Pharmacokinetics of Levofloxacin in Children Treated for Exposure to Drug-Resistant Tuberculosis

Levofloxacin is used to treat and prevent drug-resistant tuberculosis in children. We assessed levofloxacin serum drug concentrations in 24 children aged 2 to 10 years who received levofloxacin-based tuberculosis preventive therapy in Karachi, Pakistan.

ABSTRACT

Levofloxacin is used to treat and prevent drug-resistant tuberculosis in children. We assessed levofloxacin serum drug concentrations in 24 children aged 2 to 10 years who received levofloxacin-based tuberculosis preventive therapy in Karachi, Pakistan. Only 9 children (37.5%) achieved adequate drug exposure. Target serum drug concentration was met in 4 (26.7%) of 15 children dosed consistently with World Health Organization recommendations and 4 (80.0%) of 5 who received higher-than-recommended doses. Levofloxacin dosing recommendations may require reevaluation.

INTRODUCTION

Levofloxacin is used to treat children with drug-resistant tuberculosis (DR-TB) that cannot be treated with first-line drugs (1). Some programs have used levofloxacin-based preventive therapy to prevent development of disease in these children because standard preventive therapy regimens based on isoniazid or rifamycins are unlikely to be effective (2–4). When using levofloxacin for prevention, dosing typically follows recommendations for treatment of disease. Currently, there are two different World Health Organization (WHO) recommendations for levofloxacin dosing in children treated for DR-TB disease. The 2014 DR-TB treatment guidelines recommend a weight-based dose of 15 to 20 mg/kg for children aged ≤5 years and 10 to 15 mg/kg for older children (5). The 2014 childhood TB guidelines recommend 7.5 to 10 mg/kg for children with DR-TB (6). However, the sufficiency of these dosing recommendations is unclear.

Published pharmacokinetic studies of children treated for DR-TB disease or exposure are limited to cohorts from the U.S.-affiliated Pacific Islands and South Africa. A population pharmacokinetic analysis from the Pacific Islands suggested that a daily dose of 15 to 20 mg/kg would be sufficient to achieve a target maximum serum concentration of 8 to 12 μg/ml (7). However, studies from South Africa showed that a dose of 15 mg/kg was insufficient to reach target drug exposure and suggested that doses closer to 30 to 40 mg/kg might be required for all but the youngest children when commonly available adult oral tablets are used (8, 9). Because these conflicting results come from two geographically distinct settings, it is not clear which findings, if either, are generalizable to other countries.

To help resolve the question of whether currently recommended levofloxacin doses are sufficient for children in Pakistan, we conducted a pharmacokinetic assessment of children receiving levofloxacin-based preventive therapy under programmatic conditions for DR-TB exposure in Karachi, Pakistan (10). Independent of the pharmacokinetic study, clinicians were treating children with daily levofloxacin doses of 15 to 20 mg/kg for children aged ≤5 years and 7.5 to 10 mg/kg for children aged >5 years. Children received either ethambutol (15 to 25 mg/kg) or ethionamide (15 to 20 mg/kg) as a companion drug, depending on which was available (6). Levofloxacin was available in 250- and 500-mg tablets (Macleods Pharmaceutical Ltd., India) and obtained via the Global Drug Facility. Oral tablets were used for older children and crushed tablets for younger children; crushing tablets has been shown to not affect bioavailability (8).

During February to December 2017, a convenience sample of children aged 2 to 10 years who had been on treatment for >1 month were enrolled in the pharmacokinetic study with written informed consent from parents/guardians. Venous samples (1 ml) were drawn immediately after medication administration and after 1, 2, and 6 h; in previously published studies of TB patients, maximum drug concentration (C_max) typically occurred 1 to 2 h postdose (7, 11, 12). Levofloxacin serum concentrations were measured at the University of Florida (Gainesville, FL), using a validated high-pressure liquid chromatography assay with tandem mass spectrometry (LC-MS/MS) detection (11). A noncompartmental pharmacokinetic analysis was performed to estimate pharmacokinetic parameters using Phoenix v7.0 (Certara LP, Princeton, NJ). Given the restricted sampling scheme, we used the simplest approach (linear up/linear down) to calculate area under the curve from 0 to 6 hours (AUC_0–6). To determine the elimination rate constant (K_e) and half-life [ln(2)/K_e], the software performed linear regression on the final 2 or 3 data points, provided that the final point(s) was lower than the prior. Half-life could not be calculated for patients whose apparent C_max occurred at 6 h. Associations between pharmacokinetic parameters and demographic covariates (age, sex, weight, and body mass index) were evaluated using bivariate or one-way analysis.

