Chronic obstructive pulmonary disease (COPD) is preventable and treatable common disease characterized by incompletely reversible and progressive development of airflow limitation (1). It was estimated that there were nearly 300 million patients with COPD worldwide in 2017 (2). COPD has caused at least 5.8% of total deaths worldwide each year and has become the third leading cause of death in 2019 (3). In China, the annual direct medical expenses for COPD were between 72 and 3,565 US dollars per capita, which accounts for about 40% of the total income of an ordinary family (4). On average, each COPD patient experiences 0.5 to 3.5 acute exacerbations per year, which is an important reason for the increase in hospitalization, disease progression and mortality as well as the decline in health (5, 6). An epidemiological study in China showed that each patient hospitalized with AECOPD spent about RMB 11,598 in treatment per year (7). Reducing the incidence of AECOPD may help slow the progression of the disease and improve the quality of life of patients (8). Therefore, targeted interventions for patients with stable COPD are particularly important.

Among the adults with a diagnosis of COPD, those who have exacerbation history, greater disease severity, higher symptom burden, significant comorbidities and higher blood eosinophil count are more likely to develop moderate-to-severe exacerbation (9, 10). Management of those modifiable risk factors is of great value for exacerbation prevention. At present, several types of drugs have been proven by different levels of evidence to reduce exacerbation frequency in patients with COPD, mainly including inhaled steroids, long-acting bronchodilators, phosphodiesterase inhibitors and mucolytics (11). Non-pharmacological therapies such as smoking cessation, vaccination, and pulmonary rehabilitation are also recommended for the prevention of AECOPD (12). Yet, 22% ~ 40% of COPD patients still experience at least one moderate or severe exacerbation each year (13). This forces us to reflect on the limitation of current preventive measures and seek new methods to prevent acute exacerbation based on risk factors.

Chronic respiratory bacterial colonization has recently been increasingly noticed to play an important role in the pathogenesis of COPD. Potential pathogenic microorganisms could be isolated from the lower respiratory tract of as many as 74% of COPD patients, which was much higher than that of healthy people (14). Similarly, up to 80% of AECOPD were related to bacteria or viruses (15). Therefore, we believe that improving the microenvironment of the lower respiratory tract may contribute to the prevention of AECOPD. Aminoglycosides conserve good susceptibility to Gram-negative bacteria, but the poor pulmonary penetration and the high frequency of side effects in systemic administration limit their clinical use for years. However, inhaled antibiotics can theoretically compensate for the shortcomings of systemic delivery, thus having a great potential for decolonization in COPD (16). Intermittent use of low-dose macrolide antibiotics, such as azithromycin, was found to reduce the incidence of AECOPD and greatly decrease the total respiratory bacterial load (17, 18). However, macrolides were more likely to act as an inflammation modulator rather than an antimicrobial agent in preventing acute exacerbation (19). Thus, the role of modulating respiratory microbiota in using antibiotic prophylaxis for AECOPD prevention has rarely been studied. Influenza vaccination has been believed to reduce the frequency of AECOPD, the number of outpatient visits, hospitalizations, and mortality (20, 21). Similarly, Streptococcus pneumoniae vaccine was associated with the reduced risk of hospitalization for patients with moderate to severe COPD (22, 23). Given that influenza virus and S. pneumococcal infections are among the most common microbial causes of AECOPD (10, 14), we wondered if the inoculation of both vaccines simultaneously might be more effective. In addition, Alexandre et al. reported that probiotics might be related to the lower prevalence of respiratory infection and the decline of respiratory colonization, which suggest the potential application of probiotics in COPD management (24, 25). In particular, oral intake of Lactobacillus rhamnosus GG has been shown to reduce the adhesion of potential pathogens in respiratory tract and reduce the pulmonary exacerbations of cystic fibrosis in several studies (26–28), which had brought attention to the application of probiotics in COPD management.

