- Email: [email protected]
Pharmacological considerations in the emergence of resistance
International Journal of Antimicrobial Agents 11 Suppl. 1 (1999) S7 – S14
Pharmacological considerations in the emergence of resistance Guy W. Amsden * The Clinical Pharmacology Research Center and Departments of Pharmacy and Medicine, Bassett Healthcare, 1 Atwell Road, Cooperstown, NY 13326, USA
Abstract Resistance to macrolides in vitro is increasingly being reported. However, there has been no corresponding increase in clinical failures noted. Lack of clinical failures due to resistance is most likely the result of the high intracellular concentrations that these drugs achieve in phagocytes. In the case of clarithromycin, concentrations in both monocytes and granulocytes fluctuate between peaks of :22–25 mg/l and troughs of :5 mg/l during a standard dosing interval. In contrast, azithromycin attains concentrations of over 60 mg/l in granulocytes and at least 100 mg/l in monocytes. After 7 days, azithromycin concentrations of \32 mg/l are still observed. These data also imply that against pathogens with increasing minimum inhibitory concentrations (MICs), macrolides with relatively lower or less sustained intracellular concentrations will become ineffective clinically much sooner than compounds, such as azithromycin, that concentrate to a high degree and are retained in white blood cells for prolonged periods. © 1999 Elsevier Science B.V. and International Society of Chemotherapy. All rights reserved. Keywords: Macrolide; Pharmacokinetics; Resistance; Azithromycin; Clarithromycin
1. Introduction The emergence of resistance to commonly used antibiotic agents among pathogenic bacteria is of increasing concern world-wide [1]. Over the last 30 years, reports of penicillin-resistant Streptococcus pneumoniae have increased in the literature. In some countries, the incidence of penicillin-resistant pneumococci is of immediate clinical concern; prevalences of over 80% have been reported from some central and eastern european nations [2], and 25% of isolates in some centres in Spain are resistant [3]. In other areas, however, levels of resistance remain lower. The increasing MICs of these agents among pathogens is of real clinical concern in view of the pharmacokinetics of b-lactam antibiotics. The b-lactams penetrate cells relatively poorly [4]. Extracellular tissue concentrations are generally in equilibrium with coinciding serum concentrations. Comparing pathogen MICs and serum concentrations of b-lactams, therefore, provide a guide to the likely in vivo effect of * Tel.: +1-607-5473399; fax: + 1-607-5476914.
treatment at the infection site. As a result, if the serum concentrations of b-lactams are below the MICs of target pathogens, their use would be expected to result in treatment failure. Although using more aggressive doses of the b-lactams can overwhelm this problem, even this tactic becomes less and less useful as MICs increase [5] In the case of macrolides and azalides, resistance has been reported in S. pneumoniae in immunocompetent patients as well as in Mycobacterium a6ium complex (MAC) isolated from AIDS patients. In vitro studies have revealed two distinct subsets of resistant pneumococcal strains, those with macrolide/azalide MICs in the range of 4–16 mg/l, and those, albeit fewer in incidence, with MICs\ 128 mg/l [6]. Mycobacterium a6ium complex isolates with clarithromycin MICs\ 32 mg/l are also being identified [7]. In light of increasing resistance to the penicillins and early-generation cephalosporins, it is important to understand the potential clinical implications of the in vitro reports of increasing macrolide/azalide MICs among these pathogens. It has been suggested by some researchers that the low serum concentrations reported
0924-8579/99/$ - see front matter © 1999 Elsevier Science B.V. and International Society of Chemotherapy. All rights reserved. PII: S 0 9 2 4 - 8 5 7 9 ( 9 8 ) 0 0 0 9 8 - 3
S8
G.W. Amsden / International Journal of Antimicrobial Agents 11 Suppl. 1 (1999) S7 – S14
with some of the macrolides and most especially with the only available azalide (a distinct macrolide subclass), azithromycin, are responsible for the increased literature reports of in vitro resistance. The theory behind this claim is that the low serum concentrations represent subinhibitory concentrations of the antibiotics and thus select for, or even induce, the emergence of the resistant populations [8,9]. Another current hypothesis is that these same low serum concentrations could also lead to a risk of clinical failure and potential increased morbidity and mortality due to lack of clearance of bacteria from the blood supply (i.e. bacteraemia). However, if one takes into account the inherent pharmacokinetic and pharmacodynamic properties of these agents, it would be evident that these fears are unfounded, and actually support the continued clinical utility of these agents. While rising MICs in the in vitro literature should be accepted as a warning, especially for some of the short half-life macrolide agents, any true significant clinical implications are yet to be realized. This is particularly true for the azalide agent azithromycin, whose pharmacological properties are especially favourable for maintaining clinical susceptibility and activity among pathogens well beyond laboratory resistance standards. The pharmacokinetics of azithromycin are characterized by low serum and high tissue concentrations [10], uptake, and retention, by phagocytic cells [11] and an extended elimination half-life [12,13]. The data presented here compare the concentrations of azithromycin and clarithromycin (based on historical data) in serum and white blood cells (WBCs) achieved after standard treatment regimens given to healthy volunteers. The importance of serum and tissue concentrations are discussed in relation to the emergence of resistance and the clinical efficacy of the macrolide class of antibiotics.
2. Methods Data for clarithromycin’s serum and white blood cell penetration profiles were abstracted from both product labeling as well as past serum pharmacokinetic and WBC penetration studies [14 – 16]. Previous work suggests that the peak serum concentration in a patient treated with 500 mg of clarithromycin every 12 h is : 2.0 mg/l at steady-state [14,15]. Based on clarithromycin’s half-life, it would be expected that the trough concentrations associated with this dosing scheme would be : 0.5 – 0.8 mg/l. White blood cell penetration research suggests that due to clarithromycin’s monobasic nature, not only does it penetrate cells quickly but it also exits them quickly, thereby preventing intracellular accumulation from rising much beyond eight times the coinciding serum concentrations [16]. Serum, granulocyte and monocyte
concentration versus time curves were then simulated based on these data for a 10-day course of the previously mentioned clarithromycin regimen. For azithromycin, 12 healthy volunteers were administered a standard 3-day course of azithromycin (500 mg daily) in a fasting state. Serum samples were obtained at frequent intervals from the subjects to assess the serum concentration versus time profile of azithromycin both during dosing and for the 7 days after dosing (total sampling time 10 days). Granulocyte and monocyte/lymphocyte cell samples were also harvested at numerous time points during the 10-day sampling period to characterize the WBC penetration and retention properties of oral azithromycin during this standard dosing regimen. Leukocytes were separated utilizing PMN Isolation Media (Robbins Scientific) with trypan blue exclusion tests for cell viability, haemocytometer counts conducted for establishing cell numbers and Wright’s stain smears performed to assess the percentages of each type of WBC isolated. All serum and WBC samples were assayed utilizing a validated high-performance liquid chromatography assay system with electrochemical detection (CVB10%). All serum and WBC data were modelled via noncompartmental techniques using the TopFit Version 2.0 software system.
