Addressing critical knowledge gaps to improve and shorten multidrug-resistant tuberculosis treatment regimens in children

[摘要] ENGLISH ABSTRACT:Multidrug-resistant (MDR) tuberculosis (TB), defined as TB disease or infection causedby Mycobacterium tuberculosis with resistance to at least both isoniazid and rifampicin,threatens global TB control, with an estimated 490,000 incident cases of MDR-TBglobally in 2016. The burden of paediatric MDR-TB has been poorly characterized todate. However, recent modeling studies estimate that there are approximately 26,000-32,000 incident MDR-TB cases in children (< 15 years of age) worldwide each year.Traditionally, treatment regimens for adults and children were constructed using aminimum of four second-line antituberculosis drugs likely to be effective, including asecond-line injectable medication, for up to 6 months, and a total duration of treatmentof up to 18-24 months. In 2016, the World Health Organization (WHO) recommended ashortened (9-12 month) treatment regimen, which still includes an injectable drug forfour months. In addition, the development and increasing use of the novel TB drugsbedaquiline and delamanid, are radically altering the MDR-TB treatment landscape,although children have lagged behind in accessing these important developments.Treatment outcomes for adults with MDR-TB have been persistently poor, with 54%successfully treated in 2014 both overall globally, and in South Africa. In contrast,treatment outcomes among children with MDR-TB are generally good, with 78-90%successfully treated under routine clinical conditions. However, current paediatricMDR-TB treatment regimens have important limitations. These current regimensremain long (9-18 months or more), which is costly and burdensome. There are alsofrequent adverse effects, including from the second-line injectable medications(amikacin, kanamycin, capreomycin) that cause permanent sensorineural hearing lossin up to 24% of children treated long-term. Additionally, the injectables are mainlygiven by painful daily intramuscular injections, resulting in trauma and distress forpatients, their caregivers and healthcare providers. Therefore, it is an urgent priority todevelop more optimal treatment regimens for children with MDR-TB that retain theirefficacy but are shorter, more child-friendly, are better tolerated, safer and which do not require the use of an injectable medication.The purpose of this doctoral research was to address critical knowledge gaps inpaediatric MDR-TB treatment, with the aim of informing more effective, safer, and morechild-friendly MDR-TB treatment strategies in children. I identified critical knowledge gaps related to the pharmacokinetics, including the effects of formulation, optimaldosing, safety, and tolerability of key second-line and novel antituberculosis drugs inchildren, and completed complementary studies on ofloxacin, levofloxacin, linezolid,amikacin and bedaquiline designed to address these knowledge gaps.In an observational study of the pharmacokinetics and safety of ofloxacin in childrenroutinely treated for MDR-TB disease or exposure, exposures after a daily 20mg/kgofloxacin dose were well below target exposures from adults receiving the routine 800mg dose. Ofloxacin was safe and well tolerated, with few musculoskeletal complaints orserious adverse events. This data adds to the evidence of the safety of fluoroquinolonesin children even with long-term use, and identifies the need to revise ofloxacinpaediatric doses.Subsequently, in this large observational study, the population pharmacokinetics oflevofloxacin among children with MDR-TB disease or exposure was characterized usingnon-linear mixed effects modeling. One hundred and nine children treated with theroutinely available adult 250 mg tablet formulation of levofloxacin at daily doses of 15mg/kg or 20 mg/kg were included. Levofloxacin's apparent oral clearance (CL/F) washigher than expected based on previously published data, possibly due to theformulation studied. Simulations using the final model targeting exposures in adultswith TB receiving 750 mg of levofloxacin identified weight-banded doses that weremuch higher than previously in use (18 mg/kg to nearly 40 mg/kg daily). It wasconcluded that levofloxacin dosing in children should be reassessed, formulation effectsexplored further, and that safety should be carefully evaluated if higher levofloxacindoses are used.Building on this data, I completed an evaluation of the safety of long-term levofloxacinin children treated for MDR-TB. Among 70 children, median age 2.1 years, treated for amedian of 11.6 months, levofloxacin was generally safe and was well tolerated. Therewere no Grade 4 or serious adverse events, and few musculoskeletal events. There wasno QT-interval prolongation and no association of QT interval with levofloxacinconcentration. This study supported the safety of long-term fluoroquinolone treatmentin children, and provided novel data on the QT prolonging effect of levofloxacin, whichis needed, as increasingly levofloxacin is being combined with other QT prolongingmedications.The effects of drug formulation