New Developments in the Management of MAC Infections

Better Prophylaxis -- and Better Treatment -- Are Now Possible

In patients with advanced HIV disease, few opportunistic infections are more worrisome than disseminated Mycobacterium avium complex. MAC is one of the most common complications of AIDS, occurring in 22% of adults with C.D.C.-defined AIDS (1) and up to 400/0 of those with end-stage disease.(2) Appearing late in the course of HIV infection,(3) MAC bacteremia produces a constellation of vague and debilitating symptoms -- chief among them persistent fevers and night sweats, fatigue, and significant weight loss.(4, 5)

As every clinician who treats HIV-infected patients knows, a number of AIDS-related opportunistic infections, including CMV and PCP, present with some combination of these same vague symptoms -- which can make the diagnosis of disseminated MAC difficult. To assist practitioners in meeting this diagnostic challenge, the editors of HIV Newsline have prepared a Pull Out and Save diagnostic algorithm, "Diagnosing and Treating Suspected MAC Disease," in this issue.

Morbidity and mortality are both high in patients who develop M. avium infection. In a 1992 study of 1,006 AIDS patients who received no prophylaxis for MAC, the incidence of disseminated bacteremia was 21% at 12 months and 41% at 24 months.(3) The median survival time in patients who developed MAC was 134 days, and the survival rate at one year was less than 20%.

Although there have been significant advances since 1992 in both the prevention and treatment of MAC, the incidence of this late-stage opportunistic infection is rising.(5, 6) Much of this increase is attributable to improved prophylaxis against other common OIs, foremost among them PCP (see "Update: Pneumocystis carinii Pneumonia," in this issue). Indeed, disseminated MAC infection was the most common cause of OI-related death in ACTG 021, a clinical trial that followed 310 AIDS patients who were receiving secondary prophylaxis against recurrent PCP.(7)

"Once MAC infection has become established, it is difficult to control -- and often impossible to eradicate -- with any combination of the drugs currently used to treat M. avium bacteremia."

Because the acid-fast bacilli that cause MAC are ubiquitous environmental organisms, commonly found in water and soil, these pathogens are almost impossible to avoid. Precautions such as drinking only boiled, filtered, or distilled water -- and carefully washing all fresh fruits and vegetables in such treated water -- will reduce the likelihood of exposure to cryptosporidiosis (see the other Pull Out and Save feature in this issue, "Avoiding Cryptosporidiosis: What the HIV- Positive Patient Needs to Know"). There is no evidence that these measures are likely to prevent exposure to MAC, however.

MAC bacilli can colonize the GI tract and respiratory tree of HIV-infected individuals, which may lead to episodic bacteremia. A popular current hypothesis is that as immune function declines, these intermittent infections evolve into sustained bacteremia and, finally, widespread invasion of tissue, especially reticular endothelial tissue.(8) Once MAC infection has become established in an immunocompromised host, it is difficult to control -- and often impossible to eradicate -- with any combination of the drugs that are currently used to treat M. avium bacteremia.

Current recommendations for MAC prophylaxis

Several years ago the U.S. Public Health Service recommended prophylaxis with rifabutin, at 300 mg/day, to prevent disseminated MAC infections in HIV-positive individuals with CD4 counts below 100 cells/mm3.(9) Two large, well-designed, parallel, double-blind, placebo-controlled clinical trials -- which ultimately enrolled 1,146 subjects with CD4 counts below 200 cells/mm3 and followed them with monthly AFB blood cultures -- have demonstrated that daily doses of rifabutin halve the risk of MAC infection in this patient population(10) (see "Recent Advances in MAC Prophylaxis," Vol. I, No. 3, pages 52-57).

In these paired trials -- one conducted in Canada, the other in the United States -- rifabutin prophylaxis was associated with significant improvements in a number of clinical parameters: fever, fatigue, alkaline phosphate levels, and Karnofsky score. These findings led the ED.A. to approve the use of rifabutin as prophylaxis against AIDS-related MAC infection.

