Effectiveness of 4-Aminopyridine for the Management of Spasticity in Spinal Cord Injury: A Systematic Review

The current review was reported in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA).14 

A comprehensive literature search was conducted on five electronic databases (PubMed, Scopus, Embase, CINAHL, PsycINFO) for articles published up to and including January 2017. The following combination of key terms was used as a search strategy for each database to identify relevant articles: (spinal cord injur^*) AND (spasticity) AND (aminopyridine). Filters were applied in each database to restrict searches to articles published in English and with only human subjects. Retrieved articles were then reviewed to identify additional articles that may not have been discovered in the initial search.

Articles were included for review if they met all of the following a priori inclusion criteria:

• sample had three or more subjects,

• 50% or more of the sample sustained an SCI

• subjects were 18 years old or older,

• subjects received 4-AP or placebo via any route, and

• spasticity was assessed before and after the intervention with a formal outcome measure.

Articles were excluded if information on subject characteristics, intervention protocols, or outcomes could not be adequately extracted.

Two independent reviewers assessed the selected articles for methodological quality (J.W., J.H.), and discrepancies were resolved by a third reviewer (A.McI.). Randomized controlled trials (RCTs) were evaluated using the Physiotherapy Evidence Database (PEDro) tool (Table 1).15 The PEDro tool consists of 11 items, each answered with “yes” (score = 1) or “no” (score = 0). The first item is not used in calculating the final score, such that the tool yields a maximum score of 10. Descriptive assessment of PEDro scores categorized them as poor (<4), fair (4–5), good (6–8), or excellent (9–10).16 

Data were extracted precisely from the studies without assumptions or simplifications. Extracted data included subject characteristics (ie, age, gender, injury duration, injury severity, injury level), sample size, study design, intervention protocols, assessment methods, side effects, adverse events, and spasticity-related outcomes. Data were collated and organized into tables that included the aforementioned categories. All studies were assigned levels of evidence using a modified Sackett scale, which simplified the original 10-level scale into five levels (Table 2).16 

Three outcome measures were used to evaluate spasticity: Ashworth Scale/Modified Ashworth Scale, Spasm Frequency Scale, and Reflex Score.

Ashworth Scale (AS)/Modified Ashworth Scale (MAS). The AS/MAS is a measure of spasticity based on subjective clinical assessments of muscle tone during passive range of motion.17,18 Each muscle is rated on a 0 to 4 scale, with 0 indicating no increase in muscle tone and 4 indicating rigidity in flexion/extension. The MAS includes an additional item (1+) to indicate resistance through less than half of the range of motion.

Spasm Frequency Scale (SFS). The SFS measures the number of sustained flexor and extensor muscle spasms per hour. All spasms are rated on a 0 to 4 scale, with 0 indicating no spasms and 4 indicating >10 spasms per hour.19 

Reflex Score (RS). The RS measures the briskness of deep tendon reflexes and accompanying clonus at the knee and ankle of both legs. Each reflex is rated on a 0 to 4 scale, with 0 indicating no detectable response and 4 indicating very brisk response with accompanying clonus.20 

The safety and tolerability of a medication depend on its reported side effects and adverse events. A side effect is considered to be “any unintended effect of a pharmaceutical product occurring at a dose normally used in [humans], which is related to the pharmacological properties of the drug.”21(p42) An adverse event is considered to be “any untoward medical occurrence that may appear during treatment with a pharmaceutical product but which does not necessarily have a causal relationship with the treatment.”21(p40) 

For this systematic review, 9 of 131 studies met inclusion criteria (Figure 1). Studies were published between 1993 and 2014. Five studies2,9,10,22,23 were conducted in the United States, and four studies11–13,24 were conducted in Canada. Six studies2,9–11,23,24 were RCTs and assigned Level 1b evidence, with PEDro scores ranging from 6 (good) to 10 (excellent). Three studies12,13,22 were pre-post tests and assigned Level 4 evidence. One RCT10 reported on two separate trials.

The total pooled sample size of all included studies was 591, with sample sizes in each study ranging from 3 to 212 (Table 3). The duration of injury ranged from 1 to 49 years. One study10 did not report the duration of injury, only that subjects were injured for more than 1.5 years. The severity of injury was measured using the American Spinal Injury Association Impairment Scale (AIS) in seven studies2,9,10,13,22–24 and the Frankel scale in two older studies,11,12 with grades ranging from A to D. One study23 did not report the specific severity scores, only that injuries were either incomplete or complete. Four studies9,10,13,24 only included subjects with incomplete injuries. The level of injury ranged from cervical to lumbar. Two studies2,22 only included subjects with paraplegia and one study13 only included subjects with tetraplegia. One study10 did not report the level of their subjects' injuries. Only one study10 required that subjects have clinically significant levels of spasticity (AS >2) for inclusion.