in pharmacokinetic studies are critically important. In alead-in pharmacokinetics study to the TB-CHAMP trial (phase 3 cluster randomized trialcomparing levofloxacin vs. placebo for prevention of TB in child contacts of MDR-TBcases), 24 children had pharmacokinetic sampling with a novel dispersible tabletformulation of levofloxacin. The levofloxacin exposures were much higher with thisnovel formulation compared to those seen in the previously reported study using theadult 250 mg levofloxacin tablet. Combining these two data sets using non-linear mixedeffects modeling identified that reduced bioavailability of the adult 250 mg tabletformulation compared to the dispersible levofloxacin tablet was the explanation for thesubstantial differences in exposures. This study highlighted the importance offormulation considerations to paediatric pharmacokinetic studies and providedpractical weight-banded dosing guidelines for use of this formulation now becomingavailable in the field.Linezolid is a key drug with an increasingly important role in the treatment of MDR-TBstrains with additional resistance and in central nervous system TB disease. Iperformed a structured review of the literature on linezolid to inform its use in childrenfor MDR-TB treatment and identify knowledge gaps for future research. Few childrentreated with linezolid for MDR-TB were described in the literature. As in adults,linezolid appeared to be effective but was associated with frequent adverse events.There was no data on linezolid pharmacokinetics in children with TB. Practical interimguidance was provided for linezolid use in children. Priority research needs identifiedincluded studying linezolid pharmacokinetics in children with TB, characterization of itssafety with long-term use, and its optimal dose for TB in MDR-TB regimens goingforward.Following on this review, an analysis of linezolid pharmacokinetics and safety fromchildren with MDR-TB was performed with data from 48 children combined from twoobservational studies using non-linear mixed effects modeling. Seventeen childrenreceived long-term linezolid and were monitored longitudinally for safety; 31 childrenonly contributed cross-sectional pharmacokinetic data after a single-dose of linezolid.After accounting for the effects of weight with allometric scaling, no other covariatessignificantly contributed to the model. Exposures were higher than expected, based onpreviously reported data. Ten of 17 participants had a linezolid related adverse event, including five Grade 3 or 4 events; anaemia was the most common event. This first dataon linezolid pharmacokinetics in children demonstrated higher than expectedexposures and frequent, serious linezolid-related adverse events, and will inform theuse and future dosing recommendations of this increasingly important antituberculosismedication in children.While drug substitutions for injectable drugs are not yet available for many children,improving the tolerability of the continued use of second-line injectable medications isan important question to address in children. A randomized two-period crossoverstudy was designed to characterize the effect of co-administration of lidocaine on thepain and pharmacokinetics of intramuscular amikacin. Children each received a dose ofamikacin with and without additional lidocaine on separate days, and were randomizedto the sequence of treatments; pain assessments and pharmacokinetic sampling wereperformed on each day. Twelve children were enrolled and completed the study. Theaddition of lidocaine reduced pain immediately after the injection, was safe, and did notaffect the pharmacokinetics of amikacin in children, and should be considered as aroutine policy in patients with MDR-TB receiving an injectable agent.The novel drug bedaquiline is increasingly used globally and in South Africa for adultswith MDR-TB, and ongoing paediatric trials will characterize the pharmacokinetics,safety and optimal dose in children. The paediatric formulation, which is beingevaluated in at least one of the ongoing paediatric trials, may not be available forroutine care for some time. In order to inform the rational use of the adult bedaquilineformulation in young children, a randomized two-period crossover study in healthyadult volunteers was designed. Adult bedaquiline tablets administered suspended inwater were bioequivalent to adult tablets swallowed whole. The suspended tabletswere also found to be acceptable and palatable to the majority of participants, animportant finding considering that crushing or suspending some tablets, such as thefluoroquinolones, reduces their palatability and acceptability substantially. This datawill accelerate access to bedaquiline for young children in research and routine care.In conclusion, this doctoral research has addressed a number of important key knowledge gaps related to more optimal paediatric MDR-TB treatment. This researchhas raised a number of follow-up questions that have informed subsequent studies that will continue to advance the field towards a goal of effective, safe, shorter MDR-TBtreatment for all children.

[发布机构] Stellenbosch University