Oral Medications

Baclofen, a derivative of gamma aminobutyric acid (GABA), is widely used as the first line of pharmacological treatment for spasticity in people with SCI (Kirshblum 1999; Taricco et al. 2006). Baclofen, also identified as Lioresal®, CIBA Ba-34647 and b-(parachlorophenyl) gamma aminobutyric acid, crosses the blood-brain barrier more readily than GABA itself and is believed to reduce spasticity by enhancing inhibitory influences on the spinal stretch reflex by increasing presynaptic inhibition (Kirshblum 1999).

In typical practice, baclofen requires a careful dose titration period with a usual maximal recommended dose of 20 mg qid (Burchiel & Hsu 2001) which is also the dosage employed in the majority of studies involving people with SCI (Aydin et al. 2005; Nance 1994). Veerakumar et al. (2015), in an analysis of a single provider’s patient database over 25 years, reported a significant but marginal increase in baclofen dosage over the 25-year span and patients with gunshot related SCI receiving earlier baclofen initiation than patients with SCI related to motor vehicle accident. Baclofen may be especially effective in reducing flexor spasms (Shahani & Young 1974; Duncan et al. 1976; Gracies et al. 1997) although these effects may also act to impair specific functional tasks such as walking or standing (Kirshblum 1999; Burchiel & Hsu 2001). The mechanism for impairment in functional tasks may require further exploration as Chu et al. (2014) did not record substantial decreases in voluntary electromyographic activity despite baclofen’s effectiveness in reduced stretch reflexes. A variety of adverse events may limit the use of baclofen including lowering of seizure threshold, sedatory effects (i.e., drowsiness), insomnia, dizziness, weakness, ataxia, anxiety and mental confusion (Hinderer 1990; Gracies et al. 1997; Kirshblum 1999; Burchiel & Hsu 2001). A significant side effect of antispasmodics, with baclofen being the most commonly used, is the potential for reduced torque during routine activities and for activity-dependent interventions such as locomotor treadmill training (Roy & Edgerton 2012; Harkema et al. 2012). Baclofen also increases cough threshold in cervical spinal cord subjects (Dicpinigaitis 2000). Sudden discontinuation or withdrawal of baclofen can result in seizures, confusion, hallucinations and rebound muscle overactivity with fever (Gracies et al. 1997). For the most part, tolerance with sustained use of baclofen is possible (Knutsson et al. 1974), but is not a major issue (Roussan et al. 1985; Gracies et al. 1997; Kirshblum 1999).

Benzodiazepines (i.e. diazepam/valium, clonazepam) are used for multiple problems encountered post-SCI such as anxiety, musculoskeletal pain and spasticity. Although Valium has been prescribed, since the 1960s (Neill et al 1964, Kerr et al 1966, Wilson & McKehnie 1966), for being superior to placebo in treatment of SCI related spasticity, benzodiazepines are not FDA approved for spasticity. However, they are sometimes prescribed for short-term treatment of spasticity (e.g. nocturnal spasticity). This class of drug acts by inhibiting afferent pathways to relax skeletal muscle or inhibiting gamma-aminobutyric acid (GABA) pre- and post-synpatically to depress the central nervous system. Given that excessive CNS depression was often cited as an unwanted side effect, diazepam is not as commonly prescribed as a first line treatment for spasticity as it once was before newer more effective classes of medications became available (e.g. baclofen). Other common side effects of benzodiazepines are ataxia, dyscoordination, fatigue, weakness, hypotension, sedation, depression, memory impairment and risk of addiction. Despite the known sedative effect of valium, it was shown to be superior to Amytal (barbiturate used as sedative or hypnotic) and placebo in reducing SCI related spasticity (Corbett et al 1972).

Author Year

Country
Research Design

Score
Total Sample Size

Methods Outcome
Baclofen
Yan et al. 2018

China

RCT

PEDro=6

N=336

Population: Baclofen (BA, n=112): Mean age=36.55±3.42yr; Gender: males=40, females=72; Level of injury: NR; Mean time since injury=211.45±25.47d; AIS scale: NR.

