Guidelines for Antimicrobial Therapy and Prophylaxis

From WiserWiki

[edit] Guidelines for Antimicrobial Therapy and Prophylaxis

Matthew E. Levison

The goal of antimicrobial therapy is to clear the tissues of the infecting organisms. This requires that (1) the organism be susceptible to concentrations of the antimicrobial agent at the site of infection, (2) the dose and route of administration result in adequate levels at the site of infection for a sufficient time, (3) local factors at the site of infection not interfere with the activity of the antimicrobial agent, (4) the presence of host defenses to facilitate microbial clearance, and (5) adjunctive therapies be used, when necessary, such as drainage of abscesses or relief of obstructed excretory sites.

[edit] ANTIMICROBIAL DRUG EFFECTS

[edit] Spectrum of Antimicrobial Activity

Table 27-1 provides the antimicrobial spectrum of selected drugs. The beta-lactams all have a β-lactam ring either with an attached five-member ring that contains a sulfur (penams, penems), oxygen (clavams, clavems), or carbon (carbapenams, carbapenems) molecule or with an attached six-member ring that contains a sulfur (cephem), oxygen (oxacephem), or carbon (carbacephem) (Fig. 27-1). Two penams, sulbactam and tazobactam, and the clavam, clavulanic acid, are inferior antimicrobial agents but do irreversibly bind some common β-lactamases, such as those of Staphylococcus aureus, Haemophilus influenzae, Escherichia coli, Klebsiella pneumoniae, and Bacteroides fragilis. Therefore they are marketed in combination with the β-lactam antibiotics, ampicillin, ticarcillin, piperacillin, and amoxicillin. This allows the β-lactam antibiotics to be active against these bacteria, which would otherwise destroy the antibiotics because of β-lactamase production.

Figure 27-1 Schematic representation of β-lactam structures.

Table 27-1 Antimicrobial Activity of Selected Drugs

✢Emerging resistance.

†Except methicillin-resistant staphylococci.

‡Enterococci are resistant.

§Newer fluoroquinolones, levofloxacin, gatifloxacin, and moxifloxacin are active against streptococci.

The cephems are either cephalosporins, produced by a Cephalosporium species, or cephamycins, produced by a Streptomyces species. The various cephalosporins are classified into four "generations," based loosely on spectrum of activity; the first-generation drugs have the greatest activity against methicillin-sensitive S. aureus, and the fourth-generation cefepime has the greatest activity against gram-negative bacilli. Cefoxitin and cefotetan, the two cephamycins, are classified with the second-generation cephalosporins, but unlike most, they have some activity against Bacteroides. Ceftriaxone and cefotaxime, third-generation cephalosporins, and cefepime are the most potent β-lactams against Streptococcus pneumoniae, including those strains with some degree of penicillin resistance.

All fluoroquinolones have a similar basic structure; structural modifications have resulted in the fluoroquinolones being categorized into generations, as with the cephalosporins. All the fluoroquinolones have excellent activity against most gram-negative bacillary pathogens, although Pseudomonas aeruginosa is the least susceptible among these microorganisms. Ciprofloxacin has the greatest potency against P. aeruginosa. The fluoroquinolones also are active against staphylococci, although less so against methicillin- resistant strains. The newer fluoroquinolones, sparfloxacin, grepafloxacin, trovafloxacin, and levofloxacin, have greater activity against S. pneumoniae. Trovafloxacin is active against obligate anaerobes, although its usefulness is greatly diminished by its hepatotoxicity. Aminoglycosides have an antimicrobial spectrum similar to that of the older fluoroquinolones.

The macrolides are active against many gram-positive cocci and bacilli, such as Corynebacterium diphtheriae, Streptococcus pyogenes, S. pneumoniae, and S. aureus, although resistance has emerged among some of these species, especially penicillin-resistant S. pneumoniae and methicillin-resistant S. aureus. The macrolides are also active against Neisseria meningitidis, N. gonorrhoeae, Campylobacter jejuni, Borrelia burgdorferi, Mycoplasma pneumoniae, Legionella pneumophila, Chlamydia species, and Bordetella pertussis. Both clarithromycin and azithromycin are active against Mycobacterium avium complex and H. influenzae. Clindamycin is active against gram-positive cocci (except enterococci) and many obligate anaerobes. Unlike the macrolides, clindamycin is not active against Mycoplasma.

Tetracyclines have broad-spectrum activity, but the development of resistance has limited their use generally to treatment of infections caused by rickettsiae, chlamydiae, mycoplasmas, Treponema pallidum, and B. burgdorferi. Trimethoprim-sulfamethoxazole (TMP-SMX) has broad-spectrum activity, but the development of resistance has limited also its use, although it remains the drug of choice for Pneumocystis carinii infection. The glycopeptides (vancomycin, teicoplanin) are generally active against gram-positive cocci and metronidazole against obligate anaerobes. Rifampin has broad-spectrum activity. Because resistance develops rapidly if used alone, however, rifampin is usually administered in combination with another antibiotic, such as an another antistaphylococcal agent, usually for staphylococcal infection involving a foreign body or bone.

[edit] Mechanisms of Action

The number of targets for the antimicrobial effects of currently available drugs are limited (Table 27-2). The β-lactams bind to penicillin-binding proteins (PBPs), which are enzymes in the cell membrane that the microorganisms require for synthesis of the peptidoglycan component of the cell wall. When bound to β-lactams, these enzymes fail to synthesize cell wall, and the organisms die. The type and number of PBPs vary among species. Gram-negative bacilli have at least six types of PBPs. Inhibition of specific PBPs results in different morphologic changes in the bacterial cell. Most β-lactams bind to PBP-3 of gram-negative bacilli, an enzyme required for cell wall synthesis between newly formed bacilli (i.e., for septation to occur), and as a result of its inhibition, long filaments are formed that die slowly. The carbapenems imipenem and meropenem and the cephalosporin cefepime bind to PBP-2 and result in aggregated cells that burst relatively rapidly. The glycopeptides vancomycin and teicoplanin also interfere with synthesis of peptidoglycan.

