Antibiotic therapy for stable non-CF bronchiectasis in adults – A systematic review

Abstract

To provide an update on efficacy and safety of antibiotic treatments for stable non-cystic fibrosis (CF) bronchiectasis (BE). Systematic review based on the Preferred Reporting Items for Systematic reviews and Meta-Analyses guidelines was done. Twenty-six studies (1.898 patients) fulfilled the inclusion criteria. Studies of inhaled tobramycin have revealed conflicting results regarding quality of life (QoL), exacerbations and admissions, but may result in sputum cultures negative for Pseudomonas aeruginosa, whereas studies investigating the effect of inhaled gentamycin have shown positive effects on sputum bacterial density, decrease in sputum cultures positive for P. aeruginosa, QoL and exacerbation rate, but no improvement in forced expiratory volume in first second (FEV₁). Oral azithromycin can reduce exacerbations, together with minor improvements in QoL and FEV₁. Furthermore, oral erythromycin reduces exacerbations, but has no effect on lung function, symptoms or QoL. Inhaled ciprofloxacin may reduce P. aeruginosa in sputum cultures, but without changes in lung function, exacerbations or QoL. Although with limited evidence, inhaled colistin may have effects on P. aeruginosa density, exacerbations and QoL, whereas studies on aztreonam revealed no significant clinical improvements in the outcomes of interest, including exacerbation rate. Adverse events, including bronchospasm, have been reported in association with tobramycin and aztreonam. Several antibiotic treatment regimens have been shown to improve QoL and exacerbation rate, whereas findings regarding sputum production, lung function and admissions have been conflicting. Evidence-based treatment algorithms for antibiotic treatment of stable non-CF BE will have to await large-scale, long-term controlled studies.

Introduction

Bronchiectasis (BE) is a chronic airway disease characterized by abnormal destruction and dilation of bronchi and bronchioles¹ leading to dysfunctional mucociliary clearance and by that retention of secretions, repeated bacterial infections, chronic inflammation and progressive tissue destruction in a vicious circle.^1–3

Non-cystic fibrosis (non-CF) BE is associated with a large variety of both local pulmonary diseases, including infections and obstructive lung diseases, and systemic diseases including immune defects and autoimmune diseases. However, in about 50% of cases, the underlying cause remains unknown.³ The gold standard diagnostic test is high-resolution computed tomography (HRCT), but the patients should, however, also have relevant clinical symptoms, which include productive cough, dyspnoea and haemoptysis.³

The overall aim of the treatment of patients with non-CF BE is to reduce symptoms, to maintain lung function and to prevent exacerbations and thereby improving the quality of life (QoL) and long-term outcome.³ However, the management strategy to reach these goals is not clearly defined.

The present review provides an update on the current available evidence on the efficacy and safety of inhaled and oral antibiotic therapy in the management of stable patients with non-CF BE.

Methods

The general principles of the Preferred Reporting Items for Systematic reviews and Meta-Analyses guidelines^4,5 were adopted to perform this review. A series of systematic searches were carried out, last updated October 2015, using the database PubMed, EMBASE, Cochrane Controlled Trials Register and Clinical Trials.gov, and were based on the following algorithm of MeSH terms: Non-CF BE and BE were searched alone and in combination with antibiotics, tobramycin, ciprofloxacin, erythromycin, gentamycin, azithromycin and aztreonam. The search was limited to English language articles, and clinical trials published solely in abstract form were excluded because the methods and results could not be fully assessed.

To be included, studies had to meet all of the following criteria: (1) published in peer-reviewed journal; (2) inclusion of adults aged >18 years; (3) BE diagnosed by HRCT in non-CF patients; (4) report at least one of the following outcomes QoL, dyspnoea, number of exacerbations, time to next exacerbation, forced vital capacity (FVC), forced expiratory volume in first second (FEV₁), sputum production, sputum bacteriology, C-reactive protein (CRP) and white blood cell count and (5) published after 1990; and not the following exclusion criteria: (1) only physiotherapy and (2) treatment of patients with acute exacerbations. Potential relevant papers to be included in the present review were assessed in detail by at least two of the authors.

A meta-analysis was not included in the present review, primarily due to the limited number of published clinical trials fulfilling the inclusion criteria within each therapy category, as well the heterogeneity of the included studies.

Results

A total of 631 potential relevant papers were identified at the initial step, of which 582 papers were excluded as they did not meet the inclusion criteria. Of the remaining 94 papers, 69 were excluded, for example, studies only investigating patient with acute exacerbation, according to the predefined inclusion and exclusion criteria. Finally, 26 trials published in 25 papers (comprising a total of 1.898 subjects) were included in the present review. Participants in the included trials were stable, but symptomatic, and fulfilled the CT criteria for a diagnosis of non-CF BE.

Aminoglycosides

Tobramycin alone or in combination with ceftazidime

Five studies, in total including 224 individuals, have examined the effect of tobramycin on non-CF BE (Table 1).

Table 1.

Overview of studies investigating the effect of aminoglycosides for the treatment of patients with stable non-cystic fibrosis bronchiectasis.

