Botulinum toxin A (BoNT-A) has found wider therapeutic acceptance in rehabilitation across a range of neurological disorders resulting in spasticity and/or dystonia and in other medical and surgical conditions. The new research shows the toxin is effective against neuropathic pain, a common debilitating condition with poor response to currently available therapy.
1. Introduction
Neuropathic pain (NP) resulting from damage or dysfunction of the peripheral or central nervous systems is one of the most common forms of pain, affecting up to 10% of the general population
[1][2]. The current evidence of treatment outcomes puts success rates at best between 30% and 50%, and existing treatments come with significant side effects
[3]. The effect of NP pervades patient sensation, thoughts, feelings, and behaviours. Treatment outcomes of these patients’ experiences have been described as “woefully inadequate”
[1][3]. Clinicians’ express difficulty in dealing with NP
[4], thus limiting the outcomes of pain relief, the healing of underlying conditions, and rehabilitation to a satisfactory functional quality of life.
Over the last two decades, botulinum toxin A (BoNT-A) has found wider therapeutic acceptance in rehabilitation across a range of neurological disorders resulting in spasticity and/or dystonia and in other medical and surgical conditions
[5][6]. Other scientific research from pharmacology, toxicology, and biology
[6][7] has identified the mechanism of BoNT-A in NP (
Figure 1) and supported its use as being safe and effective for an increasing number of applications, including a number of conditions with neuropathic pain. Multiple randomised controlled trials (RCTs)
[8][9][10][11][12][13][14][15][16][17][18][19][20][21][22][23][24][25][26] and systematic reviews
[27][28][29][30][31][32][33] have added evidence, specifically on NP. Despite this substantial body of evidence, BoNT-A has not been considered as adequate for the first-line treatment of NP, and it is reasonable and important for rehabilitation medicine physicians to ask why this is so.
Figure 1. The mechanism and effect of botulinum toxin in neuropathic pain. Botulinum toxin reversibly inhibits the release of acetylcholine from the presynaptic vesicle and causes local chemodenervation resulting in reduced muscle contraction. The possible mechanisms of action on pain involves (i) retrograde axonal transport of toxin; (ii) inhibition of neuropeptides, such as substance P, calcitonin gene-related protein (CGRP), and glutamate; and (iii) deactivation of Na channel. All prevent peripheral and central sensitisation.
To date, published RCTs on NP have shown efficacy and safety but have failed to provide adequate supportive evidence addressing the diagnosis of NP, the effective optimum dose, administration in different NP conditions, the duration of effectiveness, quality of life and functional ability, and other data that would underpin NP management guidelines. The limited design and small sample size of the RCTs led Shackleton et al.
[28] to conclude that the promising pain control results shown in medical and surgical use in recent years required further well-designed placebo-controlled trials, not only to support BoNT-A as a backup treatment but also for first-line pain relief. Croford
[3] has further argued that new treatments for chronic pain are of the utmost urgency. To design an appropriate guidelines trial, we require more detail on the limitations of the study evidence to date.
2. Final Visual Analogue Scale (VAS) Measures
The Visual Analog Scale (VAS) is discussed in some detail because of its consistent use as an outcome measure in the reviewed studies and, secondly, because of its importance of identifying bias in research and research reporting.
The mean final VAS and the standard deviation of the final VAS for placebo and botulinum toxin groups were pooled across 17 studies using a random effects meta-analysis model. Heterogeneity in the study estimates was assessed using the I-squared statistic (88.1%) and Cochran’s Q p value (<0.0001), which showed considerable heterogeneity leading to the use of random effects models in analyses. The overall mean difference in the final VAS across the studies was 2.59 (95% confidence interval (CI): 1.79, 3.38) (Figure 2), identifying a mean higher VAS outcome pain score in the placebo group overall.
2.1. VAS Difference
The mean VAS differences (final minus baseline) and the standard deviation of VAS difference for the BoNT-A and placebo groups were pooled across 7 of the 17 studies. Heterogeneity analysis using the I2 statistic (88.2%) showed substantial heterogeneity. Overall, the mean difference in the final VAS units across the studies was 2.34 (95% CI: 1.07, 3.61), identifying a higher mean VAS outcome pain score in the placebo group.
2.2. 50% Reduction in VAS
The proportion of patients with at least a 50% reduction in VAS between baseline and final time periods in the BoNT-A and placebo groups was pooled across six studies using a random effects meta-analysis model. Heterogeneity in the study estimates was assessed using the I2 statistic (44.9%) and showed moderate heterogeneity. The mean relative risk across the studies was 4.90 (95% CI: 2.00, 6.13), identifying a higher mean VAS outcome pain score in the placebo group.
2.3. Pain Frequency (Number of Neuralgia Attacks in a Single Day)
The mean pain frequencies and standard deviations for the BoNT-A and placebo groups were pooled across three studies. Heterogeneity in the study estimates was assessed using the I2 statistic and showed moderate heterogeneity (49.7%). The overall mean difference in pain attacks across the studies was 24.47 (95% CI: 19.09, 29.86), identifying a higher mean frequency of pain in the placebo group.
2.4. Other Relevant Outcomes
Of the other study outcomes (sleep—three studies, anxiety—two studies, depression—two studies, and mental and physical health—two studies) included in the RCTs, there were no statistically significant differences observed between the BONT-A and placebo groups in the meta-analysis (Figure 3, Figure 4 and Figure 5).
Figure 4. Hospital anxiety and depression score.
Figure 5. Quality of life.
2.5. Final VAS Funnel Plot
For the comparison of the BoNT-A and placebo final VAS (based on
Figure 1), a funnel plot assessed publication bias for the final continuous VAS outcome (
Figure 6). The funnel plot shows the standard error of mean difference on the
y-axis and the mean difference on the
x-axis. As most of the values fall outside the “funnel”, there is a substantial degree of study bias (Egger’s test: R = 2.73,
p = 0.002). Oher likely sources of this bias are the small and variable sample sizes of the RCTs and the variability across other elements of RCT composition, including the sample (inpatients, outpatients, or both)
[34].
This entry is adapted from the peer-reviewed paper 10.3390/toxins14010036