4-AP was administered orally in five studies,9,10,13,23,24 intravenously in three studies,2,11,12 and intrathecally in one study22 (Table 4). Intrathecal 4-AP was infused at 5 μg/h for a period of 5 hours.22 The study was a pre-post test and did not report criteria regarding the use of concomitant medications. Intravenous 4-AP was infused up to a maximum dosage of 15 to 30 mg over a period of 2 to 4 hours. Two studies2,11 were RCTs and one study12 was a pre-post test. Both RCTs were crossover trials, in which all subjects received both 4-AP and placebo. In all three studies, subjects were prohibited from taking concomitant antispasmodic medications during the trial.

Oral 4-AP was provided in sustained-release (SR) tablets in three studies9,10,24 and immediate-release (IR) tablets in two studies.13,23 The target dosage ranged from a total of 20 to 80 mg daily. The duration of intervention ranged from 2 to 16 weeks. Three studies9,10,23 incorporated 2-week titration periods, before and after attaining the target dosage, into the total intervention period. Four studies9,10,23,24 were RCTs and one study13 was a pre-post test. The control groups in two RCTs9,10 received placebo, while the control group in one RCT23 received low-dose 4-AP. One RCT24 was a crossover trial, in which all subjects received both 4-AP and placebo. Two RCTs9,23 had two unique intervention groups. In three studies,13,23,24 subjects were prohibited from taking concomitant antispasmodic medications during the trial. In two studies,9,10 the majority of subjects were taking concomitant antispasmodic medications during the trial, as long as they were taken consistently and had been doing so for more than 3 weeks. One study10 stratified group allocation by concomitant medication use and accounted for the effects during statistical analysis.

Spasticity was a primary outcome measure in four studies2,9,10,24 and a secondary outcome measure in five studies11–13,22,23 (Table 4).

Ashworth Scale (AS)/Modified Ashworth Scale (MAS). The AS/MAS was used in seven studies.2,9,10,13,22–24 Halter et al22 reported that two subjects(33%) showed increases in AS score after intrathecal 4-AP. In one study of intravenous 4-AP, Donovan et al2 reported mean changes in the intervention and control groups for AS score on the right side (+0.34 vs +0.34) and left side (+0.50 vs +0.50) post intervention, but the differences were not significant within or between groups (p > .05).

In a small pre-post test of oral 4-AP IR, Potter et al13 reported that two subjects (66%) showed reductions in MAS score post intervention. In a trial of oral 4-AP SR, Potter et al24 found that a greater proportion of the intervention group showed greater reduction in MAS scores than the control group post intervention, although the difference was not significant (15% vs 4%; p = .180). Analysis of mean MAS scores, however, revealed a statistically significant (p < .05) and clinically meaningful (Δ = ±1) reduction that was only attributable to 4-AP. In a trial of oral 4-AP IR, Segal et al23 reported that all subjects showed a significant reduction in MAS score post intervention (2.6 ± 1.5 to 1.9 ± 1.7; p = .04); however, there were no significant differences between intervention and control.

In a phase II trial of oral 4-AP SR, Cardenas et al9 reported that the intervention groups showed lower mean AS scores than the control group post intervention, for both 25 mg bid (1.0 vs 1.2; p = .04) and 40 mg bid (1.1 vs 1.2; p = .23). While these differences were not significant (p > .025), post hoc analysis of subjects with baseline AS >1 revealed that the 25 mg bid intervention group showed significantly greater reduction in AS scores than the control group (p = .02). As such, for two phase III trials, Cardenas et al10 selected subjects with clinically significant levels of spasticity at baseline (AS >2). The intervention groups had greater reductions in mean AS scores compared to the control groups in both trials, but these differences were not significant (p > .05).

Spasm Frequency Scale (SFS). The SFS was used in three studies,9,10,24 all of which examined oral 4-AP. Potter et al24 found that a greater proportion of the control group showed reduction in SFS score than the intervention group post intervention, although the difference was not significant (23% vs 12%; p = .317). Cardenas et al9 reported no significant differences in SFS scores within or between the intervention and control groups post intervention (p > .025). In two trials, Cardenas et al10 found reductions in overall SFS scores in the intervention and control groups. The intervention group showed significantly greater reduction in mean upper extremity (UE), but not lower extremity (LE), SFS score than the control group after one of the two trials (0.13 ± 0.05 vs 0.02 ± 0.05; p = .044).

