Abstract
Background
Tuberculosis (TB) is the leading single bacterial cause of death worldwide. In 2023, approximately 400 000 people developed multidrug- and rifampicin-resistant TB (MDR/RR-TB), which complicates treatment. TB preventive treatment (TPT) is a critical strategy to prevent the progression from TB infection to TB disease among those at risk. In February 2024, based on data from 2 randomized controlled trials, levofloxacin was strongly recommended by the World Health Organization (WHO) as a TPT option in people of all ages exposed to MDR/RR-TB. There are uncertainties about the optimal dosing of levofloxacin in children and adolescents when using dispersible and solid formulations. We used pharmacokinetic modeling and simulations to determine the best dosing strategy in people aged up to 19 years for both formulations of levofloxacin.
Methods
A previously developed population pharmacokinetic model of levofloxacin in children (0.2–16.8 years) was used and applied to new WHO harmonized weight bands. Simulations were conducted using demographic data from countries with the highest incidence of RR- or MDR-TB. Two currently available levofloxacin formulations (100 mg pediatric, dispersible tablets and 250 mg solid tablets) were considered.
Results
A dosing regimen by weight band was developed for levofloxacin when used as TPT in people aged 0–19 years exposed to MDR/RR-TB. Doses correspond to 8–33 mg/kg for the 100 mg dispersible tablets and 10–42 mg/kg for 250 mg solid tablets. These doses achieve adequate adult target exposure levels.
Conclusions
Pragmatic, weight-band dosing strategies help simplify the administration of MDR/RR-TB TPT and have been included in WHO guidance.
Keywords: prevention, tuberculosis, levofloxacin, children
Using actual population demographic data and pharmacokinetic modeling and simulations, we propose an optimal dosing strategy of levofloxacin for tuberculosis prevention treatment in children. Our dosing strategy aligns with harmonized WHO weight bands and facilitates the use of 2 different formulations.
Graphical Abstract
Graphical Abstract.
Workflow summary for optimizing dosing strategies utilizing pharmacokinetic modeling and simulation. Step 1: A robust population pharmacokinetic (popPK) model is developed using data from an individual participant data meta-analysis, a pooled study, or a large database encompassing a diverse population. Step 2: Then, a reasonable target exposure is selected to ensure efficacy. Step 3: Demographic data acquired from countries with the highest incidence of rifampicin-resistant or multidrug-resistant tuberculosis are utilized. Step 4: The dosing simulations for an equitable dosing strategy utilize a rigorously constructed popPK model, which is based on a comprehensive range of age and weight groups, ensuring that dose optimizations account for diverse demographic populations.
Tuberculosis (TB) is the leading single bacterial cause of death worldwide, making it a significant global health concern [1]. The World Health Organization (WHO) estimates that 400 000 people developed multidrug-resistant or rifampicin-resistant TB (MDR/RR-TB) globally in 2023. MDR/RR-TB requires treatment with regimens that are more complex and expensive than drug-susceptible TB, making it very important to prevent it. WHO recommends the early identification and investigation of all household contacts of people with TB and MDR/RR-TB, especially those aged <5 years, who have a higher risk to progress from TB infection to disease [2, 3]. TB preventive treatment (TPT) can prevent progression to TB disease in a person infected with TB, and is a critical component of Pillar 1 of the WHO End TB strategy [4].
Levofloxacin is a fluoroquinolone antibiotic, commonly used in the treatment of active TB. The pharmacokinetics of levofloxacin are characterized by high bioavailability and extensive tissue penetration, including the lungs. Levofloxacin is minimally metabolized and is primarily eliminated unchanged in the urine [5].
The efficacy and safety of 6 months of levofloxacin dosed at 15–20 mg/kg once daily as TPT was recently evaluated in the V-QUIN and TB-CHAMP trials. A pooled analysis of data from both trials showed that levofloxacin reduced the risk of developing TB disease by 1 year by 60% (hazard ratio, 0.40 [95% confidence interval, .17–.90]) [6–8]. The V-QUIN trial, conducted in Vietnam, enrolled 2041 contacts of people with MDR/RR-TB, mostly adults and adolescents [7]. TB-CHAMP was a phase 3, double-blind, placebo-controlled trial that compared the efficacy and safety of levofloxacin versus matched placebo in children (<10 years) and adolescents (10–18 years) in South Africa (ISRCTN92634082) [6]. TB-CHAMP included 922 children and adolescents aged <18 years, of which 90% were <5 years of age. The dose of levofloxacin for the trial was selected based on previous pediatric pharmacokinetic (PK) studies in children (0.2–6 years of age) with MDR/RR-TB disease and implemented in the trial using the 250 mg solid levofloxacin tablet (Macleods Pharmaceuticals, India) [9, 10]. A PK study done ahead of the TB-CHAMP trial determined that exposures when using the dispersible formulation would be much higher for the same milligram dose with this formulation [9, 11]. Owing to concerns about safety, the dispersible formulation of levofloxacin was not used in the trial. The solid tablets were found to be extremely safe and well tolerated in TB-CHAMP.
