Principles of Antibiotic Use and the Selection of Empiric Therapy for Pneumonia
Antibiotics are the foundation of therapy for respiratory tract infections. This approach varies with the type of pneumonia, age of the affected patient, presence of various comorbid illnesses and risk factors for infection by specific pathogens, and the severity of the acute illness. For most of the patients, initial therapy is empiric, aimed at a broad spectrum of potential pathogens (see Chapter 122). Once culture data become available, therapy can be pathogen-specific making it possible to de-escalate to fewer drugs with a narrower antimicrobial spectrum.1,2 In some cases, initial empiric therapy must be continued because no etiologic pathogen is identified (see Chapter 123).
When a pathogen is defined, the term “appropriate” refers to the use of at least one antimicrobial agent that is active in vitro against the etiologic pathogen.1 The term “adequate” includes not only appropriate therapy, but also the use of that agent in the correct dose, via the right route, given in a timely fashion, and with penetration to the site of infection. Timely and appropriate antibiotic therapy can improve survival in patients with community-acquired pneumonia (CAP) and nosocomial pneumonia or hospital-acquired pneumonia (HAP) and the benefits are most evident in patients who are not otherwise terminally ill.1,3–5 In general, with severe illness, those receiving antibiotics earlier in their course have a lower mortality than those receiving delayed therapy, with the risk of death rising 7% to 8% for each hour of delay of therapy during the first 6 hours of septic shock and hypotension.6
The term HAP encompasses pneumonia in nonventilated patients, ventilated patients, and those with healthcare-associated pneumonia (HCAP) (see Chapter 129). Ventilator-associated pneumonia (VAP) is nosocomial pneumonia that develops after at least 48 hours of preceding mechanical ventilation, and thus some ventilated HAP patients may not have VAP, since they do not satisfy this definition of prolonged preceding mechanical ventilation. HCAP includes those coming from nursing homes, those in the hospital for more than 2 days in the past 90 days, those with a history of regular visits to places like dialysis or infusion centers, or those getting home wound care. Because of their exposure to the healthcare environment, some patients with HCAP are at risk for infection with multidrug-resistant (MDR) pathogens, although not all of these patients are at the same risk. Presence of this entity has blurred the distinction between CAP and HAP, since HCAP patients may reside in the “community,” but be infected with organisms very similar to those present in patients with HAP.
In the setting of CAP, effective initial antibiotic therapy is associated with a marked improvement in survival, compared to ineffective therapy, particularly in patients with severe illness.3,7 In several studies, identification of the pathogens causing severe CAP did not lead to an improved survival rate, while the use of a broad-spectrum, empiric regimen directed at likely pathogens reduced mortality (see Chapter 128). Further, patients with CAP have reduced mortality if initial antibiotic therapy is provided within 4 hours of arrival to the hospital. In the treatment of VAP, data show that appropriate therapy should be given as soon as the infection is clinically identified and lower respiratory tract samples have been collected for culture. Delay of even 24 hours in starting therapy is an important mortality risk factor in VAP.1,5
Even with the use of correct agents, not all patients recover. The fact that some HAP patients die in spite of microbiologically appropriate therapy is a reflection of the degree of antibiotic efficacy, as well as a reflection of host response capability (which may in part have a genetic determination). In some HAP patients, death is the result of the serious underlying illness; the percentage of deaths that occur because of infection, termed “attributable mortality” of HAP, has been estimated to be as high as 50% to 60%.8 However, the timely use of appropriate antimicrobial therapy can reduce this attributable mortality to as low as 20%, and recent studies suggest that excess mortality in VAP may be even lower.8,9
In recent years, a number of guidelines for empiric therapy for both CAP and HAP have been developed, but several caveats should be remembered. First, although current guidelines are evidence-based, outcome studies are required to demonstrate their utility in clinical practice. Second, guidelines must be re-evaluated relative to local patterns of antibiotic susceptibility. In the case of CAP, the emergence of penicillin-resistant pneumococcus, community-acquired methicillin-resistant Staphylococcus aureus (MRSA), and epidemic viral illness (influenza, SARS) may affect the selection of initial therapy. In the setting of HAP, each hospital has a unique flora and antimicrobial susceptibility patterns; knowledge of such patterns is essential for selection of the optimal agents.
The majority of data establishing “standard” antibiotic doses are derived from studies in healthy volunteers or ward patients, which have been extrapolated into a wide range of patients, including the elderly, the obese, and the critically ill, with differing clearances of these agents. Critically ill patients may manifest significantly altered end-organ function whereas those with sepsis may have a hyperdynamic circulation and augmented renal clearance (ARC), necessitating higher than normal doses of antibiotics to achieve effective serum and intrapulmonary concentrations. In these patients, as well as in obese individuals who have an altered volume of distribution (Vd) of some drugs, subtherapeutic concentrations may result in treatment failure or in selection of antimicrobial-resistant organisms.10
In this chapter, the principles underlying antibiotic use are examined, followed by a discussion of the commonly used antibiotics for respiratory tract infections, and the principles of empiric therapy of both CAP and HAP. Further, pharmacokinetics and pharmacodynamics (PK/PD) considerations are a theoretical way to optimize the delivery of antibiotics; recent data indicate that the use of such “optimized therapy” may have clinical benefits.11 To preserve antibiotics for future use, it is necessary to understand mechanisms of antibiotic resistance and to practice antimicrobial stewardship by focusing on proper antibiotic dosing, de-escalation therapy, the development of local protocols for antibiotic choice, but not necessarily by restricting access to potentially effective broad-spectrum agents for seriously ill patients.
PRINCIPLES OF ANTIBIOTIC USE
Antibiotics interfere with the growth of bacteria using various mechanisms including undermining the integrity of their cell wall, interfering with bacterial protein synthesis, or common metabolic pathways.12 The mechanism of activity can be used to define whether an agent is bacteriostatic or bactericidal, terms discussed in the section “Mechanism of Action.”
MECHANISM OF ACTION
Bactericidal antibiotics kill bacteria, generally by inhibiting cell-wall synthesis or by interrupting a key metabolic function of the organism. Agents of this type include the penicillins, cephalosporins, aminoglycosides, fluoroquinolones, vancomycin, daptomycin, colistin/polymyxin, rifampin, and metronidazole. Bacteriostatic agents are antibiotics that inhibit bacterial growth but generally do not interfere with cell-wall synthesis, and rely on host defenses to eliminate microbes. Agents of this type include the macrolides, tetracyclines, sulfa drugs, chloramphenicol, linezolid, and clindamycin.
