REVIEW ARTICLE
A Critical Appraisal of the Evidence for
Botulinum Toxin Type A in the Treatment for
Cervico-Thoracic Myofascial Pain Syndrome
Mehul J. Desai, MD, MPH; Tatyana Shkolnikova, MD; Andrew Nava, MD;
Danielle Inwald, MD
The George Washington University Spine & Pain Center, Washington, District of Columbia,U.S.A.
& Abstract: Myofascial pain syndrome (MPS) is a musculo-
skeletal condition characterized by regional pain and muscle
tenderness associated with the presence of myofascial trigger
points (MTrPs). The last decade has seen an exponential
increase in the use of botulinum toxin (BTX) to treat MPS. To
understand the medical evidence substantiating the role of
therapeutic BTX injections and to provide useful information
for the medical practitioner, we applied the principles of
evidence-based medicine to the treatment for cervico-tho-
racic MPS. A search was conducted through MEDLINE (Pub-
Med, OVID, MDConsult), EMBASE, SCOPUS and the Cochrane
database for the period 1966 to 2012 using the following
keywords: myofascial pain, muscle pain, botulinum toxin,
trigger points, and injections. A total of 7 trials satisfied our
inclusion criteria andwere evaluated in this review. Although
the majority of studies found negative results, our analysis
identified Gobel et al.’s as the highest quality study among
these prospectively randomized investigations. This was due
to appropriate identification of diagnostic criteria, excellent
study design and objective endpoints. The 6 other identified
studies had significant failings due to deficiencies in 1 or
more major criteria. We conclude that higher quality, rigor-
ously standardized studies are needed to more appropriately
investigate this promising treatment modality. &
Key Words: botulinum toxin, myofascial pain syndrome,
neck pain, nociceptors, myofascial trigger points, review
INTRODUCTION
Myofascial pain syndrome (MPS) is a musculoskeletal
condition characterized by regional pain and muscle
tenderness associated with the presence of myofascial
trigger points (MTrPs). Clinically, these MTrPs are
focally hypersensitive taut bands (TBs), which produce a
local twitch response (LTR) and a typical referral
pattern on palpation.1,2 MTrPs are easily diagnosed
via physical examination by expert examiners; however,
inter-rater reliability of detection is poor and tender
points are often misdiagnosed as MTrPs.3
MPS affects up to 95% of patients with chronic pain
disorders and is a common finding in patients at
specialty pain medicine centers.4 It has also been
determined that neck and back pain affects up to 70%
of the general adult American population during their
lifetime and is considered the leading cause of job-
related disability in the United States.5 Traditional
treatment for MPS includes oral pharmacotherapy
(most commonly nonsteroidal anti-inflammatory drugs
[NSAIDs], steroids, muscle relaxants, antidepressants,
opioids, and vasodilators), injections (local anesthetic
Address correspondence and reprint requests to: Mehul J Desai, MD,MPH; GW Spine & Pain Center, 2131 K Street NW, Suite 600, WashingtonDC, 20037, U.S.A. E-mail: [email protected]
Submitted: December 21, 2012; Revision accepted: April 12, 2013DOI. 10.1111/papr.12074
© 2013 World Institute of Pain, 1530-7085/13/$15.00
Pain Practice, Volume ��, Issue �, 2013 ��–��
with or without steroid), physical therapy (PT), and dry
needling (TDN). These traditional methods have been
employed with varying degree of long-term benefit and
have demonstrated only transient pain relief. This has
driven the need for further research and understanding
of the underlying mechanism of MPS and the need for
the development of targeted treatments.
