Journal of Antimicrobial Chemotherapy (2007) 60, 1398–1401 doi:10.1093/jac/dkm393 Advance Access publication 20 October 2007 Elevated gatifloxacin and reduced concentrations in a single-dose interaction study amongst healthy volunteers

Helen McIlleron1*, Jennifer Norman1, Thomas P. Kanyok2, P. Bernard Fourie3, John Horton4 and Peter J. Smith1 Downloaded from https://academic.oup.com/jac/article/60/6/1398/824439 by guest on 27 September 2021 1Division of Clinical Pharmacology, Department of Medicine, Groote Schuur Hospital, University of Cape Town, South Africa; 2Tropical Diseases, Special Programme for Research and Training, World Health Organization, Geneva, Switzerland; 3Medicine in Need, Cambridge, MA, USA; 4Tropical Projects, Hitchin, UK

Received 8 August 2007; returned 7 September 2007; revised 21 September 2007; accepted 22 September 2007

Objectives: Pharmacokinetic drug–drug interactions were investigated between the fluoroquinolone gatifloxacin and a fixed dose combination (FDC) of rifampicin, and . Patients and methods: The single-dose pharmacokinetics of the four drugs was evaluated in an open- label three-way cross-over study amongst 22 healthy volunteers following administration of gatifloxa- cin, the FDC or the two products together. Results: Modest but potentially important drug–drug interactions affecting gatifloxacin and rifampicin concentrations were detected. The elimination rate of gatifloxacin was reduced such that the AUC from 0 h to infinity was increased with a geometric mean ratio (GMR) [90% confidence interval (CI)] of 1.14 (1.10, 1.18). Conversely, the AUC from 0 h to infinity for rifampicin was reduced (GMR: 0.81, 90% CI: 0.81, 0.96) when rifampicin, isoniazid and pyrazinamide were given together with gatifloxacin. Conclusions: Studies in patients including pharmacokinetic evaluation at steady state, efficacy and toxicity are required to determine the importance of the interactions for use of the combination of gatifloxacin, rifampicin, isoniazid and pyrazinamide in the treatment of tuberculosis.

Keywords: pharmacokinetics, pyrazinamide, isoniazid, Mycobacterium tuberculosis

Introduction Materials and methods More effective antitubercular regimens that will allow reduced The study was approved by the University of Cape Town Research treatment durations are urgently needed. Promising activities have Ethics Committee (REC REF: 005/2004) and the Secretariat been demonstrated for moxifloxacin and gatifloxacin in vitro in Committee on Research Involving Human Subjects of the World murine models and in extended bactericidal studies.1–3 Four Health Organization (RPC 077). Twenty-four volunteers were month regimens including gatifloxacin or moxifloxacin in combi- enrolled after giving their written informed consent to participate. nation with rifampicin, pyrazinamide and isoniazid or They had normal findings upon medical history, physical examin- are currently being evaluated in patients. Pharmacokinetic inter- ation and laboratory testing (full blood count, serum chemistry; actions within these treatment regimens have the potential to hepatitis B surface antigen; urinary pH, protein, glucose, blood compromise their efficacy or safety. Identification of a pharmaco- and screen for drugs of abuse). They had not taken prescribed medication in the 2 weeks before the study, or over-the-counter kinetic basis for altered drug activity may facilitate the rational preparations (except paracetamol) in the week before the study, development of drug combinations and doses. We investigated the or smoked or donated blood in the 2 months before the study, single-dose pharmacokinetic interactions between gatifloxacin and or consumed .6 U alcohol/day. The women were using con- a fixed dose combination (FDC) comprising rifampicin, isoniazid traception measures. Pregnant or breast-feeding women were not and pyrazinamide in healthy volunteers. enrolled.

