Programmable pumps can be surgically implanted for the treatment of spasticity in SCI. The most commonly delivered drug is intrathecal baclofen (ITB). Many of the studies looking at intrathecal baclofen use for spasticity combine multiple causes of spasticity (SCI, multiple sclerosis and cerebral palsy) which makes the results difficult to interpret for SCI specifically. Several of the studies in this section include studies where fewer than 50% of the patients have SCI. While these individual studies may not meet the formal SCIRE criteria, their importance for inclusion emerged by representing a larger number of patients with spinal cord injuries when grouped together.
Outcome measures for intrathecal baclofen include direct spasticity measures such as the Ashworth and MAS and spasm frequency scales, and indirect measures such as functional outcome measures, complication rates and quality of life as well as cost-benefit analyses.
While oral baclofen can be useful in the treatment of spasticity, the use of high doses can lead to adverse effects, most commonly over sedation. Delivering baclofen directly into the cerebral spinal fluid allows a higher concentration of baclofen administration to the spinal cord with fewer systemic side effects. ITB is most effective in treating lower extremity spasticity and less so for upper extremity spasticity. However, the location of the intrathecal catheter tip can be adjusted at the time of surgical implantation depending on the clinical presentation with higher tip locations (cephalad at T6, compared to T10-L2) being used in patients with higher injuries (Burns & Meythaler 2001). The pump can be programmed to provide a steady dose of intrathecal baclofen throughout the day or programmed to include boluses at certain times of the day. ITB is usually only considered after one year post-SCI.
Potential complications from intrathecal baclofen treatment include overdose, withdrawal and surgical complications. Disruption or malfunction of the catheter-pump system is a common cause of withdrawal and can result in an acute life-threatening baclofen withdrawal syndrome. The signs and symptoms of acute intrathecal baclofen withdrawal include increased spasticity, itching, fever, altered mental status, rhaobdomyolysis, seizures, reversible cardiomyopathy, and death.
There were six studies employing an RCT design to evaluate the effects of intrathecal baclofen in SCI. Although these studies are small and combine different etiologies of spasticity, they do provide a limited body of level 1 evidence to support the use of intrathecal baclofen test doses to decrease spasticity in SCI as measured by Ashworth scale and spasm frequency score (Penn et al. 1989; Loubser et al. 1991; Coffey et al. 1993; Nance et al. 1995; Ordia et al. 1996, Hugenholtz et al. 1992).
Only one study (Kravitz et. al. 1992) examined the effects of intrathecal baclofen on the neurophysiological effects on mucle. They looked at EMG activity in six subjects in a double-blind placebo control crossover design. All six subjects had spinal cord injury. They noted a significant reduction in nocturnal EMG activity while on intrathecal baclofen. This was a small study and the study did not look at any correlation with spasticity or daily activity outcomes.
It is unlikely that RCTs will be undertaken to examine the long-term effectiveness of intrathecal baclofen given the effectiveness of test doses. However, several level 4 studies support the long term use of intrathecal baclofen to decrease spasticity with the most frequently used outcome measures being the Ashworth scale and spasm frequency scales (Heetla et al. 2009; Ochs et al. 1989; Penn et al. 1989; Broseta et al. 1990; Loubser et al. 1991; Penn et al. 1992; Coffey et al. 1993; Abel & Smith 1994; Nance et al. 1995; Ordia et al. 1996; Korenkov et al. 2002; Plassat et al. 2004; Boviatsis et al. 2005; Avellino Loeser 2000; Heetla et al. 2010 and Kawano et. al. 2018). Most recently Kawano conducted a retrospective chart review on 34 individuals with spasticity due to various spinal cord disorders. All the subjects were recieving long-term baclofen. 29 of the 34 subjects had spinal cord injury. The data collected showed an average reduction in the ashworth’s score of 1.92 and an average dose of intrathecal baclofen. The average followup time period was greater than 6 years and the average dose fo baclofen was 230.5 ucg per day. This review supported low complication rates and a stable dose of intrathecal baclofen in most of the patients. This study did not included any references to function or quality of life. The effects of intrathecal baclofen are more pronounced in the lower extremities than the upper extremities (Korenkov et al. 2002).