In February 2024, after critical review of these trial results by a guideline development group, WHO released a strong recommendation for the use of 6 months of levofloxacin as TPT in contacts exposed to MDR/RR-TB [4, 12].
When translating findings from clinical trials into clinical guidelines and policy-level recommendation, caution should be exercised to ensure that the dosing strategy achieves appropriate exposure in all weight and age groups. Simulations based on robust population PK models, all available formulations, and large, representative, demographic databases can help optimize drug exposure in different subgroups and inform pragmatic dosing strategies that can be implemented in relevant settings [13, 14]. This is particularly important in young children, because of the influence of age (maturation) as well as anthropometric factors, highlighting the usefulness of advanced modeling techniques that can accurately predict drug behavior across this diverse population. In addition, WHO has recently moved toward the use of harmonized weight bands to facilitate optimal treatment across different disease areas [15]. The dispersible pediatric formulation of levofloxacin (100 mg scored dispersible tablet) has improved bioavailability compared with the crushed, nondispersible adult formulations (250 mg and 500 mg tablets) [16], and dosing requirements can be different for each formulation. By aligning dosing strategies with harmonized weight bands, we can optimize the use of each formulation, thereby enhancing drug efficacy and safety.
This manuscript describes the work that informed the dosing strategy for levofloxacin TPT in the recently published WHO operational handbook on TB prevention [3], in support of the practical implementation of the strong recommendation for use of levofloxacin as TPT. An established PK modeling and simulation framework was used to determine the most effective dosing strategy for levofloxacin TPT in children and adolescents aged 0–19 years.
METHODS
Dataset Used for Simulation
Children <5 Years of Age
Demographic characteristics of nationally representative children aged <5 years were obtained from the Demographic and Health Surveys (DHS) Program database [17] . The DHS Program is a standardized international survey that collects data on health and nutrition from >90 countries with a large sample size including child-specific sex, age, height, weight, and nutritional z-scores (height-for-age z-score, weight-for-age z-score, weight-for-height z-score, and body mass index-for-age z-score) for applicable age ranges. Children with missing values for those parameters and implausible values (z-scores < −6 or >6) were excluded from analysis. DHS data were not available for China and the Russian Federation, so their data for children <5 years of age were obtained from the China Health and Nutrition Survey and the Russia Longitudinal Monitoring Survey [18, 19]. Demographic data for the included countries are summarized in Table 1.
Table 1.
Overview of the Demographic Characteristics of the Dataset Used for Simulations, Including Children <5 Years Old
| Country | India (n = 224 819) | Philippines (n = 16 120) | Russia (n = 9793) | Indonesia (n = 4393) | China (n = 1372) |
|---|---|---|---|---|---|
| Female sex | 108 565 (48) | 7807 (48) | 4851 (50) | 2149 (49) | 610 (44) |
| Age, y, median (range) | 2.5 (0–4.9) | 2.7 (0–4.9) | 2.4 (0.1–4.9) | 2.4 (0–4.9) | 2.7 (0–4.9) |
| Weight, kg, median (range) | |||||
| All age groups | 10.6 (1.8–33.7) | 11.4 (2.2–32.6) | 13.0 (2.4–32.0) | 11.1 (2.3–27.2) | 13.3 (3.0–30.1) |
| <1 y | 6.4 (1.8–15.1) | 6.9 (2.2–12.6) | 7.4 (2.4–15.0) | 6.8 (2.3–13.4) | 8.0 (3.0–13.0) |
| 1 to <3 y | 9.8 (4.2–23.1) | 10.2 (5.4–22.5) | 12.0 (5.0–25.0) | 10.2 (4.8–19.4) | 12.0 (6.3–18.2) |
| 3 to <5 y | 13.0 (6.3–33.7) | 13.6 (7.7–32.6) | 16.0 (8.5–32.0) | 13.8 (8.0–27.2) | 15.8 (10.0–30.1) |
| Underweighta | 76 228 (34) | 3556 (22) | 353 (4) | 850 (19) | 60 (4) |
| Wastedb | 44 334 (20) | 1077 (7) | 864 (9) | 420 (10) | 58 (4) |
| Stuntedc | 86 161 (38) | 5550 (34) | 962 (10) | 1575 (36) | 173 (13) |
| Malnourishedd | 122 062 (54) | 6531 (41) | 1837 (19) | 1949 (10) | 231 (17) |