Botulinumtoxin A (BTI, n=112): Mean age=36.95±7.12yr; Gender: males=36, females=68; Level of injury: NR; Mean time since injury=207.45±20.49d; AIS scale: NR.

Placebo (control group or CG, n=112): Mean age=35.47±2.21yr; Gender: males=30, females=82; Level of injury: NR; Mean time since injury=205.98±16.45d; AIS scale: NR.

Intervention: Participants were randomized to receive one of three interventions, 1) Baclofen (BA), 2) botulin umtoxin A (BTI), and 3) placebo (CG). The BA group received 5, 10, 15, and 20mg of baclofen 3x/d in the 1st, 2nd, 3rd, and 4th wk, respectively. The BTI group received a local intramuscular injection of 500 U of botulin umtoxin A under EMG guidance. All groups including the placebo group received physical therapies including locomotor training and intensive task-specific training for 6 wk. Outcome measures were assessed at 2, 4, and 6 wk from the initiation of the intervention.

Outcome Measures: Muscle tone was assessed for the thumb, wrist, and fingers using the modified Ashworth score (mAS). Other outcomes included the Disability Assessment Scale (DAS), mMRC, and Barthel Index.

1.     Compared to baseline, BA and BTI had significantly improved mAS scores at the 2-wk follow up (p=0.003; p=0.02, respectively). CG showed no significant improvement at 2 wk.

2.     At 4 wk, CG, BA, and BTI showed significant improvements in mAS compared to baseline (p<0.001 for all). 3.     At wk 6 BA had no significant improvement in mAS when compared to baseline (p>0.05). However, BTI had significantly improved mAS scores compared to baseline (p=0.02).

4.     Both BA and BTI resulted in improvements in Barthel Index.

Luo et al. 2017

China

RCT

PEDro=6

N=150

Population: Baclofen (BAC, n=75): Mean age=36.6±1.7yr; Gender: males=20, females=55; Level of injury=NR; Mean time since injury=NR; AIS scale=NR; Mean dosage=24.33±12.5 mg/d.

Tolperisone (TOL, n=75): Mean age=35.5±1.5yr; Gender: males=27, females=48; Level of injury=NR; Mean time since injury=NR; AIS scale=NR; Mean dosage=378.2±102.1 U.

Intervention: Participants were randomly assigned to one of two groups; baclofen (BAC) or tolperisone (TOL). The BAC group received baclofen. Dosage was initiated at 5-10 mg 2-3x/d and was gradually increased by 5-10 mg/wk up to 80 mg/d. The TOL group received tolpersone. Dosage was initiated at 150-450 mg/d and increased to 600 mg/d. Both groups received treatment for 6 wk. Outcome measures were assessed at baseline and at wk 2, 4, and 6.

Outcome Measures: Modified Ashworth Scale (MAS), Barthel Index, Coefficient of efficacy.

1.     There were no significant between-group differences in mAS at baseline. BAC had significantly lower mAS scores compared to TOL at wk 2 and 4 (p=0.003; p=0.02, respectively), but by wk 6 there was no significant difference between the two groups (p>0.05).

2.     Both groups showed significant within-group improvements in mAS over the 6 wk (p<0.05 for both). BAC showed a significant improvement in mAS at wk 2 and then remained consistent. TOL showed a significant improvement at wk 2 and 6

3.     MRC improvement in both groups by wk 6. Barthel index improved in both groups by wk 6, but faster and to a greater extent in the TOL group. The BAC group had more side effects.

Chu et al. 2014

USA

RCT Crossover

PEDro=7

N=10

Population: Mean age: 48.9 yr; Gender: males=10, females=0; Level of severity: AIS C=4, AIS D=6; Mean time since injury: 138.7 mo.

Intervention: Individuals were randomly allocated to the order in which they received oral administration of baclofen (30 mg), tizanidine (4 mg), and placebo (10 mg). Assessments were done at baseline and 90 to 120min after the administration of each drug.