We stratified the analysis by the actual dose received on the day of sampling. Doses of 15 to 20 mg/kg for children aged 2 to 5 years or 7.5 to 15 mg/kg for children aged >5 years were classified as consistent with WHO recommendations. Children were classified as having achieved adequate drug exposure if they achieved a C_max of ≥8 μg/ml, which is the lower limit of the normal C_max achieved by adults receiving a standard 750- to 1,000-mg dose (12). Statistical analysis was performed using SAS version 9.4 and JMP v13.2 software (SAS Institute, Cary, NC). The study was approved by the Institutional Review Board of Interactive Research and Development.

In total, 24 children were enrolled (Table 1). All children were contacts of HIV-negative index patients; children themselves were not tested for HIV given the low prevalence in Pakistan (0.1% among adults) and its concentration among key adult populations (13). Among children aged 2 to 5 years, 9 (69.2%) received a levofloxacin dose of ≥15 mg/kg on the day of the sampling. The other 4 children (30.8%) received a lower dose because they had experienced nausea or vomiting when dosed at the upper end of the range; given the pill sizes available, reducing the dose meant that these children received less than the recommended weight-based dose. Furthermore, 5 children (38.5%) in this age group received a dose of >20 mg/kg because available pill sizes precluded a dose within the target range. Among children aged 6 to 10 years, 4 (36.4%) received a dose between 7.5 and 10 mg/kg and 7 (63.6%) received a dose of 10 to 15 mg/kg.

TABLE 1.

Patient characteristics, by age group

Characteristic	Data according to age group:
	All participants (n = 24)	2–5 yrs (n = 13)	6–10 yrs (n = 11)
Age (yr) (median [IQR])	5 (4 to 9)	4 (2 to 5)	9 (7 to 9)
Male sex (n [%])	17 (70.8)	8 (61.5)	9 (81.8)
Weight (kg) (median [IQR])	16.3 (13.3 to 20.8)	13.5 (11.5 to 15.0)	21.0 (19.0 to 26.0)
BMI (kg/m2) (median [IQR])	14.1 (13.1 to 15.1)	13.3 (13.0 to 15.3)	14.3 (13.9 to 15.0)
BMI-for-age Z-scorea (median [IQR])	−1.2 (−2.5 to −0.5)	−1.5 (−3.0 to −0.4)	−1.2 (−2.0 to −0.4)
Height (cm) (median [IQR])	107 (95 to 118)	100 (90 to 106)	117 (115 to 124)
Treatment
Ethambutol for adjunct therapy (n [%])	17 (70.8)	9 (69.2)	8 (72.7)
Ethambutol dose (mg/day) (median [IQR])	300 (200 to 400)	200 (200 to 300)	400 (350 to 400)
Ethionamide for adjunct therapy (n [%])	7 (29.2)	4 (30.8)	3 (27.3)
Ethionamide dose (mg/day) (median [IQR])	250 (250 to 250)	250 (250 to 250)	250 (125 to 250)
Levofloxacin dose (mg/day)b (median [IQR])	250 (250 to 250)	250 (250 to 250)	250 (250 to 250)
Levofloxacin dose (mg/kg daily)b (median [IQR])	12.8 (9.5 to 18.5)	17.9 (13.9 to 21.7)	11.6 (9.4 to 12.5)
Treatment successfully completed (n [%])	22 (91.7)	12 (92.3)	10 (90.9)

^a8 children had a body mass index (BMI)-for-age Z-score of less than −2.0.

^bDoses reflect what was recorded on the day of the pharmacokinetic sampling.

