American Journal of Respiratory and Critical Care Medicine

Rationale: Frequent chronic obstructive pulmonary disease (COPD) exacerbations are a major cause of hospital admission and mortality and are associated with increased airway inflammation. Macrolides have airway antiinflammatory actions and may reduce the incidence of COPD exacerbations.

Objectives: To determine whether regular therapy with macrolides reduces exacerbation frequency.

Methods: We performed a randomized, double-blind, placebo-controlled study of erythromycin administered at 250 mg twice daily to patients with COPD over 12 months, with primary outcome variable being the number of moderate and/or severe exacerbations (treated with systemic steroids, treated with antibiotics, or hospitalized).

Measurements and Main Results: We randomized 109 outpatients: 69 (63%) males, 52 (48%) current smokers, mean (SD) age 67.2 (8.6) years, FEV1 1.32 (0.53) L, FEV1% predicted 50 (18)%. Thirty-eight (35%) of the patients had three or more exacerbations in the year before recruitment, with no differences between treatment groups. There were a total of 206 moderate to severe exacerbations: 125 occurred in the placebo arm. Ten in the placebo group and nine in the macrolide group withdrew. Generalized linear modeling showed that the rate ratio for exacerbations for the macrolide-treated patients compared with placebo-treated patients was 0.648 (95% confidence interval: 0.489, 0.859; P = 0.003) and that these patients had shorter duration exacerbations compared with placebo. There were no differences between the macrolide and placebo arms in terms of stable FEV1, sputum IL-6, IL-8, myeloperoxidase, bacterial flora, serum C-reactive protein, or serum IL-6 or in changes in these parameters from baseline to first exacerbation over the 1-year study period.

Conclusions: Macrolide therapy was associated with a significant reduction in exacerbations compared with placebo and may be useful in decreasing the excessive disease burden in this important patient population.

Clinical trial registered with www.clinicaltrials.gov (NCT 00147667)

Scientific Knowledge on the Subject

Frequent chronic obstructive pulmonary disease (COPD) exacerbations are a major cause of hospital admission and mortality and are associated with increased airway inflammation. Macrolides have airway antiinflammatory actions and may reduce the incidence of COPD exacerbations.

What This Study Adds to the Field

Macrolide therapy was associated with a significant reduction in COPD exacerbations compared with placebo and may be useful in decreasing the excessive disease burden in this patient population.

Patients with chronic obstructive pulmonary disease (COPD) are prone to frequent exacerbations that are a major cause of hospital admission, mortality, primary care visits, and impaired health status and have significant health economic consequences (13). Patients with frequent exacerbations have increased airway inflammation in the stable state (4) and a faster decline in lung function and thus COPD exacerbations affect disease progression, with the contribution of frequent exacerbations to FEV1 decline being on the order of 25% of total decline (5). However, antiinflammatory therapy in COPD with steroids has only a relatively small effect on airway inflammation; there is evidence from the TORCH (TOwards a Revolution in COPD Health) Study, but not from other, smaller studies (611), that steroids slow disease progression.

Macrolides have both antibacterial and antiinflammatory activity (1214). It has been suggested that the antiinflammatory activity is not related to the antibacterial action because the antiinflammatory effect is seen at low concentrations, below the minimal inhibitory concentration for airway bacteria (12). Human rhinovirus (HRV) is the commonest trigger of a COPD exacerbation (15); HRV infections account for the more severe COPD exacerbations (15) and macrolides may reduce airway cytokine production caused by HRV (16), thus reducing the susceptibility to exacerbation. Hence macrolides may reduce the frequency of COPD exacerbations, and thus disease progression as well. Macrolides are now used long term in conditions such as cystic fibrosis to reduce airway inflammation (17, 18).

We report in this article on a single-center, randomized, double-blind, placebo-controlled trial of a macrolide, erythromycin, in patients with moderate to severe COPD to test the hypothesis that regular therapy with macrolides reduces exacerbation frequency. Data from this study have been previously presented in abstract form at international meetings (19, 20).

We recruited 115 patients with COPD from outpatient chest clinics of the London Chest Hospital (London, UK) and the Royal Free Hospital (London, UK). Ethical permission for the study was obtained from St. Bartholomew's and the London Hospitals Trust as well as the Royal Free Hospital Trust Ethics committees.

Study Design

The study was a single-center, double-blind, randomized, placebo-controlled trial of erythromycin stearate (250 mg twice daily; Abbott Laboratories [Abbott Park, IL], supplied by DHP Clinical Trial Solutions [Powys, Wales, UK]). There was a 1-month run-in period followed by a 1-year treatment period. Coprimary outcome measures were exacerbation frequency and airway inflammation.

Subjects

Patients with COPD had influenza vaccination as recommended. All patients completed daily diary cards for changes in respiratory symptoms, and all patients were trained to report exacerbations to our research team, as soon as possible after onset of symptoms and before therapy was started as we have previously reported (15).

Inclusion and Exclusion Criteria

Patients were included in this study if they had moderate to severe COPD with FEV1 between 30% and 70% predicted, FEV1 reversibility of less than 15% and/or less than 200 ml to β2-agonists, were past or present cigarette smokers, and had no antibiotics or oral steroids during the run-in period. Patients were excluded if there was a history of asthma, bronchiectasis, neoplasia or other significant respiratory disease, unstable cardiac status (e.g., cardiac failure, prolonged QTc interval, cardiac arrhythmia), a history of macrolide allergy, or a history of hepatic impairment defined as abnormal liver function tests. Patients were excluded if they were taking drugs that could adversely interact with macrolides and for which therapeutic monitoring could not be undertaken. Most patients were taking inhaled steroids at recruitment. Over the study period no change in any therapy with antiinflammatory activity was allowed unless there was a clinical necessity, in which case the patient was then excluded from the study.

