Serum H 2 S as an indicator of exacerbation and pulmonary arterial hypertension in chronic obstructive pulmonary disease

Background and objective: Hydrogen sulfide (H 2 S) is a potential gasotransmitter produced by respiratory and vascular smooth muscle cells. Although some studies have examined H 2 S in chronic obstructive pulmonary disease (COPD), few have investigated its role in COPD exacerbations and pulmonary arterial hypertension (PAH). This study aimed to compare serum H 2 S levels between COPD patients and healthy controls, examine associations between H 2 S and COPD severity/exacerbations/PAH, and compare H 2 S levels between smokers and nonsmokers. Methods: Serum H 2 S was measured in 16 patients with stable COPD and 34 with acute exacerbations. COPD severity was classified using GOLD stages. Arterial blood gases, pulmonary arterial pressure by echocardiography, and clinical variables were assessed. Multivariable regression analyzed factors influencing H 2 S. ROC curves evaluated the diagnostic utility of H 2 S for exacerbations and PAH. Results: Serum H 2 S was lowest in GOLD stage IV patients compared to stages II and III. Levels were significantly lower in acute exacerbations versus stable COPD. COPD patients with PAH had lower H 2 S than those without. Arterial pH, FEV1, and FEV1/FVC positively associated with H 2 S, while Pa CO 2 , severity, exacerbations, and PAH negatively influenced H2S. Optimal H2S cutoffs for indicating exacerbations and PAH were <46.7 µ mol/L. Conclusions: Serum H 2 S may be a useful indicator of exacerbations and PAH in COPD patients. ABSTRACT


INTRODUCTION
Chronic obstructive pulmonary disease (COPD) is a progressive inflammatory disease of the lungs characterized by dynamic hyperinflation, parenchymal damage and irreversible airflow limitation. 1 According to the World Health Organization global burden of disease studies, COPD ranks third in terms of mortality and morbidity globally, and is expected to step up to the second leading cause of mortality by 2030. 2 Episodes of exacerbation are one of the major physiological manifestations of COPD which is often associated with disease severity, cardiac comorbidity, impaired quality of life and mortality. 3Pulmonary arterial hypertension (PAH) is a comorbidity in COPD that is observed among patients in advanced stages of COPD, presumably when the gas exchange is severely compromised.In addition to that, hypoxic vasoconstriction, collagen deposition, remodeling of the pulmonary arteries and emphysematous changes in the capillary bed eventually lead to pulmonary hypertension and right ventricular dysfunction. 4,5 recent years, a number of molecules have been investigated to play crucial role in COPD pathophysiology.Hydrogen sulfide (H2S) is one of those molecules.H2S is a gas with a typical rotten egg smell, and it is a potential gasotransmitter synthesized in the pulmonary arterial and airway smooth muscle cells, endothelial cells and primary fibroblast by cystathionine-γ-lyase (CSE), cystathionine-β-synthase (CBS) and 3-mercaptopyruvate sulfur transferase in mammals. 68][9] In clinical studies, some researchers demonstrated higher levels of serum H2S in COPD in compared to healthy non-smokers although H2S level was reduced in patients with acute exacerbation. 3,10The levels of serum H2S have also been tested in certain cardiovascular conditions associated with COPD. 11Although there is only one study that demonstrated a lower serum H2S among the COPD patients with pulmonary arterial systolic pressure ≥35mmHg. 10No other study describes serum H2S levels in pulmonary arterial hypertension.
In this study, we aimed at investigating the serum H2S level in patients with COPD and COPD patients with PAH.We also evaluated whether serum H2S level could indicate the occurrence of exacerbation and PAH in COPD.

Study participants
Fifty COPD patients attending the indoor and outdoor specialty clinics of the Department of Respiratory Medicine of the NRS Medical College and Hospital in Kolkata were recruited for the study.Among the 50 patients, 23 had GOLD Stage 4 disease, 13 patients had stage 3, and 14 patients had Stage 2 disease.Among the cases, 44 were smokers and 4 were exposed to ETS.Parameters were matched with age-and sex-matched apparently healthy controls selected from relatives of the patients and hospital employees.The study was conducted over a period of one year from April 2014 to March 2015.Among the controls, 13 were smokers and 37 were non-smokers.