Based on the previous evidence, we hypothesized that the preventive use of influenza-S. pneumoniae vaccines, oral probiotics or inhaled antibiotics (amikacin) during the stable phase of COPD could contribute to reduced respiratory colonization and improved airway microenvironment, so as to delay the progression of disease and improve life quality. Therefore, an exploratory prospective, randomized controlled trial was done to assess efficacy and safety of the three interventions in preventing AECOPD. This study also aimed at verifying the possible improvement of these preventive measures on the respiratory symptoms of patients with COPD.

This pilot study sought to determine whether modulating respiratory microbiota during stable stage of COPD could prevent acute exacerbation. As multiple studies have confirmed (41–43), we found that conventional treatment in accordance with the GOLD 2019 report, including standardized drug treatment and tobacco cessation support, with or without the other three interventions vaccination could significantly decrease the frequency of moderate-to-severe AECOPD. All the four groups had reduced CAT scores and mMRC scores at the 6-month follow-up, but the differences were not all significant. More importantly, the additional administration of influenza-S. pneumoniae vaccinations and long-term oral probiotic LGG, respectively, postponed the next onset of AECOPD by 41.5 days (p = 0.044) and 50.6 days (p = 0.017). The aerosol inhaled amikacin showed the same tendency to delay the exacerbation COPD, but no statistically significant difference was detected (p = 0.100), perhaps due to the fact that not all subjects enrolled in this group had high bacterial burden in lower respiratory tract.

The frequency of AECOPD in subjects who were given the influenza-S. pneumoniae vaccination decreased the most during the follow-up period compared to the year before enrollment (1.42 ± 1.79 per year, p = 0.002), suggesting that dual vaccination might be more effective in preventing AECOPD than the other interventions. We speculated that both the oral probiotics and the aerosol inhaled amikacin lasted only 3 months, whereas the validity period of vaccination could usually be maintained for more than 1 year covering the entire follow-up period, thus making it more advantageous in this study. It remains to be verified whether prolonging the intervention time of the oral probiotics and the aerosol inhaled amikacin help enhance their efficacy. Moreover, it has been reported that patients with influenza infection may be more susceptible to infections of other pathogens, such as S. pneumococcal, the mechanism of which is thought to be related to the extensive respiratory epithelial damage caused by the direct effects of the virus, the effects of induced interferon and the actions of cytotoxic T-cells after influenza virus infection (44). Compared with receiving influenza vaccine or S. pneumococcal vaccine alone, the concomitant injection of both showed additive effects in reducing the incidence of pneumonia, all-cause mortality, all-cause hospitalizations and inpatient expenditures of all diseases among the elderly (45, 46). Our study demonstrated the significant benefit of influenza-S. pneumococcal vaccination in reducing the frequency of AECOPD. Nevertheless, the additive effect of dual influenza and S. pneumococcal vaccination compared with separate administration in AECOPD prevention needs to be further verified, which will help to clarify the optimal vaccination mode in stable COPD.

Previous studies have found the crucial bidirectional connection between the intestinal microbiota and the lungs, namely the “Gut-Lung axis” (47). In respiratory infection diseases, modulating gut microbiota by oral probiotics could reduce the duration of intensive care units (ICU) admission, the severity of the common cold and the incidence of ventilator-associated pneumonia (VAP) and upper respiratory infections (48). As for chronic respiratory diseases, oral probiotics could significantly improve allergic rhinitis via improving at clinical signs, decreasing the rate of exacerbation, and reducing the use of relieving medication (24, 49). However, despite improving serum inflammatory factors and cytokines, oral probiotics could not significantly improve signs and symptoms in asthma (50). Although a reduction in alveolar inflammatory cells infiltration and subsequent lung damage was observed in emphysema mice receiving oral probiotics (51), studies evaluating the clinical effects of oral probiotics on COPD are still limited. This is the first clinical trial to report that oral probiotic LGG can delay the occurrence of AECOPD, reduce the frequency of AECOPD, and improve symptoms in patients labeled GOLD D, suggesting that COPD is also a chronic respiratory disease that can benefit from intestinal microbiota regulation. Better understanding of the “Gut-Lung axis” is needed to design gut microbiota-associated strategies for the treatment and prevention of COPD.

