Chapter 72 Emerging Therapies for Spinal Cord Injury
The reversal of paralysis following spinal cord injury (SCI) represents one of the greatest challenges in all of neuroscience. Despite significant improvements in the early medical and surgical management of SCI, along with a greatly improved understanding of SCI pathophysiology, there remains no effective treatment to improve neurologic outcomes following SCI. SCI destroys neuronal connectivity by severing descending motor and ascending sensory axonal pathways. Such damage often results in permanent paralysis and loss of sensation below the level of injury. Although SCI causes loss of neurons and glial cells at the lesion site, functional deficits result primarily from loss of white matter axons by direct trauma and from progressive damage to initially intact axons by complex secondary injury mechanisms.1 To date, numerous potential therapies have been investigated in prospective, randomized, controlled clinical trials, yet all have failed to demonstrate neurologic benefit despite evidence that these agents demonstrated benefit in preclinical animal studies of SCI. This chapter will review previous clinical trials as well as major recent preclinical advances, in pharmacologic and cell-based therapeutic approaches to SCI.
Spinal Cord Injury Epidemiology
SCI is associated with severe physical, psychological, social, and economic burdens on patients and their families. With an annual incidence rate of 15 to 40 persons per million, it has been estimated that at least 10,000 North Americans will suffer an SCI each year.2 With the average acute-care and rehabilitation charges of approximately $60,000 for each of these phases of care, the societal expense associated with the medical management, as well as lost earnings of this disorder over the course of one’s life, is significant.3 It has been estimated that the lifetime cost of medical care and other injury-related expenses for a 25-year-old patient with SCI who suffers high cervical quadriplegia is approximately $3 million.4 SCI occurs predominantly in young, otherwise healthy individuals, with injury occurring with the greatest frequency in those between 15 and 25 years of age; the male-to-female ratio for SCI is approximately 4 to 1.5 Common causes of SCI in the United States are motor vehicle accidents (50%), falls and work-related injuries (30%), violent crime (11%), and sports-related injuries (9%).6 Injuries most commonly occur in the cervical spine and are associated with the most devastating neurologic impairments (e.g., quadriplegia). A recent report from the U.S. National Spinal Cord Injury Database found that 56% of all SCI cases occur in the cervical spine.
Pathophysiology of Spinal Cord Injury
A detailed understanding of the pathophysiologic processes that occur following SCI is paramount to the development of effective therapies for SCI. The pathophysiology of SCI is best described as biphasic, consisting of a primary and a secondary phase of injury.7 The primary phase involves the initial, immediate mechanical injury during which failure of the spinal column occurs, and includes compression, contusion, shear, or laceration due to penetrating injury and acute stretching of the spinal cord as a result of vertebral distraction or sudden acceleration-deceleration of the spinal column. The most common underlying primary injury mechanism results from the acute compression and contusion of the spinal cord due to bone or disc displacement within the spinal column during fracture-dislocation or burst fracture of the spine.2 The primary mechanical trauma to the spinal cord, along with subsequent persistent compression, triggers a complex and delayed pathologic cascade, termed the secondary injury phase, which involves vascular dysfunction, edema, ischemia, excitotoxicity, electrolyte shifts, free radical production, inflammation, and delayed apoptotic cell death. Although neurologic deficits are present immediately following the initial injury, the secondary injury phase results in a protracted period of neuroinflammation and tissue destruction. The spatial extent of secondary injury events spreads both radially and longitudinally along the spinal cord in a rostral-to-caudal manner, resulting in neuronal and glial cell death. The end result is cavitation of central gray matter along with partial or complete loss of adjacent white matter tracts.
