MAC Treatment: State of the Art

Source: Chaisson RC, et al. Clarithromycin and ethambutol with or without clofazimine for the treatment of bacteremic Mycobacterium avium complex disease in patients with HIV infection. AIDS 1997;11:311-317.

Chaisson and colleagues at 22 centers in the United States randomly assigned 106 adults with HIV infection and MAC bacteremia to receive a combination of clarithromycin (500 mg twice daily) plus ethambutol (800-1000 mg once daily based on weight) with or without clofazimine (100 mg once daily) for six months. Patients assigned to the three-drug combination were more likely to be black or Hispanic and were more likely to have higher baseline MAC colony counts (3.3log10 cfu/mL) than those receiving two drugs (2.2log10 cfu/mL; P = < 0.001).

Nevertheless, the clinical and bacteriological responses within the two arms were similar, with one important exception: a higher rate of mortality was observed in patients receiving the three-drug combination compared with those receiving two agents. A total of 21/55 (38%) vs. 31/51 (61%) patients receiving two vs. three drugs died during the study (P = 0.032). In Kaplan Meier analysis, the median time to death was significantly decreased in patients receiving three agents compared with those receiving two (P = 0.012). Multivariate analysis revealed that a higher level of mycobacteremia at baseline and assignment to clofazimine were independent predictors of mortality.

The bacteriological and clinical response was otherwise comparable between the two treatment arms: 65% of patients receiving two agents vs. 54% of those receiving three agents became blood culture negative (P = NS). The median time to complete suppression of mycobacteremia was 58 days for patients receiving two drugs and 63 days for patients receiving three drugs (P = 0.40, log rank). The rate of clinical response was also similar.


Despite the recent completion of several large-scale MAC treatment trials, it’s safe to say that the optimal antimycobacterial regimen for the treatment of MAC bacteremia in AIDS has not been determined. The best regimen currently appears to be a combination of clarithromycin and ethambutol, which results in initial clearance of the organism from the bloodstream in approximately 65-74% of patients within 16 weeks. Up to 68% of these patients will subsequently fail therapy, typically after the first four months of therapy, largely due to the emergence of clarithromycin-resistant organisms. Whether the addition of a third agent to the combination of clarithromycin and ethambutol is of benefit, and which agent should be selected, remains to be determined, although clofazimine no longer appears to be a valuable choice and may, in fact, result in increased mortality.

The recent MAC treatment studies have yielded some interesting observations:

1. The rates of clearance of bacteremia in all treatment arms containing clarithromycin in each of the recent trials appear comparable.

2. The addition of a second agent to clarithromycin therapy does not necessarily result in a faster rate of clearance of bacteremia but does result in its more durable suppression.

3. The combination of clarithromycin and ethambutol has been central to the most effective (e.g., most durable) combination regimens studied to date.

4. None of the existing data provides a basis for the use of clofazimine in the treatment of MAC, and its use may result in increased mortality.

5. Whether a third agent—and which of the remaining agents—can enhance the durability of the response has not been determined.

Chaisson and colleagues found that in patients who received clarithromycin and ethambutol with or without clofazimine for six months,1 no apparent difference between the two treatment regimens could be detected with regard to the change in mycobacterial colony counts, the proportion of patients who became culture negative, the time to culture negativity, nor in the reduction of clinical symptoms.

Comparable rates of bloodstream clearance have been seen in two other studies. In the Canadian HIV Trials Network study of 229 patients randomly assigned to one of two regimens (either clarithromycin, ethambutol, and rifabutin or ethambutol, clofazimine, rifampin, and ciprofloxacin), 69% of those receiving the clarithromycin-containing regimen vs. 29% receiving the non-clarithromycin-containing regimen were blood culture-negative at three months.2 The California Collaborative Treatment Group randomly assigned 100 patients a higher dose of clarithromycin (1000 mg twice daily) plus clofazimine (100 mg once daily) with or without ethambutol (800 mg once daily).3,4 During the first three months of therapy, up to 77% and 74% of those receiving three vs. two drugs became culture negative, respectively. The median time to bacteriologic "response" (defined as at least a 2.0log10 decrease or negative cultures in at least 2 consecutive blood cultures) was four weeks in both groups. (Approximately one-half of the patients in each arm decreased their dose of clarithromycin to 500 mg daily because of intolerance.)