...... *Corresponding author. Tel: þ27-21-406-6292; Fax: þ27-21-448-1989; E-mail: [email protected]

...... 1398 # The Author 2007. Published by Oxford University Press on behalf of the British Society for Antimicrobial Chemotherapy. All rights reserved. For Permissions, please e-mail: [email protected] Gatifloxacin interaction with TB drugs

Single doses of each of three treatments (gatifloxacin 400 mg in maximum observed drug concentration (Cmax), time to Cmax (Tmax), w one Gatispan 400 tablet; rifampicin 600 mg, isoniazid 300 mg and plasma half-life (t1/2) associated with the terminal slope of the semi- pyrazinamide 1600 mg as four FDC tablets of AkuriT-Zw; and logarithmic concentration–time curve and AUC extrapolated to w Gatispan together with the FDC in the same doses; both products infinity (AUC0–1). manufactured and supplied by Lupin Ltd, India) were given with Stata version 8.2 (Stata Corp., College Station, TX, USA) was 240 mL of water under fasting conditions. The treatments were used for statistical tests and to summarize results. Wilcoxon separated by 2 weeks. The sequence of treatments was randomized. matched pairs sign-rank test was used to detect significant differ- Venous samples collected in heparinized tubes before and at 0.5, 1, ences between untransformed pharmacokinetic measures. Geometric 1.5, 2, 2.5, 3.5, 5, 8, 12, 24, 36 and 48 h after dosing were placed in mean ratios [90% confidence intervals (CIs)] were calculated to crushed ice before separation of the plasma by centrifugation (750 g compare Cmax and AUC0–1 for the combined treatments with those for 10 min). Within 1 h of sampling, plasma was stored at 2808C for the gatifloxacin tablet and the FDC when given alone. until analysis. Plasma drug concentrations were quantified by tandem HPLC mass spectrometry (Applied Biosystems API 2000). A 202.1 mm Downloaded from https://academic.oup.com/jac/article/60/6/1398/824439 by guest on 27 September 2021 Hypersil Gold C18 column (Thermo, MA, USA) was used for Results rifampicin and the racemic mixture of gatifloxacin and a 202.1 mm Betasil silica column (Thermo) for isoniazid and pyra- Twelve females and 10 males with median weight 62.5 kg zinamide. The mobile phase for gatifloxacin and rifampicin com- (range: 52–76), height 167.5 cm (range: 153–184) and age 22.5 prised a gradient from 10% to 90% acetonitrile in 0.1% formic acid years (range: 20–48) completed the study. Two participants with a 5 min run time. For isoniazid and pyrazinamide, an isocratic withdrew before receiving study treatment. elution using 80% acetonitrile in 0.1% formic acid was used. The Adverse events (frequency) included nausea or vomiting (7), flow rate was 0.3 mL/min and the injection volume was 5 mL. headache (5), dyspepsia (3), loose stools (3), skin tingling (3), Moxifloxacin served as internal standard for gatifloxacin, flushing (2), rash at cannula or dressing site (3), generalized rash for rifampicin and for isoniazid and pyrazinamide. (2), drowsiness (2), myalgia (2), altered taste or smell (2), faint- Selected reaction monitoring transitions of [M-H]þ precursor ions to product ions were gatifloxacin m/z 376.2–261.3; moxifloxacin m/z ness (2), fever (1) and lower back pain (1). They were mild and 402.1–261.4; rifampicin m/z 823.5–791.4; rifapentine m/z 877.2– self-limiting; none necessitated treatment interruption. Fifteen 845.3; isoniazid m/z 138.0–121.2; pyrazinamide m/z 124.1–81.1 (44%), 13 (38%) and 6 (18%) of the adverse events related or and sulfamethoxazole m/z 254.0–92.2. Plasma protein was precipi- possibly related to the study treatments were associated, respect- tated with 3 vol of acetonitrile containing the internal standard. ively, with the FDC alone, gatifloxacin and the FDC together Samples were vortexed and centrifuged for 5 min at 750 g. and gatifloxacin alone. Supernatant (5 mL) was injected into the column. Standard curves Absorption of gatifloxacin was delayed when the products were linear in the ranges 0.1–30 mg/L for rifampicin, 0.1–15 mg/L were given together, with a median Tmax of 2.25 h [interquartile for gatifloxacin and isoniazid and 0.2–70 mg/L for pyrazinamide. range (IQR): 1.39, 3.50] versus 1.5 h (IQR: 0.88, 2.13) Quality control samples covering the ranges were included with (P ¼ 0.037). Cmax of gatifloxacin was not affected (Table 1); each run. Inter- and intra-day coefficients of variation were below however, t1/2 was prolonged [median 7.17 h (IQR: 6.24, 8.09) 10%. The lower limit of quantification was set at 0.2 mg/L for pyra- versus 6.74 h (IQR: 5.77, 8.13); P ¼ 0.023], resulting in higher zinamide and 0.1 mg/L for rifampicin, gatifloxacin and isoniazid. AUC0–1 values (Table 1 and Figure 1) with FDC Drug concentrations below the validated ranges were treated administration. as missing data. Non-compartmental analysis using WinNonlin Absorption of rifampicin was delayed in the presence of version 3.3 (Pharsight Corp., Mountain View, CA, USA) described gatifloxacin [median Tmax 2.50 h (IQR: 2.00, 3.50) versus

