The effects of thalidomide and minocycline on taxol-induced hyperalgesia in rats

1. Introduction

Chemotherapy-induced pain is the most common treatment-limiting complication encountered by cancer patients receiving taxane-, vinca alkaloid-, or platin-based chemotherapy (Forman, 1990; Forsyth et al, 1997; Quasthoff and Hartung, 2002; Verstappen et al, 2003). Symptoms occur in a stocking and glove distribution often early in treatment and the final overall incidence and severity is directly related to the initial and cumulative drug doses, the co-administration of other chemotherapeutic agents and the presence of other diseases like diabetes or alcoholism (Chaudhry et al, 2003; Connelly et al, 1996; New et al, 1996; Rowinsky et al, 1993; Rowinsky and Donehower, 1995; Tuxen and Hansen, 1994). Chemotherapy-induced pain improves or completely resolves in 51% of patients, but becomes chronic in the remainder, affecting the quality of life and return to productivity in cancer survivors (Connelly et al, 1996). Psychophysical tests in patients with taxol-induced pain reveal marked impairment of Aβ function, less severe impairment of Aδ function, but preservation of C-fiber function and paradoxical burning pain to skin cooling (Dougherty et al, 2004; Dougherty et al, 2006).

The mechanism by which chemotherapy drugs produce pain remains unclear. Hypothetically, drugs like taxol or vincristine might damage nerves through their antineoplastic mechanism, binding to microtubules in axons, presumably altering axonal transport and subsequently inducing neuropathology (Tanner et al, 1998). However, no direct evidence supports this contention and such a mechanism is difficult to reconcile with the preferential impairment of nerve fiber function shown in psychophysical tests. Changes in spinal neuron processing of sensory information in chemotherapy-treated rats, cells that are protected from chemotherapy drugs by the blood brain barrier, suggests that other effects of these drugs might contribute to the generation of pain (Weng et al, 2003). For example, paclitaxel up-regulates multiple transcription factors (Zaks-Zilberman et al, 2001) enzymes, and cytokines, such as nuclear factor kappa-B (NF-κB) and tumor necrosis factor-alpha (TNF-α) that contribute to both neuropathic and inflammatory pain (Bao et al, 2001; Empl et al, 2001; Manni and Aloe, 1998; Opree and Kress, 2000; Sommer et al, 1998b). These effects suggest the possible involvement of alterations in glial functions in taxol pain like previously demonstrated for vincristine-induced pain (Sweitzer et al, 2006).

Two drugs, minocycline and thalidomide, each of which have profound immunomodulatory effects (Raghavendra et al, 2003) are evaluated for efficacy in protecting rats from taxol-induced pain in this study. Each drug has a well known primary pharmacologic effect, minocycline is an antimicrobial agent structurally related to tetracycline (Kriz et al, 2002; Wells et al, 2003; Yrjanheikki et al, 1999), and thalidomide was introduced as a sedative-hypnotic (Calabrese and Fleischer, 2000). However, minocycline also decreases the release of multiple pro-inflammatory cytokines (Amin et al, 1996; Kriz et al, 2002; Wells et al, 2003; Yrjanheikki et al, 1999), and thalidomide both suppresses the synthesis and release of pro-inflammatory cytokines as well as increases the release of anti-inflammatory cytokines (Ribeiro et al, 2000; Sommer et al, 1998a; Sommer et al, 2001b; Wagner et al, 1998). Each have previously also shown efficacy in the treatment of neuropathic pain in rats and/or mice (Bennett, 1999; Ledeboer et al, 2005; Narita et al, 2006).

2. Results

A total of 50 rats were used in the study. Among these, the mortality rate with taxol chemotherapy was zero. Moreover, none of the animals showed any failure to groom and all gained weight, albeit some more slowly than in untreated rats (see below). No animals showed any gross neurological abnormalities or any sign of self-mutilating behavior or autotomy. Animals were divided into five treatment groups, vehicle (n= 10); taxol (n=8); thalidomide (n=8); taxol-thalidomide (n=8); minocycline (n=8); and taxol-minocycline (n=8).

