|
|
| |||||||||||||||||||||||||||||||||||||||||||
Background Disseminated Mycobacterium avium complex infection eventually develops in most patients with the acquired immunodeficiency syndrome (AIDS). This infection results in substantial morbidity and reduces survival by about six months.
Methods We conducted two randomized, double-blind, multicenter trials of daily prophylactic treatment with either rifabutin (300 mg) or placebo. All the patients had AIDS and CD4 cell counts
Results In the first trial, M. avium complex bacteremia developed in 51 of 298 patients (17 percent) assigned to placebo and 24 of 292 patients (8 percent) assigned to rifabutin (P<0.001). In the second trial, bacteremia developed in 51 of 282 patients in the placebo group (18 percent) and 24 of 274 patients in the rifabutin group (9 percent) (P = 0.002). Rifabutin significantly delayed fatigue, fever, decline in the Karnofsky performance score (by
Conclusions Rifabutin, given prophylactically, reduces the frequency of disseminated M. avium complex infection in patients with AIDS and CD4 counts
200 per cubic millimeter. The primary end point was M. avium complex bacteremia as assessed monthly by blood culture. The secondary end points were signs and symptoms associated with disseminated M. avium complex infection, adverse events, hospitalization, and survival. 
20 percent), decline in the hemoglobin level (by more than 10 percent), 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 (P = 0.086). The distribution of minimal inhibitory concentrations of rifabutin among the isolates of M. avium complex did not differ significantly between the treatment groups. 200 per cubic millimeter.
Disseminated M. avium complex infection contributes substantially to morbidity and is associated with decreased survival5. Fever, weight loss, diarrhea, anemia, and elevated alkaline phosphatase concentrations are significantly more common in patients with M. avium complex bacteremia than in those without it6. Disseminated M. avium complex infection appears to reduce the survival of patients with AIDS by approximately six months7. Although several reports have suggested that treatment of disseminated infection may reduce symptoms or prolong survival, attempts to eradicate the infection have generally been unsuccessful8,9,10,11,12,13.
Rifabutin is a semisynthetic rifamycin with demonstrated activity against M. avium complex, both in vitro and in animal models14,15,16. A randomized study of three doses of rifabutin as part of multidrug regimens to treat pulmonary M. avium complex in patients without AIDS showed that 300 mg per day produced a better clinical response than 150 mg per day17. In patients with AIDS who received either 300 mg or 150 mg of rifabutin per day as part of multidrug regimens to treat disseminated M. avium complex, survival was better among patients receiving the higher dose18. In another study of 300 to 900 mg of rifabutin per day, the drug delayed the onset of M. avium complex infection in patients with AIDS, as compared with historical controls19. These studies showed rifabutin to be well tolerated and possibly effective against M. avium complex, and they provided the scientific basis for our placebo-controlled trial of 300 mg of rifabutin per day as prophylaxis against disseminated M. avium complex infection in patients with AIDS and CD4 counts 200 cells per cubic millimeter.
Methods
Study Design
We conducted two identically designed prospective, randomized, placebo-controlled, double-blind, multicenter trials of rifabutin for the prevention of M. avium complex bacteremia in patients with AIDS. These trials were identified as study 023 and study 027. The purpose of the replication was to minimize the possibility that a few random events in a single trial would influence the results. HIV-infected men and nonpregnant women at least 18 years of age were eligible for the study if they met the following criteria: a previous AIDS-defining event20 other than disseminated M. avium complex infection, a CD4 lymphocyte count 200 per cubic millimeter, a negative tuberculin skin test ( 5 mm of induration), at least four weeks of therapy with either zidovudine or didanosine, and anti-pneumocystis prophylaxis for at least four weeks before enrollment. Two blood cultures and one stool culture for M. avium complex were performed in each patient before entry into the study, and patients were excluded if M. avium complex was isolated at either site. They were also excluded if they had a previous mycobacterial infection, antimycobacterial therapy during the four weeks before study entry, a hematocrit below 29 percent, an absolute neutrophil count below 1000 per cubic millimeter, a platelet count below 75,000 per cubic millimeter, aminotransferase levels 5 times the upper limit of normal, an alkaline phosphatase level 3 times the upper limit of normal, a bilirubin level above 2.5 mg per deciliter (42 µmol per liter), a serum creatinine level above 1.5 mg per deciliter (170 µmol per liter), or a history of allergy to any rifamycin. The protocol was approved by the institutional review board at each participating center. Written informed consent was obtained from each patient before enrollment.