Recruitment and Run-in Period

The trial profile is shown in Figure 1. Patients were recruited if they fulfilled the inclusion criteria and were reviewed at the start of the run-in period when diary cards were provided as previously described (3). Patients were reviewed at the end of the run-in period, at the randomization visit. A run-in period was chosen so that patients would be free of exacerbation symptoms before baseline sampling. Because physiological changes associated with an exacerbation return to baseline in 70 to 74% of patients by about 4 weeks (19), a run-in period of 1 month was chosen. At recruitment the following were noted: baseline FEV1, number of exacerbations in the previous 12 months, and age. Baseline samples of sputum and blood were also taken.

Randomization Visit

Figure 1 shows that 6 of the patients recruited could not be randomized; thus 109 patients were randomized in all. At the randomization visit, spirometry was measured and the number of exacerbations in the previous 12 months was noted. Computer-generated randomization numbers were stored in sealed envelopes. Placebo and erythromycin (250 mg) were concealed in identical capsules. Medication was randomized before commencement of the study by the hospital pharmacy, independently of trial staff. Randomization was taken in blocks of 10 (5 placebo, 5 erythromycin) and patients were automatically dispensed the next allocated treatment (containing either erythromycin 250 mg twice daily or placebo twice daily for 3 mo and repeated at each clinic visit up to 1 yr). Unblinding occurred after data entry.

Follow-up of Patients, Stable Samples, and Withdrawal from the Study

Spirometry, sputum testing for bacteria, and spontaneous sputum and blood testing for inflammatory markers were performed at each follow-up visit. At these visits, diary cards were reviewed for exacerbations. Patients were also instructed to report to clinic within 48 hours of the onset of respiratory symptoms for detection of exacerbations as in our previous work (15). All initial measures were repeated at the final visit at 12 months. For purposes of analysis, a patent was said to have stable COPD if the patient was asymptomatic. A sample taken at this time was called a stable sample. All patients were followed up even if they left the study and even if they prematurely discontinued study treatments and returned to cohort follow-up. Deaths were monitored for the 1-year duration of the study for all patients.

We instructed patients and their primary care physicians that any exacerbations deemed to require antibiotics should receive penicillin or ciprofloxacin. If patients were admitted to hospital they were required to call a member of the study team, who would then contact the clinical team and convey the preceding instructions. If in spite of this the clinical team concluded that the use of a macrolide or quinolone other than ciprofloxacin was indicated, then the patient was withdrawn from the study. Patients were also withdrawn if they developed any evidence of macrolide toxicity.

Definition and Treatment of Exacerbations

A moderate exacerbation was defined as a sustained worsening of baseline respiratory symptoms for at least 2 days that required treatment with oral corticosteroids (prednisolone) and/or antibiotics, and a severe exacerbation was defined by the requirement for admission to hospital. Information about exacerbations was collected during clinic visits, and any patient experiencing worsening respiratory symptoms was instructed to contact the investigator immediately and to report to the study clinic as soon as possible as previously described (15, 21). Exacerbations were adjudicated in a blinded manner by one of the authors who was not involved in data collection during the course of the study.

The date of onset of an exacerbation was defined as the day of the first recorded symptom of that exacerbation and the end date of the exacerbation was defined as the day of disappearance of all symptoms associated with the exacerbation for at least two consecutive days. The duration of the exacerbation was defined as the number of days from onset to end. The procedure for management of an exacerbation when patients had deteriorating respiratory symptoms has been previously described (15). Exacerbation frequency was defined as the number of moderate to severe exacerbations over the 12-month follow-up period. Primary compliance data were collected from patient-completed diary card entries, which were checked against pill counts, and side effects were also recorded on diary cards.

Laboratory Analysis of Blood and Sputum Samples

Serum and sputum samples were obtained at each visit. Serum samples were analyzed for IL-6 by ELISA (R&D Systems, Abingdon, UK). Serum C-reactive protein (CRP) was measured with an Olympus luminometric analyzer (Olympus Life and Material Science Europa GmbH, Hamburg, Germany). Sputum samples were sent for identification and sensitivity testing according to our previously published methodology and according to British Society for Antimicrobial Chemotherapy guidelines (22, 23). Sputum levels of IL-6, IL-8, and myeloperoxidase (MPO; EMD Biosciences, San Diego, CA) were determined by ELISA as previously described (22).

Adverse Events Monitoring

On recruitment all patients had an ECG, liver and renal function, and complete blood count measurements and if patients were taking theophyllines then theophylline levels were measured. ECGs were repeated at screening and at 1 and 3 months to check the QTc interval. If the QTc interval was abnormal the patient would be withdrawn. ECGs were performed thereafter only if a patient's symptoms were suggestive of unstable cardiac disease. Liver function testing of patients was repeated on follow-up visits and theophylline levels of all patients taking theophyllines were checked. Patients also recorded on daily record cards symptoms that they believed were related to therapy, for example, upper gastrointestinal, lower gastrointestinal, rash, and so on, and these were reviewed at the clinic. If a patient had an adverse event that was thought to be drug related and that did not resolve, then the patient was withdrawn from the study. There was no routine screening of hearing as this was not thought to be a significant side effect at this dose of macrolide (2426).