Drug	Study/author	Design	No. of subjects	Therapy	Duration of treatment	Primary outcome	Sputum weight	FEV1	FVC	QoL	Exacerbations	Admissions	Antibiotic use	Bacterial colonies	Adverse events
Ceftazidim and tobramycin	Orriols et al. (1999)6	Randomized, case–control, unblinded	15	Inhalation (nebulized) 100mg/100 mg b.i.d.	12 months	Number and days of admissions		No change	No change			Reduction	No change
Tobramycin	Barker et al. (2000)7	Randomized, placebo- controlled, double-blind	74	Inhalation (nebulized) 300 mg b.i.d.	4 weeks	Sputum Pseudomonas aeruginosa density		No change	No change	Some improvement				Improved	No change
	Scheinberg et al. (2005)8	Open-label, controlled follow-up	31	Inhalation (nebulized) 300 mg b.i.d.	2 weeks (3 cycles)	Clinical response				Improved	No change			Improved	Some
	Drobnic et al. (2005)9	Randomized, placebo-controlled, double-blind, crossover	30	Inhalation (nebulized) 300 mg b.i.d.	6 months	Number of exacerbations and days of admission		No change	No change	No change	No change	Improved	No change	Some improvement	Some
	Couch et al. (2009)10	Randomized, placebo-controlled, single-blind	74	Inhalation (nebulized) 300 mg b.i.d.	4 weeks	Sputum P. aeruginosa density		No change	No change	No change				Improved	Some
Gentamycin	Lin et al. (1997)11	Randomized, placebo-controlled, double-blind	28	Inhalation (aerosol) 40 mg b.i.d.	3 days	Sputum production				Improved
	Murray et al. (2011)12	Randomized, placebo-controlled, single-blind	57	Inhalation (nebulized) 80 mg b.i.d.	12 months	Sputum bacteriology	No change	No change	No change	Improved	Improved			No change
	Chalmers et al. (2012)13	Randomized, placebo-controlled, single-blind	57	Inhalation (nebulized) 80 mg b.i.d.	12 months	Association between bacterial load and airway inflammation								Some improvement

FEV₁: forced expiratory volume in first second; FVC: forced vital capacity; QoL: quality of life.

In 1999, Orriols et al.⁶ published a prospective, non-blinded pilot study, where patients were randomized to either nebulized inhalations of tobramycin 100 mg and ceftazidime 1000 mg twice daily for 12 months (n = 7) or usual symptom-driven treatment (n = 8). A reduction in number of hospital admissions (0.6 vs. 2.5; p = 0.02) and days of admission (13 vs. 58; p = 0.03) was found for patients in the tobramycin–ceftazidime group compared to the controls. No significant differences were found in lung function and oral use of antibiotics. The authors concluded that patients could benefit from long-term inhaled antibiotic therapy but did not report analyses addressing changes in antimicrobial resistance.

In 2000, Barker et al.⁷ examined the ability of tobramycin to eradicate or reduce the density of Pseudomonas aeruginosa in individuals with BE. In this controlled, double-blind, phase II study, 74 patients were randomized to either 300 mg nebulized tobramycin (n = 37) or placebo (n = 37) twice daily for 28 days. At the 6-week follow-up, sputum culture was negative for P. aeruginosa in 13 of the patients receiving tobramycin but in none of the controls. Of the 13 patients, 12 were classified by the investigators as having improved medical condition, and the odds ratio for improved medical condition was 2.7 (95% confidence interval (CI) 1.1–6.9) for tobramycin-treated patients compared to controls. No significant differences were found in lung function or total reports of adverse events but tobramycin-treated patients reported more dyspnoea, chest pain and wheezing (p = 0.01 for all three comparisons). The authors concluded that tobramycin markedly reduces sputum P. aeruginosa density, but further studies of clinical efficacy and safety are warranted. This relatively large study appears, however, limited by not using a validated measure of changes in the patients’ condition.

In 2005, Scheinberg et al.⁸ published an uncontrolled, follow-up trial on the effect of tobramycin in patients with non-CF BE and a history of colonization with P. aeruginosa. Thirty-one of the 41 patients enrolled completed three cycles of 2 weeks treatment with 300 mg nebulized tobramycin twice daily, followed by no treatment for 2 weeks, resulting in a total treatment period of 12 weeks. At week 10, an improvement was observed in pulmonary total symptom severity score (mean ± standard deviation (SD) 1.5 ± 3.4; p = 0.006) and St George’s Respiratory Questionnaire (SGRQ) score (mean change ± SD – 9.8 ± 13.9; p < 0.001) compared with baseline. Six patients were considered eradicated for P. aeruginosa at week 12 or at withdrawal. Patients with sputum cultures negative for P. aeruginosa had less improvement in symptom severity score than patients with positive sputum cultures (−2.2 ± 3.9 vs. − 0.8 ± 3.2) and SGRQ (−3.3 ± 5.5 vs. −9.9 ± 14.0). Adverse event regardless of causality was reported in 95.1% of the patients, most commonly cough (43.9%), dyspnoea (34.1%) and increased sputum (29.3%). They concluded that patients with BE most likely will benefit from treatment with tobramycin, if tolerating side effects. However, the value of the study is limited by the open-label, non-placebo-controlled design.

A double-blind, placebo-controlled crossover trial in patients with BE and chronic colonization with P. aeruginosa was published by Drobnic et al.⁹ Thirty patients were randomized to receive 300 mg nebulized tobramycin or placebo twice daily and vice versa for 6 months. No difference was found in exacerbation rate, but treatment with tobramycin significantly reduced number of hospital admissions (mean ± SD 0.2 ± 0.4 vs. 0.8 ± 1.2; p = 0.04) and days of admission (mean ± SD 2.1 ± 5.0 vs. 12.7 ± 21.8; p < 0.05) compared with placebo. Tobramycin also decreased the density of P. aeruginosa, (p = 0.038 in the first-period parallel analysis), whereas no difference was observed in QoL, lung function or use of other oral or parental antibiotics. Bronchospasm induced by inhalation of tobramycin was reported in three patients, and all withdrew from the study. Limitations of the study include the relatively low number of patients included and duration of each treatment cycle.