The terms bactericidal and bacteriostatic are broad categorizations, and may not apply for a given agent relative to all organisms, with certain antimicrobials being bactericidal for one bacterial pathogen but bacteriostatic to another.15 The use of a specific agent is dictated by the susceptibility of the causative organism(s) in a given anatomic location to individual antibiotics. When neutropenia or immune compromise are present or if there is accompanying endocarditis or meningitis the use of bactericidal agents is preferred (see Chapter 123). Thus, for most patients with pneumonia, it is not essential to choose a bactericidal agent. One additional consideration is that certain organisms such as S. aureus can produce toxins, and the optimal agent must be able to kill the bacteria and also to inhibit the production of disease-mediating toxins. Agents that inhibit bacterial toxin production include clindamycin, rifampin, and linezolid.
The Minimum Inhibitory Concentration (MIC) is the minimum antibiotic concentration necessary to inhibit the growth of 90% of a standard sized inoculum, leading to no visible growth in a broth culture. The MIC is used to define the sensitivity of a pathogen to a specific antibiotic, under the assumption that the concentration required for killing (the MIC) can be reached in the serum in vivo. This term must be interpreted cautiously in the treatment of pneumonia, because the clinician must consider the MIC data in light of the penetration of an agent into lung tissues, with some agents achieving higher than serum levels at respiratory sites of infection and others reaching lower levels.13,14
The Minimum Bactericidal Concentration (MBC) is the minimum concentration needed to cause a 3-logarithmic decrease (99.9% killing) in the size of the standard inoculum, and generally all pathogenic bacteria are killed at this concentration.
The Mutation Prevention Concentration (MPC) is the lowest concentration of an antimicrobial agent that prevents bacterial colony formation from a culture containing greater than 1010 bacteria.13 At lower than MPC concentrations, spontaneous mutants can persist and be enriched among the organisms that remain during therapy. The concept has been most carefully studied with the pneumococcus and fluoroquinolones. In general the MPC is higher than the MIC, implying that it is possible to use an antimicrobial to successfully treat a clinical infection but with inadequate levels to eradicate the bacteria. These remaining organisms may provide a basis for the emergence of antimicrobial-resistant bacteria.
IMPORTANT PHARMACOKINETIC–PHARMACODYNAMIC VARIABLES
Pharmacokinetics refers to the absorption, distribution, and elimination of a drug in the body, which can be used to describe the concentration of a drug in the serum. Pharmacodynamics refers to the physiological and biochemical effects of a drug on the body or any organisms within the body, the mechanisms of action and the relationship between a drug’s concentration and effect. Bacterial killing is affected by the way in which an antibiotic reaches the site of infection considering the frequency of administration and dose administered, thus defining a close relationship between PK/PD. There are three types of bactericidal antibiotics.13
First, are those that kill in relation to the duration in which the nonprotein-bound free concentration of the agent stays above the MIC of the target organism, f T >MIC (Time-dependent killing).16 Examples include the β-lactams, carbapenems, aztreonam, macrolides, and linezolid. For maximal efficacy, β-lactam concentrations should be maintained at four times the MIC, for approximately 40% to 70% of the dosing interval, depending on the agent being employed.14 Second, are agents that kill in relation to how high the peak concentration is, relative to the MIC, defined by the Cmax:MIC ratio (Concentration-dependent killing).17 Examples include the aminoglycosides and fluoroquinolones, as well as the ketolides and daptomycin. Finally, agents like the glycopeptides and fluoroquinolones also have efficacy defined by the area under the concentration– time curve divided by the MIC (AUC: MIC ratio, or AUIC) to define the optimal PK/PD index.18,19
To optimize time-dependent killing, dosing should be chosen to achieve the maximal time above the MIC of the target organism; this can be done with continuous or prolonged (over 3–4 hours) infusions. The rate of killing is saturated once the antibiotic concentration exceeds four times the MIC of the target organism. In spite of these considerations, for many organisms, the concentration of the antibiotic only needs to be above the MIC for 40% to 50% of the dosing interval, and possibly for as little as 20% to 30% of the interval as for the carbapenems.13 For the time-dependent killing drugs listed above, the pharmacodynamic parameter that best predicts clinical efficacy is the time above the MIC.
The target AUIC for gram-negative bacteria is 125 or greater, whereas for most antibiotics that treat the pneumococcus, the target value is at least 30. Some studies have shown that aiming for Cmax/MIC target of 12 for fluoroquinolones against pneumococcus is optimal. Appropriate use of these agents would entail infrequent administration but with high doses, which is the underlying principle behind the once-daily administration of aminoglycosides.
The Post-Antibiotic Effect (PAE) maintains the efficacy of the antibiotic after the serum (or lung) concentrations have fallen below the MIC of the target organism. Thus, with once-daily aminoglycoside dosing regimens, the patient achieves a high peak concentration (maximal killing), and a low trough concentration (minimal nephrotoxicity) with the PAE achieving continued antimicrobial effects. While most agents exhibit a PAE against gram-positive organisms, prolonged PAEs against gram-negative bacilli are achieved by the aminoglycosides and fluoroquinolones.13 For pneumococcus, a PAE exists for the macrolides/azalides, clindamycin, vancomycin, quinupristin/dalfopristin, tetracyclines, and the oxazolidinones (such as linezolid). Most of the agents that kill in a concentration-dependent fashion have a prolonged PAE. Agents with little or no PAE against gram-negatives are generally also agents that kill in a time-dependent fashion and are administered several times daily. The β-lactams (including the penicillins, cephalosporins, and monobactams) generally have little or no PAE against gram-negatives; one notable exception is imipenem, which has a modest PAE against Pseudomonas aeruginosa. In clinical practice, the use of once-daily aminoglycoside dosing has had variable benefits in both efficacy and toxicity, but the advent of this type of dosing regimen follows from an understanding of pharmacodynamic principles. Postantibiotic Leukocyte Enhancement (PALE) refers to the ability of functioning leukocytes to kill organisms in the postantibiotic phase of growth. Thus, when the patient has functioning neutrophils, the PAE of some agents is extended by their PALE.
Inflammatory effects. Recently, some investigators have suggested that antibiotics should be selected based on whether they stimulate inflammation and cytokine production in response to the presence of the bacterial cell-wall lysis products.15 Certain antibiotics liberate bacterial cell-wall products that interact with cytokine-producing cells, stimulating the production of high levels of cytokines such as tumor necrosis factor. In theory, this could lead to the development or the worsening of sepsis syndrome. This may be a consideration in patients with Pneumocystis jiroveci pneumonia or pneumococcal meningitis who may benefit from the use of corticosteroids with antimicrobial therapy in these infections. In general, the clinical role of cytokine release is uncertain, but bactericidal antibiotics tend to produce a greater host inflammatory response than bacteriostatic agents. Antibiotics that are cell-wall active, and that kill slowly, have been associated with the greatest cytokine release.
