Bronchodilator reversibility in chronic obstructive pulmonary disease: use and limitations

Bronchodilator reversibility in chronic obstructive pulmonary disease: use and limitations

Review Bronchodilator reversibility in chronic obstructive pulmonary disease: use and limitations Peter M A Calverley, Paul Albert, Paul P Walker Lan...

386KB Sizes 0 Downloads 11 Views

Review

Bronchodilator reversibility in chronic obstructive pulmonary disease: use and limitations Peter M A Calverley, Paul Albert, Paul P Walker Lancet Respir Med 2013; 1: 564–73 Published Online June 18, 2013 http://dx.doi.org/10.1016/ S2213-2600(13)70086-9 School of Ageing and Chronic Disease, University of Liverpool, Liverpool, UK (Prof P M A Calverley DSc); and Aintree Chest Centre, University Hospital Aintree, Liverpool, UK (P M A Calverley, P Albert MD, P P Walker MD) Correspondence to: Prof Peter M A Calverley, Clinical Science Centre, University Hospital Aintree, Longmoor Lane, Liverpool L9 7AL, UK [email protected]

The change in forced expiratory volume in 1 s (FEV₁) after administration of a short-acting bronchodilator has been widely used to identify patients with chronic obstructive pulmonary disease (COPD) who have a potentially different disease course and response to treatment. Despite the apparent simplicity of the test, it is difficult to interpret or rely on. Test performance is affected by the day of testing, the severity of baseline lung-function impairment, and the number of drugs given to test. Recent data suggest that the response to bronchodilators is not enhanced in patients with COPD and does not predict clinical outcomes. In this Review we will discuss the insight that studies of bronchodilator reversibility have provided into the nature of the COPD, and how the abnormal physiology seen in patients with this disorder can be interpreted.

Introduction Chronic obstructive pulmonary disease (COPD) is the term used to describe a persistent and generally progressive illness that results from a respiratory bronchiolitis, beginning in the most peripheral airways,1 and that is often accompanied by alveolar loss (emphysema).2 These pathological changes delay lung emptying at rest and even more so during exercise3 and are characterised by the presence of abnormal obstructive spirometry, customarily defined as a reduction in the forced expiratory volume in 1 s (FEV1)/forced vital capacity (FVC) ratio to 0·7 or less. However, this criterion is inappropriate for use in elderly patients4 and has led some experts to advocate the use of an age-related lower limit of normal for this ratio.5,6 The degree to which the

Key messages • Small changes in FEV1 after a bronchodilator in individuals with mild chronic obstructive pulmonary disease (COPD) can challenge the diagnosis of COPD • The absolute change in FEV1 post-bronchodilator varies with the day of testing, the number of bronchodilator drugs used to test, and drug dose • The absolute increase in FEV1 post-bronchodilator is similar in individuals with moderate COPD and healthy smokers • The absolute increase in FEV1 post-bronchodilator decreases as baseline FEV1 decreases and so does the chance of being classified as reversible • Although the average number of reversible patients in a COPD population is stable over time, individuals change their reversibility status on repeat testing • Reversibility status does not identify patients with a different clinical course or response to treatment nor does it add to the baseline FEV1 in the prediction of patients with more rapid disease progression • In patients with clinical and spirometric evidence of COPD, reversibility testing adds little to management. However, in patients who have atypical clinical features, reversibility testing is still warranted

564

FEV1 falls below the predicted normal value has been used as a marker of disease severity, although recent treatment guidance has emphasised the need to base decisions about management on clinical factors as well as spirometric impairment.7 Theoretically, spirometric data should be measured after an inhaled bronchodilator is given. This approach has been applied in recent international studies of COPD prevalence8,9 and remains the recommended method to assess disease severity.7 COPD differs from bronchial asthma, which is also associated with obstructive spirometry, because patients with COPD do not show substantial variability in lung function either spontaneously or in short-term response to treatment. However, this has not prevented many clinicians and investigators10 from trying to identify discrete phenotypes of patients with COPD who show a greater than anticipated change in spirometry after treatment with a short-acting bronchodilator drug. A positive response in the bronchodilator reversibility test has been suggested to identify differences in the clinical course and therapeutic response of patients with COPD that have not been prospectively established. Nobel prize-winning psychologist Daniel Kahneman noted that it is easier to substitute the answer to an easy question when a hard one is asked. 11 In this case, doctors have extrapolated their observation of an immediate change in lung function after administration of bronchodilators into an expectation that this will predict clinical outcome. We believe that there is now good evidence that this is not the case. The topic of bronchodilator reversibility in patients with COPD has been reviewed previously,12,13 but recently published data allow this commonly used mode of testing to be set into a wider context. We clarify what can and cannot be ascertained with confidence from this widely used but surprisingly confusing test.

How to define bronchodilator reversibility The term bronchodilator reversibility implies the complete or near complete correction of an obstructive spirometric abnormality. Such changes can occur in patients with bronchial asthma but are not always seen. Normalisation www.thelancet.com/respiratory Vol 1 September 2013

Review

of spirometry after administration of a bronchodilator is not seen in patients with COPD unless the baseline FEV1 is close to the predicted normal value before the drug is given. In this case, a small absolute increase in FEV1 can mean that the post-bronchodilator value is no longer below the predicted normal value or associated with an FEV1/FVC ratio less than 0·7. This effect has been noted in population studies that identified many people with mild airflow obstruction.14 These data show how small changes in absolute volume can affect a binary categorical variable such as reversibility in patients with an FEV1/FVC ratio close to the threshold value. Such a complete normalisation of lung function is not (by definition) recorded in more advanced disease because a patient showing this response on spirometric testing would be regarded as asthmatic. However, many patients with COPD show some improvement in FEV1, FVC, or both, after a bronchodilator, and this response to treatment can reach the thresholds believed to represent reversibility. Moreover, almost all studies of COPD have focused on FEV1 as the defining variable rather than in conjunction with changes in other variables such as FVC or the FEV1/FVC ratio. In the past, it has been difficult to determine the size of a change in lung function after a bronchodilator in healthy individuals. Initial estimates were derived from patients with less severe COPD or modest sized Caucasian population samples.15,16 This problem has been overcome by the Burden of Obstructive Lung Disease study,17 which

reported the response to inhaled salbutamol in representative population samples from 14 centres worldwide. The results of the study suggest that an increase in FEV1 of more than 300 mL is very unlikely to arise by chance, and that this equates to a 12% increase from baseline in people with normal lung function, as suggested earlier by expert groups.18 A high degree of variation in FVC measurements was noted between centres, which increases the uncertainty of the significance of a positive test based on this measurement. Similar absolute changes in FEV1 were seen in the healthy comparator data reported by the Evaluation of COPD Longitudinally to Identify Predictive Surrogate Endpoints (ECLIPSE) investigators.19 However, in the ECLIPSE study, the change in FEV1 recorded in the healthy smoker group was similar to that reported in the patients with COPD who were the main focus of the ECLIPSE programme (figure 1). The definition of bronchodilator reversibility has three components. The first is the assessment of the short-term (<60 min after the test) change in lung function, usually FEV1, after inhalation of a short-acting bronchodilator drug. The short-acting β agonist salbutamol or the short-acting antimuscarinic agent ipratropium, or both, are the most common test drugs. Second, the change in FEV1 needs to be greater than what would be expected by chance; the usual way to express this is as a change greater than 12% of the baseline value. However, this calculation implies that the change in FEV1 after a

