Characterization of oxaliplatin-induced chronic painful peripheral neuropathy in the rat and comparison with the neuropathy induced by paclitaxel

Abstract

Anti-neoplastic agents in the platinum-complex, taxane, vinca alkaloid, and proteasome-inhibitor classes induce a dose-limiting, chronic, distal, symmetrical, sensory peripheral neuropathy that is often accompanied by neuropathic pain. Clinical descriptions suggest that these conditions are very similar, but clinical data are insufficient to determine the degree of similarity and to determine if they share common pathophysiological mechanisms. Animal models do not have the limitations of clinical studies and so we have characterized a rat model of chronic painful peripheral neuropathy induced by a platinum-complex agent, oxaliplatin, in order to compare it with a previously characterized model of chronic painful peripheral neuropathy induced by a taxane agent, paclitaxel. The oxaliplatin model evokes mechano-allodynia, mechano-hyperalgesia, and cold-allodynia that have a delayed onset, gradually increasing severity, a distinct delay to peak severity, and duration of about 2.5 months. There is no effect on heat sensitivity. Electron microscopy (EM) analyses found no evidence for axonal degeneration in peripheral nerve, and there is no upregulation of activating transcription factor-3 in the lumbar dorsal root ganglia. There is a statistically significant loss of intraepidermal nerve fibers in the plantar hind paw skin. Oxaliplatin treatment causes a significant increase in the incidence of swollen and vacuolated mitochondria in peripheral nerve axons, but not in their Schwann cells. Nerve conduction studies found significant slowing of sensory axons, but no change in motor axons. Single fiber recordings found an abnormal incidence of A- and C-fibers with irregular, low-frequency spontaneous discharge. Prophylactic dosing with two drugs that are known to protect mitochondria, acetyl-l-carnitine and olesoxime, significantly reduced the development of pain hypersensitivity. Our results are very similar to those obtained previously with paclitaxel, and support the hypothesis that these two agents, and perhaps other chemotherapeutics, produce very similar conditions because they have a mitotoxic effect on primary afferent neurons.

Fig. 1

Body weights for vehicle-treated and oxaliplatin-treated rats used in the behavioral time-course assays for mechano-allodynia and mechano-hyperalgesia (see next figure). Mean ± SEM (error bars are smaller than the symbols); n = 12/group. BL: Baseline weight on the day of the first injection. The difference between groups is statistically significant from the day of the 4^th injection onwards (two-way ANOVA followed by Bonferroni-corrected t-tests for each time point).

Fig. 2

Time courses for oxaliplatin-evoked (A) mechano-allodynia, (B) mechano-hyperalgesia, (C) cold-allodynia, and (D) heat sensitivity. A and B are the same animals (n = 8/group), C and D are separate groups (n = 12 and n = 9, respectively). The last oxaliplatin injection was on D4. Note the delays to peak pain severity for mechano-allodynia, mechano-hyperalgesia, and cold-allodynia. Oxaliplatin treatment had no effect on heat pain thresholds. BL: Baseline test before the first injection. Means ± SEM. *, ** p < 0.05, < 0.01 (repeated measures ANOVA with Dunnett’s test for post hoc pair-wise comparisons to pre-injection baseline). There were no significant variations over time for the vehicle-treated groups in A, B, and D. However, the vehicle-treated group in C had a small but statistically significant increase in sensitivity to cold; the reason for this is not known. Nevertheless, the oxaliplatin-treated group was significantly more sensitive in comparison to the vehicle-treated group (Bonferroni-corrected t-tests for all time points).

Fig. 3

Anatomy. Counts of peripheral nerve A-fibers (A) and C-fibers (B) in vehicle-treated and oxaliplatin-treated rats (n = 4/group) sacrificed on D35. Means ± SEM. The between group differences are not statistically significant (t-tests). (C) Counts of intraepidermal nerve fibers (IENF) per cm of epidermal border in vehicle-treated and oxaliplatin-treated rats. Oxaliplatin treatment evoked a significant loss of IENFs. * p < 0.05 (t-test). Inset: PGP9.5-stained IENFs in the skin of the plantar hind paw of an oxaliplatin-treated rat. IENFs arise from subepidermal nerve fascicles, cross the epidermal basal lamina, and form the afferent’s terminal receptor arbor by branching within the epidermis and issuing en passant and terminal receptor boutons. (D) Sections from the L5 DRG from a rat with an ipsilateral sciatic nerve transection (left) and from an oxaliplatin-treated rat (right). ATF-3-positive nuclei are stained red, Nissl substance is stained green. Scale bar = 30 μm. No ATF-3-positive DRG cells were found in oxaliplatin-treated rats.

Fig. 4

Atypical (swollen and vacuolated) mitochondria in A-fibers and C-fibers and their respective Schwann cells in vehicle-treated and oxaliplatin-treated rats. (A) Portion of a myelinated axon from an oxaliplatin-treated rat. As shown here, axons usually had a mixture of normal (barred arrows) and swollen and vacuolated (plain arrows) mitochondria. (B) Portion of a Remak bundle from an oxaliplatin-treated rat showing C-fibers with normal (barred arrows) and swollen and vacuolated (plain arrows) mitochondria. The uppermost barred arrow points to a normal mitochondrion in the Schwann cell’s cytoplasm. Scale bar = 0.5 μm. (C) The percentage (means ± SEM) of swollen and vacuolated mitochondria in A-fiber and C-fiber axons in oxaliplatin-treated rats was significantly greater than in vehicle-treated rats. ** p < 0.01 (t-tests). (D) The percentage (means ± SEM) of swollen and vacuolated mitochondria in the cytoplasm of myelinating and non-myelinating (i.e., Remak bundle) Schwann cells in vehicle-treated and oxaliplatin-treated rats were low and not significantly different (t-tests).

Fig. 5

Effects of acetyl-L-carnitine (ALCAR) on oxaliplatin-evoked painful peripheral neuropathy. The vehicle-treated group developed the expected statistically significant mechano-allodynia (A) and mechano-hyperalgesia (B). ALCAR treatment significantly reduced the severity of mechano-allodynia and mechano-hyperalgesia and the effect persisted for at least three weeks after the last ALCAR injection. Right: AUC: area-under-the-curve values. ** p < 0.001 (t-test).

Fig. 6

Effects of olesoxime on oxaliplatin-evoked painful peripheral neuropathy. The vehicle-treated group developed the expected statistically significant mechano-allodynia (A) and mechano-hyperalgesia (B). Olesoxime treatment significantly reduced the severity of mechano-allodynia and mechano-hyperalgesia and the effect persisted for at least three weeks after the last olesoxime injection. Right: AUC area-under-the-curve values. ** p < 0.001 (t-test).