Drug recovery in paediatric tuberculous meningitis

Tshavhungwe, Mvuwo phophi

Drug recovery in paediatric tuberculous meningitis

Thesis / Dissertation

2025

Publisher

University of Cape Town

Department

Division of General Surgery

Faculty

Faculty of Health Sciences

Abstract

BACKGROUND: Tuberculosis (TB) remains the single leading infectious disease killer across the world. Infection of the central nervous system (CNS) is the most lethal form of the disease. There has been little adjustment to the standard drug regimens over several decades, but there remain many unknowns about their effectiveness, especially because penetration across the blood-brain barrier (BBB) has been difficult to study. Rifampicin (RIF) is a leading drug in the management of TBM but concentrations in ventricular cerebrospinal fluid (VCSF) have not been studied in TBM patients, and emerging evidence suggests that lumbar spinal CSF (LCSF) concentrations of substances may not fully reflect brain concentrations. Finally, there are no data on brain extracellular fluid (ECF) concentrations of TB drugs in humans. AIMS AND OBJECTIVES The aim was to produce the first pilot data of RIF concentrations in VCSF in children with TBM and to explore the use of microdialysis (MD) to detect RIF in brain ECF. The primary objective was to measure RIF total concentrations in multiple compartments, namely plasma, LCSF, VCSF, and brain ECF. The secondary objectives were 1) to examine RIF concentrations in paired, time-linked L- and VCSF of TBM patients (samples taken at the same time), and 2) to determine total protein concentrations in time-linked L-and VCSF. METHODS: This study prospectively recruited children with definite or probable TBM in a descriptive cross-sectional study design at a university-affiliated paediatric hospital in Cape Town. Patients were treated with standard TB drug regimens. Sampling was performed after standard drug administration. Blood samples were taken from routine procedures or long lines or arterial lines. LCSF was sampled from scheduled clinically indicated procedures. VCSF was accessed from external ventricular drains (EVD) or CSF shunt insertion. In a subgroup of patients, VCSF and LCSF were taken at the same time (usually from clinically indicated column tests). A subgroup of patients underwent bedside MD monitoring of brain chemistry for clinical purposes. Remnant fluid was stored and examined offline for RIF concentrations in brain ECF. RESULTS: A total of 61 children who fulfilled the definition of definite or probable TBM and who had samples analysed for RIF were recruited to the study; the median age was 2.3 years and 7% were human immunodeficiency (HIV) positive. All patients had hydrocephalus (HCP) and most presented in stage 2 and 3. Plasma concentrations peaked at 2 hours (hrs) post-dose while CSF values peaked after 4 hrs. Below level of quantification percentages were 8%, 15%, 16% and 22% respectively for plasma, LCSF, VCSF, and ECF respectively. The median peak concentration in plasma was 5.2 μg/mL. By comparison, LCSF concentrations at 4 and 6 hrs were 0.23 and 0.14 μg/mL and VCSF concentrations were 0.15 μg/mL and 0.14 μg/mL respectively. CSF concentrations were at most 5% of peak plasma concentrations. RIF was detectable in MD samples (n=74) but showed the lowest concentrations. ECF concentrations ranged from 0.01-0.04 μg/mL. In the first 10 hrs of sampling, brain ECF RIF concentrations were 13-26% of corresponding VCSF concentrations. In a subgroup of patients with time-linked LCSF and VCSF samples (n=28), median RIF concentrations were significantly lower in the VCSF:133 ng/mL (range, 7.40-937 ng/mL) versus LCSF: 299 ng/mL (range, 5-1080 ng/mL) as were protein concentrations: 1.3 g/L (0.17-7.44 g/L) and 6.0 g/L (range, 0.68-51.8 g/L). LCSF showed higher RIF concentrations in 64% of paired samples, and higher protein concentrations in all paired samples. There were no apparent differences in the time-to-peak drug concentrations in the two CSF compartments. CONCLUSION: Our data adds to the limited knowledge of CNS distribution of RIF, an important drug in TBM management. We report RIF concentrations in VCSF for the first time in patients with TBM. Intriguingly, the paired sample analysis shows higher concentrations in LCSF, which may reflect a sump effect or differential permeability in the spinal blood vessels. Given that the current knowledge of RIF in TBM patients is largely based on spinal CSF studies, it is important to note that this may overestimate concentrations in VCSF. We also demonstrated for the first time the capacity to serially sample RIF directly from brain ECF, where concentrations were lower than in VCSF. This may be technical but also may reflect a difference between total and free drug concentrations, given the differential barrier properties of the BBB (former) and the blood-CSF barrier (BCSFB) (latter). This study contributes timely and important data in an era of focus on maximising RIF exposure by increasing drug dosing.

Keywords

Tuberculous

Meningitis

Reference:

Collections

PhD / Doctoral

Full item page