Published ahead of print on July 22, 2005, doi:10.1164/rccm.200502-218OC
© 2005 American Thoracic Society doi: 10.1164/rccm.200502-218OC Heterogeneity of Treatment Response to Azithromycin in Patients with Cystic FibrosisDepartment of Pediatrics, Columbia University, New York, New York; Departments of Pediatrics, University of Washington, Seattle; Statistical Analysis Unit, Cystic Fibrosis Therapeutics Development Network Coordinating Center, Seattle, Washington; and Cystic Fibrosis Foundation, Bethesda, Maryland Correspondence and requests for reprints should be addressed to Lisa Saiman, M.D., M.P.H., Columbia University College of Physicians and Surgeons, 622 West 168th Street, PH4W-470, New York, NY 10032. E-mail: LS5{at}columbia.edu
Rationale: We recently reported a randomized, placebo-controlled trial of azithromycin in patients with cystic fibrosis (CF) that demonstrated a 6.2% improvement in the 168-d relative change in FEV1 among azithromycin participants compared with placebo participants. Objectives: In the current analyses, heterogeneity of treatment response and the association between FEV1 and the risk of pulmonary exacerbations were investigated.
Methods: The time to first pulmonary exacerbation, hospitalization rates, and antibiotic use were compared between participants categorized by their relative change in FEV1 % predicted ( Measurements: All available data from the 185 randomized participants in the azithromycin trial were included in these analyses.
Main Results: Compared with placebo participants, a reduced risk of pulmonary exacerbations was observed both among azithromycin participants with Conclusions: Azithromycin participants experienced benefits in exacerbation parameters regardless of FEV1 response or subgroup. These data have implications for clinical practice and the design of clinical trials.
Key Words: antibiotic azithromycin cystic fibrosis macrolide Pseudomonas aeruginosa Over the past several years, there has been increased interest in the use of macrolide antibiotics in patients with cystic fibrosis (CF). Three recent placebo-controlled trials have shown clinical benefits from prolonged usage of azithromycin (13). Despite somewhat different patient populations and treatment regimens, azithromycin was associated with an improvement in lung function in all three trials; the mean relative improvement in FEV1 % predicted ranged from 3.6 to 6.2% after 3 to 6 mo of treatment (13). In addition, the trial conducted in the United States demonstrated that azithromycin reduced the frequency of pulmonary exacerbations. In the current analysis, we examined the heterogeneity of the treatment response to azithromycin among the 185 participants enrolled in the U.S. trial (3). Our first objective was to describe the heterogeneity of treatment response in terms of both pulmonary function and pulmonary exacerbations. The second objective was to investigate if a reduced risk of pulmonary exacerbations correlated with an improvement in pulmonary function. The third objective was to describe both the pulmonary function and pulmonary exacerbation responses to azithromycin among subgroups of participants defined by relevant clinical and physiologic parameters. The results of the current study are intended to provide guidance for monitoring the response of patients with CF to azithromycin treatment, and to inform the design of future clinical trials. Some of the results of this study have been previously reported in the form of a symposium proceeding (4).
Study Design The methods and results of the randomized, placebo-controlled, multicenter trial of azithromycin in patients with CF chronically infected with Pseudomonas aeruginosa have been recently published (3). Briefly, 185 participants were randomized with stratification by FEV1 % predicted (5) ( 60 vs. < 60%), weight ( 40 or < 40 kg), and study site. Participants who weighed 40 kg or more received 500 mg of azithromycin or placebo, and participants who weighed less than 40 kg received 250 mg of azithromycin or placebo. Drugs were administered on a Monday-Wednesday-Friday schedule for 24 wk (168 d). Participants continued their ongoing use of inhaled tobramycin solution, recombinant human dornase alfa (rhDNase) solution, and/or ibuprofen, but could not initiate these medications within 60 d of the screening visit for the trial. The pattern of use of inhaled tobramycin and rhDNase both before and during the study was further characterized by a structured questionnaire to the research coordinators at each site. The primary endpoint of the trial was the 168-d change in FEV1, and the secondary endpoints included time to first pulmonary exacerbation (defined as the use of antipseudomonal intravenous antibiotics or 7 d of oral quinolone antibiotics), hospitalization rate, and antibiotic usage.
