This policy was originally created in 2013 with searches of the MEDLINE database through May 2012. Following is a summary of the key literature to date.
Evaluation of the tests for diagnosis requires that the test findings are reproducible on test-retest and that the test is reasonably accurate compared to a validated reference standard. Assessment of the clinical role of exhaled nitric oxide (NO) and exhaled breath condensate (EBC) tests requires controlled studies of those managed conventionally compared to those whose management was additionally directed by test measurements.
Exhaled Nitric Oxide
Reproducibility of FeNO measurements
In 2010, Selby and colleagues published a study from the UK that evaluated the reproducibility of exhaled NO measurements in young people. (2) The study included 494 teenagers, aged 16-18 years, from an unselected birth cohort and 65 asthma patients between the ages of 6 and 17 years. Paired readings were obtained from each participant. The mean within-participant difference in FeNO (second reading minus the first reading) was 1.37 ppb (95% confidence interval [CI]: -7.61 to 10.34 ppb); this difference was statistically significant, p less than 0.001. When participants with high FeNO values (above 75 ppb) were excluded, there was a lower mean within-participant difference, 0.90 ppb (95% CI:-4.89 to 6.70 ppb). Among the 71 participants with asthma, the mean within-participant difference in FeNO in the 2 measurements was 2.37 ppb (95% CI: -11.38 to 16.12 ppb). When FeNO values were categorized as low, normal, intermediate, or high (using different values for participants younger than age 12 years and 12 years or older), the findings were reproducible. That is, there were no statistically significant differences in the categorization using the first and second measurement.
Using FeNO in the diagnosis of asthma in individuals with signs or symptoms of asthma
Several prospective studies have evaluated FeNO devices for diagnosing asthma; studies have generally focused on identifying the optimal FeNO cutpoint. Two studies were identified that used the NIOX MINO device (Aerocrine). In 2009, Schneider and colleagues in Germany published data on 160 patients with symptoms suspicious of obstructive airway disease. (3) All patients underwent measurement of exhaled NO. The reference standard was a stepwise series of tests, beginning with spirometry. Those with forced expiratory volume in one second (FEV1) less than 80% of predicted or FEV1/vital capacity (VC) ratio of 0.70 or less were referred to bronchodilator reversibility testing. Otherwise, patients received bronchial provocation with methacholine. Patients were classified as having asthma when: 1) bronchodilation testing found a change in FEV1 was at least 12% compared to baseline, and at least 200 mL, and lung volumes returned to predicted normal range; 2) bronchial provocation found a 20% decrease in FEV1 from the baseline value after inhaling methacholine stepwise until the maximum concentration. Exhaled NO test findings were compared to the final diagnosis status. According to standard testing, 75 (46.9%) of the patients had asthma. Receiver operating characteristic (ROC) analysis found the highest sum of sensitivity and specificity of exhaled NO at a cut-off of 46 ppb. Among patients with unsuspicious spirometry findings (n=101), 49 had asthma. The optimal cut-off of exhaled NO in this subgroup was also 46 ppb; the sensitivity of exhaled NO was 35%, and the specificity was 90%.
The second study using the NIOX MINO was published in 2010 by Pedrosa and colleagues in Spain. (4) The study included 114 individuals at least 14 years-old who had symptoms consistent with asthma, with or without rhinitis symptoms, and had normal parameters on spirometry and a negative bronchodilator test. Definitive diagnosis was based on symptom assessment and a positive methacholine bronchial challenge test. Individuals underwent FeNO assessment (flow rate of 50 mL/s) just before the methacholine inhalation challenge test. According to challenge test findings, 35 patients (31%) were diagnosed with asthma. FeNO levels were significantly higher in individuals diagnosed with asthma (mean 58 ppb) than in non-asthmatics (mean ppb 30 ppb); p<0.001. Using ROC analysis, the cut-off point with maximum sensitivity (74.3%) and specificity (72.5%) for diagnosing asthma was a FeNO value of 40 ppb.
