Effects of long-term tobramycin inhalation solution (TIS) once daily on exacerbation rate in patients with non-cystic fibrosis bronchiectasis

The BATTLE study was a multicenter, double-blind, placebo-controlled randomized controlled trial, conducted in 6 hospitals in the Netherlands between September 2016 and December 2019. Full details of the study design are available in the Additional file 1. Documented approval from the Independent Ethics Committees and Institutional Review Boards was obtained from all participating centers before start of the study and after the interim analysis.

Patients with proven bronchiectasis on HRCT, aged ≥ 18-year-old, chronic respiratory symptoms and at least two respiratory exacerbations treated with antibiotics and/or prednisolone in the preceding year were recruited from the outpatient clinic. All participants had one or more positive sputum cultures for gram-negative pathogens or S. aureus in the preceding year, as well as one positive sputum culture with the predefined pathogens at baseline. Patients with known CF, active allergic bronchopulmonary aspergillosis, tuberculosis or non-tuberculous mycobacterial infection were excluded. Other exclusion criteria were the use of maintenance antibiotics, except for maintenance treatment with macrolides if treatment was not initiated within 1 month prior to study entry. We also excluded patients treated with prednisolone > 10 mg per day for > 1 month, and/or patients treated with mucolytics. For safety reasons, patients with chronic renal insufficiency (GFR < 30 ml/min), earlier diagnosed tinnitus, hearing impairment, balance disorders or neuromuscular disorders were excluded. An overview of the complete inclusion and exclusion criteria are shown in Additional file 1.

After the run-in period of 4 weeks, stable bronchiectasis patients were randomly assigned (1:1) to receive TIS 300 mg/5 ml OD or placebo (NaCl 0.9%) OD for 52 weeks by using the InnoSpire Deluxe compressor (Philips Respironics) with a SideSteam Plus nebulizer with filter and mouthpiece.

All patients underwent a tolerance test with spirometry before and after the first dose of the study medication to assess the occurrence of local intolerability. Airway hyperresponsiveness was defined as a decrease in FEV₁% of predicted of 20% following the study medication and/or saturation < 90%, and/or signs of bronchospasm. Study visits were planned at the outpatient ward every 3 months, and clinical, QoL questionnaire’s, bacteriological and laboratory examinations were performed as well as spirometry. The study medication was delivered from a central pharmacy to each study center in batches, stored at 2–8 C. Medication was dispensed to the patient at each visit for a period of 3 months. The empty ampules were collected at each visit in blinded sachets. Patients were asked to use the diary card weekly to examine the respiratory symptoms and if so, to notice an exacerbation. The final visit was conducted 4 weeks (run-out period) after the end of the treatment period. See Additional file 1 for an overview of the study schedule.

The primary outcome of the study was the number of exacerbations during the 1-year treatment period. A protocol defined pulmonary exacerbation (PDPE) was defined as the presence of three or more of the following symptoms or signs for at least 24 h: 1. increased cough; 2. increased sputum volume and/or purulence; 3. haemoptysis; 4. increased dyspnoea; 5. increased wheezing; 6. fever (> 38.5 °C) or malaise AND the treating physician agreed that antibiotic and/or prednisolone therapy was required. Secondary outcomes were time to next exacerbation, change in lung function (FVC% predicted, FEV1% predicted) and QoL measurements based on lower respiratory tract infections- visual analogue scale (LRTI-VAS), Quality of Life-Bronchiectasis (QOL-B) and the Leicester cough score [24‐26]. In addition, change in biomarkers, liver function and renal function, and bacterial diversity in sputum with the development of tobramycin resistance were evaluated in both groups. Additional safety assessments were assessed in both groups. For an overview of the study assessments see Additional file 1.