Outcome Measures: Ankle stretch reflex torque, Isokinetic knee extension torque, Isometric knee extension torque.

1.     There was a significant decrease in stretch reflex torque after tizanidine (p=0.034) but not baclofen (p=0.116) compared to placebo.

2.     Peak knee flexion torque during extension decreased significantly after baclofen (p<0.001) but not after tizanidine (p=0.20) when compared to placebo.

3.     Peak knee extension torque during flexion decreased significantly after baclofen (p=0.014) and tizanidine (p<0.001) compared to placebo.

4.     No significant changes in isokinetic knee torque were shown for either drug compared to placebo (p=0.179).

5.     Knee flexion torque significantly increased after tizanidine compared to placebo (p=0.033).

6.     Compared to placebo, there was a significant increase in isometric knee extension torque for both baclofen (p<0.001) and tizanidine (p=0.001).

7.     Changes in peak torque for baclofen (p=0.066) and tizanidine (p=0.99) did not differ significantly from placebo.

Nance et al. 2011

USA

RCT Crossover

PEDro=7

N=37

Population: Age range: 35-43 yr; Gender: males=25, females=15; Injury etiology: traumatic SCI=34, non-traumatic=3 SCI; Level of injury: AIS A=15, other=22.

Intervention: Individuals received a sequence of extended-release arbaclofen placarbil tablets 10, 20 or 30 mg or a placebo every 12hr for 26 days for each sequence.

Outcome Measures: Ashworth Scale (AS), Patient-rated Severity of Spasticity Scale.

1.      Arbaclofen placarbil significantly improved AS scores compared to placebo over the dosing interval; least-squares mean reduction versus placebo was 0.60 for 20 mg (p=0.0059) and 0.88 for 30 mg (p=0.0007).

2.      The difference was significant for the pre-morning dose time point, 12 hr after the prior evening dose, indicating that efficacy was maintained throughout the dosing interval.

3.      Treatment differences for arbaclofen placarbil 10 mg versus placebo were not significant.

4.      Severity of spasticity ratings were significantly reduced for the combined 20/30mg group versus placebo (p=0.018).

5.      No statistically significant differences between arbaclofen placarbil and placebo were observed for muscle strength.

Aydin et al. 2005

Turkey

RCT

PEDro=6

N=41

Population: SCI (n=21): Level of severity: complete, incomplete; Injury etiology: trauma=41; Chronicity=chronic. Healthy controls (n=20).

Intervention: Either oral baclofen (titrated up to 80 mg/day) for 8 wk or TENS for 15 min/day for 15 days.

Outcome Measures: Spasm Frequency Scale (SFS), Painful Spasm Scale, Ashworth Scale (AS), Various clinical (clonus, deep tendon reflexes, response to plantar stimulation) or electrophysiologic measures (H-reflex latency and amplitude, H/M ratio) of spasticity as well as measures of function (FIM and FDS).

1.     For both intervention groups a significant improvement was noted post treatment in the lower limb Ashworth score (p=0.011 baclofen group and p=0.020 TENS group), SFS (p<0.014 for both groups), deep tendon reflex score (p<0.025 for both groups) as well as in measures of disability (FIM-baclofen group p=0.005, TENS group p=0.003; FDS–baclofen group p=0.004, TENS group p=0.003.

2.     There were only small (statistically non-significant) changes in electrophysiologic variables with either baclofen or TENS, other than a significant reduction in H-reflex maximal amplitude (p=0.032) 24 hr after the final session of TENS. This reduction was even more apparent when tested only 15 min after the last treatment (p=0.026).

Effect Sizes: Forest plot of standardized mean differences (SMD±95%C.I.) as calculated from pre- and post-intervention data.

Hinderer et al. 1990

USA

RCT

PEDro=9

N=5

Population: Gender: males=5, females=0; Level of severity: complete, incomplete; Cause of injury: trauma; Chronicity: chronic.