Statistical Analysis

From our previous work on the London COPD cohort, we have found that the average exacerbation frequency is three exacerbations per patient per year (3). To reduce exacerbation frequency from 3 per year to 1.5 per year with 5% significance and 90% power, 58 patients had to be studied in each of the active and placebo groups. However, we have observed an average dropout rate of 15% per year in our cohort. Hence a target of 68 patients was required in each arm of the study and so we expected to recruit 136 patients with COPD in total. Statistical analysis was performed after unblinding.

Data for continuous variables was expressed as means (standard deviation [SD], standard error [SE], or 95% confidence interval [95% CI]) and discrete data were summarized as number (percent). Associations were considered statistically significant at the 5% level.

To determine the independent effects of the drug on the frequency of moderate to severe COPD exacerbations, multivariate analysis was performed with the number of moderate to severe exacerbations as the outcome variable, using a generalized linear model for a Poisson distribution with log of time on treatment as an offset variable and covariates of baseline smoking status, sex, disease severity (percentage of predicted FEV1 at baseline), number of exacerbations in the 12 months before screening (dichotomized as frequent or infrequent), and age. Time to first exacerbation between the two treatment arms was examined by Kaplan-Meier survival analysis, using all patients, and statistical significance was determined by log-rank test.

Exacerbation duration between the placebo and macrolide arms of the study was analyzed with a generalized linear model for Poisson distribution. For each exacerbation, treatment was binary-coded as with or without oral steroids. The model allowed for oral steroid treatment of each exacerbation, arm of the study, and baseline covariates of smoking status, number of exacerbations in the 12 months before screening (dichotomized as frequent or infrequent, i.e., baseline exacerbation rate), age, and disease severity (percentage of predicted FEV1 at baseline) with log of time on treatment as offset variable.

Serial stable FEV1, serum CRP, serum IL-6, sputum IL-6, sputum IL-8, and sputum MPO data over the year were compared by linear mixed models analysis with visit number as the repeated measure and the respective physiological or inflammatory marker as dependent variable. The comparison was performed between treatments with covariates of baseline exacerbation rate, baseline smoking status, disease severity, age, and sex. A similar model was used to examine differences between macrolide and placebo for changes between baseline parameters and first exacerbation.

All analyses used the intent-to-treat population, defined as all randomized patients who received at least one dose of study medication. The study was designed to show the superiority of either arm and used a two-sided test at the 5% level of significance. Analyses were performed with SPSS version 12 for Windows (SPSS, Chicago, IL), apart from the generalized linear modeling and survival analyses, which were performed with SPSS version 15.

Recruitment and Baseline Data

Although our sample size calculation required that we recruit a total of 136 patients, we recruited only 115 patients to the study before the preassigned date of study closure of March 2006. Figure 1 shows that there were an equivalent number of withdrawals in each arm of the study and in all 90 patients completed the study. Median compliance was 99%. The 109 patients had a mean age 67.2 (8.6) years, with 63% males. The mean FEV1 was 1.32 (0.53) L, FVC 2.67 (0.87) L, and FEV1% 50.0 (18.0)%. Thirty-eight patients had three or more exacerbations in the year before recruitment. Table 1 shows that there was no difference between the two treatment groups in any of these parameters.

TABLE 1. BASELINE VARIABLES FOR THE TWO TREATMENT GROUPS



Drug
Placebo
Macrolide
Continuous Variable
n
Mean
SD
n
Mean
SD
Age, yr5667.799.085366.548.10
FEV1, L561.360.55531.270.51
FEV1, % predicted5650.5518.875349.2517.30
FVC, L562.710.97532.630.79
FEV1/FVC, %5650.9013.525348.9212.84
Serum CRP, mg/L547.727.33527.976.22
Serum IL-6, pg/ml546.5616.62515.747.05
Sputum IL-6, pg/ml50188.34169.2838171.37165.91
Sputum IL-8, pg/ml502,806.001,348.57393,078.091,302.08
Sputum MPO, ng/ml479.9512.243913.1818.50
Discrete Variable
n
%

n
%

Smoker25452751
Male sex36643362
Three or more exacerbations in the year before recruitment19341936
Inhaled steroids44774177
LABA34613566
LAMA21381528
Theophylline
8
14

4
7.5

Definition of abbreviations: CRP = C-reactive protein; LABA = long-acting bronchodilator; LAMA = long-acting antimuscarinic; MPO = myeloperoxidase.

n = 109 patients with COPD (P > 0.05 for all comparisons between placebo and macrolide groups). Percentages are shown to two significant figures.

All patients were current smokers (n = 52) or ex-smokers and had a mean (SD) duration of 50.5 (36.1) and 52.8 (31.7) pack-years of smoking in the placebo and macrolide groups, respectively. Twelve patients were taking oral theophylline tablets (4 in the macrolide group). Table 2 shows that there was no difference in the number of patients taking inhaled steroids between the two arms of the study.