The pathophysiology of SCI has been extensively studied in many experimental animal models, including mice, rats, cats, and primates, in order to understand the cellular and molecular mechanisms of tissue damage and the consequent disability.2 Traumatic SCI initiates a series of destructive cellular inflammatory processes which accentuate tissue damage at and beyond the original site of trauma, and cystic cavitation inevitably occurs due to these secondary injury events. Contusion SCI in animals produces a predictable pattern of progressive injury resulting in neuronal and glial cell death, vascular injury, axonal destruction, and demyelination that is analogous to the pattern in human spinal cord contusion injury, the most common type of the human SCI (Fig. 72-1).7,8 The progressive expansion of the injury from gray to white matter causes secondary damage to initially intact axons within hours to several weeks after injury. The cellular inflammatory response, predominated by macrophages, has been implicated as the primary mediator of progressive secondary injury. The major targets for spinal cord repair include severed and/or demyelinated axons, inflammatory cells and proinflammatory cytokines, and glial scar components.
Emerging Preclinical Spinal Cord Injuries Therapies
The failure of axonal regeneration after CNS injury is the result of two distinct processes: the limited intrinsic regenerative potential of CNS neurons and the inhibitory extrinsic environment of the injured CNS. However, CNS neurons may regenerate when provided with a “permissive” substrate that promotes axonal growth. The ability of CNS neurons to regenerate over long distances after axotomy was first established through pioneering work in which injured CNS axons were found to regenerate into and through growth-permissive peripheral nerve (PN) grafts.9 The therapeutic potential of PN grafts for use in SCI was then demonstrated by Cheng et al. in 1996,10 who reported that transplantation of multiple PN grafts stabilized with fibrin glue containing acidic fibroblast growth factor into a rat spinal cord transection model promoted hindlimb functional recovery.
Experimental therapies for SCI can be grouped into several subtypes, based on their general mechanism of action. These general treatment subcategories include therapies that (1) promote neuroprotection, (2) stimulate intrinsic axonal regrowth, (3) enhance remyelination, and (4) remove or block inhibitory molecules within damaged myelin and within the astroglial scar. To date, researchers have identified and characterized a number of molecular signals associated with axonal growth following injury and have developed drug therapies that are capable of stimulating axonal regeneration. Several regeneration-associated genes have been found to be up-regulated after axonal injury, including L1, c-fos, and c-jun, and the 43-kD growth-associated protein (GAP43). Despite these findings, the degree and extent of their up-regulation is apparently insufficient to promote a strong regenerative response in the CNS. Numerous reports have implicated a decrease in the levels of intracellular cAMP after CNS injury as a major cause of intrinsic regenerative failure. The elevation of intracellular cAMP levels by cAMP analogues and/or the phosphodiesterase inhibitor, rolipram, has been shown to increase axonal sprouting and reduce the effects of myelin-associated inhibitors.11,12
Multiple inhibitory molecules exist that make the injured CNS a nonpermissive environment for axonal growth. These molecules may be categorically divided into myelin-associated inhibitors and inhibitors associated with the astrocytic glial scar. Recently, major advances in molecular neuroscience have led to the identification of several growth-inhibitory molecules associated with myelin (Nogo, myelin-associated glycoprotein, and oligodendrocyte myelin glycoprotein) and the glial scar (chondroitin sulfate proteoglycans, CSPGs), as well as their receptors and downstream signaling pathways, which make up the inhibitory extrinsic environment of the injured CNS. This chapter will review several recently developed targeted molecular strategies that enable axons to overcome inhibitory influences of myelin-associated proteins and astroglial scar components.
