Antibiotic Kinetic and Dynamic Attributes for Community-Acquired Respiratory Tract Infections

...PRESENTATIONS... Antibiotic Kinetic and Dynamic Attributes for Community-Acquired Respiratory Tract Infections David P. Nicolau, PharmD Presentation Summary Factors, including the age of the treatment population, this population s multiple comorbidities, the greater severity of their illness, and the considerable change in the pathogens in their epidemiology and resistance patterns, affect the management of community-acquired respiratory tract infections in the outpatient setting. Moreover, outcome is affected by the host, etiologic agent, and the selection of the antimicrobial treatment. The challenge of selecting appropriate antimicrobial therapy is often at the discretion of the prescriber. The nature of respiratory tract infections and the relationship of antimicrobial therapy to the resolution of infection are described. The probability of a good outcome in the treatment of respiratory tract infections (RTIs) is based on a variety of factors: the condition of the host, etiologic agent, and the selection of antimicrobial treatment. At present several factors affect the management of communityacquired RTIs in the outpatient setting: the age of the treatment population, this population s multiple comorbidities, the greater severity of their illness, and the considerable change in the pathogens in their epidemiology and resistance patterns. Although the clinician has no control over those factors, the selection of antimicrobial therapy is often at the discretion of the prescriber. Currently, in this era of medical management, clinicians recognize that RTI is not a static process but has a beginning, middle, and end that is best characterized as a dynamic process. The relationship of antimicrobial therapy to the resolution of infection may also be described as a series of dynamic effects (ie, pharmacodynamics, toxicodynamics, and econodynamics), which contribute to the clinical and economic outcomes of the infection process. The major treatment options for community-acquired RTIs include cephalosporins, penicillins and their derivatives (with or without clavulanic acid), macrolides, tetracyclines, and fluoroquinolones. Of these, the macrolides clarithromycin and azithromycin compose a major portion of antibiotic usage from both an economic and clinical practice stand- S1202 THE AMERICAN JOURNAL OF MANAGED CARE DECEMBER 2000

... ANTIBIOTIC KINETIC AND DYNAMIC ATTRIBUTES... point. Over the past decade, clinicians have been able to optimize antimicrobial therapy based on contemporary pharmacodynamic principles, that is, the relationship between antimicrobial concentration and bactericidal effect. Results have shown that antibiotics eradicate bacteria by an interaction of microbiologic activity and pharmacokinetics (PK). Both properties as well as the difference in pharmacodynamic (PD) interaction between the drug and bug for various classes of antibiotics should be considered in drug selection decisions. The application of these principles provides an opportunity for using available antimicrobials in a manner that promotes maximal outcomes, while providing benchmarks for the evaluation and clinical development of novel therapies. At present, antibiotic selection for the treatment of both upper and lower RTIs remains a topic of considerable debate. For example, clinicians should consider if antibiotics are indeed indicated for the treatment of acute otitis media, realizing that the definitions applied to the term appropriate use of antibiotics may be very different depending on who defines the term. In our current medical system, this term could be defined by no less than 5 entities: patient, prescriber, payer, institutional provider, and the pharmaceutical industry. Each of these definitions may be quite different, having been derived from evaluations of the available medical literature or perceptions concerning efficacy and costs, and clinicians must recognize that these differences exist. As an understanding of the interactions between these entities is enhanced, there will be opportunities to work toward a common and unified definition that