Table 27-2 Mechanisms of Action and Resistance of Selected Drugs

The macrolides (e.g., erythromycin, clarithromycin, azithromycin, dirithromycin), clindamycin, streptogramins (e.g., Synercid), chloramphenicol, tetracyclines (e.g., tetracycline, doxycycline, minocycline), and aminoglycosides interfere with protein synthesis at the ribosomal level. Trimethoprim and sulfonamides interfere with purine synthesis. Rifampin inhibits ribonucleic acid (RNA) synthesis by binding to deoxyribonucleic acid (DNA)–dependent RNA polymerase. The fluoroquinolones interfere with DNA folding by binding to DNA gyrase and topoisomerases.

[edit] Mechanisms of Resistance

Microorganisms are inherently resistant to some of the antimicrobial agents; the resistance is a characteristic of the species. In addition, because of selective pressure from exposure to specific agents, microorganisms can acquire resistance to the antimicrobial agent. Resistance develops by mutation, with different genes having different spontaneous mutation rates, or by acquisition of DNA from other microorganisms.

Antimicrobial agents have been produced by microorganisms for many millennia. Microorganisms have survived because of their ability to evolve mechanisms to resist these agents' activity. Humans have produced these drugs for less than 50 years. As a result of the selective pressure of increasing massive exposure of microorganisms to antimicrobial agents that are used in clinical practice, animal husbandry, and agriculture, the emergence of multidrug resistance among human pathogens has become a major problem. Multidrug resistance is now frequently found in the following pathogens: pneumococci, gonococci, enterococci, staphylococci, salmonellae, shigellae, Campylobacter, tubercle bacillus, and nosocomial gram-negative bacilli.

A decrease in the use of specific antimicrobial drugs may lower selective pressure and may be associated with loss of the acquired resistance. Since acquired antimicrobial resistance is an evolving process, selection of appropriate antimicrobial therapy requires knowledge of current antimicrobial susceptibilities of the community-acquired and nosocomial pathogens. Such data should be published regularly by the local hospital's clinical microbiology laboratory. As a corollary, antimicrobial recommendations for many infections must be revised frequently as acquired resistance patterns change. Table 27-2 lists some common mechanisms of resistance.

In vitro susceptibility testing can usually detect microbial resistance. Routine testing may fail to detect some types of resistance, however, because of low concentrations of enzymes capable of inactivating the antimicrobial agent (e.g., β-lactamase) or low numbers of a resistant subpopulation of a mutant in vitro. Nevertheless, their presence is sufficient in the infected patient to result in clinical resistance. For example, antibiotic resistance in Serratia marcescens, Enterobacter cloacae, Citrobacter freundii, Morganella morganii, P. aeruginosa, and Acinetobacter calcoaceticus has been attributed to two related mechanisms: inducible production of chromosomal-encoded β-lactamases and selection of mutants that have lost the genes that control expression of β-lactamase production. This group of organisms has a relatively high mutation rate for loss of the genes that repress β-lactamase production in the absence of a β-lactam agent and that allow β-lactamase production in the presence of a β-lactam agent. The mutation results in continuous production of large amounts of β-lactamase (stable derepression). The derepressed mutants are resistant to third-generation cephalosporins, aztreonam, and broad-spectrum penicillins. In addition, these chromosomal-encoded, inducible β-lactamases are not inhibited by clavulanic acid, sulbactam, or tazobactam, the so-called β-lactamase inhibitors.

Derepressed mutants are present in the dense bacterial populations of infected tissue at the initiation of antibiotic therapy. Selection of the derepressed mutants in the presence of the β-lactam antibiotic is especially a problem in severely immunocompromised patients, whose defective host defenses are unable to control the growth of the few resistant mutants; this apparently accounts for emergence of resistance during therapy in these patients. The only β-lactams that maintain activity against the derepressed mutants are the carbapenems (imipenem, meropenem) and fourth-generation cephalosporin (cefepime). The fluoroquinolones and aminoglycosides may retain activity against these mutants. TMP-SMX may also remain active against these gram-negative bacilli, except P. aeruginosa, which is inherently resistant to TMP-SMX.

Another example in which in vitro testing may fail to predict in vivo resistance is production of plasmid-encoded, extended-spectrum β-lactamases (ESBLs). Nosocomial strains of K. pneumoniae and to a lesser extent E. coli have acquired these ESBLs that inactivate all third-generation cephalosporins, especially ceftazidime, and the monobactam aztreonam. These strains are also frequently resistant to the fluoroquinolones and TMP-SMX. ESBLs are inactivated to a variable extent by sulbactam, clavulanic acid, and tazobactam. Imipenem, meropenem, and to a lesser degree cefepime are most reliable antimicrobial agents against these strains.

[edit] Bacteriostatic vs. Bactericidal Activity

Antimicrobial agents are classified into two major groups: bacteriostatic (i.e., inhibit growth of but do not kill microorganisms) and bactericidal (Box 27-1). Bacteriostatic agents are sufficient to treat most infections, but infections in patients with impaired host defenses (e.g., neutropenia) and at sites of impaired host defenses (e.g., endocarditis, meningitis) require therapy with bactericidal agents to clear pathogens from the site of infection. Bactericidal agents are capable of clearing pathogens from tissues in the absence of host defenses, whereas the residual organisms regrow once bacteriostatic therapy is stopped.

The type of antimicrobial activity of some agents may be microorganism specific. For example, the macrolides, clindamycin, streptogramins, chloramphenicol, and tetracyclines are generally bacteriostatic, although bactericidal activity may occur under certain conditions or against specific microorganisms. Although generally bactericidal, the penicillins are bacteriostatic against enterococci. The activity may also vary with the concentration of an antimicrobial agent; that is, the agent may be bacteriostatic at low concentrations and bactericidal at higher concentrations. The β-lactam antibiotics have a bactericidal effect on only exponentially growing bacteria, which express PBPs, whereas the fluoroquinolones have a bactericidal effect on both exponentially growing and nongrowing microorganisms. Growing microorganisms are found in young cultures and early infection. The majority of microorganisms in most well-established infections are nongrowing. The rapidity and extent of bactericidal activity for the aminoglycosides, fluoroquinolones, and metronidazole are concentration dependent; that is, the higher the concentration, the greater the rate of bactericidal activity. The bactericidal activity of the β-lactams and vancomycin is slow, and the rate of bactericidal activity is minimally enhanced with increasing drug concentrations.