The pathophysiology of MTrPs is complex and
involves interactions between numerous mechanisms,
which results in the peripheral sensitization of local
muscle nociceptors.6 The inciting event in MTrP for-
mation is muscle overload whereby contractile forces
are irregularly distributed within a hypoperfused mus-
cle.7 Localized areas of muscle injury and ischemia
develop, causing low tissue pH and hypoxia. Low pH
inhibits acetylcholinesterase (AChE), increasing the
levels of acetylcholine (Ach) in the synapse, which
results in further local tenderness.7 Nociceptive termi-
nals in muscle tissue have a multitude of different
receptors in its membranes, including matched receptors
for biochemicals that are released from damaged or
hypoxic tissue such as CGRP (calcitonin gene–relatedpeptide), bradykinin, serotonin, protons (pH), and
prostaglandins.6 CGRP further increases ACh release
by the motor endplate (MEP). Continuous activation of
nociceptors by these and other endogenous substances
also leads to central sensitization of dorsal horn
neurons.6 Increased ACh concentration produces abnor-
mal MEP activity that leads to intense, self-sustaining
focal sarcomere compression under the MEP.7 Com-
pression further constricts local capillaries, increasing
hypoperfusion and ischemia. The continued presence of
these inflammatory mediators and other biochemicals
may be necessary for persistent pain conditions such as
MPS, resulting in an ongoing positive feedback loop.6
This supports Simon’s integrated hypothesis of local
ischemia and hypoxia, according to which the central
MTrP has multiple muscle fibers releasing excessive
ACh from dysfunctional MEPs.8 Therefore, if pain
arises in part due to excessive release of ACh from
dysfunctional MEPs, then blocking ACh with an agent
like botulinum toxin (BTX) may reduce the symptoms of
MPS.
BTX is a potent neurotoxin produced by the bacte-
rium Clostridium botulinum. It causes flaccid muscle
paralysis by blocking ACh release at the neuromuscular
junction (NMJ). BTX is thought to exert an analgesic
effect via muscle-relaxant properties and also directly
via inhibition of nociceptive neuropeptides.9 BTX is
commercially available in several subtypes, including
Type A and B (Botox [onabotulinumtoxinA] – Allergan,Inc., Irvine, CA, USA; Myobloc [rimabotulinumtoxinB]
– Solstice Neurosciences, Inc., South San Francisco, CA,
USA; Dysport [abobotulinumtoxinA] – Ipsen Biopharm
Ltd., Wrexham, UK; Xeomin [incobotulinumtoxinA] –Merz Pharma GmbH & Co KGaA, Frankfurt,
Germany), which are currently FDA-approved in the
United States to treat glabellar lines, cervical dystonia,
blepharospasm, strabismus, and severe primary axillary
hyperhidrosis that is inadequately managed by topical
agents. The off-label use of BTX in musculoskeletal
conditions currently extends to plantar fasciitis, pirifor-
mis syndrome, lateral epicondylitis, and MPS. Recently,
BTX has been shown in animal models to reduce the
levels of substance P and CGRP, key biochemicals
involved in pain mechanistic pathways.10 These results
have not been replicated in human models; however,
given the key role of these and other cytokines and
neuropeptides in MPS, the use of BTX is a natural
outlet.6
To understand the medical evidence substantiating
the role of therapeutic BTX injections, we applied the
principles of evidence-based medicine to the treatment
for cervico-thoracic MPS. Our purpose was to analyze
the science underlying the utility of BTX in treating
cervico-thoracic MPS in order to provide useful infor-
mation for the medical practitioner.
METHODS
We searched MEDLINE (PubMed, OVID, MDCon-
sult), EMBASE, SCOPUS, and the Cochrane database
for the period 1966 to 2012 using the following
keywords: myofascial pain, muscle pain, botulinum
toxin, trigger points, and injections. In addition, we
examined the references cited in these studies for those
key words. We did not review abstracts or unpublished
studies. Selected studies had to fulfill the following
criteria: (1) enrolled subjects had a chief complaint of
cervico-thoracic MPS, (2) each trial involved a prospec-
tive methodology, (3) 1 treatment arm received 1 or
more BTX injections, (4) the study must have been
published in English, and (5) the study had to be double-
blinded. We excluded studies that did not document
MTrPs, as well as nonspecific diagnoses such as low
back pain. We also excluded studies with small sample
sizes (n < 10), which could lead to erroneous treatment
outcomes.11,12
Articles meeting the study criteria were critically
evaluated. Data were compiled for each of the
2 � DESAI ET AL.
