Intermittent prophylactic pulsed treatment with moxifloxacin resulted in a 19% reduction in the odds of exacerbation in the ITT population and a 25% reduction in the PP EOT population in this study. This corresponds to NNTs of 28 and 19, respectively. Pre-specified subgroup analyses demonstrated that the reduction in exacerbations with moxifloxacin was seen in COPD of all severity categories, among smokers and ex-smokers, and in patients receiving concomitant COPD treatments including inhaled steroids and long-acting bronchodilators. In a post-hoc analysis, a larger (45%) reduction in the odds of exacerbation, corresponding to an NNT of 7, was seen among patients who reported purulent/mucopurulent sputum at baseline. The extent of reduction in exacerbations in these patients is similar to that seen with long-acting bronchodilators and inhaled steroids in recent clinical trials in COPD [21–26]. In a study with bronchoscopic sampling, sputum purulence was shown to be a reliable indicator of significant bronchial infection in patients with exacerbations of COPD . It is possible that sputum purulence is also a marker for chronic bronchial infection in stable COPD, thereby explaining the greater benefit with prophylactic antibiotic treatment in these patients.
Intermittent moxifloxacin did not significantly improve overall health status, reduce rates of hospitalization or mortality, or slow the ongoing decline in lung function in COPD. The size and duration of the study were not adequate for these secondary end points. In the patient population recruited in this study, hospitalization and mortality rates were low, making it difficult to observe a difference between the two treatment groups. In terms of lung function, it is, perhaps, ambitious to expect improvements in underlying COPD in such a relatively short study of only six 5-day courses of an antibiotic.
There were more adverse events with moxifloxacin than with placebo, however, the overall levels of drug-related adverse events were low in both groups (9.3% with moxifloxacin and 3.8% with placebo) and comparable to findings in another large study with moxifloxacin in AECOPD . The most common treatment-related adverse events related to the gastrointestinal tract, which also resulted in more premature terminations (3.6% vs 1.8%, in the moxifloxacin- and placebo-treated groups, respectively; p = 0.07). C. difficile infection was not reported in any of the patients taking moxifloxacin.
A major concern with antibiotic use is the emergence of resistant organisms. Regular monitoring of sputum flora and, in a subgroup, of faecal flora was therefore an essential part of the PULSE study. There was a transient increase in MIC of one strain of S. pneumoniae and three isolates of S. aureus. However, these isolates did not persist or cause exacerbations, and resistance emergence was not seen among patients in either moxifloxacin or placebo arms. The dosing regimen in PULSE was based on pharmacodynamic/pharmacokinetic principles, with intermittent dosing of a potent antibiotic at full doses, rather than use of therapeutic or subtherapeutic doses for prolonged periods. We speculate that such a dosing regimen was related to the lack of resistance emergence seen in this study although longer periods of observation are needed to confirm this. Patients who had colonization of moxifloxacin-resistant P. aeruginosa at baseline were excluded from this study but those with P. aeruginosa sensitive to moxifloxacin were included. The detection of P. aeruginosa with a higher MIC during the study than those present at screening/randomization suggests that patients colonized by this organism in the airway should not be considered for the therapeutic regimen described in this study. Further long-term studies are required to determine whether this type of intermittent preventive antibiotic therapy circumvents the development of resistant organisms.
There are three main limitations of the present study. The first is that there were fewer than expected exacerbations, despite the use of a wider definition of an exacerbation to maximize the number of events; only 0.88 were reported in the placebo group compared with the predicted 1.4 exacerbations per patient during the 48-week study period, with approximately 50% of the placebo group experiencing no exacerbations during 48 weeks. This 'placebo' effect of reduced exacerbations has been observed in other clinical trials and is probably related to closer monitoring and better compliance with maintenance COPD treatments . Secondly, maintenance therapies for COPD, such as long-acting bronchodilators and inhaled steroids, which could affect frequency of exacerbations, were not standardized across all patients in this study. However, the frequency of such treatments did not differ between the treatment arms and benefit with moxifloxacin was seen in subgroups receiving such concomitant treatment. Finally, it is possible that some exacerbations were unreported due to the absence of daily monitoring [4, 28].
The use of inhaled steroid and long-acting bronchodilators, both anticholinergic and beta-agonists, does result in a reduction in exacerbations. However, in a recent trial, when inhaled salmeterol/fluticasone was added to tiotropium, there was no additional reduction in frequency of exacerbations . Current optimal therapy for COPD often does not result in an exacerbation-free patient. Increased incidence of pneumonia with inhaled steroids is an adverse effect when these drugs are used to reduce exacerbations [22, 30, 31]. Alternative approaches to reducing exacerbations are therefore required in some patients with COPD.