Cerebrospinal Fluid and Tuberculous Meningitis

To the Editor—We read with interest the recent publication by Kempker et al regarding the cerebrospinal fluid (CSF) penetration of second-line tuberculosis (TB) drugs [1]. The authors conducted serial serum and CSF sampling in patients treated for tuberculous meningitis (TBM) to measure the concentrations of linezolid, cycloserine, clofazimine, delamanid, and bedaquiline. CSF concentrations of clofazimine, delamanid, and bedaquiline were below the limit of detection in the CSF, and the authors concluded that the utility of these drugs for treatment of TBM is uncertain.

TBM, contributing 1%–2% of the annual global TB burden, is the form of TB most likely to cause severe disability or death [2]. As many TBM trials are focused on optimizing the dose of first-line TB drugs, we applaud the authors for exploring novel drugs. Bedaquiline was first approved by the United States Food and Drug Administration in 2012 and was first measured in CSF of a patient with multidrug-resistant TBM in 2016; in that study, bedaquiline was below the limit of quantification at all time points [3]. There are 2 challenges with measuring bedaquiline in CSF. First, bedaquiline has high protein binding (>99.9%) and adsorbs to collection, storage, and analysis equipment [4]. Second, only unbound drug is available to penetrate the blood–brain and blood–CSF barriers, so the limits of quantification of an assay must be low enough to measure the drug that diffuses freely into CSF, which would be roughly 1/1000th of total drug concentrations in plasma. The authors used a lower limit of quantification of 0.01 μg/mL, which is too high for the expected CSF concentrations for a plasma range of 0.36–3.15 µg/mL the authors found in their study participants.

Our research group investigated bedaquiline concentrations in CSF in participants with pulmonary TB in a clinical trial [5]. We collected CSF directly from the lumbar puncture needle into low-binding tubes, tubes were lined with 5% bovine serum albumin, and the lower limit of quantification was set to 0.0001 µg/mL. Bedaquiline and its primary metabolite were measurable in all samples at all time points, ranging from 0.45 to 3.79 ng/mL, which was approximately 0.01% of the plasma concentrations, suggesting that unbound bedaquiline in plasma diffuses freely into the CSF. It is likely that bedaquiline readily enters brain tissue due to its high lipophilicity. Bedaquiline penetrates the brain tissue of rodents in concentrations exceeding the minimum inhibitory concentration, with a radiolabeled bedaquiline area under the curve of tissue to plasma ratio of 14.81% when dosed intravenously [6, 7]. In case reports, the addition of bedaquiline to a TBM regimen has improved clinical course, though data are sparse [8].

Bedaquiline does not enter the central nervous system in total drug concentrations similar to those in the plasma; however, this is expected and the low concentrations of active drug found in CSF should not deter researchers from further exploration of its use in TBM. Whether the concentrations of bedaquiline in the central nervous system are enough to improve TBM outcomes is unknown. We encourage researchers to evaluate its use in clinical practice and to include bedaquiline-containing regimens in clinical trials for patients with TBM.