Intervention: Baseline placebo period of varying length (2.5-4.5 wk), followed by a 2 wk dose titration period of baclofen at half target dose (40 mg/day), followed by 2.5-4.5 wk of 80 mg/day.

Outcome Measures: Viscous and elastic stiffness as assessed by measuring viscous and elastic torque responses to a sinusoidal ankle perturbation of 5° at 3 to 12 Hz. Testing occurred 2x/wk for 9 wk.

1.     No systematic effect of baclofen was noted. Of 300 total comparisons made, only 1 comparison reached significance, with an increased viscous stiffness apparent at a frequency of 4 cycles/sec when comparing placebo with initiation of baclofen at 40mg per day (p<0.05).

2.     Visual inspection of the results for individual subjects showed no evidence for a therapeutic response of baclofen that might not have been demonstrated by group statistical analysis.

Duncan et al. 1976

USA

RCT

PEDro=8

N=22

Population: SCI (n=11), MS (n=11), 3 dropouts (etiology unknown).

Intervention: Either oral baclofen (titrated up to 100 mg/day) for 4 wk or identical looking placebo.

Outcome Measures: Self-report of # of spasms, nocturnal awakenings (daily) and global impression of treatment (at end of each intervention period). Clinician also provided global impression (at end of intervention period) and assessed resistance to movement and rated change on 5-point scale (weekly). Also rated clonus, impressions of pain, use of limbs and transfer activity (weekly).

1.     Number of spasms was significantly reduced with baclofen versus placebo (p<0.01) as was number of nocturnal awakenings (p<0.01).

2.     11 of 22 subjects demonstrated less resistance to passive movement by at least 2 grades on the initial 5-point scale with baclofen versus 1/22 with placebo and this was significant (p<0.01).

3.     No improvement in gait was seen in any of those who could walk (n=8) nor were any improvements seen in tendon jerks, strength or voluntary movement.

4.     In 9 cases (41%) both individuals and clinicians felt continued use of baclofen was warranted.

5.     15 subjects identified mild side effects while on baclofen (4 on placebo). All were deemed insignificant.

Burke et al.1971

Australia

RCT

PEDro=7

N=6

Population: Injury etiology:  traumatic SCI=6; Level of injury: tetraplegic=6; Level of severity: complete, incomplete; Chronicity: chronic.

Intervention: Placebo or active drug (CIBA 34,647-Ba) was titrated to a maximum of 60mg daily over a period of 2 wk in a crossover, double-blind design.

Outcome Measures: Surface slope of EMG (quadriceps) versus velocity relationship associated with passive flexion of the knee.

1.     No group statistical results were provided.

2.     All 6 subjects had a reduced EMG/velocity ratio for any given speed tested with baclofen versus placebo (e.g., decreased to 37.5% (range 0%-67%) at a velocity of 200°/sec).

3.     All subjects displayed clinical effects with baclofen such as reduced stretch reflex responses.

Dicpinigaitis et al. 2000

USA

Prospective Controlled Trial

N=24

Population: Intervention group (n=12): Mean age: 39.2 yr; Level of injury: C=12. Control group (n=12): Mean age: 43.2y r; Level of injury: C=12.

Intervention: Both groups underwent Capsaicin cough challenge testing. The intervention group consisted of individuals receiving baclofen for the relief of muscle spasm while the control group did not.

Outcome Measures: Cough thresholds.

1.     Individuals in the intervention group had significantly higher cough thresholds than the control group in two or more coughs (p=0.009) or 5 or more coughs (p=0.024).
Veerakumar et al. 2015

USA

Cohort

N=115

Population: Mean age: 29.0 yr; Gender: males=97, females=18; Injury etiology: traumatic SCI=115; Level of injury: cervical=52, thoracic=59, lumbar=2, unknown=2.

Intervention: Chart review.

Outcome Measures: Oral baclofen use and dosage.

 

1.     53% (n=61) received oral baclofen.