suits the needs of both the infected individual and the ecosystem, which may be altered by unnecessary antibiotic exposure. Although these issues must be recognized in any discussion of antibiotic usage, this article will focus on the selection of an optimal therapy under the assumption that the initiation of antibiotic therapy is deemed appropriate. Host-Bug-Drug Triad After the decision has been made to initiate antibiotic therapy, the clinician must realize that the goal of therapy is to maximize the probability of a good outcome and that this process may be made easier by understanding the interactions of the hostbug-drug triad. Today, many patients treated for acute exacerbation of chronic bronchitis or communityacquired pneumonia are older and have multiple comorbidities that contribute to poor clinical outcomes. Furthermore, as described elsewhere in this supplement, the pathogens commonly associated with these types of infections have undergone considerable changes in their susceptibility to frequently utilized antibiotics. Consequently, if optimal clinical outcomes are to be maintained, clinicians must be aware of these contemporary issues and the resultant implications of the selection of antibiotic therapy. In addition, the medical community recognizes the infection and recovery process as a dynamic event that changes in severity over the course of the process. As a result, the rate and extent of infection resolution may be defined by 3 dynamic principles: PD, toxicodynamics, and econodynamics. Among the antibiotics, PD typically addresses the relationship between their concentration and their capacity to kill bacteria. However, a current and more comprehensive approach to the selection of an antibiotic also considers its toxicodynamics or the effect that varying its concentrations has on toxicity. Econodynamics, a perspective that has become increasingly more difficult to ignore, is the study of the cost of therapy and its association with VOL. 6, NO. 23, SUP. THE AMERICAN JOURNAL OF MANAGED CARE S1203

... PRESENTATIONS... the rate and extent of outcome, realizing that the utilization of resources may be very different over the course of infection. Optimizing Antimicrobial Therapy The selection of the optimal antimicrobial for any clinical scenario is based on the composite assessment of the microbiologic activity, PK, and drug interaction/adverse event profile of the agents under consideration. Although the microbiologic spectrum and PK profile are often points of differentiation for antimicrobials, current understanding of a drug s effectiveness is based on its PD profile, which simultaneously assesses microbiologic activity and PK. 1-3 Therefore, the PD assessment of potential therapies integrates patient variables, such as PK parameters, with pathogen variables (antibiotic susceptibility), to assess anticipated antimicrobial activity in vivo. The agent s PK profile is responsible for delivering sufficient concentrations to the site of infection, but it is the microbiologic susceptibility and, more specifically, the concentration of drug relative to the microbiologic activity (ie, the minimum inhibitory concentration [MIC]) that results in the PD profile of the agent and ultimately influences the rate and extent of bactericidal activity. However, different antibiotics exhibit their bactericidal activity according to different PD profiles, for example, the effects of some agents are concentration dependent whereas the effects of others are time dependent. The concentration-dependent agents exhibit an increase in rate and extent of bactericidal activity as concentration increases. Antibiotics that display concentration-dependent bactericidal activity include the fluoroquinolones and aminoglycosides. These concentration-dependent effects, as highlighted in the work of Craig and Ebert, reveal that the rate and extent of Pseudomonas aeruginosa kill increases as the exposure concentration increases by multiples of the MIC of the pathogen. 4 Other agents, such as β-lactams and the Figure 1. Concentration-Dependent and Concentration-Independent Bactericidal Activity Against Pseudomonas aeruginosa Log 10 CFU/mL 9 8 7 6 5 4 3 2 Tobramycin Ciprofloxacin Ticarcillin Control 1/4 MIC 1 MIC 4 MIC 16 MIC 64 MIC 0 2 4 6 0 2 4 6 0 2 4 6 8 Time (hours) CFU = colony forming unit; MIC = minimum inhibitory concentration. Source: References 3, 4. S1204 THE AMERICAN JOURNAL OF MANAGED CARE DECEMBER 2000