| Country | Pakistan (n = 4083) | Myanmar (n = 4185) | Nigeria (n = 11 305) | South Africa (n = 1073) | Vietnam (n = 3549) |
|---|---|---|---|---|---|
| Female sex | 2020 (49) | 2025 (48) | 5578 (49) | 527 (49) | 1742 (49) |
| Age, y, median (range) | 2.4 (0–4.9) | 2.5 (0–4.9) | 2.3 (0–4.9) | 2.5 (0–4.9) | 2.5 (0–4.9) |
| Weight, kg, median (range) | |||||
| All age groups | 11.1 (2.3–23.9) | 11.0 (2.2–23.7) | 10.9 (2.2–25.5) | 12.5(2.9–27.1) | 11.7 (2.5–28.9) |
| <1 y | 6.6 (2.3–12.2) | 7.0 (2.2–12.2) | 6.7 (2.2–13.5) | 7.2 (2.9–13.4) | 7.2 (2.5–13.1) |
| 1 to <3 y | 10.2 (4.7–19.5) | 10.1 (5.5–16.0) | 10.1 (5.2–18.6) | 11.5 (7.3–19.5) | 10.8 (6.0–23.0) |
| 3 to <5 y | 13.7 (7.7–23.9) | 13.4 (7.6–23.7) | 14.0 (6.6–25.5) | 15.3 (10.2–27.1) | 14.2 (8.9–28.9) |
| Underweighta | 809 (20) | 741 (18) | 2189 (19) | 52 (5) | 380 (11) |
| Wastedb | 311 (8) | 257 (6) | 743 (7) | 29 (3) | 141 (4) |
| Stuntedc | 1430 (35) | 1285 (31) | 3706 (33) | 235 (22) | 759 (21) |
| Malnourishedd | 1673 (41) | 1555 (37) | 4253 (38) | 263 (25) | 903 (25) |
Data are presented as No. (%) unless otherwise indicated.
aDefined as weight-for-age z-score < −2.0.
bDefined as weight-for-height z-score < −2.0.
cDefined as height-for-age z-score < −2.0.
dDefined as weight-for-age, weight-for-height, height-for-age, or body mass index-for-age z-score < −2.0.
The top 10 countries with the highest total incidence of MDR/RR-TB based on the 2023 WHO global tuberculosis report [1] were selected. According to this report, the largest share of incident MDR/RR-TB cases in 2022 were in India (27% of global cases), the Philippines (7.5% of global cases), and the Russian Federation (7.5% of global cases). These are followed by Indonesia, China, Pakistan, Myanmar, Nigeria, Ukraine, and South Africa. Given that demographic data from Ukraine were not available, it was replaced by Vietnam, which ranks as 11th in global incidence of MDR/RR-TB.
Children 5–19 Years of Age
The dataset used for dosing simulations on children and adolescents aged ≥5 years was built by combining demographics from our internal data repository of previous studies on children and adolescents with MDR/RR-TB. These data were based on trials and observational PK research conducted in countries ranked among the 10 with the highest global MDR/RR-TB burden. Their demographics are available in Supplementary Table 1.
Population Pharmacokinetic Model
Published population PK models of levofloxacin for children via PubMed search were reviewed using the following search terms: (“child*” OR “neonat*” OR “pediatr*” [All Fields]) AND “levofloxacin” [All Fields] AND “pharmacokinetic*” [All Fields] AND “model*” [All Fields]. Six population PK models of levofloxacin in children were identified from this search [9–11, 16, 20, 21].
The model by White et al [16] was used, given the fact that it is an individual participant data meta-analysis, encompassing PK data from 242 children and adolescents in 5 different pediatric PK studies of levofloxacin for TB. While 3 of these studies had been used to build population PK models elsewhere [9–11], 2 of these studies had not previously been analyzed by compartmental analysis and contributed data from children outside Africa [22] and adolescents (MDR-PK2 Study). The population used in building this PK model included those who received levofloxacin for treatment of MDR-TB disease as well as TPT, and its age and weight range spanned 0.2 to 16.3 years and 4.0 to 48.5 kg, respectively. The effect of the indication (levofloxacin used for MDR-TB or TPT) on the primary PK parameters was explored but was not identified as a significant covariate. The model accounted for the effect of weight and young age by allometric scaling of clearance and volume of distribution, and by including a maturation function of clearance with age, with an estimated Hill coefficient that determines the shape of the curve of 3.46, and a postmenstrual age (gestational + postnatal age) at which 50% of maturation is reached at 8.94 months. The levofloxacin formulations included in the analysis were the 100 mg dispersible tablets and the 250 mg and 500 mg conventional tablets. Relative bioavailability of the conventional tablets was estimated to be 28.7% lower than that of 100 mg dispersible tablets [16].