2.     No significant differences in terms of cause (p=0.17) or level of injury (p=0.65) between those who were and were not prescribed oral baclofen.

3.     Patients given oral baclofen were significantly younger at time of injury than those not prescribed the medication (p=0.03).

4.     Other antispasmodics prescribed included diazepam and tizanidine. Individuals receiving baclofen alone had significantly higher dosages of baclofen then individuals on multiple antispasmodics (p<0.05).

5.     Increases in yearly baclofen dosage, from time of injury, were seen for the whole sample (1.26 mg/yr, p=0.01), motor vehicle accidents (4.99 mg/yr, p=0.0001), thoracic-spine injuries (1.97 mg/yr, p=0.238), gunshot wounds (0.99 mg/yr, p=0.032), and individuals prescribed baclofen as the sole antispastic medication (1.29 mg/yr, p=0.005).

6.     No factors were found to be significantly associated with the baclofen dosage slope.

Nance 1994

Canada

Pre-Post

N=25

Population: Gender: males=25, females=0; Injury etiology: SCI=25; Level of injury: paraplegia, tetraplegia; Level of severity: complete, incomplete.

Intervention: 1 wk up-titration, 1 wk target dose (0.05 mg bid clonidine; 4 mg qid cyproheptadine; 20 mg qid baclofen), 1 wk down-titration.

Outcome Measures: Ashworth Scale (AS), Pendulum test, VII.

1.     A significant reduction in spasticity was seen with baclofen in all 3 outcome measures-as with the other 2 drugs tested (p<0.0001).

2.     Generally, baclofen results were among the most improved as compared to the other 2 drugs although this was only significant for the pendulum test (p=0.06) and VII (p<0.0007–along with cyproheptadine).

Benzodiazapines
Corbett et al. 1972

England

RCT Crossover

PEDro=7

NInitial=22

NFinal=9

Population: Traumatic SCI(n=22): Mean age=NR; Gender: males=20, females=2; Level of injury: NR; Lesion type: complete=14, incomplete=8; Time since injury=>4 mo; AIS scale: NR.

Intervention: This RCT crossover consisted of three conditions; 1) Valium 5 mg, 2) Amytal 30 mg, and 3) Placebo. The washout period consisted of 3 days. Participants received one tablet on the first day, one b.d. on the second day, and one t.d.s. for the remaining 3 days. The trial lasted 6 wk for each participant. Outcome measures were assessed by a senior doctor, two physiotherapists (one who treated the individual and who did not), a junior doctor, and the patient. Measures were taken on a daily basis with the exception of the senior doctor and treating physiotherapist who made measures once or twice per wk.   

Outcome Measures: Spasticity: subject assessment (worse, no effect, better, much better).

1.     Valium was significantly more effective at reducing spasticity compared to both amytal and placebo when assessed by the senior doctor (n=19, p<0.02).

2.     When assessed by the junior doctor, valium was significantly more effective at reducing spasticity compared to amytal (n=9, p<0.05).

3.     Valium was significantly more effective at reducing spasticity compared to placebo when assessed by the treating physiotherapist (n=11, p<0.05). 4.     There were no significant differences between the three drugs when assessed by the non-treating physiotherapist and individual (n=11, p>0.05).

Neill et al. 1964

UK

Crossover

NInitial=21

NFinal=20

Population: SCI (n=21): Mean age=41.19±13.86yr; Gender: males=16, females=5; Etiology: SCI=14, Other=7; Level of injury: C=11, T=4; Lesion type: complete=10, incomplete=5; Time since injury=NR; AIS scale: NR.

Intervention: Participants received two bottles of 250 tablets. Each bottle was given for 2 wk and consisted of either 2 mg diazepam or placebo. All participants received one bottle with diazepam and one with placebo. The starting dose consisted of two tablets every 6 hours. Dosage was increased to 3 tablets every 6hr after 1 week. Outcome measures were assessed prior to the intervention, as well as twice weekly. Assessments were completed by a senior and junior doctor, a physiotherapist, and the patient.

