Review article

Clarithromycin against Mycobacterium avium complex infections

L.B. Heifets

National Jewish Center for Immunology and Respiratory Medicine and Department of Microbiology, University of Colorado Health Sciences Center

Source: Tuber Lung Dis 1996 Feb;77(1):19-26

Reproduced with the kind permission of the author

Summary

The turning point in antimicrobial therapy of Mycobacterium avium infections came with the development of two new macrolides, clarithromycin and azithromycin. Controlled clinical trials, the first ever conducted with any agent among patients with M. avium infection, indicated the high efficiency of clarithromycin, in either acquired immune deficiency syndrome (AIDS) patients having a disseminated infection or non-AIDS patients with localized pulmonary disease. Monotherapy with clarithromycin resulted in elimination of bacteremia in almost all patients with disseminated infection, which is inevitably followed by a relapse of bacteremia in patients who survived long enough to reach this event. The strains susceptible to clarithromycin isolated before therapy contained 10^-8 or 10^-9 resistant mutants, and the relapses of bacteremia were caused by multiplication of these pre-existing mutants. Clarithromycin-resistance was associated with a mutation in the 23S rRNA gene. Cross-resistance between clarithromycin and azithromycin was confirmed with laboratory mutants and clinical isolates. At least two methods for determining the susceptibility of the M. avium isolates to clarithromycin are available: one is minimum inhibitory concentration (MIC) determination on Mueller-Hinton agar (pH 7.4) supplemented with 10% Oleic acid-albumin-dextrose catalase, the other is MIC determination in 7H12 broth, also at pH 7.4. The breakpoints for 'susceptible' for these methods are <= 8.0 µg/ml and <= 2.0 µg/ml, respectively. The breakpoints for 'resistant' are > 128 µg/ml for the agar method and >32.0 µg/ml for the broth method. The predictability value of MIC determination was confirmed by comparing the test results with the patients' clinical and bacteriological response to therapy. The remaining major problem in the therapy of the M. avium infections is a selection of companion drugs to be used in combination with clarithromycin (or azithromycin) to prevent the emergence of the macrolide-resistance. A number of clinical trials are now in progress to find a solution to this problem.

Introduction

Mycobacterium avium complex (MAC or MAI) includes M. avium and M. intracellulare and is the most frequent clinical isolate among more than 25 non-tuberculous mycobacterial species that can be found in human specimens. It is ubiquitous in the environment ^1-3, and the infections caused by these organisms are not transmissible from person to person. Before the acquired immune deficiency syndrome (AIDS) epidemic, the MAC infection most often manifested in humans was a localized pulmonary disease^3-7 which occurred in the USA in approximately 2000 patients annually. This contrasts sharply with the 24,000 to 39,000 patients with AIDS, who during 1991-1992 had disseminated MAC infection, a cumulative incidence of 15 to 24%, but the probability of this infection can be much higher in patients with CD4 lymphocyte counts less than 100 or 50 per mm.^3,8-12 The proportions of M. avium and M. intracellulare in patients with pulmonary disease reflect the distribution of these species in the environment of different geographic areas.^5,12-15 Most of the severe cases of pulmonary disease referred to the National Jewish Center in Denver were caused by M. intracellulare.¹⁶ This 'selection' of patients who usually failed to respond to initial chemotherapy corresponds with the suggestion¹⁷ that M. intracellulare is more resistant than M. avium to many of the conventional antituberculosis drugs. At the same time, isolates from AIDS patients with disseminated infection were almost exclusively M. avium^18,19. These differences between AIDS and non-AIDS patients should be taken into account when analyzing the effectiveness of treatment of the MAC infections in these two categories of patients. Antimicrobial therapy of pulmonary MAC disease often imitated drug regimens recommended for M. tuberculosis, for example isoniazid+streptomycin+ PZA, and usually resulted in failure.^20,21 Retrospective analysis of treatment regimens that included rifampin and ethambutol have indicated some success.^7,22-24 Diversity in views and uncertainty in selection of appropriate treatment regimens have been addressed in a number of reviews on the subject.^7,10,25-31

The turning point in antimicrobial therapy of MAC infections came with the introduction of clarithromycin and azithiomycin and their evaluation in controlled clinical trials in patients with disseminated M. avium infection. The literature about clarithromycin, its general properties, pharmacology, and activity against other bacteria is extensive, and has been addressed in a number of reviews.^32-39 Therefore, this review has been limited to only one issue: the antimicrobial activity of this macrolide against MAC in various experimental conditions, with emphasis on correlation of these data with its efficacy in patients.