[edit] Postantibiotic Effect

Drugs may also exhibit persistent suppression of microbial growth as a result of transient drug exposure after removal of the drug, the so-called postantibiotic effect (PAE). Drugs that exhibit concentration-dependent bactericidal activity, (aminoglycosides, fluoroquinolones, metronidazole) also exhibit PAE against susceptible organisms, and the duration of their PAE is also concentration dependent. In contrast, most β-lactams, except the carbapenems and cefepime, exhibit no PAE against gram-negative bacilli and relatively short PAEs against gram-positive cocci. Consequently, the bactericidal activity of β-lactams and vancomycin depends on the time the drug concentration exceeds the minimal concentration that inhibits microbial growth (i.e., time-dependent bactericidal activity).

[edit] Inoculum Effect

Activity of certain antimicrobial agents may also depend on the bacterial density, being reduced by dense bacterial populations or enhanced by sparse bacterial populations, the so-called inoculum effect. Dense populations can be less susceptible to antimicrobial agents because of (1) the predominance of nongrowing organisms in dense populations, (2) the high concentration of certain bacterial products in dense populations that inactivate the antimicrobial agents (e.g., β-lactamases), or (3) a greater likelihood that sub populations of resistant mutants will be present in dense populations that can emerge with antimicrobial therapy.

[edit] Synergy and Antagonism

A greater rate and extent of bactericidal activity seen with combinations of two antimicrobial agents, in contrast to the activity exhibited by each agent alone, are called synergy. For example, a combination of bacterial cell wall active agents (e.g., penicillin, ampicillin, vancomycin) alone is at best only slowly bactericidal against enterococci, and an aminoglycoside alone exhibits only inhibitory activity; the combination of the cell wall active agent with an aminoglycoside, however, results in rapid bactericidal activity. This results from enhanced bacterial penetration of the aminoglycoside in the presence of the cell wall active agent. Similarly, synergy has been shown with cell wall active agents/aminoglycoside combinations against hemolytic streptococci, S. aureus, and many gram-negative bacilli. Synergy has been defined as a 2 log 10 or greater or 99% reduction in bacterial count after overnight incubation with the combination vs. that with each of the agents alone.

Combinations of some agents have been found to be antagonistic. For example, β-lactam's bactericidal effect, which requires growing organisms, may be converted to a bacteriostatic effect when combined with another agent that prevents microbial growth.

[edit] Minimum Inhibitory and Bactericidal Concentrations

Antimicrobial activity is measured routinely in vitro in terms of the (1) lowest concentration of the drug that inhibits the growth of 10⁵ colony-forming units (CFUs)/ml of broth of a specific microorganism in the exponential phase of growth after overnight incubation, or minimum inhibitory concentration (MIC), and (2) the lowest concentration that lowers the inoculum by 99.9% (a 3 log₁₀ fall in bacterial count) after overnight incubation, or minimum bactericidal concentration (MBC). The MIC and MBC determinations are performed in serial twofold dilutions of the drug in broth and have an error of plus/minus one dilution. In practice the MIC requires only inspection of the broth culture for the development of turbidity. The lowest drug concentration that prevents visible growth (i.e., turbidity) of an initially clear bacterial suspension (i.e., growth of at least 1 log₁₀, from 10⁵ to 10⁶ CFU/ml) is the MIC. MBC determination requires the more laborious quantification of bacteria remaining in the visibly clear suspension in broth after overnight incubation.

The in vitro conditions of the MIC and MBC determination do not necessarily mimic in vivo conditions. For example, MIC and MBC, which are determined after overnight incubation, reflect a specific time point and do not provide information on the time course of antimicrobial activity. In addition, the in vitro drug concentrations remain constant throughout the incubation period, unlike the varying drug concentrations in vivo. MIC and MBC are measured against a standard inoculum (10⁵/ml) that does not necessarily correspond to bacterial densities at the site of infection (10^8-10/gm). Also, the inoculum is in the exponential phase of growth, unlike the majority of organisms in an established infection, which are nongrowing.

Antimicrobial susceptibility testing is done for most clinical isolates, unless the organism has predictable susceptibility to drugs of choice. The National Committee for Clinical Laboratory Standards (NCCLS) has provided guidelines for performance of susceptibility tests and interpretation of the results. Microorganisms are considered sensitive if their MICs are below a breakpoint concentration and resistant if the MICs are above this concentration. Breakpoint concentrations are related to serum levels achieved with standard dosing, except for the few antimicrobial agents used exclusively for treatment of lower urinary tract infection, when breakpoint concentrations are related to urinary levels. Determination of breakpoint concentrations is complicated, and results of in vitro testing may not adequately predict clinical outcome, which also depends on pharmacokinetic and pharmacodynamic factors. For example, the peak drug concentration relative to the MIC for concentration-dependent drugs and time the serum drug concentrations exceed the MIC for time-dependent drugs have been correlated with drug efficacy in clinical trials.

[edit] THERAPEUTIC APPROACHES

Use of antimicrobial agents may be empiric, pathogen directed, or prophylactic.

[edit] Empiric Therapy

Empiric therapy is usually the initial use of an antimicrobial agent, when the patient is first seen and judged to be critically ill. Patients who require immediate antimicrobial therapy include elderly persons, moderately to severely ill patients with a focal infection, septic patients, febrile neutropenic patients, and those with acute endocarditis or meningitis. At this time the pathogen has not been identified, but any delay in initiation of appropriate antimicrobial therapy would be life threatening. In such patients the antimicrobial regimen should be broad in spectrum, that is, active against all the possible pathogens causing the patient's illness, especially those likely to be rapidly fatal if untreated. The antimicrobial regimen should also be bactericidal because (1) bacteriostatic agents require host defenses to clear the pathogen from tissues, and (2) host defense in critically ill patients may be not adequate to clear the tissues of the pathogens.

Clues to possible pathogens may be present. For example, the presence of focal infection (e.g., pneumonia, urinary or biliary tract, secondary intraabdominal, meningitis) may suggest certain possible pathogens that cause infection more often at the specific site. A history of prior antibiotic use may indicate that the infection is caused by pathogens resistant to the previous antimicrobial regimen. Institutional-acquired infections are more likely to be caused by multidrug-resistant staphylococci or gram-negative bacilli.

