Synergistic Activity of R Combined with Pyrazinamide against Murine Tuberculosis

Transcription

1 ANTIMICROBIAL AGENTS AND CHEMOTHERAPY, Mar. 2007, p Vol. 51, No /07/$ doi: /aac Copyright 2007, American Society for Microbiology. All Rights Reserved. Synergistic Activity of R Combined with Pyrazinamide against Murine Tuberculosis M. Ibrahim, 1 K. Andries, 2 N. Lounis, 2 A. Chauffour, 1 C. Truffot-Pernot, 1 V. Jarlier, 1 and N. Veziris 1 * Laboratoire de Bactériologie, Faculté demédecine Pitié-Salpêtrière, Université Pierre et Marie Curie Paris 6 and Centre National de Référence de la Résistance des Mycobactéries aux Antituberculeux, Groupe Hospitalier Pitié-Salpêtrière, Assistance Publique Hôpitaux de Paris, Paris, France, 1 and Antimicrobial Research, Tibotec Pharmaceuticals, Ltd., Johnson & Johnson, Turnhoutseweg 30, 2340 Beerse, Belgium 2 Received 20 July 2006/Returned for modification 8 September 2006/Accepted 22 November 2006 In previous studies, the diarylquinoline R (also known as TMC207) was demonstrated to have high bactericidal activity when combined with first- or second-line antituberculous drugs. Here we extend the evaluation of R in the curative model of murine tuberculosis by assessing the activities of one-, two-, and three-drug combinations containing R and isoniazid (INH), rifampin (RIF), pyrazinamide (PZA), or moxifloxacin (MXF) in the setting of a high initial bacillary load (7.2 log 10 CFU). Two months of treatment with the combinations R PZA, R PZA-INH, R PZA-RIF, or R PZA-MXF resulted in culture-negative lung homogenates in 70 to 100% of the mice, while mice treated with INH-RIF-PZA (the reference regimen) or RIF-MXF-PZA remained culture positive. Combinations including R but not PZA (e.g., R INH-RIF and R MXF-RIF) were less active than R PZA-containing regimens administered either alone or with the addition of INH, RIF, or MXF. These results reveal a synergistic interaction between R and PZA. Three-drug combinations containing these two drugs and INH, RIF, or MXF have the potential to significantly shorten the treatment duration in patients, provided that these results can be confirmed in long-term experiments including periods of relapse. Tuberculosis (TB) is the second-ranked infectious disease leading to mortality, after AIDS (1, 17). It is estimated that one third of the world s population is infected with Mycobacterium tuberculosis 30 million people have active disease, 8 million develop new disease yearly, and 2 to 3 million die annually from TB (17). Although the existing standard regimen is very active against TB, the long treatment duration (6 months), the toxicity, and the potential for drug-drug interactions, particularly in the setting of antiretroviral treatment, are all factors underlining the need for new antituberculous drugs. Priorities for TB drug development are the following: (i) shorten treatment duration, (ii) increase compliance by enabling intermittent therapy, and (iii) identify drugs with a novel mechanism of action to ensure activity against drug-resistant Mycobacterium tuberculosis. Many attempts have been made to discover new antituberculous drugs (2). However, no new agents have been introduced into the first-line regimen since the discovery of rifampin (RIF) (13, 15), even though the results of recent trials on moxifloxacin (MXF) are promising (12). R (also known as TMC207) belongs to a new family of antituberculosis drugs, the diarylquinolines (1). The compound inhibits the ATP synthase of Mycobacterium tuberculosis, a mechanism of action that differs from that of other antituberculous drugs. Consequently, R is active against both sensitive and resistant Mycobacterium tuberculosis strains. * Corresponding author. Mailing address: Laboratoire de Bactériologie, Faculté de Médecine Pitié-Salpêtrière, 91 Boulevard de l Hôpital, Paris Cedex 13, France. Phone: (33) Fax: (33) veziris@chups.jussieu.fr. Published ahead of print on 18 December When R was