Patient with COPD (%)

35 30 25 20 15 10 5 0

Smoker controls (%)

Patient group

35 30 25 20 15 10 5

Non-smoker controls (%)

0 35 30 25 20 15 10 5 0 –0·65 –0·55 –0·45 –0·35 –0·25 –0·15 –0·05 0·05

0·15

0·25

0·35

0·45

0·55

0·65

0·75

0·85

0·95

1·05

1·15

1·25

1·35

Change in FEV1 (L) post-bronchodilator

Figure 1: Distribution of absolute changes in FEV1 after salbutamol in ECLIPSE participants Frequency distribution of change in FEV1 after 400 μg inhaled salbutamol in 1831 patients with chronic obstructive pulmonary disease (top), 285 healthy smokers (mid), and 228 non smokers (lower). Data from Albert and colleagues.19 FEV₁=forced expiratory volume in 1 s. FVC=forced vital capacity.

www.thelancet.com/respiratory Vol 1 September 2013

565

Review

bronchodilator will vary with lung size, something for which no real supporting data exist. Hence, if the baseline value is high, a large absolute change in FEV1 is needed for a positive response, but small changes close to the between-test variability in the measurement can be enough to suggest reversibility when the pre-test FEV1 is low. To overcome this requirement, a third pragmatic component was added and after the Intermittent Positive Pressure Breathing (IPPB) study,20 namely a requirement of at least a 12% baseline increase plus an additional 200 mL absolute change. Concerns about the overdiagnosis of reversibility on the basis of a percentage change from baseline in COPD led to a proposal by European investigators that reversibility is present when the FEV1 changes by 9% of the predicted value for that patient. This approach was initially supported by the European Respiratory Society21 because the use of a simple percentage increase removes the dependence of the result on the pre-test value of FEV1. However, partly because of the relative complexity of the definition, recommendations have reverted to the original definition described above.22 The absence of bronchodilator reversibility has been an important feature in COPD treatment trials in Europe23,24 but not North America.25,26 In North America, patients have previously been selected for trials because they showed an acute response to treatment.27 Patients with reversible COPD were more likely to report a response to oral corticosteroids than were patients without reversible COPD;28 however, this form of response testing was not a specific marker of disease progression on longer term follow up.29 Reversibility status has been linked to specific chromosomes in genetic studies of COPD,30 and findings from some studies show that patients with reversible COPD have an accelerated loss of lung function compared with those who do not have reversible COPD.31-33 However, to understand these associations and develop clinically useful recommendations based on the results of reversibility testing in individual patients has proven to be much harder than first believed.

Drawbacks of reversibility tests in COPD Large clinical studies have examined the usefulness and performance characteristics of bronchodilator reversibility testing across a wide range of COPD severities defined by spirometry.14,17,19,20,31,33–38 The table describes the characteristics of several of the most important studies. These studies have identified issues in many key areas, which are discussed in turn.

Drug, drug dose, and timing of test The most widely used drug is salbutamol in a dose of 200–400 μg with a 15 min delay before re-testing. This approach is pragmatic and uses the shortest time between tests consistent with the known pharmacological actions of this drug. However, larger doses of β agonist will produce small additional increments in FEV1 and will 566

thereby change the number of patients with a reversible response.39 Similar considerations apply to short-acting antimuscarinics that have slightly slower onsets of action and therefore necessitate a greater delay before repeating spirometry testing.40 Combination of drugs from each drug class further increases the responder rate (figure 2).34,41 Waiting longer before re-testing FEV1 after the test dose will probably result in the detection of a greater change in FEV1 in some patients and hence increase the number of patients regarded as having a reversible response. How often this situation arises and what effect it might have with different drugs and different background disease severity has not been systematically explored. Differences in test protocol contribute to differences in the reported prevalence of reversible disease, which varies between 5% in the Lung Health Study population33 to more than 50% in the UPLIFT trial.42 Another important factor to consider in the interpretation of clinical trial data is how well patients adhere to omitting their normal bronchodilator therapy before spirometric testing. If the test drug is given on top of an active bronchodilator, the FEV1 change is likely to be less than the change seen in a patient who has successfully omitted their usual bronchodilator treatment. The importance of this effect has not been explored. However, patients in the ECLIPSE study who were asked to omit their usual tiotropium for 24 h did not show any differences in response from those not taking this drug.19

Reproducibility of reversibility status To be clinically useful, the attribution of a characteristic to an individual (the phenotype) needs to be stable from day to day, and in this regard bronchodilator reversibility fails. Although the percentage of patients classified as responders in any population is stable when re-tested over 2–12 months, figure 3 shows how the individual response varies between occasions.34 Changes in the definition of a positive response, whether by using the percentage predicted method described previously or methods requiring a larger change in absolute volume (eg, a 400 mL change in FEV1), do not eradicate this tendency to individual variation. However, the definition used will determine the percentage of the population regarded as responsive.19 Therefore, treatment guidelines and management strategies no longer advocate bronchodilator testing as part of routine assessment of patients with COPD.32,43

Prediction of bronchodilator-responsiveness The most consistent predictor of reversibility status is the pre-test FEV1. Generally, the lower the FEV1, the lower the chance a patient has of having a positive reversibility test, irrespective of whether one or two drugs are used in testing.19,34,35 Data from the ECLIPSE study19 showed that although having a lower pre-test FEV1 increased the chance of exceeding the 12% change threshold, few of the patients with the worst pre-test spirometry had a www.thelancet.com/respiratory Vol 1 September 2013

