© 2002 American Thoracic Society
Effectiveness of Fluticasone Propionate and Salmeterol Combination Delivered via the Diskus Device in the Treatment of Chronic Obstructive Pulmonary DiseaseDartmouth Hitchcock Medical Center, Lebanon, New Hampshire; and GlaxoSmithKline, Research Triangle Park, North Carolina Correspondence and requests for reprints should be addressed to to Donald A. Mahler, M.D., Dartmouth Hitchcock Medical Center, Section of Pulmonary and Critical Care Medicine, 1 Medical Center Drive, Lebanon, NH 03756–0001. E-mail: donald.a.mahler{at}hitchcock.org ABSTRACT This randomized controlled trial examined the benefits of combining an inhaled corticosteroid, fluticasone propionate (F), with an inhaled long-acting ß2-agonist, salmeterol (S), to treat the inflammatory and bronchoconstrictive components of chronic obstructive pulmonary disease (COPD). A total of 691 patients with COPD received the combination of F and S (FSC), S (50 mcg), F (500 mcg), or placebo twice daily via the Diskus device for 24 weeks. A significantly greater increase in predose FEV1 at the endpoint was observed after FSC (156 ml) compared with S (107 ml, p = 0.012) and placebo (-4 ml, p < 0.0001). A significantly greater increase in 2-hour postdose FEV1 at the endpoint was observed after treatment with FSC (261 ml) compared with F (138 ml, p < 0.001) and placebo (28 ml, p < 0.001). There were greater improvements in the Transition Dyspnea Index with FSC (2.1) compared with F (1.3, p = 0.033), S (0.9, p < 0.001), and placebo (0.4, p < 0.0001). The incidence of adverse effects (except for an increase in oral candidiasis with FSC and F) was similar among the treatment groups. We conclude that FSC improved lung function and reduced the severity of dyspnea compared with individual components and placebo.
Key Words: chronic obstructive pulmonary disease • ß2-agonist • inhaled corticosteroid • Transition Dyspnea Index Chronic obstructive pulmonary disease (COPD) is a disease state that is characterized by inflammation and airflow obstruction leading to chronic bronchitis and/or emphysema (1). Patients with COPD experience exertional breathlessness caused by mucous secretion, edema of the airway wall, a loss of attachments to the terminal airways, and bronchoconstriction. As only smoking cessation has been shown to modify the progressive decline in lung function in patients with COPD, pharmacologic therapy has focused on the treatment of airway obstruction to improve symptoms, primarily dyspnea, as well as health status (2). Maintenance treatment with bronchodilators, including ß2-agonists, anticholinergic agents, and theophylline, has long been the standard of care, as these agents treat the bronchoconstriction associated with COPD. In the United States, the long-acting ß2-agonists salmeterol (S) (Serevent Diskus; GlaxoSmithKline, Research Triangle Park, NC) and formoterol are currently approved for long-term, twice-daily maintenance treatment of bronchoconstriction associated with COPD. However, many patients remain symptomatic despite optimal use of bronchodilators. The presence of inflammation in the airways supports the rationale for instituting antiinflammatory therapy with an inhaled corticosteroid (3). Inhaled corticosteroid use has been shown to reduce inflammation associated with COPD (4–8). These studies have shown that inhaled corticosteroids, including fluticasone propionate (F) (Flovent Diskus; GlaxoSmithKline) (5, 8), have resulted in improvements in a variety of inflammatory indices in bronchoalveolar lavage (4, 7) and in sputum (5, 6, 8). Long-term studies of inhaled corticosteroids in patients with COPD have demonstrated improvements in FEV1 (9–11) and reductions in exacerbations (9, 11) and rate of decline in health status (11). To date, no medication has been shown to alter the rate of decline of FEV1 (9–12). These scientific and clinical data provided the foundation for the recent evidence-based Global Initiative for Chronic Obstructive Lung Disease Guidelines, sponsored by the National Heart, Lung, and Blood Institute and the World Health Organization (2). These guidelines recommend the use of inhaled corticosteroids in patients with moderate to severe COPD with significant symptoms who demonstrate a positive response to inhaled corticosteroid therapy or who experience repeated exacerbations (2). The combination of F and S (FSC) (Advair Diskus; GlaxoSmithKline) would be expected to treat both the inflammatory and bronchoconstriction aspects of COPD. Delivery of these medications together as an inhalation powder in the Diskus device, administered as one inhalation twice daily, may help simplify treatment regimens and may also improve patient adherence compared with currently available COPD treatments. The purpose of this study was to assess the effectiveness and safety of FSC and its individual components in the treatment of COPD. Specifically, our hypothesis was that combination therapy with FSC would increase airflow, reduce symptoms (including dyspnea), and improve health status over a 24-week period compared with the individual components and placebo. METHODS