combined with the standard first-line antituberculosis drugs RIF, isoniazid (INH), and pyrazinamide (PZA), the mice were rendered culture negative after 2 months of treatment in the setting of an initial bacillary load of 6 log 10. This promising result suggested that combination therapy including R has the potential to reduce the duration of TB treatment. We carried out an experiment with a high initial bacillary load in order to determine more extensively the most-promising R containing drug combinations. Two- and threedrug combinations with first-line antituberculous drugs or MXF were systematically tested. Particular attention was paid to the combination of R and PZA, because our previous data suggested that these drugs might interact synergistically. MATERIALS AND METHODS Antimicrobial agents. R was synthesized by Johnson & Johnson (Beerse, Belgium), while the other compounds were purchased from the following manufacturers: INH from Laphal (Allauch, France), RIF and PZA from Aventis (Antony, France), and MXF from Bayer (Puteaux, France). Infection of mice. The H37Rv strain of Mycobacterium tuberculosis was grown on Lowenstein-Jensen medium. Colonies were subcultured in Dubos broth (Diagnostics Pasteur, Paris, France) for 7 days at 37 C. The turbidity of the resulting suspension was adjusted with normal saline to match that of a standard 1-mg/ml suspension of Mycobacterium bovis BCG and was further diluted with normal saline to obtain a 0.2-mg/ml suspension for mouse inoculation. Four hundred forty female 4-week-old Swiss mice purchased from the Janvier Breeding Center (Le Genest-Saint-Isle, France) were intravenously infected in the tail vein with 0.5 ml of bacterial suspension containing approximately CFU of Mycobacterium tuberculosis H37Rv. Chemotherapy. After 2 weeks of infection, the mice were randomly allocated to 21 groups. A negative control group consisted of 40 mice infected but left untreated. Five groups received monotherapy with RIF, INH, PZA, MXF, or R Eight groups received combinations of two of the above-mentioned antibiotics (RIF-INH, RIF-PZA, INH-PZA, RIF-MXF, R RIF, R INH, R207910Z-PZA, 1011

2 1012 IBRAHIM ET AL. ANTIMICROB. AGENTS CHEMOTHER. TABLE 1. Bacterial counts in the lungs after monotherapy given 5 days/week for murine TB a or R MXF), and seven groups received three-drug combinations (RIF-INH- PZA, RIF-MXF-PZA, R INH-RIF, R MXF-RIF, R INH- PZA, R RIF-PZA, or R MXF-PZA). Treatment was delayed until 2 weeks after infection to achieve a large and established bacterial population. Compound R was prepared monthly in a hydroxypropyl- -cyclodextrin solution and kept at 4 C. Other drug suspensions were prepared weekly and kept at 4 C. All drugs were administered by oral gavage 5 days per week. RIF was administered 1 h before other drugs to avoid drug-drug interactions (3, 5, 11). The dose of the drugs was selected to provide areas under the concentration-time curve (AUC) for mice that were comparable to those achievable for patients at the usual dosing (11, 14). The following doses were selected: R207910, 25 mg/kg of body weight; RIF, 10 mg/kg; PZA, 150 mg/kg; MXF, 100 mg/kg; and INH, 25 mg/kg. All surviving mice were killed after 2 months of treatment. Assessment of infection and treatment. To provide baseline values before initiation of chemotherapy, 10 control mice were killed on day 1 after infection, and 20 were killed on day 14 (days 13 and 0, respectively, in relation to the initiation of treatment). In each treatment group, 10 mice were sacrificed after 1 month of treatment, and 10 were sacrificed after 2 months of treatment. The severity of infection and the treatment effect were assessed by survival rate, spleen weight, gross lung lesions (0, no lesions;, fewer than 10 tubercles;, 10 to 50 tubercles;, more than 50 tubercles), and the numbers of CFU in the lungs. Lungs were aseptically removed and homogenized by a standard procedure (4). The enumeration of CFU was done