Outcome Measures: Spasticity: assessed by passive movement of limbs [better (+1), worse (-1), or same (0) as initial condition].

1.     13/20 participants self-reported an improvement with diazepam, however, no improvement on placebo.

2.     5/20 participants self-reported no improvement in spasticity on either diazepam or placebo.

3.     2/20 participants self-reported the best benefit from the placebo.

4.     8, 3, and 9 participants had preference for diazepam, placebo, and neither pill, respectively.

5.     According to the observers’ assessments, 13/20 participants showed a significant improvement in spasticity when on diazepam (difference of 5 or more cumulatively). One individual showed significant improvement on placebo.

Discussion

Oral baclofen

Despite the general acceptance and clinical experience of using oral baclofen to reduce spasticity in people with SCI, at least two systematic reviews have noted a relative paucity of high quality studies (i.e., RCTs) demonstrating specific or comparative efficacy (Chou et al. 2004; Taricco et al. 2006). Taricco et al. (2006) conducted a Cochrane Review of all pharmacological interventions for spasticity following SCI. Only one study examining the effect of oral baclofen (Burke et al. 1971) met the review inclusion criteria (i.e., RCT with at least 50% of participants with SCI published up to July 2004). The reviewers deemed this study to have been relatively poor quality with small sample size (n=6) so did not provide a positive assessment of the efficacy of oral baclofen.

Since the 2006 Cochrane Review, two additional RCTs (Aydin et al. 2005, n=21; Chu et al. 2014, n=10) published demonstrating effective reduction in spasticity with oral baclofen as measured using the Ashworth Scale, Spasm Frequency Scale, deep tendon reflex score, FIM and Functional Disability Scores, and electromyographical measures. An earlier RCT (Duncan et al. 1976) demonstrated reduced spasticity as measured by the Ashworth Scale and Spasm Frequency Scale. Further support for the efficacy of oral baclofen was provided by a pre-post study by Nance (1994) in which baclofen was compared to clonidine and cyproheptadine in 25 subjects with SCI. In general, all three agents were shown to be effective in relieving spasticity with baclofen among the most effective for each of the measures. Chu et al. (2014; N=10) compared baclofen and tizanidine and reported that baclofen was preferentially effective for flexors and tizanidine for extensors. This finding suggests that tailoring of antispastic drug therapy to spasticity characteristics of individual patients may be possible.

Yan et al (2018) directly compared Baclofen (BA) to botulinumtoxin A (BTI) and reported trial participants’ modified Ashworth scores (MAS) reflected significant anti-spasmodic performance of both BA and BTI at 2 weeks (p=0.003 and p=0.02, respectively) compared to baseline. At 4 weeks, the control group’s MAS score was also significantly decreased (p<0.001 for all 3 treatments). However, at 6 weeks post treatment, only the BTI group had significantly reduced spasticity (p>0.05) compared to baseline. These results confirm the known immediate efficacy of BA that could also be susceptible to resistence development. On the other hand, this trial demonstrates the persistence of BTI neurotransmitter release inhibition. Interestingly, only the BA group was significantly improved functionally as assessed by the Disability Assessment Scale (DAS; p=0.05). However, the DAS was specifically developed for use in stroke, not SCI. Both BA and BTI administration produced side effects such as asthenia and sleepiness, and bronchitis and elevated blood pressure, respectively.

Luo et al (2017) compared baclofen to tolperisone (a centrally acting muscle relaxant) for the treatment of spasticity secondary to SCI. As with the comparison to BTI, after initial significant efficacy, MAS continued to significantly decline b6 week 6 only with the comparator (tolperisone). Thomas et al 2010 attributes long-term use of baclofen with reduced muscle activity and maximal tetanic forces. Accordingly, Luo et al (2017) surmise that long-term use of baclofen weakens the whole muscle that in turn introduces fatigue.