Before starting antimicrobial therapy, blood should be cultured and exudates and appropriate body fluids examined microscopically and cultured to determine the causative pathogen. Gram's stain of exudate or body fluid may quickly indicate the causative pathogen. Cultures obtained after initiation of antimicrobial therapy may not reliably yield the causative pathogen, and because of subsequent alteration of bacterial flora of mucosal surfaces and wounds, growth from posttreatment cultures of sputum or exudates may actually be misleading.

Selection of appropriate antimicrobial agents for empiric therapy requires consulting the antimicrobial susceptibility patterns of community-acquired and nosocomial pathogens regularly published by the local hospital's clinical microbiology laboratory. Once results of pretreatment cultures are available, the empiric regimen can be altered according to results of susceptibility tests to the narrowest spectrum, least toxic, and least costly agent among those of comparable efficacy available for the particular type of infection.

[edit] Pathogen-directed Therapy

Pathogen-directed therapy is used when the pathogen is known and laboratory testing has determined its antimicrobial susceptibility.

[edit] Prophylactic Use

Prophylaxis may be primary or secondary. Primary prophylaxis attempts to prevent a pathogen from causing disease in a patient who has never been infected by that pathogen. Examples include perioperative antibiotic administration to prevent surgical wound infection and antibiotic administration immediately before procedures likely to induce bacteremia with microorganisms that cause endocarditis in patients with certain cardiac conditions. Secondary prophylaxis attempts to prevent a reinfection or relapse. For example, antibiotics can be used to prevent reinfection with group A streptococcal pharyngitis and consequent recurrence of rheumatic carditis in a patient with a history of rheumatic carditis. Antibiotics can also be used to prevent clinical relapse of a pathogen by eradication of clinically latent infection (e.g., isoniazid to prevent tuberculosis in a patient who has a positive tuberculin skin test) or only to suppress the clinical emergence of latent infection (e.g., recurrence of P. carinii pneumonia in patients with AIDS).

[edit] Combination Therapy

An antimicrobial regimen may either involve a single agent or a combination of two or more agents. Combinations are used (1) to broaden the spectrum in an empiric regimen, (2) to treat polymicrobial infection, (3) to prevent emergence of antimicrobial resistance (because emergence of mutants resistant to one of the antimicrobial agents is more likely than emergence of a doubly resistant mutant), and (4) to improve the rate and extent of bactericidal activity (i.e., synergy). Therapy with an agent active against E. coli, combined with another agent that is active against anaerobes (e.g., B. fragilis), is required for treatment of secondary intraabdominal infection, although monotherapy is now possible with the development of single agents with activity against both E. coli and anaerobes. Combination therapy with an antienterococcal β-lactam plus an aminoglycoside is required to achieve bactericidal activity for treatment of enterococcal endocarditis. An antistreptococcal β-lactam/aminoglycoside combination has been used to shorten the duration of antimicrobial therapy of streptococcal endocarditis. Similarly, an anti–P. aeruginosa β-lactam/aminoglycoside combination has been used to treat severe P. aeruginosa infections. Rifampin is frequently used in combination with an antistaphylococcal agent for staphylococcal infection involving a foreign body (e.g., prosthetic joint, cardiac valve) or vascular graft.

[edit] Duration of Therapy

The optimal duration for many infections is unknown but should be the shortest time necessary to eradicate the pathogen from the site of infection and prevent relapse. Duration will vary with (1) the rate of clearance, which is characteristic for the specific antimicrobial agent; (2) the presence of effective host defenses at the site of infection, which may enhance antimicrobial efficacy; (3) the age of the infection, which determines the number of growing organisms and the bacterial density; and (4) the ability to eliminate local conditions at the site of infection that favor microbial survival (e.g., foreign bodies, renal or biliary stones, dead bone and other necrotic tissue, continued contamination) or that interfere with the activity of the antimicrobial agents used. For example, only a short course of antimicrobial therapy (about 24 hours) is required for sterile peritonitis that occurs around an infected but resected intraabdominal organ (e.g., appendix, gallbladder) or after adequate surgical early debridement and closure of traumatic perforation of the bowel, whereas weeks of antimicrobial therapy combined with drainage and debridement may be required once intraperitoneal abscesses have formed.

The duration of antimicrobial therapy depends on severity of infection, clinical response to therapy, and normalization of the white blood cell count. Once the patient can tolerate oral therapy, well-absorbed antimicrobial agents can be given orally rather than intravenously, if oral agents are available that have antimicrobial efficacy comparable to that of the intravenous (IV) regimen.

[edit] Adjunctive Therapy

Additional management includes drainage, debridement, relief of obstruction, removal of foreign bodies, and restoration of host defenses. Closed-space infections such as abscesses are difficult to eradicate with antimicrobial therapy without drainage. The poor drug entry and the presence of antimicrobial-inactivating enzymes and other substances, an acidic anaerobic environment, and high microbial density with predominantly nongrowing organisms in abscesses impair the therapeutic efficacy of many antimicrobial drugs. The presence of foreign bodies, stones, dead bone, and other nonviable tissue favors persistence of organisms.

[edit] Cost

Antibiotics constitute 20% to 30% of hospital pharmacy costs. Cost of these drugs includes not only the purchase price, but also the costs of pharmacy and nursing time, supplies required for drug preparation and administration, untoward consequences of their use (e.g., emergence of antimicrobial resistance), and other adverse drug reactions. Costs can be reduced by switching from IV to oral therapy, decreasing the frequency of dosing, switching from multidrug to single-drug therapy, and using drugs or drug combinations with less potential adverse reactions. Oral antimicrobial therapy at home is least costly, but outpatient IV therapy is still less costly than receiving antimicrobial therapy in the hospital. Outpatient infusion therapy can be delivered at an infusion center, by a visiting nurse at home, or by self-administration at home. Outpatient infusion ideally should be with agents requiring infrequent dosing; antibiotics that require frequent dosing should be given by more expensive electronic infusion pumps, if drug stability and solubility are not problems.

[edit] PHARMACOKINETICS

Pharmacokinetics is the study of the time course of a drug's disposition in the body, which is usually described in terms of the drug's concentration in serum because of the relative ease of measurement in this body fluid. The therapeutic effect, however, depends on the time course of the drug concentration at the site of infection.