as previously described (8), but for mice receiving R containing drug combinations, both lungs were harvested, and undiluted lung homogenates were cultivated to increase the sensitivity and to assess the culture negativity more stringently. Statistical analysis. Mean CFU counts were compared using the Mann- Whitney U test. Differences were considered significant at the 95% level of confidence. Proportions of mice with positive cultures after 2 months of treatment were compared using the chi-square test. RESULTS J R M H Z a Groups were sorted based on efficacy after 2 months. Survival rate. All untreated mice died between day 5 and day 35. Monotherapy with PZA alone did not completely prevent mortality, and 7 out of the 20 mice treated with this drug died of TB on days 8, 10, 11, 12, 15, 16, and 22, respectively. Few mice died in the other treated groups: one mouse died on day 8 in the RIF-treated group, one mouse on day 1 in the INH-treated group, one mouse on day 5 in the RIF-PZA-treated group, three mice on days 2 and 3 in the RIF-MXF-treated group, one mouse on day 16 in the R INH-treated group, two mice on days 8 and 16 in the R MXF-treated group, one mouse on day 15 in the RIF-INH-PZA-treated group, three mice on days 4, 5, and 19 in the R INH-PZA-treated group, one mouse on day 40 in the R MXF-RIF-treated group, and one mouse on day 8 in the R MXF-PZA-treated group. During the first 10 days of treatment, mortality was attributed to TB, as mice harbored lung lesions (score, ) and had low TABLE 2. Bacterial counts and proportion of mice with positive cultures in the lungs after treatment with two-drug combinations given 5 days/week for murine TB a J JZ , JR JH JM RZ RM RH HZ a Groups were sorted based on efficacy after 2 months. R monotherapy was added for comparison. body weights and important splenomegaly. The five deaths after day 10 were attributed to gavage accidents. Mean spleen weights. The mean spleen weight ( standard deviation) of infected mice increased more than fourfold during the first 2 weeks after infection (from mg to mg). One month of treatment with any regimen, except PZA monotherapy, resulted in a significant decrease in mean spleen weight, from 524 to less than 350 mg. The second month of treatment did not further reduce these spleen weights. Gross lung lesions. At the start of treatment, 2 weeks after the infection, all mice had developed massive gross lung lesions (score, ). All treated mice had less-severe lesions after 1 month of therapy. The reduction of lesions was more and less pronounced in the groups receiving R alone and PZA alone, respectively, than in those receiving monotherapy with RIF, INH, or MXF. All combinations of two or three drugs were able to markedly reduce the number of gross lung lesions after 1 month of therapy. An additional month of monotherapy with R207910, RIF, INH, or MXF further reduced the gross lung lesions but did not cure them completely, whereas mice treated with PZA alone developed lung lesions (score, ) that were more severe than at the start of treatment. All combinations of drugs were able to further decrease the number of gross lung lesions that were initially present at 1 month. Enumeration of CFU in the lungs. The mean CFU count in the lungs was 6.0 log 10 the day after inoculation and reached 7.2 log 10 2 weeks later. R alone reduced the CFU count by 3.1 log 10 during the first month of treatment and by 5.0 log 10 after 2 months (Table 1). At the end of the 2 months of treatment, the bacillary load was significantly lower in mice treated with R alone than in mice treated with the other antibiotics (RIF, INH, MXF, or PZA, P ). The activity of R was not increased by adding RIF, INH, or MXF. In contrast, the addition of PZA significantly enhanced the activity of R207910, as shown in Table 2. After only 1 month of treatment, the CFU decrease was 5.6 log in mice treated with R PZA versus 3.1 log in mice treated with R (P 0.02) and 1.1 log in mice treated with PZA alone. After 2