The patients were randomly assigned to receive either 300 mg of rifabutin taken once daily or matching placebo, according to a prospectively generated randomization schedule. Separate schedules were generated for each investigational site, with randomly permuted blocks of four within which two patients were assigned to rifabutin and two to placebo.
Patients were evaluated at base line, one week and one month after enrollment, and at monthly intervals thereafter while they were receiving double-blind therapy. At each visit a history of adverse events, hospitalizations, intercurrent illnesses, and M. avium complex-related symptoms was obtained, and a Karnofsky performance score was determined. Changes in concomitant medications were recorded, compliance with the protocol was assessed, and a physical examination was performed. A complete blood count, urinalysis, blood chemistry panel, and blood culture for M. avium complex were also performed at each monthly visit.
M. avium complex was isolated from blood in 7H12 or 7H13 broth by a radiometric method21 at either Metpath Laboratories (Teterboro, N.J.) or the Provincial Laboratories of Public Health for Northern Alberta (Edmonton). The isolates of M. avium complex recovered were tested for antimicrobial susceptibility at the National Jewish Center for Immunology and Respiratory Medicine, Denver. This test was performed in 7H12 broth by a radiometric method described by Heifets22.
During the study the patients were not permitted to take other experimental medications or medications intended to treat or prevent mycobacterial infections. Therapy with rifampin, ethambutol, clofazimine, amikacin, or streptomycin was specifically prohibited. Up to 14 days of ciprofloxacin treatment was allowed for acute infections. Concomitant medications were not otherwise restricted.
Patients received double-blind study medication until any one of the following events occurred: the culture of M. avium complex from blood or bone marrow, the isolation of M. tuberculosis from any site, the administration of antimycobacterial therapy, serious or life-threatening toxicity either attributable to the study drug or of unknown cause, failure or inability of the patient to comply with the protocol, voluntary withdrawal, death, or termination of the study.
Patients in whom M. avium complex bacteremia developed were eligible to receive open-label rifabutin therapy on a compassionate-use basis. The concurrent use of other antimycobacterial drugs was recommended. Patients in whom M. avium complex bacteremia had not developed when the studies were completed were eligible to receive open-label rifabutin prophylaxis on a compassionate-use basis. Patients who elected to receive rifabutin were followed monthly; others were followed every three months.
Statistical Analysis
The primary end point in each study was the development of M. avium complex bacteremia. The time to this end point was calculated by the Kaplan-Meier method. The log-rank test was used to compare the time to the development of bacteremia in the two groups. Relative risks (i.e., hazard ratios) and the influence of covariates on these risks were calculated by the Cox proportional-hazards method. All statistical tests were two-tailed.
The statistical analyses were performed twice: first, in an intention-to-treat analysis of all patients randomly assigned to receive the study drug, except those with M. avium complex bacteremia at base line; and then, with all observations censored 30 days after the last dose of the study drug -- that is, an analysis limited to the double-blind phase. In the latter analysis, patients in whom M. avium complex bacteremia developed more than 30 days after the last dose of medication given in the double-blind phase were considered not to be infected (i.e., their data were censored) at the end of the treatment period. These primary evaluations of efficacy were first performed independently within each of the two trials, and then for the combined study populations.