TABLE 2. MULTIVARIATE ANALYSIS BY INTENT TO TREAT WITH EXACERBATION FREQUENCY (OUTCOME VARIABLE) IN 109 PATIENTS WITH CHRONIC OBSTRUCTIVE PULMONARY DISEASE TREATED WITH MACROLIDE OR PLACEBO




95% CI

Parameter
Rate Ratio
Lower
Upper
P Value
Macrolide0.6480.4890.8590.003
Frequent exacerbations at baseline2.6792.0083.575<0.001
Current smoker1.0560.7891.4140.715
Male sex0.8840.6601.1840.407
Regression Coefficient
Age, yr−0.015−0.0320.0020.086
FEV1% at baseline
−0.008
−0.016
0.000
0.046

Definition of abbreviation: 95% CI = 95% confidence interval.

Rate ratios are shown for binary variables and regression coefficients (B) are shown for continuous variables with 95% confidence intervals. See text for details.

The number of patients attending follow-up visits at 1, 3, 6, 9, and 12 months, respectively, was as follows: 83, 85, 92, 85, and 89 (patient 90 attended for follow-up at more than 4 wk after the end of the study; diary cards were collected but the patient was not sampled).

Exacerbation Frequency

There were a total of 206 moderate to severe exacerbations, of which 125 occurred in the placebo arm and 81 in the macrolide group. This shift of the exacerbation frequency curve to the left (less moderate to severe exacerbations) in the macrolide group is shown in Figure 2. There were 14 (11.2%) hospitalizations for COPD exacerbation in the placebo group and 6 (7.4%) in the macrolide group. There was a median (interquartile range) exacerbation frequency of 2 (0.25, 3.75) in the placebo arm and 1.00 (0.00, 2.00) in the macrolide arm of the study (P = 0.006; Mann-Whitney test). Table 2 shows that the exacerbation frequency was significantly reduced in the macrolide arm of the study, with a rate ratio of a moderate/severe exacerbation being 0.648 compared with placebo, and that exacerbations occurring during the study were more frequent in patients with a history of frequent exacerbations at baseline or lower FEV1% predicted at recruitment. Sex and smoking status did not affect exacerbation frequency. Kaplan-Meier survival analysis showed that the median time to the first exacerbation in the macrolide arm was 271 days versus 89 days in the placebo arm (P = 0.020; log-rank test) (Figure 3).

Exacerbation Duration and Inflammation at Exacerbation

The duration of the exacerbation could be calculated for a total of 168 exacerbations (97 placebo, 71 macrolide). Median duration of exacerbations in the placebo arm was 13 (624) days, and in the macrolide arm of the study it was 9 (614) days (P = 0.036; Mann-Whitney test). Figure 4 illustrates this difference by showing that there was a greater proportion of short-duration exacerbations in patients taking macrolide compared with those taking placebo. For example, 63% of exacerbations had ended by Day 10 in the macrolide arm as opposed to 40% in the placebo arm. Poisson regression with duration of exacerbation as outcome variable revealed that patients in the macrolide arm were likely to have a shorter duration exacerbation (regression coefficient [B] = −0.286, P < 0.001). Other significant factors in the regression were a history of three or more exacerbations in the year before recruitment (B = −0.301, P = < 0.001) and oral steroid treatment at exacerbation (B = 0.177, P < 0.001).

There was no difference between drug and placebo arms in terms of changes at first exacerbation in FEV1, serum CRP, serum IL-6, or sputum inflammatory markers. Details are shown in the online supplement.

Stable Spirometry and Inflammatory Markers over 1 Year

Stable data for months 0, 1, 3, 6, 9, and 12 were analyzed for trends over time for FEV1 and serum and sputum inflammatory markers. Table 3 shows the data for 0 and 12 months only. Linear mixed models analysis for the sequential data over the year showed no significant change in any parameter with respect to time. The P values for this analysis are shown in Table 3.

TABLE 3. ANALYSIS OF STABLE SPIROMETRY AND INFLAMMATORY MARKERS OVER 1 YEAR


Group

Time

FEV1 (L)

Serum IL-6 (pg/ml)

Serum CRP (mg/L)

Sputum IL-6 (pg/ml)

Sputum IL-8 (pg/ml)

Sputum MPO (ng/ml)
PlaceboBaseline
n565454505047
Mean1.336.547.631952,88410.2
95% CI1.19–1.473.08–9.995.76–9.50147–2422,511–3,2575.61–14.8
12 mo
n454042313130
Mean1.258.319.012233,26518.9
95% CI1.08–1.423.87–12.765.55–12.46164–2822,763–3,76810.0–27.7
MacrolideBaseline
n535152383939
Mean1.255.667.751743,09513.3
95% CI1.11–1.392.11–9.215.89–9.64120–2282,678–3,5148.3–18.2
12 mo
n443841252525
Mean1.136.096.471283,13816.1
95% CI0.96–1.301.53–10.662.98–9.9663–1942,576–3,6996.3–25.8
P value (treatment × time)
0.966
0.884
0.812
0.514
0.551
0.201

Definition of abbreviations: CRP = C-reactive protein; MPO = myeloperoxidase; 95% CI = 95% confidence interval.

Shown are data only for the baseline visit (visit 0) and final (12 mo) visit. The estimated means are displayed together with the 95% confidence intervals. The P value for interaction between time and treatment over the 12-month period is shown for each variable. The table reveals that there is no difference in trend between the two arms of the study for the parameters shown. (Statistical modeling is by linear mixed models for visits 0, 1, 3, 6, 9, and 12; see text for details.) Note that the means shown here are means estimated from the multivariate model used for this analysis. Thus they differ slightly from the mean baselines shown in Table 1.

