This section contains the following sub-sections:

• Research into epidemiology of Mycobacterium avium intracelllare.

• Treatment with macrolide-containing regimens.

• Non-macrolide treatment regimens.

• Research into interrupting Mycobacterium avium intracellulare prophylaxis.

• Prophylaxis in the age of highly-active antiretroviral therapy.

• Prophylaxis prior to highly-active antiretroviral therapy.

• Drug interactions.

• References.

Research into the epidemiology of Mycobacterium avium intracellulare

von Reyn (2002) conducted a prospective study to identify sources of MAI infection among HIV-infected people with CD4 counts below 100. Increased risk was associated with PCP (hazard ratio 1.8), consumption of spring water (HR 4.9), consumption of raw seafood (HR 34.3), and gastrointestinal endoscopy (HR 2.9).

Ives (2001) conducted a retrospective analysis of the impact of highly active antiretroviral therapy (HAART) on the incidence of opportunistic infections among HIV patients at a London clinic. Comparing incidence rates 1990-1995 and 1996-1998, the incidence of MAI fell by 18% - from 1.58 to 1.29 cases per 100 patient-years.

Rossi (2001) analysed the Swiss HIV Cohort data and found 398 people developed MAI between 1990 and 1999. During this period, 3208 people had CD4 counts below 50 with follow-up time of 6004 years. Incidence fell from 8.8 to 1.4 per 100 person-years over the period (p<0.001). Diagnosis after 1995 was most strongly associated with improved survival.

Tumbarello (2001) reported that the incidence of MAI was 3.7 per 100 person-years prior to the introduction of highly active antiretroviral therapy (HAART) and fell to 0.9 per 100 person-years in the HAART era.

Horsburgh (1994) reported that in a case-control study of 83 people with MAI and CD4 counts below 50, CD4 count below 25 and consumption of hard cheese were associated with disease, while daily showering was protective.

Montecalvo (1994) reported that MAI could be isolated from 32% of domestic tap water samples taken from the homes of 81 people with HIV in New York. 3 samples were of the same serovar as MAI isolates from the residents.

Chin (1994) found that among 184 people with CD4 counts below 50, the risk of MAI bacteraemia was (approximately) 60% within 1 year for people with MAI in either the respiratory or gastro-intestinal tract and was greater than for those without MAI in these sites (relative hazards for respiratory and gastro-intestinal tract, 2.3 and 6.0 respectively). Both respiratory tract specimen and stool culture had poor sensitivities (22% and 20%, respectively) but good positive predictive values (approximately 60%) for bacteraemia.

Nightingale (1992) prospectively monitored 1,006 patients with AIDS for the development of MAI bacteraemia. Among patients who survived for 2 years after an AIDS diagnosis, 43% had developed MAC bacteraemia. The risk of developing MAC bacteraemia increased as the CD4 count fell, and most cases occurred in patients with CD4 counts below 100.

Treatment with macrolide-containing regimens

Preferred treatment regimen for MAI bacteraemia is clarithromycin (500mg twice daily) plus ethambutol (15mg/kg/day). Azithromycin (600mg/day) instead of clarithromycin is also acceptable, with or without rifabutin (300-600mg/day or adjusted if taking concurrent protease inhibitor or NNRTI). Amikacin (10-15mg/kg IV) or ciprofloxacin (500-700mg twice daily) may also be added to a combination regimen for MAI.

Chaisson (1994) enrolled 108 patients with DMAI in ACTG 157, a dose-comparison study of clarithromycin (500, 1000, or 2000mg twice a day). An interim analysis was performed on the 72 participants, with evaluation based on bacteriologic and clinical assessments. Eradication (99% reduction in bacterial count) or partial eradication (90% reduction) occurred in 44% of the low-dose group, 78% in the middle-dose group and 94% at the higher dosage level after 2 weeks. At four weeks the differences between the groups were reduced, and at 8 weeks all regimens were equally efficacious (approx. 80%) in bacterial suppression. Fevers and night sweats improved in 76%, 47% and 75% of participants. Survival at early stages of treatment was significantly poorer at the higher dose; 6% of those receiving the low-dose died during the first 12 weeks of therapy, compared with 29% of the 1000mg group (p = 0.01). Adverse effects, predominantly GI, occurred in 90% of participants, leading to discontinuation in 3 patients (500mg), 2 patients (1000mg) and 11 patients (2000mg). In vitro resistance to clarithromycin developed in 16 participants, and was accompanied by a return of clinical symptoms and/or increased colony-forming units in some patients.