A 2009 study by Sivan and colleagues in Israel used a different device (CLD88 FeNO analyzer; EcoMedics; Switzerland) that is not FDA-approved. (5) The study included 150 children aged 18 years or younger who were referred for evaluation of possible asthma. It evaluated the diagnostic yield of exhaled NO test findings compared to sputum eosinophil count, the ‘gold standard” for assessment of eosinophilic inflammation of the airways. Final assessment of asthma status was done by a pediatric pulmonologist after at least 18 months of follow-up. Receiver operating curves (ROCs) were used to determine the optimal cut-off points for the exhaled NO test in diagnosing asthma. The area under the ROC for exhaled nitric oxide versus eosinophil percent was 0.886. The best cut-off was 18 ppb, which provided 82% sensitivity and 84% specificity. This was similar to the best cut-off for eosinophil count, 2.7%, which had 85% sensitivity and 89% specificity. A 2010 study conducted in Italy included 280 children with asthma, allergic rhinitis or both. (6) The authors used ROC analysis and found that the optimal cut-off for discriminating between patients with bronchial hyperactivity from those with absent or borderline bronchial hyperactivity was 32 ppb of NO.
The cut-off level of exhaled NO also varied in earlier studies. A 2005 study by Berkman and colleagues studied 95 asthmatic and non-asthmatic patients without age limitation and found that a cut-off of FeNO over 7 ppb best differentiated between the 2 groups. (7) In 2003, Dupont and colleagues evaluated 240 non-smoking corticosteroid-naïve patients without age limitations and found the optimal cut-off was 16 ppb exhaled NO. (8)
Using FeNO in the diagnosis of respiratory disorders other than asthma
Rouhos and colleagues in Finland published a study in 2011 on repeatability of FeNO measurements in 20 patients with stable chronic obstructive pulmonary disease (COPD) and 20 healthy controls. (9) FeNO was measured 3 times in each individual; a baseline measurement and measurements 10 minutes and 24 hours after baseline. In COPD patients, median FeNO values were 15.2 ppb at baseline, 17.4 ppb 10 minutes later, and 14.5 ppb 24 hours later. In healthy controls, corresponding median FeNO values were 15.6 ppb, 19.6 ppb, and 15.7 ppb. Differences between the baseline and 24-hour measurements in both groups were not statistically significant. FeNO values 10 minutes after baseline were significantly higher than the 24 hour measurement in both groups; the authors attributed this difference to the fact that patients did not rinse their mouths with sodium bicarbonate between the baseline and 10-minute measurements.
Using FeNO to guide treatment decisions in patients with asthma:
In 2005, a TEC Assessment was published on exhaled NO monitoring for guiding treatment decisions in patients with chronic asthma. (10) The assessment identified 2 randomized controlled trials (RCTs); both were published in 2005. Smith and colleagues reported that equivalent outcomes (e.g. exacerbations, pulmonary function) were achieved in the group managed using exhaled NO measurements compared to the group managed using conventional guidelines (11) The FeNO group, however, used lower doses of inhaled corticosteroids at the end of the study. Pijnenburg and colleagues found similar changes in steroid dose and FEV1 in groups managed with and without FeNO measurements. (12) Bronchial hyperreactivity, an intermediate outcome, improved more in the FeNO group. The TEC Assessment concluded that the available evidence did not permit the conclusion that use of NO monitoring to guide treatment decisions in asthma leads to improved outcomes.
A 2009 Cochrane review identified studies comparing outcomes in asthma patients whose medication adjustments were managed based on FeNO levels versus managed based on clinical symptoms (with or without spirometry/peak flow). (13) Six RCTs were identified, including the 2 that had been discussed in the TEC Assessment, as well as 4 additional studies, Shaw et al. 2007, (14) Fritsch et al. 2006, (15) Szefler et al. 2008, (16) and de Jongste et al. 2009. (17) Four studies included children or adolescents, 1 included only adults and the sixth included both adolescents and adults. Two studies were double-blind and the other 4 were single-blind. Five studies used hospital-based FeNO measurements, and one used a portable at-home NO analyzer. Four studies measured FeNO at a flow rate of 50 mL/s.