Patients
Study Design The primary efficacy measures were change in predose FEV1 values for the comparisons of FSC to S and of F to placebo, and change in 2-hour postdose FEV1 values for the comparisons of FSC to F and of S to placebo. Other efficacy parameters included morning peak expiratory flow (PEF), supplemental albuterol use, dyspnea assessed by the Transition Dyspnea Index (TDI) (13), the Chronic Bronchitis Symptom Questionnaire (14, 15), and exacerbations defined by treatment. Health status was evaluated using the Chronic Respiratory Disease Questionnaire (CRDQ) (16). Safety was assessed using adverse events reporting, electrocardiograms, 24-hour Holter monitoring, vital signs, clinical laboratory evaluations, oropharyngeal examination, and short cosyntropin (Cortrosyn for injection; Organon, Inc., West Orange, NJ) stimulation testing (17). The protocol was approved by the institutional review boards for each site, and all patients provided written informed consent.
Statistical Methods Differences between treatments at the endpoint and all other time points in the change from baseline in predose and postdose FEV1, the Chronic Bronchitis Symptom Questionnaire, and the CRDQ were estimated and analyzed using contrasts from analysis of covariance adjusting for baseline and investigator. Estimation and analysis of differences in the Baseline Dyspnea Index/TDI were performed using contrasts with analysis of variance. Time to exacerbation was analyzed using Wald chi-square tests based on a Cox proportional hazards model with age and baseline FEV1 as covariates. Overall average and monthly average morning PEF, nighttime awakenings, and albuterol use were analyzed using the van Elteren modification of the Wilcoxon test (18). RESULTS
Study Population
Treatment compliance was high and comparable across the treatment groups. Approximately 90% of the patients had an individual adherence of 80% or more as measured by the dose counter on the Diskus device.
Lung Function
Two-hour postdose FEV1. A significantly greater increase in 2-hour postdose FEV1 at the endpoint was observed after treatment with FSC (261 ml) compared with F (138 ml, p < 0.001) and placebo (28 ml, p < 0.001) (Figure 2) . The increase in 2-hour postdose FEV1 at the endpoint after treatment with FSC was 24.2% over baseline. Significantly greater increases in 2-hour postdose FEV1 were observed at Day 1 and throughout the study during treatment with FSC compared with F. Significantly greater increases in 2-hour postdose FEV1 were observed for the S treatment group versus placebo (233 versus 28 ml, respectively, p  0.024) at the endpoint and at all assessment points throughout the study. Estimated differences between treatment groups at the endpoint are summarized in Table 2.
Response by reversibility. Patients who demonstrated reversibility of airway obstruction with albuterol had slightly better improvements in predose FEV1, 2-hour postdose FEV1, and TDI with individual components as well as with FSC. Although the magnitude of effect in nonreversible patients was somewhat lower, the effects were still clinically meaningful (Table 3) .
Morning PEF rate. Increases in morning PEF on Day 2, approximately 24 hours after the initiation of treatment, were greater for FSC compared with F, S, and placebo (p < 0.005) (Figure 3) . Greater increases in morning PEF were observed throughout the 24-week treatment period with FSC compared with F, S, and placebo. The mean overall change from baseline in morning PEF with FSC treatment (31.9 L/min) was greater than the sum of the mean changes from baseline observed with the individual components, 12.9 and 16.8 L/min for F (p < 0.001) and S (p < 0.001), respectively. Mean overall changes from baseline were also significantly greater for both F and S versus placebo (p < 0.001).
Dyspnea The mean Baseline Dyspnea Index scores for the placebo, FSC, F, and S groups were 5.8, 6.2, 6.0, and 5.9, respectively, indicating a moderate level of dyspnea for most patients at the beginning of the treatment period. At the endpoint, the mean TDI score for treatment with FSC (2.1) was greater than that after treatment with F (1.3, p = 0.033) and was significantly greater than that after treatment with S (0.9, p < 0.001) and placebo (0.4, p < 0.001) (Figure 4) . The estimated difference between FSC and S and between FSC and placebo exceeded the predefined criteria for clinical importance of 1 unit or more (13, 19). Significant differences in TDI scores were observed as early as Week 1 with FSC as compared with F, S, and placebo.