As was noted previously, the use of antibiotics that inhibit protein synthesis (linezolid, clindamycin) may have an advantage in toxin-mediated illnesses, such as those caused by certain strains of S. aureus, when compared with cell-wall active bactericidal antibiotics.20
FACTORS AFFECTING ANTIBIOTIC CONCENTRATIONS
Alterations in lung penetration, protein binding, and changes in drug Vd and clearance (Cl), can significantly influence antibiotic concentrations achieved at a site of infection and, therefore, the probability of microbiological success.10 The pharmacokinetic profile of piperacillin was explored in a small cohort of septic and critically ill patients, demonstrating a wide distribution of values (up to 14-fold variation) for drug clearance, Vd, and half-life21 despite “normal” biochemical renal indices in all patients. Similar data are available for other antibiotics in this setting,22 confirming that in a significant fraction of critically ill patients, drug elimination is likely to be markedly different from healthy volunteers, and thus it is challenging to define optimal dosing in these patients.
Beyond the immediate threat of treatment failure, subtherapeutic antibiotic concentrations are a risk factor for the selection of drug-resistant organisms.23 Tigecycline resistance was identified within months of the commercial release of this antibiotic, and this may have been related to suboptimal concentrations at the site of infection.24 In this respect, accurately identifying patients at risk of suboptimal drug exposure for a given causative pathogen remains a difficult clinical problem. While therapeutic drug monitoring (TDM) provides the most robust means of identifying subtherapeutic concentrations, its availability for all classes of antibiotic is largely limited to specialized centers.25–27 In addition, changing organ function during the course of critical illness will affect antibiotic PK parameters, necessitating a regular review of dosing requirements.
PENETRATION INTO THE LUNG
A key requirement for the successful treatment of pneumonia is the achievement of therapeutic drug concentrations at the site of infection; in pneumonia, the targets are the lung parenchyma and alveoli (Table 125-1).28 The drug concentration within the epithelial lining fluid (ELF) has been the most commonly used surrogate in pneumonia therapy. Research over a number of years has established that there is wide variability both within and between antibiotic classes in terms of ELF penetration.29 The physiochemical properties that govern drug movement across biological membranes include the degree of ionization, fat solubility, molecular weight, and protein binding. Despite accounting for such characteristics, many of the differences between antibiotics in ELF to plasma concentration ratios (CELF:CPlasma) remain poorly understood.29 Sputum and bronchial concentrations may be the most relevant measurements for bronchial infections, whereas the concentrations in lung parenchyma, ELF, and cells such as macrophages and neutrophils may be more important for pneumonia. The localization of the pathogen may also be important; intracellular organisms such as Legionella pneumophila and Chlamydophila pneumoniae may be best eradicated by agents that achieve high intracellular concentrations in macrophages.
Macrolides have good penetration into ELF (e.g., CELF:CPlasma >10 for azithromycin),29 whereas β-lactams have varying penetration ratios,30 without a predictable pattern. Vancomycin demonstrates limited lung penetration (CELF:CPlasma <1),31 while both linezolid and fluoroquinolones are reported to have ELF concentrations equivalent to or greater than those in plasma (CELF:CPlasma ≥1).32–34 Specific data from critically ill subgroups are generally limited, and in those where results have been reported, significant interpatient variability has been noted.35
The local concentration of an antibiotic must be considered in light of the activity of the agent at the site of infection. For example, antibiotics can be inactivated by certain local conditions. Aminoglycosides have reduced activity in acidic pH, which may be present in infected lung tissues. In addition, some bacteria develop resistance by producing destructive enzymes (such as β-lactamases), by altering the permeability of the outer cell wall, by changing the target site of antimicrobial action, or by pumping (efflux) of the antimicrobial from the interior of the cell. In these conditions a high local concentration of antimicrobial may help to offset bacterial resistance mechanisms.
The concentration of an antibiotic in lung parenchyma depends on the penetration through the walls of the bronchial capillaries, which have a fenestrated endothelium. Antibiotics penetrate in proportion to their molecular size and protein binding with small molecules that are not highly protein bound passing readily into the lung parenchyma. When inflammation is present, penetration is increased. Antibiotics reaching the ELF must pass through the pulmonary vascular bed with a nonfenestrated endothelium. This presents an advantage for lipophilic agents generally not inflammation-dependent.
Lipophilic agents include chloramphenicol, the macrolides (including the azalides and ketolides), linezolid, clindamycin, the tetracyclines, the quinolones, and trimethoprim–sulfamethoxazole (TMP-SMX). Agents that are poorly lipid soluble are inflammation-dependent for entry into the ELF; these include the penicillins, cephalosporins, aminoglycosides, vancomycin, carbapenems, and monobactams.
Active transport can facilitate antibiotic entry into lung tissue and phagocytes. Agents that are concentrated in phagocytes in this manner include the macrolides, clindamycin, and the fluoroquinolones. Antibiotics, such as the β-lactams, that are not concentrated in phagocytes by active transport remain in the extracellular space, which constitutes 40% of the weight of bronchial tissue; thus, penicillins achieve only about 40% of the serum level in lung tissue.
Drugs that penetrate well into the sputum or bronchial tissue include the quinolones, the newer macrolides and azalides (azithromycin and clarithromycin), the ketolides (telithromycin), the tetracyclines, clindamycin, and TMP-SMX. On the other hand, the aminoglycosides, vancomycin, and to some extent the β-lactams, penetrate less well into these sites. With the use of once-daily aminoglycoside dosing, high peak serum concentrations can be achieved, but the alveolar lining fluid concentration in patients with pneumonia is only 32% of the serum level over the first 2 hours; the two sites have more similar concentrations later in the dosing interval.36 Since aminoglycosides require high peak concentrations for optimal killing, their poor penetration with systemic administration often impacts efficacy suggesting a potential role for delivery by the aerosol route.
VOLUME OF DISTRIBUTION
The Vd defines the relationship between the drug concentration observed in plasma and the dose administered. Values exceeding 3 L imply a distribution outside the plasma. Hydrophilic drugs that are poorly lipid soluble diffuse freely into interstitial fluid but do not penetrate cells. Thus only the free, nonprotein-bound drug can be distributed outside the plasma. Aminoglycosides and β-lactams are hydrophilic and will distribute into the extracellular space (~0.6 L/Kg).37 Lipophilic agents such as the macrolides and quinolones are extensively distributed in body tissues, and serum levels underestimate the effect at the site of infection. This observation explains the efficacy of azithromycin, which achieves relatively low serum levels with high intracellular concentrations in phagocytes and can treat pneumonia.