3. Results
3.1. Serum concentrations The serum concentration versus time profiles (Fig. 1) for both clarithromycin and azithromycin are much as would be expected from past literature reports. Clarithromycin, when dosed 500 mg every 12 h, achieves peak serum concentrations that approach values which are a log-fold higher than those associated with azithromycin (2–2.5 vs. 0.5–0.7 mg/l, respectively). Trough concentrations with clarithromycin fall to : 0.5–0.8 mg/l by the end of the dosing interval, whereas those for azithromycin fall to below 0.1 mg/l within a short period of time after each dose and are almost undetectable at the day 10 final sampling point.
3.2. White blood cell concentrations As demonstrated in Figs. 2 and 3, clarithromycin achieves peak intracellular concentrations of :20–25 mg/l, which coincide with the achievement of peak serum concentrations. This also holds true at the end of the 12-h dosing interval, when trough WBC concentrations decline to :5 mg/l. In contrast, azithromycin achieves a much higher degree of intracellular penetration, and retains it for an extended period. Peak granulocyte concentrations are :85 mg/l and only decline to
G.W. Amsden / International Journal of Antimicrobial Agents 11 Suppl. 1 (1999) S7 – S14
Fig. 1. Macrolide concentration in serum.
Fig. 2. Macrolide concentration in granulocytes.
Fig. 3. Macrolide concentration in monocytes.
S9
S10
G.W. Amsden / International Journal of Antimicrobial Agents 11 Suppl. 1 (1999) S7 – S14
: 32 mg/l at the end of the 10-day sampling period. Monocyte penetration is even more impressive, with peak concentrations of over 100 mg/l being achieved that again fall to : 32 mg/l by the end of the sampling period.
4. Discussion
4.1. Low serum concentrations Serum concentrations of both clarithromycin and azithromycin measured for 10 dosing/post-dosing days are low as compared to many other antibiotics used commonly to treat community-acquired respiratory tract infections. Serum concentrations fluctuate between peaks of 2.0 mg/l and troughs of : 0.5 mg/l following dosing of 500 mg clarithromycin every 12 h. These concentrations, including the trough concentrations, are well above the average MIC of sensitive isolates of S. pneumoniae as shown in Fig. 1. The concentrations of azithromycin peak at a lower level of 0.5 – 0.7 mg/l, and become almost undetectable within a matter of a few days after a standard 3-day course. In contrast to clarithromycin, the serum concentrations achieved with this regimen are infrequently above the average MIC of sensitive S. pneumoniae isolates, which would lead a clinician to believe that the drug should be clinically ineffective for this organism. From a comparison of the serum concentration profiles of both clarithromycin and azithromycin with a pneumococcal MIC of 8.0 mg/l, as has been reported in the resistance literature, it would be expected that both drugs would fail clinically. However, since it is the intracellular concentrations of the macrolides and azalide that are important in the clinical situation, it is apparent why there have not been any wide-scale clinical failures associated with these drugs and this organism reported in the scientific literature. It has been suggested [8,9,17] that the particularly low concentrations of azithromycin in the serum of patients could act as a trigger for the emergence of resistance amongst pathogens. The fact that low concentrations of azithromycin can be detected in the blood for extended periods after the completion of dosing (up to 10 days in this study, over 6 weeks in others [9]) is also of no relevance to the emergence of resistance. In a patient who has been treated successfully, both the blood and the infection site would return to their normal sterile environment (i.e. any pathogens would have been cleared by the WBC and intracellularly exposed to static/cidal concentrations of azithromycin [see below]). There would, therefore, be no pathogens exposed to the azithromycin which was slowly exuded from the white blood cells and tissue sites after curing the patient. Even if bacteria did remain in the blood, in vitro and in vivo
studies, in which pathogens were exposed to sub-inhibitory concentrations of azithromycin [18,19], failed to select for azithromycin resistance. One of these recent studies [18] compared the potential for subinhibitory concentrations of azithromycin, ampicillin and cefaclor to select for resistance in seven clinical pneumococcal isolates. After five serial passages, azithromycin was not shown to result in the development of resistance. In contrast, after exposure to subinhibitory concentrations of either ampicillin or cefaclor, resistance emerged in six of seven strains. Emergence of resistance demonstrated in the normal flora of the oropharynx or gastrointestinal tract during or after a course of antibiotic is common to all marketed antimicrobial agents [17]. However, a common feature of this resistance is its transient nature, and the lack of clinical impact that it has in causing subsequent infections in immunocompetent hosts. The low concentrations of macrolides and azalides attained in the serum may suggest implications related to possible failure of treatment due to bacteraemia. However, these agents, particularly the azalide azithromycin, are actually present in high concentrations, having been avidly taken up by the circulating WBCs. These will phagocytize and clear any pathogens from the blood, thereby bringing them into contact with very high static/cidal concentrations of azithromycin locally. Suggestions that macrolides or azithromycin increase the risk of bacteraemia, or clinical failure due to bacteraemia [20], are not borne out by 40 years of published reports concerning erythromycin, or over 10 years of use of the new macrolides and azithromycin. Failure of treatment with macrolide or azalide antibiotics due to bacteraemia is actually very rare, and those cases which have been documented have usually been due to pathogens not originally susceptible to macrolides or azalides (e.g. Klebsiella spp., Proteus spp.).