At the endpoint, TDI scores were significantly greater for F (1.3, p = 0.002), but not S, compared with placebo. Estimated differences between treatment groups at the endpoint are summarized in Table 2.
Supplemental Albuterol Use
A significant increase in the overall percentage of nights with no awakenings requiring albuterol was observed for treatment with FSC, F, and S compared with placebo (p < 0.001).
Health Status
Other Endpoints
Safety A total of 515 patients (75%) experienced at least one adverse event during the study (Table 5) . A greater percentage of patients in the F and the FSC groups experienced candidiasis (mouth/throat) based on visual inspection compared with the placebo and S groups, as expected with the use of an inhaled corticosteroid.
Three patients died during the study; all were in the placebo group. None of the deaths was considered by the investigator to be related to the study drug. The incidence of serious adverse events was low and similar across the treatment groups (4–7%). Morning plasma cortisol concentrations were obtained before administration of the morning dose of study medication at baseline and then again at end of study. At end of study, the prestimulated morning plasma cortisol concentrations were similar across the treatment groups (Table 6) . The number of subjects with an inappropriate cortisol response to stimulation with cosyntropin at Day 1 was zero in the placebo group, zero in the F group, two in the S group, and one in the FSC group. At the endpoint, the number of subjects with an abnormal response was zero in the placebo group, one in the F group, one in the S group, and one in the FSC group.
The incidence of clinically significant electrocardiogram abnormalities was comparable among the treatment groups (four patients in the placebo group, four patients in the F group, two patients in the S group, and three patients in the FSC group). No treatment-related effects on vital signs, QTc, or cardiac rate were observed. The incidence of ventricular ectopic events was similar in the placebo group compared with the active treatment groups at screening and at Week 4. Only five patients experienced significant changes in Holter monitor results (one patient each in the placebo, S, and FSC groups and two patients in the F group). DISCUSSION The major findings of this study were as follows: (1) FSC provided significantly greater improvements in predose and 2-hour postdose FEV1 as well as in morning PEF compared with individual agents or placebo. (2) FSC significantly reduced the severity of dyspnea as measured using the TDI compared with S and placebo. (3) There were significant reductions in albuterol use as a rescue medication with FSC compared with F and placebo. (4) FSC resulted in significant increases in health status, as measured using the CRDQ compared with F and placebo. (5) There were no differences in adverse events (except for oral candidiasis in the FSC and F groups) among treatment groups. The improvements in FEV1 were apparent with F, S, and FSC after the first week of treatment and continued during the subsequent 24 weeks of the study (Figures 1 and 2). The predose FEV1 at the endpoint with S was 107 ml (Figure 1), which is comparable to the results previously observed at 12 weeks with this long-acting ß2-agonist (20, 21). The higher predose FEV1 values observed with FSC (156 ml) reflect the additional contribution of F, which is presumably caused by an antiinflammatory effect. In a study of 80 patients with COPD, Cazzola and colleagues (22) investigated the addition of F (500 mcg twice daily) to S, and the results of combination therapy suggested a progressive improvement in lung function over a 3-month period compared with S alone. These results observed by Cazzola and associates (22) support our findings with combination therapy. The enhanced 2-hour postdose FEV1 with FSC (261 ml) illustrates the additional bronchodilation with S. For example, previous studies have demonstrated that S has its peak bronchodilator effect at 2 hours after the dose (20, 21, 23). Endpoint analysis for both predose and postdose FEV1 was performed to ensure that the patients prematurely withdrawing from the trial did not impact the robustness of the FEV1 results. The appropriateness of this analysis was supported by evaluating the data using alternative methods of handling dropouts, including multiple imputation, analysis of only completers, and recursive regression imputation. The significant improvement in morning PEF occurred within 24 hours after beginning treatment (Figure 3) and was higher than salmeterol treatment alone, suggesting an important early contribution from the inhaled corticosteroid. Thus, combination therapy with FSC provided rapid and consistent bronchodilation (predose and 