Review

Number of Method of patients reversibility

COPD severity

Reversibility criteria

Clinical predictors of reversibility

Clinical outcomes associated with reversibility

Mean prebronchodilator FEV₁=1·25 L (SD 0·49), 45% (SD 15) predicted; GOLD II=848 (46%) GOLD III=750 (41%) GOLD IV=233 (13%)

ATS/ERS: FEV₁ increase ≥12% plus 200 mL

GOLD stage: GOLD II mean 0·16 mL (SD 0·17) FEV₁ increase; GOLD III mean 0·1 mL (SD 0·13) FEV₁ increase; GOLD IV 0·05 mL (SD 0·08) FEV₁ increase Gender mean for men 0·13 mL (SD 0·16) FEV₁ increase vs women 0·1 mL (SD 0·12) increase; odds ratio (OR) 1·79 (95% CI 1·37–2·31) Extent of emphysema on CT: association r=0·09; p<0·001; no association with age, smoking status, or cigarette pack-years

Decreased FEV₁: additional 17 mL (SD 4) annual decrease in FEV₁ in reversible patients but mean baseline FEV₁ 220 mL (SD 22) higher

Three criteria used: FEV₁ absolute change in m, change expressed as a percentage of the pre-bronchodilator value, and change expressed as a percentage of predicted normal FEV₁

Methacholine reversibility: p<0·001 on all Decreased FEV₁, no association when baseline data excluded from assessment three criteria; age: p<0·001 all three criteria; pack-years: p<0·001 all three criteria; gender: no association; no association with quit status (sustained quitter, intermittent quitter, and continued smoker)

ECLIPSE19,32 1831

15 min after 400 μg salbutamol

LUNG HEALTH33

4194

10 min after Mean pre-bronchodilator 200 μg FEV₁=2·64 L (SD 0·6), isoprotenerol 75% (SD 9) predicted; mean postbronchodilator FEV₁=2·75 L (SD 0·6), 79% (SD 9) predicted

ISOLDE34

660

30 min after 400 μg salbutamol, then 30 min after 80 μg ipratropium

Mean prebronchodilator FEV₁=1·28 L (SD 0·46), 46% (SD 15) predicted

Three criteria used: FEV₁ absolute No association between absolute change change in mL, change expressed in FEV₁ (mL) vs smoking status, atopy, or gender as a percentage of the pre-bronchodilator value, and change expressed as a percentage of predicted normal FEV₁

No association between absolute change in FEV₁ (mL) vs FEV₁ decrease, worsening of health status measured by SGRQ, or exacerbation rate

UPLIFT35,37

5756

60 min after 80 μg ipratropium then 30 min after 200 μg salbutamol

Mean pre-bronchodilator FEV₁=1·1 L, 39% (SD 12) predicted; mean post-bronchodilator FEV₁=48% (SD 13) predicted; GOLD II=47%, GOLD III=45%, GOLD IV=9%

Three criteria used: criteria a FEV1 increase ≥12% plus 200 mL (52% reversible by this criteria), criteria b FEV₁ increase by more than 15% over baseline (66% reversible by this criteria), criteria c ≥10% absolute increase in % predicted FEV₁ (39% reversible by this criteria)

Criteria a: age 64 years (SD 8) reversible group vs 65 years (SD 8) irreversible (p<0·001); gender 79% reversible patients male vs 69% irreversible patients male (p<0·001); pack-years 50 years (SD 29) reversible patients vs 47 years (SD 27) irreversible (p<0·001) Criteria b: BMI 25·9 (SD 5) reversible patients vs 26·3 (SD 5·2) irreversible (p<0·001); present smoking status 29% reversible patients smoked vs 34% irreversible (p<0·001) Criteria c: no significant association including duration of COPD

Criteria a: health status measured by SGRQ 44 (SD 17) reversible patients vs 48 (SD 17) irreversible (p<0·001); exacerbation rate 1 year or longer 67·9% reversible patients vs 68·5% irreversible; all-cause mortality 10·8% reversible group vs 16·2% irreversible Criteria b: health status measured by SGRQ no association; exacerbation rate 1 year or longer 67·6% reversible patients vs 68·5% irreversible; all-cause mortality 12·5% reversible group vs 15·1% irreversible Criteria c: health status measured by SGRQ 44 (SD 17) reversible patients vs 47 (SD 17) irreversible (p<0·001); exacerbation rate 1 year or longer 67·5% reversible patients vs 69·3% irreversible; all-cause mortality 9·4% reversible group vs 15·9% irreversible

PLATINO38

5571

15 min after 200 μg salbutamol

Mean FEV₁ 2·49 L, 378 of 5571 (7%) patients reversible

ATS/ERS: FEV₁ increase ≥12% plus 200 mL

.. 15% patients taking inhaled therapy (defined as having taken a bronchodilator or inhaled corticosteroid within the previous 1 year) reversible vs 7% patients not taking therapy (p<0·001)

NETT36

544

15 min after 116 μg salbutamol

Mean prebronchodilator FEV₁=24% (SD 7) predicted

ATS/ERS: FEV₁ increase ≥12% plus 200 mL (121 of 544 [22%] patients met criteria on at least one occasion); FEV₁ ≥400 mL (10 of 544 [2%] patients met criteria on at least one occasion)

.. Patients who met ATS/ERS reversibility criteria on one or more occasions (reversible) vs those who were never reversible: no association with age; TLC (post-bronchodilator; 126% predicted in reversible patients vs 130% irreversible, p<0·005); gender (11% female reversible vs 29% male reversible) OR 2·37 (1·37–4·12); p=0·002; DLco 32% (SD 10) reversible patients vs 28 (10)% irreversible (p<0·001); no association with pack-years or extent of emphysema on CT

IPPB20

985

250 μg Pre-bronchodilator isoproterenol FEV₁=1·03 L (36·1% predicted), mean FEV₁ improvement with bronchodilator=14·5%

FEV₁ increase >12%

..

Patients who showed >12% improvement in FEV₁ (reversible): FEV₁ decrease: mean 52 mL annual decrease in FEV₁ in irreversible patients vs 27 mL annual decrease in reversible; no association with mortality or hospitalisation

Results from and response criteria used in selected studies of bronchodilator reversibility in patients with COPD, including the association between bronchodilator reversibility status and subsequent clinical outcomes. The missing data in the right columns is because these studies did not relate reversibility testing to clinical outcomes. COPD=chronic obstructive pulmonary disease. FEV1=forced expiratory volume in 1 s. ATS/ERS=American Thoracic Society/European Respiratory Society. SGRQ=St George’s Respiratory Questionnaire. TLC=total lung capacity. DLco=diffusion coefficient for carbon monoxide.