Considering the multifactorial nature of secondary pathologic events attributed to SCI, drug cocktails with multifaceted modes of neuroprotection and axonal growth promotion will likely be useful in preventing or limiting secondary injury progression. Because of the numerous complex pathophysiologic changes that occur after SCI, successful repair may involve strategies that combine the use of neuroprotective drug therapies with those that promote axonal regeneration, such as cell transplantation, genetic engineering to increase neurotrophic factors, neutralization of inhibitory factors, limiting glial scar formation, and neurorehabilitation. Furthermore, neuroprotective drug therapies that function at later time points following SCI will likely have greater clinical relevance than those with small therapeutic windows. There is also a requirement to establish methods for the systematic evaluation of preclinical therapeutic outcomes from the various research SCI centers. In the evaluation of future potential SCI therapies, a clinical basis for the implementation of novel therapies will be required. Refined clinical outcome measures as well as neurophysiologic measures are needed for a precise qualitative and quantitative assessment of spinal cord function in SCI patients, especially at an early stage after injury. This will allow for better detection of improvements in functional recovery and will enable clinicians to precisely monitor the effects of novel therapies.3,13
Neuroprotective Strategies
Methylprednisolone
The most extensively investigated pharmacologic intervention for SCI is the glucocorticoid methylprednisolone sodium succinate (MPSS, Solu-Medrol). To date, three major clinical trials evaluating the use of MPSS in acute SCI have been performed. The National Acute Spinal Cord Injury Study (NASCIS I) compared the efficacy of administering an intravenous (IV) loading dose of MPSS (100 or 1000 mg) and a dose of either 25 or 250 mg every 6 hours thereafter for 10 days to SCI patients who had motor or sensory deficits below the level of injury.14 This study demonstrated that higher-dose MPSS was no more effective than the lower dose in reducing neurologic deficits between the two groups at 6 weeks, 6 months, and 1 year after injury. Although a placebo group was not included in the NASCIS I because of ethical concerns, the use of MPSS became much less prevalent after this trial.
Subsequent animal studies demonstrated that the effective neuroprotective dose of MPSS was much higher than that used in NASCIS I.15 Therefore, following the NASCIS I study, the NASCIS II trial was conducted to compare the efficacy of high-dose MPSS administered as a 30-mg/kg bolus over the first hour, followed by an infusion of 5.4 mg/kg/hour over the next 23 hours, to that of placebo and that of the opioid receptor antagonist naloxone.16,17 The trial was also designed to examine the effects of different timing schedules of drug administration following SCI. An analysis of all patients entered into the trial failed to demonstrate a significant difference in motor and sensory function among the treatments. However, upon stratification of the data on the basis of time to loading dose (<8 hours or >8 hours from SCI) and adjustment for the severity of SCI (complete vs. incomplete), analysis of the results from the NASCIS II study at the 6-week and 6-month follow-up revealed that patients who were treated with MPSS within the first 8 hours of injury had significantly improved motor and sensory function, compared to patients who received placebo, naloxone, or MPSS at later times. Furthermore, these differences remained significant 1 year after injury.18 None of the differences in patients taking naloxone or in patients treated with MPSS more than 8 hours after SCI were statistically significant.
Despite the beneficial therapeutic effects demonstrated with MPSS treatment, the results from the NASCIS II trial have not been universally accepted.19,20 Concerns about the small sample population size for the groups showing beneficial effects, nonstandardized performance of medical and surgical protocols by participating centers, and lack of defined functional outcome measures to assess whether the improvements seen with MPSS treatments were correlated with clinical significance have reduced enthusiasm for the indication of MPSS for acute SCI. Additional criticisms remained because of concerns about the adverse effects of steroids and the small therapeutic effects that were noted over the short (6-month) follow-up period used as the basis for conclusions about the trial, as well as the differential loss to follow-up between MPSS-treated and untreated patients. The NASCIS III trial resolved some of these issues with the inclusion of a functional independence measure.21 All patients in this study received an initial MPSS bolus of 30 mg/kg before being randomized to one of three treatment arms: (1) MPSS infusion of 5.4 mg/kg/hour over 24 hours, (2) MPSS infusion of 5.4 mg/kg/hour over 48 hours, or (3) tirilizad mesylate (TM, a lazaroid with inhibitory effects on lipid peroxidation without glucocorticoid side effects) 2.5-mg/kg bolus every 6 hours for 48 hours. The results from this third study concluded that no benefit was associated with extending MPSS treatment beyond 24 hours if MPSS had been administered within the first 3 hours of SCI, while patients who started MPSS therapy at 3 to 8 hours demonstrated improvements in motor capabilities if drug infusion was continued for 48 hours in comparison to 24 hours of infusion.