Table 1. Median (IQR) for Cmax and AUC0–1 after oral administration of a single dose of gatifloxacin (400 mg) given alone, after a single dose of rifampicin (600 mg), isoniazid (300 mg) and pyrazinamide (1500 mg) in FDC given alone, and after oral administration of the two products together [the GMR with 90% CI is given for comparison of the combined treatment values with those of gatifloxacin and the FDC (rifampicin, isoniazid plus pyrazinamide)]

Gatifloxacin plus FDC GMR (90% CI)

gatifloxacin alone Gatifloxacin Cmax (mg/L) 3.62 (3.13, 4.05) 3.39 (3.07, 3.59) 0.98 (0.90, 1.07) a Gatifloxacin AUC0–1 (mg.h/L) 32.89 (29.79, 35.27) 36.27 (31.92, 43.61) 1.14 (1.10, 1.18) FDC alone a Rifampicin Cmax (mg/L) 13.90 (11.70, 15.20) 10.20 (9.28, 12.60) 0.80 (0.71, 0.90) a Rifampicin AUC0–1 (mg.h/L) 95.59 (83.18, 114.44) 82.38 (68.45, 113.47) 0.88 (0.81, 0.96) Isoniazid Cmax (mg/L) 3.75 (3.09, 4.98) 3.63 (2.81, 4.40) 0.86 (0.74, 1.00) Isoniazid AUC0–1 (mg.h/L) 10.73 (8.61, 25.27) 9.67 (7.47, 25.79) 0.95 (0.89, 1.02) Pyrazinamide Cmax (mg/L) 38.10 (32.10, 40.00) 35.70 (31.00, 38.60) 0.98 (0.93, 1.05) Pyrazinamide AUC0–1 (mg.h/L) 538.88 (461.13, 580.69) 550.17 (480.96, 593.07) 1.03 (1.00, 1.07) aP , 0.05 using Wilcoxon matched pairs sign-rank test.

1399 McIlleron et al.

modestly increased [median 8.61 h (IQR: 7.48, 10.29) versus 7.24 h (IQR: 6.54, 10.31); P ¼ 0.027].