Horsburgh (1991) reported that a group of 39 patients with untreated disseminated MAI (DMAI) had significantly shorter survival, mean of 5.6 +/- 1.1 months (median 4 months), than 39 matched patients with AIDS but without DMAI, mean 10 .8 +/- 1.3 months (median 11 months). The survival of 16 additional patients with DMAC who received anti-mycobacterial therapy, mean of 9.5 +/- 1.4 months (median 8 months), was not significantly shorter than that of an additional 16 matched control subjects, mean 11.7 +/- 1.9 months (median 11 months). Patients with treated DMAC survived significantly longer than those with untreated DMAI. It is concluded that untreated DMAC significantly shortens survival, and that patients with DMAI who receive anti-mycobacterial therapy do not experience the shortened survival seen in untreated DMAI.

Dautzenberg (1991) reported on a multicentre open trial of clarithromycin in 77 people with late-stage AIDS. M. avium was eradicated from blood cultures in 11 (63%) of 16 evaluable patients receiving daily doses or 500 or 1000mg and in 45 of 46 (98%) of those receiving 1500 or 2000mg. Eradication after 2 months was influenced by continuity of drug treatment; 36 of 42 patients with no relapse had received continuous treatment versus six of 14 patients whose drug treatment had been stopped for 7 days or longer. After 2 to 7 months of treatment, acquired resistance associated with relapse was observed. Drug side effects were elevated liver enzyme levels (26%) and impaired hearing (4%). Concomitant AIDS drugs had no favourable effect on outcome and may have worsened liver toxicity.

Young (1991) conducted an uncontrolled phase I study using azithromycin monotherapy for DMAI in 24 patients with AIDS. 3 groups of patients with positive blood cultures for M.avium received 500mg oral azithromycin daily for 10 days (group 1), 20 days (group 2), or 30 days (group 3). The mean concentration of mycobacteria in blood decreased in each of the groups, with the greatest reduction (-1.37 log) occurring in the 30-day treatment group. In the groups treated for 20 to 30 days, fever, night sweats and splenomegaly were all reduced. Fatigue, weight loss, Karnofsky score and appetite did not improve. Diarrhoea was the most common side effect, but it did not lead to discontinuation of therapy. Preliminary in vitro data suggest that some patients may have developed MAI strains resistant to azithromycin during treatment.

Koletar (1999) randomised 88 patients with MAI to receive a daily dose of either 600mg or 1200mg of azithromycin. 62 people completed 6 weeks of treatment. At 6 weeks 56% of blood cultures were sterile in the 600mg group compared with 42% of the 1200mg group (nss). Side effects, predominantly gastrointestinal, were more common at the higher dose.

Ives (1995) found that prior to the availability of clarithromycin and azithromycin 61.5% of people with MAI received anti-mycobacterial treatment compared with 92% afterwards. Median survival of treated people was 255 versus 145 days for untreated people. Median survival of macrolide-treated people was 284 versus 168 days for people receiving treatment without a macrolide.

May (1997) enrolled 144 eligible people with MAI bacteraemia in a trial comparing two clarithromycin-based combinations. Participants received clarithromycin (2000mg/day for two months then 100mg/day) plus clofazimine (200mg/day then 100mg/day), or clarithromycin, rifabutin (450mg/day) and ethambutol (1200mg/day). There was no difference between the groups in terms of the proportion who were MAC bacteraemic at months 2 or 6, or in survival. However, the number of relapses between months 2 and 6 were 6 and 22, a significant difference favouring the triple combination. Resistance to clarithromycin was also significantly more likely in the clarithromycin plus clofazimine arm (21 versus 2 cases).

Cohn (1999) enrolled 85 people with MAC in a study comparing four three-drug regimens: clarithromycin (500mg or 1000mg twice daily), plus ethambutol (800 - 1200mg/day), plus rifabutin (300mg/day) or clofazimine (100mg/day). The clarithromycin dose comparison was stopped after a significantly higher death rate was observed among people receiving the higher dose after a mean follow-up of 4.4 months (43% versus 22%). There were no differences in adverse events, and after 2 months clinical improvements and microbiologic responses were similar.

Shafran (1996) enrolled 229 people with MAI bacteraemia in a randomised open-label trial comparing 16 weeks treatment with a four-drug combination - ciprofloxacin (750mg twice daily), ethambutol (15mg/kg/day), rifampicin (600mg/day) and clofazimine (100mg/day) - versus a three-drug combination of clarithromycin (1000mg twice daily), rifabutin (600mg/day) and ethambutol (15mg/kg/day). The dose of rifabutin was lowered to 300mg/day after 24 cases of uveitis were reported. Among those treated for at least 4 weeks, clearance of bacteraemia was seen in 78% of the three-drug arm and 40% of the four-drug arm. Median survival times were 8.6 and 5.2 months respectively.