Research on the neuroprotective mechanism of MPSS in experimental SCI models has demonstrated that in addition to possessing antioxidant properties, MPSS reduces tumor necrosis factor–α (TNF-α) protein synthesis and nuclear factor kappa B (NF-κB) activity, effects that may account for the anti-inflammatory actions exerted by MPSS.22 However, MPSS treatment is not without adverse effects; high-dose MPSS has been associated with an increased prevalence of wound infections, pneumonia, sepsis, and death due to respiratory complications.16,21 Furthermore, the anti-inflammatory actions of high-dose MPSS may also have detrimental effects on neuronal regeneration and axonal sprouting while exacerbating postischemic necrosis. These adverse effects may explain the ineffectiveness of MPSS treatment started beyond 8 hours post-SCI observed in the NASCIS II study. The guidelines committee of the American Association of Neurological Surgeons and Congress of Neurological Surgeons Joint Section on Disorders of the Spine and Peripheral Nerves, upon reviewing the evidence regarding the use of MPSS in the treatment of acute SCI in adults, concluded that its use could be supported only at the level of a treatment option (Joint Section on Disorders of the Spine and Peripheral Nerves of the AANS/CNS, 2002). However, criticisms directed toward the NASCIS II and III trials must be balanced by the current lack of alternative neuroprotective strategies for acute SCI. Therefore, MPSS administration may remain justified for acute SCI (within 8 hours) in nondiabetic and nonimmunocompromised patients, given the severity of SCI deficits and current lack of alternatives. Although many neurosurgeons continue to administer steroids, the rationale for this administration appears to have changed over recent years, according to a survey published in 2006, which revealed that the majority of respondents continue to administer methylprednisolone but that this is motivated predominantly by fear of litigation.23
21-Aminosteroids
The 21-aminosteroids (lazaroids) are synthetic glucocorticoid analogues that are capable of inhibiting lipid peroxidation without activating glucocorticoid receptors. The 21-aminosteroid TM utilizes three neuroprotective mechanisms: antioxidation, preservation of endogenous vitamin E, and membrane stabilization through the inhibition of lipid peroxidation.15 TM was administered in comparison to MPSS in the NASCIS III clinical trial and was shown to be as effective as the 24-hour MPSS regimen.21 However, because TM was not shown to have superior efficacy in comparison to MPSS, and given the absence of proven benefit over placebo controls, 21-aminosteroids have not been adopted for neuroprotection in clinical SCI.
GM-1 Ganglioside
The gangliosides are a group of sialic acid–containing glycosphingolipids located in high concentrations in the outer membranes of nervous tissue. A small randomized, placebo-controlled, double-blind trial, the Maryland monosialotetrahexosylganglioside (GM-1) Ganglioside Study, was conducted to investigate the efficacy of GM-1 in patients with cervical and thoracic SCIs.24 The test drug protocol consisted of either 100 mg of GM-1 or placebo control administered once daily via IV infusion for a total of 18 to 32 doses, the first dose beginning within 72 hours of the onset of SCI. Results from this study demonstrated that GM-1 improved the motor and sensory recovery of lower extremities only, suggesting that GM-1 had an ability to enhance the function of axons traversing the site of injury but had no effect on the gray matter at the level of trauma. Furthermore, unlike the dosing regimen of MPSS, which appears to be effective when initiated within the first 8 hours of SCI, GM-1 demonstrated neuroprotective effects following initiation of treatment 48 hours post-SCI. The results from this study led to the initiation of a larger clinical trial comparing the effects of low- and high-dose GM-1 to placebo.