Couch et al.¹⁰ further investigated the effect and safety of aerosolized tobramycin in a placebo-controlled, single-blind, multicentre study in BE patients colonized with P. aeruginosa published in 2001. A total of 74 patients were randomized to either 300 mg aerosolized tobramycin solution (n = 37) or placebo (n = 37) twice daily for 4 weeks. Sputum density of P. aeruginosa was reduced significantly in patients treated with tobramycin (mean change in log₁₀ colony-forming units (CFU)/g sputum −4.8, p < 0.001 and −4.5, p < 0.001 after 2 and 4 weeks of treatment, respectively) and 2 weeks after ended treatment 17 patients had sputum negative for P. aeruginosa, whereas no such changes were found in the placebo group; and lung function was unchanged in both groups. Isolates of P. aeruginosa resistant to tobramycin developed in 3 of 36 patients on tobramycin and 1 of 34 patients on placebo. Adverse events were reported more often in the intervention group, although overall frequency of adverse events was not reported. Two weeks after end of study treatment, the general health was rated, by the health care professional, as improved, worse or unchanged in 62%, 22% and 16%, respectively, in the intervention group and 38%, 13% and 49%, respectively, in the placebo group. Patients treated with tobramycin in whom P. aeruginosa was considered eradicated were more likely to have improved health status (12 of 13 patients) compared to patients with positive sputum cultures (11 of 24 patients). The authors conclude that inhaled tobramycin in patients with non-CF BE colonized with P. aeruginosa reduce sputum density of P. aeruginosa and improve general health status. However, the short follow-up period is a limitation, also with regard to incidence of re-colonization, and, furthermore, the term general health status is not further specified.

In conclusion, inhaled tobramycin can be used to achieve sputum cultures negative for P. aeruginosa in non-CF BE patients. However, the available studies reveal non-consistency with regard to the effect on symptoms or QoL. Only one study, unblinded and small, has shown an effect of exacerbations requiring hospital admission and duration of admission, and none have shown an effect on lung function. Furthermore, adverse effects, primarily therapy-induced bronchospasm, appear to be a potential limitation for use of inhaled tobramycin in this group of patients.

Gentamycin

Three studies, including 142 subjects in total, have examined the effect of inhaled gentamycin in non-CF BE (Table 1).

In 1997, Lin et al.¹¹ published a small, double-blind, placebo-controlled study on the effect of inhaled gentamicin, where 28 patients were randomized to receive aerosolized gentamycin 40 mg or 0.45% saline twice daily for 3 days. The group receiving active treatment (n = 13) had a significant decrease in daily sputum production (95 mL to 58 mL, p < 0.01) and an increase in peak expiratory flow rate (PEF) (184–216 L/min), 6-min walking distance (325–408 m, p < 0.05) and subjective improvements in dyspnoea measured by the Borg scale. No improvements were seen in the placebo group. The study is limited by a very short duration of treatment of only 3 days; and it also seems more appropriate to have used FEV1 and FVC as a measure of lung function instead of PEF.

In 2011, Murray et al.¹² published a randomized, controlled, single-blind study assessing the efficacy of nebulized gentamycin therapy over 1 year with follow-up every 3-month and after a treatment free 3-month period. Fifty-seven patients were randomized to receive nebulized 80 mg gentamycin (n = 27) or 0.9% saline (n = 30) twice daily for 12 months. At the end of treatment, the gentamycin group had significantly lower bacterial density in sputum compared with the saline group (3.0, CI 1.0–5.9 log₁₀ CFU/mL vs. 7.7, CI 7.3–8.2 log₁₀ CFU/mL; p < 0.0001). However, there was no difference between the groups at the end of the 3-month treatment free follow-up period. Sputum cultures negative for P. aeruginosa was achieved in 4 of 13 patients in the gentamycin group colonized with P. aeruginosa at the end of the treatment period. Likewise, at that time point, fewer patients treated with gentamycin had purulent sputum (p < 0.03) but still with no difference after the 3-month treatment free follow-up period. Significant clinical improvements for patients treated with gentamycin were also observed for QoL, assessed by LCQ (≥1.3 unit improvement) and SGRQ (≥4 unit improvement), compared to the saline group (81 vs. 20, p < 0.01; 88 vs. 19, p < 0.004, respectively), but only sustained at follow-up for the SGRQ score. Fewer patients in the gentamycin group experienced exacerbations compared to the saline group (0 (0–1) vs. 1.5 (1–2); p < 0.0001), and an increase was also observed in median time (days) to first exacerbation (120 (87–162) vs. 62 (21–122); p = 0.02). There was no significant difference in quantitative bacteriology, 24-hour sputum volume, lung function or development of resistant strains of P. aeruginosa. With regard to adverse events, 22% of the original cohort receiving gentamycin experienced bronchospasm (7 of 32) compared to only 6% (2 of 33) in the original cohort receiving saline. The authors concluded that if nebulized gentamycin is to be effective it should be given continuously, since the effect decline rather quickly after end of treatment. However, further studies will be needed to answer the important question of effect of treatment cycles versus continuous treatment.