When findings for the 2 studies that included adults and/or adolescents were pooled (Shaw et al. 2007 and Smith et al. 2005), total n=197, there was not a significant difference in the number of patients experiencing an exacerbation (odds ratio [OR]: 0.85, 95% CI: 0.30 to 2.43) or the occurrence of any exacerbation (mean difference: -0.14, 95% CI: -0.41 to 0.12). There was also no significant difference in symptom scores (mean difference of -0.14 [95% CI: -0.42 to 0.14]). Findings from 3 of the 4 pediatric trials were pooled, total n=782 (Pijnenburg et al. 2005,  Szefler et al. 2008,  and de Jongste et al. 2009 ). As with the adult studies, there was not a significant difference in the number of patients experiencing an exacerbation (OR=0.75, 95% CI: 0.55 to 1.01). Another pooled analysis of these 3 studies found a statistically significantly higher dose of inhaled corticosteroid at the final study visit in patients managed using exhaled NO levels (mean difference: 140.2, 95% CI: 28.9-251.4). A pooled analysis of 2 of the studies (Szefler et al. 2008 and de Jongste et al. 2009, total n=631) did not find a significant difference in symptom scores when patients were managed with and without measurement of exhaled NO (mean difference: 0.04, 95% CI: -0.11 to 0.20). Findings on the number of patients experiencing an exacerbation were not pooled for the pediatric studies. The authors of the Cochrane review concluded, “Tailoring the dose of inhaled corticosteroids based on exhaled nitric oxide in comparison to clinical symptoms was carried out in different ways in the six studies and found only modest benefit at best and potentially higher doses of inhaled corticosteroids in children. The role of utilizing exhaled nitric oxide to tailor the dose of inhaled corticosteroids cannot be routinely recommended for clinical practice at this stage and remains uncertain.”
A 2010 review by Barnes and colleagues stated that the published trials comparing exhaled NO measurement as an add-on to clinical guideline management have had substantial design issues that may limit their validity. (18) For example, the authors questioned the adequacy of the cutpoints used for determining a positive test. They pointed out that, due to inter-individual variability in exhaled NO levels, use of individual cutpoints, as determined by baseline assessment, or measuring change from baseline may be more valid than use of a fixed cutpoint. In addition, the authors stated that the choice of outcome variables is important. They asserted that asthma exacerbations may be the most relevant primary outcome, but that was not consistently the case in the published studies. The authors concluded, “The true value of FeNO in improving asthma control and reducing exacerbations has yet to be tested rigorously.”
In 2011, Powell and colleagues in Australia published a double-blind RCT evaluating FeNO for guiding treatment decisions in pregnant non-smoking women with asthma. (19) Eligibility included being between 12 and 20 weeks’ of gestation and using inhaled therapy for asthma within the past year. Women were randomized to a FeNO algorithm to adjust therapy (n=111) or a clinical guideline algorithm that did not include FeNO measurement (n=109). The FeNO algorithm appeared to be devised by the study investigators. According to the algorithm, the cut-off for reducing the dose of inhaled corticosteroids was less than 16 ppb, and the cut-off for dose increase was at least 30 ppb. Both treatment groups also had their symptoms assessed by the Asthma Control Questionnaire (ACQ), and ACQ scores were utilized in both medication adjustment algorithms. A total of 203 of 220 women (92%) completed the study; analysis was intention to treat. The primary study outcome was the total number of asthma exacerbations during pregnancy (and after study enrollment) for which the patient sought medical attention. The mean total exacerbation rate was significantly lower in the FeNO group (0.29 per pregnancy) compared to the control group (0.62 per pregnancy), p=0.01. Overall, 28 (25%) of women in the FeNO group and 45 (41%) in the control group had at least one exacerbation; the difference between groups was statistically significant, p=0.01. Among the secondary outcomes, there were significantly fewer unplanned doctors visits in the FeNO group (mean of 0.26 per patient) than the control group (mean of 0.56 per patient), p=0.002.
