Milda Girdenytė: Pharmacokinetic Mechanisms in the Development and Prevention of Chemotherapy-Induced Peripheral Neuropathy

Abstract

Chemotherapy-induced peripheral neuropathy (CIPN) is a major dose-limiting adverse effect of anticancer microtubule-targeting agents (MTAs). However, the pharmacologically active unbound exposure of MTAs at CIPN-relevant sites, including peripheral and central nervous system tissues and skeletal muscle, and its relationship to neurotoxicity and biomarkers remain incompletely characterized. The overall aim of this thesis was to investigate MTAs distribution and intra- and extracellular exposure at CIPN-relevant sites using preclinical models and integrating these findings with pharmacokinetic-pharmacodynamic (PKPD) modelling and biomarker evaluation to improve understanding and monitoring of CIPN in patients.

Particular emphasis was placed on paclitaxel-induced peripheral neuropathy (PIPN), with focus on drug formulation, translational pharmacokinetics, and serum neurofilament light chain (sNfL) as a biomarker of axonal injury. Albumin-bound and micellar paclitaxel formulations showed greater unbound distribution across endothelial barriers at CIPN-relevant sites, expressed as the unbound tissue-to-plasma concentration ratio (Kp,uu,tissue), than paclitaxel formulated with Cremophor EL. Results provide a plausible mechanistic explanation for the higher incidence of CIPN with non-Cremophor EL formulations. Based on PIPN site pharmacokinetic data, a semi-mechanistic PKPD model was developed linking sNfL kinetics to intracellular paclitaxel-tubulin binding in dorsal root ganglia, peripheral nerves, and brain. The framework was translated to humans, where sNfL dynamics were related to the Total Neuropathy Score-clinical version in breast cancer patients. sNfL showed moderate predictive performance for CIPN, highlighting the multifactorial nature of PIPN.

The thesis was subsequently extended to cabazitaxel, docetaxel, vinorelbine, and eribulin to investigate whether differences in CIPN prevalence could be related to MTA distribution at CIPN-relevant sites. Despite differences in the time required to reach steady-state, governed by tubulin abundance and transporter-mediated distribution across endothelial and parenchymal barriers, all MTAs showed a similar distribution pattern, with greater distribution into PNS than CNS tissues. Thus, dissimilarities in clinical unbound plasma exposure, rather than in tissue distribution patterns, may be important determinants of intracellular exposure and neurotoxicity.

Together, the translational framework developed in this thesis links drug exposure, intracellular target engagement, and sNfL dynamics from preclinical to clinical settings. It provides a rationale for PK-informed CIPN assessment and for the development of novel MTAs with reduced neurotoxicity.