The randomized double-blind Sygen Multicenter Acute Spinal Cord Injury Study was completed in the late 1990s and reported in 2001.25 In this landmark clinical trial, all 797 patients received a 30-mg/kg bolus of MPSS, followed by a 5.4-mg/kg/hour infusion of MPSS for 23 hours initiated within the first 8 hours following SCI, as the standard of care based on the initial report from the NASCIS II. Placebo, low-dose GM-1 (300 mg loading dose followed by 100 mg/day for 56 days), or high-dose GM-1 (600 mg loading dose followed by 200 mg/day for 56 days) was then randomly assigned following the completion of corticosteroid administration to avoid unwanted drug interactions. Recovery patterns were measured by using the American Spinal Injury Association (ASIA) and modified Benzel classification neurologic examination scales.26 The primary efficacy assessment assigned to this trial was termed marked recovery, defined as the proportion of patients who demonstrated a two-grade improvement from baseline examination. Although the primary outcome for this trial was negative, patients who received either low- or high-dose Sygen showed an accelerated motor recovery over the first 3 months postinjury, regardless of their initial severity. There was also a trend toward improved bowel/bladder function, sacral sensation, and anal contraction.26 Interestingly, the placebo group of this trial failed to confirm the neurologic improvements noted with MPSS for the NASCIS II trial. To our knowledge, no future clinical trials with Sygen are currently planned.
Opioid Antagonists
Following SCI, the release of the endogenous opioid peptide dynorphin A is increased at the injury site along with kappa opioid receptor binding capacity. Although dynorphin A and other related opiates have well-defined roles as inhibitory neurotransmitters in pain reduction, evidence also exists to support the hypothesis that exposure to high concentrations of dynorphins can induce hyperalgesia and allodynia and may contribute to neurodegeneration. Moreover, sustained exposure to dynorphin-derived peptides is neurotoxic, and activation of the kappa-opioid receptor subtype has been associated with reduced spinal cord blood flow, which may contribute to secondary injury after SCI.27 The use of opioid antagonists for treatment of SCI is controversial. While the nonselective opioid receptor antagonist naloxone has demonstrated an ability to improve spinal cord conduction, reduce edema, and decrease allodynia in animal models of SCI, results from the NASCIS II trial showed that naloxone given as an IV bolus of 5.4 mg/kg followed by an infusion of 4 mg/kg for 23 hours failed to provide benefit.16 However, further analysis of the effects of timing for initiating naloxone treatment in patients with incomplete SCI showed that naloxone dosing within the first 8 hours of SCI resulted in significant motor recovery below the level of injury.28 Therefore, further refinement of naloxone-dosing regimens in experimental SCI models may be beneficial.
Thyrotropin-Releasing Hormone
TRH is a tripeptide (Glu-His-Pro) that possesses numerous physiologic actions, in addition to its well-characterized actions on the pituitary gland. TRH and its analogues are capable of antagonizing the effects of endogenous opioids, platelet-activating factor, peptidoleukotrienes, and excitatory amino acids, all of which have been implicated in the biochemical events of secondary SCI.29 In response to the overwhelming therapeutic benefit demonstrated by TRH in animal models of SCI, a small clinical trial consisting of 20 patients was designed to assess the safety and potential efficacy of TRH in patients with complete and incomplete SCI.30 The patients were randomized on the basis of the severity of their SCI (complete or incomplete) in a double-blinded fashion to receive either a 0.2-mg/kg bolus of TRH followed by a 0.2-mg/kg/hour infusion of TRH over 6 hours or an equal volume of saline vehicle placebo. The results of the 4-month follow-up revealed statistically significant improvements in neurologic and sensory function. However, this study was hampered by the small patient population that was included in the study and the high variability of neurologic scores obtained within the placebo group. Results for complete SCI patients treated with TRH were negative. Although 1-year data were available, a relatively high number of patients were lost to follow-up, and the data were deemed inconclusive. Considering the amount of evidence in support of neuroprotective effects generated by TRH and TRH analogues from previous preclinical studies and the results obtained from this small clinical trial, larger and more extensive trials with TRH may be beneficial.