The Powell study demonstrates a potential benefit to using a treatment algorithm that incorporates FeNO levels. However, this trial is prone to many of the same limitations as previous trials of FeNO management algorithms. Most importantly, patients in each group end up on differing regimens of medications according to the algorithm followed. It is then difficult to isolate the effect of the algorithm from the efficacy of the medications themselves. For example, if a FeNO algorithm uses a lenient cut-off point for increasing inhaled corticosteroids, then the FeNO group will likely end up on higher doses of inhaled steroids. Improved outcomes are then more likely to be due to the efficacious effect of inhaled steroids, rather than the inclusion of FeNO in the algorithm. In the Powell study, (19) the cut-off point for increasing inhaled steroids was lowered compared to previous algorithms, thus resulting in more patients being started on inhaled steroids. Together with this, the control group was treated by an algorithm that differed from current treatment guidelines in at least 2 important ways, both which resulted in less intensive treatment compared to treatment guidelines. The net effect of these algorithms was that more patients in the FeNO group received both long-acting beta-agonists and inhaled corticosteroids, although patients treated with inhaled steroids in the control group were treated at higher doses. Therefore, the differences in outcomes may be due to differences in treatment regimens that could have been achieved with or without the use of FeNO in the guidelines.
Using FeNO to guide treatment decisions in patients with respiratory disorders other than asthma
One RCT, a double-blind cross-over study by Dummer and colleagues, evaluated the ability of exhaled NO test results to predict corticosteroid response in COPD. (20) The study included 65 patients with COPD who were 45 years or older, were previous smokers with at least a 10-pack a year history, had persistent symptoms of chronic airflow obstruction, had a post-bronchodilator forced expiratory volume in one second/forced vital capacity ratio (FEV1/FVC) of less than 70% and a FEV1 of 30–80% predicted. Patients with asthma or other co-morbidities and those taking regular corticosteroids or had used oral corticosteroids for exacerbations more than twice during the past 6 months were excluded. Treatments, given in random order, were 30 mg/d of prednisone or placebo for 3 weeks; there was a 4-week washout period before each treatment. Patients who withdrew during the first treatment period were excluded from the analysis. Those who withdrew between treatments or during the second treatment were assigned a net change of zero for the second treatment period. Fifty-five patients completed the study. Two of the 3 primary outcomes, 6-minute walk distance (6MWD and FEV1) increased significantly from baseline with prednisone compared to placebo. There was a non-significant decrease in the third primary outcome, score on the St. George’s Respiratory Questionnaire (SGRQ). The correlation between baseline fraction of exhaled NO was not significantly correlated with change in 6MWD (r=0.10, p=0.45) or SGRQ (r=0.12, p=0.36) but was significantly related to change in FEV1 (r=0.32, p=0.01). At the optimal fraction of exhaled NO cut-off of 50 ppb, as determined by ROC analysis, there was a 29% sensitivity and 96% specificity for predicting a 0.2-liter increase in FEV1. (A 0.2-liter change was considered to be the minimal clinically important difference.) The authors concluded that exhaled NO is a weak predictor of short-term response to oral corticosteroid treatment in patients with stable, moderately severe COPD and that a normal test result could help clinicians decide to avoid prescriptions that may be unnecessary; only about 20% of patients respond to corticosteroid treatments. Limitations of the study include that the response to treatment measured was short term, and this was not a trial of management decisions based on exhaled NO test results.