... ANTIBIOTIC KINETIC AND DYNAMIC ATTRIBUTES... macrolides, exhibit different bacterial killing attributes. After these drugs achieve concentrations of 2 to 4 times above the MIC, further increases in concentration do not yield greater rates of bacterial killing. 4 Figure 1 illustrates the concentration-independent bactericidal activity concept for ticarcillin, tobramycin, and ciprofloxacin against P aeruginosa. Based on in vitro data, PD models have been developed during the past decade and utilized to assist with the prediction of clinical effectiveness for a variety of antimicrobials, most notably the β-lactams, aminoglycosides, and the fluoroquinolones. For the fluoroquinolones, the area under the concentration-time curve:mic (AUC/MIC) ratio is widely regarded as the PD parameter which is most predictive of bactericidal activity and survival in in vivo models of infection and clinical outcomes in humans (Figure 2). 5 As a result of in vitro modeling data and the apparent clinical success of the newer respiratory fluoroquinolones (ie, grepafloxacin, sparfloxacin, levofloxacin, gatifloxacin) in clinical practice, a target AUC/MIC of approximately 40 may be reasonable for community-acquired infections in persons with normal host defenses. 6 Figure 3 illustrates the range of fluoroquinolone AUC/MIC ratios for Streptococcus pneumoniae, which are anticipated based on available PK and microbiologic profiles of the selected agents. 6 The AUC/MIC ratio for ciprofloxacin is minimal, and the exposure provided is fairly limited. These data are consistent with clinical experience showing that ciprofloxacin is not very effective against S pneumoniae and therefore is not a preferred treatment for this commonly occurring communityacquired RTI pathogen. 7-9 The use of levofloxacin 500 mg once daily has raised some concern among clinicians because more than 20% of patients using this drug may achieve less than adequate AUC/MIC ratios. Nevertheless, levofloxacin is clinically Figure 2. Pharmacodynamic Interactions: Integration of Pharmacokinetics and Microbiologic Activity Concentration Peak/MIC Time MIC Time AUC/MIC MIC AUC = area under the concentration-time curve; MIC = minimum inhibitory concentration. Figure 3. Pharmacodynamic Profile of Selected Fluoroquinolones Against Streptococcus pneumoniae AUC/MIC 400 350 300 250 200 150 100 50 0 20-44 Ciprofloxacin 750 mg 24-149 Levofloxacin 500 mg 69-275 Trovafloxacin 200 mg 65-212 Gatifloxacin 400 mg 188-377 Moxifloxacin 400 mg AUC = area under the concentration-time curve; MIC = minimum inhibitory concentration. Source: Adapted from reference 6. VOL. 6, NO. 23, SUP. THE AMERICAN JOURNAL OF MANAGED CARE S1205

... PRESENTATIONS... Figure 4. Clinical Outcome and Microbiologic Eradication Rates: Gatifloxacin or Levofloxacin for the Management of CAP Response Rate (%) 100 90 80 70 60 50 40 30 20 10 0 156/163 96 166/176 94 Clinical Success Gatifloxacin (400 mg qd) 122/125 98 106/114 93 Microbiologic Eradication qd = every day; CAP = community-acquired pneumonia. Source: Reference 11. Log Reduction in CFU 12/12 100 13/16 81 S pneumoniae Eradication Levofloxacin (500 mg qd) Figure 5. Pharmacodynamic Profile of Clarithromycin and Azithromycin in an H. influenzae Pulmonary Infection Model: Day 3 of Therapy 7 6 5 4 3 2 1 0 Clarithromycin Azithromycin 0 100 200 300 400 500 AUC/MIC Ratio in Lung AUC = area under the concentration curve; CFU = colony forming unit; MIC = minimum inhibitory concentration. Source: Adapted from reference 16. effective among many older patients because their age-related reductions in the drug s elimination has resulted in sufficiently high AUC/MIC values. 10 However, the full advantage of agents with AUC/MIC values far in excess of 40 is limited by the lack of comparative trials among compounds of this class. One study comparing the clinical efficacy in communityacquired pneumonia of 2 quinolones, gatifloxacin and levofloxacin, demonstrated that although the clinical success among the agents was very similar, eradication of the pneumococci may be considerably different (Figure 4). 