[edit] Absorption

Most antimicrobial agents are administered intermittently at fixed dosing intervals. The extent of absorption from the site of drug administration is measured by the fraction of the dose absorbed (bioavailability), the maximum serum concentration (C_max), and the time to reach the maximum concentration after administration (T_max). For most drugs the bioavailability and T_max are independent of dose, whereas C_max is dose dependent. After IV bolus administration, absorption is assumed to be rapid and complete. In contrast, absorption after oral or intramuscular (IM) administration is variably slower and incomplete, and the T_max is delayed (usually 1 to 2 hours) and the C_max is lower, because the drug is being eliminated before absorption from the gastrointestinal (GI) tract or from the IM site is complete. Bioavailability after oral absorption can be compromised by GI dysfunction, such as hypochlorhydria (gastric acidity is required for effective absorption of cefuroxime axetil, cefpodoxime proxetil, itraconazole, and ketoconazole), vomiting, rapid intestinal motility, short-gut syndrome, and ileus. Food may interfere with GI absorption of some antimicrobial agents, and divalent and trivalent cations may interfere with absorption of several fluoroquinolones and tetracyclines by chelation. However, the bioavailability after proper oral administration of flu oroquinolones, doxycycline, TMP-SMX, fluconazole, rifampin, and metronidazole is excellent, and their C_max after oral administration approximates that achieved after IV administration. IV administration is used for life-threatening infection because GI function is likely to be impaired in this situation. Many mild infections can be treated entirely with oral agents, or their course of antimicrobial therapy can be completed with oral agents once the patient's condition is no longer critical and the GI tract is functional, if oral agents are available to which the pathogen is sufficiently susceptible. Otherwise, the course of antimicrobial therapy can be completed with IV administration at home once the patient is stabilized.

[edit] Maximum Serum Concentration.

C_max occurs at the completion of an IV infusion or absorption from the GI tract or IM site. The C_max depends on the size of the dose administered, the amount of the drug eliminated during infusion or absorption, and its volume of distribution (V_d). C_max is higher if the dose is larger, if the rate of infusion or absorption is faster, or the V_d is smaller. Too rapid IV infusion (i.e., less than 1 to 2 hours) is not used for certain drugs to avoid toxic reactions, such as chills and fever with amphotericin B and "red-man's" syndrome with vancomycin. Otherwise, IV infusions of antimicrobials are usually given over 20 to 30 minutes.

[edit] Volume of Distribution.

V_d is a proportionality constant that relates the total amount of drug in the body to the serum concentration (e.g., V_d = IV bolus dose + C_max), as if the drug were present throughout the body at the same concentration as found in serum. It is a theoretic value and does not correspond to any actual body compartment. V_d is useful to compare distribution characteristics of different drugs. Drugs that distribute primarily in extracellular fluid (ECF) (e.g., β-lactams, aminoglycosides) will have a relatively small V_d (20% to 30% of lean body weight, or about 15 to 20 L), although patients with an expanded ECF volume (e.g., congestive heart failure, fluid overload, extensive burns, abundant ascites, sepsis) will have a larger V_d and require higher doses to achieve desired serum levels. Drugs that distribute intracellularly (e.g., azithromycin, clarithromycin, fluoroquinolones, rifampin, clindamycin) will have a large V_d (40 L or greater).

[edit] Distribution

A drug is initially distributed after administration throughout the blood volume and tissues in rapid equilibrium with blood, such as highly perfused tissues of the heart, lungs, liver, and kidneys. Active transport pumps from the systemic circulation are present only at excretory sites, such as the liver or kidneys. Otherwise, antimicrobial drugs move from blood to the ECF of tissues by passive diffusion along a concentration gradient. Levels of antimicrobial drugs in interstitial fluid are at best equal to or lower than peak serum levels. Factors that favor transfer of drug from the blood into interstitial fluid include (1) high blood flow and large surface area of the vascular bed of the tissue, (2) absence of endothelial tight junctions and presence of capillary pores in the tissue's vascular bed, (3) high serum drug levels, and (4) low serum protein binding. In tissues such as lung that are well perfused by capillary beds fenestrated by pores, the level of drug in the ECF is similar to free drug levels in serum. Diffusion of drug into closed-space infection (e.g., abscess, empyema) is impaired because the ratio of surface area of the vascular bed to volume of the space is low.

Drugs have difficulty diffusing into tissues that have tight junctions and the absence of capillary pores, such as the eye, prostate, and brain. These sites have an endothelial lipid membrane barrier that limits passive diffusion of hydrophilic drugs such as the β-lactams, although these drugs are able to penetrate the cerebrospinal fluid (CSF) to some extent in the presence of intense inflammation. In addition, the continual formation of CSF lowers CSF drug levels by dilution, and active pumps of organic acids, such as β-lactams, in the eye and choroid plexus lower drug levels in vitreous fluid and CSF, respectively. Lipophilic drugs (e.g., doxycycline, metronidazole, rifampin, trimethoprim, chloramphenicol) readily cross lipid membranes. Lipophilic drugs also enter the prostate and the intracellular fluid (ICF) by another related mechanism, ion trapping, which depends on the pK_a of the drug and ionic charge of the drug molecule at the different pH values of the fluid on either side of the lipid membrane. Weak bases (e.g., macrolides, fluoroquinolones, trimethoprim, clindamycin) are unionized (nonionized) at the pH of serum, and being lipid soluble, the unionized drug diffuses into the cell. Once within the cell, the drug becomes charged at the more acid pH of ICF and is less able to diffuse out. Most β-lactams are weak acids and thus are unionized in the acidic ICF and diffuse more readily out of the cell.

Because capillary pores do not permit passage of serum protein into extravascular sites, only that portion of a drug not bound to serum protein is free to diffuse into tissues. Highly serum protein-bound drugs (generally defined as being 90% or greater protein bound), such as nafcillin, cefazolin, and ceftriaxone, have less free drug to diffuse into tissues than drugs with lower serum protein binding. Protein binding also decreases the antimicrobial activity of drugs in serum. Only the drug's free, unbound portion in serum is active against bacteria. In general, for highly protein-bound drugs, C_max will overestimate antimicrobial activity that is based on MIC testing in a protein-free system, such as broth. The presence of certain serum factors, however, may enhance the anti microbial effect with some drug-bacteria combinations.