Table: Selected studies reporting bronchodilator reversibility testing in chronic obstructive pulmonary disease

www.thelancet.com/respiratory Vol 1 September 2013

567

Review

FEV1 pre-bronchodilator FEV1 post salbutamol or ipratropium FEV1 post both

1·6

FEV1 (L)

1·5 1·4 1·3 1·2 1·1 Visit 0 (salbutamol)

Visit 1 (ipratropium)

Visit 2 (both)

Figure 2: Absolute change in FEV1 after salbutamol and ipratropium Mean (SE) FEV1 before and after 400 μg salbutamol, 80 μg ipratropium, or both on three occasions at monthly intervals in 660 stable patients with chronic obstructive pulmonary disease. Reproduced with permission from Calverley and colleagues,34 by permission of the BMJ Group. FEV₁=forced expiratory volume in 1 s.

200 mL increase in FEV₁ post salbutamol, the absolute change in FEV₁ ranging from 160 ml in GOLD grade 2 patients to 50 mL in GOLD 4. Findings from other studies have shown a similar GOLD grade-related reduction in reversibility.17,35 Many other variables may be useful for the identification of patients likely to respond to bronchodilator drugs. The presence of sputum eosinophilia might identify patients with COPD who are more likely to respond to oral corticosteroids,44 although whether such a response predicts longer term clinical outcomes has been challenged.29 No consensus exists on the prevalence of sputum or blood eosinophilia or both in clinically diagnosed patients with COPD. Furthermore, extensive data are not available in patients with COPD on the day to day variation of measures of eosinophilia. Neither gender, present smoking status, nor atopy consistently distinguish responsive from unresponsive patients.34 The National Emphysema Treatment Trial (NETT)36 investigators found that the presence of CT-defined emphysema reduced the chance of a patient being classed as reversible. However, this study recruited patients with a much lower baseline FEV1 and a high likelihood of emphysema. The relation between reduction in CT-defined lung density and reversibility status was much weaker in the ECLIPSE study of patients in secondary care clinics.19 The table shows a selection of studies that examined predictors of bronchodilator responsiveness. These data clearly suggest that apart from pre-bronchodilator FEV1, no consistent or reliable predictor of response, and certainly no predictor that is useful in day-to-day clinical practice exists.

Clinical implications of reversibility testing In view of the poor day to day reproducibility of reversibility status, the fact that reversibility status does not predict clinical progress in patients with COPD is not surprising.19,34 The table shows the main outcomes of reversibility testing in stable COPD considered together 568

with the studies in which they were reported. Two of these merit special attention. First, to see if patients who never showed a reversible response to bronchodilators differed from those who sometimes had a positive reversibility test, the ECLIPSE investigators compared 1362 patients who never met the standard criterion of reversibility in 4 tests over 1 year with 227 who did show a response to bronchodilators on one or more occasions. With this strict and rather clinically impractical definition, no association existed between reversibility and either baseline health status, change in health status over time, or mortality. However, patients who never responded were more likely to exacerbate in the 2 years of follow up after testing. This finding was attributable to the confounding effects of pre-bronchodilator lung function because the patients with decreased pre-test FEV1 were also less likely to reverse and had more exacerbations than patients not showing reversibility. The relation between reversibility status and exacerbations was lost when allowance was made for baseline FEV₁. Data from earlier studies suggested that reversibility status predicted the rate of decrease in FEV1, but this was only true for decrease in pre-bronchodilator FEV1.31,45 The ECLIPSE data suggest that the association between decrease in post-bronchodilator FEV1 over time and reversibility was more robust, but was again confounded by patients with less severe lung function impairment showing both a more rapid loss of lung function and an increased chance of being classed as reversible.32,46 The poor clinical stability of the reversibility test in patients in GOLD 2 (unpublished data) makes it more likely that decline in lung function is mainly driven by pre-bronchodilator FEV1 rather than reversibility status.

Physiological mechanisms associated with bronchodilator responsiveness Whatever their mode of action, inhaled bronchodilators produce rapid relaxation of airway smooth muscle and improve the expiratory flow rate. In COPD, in which increased peripheral airways resistance and a variable loss of lung elastic recoil exists, expiratory flow limitation (a disorder in which the expiratory flow rate cannot be increased or falls despite increasing effort) often occurs during tidal breathing.47,48 Flow limitation develops earlier during a forced expiratory manoeuvre in patients with COPD than in healthy individuals who only have flow limitation at low lung volumes. In patients with COPD, inhaled bronchodilators decrease both total and respiratory system resistance, as shown during tidal breathing with the effort-independent forced oscillation method.49,50 However, bronchodilators have much less effect on expiratory than inspiratory resistance, and hence FEV1, when flow limitation is present and so seem to be less effective in these patients.50 The most important physiological effect of bronchodilators is to decrease end-expiratory lung volume usually without any effect on www.thelancet.com/respiratory Vol 1 September 2013

Review

1831

Visit 1: 76% not reversible

1394

Visit 2: 77% not reversible

Visit 3: 77% not reversible

Visit 4: 79% not reversible

Not reversible Reversible

437

1167

1019

923

96

227

148

109

238

147

39

111

80

36

37

199

162

43

123

76

39

43

76

33

48

123

28

51

72

Figure 3: Repeated reversibility testing over one year in stable COPD Reproducibility of reversibility classification (American Thoracic Society/European Respiratory Society criteria) in 1831 patients with chronic obstructive pulmonary disease on four occasions over 1 year. The percentage not reversible at each visit is shown on the left. Reproduced from Albert and colleagues,19 by permission of the BMJ Group.