Additionally, for children aged ≥5 years, PK models identified in the literature built with adult data [23, 24] were used and explored for dosing simulations. The results obtained after the simulation with the adult models were then compared with those obtained from the model developed by White et al [16]. This comparison was motivated by the fact that a relatively small number of patients in the higher weight bands was included in the pediatric model developed, and because of uncertainty of whether adolescent exposures are better predicted with pediatric or adult models. Parameter estimates of these models are summarized in Supplementary Table 2.
Exposure Target
Given that the TB-CHAMP trial demonstrated that levofloxacin is effective and safe in preventing TB in children using a 15–20 mg/kg daily dose [25], the median calculated exposure from TB-CHAMP was selected as a reasonable target to be achieved [15]. The model developed by White et al was used to simulate the doses of levofloxacin with the 250 mg conventional tablet that was used in TB-CHAMP [16, 25] (Supplementary Table 3). This was then applied to the DHS data from South Africa, thus extrapolating the trial findings to a child population structure that is more representative of the country's demography. Drug exposure, measured as the area under the concentration–time curve at steady state over a dosing interval (AUC0–24), was then evaluated.
Dosing Simulations
Monte Carlo simulations were performed using the population PK model mentioned above. The number of simulations was determined to include approximately 50 000 children from each country, except for India, where all children included in the DHS dataset were used. Doses were assigned to new WHO harmonized weight bands [15]. Distribution of AUC0–24 was visualized according to weight bands, age, nutritional status (based on weight-for-age z-score), and country. The dosing strategy was targeted to reach the simulated median AUC0–24 calculated using the dosing strategy used in the TB-CHAMP trial.
Simulation and dose optimization for the 100 mg dispersible tablet and 250 and 500 mg conventional solid tablets were performed for children <5 years of age. Only 250 and 500 mg conventional solid tablets were considered for those ≥5 years of age, since it was considered that they would be able to swallow whole tablets. The proposed dosing was intended to be practical to implement. Therefore, dose optimization was performed in increments of 0.5 tablet. For 250 or 500 mg tablets, increments smaller than 0.5 tablet were explored for those who need doses <125 mg to achieve optimal exposure. For younger children, a mixed age- and weight-based approach to dosing was considered to account for the rapid change in maturation that significantly affects levofloxacin's PK in these younger population.
Sensitivity Analysis
To evaluate the robustness of the proposed dosing strategy for younger children, dosing simulations using the 5th and 95th percentiles of the estimated PM50 (age at which clearance reaches 50% of the adult value) and Hill coefficient values by White et al [16] were done. The results obtained with this set of parameters were then compared to those obtained using median estimates.
Software Used
The software NONMEM, version 7.5.1 (Globomax, Hanover, Maryland, USA) was used for the simulations and was installed on a Mac platform using GCC 2.96 under Red Hat Linux 9. The statistical package R version 3.4.3 (http://www.r-project.org/) was used for all exploratory analyses, figure creation, and postprocessing of the NONMEM output. The model codes used are provided in the Supplementary Material.
RESULTS
Exposure Target
The median AUC0–24 of dosing strategy in the TB-CHAMP trial calculated by the White et al [16] model was 60.7 mgh/L. A significant disparity in exposure was observed in the lowest weight band (3–4.9 kg), resulting in particularly low exposure in those >3 months of age in this weight band.
Dosing Simulations
Children <5 Years Old
Simulations with the proposed dosing strategy were done using the DHS dataset, overall, and by country. The final dosing strategy proposed is detailed in Table 2. Following this dosing approach, the median AUC0–24 observed in the TB-CHAMP trial was reached across all weight bands for both the 250 mg conventional tablets and the 100 mg dispersible tablets (Figure 1). The proportion of children who attained the exposure target according to nutritional status in each country is shown in in Figure 2. Additionally, the target attained for the overall population per nutritional status and weight band is shown in Supplementary Figures 1 and 2, respectively.
Table 2.
Proposed Levofloxacin Doses by Weight Band
| Weight Band | 250 mg NDT | 500 mg NDT | 100 mg DT | ||||||
|---|---|---|---|---|---|---|---|---|---|
| mg | mg/kg | No. of Tablets | mg | mg/kg | No. of Tablets | mg | mg/kg | No. of Tablets | |
| 3–5.9 kg, <3 m | 62.5 | 11–21 | 1/4 | … | … | … | 50 | 8–17 | 1/2 |
| 3–5.9 kg, ≥3 m | 125 | 21–42 | 1/2 | … | … | … | 100 | 17–33 | 1 |
| 6–9.9 kg, <6 m | 125 | 13–21 | 1/2 | … | … | … | 100 | 10–17 | 1 |
| 6–9.9 kg, ≥6 m | 250 | 25–41 | 1 | … | … | … | 150 | 15–25 | 1 + 1/2 |
| 10–14.9 kg | 250 | 17–25 | 1 | … | … | … | 200 | 13–20 | 2 |
| 15–19.9 kg | 375 | 19–25 | 1 + 1/2 | … | … | … | 250 | 13–17 | 2 + 1/2 |
| 20–24.9 kg | 375 | 15–19 | 1 + 1/2 | … | … | … | 300 | 12–15 | 3 |
| 25–29.9 kg | 500 | 17–20 | 2 | 500 | 17–20 | 1 | 350 | 12–14 | 3 + 1/2 |
| 30–34.9 kg | 500 | 14–17 | 2 | 500 | 14–17 | 1 | … | … | … |
| 35–49.9 kg | 500 | 10–14 | 2 | 500 | 10–14 | 1 | … | … | … |
| ≥50 kg | 750 | <15 | 3 | 750 | <15 | 1 + 1/2 | … | … | … |
Bold text represents the preferred formulation to be used in the specific weight bands.