11 Even though comparative data regarding these 2 compounds are limited by the sample size of proven pneumococcal infection, the microbiologic eradication rate of gatifloxacin 100% (12/12) appears to be substantially greater than levofloxacin 81% (13/16). Prediction is based on the PD profile of these 2 fluoroquinolones for this organism. These data suggest that the improved PD profile of gatifloxacin leads to higher pneumococcal eradication and minimizes the potential for the development of resistant isolates. For the β-lactams and macrolides (eg, erythromycin and clarithromycin), in vivo animal and clinical outcome data suggest that the PD parameter most closely related to outcome is the time the serum concentration remains above the MIC (time [T] > 1, 12-15 MIC). More specifically, in otitis media, the PD indices of T > MIC and the middle ear fluid:mic ratio appear to predict bacteriologic efficacy with similar accuracy. These same data reveal that adequate bacterial killing is present when the T > MIC in serum is maintained for 40% to 50% of the dosing interval for these agents. 15 Macrolide PD at Infection Site The similar distribution characteristics in serum and interstitial fluid for the β-lactams explains the PD rela- S1206 THE AMERICAN JOURNAL OF MANAGED CARE DECEMBER 2000

... ANTIBIOTIC KINETIC AND DYNAMIC ATTRIBUTES... tionship between the duration of exposure in serum and the magnitude of the bactericidal effect. However, the macrolide distribution profile is somewhat different, as macrolides tend to concentrate not only within the interstitial space but also within the intracellular space. This difference in antimicrobial distribution and its effects on an agent s in vivo bactericidal activity are evident in the data of Alder et al. 16 In this study the investigators evaluated the in vivo bactericidal activity of clarithromycin and azithromycin in a rat model of Haemophilis influenzae pneumonia by assessing the reduction in the bacterial density of the lung. Although the overall drug exposure (ie, AUC/MIC) in the lung is quite similar between the 2 agents, clarithromycin has a more profound bactericidal effect after 3 days of therapy (Figure 5). This difference in the observed bactericidal effect may be related to the differing distribution characteristics of the agents: Clarithromycin is distributed into both the interstitial and intercellular space. Azithromycin is sequestered in the intracellular space and is unavailable to exert its maximal bactericidal activity on the H influenzae that predominately reside in the interstitial space. By day 7, the reduction in bacterial density within the lung is similar for the 2 agents, presumably as a result of intracellularly sequestered azithromycin released back into the interstitial space. Although the difference in the distribution characteristics of clarithromycin and azithromycin appears to have a profound effect on the rate of bacterial killing in this in vivo model, additional studies are required to determine if observation will correlate to an improvement in the rate of response for patients undergoing clarithromycin treatment. Current understanding of antimicrobial PD can also assist with the explanation of the obvious in vitro and in vivo paradigm regarding the macrolide antibiotics and the pneumococci. Despite the apparent widespread emergence of macrolideresistant S pneumoniae, the clinical efficacy of clarithromycin has been essentially unchanged. 17 The clinical effectiveness of an antimicrobial is the result of not only its degree of susceptibility but also the amount of antibiotic that reaches the organism. A recent surveillance study (see Brueggemann and Doern article in this supplement) indicated that the incidence of macrolide-resistant S pneumoniae has increased. Current rates of resistance approach 26%, as currently defined by the National Committee for Clinical Laboratory Standards. However, macrolide resistance is not one entity as generally defined but the result of 2 very different mechanisms of resistance. In the United States, the more prominent of the 2 mechanisms is a result of the efflux pump encoded by the mefa gene and appears to account for approximately 75% of macrolide resistance observed in S pneumoniae. 