Unfortunately, these two clinical trials did not demonstrate a survival benefit for rifabutin prophylaxis. This unanticipated and somewhat disconcerting finding may be attributable to the fact that prophylaxis against other OIs was not uniform at the participating centers -- which made it difficult for investigators to gauge the degrees to which PCP, CMV, toxoplasmosis, and invasive fungal diseases may have contributed to the deaths of rifabutin-treated patients.

Moreover, investigators who participated in these trials recorded a number of significant drug-drug interactions between rifabutin and other medications commonly taken by patients with advanced HIV disease. Rifabutin is a potent inducer of the cytochrome P-450 system, and when the drug is taken in combination with drugs that are also metabolized by hepatic enzymes, notably clarithromycin and zidovudine, serum levels of these other drugs drop significantly.

Fluconazole, by contrast, is a potent inhibitor of the P-450 system, and consequently it delays metabolism of rifabutin -- thereby increasing serum levels of the latter drug.(11) Protease inhibitors also increase serum levels of rifabutin. For that reason, patients taking indinavir should reduce their rifabutin dose by half, and patients taking ritonavir should reduce the dose by 75% or avoid rifabutin altogether.(12, 13)

Another probable consequence of the drug-drug interactions associated with rifabutin therapy is the incidence of acute uveitis reported in AIDS patients treated with rifabutin in combination with clarithromycin and ethambutol, especially when fluconazole is also used.(14)

In patients with late-stage HIV disease, many of whom are taking one or more of these drugs, the potential for drug-drug interactions with rifabutin -- and the necessity of adjusting dosages to compensate for those interactions -- complicate an already complicated clinical picture. It is hardly surprising, then, that the incidence of rifabutin-induced drug-drug interactions -- coupled with the absence of a demonstrable survival benefit to rifabutin prophylaxis -- prompted researchers to investigate the safety and efficacy of other drugs as prophylaxis against disseminated MAC infection.

New approaches to MAC prophylaxis: Clarithromycin

Two new macrolide agents, clarithromycin and azithromycin, have shown good in vitro and in vive activity against M. avium(15-18) -- which makes them logical candidates for MAC prophylaxis Clarithromycin is the more active against M. avium,(17) and was the first macrolide to be approved by the ED.A. for the treatment of MAC. As such, it is a mainstay of combination therapy for MAC bacteremia.

A recently completed clinical trial assessed the effectiveness of clarithromycin monotherapy as prophylaxis against MAC in patients with advanced HIV disease.(19) This prospective, double-blind, and placebo-controlled study began enrolling subjects at 66 centers in the United States and Europe in November of 1992. By July of the following year, when enrollment closed, the investigators had randomized 341 patients to active treatment (500 mg of clarithromycin b.i.d.) and a like number to a look-alike placebo. All 682 study subjects underwent baseline and regular follow-up clinical examinations for signs and symptoms of disseminated MAC infection and blood cultures for the presence of M. avium. Fourteen members of the original study cohort were later found to have positive pretreatment cultures, and they were excluded from analysis.

The 333 patients who remained in the treatment arm of the study were followed for a mean of 427 days; the 334 patients in the no-treatment arm, for a mean of 402 days. Drop-out rates were similar in both arms: 18% of the patients receiving clarithromycin, and 17% of those receiving placebo, withdrew from the study because of adverse events. In each cohort 12% of the original participants were lost to follow-up.

Disseminated MAC infections developed in 19 of the 333 patients who received prophylactic clarithromycin (6%), and in 53 of the 334 patients who received the placebo (16%). In both arms infection was most likely to occur in patients with the lowest CD4 counts. Among treated patients, for example, 11 of the 19 cases of MAC occurred in subjects with CD4 counts less than 10 cells/mm3; in the placebo arm, 38 of the 53 infections developed in subjects who had CD4 counts less than 19 cells/mm3. All of the study subjects who developed positive cultures had CD4 counts less than 50 cells/mm3.