In the analyses of secondary efficacy end points, data from the two studies were combined. The effect of rifabutin was assessed by comparing the time to the initial development of each end point in the patients assigned to rifabutin or placebo with use of the log-rank test. In the case of fever, fatigue, night sweats, diarrhea, or abdominal pain, the end point was considered to have occurred if it was reported to have been moderate, severe, or very severe. In the case of laboratory measures, the end point was determined to have occurred if the abnormality was considered serious or life-threatening.
When these studies were designed, the incidence of M. avium complex bacteremia in the placebo group was projected to be 30 percent at two years. A clinically meaningful result was taken to be a 50 percent decrease in the incidence of bacteremia in the treatment group, to 15 percent after two years. A study in which 68 cases of M. avium complex bacteremia occurred would have a 90 percent power to detect this difference by the log-rank test at a two-tailed significance level of 0.05. To accumulate this number of events within a two-year period, an enrollment of up to 750 patients (375 per group) in each study was permitted. The target sample size, however, involved the number of cases of M. avium complex bacteremia rather than the number of patients enrolled.
Results
Study Populations
From February 1990 through October 1991, 31 centers in the United States enrolled 590 patients in study 023. Two hundred ninety-eight patients were randomly assigned to receive placebo and did so for a mean of 214 days (range, 6 to 621). Two hundred ninety-two patients were randomly assigned to rifabutin and received that drug for a mean of 231 days (range, 1 to 595). The base-line characteristics of these 590 patients are presented in Table 1. Their median age was 37 years; the majority were white and male. The median CD4 cell count at base line was slightly higher in the rifabutin group (P = 0.048), but when the patients were grouped according to the range of their CD4 cell counts (e.g., 0 to 24 cells per cubic millimeter), the distribution did not differ significantly between the placebo group and the rifabutin group by the chi-square test. There were no other significant differences between the two groups with respect to base-line characteristics.
|
M. avium Complex Bacteremia
In study 023, five of the patients randomly assigned to placebo and five of those randomly assigned to rifabutin were found to have M. avium complex bacteremia at base line; these events were excluded from subsequent analysis. Among the 298 patients randomly assigned to placebo, M. avium complex bacteremia developed in 51 (17 percent), as compared with 24 (8 percent) of the 292 patients randomly assigned to rifabutin (Table 2). According to the intention-to-treat analysis, the relative risk of M. avium complex bacteremia in the patients who received rifabutin as compared with placebo was 0.43 (95 percent confidence interval, 0.26 to 0.70; P<0.001). When events that occurred more than 30 days after the last dose of the study drug were excluded, there were 41 patients with M. avium complex bacteremia in the placebo group and 16 in the rifabutin group. In this analysis limited to the double-blind phase, the relative risk of bacteremia with rifabutin was 0.36 (95 percent confidence interval, 0.20 to 0.64; P<0.001).
|
A summary of these results in the combined study populations appears in Figure 1. In both the separate and the combined populations, covariate adjustment of the proportional-hazards analysis for age, race, sex, base-line CD4 count, and duration of AIDS diagnosis did not affect the statistical significance of the treatment effect. The median CD4 cell count in the placebo group at the time of M. avium complex bacteremia was 12 per cubic millimeter (range, 1 to 168), and the median CD4 count in the patients assigned to rifabutin at the time bacteremia developed was 15 per cubic millimeter (range, 2 to 198). The patients lost to follow-up during the study were distributed equally between the placebo group and the rifabutin group (7 percent and 6 percent, respectively), as were the patients who were excluded from the study because of the concomitant use of unapproved medications (4 percent and 3 percent).
|
Clinical Manifestations of M. avium Complex Infection
Signs, symptoms, and laboratory abnormalities associated with M. avium complex bacteremia that were significantly reduced by rifabutin prophylaxis included moderate, severe, or very severe fever and fatigue; a decrease of 20 or more points from base line in the Karnofsky performance score; anemia, defined as a 10 percent decrease in the hemoglobin level from base line; elevation of the alkaline phosphatase level, defined as a 50 percent increase from base line; and hospitalization for any cause within 30 days of the last dose of the study drug (Table 3). One hundred eighty patients assigned to rifabutin were hospitalized at least once during the trial, as compared with 218 patients receiving placebo. The occurrence of weight loss of 5 percent or more from the base-line value was not significantly reduced in the rifabutin group, however, nor was the incidence of night sweats, abdominal pain, or diarrhea.
