Validation study of asthma screening criteria based on subjective symptoms and fractional exhaled nitric oxide

Background

In the latest Global Initiative for Asthma guideline, neither sputum eosinophilia nor fractional exhaled nitric oxide (FeNO) have been evaluated prospectively as an aid in asthma diagnosis, but these measurements are being evaluated for potential use in determining optimal treatment.

Objective

To report criteria for screening asthma using subjective symptoms and FeNO levels and results of a prospective validation study using these criteria.

Methods

Sixty-one outpatients with recurrent cough, wheezing, or dyspnea underwent measurements of FeNO levels, pulmonary function, methacholine airway responsiveness, and inflammatory cells in induced sputum. The sensitivity, specificity, and concordance achieved using the FeNO-based criteria (at least 1 of the following subjective symptoms: recurrent cough, wheezing, or dyspnea and/or FeNO level ≥40 ppb) were analyzed and compared with the values obtained using conventional asthma diagnostic criteria, which includes subjective symptoms with any 2 of the following conditions: airway hyperresponsiveness, reversible airflow limitation, and eosinophilia in induced sputum.

Results

Of the 61 patients, 41 were diagnosed as having asthma by the conventional criteria, and 33 were diagnosed as having asthma by the FeNO-based criteria, which showed a sensitivity of 78.6%, a specificity of 89.5%, and a concordance rate of 0.62. Nine of 42 patients were misdiagnosed as not having asthma by FeNO-based criteria (mean [SD] FeNO level, 23.9 [8.0] ppb). Seven of 9 patients were diagnosed as having nonatopic asthma according to IgE levels.

Conclusions

Asthma may be accurately diagnosed in daily practice on the basis of subjective symptoms and FeNO levels, particularly in atopic patients.

Introduction

The prevalence of bronchial asthma is continually increasing and becoming a significant socioeconomic burden throughout the world. Thus, several asthma guidelines have been published for standardizing asthma diagnosis and management.1, 2 However, accurate diagnosis of asthma in daily practice continues to remain difficult for physicians.

Bronchial asthma is characterized by pathophysiologic conditions, such as reversible airway obstruction and nonspecific airway hyperresponsiveness, based on chronic airway inflammation, featured primarily in eosinophils, lymphocytes, and mast cells. Asthma is usually diagnosed on the basis of subjective symptoms and the existence of both reversible airway obstruction and airway hyperresponsiveness. Symptoms, however, are subjective and can be overestimated or underestimated by patients. Although pulmonary function testing is objective and simple, the sensitivity is low³ because many patients with mild asthma exhibit normal pulmonary function, which makes the assessment of reversibility difficult. Testing for airway hyperresponsiveness produces highly sensitive results of diagnostic value, but this test can produce false-positive results and is time-consuming, is relatively invasive, and sometimes triggers asthmatic attacks.⁴ Although these 2 markers exhibit some correlations with airway inflammation, they do not directly reflect airway inflammation. However, because the basic feature of bronchial asthma is chronic airway inflammation, direct evidence showing this condition would be of high diagnostic value. Biopsies of airway mucosa and bronchoalveolar lavage under bronchoscopy can determine airway inflammation, but these invasive procedures are not suitable for daily practice. Analysis of induced sputum using a hypertonic saline inhalation is currently used to confirm airway inflammation. This approach is relatively noninvasive but is time-consuming and costly. Sputum induction, moreover, does not necessarily yield high-quality samples, handling and reading of samples are sometimes difficult, and hypertonic saline inhalation can trigger asthmatic attacks, making the procedure unfeasible for routine diagnosis.

Fractional exhaled nitric oxide (FeNO) has recently been developed as a convenient, sensitive, and noninvasive marker for monitoring allergic airway inflammation. Studies report that FeNO levels are higher in asthmatic patients,5, 6 and FeNO has been correlated with the severity of airway inflammation,⁷ eosinophil counts in induced sputum, airway hyperresponsiveness, and pulmonary function.5, 6, 8 FeNO levels also decrease with inhaled steroid treatment.9, 10 FeNO thus shows promise as a diagnostic marker for bronchial asthma.¹¹ However, this marker is also elevated in other diseases, including allergic rhinitis, bronchiectasis, and respiratory tract infections¹²; therefore, an accurate comparison of diagnosis using FeNO levels and conventional asthma diagnostic criteria is required to determine whether FeNO levels are suitable for clinical diagnosis of asthma. Most studies attempting such comparisons, however, have been retrospective,13, 14 and little systematic, prospective research is available. The latest Global Initiative for Asthma guideline indicates that neither sputum eosinophilia nor FeNO has been evaluated prospectively as an aid in asthma diagnosis.¹

In 3 of our previous clinical studies,5, 6, 8 we showed that FeNO has significant value for diagnosing bronchial asthma. Surprisingly, as shown in Figure 1, FeNO cutoff levels for diagnosing asthma obtained from these 3 independent studies were approximately 40 ppb. Therefore, we propose the following asthma screening criteria: (1) recurrent cough, wheezing, or dyspnea; (2) FeNO levels of 40 ppb or higher; and (3) exclusion of other lung diseases. This study was designed to prospectively validate the criteria by comparing diagnostic results with those obtained using conventional asthma diagnostic criteria.