following categories: inclusion criteria, randomization
protocol (if appropriate), total number of subjects
enrolled initially and at final analysis, statistical
analysis used, documentation of technique, outcome
measures, follow-up intervals, results (positive or
negative), funding sources, and reported complica-
tions. The quality assessment systems developed by the
Agency for HealthCare Research and Quality (AHRQ)
(Table 1) and the Cochrane Database (Table 2) were
applied to the data.13 Two review authors indepen-
dently screened the trials and assessed the quality and
analysis of the results. If consensus was not reached,
data from the trial in question were reviewed as a
team again until the review authors resolved the
contentious issue.
A study was considered positive if the authors
concluded that BTX injection was more effective than
the reference/control treatment. Usually, this conclusion
was predicated on the conventional 5% level (P = 0.05)
for detecting a statistically significant difference in the
primary outcome measures. In a negative study, the
authors concluded either no difference between the
study treatments, or that better results were observed in
the reference/control treatment group.
RESULTS
A total of 7 trials satisfied the inclusion criteria,
were evaluated in this review, and are summarized in
Table 3.14–20 All 7 were published between 1998 and
2011 in peer-reviewed journals. Non-English studies
were excluded. All of the studies were prospective,
randomized, and double-blinded. One of the studies
used crossover design.14 All 7 studies incorporated a
placebo-controlled group using saline.
Methodology
In all 7 studies, the presenting complaint was cervico-
thoracic muscle pain. The duration in the studies varied
from 2 to 6 months. Gobel et al. specified that the
duration of symptoms criteria not exceed 24 months
with a minimum of 6 months of symptoms prior to
enrollment. Wheeler et al.16, however, recommended
average pain duration of 8.6 years prior to study
enrollment.
Inclusion criteria included (1) a regional pain com-
plaint, (2) pain or paresthesia in the typical distribution
Table 1. Agency for Healthcare Research and Quality(AHRQ) Domains and Elements for Randomized Con-trolled Trials*
Domain Elements†
Study Question Clearly focused and appropriate questionStudy Population Description of study population
Specific inclusion and exclusion criteriaSample size justification
Randomization Adequate approach to sequence generationAdequate concealment method usedSimilarity of groups at baseline
Blinding Double-blinding to treatment allocationInterventions Intervention(s) clearly detailed for all study groups
Compliance with interventionEqual treatment of groups except for interventions
Outcomes Primary and secondary outcome measures specifiedAssessment method standard, valid and reliable
Statistical Analysis Appropriate analytic techniques that address studywithdrawals, loss to follow-up, missing data, andintention to treat
Power CalculationAssessment of confoundingAssessment of heterogeneity, if applicable
Results Measure of effect for outcomes and appropriate
measure of precisionProportion of eligible subjects recruited into studyand followed up at each assessment
Discussion Conclusion supported by results with possiblebiases and limitations taken into consideration
Funding orSponsorship
Type and sources of support for study
*Adapted fromCenter for Regulatory Effectiveness prepared for Agency for HealthcareResearch and Quality, US. Department of Health and Human Services.†Elements appearing in italics are those with an empirical basis. Elements appearing inbold are those considered essential to give a system a full Yes rating for the domain.
Table 2. Methodological Quality Criteria List of CochraneMusculoskeletal Review Group*
Patient Selection1. Treatment AllocationWas the method of randomization describedand adequate?Was the treatment allocation concealed?
Yes No Don’tKnow
2. Were the groups similar at baseline regardingthe most important prognostic indicators?
Yes No Don’tKnow
Intervention3. Was the care provider blinded? Yes No Don’t
Know4. Was controlled for co-interventions whichcould explain the results?