Over the course of the treatment and follow-up periods, there were 107 deaths from all causes among the patients who received clarithromycin prophylaxis for MAC (32%), versus 137 deaths among the patients who did not receive prophylaxis (41%). The overall difference in mortality between the two groups was 26%, with the greatest survival benefit seen during the first 12 months of the trial (Figure 1). The investigators, Pierce et al., observed that roughly two-thirds of the 244 recorded deaths in the study cohort occurred a month or more after the study medication had been withdrawn.

"Patients who received clarithromycin prophylaxis were less likely to require hospitalization for any reason during the course of treatment and follow-up. The investigators speculate that this 23% reduction in risk of hospitalization -- like the 26% reduction in risk of death -- may be attributable to clarithromycin's activity against other OIs."

One noteworthy finding of this study was that patients who received prophylactic therapy were less likely to require hospitalization for any reason during the course of treatment and follow-up (49% versus 57%). The investigators speculate that this 23% reduction in risk of hospitalization -- like the 26% reduction in risk of death -- may be attributable to clarithromycin's activity against other pathogens, such as pneumonia and giardia, that commonly infect profoundly immunocompromised patients. Alternatively, these risk reductions may simply reflect the overall benefits of avoiding MAC infection itself.

These findings led the investigators to conclude that clarithromycin monotherapy was well tolerated, reduced the incidence of MAC infection and hospitalization as well as the development of other complications of HIV infection, and prolonged survival in AIDS patients. These findings also led a national task force on pharmacotherapy in patients with late-stage HIV disease to recommend that clarithromycin be considered as an alternative to rifabutin as MAC prophylaxis.(20) The committee also recommended that the threshold for prophylactic treatment be lowered from a CD4 count of 100 cells/mm3 to 50-75 cells/mm3.

"Clarithromycin prophylaxis is well tolerated, reduces the incidence of MAC infection and hospitalization as well as the development of other complications of HIV infection, and prolongs survival in AIDS patients. These findings have led a task force on pharmacotherapy in patients with late-stage HIV disease to recommend that clarithromycin be considered as an alternative to rifabutin as MAC prophylaxis."

A second trial of clarithromycin prophylaxis for MAC has been completed, and it provides even more convincing evidence of the long-term benefit of prophylaxis with this agent for patients with very advanced HIV disease. In this trial, conducted by the National Institute of Allergy and Infectious Diseases AIDS Clinical Trials Group (ACTG 196) and the Community Program on Clinical Research in AIDS (CPCRA 009), 1216 patients with a median absolute CD4 count of 28 cells/mm3 were randomly assigned to receive rifabutin (300 to 450 mg/day), clarithromycin (500 mg twice daily), or the combination of both agents.(22) Median follow-up of patients enrolled in ACTG 196/CPCRA 009 was 589 days -- longer than in any other MAC prophylaxis trial ever conducted -- and overall survival was approximately 50%. In an intent-to-treat analysis, only 9% of the patients assigned to clarithromycin developed disseminated MAC, compared to 15% assigned to rifabutin (p < .001).

In this trial the combination of clarithromycin and rifabutin was not significantly better than clarithromycin alone. Based on previous natural history studies(3) and placebo-controlled prophylaxis trials, the expected incidence of disseminated MAC in this trial group -- in the absence of any MAC prophylaxis -- would have been 36% at 19 months. Thus, long-term clarithromycin prophylaxis administered to 100 patients would be expected to prevent 27 cases of disseminated MAC over the median lifetime of the population at risk.

New approaches to MAC prophylaxis: Weekly azithromycin

In a double-blind, multicenter clinical trial recently reported by the California Collaborative Treatment Group, 693 study participants were randomized to daily rifabutin (300 mg), weekly azithromycin (1200 mg, given as two 600-mg lactose-free tablets), or both drugs at the same dosages.(23) Azithromycin was chosen because of its demonstrated effectiveness in animal models(24) and in smaller clinical trials.(18, 25) Because azithromycin has unique tissue-binding properties and persists in phagocytic cells for six days or longer,(26-28) it was possible to administer the drug on an intermittent rather than a daily basis.