Arbaclofen placarbil (AP) is a prodrug of R-baclofen in an extended-release oral formulation that is well absorbed throughout the gastrointestinal tract (Lal et al. 2009). Efficacy and safety of AP was studied (Nance et al. 2011) and results indicated that 20/30mg of AP every 12 hours for 26 days, significantly improved Ashworth scores and reduced severity of spasticity ratings compared to the placebo group. However, there was no significant difference in muscle strength between the AP and placebo group. Chu et al. (2014) also reported that neither baclofen or tizanidine had a negative impact on strength and postulated that patient reported weakness secondary to drug administration by be due to decreased motivation secondary to drowsiness.

In contrast to these studies, a counter-therapeutic response to baclofen was found by Hinderer (1990). In this RCT (n=5) the effect of baclofen on spasticity was studied by examining the viscous stiffness (resistance torque) following a 5° sinusoidal ankle perturbation at 3-12 Hz. No difference was noted between baclofen and placebo on this measure. No other outcome measures were assessed. Chu et al. (2014) noted that baclofen had a stronger inhibitory effect on knee flexors. These studies illustrate one of the limitations in establishing the efficacy for any spasticity-relieving agent–the heterogeneity of reported outcomes and outcome measures used across studies (Chou et al. 2004; Taricco et al. 2006). Spasticity is multi-dimensional with a variety of clinical manifestations and much day-to-day and diurnal variation within an individual. A battery of measures is needed to obtain valid and reliable measurement of spasticity within a given trial (Priebe et al. 1996). The range of studies outlined in the present review demonstrates various physiological, clinical and functional measures, yet there is minimal consistency of outcome measure selection across trials.

Interesting findings from a retrospective analysis of a single provider’s 25-year patient database (Veerakumar et al. 2015, n=115) revealed that baclofen use varied with etiology of the SCI, time since injury and concomitant antispasmodic use. These findings provide further evidence that baclofen use requires careful dose titration and monitoring based on the unique spasticity profile of individual patients.

Benzodiazepines

Although Valium proved to be superior to Amytal and placebo for controlling SCI related spasticity, only 2/22 participants did not require other treatment methods (e.g. physiotherapy, hydrotherapy) to fully control their spasticity (Corbett et al 1972, RCT). Neill et al (1964) confirmed anti-spasmodic efficacy (compared to placebo) on 13/20 people with SCI and the prominence of drowsiness as the most frequent side effect. However, Neill et al concluded that the greatest benefit was recorded in participants with traumatic cervical SCI. This latter claim should be interpreted with caution given the small cohort of participants (N=20). Diazepam’s effectiveness for treatment of spasticity in people with SCI was further supported with an observational survey of 35 participants where 30 reported good to excellent relief. Only 3 participants complained of drowsiness but did not require discontinuation or diminution of the dosage.

Conclusions

There is Level 1a evidence that oral baclofen improves muscle spasticity secondary to SCI. This conclusion is based on the results from eight RCTs (Yan et al. 2018, Luo et al. 2017, Chu et al. 2014; Nance et al. 2011; Aydin et al. 2005; Duncan et al. 1976; Burke et al. 1971, Jones et al. 1970) although is minimally muted by a single negative finding from one small RCT (Hinderer et al. 1990) with an overall lack of homogeneity in outcome measures and study participants. Additional evidence from a prospective controlled trial (Dicpinigaitis et al. 2000), a cohort (Veerakumar et al. 2015) and pre-post study (Nance 1994) also provide support for the use of oral baclofen in reducing spasticity.

There is Level 1b evidence (Yan et al. 2018, N=336: Luo et al. 2017, N=150) supporting the immediate effect of baclofen for the treatment of spasticity but that at 6 weeks post treatment, baclofen is inferior to botulinumtoxin A and tolperisone.

There is level 1b evidence (Corbett et al. 1972, RCT, N=9) supported by 2 other trials ( level 2 evidence (Level 2, Neill et al 1964, Cohort, N=20) confirming that valium (diazepam) is effective in decreasing spasticity secondary to SCI.