Bacteriology

Three hundred and thirty-nine sputum samples were taken for analysis for bacterial species from patients who spontaneously produced sputum. Of these, 73 were at exacerbation and the rest at baseline or follow-up. All Haemophilus influenzae were resistant or assumed constitutionally resistant to erythromycin. Detection of H. influenzae was positive in 27% stable samples and in 40% of exacerbation samples, and the corresponding distribution for Streptococcus pneumoniae was 7 and 10%, respectively. There was no difference in detection rate for any organism between the two arms of the study at any of the follow-up time points (P > 0.05 in all cases).

Sensitivity testing showed that at baseline 69 patients (placebo [P] = 39; macrolide [M] = 30) produced sputum samples, of which 33 showed no significant growth. Of those in which a bacterial pathogen was found there were H. influenzae (22, all resistant [P = 12, M = 10]), S. pneumoniae (6, all sensitive [P = 5, M = 1]), and Mycobacterium catarrhalis (3, all sensitive [P = 2, M = 1]). At the 12-month follow-up visit there were 43 sputum samples (P = 20, M = 23), with no significant growth in 26. Of those in which pathogens were detected there were H. influenzae (4 [P = 3, M = 1]), S. pneumoniae (3 [P = 2, all sensitive; M = 1, resistant]), and M. catarrhalis (3, two sensitive and sensitivity unknown in one [P = 2]).

Adverse Events Profile

Table 4 shows that there was no significant difference in side effects between patients in the two arms of the study and that the frequency of side effects was low in both arms of the study. In some cases more than one side effect occurred in the same patient, for example, the occurrence of an upper gastrointestinal symptom in a patient who also had tinnitus (classified as “other” in Table 4). Of the 12 patients taking theophyllines, no serum theophylline levels above the therapeutic range were detected during the study.

TABLE 4. NUMBER OF PATIENTS WITH ADVERSE EVENTS DURING THE 1-YEAR STUDY


Side Effect

Placebo

Macrolide
Upper GI55
Lower GI33
Rash23
Other
2
3

Definition of abbreviation: GI = gastrointestinal.

P > 0.05 in all case for comparison between placebo (n = 56) and macrolide (n = 53). Upper GI effects were nausea, vomiting or dyspepsia. Lower GI effects were diarrhea or cramps.

This is the first 12-month randomized controlled study of the effect of macrolide therapy in COPD. The results show a significant effect of low-dose macrolide therapy, reducing exacerbation frequency and severity in patients with moderate to severe COPD. The treatment was well tolerated over the 1-year study period. The study groups were similar with respect to age, FEV1%, sex, sputum inflammatory markers, and exacerbation frequency in the year before recruitment. Despite the fall in exacerbation frequency, we found no differences between the two arms of the study in FEV1 decline, airway or sputum inflammatory markers, or bacteria isolated over the 12-month study period. The duration of the exacerbations in the macrolide arm of the study was significantly shorter than in the placebo arm.

Oral macrolides have been shown to affect outcomes in two diseases that involve chronic airway obstruction: diffuse panbronchiolitis and cystic fibrosis (CF). In diffuse panbronchiolitis, for which erythromycin has been the most commonly used drug, there has been a highly significant improvement in survival (27) and symptoms (2830). Studies in CF have used mostly azithromycin and found improvement in morbidity, with the randomized controlled trials all showing a significant decrease in CF exacerbations in the macrolide arm in both children (18, 31, 32) and young adults (17). We and others have previously shown that respiratory viruses, mainly rhinoviruses, are associated with up to 50% of COPD exacerbations (15, 3335). In addition, erythromycin had been shown to decrease the inflammatory response to rhinovirus in vitro (16). Thus, for this reason, we chose to study erythromycin in the patients described here. We have found that erythromycin use was associated with a 35% fall in the rate ratio of moderate to severe exacerbations compared with the placebo arm patients. Our results are also similar to those of an open label study of 109 patients with COPD from Suzuki and colleagues, who found that erythromycin use was associated with a decrease in exacerbation frequency (36). The rate ratio for a moderate to severe exacerbation of COPD in the macrolide arm of our study was of similar magnitude to that of exacerbations in the azithromycin-treated arm in children with cystic fibrosis (32).

Previous studies have shown that inhaled steroid use is associated with about a 25% reduction in the mean exacerbation frequency in patients with moderate to severe COPD (6, 37). There was no statistically significant difference between the macrolide and placebo arms of our study in inhaled steroid use or dosage. Nevertheless, we found a significant reduction in exacerbation rate in the macrolide arm compared with the placebo arm. Our data would appear to suggest that the effect of the macrolide, erythromycin, occurred on top of any effect that inhaled steroid may have had on reducing exacerbation frequency. The mechanism by which steroids affect exacerbations may involve action at the nuclear level, but the precise mechanism by which macrolides affect exacerbation frequency is unknown. However, our study was not designed to test the hypothesis of an additive effect of macrolides on inhaled steroid therapy.

