Test of reversibility of airway obstruction: Time for a review?

The most commonly used pulmonary function test in clinical practice is spirometry. In patients found to have airway obstruction, evaluation of acute response to bronchodilators (the test of reversibility of airway obstruction) is a commonly used procedure in clinical and research studies. The results of this test are used to take decisions on treatment, inclusion in/exclusion from drug trials and other research studies, and for diagnostic labeling [asthma vs. chronic obstructive pulmonary disease (COPD)]. Usually, the forced expiratory volume in 1 s (FEV1) or forced vital capacity (FVC) values before and after administration of the bronchodilator are compared and the change is computed. While there is no clear consensus on what constitutes reversibility or a significant acute response to a bronchodilator, the most often used criteria for a bronchodilator response (BDR) are those recommended by the American Thoracic Society (ATS), i.e. an increase of 12% and 200 ml in FEV1 or FVC over the baseline value.[1] In an analysis of the Burden of Obstructive Lung Disease (BOLD) study data, the estimated 95th percentiles (95% CI) for a change after bronchodilator were found to be 284 ml (263-305) and 12.0% (11.2- 12.8%) increase in FEV1 over baseline in 3922 healthy, never smokers. The corresponding mean changes in FVC were 322 ml (271-373) and 10.5% (8.9-12.0%) relative to baseline value.[2] Therefore, a positive BDR exceeding the threshold defined by the ATS criteria may be considered to be a clinically significant improvement, not likely to be observed in healthy normal subjects.

Evidence over the last few years has raised questions over the application of the test of reversibility as a decision-making tool in areas where it has enjoyed traditional acceptance. It is well known that airway obstruction in asthma usually shows a good response to bronchodilators while that in COPD is generally poor. The above cut-off is therefore often used to diagnose these diseases. Generations of students have been taught that asthma and COPD can be differentiated by the test of BDR, i.e., an increase of 12% and 200 ml in FEV1 or FVC over the baseline value or more than this being diagnostic of the former and a lesser response diagnosing COPD. However, it has been pointed out in literature that “reversibility versus irreversibility” is not an appropriate approach in making a distinction between the two diseases, as “irreversible” obstruction is well known in asthma and many COPD patients have a substantial reversibility.[34]

A few years back, we showed that bronchodilator responsiveness in asthma and COPD differs not only quantitatively but also in the pattern that may be related to differences in the pathophysiology of the two diseases. In COPD, an exclusive FVC response (post-bronchodilator increase only in FVC) was the predominant response, seen in a substantial number of patients, while an exclusive FEV1 response (post-bronchodilator increase only in FEV1) was uncommon. Asthmatics tended to respond with an increase in both FVC and FEV1.[5] More recently, we have shown that the test of acute bronchodilator responsiveness has limited diagnostic value in separating asthma and COPD. Positive predictive values of 81% observed in this study with the ATS criteria implied that one can be sure of a diagnosis of asthma to this extent when the BDR exceeds this threshold and the negative predictive values of 64% indicated the probability of a diagnosis of COPD when the response was less than the threshold. Thus, this cut-off left a substantial scope for misdiagnosis.[6] These results were the consequence of a substantial overlap in the response to bronchodilator between asthma and COPD, and the fact that a fairly large proportion of subjects with asthma did not show acute bronchodilator responsiveness while a large proportion of COPD patients had a good reversibility. Given the limited diagnostic value of this test, there appears to be little justification for using a particular threshold of response for differentiating asthma from COPD or as an inclusion or exclusion criterion for studies on patients with asthma and COPD.

The recent updates of the guidelines on asthma and COPD have also emphasized that the diagnosis of these diseases is largely clinical and the results of spirometry need to be interpreted in the clinical context.[7-9] The National Institute of Clinical Excellence (NICE) COPD guidelines update of 2010[7] and the Global Initiative for Chronic Obstructive Lung Disease (GOLD) revised update of 2011[8] clearly state thus: “The degree of reversibility of airflow limitation is no longer recommended for diagnosis or predicting long-term response to bronchodilators or corticosteroids.” The Global Initiative for Asthma (GINA) update of 2011 recognizes that while a response to bronchodilator greater than the ATS cut-off supports the diagnosis of asthma in the clinical context, the test lacks sensitivity.[9] The BOLD data on worldwide distribution of BDR cited above revealed that 11.1% of patients of COPD GOLD stage 2 and above exceeded the absolute volume threshold and 30.8% exceeded the percent change over baseline threshold for FEV1. For FVC response, the corresponding figures were 22.6% and 28.6%, respectively.[2] This data effectively buries the long-standing belief that patients of COPD lacked a significant BDR.

Bronchodilators are the treatment of choice in COPD. It is the experience of every physician caring for such patients that these drugs produce a variable but significant improvement in a proportion of patients over time. This improvement may be evident in an improved FVC or inspiratory capacity (IC) (reflecting reduced air trapping and hyperinflation) as well as in patient-centered outcomes such as dyspnea, quality of life, and exercise tolerance. It would indeed be paradoxical to label bronchodilators as the mainstay of pharmacological treatment if COPD was not responsive to bronchodilators. Only we are unable to demonstrate a BDR during acute testing of reversibility of airway obstruction in a large proportion of patients. Even this acute BDR is not reproducible. A subject labeled as “nonresponder” on one occasion may show a BDR on another occasion and would be classified as a responder.[10] It is also erroneously believed that patients of COPD showing a BDR after administration of a bronchodilator will respond better to treatment. This implies that the test can pick up a distinct phenotype. Both the GOLD and NICE guidelines have stated that this has not been substantiated in studies. The long-term response to bronchodilators cannot be predicted by the acute BDR during the test of reversibility. As observed in a recent study, patients with a BDR do not differ in mortality, hospitalization, or exacerbation experience from “irreversible” patients. Thus, bronchodilator reversibility status does not distinguish clinically relevant outcomes, making it an unreliable phenotype.[11]

Often, the ATS criteria are used for inclusion and exclusion of subjects in research studies on asthma and COPD. Given the limitations of the test of reversibility as a diagnostic tool, application of these criteria no longer seems necessary. The diagnosis needs to be made as described in the recent updates of the guidelines on asthma and COPD. Similarly, the choice of treatment also should not take into account the acute BDR.

It is clearly time to review how we have been applying the test of reversibility of airway obstruction in clinical usage and research. The traditional applications lack evidence to support these. The limitations of a BDR need to be recognized. However, it needs to be emphasized that administration of bronchodilator is still required as part of pulmonary function evaluation. The diagnosis and classification of severity of COPD is based on post-bronchodilator spirometry. In diagnosing COPD, the change after bronchodilator itself is not relevant, but only the post-bronchodilator values are taken into consideration. The way we use the information obtained from the test of reversibility of airway obstruction requires a change consistent with the current evidence and consensus.