Efficacy in patients

The first clarithromycin clinical trial was conducted in1988-1989 in France.⁴⁰ It was the first double-blind placebo-controlled randomized trial ever performed in patients with M. avium infection. During the first 6 weeks of this pilot study (Phase I), eight patients received 1000 mg twice a day, while 7 patients received placebo. During the second 6 week period of observation (Phase II), the patients were crossed over: eight patients of the first group started receiving placebo, and only three patients of the second group started receiving clarithromycin. During this second period of observation, patients of both groups were administered four additional antituberculosis drugs (isoniazid, rifampin, ethambutol, clofazimine). The number of colony forming units (CFU)/ml in blood was increased in all patients receiving placebo during Phase I by a mean of 1.55 log₁₀, while the number of CFU/ml in those receiving clarithromycin decreased by a mean of 2.65 log₁₀, with complete eradication of bacteremia in seven of eight patients. During Phase II the relapse of bacteremia or increase of the number of CFU/ml was observed in four of eight patients after they were crossed over from clarithromycin to placebo, while a substantial decrease in the number of CFU/ml occurred in three patients treated during Phase II with clarithromycin. This short-term trial provided the first clear evidence on the effectiveness of clarithromycin in the therapy of disseminated M. avium infection in patients with AIDS. It also demonstrated that the four-drug combination (isoniazid+ rifampin+ethambutol+clofazimine) was not effective. Similar results were obtained in another short-term randomized pilot trial in which four patients received clarithromycin and five patients were treated with isoniazid+ethambutol+clofazimine.⁴¹ These randomized pilot trials were followed by a prospective compassionate-use study to assess the long-term efficacy of clarithromycin in AIDS patients with disseminated M. avium infection.⁴² Eradication of MAC bacteremia was achieved in 11 of 16 patients receiving 500 or 1000 mg/day and in 45 of 46 patients receiving 1500 or 2000 mg of clarithromycin along with various combinations of conventional antimycobacterial agents. A relapse of bacteremia occurred in 14 of 56 patients, and in 9 of 14 the isolates became resistant to clarithromycin. The results of the subsequent observations conducted in France⁴³ have shown that the initial eradication of bacteremia occurred in 93% of 173 AIDS patients treated with clarithromycin in combination with other drugs. The resistance associated with relapses of bacteremia was registered in 23% of 136 patients after 2 to 7 months of therapy. It is suggested that the only companion drug which appeared to protect against emerging resistance to clarithromycin was ethambutol, resulting in a twofold decrease in the relapse rate.⁴³

The first large-scale randomized, double-blind dose-ranging trial with clarithromycin was conducted in the USA in 1991-1992.⁴⁴ The 154 AIDS patients were divided into three groups (53, 51, and 50 patients) to receive 500 mg, 1000 mg, or 2000 mg twice daily for 12 weeks. Clarithomycin monotherapy was effective in more than 99% of patients, resulting in elimination of the initial bacteremia, the symptoms associated with the MAC infection, and in a longer period of survival than could be expected for this type of patients. An average decline in the intensity of bacteremia was 2.7 log₁₀ of CFU/ml. Although the elimination of the initial bacteremia in patients receiving 500 mg occurred later than in the other two groups, the clinical and bacteriological response in general to this dose was the same or better than that seen in the other groups. The adverse reactions were highest among patients receiving 2000 mg (40%) compared to 16% and 15% in the other two groups. The proportion of patients who failed to complete 12-wk therapy was also the highest among patients receiving 2000 mg (58%) compared to 43% and 28% in groups treated with 1000 mg and 500 mg respectively. Based on these and other data, the dose 500 mg b.i.d. has been approved for therapy of disseminated M. avium infection in AIDS patients.