3 VOL. 51, 2007 R AND PYRAZINAMIDE AGAINST MURINE TUBERCULOSIS 1013 TABLE 3. Bacterial counts and proportion of mice with positive cultures in the lungs after treatment with three-drug combinations given 5 days/week for murine TB a J JZ JZM c 78 JZR d 70 JZH e 78 JRH JRM RMZ RHZ a Groups were sorted based on efficacy after 2 months. R monotherapy and R PZA therapy were added for comparison. c Two mice out of nine with positive cultures, one with 2 CFU, the other with 1 CFU. d Three mice out of 10 with positive cultures, one with 5 CFU, the other two with 1 CFU. e Two mice out of nine with positive cultures, one with 39 CFU, the other with 1 CFU. months of treatment, the CFU count had dropped to an undetectable number in mice treated with R PZA ( 7.2 log) versus 2.3 log in mice treated with R Among the two-drug combinations not containing R207910, only RIF-PZA matched the bactericidal activity of R monotherapy after 2 months (Table 2). Considering the combinations of three drugs, the WHO s standard three-drug combination regimen of RIF-INH-PZA reduced the bacillary load by 3.4 log 10 CFU after 1 month of treatment and by 5.0 log 10 CFU after 2 months. The CFU decrease with RIF-MXF-PZA was 0.7 log less than that with RIF-INH-PZA after 1 month (P ) but 0.9 log higher than that with RIF-INH-PZA after 2 months (P ) (Table 3). The addition of RIF, INH, or MXF to the two-drug combination R PZA did not increase its bactericidal activity (P 0.05). R MXF-RIF and R INH- RIF were equipotent after 1 and 2 months of treatment (3 log 10 reduction in CFU per month) and matched the potency of R monotherapy after 1 month but were more active than R after 2 months of treatment (P and P 0.025, respectively). Importantly, the three-drug combinations containing R but not PZA (R INH-RIF and R MXF-RIF) were less active than the two-drug combination R PZA after 1 (P and P ) and 2(P and P ) months of treatment. Finally, all the three-drug combinations containing R and PZA (R INH-PZA, R RIF-PZA, and R MXF- PZA) were more active after 1 and 2 months than were the two control regimens RIF-INH-PZA and RIF-MXF-PZA (P 0.01), but they were not more active than the two-drug combination R PZA (P 0.05). At the end of the 2 months of therapy, the percentage of mice with negative lung cultures was 100% for the R PZA group and ranged between 70 and 78% for the groups where INH, RIF, or MXF was added to R PZA. For all other two- or three-drug combinations, the percentage of mice with negative lung cultures ranged between 0 and 30%. The percentages of mice with positive lung cultures after 2 months of treatment were significantly higher for the two control groups (INH-RIF-PZA and MXF-RIF-PZA) (P ) than for the R PZA group (P 0.001). The percentages of mice with positive cultures were also significantly higher when R was used without PZA (R PZA versus R INH-RIF, P 0.001; R PZA versus R MXF-RIF, P ). DISCUSSION In the first study on the efficacy of R in mice, the drug combinations R INH-PZA, R RIF-PZA, and R INH-RIF-PZA were shown to reduce the bacillary load by approximately 3 log 10 CFU per month and to render the lung culture negative after 2 months of treatment (1). However, in that first study, the initial bacillary load was only 6 log 10 CFU and the number of regimens tested was limited. In the present experiment, we increased the bacillary load 20- fold, and the whole lungs were harvested and tested at the end of treatment in order to be able to evaluate the activity of R in the setting of a high initial bacillary load. The selected doses were chosen in order to obtain pharmacokinetic parameters equivalent to those obtained in humans. Since R has not yet been tested in tuberculosis-infected patients, we used the same dose as in our first experiment, i.e., 25 mg/kg for 5 days a week, a dose which leads to plasma levels in mice equivalent to those reached in healthy human volunteers and which