Abbreviations: DT, dispersible tablet; NDT, nondispersible tablet.
Figure 1.
Exposures measured as area under the concentration–time curve at steady state (AUC0–24) in children <5 years old, simulated using the proposed dosing schedule with the WHO harmonized weight bands using the 100 mg dispersible tablets and the 250 mg nondispersible tablets.
Figure 2.
Target attainment by nutritional status in children <5 years old split by country, considered the median area under the concentration–time curve at steady state (AUC0–24) as simulated in TB-CHAMP (60.7 mgh/L) as the target. Blue line represents the dispersible tablets, and red line represents the nondispersible tablets. Gray bars represent the distribution of the population per country in the dataset used for simulation. Green shaded area represents a 10% upper and lower margin of the median exposure, showing 40%–60% target attainment.
Sensitivity Analysis
The simulations for children aged <5 years were repeated using the 5th and 95th percentile values for PM50 and Hill coefficient from the White et al model [16]. The range of AUCs obtained using the 5th percentile values of the parameters was similar for all subgroups, as seen in Figure 3. This was also true for AUCs obtained using the 95th percentile values of the parameters, except for those weighing 3.0–5.9 kg and aged ≤3 months. For this specific subset of children, the median AUC0–24 value was higher than that in other scenarios, approaching the defined target for TB treatment in adults (101 mgh/L) [26]. Therefore, differences in exposures due to potential differences in maturation functions were considered acceptable. In addition, a comparison of the maturation function with that of White et al [16] is shown with respect to the maturation function of other published manuscripts [23, 24] in Supplementary Figure 3.
Figure 3.
Sensitivity analysis on the exposures measured as area under the concentration–time curve at steady state (AUC0–24) in children <5 years old simulated using the proposed dosing schedule and the 5th, 50th, and 95th percentiles of the maturation function of the model developed by White et al [16] with the WHO harmonized weight bands using nondispersible tablets (upper panel) and dispersible tablet (lower panel).
Children 5–19 Years Old
Only the conventional tablets (250 and 500 mg) were considered when simulating this population. The increments considered in this cohort was done in half of a tablet when needed, preferring increases in whole tablets. The models used were the White et al [16] model and the 2 adult models identified in the literature [23, 24]. The decision to use the adult model was based on the question of whether adolescents (10–19 years) are either accurately predicted by pediatric models or better predicted by adult models. Simulations with the pediatric model [16] rendered similar exposures in the weight bands of 25–35 kg when compared to exposures obtained with the adult models [23, 24], but lower exposures in >35 kg. Exposure measured as AUC0–24 seemed to be adequate when using the adult models [23, 24], and therefore doses were not further increased in these weight bands. The proposed dosing is indicated in Table 2. Following this dosing approach, adequate exposures were reached in all weight bands (Figure 4). In addition, these doses align well with the doses tested in the V-QUIN trial, which proved successful in preventing TB disease using levofloxacin in adults and adolescents, albeit with higher reported adverse events compared with TB-CHAMP, supporting the selection of the doses in the higher weight bands.
Figure 4.
Exposures measured as area under the concentration–time curve at steady state (AUC0–24) in children ≥5 years old simulated using the proposed dosing schedule with the WHO harmonized weight bands using the nondispersible tablets. Exposures were simulated using the models by White et al [16] (A), Peloquin et al [23] (B), and van der Elsen et al [24] (C).
Dosing Tables
As a result of the simulation exercise done, and evaluating the exposure obtained overall and on a country level, dosing was proposed in the WHO harmonized weight bands. These included doses of 10–42 mg/kg for the 250 mg tablet and 8–33 mg/kg for the 100 mg dispersible tablet. The proposed dosing across weight bands is shown in Table 2.