Vd is also affected by obesity. Dosing based on ideal body weight may lead to underdosing in general, while for hydrophilic antibiotics, dosing based on total body weight may result in excessive drug levels. In critical illness, the Vd of many agents can differ from that in healthy subjects, primarily as a reflection of changes in microvascular function. Also volume resuscitation, an ubiquitous intervention in critical illness, may increase Vd for hydrophilic agents in the face of increased capillary permeability and interstitial edema.38 Sánchez et al.39 described the hemodynamic changes observed after fluid resuscitation for septic shock, showing that antibiotics such as the β-lactams, aminoglycosides, and glycopeptides “go where the water goes,” meaning that in critical illness, these drugs will be preferentially “dragged” into the interstitial space by extravascular movement of free fluid. The Vd of such drugs is much higher than in the absence of “leaky” capillaries. This effect is associated with decreased serum concentrations for any given dose with thus a larger dose required to compensate for the increased Vd. As patients improve, fluid redistribution from tissues into the vascular space occurs, resulting in a significant decrease in the Vd.40 Thus for drugs that distribute into extravascular interstitial fluid, an initial loading dose41 is required to adequately “fill” these compartments and the Vd will change in parallel with the patients’ physiology. This may explain why higher doses of vancomycin were needed to reach target serum concentrations in ICU patients in whom standard “ward” doses are likely to be inadequate. A larger Vd is also well described for β-lactams.42
PROTEIN BINDING
Many antibiotics demonstrate plasma protein binding, to either albumin or alpha-1-acid glycoprotein (AAG)37 which can be disturbed in critical illness with many patients manifesting hypoalbuminemia.43 The free fraction of the drug is eliminated with low plasma protein concentrations resulting in an increase in Vd and CL for highly protein-bound agents such as flucloxacillin, ertapenem, ceftriaxone, and teicoplanin.43,44 Measuring the free drug concentration can be technically difficult, thus most published research has focused on total serum levels without correction for protein binding based on data from healthy volunteers.
DRUG CLEARANCE
Drug Cl represents the volume of plasma cleared of a specific pharmaceutical per unit time, and can involve organ independent enzymatic degradation, hepatic metabolism, or renal elimination of the parent compound or metabolites. During critical illness many aspects of these elimination pathways are disturbed. There are changes in major organ blood flow, saturation of enzyme systems, or complex drug–drug interactions.37 Few data are available that describe the impact of changes in hepatic function on antibiotic PK–PD.
For many antibiotics (such as the aminoglycosides, β-lactams, and glycopeptides), renal drug clearance is directly proportional to Creatinine clearance (CrCl).37 With the administration of large volumes of fluid and the use of vasoactive agents in critically ill patients with sepsis, trauma, burns, and in neurosurgery patients, there is often an increased cardiac output and organ perfusion with enhanced delivery of solute to drug eliminating organs leading to the phenomenon of ARC. With ARC, serum creatinine concentrations are almost always within the “normal range,” while measured creatinine clearance is elevated, even as high as 200% of normal.44 In such patients the use of an isolated serum creatinine concentration to extrapolate GFR is not possible; dosing derived from noncritically ill patient populations may fail to account for ARC.45 In this setting, a timed urinary CrCl provides a simple, cheap, and repeatable estimate of renal drug elimination,46 and affords an alternate means of identifying such patients. Recent research has confirmed an association between ARC and suboptimal trough β-lactam concentrations in critical illness,47 thereby impacting the likelihood of achieving the necessary fT > MIC targets. From an antibacterial prescribing perspective, ARC will lead to an increased rate of drug elimination, and may necessitate more frequent intermittent doses or administration by continuous or extended infusion.22
A paucity of data to guide dose selection is also present when patients receive extracorporeal renal replacement therapy (RRT). In such instances, native renal drug elimination is limited (although not completely absent), whereas nonrenal mechanisms may assume a greater role in total body clearance. As a consequence, highly variable antibiotic concentrations have been observed in patients receiving RRT.48 The variety of RRT modalities available further complicates such determinations due to variations in the parameters employed (including blood flow rates, effluent flow rates, and membrane pore size) and the unique physiochemical characteristics of each drug.49 Not surprisingly, optimal dosing strategies in RRT remain largely uncertain.50
ANTIMICROBIALS USED IN THE THERAPY OF RESPIRATORY TRACT INFECTIONS
A variety of classes of antibiotics are used in treatment of respiratory tract infections. Each is discussed below.
MACROLIDES (INCLUDING AZALIDES)
Macrolides are bacteriostatic agents that bind to the 50 S ribosomal subunit of the target bacteria and inhibit RNA-dependent protein synthesis. They have good activity against pneumococci as well as “atypical” pathogens (C. pneumoniae, Mycoplasma pneumoniae, Legionella). Older macrolides including erythromycin are not active against Haemophilus influenzae and have poor intestinal tolerance so that prolonged therapy is difficult. The newer agents in this class include azithromycin (also referred to as an azalide) and clarithromycin. These agents have enhanced activity against H. influenzae (including β-lactamase producing strains), although on an MIC basis, azithromycin is more active. Erythromycin is active against Moraxella catarrhalis although the new agents have enhanced activity against this pathogen. Among the new macrolides, azithromycin is more active than clarithromycin against not only H. influenzae and M. catarrhalis, but also M. pneumoniae. On the other hand, clarithromycin is more active than azithromycin against Streptococcus pneumoniae, Legionella, and C. pneumoniae.
Both of the newer agents have better intestinal tolerance than erythromycin and penetrate well into sputum, lung tissue, and phagocytes. Clarithromycin, which has an active 14-hydroxy metabolite that is antibacterial, is administered twice a day orally at a 500-mg dose for 7 to 10 days in the treatment of CAP and acute exacerbations of chronic bronchitis (AECB). A new preparation of extended-release clarithromycin is administered as a 1000-mg dose once daily and has been effective as a 7-day course of therapy for AECB. Azithromycin has a longer half -life than clarithromycin, and concentrates in tissues, achieving very low serum levels when administered orally. The dosing regimen for CAP is usually 500 mg daily for 3 days in outpatients, but an extended release preparation allows the administration of 2000 mg as a one-time dose for CAP. For the hospitalized patient, an intravenous preparation of azithromycin is available and is dosed as 500 mg daily, with the duration defined by the clinical course of the patient, but usually for 7 to 10 days.3 Because of its intravenous administration, the serum levels achieved have been adequate for the therapy of bacteremic pneumococcal pneumonia.51
Clinical studies of CAP have consistently shown a benefit of using macrolide therapy, usually in conjunction with a β-lactam, but the mechanism for this favorable effect is not known.52,53 Speculation has included the possibility of atypical pathogen coinfection, a possibility supported by studies that have found the benefit of the addition of a macrolides to vary over the course of time. Another explanation is the anti-inflammatory effects possessed by macrolides, which may explain their benefit in improving quality of life in patients with cystic fibrosis. Macrolides have a myriad of other effects, including interference with “quorum sensing” between bacteria, which could inhibit the in vivo proliferation of P. aeruginosa after colonization has occurred. A recent study has shown that azithromycin may prevent progression of pseudomonal colonization to VAP.54
Although macrolides remain an important therapeutic option for community respiratory tract infections, pneumococcal resistance is becoming increasingly common, resistance may have plateaued at ≈30% in the United States55,56 especially in patients who have received an agent of this class in the past 3 months.57 In addition, macrolide resistance can also coexist with penicillin resistance with up to 30% to 40% of penicillin-resistant pneumococci also erythromycin-resistant. The clinical relevance of these in vitro findings remains to be defined. However, there are two forms of pneumococcal macrolide resistance, one involving efflux of the antibiotic from the bacterial cell and the other involving altered ribosomal binding of the antibiotic. The former mechanism is associated with much lower levels of resistance than the latter and is present in two-thirds of the macrolide-resistant pneumococci in the United States. The latter form of resistance is less common and, when present, makes macrolide therapy for pneumococcal infection ineffective.
