© 2004 American Thoracic Society
Fluoroquinolones in the Treatment of TuberculosisA Study in MiceDepartment of Cellular & Molecular Medicine St. George's Hospital Medical School London, United Kingdom Nuermberger and coworkers (1) report the results of an important experiment in murine tuberculosis in the current issue of the Journal (pp. 421–426). Groups of mice were infected by the airborne route and were then treated for 6 months with regimens consisting of a 2-month initial intensive phase followed by a continuation phase in which pyrazinamide was omitted to simulate conventional practice in the treatment of pulmonary tuberculosis. The standard regimen started with isoniazid, rifampin, and pyrazinamide, and continued with isoniazid and rifampin. In other regimens, the fluoroquinolone moxifloxacin was either added throughout to the standard regimen or was substituted in three separate regimens for the isoniazid, rifampin, or pyrazinamide component of the standard regimen. The sterilizing effects of treatment were assessed as the rate at which counts of viable tubercle bacilli in the lungs and spleens fell during the 6 months of treatment and by relapses in groups left untreated for a further 3 months. The addition of moxifloxacin to the standard regimen only increased the sterilizing activity to a small extent. Furthermore, substitution for rifampicin or pyrazinamide led to lower sterilizing activity than in the standard regimen. These results suggest that moxifloxacin only increases sterilizing activity to a limited extent, an increase smaller than is obtained with either rifampicin or pyrazinamide. They suggest, but certainly do not prove, that the current three clinical trials with fluoroquinolones, two with moxifloxacin (2) and the Oflotub study with gatifloxacin, supported by the European Commission and the World Health Organization, might be unsuccessful in providing evidence sufficient for shortening the treatment period. In all of these studies the fluoroquinolone is added to the basic isoniazid, rifampin, and pyrazinamide regimen, in substitution for ethambutol in the initial phase. A remarkable and important result of the mouse study was that the sterilizing activity of the regimen of moxifloxacin, rifampicin, and pyrazinamide, with isoniazid omitted, was much greater than that of the standard regimen including isoniazid. This result suggests that this regimen could shorten treatment from 6 months to 4 months or even less. Repetitions of the study are being undertaken both to confirm that shortening treatment would not be accompanied by a higher relapse rate in mice, and also to extend the findings to another fluoroquinolone gatifloxacin. What is the explanation for this unexpected finding? The findings might be due to antagonism or interaction between isoniazid and rifampin/pyrazinamide, or between isoniazid and moxifloxacin. There are numerous known interactions between isoniazid, its metabolites, and other drugs (3), and absorption of antituberculosis drugs, particularly rifampin, is altered when multiple drugs are given together by the oral route to mice (4). Although drug interactions affecting the availability of rifampin and pyrazinamide have been excluded by estimations of blood levels of both drugs in the mice, no such studies were done on moxifloxacin. Estimations of its blood levels should be done to rule out a decrease in moxifloxacin concentrations when isoniazid is also in the regimen. In their discussion, Nuermberger and coworkers (1) favor the hypothesis that there might be an antagonism between the sterilizing activity of isoniazid and the activities of rifampin and pyrazinamide, as has been shown to occur during the first 14 days of treatment between ethambutol and the other antituberculosis drugs (5). Might isoniazid be antagonistic because it is also a bacterial cell wall inhibitor (6)? If there is a true drug–drug interaction, it should be detectable in vitro. Experiments to measure the in vitro sterilizing activity of various relevant drug combinations against old (30-day and 100-day), micro-aerophilically adapted cultures are currently under study at the Tuberculosis Research Centre, Chennai. These experiments simulate some of the in vitro tests that have been used for the measurement of the sterilizing activities of several fluoroquinolones (7), and may disclose a relevant interaction. Would any brave clinician try out the regimen with moxifloxacin replacing isoniazid? Probably not, as every initial treatment regimen used during the past 50 years has included isoniazid. Further important evidence, which could alter opinion, might be obtained from clinical studies with a regimen containing a fluoroquinolone, in which patients with strains initially resistant are not excluded. From the mouse study, the response of these resistant patients should be better than those who start with sensitive organisms. Such data will become available from the Oflotub study, which is going to compare standard treatment with a regimen to which gatifloxacin will be added. It might be easier than we think to lose isoniazid from initial treatment regimens. Thus, the evidence, from extended studies of early bactericidal activity (5) and from the unimpaired results with strains resistant only to isoniazid (8), suggests that, apart from an initial bactericidal kill during the first 2 days, isoniazid contributes nothing to the sterilizing activity of standard therapy; all of the sterilizing activity seems due to rifampin and pyrazinamide. To take a highly speculative look into the future, the removal of isoniazid from first-line regimens would leave it available instead as the basis of reliable, though lengthy treatment of patients with strains initially resistant only to rifampin. In the absence of isoniazid from the continuation phase and the consequent stimulus to provoke isoniazid resistance, such strains would gradually replace multiple resistant strains that are currently resistant to both rifampin and isoniazid, and their treatment would then become solid and safe. We might then have achieved two major aims in current therapy (9)—shortening first-line therapy and improving the treatment of resistant tuberculosis. FOOTNOTES Conflict of Interest Statement: D.A.M. has no declared conflict of interest. REFERENCES
This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