In 2012, Chalmers et al.¹³ published a secondary analysis of the study performed by Murray et al. in 2011, mentioned above,¹² focusing on the effect of gentamycin on sputum bacteriology, airway and systemic inflammation in patients with stable non-CF. As described above, 57 patients were randomly assigned to receive nebulized gentamycin 80 mg (n = 27) or 0.9% saline (n = 30) for 12 months with further follow-up 3 months after end of treatment. Another 15 patients, with bacterial load greater than or equal to 1 × 10⁷ CFU/mL when stable, were treated with intravenous antibiotics for 14 days (dose not reported) and matched to a control group of 11 stable patients not receiving any treatment. There was no reduction in bacterial load. The authors conclude that short- and long-term antibiotic therapy reduces markers of airway and systemic inflammation (intercellular adhesion molecule 1 (ICMA-1), E-selectin, vascular cell adhesion molecule 1 (VACM-1), interleukin-8 (IL-8), neutrophil elastase activity, tumor necrosis factor-α, IL-1-β and myeloperoxidase activity).

In conclusion, treatment with nebulized gentamycin may reduce inflammatory markers, reduce bacterial load and improve QoL, but so far, there is no evidence that it can reduce sputum volume or improve lung function.

Macrolides

Azithromycin

Six studies, in total including 348 subjects, have explored the effect of the macrolide azithromycin, administered orally, for the treatment of non-CF BE (Table 2).

Table 2.

Overview of studies investigating the effect of macrolides for the treatment of patients with stable non-cystic fibrosis bronchiectasis.

Drug	Study/author	Design	No. of subjects	Therapy	Duration of treatment	Primary outcome	Sputum weight	FEV1	FVC	QoL	Exacerbations	Admissions	Antibiotic use	Bacterial colonies	Adverse events
Azithromycin	Davies et al. (2004)14	Prospective, non- randomized, follow-up	33	Oral 750 mg o.d. (2 days), oral 250 mg 3 times/week	Unfixed period	Clinical response and sputum volume		No change	No change	Improved	Improved
	Cymbala et al. (2005)15	Randomized, open, case–control, crossover	11	Oral 500 mg b.i.d.	6 months	Exacerbation rate	Improved	No change	No change	Improved		Improved
	Anwar et al. (2008)16	Retrospective follow-up	50	Oral 250 mg 3 times/week	Minimum 3 months	Exacerbation rate	Improved	Improved	No change	No change	Improved			Improved
	Wong et al. (2012)17	Randomized, placebo-controlled, double-blind	141	Oral 500 mg 3 times/week	6 months	Exacerbation rate		No change		No change	No change
	Altenburg et al. (2013)18	Randomized, placebo-controlled, double-blind	83	Oral 250 mg o.d.	12 months	Numbers of infective exacerbation		Improved		Improved	Improved			No change
	De Diego et al. (2013)19	Randomized, placebo-controlled, double-blind	30	Oral 250 mg 3 times/week	3 months	Oxidative stress markers	Improved	No change	No change	Improved	Improved
	Coeman et al. (2011)20	Retrospective follow-up	61	Add-on azithromycina	3–8 weeks	Clinical response				Some improvement
Erythromycin	Tsang et al. (1999)21	Randomized, placebo-controlled, double-blind	21	Oral 500 mg b.i.d.	8 weeks	Sputum volume and lung function	Improved	Improved	Improved					No change
	Serisier et al. (2011)22	Uncontrolled, follow-up, open	24	Oral 250 mg o.d.	Minimum 12 months	Number of exacerbations		No change			Improved
	Serisier et al. (2013)23	Randomized, placebo-controlled, double-blind	117	Oral 400 mg b.i.d.	48 weeks	Exacerbation rate	Improved	Improved		No change	Improved		No change

FEV₁: forced expiratory volume in first second; FVC: forced vital capacity; QoL: quality of life.

^aOral clarithromycin 500 mg or 250 mg once daily, and oral azithromycin 500 mg once daily or azithromycin 500 mg or 250 mg three times a week as add-on.

In 2004, Davies and Wilson¹⁴ published a prospective, non-randomized trial of prophylactic treatment with azithromycin, where 33 patients received 500 mg once daily for 6 days, 250 mg once daily for 6 days and then 250 mg three times weekly for an unfixed period of time (mean 20 months, range 4–38). Patients completing at least 4 months of treatment had a significant reduction in infective exacerbations (mean 0.71 per month vs. 0.13 per month, p < 0.001) and treatment with intravenous antibiotics (mean 0.08 courses vs. 0.003 courses per month, p < 0.001) compared to pretreatment. Symptoms were significantly reduced, using a 5-point scale including sputum, cough, fatigue, exercise tolerance, wheezing and breathlessness, but no improvement in lung function. The study is limited by non-controlled design, few patients and wide variation in treatment period, and a very modest, although statistical significant, changes courses of antibiotics.

Cymbala et al.¹⁵ published a pilot crossover study including 11 patients with non-CF BE in 2005. In random order, patients received standard care or standard care plus azithromycin 500 mg twice weekly for 6 months or vice versa (with washout period). Azithromycin significantly reduced the number of exacerbations (5 vs. 16, p = 0.02). Mean 24-hour sputum production was decreased (15%, p = 0.005) during the active treatment. Patients reported increased energy and QoL during treatment with azithromycin, although quantitative measurements were not reported. No difference in lung function. The authors conclude that add-on azithromycin may be beneficial, but further studies are needed.