2-hour postdose FEV1 and morning PEF) throughout the 6 months of the study compared with individual medications and placebo. Improvements in lung function were observed in patients who had reversible as well as nonreversible changes in FEV1 after acute administration of 400 mcg of albuterol (Table 3). Any increase in expiratory airflow may contribute to improvements in clinical outcomes. In this trial, we measured changes in dyspnea from baseline with the multidimensional TDI. The increase in the TDI score with FSC at the endpoint (2.1) was substantial and exceeded the responses with S (0.9), F (1.3), and placebo (0.4). Furthermore, the estimated differences between FSC and S (1.2) and placebo (1.7) exceeded the change of one unit considered to be the minimum clinically important difference (13, 19), whereas the estimated difference between FSC and F (0.7) approached this one-unit criterion. Greater reductions in the use of rescue albuterol with active treatments compared with placebo throughout the study further substantiate the clinical improvement in dyspnea. The improvements in TDI scores seen with FSC are considerably greater than those reported for any medication studied to date in a major clinical trial in patients with COPD (20, 21, 24–26). Although improvements in the TDI score were greater in the reversible patients, FSC also provided clinically important reductions in dyspnea in those patients considered to be nonreversible (Table 3). Although we did not investigate the mechanisms for the observed improvement in dyspnea with FSC in this study, it is reasonable to speculate on possible explanations. Previous studies with other bronchodilators, particularly albuterol (27) and ipratropium bromide (28), have shown that the improvements in breathlessness with these medications were due in part to the reduced dynamic hyperinflation that frequently develops in patients with COPD during exertion. The improved respiratory mechanics would diminish elastic recoil and increase the length of the vertical muscle fibers of the diaphragm. These changes would modulate inspiratory effort and reduce the intensity of breathlessness (28, 29). In another study, Ayers and colleagues (30) reported that S (two puffs, 42 mcg) and ipratropium (four puffs, 72 mcg) had similar effects on inspiratory capacity (used to measure dynamic hyperinflation) and ratings of dyspnea during submaximal cycle ergometry, although a placebo arm was not included for comparison with the bronchodilators. Therefore, it is likely that S, as a component of FSC, improved dynamic hyperinflation in our patients, which could have contributed to the observed reduction in dyspnea with combination therapy. Furthermore, Weber's law states that the perception of breathlessness must be increased by a constant fraction of its background value to produce a minimum or perceived change (31). Thus, a greater increase in airway resistance would be required after bronchodilation with FSC for the individual patient to notice a change in the severity of dyspnea. In the Lung Health Study II, inhaled triamcinolone reduced airway reactivity in response to methacholine in those with mild COPD (12). These patients, most of whom continued to smoke cigarettes, "had fewer respiratory symptoms during the course of the study." Any similar reduction in airway reactivity caused by inhaled F, a component of FSC, might also be expected to improve breathlessness. Thus, it is plausible that the combination of S (bronchial smooth muscle relaxation) and F (reduction in edema and inflammation within the wall of the airway) as part of combination therapy could work by different mechanisms to enhance airflow, reduce hyperinflation, and improve dyspnea as related to activities of daily living. Certainly, additional studies are needed to investigate these and other possible mechanisms whereby FSC reduces dyspnea. Health status generally slowly deteriorates in patients with COPD over time (11, 32). In the 6 months of this study, we observed that the score for the CRDQ actually increased with placebo treatment (5.0). The reasons for the increase in the placebo group could be due to patients receiving better medical attention when involved in a clinical trial and/or the increased use of albuterol in the placebo versus the active treatment groups. It is also likely that health status may have declined if the trial lasted for a longer time period. In this trial, we observed greater increases in the scores on the CRDQ with FSC (10.0) compared with S (8.0) and F (4.8) alone. Various investigators have reported that S (50 mcg) twice daily improved health status using either the CRDQ (20, 21, 26) or the St. George's Respiratory Questionnaire (33), although the differences compared with placebo were not always statistically