DISCUSSION
Evidence from 2 recent randomized controlled trials in children and adults, TB-CHAMP [6] and V-QUIN [7], has recently informed updated WHO recommendations on the use of 6 months of levofloxacin as an option for TPT among people exposed to MDR/RR-TB [4]. Levofloxacin-based TPT is well tolerated and anticipated to reduce treatment burden of MDR/RR-TB disease, which remains substantial. The risk of progression from TB infection to TB disease is higher in young children [27], making it crucial to reach adequate levofloxacin exposure when giving TPT to this population [28, 29]. For children, it is also essential to have suitable formulations that are acceptable for them and easy for them to take and caregivers to administer, ensuring that they receive the full therapeutic benefit with the lowest possible treatment burden. Because the titration to the optimal dose is easier with the dispersible tablet, and because this formulation is more acceptable in younger children [28, 29], it should be prioritized for use, whenever available.
Developing a practical dosing guidance for the levofloxacin-based TPT regimen using currently available formulations is key to optimize target levofloxacin exposures, while ensuring adherence and reducing dosing errors. The dosing strategy proposed here is framed in the context of the harmonized weight bands being used by WHO to simplify dosing and minimize errors, especially with simultaneous treatments for multiple diseases and comorbidities [15]. This new approach aims for a dosing schedule based on body weight. In the case of levofloxacin TPT, an additional stratification according to age was only required for children weighing 3.0–9.9 kg, given that drug metabolism in the first 6 months of life may not be mature enough and thus cause toxicity. This would provide a more appropriate dose to infants aged <6 months who have a low weight-for-age. The addition of an age stratification on the lower weight bands is included in WHO dosing guidance for other TB medicines including bedaquiline, delamanid, and rifapentine [3, 30].
Our approach to determining an adequate dosing strategy in children is fundamentally different from directly adopting a specific dose per weight (mg/kg/day) as tested in clinical trials. Simple adoption of uniform dose per weight across different weights and ages is not congruent with the rules of allometric scaling [31] as well as the ontogeny of drug metabolism that strongly influences the drug exposure in young children [32]. First, we used the population PK model that was built on the most comprehensive age and weight group [16]. Second, we used the demographics of children from countries with the highest burden of RR- and MDR-TB to perform dosing simulation and dose optimization. The model used in this dosing simulation uses maturation function of clearance to account for an effect of age, so it was important to perform dosing simulation on a dataset of children with a combination of age and weight frequently observed in high-burden TB settings, where TPT will be most frequently administered.
We also performed a sensitivity analysis using different parameter estimates of the maturation function to ensure that younger children in this weight band would have adequate exposure to the drug even if the maturation function was very different from the one estimated. This analysis showed that exposure among younger children would be comparable or higher than in older children in the same weight band. Although the exposure among those <3 months of age and weighing 3–5.9 kg is estimated to be the highest of all groups when assuming the smaller impact of maturation function, this approximates the exposure target of levofloxacin for treatment of TB disease (101 mgh/L) [26]. Safety data for this exposure target are also available in a larger number of children [33].
Our work has limitations. First, the exposure target in this study is calculated rather than observed. This approach was used because drug exposure was not measured as part of the TB-CHAMP trial, even though dosing was informed from a lead-in PK study prior to opening the randomized trial [9]. The inclusion of PK measurements in clinical trials helps establish exposure-outcome relationships upfront and facilitate timely optimization of the dosing strategy across populations. Second, the dosing simulation was performed in children from the DHS and other demographic surveys. Given that child nutrition is likely to have improved since the time that these demographic data were collected, it is possible that prevalence of malnutrition in comparison to the current populations was overestimated. Nevertheless, prevalence of malnutrition was similar between the subset of our dataset from South Africa and participants of TB-CHAMP.
In conclusion, the findings from this study provide important insights into an optimal and pragmatic dosing strategy for TPT with levofloxacin and have contributed to the WHO updated dosing guidance. The approach used in this work leverages results from clinical trials, large-scale demographic databases, and PK information to ensure an optimal dosing strategy of drugs and improve global health outcomes. Practical dosing strategies that support adherence are vital for the success of TPT in children and the reduction of disease transmission and progression. Making dispersible levofloxacin more affordable would enhance further its uptake by communities at risk. The widespread implementation of safe and effective TPT for MDR/RR-TB in children is another tool to help reduce the burden of MDR/RR-TB disease and death in the world.
Supplementary Material
Notes
Author contributions. B. P. S., R. M., M. S., and R. M. S. contributed to the analysis plan. B. P. S., R. M., and M. S. prepared the data for analysis. B. P. S., with support from R. M., M. S., and R. M. S., performed the modeling and simulation work. B. P. S., R. S., and M. S. drafted the figures, tables, and manuscript, with critical editorial support from A. C. H., T. M., A. K., D. F., Y. W. and R. M. S.
Disclaimer. A. K., D. F., and T. M. are WHO staff members. The views expressed in this article are those of the authors and do not necessarily represent the views, decisions, or policies of the WHO.
Financial support. None.
All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.