Inclusion and exclusion criteria
Patients were excluded if they a) had clinical, laboratory or radiological evidence of pneumonia, pulmonary tuberculosis, bronchiectasis, bronchogenic carcinoma, asthma, interstitial lung disease or pregnancy; b) were on H2S donor or H2S inhibitor drugs; or c) did not provide informed consent.
After all exclusions, 50 patients were recruited for the study.The diagnosis of COPD was made according to the Global Initiative for Chronic Obstructive Lung Disease (GOLD) guidelines. 12tients with a history of chronic respiratory symptoms such as dyspnea, coughing, sputum production or both, a smoking history of >10 pack-years, and a post-bronchodilator FEV1/FVC ratio <70% were recruited as cases.Disease severity of the patients was assessed by GOLD stages I to IV. Patients were categorized as stable COPD (S-COPD) if they did not have any symptoms of exacerbation in the last 4 weeks, or with acute exacerbation (AECOPD) if they reported new or worsening respiratory symptoms (cough, dyspnea and/or sputum production) for two or more consecutive days. 13

Sample size
The average patient volume of COPD in outpatient departments and those admitted with AECOPD was calculated for the past 3 years, and since the prevalence of COPD is 9.23% in this region of India, the sample size was calculated with 10% allowable error.Due to funding and time constraints, a larger sample size could not be obtained.We planned to work with a bigger sample in the future over a longer timeframe.

Echocardiography
Doppler echocardiography (DE) was performed to measure the pulmonary arterial systolic pressure (sPAP) in a non-invasive way by using conventional echocardiographic equipment (GE Healthcare, Model: Vivid T8 and P3, USA) with 3Hz transducer mode.Pericardium and cardiac functions including the valvular anatomy and function and chamber size, were assessed from the echocardiograms.Tricuspid regurgitant (TR) flow was identified by color flow Doppler procedure and the TR jet was obtained in Parasternal longaxis/RV inflow, parasternal short axis at the level of aortic valve, apical four-chamber, and subcostal views.The measurement of modal peak velocity was obtained by maximum TR jet velocity where the Doppler envelop was complete.[16][17]

Arterial blood gas analysis
Arterial blood gas samples were analyzed in an ABL90 Flex blood analyzer (Radiometer, Bengaluru, India) according to established guidelines. 18The pH, partial pressure of oxygen (PaO2) and carbon dioxide (PaCO2) were recorded.

Measurement of serum H2S
Serum H2S was measured according to a protocol reported elsewhere with minor modifications and standardization. 19,20This spectrophotometric method is based on the formation of methylene blue from the reactions of a sulfide salt with N,N-dimethyl-pphenylenediamine sulfate in the presence of oxidizing Fe 3+ in an acidic medium.The absorbance of methylene blue was measured at 670nm wavelength using a UV-Vis dual beam spectrophotometer (Systronics India Ltd, New Delhi, India).

Method of measurement of H2S concentration in serum:
This spectrophotometric method involves the reaction of sulfide with N,N-dimethyl-pphenylenediamine sulfate in the presence of the oxidizing agent Fe 3+ in hydrochloric acid to form methylene blue, which is read at 670nm.Assay procedure: Seventy-five microliters of serum was added to 425 microliters of phosphate buffered saline (PBS) and 250 microliters of 10% trichloroacetic acid in a capped glass tube.It was then centrifuged at 3000 rpm for 30 minutes.The supernatant was taken in another glass tube, and 250 microliters of 1% zinc acetate, 133 microliters of 20 millimolar N,N-dimethyl-p-phenylenediamine sulfate in 7.2 mM HCl, 133 microliters of 30 millimolar FeCl3 in 1.2 mM HCl, and 60 microliters of 10% NaOH were added, capped, and incubated for 10 minutes at room temperature.All samples were assayed in triplicate, and serum H2S levels were calculated against a calibration curve prepared with 25-250 micromol/L concentrations of sodium sulfide (NaHS, Sigma-Aldrich, MO, USA) (previously published by some of the authors in a different H2S study).The intra-assay and inter-assay variations of this method were 7.576 and 3.944 respectively, and the maximum sensitivity was up to 25 micromol/L. 21

Statistical analysis
Data were depicted as mean and SD for normally distributed variables and median with interquartile range (IQR) for non-normally distributed variables.To compare serum H2S levels between S-COPD and AECOPD and between COPD patients with and without PAH, student's t-tests were performed.One-way analysis of variance (ANOVA) was used to test serum H2S levels across GOLD stages, and Bonferroni's test was used to assess intergroup variation.To evaluate the association between serum H2S level and parameters related to COPD and PAH, we first tested the bivariate relationship between serum H2S and each clinical parameter using student's t-test and Spearman's rank order correlation as appropriate.Second, to assess associations between possible confounders such as age, sex, smoking status, BMI, and serum H2S and all clinical parameters, we performed the aforementioned tests (Student's t-test, Spearman's rank order correlation, and Wilcoxon signed rank test if required).Third, a multivariable linear regression model was constructed for serum H2S against each clinical variable if those variables exhibited a p-value <0.2 in the corresponding bivariate analysis.Covariates were included in the model using a step-forward and stepbackward selection process if (i) they related to both the exposure and outcome in bivariate analysis, (ii) caused >10% change in the regression coefficient estimates of the remaining variables in the multivariable models, or (iii) were found to associate with serum H2S level in previously published literature.Regression diagnostics were performed to assess model goodness-of-fit and eliminate any possible collinearity between clinical variables.
Receiver operating characteristic (ROC) curves were constructed to determine the predictive value of serum H2S for acute exacerbations and occurrence of PAH in COPD.For all analyses, a two-tailed p-value <0.05 was considered significant.All analyses were performed in STATA V.12.0 (Stata Corp, College Station, Texas, USA).