We found no difference in FEV1 between the two arms of the study over the 1-year period. Previous studies of lung function changes in CF have been equivocal, with macrolide use favoring small differences in three studies (17, 18, 31) but not in a fourth (32). The study by Wolter and colleagues found no evidence of a decline in FEV1 in the macrolide arm but a 3.6% decline in FEV1% in the placebo arm (17). Apart from the TORCH Study (38), most of the more recent long-term studies of COPD have not found significant differences between interventional and placebo arms in FEV1 decline (6, 8, 37, 38). In a 4-year study, we showed that an exacerbation frequency greater than about three per year is associated with a greater rate of FEV1 decline by about 25% (5), which has been supported by data from the Lung Health Study (39). Thus it can be hypothesized that if exacerbations decreased in the macrolide arm, there would be a corresponding decrease in the rate of FEV1 decline. However, our study was not powered to detect such a difference in FEV1 decline.

We have previously shown that lower airway bacterial colonization in patients with COPD is related to exacerbation frequency (22). There is legitimate concern regarding the development of resistance with any trial of long-term antibiotics. However, in this study the microbiological profile of sputum was not influenced by the use of erythromycin. Only one case of erythromycin resistance occurred in the study in the macrolide arm at 12 months, but we do not make any definitive inference from this result because of the small size of our study. Similar results have been reported in adults and children with CF (17, 18, 32), although macrolide-resistant S. aureus has also been found in these patients (40).

In vitro studies of the macrolide erythromycin have shown a reduction in the inflammatory response to human rhinovirus (16), and Jang and colleagues have shown that clarithromycin is associated with a decrease in rhinovirus titer in infected A549 pulmonary epithelial cells (41). There is also evidence that macrolides decrease neutrophil activity through lowered oxidant production (42), decreased bacterial adhesion (30), increased bactericidal activity (13, 43), and antichlamydial activity (44). There are few data on systemic inflammatory markers in macrolide studies of patients with CF, with only one study showing a significant fall in CRP in adults with CF when treated with macrolides (17). We found that inflammatory markers measured in sputum and blood remained stable throughout the 12-month period in both arms.

Sputum inflammatory markers may show significant variability and it is possible that by using induced sputum samples, this variability may have been reduced, although one study has shown consistency between levels of inflammatory markers in spontaneous and induced sputum (4). However, previous studies of inhaled corticosteroids on sputum markers have not shown effects on reducing inflammation (7), although studies of the effect of inhaled corticosteroid therapy on airway biopsies have shown some reduction in airway inflammatory cells (10, 11, 45). Thus adequately powered airway biopsy studies of the effects of long-term macrolides on airway inflammation in COPD are now required.

We found that exacerbations in the macrolide arm were shorter in duration and therefore perhaps less severe (21), compared with the placebo arm, even after allowance for oral steroid treatment at exacerbation. There was also no difference in inflammatory markers at exacerbation for the two treatment arms. This is in keeping with our previous findings that exacerbation clinical severity is not closely related to the degree of inflammatory change at exacerbation (4).

The drug was well tolerated, with similar adverse events in both groups, and these results were similar to those described in other long-term studies of CF (17, 32). All patients at recruitment had normal QTc intervals as determined by electrocardiography. There were no treatment-emergent changes in QTc interval or liver function tests. Median compliance was high during the study, at 99%, although compliance was assessed by counting tablets at every visit and may have been overestimated. Although no significant difference was found between the two arms of the study in terms of inflammatory markers and bacterial agents, it is important to note that our study was not powered to detect a difference in these parameters. A further limitation of our study is that it would have been useful to correlate changes observed over the year with quality of life data, which were not systematically measured during the study.

In conclusion, we have shown that the macrolide erythromycin, at a dosage of 250 mg twice daily, is associated with a significant reduction in moderate to severe exacerbations in patients with moderate to severe COPD. There was no corresponding effect on FEV1 or on airway or systemic inflammatory markers. There was a statistically significant decrease in exacerbation duration associated with macrolide use. Macrolides have a role in COPD and may be used to augment therapy in patients with moderate to severe COPD.

The authors thank Dr. Mark Wilks and Dr. Angela Whiley (St. Bartholomew's and the London NHS Trust), Dr. Tim McHugh (Microbiology Department, Royal Free and University College Medical School), and also the Biochemistry Department of the Royal Free Hospital (Hampstead, London, UK).