is well tolerated (1). The present results confirm that monotherapy with R is more active than monotherapy with any of the currently available drugs, decreasing the bacillary load by approximately 5 log 10 CFU in 2 months versus 3 log 10 CFU for RIF, which has the highest activity after R In addition, the results confirm that R monotherapy is as active as the standard regimen RIF-INH-PZA. This new study brings important results concerning R activity. The most important result is the dramatic activity of the combination of R with PZA. Indeed, R PZA reduced the bacillary load by 5.6 log 10 CFU after 1 month of treatment, a figure higher by more than 2 log 10 CFU than that obtained with the most effective drug combination that did not contain R207910, i.e., RIF-INH-PZA. R PZA was the only drug combination able to render 100 negative after 2 months. All the other R containing two-drug combinations led to negative cultures for only 20 to 30% of the mice. Such a level of efficacy has so far been obtained only after 4 months of therapy with the three-drug combination RIF-MXF-PZA (11). Thus, R and PZA clearly act synergistically against Mycobacterium tuberculosis in the murine model. We did not investigate the basis for this synergism in the current study, but N,N -dicyclohexylcarbodiimide, an aspecific inhibitor of ATP synthase, the enzyme targeted by R207910, has been shown to interact synergistically with PZA in vitro (19). PZA is known to disrupt the membrane potential (19) which is required by ATP synthase to generate ATP. Given that PZA indirectly inhibits the synthesis of ATP, the synergism of this drug with R207910, a specific ATP syn-

4 1014 IBRAHIM ET AL. ANTIMICROB. AGENTS CHEMOTHER. thase inhibitor, is not surprising. However, the level of synergism in vivo is remarkable. There has been concern about the safety of PZA in humans, as a number of patients receiving RIF-PZA chemoprophylaxis developed severe hepatitis (10). As a consequence, there has been an attempt in the mouse model to reduce the duration of PZA treatment (12). The synergism between R and PZA in the murine model was observed when both drugs were given 5 days a week over 2 months. Due to the toxicity concern with the use of PZA, the minimum dosing and duration of PZA treatment required to obtain synergism deserve further specific evaluation. It will be particularly interesting to evaluate synergism in intermittent regimens, as the pharmacokinetics of R allow intermittent administration. Another point assessed in the present study was the interaction of R with the other drugs. Based on our previous results, we were particularly interested in assessing a possible antagonism between R and RIF (1). Indeed, in our previous study, R RIF-PZA seemed to be a little less active than R INH-PZA after 1 month of treatment but rendered lung cultures negative after 2 months. In the present study, the two-drug combination R RIF reduced the bacillary load by 1 log 10 less than the two-drug combination R INH after 1 month of treatment; however, this difference was not statistically significant. After 2 months of treatment, the bacillary loads were the same in these two groups. Thus, a significant antagonism between R and RIF was not confirmed by this study. Based on these results, we can expect that if both R and RIF were used during the full duration of treatment for TB, the benefit of adding the powerful sterilizing activity of RIF to R would exceed a possible small antagonism between these two drugs during the very beginning of treatment. In the present study, the addition of RIF, INH, or MXF to the R PZA combination could not further increase the activity after 2 months of treatment. In terms of percentage of culture-positive mice, the addition of INH, RIF, or MXF to R PZA even seemed to be slightly deleterious, although the differences were not statistically significant. However, a potential benefit of combining R PZA with a third drug cannot be