Another macrolide, azithromycin, has also shown efficacy in the treatment of disseminated M. avium infection in AIDS patients.⁴⁵ Although its reported ability to decrease the intensity of bacteremia was much lower than that of clarithromycin, an average of only 1.4 log₁₀, the conclusion as to which of these two macrolides is more effective can be reached only in a controlled, randomized comparison trial. Either of these macrolides has been recommended as a mandatory element in the therapy of M. avium infection in AIDS patients, in combination with other agents.⁴⁶

Clarithromycin monotherapy also appeared to be effective in non-AIDS patients with localized pulmonary disease caused by either M. intracellulare or M. avium. In a non-randomized trial of 19 patients who had pretreatment isolates susceptible to clarithromycin and completed a minimum of 4 months monotherapy (500 mg b.i.d.), 11 converted to negative sputum culture and only two of them relapsed with the emergence of clarithromycin-resistance.³¹ In addition, four other patients also had clearly favorable response to therapy with a substantial decline in the bacterial load in sputum. Besides these 15 patients with a definite response to therapy, two other patients were considered as indeterminant responders, and two were non-responders.

Bacteriological response to therapy and the issue of drug resistance

Emergence of resistance of the MAC isolates to clarithromycin (minimum inhibitory concentration ([MIC] > 32 µg/ml) was registered in 46% of 152 patients, and was always associated with recrudescence of symptoms and relapse of bacteremia.⁴⁴ This rate does not reflect the actual probability of these events, since many patients discontinued therapy before the emergence of drug resistance could take place. Only 39 patients in this trial received 500 mg or 1000 mg for a period of 150 days or longer. One patient did not respond to therapy, and therefore only 38 patients were included in a detailed analysis of the bacteriological response to the clarithromycin monotherapy.⁴⁷ The time required to achieve a negative culture or the lowest level of bacteremia depended on the intensity of the pretreatment bacteremia, ranging from 35.1 to 44.0 to 59.3 and 73.3 days for those whose CFUs per ml of blood were in a range <= 100, 101-1000, 1001-10000, and > 10000, respectively. The average period of remission was 76.7 days, the longest of which (101.3 days) occurred among those with the lowest number of CFU/ml before treatment and shortest (42.7 days) among those with the largest number of CFU/ml before therapy. The relapse of bacteremia took place in all 38 patients after 116 to 136 days of treatment, and it has always been associated with the emergence of resistance to clarithromycin. It can be concluded from this analysis that the emergence of drug resistance is inevitable if the period of clarithromycin monotherapy is long enough for these events to take place.

Multiple strains isolated during the period of decline of bacteremia were susceptible to 0.12 to 2.0 µg/ml, while those isolated during the relapse of bacteremia had an MIC of clarithromycin > 256.0 µg/ml when tested in 7H12 broth at pH 7.3-7.4. Direct susceptibility testing with blood specimens obtained during the relapse showed that up to 100% of the bacterial population consisted of these highly resistant bacteria, indicating that the susceptible population was eliminated during the initial phase of therapy.⁴⁷ Therefore, we concluded that the emergence of resistance was the result of multiplication of pre-existing resistant mutants, found with a frequency of 10^-8 and 10^-9 in M. avium strains susceptible to clarithromycin. The origin of the resistant strains isolated during the relapse of bacteremia was confirmed as being from the initial population by the restriction fragment length polymorphism testing of both in two laboratories.^48,49 These clarithromycin-resistant isolates were not different in their drug susceptibility/resistance pattern to the conventional antimycobacterial drugs, but showed clear cross-resistance to azithromycin.⁴⁷ Based on these data, it is speculated in this report that resistance to these macrolides may have a different mechanism than resistance of MAC to the conventional antimycobacterial drugs, which is usually associated with cell wall permeability. Resistance to clarithromycin in M. intracellulare has been identified as being associated with a mutation in the 23S rRNA position adenine 2058.⁵⁰ Resistance in M. avium isolated from the 38 patients was also found to be related to the mutation in the gene coding in 23S rRNA positions A2058/2059.⁵¹