Patients were enrolled in this trial beginning in late 1992, and they were followed for a median of 514 days. At entry, all study subjects had CD4 counts below 100 cells/mm3 (median: 40 cells/mm3) and blood cultures that were negative for M. avium. (Cultures that remained free of growth for eight weeks were deemed negative.) All enrolled patients were checked at monthly intervals for signs and symptoms of MAC, other OIs, compliance with therapy, and adverse reactions to that therapy. Study medications were discontinued if patients developed disseminated MAC infection or exhibited toxic side effects of therapy.

When investigators completed their intention-to-treat analysis, they found that over the course of the study MAC infections had developed in 23% of the patients treated with daily rifabutin (52 of 223), in 14% of the patients treated with weekly azithromycin (31 of 223), and in 8% of those given both drugs (18 of 218). The one-year cumulative risk of disseminated MAC infection, adjusted for baseline CD4 count, was 47% lower for patients taking azithromycin alone than for patients taking rifabutin alone, and it was 72% lower for patients taking both drugs than for those taking only rifabutin (Figure 2).

Respiratory-tract infections developed twice as often in subjects treated with rifabutin alone as in those who got azithromycin, alone or in combination. This is an unsurprising finding, given the known efficacy of azithromycin against such common respiratory- tract pathogens as Streptococcus pneumoniae, Haemophilus influenzae, and Mycoplasma pneumoniae. Overall, however, adverse events were more common in patients who received azithromycin than in those who got rifabutin. Gastrointestinal symptoms accounted for most of the observed difference. In general, these symptoms were mild and were limited to the day the medication was taken. Permanent discontinuation of therapy due to GI symptoms was virtually identical in all three treatment arms.

"Azithromycin is more effective than rifabutin as single-agent MAC prophylaxis, and it does not interact with other commonly-prescribed AIDS medications."

Noting that azithromycin therapy leads to high and persistent concentrations of active drug in the macrophages,(26, 27) the very cells infected by M. avium complex, the investigators concluded that intermittent azithromycin monotherapy may provide as constant a protective effect as daily dosing with other prophylactic agents. The authors made this recommendation based on their own findings and those of a recently-concluded, placebo-controlled trial that demonstrated a 60% reduction in risk of MAC in patients who received once-weekly azithromycin compared with those who got placebo.(29) As they observed, azithromycin is not only more effective than rifabutin as single-agent prophylaxis, it does not interact with other common AIDS medications.

The fact that azithromycin can be administered once weekly is seen by members of the California Collaborative Treatment Group as a particular advantage, especially in patients with well-advanced HIV disease, in whom polypharmacy and compliance with complex dosing regimens are already matters of considerable concern. The addition of daily rifabutin to a prophylactic regimen of weekly azithromycin clearly confers additional benefits in terms of reducing the risk of MAC infection, but it has little or no impact on survival: there were 83 deaths among patients assigned to azithromycin versus 85 among those assigned to rifabutin and 81 among those on combination therapy.

Before prescribing such combination therapy as prophylaxis against disseminated MAC infections, clinicians will want to consider the additional cost, the additional complications (daily versus weekly dosing, and 16 tablets per week versus 2), and the likelihood that combination therapy may be less well tolerated and may cause drug interactions with rifabutin.

"Adding daily rifabutin to a prophylactic regimen of weekly azithromycin reduces the risk of MAC infection, but it has little or no impact on survival. Before prescribing such combination therapy clinicians will want to consider the additional cost, the additional complications, and the likelihood that combination therapy may be less well tolerated and may cause drug interactions."

In comparing the results of the clarithromycin and azithromycin prophylaxis trials, it is important to note that the breakthrough infection rate in ACTG 196/CPCRA 009 for patients treated with clarithromycin alone was as low as that for patients taking the azithromycin/rifabutin combination in the California Collaborative Treatment Group trial. The latter patients began with a higher median CD4 count and were followed for a shorter time -- and thus were probably at lower overall risk.