In 2008, Anwar et al.¹⁶ published a retrospective follow-up study on long-term low-dose azithromycin. A total of 50 patients received oral azithromycin 250 mg thrice weekly for more than 3 months (mean 9 ± 8 months). There was a significant reduction in exacerbation rate (0.8 (SD 0.3) vs. 0.4 (SD 0.5), p < 0.001) and number of positive sputum cultures (42 vs. 13, p < 0.005) compared to pretreatment. Thirty-six of the 50 patients no longer had sputum production. Lung function could be assessed for only 29 patients and showed a statistically significant improvement in FEV₁ (1.56 vs. 1.64 l, p = 0.005) but no significant change in FVC (2.51 vs. 2.60 l, p = 0.13). The authors concluded that azithromycin should be considered as maintenance therapy for patients with frequent exacerbations. However, study design, modest improvements and incomplete follow-up of enrolled patients limit the validity of the study.

A randomized, double-blind, placebo-controlled trial was published by Wong et al. in 2012.¹⁷ A total of 141 patients were assigned to receive 500 mg azithromycin (n = 71) or placebo (n = 70) three times a week for 6 months, with follow-up after 6 and 12 months. After 6 months, the exacerbation rate was 0.6 and 1.6, respectively, in the intervention and placebo group (rate ratio 0.4, 95% CI 0.3–0.6, p < 0.0001). There was no significant difference in FEV₁ or QoL, assessed by the SGRQ, from baseline in any of the two groups. After 12 months, no significant difference was found in exacerbation rate, FEV₁ or QoL. The authors concluded that azithromycin is an option for prevention of exacerbations in non-CF BE with at least one exacerbation in the past year. However, their conclusion may seem optimistic since they, apart from no difference in FEV₁ or QoL, did not observe a significant effect on exacerbation rate after 12 months.

In 2013, Altenburg et al.¹⁸ published a randomized, double-blind, placebo-controlled study on long-term azithromycin treatment (BAT trial). Eighty-three patients were randomized to either 250 mg azithromycin (n = 43) or placebo (n = 40) once daily for 12 months. At end of treatment, they found an absolute risk reduction of exacerbation in the intervention group of 34% (95% CI 14–53%). The number of patients needed to treat to prevent an exacerbation was three. The FEV₁%predicted (FEV₁%pred) significantly increased in the azithromycin group with 1.0% per 3 months compared to a decrease of 0.1% per 3 months in the placebo group (p < 0.05). QoL measured by SGRQ and the lower respiratory tract infection visual analogue scale (LRTI-VAS) improved significantly more in the intervention group than the placebo group (SGRQ: −6.09 vs. −2.06 per 6 months, p < 0.05, giving an average decrease of 2 × 6.09 = 12.18 after 1 year; LRTI-VAS: 1.11 vs. 0.056 per 3 months, p = 0.047, which the authors state giving an average decrease of 4 × 1.11 = 4.44 after 1 year). There were no significant changes in inflammatory markers or microbiology. The authors conclude that azithromycin lowers the rate of exacerbations, which may also be the reason for the observed improvement in QoL.

Diego et al.¹⁹ published a randomized, investigator-blinded, controlled trial on the effect of azithromycin in patients with non-CF BE in 2013. Thirty patients were allocated to either standard care (n = 14) or standard care plus oral 250 mg azithromycin thrice weekly (n = 16) for 3 months. After ended treatment, patients receiving azithromycin, in comparison with controls, had a significant reduction in exacerbation rate (mean (SD) 0.1 (0.6) vs. 1.2 (0.6); p < 0.05), sputum volume (mean (SD) 2.1 (3.4) mL vs. 8.9 (1.8) mL; p < 0.05) and level of dyspnoea measured by the Borg scale (mean (SD) 0.4 (0.1) vs. 0.1 (0.2)). QoL, measured by SGRQ, improved in patients on azithromycin (mean (SD) −7.9 (3.1) vs. 4.1 (3.8); p > 0.05), but at the subcomponent level with a significant improvement only in the symptom component. There was no difference in lung function. Authors conclude that azithromycin has clinical benefits in patients with non-CF BE. The study is limited by a small number of participants, not having a placebo group and not being double blind.

In conclusion, there is good evidence that long-term low-dose azithromycin can reduce the number of exacerbations in patients with non-CF BE. Furthermore, two of seven studies also found an improvement in FEV₁%pred and four of five studies with symptom score and QoL as end points reported an improvement for patients treated with azithromycin.

Coeman et al.²⁰ published a retrospective, observational cohort study in 2011, where 108 non-CF patients with asthma and/or BE received clarithromycin 500 mg (n = 95) or 250 mg (n = 2) once daily, or azithromycin 500 mg (n = 2) once daily or azithromycin 500 mg (n = 5) or 250 mg (n = 4) three times a week for 3–8 weeks as add-on to standard care. The study does not report any randomization of patients or reasoning behind the differences in treatment duration and dose. In patients with BE only (n = 61), a statistical significant reduction was reported in total symptom score (dyspnoea (0–2) + cough (0–2) + nocturnal cough (0–2) + sputum purulence/volume (0–2) (4.6 to 1.9, p < 0.001). The cohort was divided into responders (n = 34), defined as patients having more than 60% improvement in total symptom score, and non-responders (n = 26), defined as patients not reaching this threshold. Responders were older (61 vs. 53 years, p = 0.004) and more likely to be male (53% vs. 27%, p = 0.043). There was no significant association between responder status and duration of disease, lung function or HRCT pathology. They concluded that clarithromycin might be useful as add-on therapy in patients with non-CF BE. The design of the study is an important limitation since it is retrospective, uncontrolled and non-blinded. Likewise, no valid conclusions can be drawn with regard to dosing; and, furthermore, the cut-off value for responders and non-responders appears not to be validated.