In vitro susceptibility tests

The activity of clarithromycin against M. avium has been evaluated in a number of studies.^52,63 These publications, as well as a number of presentations at scientific meetings, have shown that the activity of clarithromycin in vitro depends on the type of medium (agar or broth) and pH conditions. The MICs in agar media were four- to 16-fold greater than the broth determined MICs.^58,59 The MICs at pH 7.4 were two-to eight-fold lower than at pH 6.6-6.8 either in agar media⁶³ or in broth.^58,62 The broth-determined MICs were 1.0 µg or less at pH 7.4, 1.0-4.0 at pH 6.8, and 8.0 and greater at pH 5.0 These data correspond with the fact that in general macrolides are more potent at a basic than at acidic pH against any organisms.^64,65 Two types of media have been recommended for determining susceptibility to clarithromycin of the MAC clinical isolates: Mueller-Hinton pH 7.4 agar supplemented with 10% OADC and 7H12 broth adjusted to pH 7.4.^40,44,58,63 The results of the previously described large-scale randomized trial showed that all patients whose isolates were susceptible to 2.0 µg/ml or less when tested in 7H12 broth at pH 7.4 radiometrically (BACTEC) responded favorably to the treatment.^44,47 On the other hand, the MICs for all isolates obtained during the relapse of bacteremia were 256.0 µg/ml or greater, and these patients did not respond to continued therapy with clarithromycin. Therefore we have suggested⁴⁷ broth-determined MIC of 2.0 µg/ml as the breakpoint for 'susceptible,' which is the same as that established for most other bacterial species by the National Committee for Clinical Laboratory Standards.⁶⁶ The breakpoint for 'resistant' could be >= 256.0 µg/ml, but for the convenience of testing we suggested classifying as resistant the MAC isolates with MIC > 32.0 µg/ml. Only two isolates obtained at the end of the decline of bacteremia from two of 154 patients had an MIC of 8.0 µg/ml; an MIC of 8.0 µg/ml was therefore suggested to classify such rare isolates as 'intermediate' or 'moderately susceptible.⁴⁷ The agar-determined MICs for strains susceptible to clarithromycin were not greater than 8.0 µg/ml,⁵⁸ and this concentration can be suggested as a breakpoint for 'susceptible' using this method. To classify an isolate resistant to clarithromycin in the agar test, a concentration of > 128.0 µg/ml would be equivalent to >32.0 µg/ml in broth. An intermediate concentration for the agar method would then be 32.0 µg/ml. These and other suggestions need to be evaluated for reproducibility of the results in a multi-laboratory study, which is currently in progress.

Activity against intracellular bacteria

It was shown that clarithromycin accumulates in polymorphonuclear leucocytes, in various tissues, and in macrophages.^67-70 The intracellular/extracellular ratio was 9.2 and 16.7 in polymorphonuclear leucocytes.^67,70 The drug concentrations in infected macrophages exceeded the extracellular contents by 17-fold.⁷⁰ Quantitative assessment of the antimicrobial activity of clarithromycin has shown that for extracellular bacteria the MICs ranged from 0.25-1.0 and minimum bactericidal concentrations (MBCs) from 16.0-64.0 µg/ml; for intracellular bacteria the MICs were 0.5-1.0 µg/ml and MBCs 16.0 or 32.0 µg/ml.⁷¹ while some of the intracellular bacteria may reside in phagosomes at neutral pH, a substantial proportion of the intracellular population may reside in phagolysosomes;⁷² the activity of the drug in an acidic environment (pH 5.0) can be significantly diminished, with MICs of up to 16.0 µg/ml and greater.⁵⁸ The balance between the accumulation of the drug in macrophages and its diminished activity in acidic conditions resulted in the fact that complete inhibition of intracellular multiplication would require a minimum of 1.0 µg/ml to be present in extracellular fluid, which is the same as for extracellular bacteria at pH 7.4. Most of the other studies on the activity of clarithromycin against M. avium residing in macrophages have been performed with a drug concentration equivalent to the peak concentration in serum (C_max), usually 4.0 µg/ml, to which the infected macrophages were exposed continuously for 7 days,^56,73-76 resulting in a decline in the number of bacteria in some of the observations. A 4-week continuous exposure of infected macrophages to 2.0 µg/ml also resulted in a substantial decline in the number of viable bacteria.⁷⁷ Apparently such models do not reflect the in vivo events, since the exposure to C_max lasts for only a short period. The potential effect of C_max was evaluated by a 2 h pulsed exposure of the infected macrophages to 3.0 µg/ml,⁷⁸ producing complete inhibition of bacterial multiplication for 4 days, but no bactericidal effect was determined. Macrolide antibiotics are generally not considered bactericidal, and despite some suggestions about the bactericidal activity of clarithromycin^59,75,76 the overall assessment is that this agent produces only an inhibitory effect against M. avium at concentrations attainable in vivo.^{34,58,60,73,79}