Fortunately, breakthrough infections attributable to macrolide-resistant mycobacteremia have been rare. In the three prophylaxis trials described above only 2% to 3% of patients given clarithromycin, and only I% of patients given azithromycin, had breakthrough bacteremia with MAC isolates that were resistant to these drugs.

New regimens for the treatment of MAC bacteremia

No prophylaxis regimen will completely prevent breakthrough infections of disseminated MAC, and once infection takes hold the clinician's only option is suppressive therapy with a combination of agents. Although there are established guidelines for the treatment of MAC infection,(9) until quite recently there had been no large-scale, well-designed comparative trials of the various multidrug regimens used to treat disseminated infection. The Canadian HIV Trials Network is the first to provide such data.(30)

In 1991, two state-of-the-art reviews(4, 31) of the extant literature on the treatment of MAC bacteremia reached the same conclusion: disseminated infections should be treated with a combination of four drugs. The recommended regimen was: rifampin (600 mg daily), clofazimine (100 mg daily), ciprofloxacin (750 mg twice daily), and ethambutol (given once daily at a dose based on the patient's weight: 800 mg for patients weighing less than 60 kg, 1200 mg for patients weighing 60 to 80 kg, and 1600 mg for patients weighing more than 80 kg).

This standard treatment protocol was developed before the availability of macrolides like clarithromycin, the most active drug used in the treatment of MAC bacteremia,(17) and before it was understood that rifabutin is effective in both prophylaxis(10, 22) and treatment(32, 33) of MAC. It included rifampin and clofazimine, both of which are active in vitro but are ineffective when they are given as monotherapy for MAC bacteremia.(34)

The Canadian group elected to compare the safety and efficacy of the standard four-drug regimen with a three-drug regimen that included clarithromycin (1000 mg twice daily), rifabutin (600 mg daily), and ethambutol at the weight-adjusted doses described above. Ethambutol was included in the three-drug regimen because earlier studies had established that ethambutol monotherapy does result in modest reductions in circulating M. avium in treated patients.(34)

Enrollment began in November of 1992, with patients randomized to one of the two treatment arms. In November of the following year, the dose of rifabutin was reduced from 600 to 300 mg daily because uveitis developed with some frequency in the three-drug group. In all, 229 patients were enrolled at the 24 participating centers, and 187 of these patients could be evaluated bacteriologically. The clinical status of all subjects was assessed at the time of enrollment and again at weeks 2, 4, 8, 12, and 16 of the study.

The two treatment arms were virtually identical with respect to sex, quantitative mycobacteremia, hemoglobin and alkaline phosphate levels, Karnofsky scores, and CD4 counts -- which were very low in both groups (median: 10 cells/mm3). Discontinuation of therapy was defined as permanently stopping treatment with at least two of the study drugs. The median times to discontinuation were 68 days in the four-drug group and 127 days in the three-drug group. Of those patients who discontinued treatment during the 16 weeks of the trial, only six patients in the four-drug cohort and two in the three-drug cohort did so because therapy proved too toxic.

The three-drug combination of clarithromycin, rifabutin, and ethambutol cleared MAC bacteremia in 69% of treated patients. In those on the four-drug regimen, by contrast, the clearance rate was only 29%. Moreover, blood was sterilized much more rapidly in the three-drug arm: 87% of those whose bacteremia was eventually cleared achieved that outcome by Week 4, versus 54% of those in the four-drug arm (Figure 3).

During the first 16 weeks of the study no relapses were seen in patients treated with the three-drug combination of clarithromycin, rifabutin, and ethambutol. By contrast, clarithromycin monotherapy leads to rapid development of resistance, which in one study was 21% at 12 weeks and eventually rose to 46%.(17) The investigators suggest that the addition of rifabutin and ethambutol to the treatment regimen reduces the incidence of clarithromycin resistance, a conclusion supported by a French study that found significantly less clarithromycin resistance in patients treated with that drug plus rifabutin and ethambutol than in patients treated with clarithromycin plus clofazimine.(35)

Although fevers did not abate and diarrhea did not resolve any more quickly with one treatment than with the other, patients in the three-drug arm experienced significantly less weight loss (0.5 kg vs. 2.5 kg). More importantly, treatment with clarithromycin, rifabutin, and ethambutol conferred a survival benefit in these very sick patients. Median survival was 8.6 months in the three-drug group, versus 5.2 months in the other group (p = 0.001) -- an improvement of 65%.