Erythromycin

Three studies, in total including 162 individuals, have investigated the effect of the macrolide erythromycin, in patients with non-CF BE (Table 2).

Tsang et al.²¹ conducted a pilot, double-blind, placebo-controlled study published in 1999. Twenty-one patients with idiopathic (not further defined) BE, and a 24-hour sputum volume above 10 mL, were recruited and treated with erythromycin 500 mg (n = 11) or placebo (n = 8) twice daily for 8 weeks. No information was given regarding randomization. Patients in the intervention group improved significantly in FEV₁ (1.1 L (95% CI 0.8–1.5) vs. 1.2 L (95% CI 0.8–1.7) after treatment, p < 0.05), FVC (1.9 L (95% CI 1.4–2.5) vs. 1.9 L (95% CI 1.4 2.8, p < 0.05) and 24-hour sputum volume (33.7 mL (95% CI 23.0–49.3) vs. 23.8 mL (95% CI 15.7–7.5, p < 0.05), whereas no improvements in these parameters were found for the placebo group. Erythromycin had no effect on bacterial density. The authors conclude that erythromycin improves lung function. However, the improvements were modest, and, on average, the included patients seem to have had very poor lung function, which may be related to the not well-defined term ‘idiopathic BE’. Also it is limiting that symptoms, number of exacerbations or QoL were not defined as outcome of interest, and not reported.

Serisier and Martin²² published an uncontrolled evaluation of the effect of a lower dose of erythromycin in 2011. For 12 months, 24 patients received 250 mg erythromycin once daily. They found a significant decrease in number of exacerbations requiring antibiotics, both oral and intravenous (median 4 (2–11) vs. 2(0–8) after treatment (95% CI 1.5–3.5, p > 0.0001) and days of treatment with antibiotics (44 (15–138) vs. 21 (0–78) (95% CI 18–40, p < 0.0001). They found no effect on FEV₁ or evidence of change in antimicrobial resistance in sputum pathogens. Being retrospective, not having a control group and including small number of patients limits the value of the study.

In 2013, Serisier et al.²³ published the Bronchiectasis and Low-dose Erythromycin Study (BLESS-study), which is a randomized, double-blind, placebo-controlled trial. Patients were allocated to 400 mg erythromycin ethylsuccinate (i.e. 250 mg erythromycin base; n = 59) twice daily or placebo (n = 58) for 48 weeks. Erythromycin significantly reduced the number of protocol-defined pulmonary exacerbations (based on modified Anthonisen criteria²⁴ (76 vs. 114; mean 1.3 (95% CI 0.9–1.7) vs. 2.0 (95% CI 1.5-–2.5), respectively, per patient per year; incidence rate ratio 0.6 (95% CI, 0.4–0.8), p = 0.003). Erythromycin significantly reduced the 24-hour sputum production (change from baseline, erythromycin mean −5.4 g (−16.0–1.1) vs. placebo mean −1.7 g (−8.2−2.6); treatment effect −4.3 g (95% CI −7.8 to −1, p = 0.01)) and was associated with a decreased decline in FEV₁%pred compared to the placebo group (mean −1.6 (SD 4.6) vs. −4.0 (SD 6.6); treatment effect 2.2 (95% CI 0.1–4.3, p = 0.04)). There was no significant effect on total days of antibiotics, symptoms measured by SGRQ, LCQ and 6-minute walking test (6MWT) or inflammatory markers. It was concluded that long-term low-dose erythromycin significantly reduces exacerbations, protects against lung function decline and reduces sputum production.

Based on the available evidence, low-dose erythromycin may reduce the number of exacerbations and decrease sputum production. There is no clear evidence of an effect on lung function, symptoms or QoL. Further studies with more participants are needed to assess the long-term effect of erythromycin in patients with non-CF BE.

Quinolones

Ciprofloxacin

Four studies, in total including 258 subjects, have investigated the effect of quinolones on non-CF BE (Table 3); in three of the studies ciprofloxacin was administered as inhalation.

Table 3.

Overview of studies investigating the effect of quinolones, polymyxins and β-lactam antibiotics for the treatment of patients with stable non-cystic fibrosis bronchiectasis.

Drug	Study/author	Design	No. of subjects	Therapy	Duration of treatment	Primary outcome	Sputum weight	FEV1	FVC	QoL	Exacerbations	Admissions	Antibiotic use	Bacterial colonies	Adverse events
Ciprofloxacin	Rayner et al. (1994)24	Retrospective, non-randomized, follow-up	10	500–1500 mg 2–3 times/day	Minimum 90days	Number of infective exacerbations		No change	No change		Improved		Improved		No change
	Serisier et al. (2013)25	Randomized, placebo-controlled, double-blind	42	Inhalation (nebulized) liposomal 150 mg/free 60 mg o.d.	28 days (3 cycles)	Sputum Pseudomonas aeruginosa density		No change		No change	No change			Improved
	Wilson et al. (2013)26	Randomized, placebo-controlled, double-blind	124	Inhalation (dry powder) 32.5 g o.d.	28 days	Sputum bacterial density		No change	No change	No change	No change			Some improvement	No change
	Antoniu et al. (2013)27	Randomized, placebo-controlled, double-blind	82	Inhalation (dry powder) 50 mg b.i.d.	28 days	Sputum bacterial density	Improved	No change	No change	No change	No change			Improved
Colomycin	Dhar et al. (2010)28	Retrospective, non- randomized, follow-up	19	Inhalation (nebulized) 1-2 megaunits b.i.d.	Minimum 6 months	Exacerbation frequency	Improved	No change			Improved	Improved
	Haworth et al. (2014)29	Randomized, placebo-controlled	144	Inhalation (nebulized), 1mio. IU b.i.d.	Up to 6 months	Time to exacerbation				Improved	No change			Improved
Aztreonam	Barker et al. (2014)30	Randomized, double-blind, placebo-controlled	540	Inhalation (nebulized) 75 mg 3 times/day	4 weeks	QoL-B-RSS				no improvement	no improvement			improved	improved

QoL: quality of life; QOL-B-RSS: quality of life-bronchiectasis respiratory symptoms score; FEV₁: forced expiratory volume in first second; FVC: forced vital capacity.

