Targeted drug delivery and drug resistant pathogens

KEYWORDS:

• Pneumonia
• amikacin
• resistance
• targeted
• mortality

Drug-resistant bacteria are a common problem in respiratory tract infections including ventilated critically ill patients. Concentrated antimicrobial therapy is inadequate in many patients and impacts mortality [1]. Targeted drug delivery in terms of inhaled antibiotics offer high pulmonary drug concentrations with minimal systemic toxicities and may be a potential therapy for ventilator-associated pneumonia (VAP) with multidrug-resistant (MDR) pathogens.

Antimicrobial resistance has complicated the use of antimicrobial drugs for treatment of infection – mild, moderate, and severe. The impact is likely greatest in those patients with moderate-to-severe infections and where treatment impacts clinical deterioration and death. In patients with mild – often self-limiting infections – the impact of ineffective antimicrobials more likely impact morbidity as the likelihood of mortality is low.

Several variables exist with the use of antimicrobial agents including spectrum of activity, drug distribution and elimination, achievable and sustainable drug concentration in various body compartments, toxicity, hypersensitivity/side effects, drug–drug interactions, bacteriostatic versus bactericidal and the likelihood for resistance selection. Clinical utility is also important and based on randomized clinical trials demonstrating non-inferiority to a comparator drug (or drugs) for specific clinical indications. Antimicrobial agents with broad spectrum activity and wide distribution in the body often possess the widest clinical utility (many clinical indications) but may also possess the greatest risks for unintended consequences including – but not always – toxicities and a potential negative impact on the body’s normal microbiota and the potential for resistance selection amongst nontargeted organisms. Duration of therapy is clearly important with shorter duration of therapy having a lesser negative impact than does longer durations. Despite the above, antimicrobial agents have cured infectious diseases in innumerable patients worldwide and continue to do so. Unfortunately, escalating trends in antimicrobial resistance and further complicated by MDR pathogens compromise use.

One potential solution to the above is the use of drugs that specifically accumulate in a body compartment with minimal or no accumulation elsewhere. Some examples of such drugs already exist: norfloxacin – a fluoroquinolone – concentrates in urine and was used primarily for treatment urinary tract infections (UTIs), however, the drug also impacted fecal flora. Similarly, nitrofurantoin is utilized for UTIs. Macrolides/azalides tend to accumulate at higher concentrations in pulmonary compartments and alveolar macrophages but serum concentrations approximate the mutant selection window (MSW) for some pathogens such that during bacteremia, the risk for resistance selection increases [2]. The MSW is bordered by the minimum inhibitory concentration (MIC) and mutant prevention concentration (MPC). The MSW has been argued to be the drug concentration range where selective amplification of bacterial cells, not inhibited by the MIC, occurs [3]. Dosing to exceed the MPC has been argued to prevent growth of organisms with reduced susceptibility to drug. Additionally, accumulation of macrolides/azalides in pulmonary compartments is not equal and agent specific differences require specific attention to appreciate the differences [4]. Just because two agents may give similar clinical outcomes does not mean the drugs are identical in terms of their specific characteristics. Appreciating the differences between individual drugs and drug classes may allow optimization of therapy and be consistent with antimicrobial stewardship principals.

While slightly different, topical drug applications [5] are targeted therapy in that higher drug concentrations are delivered to a specific anatomical location without systemic distribution. Topical therapy has been used to treat chronic wounds [6], used in ophthalmology [7], and used to treat uncomplicated vaginal candidiasis [8,9]. Other topical applications have also been described and at present seem more popular in primary care settings.

VAP is a life-threatening complication of mechanical ventilation and is associated with high mortality [1,10]. Pathogens associated with VAP include Pseudomonas aeruginosa, E. coli, Klebsiella pneumoniae, Acinetobacter spp., Staphylococcus aureus, Enterococcus faecium, and Enterobacter spp. – the ESKAPE group of organisms – and collectively, these are responsible for 80% of all cases [11].

Currently, antimicrobial resistance is problematic among ESKAPE organisms and MDR phenotypes have been repeatedly described. For example, Slavcovici et al. reported that ceftazidime, ciprofloxacin, and gentamycin co-resistance was found at 48% [12]. Highly resistant pathogens such as Acinetobacter baumanni and P. aeruginosa that were carbapenem resistant and extended-spectrum beta-lactamase (ESBL)-producing Klebsiella spp. were identified or more commonly seen in VAP and hospital-acquired pneumonia. For such MDR pathogens, therapeutic options are few and may be limited to tigecycline or colistin. Tigecycline does not cover all of the ESKAPE organisms and colistin is associated with renal and neurotoxicity [13,14]. Furthermore, tigecycline has been associated with increased mortality when used as monotherapy [15,16] and the description of transmissible colistin resistance from human and veterinary medicine and in food [17–19] has caused considerable concern with this often referred to ‘last resort’ antimicrobial. At present, novel drug development is unlikely to bring forth a new agent to address these concerns so where do we go from here?