The rates of complete suppression of bacteremia for each of these three studies for all arms containing clarithromycin were comparable, irrespective of whether patients received one or two additional agents. There is, therefore, no evidence that adding a drug or drugs to clarithromycin significantly hastens sterilization of blood cultures. Thus, given the current therapeutic armamentarium, the kinetics of mycobacterial clearance are dependent on the activity of clarithromycin, and the time to clearance of bacteremia will be comparable for all clarithromycin-containing regimens.

This is likely due to the substantially lower anti-MAC activity of the other agents. Clarithromycin, and possibly azithromycin, are currently the most active agents against MAC, resulting in the complete suppression of bacteremia in 45-65% of patients within 4-6 weeks. We have demonstrated that ethambutol, when administered as a single agent, resulted in a statistically significant reduction of bacteremia of 0.6 logs within four weeks in AIDS patients with MAC bacteremia.5 In the same study, clofazimine and rifampin were at best bacteriostatic, having no significant effect on the level of mycobacteremia.

This is no doubt an over-simplification, but the aim of therapy (and of treatment studies) should be the enhancement not of just the initial rate of response, but also of the durability of that response. The ability to eliminate tissue infection, and hence prevent recrudescent infection, most likely depends on other factors—the total amount of tissue infection, the susceptibility pattern of the strain (or strains), the antimycobacterial activity of the combination therapy, the ability to achieve good tissue concentrations of drug, as well as host immune factors—but essentially boils down to one thing: the elimination of replicative tissue infection before the emergence of resistant clones of mycobacteria.

So, what happened after the initial period of suppression in these trials? In the Chaisson study, five patients in each arm (16% and 23% for the 2- vs 3-drug regimens) had recrudescent bacteremia within the study period, at least one of which was due to clarithromycin-resistant organisms.1 In the CCTG study, the rate of relapse was determined only for those patients who initially "responded" to therapy.3,4 For patients who received clarithromycin, ethambutol, and clofazimine, the estimated risk of relapse was 5% and 12% at 24 and 36 weeks, respectively. In contrast, for patients who received clarithromycin and clofazimine, the rate of relapse was 22% and 68% at 24 and 36 weeks, respectively (P = 0.004). All of the isolates were resistant to clarithromycin. The median time to the development of clarithromycin resistance was 16 weeks in patients receiving clarithromycin plus clofazimine but was 40 weeks in those receiving all three agents. The median survival was 31 weeks vs. 39 weeks for patients receiving two vs. three drugs (P = NS).

At first glance, one might conclude from the CCTG data that the use of three agents is superior to two. Look again. If one concludes that clofazimine contributed little to the treatment of MAC in the Chaisson study,1 and had no effect on bacterial colony counts in the single agent study,5 then the relative success of the three drug regimen in the CCTG study was likely due to the addition of ethambutol.

In fact, the initial response to treatment and the subsequent rate of failure for patients who received clarithromycin and clofazimine in the CCTG study in many ways mirrors that observed when clarithromycin was used as a single agent. In a randomized dose ranging study of clarithromycin (ACTG 0157), the median time to clearance of bacteremia was 43 days in patients receiving clarithromycin 1000 mg twice daily.6 Recrudescence of bacteremia, associated with the development of high-level resistance to clarithromycin, occurred in 46% of patients. The median time to development of clarithromycin resistance, which was dependent on the baseline colony count at entry to study, varied between 14 and 19 weeks.

Furthermore, the finding of a higher rate of mortality with the administration of clofazimine in the Chaisson study cannot be easily dismissed, although a biological basis for this effect is not readily apparent. Both higher baseline colony counts and the use of clofazimine were independently associated with a higher risk of death. That higher colony counts may contribute to lower survival is not unexpected and has been shown in other studies.7 Increased mortality was also seen when clofazimine was used in the Shafrin study,2 although this may have been more a function of the lack of clarithromycin in the regimen. The CCTG results are no less reassuring; although survival rates were not significantly different between the two treatment arms, clofazimine was used in both.

Examination of the relationship of mortality to the initial level of mycobacteremia is revealing in the study by Chaisson et al: a statistically significant difference in mortality was observed for patients whose initial level of bacteremia was between 100/mL and 1000/mL and who received two agents (31%) vs. those who received three agents (75%) (P = 0.05).1 No significant difference in mortality was observed between the two treatment arms in those patients whose initial level of mycobacteremia was less than 100/mL or greater than 1000/mL. In other words, clofazimine did not appear to affect the rate of mortality in patients with very low or very high initial levels of mycobacteremia. Multivariate analysis revealed, however, that these two variables (administration of clofazimine and high initial colony counts) were independent predictors of mortality.