Chaisson (1997) studied the efficacy of clarithromycin (500mg twice daily) and ethambutol (15mg/kg/day) with or without clofazimine (100mg/day) for treating 106 people with MAI. 89 were bacteraemic at baseline; 65% of the 2-drug arm and 54% of the 3-drug arm become culture negative. The proportions of participants who withdrew due to adverse events were 13% and 22% respectively. 38% of 2-drug recipients and 61% of 3-drug recipients died, and time to death was shorter in the 3-drug arm.

Dube (1997) studied the effectiveness of clarithromycin (1000mg twice daily) plus clofazimine (100mg/day) with or without the addition of ethambutol (800mg/day) for treating MAI bacteraemia and preventing relapse in 80 bacteraemic people. Initial response was similar for the two drug and three drug arms. However, the estimated risk of relapse at 36 weeks were 68% for the two-drug arm versus 5% for the triple-drug arm. Of 27 positive cultures following an initial response, 21 (78%) were clarithromycin-resistant.

Dunne (2000) enrolled 246 people with MAI into a randomised, double-blind study of azithromycin (250mg daily) vs azithromycin (600mg daily) vs clarithromycin (500mg twice daily) each combined with ethambutol for 24 weeks. The 250mg dose was dropped after interim analysis showed a lower clearance rate of bacteria. Response rate at in the azithromycin and clarithromycin groups respectively at 24 weeks was 46% and 56% (p=0.24) while approximately 60% of both groups had at least one negative culture. Relapse rate was 39% in the azithromycin group and 27% in the clarithromycin group.

Kemper (1994b) noted that although it is generally assumed that MAI bacteraemia, once it develops, is unremitting, and on the basis of this presumption, changes in the level of mycobacteraemia are used to gauge therapeutic response, in 7 out of 60 (12%) people enrolled in a prospective trial of MAI bacteraemia and AIDS, bacteraemia became undetectable before the initiation of anti-mycobacterial therapy. Patients with transient bacteraemia reported fewer and shorter symptoms and survived longer than those with sustained bacteraemia (59 vs. 31 weeks). There was no difference in the duration of AIDS, CD4 cell count, haematocrit, or body weight between groups. Two additional people with transient bacteraemia were identified outside this study setting. Despite disappearance of detectable mycobacteraemia and subsequent administration of anti-mycobacterial agent(s), bacteraemia once again became detectable in 6 patients 4-45 weeks after their negative pretreatment cultures.

Non-macrolide treatment regimens

Research into interrupting Mycobacterium avium intracellulare prophylaxis

Prophylaxis in the age of highly-active antiretroviral therapy

Phillips (2002) reviewed all 587 people in one Canadian province with CD4 counts below 75 between Jan 1996-Dec 1997. 383 (65%) were taking HAART. 316 received azithromycin (AZM) (median CD4 count 30) and 271 no prophylaxis (median CD4 count 40). Previous time on anti-HIV therapy was 11 months in the AZM group vs 2 months in the no prophylaxis group. Total incidence of new cases of MAI after 12 months was 3.2% in the AZM group (10 cases) and 8.6% in the no prophylaxis group (12 cases). 23% of cases were a form of immune reconstitution disease - HAART-associated lymph node inflammation. Prophylaxis reduced the risk of MAI events compared with no prophylaxis. No AZM resistance was detected among those who developed MAI while on prophylaxis.

Currier (2000) reported a randomised, double-blind, placebo-controlled study of azithromycin (1200mg weekly) in 643 people with previous CD4 cell counts below 50 (mean 20), with current CD4 counts over 100 (mean 219-235) due to combination antiretroviral therapy. During a median of 16 weeks follow-up, 2 cases of MAI had occurred both in the placebo arm and no cases in the prophylaxis arm. Both cases were atypical, occurring in the spine.

Mary-Krause (2000) reviewed 6 month data on people with CD4 counts below 100 during 2 periods: 1992-1995 and 1996-1998. Data for the early 90s vs late 90s was as follows: cases of MAI, 1710 vs 453; incidence of MAI 8.4 vs 2.8; and median CD4 count 8 vs 12. HAART reduced the risk of MAI by 73%. Prophylaxis was not associated with reduced risk of MAI.

Prophylaxis prior to highly-active antiretroviral therapy

Benson (2000) enrolled 1178 people with CD4 counts below 100 in a study comparing clarithromycin (500mg twice daily), rifabutin (450mg daily) and the combination of both drugs as MAI prophylaxis (ACTG 196/CPCRA 009). The incidence of MAI disease was significantly reduced among people received clarithromycin (9%, 15% and 7% respectively). Combination recipients were significantly more likely to discontinue due to toxicity (16%, 18% and 21% respectively). There was no difference in survival. Among participants who developed MAI , treatment with rifabutin was not associated with resistance, but 29% of breakthrough isolates in the clarithromycin monotherapy arm and 25% in the combination arm were resistant to clarithromycin.