In 1994, Rayner et al.²⁵ published a small, retrospective study of 10 patients, receiving oral ciprofloxacin 500–1500 mg, two to three times daily for minimum 90 days (mean duration 412, range 90–860). Compared to the 12 months before treatment, patients had a decrease in number of exacerbations (6.2 ± 2.9 vs. 0.5 ± 0.5; p < 0.01) and number of hospitalizations during the treatment period (1.6 ± 0.7 vs. 0.3 ± 0.5, p < 0.01). There was a significant increase in peak flow and decrease in residual volume (252 ± 92 L/min vs. 326 ± 103 L/min, p < 0.01 and 155 ± 53%pred vs. 131 ± 46%pred, p < 0.05, respectively) but no change in FEV₁, FVC, total lung capacity or haematology. Prior to treatment with ciprofloxacin, five patients had sputum positive for P. aeruginosa. After treatment, two had sputum cultures negative for P. aeruginosa, whereas to had sputum culture with P. aeruginosa resistant to ciprofloxacin. No serious adverse effects were reported. The authors concluded that long-term oral ciprofloxacin is well tolerated and can improve symptoms and reduce hospital admissions, but the development of ciprofloxacin resistance is of concern. The study is, however, limited by being retrospective, having a small number of participants, no control group and a very wide range of dose and duration of treatment with oral ciprofloxacin.

In 2013, Serisier et al.²⁶ published a larger phase II, randomized, double-blind, placebo-controlled trial of inhaled ciprofloxacin comprising 42 patients positive for P. aeruginosa. The intervention group received three treatment cycles of 28 days on/28 days off dual release ciprofloxacin inhalation containing 150 mg liposomal ciprofloxacin and 60 mg free ciprofloxacin (n = 20), whereas the control group was treated with placebo (n = 19). They found a significant reduction in P. aeruginosa density after 28 days in the intervention group compared to the placebo group (−4.2 ± 3.7 log₁₀ CFU/g vs. 0.1 ± 3.8 log₁₀ CFU/g; p = 0.002). In the ‘off-therapy’ periods, there was an increase in sputum P. aeruginosa density towards baseline. There were no significant differences in number of exacerbations, FEV₁ or QoL, measured by SGRQ, and 6MWT. The authors concluded that ciprofloxacin significantly reduces P. aeruginosa density. The reduction of P. aeruginosa is relevant, since studies have previously shown that presence of P. aeruginosa is associated with worse QoL,³¹ more exacerbations³² and faster lung function decline.³³ However, in this study by Serisier et al.,²⁶ no association was found between eradication and improvements in outcome.

In 2013, Wilson et al.²⁷ published a phase II, multicentre, randomized, double-blind, placebo-controlled trial, where patients received either 32.5 mg ciprofloxacin dry powder inhalation (n = 60) or placebo (n = 64) twice daily for 28 days. Follow-up was at day 28, 42, 56 and 84. Patients receiving ciprofloxacin had a significant reduction in sputum bacterial load at day 28 (−3.6 log₁₀ CFU/g, range −9.8–5.0 log₁₀ CFU/g) compared to the placebo group (−0.3 log₁₀ CFU/g, range −8.0–5.3 log₁₀ CFU/g, p < 0.001). However, this difference was not present at follow-up after 42 days. In the inhaled ciprofloxacin group, 14 (35%) subjects had negative sputum cultures at day 28 compared to 4 (8%) in the placebo group (p = 0.001), but not after 42 days. There was no significant difference in number of exacerbations, use of antibiotics, lung function, CRP, QoL (SGRQ) or reports of adverse effects including bronchospasm. The authors concluded that treatment with ciprofloxacin dry powder inhalation is well tolerated and significantly reduces total bacterial sputum load, but further studies are required regarding long-term effects. The study showed that if inhalation with ciprofloxacin is to be used to reduce sputum bacteriology it should possibly be given continuously, although it had no effect on QoL or number of exacerbations.

A phase II, randomized, double-blind, placebo-controlled study was published in 2013 by Antoniu and Azoicai²⁸ Eighty-two patients were treated with either inhaled dry powder ciprofloxacin 50 mg (n = 37) or placebo (n = 45) twice daily for 28 days. The follow-up period was 56 days. Patients receiving active treatment had a significant decrease in sputum bacterial density compared to the placebo group at the end of the treatment period (log₁₀ CFU/g −3.6 vs. −0.3, p < 0.001) but with no significant difference at follow-up. They also had a higher percentage bacterial eradication at end of therapy, compared to baseline (35% vs. 8%, p = 0.001), and a larger decrease in sputum volume (18% vs. 10%) and blood neutrophils (−0.36 × 10⁶/mL vs. 0.60 × 10⁶/mL, p = 0.0014). There were no significant differences in number of patients with at least one exacerbation, lung function or adverse effects. An insignificant improvement in QoL was reported for the ciprofloxacin group, but the method of assessment was not specified further.