The systematic review by McIntyre et al. (2014) summarizes the evidence from a pooled subject sample size of 162 patients. All eight of the studies included in the review provided level 4 evidence that ITB decreases spasticity with a decrease in MAS from 3.1 to 4.5 at baseline to 1.0 to 2.0 at follow-up. Over time, there was an increase in the range of average daily dosing in these subjects from 57 to 187 mcg/day at baseline to 219 to 536 mcg/day at follow-up. Follow-up periods ranged from five to 41 months.
The effects of intrathecal baclofen on functional outcomes are much harder to summarize. Most studies are observational, pre-post studies with small numbers of SCI patients grouped in combination with several other diagnoses (most often MS). In addition, there is a lack of standardized outcome measures used to study functional outcomes. Finally, the majority of studies are not stratified by SCI level or ASIA impairment scale (AIS).
There are several observational studies looking at the short-term and long-term complication rates seen with intrathecal baclofen. Overall, complication rates are low and can be categorized as medication related or pump related. However, complications can be severe and include death (Heetla et al. 2009; Loubser et al. 1991; Penn et al. 1992; Coffey et al. 1993; Abel & Smith 1994; Nance et al. 1995; Azouvi et al. 1996; Ordia et al. 1996; Stempien & Tsai 2000; Korenkov et al. 2002; Plassat et al. 2004; Avellino & Loeser 2000; Heetla et al. 2010).
Borrini et al. (2014) conducted a prospective study of adverse events in a mixed group of patients with SCI (42%), multiple sclerosis (28%) and other diagnoses resulting in spasticity. Among 128 patients with pre-existing pumps and 30 patients undergoing new pump implantation, they found an overall incidence rate of event of 0.023 adverse events per month. The majority of adverse events (58%) occurred within the first month post surgery (new implantation or pump replacement). Furthermore, in a retrospective review of 130 patients with SCI (38%) and multiple sclerosis (62%), Draulans et al. (2013) found an adverse event rate of 0.011 events per month with the majority (75%) of complications related to catheter problems. Finally, Stetkarova et al. (2015) reported a complication rate of 0.01 per month with an acute withdrawal rate of 0.004 per month.
Tolerance to intrathecal baclofen has also been observed (Ochs et al. 1989; Coffey et al. 1993; Abel & Smith 1994; Ordia et al. 1996; Heetla et al. 2009). Intrathecal baclofen has been shown to decrease sexual function as measured by self-reported penile rigidity, duration of erection and ejaculation. The effect of intrathecal baclofen on ejaculation appears to be reversible based on a small number of cases (Denys et al. 1998).
In a recent pre-post study, Calabro et al. (2014) demonstrated a decrease in erectile function in men following intrathecal baclofen pump implantation, especially at higher dosages.
Overall, there is level 4 evidence to suggest that functional outcomes as measured by scales such as Barthel index scale and FIM improve with intrathecal baclofen (Parke et al. 1989; Broseta et al. 1990; Nance et al. 1995; Azouvi et al. 1996; Ordia et al. 1996; Plassat et al. 2004; Boviatsis et al. 2005). However, it is notable that Zahavi et al. (2004) reports a small statistically significant deterioration in disability as measured by the expanded disability status scale, ambulation index and incapacity status scale. The article notes that this may not be a clinically significant deterioration (Zahavi et al. 2004). Loubser reports the potential for decreased functional outcomes especially with respect to ambulatory status in patients who may depend on their spasticity for ambulation (Loubser et al. 1991).
Finally, in a recent qualitative study by Gunnarsson and Samuelsson (2015), patients reported an overall satisfaction with intrathecal baclofen treatment and all 14 patients stated that they would undergo pump implantation again.
There has been one prospective study that performed a cost analysis for intrathecal baclofen pumps but this study did not meet SCIRE criteria for inclusion (i.e., <50% SCI subjects). However, without other cost analysis studies involving SCI patients, findings by Postma (1999) are presented here since intrathecal baclofen cost effectiveness is an important consideration. Postma (1999) studied 33 subjects with MS and SCI and found subjects who received a pump had higher direct medical costs than subjects who did not receive a pump. However, Postma (1999) concluded that the improvement in quality of life in subjects who received a pump more than justified the direct costs associated with the pump.