|
At the time of the final analysis, a total of 426 patients (200 in the rifabutin group and 226 in the placebo group) were known to have died. There was no significant difference between the groups with respect to overall survival (P = 0.137). When the analysis was restricted to survival during the double-blind phase (by censoring the data 30 days after the last dose of study drug), there were 33 deaths in the rifabutin group and 47 deaths in the placebo group (Figure 2). The analysis limited to the double-blind phase yielded a relative risk of death of 0.68 with rifabutin treatment (95 percent confidence interval, 0.43 to 1.06; P = 0.086).
|
Overall, the incidence rate for the occurrence of at least one adverse event was similar in the two groups (placebo, 50 percent; rifabutin, 51 percent). These adverse events do not include orange discoloration of urine, an anticipated side effect of the rifamycins, which was observed in 6 percent of the placebo group and 30 percent of the rifabutin group. Therapy was discontinued because of adverse events in 16 percent of the rifabutin group and 8 percent of the placebo group. The primary reasons for the discontinuation of rifabutin were rash (4 percent), gastrointestinal intolerance of the drug (3 percent), and neutropenia (2 percent). Myalgias, eructation, and dysgeusia, although reported in 3 percent or less of the patients receiving rifabutin, were more common in that group than in the placebo group.
Drug Susceptibility
The minimal inhibitory concentration of rifabutin was determined for 59 of the 102 first M. avium complex isolates from patients randomly assigned to placebo and for 29 of the 48 first isolates from patients randomly assigned to rifabutin. In the others, organisms could not be retrieved for analysis. The distribution of the minimal inhibitory concentrations of rifabutin among the organisms isolated from patients did not differ significantly between groups.
Discussion
These two placebo-controlled trials demonstrate that prophylaxis against disseminated M. avium complex infection can be achieved in patients with AIDS. The reduction in bacteremia, the primary end point of this trial, was accompanied by several clinical benefits. Fever and fatigue, which contribute substantially to morbidity, were significantly reduced in patients randomly assigned to treatment with rifabutin. Hospitalizations, which increase the cost of care in addition to reducing the quality of life, were also significantly reduced among patients assigned to rifabutin, as was the incidence of anemia, which can limit antiretroviral therapy and impair the patient's activity. If rifabutin had merely inhibited the growth of M. avium complex in blood cultures without preventing infection in the body, these beneficial effects should not have been observed.
Thirty-five patients who received rifabutin had M. avium complex bacteremia nonetheless. Failure to comply with the protocol, inadequate drug absorption, and infection with rifabutin-resistant organisms may have caused some failures of prophylaxis, but there was no evidence that resistance to rifabutin emerged during prophylaxis.
Rifabutin was well tolerated. Adverse events, notably rash, gastrointestinal intolerance to the drug, and neutropenia, were infrequent, and no hepatotoxicity was detected. In fact, elevation of alkaline phosphatase was significantly more common in the placebo group than in the rifabutin group, possibly because rifabutin prevented hepatic dysfunction due to disseminated M. avium complex infection.
Although we could demonstrate no overall survival benefit with rifabutin prophylaxis in the blinded analysis, there was a trend toward improved survival with rifabutin in the analysis limited to the double-blind phase. Because patients were screened monthly for M. avium complex bacteremia, even when they were asymptomatic, disseminated infection with M. avium complex may have been diagnosed earlier than in routine clinical practice. The combination of early intervention and the use of newly available therapies, including the macrolides as well as rifabutin, may have reduced mortality from disseminated infection in those who contracted it. Furthermore, all patients in whom M. avium complex bacteremia had not developed by the completion of the study were offered rifabutin prophylaxis, which may have further reduced differential mortality.