It is now clear that: 1) a macrolide, such as clarithromycin or azithromycin, is absolutely necessary in the treatment of disseminated MAC; 2) at least one additional drug should be added; and 3) the best choices for additional drugs are ethambutol and rifabutin.

What is not clear is whether the addition of ethambutol alone, rifabutin alone, or both will maximize efficacy and minimize toxicity. A recent trial compared a treatment regimen of clarithromycin plus clofazamine to a regimen of clarithromycin plus clofazamine plus ethambutol as initial treatment for disseminated MAC.(36) In this trial, 69% of patients in both arms had an initial microbiologic clearing of MAC from the blood. However, after 36 weeks of therapy the risk of relapse was 68% in the two-drug arm versus only 5% in the three-drug arm (P = .004).

Another recent trial has demonstrated that clofazamine is of no benefit in the initial treatment of disseminated MAC infection. Thus, results with the two-drug combination of clarithromycin and ethambutol compare quite favorably with the results reported by the Canadian HIV Trials Network for the three-drug combination of clarithromycin, rifabutin, and ethambutol. An NIAID AIDS Clinical Trials Group study is currently comparing the efficacy of clarithromycin/ethambutol versus clarithromycin/rifabutin versus clarithromycin/ethambutol/rifabutin as treatment regimens for disseminated MAC. Until the results of that trial are available, no firm recommendation can be made regarding which regimen should be used. In clinical practice, however, most AIDS specialists are using the clarithromycin/ethambutol combination because of its simplicity and reduced risk of side effects.


The advent of antimycobacterial macrolide therapy has greatly improved prophylaxis and treatment of disseminated MAC infection. Although questions still remain about which regimens are optimal, several general recommendations can be made:

Patients with CD4 counts below 50 cells/mm3 who do not have evidence of active mycobacterial disease should receive prophylaxis with either clarithromycin (500 mg twice daily) or azithromycin (1200 mg once weekly) -- which can be administered with rifabutin, (300 mg daily).

Optimal treatment of disseminated MAC should begin with clarithromycin (500 mg twice daily) plus ethambutol (approximately 15 mg/kg/day). Rifabutin may be added to this regimen, but the exact rifabutin dosage will depend on the patient's concurrent medications, which could result in drug-drug interactions.


1. Jones JL, Hanson DL, Chu SY, Fleming FL, Hu DJ, Ward JW. Surveillance of AIDS-defining conditions in the United States. AIDS 1994; 8: 1489-93.

2. Inderlied CB, Kemper CA, Bermudez LEM. The Mycobacterium avium complex. Clin Microbiol Rev 1993; 3: 266-310.

3. Nightingale SD, Byrd LT, Southern PM, Jockusch JD, Cal SX, Wynne BA. Incidence of Mycobacterium avium-intra-cellulare complex bacteremia in human immunodeficiency virus-positive patients. J Infect Dis 1992; 165: 1082-5.

4. Horsburgh CR Jr. Mycobacterium avium complex infection in the acquired immunodeficiency syndrome. N Engl J Med 1991; 324: 1332-8.

5. Havlik JA Jr, Horsburgh CR Jr, Metchock B, Williams PP, Fann SA, Thompson SE III. Disseminated Mycobacterium avium complex infection: Clinical identification and epidemiological trends. J Infect Dis 1992; 165: 577-80.

6. Bacellar H, Munoz A, Hoover DR, et al. Incidence of clinical AIDS conditions in a cohort of homosexual men with CD4 + cell counts < 100/mm3: Multicenter AIDS Cohort Study. J lnfect Dis 1994; 170: 1284-7.

