCNP levels in patients with COPD is not well understood. Natriuretic peptides are released in response to hemodynamic stress and help regulate intravascular volume homeostasis8,11. According to literature data, the expression and secretion of CNP is regulated by various cytokines and growth factors involved in vascular remodeling and inflammation. Among them are tumor necrosis factor-α, lipopolysaccharides, the main fibroblast growth factor, interleukin-1, transforming growth factor beta and thrombin14,21,22. CNP secretion, which is increased in response to shear stress23,24 has an anti-remodeling effect on vascular smooth muscles and endothelial cells25.

Endothelial CNP is a useful biomarker for cardiovascular diseases26. In particular, its elevated urine level in acute decompensated heart failure is a predictor of an unfavorable outcome26,27. CNP regulates the platelet and leukocyte reactivity28the growth of endothelial and vascular smooth muscle cells29,30as well as cardiac hypertrophy and myocardial damage during ischemia-reperfusion31,32,33.

An increased CNP level has been observed in chronic renal failure34 in sepsis22after angioplasty35 in atherosclerotic aortic stenosis36 in chronic heart failure and diabetic cardiomyopathy in a genetic mouse model37. In animal models, the inflammatory cytokines, such as IL-1, TNF-α and endotoxins, are able to release CNP from endothelial cells and thereby affect the local vascular tone38.

This study has some limitations. First, the relatively small number of patients involved in the present study is of concern in the proposal of a specific cut-off level for use in the detection of PH in clinical practice. Second, SPAP was evaluated with echocardiography instead of pulmonary artery catheterization. But the current clinical practice for the diagnosis and assessment of the pulmonary hypertension associated with chronic pulmonary diseases is based on the noninvasive methods such as echocardiography3. Third, we examined the value of a new biomarker in a relatively narrow population of PH – only patients with COPD without renal impairment were included in our study. And finally, we did not verify the outcome of subjects with high plasma NT-proCNP levels. This means that the efficacy of early diagnosis and intervention based on measurement of plasma NT-proCNP levels is still unknown.

Prospective larger studies with follow-up observation are required to further delineate the importance of plasma NT-proCNP, and therefore this marker can be more widely used in clinical practice. Despite these limitations, our study is the first attempt to evaluate the diagnostic role of NT-proCNP levels in patients with PH associated with COPD.

Conclusions

Our study demonstrated that pulmonary hypertension in patients with COPD was associated with higher plasma NT-proCNP level. In addition, plasma levels of NT-proCNP were significantly correlated with systolic pulmonary artery pressure in COPD patients. The plasma NT-proCNP appears to be a promising biomarker for the management of patients with COPD. NT-proCNP levels may be used to guide therapy in the conditions associated with PH. This pilot study suggests that further work is required to establish the role of NT-proCNP in the pulmonary hypertension associated with chronic pulmonary diseases.