Tissue is mainly cellular, and intracellular water accounts for most of its volume. Only a small portion of tissue volume is ECF. For drugs distributed only in ECF (e.g., β-lactams, aminoglycosides), total tissue levels (ICF plus ECF levels) may be only a small fraction of the drug's free serum levels, although ECF levels actually may be similar to free serum levels. For drugs distributed in both ICF and ECF (e.g., fluoroquinolones, rifampin, clindamycin), tissue levels approach or exceed the drug's free serum levels. Bacteria in tissues may be located in ECF, phagocytes, or both. However, ICF location of the pathogen may not correspond exactly to the ICF location of the drug, or the intracellular conditions may interfere with the drug's antimicrobial activity, in which case the desired antimicrobial effect may not necessarily occur.

[edit] Time Course of Serum Levels

Serum levels decline after the peak concentration, and when plotted vs. time on semilog scale, the rate of decline may have one or more phases (Fig. 27-2). The initial rapid portion, called the α-phase, results from mixing with blood and diffusion into those tissues in rapid equilibration with blood. The rate of decline of the subsequent portion, called the β-phase, is slower and is mainly caused by elimination from the body through metabolism or excretion. An additional slower decline, or γ-phase, is seen with some drugs at low serum levels and is mainly caused by slow release of drug from secluded tissue foci.

Figure 27-2 Concentrations of drug in serum after administration (arrow) showing α-, β-, and γ-phases of serum concentration curve and β-phase serum half-life (T_½).

The rate of elimination, or β-phase, is constant for most antimicrobial agents. The elimination of these drugs is described by their elimination half-life (T_½, or the time required to eliminate 50% of the drug present (equals 0.693/slope of β-phase serum level curve, or elimination rate constant). Elimination is almost complete in four to five half-lives. Most β-lactams have serum T_½ of 2 hours or less and are usually given every 4 to 8 hours. The exception is ceftriaxone, with a T_½ of 6 to 8 hours, which is usually dosed every 24 hours. The older quinolones (e.g., norfloxacin, ciprofloxacin, enoxacin) have serum T_½ of 4 hours and are dosed every 12 hours. Structural modifications have resulted in a longer serum T_½ for sparfloxacin, levofloxacin, gatifloxacin, and moxifloxacin, which allows dosing every 24 hours, and decreased renal excretion for moxifloxacin.

In contrast to most antimicrobial agents, the rate of elimination for the ureidopenicillins (e.g., piperacillin, azlocillin, mezlocillin) is dose dependent; the larger the dose (e.g., 5 rather than 3 gm), the longer is the T_½, and the less frequently these drugs need to be administered (e.g., every 8 rather than every 4 hours, respectively).

[edit] Area Under Curve.

Another important pharmacokinetic parameter is the area under the concentration curve (AUC) over the dosing interval. AUC integrates both time and intensity of drug concentrations. AUC depends on the dose, the elimination rate constant (k), and the V_d:AUC = Dose/ V_d × k, or Dose × T_½/ V_d × 0.693.

[edit] Elimination

Renal excretion may be through glomerular filtration, tubular secretion, or both. Elimination by glomerular filtration is passive diffusion of the portion of the drug not bound to serum protein. Glomerular clearance is equal to the percentage of free drug times glomerular filtration rate (GFR). Estimated GFR equals (140 − age in years) (ideal body weight in kg)/(serum creatinine in mg/dl) × 72 in males or 0.85 × male value in females. Tubular secretion is an active transport process. Probenecid inhibits the tubular secretion of organic acids, such as many of the β-lactams, and prolongs their T_½. Although hemodialysis removes variable amounts of many antimicrobial agents, peritoneal dialysis removes very little of most antimicrobial agents. Drugs instilled into the peritoneal cavity, however, are rapidly absorbed over the large expanse of the peritoneal surface into the systemic circulation, so that serum levels achieved will be equal to the concentration of the drug added to the peritoneal dialysis fluid.

Some relatively lipid-soluble antimicrobial agents (e.g., rifampin, macrolides, imidazole antifungals) are metabolized to more polar, less lipid-soluble, and more readily excreted products that also often have reduced antimicrobial effects and toxicity. The enzyme systems involved with metabolism of many of these antimicrobial drugs are located in the liver microsomal, smooth endoplasmic reticulum. A drug may enhance its own metabolism by stimulating the activity of these enzymes, or the activity can be induced by other drugs. The rate of metabolism may also be influenced by competing endogenous and exogenous substances. Induction and competition may result in complex interactions with other drugs and endogenous substances. For example, rifampin, which competes with bilirubin for biliary excretion, initially will elevate serum bilirubin levels, until bilirubin glucuronide production and its excretion in bile increase, as a result of enzyme induction during the first 6 days of treatment, and serum bilirubin levels return to normal. Similarly, rifampin enhances the metabolism of several other drugs, such as prednisone, the sulfonylureas, warfarin, and ketoconazole.

[edit] Loading Doses

Therapy of critically ill patients requires that antimicrobial levels be established at the site of infection as quickly as possible. If the dosing interval is less than four to five half-lives, a progressive increase in serum levels occurs until a steady state is reached (Fig. 27-3). Steady state occurs when the amount of drug administered during the dosing interval equals the amount eliminated. The time required to reach a steady state is also about four to five half-lives. If a drug is given by continuous infusion rather than by intermittent administration, the serum levels rise slowly until equilibration is reached. As with intermittent administration, equilibration occurs between four and five half-lives. If therapeutic levels will only be reached at steady state (i.e., in four to five half-lives), a larger-than-usual dose (a loading dose) of the antimicrobial agent must be administered to achieve therapeutic serum levels with the first dose (Fig. 27-3). A loading dose is recommended for a few of the antimicrobial agents for which the dosing interval is less four to five half-lives (Table 27-3). As with most antimicrobial agents, if the drug is administered at dosing intervals equal to or greater than four half-lives, subsequent doses result in little or no accumulation between doses (i.e., each dose results in serum levels that are identical with those obtained after the initial dose). In this case, each dose should be sufficient to achieve therapeutic serum levels, and no loading dose is necessary.

