Polyneuropathy is a frequent and potentially severe side effect of clinical tumor chemotherapy. The goal of this study was to characterize paclitaxel-, cisplatin-, vincristine- and bortezomib-induced neuropathy in C57BL/6 mice with a comparative approach. The phenotype of the animals was evaluated at four time points with behavioral and electrophysiological tests, followed by histology. Treatment protocols used in this study were well tolerated and induced a sensory and predominantly axonal polyneuropathy. Behavioral testing revealed normal motor coordination, whereas all mice receiving verum treatment developed mechanical allodynia and distinct gait alterations. Electrophysiological evaluation showed a significant decrease of the caudal sensory nerve action potential amplitude for all cytostatic agents and a moderate reduction of nerve conduction velocity for cisplatin and paclitaxel. This finding was confirmed by histological analysis of the sciatic nerve which showed predominantly axonal damage: Paclitaxel and vincristine affected mostly large myelinated fibers, bortezomib small myelinated fibers and cisplatin damaged all types of myelinated fibers to a similar degree. Neuropathic symptoms developed faster in paclitaxel and vincristine treated animals compared to cisplatin and bortezomib treatment. The animal models in this study can be used to elucidate pathomechanisms underlying chemotherapy-induced polyneuropathy and for the development of novel therapeutic and preventative strategies.

Chemotherapy-induced central nervous system (CNS) neurotoxicity presents an unmet medical need. Patients often report a cognitive decline in temporal correlation to chemotherapy, particularly for hippocampus-dependent verbal and visuo-spatial abilities. We treated adult C57Bl/6 mice with 12 × 20 mg kg−1 paclitaxel (PTX), mimicking clinical conditions of dose-dense chemotherapy, followed by a pulse of bromodesoxyuridine (BrdU) to label dividing cells. In this model, mice developed visuo-spatial memory impairments, and we measured peak PTX concentrations in the hippocampus of 230 nm l−1, which was sevenfold higher compared with the neocortex. Histologic analysis revealed a reduced hippocampal cell proliferation. In vitro, we observed severe toxicity in slowly proliferating neural stem cells (NSC) as well as human neuronal progenitor cells after 2 h exposure to low nanomolar concentrations of PTX. In comparison, mature post-mitotic hippocampal neurons and cell lines of malignant cells were less vulnerable. In PTX-treated NSC, we observed an increase of intracellular calcium levels, as well as an increased activity of calpain- and caspase 3/7, suggesting a calcium-dependent mechanism. This cell death pathway could be specifically inhibited with lithium, but not glycogen synthase kinase 3 inhibitors, which protected NSC in vitro. In vivo, preemptive treatment of mice with lithium prevented PTX-induced memory deficits and abnormal adult hippocampal neurogenesis. In summary, we identified a molecular pathomechanism, which invokes PTX-induced cytotoxicity in NSC independent of cell cycle status. This pathway could be pharmacologically inhibited with lithium without impairing paclitaxel’s tubulin-dependent cytostatic mode of action, enabling a potential translational clinical approach.