ruled out for two reasons. First, in a model that assesses sterilizing activity, such as in a long-term relapse experiment in the murine model, R PZA-MXF or RIF may still turn out to be more active than R PZA because of the sterilizing activity of RIF and MXF. In a study by Nuermberger and colleagues, RIF-INH-PZA led to negative cultures after 4 months of treatment of mice infected by inhalation, but an additional 2 months of treatment was needed to sterilize these mice and prevent relapses after the discontinuation of treatment. In contrast, using the RIF-MXF-PZA combination, the times required to render the mice culture negative and to prevent relapses were the same, i.e., 4 months (12). Whether R PZA-containing drug combinations will prevent relapses as soon as organs are culture negative (as with RIF-MXF-PZA) or a longer duration of treatment will be required (as with RIF-INH-PZA) remains to be determined. The second reason that can justify the addition of a third drug to R PZA is that in humans, in contrast to what is observed in the mouse model, PZA is not able to prevent the selection of mutants resistant to the companion drug (6). Consequently, a two-drug combination containing R and PZA could lead to treatment failure due to the selection of R resistant mutants. The results of the present study also raise the possibility of developing regimens without PZA or without INH and PZA. Although less active than PZA-containing regimens, the R RIF-INH and R RIF-MXF drug combinations were at least as active as the two control regimens RIF- INH-PZA and RIF-MXF-PZA. This possibility may be attractive because of the risk for severe hepatitis with PZA and the high rate of resistance to INH in many countries (7, 18). Finally, our results suggest that it may be possible to develop regimens without INH and RIF (e.g., R MXF-PZA). Such regimens would be an important alternative to the currently used combination of aminoglycoside, fluoroquinolone, ethionamide, and pyrazinamide given for 18 months, as recommended by the WHO, to multidrug-resistant (MDR) TB patients. The most effective regimen not containing INH and RIF described so far is the four-drug combination of amikacinethionamide-mxf-pza, a combination that needs to be given for 9 months to render mice culture negative (16). Interestingly, the bactericidal activity of R MXF-PZA seems equal to the bactericidal activity of R MXF-PZA-amikacin-ethionamide, a point that opens the way for an effective regimen against MDR TB without an injectable agent, which would greatly simplify the treatment (9). Moreover, in the present study, the fully orally administered R MXF- PZA regimen was more active than RIF-INH-PZA after 2 months, raising hope for an alternative oral treatment of short duration for MDR TB. In conclusion, the unprecedented activity of the combination of R and PZA demonstrated in the mouse model opens several possibilities for treating susceptible or MDR TB with shorter regimens than those currently recommended by the WHO. Further relapse experiments using R PZA-containing regimens associated with RIF for susceptible TB or with MXF and/or ethionamide for MDR TB will help to identify the most interesting regimens to be tested in clinical trials. ACKNOWLEDGMENTS This work was supported by a grant from Johnson & Johnson Pharmaceuticals, Beerse, Belgium. The laboratory is supported by annual grant EA 1541 from Université Paris 6, Pierre et Marie Curie, Paris, France. We thank Solene Maucolin for technical assistance. The animal experiment guidelines of the Faculté demédecine Pitié- Salpêtrière were followed. REFERENCES 1. Andries, K., P. Verhasselt, J. Guillemont, H. W. Gohlmann, J. M. Neefs, H. Winkler, J. Van Gestel, P. Timmerman, M. Zhu, E. Lee, P. Williams, D. de Chaffoy, E. Huitric, S. Hoffner, E. Cambau, C. Truffot-Pernot, N. Lounis, and V. Jarlier A diarylquinoline drug active on the ATP synthase of Mycobacterium tuberculosis. Science 307: Ballell, L., R. A. Field, K. Duncan, and R. J. Young New smallmolecule synthetic antimycobacterials. Antimicrob. Agents Chemother. 49: Dhillon, J., J. M. Dickinson, K. Sole, and D. A. Mitchison Preventive chemotherapy of tuberculosis in Cornell model mice with combinations of rifampin, isoniazid, and pyrazinamide. Antimicrob. Agents Chemother. 40: Grosset, J., C. Truffot, J. Fermanian, and H. Lecoeur Sterilizing activity of the main drugs on the mouse experimental tuberculosis. Pathol. Biol. 30: (In French.) 5. Grosset, J., C. Truffot-Pernot, C. Lacroix, and B. Ji Antagonism

