Defining the role of exhaled nitric oxide measurements in COPD

The fraction of nitric oxide in exhaled breath (FENO) is a biomarker that enables the clinician to assess airway inflammation. Conventionally, FENO aids the distinction between eosinophilic and non-eosinophilic airway inflammation.1 More importantly, in part because the former responds more favourably to treatment with corticosteroid than the latter, FENO may be regarded as a marker of potential steroid responsiveness.2 In adults, an FENO above approximately 50 parts per billion (ppb) identifies patients who are likely to respond to corticosteroid, while an FENO below 25 ppb identifies those who are not, irrespective of the diagnostic label.1, 2 Much work has been done to establish the role of FENO in asthma. Less has been done in COPD but here, too, FENO may have a useful role, given that eosinophilic airway inflammation (and hence potential steroid responsiveness) is present in a subgroup of stable patients,3-5 and also with acute exacerbations.6, 7 The levels of FENO in patients with COPD are only occasionally elevated compared with controls.8, 9 However, even though levels may be within the ‘normal’ range, there is a significant relationship between baseline FENO and the subsequent increase in FEV1 after bronchodilator10 or inhaled corticosteroid.11, 12 The relationship between FENO and corticosteroid response has been further defined in two recent studies demonstrating that, in stable COPD, FENO predicts a clinically significant increase in FEV1 in response to short-term treatment with corticosteroid.13, 14 A high FENO (>50 ppb) predicts steroid responsiveness only modestly (positive predictive value 67%). However, a low FENO (<25 ppb) has a high negative predictive value (87%) and is therefore a useful predictor of the absence of a response. This is helpful to the clinician when deciding whether or not to prescribe inhaled corticosteroid, and offers an improvement on the current situation, where treatment is started empirically often to no good effect. While FENO is known to be elevated in acute exacerbations of COPD,15 no clinical application for this observation has yet been determined. In this issue of Respirology, Antus et al. explored the potential role of FENO in predicting the response to treatment in exacerbations of COPD requiring admission to hospital.16 They measured FENO and lung function in 58 patients on admission and again at discharge. Treatment was not standardized: all patients received β-agonist/anticholinergic bronchodilators and systemic corticosteroid, 48 (83%) received theophylline and 30 (52%) received antibiotics. Nonetheless, their results were striking: FENO on admission was a strong predictor of a clinically significant increase in FEV1 following treatment (area under the receiver operator characteristic curve = 0.82). An FENO below the optimum cut-point of 27 ppb predicted the absence of any significant improvement in lung function, with a high negative predictive value of 85%. These novel findings suggest the possibility of a role for FENO in predicting the response to treatment in acute exacerbations of COPD. They also raise additional questions: what pathology underlies an elevated FENO at exacerbation? Is FENO acting as a surrogate marker for eosinophilic inflammation or an alternative aspect of airway pathophysiology? Is FENO predicting the response to corticosteroid in this situation? Acute exacerbations of COPD are frequently precipitated by the advent of acute viral infection, as highlighted by Mohan et al. in this issue of Respirology.17 Simultaneously, they are associated with an increase in eosinophilic airway inflammation, and it has been proposed that this is most likely driven by viral infection18—the same mechanism may also occur in stable COPD, given the evidence for viral persistence in some patients.19In vitro data indicate that infection with rhinovirus, respiratory syncytial virus and influenza A, which are known to cause COPD exacerbations, promotes eosinophilic inflammation by releasing eosinophil chemoattractants and activators.20-23 In the case of respiratory syncytial virus, increased activation of inducible nitric oxide synthase and production of nitric oxide have also been shown to occur.24 Thus the increase in FENO in acute exacerbations of COPD may be the direct result of viral infection, an indirect marker for eosinophilic inflammation triggered by viral infection or, as in stable COPD, an independent marker of potential steroid responsiveness. What remains unclear is whether the increase in FENO seen in acute exacerbations predicts the long-term efficacy of inhaled corticosteroid treatment. This information would be clinically helpful: there is an opportunity to reassess the need for long-term treatment when a patient presents acutely. Although treatment of acute exacerbations of COPD with corticosteroid causes a reduction in sputum eosinophils,25 the utility of sputum eosinophilia at exacerbation as a predictor of corticosteroid response remains unknown. Likewise, the utility of FENO at exacerbation as a predictor of long-term corticosteroid response has yet to be defined. The paper by Antus et al. provides a further piece in the jig-saw, as well as an incentive to pursue the decade-old vision of targeted corticosteroid treatment for COPD.