Details of the randomization process, sample size calculations and statistical analyses were described at the online Additional file 1. Using an expected baseline exacerbation frequency and effect size, a power analyses was done based on a Poisson regression model. After inclusion and follow up of 50% of the randomized patients an interim analysis was conducted to test the assumptions used in the sample size calculation. The interim analysis revealed that the assumptions made for the sample size calculation were appropriate. However, the drop-out rate was 34%, which was higher than the expected drop-out rate of 30% and made it necessary to increase the sample size by 6 patients.

The primary efficacy analysis was performed on the modified intention to treat (mITT) population. Patients who were randomized but dropped out directly after the tolerance test due to a moderate-severe bronchial obstruction, or during the first 2 weeks of the study treatment, were excluded from the mITT analysis. All randomized patients who received and completed treatment according to the study protocol for at least 9 months were included in the per protocol (PP) analysis. Early termination of the study or incomplete data collection was defined as non-evaluable. Statistical analysis was performed using SPSS version 25. The analyses were performed on the mITT- and the PP-population; discrete variables were presented as counts (percentage) and continuous variables as means with standard deviation (SD) if normally distributed and medians with interquartile range (IQR) if not normally distributed. Between groups differences were tested using the chi- square or Fisher exact test if appropriate in case of nominal or ordinal variables. In case of continuous variables, the student T-test or the Mann–Whitney U Test depending on the distribution was used. The effect of TIS as compared to placebo on exacerbation frequency was analyzed by means of negative binomial regression analysis correcting for the exacerbation rate in the year prior to the study. The association between use of TIS and exacerbation rate was expressed as a rate ratio with 95% confidence interval (RR (95%CI)). Linear mixed model analysis was used to analyze the effects on lung function and QoL over time. Results were presented as differences with p-values. Time to first exacerbation during the treatment period was analyzed using a Cox regression analysis. Resulting associations were presented as hazard ratios with 95% confidence intervals (HR (95%CI). A p-value < 0.05 was considered statistically significant.

A total of 58 patients were randomly assigned to receive either TIS or placebo OD and were included in the ITT-population. A total of 6 patients (3 patients in each group) dropped out in the first 2 weeks of study treatment (Fig. 1). These patients were excluded from the mITT population. Baseline patient characteristics of the mITT population were well balanced between TIS and placebo (Table 1), except for maintenance stable doses of prednisolone < 10 mg, which was significantly more frequent in the TIS population (p = 0.04). All patients passed the tolerance test at visit 1, with no occurrence of severe bronchus obstruction (defined as a decrease in FEV₁% of predicted of > 20% and/or saturation < 90%) after the first dose of the study medication. Treatment compliance was high; from empty ampules count, we estimated that patients adhered 94% over the time in the TIS group and 95% in the placebo group.

In the mITT population, a total of 99 protocol defined exacerbations were reported during the study, 41 (41.4%) of which occurred in the TIS-treated patients and 58 (58.6%) in the placebo-treated patients. During the study, a median number of 2 (IQR 0–2) exacerbations was found in the TIS treatment group and 2 (IQR 1–3) in the placebo treatment group. Negative binomial regression analysis correcting for baseline exacerbation rate showed a RR of 0.74 (95% CI 0.49–1.14) suggesting a lower, non-significant exacerbation incidence rate in TIS group (p = 0.15). For the PP-population similar results in exacerbation incidence were found, with a RR of 0.77 (95% CI 0.42–1.17) (Additional file 1).

The total number of exacerbations of the mITT population also included number of exacerbations during the study period from patients who terminated the study medication.

The secondary endpoint time to first exacerbation differed between both groups with a mean of 29 (SD19.6) weeks for patients treated with TIS and a mean of 21 (SD18.7) weeks for the placebo-treated patients, resulting in a hazard ratio of 0.64 (p = 0.15) (Fig. 2).

The longitudinal analysis of lung function in the mITT-population showed no statistically significant differences between both groups (Additional file 1). Comparisons within the TIS treatment group showed a non-significant improvement in FEV₁% of predicted with a mean of 65.9% (SD 24.9) at baseline and a mean of 72.7% (SD 23) at the end of the study (p = 0.58). A non-significant improvement was also found in the placebo population, with a mean of 70.5% (SD 24.0) at baseline and 73.2% (SD 23.5) at the end of the study (p = 0.21). The FVC % of predicted was stable during the study for both groups (Table 2).

