Inhaled amikacin for pneumonia treatment and dissemination prevention: an experimental model of severe monolateral Pseudomonas aeruginosa pneumonia

Current guidelines for the management of hospital-acquired pneumonia (HAP) and ventilator-associated pneumonia (VAP) stress the need for preventive strategies and accurate etiologic diagnosis [1, 2]. Unfortunately, routinely employed antibiotic agents are often ineffective against gram-negative pathogens, with multidrug resistance (MDR) a common problem, even in combination therapy [3]. The treatment of severe gram-negative pneumonia therefore remains a major challenge [3, 4].

Pseudomonas aeruginosa pneumonia is usually treated with a combination of intravenous antibiotics to ensure the adequate treatment of MDR isolates [1, 2]. However, insufficient lung distribution and adverse side effects limit this approach [5], leading to inhaled antibiotics gaining increasing attention as a site-specific treatment [6]. Inhaled has have improved bactericidal properties by ensuring high concentrations in tracheal secretions and epithelial lining fluid (ELF) [7, 8], while lowering systemic concentrations to curtail its nephrotoxic and ototoxic effects [9, 10]. Experimental and clinical studies have elicited conflicting results when assessing inhaled AMK in VAP [7, 11‐19]. Unfortunately, the latest two randomized clinical trials of inhaled AMK have failed to demonstrate any benefit in primary outcomes (i.e., change in clinical pulmonary infection score and survival at days 28–32, respectively) for mechanically ventilated patients with pneumonia [14, 19]. The most recent management guidelines for HAP/VAP advise against the routine use of adjunctive inhaled therapy, recommending it only for susceptible MDR bacteria [2, 20]. Given that study design, dosing, and nebulization technique may have obscured the impact of inhaled AMK on VAP dissemination and the emergence of resistance [21, 22] further research is warranted.

This study describes the development of a porcine model of monolateral severe P. aeruginosa pneumonia resistant to AMK and susceptible to MEM, which we used to explore the inhaled amikacin therapy. Our findings demonstrate that MEM treatment drove a reduction in lung tissue concentration of P. aeruginosa, and that adding inhaled AMK as adjunctive therapy only reduced the bacterial burden of tracheal secretions. Unfortunately, we found no effect of the inhaled therapy on preventing dissemination compared with systemic monotherapy, although histological analysis revealed significantly fewer signs of pneumonia than in the CONTROL group.

Our results are consistent with the latest data from randomized controlled trials [14, 19]. Indeed, in the IASIS trial [14] found only a marginal effect with inhaled adjunctive amikacin/fosfomycin and only in tracheal aspirate samples, of which significantly fewer were positive on days 3 and 7 compared to placebo. In the INHALE trial [19], which used the same dosage of AMK (i.e., 400 mg every 12 h), did show more frequent eradication among patients infected with P. aeruginosa and treated with inhaled AMK; however, this did not translate to improved survival. The reason why inhaled antibiotics do not prove benefits may be related to the susceptibility of pathogens [25]. All enrolled patients in both trials were infected by pathogens susceptible to the intravenous antibiotics, as P. aeruginosa was susceptible to meropenem in our model. Furthermore, P. aeruginosa was resistant to AMK. Therefore, inhaled adjunctive therapy, even if effective, was unlikely to have a measurable effect. In a recent meta-analysis of six randomized controlled trials, inhaled adjunctive therapy achieved higher clinical resolution (odds ratio, 1.96; 95% CI 1.30–2.96) in patients with pneumonia due to MDR pathogens, albeit not in those with susceptible bacteria [26].

Even with 72 h of IV MEM therapy, inhaled AMK suppressed the emergence of the MEM-resistant subpopulation compared with systemic MEM therapy alone. The most recent clinical studies indicated that inhaled treatment may hinder the development of resistance [27]. In our study, only one animal in the AMK group developed MEM resistance, and of note, this was in the only animal in which P. aeruginosa colonized the left non-infected lung, suggesting a role for acquired resistance.

As expected, tracheal secretions revealed high AMK concentrations and MEM concentrations below the limit of detection (0.10 mg/L). AMK was not detected in plasma though out the experiment, suggesting poor AMK translocation from the lungs into the bloodstream, reinforcing the idea that using such drugs can prevent systemic toxicity [18]. In contrast, high AMK concentrations in tracheal secretions correlate with rapidly sterilized bronchial secretions. These results should not be neglected as they suggest a favorable prophylactic effect on the progression from ventilator-associated tracheobronchitis (VAT). Indeed, Palmer et al. showed a faster resolution of signs of infection when assessing the effects of adjunctive nebulized antibiotic therapy in patients with VAT [28].