Figure 27-3 Concentration of drug in serum during administration by intermittent infusion every T_½ with (A) and without (A') a loading dose or by continuous infusion (B).

Table 27-3 Recommended Loading Doses for Antimicrobial Agents

†When metronidazole is used to treat serious infections caused by anaerobic microorganisms.

[edit] Dose Modification.

A loading dose is also necessary when doses are reduced to compensate for reduced elimination (e.g., renal failure). The loading dose is equal to the usual dose given to a patient with normal drug elimination in order to achieve an initial C_max and trough level that fall within the therapeutic range. The size of the maintenance dose, the dosing interval, or both can be altered to prevent drug accumulation in these patients. The pharmacodynamic properties of the drug to be used will influence which modification is preferable. Extension of the dosing interval to about four to five half-lives, while maintaining the size of subsequent doses equal to the loading dose, results in the desired C_max but possibly prolonged periods with subtherapeutic levels; this is the preferred option for drugs exhibiting concentration-dependent pharmacodynamics, such as aminoglycosides and fluoroquinolones. Alternatively, lowering the maintenance dose, while maintaining the normal dosing interval, results in more constant levels with less fluctuation between C_max and trough concentrations; this is the preferred option for drugs exhibiting time-dependent pharmacodynamics, such as β-lactams and vancomycin (see next section).

Although with renal failure the creatinine clearance can be used to judge dose modification for drugs excreted by the kidney, with impairment of hepatic function no such measure of hepatic function exists for drugs excreted by the liver. Drug levels should be determined in serum at peak and trough concentrations at steady state (four to five half-lives), if possible, when doses are reduced for renal failure and should be repeated when the dose or renal function has changed, because recommendations for dose reduction are approximations. Determination of levels is also important in patients with normal elimination for drugs with therapeutic levels that are close to the toxic levels or with individual variability in pharmacokinetics (e.g., aminoglycosides, vancomycin). For example, vancomycin and teicoplanin have more rapid drug elimination and may have larger volumes of distribution in IV drug users, which necessitate determination of peak and trough serum levels of these drugs to ensure a therapeutic effect.

[edit] PHARMACODYNAMICS

Pharmacodynamics describes the antimicrobial effect at the site of infection in relation to the concentrations of the antimicrobial agent during therapy. After a limited exposure of microorganisms to an antibacterial agent, such as after intermittent drug administration, a variable portion of the microbial population persists. The size of this population when the next dose is given will depend on (1) the size of the initial population, (2) the potency (MIC and MBC) and pharmacokinetic characteristics of the antimicrobial agent, (3) the rate and extent of any bactericidal effect, (4) the presence of a postantibiotic effect, and (5) the rate of regrowth of persistent organisms. If doses are spaced too far apart, the residual bacterial count may increase in the later portion of each dosing interval; thus the bacterial count could equal or exceed the count at the beginning of the dosing interval. The factors that allow infrequent dosing without loss of efficacy involve pharmacokinetic variables in relation to antimicrobial effects.

[edit] Time-dependent Bactericidal Action

Bacterial killing by β-lactam antibiotics and vancomycin is not enhanced by increasing drug concentrations above the MBC, and the bactericidal action of these drugs is relatively slow. Consequently, there will be a relatively large residual population when levels fall below the MBC. Once the levels fall below the MIC, persistent suppression of growth may occur either from inhibitory activity of sub-MIC residual drug levels against low bacterial densities (i.e., inoculum effect at both low drug and bacterial levels) or local host defenses. After levels at the site of infection fall below the MIC, the residual population can regrow quickly, because there is either no PAE for most β-lactams against gram-negative bacilli or relatively short PAE against gram-positive cocci. The rate of regrowth depends on many factors, including the inherent doubling time of the microorganism, the availability of nutrients in the infected tissues, and the adequacy of host defense mechanisms. In the absence of host defenses, microorganisms can increase in vivo exponentially at a rate similar to that which occurs in vitro.

Some regrowth may restore susceptibility of the bacterial population to the bactericidal effect of β-lactam antibiotics. Ideally, the next dose is given before significant regrowth occurs. For drugs such as the β-lactams and vancomycin, with concentration-independent bactericidal activity and no or small PAEs, dosing strategies that maximize duration of exposure, such as smaller fractions of the total daily dose given at frequent intervals or use of β-lactams with long serum T_½ (e.g., ceftriaxone, with T_½ of 6 to 8 hours), would be efficacious. An effective dosing regimen for time-dependent antibiotics requires that serum drug concentrations exceed the MIC of the causative pathogen for at least 40% to 50% of the dosing interval. The time above the MIC can be used to compare the effectiveness of different time-dependent antibiotics. As a corollary, drugs within a class having the greater potency (i.e., a lower MIC) will be anticipated to have longer time above MIC ratios and therefore greater effectiveness.

[edit] Concentration-dependent Bactericidal Action

For drugs with concentration-dependent bactericidal action, such as aminoglycosides, fluoroquinolones, and metronidazole, the rate of bactericidal activity will be greatest at the C_max. As drug concentration decreases, bactericidal activity will decrease. Higher doses will increase not only the rate of reduction of bacteria, but also the length of time of drug exposure to bactericidal concentrations. This dependence on both the magnitude and the duration of exposure of bactericidal concentrations implies that concentration-dependent drugs are influenced by the C_max and the AUC for a particular dose. For drugs with time-dependent activity, however, the rate of reduction of bacteria is constant, and the extent of bactericidal activity will depend solely on the duration of drug exposure.

After drug levels at the site of infection fall below the MBC but still exceed the MIC, bacterial counts may continue to fall because of local host defenses (e.g., phagocytes, antibody, complement), or the residual population may remain stable because of the inhibitory effect of the antimicrobial agent. Once residual drug is entirely eliminated from tissues, suppression of growth may persist because of a PAE, the duration of which is also concentration dependent for aminoglycosides and fluoroquinolones; the higher the drug concentration, the longer the PAE. Eventually the PAE will wane, and the residual organisms will begin to regrow.