Hewitt (1999) conducted a non-randomised, retrospective study of 192 people with CD4 counts below 100. 84 took 500mg clarithromycin once daily, 47 took 300mg rifabutin once daily and 61 took no prophylaxis. The incidence of MAI was significantly reduced among people receiving clarithromycin compared with rifabutin or no prophylaxis.

Oldfield (1998) enrolled 182 people with CD4 counts below 100 in a trial comparing azithromycin (1200mg once weekly) with placebo as MAI prophylaxis. After mean follow-up of approximately one year, intent-to-treat analysis showed 10.6% of azithromycin recipients developed MAI infection, versus 24.7% of placebo recipients. Of patients who died during the study 4/38 (10.5%) azithromycin recipients and 12/38 placebo recipients (31.6%) had a recurrence of MAI prior to death (p = 0.025). Overall, there was no difference between the two groups in terms of number of deaths or time to death. 78.9% of azithromycin recipients reported gastrointestinal toxicities, versus 27.5% of the placebo group.

Bishai (1996) reported a case in which rifabutin induced rifampicin-resistant TB in an individual with AIDS.

Havlir (1996) compared the effectiveness of azithromycin monotherapy (1200mg weekly), rifabutin monotherapy (300mg daily) or a combination of both drugs at preventing MAI bacteraemia in 693 people with CD4 counts below 100. In an intent-to-treat analysis, the proportions who had developed MAI after a median of 514 days follow-up were 13.9% in the azithromycin arm, 23.3% in the rifabutin arm and 8.3% in the combination arm. The proportions reporting dose-limiting side effects were 13%, 16% and 23% respectively. No difference in survival was seen. Resistance was rare; only 11% of azithromycin breakthrough isolates were resistant to azithromycin.

Pierce (1996) enrolled 667 people with CD4 counts below 100 in a trial comparing clarithromycin (500mg twice daily) with placebo as MAI prophylaxis. During the study and follow-up, 6% of the clarithromycin recipients developed MAI, compared with 16% in the placebo group. At 10 months death rates were 32% and 41% respectively. 11/19 clarithromycin recipients who developed MAI had high-level resistance to clarithromycin.

Moore (1995) re-analysed the pooled data from long-term follow-up of these two trials. Receipt of rifabutin was associated with a 20% risk of dying at one year, and a 31% risk at 1.5 years. Receipt of placebo was associated with a 28% risk at one year, and a 45% risk at 1.5 years. For participants with initial CD4 counts below 50, the relative hazard of dying while receiving rifabutin was 0.75, and 0.60 for participants with initial counts above 50.

Revicki (1995) analysed data on 1104 people with baseline CD4 counts below 200 who were randomised to rifabutin or placebo to prevent MAI. 9% of rifabutin recipients developed MAI compared to 18% of the placebo group (p<0.001). Quality time without symptoms and toxicity was 325 days in the rifabutin group and 309 days in the placebo group (p<0.05).

Abrams (1993) reported a trial of clofazimine (50mg by mouth every day) for MAI prophylaxis which was stopped after seven of 56 participants on clofazimine developed evidence of DMAI in blood cultures, compared with six of 54 untreated people.

Nightingale (1993b) conducted two randomized, double-blind, multicentre, placebo-controlled studies which both enrolled people with AIDS and CD4 counts below 200. Rifabutin monotherapy was administered at a dose of 300mg/day. In the first study (Adria 087027) MAI bacteraemia developed in 51 of 298 placebo recipients (17%) and 24 of 292 rifabutin recipients (8%). In the second trial Adria (087823), bacteraemia developed in 51 of 282 people in the placebo group (18%) and 24 of 274 people in the rifabutin group (9%). Rifabutin significantly delayed fatigue, fever, and decline in the Karnofsky performance score, decline in the haemoglobin level, elevation in alkaline phosphatase and hospitalization. The incidence of adverse events was similar with rifabutin and placebo. Overall survival did not differ significantly between the two groups, although there were fewer deaths with rifabutin (33) than with placebo (47) during the double-blind phase.

Drug interactions

Many anti-HIV drugs and other treatments interact with rifabutin and concurrent dosing may require a dose reduction in rifabutin. For recommended dosing see Rifabutin in Drugs used by people with HIV: Antibiotic drugs.

Clarithromycin should not be taken by a person also taking efavirenz. Clarithromycin levels are increased by conadministration of indinavir, ritonavir and saquinavir.