Performance of fractional exhaled nitric oxide in predicting response to inhaled corticosteroids in chronic cough: a meta-analysis

Background. Chronic cough is a disabling condition with a high proportion of diagnostic and therapeutic failures. Fractional exhaled nitric oxide (FeNO) has been considered a useful biomarker for predicting inhaled corticosteroids (ICS) response. We evaluated the relationship between FeNO and ICS response in chronic cough by performing a systematic review with meta-analysis. Methods. PubMed, Web of Science, Scopus and EMBASE databases were systematically searched. Differences were expressed as Odds Ratio (OR) with 95% confidence intervals (95%CI). Pooled sensitivity, specificity, positive (PLR) and negative likelihood ratio (NLR), and area under the hierarchical summary receiver operating characteristic curve (HSROC(AUC)) were estimated. Results. Nine articles on 740 chronic-cough patients showed that the response rate to ICS was 87.4% (95%CI: 83.8–91.0) in 317 patients with a high FeNO and 46.3% (95%CI: 41.6–51.0) in 423 controls, with an attributable proportion of 47.0% and a diagnostic OR of 9.1 (95%CI: 3.7–22.4, p < .001). The pooled estimate of diagnostic indexes resulted in a sensitivity of 68.5% (95%CI: 46.7–84.4) and specificity of 81.9% (95%CI: 63.0–92.3), with a HSROC(AUC) of 0.82 (95%CI: 0.64–0.90). In a realistic scenario with a pre-test probability set at 30%, based on a pooled PLR of 3.79 (95%CI: 1.24–7.47) and NLR of 0.38 (95%CI: 0.22–0.66), the post-test probability was 62% with a high FeNO and 14% if the test was negative. Subgroup analyses confirmed a better performance for the recommended FeNO cut-off greater than 25 ppb. Meta-regression and sensitivity analyses showed no impact of major demographic and clinic variables on results. Conclusions. KEY MESSAGES: Chronic cough is a disabling condition with a high proportion of diagnostic and therapeutic failures. Fractional exhaled nitric oxide (FeNO) may be a useful biomarker for identifying chronic cough patients who respond to steroid treatment. A FeNO cut-off lower than 25 ppb should be considered irrelevant for this clinical application.