Rifamycins are known to have pharmacokinetic interactions with other drugs23. Although rifabutin has less activity in this regard in normal subjects than rifampin,24 rifabutin has been found to cause a 32 percent decrease in the area under the curve of the blood zidovudine concentration plotted against time25. The clinical importance of this finding is uncertain. However, the data on both safety and survival from these prophylaxis trials support the conclusion that rifabutin is safe as well as effective in patients with AIDS.
Using rifabutin in persons with active tuberculosis may promote the development of rifampin-resistant strains26. It is therefore imperative that active tuberculosis be excluded in an HIV-infected patient before rifabutin monotherapy is begun. Since most strains of M. tuberculosis are inhibited by rifabutin,27 rifabutin prophylaxis may also prevent the development of tuberculosis in those without active disease. However, the incidence of tuberculosis in this study population was too low to test this hypothesis.
The studies reported here show that rifabutin was safe, well tolerated, and effective in reducing the incidence of M. avium complex bacteremia and its associated morbidity. On the basis of these results, rifabutin can be recommended for the prevention of disseminated M. avium complex infection in patients with AIDS who have CD4 counts 200 cells per cubic millimeter.
Supported by grants from Adria Laboratories and by the Canadian HIV Clinical Trials Network. Dr. Cameron is a Career Scientist of the Ontario Ministry of Health.
Source Information
From the University of Texas Southwestern Medical Center, Dallas (S.D.N.); Ottawa General Hospital, Ottawa, Ont. (D.W.C.); the Veterans Affairs Medical Center, Washington, D.
(F.M.G.); the Veterans Affairs Medical Center, San Francisco (P.M.S.); Public Health Administration and Disease Control, Denver (D.L.C.); Johns Hopkins School of Medicine, Baltimore (R.E.C.); Community Research Initiative of South Florida, Miami (P.D.S.); Community Research Initiative, New York (B.B.); the Pennsylvania Hospital, Philadelphia (J.J.S.); San Diego Community Research Group, San Diego, Calif. (D.D.P.); TheraFirst Medical Center, Fort Lauderdale, Fla. (A.L.); the Long Island Jewish Medical Center, New Hyde Park, N.Y. (F.P.S.); and private practice in Annandale, Va. (L.J.E.), New York (D.L.K.), and Atlanta (W.G.W.). Additional clinical investigators and study centers are listed in the Appendix.
Address reprint requests to Dr. Gordin at the Veterans Affairs Medical Center, 50 Irving St., NW, Washington, DC 20422.
References
In addition to the authors, the following principal clinical investigators and institutions participated in this study.
D. Armstrong, Memorial Sloan-Kettering Cancer Center, New York; B. Bernstein, Medical College of Wisconsin, Milwaukee; C.L. Besch, Tulane University Medical Center, New Orleans; C. Brosgart, East Bay AIDS Center, Berkeley, Calif.; A. Burnside, Jr., Columbia, S.C.; J. Carey, Cleveland; P. Casner, Texas Tech University, El Paso; T. Chew, St. Francis Memorial Hospital, San Francisco; P. Cimoch, Center for Special Immunology, Irvine, Calif.; C. Cohen, New England Community Research Initiative, Boston; M. Conant, San Francisco; D. Conway, Community Research Initiative of Toronto, Toronto; L. Crane, Comprehensive AIDS Alliance of Detroit, Detroit; S. Deresinski, AIDS Community Research Consortium, Redwood City, Calif.; I. Fong, St. Michael's Hospital, Toronto; J. Galpin, Harbor General