significant. In the Inhaled Steroids in Obstructive Lung Disease trial, Burge and colleagues (11) found that F (500 mcg) twice daily slowed the decline in health status (measured on the St. George's Respiratory Questionnaire) compared with the decline with placebo over a 3-year time period. Our present data with combination therapy are quite consistent with these previous results for S and F used individually. These collective findings support the benefits of FSC for improving health status in patients with COPD. Exacerbation rates have been shown to be reduced in previous studies evaluating treatment with inhaled corticosteroids (9, 11). However, in this study, patients who experienced an exacerbation resulting in hospitalization or treated with an inhaled or oral corticosteroid during the study period were withdrawn to ensure that the concomitant use of these agents did not confound our ability to assess treatment effects on the primary efficacy measures of FEV1. This precluded examination of the effect of the treatment on rate of exacerbations. Other study design issues that limited the examination of a treatment effect on exacerbations include not requiring a history of exacerbations as an entry criteria, relatively short treatment duration, and inadequate sample size to discern significant treatment differences. Future studies are needed to confirm the positive findings seen with inhaled corticosteroids in previous trials (9, 11). The safety profile of FSC was consistent with that observed with the administration of both S plus F and was not different from that for the components alone. The overall adverse event profile of the combination therapy showed no new or unexpected adverse events. There was no evidence that treatment with FSC was associated with any increased risk of clinically relevant hypothalamic–pituitary–adrenal axis suppression, as assessed by cosyntropin stimulation testing, compared with treatment with F, S, or placebo. No unexpected cardiovascular effects, as assessed by Holter monitoring and routine electrocardiogram, were observed in those patients receiving FSC compared with patients receiving S or placebo. What are the clinical implications of the findings of this study? Based on the Global Initiative for Chronic Obstructive Lung Disease Guidelines, bronchodilator medications are indicated for the treatment of symptoms associated with COPD, whereas inhaled corticosteroids are indicated for airflow limitation that is responsive to a trial period of inhaled glucocorticoid treatment (2). When combined together, these agents treat both the bronchial smooth muscle constriction and the airway inflammation that are major components causing airflow limitation in this disease. Previous studies have shown both physiologic and/or clinical benefits of inhaled corticosteroids in patients with COPD (9, 11, 12, 22). Our results demonstrate that FSC not only improved airflow obstruction, but also provided clinical benefits as manifested by reduced severity of dyspnea, reduced use of rescue albuterol, and improved health status. We conclude that FSC can provide a useful treatment option for symptomatic patients with COPD. Acknowledgments The authors thank the following investigators who participated in this trial: Theodore Amgott, M.D.; Kenneth C. Anderson, M.D.; Janet Au, M.D.; Michael Baron, M.D.; Rebecca Bascom, M.D., M.P.H.; Robert Benkert, M.D.; Horst H. Blumberg, M.D.; Daniel Brune, M.D.; Raymond Casciari, M.D.; Mark Chesnutt, M.D.; Dennis Clifford, M.D.; Gene L. Colice, M.D., Luther Corley, M.D.; Leonard Cosmo, M.D.; Gilbert D'Alonzo, D.O.; James Donohue, M.D.; Donald Dvorin, M.D.; Peter Economou, M.D.; Travis Ellison, M.D.; Gregory Fino, M.D.; Harry Geisberg, M.D.; Glenn Giessel, M.D.; David Handshoe, M.D.; Stanley R. Horner, D.O.; Frank Horton, M.D.; James Hoyt, M.D.; William Jannetti, M.D.; Michael Kallay, M.D.; Gilbert Kuhn, M.D.; Michael Lawrence, M.D.; Bernard Levine, M.D.; Michael Littner, M.D.; Richard Lockey, M.D.; Robert Lodato, M.D., Ph.D.; Melvin Morganroth, M.D.; John Murray, M.D., Ph.D.; Robert Nathan M.D.; Gregory Neagos, M.D.; Harold S. Nelson, M.D.; Ronald Ovetsky, M.D.; William Pinkston, M.D.; Albert Razzetti, M.D.; Lawrence Repsher, M.D.; Anthony Rooklin, M.D.; Steven Sahn, M.D.; Gilbert Salazar, M.D.; Paul Scanlon, M.D.; Thomas Siler, M.D.; Irwin Spirn, M.D.; Randall Stoltz, M.D.; Mary Strek, M.D.; James Taylor, M.D.; Raymond Tidman, M.D.; John Votto, D.O.; Michael Weinstein, M.D.; Idelle Weisman, M.D.; Jan Westerman, M.D.; Martha White, M.D.; Warren Whitlock, M.D.; and John Winder, M.D. FOOTNOTES Received in original form December 18, 2001; accepted in final form July 23, 2002 REFERENCES
This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
0.001). SEMs ranged from 2.6 to 6.3 L/min.