TETRACYCLINES
The tetracyclines are bacteriostatic agents that act by binding the 30S ribosomal subunit and interfering with protein synthesis. These agents can be used in CAP because they are active against H. influenzae and atypical pathogens, but in the United States, pneumococcal resistance to tetracyclines may be approaching 20%, and may exceed 50% among organisms with high-level penicillin resistance. Photosensitivity is the major side effect, limiting the use of these agents in sun-exposed patients. This drug still remains as a good alternative to treat patients with suspected atypical pneumonia infection in those with electrocardiographic QT prolongation.
TIGECYCLINE
Tigecycline is the first clinically available drug from a new class of antibiotics, the glycylcyclines. It is structurally similar to the tetracyclines, and contains a central four-ring carbocyclic skeleton, but is a derivative of minocycline. Tigecycline is bacteriostatic and is a protein synthesis inhibitor that acts by binding to the 30 S ribosomal subunit of bacteria, thereby blocking entry of aminoacyl-tRNA into the ribosome during prokaryotic translation.
This drug is approved to treat complicated skin and skin-structure infections and complicated intra-abdominal infections, but can also be used in CAP caused by S. pneumoniae, H. influenzae, and L. pneumophila. With the exception of gastrointestinal adverse events, tigecycline was generally well tolerated. It is not approved for the therapy of MRSA pneumonia or HAP. In a study by Freire et al.58 tigecycline was compared with imipenem/cilastatin in a multicenter, randomized, double-blind study of HAP in 945 patients. Tigecycline at a dose of 100 mg per day was found to be noninferior to imipenem/cilastatin in a modified intent-to-treat analysis but not in the clinically evaluable population and not in VAP. Some data suggest that efficacy may have been limited by the relatively low dose studied. Guner et al.59 compared tigecycline alone or in combination therapy in the treatment of carbapenem-resistant Acinetobacter baumannii, a major MDR VAP pathogen. This retrospective study of 33 patients showed a 30-day overall mortality rate and attributable mortality rates of 57.6% and 24.2%, respectively. Tigecycline should not be used as monotherapy of VAP, but could be part of a combination when treating resistant Acinetobacter strains.
TRIMETHOPRIM–SULFAMETHOXAZOLE
This combination antibiotic is the mainstay of therapy against Pneumocystis pneumonia. In the past, given its antimicrobial spectrum, ease of use and low cost, it was an effective agent for CAP and AECB. It has bactericidal activity against the pneumococcus, H. influenzae and M. catarrhalis, but not against atypical pathogens. Recently, it has become less popular because of the emergence of pneumococcal resistance at rates of at least 30%, since 80% to 90% of organisms that are penicillin resistant are also resistant to TMP-SMX. The sulfa component of the drug inhibits the bacterial enzyme responsible for forming the immediate precursor of folic acid, dihydropteroic acid. Trimethoprim is synergistic with the sulfa component and inhibits the activity of bacterial dihydrofolate reductase. TMP-SMX is available in a fixed combination of 1:5 (TMP:SMX), and is dosed as either 80/400 mg or 160/800 mg orally twice a day for 10 days, but the dosage should be adjusted in renal failure. An intravenous preparation is also available. Side effects generally result from the sulfa component and include rash, GI upset, and occasional renal failure (especially in elderly patients).
BETA-LACTAM ANTIBIOTICS
β-Lactam antibiotics are bactericidal agents that interfere with the synthesis of bacterial cell–wall peptidoglycans by binding to bacterial penicillin–binding proteins. They share a β-lactam ring with structural modifications determining classification: (a) penicillins with the β-lactam ring is bound to a five-membered thiazolidine ring; (b) cephalosporins with the β-lactam ring bound to a six-membered dihydrothiazine ring; (c) carbapenems with modifications in the thiazolidine ring (imipenem, ertapenem, and meropenem); (d) monobactams with absence of a second ring structure found in the carbapenems (aztreonam). These agents can also be combined with β-lactamase inhibitors such as sulbactam, tazobactam, or clavulanic acid, to create the β-lactam/β-lactamase inhibitor drugs, which may overcome the bacterial resistance mechanism of β-lactamase production.
The penicillins used for respiratory tract infections include the natural penicillins (penicillin G and V), the aminopenicillins (ampicillin, amoxicillin), the anti-staphylococcal agents (nafcillin, oxacillin), the anti-pseudomonal agents (piperacillin, ticarcillin), and the β-lactam/β-lactamase inhibitor combinations (ampicillin/sulbactam, amoxicillin/clavulanate, piperacillin/tazobactam, and ticarcillin/clavulanate). Among the anti-pseudomonal penicillins, piperacillin is the most active agent.
The cephalosporins span four generations. The earlier agents were generally active against gram-positive bacteria without activity against the more complex gram-negatives or anaerobes, and were susceptible to destruction by bacterial β-lactamases. The newer generation agents are generally more specialized, with broader-spectrum activity and with resistance to breakdown by bacterial enzymes. The second generation and newer agents are resistant to bacterial β-lactamases, but recent data suggest that cefuroxime may not be an optimal pneumococcal agent if resistance is present.60 The third-generation agents such as ceftriaxone and cefotaxime are active against penicillin-resistant pneumococci, while ceftazidime is not reliable against the pneumococcus but is active against P. aeruginosa. The third-generation agents may induce β-lactamases among certain gram-negatives (especially the Enterobacteriaceae spp.) and thus may promote the emergence of resistance during monotherapy. The fourth-generation agent, cefepime, is active against pneumococci and P. aeruginosa but is also less likely to induce resistance among the Enterobacteraceae than the third-generation agents. Ceftaroline is a new, advanced-generation cephalsosporin that has been approved for the therapy of CAP; although it has in vitro activity against MRSA, it has not been studied in a randomized controlled trial for pneumonia due to this pathogen and its efficacy in this setting is unknown.