4.2. High intracellular concentrations The WBC concentrations following dosing with approved regimens of both azithromycin and clarithromycin are much greater (generally 1–2 logs) than those recorded in the serum. Concentrations of azithromycin in both granulocytes and monocytes are much higher than those achieved with clarithromycin. All the macrolides and the only existing azalide, azithromycin, enter phagocytic cells. As a result of the presence of basic amino groups, they concentrate in the most acidic organelles in the phagocyte, the lysosomes. The macrolides are monobasic (Fig. 4), as a result, it is much easier for them to go back and forth between a non-ionized and ionized state, thereby giving them a much more fluid influx into and efflux from the phagocytes. This is the main reason for the limited and
Fig. 4. Structures of macrolides — erythromycin and clarithromycin.
G.W. Amsden / International Journal of Antimicrobial Agents 11 Suppl. 1 (1999) S7 – S14
S11
S12
G.W. Amsden / International Journal of Antimicrobial Agents 11 Suppl. 1 (1999) S7 – S14
Fig. 5. Structure of azalide — azithromycin.
non-accumulating phagocyte concentrations demonstrated with these agents, as well as for the fluctuating peaks and troughs seen between doses, as described above. In contrast, azithromycin is dibasic in nature (Fig. 5) [21], and thus is less likely to go back to an ionized state and pass out of the lysosome, and eventually the cell. This explains the higher concentrations achieved by azithromycin in the WBC compared with erythromycin and clarithromycin, as well as its continued presence at high concentrations even several days after the end of dosing. The intracellular concentrations of the antibiotics are clinically important. It has been described in the literature that the activity, or pharmacodynamics, of the macrolides and azalide are optimized by maximizing the concentration of the antibiotic for as long as possible (i.e. maximize time of AUC \ MIC) at the sites where the bacteria are located [22]. In this way, the macrolides and the azalide are much like b-lactams, which are also concentration-independent killers. It is important to remember that the WBCs are the final sites of bacterial clearance, whether the pathogen is located in a tissue site or the blood. Macrophages and granulocytes are, therefore, the ideal sites for an antibiotic to exert its action. An additional consideration is the fact that the lysosomes, with their stores of drug, actually merge with the phagosomes that are storing the engulfed bacteria, thereby forming phagolysosomes and exposing the stored bacteria to the very high local concentrations of drug that are shown in Figs. 2 and 3. As the data show, azithromycin phagocyte concentrations remain at or above 32 mg/l for at least 10 days after the start of a standard 3-day regimen. This inherently allows azithromycin to maximize its activity against not only sensitive pathogens, but also pathogens with MICs that are increasing to as high as 32 mg/l (S. pneumoniae strains are classified as resistant in the laboratory at an MIC of 2 mg/l). Clarithromycin
achieves peak cellular concentrations of : 20–25 mg/l that decline to : 5 mg/l immediately prior to the next dose administered. As a result, even though clarithromycin is fully active against sensitive pneumococcal pathogens, once MICs increase much above 4–8 mg/l, the pharmacodynamics of the drug are no longer optimal and clinical failures may ensue. This is not unexpected, and has been demonstrated with the macrolides and azalide against MAC for some time now. The average MICs for MAC to clarithromycin range from 2 to 16 mg/l and from 4 to 16 mg/l to azithromycin. In clinical trials, resistance leading to failures has been significantly more prominent with clarithromycin than with azithromycin (30–50% vs 11%, respectively) [23,24]. If the concentrations that can be achieved in monocytes/macrophages for both drugs are taken into consideration, this higher incidence of resistance with clarithromycin against MAC is quite understandable, and may predict the situation that could be encountered in the future with other pathogens with rising MICs. 5. Conclusions Pharmacological considerations in the emergence of resistance indicate that it is the concentrations of macrolides and azalide which are achieved in the WBCs that are clinically important. Higher concentrations at the site of eradication are less likely to select for resistance. Thus the clinical usefulness of those macrolides which achieve relatively low intracellular concentrations, such as erythromycin, clarithromycin and roxithromycin, will most likely be affected first by the development of resistance. As the MICs increase, these drugs will not be as effective as dirithromycin or azithromycin, compounds with enhanced cellular penetration and prolonged retention, and clinical failures may result. Although dirithromycin penetrates cells
G.W. Amsden / International Journal of Antimicrobial Agents 11 Suppl. 1 (1999) S7 – S14
more extensively than other macrolides and has a prolonged retention time within the cells, it does so less than azithromycin and thus will be affected by resistance earlier. Resistance to macrolides is primarily a result of target site modification (methylation of the ribosomal binding site mediated by the erm gene) or efflux of the antibiotic (mediated by mefE) [25 – 27]. Since resistance to b-lactam antibiotics is largely a result of similar mechanisms, and can be overcome with higher concentrations of the drug [5], it is likely that any resistance to the macrolides or azalide which does become clinically important in the future may be surmounted in the same way. The relevance of in vitro susceptibility breakpoints is also called into question for the macrolide/azalide antibiotics. First, standardized methods have to be developed for clinicians to receive data of consistent quality. Currently, the macrolide/azalide susceptibility data are varied due to the use of different techniques (i.e. E-test vs broth microdilution, with or without CO2) that result in susceptibilities that can range widely for the same isolate [28]. Second, once these techniques are standardized, clinical breakpoints that are based on the pharmacokinetics and pharmacodynamics of the drug need to be validated and instituted to allow clinicians to interpret the MIC results so that they relate to their patient and not to a test tube. As an example, this would raise pneumococcal susceptibility breakpoints for azithromycin from 2 to 32 mg/l, and from 1 mg/l for clarithromycin to 4– 8 mg/l. The in vivo pharmacokinetics of the azalide azithromycin maximize drug dynamics in the clinical setting. With other macrolides, the lower intracellular concentrations mean that the reports of resistance should be viewed as a warning, as their clinical efficacy could be compromised in the near future. The increased resistance of MAC to clarithromycin in prophylaxis trials in the HIV population is most likely predictive of the situation that may arise with other pathogens, as the MIC values begin to rise.