Clofazimine has known anti-inflammatory properties and has been used in a number of chronic autoimmune skin disorders. Clofazimine inhibits in vitro lymphocyte response to mitogens.8 A recent paper suggests that it may suppress alloreactive T-cell function in chronic graft-versus-host-disease (GVHD).9 Twenty-two patients with chronic GVHD received clofazimine 300 mg orally once daily for three months and then 100 mg once daily. More than 50% had a complete or partial response in their clinical symptoms, and one-third were able to reduce their dose of other immunosuppressive medication. The purported effect on T-cell function has obvious implications for patients with HIV and deserves further study.

In the study by Chaisson et al, the duration of survival also appeared to be substantially improved in patients who continued to receive antiretroviral therapy: the median survival was 29 weeks in patients who discontinued antiretroviral therapy vs. greater than 66 weeks in those who continued it (P = 0.07).1 The use of highly active antiretroviral therapy in patients with MAC bacteremia is likely very important to the overall response to antimycobacterial therapy.

Finally, no regimen works well if it’s poorly tolerated. In each of these treatment studies, 10-16% of the patients had treatment-limiting toxicity. Finding means to enhance patient tolerability is key to the success of the regimen, especially during the first 2-3 weeks of therapy when patients are most symptomatic from their MAC disease.

What happens to the other 15-35% of patients who apparently tolerate the regimen but don’t respond? Lack of adherence to the regimen or a lack of absorption may be responsible for many of these failures. Initial pre-treatment resistance to clarithromycin has not yet been an issue, although it may become one in the future.

At present, a combination of clarithromycin and ethambutol should be considered the core of any treatment regimen. Based on theoretical considerations, the use of a third agent should still be considered. Available options include rifabutin, amikacin, streptomycin, and a quinolone. The L-isomer of ofloxacin, levofloxacin, is highly active against MAC and may be a better choice than either ofloxacin or ciprofloxacin. Azithromycin may be an acceptable alternative to clarithromycin, although treatment data are limited, and its use in non-HIV infected patients with pulmonary MAC was associated with a 41% incidence of ototoxicity.10

It’s apparent that the identification of at least one additional agent active against MAC is crucial to our control of this infection. We are currently performing a Phase I study to examine the safety of liposomal-encapsulated amikacin. In addition, immune-based therapies, such as the addition of GM-CSF, 11,12 may provide some relief, especially in those with recrudescent infection due to clarithromycin-resistant organisms.


1. Chaisson RC, et al. AIDS 1997;11:311-317.

2. Shafran SD, et al. NEJM 1996;335:377-383.

3. Dube MP, et al. A randomized study of clarithromycin plus clofazimine, with or without ethambutol, for treatment and prevention of relapse of disseminated MAC in AIDS [abstract I201]. In: Program and abstracts of the 35th Interscience Conference on Antimicrobial Agents and Chemotherapy (San Francisco, CA). Washington, DC: American Society for Microbiology, 1995:241.

4. Dube M, et al. Prevention of relapse of MAC bacteremia in AIDS: A randomized study of clarithromycin plus clofazimine, with or without ethambutol [abstract 206]. The 3rd Conference on Retroviruses and Opportunistic Infections(Washington DC). Infectious Disease Society of America, 1996:91.

5. Kemper CA, et al. J Infect Dis 1994;170:157-164.

6. Chaisson RE, et al. Ann Intern Med 1994;121:905-911.

7. Kemper CA, et al. J Infect Dis 1994;170:488-493.

8. Van Rensburg CEJ, et al. Antimicrob Agents Chemother 1982;21:693.

9. Lee BSJ, et al. Blood 1997; 89:2298-2302.

10. Griffith DE, et al. Clin Infect Dis 1996;23:983-989.

11. Bermudez LE, et al. Biotherapy 1995;4:3-10.

12. C. Kemper, et al. Immunomodulatory therapy of Mycobacterium avium (MAC) bacteremia in AIDS with rhGMCSF [abstract G109]. In: Program and abstracts of the 35th Interscience Conference on Antimicrobial Agents and Chemotherapy (San Francisco, CA). Washington, DC: American Society for Microbiology, 1995:177.