Preclinical development and characterisation of PP353, a formulation of linezolid for intradiscal administration

Abstract

Introduction: Bacterial infection of the intervertebral disc can lead to vertebral endplate edema known as Modic changes, with associated chronic low back pain. Oral antimicrobial therapy has shown efficacy but relies on prolonged dosing and may not be optimal in terms of patient outcome, side effects, or antibiotic stewardship. There is no antibiotic formulation approved for intradiscal administration. Here, we describe the development and preclinical characterization of a formulation of linezolid, a suspension of 50 mg/mL micronized powder, for intradiscal administration.

Methods: Micronization, particle size analysis, Franz cell diffusion assays, ex vivo bioassay, and estimates of gelling temperature were used to optimize the composition and properties of the formulation. Performance of the formulation was assessed using sheep to characterize the pharmacokinetics and a model of intradiscal infection was developed to demonstrate efficacy. Suitability for human administration was demonstrated in a Good Laboratory Practice (GLP) local tolerance study.

Results: Micronized linezolid, formulated as a powder suspension using a vehicle containing poloxamer 407 and iohexol, provided a temperature-dependent radio-opaque gel that was suitable for image-guided percutaneous intradiscal administration. Efficacy in a sheep model of intradiscal Staphylococcus aureus infection was demonstrated. The formulation provides a high level of sheep disc tissue exposure, with C_max of 6500 μg/g and limited systemic exposure, with a plasma C_max of 0.04 μg/mL per 0.1 mL dose (5 mg of linezolid). Deconvolution of plasma linezolid pharmacokinetics correlated with linezolid remaining in the disc over time. Observations from a GLP local tolerance study with the linezolid formulation were of a minor nature and related to the intradiscal administration procedure.

Conclusions: Linezolid can be formulated for image-guided percutaneous intradiscal administration. The formulation is now in a Phase 1b clinical trial to evaluate safety, pharmacokinetics, and efficacy in patients with CLBP and suspected bacterial infection.

Keywords: Modic; intradiscal; linezolid; pharmacokinetics; vertebrogenic back pain.

FIGURE 1

(A) Effect of linezolid particle size on linezolid release (mean ± SD). (B) Effect of linezolid and poloxamer 407 concentration on linezolid release (mean ± SD).

FIGURE 2

Design of the sheep intradiscal infection efficacy study.

FIGURE 3

In vivo efficacy of intradiscal linezolid. Bacterial burden values are geometric mean ± geometric SD. The potentially sterile discs were plotted as 10 CFU/g disc as an estimate of the limit of detection. The Mann‐Whitney Rank Test p‐values for colocalized linezolid and non‐colocalized linezolid dosing compared to vehicle controls are 0.002 and 0.003 respectively.

FIGURE 4

Lateral view of the X‐ray image of the sheep lumbar spine after intradiscal administration of PP353.

FIGURE 5

(A) Pharmacokinetics of PP353 in plasma and nucleus pulposus (mean ± SD). Linezolid concentrations in plasma after intradiscal (i.d.) administration of linezolid (5 mg/disc, 4 discs per animal) to sheep. Plasma concentrations over the first 24 h are mean of 6 animals and at 48 h 4 animals. The intradiscal concentrations are mean of 8 discs, 4 discs in each of 2 animals. (B) Plasma pharmacokinetics of linezolid after Intradiscal and intramuscular administration of PP353 (mean ± SD). Linezolid concentrations in plasma after intradiscal and intramuscular administration of linezolid (5 mg, 2 discs or 2 i.m. sites per sheep; 10 mg total) to sheep. Plasma concentrations up to 16 h are mean of 9 animals and beyond 16 h the mean of 6 animals.

FIGURE 6

Correlation between intradiscal linezolid concentration in the non‐GLP study estimated by deconvolution of plasma concentration and that measured in disc tissues. Note that this is the estimate for 4 discs receiving the targeted 5 mg linezolid per disc.