Results from the few previous studies of long-term inhaled antibiotics in patients with stable COPD have been discouraging. A clinical trial (NCT00739648) conducted in the United States observed no changes in exacerbation rate after levofloxacin inhalation. Bruguera-Avila N and colleagues found that long-term inhalation of colistin for at least 3 months was not associated with the number of AECOPD cases not requiring admission in COPD patients with bronchial colonization by Pseudomonas aeruginosa but could decrease the hospitalization and the length of hospital stay (52). These results suggested the necessity of selecting sensitive antibiotics and the possibility that inhaled antibiotics may be more effective in preventing severer AECOPD. Most of the stable COPD patients have airway bacterial colonization dominated by Gran-negative bacteria, including Moraxella pneumoniae, S. pneumoniae and P. aeruginosa (53), which can be covered by amikacin. Our study found that inhaled amikacin did not significantly delay the occurrence of AECOPD. However, in the subgroup analysis, patients with high symptom burden had prolonged onset of moderate-to-severe AECOPD after receiving any of the three interventions, including aerosol amikacin (237.3 vs. 198.2 days), compared to the conventional group. Additionally, only the amikacin group had a significant improvement in CAT scores at the 6-month follow-up (18.64 vs. 21.71). It suggests that patients with high symptom burden may represent a group of people with specific airway microbiota and high load of respiratory tract colonization. These patients may be more likely to benefit from appropriate inhaled antibiotics. Thus, the inclusion criteria of future studies on the nebulized antibiotics against respiratory decolonization should emphasize the isolation of bacteria (such as P. aeruginosa) or the manifestations of bronchitis (such as cough and sputum).

The COVID-19 pandemic has become a major obstacle to conducting this study, thus resulting in several limitations. Due to the social policies for epidemic control, part of the follow-up was completed through telephone visits, which may affect the accuracy of the outcome assessment. However, the onset of moderate-to-severe AECOPD is a relatively objective event according to the criteria defined in this study, so we believe that the primary outcome data collected are reliable. Without the onsite visit, we were unable to collect blood and sputum samples, which made it impossible to complete the planned microbial study and restricted the analysis of existing clinical indicators. In addition, the community epidemic prevention measures, such as wearing a mask, carried out in Shanghai and Guangzhou since the beginning of 2020 resulted in a visual reduction in respiratory infections (e.g., flu) and AECOPD, which could reduce the effects of the three interventions.

For moderate-to-very severe COPD patients with a history of moderate-to-severe exacerbations, the combined vaccination (influenza and Streptococcus pneumoniae vaccine) and continuous oral probiotic LGG for 3 months can significantly delay the next moderate-to-severe exacerbation. Similarly, for patients with high symptom burden, the aerosol inhaled amikacin also significantly delayed the next moderate-to-severe exacerbation.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The trial has been approved in the Ethics Committee of Zhongshan Hospital of Fudan University (B2017-197R). The studies were conducted in accordance with the local legislation and institutional requirements. The participants provided their written informed consent to participate in this study.

J-lH: Conceptualization, Data curation, Formal analysis, Funding acquisition, Investigation, Methodology, Project administration, Resources, Visualization, Writing – original draft, Writing – review & editing. Z-fY: Conceptualization, Data curation, Formal analysis, Funding acquisition, Investigation, Methodology, Project administration, Resources, Writing – review & editing. Q-jC: Conceptualization, Data curation, Formal analysis, Funding acquisition, Investigation, Methodology, Writing – review & editing. Y-pH: Conceptualization, Data curation, Formal analysis, Funding acquisition, Investigation, Methodology, Validation, Writing – review & editing. Z-tL: Data curation, Investigation, Project administration, Writing – review & editing. R-rD: Data curation, Investigation, Project administration, Writing – review & editing. B-fH: Formal analysis, Investigation, Visualization, Writing – review & editing. Y-xW: Formal analysis, Investigation, Software, Validation, Writing – review & editing. JZ: Conceptualization, Data curation, Formal analysis, Funding acquisition, Investigation, Methodology, Project administration, Resources, Supervision, Visualization, Writing – original draft, Writing – review & editing.