References [1] Jones RN. Can antimicrobial activity be sustained? An appraisal of orally administered drugs used for respiratory tract infections. Diagn Microbiol Infect Dis 1997;27:21–8. [2] Appelbaum PC, Gladkova C, Hryniewicz W, et al. Carriage of antibiotic-resistant Streptococcus pneumoniae by children in eastern and central Europe—a multicenter study with use of standardized methods. Clin Infect Dis 1996;23:712–7. [3] Baquero F. Trends in antibiotic resistance of respiratory pathogens: an analysis and commentary on a collaborative surveillance study. J Antimicrob Chemother 1996;38 (Suppl A):117–32. [4] Chambers HF, Neu HC. Penicillins. In: Mandell GL, Bennett JE, Dolin R, editors. Principles and Practice of Infectious Diseases, 4th edn. NewYork: Churchill Livingstone, 1995:233 – 46.
S13
[5] Pallares R, Lin˜ares J, Vadillo M, et al. Resistance to penicillin and cephalosporin and mortality from severe pneumococcal pneumonia in Barcelona, Spain. N Engl J Med 1995;333:474–80. [6] Ballow CH, Jones RN, Johnson DM, Deinhart JA, Schentag JJ, SPAR Study Group. Comparative in vitro assessment of sparfloxacin activity and spectrum using results from over 14,000 pathogens isolated at 190 medical centers in the USA. Diagn Microbiol Infect Dis 1997;29:173 – 86. [7] Bermudez LE, Petrofsky M, Kolonoski P, Young LS. Emergence of Mycobacterium a6ium populations resistant to macrolides during experimental chemotherapy. Antimicrob Agents Chemother 1998;42:180 – 3. [8] Bauernfeind A, Jungwirth R, Eberlein E. Comparative pharmacodynamics of clarithromycin and azithromycin against respiratory pathogens. Infection 1995;23:316 – 21. [9] Cauchie P, Hubloux A, Crokaert F. A phase I determination of azithromycin concentrations in plasma during a six-week period. Presented at the 4th International Conference on the Macrolides, Azalides, Streptogramins and Ketolides, Abstract no. 10.10. Barcelona, Spain. January 21 – 23 1998. [10] Foulds G, Madsen P, Cox C, Shepard R, Johnson R. Concentration of azithromycin in human prostatic tissue. Eur J Clin Microbiol Infect Dis 1991;10:868 – 71. [11] Gemmell CG. Macrolides and host defences to respiratory tract pathogens. J Hosp Infect 1991;19 (Suppl) A:11 – 19. [12] Foulds G, Shepard RM, Johnson RB. The pharmacokinetics of azithromycin in human serum and tissues. J Antimicrob Chemother 1990;25 (Suppl A):73 – 82. [13] Krohn K. Gynaecological tissue levels of azithromycin. Eur J Clin Microbiol Infect Dis 1991;10:864 – 8. [14] Clarithromycin Product Information, Abbott, 1997. [15] Amsden GW, Ballow CH, Forrest A. Comparison of the plasma, urine and blister fluid pharmacokinetics of clarithromycin and azithromycin in normal subjects. Clin Drug Invest 1997;13:152– 61. [16] Ishiguro M, Koga H, Kohno S, Hayashi T, Yamaguchi K, Hirota M. Penetration of macrolides into human polymorphonuclear leucocytes. J Antimicrob Chemother 1989;24:719– 29. [17] Guggenbichler JP, Kastner U. The influence of macrolide antibiotics on the normal flora. Infect Med 1998;15 (Suppl A):15–22. [18] Hyatt JM, Schentag JJ. A comparison of resistance emergence rates in Streptococcus pneumoniae following exposure to subinhibitory or higher concentrations of azithromycin, ampicillin or cefaclor. Presented at the 4th International Conference on the Macrolides, Azalides, Streptogramins and Ketolides, Abstract 3.14. Barcelona, Spain, January 21 – 23 1998. [19] Retsema JA, Girard AK, Brennan LA, Cimochowski CR, Faiella JA. Lack of emergence of significant resistance in vitro and in vivo to the new azalide antibiotic azithromycin. Eur J Clin Microbiol Infect Dis 1991;10:843 – 6. [20] Jackson MA, Burry VF, Olson LC, Duthie SE, Kearns GL. Breakthrough sepsis in macrolide-resistant pneumococcal infection. Pediatr Infect Dis J 1996;15:1049 – 51. [21] Bright GM, Nagel A, Bordner K, et al. Synthesis and in vitro and in vivo activity of novel 9-deoxo-9a-aza-9a-homoerythromycin A derivatives: a new class of macrolide antibiotics, the azalides. J Antibiot 1988;41:1029 – 47. [22] Craig WA. Postantibiotic effects and the dosing of macrolides, azalides, and streptogramins. In: Zinner SH, Young LS, Acar JF, Neu HC, editors. Expanding Indications for the New Macrolides, Azalides, and Streptogramins. NewYork: Marcel Dekker 1997:27 – 38. [23] Goujard C, Lebrun L, Doucet-Populaire F, Vincent V, Nordmann P. Clarithromycin-resistant Mycobacterium a6ium strain in a clarithromycin-naive AIDS patient. Clin Infect Dis 1998;26:186.
S14
G.W. Amsden / International Journal of Antimicrobial Agents 11 Suppl. 1 (1999) S7 – S14
[24] Amsden GW, Peloquin CA, Beming SE. The role of advanced generation macrolides in the prophylaxis and treatment of Mycobacterium a6ium complex (MAC) infections. Drugs 1997;54:69 – 80. [25] Leclercq R, Courvalin P. Bacterial resistance to macrolide, lincosamide, and streptogramin antibiotics by target modification. Antimicrob Agents Chemother 1991;35:1267–72. [26] Leclercq R, Courvalin P. Intrinsic and unusual resistance to macrolide, lincosamide, and streptogramin antibiotics in bact-
.
eria. Antimicrob Agents Chemother 1991;35:1273 – 6. [27] Tait-Kamradt A, Clancy J, Cronan M, et al. mefE is necessary for the erythromycin-resistant M phenotype in Streptococcus pneumoniae. Antimicrob Agents Chemother 1991;41:2251–5. [28] Gerardo SH, Citron DM, Claros MC, Goldstein EJC. Comparison of Etest to broth microdilution method for testing Streptococcus pneumoniae susceptibility to levofloxacin and three macrolides. Antimicrob Agents Chemother 1996;40:2413– 5.