Discussion The pharmacokinetic interactions between the FDC (containing rifampicin, isoniazid and pyrazinamide) and the tablet formu- lation of gatifloxacin were not anticipated. The pharmacokinetics of gatifloxacin following a single 400 mg dose was similar to that reported previously.4,5 Gatifloxacin concentrations were modestly increased when it

was given with the FDC. As 80% to 95% of the gatifloxacin Downloaded from https://academic.oup.com/jac/article/60/6/1398/824439 by guest on 27 September 2021 dose is excreted unchanged in the urine, metabolically based 5 drug–drug interactions are unlikely. The prolonged t1/2, there- fore, suggests a renal mechanism. Probenecid has the potential to reduce renal excretion of gatifloxacin, suggesting that tubular secretion contributes to its elimination.5 This secretory mechan- ism is shared by many weak acids. Thus, the acid metabolites of pyrazinamide or isoniazid might compete with gatifloxacin for renal tubular secretion, accounting for the retarded elimination of gatifloxacin observed. That the change in gatifloxacin and pyrazinamide concentrations were correlated supports this hypothesis; the difference between the AUC0–1 for gatifloxacin with the FDC and the AUC0–1 for gatifloxacin without the FDC and the difference between the AUC0–1 for pyrazinamide with gatifloxacin and the AUC0–1 without gatifloxacin were associated (Spearman’s r ¼ 0.451; P ¼ 0.035). The AUC was increased by only 14% in this study. However, adverse effects of gatifloxacin might be more frequent with higher gatifloxacin concentrations,6 and it is a concern that accumulation of gati- floxacin may occur more readily in patients at risk (e.g. those with renal impairment). Whether modestly increased concen- trations of gatifloxacin will enhance the activity of a multidrug antitubercular treatment regimen is not known. However, an early bactericidal activity study amongst tuberculosis patients treated with 400 mg daily found a correlation between the decline in bacilli during the first 2 days and the AUC to 24h/MIC.3 The concentrations of rifampicin were in keeping with other Figure 1. Individual changes in the AUC0–1 for gatifloxacin (a) and rifampicin (b), when gatifloxacin 400 mg and the FDC product (FDC; single-dose studies in healthy volunteers and higher than those rifampicin 600 mg, isoniazid 300 mg and pyrazinamide 1600 mg), respectively, reported in patients. This can be attributed, in part, to autoinduc- were given alone and together. tion of metabolizing enzymes and P-glycoprotein with repeated doses. The modest reduction in rifampicin concentrations with gatifloxacin co-administration appears to be a result of reduced 2.00 h (IQR: 1.42, 2.50); P ¼ 0.010], and rifampicin concen- absorption as the t1/2 was not affected. The importance of the trations were reduced (Table 1). AUC0–1 was 12% lower on interaction needs to be assessed in the context of a growing con- average, but the individual reductions varied; three participants sensus that the currently used doses of rifampicin are at the had reductions of more than 35% (Figure 1b). The t1/2 of rifam- bottom of its effective range. Although the optimum range for picin was similar with or without concomitant gatifloxacin rifampicin concentrations in tuberculosis patients has not been [median 3.63 h (IQR: 3.10, 4.18) versus 3.91 h (IQR: 3.57, defined, higher doses are associated with improved early bacteri- 4.10); P ¼ 0.789]. cidal activity and better treatment results.7,8 Wide inter-patient Although isoniazid concentrations tended to be marginally variability in rifampicin concentrations and the low rifampicin lower (Table 1), and the t1/2 slightly reduced when the FDC and concentrations reported in several patient studies add to concerns gatifloxacin were co-administered [median 1.66 h (IQR: 1.31, that an average reduction of 12% in the AUC might result in cri- 9–11 4.37) versus 1.71 h (IQR: 1.35, 4.09); P ¼ 0.047], the Cmax and tically low rifampicin concentrations in some patients. AUC0–1 values were not significantly different. Gatifloxacin and rifampicin have important sterilizing Pyrazinamide concentrations were less variable than those of activities against Mycobacterium tuberculosis and the potential the other drugs. Cmax and AUC0–1 were not affected when the to shorten the time to eradication of the pathogen. It is not two products were given together (Table 1), although t1/2 was known whether increased concentrations of gatifloxacin could