AMK concentrations in the ELF were significantly higher than MEM concentrations, but we still below the MIC. The large difference between the MICs of MEM and AMK (i.e., 0.5 vs 256 mg/L, respectively) means that although the AMK concentration in the ELF reached higher figures, the maximum concentration (at least 10 times the MIC of the infecting pathogen) was not achieved. By contrast, the MEM free fraction concentration remained above the MIC achieved 100% of the time in the ELF.

We also measured the AMK and MEM levels in both the infected and non-infected lungs, and found similar concentrations in the ELF bilaterally. However, when we measured the AMK concentrations in tissue samples, we found a non-significant trend of higher antimicrobial concentration in the more preserved areas. This indicated that inhaled AMK did not efficiently reach poorly aerated lung parenchyma. The deposition of inhaled drugs in the lungs and in areas of pneumonia with loss of aeration is often questionable and may constitute a major limitation when using this approach to treat ventilated patients [25]. Indeed, previous studies have shown that inhaled antibiotics may not reach consolidated lung segments [12, 15], with research by Elman et al. in infected piglets revealing that the pulmonary concentration of inhaled AMK was reduced in cases with extensive parenchymal infection [12]. In our setting, probably the infected lung has different lung characteristics with a lower compliance and higher airway resistance reducing drug distribution [29].

Notably, we did not change the ventilator settings during nebulization, including the humidification and ventilator circuit, according to the manufacturer’s instructions. The European Investigators Network for Nebulized Antibiotics in Ventilator-associated Pneumonia agreed on specific recommendations for ventilator settings [30], but we did not follow all of these in our protocol. Specifically, the recommendations specify that the mesh nebulizer should be placed 10–15 cm before the Y-piece on the inspiratory limb, whereas we placed it only a few centimeters from the Y-piece [31]. Also, turning off active humidification is recommended during nebulization to avoid hygroscopic growth and a rainout effect in the circuits and the endotracheal tube [32]. Although the penetration of AMK may be modified by using our ventilator configuration, the impact of this was not investigated.

As for how the study treatments affected other clinical, pulmonary mechanics, and hemodynamics variables, we found that even the most efficacious treatment had minimal impact. Only oxygenation was improved in treated animals, but without any additional benefit associated with the use of inhaled AMK, though it did have a slight effect on the systemic IL-1β downregulation.

Finally, no adverse effects were reported. Creatinine levels were higher in both the MEM and MEM + AMK groups than in the CONTROL group, but without significant differences between the treatment groups, or indeed, evidence of nephrotoxicity [17, 18].

Our findings help to delineate where inhaled antibiotics might have utility in the management of VAP or ventilated HAP, whereas they may be limited to patients with difficult-to-treat pathogens. Although their impact on preventing dissemination seems to be trifling, the high efficacy on tracheal secretions may provide a therapeutic opportunity for VAT. Second, as observed over the 100 h of this study, the use of a nebulized antibiotic may impede resistance to IV antibiotics in selected high-risk patients or in intensive care units with high MDR rates [33]. Third, future studies should explore the optimal method for measuring inhaled antibiotic concentrations to ensure adequate dosage regimens to reach distal portions of highly infected pulmonary regions [34].

This study presents some limitations that deserve further discussion. First, we used a 72-h course of therapy, which is unlikely in the most probable clinical scenario. Second, the susceptibility of P. aeruginosa to MEM makes it difficult to show the window of efficacy for inhaled AMK. Furthermore, only one strain was tested. Third, pharmacokinetic analyses were performed for only 4 h after dose administration, limiting the picture of antimicrobial distribution to a brief period. Moreover, the differences between ELF and tissue samples may suggest that the former are not the best surrogate of pulmonary concentrations, particularly for inhaled drugs with a high risk of bronchial tree contamination and where their deposition is important [34, 35]. The AMK dose may also be questioned because it is a concentration-dependent antibiotic, and twice-daily administration similar to the IASIS trial may have produced variable ELF concentrations [14, 19]. Finally, we studied young animals with no comorbidities under deep sedation. Moreover, anatomical differences of the tracheobronchial tree of piglets from human anatomy are critical factors that may affect lung deposition of inhaled particles [36].