For 19 children (79.2%), C_max occurred at 1 or 2 h. Only 9 (37.5%) of the 24 children achieved adequate drug exposure (Table 2). Adequate drug exposure was achieved in 4 (26.7%) of 15 children dosed consistently with WHO recommendations and 4 (80.0%) of 5 who received higher-than-recommended doses. Both C_max (P = 0.0493) and AUC_0–6 (P = 0.0405) were lower with increasing body weight, and weight explained roughly 13% to 14% of the variability in each parameter (adjusted R² = 0.13 and 0.14, respectively). Other demographic covariates were not predictive of C_max or AUC_0–6. A rough estimate of elimination half-life was a median of 3.5 h (interquartile range [IQR], 3.1 to 4.3 h), but this was limited by the short sampling window.

TABLE 2.

Levofloxacin pharmacokinetic parameters, by age group and levofloxacin dose received

Parameter	All children ageda:		Age group, by dosea
			2–5 yrs		6–10 yrs
	2–5 yrs (n = 13)	6–10 yrs (n = 11)	15–20 mg/kg (n = 4)	>20 mg/kg (n = 5)	7.5–10 mg/kg (n = 4)	10–15 mg/kg (n = 7)
Levofloxacin dose (mg/kg)	17.9 (7.8–35.7)	11.6 (7.6–13.2)	17.9 (16.3–19.2)	23.8 (20.8–35.7)	9.3 (7.6–9.6)	12.2 (11.4–13.2)
AUC0-6 (μg · h/ml)	33.5 (16.7–68.9)	16.8 (7.7–47.5)	35.3 (21.6–68.9)	36.6 (16.7–41.5)	13.4 (9.5–24.9)	21.8 (7.7–47.5)
Tmax (h)	2.0 (1.0–6.0)	2.0 (1.0–6.0)	1.5 (1.0–2.0)	1.0 (1.0–6.0)	4.0 (2.0–6.0)	2.0 (1.0–6.0)
Cmax (μg/ml)	8.2 (4.6–21.0)	4.8 (2.3–14.3)	7.9 (6.0–21.0)	8.5 (5.1–12.5)	3.2 (2.4–5.4)	5.5 (2.3–14.3)
Children reaching target Cmax of 8 μg/ml (n [%])	7 (53.8)	2 (18.2)	2 (50.0)	4 (80.0)	0 (0.0)	2 (28.6)

^aAll values shown are medians (ranges) unless otherwise noted. Dose-stratified results not shown for four children aged 2–5 years who received a dose of <15 mg/kg due to side effects, because the dose received was lower than the WHO recommendation (5).

In our cohort of children treated for DR-TB exposure, most of the children who received WHO-recommended levofloxacin doses did not achieve adequate drug exposure. Our findings are thus consistent with reports from South Africa (8, 9) and different from reports from the U.S.-affiliated Pacific Islands (7). Given that prior reports are only available from these two settings, our findings from a geographically distinct setting provide important evidence for the need to reconsider levofloxacin dosing recommendations for treating DR-TB disease or exposure in children. The poor pharmacokinetic target attainment observed among older children supports the recommendation by the Sentinel Project on Pediatric Drug-Resistant Tuberculosis that a dose of 15 to 20 mg/kg be applied to children of all ages (14). Moreover, although our sample was small, the higher drug exposures observed among children receiving doses of >20 mg/kg support the suggestion that doses closer to 30 mg/kg may be necessary when adult tablets are used (8).

Our study was limited by a small sample size, an absence of children aged <2 years, and a lack of time points beyond 6 h. Furthermore, a number of children in our study received lower-than-recommended doses because available pill sizes left little room for adjustment for children who experienced adverse events. Nevertheless, our results add to the evidence that current levofloxacin dosing recommendations for children with DR-TB may be insufficient where commonly available oral tablets are used. While a recently developed pediatric dispersible formulation has been shown to achieve higher drug exposure in young children (15), many programs currently have access only to adult oral tablets. If higher doses are required, it will be necessary to determine how to manage adverse events at higher doses. Thus, there is a need to reevaluate the current levofloxacin dosing recommendations for children being treated for DR-TB disease and exposure to ensure that they are treated appropriately.