Glutamate Receptor Antagonists
The class of glutamate receptor antagonists readily crosses the blood–spinal cord barrier following systemic administration. However, given the fact that glutamate is a key neurotransmitter possessing a variety of physiologic functions, blockade of glutamatergic synaptic transmission results in unwanted psychomimetic effects. Gacyclidine, a potent and specific phencyclidine analogue noncompetitive N-methyl D-aspartate receptor antagonist, was previously shown to provide neuroprotection through improvement of the functional, histologic, and electrophysiologic status of the injured rat spinal cord following experimental contusion with an absence of dose-related adverse effects.31 These findings prompted initiation of a randomized double-blind phase II clinical trial with more than 200 patients recruited to test the efficacy of 0.005, 0.01, and 0.2 mg/kg gacyclidine administered via IV infusion within 2 hours of SCI and with a second dose given within the next 4 hours.32 The subjects of this trial were divided into four strata based on the level of SCI (cervical or thoracic) and the severity of injury (complete or incomplete). The study results, reported for the 1-month follow-up, demonstrated improved overall ASIA scores. However, results for the follow-up at 12 months posttreatment failed to support long-term benefit on neurologic scores. Despite evidence of a trend for improved motor function in the incomplete cervical strata receiving higher gacyclidine dosing, further development of this drug for SCI has been halted.
Calcium Channel Blockade
Calcium influx is believed to be an important mediator of excitotoxic intracellular damage and vasospasm-induced ischemia. Consequently, Ca2+-channel blockers are prescribed in the treatment of systemic hypertension, coronary artery disease, stroke, cardiac arrhythmias, and vasospasm. Vasospasm and the release of endogenous vasoactive amines have been suggested as important contributing factors to the pathobiology of SCI.2 The dihydropyridine-sensitive (L-type) Ca2+-channel blocke nimodipine represents the most widely investigated therapy for vasospasm. In a rat model of severe spinal cord compression, nimodipine, alone or in combination with dextran, was examined for effects on spinal cord blood flow and spinal cord axonal function.33 The results of this study demonstrated that the administration of nimodipine alone was associated with systemic hypotension due to vasodilation. However, the combination of nimodipine with dextran resulted in increased cord blood flow within the lesion epicenter and improved axonal function in the motor and somatosensory tracts of the cord.
In 1996, a randomized clinical trial of 100 acute SCI patients was initiated, comparing the safety, efficacy, and neurologic outcome following administration of nimodipine (0.015 mg/kg/hour over 2 hours followed by 0.03 mg/kg/hour for 7 days), MPSS (NASCIS II dosing regimen), both treatments, or placebo control.34 Results from the 1-year follow-up demonstrated no neurologic benefit beyond the natural course of recovery within the placebo group. The authors also examined the results based on the timing of surgical spinal decompression and stabilization and found no correlation with recovery of neurologic function and the timing of surgery, although the patients were not randomized on the basis of timing of surgery. Although the data from this SCI clinical trial were not suggestive of nimodipine-mediated therapeutic effects, the conclusions made by the authors were susceptible to type II errors owing to the small patient population.
Sodium Channel Blockade
After SCI, there is a deleterious accumulation of intracellular sodium.35 Injury causes failure of the Na+/K+-ATPase and accumulation of axoplasmic sodium through noninactivating Na+ channels, which, together with membrane depolarization, promotes reverse Na-Ca exchange and axonal calcium overload. Thus, administration of sodium channel blockers may prevent the ensuing calcium-induced cell injury or death. Pharmacologic antagonism of voltage-gated Na+ channels has been demonstrated to prevent axonal degeneration, preserve the function of injured spinal cord white matter tracts, and reduce damage to myelin.36 For example, the focal administration of tetrodotoxin directly into the contused rodent spinal cord significantly reduced axonal loss and axoplasmic pathology, as compared to vehicle-treated animals.36 The benzothiazole anticonvulsant Na+ channel blocker riluzole is neuroprotective and promotes functional neurologic recovery following SCI in rodents.37 Riluzole exerts neuroprotective properties in the injured spinal cord following systemic administration by sparing gray and white matter rostral-to-caudal to the injury epicenter,37 a property that appears to be due to its ability to decrease the levels of intracellular sodium and calcium. The use of riluzole as a therapy for SCI is potentially feasible, as it has already received approval from the Food and Drug Administration (FDA) for treatment of amyotrophic lateral sclerosis. Furthermore, riluzole has been used as an adjunctive therapy in combination with MPSS.38 However, despite the current in vivo evidence in support of riluzole as a therapeutic drug for SCI, efforts to proceed with clinical testing of this drug are lacking.