18 Although this mechanism can produce MICs ranging from 1 to 32 µg/ml, modal MIC values are generally 2 to 4 µg/ml. The other mechanism that accounts for the remaining macrolide-resistant isolates is mediated by the production of a ribosomal methylase, resulting in MIC values 64 µg/ml. However, resistance is an epidemiologic phenomenon. The clinical effectiveness of an antimicrobial is the result of not only its degree of susceptibility but also the amount of antibiotic that reaches the organism. Although the issue of macrolide resis- VOL. 6, NO. 23, SUP. THE AMERICAN JOURNAL OF MANAGED CARE S1207

... PRESENTATIONS... Figure 6. Intrapulmonary Disposition of Clarithromycin and Azithromycin in Patients Undergoing Bronchoscopy Total Antibiotic in Activity (µg/gm or µg/ml) 700 600 500 400 300 200 40 20 0 68 41.3 34.4 26 15 2.5 0.1 1 2.1 0.1 2.2 1.8 0.04 1 Clari 610 Plasma ELF Alveolar Macrophage 254 Azi Clari Azi Clari Azi 4 8 Hours 12 Azi = azithromycin; Clari = clarithromycin; ELF = epithelial lining fluid. Source: Adapted from reference 16. Figure 7. Clarithromycin Pharmacodynamics at the Infection Site: Epithelial Lining Fluid MIC or Concentration (µg/ml) 100.00 10.00 1.00 0.10 0.01 Resistant (ERM) Clarithromycin 500 mg Resistant (Efflux) Susceptible 1 2 3 4 5 6 7 8 9 10 Day of Dosing MIC = minimum inhibitory concentration. 242 8% 18% 74% 35 % S pneumoniae in MIC Range tance has been considered, drug exposure at the site of infection must also be taken into account. Drug concentration at the site of pulmonary infection can be assessed by measuring the concentration of antibiotic in the epithelial lining fluid (ELF) of patients undergoing bronchoscopy. ELF has been selected for study because these concentrations may most closely resemble the concentration of antibiotic in the interstitial space (the site of residence for pathogens such as the pneumococci). 19 As a result of bronchoscopy studies, such as the one conducted by Rodvold et al, 19 the intrapulmonary disposition of clarithromycin and azithromycin is more clearly defined. As displayed in Figure 6, the intrapulmonary disposition of these agents is markedly different when considering achievable ELF concentrations; concentrations at this site generally range from 34 to 15 µg/ml over the conventional 12-hour dosing interval for clarithromycin. In an attempt to better define the potential utility of clarithromycin in the wake of its reported pneumococcal resistance rate, PK and microbiologic profiles must be integrated at the site of infection. When this PD profiling is undertaken (Figure 7), it is evident that concentrations at the site of infection are higher than most, if not all, of the efflux mediated resistance. As a result, the effective resistance rate for clarithromycin is approximately equal to the estimated resistance of ribosomal methylation (~ 8%). Drug concentrations are well above the inhibitory levels required for both the susceptible and efflux mediated resistant isolates. The discordance between the incidence of macrolideresistant pneumococci and clinical outcomes is confusing for the practicing clinician, but the utilization of PD principles yields a reliable explanation that builds on theory and is supported by good clinical outcomes with the compound. Clinical Application The challenge of prescribing the most effective antibiotic for each patient will increase as resistance continues to influence both the uti- S1208 THE AMERICAN JOURNAL OF MANAGED CARE DECEMBER 2000