Mouse Model

Clarithromycin has shown higher activity than any other agent tested against MAC infection in beige mice, resulting in a dramatic (up to 2 log₁₀ or more) decrease in the number of bacteria in the lung, spleen, liver and bone marrow.^52,56,80,51 This decline in the bacterial load has at times been classified as 'bactericidal' and at times as 'sterilizing.' The term 'sterilizing' is probably more appropriate for in vivo events, since it is not necessarily related to direct killing. The effectiveness of clarithromycin in beige mice was clearly dose-dependent, and dependent also on the duration of treatment.^80,81 Besides the sterilizing activity, another similarity to the events that took place in patients was emergence of resistance to MAC after prolonged monotherapy.⁸² These observations confirmed the validity of the beige mouse model for testing of the potential activity of antimicrobial agents against MAC infection.

Companion drugs

The events that took place in patients with disseminated M. avium infection treated with clarithromycin for a prolonged period of time have a striking similarity to the 'fall and rise phenomenon' described in the early years of monotherapy of tuberculosis with streptomycin or isoniazid.^83-86 These data led to the principle of multi-drug therapy of tuberculosis with the initial purpose of preventing the emergence of drug resistance. It is very likely that the same approach will work in preventing clarithromycin resistance in patients with MAC infections. In searching for effective companion drugs, various combinations were tested in vitro, in macrophages, in mice and in limited clinical observations. Despite claims by some authors about a synergistic interaction between clarithromycin and other agents, the overall assessment of these interactions in vitro and in macrophages has shown only an additive effect.^{59,75,77,81,87-89} The beige mice models have been used to justify the use of amikacin in combination with other drugs.^90-94 Comparison of the activity of various agents including clarithromycin in the beige mouse model has shown that amikacin was almost as active as clarithromycin, while rifampin and rifabutin were totally inactive, and ethambutol and spafflokacin exhibited a limited inhibitory effect.⁸⁰ The probability of emerging resistance to clarithromycin decreased when mice were treated with clarithromycin+ amikacin, but not with a combination of clarithromycin and minocycline.

Observations in small groups of patients on the efficacy of various drugs that could be considered candidates for combination with clarithromycin have shown a broad diversity in results and interpretations.^{25,26,90,95-99} Ethambutol reduced the intensity of MAC bacteremia in AIDS patients by only about 0.6 log₁₀ during one month of treatment, and a combination of this agent with clofazimine and rifampin resulted in a high rate of toxic side effects.¹⁰⁰ In the same trial, rifampin monotherapy was not effective at all. The clinical efficacy of rifabutin has been evaluated in combination with other drugs, showing various percentages of patients with conversion to negative blood culture.^{11,95,101-104} A decrease in the intensity of bacteremia by 1.4-1.5 log₁₀ or conversion to negative blood cultures was reported to occur in small groups of patients treated with a combination of rifampin+ethambutol+ciprofloxacin+amikacin,⁹⁰ or rifampin+ethambutol+ciprofloxacin+clofazimine.¹⁰⁵ These mostly non-randomized trials have not provided clear conclusions as to the choices to be made, and now a number of controlled clinical trials to test the effect of various drug combinations with clarithromycin are in progress. As stated above, the relapse of bacteremia associated with emerging drug resistance occurred in all patients who were on clarithromycin monotherapy over a prolonged period. It can be suggested therefore that detection of a treatment regimen effectively preventing relapse of bacteremia will require a prolonged period of observation of not less than 4-5 months.

Acknowledgements

Our studies summarized in this review have been supported by grants from Abbott Laboratories (Abbott Park, IL). I thank P. Lindholm-Levy for editorial help and C. Queen for preparation of the manuscript.

Correspondence to:
L. B. Heifets MD
National Jewish Center,
1400 Jackson St, Denver,
CO 80206,
USA
Tel: +303 398-1384;
Fax: +303 398-1953;
Internet: heifetsl@njc.org.

References