1. Strachan DP. Epidemiology: a British perspective. In: Calverley PMA, Pride NB, editors. Chronic obstructive pulmonary disease. London: Chapman & Hall; 1995. pp. 57–62.
2. Soler-Cataluna JJ, Martinez-Garcia MA, Roman Sanchez P, Salcedo E, Navarro M, Ochando R. Severe acute exacerbations and mortality in patients with chronic obstructive pulmonary disease. Thorax 2005;60:925–931.
3. Seemungal T, Donaldson GC, Paul EA, Bestall JC, Jeffries DJ, Wedzicha JA. Effect of exacerbation on quality of life in patients with chronic obstructive pulmonary disease. Am J Respir Crit Care Med 1998;157:1418–1422.
4. Bhowmik ST, Sapsford RJ, Wedzicha JA. Relation of sputum inflammatory markers to symptoms and physiological changes at COPD exacerbations. Thorax 2000;55:114–120.
5. Donaldson GC, Seemungal TAR, Bhowmik A, Wedzicha JA. The relationship between exacerbation frequency and lung function decline in chronic obstructive pulmonary disease. Thorax 2002;57:847–852.
6. Calverley PM, Anderson JA, Celli B, Ferguson GT, Jenkins C, Jones PW, Yates JC, Vestbo J; TORCH investigators. Salmeterol and fluticasone propionate and survival in chronic obstructive pulmonary disease. N Engl J Med 2007;356:775–789.
7. Keatings VM, Jatakanon A, Wordsell YM, Barnes PJ. Effects of inhaled and oral glucocorticoids on inflammatory indices in asthma and COPD. Am J Respir Crit Care Med 1997;155:542–548.
8. Burge PS, Calverley PMA, Jones PW, Spencer S, Anderson JA, Maslen TK. Randomised, double blind, placebo controlled study of fluticasone propionate in patients with moderate to severe COPD: the ISOLDE trial. BMJ 2000;320:1297–1303.
9. Lung Health Study Research Group. Effect of inhaled triamcinolone on the decline in pulmonary function in chronic obstructive pulmonary disease. N Engl J Med 2000;343:1902–1909.
10. Barnes NC, Qiu Y, Pavord ID, Parker D, Davis PA, Zhu J, Johnson M, Thomson NC, Jeffery PK; on behalf of the SCO30005 Study Group. Antiinflammatory effects of salmeterol/fluticasone propionate in chronic obstructive lung disease. Am J Respir Crit Care Med 2006;173:736.
11. Bourbeau J, Christodoulopoulos P, Maltais F, Yamauchi Y, Olivenstein R, Hamid Q. Chronic obstructive pulmonary disease: effect of salmeterol/fluticasone propionate on airway inflammation in COPD: a randomised controlled trial. Thorax 2007;62:938–943.
12. Peckham DG. Macrolide antibiotics and cystic fibrosis. Thorax 2002;57:189–190.
13. Ishizawa K, Suzuki T, Yamaya M, Jia YX, Kobayashi S, Ida S, Kubo H, Sekizawa K, Sasaki H. Erythromycin increases bactericidal activity of surface liquid in human airway epithelial cells. Am J Physiol Lung Cell Mol Physiol 2005;289:L565–L573.
14. Hoiby H. Diffuse panbronchiolitis and cystic fibrosis: East meets West. Thorax 1994;49:531–532.
15. Seemungal TAR, Harper-Owen R, Bhowmik A, Moric I, Sanderson G, Message S, MacCallum P, Meade TW, Jeffries DJ, Johnston SL, et al. Symptoms, inflammatory markers and respiratory viruses in acute exacerbations and stable COPD. Am J Respir Crit Care Med 2001;164:1618–1623.
16. Suzuki T, Yamaya M, Sekizawa K, Hosoda M, Yamada N, Ishizuka S, Yoshino A, Yasuda H, Takahashi H, Nishimura H, et al. Erythromycin inhibits rhinovirus infection in cultured human tracheal epithelial cells. Am J Respir Crit Care Med 2002;165:1113–1118.
17. Wolter J, Seeney S, Bell S, Bowler S, Masel P, McCormack J. Effect of long term treatment with azithromycin on disease parameters in cystic fibrosis: a randomized trial. Thorax 2002;57:212–216.
18. Equi A, Balfour-Lynn IM, Bush A, Rosenthal M. Long term azithromycin in children with cystic fibrosis: a randomised, placebo-controlled crossover trial. Lancet 2002;360:978–984.
19. Seemungal TAR, Wilkinson TMA, Perera W, Hurst JR, Wedzicha JA. Long term macrolide therapy decreases exacerbations in COPD [abstract]. Am J Respir Crit Care Med 2007;175(Suppl):A764.
20. Wilkinson TMA, Seemungal TAR, Sapsford R, Hurst JR, Perera W, Wedzicha JA. Effect of long term erythromycin in COPD trial (ELECT): exacerbations and inflammation [abstract]. Thorax 2007;62(Suppl 3):S115.
21. Seemungal TAR, Donaldson GC, Bhowmik A, Jeffries DJ, Wedzicha JA. Time course and recovery of exacerbations in patients with chronic obstructive pulmonary disease. Am J Respir Crit Care Med 2000;161:1608–1618.
22. Patel SI, Seemungal TAR, Wilks M, Lloyd-Owen S, Donaldson GC, Wedzicha JA. Relation of bacterial colonisation and the frequency, character and severity of COPD exacerbations. Thorax 2002;57:759–764.
23. Andrews JM; for the BSAC Working Party on Susceptibility Testing. BSAC standardized disc susceptibility testing method (version 4). J Antimicrob Chemother 2005;56:60–76.
24. Haydon RC, Thelin JW, Davis WE. Erythromycin ototoxicity: analysis and conclusions based on 22 case reports. Otolaryngol Head Neck Surg 1984;92:678–684.
25. Swanson DJ, Sung RJ, Fine MJ, Orloff JJ, Chu SY, Yu VL. Erythromycin ototoxicity: prospective assessment with serum concentrations and audiograms in a study of patients with pneumonia. Am J Med 1992;92:61–68.