Pharmacological considerations in the emergence of resistance Guy W. Amsden * The Clinical Pharmacology Research Center and Departments of Pharmacy and Medicine, Bassett Healthcare, 1 Atwell Road, Cooperstown, NY 13326, USA
Abstract Resistance to macrolides in vitro is increasingly being reported. However, there has been no corresponding increase in clinical failures noted. Lack of clinical failures due to resistance is most likely the result of the high intracellular concentrations that these drugs achieve in phagocytes. In the case of clarithromycin, concentrations in both monocytes and granulocytes fluctuate between peaks of :22–25 mg/l and troughs of :5 mg/l during a standard dosing interval. In contrast, azithromycin attains concentrations of over 60 mg/l in granulocytes and at least 100 mg/l in monocytes. After 7 days, azithromycin concentrations of \32 mg/l are still observed. These data also imply that against pathogens with increasing minimum inhibitory concentrations (MICs), macrolides with relatively lower or less sustained intracellular concentrations will become ineffective clinically much sooner than compounds, such as azithromycin, that concentrate to a high degree and are retained in white blood cells for prolonged periods. © 1999 Elsevier Science B.V. and International Society of Chemotherapy. All rights reserved. Keywords: Macrolide; Pharmacokinetics; Resistance; Azithromycin; Clarithromycin
1. Introduction The emergence of resistance to commonly used antibiotic agents among pathogenic bacteria is of increasing concern world-wide [1]. Over the last 30 years, reports of penicillin-resistant Streptococcus pneumoniae have increased in the literature. In some countries, the incidence of penicillin-resistant pneumococci is of immediate clinical concern; prevalences of over 80% have been reported from some central and eastern european nations [2], and 25% of isolates in some centres in Spain are resistant [3]. In other areas, however, levels of resistance remain lower. The increasing MICs of these agents among pathogens is of real clinical concern in view of the pharmacokinetics of b-lactam antibiotics. The b-lactams penetrate cells relatively poorly [4]. Extracellular tissue concentrations are generally in equilibrium with coinciding serum concentrations. Comparing pathogen MICs and serum concentrations of b-lactams, therefore, provide a guide to the likely in vivo effect of * Tel.: +1-607-5473399; fax: + 1-607-5476914.
treatment at the infection site. As a result, if the serum concentrations of b-lactams are below the MICs of target pathogens, their use would be expected to result in treatment failure. Although using more aggressive doses of the b-lactams can overwhelm this problem, even this tactic becomes less and less useful as MICs increase [5] In the case of macrolides and azalides, resistance has been reported in S. pneumoniae in immunocompetent patients as well as in Mycobacterium a6ium complex (MAC) isolated from AIDS patients. In vitro studies have revealed two distinct subsets of resistant pneumococcal strains, those with macrolide/azalide MICs in the range of 4–16 mg/l, and those, albeit fewer in incidence, with MICs\ 128 mg/l [6]. Mycobacterium a6ium complex isolates with clarithromycin MICs\ 32 mg/l are also being identified [7]. In light of increasing resistance to the penicillins and early-generation cephalosporins, it is important to understand the potential clinical implications of the in vitro reports of increasing macrolide/azalide MICs among these pathogens. It has been suggested by some researchers that the low serum concentrations reported
0924-8579/99/$ - see front matter © 1999 Elsevier Science B.V. and International Society of Chemotherapy. All rights reserved. PII: S 0 9 2 4 - 8 5 7 9 ( 9 8 ) 0 0 0 9 8 - 3
S8
G.W. Amsden / International Journal of Antimicrobial Agents 11 Suppl. 1 (1999) S7 – S14
with some of the macrolides and most especially with the only available azalide (a distinct macrolide subclass), azithromycin, are responsible for the increased literature reports of in vitro resistance. The theory behind this claim is that the low serum concentrations represent subinhibitory concentrations of the antibiotics and thus select for, or even induce, the emergence of the resistant populations [8,9]. Another current hypothesis is that these same low serum concentrations could also lead to a risk of clinical failure and potential increased morbidity and mortality due to lack of clearance of bacteria from the blood supply (i.e. bacteraemia). However, if one takes into account the inherent pharmacokinetic and pharmacodynamic properties of these agents, it would be evident that these fears are unfounded, and actually support the continued clinical utility of these agents. While rising MICs in the in vitro literature should be accepted as a warning, especially for some of the short half-life macrolide agents, any true significant clinical implications are yet to be realized. This is particularly true for the azalide agent azithromycin, whose pharmacological properties are especially favourable for maintaining clinical susceptibility and activity among pathogens well beyond laboratory resistance standards. The pharmacokinetics of azithromycin are characterized by low serum and high tissue concentrations [10], uptake, and retention, by phagocytic cells [11] and an extended elimination half-life [12,13]. The data presented here compare the concentrations of azithromycin and clarithromycin (based on historical data) in serum and white blood cells (WBCs) achieved after standard treatment regimens given to healthy volunteers. The importance of serum and tissue concentrations are discussed in relation to the emergence of resistance and the clinical efficacy of the macrolide class of antibiotics.
2. Methods Data for clarithromycin’s serum and white blood cell penetration profiles were abstracted from both product labeling as well as past serum pharmacokinetic and WBC penetration studies [14 – 16]. Previous work suggests that the peak serum concentration in a patient treated with 500 mg of clarithromycin every 12 h is : 2.0 mg/l at steady-state [14,15]. Based on clarithromycin’s half-life, it would be expected that the trough concentrations associated with this dosing scheme would be : 0.5 – 0.8 mg/l. White blood cell penetration research suggests that due to clarithromycin’s monobasic nature, not only does it penetrate cells quickly but it also exits them quickly, thereby preventing intracellular accumulation from rising much beyond eight times the coinciding serum concentrations [16]. Serum, granulocyte and monocyte
concentration versus time curves were then simulated based on these data for a 10-day course of the previously mentioned clarithromycin regimen. For azithromycin, 12 healthy volunteers were administered a standard 3-day course of azithromycin (500 mg daily) in a fasting state. Serum samples were obtained at frequent intervals from the subjects to assess the serum concentration versus time profile of azithromycin both during dosing and for the 7 days after dosing (total sampling time 10 days). Granulocyte and monocyte/lymphocyte cell samples were also harvested at numerous time points during the 10-day sampling period to characterize the WBC penetration and retention properties of oral azithromycin during this standard dosing regimen. Leukocytes were separated utilizing PMN Isolation Media (Robbins Scientific) with trypan blue exclusion tests for cell viability, haemocytometer counts conducted for establishing cell numbers and Wright’s stain smears performed to assess the percentages of each type of WBC isolated. All serum and WBC samples were assayed utilizing a validated high-performance liquid chromatography assay system with electrochemical detection (CVB10%). All serum and WBC data were modelled via noncompartmental techniques using the TopFit Version 2.0 software system.