5 VOL. 51, 2007 R AND PYRAZINAMIDE AGAINST MURINE TUBERCULOSIS 1015 between isoniazid and the combination pyrazinamide-rifampin against tuberculosis infection in mice. Antimicrob. Agents Chemother. 36: Grosset, J., C. Truffot-Pernot, S. Poggi, H. Lecoeur, and C. Chastang Prevention of resistance to rifampicin by pyrazinamide in experimental tuberculosis in the mouse. Rev. Mal. Respir. 2: (In French.) 7. Ijaz, K., J. A. Jereb, L. A. Lambert, W. A. Bower, P. R. Spradling, P. D. McElroy, M. F. Iademarco, T. R. Navin, and K. G. Castro Severe or fatal liver injury in 50 patients in the United States taking rifampin and pyrazinamide for latent tuberculosis infection. Clin. Infect. Dis. 42: Lecoeur, H. F., C. Truffot-Pernot, and J. H. Grosset Experimental short-course preventive therapy of tuberculosis with rifampin and pyrazinamide. Am. Rev. Respir. Dis. 140: Lounis, N., N. Veziris, A. Chauffour, C. Truffot-Pernot, K. Andries, and V. Jarlier Combinations of R with drugs used to treat multidrugresistant tuberculosis have the potential to shorten treatment duration. Antimicrob. Agents Chemother. 50: McElroy, P. D., K. Ijaz, L. A. Lambert, J. A. Jereb, M. F. Iademarco, K. G. Castro, and T. R. Navin National survey to measure rates of liver injury, hospitalization, and death associated with rifampin and pyrazinamide for latent tuberculosis infection. Clin. Infect. Dis. 41: Nuermberger, E. L., T. Yoshimatsu, S. Tyagi, R. J. O Brien, A. N. Vernon, R. E. Chaisson, W. R. Bishai, and J. H. Grosset Moxifloxacin-containing regimen greatly reduces time to culture conversion in murine tuberculosis. Am. J. Respir. Crit. Care Med. 169: Nuermberger, E. L., T. Yoshimatsu, S. Tyagi, K. Williams, I. Rosenthal, R. J. O Brien, A. A. Vernon, R. E. Chaisson, W. R. Bishai, and J. H. Grosset Moxifloxacin-containing regimens of reduced duration produce a stable cure in murine tuberculosis. Am. J. Respir. Crit. Care Med. 170: Stover, C. K., P. Warrener, D. R. VanDevanter, D. R. Sherman, T. M. Arain, M. H. Langhorne, S. W. Anderson, J. A. Towell, Y. Yuan, D. N. McMurray, B. N. Kreiswirth, C. E. Barry, and W. R. Baker A small-molecule nitroimidazopyran drug candidate for the treatment of tuberculosis. Nature 405: Truffot-Pernot, C., B. Ji, and J. Grosset Activities of pefloxacin and ofloxacin against mycobacteria: in vitro and mouse experiments. Tubercle 72: Tyagi, S., E. Nuermberger, T. Yoshimatsu, K. Williams, I. Rosenthal, N. Lounis, W. Bishai, and J. Grosset Bactericidal activity of the nitroimidazopyran PA-824 in a murine model of tuberculosis. Antimicrob. Agents Chemother. 49: Veziris, N., C. Truffot-Pernot, A. Aubry, V. Jarlier, and N. Lounis Fluoroquinolone-containing third-line regimen against Mycobacterium tuberculosis in vivo. Antimicrob. Agents Chemother. 47: World Health Organization Global tuberculosis control: surveillance, planning, financing. WHO report WHO/HTM/TB/ World Health Organization, Geneva, Switzerland. 18. Yee, D., C. Valiquette, M. Pelletier, I. Parisien, I. Rocher, and D. Menzies Incidence of serious side effects from first-line antituberculosis drugs among patients treated for active tuberculosis. Am. J. Respir. Crit. Care Med. 167: Zhang, Y., M. M. Wade, A. Scorpio, H. Zhang, and Z. Sun Mode of action of pyrazinamide: disruption of Mycobacterium tuberculosis membrane transport and energetics by pyrazinoic acid. J. Antimicrob. Chemother. 52: