Share this article on:

Rifabutin but not clarithromycin prevents cryptosporidiosis in persons with advanced HIV infection

Fichtenbaum, Carl J.a; Zackin, Robertb; Feinberg, Juditha; Benson, Constancec; Griffiths, Jeffrey K.dthe AIDS Clinical Trials Group New Works Concept Sheet Team 064

Clinical Science

Background Macrolides have been reported to be effective for the prevention of cryptosporidiosis in persons with HIV infection.

Objective To evaluate the efficacy of clarithromycin and rifabutin for the prevention of cryptosporidiosis in persons with advanced HIV infection.

Design Cross-protocol analysis involving 2288 individuals with a history of a CD4 cell counts of ≤ 100 × 106 cell/l who were enrolled in two prospective clinical trials to prevent Mycobacterium avium complex (MAC) infection and cytomegalovirus (CMV) end-organ disease.

Interventions Clarithromycin 500 mg twice daily, rifabutin 300–450 mg daily, the combination of the two or no MAC prophylaxis.

Main outcome measurement Laboratory-confirmed cryptosporidiosis. Subjects were analyzed in an intent-to-treat and as-treated manner using time-to-event analyses (Cox proportional hazards models).

Results The median length of follow up was 463 days. The median CD4 count at entry was 29 × 106 cell/l (range 0–182). There were 60 cases of cryptosporidiosis during the prospective observational period, with an event rate of 2.2 per 100 person-years. In the intent-to-treat [relative risk (RR) 0.50; 95% confidence interval (CI) 0.26–0.96;P = 0.04] and as-treated (RR 0.42; 95% CI 0.20–0.91;P = 0.03) analyses, rifabutin alone was significantly associated with a lower rate of cryptosporidiosis. Clarithromycin alone was not protective in similar analyses (P = 0.98 and 0.90, respectively).

Conclusions In doses used to prevent MAC infection, rifabutin but not clarithromycin decreases the risk of developing cryptosporidiosis in persons with advanced HIV infection who are not receiving potent combination antiretroviral therapy.

From the aUniversity of Cincinnati College of Medicine, Cincinnati, bHarvard School of Public Health, Boston, Massachusetts, the cUniversity of Colorado Health Sciences Center, Denver, Colorado and dTufts University School of Medicine, Boston, Massachusetts. Requests for reprints to Dr C. J. Fichtenbaum, University of Cincinnati College of Medicine, Infectious Diseases Center, Mail location 0405, Eden Avenue and Albert Sabin Way, Cincinnati, Ohio 45267-0405, USA.

Received: 15 February 2000;

revised: 14 August 2000; accepted: 23 August 2000.

Sponsorship: This work was supported, in part, by NIH grant AI-38858 to the AIDS Clinical Trials Group and AI-25897 to the University of Cincinnati College of Medicine.

Back to Top | Article Outline


Cryptosporidiosis causes diarrhea in persons with advanced HIV infection [1,2]. In the era before potent combination antiretroviral therapy, the incidence of cryptosporidiosis in persons with advanced HIV infection was 3–5%[3,4]. Although potent combination antiretroviral therapy has led to a dramatic decline in the incidence of cryptosporidiosis, this remains an important opportunistic infection in areas where antiretroviral therapeutic options are limited and it may remain a threat to those with multiply resistant HIV [5,6].

There is no proven effective specific therapy for Cryptosporidium parvum infection [7]. Potent combination antiretroviral therapy appears to decrease shedding of C. parvum and shorten the duration of diarrhea in persons with cryptosporidiosis [5,6,8]. This is believed to be a non-specific effect involving the reconstitution of host immunity [8]. A number of medications have demonstrated activity against C. parvum in vitro and in animal models [7]. Macrolide antibiotics have been reported to be effective in preventing and treating cryptosporidiosis [9–16]. Therefore, we evaluated whether clarithromycin, rifabutin, or the combination were effective in preventing cryptosporidiosis in persons with advanced HIV infection.