Characteristics of the patients
The baseline demographic characteristics of the patients are presented in Table 1.There were 45 male patients and 5 female patients, all with a mean body mass index of 21.1±3.7 kg/m 2 .According to the GOLD staging for severity, 28%, 26% and 46% of the patients belonged to GOLD-II, III and IV, respectively (no patients were in GOLD-I).Thirty-four patients (68%) had AECOPD and 16 (32%) had stable COPD (S-COPD).ABG analysis demonstrated a mean arterial blood pH of 7.36±0.09,PaO2 of 76.5±17.8mmHg and PaCO2 of 67.5±20.2mmHg.The patients had a mean pulmonary arterial systolic pressure (sPAP) of 36.3±16.4mmHg obtained from the DE test and 37 (74%) patients were diagnosed with PAH.Mean serum H2S of the patients was 40.4±15.1µmol/L.

Relationship between serum H2S and clinical parameters
In Table 2, we present the relationships between serum H2S and the clinical parameters relevant to COPD.Arterial blood pH was found to have a strong correlation with serum H2S level (p<0.001).Both FEV1 (%predicted) and FEV1/FVC ratio demonstrated positive associations with serum H2S and the estimates did not change after adjusting for potential confounders such as age, gender, smoking history and BMI.PaCO2 demonstrated a negative association (regression coefficient β: -0.46; 95% confidence interval: -0.61 to -0.31) with serum H2S in a multivariable model after adjusting for potential confounders.Serum H2S level was found to be negatively associated with the GOLD severity stages (-9.4;-13.3 to -5.4) and acute exacerbation in COPD (-10.5;-18.8 to -2.3).We also observed that the cooccurrence of PAH also negatively influenced serum H2S level significantly (-19.2;-26.3 to -12.1).

Serum H2S and sPAP
We found that serum H2S level was negatively influenced by sPAP (R2: 0.57; p<0.001, Figure 3).In an unadjusted model, we observed a 0.69 µmol/L decline of serum H2S with a per unit increase of sPAP, the estimate remained unchanged upon adjustment for potential confounders (-0.63; -0.83 to -0.43).

H2S ratio as indicative of an exacerbation and PAH
We constructed a ROC curve to determine the cut-off level of serum H2S as an indicator of acute exacerbation from stable COPD (Figure 4).Since higher serum H2S level was associated with more stable COPD (conversely, reduced exacerbation), the ROC was constructed using negative H2S as the indicator variable.The optimal cut-off level of H2S for indicating an exacerbation was <46.7 µmol/L (area under the curve: 0.730, 95% CI: 0.59-0.87).A similar ROCwas constructed to determine PAH in COPD (Figure 5).We found an optimal cut-off value for H2S indicating the possible occurrence of PAH <46.7 µmol/L (AUC: 0.917; 95%CI: 0.84-0.99).