Reflex Score (RS). The RS was used in three studies,2,11,12 all of which examined intravenous 4-AP. Hansebout et al11 reported that four subjects (50%) showed reductions in RS post intervention, while Hayes et al12 reported that two subjects (33%) showed reductions in RS. Donovan et al2 reported mean changes in the intervention and control groups for RS score on the left side (+0.25) and right side (+0.17) respectively post intervention, but the differences were not significant within or between groups (p < .05).

In earlier studies of oral 4-AP,13,23,24 there was inconsistent reporting of side effects and adverse events. Potter et al24 reported that subjects (n = 29) experienced side effects of dizziness (17%), pneumonia (3%), and urinary tract infection (3%) with 4-AP SR (12.5–17.5 mg bid), while no adverse events were experienced. Segal et al23 noted that dizziness and abdominal distress were common though transient side effects of 4-AP IR (30 mg qd), but did not specify the respective rates. Potter et al13 did not report any information regarding side effects or adverse events of 4-AP IR (10 mg bid/tid).

The more recent trials of oral 4-AP SR by Cardenas et al9,10 did not distinguish between side effects and adverse events, referring to all such occurrences by the latter term. Cardenas et al9 performed statistical analyses comparing the rate of adverse events in the placebo group (n = 31) to those in the low-dose 4-AP SR (25 mg bid; n = 30) and high-dose 4-AP SR (40 mg bid; n = 30) groups. In the low-dose group, there were significantly higher rates of generalized pain (37% vs 13%) and abdominal pain (23% vs 3%) than the placebo group, respectively. In the high-dose group, there were significantly higher rates of insomnia (43% vs 10%), dizziness (40% vs 6%), paresthesia (27% vs 3%), abdominal pain (23% vs 3%), anxiety (23% vs 3%), and nervousness (23% vs 0%) than the placebo group, respectively. The rate of dropouts due to adverse events was significantly higher in the high-dose group (37%) than in the low-dose (10%) and placebo (7%) groups.

In two separate trials, Cardenas et al10 compared 4-AP SR (25 mg bid; n₁ = 114, n₂ = 103) to placebo (n₁ = 98, n₂ = 100). The rate of adverse events was greater with 4-AP than placebo in both the first (43.0% vs 24.5%) and second (48.5% vs 25.0%) trials. The most common of these events in all subjects who received 4-AP were urinary tract infection (25.8%), hypertonia (21.7%), and dizziness (14.7%), while the least common were urinary incontinence (4.6%), peripheral edema (4.6%), and pharyngitis (2.8%). The rate of discontinuation due to adverse events was greater with 4-AP than placebo in both the first (16.7%vs 3.1%) and second (15.5% vs 3.0%) trials. The most common reasons for discontinuation in all subjects who received 4-AP were dizziness (5.1%) and hypertonia (2.8%). Only one death was reported between the two trials, which occurred in a subject receiving placebo due to a preexisting medical condition.

In the three studies of intravenous 4-AP,2,11,12 there were no serious adverse events, but localized pain at the site of infusion was common. Hansebout et al11 reported that five of eight subjects (62.5%) experienced such pain, and Hayes et al12 found it in all six subjects (100%). Donovan et al2 noted that 9 of 12 subjects (75%) experienced localized pain only during 4-AP infusion. In the one study22 of intrathecal 4-AP, there were no serious adverse events, but four of six subjects (66.7%) each experienced one minor side effect (ie, pain, spasm, headache, and dysesthesia).

For the last 25 years, multiple reports regarding the effects of 4-AP on spasticity in SCI have emerged. A total of five small studies,11–13,23,24 utilizing intravenous or compounded 4-AP oral tablets, have reported varying improvements in spasticity secondary to SCI. Spasticity assessments were all secondary endpoints with the exception of one study24 that collected the MAS as a co-primary endpoint. When efficacy was reported in a subset of subjects, the percentage of subjects experiencing improvement was consistent with larger scale, later studies9,10 (≥30%). Two other small studies utilizing intravenous2 or intrathecal22 4-AP reported no significant reduction in spasticity secondary to SCI.