Several additional endpoints were investigated in the current analyses. Heterogeneity of pulmonary function response was assessed by categorizing participants into different levels of response to treatment, defined by their relative change in FEV1 % predicted at Day 168. Heterogeneity of treatment response, as determined by secondary outcomes, such as pulmonary exacerbations experienced during the trial, as well as the rate of hospitalization and various antibiotic use parameters, were characterized descriptively. To determine if pulmonary function response correlated with heterogeneity of response to these secondary outcomes, the time to first pulmonary exacerbation, hospitalization rate, and antibiotic use were compared between groups of participants categorized by their relative change in FEV1 % predicted (
The pulmonary function and pulmonary exacerbation responses were also compared in several subgroups of participants, including those defined by the following: (1) long-term use of inhaled tobramycin solution (yes/no), defined as alternating cycles of 1 mo of use followed by 1 mo of nonuse for
Statistical Analysis
Descriptive statistics were used to summarize study endpoints with corresponding 95% confidence intervals (CI). Time to first exacerbation was modeled using Cox proportional hazards regression and graphically displayed using Kaplan-Meier estimates.
Heterogeneity of Treatment Response Pulmonary function. As previously published, at Day 168, the 87 participants randomized to azithromycin had a 0.094 L increase in FEV1 (95% CI, 0.020.17) and a relative improvement in FEV1 % predicted of 6.2% (95% CI, 2.69.8%) compared with the 98 participants randomized to placebo (3). Despite this average improvement in FEV1, heterogeneity was observed in pulmonary function response among the individual participants, as shown in Figure 1. In the azithromycin group, the 168-d relative change in FEV1 % predicted ranged from a decline of greater than 15% to an improvement of 15% or more. Despite this heterogeneity, 26% (95% CI, 1836%) of the azithromycin group experienced a 10% or greater improvement in FEV1 % predicted compared with only 12% (95% CI, 720%) of the placebo group. Furthermore, 13% of azithromycin participants had a 15% or greater improvement (95% CI, 822%) compared with 0% of placebo participants (95% CI, 04%). In contrast, 18% (95% CI, 1127%) of the placebo group had 10% or greater decline in the 168-d relative change in FEV1 % predicted as compared with 10% (95% CI, 518%) of the azithromycin group.
Pulmonary exacerbations. The azithromycin group had a significantly lower risk of pulmonary exacerbation compared with the placebo group (relative risk [RR] = 0.65; 95% CI, 0.440.95) (3). In addition, the azithromycin group experienced significant reductions in the proportion of participants hospitalized (relative reduction 0.46; 95% CI, 0.040.85), the average number of oral quinolone antibiotic courses (difference between treatment groups, 0.5 courses; 95% CI, 0.8 to 0.2), and the average number of nonquinolone oral antibiotic days (difference between treatment groups, 8.0 d; 95% CI, 15.2 to 0.82) (3) As with pulmonary function response, heterogeneity was observed among the study participants in terms of the number of pulmonary exacerbations experienced during the 168-d intervention period, as shown in Figure 2. However, the treatment response, as determined by pulmonary exacerbations, was less heterogeneous than the relative change in FEV1. Whereas 79% (95% CI, 7086%) of the azithromycin participants experienced 01 pulmonary exacerbation during the study, only 57% (95% CI, 4766%) of the placebo participants experienced 01 exacerbation.
Association between Pulmonary Function and Exacerbation Response As shown in Figure 3, both azithromycin participants who had a 168-d relative change in FEV1 % predicted of 5% and those with a relative change of < 5% experienced a lower risk of pulmonary exacerbations compared with the respective placebo participants. Among azithromycin participants with a 168-d response in FEV1 of 5%, the decreased risk of exacerbations (RR = 0.70; 95% CI, 0.381.31) was similar to that of the azithromycin participants with a 168-d response of < 5% (RR = 0.61; 95% CI, 0.371.01). Thus, on average, azithromycin participants who did not experience an improvement in pulmonary function still experienced clinical benefits, as measured by a reduction in pulmonary exacerbations. These results are supported by additional analyses displayed in Table 1, which shows the improvements in several secondary exacerbation parameters by the two FEV1 response groups. Regardless of pulmonary function response, azithromycin participants experienced a decrease in hospitalization and oral quinolone and oral nonquinolone antibiotic use.
Subgroup Analyses Subgroup analyses further demonstrated the discordance between the pulmonary function and pulmonary exacerbation responses to azithromycin. The 168-d relative change in FEV1 % predicted observed among various subgroups of participants is displayed in Table 2, and the RR of exacerbations among these subgroups is shown in Table 3.