In conclusion, ciprofloxacin is the only quinolone examined for the treatment of stable non-CF BE. The available studies suggest that inhaled ciprofloxacin is safe to use and can reduce the density of P. aeruginosa and also bacterial load in general. There is, so far, no convincing evidence for an effect on lung function, number of exacerbations or QoL.

Polymyxins

Colistin

The effects of colistin on patients with non-CF BE and chronic colonisation with P. aeruginosa have been investigated in two studies, in total including 163 individuals (Table 3).

In 2010, Dhar et al.²⁹ published a retrospective, non-randomized study including 19 patients colonized with P. aeruginosa, not receiving macrolide treatment within minimum 4 weeks. Patients received nebulized colistin 1–2 mega units twice daily for minimum 6 months (mean 21 months, range 6–39 months). At the end of the treatment period, the patients had a significant decrease in number of exacerbations per year (mean 8 vs. 3, p < 0.001), in admissions per year (mean 3 vs. 1, p > 0.002), P. aeruginosa sputum positivity per year (mean 4.2 vs. 0.5, p < 0.001) and self-reported sputum volume in mL (31 vs. 14, p < 0.001) compared to before treatment. There was no difference in FEV₁. The small number of participants, being retrospective and not having a control group is important limitations.

Haworth et al.³⁰ published a randomized, placebo-controlled study in 2014 including 144 patients enrolled within 21 days of completing a course of anti-pseudomonas antibiotics for an exacerbation. Patients were randomized to receive colistin 1 million IU (n = 73) or placebo (0.45% saline; n = 71) twice daily via nebulizer for up to 6 months. The study did not reach its primary end point as no significant difference in median time to exacerbation was found between the two groups. However, a post hoc analysis revealed that adherent patients, recorded by the nebulizer, receiving active treatment had an increased median time to exacerbation (165 days vs. 111 days, p = 0.04) compared to those on placebo. This subgroup of enrolled patients also had a reduced density of P. aeruginosa after 4 and 12 months (p = 0.001 and p = 0.006) and improved QoL, measured by SGRQ, after 26 weeks (p = 0.006), compared to the placebo group. No safety concerns were reported.

In conclusion, the two studies suggest an improvement in exacerbation rate, QoL and P. aeruginosa density in patients receiving colistin and being adherent. However, the evidence obtained so far is limited, and primarily based on a post hoc analysis, more studies on the effect of colistin in this group of patients are, therefore, needed before valid conclusions can be drawn.

β-Lactam antibiotics

Aztreonam

The safety and efficacy of inhaled aztreonam for the treatment of non-CF BE has been investigated in two trials comprising 540 patients (Table 3).

In the AIR-BX1 and AIR-BX2 trials, Barker et al.³⁴ investigated the clinical effects of inhaled aztreonam in patients with non-CF BE. The two double-blind, multicentre, randomized, placebo-controlled trials comprised 540 patients and was reported as a pooled analysis. In both trials, patients were randomly assigned aztreonam for inhalation solution 75 mg or placebo three times daily using eFlow nebulizer for two 4-week courses, each followed by a period of 4 weeks of treatment. The primary outcome was change from baseline in the QoL-bronchiectasis respiratory symptoms score (QOL-B-RSS) after 4 weeks of treatment. The pooled analysis showed that compared with placebo, treatment with aztreonam did not lead to clinically significant improvements in QOL-B-RSS score. In addition, no improvement was observed in time to first exacerbation. On the other hand, compared to placebo, aztreonam was observed to significantly reduce the sputum microbial density especially gram-negative organisms including Pseudomonas. Compared to placebo, treatment-related adverse events were reported more often by aztreonam-treated patients compared with the placebo group. However, the reasons for the disagreement between microbiological and clinical end points remain unclear.

To conclude, although aztreonam had positive impact on microbiological parameters, no evidence for an effect on clinical outcomes was found. So, therefore, the efficacy of inhaled aztreonam in patients with non-CF BE will have to be studied in further trials, probably with different design and dosing schedule, before valid conclusions can be drawn.

Discussion, conclusions and future directions

Several studies addressing the treatment with antibiotics of stable patients with non-CF BE have, although often based on relatively small studies, reported positive impact on a number of relevant outcome measures for a number of different treatment regimens. However, within each group of antibiotics, the available studies have often not reported consistent findings.

A very important outcome of studies assessing strategies for management of non-CF BE is impact on exacerbation rate.² However, unfortunately so far, the available studies of antibiotic therapy in this group of patients have shown not only inconsistent findings but also for many of the antibiotics used in the trials no effect on exacerbation rate. Future studies with exacerbation rate as the primary outcome are, therefore, of outmost importance. Furthermore, these studies should also address the issue of continuous therapy versus cycles of active therapy and, preferably, be of at least 12 months duration.

In line with the above considerations, future studies of antibiotic therapy for non-CF BE should probably also focus more on patient-related outcomes, including symptoms and QoL, instead of focusing on eradication of bacterial colonisation, as the latter not necessarily has an impact on morbidity and long-term outcome.

In conclusion, the present systematic review of therapy for stable non-CF BE revealed limited amount of evidence for various antibiotic regimens. There is, therefore, clearly an urgent need for high-quality, randomized, placebo-controlled, long-term studies, probably not least regarding the efficacy of erythromycin, gentamycin and colistin, focusing on prevention of exacerbations and patient-related outcomes.