For concentration-dependent drugs, dosing strategies that maximize the intensity of drug exposure (e.g., giving total daily dose as single dose every 24 hours rather than giving smaller divided doses) would increase the C_max and allow for comparable efficacy at greater convenience and lower cost if adverse effects were not also concentration dependent. Dose-dependent toxicity was once thought to limit giving the total daily dose of an aminoglycoside as a single dose every 24 hours. Data from both animal models and human trials, however, suggest dosing regimens that provide very high peak aminoglycoside concentrations relative to the MIC and prolonged periods of subinhibitory aminoglycoside concentrations may be equally or more effective, without excessive toxicity, than regimens that provide lower peaks but more persistent inhibitory concentrations. These clinical studies, however, have primarily involved combination therapy with other antimicrobial agents or treatment of less severely ill patients or less virulent pathogens.

An effective dosing regimen for concentration-dependent antibiotics requires that either the 24h-AUC/MIC be at least 125 for gram-negative bacilli and probably 25 to 50 for S. aureus and S. pneumoniae or the C_max/MIC of the causative pathogen be at least 10. These ratios can be used to compare the effectiveness of different concentration-dependent antibiotics. As a corollary, drugs within a class having the greater potency (i.e., lower MIC) will have higher AUC/MIC or C_max/MIC ratios and therefore will be anticipated to have greater effectiveness. It will be clear that an infection from susceptible pathogens with relatively high MICs may not be adequately treated with standard dosing of the antimicrobial agent. For example, gentamicin-susceptible strains of P. aeruginosa with gentamicin MICs of 1 to 4 μg/ml may respond suboptimally to standard dosing regimens that provide mean peak serum gentamicin levels of 6 μg/ml. Similarly, ciprofloxacin-susceptible strains of P. aeruginosa with ciprofloxacin MICs of 0.5 to 1 μg/ml may respond suboptimally to standard dosing regimens that provide peak serum ciprofloxacin levels of about 3 to 4 μg/ml. Levofloxacin-susceptible strains of S. pneumoniae with levofloxacin MICs of 1 to 2 μg/ml may respond suboptimally to standard dosing regimens that provide peak serum levofloxacin levels of about 5 μg/ml.

[edit] Bacteriostatic Activity

The macrolides, clindamycin, and the tetracyclines exhibit little if any concentration-dependent killing. These drugs produce prolonged PAE, however, which allows them to be efficacious when concentrations exceed the MIC for less than 50% of the dosing interval.

[edit] Rate and Extent of Bactericidal Effect

Higher rates of bactericidal action result in lower residual bacterial counts and longer intervals before significant regrowth occurs. With concentration-dependent drugs, maximizing the AUC/MIC or C_max/MIC ratio will maximize the rate and extent of bactericidal activity. Similar considerations occur with the use of synergistic drug combinations (i.e., use of two drugs that exert significantly more rapid and extensive bactericidal action in combination than if used alone). Synergistic combinations to clear more rapidly the tissues of the infecting microorganism have been used to shorten the course of therapy for α-hemolytic streptococcal endocarditis (i.e., penicillin or ceftriaxone plus gentamicin for 2 weeks vs. penicillin or ceftriaxone alone for 4 weeks) and for uncomplicated methicillin-sensitive S. aureus, right-sided endocarditis (nafcillin plus gentamicin for 2 weeks vs. nafcillin alone for 4 weeks).

[edit] Prevention of Resistance

Dense populations of bacteria in tissue likely contain subpopulations of bacteria with relatively higher MICs. The likelihood that resistant subpopulations will emerge on antimicrobial therapy depends on (1) the propensity for resistance within the population, (2) the ability of host defenses to control the resistant microorganisms, and (3) the magnitude of the antimicrobial drug levels at the site of infection. Drug levels should be at least eight times the MIC to prevent emergence of resistant subpopulations. This can be accomplished by using a single daily aminoglycoside dose, the most potent fluoroquinolone, or high doses of a β-lactam.

[edit] ADDITIONAL READINGS

• S Alvarez-Elcoro, MJ Enzler: The macrolides: erythromycin, clarithromycin and azithromycin. Mayo Clin Proc 1999; 74:613.
• M Barza: Pharmacologic principles. Gorbach SL Bartlett JG Blacklow NR Infectious diseases. Philadelphia: Saunders; 1992:
• C Carbon: Pharmacodynamics of macrolides, azalides, and streptogramins: effects on extracellular pathogens. Clin Infect Dis 1998; 27:28.
• WA Craig: Pharmacokinetic/pharmacodynamic parameters: rationale for antimicrobial dosing in mice and men. Clin Infect Dis 1998; 26:1.
• GL Drusano: Human pharmacodynamics of beta-lactams, aminoglycosides and their combination. Scand J Infect Dis Suppl 1991; 74:235.
• RS Edson, CL Terrell: The aminoglycosides. Mayo Clin Proc 1999; 74:519.
• L Estes: Review of pharmacokinetics and pharmacodynamics of antimicrobial agents. Mayo Clin Proc 1998; 73:1114.
• WC Hellinger, NS Brewer: Carbapenems and monobactams: imipenem, meropenem, and aztreonam. Mayo Clin Proc 1999; 74:420.
• MJ Ingerman, PG Pitsakis, AF Rosenberg,et al.: The importance of pharmacodynamics in determining the dosing interval in therapy for experimental Pseudomonas endocarditis in the rat. J Infect Dis 1986; 153:707.
• MK Lacy, DP Nicolau, CH Nightingale, R Quintiliani: The pharmacodynamics of aminoglycosides. Clin Infect Dis 1998; 27:23.
• ME Levison: Pharmacodynamics of antimicrobial agents: bactericidal and post-antibiotic effects. Infect Dis Clin North Am 1995; 15:518.
• H Lode, K Borner, P Koeppe: Pharmacodynamics of fluoroquinolones. Clin Infect Dis 1998; 27:33.
• WF Marshall, JE Blair: The cephalosporins. Mayo Clin Proc 1999; 74:187.
• SL Preston,et al.: Pharmacodynamics of levofloxacin: a new paradigm for early clinical trials. JAMA 1998; 279:125.
• JD Turnidge: The pharmacodynamics of β-lactams. Clin Infect Dis 1998; 27:10.
• AJ Wright: The penicillins. Mayo Clin Proc 1999; 74:290.