The largest treatment response in terms of pulmonary function response was observed among participants who were not using long-term inhaled tobramycin at enrollment (Table 2). Of note is the observation that participants who were long-term users of aerosolized tobramycin had worse pulmonary function at baseline than nonusers, with average FEV1 % predicted values of 65 and 76%, respectively. The subgroup analyses based on long-term use of rhDNase showed that both long-term users and nonusers of rhDNase had similar improvement in pulmonary function. Likewise, the pulmonary function response was similar in both severity-of-disease subgroups. Finally, the subgroup that was homozygous for F508 showed a greater pulmonary function response to azithromycin than the subgroup that was heterozygous for F508. The subgroup analysis suggested that, in general, participants with more severe baseline lung disease experienced a greater reduction in exacerbations compared with participants with less severe baseline lung disease. Although tobramycin users were 1.74 times more likely (95% CI, 1.162.60) to have a pulmonary exacerbation compared with nonusers, which is consistent with the observation that users had more severe lung disease at baseline, azithromycin was associated with a slightly greater decrease in the risk of exacerbations among tobramycin users than among non-users (Table 3). Long-term rhDNase users had more severe pulmonary disease than nonusers at baseline (FEV1% predicted of 66.3 and 77.3%, respectively), and experienced a greater reduction in pulmonary exacerbations than nonusers (Table 3). Similarly, a greater reduction in the RR of exacerbation was observed among those with baseline FEV1 % predicted of less than 60% when compared with those with baseline FEV1 % predicted of 60% or greater (Table 3). However, regardless of genotype, participants in the azithromycin group had a decreased risk of exacerbations as compared with the placebo group. Notably, the baseline lung function of heterozygous versus homozygous participants was comparable: 68 versus 71%, respectively.
Three randomized, placebo-controlled trials have shown an improvement in pulmonary function in patients with CF treated with azithromycin (13). A recent Cochrane review of these studies concluded that azithromycin was associated with improvements in pulmonary function at 1- and 6-mo intervals, although questions remained about the efficacy and safety of long-term use of azithromycin (6). The azithromycin trial conducted in the United States showed heterogeneity of response in both pulmonary function and the secondary outcome measures of pulmonary exacerbations. To determine if improved lung function was predictive of improvements in pulmonary exacerbation parameters, we explored the time to exacerbation in participants with varying treatment response in terms of FEV1. Regardless of relative FEV1 response, participants in the azithromycin group had a decreased risk of pulmonary exacerbations, decreased hospitalization rate, and decreased oral antibiotic use. These observations suggested that other factors, such as stabilization in lung function, improved weight gain, or unmeasured factors, such as a reduction in inflammation (7) or an improvement in sputum rheology, may contribute to the reduced risk of pulmonary exacerbations (8). Thus, the clinical assessment of response to azithromycin should encompass more than pulmonary function and be of sufficient duration to monitor treatment parameters that relate to pulmonary exacerbations.
In efforts to further understand the heterogeneity of treatment effect, the impact of azithromycin was further explored among clinically relevant subgroups. These analyses demonstrated the impact of baseline lung function. In the two subgroup analysesusers versus nonusers of long-term inhaled tobramycin and users versus nonusers of long-term rhDNaseusers had worse baseline lung function and experienced a greater reduction in the risk of pulmonary exacerbations when azithromycin participants were compared with respective placebo participants. Similarly, participants with worse baseline severity of disease (i.e., participants with FEV1 < 60% predicted) experienced a reduced risk of exacerbations. Finally, those participants who were homozygous for It is critical to recognize that our initial study was not adequately powered to determine the statistical significance of these subgroup analyses, and that the analyses presented are descriptive. Nevertheless, our analyses did uncover some interesting trends and provided some potential explanations for the heterogeneity of the treatment effect that should be confirmed in future studies. These data have implications for the design and analyses of future clinical trials for patients with CF. Pulmonary function and exacerbation parameters should both be monitored to determine response to a therapy. Response to one parameter may not predict response to the other. Lung function may be an inadequate surrogate marker for other clinical benefits when designing clinical trials for patients with CF. Furthermore, careful consideration should be given to the potential impact of enrolling patients with CF in clinical trials when those patients are being treated with long-term inhaled tobramycin and rhDNase, as the use of these therapies appears to be associated with worse lung function and more frequent exacerbations. However, as evidenced, the results presented here demonstrate that, whereas patients with CF using these long-term therapies did not have the largest response to treatment in terms of improvement in lung function, they still derived improvements in other clinical outcomes related to pulmonary exacerbations. In conclusion, all subgroups experienced improvements in lung function when azithromycin participants were compared with respective placebo participants. Although worse baseline lung function was associated with the use of concomitant treatments, such as inhaled tobramycin and rhDNase, and an increased risk of pulmonary exacerbations, increased severity of disease at baseline was also associated with greater reductions in pulmonary exacerbations. Importantly, our analyses suggest that patients with CF without an improvement in lung function could still experience a lower risk of pulmonary exacerbations associated with azithromycin.
This study was supported by the Cystic Fibrosis Foundation. Members of the Macrolide Study Group are listed at the end of the article. Conflict of Interest Statement: None of the authors have a financial relationship with a commercial entity that has an interest in the subject of this manuscript.