7. Hardy WD, Feinberg J, Finkelstein DM, et al. A controlled trial of trimethoprim-sulfamethoxazole or aerosolized pentamidine for secondary prophylaxis of Pneumocystis carinii pneumonia in patients with acquired immune-deficiency syndrome: ACTG 021. N Engl J Med 1992; 327: 1842-7.

8. Torriani F, Behling C, McCutchan JA, Haubrich R, Keays L, Havlir D. Disseminated Mycobacterium avium complex (MAC): Correlation between blood cultures and tissue burden in AIDS patients. Second National Conference on Human Retroviruses and Related Infections, Washington, D.C., 1995. Abstract 115.

9. Public Health Task Force on Prophylaxis and Therapy for Mycobacterium avium Complex. Recommendations on prophylaxis and therapy for disseminated Mycobacterium avium complex for adults and adolescents infected with human immunodeficiency virus. MMWR 1993; 42 (RR-9): 14-20.

10. Nightingale SB, Cameron DW, Gordin FM, et al. Two controlled trials of rifabutin prophylaxis against Mycobacterium avium complex infection in AIDS. N Engl J Med 1993; 329: 828-33.

11. Trapnell CB, Narang PK, Li R, Lavelle JP. Increased plasma rifabutin levels with concomitant fluconazole therapy in HIV-infected patients. Ann Intern Med 1996; 124: 573-6.

12. The Indinavir Pharmacokinetic Study Group. Indinavir (MK 639) drug interaction studies. Eleventh International Conference on AIDS, Vancouver, B.C., 1996. Abstract.

13. Alien C, Cavanaugh JH, Shi H, Hsu A, Granneman GR, Leonard J. Assessment of multiple doses of ritonavir on the pharmacokinetics of rifabutin. Eleventh International Conference on AIDS, Vancouver, B.C., 1996. Abstract.

14. Shafran SD, Deschenes J, Miller M., et al. Uveitis and pseudojaundice during a regimen of clarithromycin, rifabutin, and ethambutol. N Engl J Med 1994; 330: 438.

15. Fernandes PB, Hardy DJ, McDaniel D, Hanson CW, Swanson RN. In vitro and in vive activities of clarithromycin against Mycobacterium avium. Antimicrob Agents Chemother 189; 33: 1531-4.

16. Dautzenberg B, Truffot C, Legris S, et al. Activity of clarithromycin against Mycobacterium avium infection in patients with the acquired immune deficiency syndrome: A controlled clinical trial. Am Rev Respir Dis 1991; 144: 564-9.

17. Chaisson RE, Benson CA, Dubé MP et al. Clarithromycin therapy for bacteremic Mycobacterium avium complex disease: A randomized, double-blind, dose-ranging study in patients with AIDS. Ann Intern Med 1994; 121: 905-11.

18. Young LS, Wiviott L, Wu M, Kolonoski P, Bolan R, Inderlied CB. Azithromycin for treatment of Mycobacterium avium-intracellulare complex (MAI) in patients with AIDS. Lancet 1991; 338: 1107-9.

19. Pierce M, Crampton S, Henry D, Heifets L, LaMarca A, et al. A randomized trial of clarithromycin as prophylaxis against disseminated Mycobacterium avium complex infection in patients with advanced acquired immunodeficiency syndrome. N Engl J Med 1996; 335: 384-91.

20. Disseminated infection with Mycobacterium avium complex. In: USPHS/IDSA guidelines for the prevention of opportunistic infections in persons infected with human immunodeficiency virus: A summary. MMWR 1995; 44 (RR-8): 11-12.

21. Heifets L, Mor N, Vanderkolk J. Mycobacterium avium strains resistant to clarithromycin and azithromycin. Antimicrob Agents Chemother 1993; 37: 2364-70.