Yes No Don’tKnow
5. Was the compliance rate (in each group)unlikely to cause bias?
Yes No Don’tKnow
6. Was the patient blinded? Yes No Don’tKnow
Outcome measurement
7. Was the outcome assessor blinded? Yes No Don’tKnow
8. Was at least one of the primary outcomemeasures applied?
Yes No Don’tKnow
9. Was the withdrawal/drop-out rate unlikely tocause bias?
Yes No Don’tKnow
Statistics10 Did the analysis include an intention-to-treatanalysis?
Yes No Don’tKnow
*Adapted from Manchikanti et al.
Botulinum Toxin in the Treatment for Myofascial Pain � 3
Table
3.Resu
ltsofPublish
edInvestigationsofBotulinum
Toxin
TypeA
intheTreatm
entofCervico-ThoracicMyofascialPain
Syndrome
Author
StudyDesign
DurationofPain
Dose
Regim
en
NConcu
rrent
Therapy
Study
Duration
Outcome
Measures
Results
AHRQ
C
Ojala
etal.
(2006)
DBRCT,
crossove
r>2month
ave
rage:
10.0
�8.6
months
Total
15–3
5U,
Mean
28�
6U
31(T
=15,
C=16)
Paracetamol,
n=13used
NSA
IDS
8weeks
(4week
treatm
ent)
SNP,PPT
Nosignificant
difference
betw
een
treatm
entand
controlgroups
7/10
4/11
Ferrante
etal.
(2005)
DBRCT
>6month
0/TP,10/TP,25/TP,
50/TP
Max5TPs
132(T1=32,
T2=34,
T3=31,
C=35)
Standardized
regim
enof
amitriptyline
10mg2hours
before
bedtime,
25mgafter
1week,ibuprofen
800mgQID,and
propoxyphene-
acetaminophen1
tabletQ4prn
12weeks
VASat
24hours,
1,2,4,6,8
and12weeks;
rescue
medication
use;PPT;SF-36
Nosignificant
difference
betw
een
treatm
entand
controlgroups
8/10
7/11
Wheeleretal.
(2001)
DBRCT
Atleast
5month
ave
ragewas
8.6
�9.6
years
231.20�
50.1
of
BTAX
206.80�
39.1
ofsaline
50(T
=25,
C=25)
Notstated
16weeks
NPAD;GAS;
SF-36;BDI
Nodifference
betw
een
treatm
entandco
ntrol
groups,howeve
rmore
AEswere
reportedby
theBTXA
thantheNS
groupduringthe4th
and8th
week.
7/10
3/11
Wheeleretal.
(1998)
DBRCT
>3month
Placebo,50Units
BTXA,100Units
BTXA
allin
2cc
ofnorm
alsaline
33(T1=11,
T2=11,
C=11)
NR
4monthsand
thenoffered
aseco
nd
injection
of100units
ofBTXA
NPAD;PPT
Nostatistical
significance
inBTX-A
ove
rplacebobutsome
subjectsasymptomatic
afterseco
ndinjection
of100uBTXA.
7/10
4/11
Gobeletal.
(2006)
DBRCT
6–2
4month
(ave
rage
18.5
month)
40U
persite
with
10most
tender
triggerpoints
145(T
=75,
C=70)
Noopioids,inva
sive
therapymethods,
NM
blocks,
corticosteroidsor
muscle
relaxa
nts.
NoTylenol,ice,
heat,massageor
rheumatism
bath
therapyonday
priorto
treatm
ent
12weeks
Proportionof
subjectswith
mildorno
pain
atweek
5;ch
angesin
pain
intensity;
numberof
pain-freedays
perweek;
pain
on
palpationof
muscles;GAS
Positive
primary
outcomeof
significantlymore
patients
onBTX-A
at
week5showedmildor
nopain
compared
withplacebo.In
additiontheBTX-A
groupalsoalloweda
significantlygreater
changefrom
baseline
inpain
intensity
duringweek5–8
and
significantlyfewer
days
perweekwith
outpain
betw
een
weeks5and12.