DISCUSSION
In this study, we observed that serum H2S level was lower among the COPD patients with acute exacerbation in compared to the stable-COPD patients.There was a consistent drop of serum H2S level in accordance with the increase of disease severity, i.e., serum H2S level was the lowest among the patients in GOLD-IV stage in compared to those in GOLD-II and GOLD-III.Among the COPD patients with co-occurrence of PAH, the serum H2S level was significantly lower than the H2S level in those COPD patients without PAH.Serum H2S level had a significant inverse correlation with PaCO2 and exacerbation and positive correlations with lung functions suggesting that serum H2S level is an important determinant of exacerbation and associated clinical manifestations in COPD.The strong inverse relation between sPAP and serum H2S also indicated that serum H2S could be a useful indicator of PAH.
The role of endogenously synthesized H2S in COPD has been reported in only a few research articles.We observed that FEV1% and FEV1/FVC were positively correlated with serum H2S which is consistent with previously published reports. 10,11Our result of decrease in level of H2S in according to the disease severity also substantiates previous report. 10Our observation of markedly reduced serum H2S level among the AECOPD patients than the S-COPD patients are also in accordance with the previously published reports. 3,10It has been demonstrated that the serum H2S drops in conditions, like chest infection. 22Also, in animal models, deficiency of H2S synthesizing enzymes was found to associate with increased airway hyper-responsiveness inflammation. 7,23These observations speak on the role of endogenous H2S in COPD pathobiology.H2S has been demonstrated to have potentially important role in vasodilation of the pulmonary arteries and reducing pulmonary arterial pressure 24,25 although much is unknown about the true mechanism.In COPD, pulmonary fibrosis and obstructive sleep apnea, the development of global pulmonary hypoxia is a clinically important perturbation that may lead to hypoxic pulmonary hypertension. 26However, there is only one study reported so far on the possible role of H2S in COPD-associated PAH.Both in our results and that of the previously published one, the COPD patients diagnosed with PAH had a lower serum H2S level than those COPD patients without PAH. 10In isolated human lungs model, exogenous administration of H2S substantially reduced pulmonary artery pressure. 25In animal model, the effect of H2S on pulmonary arteries has been well demonstrated.In an artificially developed PAH model in rats, H2S donor was observed to reduce pulmonary arterial pressure and lowered the relative medial thickness (RMT) and relative medial areas (RMA). 6H2S also reduced inflammation in the pulmonary vasculature by reducing the intracellular cell adhesion molecule-1 (ICAM-1) in the pulmonary vasculature. 6 are aware of the potential limitations of this study.One of the major limitations was the use of Doppler echocardiography to measure the sPAP instead of right heart catheterization (RHC) although DE is a common measure to estimate the sPAP and has high correlation with the measurement done by RHC. 17 DE has been shown to have a lesser precision in estimating sPAP in compared to RHC in certain clinical conditions such as in advanced pulmonary diseases.However, RHC is an invasive procedure with several fatal complications 27 and in advanced-stage COPD with acute exacerbation and/or hypertension in which the patients become hemodynamically unstable, such invasive diagnostic procedures may impart severe adverse health effects.Furthermore, in countries like India where there is a lack of consensus guideline of medical procedures, add-on tests are often disregarded by the patients' families primarily because of the costs of the procedure as most of the individuals are not benefited by social security system or health insurance policies.However, to overcome the potential procedural limitation and to nullify any suspected case of over or underestimation, each DE result was scrutinized by two blinded independent experts and the scores of sPAP obtained from them were tallied using Cohen's kappa for an inter-rater agreement and found a very high agreement (κ= 0.93, p<0.001) between the observations.Medication can also alter the H2S level in acute exacerbated COPD patients.Although the effects of COPD medications such as cholinergic antagonists and β-adrenergic agonists on endogenous H2S is unclear, corticosteroids might impair the activity of H2S synthesizing enzyme cystathionine γ-lyase (CSE), thus attenuating the H2S production. 28,29In our study inhaled corticosteroids (ICS) were prescribed to the AECOPD patients as per the GOLD guidelines, therefore the effects of ICS on reduced H2S cannot be ignored.
1][32][33][34] Studies on animal models and isolated human lung segments have demonstrated that H2S donors ameliorates oxidative stress, airway inflammation, and remodeling 35 and provides protection against emphysema and pulmonary hypertension. 25,35However, therapeutic potential, efficacy and safety of exogenously administered H2S-donors or H2S augmenters in human systems are yet to be validated through clinical studies.

Study limitations
H2S has been assayed with crude form of chromogen in spectrophotometer, this study can be done by mass spectrometry if it would have been available.Another limitation is use of non-invasive color Doppler echocardiography to measure sPAP instead of right heart catheterization, an invasive procedure.The study was done on a fewer number of subjects due to inadequate funds and time constraints.Some medications like corticosteroids may have interference with the H2S synthesizing enzymes which may have affected the levels of serum H2S.A patient registry and tracking system is lacking in most of the health-care settings in India, therefore, the history, course and progression of the disease remain undetermined in most of the cases.Finally, we could not perform high-resolution CT scans of all the patients which could be a useful instrument to quantify and grade the structural changes of the lungs.It was a pilot project to be followed up by a more detailed study recruiting more cases and controls.

Figure 1 Figure 2
Figure 1 Concentration ofserum H2S across COPD stages, GOLD-II, III and IV.* indicates difference between GOLD-II and IV and # indicates difference between GOLD-III and IV.Level of significance p<0.05.(For abbreviation, see abbreviations list.)

Figure 4
Figure 4 Receiveroperator characteristics (ROC) curve for H2S to predict the risk ofexacerbation in COPD.

Figure 5
Figure 5 Receiver operatorcharacteristics (ROC) curve for H2S to predict the risk ofdeveloping PAH among patients with COPD.