3. Results
3.1. Serum concentrations The serum concentration versus time profiles (Fig. 1) for both clarithromycin and azithromycin are much as would be expected from past literature reports. Clarithromycin, when dosed 500 mg every 12 h, achieves peak serum concentrations that approach values which are a log-fold higher than those associated with azithromycin (2–2.5 vs. 0.5–0.7 mg/l, respectively). Trough concentrations with clarithromycin fall to : 0.5–0.8 mg/l by the end of the dosing interval, whereas those for azithromycin fall to below 0.1 mg/l within a short period of time after each dose and are almost undetectable at the day 10 final sampling point.
3.2. White blood cell concentrations As demonstrated in Figs. 2 and 3, clarithromycin achieves peak intracellular concentrations of :20–25 mg/l, which coincide with the achievement of peak serum concentrations. This also holds true at the end of the 12-h dosing interval, when trough WBC concentrations decline to :5 mg/l. In contrast, azithromycin achieves a much higher degree of intracellular penetration, and retains it for an extended period. Peak granulocyte concentrations are :85 mg/l and only decline to
G.W. Amsden / International Journal of Antimicrobial Agents 11 Suppl. 1 (1999) S7 – S14
Fig. 1. Macrolide concentration in serum.
Fig. 2. Macrolide concentration in granulocytes.
Fig. 3. Macrolide concentration in monocytes.
S9
S10
G.W. Amsden / International Journal of Antimicrobial Agents 11 Suppl. 1 (1999) S7 – S14
: 32 mg/l at the end of the 10-day sampling period. Monocyte penetration is even more impressive, with peak concentrations of over 100 mg/l being achieved that again fall to : 32 mg/l by the end of the sampling period.
4. Discussion
4.1. Low serum concentrations Serum concentrations of both clarithromycin and azithromycin measured for 10 dosing/post-dosing days are low as compared to many other antibiotics used commonly to treat community-acquired respiratory tract infections. Serum concentrations fluctuate between peaks of 2.0 mg/l and troughs of : 0.5 mg/l following dosing of 500 mg clarithromycin every 12 h. These concentrations, including the trough concentrations, are well above the average MIC of sensitive isolates of S. pneumoniae as shown in Fig. 1. The concentrations of azithromycin peak at a lower level of 0.5 – 0.7 mg/l, and become almost undetectable within a matter of a few days after a standard 3-day course. In contrast to clarithromycin, the serum concentrations achieved with this regimen are infrequently above the average MIC of sensitive S. pneumoniae isolates, which would lead a clinician to believe that the drug should be clinically ineffective for this organism. From a comparison of the serum concentration profiles of both clarithromycin and azithromycin with a pneumococcal MIC of 8.0 mg/l, as has been reported in the resistance literature, it would be expected that both drugs would fail clinically. However, since it is the intracellular concentrations of the macrolides and azalide that are important in the clinical situation, it is apparent why there have not been any wide-scale clinical failures associated with these drugs and this organism reported in the scientific literature. It has been suggested [8,9,17] that the particularly low concentrations of azithromycin in the serum of patients could act as a trigger for the emergence of resistance amongst pathogens. The fact that low concentrations of azithromycin can be detected in the blood for extended periods after the completion of dosing (up to 10 days in this study, over 6 weeks in others [9]) is also of no relevance to the emergence of resistance. In a patient who has been treated successfully, both the blood and the infection site would return to their normal sterile environment (i.e. any pathogens would have been cleared by the WBC and intracellularly exposed to static/cidal concentrations of azithromycin [see below]). There would, therefore, be no pathogens exposed to the azithromycin which was slowly exuded from the white blood cells and tissue sites after curing the patient. Even if bacteria did remain in the blood, in vitro and in vivo
studies, in which pathogens were exposed to sub-inhibitory concentrations of azithromycin [18,19], failed to select for azithromycin resistance. One of these recent studies [18] compared the potential for subinhibitory concentrations of azithromycin, ampicillin and cefaclor to select for resistance in seven clinical pneumococcal isolates. After five serial passages, azithromycin was not shown to result in the development of resistance. In contrast, after exposure to subinhibitory concentrations of either ampicillin or cefaclor, resistance emerged in six of seven strains. Emergence of resistance demonstrated in the normal flora of the oropharynx or gastrointestinal tract during or after a course of antibiotic is common to all marketed antimicrobial agents [17]. However, a common feature of this resistance is its transient nature, and the lack of clinical impact that it has in causing subsequent infections in immunocompetent hosts. The low concentrations of macrolides and azalides attained in the serum may suggest implications related to possible failure of treatment due to bacteraemia. However, these agents, particularly the azalide azithromycin, are actually present in high concentrations, having been avidly taken up by the circulating WBCs. These will phagocytize and clear any pathogens from the blood, thereby bringing them into contact with very high static/cidal concentrations of azithromycin locally. Suggestions that macrolides or azithromycin increase the risk of bacteraemia, or clinical failure due to bacteraemia [20], are not borne out by 40 years of published reports concerning erythromycin, or over 10 years of use of the new macrolides and azithromycin. Failure of treatment with macrolide or azalide antibiotics due to bacteraemia is actually very rare, and those cases which have been documented have usually been due to pathogens not originally susceptible to macrolides or azalides (e.g. Klebsiella spp., Proteus spp.).
