Linezolid Population Pharmacokinetic Model in Plasma and Cerebrospinal Fluid Among Patients With Tuberculosis Meningitis

Abstract

Background: Linezolid is evaluated in novel treatment regimens for tuberculous meningitis (TBM). Linezolid pharmacokinetics have not been characterized in this population, particularly in cerebrospinal fluid (CSF), as well as, following its co-administration with high-dose rifampicin. We aimed to characterize linezolid plasma and CSF pharmacokinetics in adults with TBM.

Methods: In the LASER-TBM pharmacokinetic substudy, the intervention groups received high-dose rifampicin (35 mg/kg) plus 1200 mg/day of linezolid for 28 days, which was then reduced to 600 mg/day. Plasma sampling was done on day 3 (intensive) and day 28 (sparse). A lumbar CSF sample was obtained on both visits.

Results: Thirty participants contributed 247 plasma and 28 CSF observations. Their median age and weight were 40 years (range, 27-56) and 58 kg (range, 30-96). Plasma pharmacokinetics was described by a 1-compartment model with first-order absorption and saturable elimination. Maximal clearance was 7.25 L/h, and the Michaelis-Menten constant was 27.2 mg/L. Rifampicin cotreatment duration did not affect linezolid pharmacokinetics. CSF-plasma partitioning correlated with CSF total protein up to 1.2 g/L, where the partition coefficient reached a maximal value of 37%. The plasma-CSF equilibration half-life was ∼3.5 hours.

Conclusions: Linezolid was readily detected in CSF despite high-dose rifampicin coadministration. These findings support continued clinical evaluation of linezolid plus high-dose rifampicin for the treatment of TBM in adults. Clinical Trials Registration. ClinicalTrials.gov (NCT03927313).

Keywords: cerebrospinal fluid; linezolid; modeling and simulation; population pharmacokinetics; tuberculosis meningitis.

Figure 1.

Representation of the final model: k_tr is the rate constant for the drug passage through the transit compartments; k_plasma-CSF is the equilibration rate constant for plasma–cerebrospinal fluid (CSF), which describes how soon the change in plasma is reflected in the CSF; k_a, absorption rate constant; PPC_plasma-CSF is the pseudo-partition coefficient, which represents the ratio of drug in CSF to the plasma.

Figure 2.

The relationship of the pseudo-partition coefficient (PPC) vs the cerebrospinal fluid (CSF) protein level via the piecewise function (broken stick). The solid line represents the median, and the shaded areas represent the uncertainty around the estimates of the breakpoint (the maximal CSF protein value at which PPC_max is reached) and the calculated slope. The dashed line depicts the extrapolated part of the PPC–CSF protein relationship for CSF protein values outside the range observed in the study cohort; the lowest observed value was 0.22 mg/mL. The ticks on the x-axis represent the values of the CSF protein observed in our cohort; CSF protein values >3 mg/mL were truncated for better figure visibility. Some ticks are overlapping because there are some duplicated CSF protein values.

Figure 3.

Simulated typical concentration-time profiles for plasma and cerebrospinal fluid (CSF) for the oral daily dose of linezolid: 1200 and 600 mg. The solid and dashed lines represent the median for the plasma and CSF, respectively, and the shaded areas represent the 90% CIs. The horizontal dotted line indicates the wild type minimum inhibitory concentration (MIC) value of linezolid for Mycobacterium tuberculosis.

Figure 4.

Secondary model-derived exposure parameters: AUC_0-24h and concentration at 24 hours postdose (C_24h) stratified by dose. Horizontal lines indicates medians. Boxes represent IQRs while whiskers are the 2.5th and 97.5th percentiles. Dots represent individual values: n = 7, 600 mg; n = 40, 1200 mg (n = 30, day 3; n = 10, day 28). AUC, area under the curve.