One promising approach for treatment of VAP are targeted drug delivery investigations of which inhaled amikacin is one. Amikacin, of course, is an older and well-known antimicrobial and shares many of the properties of the aminoglycoside class of drugs including renal and otic toxicity when administered systemically [20]. For intubated and mechanically ventilated patients, it is a viscous solution that is nebulized and the particle diameter of 2–5 µm is essential for delivering to the deep lung. The benefits of such a delivery system are high pulmonary drug concentrations with minimal systemic distribution [1]. Typical systemic toxicities associated with aminoglycosides are absent with the formulation and the high pulmonary drug concentrations exceed both MIC and MPC values for ESKAPE pathogens including those testing resistant by established breakpoints.

Recently, Wunderink (2016) and Kollef (2017) debated the routine use of inhaled antibiotics for lower respiratory tract infections in the ICU setting with Wunderink arguing ‘Yes’ and Kollef arguing ‘No’ [21,22]. Wunderink identified inappropriate initial therapy and ineffective therapy for MDR pathogens as two main issues of concern for hospital-acquired and VAP. As argued, P. aeruginosa and Acinetobacter spp. are clearly MDR pathogens as are Enterobacteriaceae (E. coli, Klebsiella spp., Enterobacter spp.) producing ESBLs or Gram-negative bacilli producing carbapenemase enzymes or even carbapenem-resistant bacteria. Combination antimicrobial therapy for MDR pathogens was argued to be necessary to avoid initial therapy that was inappropriate. Beta-lactams (penicillin, cephalosporins, carbapenem, monobactam) and fluoroquinolone resistance has elevated aminoglycosides to monotherapy in situations where resistance to the other agents used in combination with aminoglycosides occurs. Aminoglycoside monotherapy for HAP/VAP is associated with high failure rates [23]. Wunderink argued that the first 3 days of therapy might be the most critical for controlling infection and early inappropriate therapy may impact outcome. Wunderink further argued that combination antimicrobial therapy, longer duration of therapy, and pharmacokinetic and pharmacodynamic optimization did not clearly lead to better outcomes. Wunderink indicated that in some HAP/VAP cases, 10⁸–10¹⁰ bacteria/ml may be present in alveolar spaces and mutant bacterial cells may be present in such high density populations. Antimicrobial therapy insufficient to block mutant growth may allow for the selective amplification of the mutant population during therapy while the susceptible cells are eliminated by the drug therapy. Such thinking has been investigated by the numerous publications on MPC and the MSW [3,24].

According to Kollef, the arguments for aerosolized antibiotics was argued to be the high lung drug concentrations and the likely substantial bacterial killing – even against organisms testing resistant by standardized susceptibility testing – but with MICs below the achievable and sustainable pulmonary drug concentration. Limited clinical experience may lead to hesitancy in adopting aerosolized antibiotic use. He argued that bacterial lower respiratory tract infections are relatively common in patients requiring mechanical ventilation and that antimicrobial agents have been used both prophylactically and therapeutically for both VAP and ventilator-associated tracheal bronchitis. He also argued that pulmonary concentrations of antibiotics are an important factor in influencing both clinical outcomes and the emergence of resistance. He also commented that in vitro studies have shown that antibiotic concentrations below the MPC but above the MIC drug concentration have a greater association with the emergence of antibiotic resistance. Variable drug concentrations in the lung are deemed to be important in this regard. He also acknowledged that the rationale for the development of aerosolized antibiotics have been borne out of MDR pathogens causing respiratory tract infections and the inadequacy of pulmonary concentrations of systemically administered drugs. As well, vibrating mesh nebulizers impact on drug particle size and location in the lung to which they may be distributed [1]. Vibrating mesh nebulizers appear to result in more consistent particle sizes. An important point was made regarding MDR bacteria with elevated MICs to carbapenems in the range of 256 µg/ml or greater. With such organisms, lung drug concentrations may need to be in the 5000–6000 or greater µg/ml range not only to deal with the bacterial pathogens but also host variables such as sputum antagonism. Kollef also argued that there was a shortage of evidence regarding the routine use of aerosolized antibiotics for the treatment of VAP and VAT. Kollef further argued that the use of aerosolized antibiotics has been expanding worldwide and raised concerns about the potential for increasing antimicrobial resistance. One example given related to the expanding use of colistin for the treatment of MDR bacteria and now the recognized the emergence of plasmid-mediated colistin resistance [19]. Further expansion of colistin resistance may give rise and spread of extremely drug-resistant bacteria. Furthermore, several questions regarding aerosolized antibiotics and their potential use for bacterial lower respiratory tract infections were summarized by Kollef and some of these questions related to which types of infections would be treated with aerosolized antibiotics, which antibiotic and dose, the potential for combination therapy, the optimal delivery device, duration of therapy, and which outcomes should be measured.