Potassium Channel Blockade
Demyelination of intact and injured axons is a prominent feature of SCI. Axonal conduction deficits that arise from demyelination appear to contribute to the neurologic outcome following SCI, although the underlying mechanisms remain unclear.39,40 It has been suggested that focal demyelination of intact axons, along with altered activity of ion channels, plays an important role in the loss of axonal conduction.41 The exposure of K+ channels as a result of demyelination results in a reduced safety factor of action potential propagation (i.e., the ratio of action current generated by an impulse to the minimum amount of action current needed to maintain conduction) across the demyelinated region of the axon. Under normal physiologic conditions, myelinated axons have a high density of voltage-gated Na+ channels at the nodes of Ranvier, while rapidly inactivating voltage-gated K+ channels are clustered at the paranodal and internodal regions beneath the myelin sheath. As such, these K+ channels normally do not play a prominent role in repolarization of the action potential, as their outward current is constrained by myelin. However, when demyelination occurs, there is no myelin to shield the capacitance of the internode, and shunting of the action current occurs due to these fast voltage-gated K+ channels. Subsequently, the action potential progressively declines, resulting in conduction failure, marked slowing of the conduction velocity, and/or the inability to sustain repetitive discharges.42
Experimental studies using K+ channel blockers on animal nerve preparations have provided the rationale for undertaking clinical trials to assess the safety and efficacy of 4-AP, a blocker of rapidly inactivating voltage-gated K+ channels, as a therapy in SCI patients. Electrophysiologic studies have shown that 4-AP is capable of restoring conduction in focally demyelinated axons, enhancing synaptic transmission in many types of neurons, and potentiating muscle contraction.42 Several preclinical trials of intravenously administered 4-AP have demonstrated transient improvements in neurologic function in patients with chronic SCI.43–45 To date, three multicenter phase II randomized clinical trials of fampridine-SR, a sustained-release oral form of the K+ channel-blocking compound 4-AP, have been conducted with SCI patients. The first of these studies was a randomized crossover study, which demonstrated that fampridine-SR improved spasticity (modified Ashworth scores), ASIA motor and sensory scores, erectile dysfunction, and bowel function compared to placebo.46 In another crossover trial involving 60 patients, there was a nonsignificant trend toward reduced erectile dysfunction and decreased spasticity.47 A double-blind, randomized, placebo-controlled, parallel-group phase II clinical trial was conducted at 11 academic rehabilitation research centers in the United States to assess the safety and efficacy of fampridine-SR in subjects with chronic SCI.48 The study enrolled 91 patients with motor-incomplete SCI, randomized to three arms: fampridine-SR 25 mg twice daily (group I), fampridine-SR 40 mg twice daily (group II), and placebo (group III) for 8 weeks. Outcome measures included ASIA and Ashworth scores, bladder and bowel management questionnaires, and Subject Global Impression (SGI). A higher discontinuation rate was seen in group II patients, compared to group I and group III patients, due to more frequent adverse side effects such as hypertonia, generalized spasm, insomnia, dizziness, pain, constipation, headache, and seizures. Overall, group I patients showed significant improvement in SGI, and subgroup analysis showed improvement in spasticity in the lower-dose fampridine-SR group compared to the placebo group.48
Erythropoietin
Erythropoietin (EPO) is a 34-kD hematopoietic glycoprotein that binds to its receptor (EPOR) to induce signals promoting survival, differentiation, and proliferation of erythroid progenitor cells.49 The expression of EPO and EPOR is widely distributed within the developing and adult human brain and spinal cord, and is up-regulated in the adult brain after injury. Peripherally administered EPO crosses the blood-brain barrier, stimulates neurogenesis and neuronal differentiation, and activates neurotrophic, antiapoptotic, antioxidant, and anti-inflammatory signaling pathways.50 EPO is the only hematopoietic growth factor whose production is regulated by hypoxia, in which low oxygen tension activates hypoxia-inducing factor-1 to up-regulate EPO gene transcription.