... ANTIBIOTIC KINETIC AND DYNAMIC ATTRIBUTES... lization of existing agents and the development of new therapeutic entities. In selecting therapies to optimize bacterial eradication, clinicians should refer to the PD principles discussed. The implementation of antibiotic therapy utilizing the wrong drug or the lowest effective dose will likely nurture the disturbing trend toward drug resistance. Summary Emerging drug resistance and costeffectiveness issues demand that clinicians take a comprehensive approach to management of communityacquired respiratory infections. The therapeutic goal when using timedependent agents, such as the penicillins, carbapenems, cephalosporins, and the macrolides, should be the utilization of a dosage that maximizes the amount of time the antibiotic level remains above the MIC at the site of infection. On the other hand, the appropriate use of concentrationdependent agents, such as the fluoroquinolones, is to maximize the AUC/MIC ratio while choosing agents that minimize the toxicodynamic profile to the host.... REFERENCES... 1. Craig W. Pharmacokinetic/pharmacodynamic parameters: Rationale for antibacterial dosing of mice and men. Clin Infect Dis 1998;26:1-10. 2. Hyatt JM, McKinnon PS, Zimmer GS, et al. The importance of pharmacokinetic/pharmacodynamic surrogate markers to outcome: Focus on antibacterial agents. Clin Pharmacokinet 1995;28:143-160. 3. Quintiliani R, Nicolau, DP, Nightingale CH. Pharmacokinetic and pharmacodynamic principles in antibiotic usage. In: Johnson JT, Yu, VL, eds. Infectious Diseases and Antimicrobial Therapy of the Ears, Nose and Throat. Philadelphia, PA: WB Saunders; 1997:48-55. 4. Craig WA, Ebert SC. Killing and regrowth of bacteria in vitro: A review. Scand J Infect Dis 1990;74(suppl):63-70. 5. Nicolau DP. Using pharmacodynamic/ pharmacokinetic surrogate markers in clinical practice: Optimizing antibiotic therapy in respiratory tract infections. Am J Health- System Pharm 1999;56(suppl 3):S16-S20. 6. Grant EM, Nicolau DP. Pharmacodynamic considerations in the selection of antibiotics for respiratory tract infections: Focus on the fluoroquinolones. Antibio Clinicians 1999;3(suppl 1):21-28. 7. Lee BL, Padula AM, Kimbrough RC, et al. Infectious complications with respiratory pathogens despite ciprofloxacin therapy. N Engl J Med 1991;325:520-521. 8. Gordon JJ, Kaufman CA. Superinfection with Streptococcus pneumoniae during therapy with ciprofloxacin. Am J Med 1990;89:383-384. 9. Righter J. Pneumococcal meningitis during intravenous ciprofloxacin therapy [comment]. Am J Med 1990;88:538. 10. Preston SL, Drusano GL, Berman AL, et al. Pharmacodynamics of levofloxacin: A new paradigm for early clinical trials. JAMA 1998;279:125-129. 11. Sullivan JG, McElroy AD, Honsinger RW, et al. Treating community-acquired pneumonia with once-daily gatifloxacin vs once daily levofloxacin. J Respir Dis 1999; 20(suppl 11):S49-S59. 12. Andes D, Craig WA. In vivo activities of amoxicillin and amoxicillin-clavulanate against Streptococcus pneumoniae: Application to breakpoint determinations. Antimicrob Agents Chemother 1998; 42:2375-2379. 13. Craig WA. Interrelationship between pharmacokinetics and pharmacodynamics in determining dosage regimens for the cephalosporins. Diagn Microbiol Infect Dis 1995;22:89-96. 14. Nicolau DP, Onyeji CO, Zhong MK, Tessier PR, Banevicius MA, Nightingale CH. Pharmacodynamic assessment of cefprozil against Streptococcus pneumoniae: Implications for breakpoint determinations. Antimicrob Agents Chemother 2000;44:1291-1295. 15. Craig WA, Andes D. Pharmacokinetics and pharmacodynamics of antibiotics in otitis media. Pediatr Infect Dis J 1996;15:255-259. 16. Alder JD, Ewing PJ, Nilius AM, et al. Dynamics of clarithromycin and azithromycin efficacies against experimental Haemophilus influenzae pulmonary infection. Antimicrob Agents Chemother 1998;42:2385-2390. VOL. 6, NO. 23, SUP. THE AMERICAN JOURNAL OF MANAGED CARE S1209

... PRESENTATIONS... 17. Gotfried MH. Comparison of bacteriologic eradication of Streptococcus pneumoniae by clarithromycin and reports of increased antimicrobial resistance. Clin Ther 2000;22:2-14. 18. Shortridge VD, Doern GV, Brueggemann AB, Beyer JM, Flamm RK. Prevalence of macrolide resistance mechanisms in Streptococcus pneumoniae isolates from a multicenter antibiotic resistance surveillance study conducted in the United States in 1994-1995. Clin Infect Dis 1999;29:1186-1188. 19. Rodvold KA, Gotfried MH, Danziger LH, Servi RJ. Intrapulmonary steady-state concentrations of clarithromycin and azithromycin in healthy adult volunteers. Antimicrob Agents Chemother 1997;41:1399-1402. S1210 THE AMERICAN JOURNAL OF MANAGED CARE DECEMBER 2000