Both Wunderink and Kollef agree on the need for additional data and evidence related to aerosolized antimicrobial agents. Clearly, clinical decisions regarding where and when to use such antimicrobial agents will be formed from clinical trial data as well as input from PK/PD studies and various microbiological measurements. From work in our own laboratory investigating amikacin and MIC, MPC, and kill experiments, we looked at both inhibitory drug concentrations (MIC, MPC) and pulmonary drug concentrations resulting in rapid bactericidal activity. For strains of ESBL positive Klebsiella pneumoniae, amikacin MIC values ranged from 0.25 to 4 µg/ml and corresponding MPC values ranged from 8 to 64 µg/ml. When kill experiments were attempted using the MIC drug concentration, insignificant or no bacterial killing was observed; however, in kill experiments that were conducted with the measured MPC drug concentration, 22–64% of bacterial cells were killed following the first 60 min of drug exposure and this increased to >99% kill following 120 min of drug exposure. When these organisms were exposed to the epithelial lung fluid drug concentration of 976 µg/ml, 96% of viable cells were killed within the first 5 min after drug exposure and this increased to >99% following 10 min of drug exposure. Very similar observations were also seen with ESBL positive strains of E. coli where 49–93% of bacterial cells were killed between 10 and 30 min of drug exposure at the MPC drug concentration and this increased to >99% kill following 60 min of drug exposure. At the maximum tissue drug concentrations 93% of bacterial cells were killed within the first 5 min of drug exposure in this increased to >99% following 10 min of drug exposure. These data clearly suggest the rapid killing of MDR Gram-negative organisms at clinically achievable drug concentrations. In vitro measurements evaluating MPC and killing of MDR P. aeruginosa strains, Acinetobacter species strains, and methicillin-susceptible and resistant S. aureus strains with clinically achievable drug concentrations are ongoing.

Our laboratory is also investigating ciprofloxacin dry powder inhaled and clinically achievable drug concentrations and doing similar experiments to those highlighted above for amikacin. With ciprofloxacin, low concentration of drug has been reported to be present systemically. Rapid bactericidal activity was seen against strains of Enterobacter cloacae, K. pneumoniae, and E. coli tested over varying organism densities ranging from 10⁶ to 10⁹ cfu/ml when exposed to reported pulmonary drug concentrations [25]. In order to address if such low concentrations may have a negative impact on resistance selection, we conducted some in vitro investigations where we exposed strains of Gram-negative bacilli to low drug concentrations of ciprofloxacin and then measured to detect increasing antimicrobial resistance. For those experiments, MIC measurements were performed on strains of E. coli, K. pneumoniae, and P. aeruginosa prior to and after exposure to 0.01–0.1 µg/ml of ciprofloxacin. In all instances, MICs were the same or within one doubling dilution for the bacterial strains exposed to the low ciprofloxacin drug concentrations.

In those experiments, exposure to low ciprofloxacin drug concentrations did not appear to select for antimicrobial resistance and this observation is consistent with the MSW and the argument that drug concentrations that fall below the MIC (the lower boundary of the MSW) would be unlikely to selected for resistance if the drug concentration is sufficiently low and not be inhibiting any cells within the population [26].

Clinical trial data will dictate where and when new antimicrobials will be used – clinical indications, disease severity, comorbidities, etc. PK/PD and microbiological measurements contribute substantial data and may often explain clinical observations. Rapid bacterial killing, including MDR strains, will clearly be important for new agents (or older agents with new indications) with indicators for use in critically ill patients. As MIC or MPC values increase, higher drug concentrations will be necessary but this may lead to higher drug related toxicities. Targeted drug delivery with high compartment drug concentrations appear necessary in today’s antimicrobial-resistant environment.

Declaration of interest

The author has no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties. Peer reviewers on this manuscript have no relevant financial or other relationships to disclose.

Additional information

Funding

This manuscript has not been funded.