Contributor Information
Belén P Solans, Department of Bioengineering and Therapeutics, Schools of Pharmacy and Medicine, University of California, San Francisco, San Francisco, California, USA; Center for Tuberculosis, University of California, San Francisco, San Francisco, California, USA.
Ryo Miyakawa, Department of Bioengineering and Therapeutics, Schools of Pharmacy and Medicine, University of California, San Francisco, San Francisco, California, USA; Center for Tuberculosis, University of California, San Francisco, San Francisco, California, USA.
Maureen Shin, Department of Bioengineering and Therapeutics, Schools of Pharmacy and Medicine, University of California, San Francisco, San Francisco, California, USA; Center for Tuberculosis, University of California, San Francisco, San Francisco, California, USA.
Anneke C Hesseling, Desmond Tutu Tuberculosis Centre, Department of Paediatrics and Child Health, Faculty of Medicine and Health Sciences, Stellenbosch University, Stellenbosch, South Africa.
Yasmine White, Department of Bioengineering and Therapeutics, Schools of Pharmacy and Medicine, University of California, San Francisco, San Francisco, California, USA; Center for Tuberculosis, University of California, San Francisco, San Francisco, California, USA.
Tiziana Masini, Global Tuberculosis Programme, World Health Organization, Geneva, Switzerland.
Avinash Kanchar, Global Tuberculosis Programme, World Health Organization, Geneva, Switzerland.
Dennis Falzon, Global Tuberculosis Programme, World Health Organization, Geneva, Switzerland.
Radojka M Savic, Department of Bioengineering and Therapeutics, Schools of Pharmacy and Medicine, University of California, San Francisco, San Francisco, California, USA; Center for Tuberculosis, University of California, San Francisco, San Francisco, California, USA.
Supplementary Data
Supplementary materials are available at The Journal of Infectious Diseases online (http://jid.oxfordjournals.org/). Supplementary materials consist of data provided by the author that are published to benefit the reader. The posted materials are not copyedited. The contents of all supplementary data are the sole responsibility of the authors. Questions or messages regarding errors should be addressed to the author.
References
- 1. World Health Organization (WHO) . Global tuberculosis report. Geneva, Switzerland: WHO, 2023. [Google Scholar]
- 2. World Health Organization (WHO) . WHO consolidated guidelines on tuberculosis. Module 4: treatment—drug-resistant tuberculosis treatment, 2022 update. 2022. Available at: http://www.ncbi.nlm.nih.gov/books/NBK588564/. Accessed 16 July 2024. [Google Scholar]
- 3. World Health Organization (WHO) . WHO operational handbook on tuberculosis. Module 1: prevention. Geneva, Switzerland: WHO, 2024. [Google Scholar]
- 4. World Health Organization (WHO) . WHO consolidated guidelines on tuberculosis. Module 1: prevention—tuberculosis preventive treatment. Geneva, Switzerland: WHO, 2024. [Google Scholar]
- 5. Fish DN, Chow AT. The clinical pharmacokinetics of levofloxacin. Clin Pharmacokinet 1997; 32:101–19. [DOI] [PubMed] [Google Scholar]
- 6. Hesseling AC, Purchase SE, Martinson NA, et al. Levofloxacin preventive treatment in children exposed to MDR tuberculosis. N Engl J Med 2024; 391:2315–26. [DOI] [PubMed] [Google Scholar]
- 7. Fox GJ, Nhung NV, Cam Binh N, et al. Levofloxacin for the prevention of multidrug-resistant tuberculosis in Vietnam. N Engl J Med 2024; 391:2304–14. [DOI] [PubMed] [Google Scholar]
- 8. Duong T, Brigden J, Simon Schaaf H, et al. A meta-analysis of levofloxacin for contacts of multidrug-resistant tuberculosis. NEJM Evid 2025; 4:EVIDoa2400190. [DOI] [PubMed] [Google Scholar]
- 9. Garcia-Prats AJ, Purchase SE, Osman M, et al. Pharmacokinetics, safety, and dosing of novel pediatric levofloxacin dispersible tablets in children with multidrug-resistant tuberculosis exposure. Antimicrob Agents Chemother 2019; 63:e01865-18. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Denti P, Garcia-Prats AJ, Draper HR, et al. Levofloxacin population pharmacokinetics in South African children treated for multidrug-resistant tuberculosis. Antimicrob Agents Chemother 2018; 62:e01521-17. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. van der Laan LE, Hesseling AC, Schaaf HS, et al. Pharmacokinetics and optimized dosing of dispersible and non-dispersible levofloxacin formulations in young children. J Antimicrob Chemother 2023; 78:2481–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. World Health Organization (WHO) . Rapid communication: tuberculosis preventive therapy. 2024. Available at: https://iris.who.int/bitstream/handle/10665/375983/9789240089723-eng.pdf?sequence=1. Accessed February 2025.