tidal expiratory flow limitation.39,50,51 This fall in lung volume allows patients to exercise for longer before dynamic hyperinflation reaches the critical point where inspiratory reserve volume is compromised and dyspnoea becomes severe.52,53 These more subtle physiological effects help to explain why changes in FEV1 after a bronchodilator that fall within the between-test reproducibility of the measurement can translate into the clinically relevant improvements in exercise capacity, health status, exacerbation frequency, and even in symptom intensity during exacerbations.42,54–56 In fact, changes in end-expiratory lung volume after administration of a short-acting bronchodilator track changes in residual volume and the change in FVC, irrespective of GOLD grade. This contrasts with FEV1 change which, as noted above, decreases in magnitude as the percent predicted pre-test value falls.19 Thus, change in FEV1 is an indirect marker of the physiologically important effect of the bronchodilator, and as disease becomes more severe, the change in end-expiratory lung volume and FEV1 become less closely related. Moreover, this differential effect on FEV1 and FVC can produce a seemingly paradoxical decrease of FEV1/FVC ratio as shown in figure 4. This discrepancy results in some patients having an isolated FVC response,57,58 which has been associated with emphysema59 as was seen in the NETT data.36 In these circumstances, the improvement in lung volume outweighs the deterioration in forced lung emptying in terms of clinical benefit. In view of the complexity of these processes, the fact that acute changes in one indirect measure of lung function post-bronchodilator do not predict clinical benefit is not surprising. The absolute change in FEV1 after a bronchodilator might decrease slightly with baseline FEV₁, but no evidence exists for an asthmatic subset of patients with COPD defined by an unexpectedly large change in lung function. Figure 1 shows the frequency distribution of www.thelancet.com/respiratory Vol 1 September 2013

FEV₁ change after treatment with salbutamol in the patients with COPD in the ECLIPSE study.19 Similar results were also reported from the UPLIFT and LHS1 studies.33,35 ECLIPSE included comparator groups of slightly younger healthy smokers and non-smokers. As noted above, no significant difference was found in the mean FEV1 change between the smokers and GOLD grade 2 patients with COPD (160 mL vs 140 mL), but the healthy individuals in ECLIPSE and other trials showed a smaller change in mean lung function in keeping with the BOLD data.17 There are several possible explanations for this finding. Airway smooth muscle tone is cholinergically mediated60 and airway inflammation, which is increased in both smokers and in patients with COPD,61,62 might enhance this neural mechanism and reduce resting airway calibre. However, many other processes, including airway wall oedema and vascular congestion, can result from an airway inflammation. In the ECLIPSE study,19 the pre-test FEV1 in the smoking group was lower than in the healthy non-smoker individuals, which supports this idea. Alternatively, the presence of panacinar emphysema might influence airway responsiveness as suggested in recent pathological studies.63 The mechanism, whatever it might be, seems to become less important as COPD worsens, probably because of the onset of fibrotic changes in the peripheral airways.2

Methodological challenges with spirometry data The measurement of spirometry involves patient cooperation, is effort dependent, and has a known within-day reproducibility in COPD.64 The 200 mL threshold incorporated as part of the reversibility definition is designed to take into account the variation in spirometric measurement and to ensure that any change reported has not happened because of chance variation in the test itself.15,65 Nonetheless, some positive tests can arise by chance, especially when the baseline FEV1 value is low. This situation will contribute to the 569

Review

between day instability of the reversibility classification in the scientific literature. A more important factor identified in both the ISOLDE34 and ECLIPSE19,32 populations is spontaneous variation in the pre-test FEV1 between test days. When spirometry was measured repeatedly, an average FEV1 value could be identified for each patient. If the FEV1 was low relative to its average value, a positive reversibility test was more likely to be reported. Conversely, there is less room for improvement after treatment in patients with a high baseline FEV1, and this was associated with a decreased chance of being classed as reversible. These data suggest that the between day variation in responder status probably results from normal physiological changes in pre-test airway calibre on a background of a low FEV1 rather than clinically important changes in airway responsiveness. A

FEV1 change (L)

0·4

Post-bronchodilator FEV1 change by GOLD status p<0·001 p<0·001

0·3 p<0·001 0·2

0·1

0

B 0·8

Post-bronchodilator FVC change by GOLD status p=0·983 p=0·865

FVC change (L)

0·6 p=0·877 0·4

0·2

0

C

Post-bronchodilator FEV1/FVC change by GOLD status p<0·001 p<0·001

FEV1/FVC change (%)

5·0

p<0·001

0

–5·0 Gold II

Gold III

Gold IV

Figure 4: Changes in spirometry after salbutamol in patients with COPD grouped by GOLD grade Changes (SD) in FEV₁ (top), FVC (middle), and FEV1/FVC (bottom) after treatment with salbutamol by GOLD grade at the first visit reported in figure 3. Data from Albert and colleagues.19 FEV₁=forced expiratory volume in 1 s. FVC=forced vital capacity.

570

Clinical and research lessons for the future Several decades of intensive study of reversibility testing in COPD has highlighted some important truths about COPD and our general approach to interpretation of spirometry data in clinical settings. First, the apparent simplicity of classification of a patient as either reversible or not on the basis of one test has proven to be an illusion. Without an understanding of the between and within test variability in FEV1, this apparently straightforward test of lung function can easily be misinterpreted. The failure to identify a consistently responsive subgroup of patients with COPD, despite repeated attempts in large populations over several decades, shows that this approach is unlikely to be of benefit in routine clinical practice, whatever the threshold chosen for significant FEV1 change or even if other volume-based measurements are substituted. In fact, the real importance of these findings is that a clinical diagnosis of COPD supported by a measurement of lung function is a robust way to identify this disorder. In an era of protocolised medicine, it is important to stress that in all studies of reversibility, patients were selected for inclusion because their physicians believed them to have COPD on the basis of clinical examination and spirometry testing. Patients who have atypical features in their history, either in terms of age of onset, family history, or in the timing of associated symptoms, might exhibit the kind of dramatic responses to treatment that are so clinically memorable. Just because reversibility assessment adds little to routine clinical practice, does not mean that the assessment is uninformative in specific situations. However, in general, the between test variability in classification of reversibility status makes this reversibility test a poor way to classify patients or select their bronchodilator treatment. Likewise, the value of post-bronchodilator spirometry in determining the population-based prevalence of COPD is considerable when the prevalence of GOLD grade 1 disease (FEV1/FVC<0·7 with an FEV1 more than 80% predicted) is high because the number of patients identified as having COPD varied significantly between pre-bronchodilator and post-bronchodilator spirometry assessments14 (ie, the use of pre-bronchodilator data only will exclude many patients with apparent COPD). Most data have been obtained through the study of patients who already have a diagnosis of COPD. Reversibility testing might be more helpful in truly treatment-naive individuals, although it seems likely that the same issues of between-test variability in lung function will still apply. Alternative tests that show the effect of spontaneous variations in airway calibre in disorders of tidal breathing might prove more helpful. The measurement of respiratory system resistance with forced oscillation methods has already identified substantial spontaneous variability in patients with asthma66,67 and recent advances in the technique mean that it might become a useful test in a research setting. In this context, this method could be applied to patients with COPD to confirm or refute www.thelancet.com/respiratory Vol 1 September 2013

