© 2003 American Thoracic Society
Rifampin and Pyrazinamide for Treatment of Latent Tuberculosis InfectionIs It Safe?Division of Pulmonary and Critical Care Medicine San Francisco General Hospital University of California San Francisco, California As tuberculosis case rates in the United States have decreased (1), the focus of prevention and control efforts has shifted toward the identification and treatment of persons with latent tuberculosis infection who are at increased risk for developing tuberculosis. For years, isoniazid has been the treatment of choice for latent tuberculosis infection. Treatment of latent tuberculosis infection with isoniazid is highly efficacious, preventing tuberculosis in up to 93% of those who receive it, depending on the duration of and adherence to therapy (2, 3). This benefit is offset, however, by isoniazid's association with hepatotoxicity, which occurs in less than 1% of patients under program conditions (4, 5). The current guidelines of the American Thoracic Society and CDC (6) recommend targeted tuberculin skin testing among persons at high risk for developing tuberculosis and if positive, treatment of latent tuberculosis infection. When the guidelines were published 2 years ago, a 2-month regimen of rifampin and pyrazinamide was offered as another option, in addition to isoniazid, for treatment of latent tuberculosis infection. Rifampin/pyrazinamide regimens have been demonstrated to be as efficacious as isoniazid and without major toxicities in human immunodeficiency virus (HIV)–infected persons in three randomized clinical trials conducted in several countries (7–9). Although the efficacy of rifampin/pyrazinamide had never been demonstrated in adults without HIV infection, experts believed that this therapy would be equally efficacious and well tolerated among this much larger group of patients with latent tuberculosis infection. Unfortunately, shortly after the American Thoracic Society/CDC recommendations were published, cases of severe liver injury, including five deaths, were reported to CDC among patients treated with the rifampin/pyrazinamide regimen (10). Revised guidelines recommended that rifampin/pyrazinamide not be used in anyone with underlying liver disease, history of alcoholism, or isoniazid-associated liver injury. Furthermore, it was recommended that no more than a 2-week supply of rifampin/pyrazinamide be dispensed and that liver enzymes be tested at baseline and after 2, 4, and 6 weeks of treatment to detect early liver injury. In this issue of AJRCCM (pp. 824–827), Stout and colleagues (11) evaluated the frequency of drug-induced hepatotoxicity and completion of therapy in 114 patients who received rifampin/pyrazinamide for 2 months for treatment of latent tuberculosis infection. The study was conducted among a group of patients at high risk for both liver disease and nonadherence to treatment; 17% had a history of heavy alcohol use and 61% were homeless. Among the 114 patients enrolled, 18 (16%) developed elevated transaminases, 9 (7.9%) had elevation of hepatic enzymes to more than five times normal, and 6 (5.3%) had symptoms of hepatitis (4 with elevation of hepatic transaminase enzymes to more than five times normal). The frequency of drug-induced hepatitis was higher than that reported previously with isoniazid therapy (4, 5). The study by Stout and colleagues (11) complements other studies (12–15) evaluating the risk of hepatotoxicity from rifampin/pyrazinamide. Despite different patient populations, three of the four studies reported strikingly similar rates of World Health Organization Grade 3 or 4 liver injury (i.e., World Health Organization Grade 3, liver enzymes more than five times normal; or Grade 4, enzymes more than ten times normal or more than five times normal with compatible symptoms [16]). Combining data from all five of these studies in 1,311 patients (including the current study by Stout) leads to a rate of Grade 3 or 4 liver injury of 5.8%. This is not a trivial problem in treating latent tuberculosis infection, especially when considering that most infected individuals have only a modest lifetime risk of developing active tuberculosis (5, 6). Two interesting questions arise from these studies. First, why are the hepatotoxicity rates so different among those without HIV infection compared with the early studies among HIV-infected persons? The most likely explanation would be that patients infected with HIV, because of their impaired host immunity, are unable to manifest the same degree of inflammation of the liver that occurs in patients without HIV infection who develop hepatitis with rifampin/pyrazinamide. Second, why are the rates of hepatotoxicity that have been observed with the two-drug regimen higher than those observed among patients with active tuberculosis who are treated with three potentially hepatotoxic drugs for 2 months (isoniazid in addition to rifampin and pyrazinamide)? Perhaps impaired host immunity, which is common in many patients with tuberculosis, plays a role here as well. Further studies, especially those examining the molecular basis for hepatotoxicity, are needed to improve our understanding of this observation. Despite the high frequency of hepatotoxicity and the fact that 61% of the patients were homeless, two-thirds of the patients in the study reported by Stout and colleagues completed therapy. Because of this relatively high completion rate, the authors concluded that the short-course regimen may be useful for high-risk, traditionally nonadherent patient groups but that careful monitoring is required. However, without a comparison group, it is not possible to conclude that this one factor, completion of therapy, would warrant the use of a more toxic regimen: the completion of therapy would have to be significantly higher than that of a comparator drug, such as isoniazid. This is not always the case. In a recently published multicenter study (12), patients were assigned in alternate weeks to receive rifampin and pyrazinamide daily for 2 months (n = 307) or isoniazid daily for 6 months (n = 282): The proportion of patients who completed therapy was 61 and 57% (p > 0.2), respectively. Given the increased costs of rifampin and pyrazinamide compared with isoniazid, the increased frequency of drug-induced hepatotoxicity and the increased monitoring required, it is unlikely that the rifampin/pyrazinamide regimen would be cost-effective for tuberculosis programs unless the adherence rates with standard isoniazid treatment regimens were exceptionally low. This study and others have demonstrated that we are still in need of new short-course treatments for latent tuberculosis infection that are efficacious, safe, and that patients can receive without intensive laboratory monitoring for toxicity. A large randomized, controlled, multicenter trial (Study 26) conducted by the CDC's Tuberculosis Trials Consortium is assessing the role of a long-acting rifamycin, rifapentine, in the treatment of latent tuberculosis infection (5). Until these results are available, it seems that we are back to where we started and that isoniazid is still the treatment of choice for most patients with latent tuberculosis infection. REFERENCES
This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