22. Benson CA, Cohn DL, Williams P, et al. A phase III prospective, randomized, double-blind study of the safety and efficacy of clarithromycin (CLA) vs. rifabutin (RBT) vs. CLA + RBT for prevention of Mycobacterium avium complex disease in HIV+ patients with CD4 copies < 100 cells/mm3. Third Conference on Retroviruses and Opportunistic Infections, Washington, D.C., 1996. Abstract.

23. Havlir DV, Dubé MP, Sattler FR, et al. Prophylaxis against disseminated Mycobacterium avium complex with weekly azithromycin, daily rifabutin, or both. N Engl J Med 1996; 335: 392-8.

24. Klemens SP Cynamon MH. Intermittent azithromycin for treatment of Mycobacterium avium infection in beige mice. Antimicrob Agents Chemother 1994; 38: 1721-5.

25. Beny A, Koletar S, Williams D. Azithromycin therapy for disseminated Mycobacterium avium-intracellulare in AIDS patients. First National Conference on Human Retroviruses and Related Infections, Washington, D.C., 1993. Abstract.

26. Girard AE, Girard D, English AR, et al. Pharmacokinetics and in vive studies with azithromycin (CP-62,993), a new macrolide with an extended half-life and excellent tissue distribution. Antimicrob Agents Chemother 1987; 31: 1948-54.

27. Gladue RP Bright GM, Isaacson RE, Newborg ME In vitro and in vivo uptake of azithromycin (CP-62,993) by phagocytic cells: Possible mechanism of delivery and release at sites of infection. Antimicrob Agents Chemother 1989; 33: 277-82.

28. Foulds G, Shepard RM, Johnson RB. The pharmacokinetics of azithromycin in human serum and tissues. J Antimicrob Chemother 1990; 25 (SupplA): 73-82.

29. Oldfield EC, Dickinson G, Chung R, et al. Once weekly azithromycin for the prevention of Mycobacterium avium complex (MAC) infection in AIDS patients. Third Conference on Retroviruses and Opportunistic Infections, Washington, D.C., 1996. Abstract.

30. Shafran SD, Singer J, Zarowny DP Phillips P, et al. A comparison of two regimens for the treatment of Mycobacterium avium complex bacteremia in AIDS: Rifabutin, ethambutol, and clarithromycin versus rifampin, ethambutol, clofazimine, and ciprofloxacin. N Engl J Med 1996; 335: 377183.

31. Ellner JJ, Goldberger MJ, Parenti DM. Mycobacterium avium infection and AIDS: A therapeutic dilemma in rapid evolution. J Infect Dis 1991; 163: 1326-35.

32. Sullam PM, Gortin FM, Wynne BA, Rifabutin Treatment Group. Efficacy of rifabutin in the treatment of disseminated infection due to Mycobacterium avium complex. Clin Infect Dis 1994; 19: 84-6.

33. Dautzenberg B, Ruf B, Esposito R, et al. Rifabutin vs. placebo in combination therapy for the treatment of disseminated MAC in HIV+ patients. 34th Interscience Conference on Antimicrobial Agents and Chemotherapy, Orlando, FL, 1994. Abstract.

34. Kemper CA, Havlir D, Haghighat D, et al. The individual microbiologic effect of three antimicrobial agents, Clofazimine, ethambutol, and rifampin, on Mycobacterium avium complex bacteremia in patients with AIDS. J Infect Dis 1994; 170: 157-64.

35. May T, Brel F, Beuscart C, et al. A French randomized open trial of 2 clarithromycin combination therapies for MAC bacteremia: First results. 35th Interscience Conference on Antimicrobial Agents and Chemotherapy, San Francisco, CA, 1995. Abstract.

36. Dubé M, Sattler F, Torriani F et al. Prevention of relapse of MAC bacteremia in AIDS: A randomized study of clarithromycin plus clofazamine, with or without ethambutol. Third. Conference on Retroviruses and Opprtunistic Infections, Washington, D.C., 1996. Abstract 206.

Mark A. Jacobson, M.D., is Associate Professor of Medicine at UCSF School of Medicine, AIDS Program, San Francisco General Hospital, San Francisco, CA.