A more recent review of ITB cost analysis, which was funded by a pump manufacturer, Medtronic, noted significant cost benefit, proportionate to the duration of pump use (Saulino et al. 2015). Again, this review did not meet the SCIRE criteria for inclusion (i.e., <50% SCI patients). To illustrate the potential cost savings of ITB treatment, Saulino et al. (2015) reported that by seven years post pump implant, the average cumulative savings was more than USD $22,000 per patient.
There are two level 1b studies examining cost-effectiveness with the usage of intrathecal baclofen (Nance et al. 1995; Ordia et al. 1996). Ordia et al. (1996) did not specify whether SCI or MS subjects were studied for cost-effectiveness, but the authors did report gross cost savings with intrathecal baclofen due to an overall reduction in hospital days post pump implantation (Ordia et al. 1996).
Nance et al. (1995) also combines MS and SCI subjects. In contrast to Ordia et al. (1996) who examined overall hospital days, Nance et al. (1995) examined only hospital days related to spasticity and found a net savings in costs related to pump implantation.
Intrathecal Baclofen Withdrawal and Overdose
With sudden withdrawal of intrathecal baclofen, there is a risk of an acute life-threatening baclofen withdrawal syndrome. The signs and symptoms of acute intrathecal baclofen withdrawal include increased spasticity, itching, fever, altered mental status, rhaobdomyolysis, seizures and death.
Withdrawal can occur with human errors in pump programming, errors in pump refills (wrong concentrations or dosages) and with mechanical failures in pumps and/or catheters. Of note, Miracle et al. (2011) reviewed varying types of imaging as an approach to the evaluation of intrathecal pump and catheter systems in the event of withdrawal symptoms. Patients with intrathecal baclofen pumps need to be educated regarding the pump alarm sounds and the signs and symptoms of baclofen withdrawal so that they can seek early assessment and treatment. The differential diagnosis for baclofen withdrawal includes autonomic dysreflexia, neuromalignant syndrome and malignant hyperthermia (Coffey 2002).
Initial treatment for intrathecal baclofen withdrawal is the reestablishment of intrathecal baclofen treatment as soon as possible. If this is not possible, then oral baclofen, dantrolene and intravenous benzodiazepines are used to help manage the withdrawal syndrome (Coffey 2002).
Acute baclofen withdrawal syndrome shares many characteristics with serotonergic syndrome. Meythaler (2003) added cyproheptadine, a serotonin antagonist, to the management of acute intrathecal baclofen withdrawal in four subjects and found improvements in signs and symptoms of withdrawal.
Intrathecal baclofen overdose is less common and is most often due to errors at the time of pump refills or programming errors (Watve 2011). Symptoms include hypotonia, hypotension, somnolence, seizures and respiratory depression. Acute overdose is managed with cessation of baclofen treatment, supportive treatment and withdrawal of cerebrospinal fluid (and subsequent reduction of circulating ITB). Once ITB treatment has been stopped and the overdose symptoms have been treated, patients need to be monitored closely for the development of of ITB withdrawal. Treatment with ITB should be resumed as early as possible
There is level 1a evidence (from six small-sample RCTs; Ordia et al. 1996; Nance et al. 1995; Coffey et al. 1993; Hugenholtz et al. 1992; Loubster et al. 1991; Penn et al. 1989) that bolus or test dose intrathecal baclofen decreases spasticity.
There is level 4 evidence (from several studies; see Table 13) that support the use of long-term intrathecal baclofen to decrease spasticity.
There is conflicting level 4 evidence (from several studies; see Table 15) that intrathecal baclofen may improve functional outcomes.
There is level 1b evidence (from two RCTs; Ordia et al. 1996; Nance et al. 1995) that intrathecal baclofen is a cost-effective intervention for treating post SCI spasticity.
There is level 4 evidence (from several studies; see Table 13) that complication rates with the long-term use of intrathecal baclofen are relatively low although complications can occasionally be severe.
There is level 4 evidence (from one case series; Meythaler et al. 2003) that adding cyprohetptadine to baclofen and benzodiazepines may be useful for the treatment of intrathecal baclofen withdrawal.
- Bolus or long-term intrathecal baclofen decreases spasticity and may improve functional outcomes with low complication rates and is a cost effective intervention.