Given the uncertainty of 4-AP's effectiveness in managing SCI-related spasticity, and thus the potential for additional study, Acorda Therapeutics conducted a phase II clinical trial (n = 91) of a 4-AP SR matrix tablet (Fampridine-SR) for 25 mg bid and 40 mg bid dosage arms compared to placebo.9 Cardenas et al9 reported that the phase II trial showed a nonsignificant but strong trend in reduction of mean AS score in the group receiving the lower of two dose levels compared to controls. Since a large proportion of subjects had little or no spasticity at baseline, a post hoc subgroup analysis performed on those receiving 25 mg with baseline AS score >1 (n = 14) revealed statistical significance for improved AS over controls (n = 16; p = .04). For a similar subgroup with AS ≥2, the magnitude of AS reduction was greater than that of the subgroup with AS >1. Despite this larger magnitude, the effect in the second subgroup did not reach significance (p > .025) when compared to controls.

Two parallel phase III trials (n₁ = 213, n₂ = 203) treated subjects with 25 mg bid or placebo, with change from baseline AS score as a co-primary endpoint.10 Cardenas et al10 reported small, nonsignificant differences in the change from baseline AS score in favor of the intervention over placebo for both phase III studies. A similar trend was also observed for the secondary endpoint (ie, SFS). However, nominal significance was achieved for the SFS UE subscale in the second phase III study.

It has been noted that higher rates of side effects at higher doses of 4-AP may have contributed to the failure of the phase II and phase III clinical trials to reveal significant improvement in spasticity secondary to SCI.5,25 Discontinuation of intervention in these trials was more than five- to six-fold greater in treated subjects than controls. The frequency of adverse events considered possibly or probably related to intervention was almost double in treated subjects compared to controls. Such observations may be the rationale for future clinical trials using derivatives of 4-AP such as 4-aminopyridine-3-methanol (4-AP-3-MeOH) that have shown a 10-fold increased potency in restoring axonal conduction with less side effects in animal studies.5,6,25,26 Consistent efficacy trends and known off-label use of 4-AP in subjects with SCI suggest a need for further study,27 whether future clinical trials be focused on 4-AP or a derivative with fewer side effects.

Different dimensions of spasticity are likely represented by different efficacy measurements as assessed by subject self-reports (ie, SFS is correlated with clonus and/or activities of daily living [ADLs] interference) and clinical examination (ie, AS/MAS reflects single-joint resistance to movement). Subject perception of changes in multiple ADLs throughout the day as compared to a single weekly clinical assessment may account for the stronger trend in SFS over AS seen in both phase III trials.

Differences between results of UE and LE SFS may reflect the differences noted by the subject when performing repetitive UE ADLs (eg, grasping, lifting, pushing, turning) compared to less repetitive LE ADLs (eg, transfers that move the LE). UE ADLs are likely more prevalent during the majority of the day for all subjects regardless of neurological level of injury. SFS for performance of UE and LE ADLs that require coordinated movement and are common across subjects may help to elucidate consistent functional improvements resulting from 4-AP administration.

Outcome measures have evolved since the phase II and III trials were conducted. Additional validated spasticity and functional measures specific to SCI, such as the Spinal Cord Assessment Tool for Spastic Reflexes (SCATS), Spinal Cord Injury Spasticity Evaluation Tool (SCI-SET), Spinal Cord Independence Measure (SCIM), Graded Redefined Assessment of Strength, Sensibility and Prehension (GRASSP), and Spinal Cord Ability Ruler (SCAR), are now available to improve the measurement of spasticity and related outcomes. As well, uncontrolled adjunctive interventions such as ongoing clinical therapies including extension to home-based practices, self-initiated activities, and individual activity levels may have impacted spasticity outcomes.

The conclusions of the current review are limited by the small number of studies that met inclusion criteria. Seven studies2,11–13,22–24 with small sample sizes reported that adequate power was not achieved. Five studies11–13,22,23 did not assess spasticity as a primary outcome. Four studies11–13,22 did not perform statistical analysis and only reported rates of improvement. There was considerable variation between the studies in terms of route, dose, and/or duration of 4-AP administration. All of these factors complicated the synthesis and interpretation of findings in the context of a systematic review.

It should also be noted that the current review only included peer-reviewed studies that were published in English. Studies published in other languages and other formats (eg, conference abstracts, grey literature) were not included. These factors rendered the review susceptible to publication bias and may have contributed to the limited number of included studies. Additionally, a meta-analysis was not performed due to insufficient raw data available in most of the studies.

There is a lack of significant evidence supporting the efficacy of 4-AP via oral, intravenous, or intrathecal administration to warrant its routine use in managing SCI-related spasticity. Factors such as different outcome measures used and higher rates of side effects at higher doses may have contributed to the obfuscation of a clear effect. Future phase III trials should utilize contemporary outcome measures validated for spasticity in SCI and should ensure that 4-AP is administered in safe and tolerable doses. Additional analyses should be conducted to account for the use of adjunctive interventions and the heterogeneity of SCI subjects.