No controlled studies were identified that evaluated the role of exhaled NO tests in the management of respiratory conditions other than asthma and COPD. A prospective uncontrolled study by Prieto and colleagues assessed the utility of exhaled oxide measurement for predicting response to inhaled corticosteroids in patients with chronic cough. (21) The study included 43 patients with cough of at least 8 weeks’ duration who were non-smokers and did not have a history of other lung disease. Patients were evaluated at baseline and after 4 weeks of treatment with inhaled fluticasone propionate 100 µg twice daily. Nineteen patients (44%) had a positive response to the treatment, defined as at least a 50% reduction in mean daily cough symptom scores. ROC analysis showed that, using 20 ppb as the FeNO cut-off, the sensitivity was 53% and the specificity was 63%. The authors concluded that exhaled NO is not an adequate predictor of treatment response.
Exhaled Breath Condensate
It appears from the published literature that exhaled breath condensate (EBC) is at an earlier stage of development compared to exhaled NO. For example, several review articles note that before routine clinical use in the diagnosis and management of respiratory disorders can be considered, the following issues must be resolved (22-25):
- Standardization of collection and storage techniques
- Effect of dilution of respiratory droplets by water vapor
- Effect of contamination from oral and retropharyngeal mucosa
- Understanding how particles/droplets form and change during exhalation before leaving the body
- Techniques of measuring concentrations of nonvolatile substances in EBC; in most cases, these concentrations are very low, which may be at the lower limits of detection of conventional analytic techniques
- Variability in EBC assays for certain substances
- Lack of gold standard for determining absolute concentrations of airway lining fluid non-volatile constituents to compare with EBC.
No controlled studies were identified that evaluated the role of EBC tests in the management of asthma or other respiratory disorders. Representative published studies include a 2009 case series investigating whether components of EBC could predict response to steroid treatment in patients with asthma. (26) Eighteen steroid-naive asthma patients were included; EBC collection, spirometry, and methacholine challenge were performed before and 12 weeks after inhaled steroid therapy (equivalent dose of 400 µg fluticasone propionate/d). Among the molecules in EBC examined, higher IL-4 and RANTES levels and lower IP-10 levels at baseline were correlated with an improvement in FEV1. The study had a small sample size, was uncontrolled, and did not address the clinical utility of EBC testing.
In addition, a 2010 study by Antus and colleagues evaluated EBC in 58 hospitalized patients (20 with asthma and 38 with COPD) and 36 healthy controls (18 smokers and 18 non-smokers). (27) The EBC pH was significantly lower in patients with asthma exacerbations (all non-smokers) at hospital admission compared to non-smoking controls (6.2 vs. 6.4, respectively, p<0.001). The pH of EBC in asthma patients increased during the hospital stay and was similar to that of non-smoking controls at discharge. Contrary to investigators’ expectations, EBC pH values in ex-smoking COPD patients (n=17) did not differ significantly from non-smoking controls, either at hospital admission or discharge. Similarly, pH values in EBC samples from smoking COPD patients (n=21) at admission and discharge did not differ significantly from smoking controls.
Practice Guidelines and Position Statements
American Thoracic Society (ATS): In 2011, the ATS published a clinical practice guideline on interpretation of FeNO levels. (28) The guideline was critically appraised using criteria developed by the Institute of Medicine (IOM), which includes 8 standards. (29) The guideline was judged to not adequately meet the following standards: Standard 3: guideline development group composition; Standard 4: clinical practice guideline-systematic review intersection; Standard 5: Establishing evidence foundation for and rating strength of recommendations; and Standard 7: external review.
The ATS guideline included the following strong recommendations (if not otherwise stated, the recommendations apply to asthma patients):
- We recommend the use of FENO in the diagnosis of eosinophilic airway inflammation (strong recommendation, moderate quality of evidence).
- We recommend the use of FENO in determining the likelihood of steroid responsiveness in individuals with chronic respiratory symptoms possibly due to airway inflammation (strong recommendation, low quality of evidence).
- We recommend accounting for age as a factor affecting FENO in children younger than 12 years of age (strong recommendation, high quality of evidence).