1400 Gatifloxacin interaction with TB drugs compensate for reduced rifampicin exposure in a combined treat- References ment regimen. Steady-state conditions and the effects of enzyme induction 1. Hu Y, Coates ARM, Mitchison DA. Sterilizing activities of fluoro- and inhibition with multiple daily doses were not studied. quinolones against rifampin-tolerant populations of Mycobacterium Volunteers given consecutive doses of rifampicin for 4 days had tuberculosis. Antimicrob Agents Chemother 2003; 47: 653–7. reduced concentrations of moxifloxacin, probably largely a 2. Nueremberger EL, Yoshimatsu T, Tyagi S et al. Moxifloxacin- result of increased expression of phase II metabolic enzymes.12 containing regimen greatly reduces time to culture conversion in murine tuberculosis. Am J Respir Crit Care Med 2004; 169: 421–6. As gatifloxacin is almost completely eliminated in the urine unchanged, repeated doses of rifampicin are unlikely to decrease 3. Johnson JL, Hadad DJ, Boom WH et al. Early and extended early bactericidal activity of levofloxacin, gatifloxacin and moxifloxacin its concentrations by those mechanisms. Other important limit- in pulmonary tuberculosis. Int J Tuberc Lung Dis 2006; 10: 605–12. ations in the application of the findings of this study to tuber- 4. Nakashima M, Uematsu T, Kosuge K et al. Single- and multiple- culosis patients include the study population, which does not dose pharmacokinetics of AM-1155, a new 6-fluoro-8-methoxy quino- reflect heterogeneous patient populations or the effects of lone, in humans. Antimicrob Agents Chemother 1995; 39: 2635–40. disease and nutritional status, and the design, which does not 5. Grasela DM. Clinical pharmacology of gatifloxacin, a new Downloaded from https://academic.oup.com/jac/article/60/6/1398/824439 by guest on 27 September 2021 allow discrimination between the effects of rifampicin, isoniazid fluoroquinolone. Clin Infect Dis 2000; 31 Suppl: 51–8. and pyrazinamide on gatifloxacin concentrations. Moreover, the 6. Ishiwata Y, Sanada Y, Yasuhara M. Effects of gatifloxacin on pharmacodynamic consequences of the altered drug concen- serum glucose concentration in normal and diabetic rats. Biol Pharm trations could not be evaluated. It is therefore necessary to Bull 2006; 29: 527–31. evaluate the pharmacokinetic interactions and the relationship of 7. Diacon AH, Patientia RF, Venter A et al. Early bactericidal the drug concentrations to their effects in patient studies. activity of high-dose rifampin in patients with pulmonary tuberculosis evidenced by positive sputum smears. Antimicrob Agents Chemother 2007; 51: 2994–6. Acknowledgements 8. Long MW, Snider DE, Farer LS. U.S. Public Health Service cooperative trial of three rifampin–isoniazid regimens in treatment of We thank clinical and laboratory staff in the Division of Clinical tuberculosis. Am Rev Respir Dis 1979; 119: 879–94. Pharmacology, University of Cape Town. 9. Gurumurthy P, Ramachandran G, Hemanth Kumar AJ et al. Decreased bioavailability of rifampin and other antituberculosis drugs in patients with advanced human immunodeficiency virus disease. Antimicrob Agents Chemother 2004; 48: 4473–5. Funding 10. McIlleron H, Wash P, Burger A et al. Determinants of rifampicin, isoniazid, pyrazinamide and ethambutol pharmacokinetics in a cohort The project was sponsored by UNDP/World Bank/WHO Special of tuberculosis patients. Antimicrob Agents Chemother 2006; 50: Programme for Research and Training in Tropical Diseases 1170–7. (TDR), World Health Organization, Geneva. 11. Tappero JW, Bradford WZ, Agerton TB et al. Serum concen- trations of drugs in patients with pulmonary tuber- culosis in Botswana. Clin Infect Dis 2005; 41: 461–9. Transparency declarations 12. Weiner M, Burman W, Luo CC et al. Effects of rifampin and multidrug resistance gene polymorphism on concentrations of moxi- None to declare. floxacin. Antimicrob Agents Chemother 2007; 51: 2861–6.

1401