Review

the existence of patients who have more asthmatic disease features; neither reversibility testing nor provocation challenge to induce bronchoconstriction can conclusively solve this issue.68 However, the available data do not point to coexisting asthma as an explanation for the response to treatment in COPD. A further result of these studies of reversibility in COPD is the light they throw on COPD pathology. Changes in inflammatory cells, especially mast cells, have been reported in airway smooth muscle in patients with less advanced COPD and in patients with asthma.63,69 However, the response to anti-inflammatory treatment in these diseases is very different. Airway smooth muscle thickness is markedly increased in patients with asthma but much less so in patients with COPD;70 this difference provides a physiologically plausible explanation for the difference in bronchodilator response seen in these two disorders. The similarity in absolute FEV1 improvement in response to bronchodilator in patients with moderate COPD and in healthy smokers of similar gender composition (and hence lung size) is striking, as is the similarity in spontaneous variation of the pre-test FEV1.19 These data suggest that airway smooth muscle behaves normally in patients with COPD without the lability that characterises patients with asthma. Further, the main abnormality in COPD is persistent airway narrowing, whether in the peripheral airways or as a result of emphysema. If, as seems to be the case, acute variation in lung function is indicative of the abolition of preserved airway smooth muscle tone, then a ceiling probably exists for the effects of bronchodilator treatment in patients with COPD, however treatment is administered. Optimisation of bronchodilator treatment with one or more drugs of different classes is an area of great interest, but this approach has restricted potential. Likewise, because of the ceiling effect, the addition of anti-inflammatory treatment can improve lung function independent of and in addition to any improvement due to bronchodilators; however, reliance on changes in the FEV1 over a few months of treatment to establish the efficacy of anti-inflammatory treatment

Search strategy and selection criteria This Review combines the research of the authors over many years with a search of Medline and Embase for articles published in English from Jan 1, 2003, to Oct 30, 2012, using the search terms “bronchodilator reversibility”, “chronic obstructive pulmonary disease” OR “chronic bronchitis” OR “emphysema”, and “reversibility testing”. Relevant references published before the search period were also included and references from relevant articles were also searched. Review articles and book chapters are cited to provide readers with more details and more references than this Review can report.

www.thelancet.com/respiratory Vol 1 September 2013

might be misplaced. Future treatment needs to address the effect of loss of airways early in the natural history of COPD when lung function declines most rapidly71 rather than focusing on the few gains in function that can be achieved by manipulation of airway smooth muscle tone. A better understanding of bronchodilator responsiveness in COPD will also lead to an improvement in the ability to interpret the clinical trial data that form the evidence base for treatment frameworks. In view of the failure of reversibility testing to identify patients with a different clinical course, the exclusion of reversible patients from clinical trials in which clinical outcomes such as exacerbation rates, health status, or mortality are the main outcomes is not sensible. Experience with bronchodilator testing has shown that the absolute increase in FEV1 after a bronchodilator is greater when the baseline FEV1 is higher. Similar effects have been reported in drug trials, even when anti-inflammatory drugs are being tested.72,73 Aggregation of studies with small but significant differences in baseline function and degree of reversibility can be misleading and emphasises the need to look at endpoints other than just FEV1 in the assessment of treatment response.

Conclusion Bronchodilator reversibility in patients with COPD is a normally distributed continuous variable and the application of specific criteria defining significant and insignificant reversibility is arbitrary. Careful study has revealed that bronchodilator reversibility varies between tests, often as a function of baseline FEV1—ie, for an individual on a specific testing day, a decreased baseline FEV1 is associated with an increased chance of significant reversibility. This finding is independent of the chosen definition of significant reversibility and associations with clinical outcomes have not been consistently shown; therefore, the presence of a positive bronchodilator response does not identify a useful phenotype. Reversibility testing should not be used to determine which patients with COPD should be prescribed either short-acting or long-acting bronchodilators, and the presence or absence of significant bronchodilator reversibility should not be used to select patients with COPD for participation in clinical trials. Contributors The authors of this Review conceived, conducted, and wrote this paper and take responsibility for its content. This Review was done independently of any pharmaceutical input and without the aid of a medical writer. This Review was not directly funded nor was ethical approval sought for this work. PMAC conceived the Review and was responsible for its overall content. PA assisted with writing, editing, and created the figures. PW assisted with writing, editing, and created the table. Conflicts of interest PMAC has in the past received funding from several pharmaceutical companies for studies of patients with COPD. PA and PW have not received any funding. All authors declare that they have no conflicts of interest in relation to this Review.