3. Discussion

Thalidomide is shown here to reduce taxol-induced mechanical hyperalgesia whereas minocycline reduces taxol-induced mechanical as well as thermal hyperalgesia. Thalidomide is currently used due to its immunomodulatory properties in the treatment of several diseases including erythemia leprosun nodosum, Bechet’s syndrome, aphtous stomatitis and graft-versus-host disease (Calabrese and Fleischer, 2000). The exact mechanism of action is not clear, but modulation of the production and release of TNF-α is thought to be especially important. Thalidomide reduces synthesis of TNF-α, IL-12 and IL-8 in culture of LPS-stimulated human monocytes and microglial cells (Lokensgard et al, 2000; Moller et al, 1997; Sampaio et al, 1991; Sampaio et al, 1993). Thalidomide also accelerates the degradation of pro-inflammatory cytokines (Moreira et al, 1993); and dampens their intracellular signaling by inhibiting the translocation of NF-κB into the nucleus (Keifer et al, 2001). These alterations of cytokine trafficking and intracellular signaling result in altered cellular immunity such that patterns of lymphocyte activity shift from Th1 to Th2 dominant, levels of IL-4 increase in the short term and those of IFN-γ decrease over the long term (Mchugh et al, 2003). Thalidomide also alters the expression of adhesion molecules (ICAM and L-selectin) in endothelial cells and in polymorphonuclear leucocytes, thus influencing tissue migration of inflammatory cells, and reduces phagocytic activity in monocytes (Barnhill et al, 1984; Geitz et al, 1996).

Minocycline is an antibiotic structurally related to tetracycline that has neuroprotective pharmacologic actions in several models of nervous system injury in addition to its antimicrobial activity. Minocycline improves locomotor rating and histopathology scores in mice with spinal cord lesion compared to non-treated controls (Wells et al, 2003) and reduces glutamate and kainate-induced exitotoxicity (Tikka et al, 2001). Like thalidomide, minocycline also inhibits LPS-induction of TNF-α in vivo and the induction by LPS of both TNF-α and IL-1β from human monocytes in vitro (Shapira et al, 1996). Finally, minocycline improves survival rate, delays onset of muscle weakness and decreases microglia activation in mice with experimentally-induced amyotrophic lateral sclerosis (Kriz et al, 2002).

Several lines of evidence suggest that the immunomodulatory effects of thalidomide and minocycline underlie their protective effects on taxol-induced neuropathic pain shown here. Tumor necrosis factor-alpha (TNF-α) and other functionally-related pro-inflammatory cytokines such as interleukin-1 beta (IL-1β) and interleukin-6 (IL-6) are becoming well recognized as a key mediators in the pathogenesis of many types of neuropathic pain (Bennett, 1999; Bennett, 2000; DeLeo and Yezierski, 2001). TNF-α, IL-1β, and IL-6 are increased in the nerves of rats with chronic constriction injury (CCI) of the sciatic nerve (Okamoto et al, 2001; Wagner and Myers, 1996a); and administration of anti-TNF or anti-IL-6 antibodies reduce the mechanical allodynia and thermal hyperalgesia of CCI neuropathy (Arruda et al, 2000; Schafers et al, 2001; Sommer et al, 1998b; Sommer et al, 2001a). Antagonists to TNF-α, IL-6 or IL-1 also prevent and reverse hyperalgesia in rats with sciatic inflammatory neuropathy (Milligan et al, 2003); and experimental autoimmune neuritis is similarly associated with TNF-α positive macrophages infiltrating nerve sheaths and again anti-TNF-α antibodies are protective in this model of autoimmune demyelination (Stoll et al, 1993). Injection of TNF-α around or into peripheral nerves or injection of either TNF-α, IL-1β or Il-8 into the hindpaws produces hyperalgesia in awake behaving rats (Ferreira et al, 1988; Sachs et al, 2002; Sorkin and Doom, 2000; Wagner and Myers, 1996a) and also produce increases in evoked-activity, decreases in mechanical activation thresholds, and increased spontaneous ongoing activity in Aδ and C fibers (Fukuoka et al, 1994; Ozaktay et al, 2002; Sorkin et al, 1997). Thalidomide reverses and prevents the hyperalgesia in CCI and spinal nerve ligation neuropathies, respectively (Bennett, 1999; Bennett, 2000), an effect shown in CCI that correlates to decreases in TNF-α levels in sciatic nerves (George et al, 2000).