4.2. High intracellular concentrations The WBC concentrations following dosing with approved regimens of both azithromycin and clarithromycin are much greater (generally 1–2 logs) than those recorded in the serum. Concentrations of azithromycin in both granulocytes and monocytes are much higher than those achieved with clarithromycin. All the macrolides and the only existing azalide, azithromycin, enter phagocytic cells. As a result of the presence of basic amino groups, they concentrate in the most acidic organelles in the phagocyte, the lysosomes. The macrolides are monobasic (Fig. 4), as a result, it is much easier for them to go back and forth between a non-ionized and ionized state, thereby giving them a much more fluid influx into and efflux from the phagocytes. This is the main reason for the limited and
Fig. 4. Structures of macrolides — erythromycin and clarithromycin.
G.W. Amsden / International Journal of Antimicrobial Agents 11 Suppl. 1 (1999) S7 – S14
S11
S12
G.W. Amsden / International Journal of Antimicrobial Agents 11 Suppl. 1 (1999) S7 – S14
Fig. 5. Structure of azalide — azithromycin.
non-accumulating phagocyte concentrations demonstrated with these agents, as well as for the fluctuating peaks and troughs seen between doses, as described above. In contrast, azithromycin is dibasic in nature (Fig. 5) [21], and thus is less likely to go back to an ionized state and pass out of the lysosome, and eventually the cell. This explains the higher concentrations achieved by azithromycin in the WBC compared with erythromycin and clarithromycin, as well as its continued presence at high concentrations even several days after the end of dosing. The intracellular concentrations of the antibiotics are clinically important. It has been described in the literature that the activity, or pharmacodynamics, of the macrolides and azalide are optimized by maximizing the concentration of the antibiotic for as long as possible (i.e. maximize time of AUC \ MIC) at the sites where the bacteria are located [22]. In this way, the macrolides and the azalide are much like b-lactams, which are also concentration-independent killers. It is important to remember that the WBCs are the final sites of bacterial clearance, whether the pathogen is located in a tissue site or the blood. Macrophages and granulocytes are, therefore, the ideal sites for an antibiotic to exert its action. An additional consideration is the fact that the lysosomes, with their stores of drug, actually merge with the phagosomes that are storing the engulfed bacteria, thereby forming phagolysosomes and exposing the stored bacteria to the very high local concentrations of drug that are shown in Figs. 2 and 3. As the data show, azithromycin phagocyte concentrations remain at or above 32 mg/l for at least 10 days after the start of a standard 3-day regimen. This inherently allows azithromycin to maximize its activity against not only sensitive pathogens, but also pathogens with MICs that are increasing to as high as 32 mg/l (S. pneumoniae strains are classified as resistant in the laboratory at an MIC of 2 mg/l). Clarithromycin
achieves peak cellular concentrations of : 20–25 mg/l that decline to : 5 mg/l immediately prior to the next dose administered. As a result, even though clarithromycin is fully active against sensitive pneumococcal pathogens, once MICs increase much above 4–8 mg/l, the pharmacodynamics of the drug are no longer optimal and clinical failures may ensue. This is not unexpected, and has been demonstrated with the macrolides and azalide against MAC for some time now. The average MICs for MAC to clarithromycin range from 2 to 16 mg/l and from 4 to 16 mg/l to azithromycin. In clinical trials, resistance leading to failures has been significantly more prominent with clarithromycin than with azithromycin (30–50% vs 11%, respectively) [23,24]. If the concentrations that can be achieved in monocytes/macrophages for both drugs are taken into consideration, this higher incidence of resistance with clarithromycin against MAC is quite understandable, and may predict the situation that could be encountered in the future with other pathogens with rising MICs. 5. Conclusions Pharmacological considerations in the emergence of resistance indicate that it is the concentrations of macrolides and azalide which are achieved in the WBCs that are clinically important. Higher concentrations at the site of eradication are less likely to select for resistance. Thus the clinical usefulness of those macrolides which achieve relatively low intracellular concentrations, such as erythromycin, clarithromycin and roxithromycin, will most likely be affected first by the development of resistance. As the MICs increase, these drugs will not be as effective as dirithromycin or azithromycin, compounds with enhanced cellular penetration and prolonged retention, and clinical failures may result. Although dirithromycin penetrates cells
G.W. Amsden / International Journal of Antimicrobial Agents 11 Suppl. 1 (1999) S7 – S14
more extensively than other macrolides and has a prolonged retention time within the cells, it does so less than azithromycin and thus will be affected by resistance earlier. Resistance to macrolides is primarily a result of target site modification (methylation of the ribosomal binding site mediated by the erm gene) or efflux of the antibiotic (mediated by mefE) [25 – 27]. Since resistance to b-lactam antibiotics is largely a result of similar mechanisms, and can be overcome with higher concentrations of the drug [5], it is likely that any resistance to the macrolides or azalide which does become clinically important in the future may be surmounted in the same way. The relevance of in vitro susceptibility breakpoints is also called into question for the macrolide/azalide antibiotics. First, standardized methods have to be developed for clinicians to receive data of consistent quality. Currently, the macrolide/azalide susceptibility data are varied due to the use of different techniques (i.e. E-test vs broth microdilution, with or without CO2) that result in susceptibilities that can range widely for the same isolate [28]. Second, once these techniques are standardized, clinical breakpoints that are based on the pharmacokinetics and pharmacodynamics of the drug need to be validated and instituted to allow clinicians to interpret the MIC results so that they relate to their patient and not to a test tube. As an example, this would raise pneumococcal susceptibility breakpoints for azithromycin from 2 to 32 mg/l, and from 1 mg/l for clarithromycin to 4– 8 mg/l. The in vivo pharmacokinetics of the azalide azithromycin maximize drug dynamics in the clinical setting. With other macrolides, the lower intracellular concentrations mean that the reports of resistance should be viewed as a warning, as their clinical efficacy could be compromised in the near future. The increased resistance of MAC to clarithromycin in prophylaxis trials in the HIV population is most likely predictive of the situation that may arise with other pathogens, as the MIC values begin to rise.