571

Review

References 1 McDonough JE, Yuan R, Suzuki M, et al. Small-airway obstruction and emphysema in chronic obstructive pulmonary disease. N Engl J Med 2011; 365: 1567–75. 2 Hogg JC. Pathophysiology of airflow limitation in chronic obstructive pulmonary disease. Lancet 2004; 364: 709–21. 3 O’Donnell DE, Webb KA. Exertional breathlessness in patients with chronic airflow limitation. The role of lung hyperinflation. Am Rev Respir Dis 1993; 148: 1351–57. 4 Hardie JA, Buist AS, Vollmer WM, Ellingsen I, Bakke PS, Morkve O. Risk of over-diagnosis of COPD in asymptomatic elderly never-smokers. Eur Respir J 2002; 20: 1117–22. 5 Mannino DM, Watt G, Hole D, et al. The natural history of chronic obstructive pulmonary disease. Eur Respir J 2006; 27: 627–43. 6 Swanney MP, Ruppel G, Enright PL, et al. Using the lower limit of normal for the FEV1/FVC ratio reduces the misclassification of airway obstruction. Thorax 2008; 63: 1046–51. 7 Vestbo J, Hurd SS, Agusti AG, et al. Global strategy for the diagnosis, management, and prevention of chronic obstructive pulmonary disease: GOLD executive summary. Am J Respir Crit Care Med 2013; 187: 347–65. 8 Buist AS, McBurnie MA, Vollmer WM, et al. International variation in the prevalence of COPD (the BOLD Study): a population-based prevalence study. Lancet 2007; 370: 741–50. 9 Menezes AM, Perez-Padilla R, Jardim JR, et al. Chronic obstructive pulmonary disease in five Latin American cities (the PLATINO study): a prevalence study. Lancet 2005; 366: 1875–81. 10 Nisar M, Earis JE, Pearson MG, Calverley PM. Acute bronchodilator trials in chronic obstructive pulmonary disease. Am Rev Respir Dis 1992; 146: 555–59. 11 Kahneman DK. Thinking Fast and Slow. New York: Farrar, Strauss and Groux, 2011. 12 Eliasson O, Degraff AC Jr. The use of criteria for reversibility and obstruction to define patient groups for bronchodilator trials. Influence of clinical diagnosis, spirometric, and anthropometric variables. Am Rev Respir Dis 1985; 132: 858–64. 13 Hanania NA, Celli BR, Donohue JF, Martin UJ. Bronchodilator reversibility in COPD. Chest 2011; 140: 1055–63. 14 Perez-Padilla R, Hallal PC, Vazquez-Garcia JC, et al. Impact of bronchodilator use on the prevalence of COPD in population-based samples. COPD 2007; 4: 113–20. 15 Enright PL, Connett JE, Kanner RE, Johnson LR, Lee WW. Spirometry in the Lung Health Study: II. Determinants of shortterm intraindividual variability. Am J Respir Crit Care Med 1995; 151: 406–11. 16 Kainu A, Lindqvist A, Sarna S, Lundback B, Sovijarvi A. FEV₁ response to bronchodilation in an adult urban population. Chest 2008; 134: 387–93. 17 Tan WC, Vollmer WM, Lamprecht B, et al. Worldwide patterns of bronchodilator responsiveness: results from the Burden of Obstructive Lung Disease study. Thorax 2012; 67: 718–26. 18 Pellegrino R, Viegi G, Brusasco V, et al. Interpretative strategies for lung function tests. Eur Respir J 2005; 26: 948–68. 19 Albert P, Agusti A, Edwards L, et al. Bronchodilator responsiveness as a phenotypic characteristic of established chronic obstructive pulmonary disease. Thorax 2012; 67: 701–08. 20 [No authors listed]. Intermittent positive pressure breathing therapy of chronic obstructive pulmonary disease. A clinical trial. Ann Intern Med 1983; 99: 612–20. 21 Quanjer PH, Tammeling GJ, Cotes JE, et al. Lung volumes and forced ventilatory flows. Report Working Party Standardization of Lung Function Tests, European Community for Steel and Coal. Official Statement of the European Respiratory Society. Eur Respir J 1993; 16: 5–40. 22 Miller MR, Hankinson J, Brusasco V, et al. ATS/ERS Task Force. Standardisation of spirometry. Eur Respir J 2005; 26: 319–38. 23 Calverley P, Pauwels R, Vestbo J, et al. Combined salmeterol and fluticasone in the treatment of chronic obstructive pulmonary disease: a randomised controlled trial. Lancet 2003; 361: 449–56. 24 Calverley PM, Boonsawat W, Cseke Z, Zhong N, Peterson S, Olsson H. Maintenance therapy with budesonide and formoterol in chronic obstructive pulmonary disease. Eur Respir J 2003; 22: 912–19.

572

25

26

27

28

29

30

31 32

33

34

35

36

37

38

39

40

41

42

43

44

45

46

Mahler DA, Wire P, Horstman D, et al. Effectiveness of fluticasone propionate and salmeterol combination delivered via the Diskus device in the treatment of chronic obstructive pulmonary disease. Am J Respir Crit Care Med 2002; 166: 1084–91. Niewoehner DE, Rice K, Cote C, et al. Prevention of exacerbations of chronic obstructive pulmonary disease with tiotropium, a once-daily inhaled anticholinergic bronchodilator: a randomized trial. Ann Intern Med 2005; 143: 317–26. COMBIVENT Inhalation Aerosol Study Group. In chronic obtructive pulmonary disease, a combination of ipratropium and albuterol is more effective than either agent alone. An 85-day multicenter trial. Chest 1994; 105: 1411–19. Nisar M, Walshaw M, Earis JE, Pearson MG, Calverley PM. Assessment of reversibility of airway obstruction in patients with chronic obstructive airways disease. Thorax 1990; 45: 190–94. Burge PS, Calverley PM, Jones PW, Spencer S, Anderson JA. Prednisolone response in patients with chronic obstructive pulmonary disease: results from the ISOLDE study. Thorax 2003; 58: 654–58. Palmer LJ, Celedon JC, Chapman HA, Speizer FE, Weiss ST, Silverman EK. Genome-wide linkage analysis of bronchodilator responsiveness and post-bronchodilator spirometric phenotypes in chronic obstructive pulmonary disease. Hum Mol Genet 2003; 12: 1199–210. Anthonisen NR, Wright EC. Bronchodilator response in chronic obstructive pulmonary disease. Am Rev Respir Dis 1986; 133: 814–19. Vestbo J, Edwards LD, Scanlon PD, et al. Changes in forced expiratory volume in 1 second over time in COPD. N Engl J Med 2011; 365: 1184–92. Anthonisen NR, Lindgren PG, Tashkin DP, Kanner RE, Scanlon PD, Connett JE. Bronchodilator response in the lung health study over 11 yrs. Eur Respir J 2005; 26: 45–51. Calverley PM, Burge PS, Spencer S, Anderson JA, Jones PW. Bronchodilator reversibility testing in chronic obstructive pulmonary disease. Thorax 2003; 58: 659–64. Tashkin DP, Celli B, Decramer M, et al. Bronchodilator responsiveness in patients with COPD. Eur Respir J 2008; 31: 742–50. Han MK, Wise R, Mumford J, et al. Prevalence and clinical correlates of bronchoreversibility in severe emphysema. Eur Respir J 2010; 35: 1048–56. Hanania NA, Sharafkhaneh A, Celli B, et al. Acute bronchodilator responsiveness and health outcomes in COPD patients in the UPLIFT trial. Respir Res 2011; 12: 6. Montes de Oca M, Tálamo C, Perez-Padilla R, et al: PLATINO Team. Use of respiratory medication in five Latin American cities: the PLATINO study. Pulm Pharmacol Ther 2008; 21: 788–93. Hadcroft J, Calverley PM. Alternative methods for assessing bronchodilator reversibility in chronic obstructive pulmonary disease. Thorax 2001; 56: 713–20. Gross NJ, Petty TL, Friedman M, Skorodin MS, Silvers GW, Donohue JF. Dose response to ipratropium as a nebulized solution in patients with chronic obstructive pulmonary disease. A three-center study. Am Rev Respir Dis 1989; 139: 1188–91. The COMBIVENT Inhalation Solution Study Group. Routine nebulized ipratropium and albuterol together are better than either alone in COPD. Chest 1997; 112: 1514–21. Tashkin DP, Celli B, Senn S, et al. A 4-year trial of tiotropium in chronic obstructive pulmonary disease. N Engl J Med 2008; 359: 1543–54. O’Reilly J, Jones MM, Parnham J, Lovibond K, Rudolf M. Management of stable chronic obstructive pulmonary disease in primary and secondary care: summary of updated NICE guidance. BMJ 2010; published online June 25. DOI:10.1136/bmj.c3134. Brightling CE, Monteiro W, Ward R, et al. Sputum eosinophilia and short-term response to prednisolone in chronic obstructive pulmonary disease: a randomised controlled trial. Lancet 2000; 356: 1480–85. Hansen EF, Phanareth K, Laursen LC, Kok-Jensen A, Dirksen A. Reversible and irreversible airflow obstruction as predictor of overall mortality in asthma and chronic obstructive pulmonary disease. Am J Respir Crit Care Med 1999; 159: 1267–71. Albert P, Calverley P. A PEACE-ful solution to COPD exacerbations? Lancet 2008; 371: 1975–76.

