Using fractional exhaled nitric oxide to guide step-down treatment decisions in asthma: practical considerations

From the authors:

We thank L. Pérez de Llano and colleagues for their comments on our study [1] and for comparing our findings with those of their prospective multicentre study describing a simple score for predicting step-down failure in adults with well-controlled asthma [2]. However, we would like to clarify that recent systematic review findings conclude that there is insufficient evidence, rather than evidence against, the ability of low fractional exhaled nitric oxide (F_ENO) to identify individuals in whom treatment can be safely stepped down [3].

We fully acknowledge that F_ENO should be interpreted in the context of other potential risk factors and that pro-active efforts should be made to identify and treat other conditions that increase F_ENO, particularly in patients whose F_ENO is high despite satisfactory medication adherence and inhaler technique.

We also acknowledge that, although there is well-established evidence that F_ENO is correlated with sputum eosinophil counts in nonsmoking patients with stable asthma [4], this correlation is not strong and it is possible for eosinophilic airway inflammation to occur in the presence of low F_ENO. However, it is neither feasible nor acceptable to obtain induced sputum samples in primary care. Blood eosinophil counts may be an adequate surrogate but this has not been demonstrated. We therefore recommend that, in community-based healthcare settings, treatment should be stepped down gradually and patients should be monitored carefully in case any reductions subsequently unmask undetected airway inflammation.

We are aware of the practical implications of variability in F_ENO measurements between different devices. However, this is unlikely to have had an undue impact on the findings of our meta-analysis, as only two included studies used Sievers rather than Aerocrine analysers [5, 6] and only three participants across these two studies had one or more exacerbations.

The study by L. Pérez de Llano and colleagues also highlights other issues that should be considered when using F_ENO to guide step-down decisions.

1) The definition of step-down failure. The primary outcome used by L. Pérez de Llano and colleagues was “loss of control”. This was defined as an Asthma Control Test score of ≤19, a decrease in forced expiratory volume in 1 s of ≥20% from baseline, or exacerbations resulting in symptom deterioration, whether or not these required treatment with systemic corticosteroids or hospitalisation. However, the primary outcome of our meta-analysis was exacerbations requiring antibiotics, systemic corticosteroids, hospitalisation or unscheduled healthcare visits. These events may relate more closely to type-2 airway inflammation than less severe events [7] and have a significant detrimental impact on health-related quality of life [8], which is in turn associated with increased asthma-related costs and healthcare resource utilisation [9]. Only ∼13% of participants who experienced loss of control in the study by L. Pérez de Llano and colleagues required systemic corticosteroids (13 out of 102 participants) and no participants were hospitalised. This would suggest that, while F_ENO may not be a reliable predictor of mild-to-moderate deteriorations in symptom control or lung function, it could still be a useful predictor of clinically important exacerbations requiring further intervention.

2) Target population characteristics. The list of possible diagnostic criteria for asthma used by L. Pérez de Llano and colleagues included F_ENO >50 ppb. However, none of the studies we included in our meta-analysis selected participants based on their F_ENO measurement at baseline. F_ENO may therefore have a greater predictive value in asthma populations that do not include individuals who have been pre-selected on the basis of already having a high F_ENO.

3) Follow-up after stepping down treatment. Whilst our meta-analysis only analysed one step-down episode per participant, L. Pérez de Llano and colleagues analysed all step-down episodes. Each participant could undergo up to three step-down episodes over a 12-month period, with a 6-month follow-up period after the step-down episode at the third study visit. However, analysis of multiple episodes per participant is likely to result in clustering of outcomes within individuals. Additionally, extended follow-up periods may result in exacerbations or “loss of control” episodes unrelated to treatment being stepped down.

In conclusion, we agree with L. Pérez de Llano and colleagues that F_ENO should not be considered in isolation to guide step-down treatment decisions. However, in real life clinical practice, it is practically impossible to control for every potential source of confounding or variability. We therefore advocate using F_ENO as part of a gradual, safely monitored step-down approach.

Shareable PDF