- We recommend that low FENO less than 25 ppb (<20 ppb in children) be used to indicate that eosinophilic inflammation and responsiveness to corticosteroids are less likely (strong recommendation, moderate quality of evidence).
- We recommend that FENO greater than 50 ppb (>35 ppb in children) be used to indicate that eosinophilic inflammation and, in symptomatic patients, responsiveness to corticosteroids are likely (strong recommendation, moderate quality of evidence).
- We recommend that FENO values between 25 ppb and 50 ppb (20–35 ppb in children) should be interpreted cautiously and with reference to the clinical context. (strong recommendation, low quality of evidence).
- We recommend accounting for persistent and/or high allergen exposure as a factor associated with higher levels of FENO (strong recommendation, moderate quality of evidence).
- We recommend the use of FENO in monitoring airway inflammation in patients with asthma (strong recommendation, low quality of evidence).
American Thoracic Society/European Respiratory Society: A 2009 statement includes the following key points on exhaled nitric oxide:
“The clinical utility of FeNO-based management strategies has not been explored extensively. Currently available evidence suggests a role in identifying the phenotype in airways disease, particularly in the identification of corticosteroid responsiveness. Due to logistic and cost issues, FeNO is the only biomarker likely to have a role in primary care-based asthma studies, although it is possible that with technological improvements, other techniques including sputum induction could have a role in the medium term.” (1)
National Heart Lung and Blood Institute (NHLBI): Their 2007 expert panel guidelines for the diagnosis and management of asthma state:
“Use of minimally invasive markers (“biomarkers”) to monitor asthma control and guide treatment decisions for therapy is of increasing interest. Some markers, such as spirometry measures, are currently and widely used in clinical care; others, such as sputum eosinophils and FeNO, may also be useful, but they require further evaluation in both children and adults before they can be recommended as clinical tools for routine asthma management (Evidence D).”
“The Expert Panel recommends some minimally invasive markers for monitoring asthma control, such as spirometry and airway hyper-responsiveness, that are appropriately used, currently and widely, in asthma care (Evidence B). Other markers, such as sputum eosinophils and FeNO, are increasingly used in clinical research and will require further evaluation in adults and children before they can be recommended as a clinical tool for routine asthma management (Evidence D).” (30)
Evaluation of exhaled nitric oxide and exhaled breath condensate are proposed as techniques to diagnose and monitor asthma and other respiratory conditions. Several prospective studies have addressed FeNO measurement; however, there is still no standardized and validated cut-off to use for diagnosing asthma. Multiple randomized controlled studies have evaluated the use of FeNO tests for the management of patients and have not consistently found improvement in health outcomes. Moreover, a 2009 Cochrane review pooling results of studies evaluating FeNO in the management of patients with asthma found a high degree of variability among studies and did not recommend routine use of FeNO in clinical practice. A 2011 RCT of pregnant women with asthma found better outcomes in the group managed using a FeNO algorithm than standard care
. . In this study, as in many others, there are concerns that differences in treatment regimens that arise as a result of different algorithms may confound the outcomes, particularly in cases where the control algorithm may lead to undertreatment.
There is less evidence on the utility of FeNO for the diagnosis and management of other respiratory disorders. There are also few studies on exhaled breath condensate evaluation for the diagnosis and treatment of asthma and other conditions. Thus, the evidence is insufficient to determine the effect of exhaled nitric oxide and exhaled breath condensate tests on health outcomes.
Disclaimer for coding information on Medical Policies
Procedure and diagnosis codes on Medical Policy documents are included only as a general reference tool for each policy. They may not be all-inclusive.
The presence or absence of procedure, service, supply, device or diagnosis codes in a Medical Policy document has no relevance for determination of benefit coverage for members or reimbursement for providers. Only the written coverage position in a medical policy should be used for such determinations.
Benefit coverage determinations based on written Medical Policy coverage positions must include review of the member’s benefit contract or Summary Plan Description (SPD) for defined coverage vs. non-coverage, benefit exclusions, and benefit limitations such as dollar or duration caps.