References [1] Jones RN. Can antimicrobial activity be sustained? An appraisal of orally administered drugs used for respiratory tract infections. Diagn Microbiol Infect Dis 1997;27:21–8. [2] Appelbaum PC, Gladkova C, Hryniewicz W, et al. Carriage of antibiotic-resistant Streptococcus pneumoniae by children in eastern and central Europe—a multicenter study with use of standardized methods. Clin Infect Dis 1996;23:712–7. [3] Baquero F. Trends in antibiotic resistance of respiratory pathogens: an analysis and commentary on a collaborative surveillance study. J Antimicrob Chemother 1996;38 (Suppl A):117–32. [4] Chambers HF, Neu HC. Penicillins. In: Mandell GL, Bennett JE, Dolin R, editors. Principles and Practice of Infectious Diseases, 4th edn. NewYork: Churchill Livingstone, 1995:233 – 46.
S13
[5] Pallares R, Lin˜ares J, Vadillo M, et al. Resistance to penicillin and cephalosporin and mortality from severe pneumococcal pneumonia in Barcelona, Spain. N Engl J Med 1995;333:474–80. [6] Ballow CH, Jones RN, Johnson DM, Deinhart JA, Schentag JJ, SPAR Study Group. Comparative in vitro assessment of sparfloxacin activity and spectrum using results from over 14,000 pathogens isolated at 190 medical centers in the USA. Diagn Microbiol Infect Dis 1997;29:173 – 86. [7] Bermudez LE, Petrofsky M, Kolonoski P, Young LS. Emergence of Mycobacterium a6ium populations resistant to macrolides during experimental chemotherapy. Antimicrob Agents Chemother 1998;42:180 – 3. [8] Bauernfeind A, Jungwirth R, Eberlein E. Comparative pharmacodynamics of clarithromycin and azithromycin against respiratory pathogens. Infection 1995;23:316 – 21. [9] Cauchie P, Hubloux A, Crokaert F. A phase I determination of azithromycin concentrations in plasma during a six-week period. Presented at the 4th International Conference on the Macrolides, Azalides, Streptogramins and Ketolides, Abstract no. 10.10. Barcelona, Spain. January 21 – 23 1998. [10] Foulds G, Madsen P, Cox C, Shepard R, Johnson R. Concentration of azithromycin in human prostatic tissue. Eur J Clin Microbiol Infect Dis 1991;10:868 – 71. [11] Gemmell CG. Macrolides and host defences to respiratory tract pathogens. J Hosp Infect 1991;19 (Suppl) A:11 – 19. [12] Foulds G, Shepard RM, Johnson RB. The pharmacokinetics of azithromycin in human serum and tissues. J Antimicrob Chemother 1990;25 (Suppl A):73 – 82. [13] Krohn K. Gynaecological tissue levels of azithromycin. Eur J Clin Microbiol Infect Dis 1991;10:864 – 8. [14] Clarithromycin Product Information, Abbott, 1997. [15] Amsden GW, Ballow CH, Forrest A. Comparison of the plasma, urine and blister fluid pharmacokinetics of clarithromycin and azithromycin in normal subjects. Clin Drug Invest 1997;13:152– 61. [16] Ishiguro M, Koga H, Kohno S, Hayashi T, Yamaguchi K, Hirota M. Penetration of macrolides into human polymorphonuclear leucocytes. J Antimicrob Chemother 1989;24:719– 29. [17] Guggenbichler JP, Kastner U. The influence of macrolide antibiotics on the normal flora. Infect Med 1998;15 (Suppl A):15–22. [18] Hyatt JM, Schentag JJ. A comparison of resistance emergence rates in Streptococcus pneumoniae following exposure to subinhibitory or higher concentrations of azithromycin, ampicillin or cefaclor. Presented at the 4th International Conference on the Macrolides, Azalides, Streptogramins and Ketolides, Abstract 3.14. Barcelona, Spain, January 21 – 23 1998. [19] Retsema JA, Girard AK, Brennan LA, Cimochowski CR, Faiella JA. Lack of emergence of significant resistance in vitro and in vivo to the new azalide antibiotic azithromycin. Eur J Clin Microbiol Infect Dis 1991;10:843 – 6. [20] Jackson MA, Burry VF, Olson LC, Duthie SE, Kearns GL. Breakthrough sepsis in macrolide-resistant pneumococcal infection. Pediatr Infect Dis J 1996;15:1049 – 51. [21] Bright GM, Nagel A, Bordner K, et al. Synthesis and in vitro and in vivo activity of novel 9-deoxo-9a-aza-9a-homoerythromycin A derivatives: a new class of macrolide antibiotics, the azalides. J Antibiot 1988;41:1029 – 47. [22] Craig WA. Postantibiotic effects and the dosing of macrolides, azalides, and streptogramins. In: Zinner SH, Young LS, Acar JF, Neu HC, editors. Expanding Indications for the New Macrolides, Azalides, and Streptogramins. NewYork: Marcel Dekker 1997:27 – 38. [23] Goujard C, Lebrun L, Doucet-Populaire F, Vincent V, Nordmann P. Clarithromycin-resistant Mycobacterium a6ium strain in a clarithromycin-naive AIDS patient. Clin Infect Dis 1998;26:186.
S14
G.W. Amsden / International Journal of Antimicrobial Agents 11 Suppl. 1 (1999) S7 – S14
[24] Amsden GW, Peloquin CA, Beming SE. The role of advanced generation macrolides in the prophylaxis and treatment of Mycobacterium a6ium complex (MAC) infections. Drugs 1997;54:69 – 80. [25] Leclercq R, Courvalin P. Bacterial resistance to macrolide, lincosamide, and streptogramin antibiotics by target modification. Antimicrob Agents Chemother 1991;35:1267–72. [26] Leclercq R, Courvalin P. Intrinsic and unusual resistance to macrolide, lincosamide, and streptogramin antibiotics in bact-
.
eria. Antimicrob Agents Chemother 1991;35:1273 – 6. [27] Tait-Kamradt A, Clancy J, Cronan M, et al. mefE is necessary for the erythromycin-resistant M phenotype in Streptococcus pneumoniae. Antimicrob Agents Chemother 1991;41:2251–5. [28] Gerardo SH, Citron DM, Claros MC, Goldstein EJC. Comparison of Etest to broth microdilution method for testing Streptococcus pneumoniae susceptibility to levofloxacin and three macrolides. Antimicrob Agents Chemother 1996;40:2413– 5.