Imipenem, doripenem, and meropenem are the carbapneems, the broadest-spectrum agents in this class, being active against gram-positives, anaerobes, and gram-negatives including P. aeruginosa. They have shown efficacy for patients with severe pneumonia, both community-acquired and nosocomial. A nonpseudomonal carbapenem, ertapenem is also available and has been used effectively in the therapy of CAP and HCAP. Aztreonam is a monobactam that is antigenically distinct from the β-lactams and can be used in penicillin allergic patients. It is only active against gram-negative organisms, having a spectrum very similar to the aminoglycosides.
The role of continuous or extended infusions for time-dependent antibiotics has been extensively studied as a means to optimize fT > MIC, with numerous studies suggesting an improved ability to achieve the required PK-PD targets,61,62 particularly with more virulent bacteria. The clinical benefit is uncertain, with only a limited number of studies demonstrating improved outcomes. Chytra et al.63 performed a randomized, open-label controlled trial, comparing continuous versus intermittent application of meropenem in critically ill patients with severe infection with approximately 50% requiring treatment for a respiratory tract infection. Continuous infusion resulted in higher microbiological success with a shorter duration and lower dose of meropenem although no differences in clinical cure rates or outcomes were observed.63 Similar data have been reported by Dulhunty et al.64 using three commonly prescribed β-lactams and showing that continuous infusion was associated with a significantly improved ability to achieve optimal PK-PD targets along with higher clinical cure rates; however, no difference was identified in ICU-free days or hospital survival. Recently, Arnold et al.65 found that 3-hour infusions of β-lactams for gram-negative infection were not more effective than 30-minute infusions based on endpoints of mortality, duration of therapy, and use of de-escalation.
The unique challenges to effective antibiotic dosing for pneumonia in critical illness have been highlighted in studies of new or emerging antibiotics. A recent clinical study of prolonged infusion and short duration (7 days) therapy of doripenem was compared to 10 days of imipenem/cilastatin for the treatment of VAP, and showed greater mortality and lower clinical cure rates in those receiving shorter durations of therapy even though the PK/PD targets might have been achieved by the extended infusion in patients with ARC.66,67 These data are consistent with an earlier trial (ClinicalTrials.gov Identifier: NCT00229008) examining the use of ceftobiprole versus ceftazidime/linezolid for the treatment of nosocomial pneumonia in which subgroup analyses, showed a trend favoring ceftazidime/linezolid in VAP patients, particularly those <45 years of age, and with a CrCl ≥150 mL/min.68 These findings illustrate the impact of ARC on drug dosing and reinforce the need to further investigate optimal dosing strategies in those with critical illness.
FLUOROQUINOLONES
Fluoroquinolones are bactericidal agents that act by interfering with bacterial DNA gyrase and/or topoisomerase IV, leading to impaired DNA synthesis repair, transcription, and other cellular processes and resulting in bacterial cell lysis. Quinolones inhibit many forms of bacterial topoisomerase enzymes including DNA gyrase. The earlier quinolones (such as ciprofloxacin and ofloxacin) are active primarily against DNA gyrase, which accounts for efficacy against gram-negative bacteria. The newer agents (levofloxacin and moxifloxacin) bind both DNA gyrase and topoisomerase IV, accounting for efficacy against even drug-resistant gram-positive organisms, including drug-resistant S. pneumoniae (DRSP). Resistance to fluoroquinolones can occur via a variety of mechanisms including mutations in the topoisomerase enzymes, by altered permeability of the bacterial cell wall or efflux of the antibiotic from inside of the bacteria.69
The quinolones kill in a concentration-dependent fashion allowing optimal antibacterial activity with infrequent dosing and high peak concentrations and high ratios of AUC/MIC or Cmax/MIC. In addition, because quinolones have a PAE against both gram-positive and gram-negative organisms, killing is maintained after local concentrations fall below the MIC of the target organism. These properties allow the fluoroquinolones to be infrequently dosed often at once daily, particularly given the relatively long half-life of the newer compounds. The factor limiting once-daily dosing for all quinolones is the toxicity associated with higher doses of some agents such as ciprofloxacin, given concerns related to neurotoxicity and possible seizures.
Two features of quinolones that make them well suited to respiratory infections are good oral bioavailability and excellent penetration into respiratory secretions and inflammatory cells within the lung, achieving local concentrations that often exceed serum levels; similar serum and tissue levels can be reached with oral or intravenous administration. Given good lung penetration, the quinolones may be clinically more effective than predicted by MIC values and may be better than other agents in prolonging the “disease-free” interval between exacerbations of COPD, a finding that has been demonstrated for moxifloxacin.70 High bioavailability allows some “borderline” patients (such as nursing home patients) with pneumonia to be managed with outpatient oral therapy and maintain high therapeutic levels in the serum. In addition, these agents permit an easy transition from intravenous to oral therapy of inpatients with pneumonia, facilitating early discharge when the patient is clinically improving.
The fluoroquinolones are active against β-lactamase producing organisms like H. influenzae and M. catarrhalis making them very useful for patients with AECB. However, the newer agents (levofloxacin and moxifloxacin) extend the activity of the quinolones with enhanced gram-positive activity as well as by being more active against C. pneumoniae and M. pneumoniae when compared to the older agents. The new agents are also highly effective against L. pneumophila and may be the drug of choice for this organism. For P. aeruginosa, as in certain patients with CAP, AECB, and HAP, only ciprofloxacin (750 mg twice daily orally or 400 mg every 8 hours intravenously) or levofloxacin (750 mg orally or intravenously daily) are active enough for clinical use.1 Since the older agents (ciprofloxacin and ofloxacin) have borderline activity against the pneumococcus, if they are used for AECB or CAP, the dose should probably be optimized to either 750 mg twice daily of ciprofloxacin for AECB or 750 mg once daily of levofloxacin for CAP or AECB.
In clinical trials, all of the newer agents have been effective in treating AECB with 5 days of therapy. In CAP, therapy is usually for 7 to 14 days, but levofloxacin at 750 mg daily can be used for 5 days. Levofloxacin, but not moxifloxacin, is renally excreted, and thus requires dosage adjustment in patients with renal insufficiency. There are no good studies of severe CAP demonstrating efficacy of any of the quinolones as monotherapy, although monotherapy has been shown to be effective in both nosocomial pneumonia and AECB. These agents also differ in the degree of protein binding. The relevance of this feature to clinical outcome is uncertain, but agents like levofloxacin and moxifloxacin are not highly protein bound and have higher free serum concentrations. Although the newer agents are highly active against both penicillin sensitive and resistant pneumococci, there is concern that widespread use may increase pneumococcal resistance to these agents. Recent data show that many pneumococci (up to 20%) have quinolone-resistance determinant genes present but have not developed full resistance.69 Risk factors for quinolone resistance among pneumococci are recent hospitalization, recent quinolone therapy, and residence in a nursing home.