Hospital, Los Angeles; J. Gill, University of Calgary Health Sciences Center, Calgary, Alta.; A. Glatt, Nassau County Medical Center, East Meadow, N.Y.; S. Hall, Veterans Affairs Medical Center, Reno, Nev.; S. Hauptman, Thomas Jefferson University Hospital, Philadelphia; J. Havlik, Grady Memorial Hospital, Atlanta; K. Henry, St. Paul-Ramsey Medical Center, St. Paul, Minn.; R. Hewitt, Erie County Medical Center, Buffalo, N.Y.; L. Hobratsch, Scott and White Hospital, Temple, Tex.; P. Jensen, Veterans Affairs Medical Center, San Francisco; N. Klimas, Veterans Affairs Medical Center, Miami; M. Kubota, Santa Rosa, Calif.; R. Lalonde, Montreal Chest Hospital Centre, Montreal; S. Lea, Waco Medical Group, Waco, Tex.; R. Luskin-Hawk, Chicago Community Program for Clinical Research on AIDS, Chicago; L. Lutwick, Maimonides Medical Center, Brooklyn, N.Y.; E. Matthew, HIV Study Group, Austin, Tex.; M. Miller, Jewish General Hospital, Montreal; S. Miller, Baylor College of Medicine, Houston; J. Montaner, St. Paul's Hospital, Vancouver, B.C.; K. Murphy, Nelson-Tebedo Community Clinic, Dallas; A. Najjar, Houston Clinical Research Network, Houston; M. Para, Ohio State University Hospitals, Columbus; G. Perez, North Jersey Community Research Initiative, Newark, N.J.; J. Phair, Northwestern University Medical School, Chicago; S. Pomerantz, Maricopa County Medical Center, Phoenix, Ariz.; A. Rachlis, Sunnybrook Health Science Center, Toronto; W. Reiter, Center for Special Immunology, Fort Lauderdale, Fla.; F. Rhame, University of Minnesota, Minneapolis; D. Romig, AIDS Wellness Program, Santa Fe, N.M.; S. Sathe, Veterans Affairs Medical Center, Brooklyn, N.Y.; S. Scheibel, Community Health Network, Rochester, N.Y.; W. Schlech, Victoria General Hospital, Halifax, N.S.; S.D. Shafran, Provincial Laboratories of Public Health for Northern Alberta, Edmonton; F. Smaill, McMaster University Medical Center, Hamilton, Ont.; D. Smith, Research Medical Center, Kansas City, Mo.; J.A. Smith, University of Texas Health Science Center, San Antonio; L. Smith, Veterans Affairs Medical Center, Albany, N.Y.; M. Thompson, AIDS Research Consortium of Atlanta, Atlanta; E. Toma, Hotel-Dieu de Montreal, Montreal; L. Waites, San Francisco; J. Wallace, Olive View Medical Center, Sylmar, Calif.; J. Weisman, Pacific Oaks Medical Group, Beverly Hills, Calif.; and N. Zide, Memorial Hospital, Hollywood, Fla.
| |||||||||||||||||||||||||||||||||||||||||||
Related Letters:
Rifabutin Prophylaxis and Uveitis
Fuller J. D., Stanfield L. E.D., Craven D. E.
Extract |
Full Text
N Engl J Med 1994;
330:1315-1316, May 5, 1994.
Correspondence
Fluconazole and Enhanced Effect of Rifabutin Prophylaxis
Narang P. K., Trapnell C. B., Schoenfelder J. R., Lavelle J. P., Bianchine J. R.
Extract |
Full Text
N Engl J Med 1994;
330:1316-1317, May 5, 1994.
Correspondence
Rifabutin Prophylaxis against Mycobacterium avium Complex Infection
Goldschmidt R. H., Hearst N., Chambers D. B., Zbrozek A. S., von Reyn C. F., Brown S. T., Arbeit R. D., Reichman L. B., McDonald R. J., Mangura B. T., Nightingale S. D., Cameron D. W.
Extract |
Full Text
N Engl J Med 1994;
330:436-438, Feb 10, 1994.
Correspondence
This article has been cited by other articles:
HOME | SUBSCRIBE | SEARCH | CURRENT ISSUE | PAST ISSUES | COLLECTIONS | PRIVACY | TERMS OF USE | HELP | beta.nejm.org Comments and questions? Please contact us. The New England Journal of Medicine is owned, published, and copyrighted © 2010 Massachusetts Medical Society. All rights reserved. |
Previous