TNF-α and pro-inflammatory cytokines may similarly mediate taxol-induced peripheral neuropathy and hyperalgesia. Taxol has LPS-mimetic activity inducing the synthesis of several proinflammatory cytokines and chemokines, including both TNF-α and IL-1β, in human monocytes, T-lymphocytes and breast cancer cell lines and in murine macrophages and breast cancer cell lines (Ding et al, 2002; Manthey et al, 1992; Moos et al, 1999; White et al, 1998; Zaks-Zilberman et al, 2001). The increased synthesis and release of TNF-α and IL-1β is independent of the effects of taxol on microtubules (Moos and Fitzpatrick, 1998), but instead is linked in rodents to an up-regluation of activity of NF-κB and microtubule-associated protein kinases that in turn is driven by activation of the toll-like 4 receptor (Byrd-Leifer et al, 2001). In humans, taxol may require an additional immune stimulus to evoked elevated levels of TNF-α and IL-1β(Allen et al, 1993), and does not appear to be linked via the toll-4 receptor (Byrd-Leifer et al, 2001; Kawasaki et al, 2001). Nevertheless, this particular species difference likely has little overall functional significance for cytokine induction during chemotherapy with taxol in humans given the abundance of antigenic stimuli present in cancer patients. More directly, minocycline reduces pro-inflammatory cytokine responses and hyperalgesia in sciatic nerve inflammation hyperalgesia (Ledeboer et al, 2005) and pro-inflammatory cascades induced by taxol treatment can be suppressed using interleukin-10 gene therapy. This manipulation results in loss of chemotherapy-induced hyperalgesia (Ledeboer et al, 2007). Finally, the observation that propentofylline, a glial modulating agent, reduces vincristine-induced pain, suggests this cell group as important in generating the pro-inflammatory cytokine responses (Sweitzer et al, 2006).

Still to be reconciled with the idea that taxol-induced neuropathic pain is caused by the induction of pro-inflammatory cytokines around nerve endings is the long-standing view that chemotherapy-induced neuropathy is caused by the well known anti-microtubule effects of these drugs. The thinking has been that chemotherapy drugs like taxol cause alterations in microtubule organization and cytoskeletal ultrastructure of axons that leads to dysfunction and hence pain (Bradley et al, 1970; Green et al, 1977; Shelanski and Wisniewski, 1969; Tanner et al, 1998; Theiss and Meller, 2000). Sciatic nerves of rats treated with very high doses of taxol (25 mg/kg) do indeed show signs of mild axonal degeneration and activated Schwann cells (Cavaletti et al, 2000); and similarly nerves in rats treated with a somewhat lower dose of taxol (15 mg/kg) also show decreased mean axonal caliber and degenerated myelin profiles in dorsal but not ventral roots (Cliffer et al, 1998). Nerve histopathology in a patient with severe taxol-induced peripheral neuropathy that occurred following 17 cycles (6603 mg) of treatment similarly showed axonal degeneration, a diminished number of myelinated fibers and Schwann cells with myelin degradation products (Sahenk et al, 1994). However, histopathology is not prerequisite for the induction of pain by chemotherapy. For example, nerves from rats with chemotherapy-induced hyperalgesia following less aggressive taxol treatment (1 mg/kg/day × 3) only showed endoneurial edema without signs of overt damage to either myelinated or unmyelinated axons (Polomano et al, 2001). Moreover, histopathology can be produced by administration of many compounds around nerves and thus does not necessarily imply that microtubules and ultrastructure was damaged by taxol. For example, in addition to hyperalgesia, intra- or epi-neural injection of TNF-α also induces leukocyte margination and migration in endothelial cells and increased number of macrophages in both the epi- and endoneurium followed by demyelination and axonal degeneration like observed with CCI and focally induced sciatic neuritis (Bennett, 1999; Bennett, 2000; Redford et al, 1995; Wagner and Myers, 1996a). Given that Schwann cells are a source of TNF-α as well as Il-1β(Skundric D S et al, 2002; Wagner and Myers, 1996b) we propose that exposure of these cells to either pro-inflammatory cytokines released from tissue macrophages or possibly to taxol itself promotes cytokine release and subsequently pain and neuropathology.

Finally, it must be also be considered that pharmacological properties of thalidomide and minocycline aside from their immunomodulatory capacity could explain the effects observed here on taxol-evoked hyperalgesia. Thalidomide first received notoriety as a nearly ideal sedative hypnotic that produced comfortable sleep without producing hangover effects {Lasagna, 1960 19821 /id}. This effect was suggested as mediated via interruption or modulation of serotonin’s actions within the CNS {Youdim, 1985 19820 /id}.

In conclusion, thalidomide and minocycline ameliorate pain-related behavior normally seen in the animals following treatment with taxol. We hypothesize that taxol causes hyperalgesia by its ability to alter the synthesis of the proinflammatory cytokines that excite sensory nerves.

4. Materials and Methods

Thirty-four male Sprague-Dawley rats were used. They were housed singly on a 12 / 12 h light/dark cycle, at a room temperature of 22° C with free access to food and water. Rats were tested by the same experimenter throughout who was blinded to drug treatment. Each rat was followed and tested for a length of 60 days.