One major distinction among the new quinolones is their profile of toxic side effects. A number of agents have been removed from clinical use because of toxicities including QT prolongation (grepafloxacin), hypoglycemia (gatifloxacin), phototoxicity (sparfloxacin), and liver necrosis (trovafloxacin). The side effect profiles of other new agents have generally been acceptable, but the risks of use should be weighed against the benefits. A study comparing moxifloxacin to levofloxacin in elderly hospitalized patients with CAP and a high incidence of heart disease, showed comparable safety, with low-frequency cardiac arrhythmias and Clostridium difficile diarrhea among both groups.71
AMINOGLYCOSIDES
The aminoglycosides are bactericidal agents that act by binding to the 30 S ribosomal subunit of bacteria, thus interfering with protein synthesis. Aminoglycosides have primarily a gram-negative spectrum of activity and are usually used in combination with other agents targeting difficult organisms such as P. aeruginosa or other resistant gram-negative bacteria. When combined with certain β-lactam agents they can achieve antibacterial synergy against P. aeruginosa. Amikacin is least susceptible to enzymatic inactivation by bacteria whereas tobramycin is more active than gentamicin against P. aeruginosa. Aminoglycosides penetrate poorly into lung tissue, and can be inactivated by the acidic environment of pneumonic lung tissue. Because of poor penetration, some investigators have used nebulized aminoglycosides for the therapy and/or prevention of gram-negative pneumonia, but clinical data are limited for this application. Clinical trials conducted for nosocomial pneumonia therapy revealed that the use of an aminoglycoside with a β-lactam antibiotic compared to a β-lactam monotherapy alone was not more effective and the combination did not prevent the emergence of Pseudomonal resistance. In the treatment of bacteremic Pseudomonal pneumonia, aminoglycoside combination therapy may be more effective than monotherapy.1,72
As discussed above, aminoglycosides kill in a concentration- dependent fashion, and can be dosed once daily to optimize killing while minimizing toxicity (primarily in renal insufficiency). In clinical practice, this has not been proven to occur, and once-daily dosing is comparable in efficacy and nephrotoxicity to multiple dose regimens.73 When aminoglycosides are used, it is necessary to monitor serum levels to minimize the occurrence of acute renal failure. Peak concentrations correlate with efficacy, but only have meaning with multiple daily doses, and their utility in once-daily regimens has not been established. Trough concentrations are monitored to minimize toxicity and probably should be followed regardless of dosing regimen.
METHICILLIN-RESISTANT S. AUREUS INFECTIONS
MRSA has emerged as an important pathogen in patients with nosocomial pneumonia, particularly VAP and in association with necrotizing post-influenza pneumonia. In the past, vancomycin was the agent used most commonly against this pathogen. However, there have been concerns about the limited efficacy of vancomycin, primarily because of poor penetration into respiratory secretions. In addition, vancomycin is dosed intermittently with the goal of maintaining trough concentrations between 15 and 20 μg/mL; in critically ill patients this level may be associated with nephrotoxicity and may not be effective for organisms with MIC values >1 μg/mL. Vancomycin is often underdosed in critically ill patients with pneumonia and ARC.74 Higher doses may be required in those receiving RRT. Use of a weight-based loading dose (approximately 35 mg/kg, using actual body weight) followed by continuous infusions (adjusted according to the CrCl) has been advocated by some investigators to rapidly achieve therapeutic vancomycin concentrations in critical illness, but without proven benefit.
Quinupristin/dalfopristin has been tested in patients with VAP and was not as effective against MRSA as vancomycin in spite of good in vitro activity. There are several other agents in various stages of development that have activity against MRSA that have not been proven to be useful for the therapy of respiratory tract infections. This includes daptomycin, which is inactivated by pulmonary surfactant, thus explaining a lack of efficacy in pneumonia therapy trials. Tigecycline and ceftaroline are available for nonrespiratory tract infections caused by MRSA, but efficacy of these agents in the therapy of MRSA pneumonia is not known.
LINEZOLID
Linezolid is the first agent in a new antibiotic class, the oxazolidinones, which act by inhibiting bacterial protein synthesis, and has activity against MRSA. Oxazolidinones bind to the 50 S ribosomal subunit preventing binding to transfer RNA and blocking formation of the 70 S initiation complex. Oxazolidinones inhibit production of antibacterial toxins such as the Panton-Valentine leukocidin, which can be produced by community-acquired MRSA strains. Linezolid is active against MRSA as well as against DRSP, and vancomycin- resistant enterococci (VRE) (both Enterococcus faecium and Enterococcus faecalis). This agent has a high bioavailability with serum levels the same with oral or iv therapy. Renal and nonrenal clearance occurs, and dosing adjustments are not needed for patients with renal failure.
Given the increased recognition of the unique PK-PD changes in critical illness, empiric use of linezolid instead of high-dose vancomycin has a theoretical advantage for the treatment of MRSA pneumonia. One prospective double-blind controlled trial involving patients with confirmed MRSA pneumonia, randomized to receive either linezolid (600 mg 12 hourly) or vancomycin (15 mg/kg every 12 hours with dose adjustment on the basis of trough levels) for 7 to 14 days, showed higher clinical success with linezolid with similar 60-day mortality rates between groups.75
Linezolid is a reasonable alternative for patients with treatment failure while receiving vancomycin, for isolates with vancomycin MIC values >2 μg/mL, with allergic reactions, or vancomycininduced nephrotoxicity.76 Side effects are not common and include nausea, diarrhea, anemia, and thrombocytopenia (especially with prolonged use). It is a weak monoamine oxidase inhibitor.
TELAVANCIN
Telavancin is a semi-synthetic derivative of vancomycin and a bactericidal agent that inhibits bacterial cell-wall synthesis and has been used to treat MRSA pneumonia, and is now approved for this use in the United States. In a prospective randomized double-blinded trial, telavancin was compared against vancomycin in the treatment of MRSA HAP in 1503 patients.77 Higher cure with monomicrobial S. aureus were seen with telavancin and it was more effective than vancomycin for MRSA with vancomycin MIC values >1 μg/mL.78
POLYMYXINS
Polymyxins are bactericidal antibiotics that work in a concentration-dependent fashion, and include polymyxin B and E, the latter known as colistin. After binding to lipopolysaccharide (LPS) in the outer membrane of gram-negative bacteria, polymyxins disrupt both the outer and inner membranes, causing increased permeability of the bacterial cell wall to other antibiotics. Gram-negative bacteria can develop resistance to polymyxins through various modifications of the LPS structure that inhibit the binding of polymyxins to LPS. Colistin is used in the treatment of gram-negative bacterial infections.
