To the Editor:
Hydrogen sulfide (H2S) has emerged as a new and important endogenous regulator of inflammation in recent years [1] and may also protect from emphysema induced by cigarette smoke exposure [2]. We have also recently shown that H2S can inhibit airway smooth muscle cell proliferation and inflammatory mediator release in vitro [3]. Serum levels of H2S positively correlate with the decline in lung function in chronic obstructive pulmonary disease (COPD) and were significantly lower in Global Initiative for Chronic Obstructive Lung Disease (GOLD) stage III patients compared with those in GOLD I [4]. Existing therapies for COPD, such as corticosteroids or long-acting anticholinergic agents, may reduce the exacerbation rate but do not significantly slow disease progression. A previous study has shown that theophylline alone had no impact on serum H2S levels and is of limited value in the management of stable COPD [5]. Interestingly, sputum H2S measured in patients with asthma correlated with sputum neutrophil counts and the degree of airflow obstruction measured by forced expiratory volume in 1 s (FEV1) % predicted [6]. Moreover, combination therapy of an inhaled glucocorticoid with low-dose theophylline has been shown to attenuate airway inflammation in patients with COPD and reverse glucocorticoid resistance [7]. We therefore investigated whether the combination of inhaled corticosteroid and low-dose theophylline, as opposed to low-dose theophylline alone, would modulate H2S levels in the lungs of COPD patients. We now report the levels of H2S assayed in sputum samples collected during this study (www.clinicaltrials.gov identifier NCT00241631), details of which have already been published [7].
Briefly, a total of 29 patients with stable COPD, 18 in GOLD II and 11 in GOLD III, provided sputum for the measurement of H2S. After an initial 2-week washout during which all inhaled corticosteroid was ceased, subjects were randomised into two treatment arms, one with inhaled fluticasone propionate (500 mg twice daily) the other without for 4 weeks; both arms were given placebo theophylline capsules. After another 2-week washout period, this was followed by a further 4-week period where both arms were given active slow-release theophylline capsules (250 mg twice daily). Induced sputum samples were obtained at the end of the initial run-in period and at the end of each 4-week treatment period. The supernatant fraction was analysed for H2S content using a method that accounts for all H2S species and those derived from it at physiological pH, as described previously [8]. The data were analysed using the statistical package SPSS (version 8; IBM, Armonk, NY, USA) incorporating mixed models to account for nonconstant variability across visits and correlation between visits within patient groups. Over the course of the study, COPD subjects as a group exhibited a significant decrease in sputum H2S levels in the placebo arm compared with the fluticasone propionate arms (fig. 1a). When patients were analysed according to disease severity, as defined by GOLD status, those patents in GOLD II showed no change in sputum H2S levels in either the placebo or fluticasone arms (fig. 1b). As a group, GOLD III COPD subjects in the placebo arm (no fluticasone) had a reduced level of sputum H2S after 4 weeks treatment with theophylline only compared with baseline (fig. 1c). In contrast, GOLD III subjects as a group in the fluticasone arm exhibited no changes in sputum H2S levels after theophylline treatment (fig. 1c). Interestingly, when sputum H2S levels were analysed within the context of changes in FEV1, those COPD patients who exhibited the greatest increase in FEV1 upon the addition of theophylline to fluticasone had very low initial sputum H2S levels. H2S levels rose significantly for all subjects in this subgroup upon addition of theophylline to fluticasone (fig. 1d). This effect was not apparent in either the placebo arm of the study or those subjects in the fluticasone arm that had an elevated baseline H2S level.
The significance of decreased sputum H2S levels in GOLD III is not clear. It may simply reflect the greater degree of oxidative burden in the GOLD III COPD lung [9] as H2S, or species derived from it at physiological pH, readily interacts with detrimental oxidant species, such as nitric oxide, peroxynitrite and hypohalous acids. Alternatively, as corticosteroids appear to induce the synthesis of endogenous H2S synthesis enzymes [10], our results may simply reflect a lower level of endogenous H2S in more severe COPD in the absence of corticosteroid. The most surprising finding was that those subjects in the fluticasone study arm who had low baseline levels of sputum H2S demonstrated the greatest increase in FEV1 upon the addition of theophylline. This suggests that low sputum H2S levels could act as a biomarker for those COPD subjects who could benefit the most from theophylline/corticosteroid combination therapy. Clearly, this small study is hypothesis generating, requires replication and should be interpreted with great caution. However, our data confirm previous evidence suggesting that theophylline alone does not regulate H2S levels [5] and also suggest that the combination of theophylline plus fluticasone at least maintains H2S levels in patients with GOLD II or III disease. Increased H2S levels in the fluticasone arm may be due to increased expression of H2S-synthesising enzymes [10]. This effect would be independent of the transrepression actions associated with loss of histone deacetylase 2 [9] and may suggest that transactivation by corticosteroids is unaffected in COPD. In conclusion, monotherapy with theophylline does not appear to have any impact on maintaining or further increasing sputum H2S levels. In combination with inhaled corticosteroid, however, while there is a trend towards increasing sputum H2S levels, this is most evident in those COPD subjects who had very low baseline sputum H2S levels and, furthermore, appeared to benefit most in relation to increases in FEV1.
Footnotes
Conflict of interest: Disclosures can be found alongside the online version of this article at www.erj.ersjournals.com
Clinical trial: This study is registered at www.clinicaltrials.gov with identifier number NCT00241631.
- Received July 30, 2013.
- Accepted January 4, 2014.
- ©ERS 2014