Back to Top | Article Outline



This was a retrospective, cross-protocol analysis of subjects from two trials: ACTG 196/CPCRA 009 and ACTG 204. ACTG 196/CPCRA 009 was a randomized trial of the safety and efficacy of clarithromycin 500 mg twice daily versus rifabutin 450 mg daily versus the combination for the prevention of disseminated Mycobacterium avium complex (MAC) infection [17]. The dose of rifabutin was lowered to 300 mg daily several months into the trial when it was recognized that high levels of rifabutin were associated with uveitis. ACTG 204 was a randomized trial of valacyclovir versus two doses of acyclovir (aciclovir) for the prevention of cytomegalovirus (CMV) end-organ disease [18]. These trials were conducted between December 1992 and June 1995, prior to the availability of more potent protease inhibitor-based combination antiretroviral therapies.

Back to Top | Article Outline

Study population

Study subjects in ACTG 196/CPCRA 009 and ACTG 204 included HIV-infected persons with a CD4 cell count ≤ 100 × 106 cell/l. In addition, subjects in ACTG 204 had to have serologic evidence of prior CMV infection. Co-enrollment was allowed and 139 subjects (6%) were enrolled in both studies.

Back to Top | Article Outline

Study definitions

Evaluation for diarrhea lasting more than 7 days in ACTG trials includes evaluation of stool specimens for ova, parasites, Clostridium difficile toxin and bacterial culture. Cryptosporidiosis was defined by the recovery of oocysts from stool. Oocysts were identified at local laboratories using either a modified acid-fast stain or immunoassay. Study records of each participant with cryptosporidiosis were reviewed to document the presence of symptoms and signs consistent with the diagnosis. All diagnoses were verified by the principal investigator (C. J. F.).

Back to Top | Article Outline

Statistical analyses

The primary analysis assessed whether clarithromycin, rifabutin or the combination prevented the development of cryptosporidiosis. To minimize the potential for bias, the entire cohort was initially analyzed. The primary analysis was performed two different ways: intent-to-treat and as-treated.

The first analysis used an intent-to-treat method for the use of clarithromycin and/or rifabutin in ACTG 204. Randomization to medications for subjects in ACTG 196 was then identified and subjects were analyzed in four different groups (any prophylaxis, rifabutin, clarithromycin and rifabutin plus clarithromycin) and compared with subjects remaining in the cohort. For example, subjects who had used rifabutin in ACTG 204 or were randomized to rifabutin in ACTG 196 were compared with the rest of the cohort (excluding those randomized to or taking combination therapy). Similarly, those taking combination therapy were compared with the rest of the cohort. The comparison group included those taking rifabutin alone, clarithromycin alone or no prophylaxis. Consequently, the size and composition of the comparison group changed depending upon the drug regimen being analyzed.

In the as-treated analysis, the comparison groups were constructed as in the intent-to-treat analysis; however, subjects in any of the prophylaxis groups were censored 28 days after the time they discontinued prophylaxis. For the as-treated analysis, only endpoints at least 28 days after initiation and within 30 days of discontinuation of medication were considered.

A secondary analysis compared subjects taking clarithromycin alone or rifabutin alone with those who never took MAC prophylaxis using intent-to-treat and as-treated methods. Subjects with cryptosporidiosis prior to study entry were excluded. Cox proportional hazards models were used to determine the significance of any protective effect of medication on the development of cryptosporidiosis.

Back to Top | Article Outline


There were 2288 subjects available for analysis. The baseline characteristics are shown in Table 1. The median follow-up for study participants was 463 days. Sixty cases (2.6%) of cryptosporidiosis were documented during the study, with a calculated rate of cryptosporidiosis of 2.2 per 100 person-years. Cryptosporidiosis-free rates were 98.9% after 1 year and 94.9% after 2 years of follow-up.

Table 1

Table 1

In the intent-to-treat and as-treated analyses, the use of any prophylaxis was significantly associated with a lower rate of cryptosporidiosis (Table 2) and the use of rifabutin alone was associated with a lower rate of cryptosporidiosis, with a reduction of approximately 50%. While the combination of rifabutin plus clarithromycin was associated with a lower risk than the use of clarithromycin alone in all three analyses, this difference was not statistically significant. A subgroup analysis comparing subjects taking rifabutin or clarithromycin alone with subjects who never took MAC prophylaxis confirmed the analysis of the entire cohort. Rifabutin alone was also associated with a reduction in risk of the development of cryptosporidiosis compared with subjects who never used any type of MAC prophylaxis in both intent-to-treat [relative risk (RR) 0.39; 95% confidence interval (CI) 0.19–0.79;P = 0.009] and as-treated (RR 0.40; 95% CI 0.18–0.85;P = 0.02) analyses. In contrast, there was no significant difference in either the intent-to-treat (RR 0.64; 95% CI 0.34–1.21;P = 0.17) or as-treated (RR 0.68; 95% CI 0.35–1.31;P = 0.25) analyses for similar comparisons of clarithromycin alone versus no MAC prophylaxis. In subgroup analysis, combination therapy with clarithromycin plus rifabutin was protective compared with no prophylaxis in the intent-to-treat analysis (RR 0.43; 95% CI 0.21–0.91;P = 0.03) but did not reach significance in the as-treated analysis (RR 0.48; 95% CI 0.21–1.06;P = 0.07). There was no difference in the median CD4 lymphocyte count or use of anti-retroviral therapy between those persons receiving or not receiving MAC prophylaxis (data not shown).