26. Davis KK, Kao PN, Jacobs SS, Ruoss SJ. Aerosolized amikacin for treatment of pulmonary Mycobacterium avium infections: an observational case series. BMC Pulm Med 2007;7:2.
27. Kudoh S, Azuma A, Yamamoto M, Izumi T, Ando M. Improvement of survival in patients with diffuse panbronchiolitis treated with low-dose erythromycin. Am J Respir Crit Care Med 1998;157:1829–1832.
28. Takeda H, Miura H, Kawahira M, Takeda H, Miura H, Kawahira M, Kobayashi H, Otomo S, Nakaike S. Long-term administration study on TE-031 (A-56268) in the treatment of diffuse panbronchiolitis. Kansenshogaku Zasshi 1989;63:71–78.
29. Yamamoto M, Kondo A, Tamura M, Izumi T, Ina Y, Noda M. Long-term therapeutic effects of erythromycin and new quinolone antibacterial agents on diffuse panbronchiolitis. Nihon Kyobu Shikkan Gakkai Zasshi 1990;28:1305–1313.
30. Kobayashi H, Takeda H, Sakayori S, Kawakami Y, Otsuka Y, Tamura M, Konishi K, Tanimoto S, Fukakusa M, Shimada K. Study on azithromycin in treatment of diffuse pan-bronchiolitis. Kansenshogaku Zasshi 1995;69:711–712.
31. Saiman L, Marshall BC, Mayer-Hamblett N, Burns JL, Quittner AL, Cibene DA, Coquillette S, Fieberg AY, Accurso FJ, Campbell PW III; Macrolide Study Group. Azithromycin in patients with cystic fibrosis chronically infected with Pseudomonas aeruginosa: a randomized controlled trial. JAMA 2003;290:1749–1756.
32. Clement A, Tamalet A, Leroux E, Ravilly S, Fauroux B, Jais J-P. Long term effects of azithromycin in patients with cystic fibrosis: a double blind, placebo controlled trial. Thorax 2006;61:895–902.
33. Seemungal TAR, Harper-Owen R, Bhowmik A, Jeffries DJ, Wedzicha JA. Detection of rhinovirus in induced sputum at exacerbation of chronic obstructive pulmonary disease. Eur Respir J 2000;16:677–683.
34. Rohde G, Wiethege A, Borg I, Kauth M, Bauer TT, Gillissen A, Bufe A, Schultze-Werninghaus G. Respiratory viruses in exacerbations of chronic obstructive pulmonary disease requiring hospitalisation: a case–control study. Thorax 2003;58:37–42.
35. Papi A, Bellettato CM, Braccioni F, Romagnoli M, Casolari P, Caramori G, Fabbri LM, Johnston SL. Infections and airway inflammation in chronic obstructive pulmonary disease severe exacerbations. Am J Respir Crit Care Med 2006;173:1114–1121.
36. Suzuki T, Yanai M, Yamaya M, Satoh-Nakagawa T, Sekizawa K, Ishida S, Sasaki H. Erythromycin and common cold in COPD. Chest 2001;120:730–733.
37. Calverley P, Pauwels R, Vestbo J, Jones P, Pride N, Gulsvik A, Anderson J, Maden C; for the TRISTAN (TRial of Inhaled STeroids ANd long-acting β2 agonists) Study Group. Combined salmeterol and fluticasone in the treatment of chronic obstructive pulmonary disease: a randomised controlled trial. Lancet 2003;361:449–456.
38. Celli BR, Thomas NE, Anderson JA, Ferguson GT, Jenkins C, Jones PW, Vestbo J, Knobil K, Yates JC, Calverley PM. Effect of pharmacotherapy on rate of decline of lung function in COPD: results from the TORCH Study. Am J Respir Crit Care Med 2008;178:332–338.
39. Kanner RE, Anthonisen NR, Connett JE. Lower respiratory illnesses promote FEV1 decline in current smokers but not ex-smokers with mild chronic obstructive lung disease: results from the Lung Health Study. Am J Respir Crit Care Med 2001;164:358–364.
40. Tramper-Stranders GA, van der Ent CK, Gerritsen SA, Fleer A, Kimpen JL, Wolfs TF. Macrolide-resistant Staphylococcus aureus colonization in cystic fibrosis patients: is there transmission to household contacts? Pediatr Infect Dis J 2007;60:665–668.
41. Jang YJ, Kwon H-J, Lee BJ. Effects of clarithromycin on rhinovirus-16 infection in A549 cells. Eur Respir J 2006;27:12–19.
42. Labro MT, Abdelghaffer H. Immunomodulation by macrolide antibiotics. J Chemother 2001;13:3–8.
43. Hodge S, Hodge G, Jersmann H, Matthews G, Ahern J, Holmes M, Reynolds PN. Azithromycin improves macrophage phagocytic function and expression of mannose receptor in COPD Am J Respir Crit Care Med 2008;178:139–148.
44. Blasi F, Damato S, Cosentini R, Tarsia P, Raccanelli R, Centanni S, Allegra L; Chlamydia InterAction with COPD (CIAC) Study Group. Chlamydia pneumoniae and chronic bronchitis: association with severity and bacterial clearance following treatment. Thorax 2002;57:672–676.
45. Hattotuwa KL, Gizycki MJ, Tareq K, Ansari W, Jeffery PK, Barnes NC. The effects of inhaled fluticasone on airway inflammation in chronic obstructive pulmonary disease: a double-blind, placebo-controlled biopsy study. Am J Respir Crit Care Med 2002;165:1592–1596.
Correspondence and requests for reprints should be addressed to J. A. Wedzicha, M.D., Academic Unit of Respiratory Medicine, Royal Free & University College Medical School, Rowland Hill Street, Hampstead, London NW3 2PF, UK. E-mail:

Related

No related items
American Journal of Respiratory and Critical Care Medicine
178
11

Click to see any corrections or updates and to confirm this is the authentic version of record