The QoL was measured during the study period by using the LRTI-VAS, the Leicester Cough questionnaire, and the QoL-B questionnaire in both groups [24‐26]. Longitudinal analysis of the QoL measurements showed no significant improvement for both the mITT and the PP population (Additional file 1).

However, within the TIS population significant improvement of the total LRTI-VAS score (improvement is reflected by a reduction in LRTI-VAS score) and the total Leicester Cough questionnaire score were found after 1-year of study treatment (p = 0.02, p = 0.02) (Table 2). These differences were not observed into the placebo population (p = 0.17; p = 0.08) (Table 2). Within groups comparisons showed no significant differences for the 8 QoL-B subscales. However, in the TIS treatment group, the QoL-B respiratory symptom scale (RSS) improved from 56.1 (SD 17.8) to 61.0 (SD 17.4) points after 1-year (p = 0.05), with also improvement of the QoL-B health perceptions scale from 42.4 (SD 16.8) to 48.1 (SD 19.7) points at the end of the study (p = 0.08), while the QoL treatment burden scale and the emotional scale decreased (p = 0.65; p = 0.05) (Table 2).

A total of 162 sputum cultures were collected during the study. At baseline, the most common isolated pathogens were H. influenzae (16 (30.8%)) and P. aeruginosa (14 (26.9%)). At the end of the study, no pathogens were isolated from sputum in 10 (38.5%) TIS-treated patients as compared to 4 (15.4%) placebo-treated patients (Table 3). In 80% of patients with P. aeruginosa chronic infection treated with maintenance TIS (n = 5 (19.2%)), no P. aeruginosa was isolated in sputum at the end of the study, as compared to 33.3% after placebo (n = 9 (34.6%)). In one (3.8%) TIS-treated patient with Escherichia Coli (E. coli) infection tobramycin resistance occurred during the study treatment. After the run-out phase, 4 weeks after the last dose of the study medication, Aspergillus fumigatus was isolated from two sputum cultures, 1 (3.8%) in the placebo population and 1 (3.8%) in the TIS population, both with no clinical signs of Aspergillus infection. No other ‘opportunistic’ pathogens were found during the study. In addition, serum inflammatory markers were collected every 3 months during the study. No differences were found in CRP, leucocytes, and eosinophil counts. (Additional file 1). During an exacerbation the inflammatory markers were not collected on regulatory base (not shown).

All patients underwent a tolerance test at the start of the study. Spirometry measurements were performed before and after the first dose of the study medication (Additional file 1). No occurrence of severe bronchus obstruction was found directly after the first dose. However, 3 (8.8%) TIS-treated patients developed respiratory symptoms in the first 4 weeks of study treatment (none in the placebo-treated patients) and withdrew from the study. Unfortunately, no spirometry was performed during the development of these respiratory symptoms. Two out of 29 (6.8%) TIS-treated patient and 3 out of 29 (10.3%) placebo-treated patients experienced no effect of the study treatment and withdrew from the study. Two out of 58 (3.4%) patients (1 TIS and 1 placebo) mentioned intensive study treatment as a reason for discontinuing the study (Fig. 1). In the mITT population, 7 (26.9%) TIS-treated patients mentioned cough as compared to 5 (19.2%) placebo-treated patients (p = 0.47). An overview of all the side effects is shown in Table 4. A total of 28 serious adverse events (SAE) were reported during the study, all related to hospital admissions, of which 24 to a pulmonary exacerbation, 2 patients with known cardiac diseases,1 patient with a near collapse and 1 patient with an anaphylactic reaction on amoxicillin clavulanate. One hundred and fifty-seven adverse events were reported, mostly related to a PDPE or non-PDPE (Additional file 1).