Translation of paclitaxel-induced peripheral neurotoxicity from mice to patients: the importance of model selection.

Abstract

Paclitaxel-induced peripheral neurotoxicity (PIPN) is a potentially dose-limiting side effect in anticancer chemotherapy. Several animal models of PIPN exist, but their results are sometimes difficult to be translated into the clinical setting. We compared 2 widely used PIPN models characterized by marked differences in their methodologies. Female C57BL/6JOlaHsd mice were used, and they received only paclitaxel vehicle (n = 38) or paclitaxel via intravenous injection (n = 19, 70 mg/kg) once a week for 4 weeks (Study 1) or intraperitoneally (n = 19, 10 mg/kg) every 2 days for 7 times (Study 2). At the end of treatment and in the follow-up, mice underwent behavioral and neurophysiological assessments of PIPN. At the same time points, some mice were killed and dorsal root ganglia, skin, and sciatic and caudal nerve samples underwent pathological examination. Serum neurofilament light levels were also measured. The differences in the neurotoxicity parameters were analyzed using a nonparametric Mann-Whitney test, with significance level set at P < 0.05. Study 1 showed significant and consistent behavioral, neurophysiological, pathological, and serological changes induced by paclitaxel administration at the end of treatment, and most of these changes were still evident in the follow-up period. By contrast, study 2 evidenced only a transient small fiber neuropathy, associated with neuropathic pain. Our comparative study clearly distinguished a PIPN model recapitulating all the clinical features of the human condition and a model showing only small fiber neuropathy with neuropathic pain induced by paclitaxel.