- 13. Van Wijk RC, Imperial MZ, Savic RM, Solans BP. Pharmacokinetic analysis across studies to drive knowledge-integration: a tutorial on individual patient data meta-analysis (IPDMA). CPT Pharmacomet Syst Pharmacol 2023; 12:1187–200. [Google Scholar]
- 14. Chua HC, Tam VH. Optimizing clinical outcomes through rational dosing strategies: roles of pharmacokinetic/pharmacodynamic modeling tools. Open Forum Infect Dis 2022; 9:ofac626. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Waalewijn H, Almett M, Wasmann RE, et al. Simplifying medicine dosing for children by harmonising weight bands across therapeutic areas. Lancet Child Adolesc Health 2025; 9:274–82. [DOI] [PubMed] [Google Scholar]
- 16. White YN, Solans BP, Denti P, et al. Pharmacokinetics and optimal dosing of levofloxacin in children for drug-resistant tuberculosis: an individual patient data meta-analysis. Clin Infect Dis 2024; 78:756–64. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. ICF . The DHS program. Available at: https://dhsprogram.com/Data/. Accessed 17 February 2024.
- 18. Carolina Population Center . China health and nutrition survey. Available at: https://www.cpc.unc.edu/projects/china. Accessed 17 February 2024.
- 19. Higher School of Economics . The Russia longitudinal monitoring survey. Available at: https://rlms-hse.cpc.unc.edu. Accessed 17 February 2024.
- 20. Savic R, Ruslami R, Hibma J, et al. Pediatric tuberculous meningitis: model-based approach to determining optimal doses of the anti-tuberculosis drugs rifampin and levofloxacin for children. Clin Pharmacol Ther 2015; 98:622–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Mase SR, Jereb JA, Gonzalez D, et al. Pharmacokinetics and dosing of levofloxacin in children treated for active or latent multidrug-resistant tuberculosis, Federated States of Micronesia and Republic of the Marshall Islands. Pediatr Infect Dis J 2016; 35:414–21. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Malik AA, Brooks MB, Siddiqui S, et al. Pharmacokinetics of levofloxacin in children treated for exposure to drug-resistant tuberculosis. Antimicrob Agents Chemother 2019; 63:e02569-18. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Peloquin CA, Hadad DJ, Molino LPD, et al. Population pharmacokinetics of levofloxacin, gatifloxacin, and moxifloxacin in adults with pulmonary tuberculosis. Antimicrob Agents Chemother 2008; 52:852–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. van den Elsen SHJ, Sturkenboom MGG, Van’t Boveneind-Vrubleuskaya N, et al. Population pharmacokinetic model and limited sampling strategies for personalized dosing of levofloxacin in tuberculosis patients. Antimicrob Agents Chemother 2018; 62:e01092-18. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Seddon JA, Garcia-Prats AJ, Purchase SE, et al. Levofloxacin versus placebo for the prevention of tuberculosis disease in child contacts of multidrug-resistant tuberculosis: study protocol for a phase III cluster randomised controlled trial (TB-CHAMP). Trials 2018; 19:693. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Peloquin CA, Phillips PPJ, Mitnick CD, et al. Increased doses lead to higher drug exposures of levofloxacin for treatment of tuberculosis. Antimicrob Agents Chemother 2018; 62:e00770-18. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Martinez L, Cords O, Horsburgh CR, et al. The risk of tuberculosis in children after close exposure: a systematic review and individual-participant meta-analysis. Lancet 2020; 395:973–84. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. Purchase SE, Wademan DT, Tshetu NL, et al. Holistic acceptability of an adult levofloxacin formulation in children and adolescents on a tuberculosis preventive treatment trial. PLoS Glob Public Health 2024; 4:e0003381. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Wademan DT, Draper HR, Purchase SE, et al. Acceptability of levofloxacin dispersible and non-dispersible tablet formulations in children receiving TB preventive treatment. IJTLD Open 2024; 1:69–75. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. World Health Organization (WHO). WHO operational handbook on tuberculosis. Module 5: management of tuberculosis in children and adolescents. 1st ed. Geneva, Switzerland: WHO, 2022. [Google Scholar]
- 31. Anderson BJ, Holford NHG. Mechanistic basis of using body size and maturation to predict clearance in humans. Drug Metab Pharmacokinet 2009; 24:25–36. [DOI] [PubMed] [Google Scholar]
- 32. Germovsek E, Barker CIS, Sharland M, Standing JF. Scaling clearance in paediatric pharmacokinetics: all models are wrong, which are useful? Br J Clin Pharmacol 2017; 83:777–90. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33. Garcia-Prats AJ, Draper HR, Finlayson H, et al. Clinical and cardiac safety of long-term levofloxacin in children treated for multidrug-resistant tuberculosis. Clin Infect Dis 2018; 67:1777–80. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.