Polymyxins can be used for pneumonia caused by strains of MDR P. aeruginosa, Acinetobacter resistant to carbapenems, or carbapenemase-producing Enterobacteriaceae. Polymyxins are not absorbed from the gastrointestinal tract and therefore they are administered intravenously or by inhalation. Elimination is primarily renal, and they are 50% protein bound. Major side effects include nephrotoxicity and neurotoxicity (neuromuscular block, confusion, ataxia, visual disturbances, and dizziness).79
A retrospective matched case-control study comparing intravenous (IV) colistin or aerosolized (AS) plus IV colistin in the treatment of VAP by MDR pathogens, in which A. baumannii was the most common pathogen, followed by Klebsiella pneumonia and P. aeruginosa showed no significant differences between the groups for eradication of pathogens, clinical cure, and mortality.80 Optimal pneumonia dosing is not fully established. Dosing is from 3 million (240 mg) units to 5 million (400 mg) units every 8 hours.
AEROSOLIZED ANTIBIOTICS FOR RESPIRATORY TRACT INFECTIONS
Lung penetration of many intravenous antibiotics, including aminoglycosides and colistin, is limited despite appropriate antibiotic administration. Increased dosages and prolonged duration of therapy used to treat resistant organisms can be associated with greater systemic toxicity with an increased risk of development of antimicrobial resistance.81,82 Aerosolized antibiotics might address these concerns by direct delivery to the site of respiratory infection, particularly for agents that penetrate poorly into the lung with systemic administration. Direct delivery of antibiotics is usually achieved by nebulization, which achieves high intrapulmonary concentrations with generally low systemic absorption, reducing the risk of systemic toxicity. The majority of studies of inhaled antibiotics have been done in nonventilated patients with cystic fibrosis, and chronic bronchial infection with P. aeruginosa, including patients with bronchiectasis. Nebulized tobramycin has been shown to improve pulmonary function, as well as decrease the density of P. aeruginosa in sputum and thus reduce the risk of hospital admission83 in cystic fibrosis.
The use of this approach in mechanically ventilated patients has been reported for patients with either infectious tracheobronchitis or VAP, particularly when the infections involve highly resistant gram-negative bacteria (P. aerogenosa, Klebsiella, and Acinetobacter) and in whom topical delivery of antibiotics might allow effective therapy of pathogens that cannot be eradicated by systemic therapy. The appropriate dose of aerosolized antibiotics is uncertain, but can be calculated using the intravenous dose for that particular medication and adding to it the amount of drug with extrapulmonary deposition. When concentration-dependent killing antibiotics are administrated, higher peak tissue concentration is associated with greater bactericidal activity, while time-dependent antibiotics are required to maintain lung tissue concentrations above the MIC of the target organism for a prolonged period.84–89
Factors that influence lung deposition of nebulized antibiotics during mechanical ventilation include aerosol particle size, type of the nebulizer, physical characteristics of the carrying gas, respirator settings, pneumonia severity, and degree of lung aeration (bronchopneumonia differs from consolidation).90–92 Ultrasonic or vibrating plate nebulizers are preferred to jet nebulizers.93–95 Aerosolized particles with median mass aerodynamic diameter of 1 to 5 μm can reach the distal bronchioles and alveolar space.96 A volume control mode of mechanical ventilation has been used with a constant inspiratory flow, but may require sedation during aerosol administration.95,97 Nebulization should be synchronized to the inspiratory cycle, while the inspiratory to expiratory ratio is maintained at 1:1, and the nebulizer can be placed in the inspiratory tubing or in line between the endotracheal tube and the Y connector.95,98,84 Conventional humidifiers need to be used in the ventilator, but humidification is discontinued during nebulization.91 While the optimal method of administration of aerosol therapy is unknown, most studies have shown that nebulization can be effective and achieve more uniform distribution than direct instillation and, with modern delivery methods, more than 50% of the nebulized dose is retained in the lung.
In experimental studies, it has been shown that the alveolar capillary membrane permeability is markedly increased during lung parenchymal infection.85,88 Thus, inhaled antibiotics like aminoglycosides and cephalosporins, but not colistin, could be absorbed into the systemic circulation, with some potential for toxicity. The BAY41–6551 (NCT01004445) drug-device investigational combination using amikacin has demonstrated high-level local drug delivery with low corresponding systemic levels.99
Several recent studies have shown that inhaled antibiotics can be used as either sole therapy or adjunctive (to intravenous) therapy for MDR gram-negative VAP. Cure rates above 60% have been reported, especially with the use of the vibrating mesh plate, small particle nebulizer. Aerosolized antibiotics have shown efficiency in studies to treat the lung infection as well as the bacterial reservoir within the biofilm of the endotracheal tube. One randomized trial compared nebulized amikacin and ceftazidime (without systemic therapy) in the treatment of Pseudomonal VAP to intravenous treatment with the same medications. The study showed that nebulized medications were able to treat VAP but prevented treatment-related acquisition of antibiotic resistance that was seen in patients receiving intravenous antibiotics.100
Another prospective randomized trial examined the impact of adjunctive inhaled amikacin, using a vibrating mesh plate nebulizer in patients with VAP, many of whom had MDR gram-negative infection, and showed that inhaled therapy led to less systemic antibiotic use and more de-escalation than adjunctive inhaled saline.99 In an older study, the addition of endotracheal tobramycin did not improve clinical outcome compared to placebo in VAP, although microbiologic eradication was significantly greater in the patients receiving aerosolized antibiotics.101 In this trial, the lack of benefit may have been related to the use of suboptimal nebulization devices; recent studies using older nebulization devices have not been successful. In general, sporadic small and uncontrolled series have shown that when patients have VAP due to MDR P. aeruginosa or Acinetobacter spp., aerosolized aminoglycosides or colistin may be helpful as adjunctive therapy to systemic antibiotics.95,102
In mechanically ventilated patients, even though local antibiotic administration has been successful, there has been concern about the emergence of MDR gram-negatives with nebulized antibiotic therapy. Most current studies of VAP using either aminoglycosides or polymyxin B have not shown resistance to emerge. One side effect of aerosolized antibiotics has been bronchospasm, which can be induced by the antibiotic or the associated diluents present in certain preparations. A specially formulated preparation of tobramycin for aerosol administration has been designed to avoid this complication.
PRINCIPLES OF THERAPY FOR RESPIRATORY TRACT INFECTIONS
Principles of therapy underlying treatment of community-acquired, healthcare-associated, and hospital-acquired pneumonia are considered below.
COMMUNITY-ACQUIRED PNEUMONIA
Empiric therapy for CAP is selected by categorizing patients on the basis of the presence or absence of cardiopulmonary disease or other specific “modifying” factors that make certain pathogens more likely, severity of illness, and place of therapy (outpatient, inpatient, ICU). Table 125-2 describes the common pathogens causing CAP in specific patient populations. Figure 125-1 describes antibiotic preferences for the treatment of CAP in an outpatient setting, based on patient characteristics and risk factors. Figure 125-2 describes the antibiotic preferences for the treatment of pneumonia for non-ICU hospitalized patients (hemodynamically stable) based on risk factors and allergy status.