* Members of the Macrolide Study Group: Cystic Fibrosis Foundation Therapeutics Development Network Coordinating Center; University of Washington, Seattle: Christopher Goss, M.D., M.P.H.; Bonnie Ramsey, M.D.; Jane L. Burns, M.D.; Richard A. Kronmal, Ph.D.; Judy Williams, M.P.H., R.N.; Children's Hospital and Regional Medical Center, Seattle: Morty Cohen, R.Ph.; Jenny Stapp, B.S.; University of Colorado, Denver: Frank J. Accurso, M.D.; Iris Osberg, M.T.; Axio Research Corporation, Seattle: Craig Johnson, M.S.; Brian Ingersoll, B.S.; Sandi Tennyson, B.S.; Baylor College of Medicine/Texas Children's Hospital, Houston: Peter Hiatt, M.D.; Charlene Hallmark, R.N.; Charles Sellers, B.B.A., R.R.T.; The Children's Hospital of Buffalo: Jack K. Sharp, M.D.; Mary Kontos, P.N.P.; Children's Hospital, Columbus: Karen McCoy, M.D.; Terri Johnson, R.N., B.S.N.; Childrens Hospital, Los Angeles: Marlyn S. Woo, M.D.; Antoniette Ziolkowski, R.N., B.S.N., M.S.N.; Daisy B. Bolduc, R.P.F.T., M.B.A.; Children's Hospital of Pittsburgh/University of Pittsburgh: Jonathan D. Finder, M.D.; Elizabeth Hartigan, R.N., M.P.H.; Children's Memorial Hospital/Northwestern University, Chicago: Steven Boas, M.D.; Manu Jain, M.D.; Eileen Potter, M.S., R.D.; Catherine Powers, C.C.R.C., R.D.; Columbia University, New York: Lynne Quittell, M.D.; Cook Children's Medical Center, Fort Worth: Maynard Dyson, M.D.; Janet Garbarz, R.N., C.P.N.P.; Emory University, Atlanta; Daniel Caplan, M.D.; Gerald Teague, M.D.; Jacqueline Geter, M.S., R.N., C.P.N.P.; Diane Morrison, R.N., B.S.N.; The Floating Hospital for Children/Tufts-New England Medical Center, Boston: William Yee, M.D.; Monica Ulles, R.N., M.S., P.N.P.; Oregon Health and Science University, Portland: Mark Chesnutt, M.D.; Lynn Oveson, R.N., M.N., A.N.P.; Svetlana Sheinkman, R.N., B.S.N.; Phoenix Children's Hospital: Peggy Radford, M.D.; Natalia Argel, B.S.N., R.N.C.; Annette Szpiszar, M.S., R.N.; Riley Hospital for Children, Indianapolis: Michelle Howenstine, M.D.; Mary Blagburn, R.N., B.S.N., C.C.R.C.; Delana Terrill, R.R.T., R.P.F.T., C.C.R.C.; Tulane University School of Medicine, New Orleans: Scott Davis, M.D.; Robert Beckerman, M.D.; Annette Broussard, R.R.T.; The University of Alabama at Birmingham/Children's Health System: Raymond Lyrene, M.D.; Valerie Eubanks, M.S., R.D.; University of Arizona Health Sciences Center, Tucson: Mark Brown, M.D.; Betty-Pauline Polanco, R.R.T.; Pat Cunningham, C.R.T.; University of Arkansas for Medical Sciences, Little Rock: Paula Anderson, M.D.; Larry P. Gann, B.S.; Cynthia Spinks, M.S.E.; April Stewart, B.S.; University of Kentucky Medical Center, Lexington: Michael Anstead, M.D.; Jamshed Kanga, M.D.; Lois Craigmyle, R.N.; University of Michigan Health Systems, Ann Arbor: Samya Nasr, M.D.; Julie Konkle, R.N.; Ermee Sakmar, R.N.; University of Nebraska Medical Center, Omaha: John Colombo, M.D.; Dee Acquazzino, B.S.; University of Utah, Salt Lake City: Barbara Chatfield, M.D.; Carmen Henshaw, R.N., B.S.N.; Deanna Allred, R.N., B.S.N.; University of Wisconsin/Children's Hospital, Madison: Michael Rock, M.D.; Tim Wolf, Pharm.D.; Vanderbilt University Medical Center, Nashville: Christopher Harris, M.D.; Stefanie F. Rushing, R.N., M.S.N.; Judy Marciel, R.N., M.S.N., C.P.N.P. Received in original form February 11, 2005; accepted in final form July 21, 2005
This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||