www.thelancet.com/respiratory Vol 1 September 2013

Review

47

48

49

50

51

52

53

54

55

56

57

58

59 60

Koulouris NG, Valta P, Lavoie A, et al. A simple method to detect expiratory flow limitation during spontaneous breathing. Eur Respir J 1995; 8: 306–13. Calverley PM, Koulouris NG. Flow limitation and dynamic hyperinflation: key concepts in modern respiratory physiology. Eur Respir J 2005; 25: 186–99. Crim C, Celli B, Edwards LD, et al. Respiratory system impedance with impulse oscillometry in healthy and COPD subjects: ECLIPSE baseline results. Respir Med 2011; 105: 1069–78. Dellaca RL, Pompilio PP, Walker PP, Duffy N, Pedotti A, Calverley PM. Effect of bronchodilation on expiratory flow limitation and resting lung mechanics in COPD. Eur Respir J 2009; 33: 1329–37. Tantucci C, Duguet A, Similowski T, Zelter M, Derenne JP, Milic-Emili J. Effect of salbutamol on dynamic hyperinflation in chronic obstructive pulmonary disease patients. Eur Respir J 1998; 12: 799–804. Laveneziana P, Webb KA, Ora J, Wadell K, O’Donnell DE. Evolution of dyspnea during exercise in chronic obstructive pulmonary disease: impact of critical volume constraints. Am J Respir Crit Care Med 2011; 184: 1367–73. O’Donnell DE, Voduc N, Fitzpatrick M, Webb KA. Effect of salmeterol on the ventilatory response to exercise in chronic obstructive pulmonary disease. Eur Respir J 2004; 24: 86–94. Hay JG, Stone P, Carter J, et al. Bronchodilator reversibility, exercise performance and breathlessness in stable chronic obstructive pulmonary disease. Eur Respir J 1992; 5: 659–64. Calverley PM, Anderson JA, Celli B, et al. Salmeterol and fluticasone propionate and survival in chronic obstructive pulmonary disease. N Engl J Med 2007; 356: 775–89. O’Donnell DE, Lam M, Webb KA. Spirometric correlates of improvement in exercise performance after anticholinergic therapy in chronic obstructive pulmonary disease. Am J Respir Crit Care Med 1999; 160: 542–49. Newton MF, O’Donnell DE, Forkert L. Response of lung volumes to inhaled salbutamol in a large population of patients with severe hyperinflation. Chest 2002; 121: 1042–50. Walker PP, Calverley PM. The volumetric response to bronchodilators in stable chronic obstructive pulmonary disease. COPD 2008; 5: 147–52. Cerveri I, Dore R, Corsico A, et al. Assessment of emphysema in COPD: a functional and radiologic study. Chest 2004; 125: 1714–18. Gross NJ, Co E, Skorodin MS. Cholinergic bronchomotor tone in COPD. Estimates of its amount in comparison with that in normal subjects. Chest 1989; 96: 984–87.

www.thelancet.com/respiratory Vol 1 September 2013

61

62 63

64 65

66

67

68

69

70 71

72

73

Di Stefano A, Capelli A, Lusuardi M, et al. Severity of airflow limitation is associated with severity of airway inflammation in smokers. Am J Respir Crit Care Med 1998; 158: 1277–85. Cosio MG, Saetta M. Evasion of COPD in smokers: at what price? Eur Respir J 2012; 39: 1298–303. Ballarin A, Bazzan E, Zenteno RH, et al. Mast cell infiltration discriminates between histopathological phenotypes of chronic obstructive pulmonary disease. Am J Respir Crit Care Med 2012; 186: 233–39. Miller MR, Hankinson J, Brusasco V, et al. Standardisation of spirometry. Eur Respir J 2005; 26: 319–38. Tweeddale PM, Alexander F, McHardy GJ. Short term variability in FEV₁ and bronchodilator responsiveness in patients with obstructive ventilatory defects. Thorax 1987; 42: 487–90. Frey U, Brodbeck T, Majumdar A, et al. Risk of severe asthma episodes predicted from fluctuation analysis of airway function. Nature 2005; 438: 667–70. Frey U, Suki B. Complexity of chronic asthma and chronic obstructive pulmonary disease: implications for risk assessment, and disease progression and control. Lancet 2008; 372: 1088–99. Hansen EF, Vestbo J. Bronchodilator reversibility in COPD: the roguish but harmless little brother of airway hyperresponsiveness? Eur Respir J 2005; 26: 6–7. Brightling CE, Bradding P, Symon FA, Holgate ST, Wardlaw AJ, Pavord ID. Mast-cell infiltration of airway smooth muscle in asthma. N Engl J Med 2002; 346: 1699–705. Moreno RH, Hogg JC, Pare PD. Mechanics of airway narrowing. Am Rev Respir Dis 1986; 133: 1171–80. Drummond MB, Hansel NN, Connett JE, Scanlon PD, Tashkin DP, Wise RA. Spirometric predictors of lung function decline and mortality in early chronic obstructive pulmonary disease. Am J Respir Crit Care Med 2012; 185: 1301–06. Rabe KF, Bateman ED, O’Donnell D, Witte S, Bredenbroker D, Bethke TD. Roflumilast—an oral anti-inflammatory treatment for chronic obstructive pulmonary disease: a randomised controlled trial. Lancet 2005; 366: 563–71. Calverley PM, Rabe KF, Goehring UM, Kristiansen S, Fabbri LM, Martinez FJ. Roflumilast in symptomatic chronic obstructive pulmonary disease: two randomised clinical trials. Lancet 2009; 374: 685–94.

573