Endogenous and exogenously administered EPO, and EPOR, have been reported to play important roles in SCI.51 EPO expression is up-regulated after SCI as part of the physiologic response to hypoxia. EPO has been shown to be neuroprotective in vitro and is capable of protecting neuronal cells from hypoxia-induced apoptosis and from excitotoxic cell death.52 The administration of exogenous recombinant human EPO (rhEPO) has been reported to produce substantial neuroprotection in animal models of SCI, spinal nerve root crush injury, transient spinal cord ischemia, and spinal cord inflammation in experimental autoimmune encephalitis. Although the mechanisms by which EPO exhibits its neuroprotective effects are not fully understood, EPO is capable of preventing apoptosis, reducing inflammation, and restoring vascular integrity. The preventive effects of EPO on neuronal apoptosis have also been demonstrated in a spinal cord compression model in rats, in which EPO administration results in inhibition of caspase-1 and caspase-3 and induction of survival proteins such as Bcl-xL. EPO administration also results in a reduction in neutrophil infiltration after SCI and has been shown to delay the postinjury increase in TNF-α, decrease interleukin-6 (IL-6) levels, and reduce apoptotic cell death.53–55 EPO has also been shown to prevent endothelial cell apoptosis, stimulate mitogenesis, and promote angiogenesis by restricting vascular endothelial growth factor–induced permeability and strengthening endothelial tight junctions. In addition to the anti-inflammatory properties of rhEPO, the inhibition of lipid peroxidation may contribute to its neuroprotective effects.56 Results of studies in animal models suggest that treatment with rhEPO may be beneficial after SCI, even when rhEPO was administered up to 24 hours after the initial injury. The delivery of rhEPO appears to protect ventral spinal cord motor neurons in an ischemic injury model in rabbits.57 Although a study of rat thoracic spinal cord contusion and clip compression injury reported substantial tissue sparing and recovery in locomotor function,55 an independent SCI research group failed to reproduce similar results, as delivery of rhEPO in the same injury paradigm failed to decrease secondary injury and cystic cavitation or to improve locomotor function.58
In a recent phase II clinical trial, intravenously administered rhEPO was shown to be safe and demonstrated a strong trend to reduce infarct size and improve clinical outcome in stroke patients.59 Although there has been great interest in conducting an SCI clinical trial with rhEPO, given the significance of initial preclinical results following experimental SCI, the use of rhEPO in clinical trials of human SCI deserves caution, and requires further investigation before implementing clinical trials. A major concern of EPO is the inadvertent and unwanted stimulation of hematopoietic activity, increasing the risk for thrombosis. Several EPO analogues have been developed in attempts to address this concern. The administration of the short-lived asialo-erythropoietin, in which sialic acid residues have been removed, has been shown to be neuroprotective in animal models of stroke, SCI, and peripheral neuropathy without causing erythrocytosis.60 In addition, asialo-erythropoietin was as effective as rhEPO in normalizing motor function after experimental SCI using a clip compression model. Other EPO analogues such as carbamylated EPO do not bind to the EPO receptor, a property that confers a loss of hematopoietic activity.61 Carbamylated EPO is capable of maintaining its associated neuroprotective properties, resulting in reduced neurologic deficit in comparison to saline or EPO in a chronic rodent model of SCI, and remained effective even when treatment was delayed for 24 hours.61