Table 2

Table 2

Back to Top | Article Outline


The results of this study suggest that rifabutin, in doses used to prevent MAC infection, was also useful in preventing cryptosporidiosis. The use of rifabutin was associated with an approximately 50% reduction in risk of developing cryptosporidiosis. However, clarithromycin had no utility in preventing cryptosporidiosis. The results were consistent using different analytic approaches on the entire cohort (intent-to-treat and as-treated models) and in a subgroup analysis comparing each drug alone with subjects who had never used MAC prophylaxis. The strengths of this study include the relatively high number of clinical endpoints (60 cases of cryptosporidiosis), randomization of half the cohort to various MAC prophylaxis regimens, and the consistent approach to identifying cryptosporidiosis across study sites.

There are several important limitations in our study. This was a post-hoc analysis to test the hypothesis that macrolides might have some clinical effect in preventing cryptosporidiosis, which was based on in vitro and animal model data not a randomized trial for the prevention of cryptosporidiosis. Secondary analyses of data collected for other reasons are often confounded by inherent biases. This potential bias was somewhat mitigated by randomization of over half the subjects to one of three approaches to MAC prophylaxis. The intent-to-treat model tends to bias the results against demonstrating an effect by virtue of its conservative assumptions that always count subjects by their assigned treatment group regardless of whether they prematurely discontinue their medication. The fact that the intent-to-treat analytic method demonstrated a significant effect with rifabutin is important. We are unaware of any biological mechanism to explain the protective effect of rifabutin against cryptosporidiosis. There are no published data of in vitro experiments or animal models that explore the effects of rifamycins on C. parvum Therefore we cannot impute a true cause and effect from this association. However, it is unlikely that a randomized clinical trial of rifabutin prophylaxis for cryptosporidiosis will be conducted in the era of potent combination antiretroviral therapy.

These findings differ from those of a recently published study of the protective effects of clarithromycin and rifabutin against cryptosporidiosis [16]. In that study of 1019 HIV-infected persons with CD4 cell counts < 75 l × 106cells/l, clarithromycin or rifabutin were each more than 75% effective in preventing cryptosporidiosis [16]. However, this study was a retrospective review of the medical records from several large medical practices that provide HIV care. The decision to initiate MAC prophylaxis was at the discretion of the treating physician. The analysis of the effect of clarithromycin or rifabutin was restricted to patients who had taken medication for at least 3 months, which potentially favors treated patients who tolerated these medications and survived beyond 3 months. It is also unlikely that the diagnostic evaluations of diarrhea were standardized within and between primary care sites. These differences may account for the disparate results of each study.

Rifabutin, in doses used to prevent MAC infection, has now been shown to decrease the risk of developing cryptosporidiosis in two studies [16]. It would be useful to study the effects of semisynthetic rifamycins, particularly rifabutin, in animal models of cryptosporidiosis. Clarithromycin was not effective in our study. There are limited human data on the protective efficacy of azithromycin [15]. Some in vitro studies and animal models suggest that azithromycin may be more likely to be protective than clarithromycin [11–15]. Further analysis of the protective effects of azithromycin seems warranted. At present, the most reliable measure for preventing cryptosporidiosis in persons with advanced HIV infection is the use of potent combination antiretroviral therapy.

In conclusion, in doses used to prevent MAC infection, rifabutin, but not clarithromycin, appears to prevent cryptosporidiosis in persons with advanced HIV infection. While a prospective randomized clinical trial would be required to confirm these results, the diminished incidence of all opportunistic infections, including cryptosporidiosis, at present makes this unlikely.

Back to Top | Article Outline


1. Andreani T, le Charpentier Y, Brouet JC. et al. Acquired immunodeficiency with intestinal cryptosporidiosis: possible transmission by Haitian Whole Blood. Lancet 1983, 1: 1187 –1191.
2. Flanigan T, Whalen C, Turner J. et al. Cryptosporidium infection and CD4 counts. Ann Intern Med 1992, 116: 840 –842.
3. Sorvillo FJ, Leib LE, Kerndt PR, Ash LR. Epidemiology of cryptosporidiosis among persons with acquired immunodeficiency syndrome in Los Angeles County. Am J Trop Med Hyg 1994, 51: 326 –331.
4. Moore RD, Chaisson RE. Natural history of opportunistic disease in an HIV-infected urban clinical cohort. Ann Intern Med 1996, 124: 633 –642.
5. Carr A, Marriott D, Field A, Vasak E, Cooper DA. Treatment of HIV-1-associated microsporidiosis and cryptosporidiosis with combination antiretroviral therapy. Lancet 1998, 351: 256 –261.
6. Foudraine NA, Weverling GJ, van Gool T. et al. Improvement of chronic diarrhoea in patients with advanced HIV-1 infection during potent antiretroviral therapy. AIDS 1998, 12: 35 –41.
7. Griffiths JK. Human cryptosporidiosis: epidemiology, transmission, clinical disease, treatment, and diagnosis. Adv Parasitol 1998, 40: 37 –85.
8. Landau A, Aaron L, Pialoux et al. Impact of antiretroviral therapy (ART) on cryptosporidiosis outcome and factors of clinical resistance.Fifth Conference on Retroviruses and Opportunistic Infections. Chicago, 1998 [abstract 480].
9. Brasseur P, Lemeteil D, Ballet J. Anti-cryptosporidial drug activity screened with an immunosuppressed rat model. J Protozool 1991, 38: 230S –231S.
10. Cama VA, Marshall MM, Shubitz LF, Ortega YR, Sterling CR. Treatment of acute and chronicCryptosporidium parvuminfections in mice using clarithromycin and 14-OH clarithromycin. J Eukaryotic Microbiol 1994, 41: 25S. 25S.
11. Rehg JE. Anti-cryptosporidial activity of macrolides inimmunosuppressed rats. J Protozool 1991, 38: 228S –230S.
12. Kimata I., Uni S., Iseki M. Chemotherapeutic effect of azithromycin and lasalocid on cryptosporidium infection in mice. J Protozool 1991, 38: 232 –233S.
13. Rehg JE. A comparison of anticryptosporidial activity of paromomycin with that of other aminoglycosides and azithromycin in immunosuppressed rats. J Infect Dis 1994, 170: 934 –938.
14. Rehg JE. Activity of azithromycin against Cryptosporidia in immunosuppressed rats. J Infect Dis 1991, 163: 1293 –1296.
15. Hicks P, Zweiner RJ, Squires J, Savell V. Azithromycin therapy forCryptosporidium parvuminfection in four children infected with human immunodeficiency virus. J Ped 1996, 129: 297 –300.
16. Holmberg SD, Moorman AC, von Bargen JC. et al. Possible effectiveness of clarithromycin and rifabutin for cryptosporidium chemoprophylaxis in HIV disease. :HIV Outpatient Study (HOPS) Investigators. JAMA 1998, 279: 384 –386.
17. Benson CA, Williams PL, Cohn DL. et al. Clarithromycin or rifabutin alone or in combination for primary prophylaxis ofMycobacterium aviumcomplex disease in patients with AIDS: a randomized, double-blind, placebo-controlled trial. J Infect Dis 2000, 181: 1289 –1297.
18. Feinberg JE, Hurwitz S, Cooper D. et al. A randomized, double-blind trial of valacyclovir prophylaxis for cytomegalovirus disease in patients with advanced humanimmunodeficiency virus infection. :AIDS Clinical Trials Group Protocol 204/Glaxo Wellcome 123–014 International CMV Prophylaxis Study Group. J Infect Dis 1998, 177: 48 –56.

macrolides; rifabutin; clarithromycin; cryptosporidosis; HIV infection

© 2000 Lippincott Williams & Wilkins, Inc.