9/10
11/11
4 � DESAI ET AL.
Table
3.(Continued)
Author
StudyDesign
DurationofPain
Dose
Regim
en
NConcu
rrent
Therapy
Study
Duration
Outcome
Measures
Results
AHRQ
C
Qeramaetal.
(2006)
DBRCT
>6month
50U
/.25mLor
0.25mLof
isotonic
salineper
triggerpoint
30(T
=15,
C=15)
Physicalandmanual
treatm
ents
were
stoppedduring
studyperiod
4weeks
NRS,
spontaneous
andevo
ked
pain;shoulder
move
ments;
interference
pattern
and
spontaneous
electrical
activityon
EMG
BTX-A
reducedMEP
activityandthe
interference
pattern
ofEMG
significantly
buthadnoeffect
on
eitherpain
(spontaneousor
referred)orpain
thresholdsco
mpared
withisotonic
saline.
7/10
7/11
Lew
Etal.
(2008)
DBRCT
Atleast
2months
50U
perinjection
site
notto
exceed
totaldose
of200U
29(T
=14,
C=15)
NR
6months
VAS;
NDI;SF-36
Nosignificant
difference
betw
een
treatm
entandco
ntrol
groupexceptin
regard
totheSF
36there
was
astatistically
significantdifference
seenin
bodilypain
at2
and4month
and
mentalhealthscale
atonemonth.
8/10
7/11
BDI,Beck
DepressionInve
ntory;BTIG,BTX-A
injection;DNG,Dry
needling;GAS,
Patientandphysician’sglobalassessmentofim
prove
ment;LIG,Lidocaineinjection;NDI,Neck
DisabilityIndex;
NHP,Nottingham
HealthProfile;NPAD,
Neck
pain
anddisabilityScale
(consistsof20itemswhichuse
avisuala
nalogscale
tomeasure
neck
pain
andassociatedproblems;PPT,P
ain
pressure
threshold
withalgometry;
PS,
Pain
score
uponpalpation0-3;S
F-36,3
6item
short
form
healthsurveyasaquality
oflife
measurement;NRS,Numericratingscale
forpain;SNP,Seve
rity
ofneck-shoulderpain;T
P,T
riggerpoint;MEP,M
otorendplate;V
AS,Visuala
nalogpain
scale
ove
rthelast24hours;C
,Coch
raneassessment
scale;AHRQ,Agency
forHealthCare
ResearchandQuality;DBRCT,Double-blindrandomizedco
ntrolledtrial:NM,Neuromuscular;NR,Notreco
rded.
Botulinum Toxin in the Treatment for Myofascial Pain � 5
of the MTrP, (3) a TB in the muscle, (4) exquisite
tenderness found in that TB, and (5) a restricted range of
motion (ROM) in the affected muscle. Exclusion criteria
included history of diffuse pain, a history of cervical disk
or bone disease, neurologic deficits, trauma to the
shoulder or neck region, previous cervical surgery, and
radiculopathy of the upper extremities. Additionally,
patients were excluded if they suffered from another
serious medical condition or had psychologic problems,
including depression.
In 3 of the 7 studies, patients were randomized via
computer-generated randomization schedules18–20 One
study utilized block randomization,14 another study
used a randomization table,15 and 1 used numerical
coding17 to randomize subjects. Wheeler16 did not
report their randomization process.
Ojala randomized 31 subjects to total doses of 15-
35 units (U) of botulinum toxin A (BTX-A, Botox)
with 5 U per MTrP for a total of 3 to 7 sites (n = 15)
or just saline (0.05 mL) per MTrP in 3 to 7 sites
(n = 16). At 4 weeks, the subjects were crossed over.
Following randomization, Ferrante performed MTrP
injections in 4 total groups. The experimental groups
received 10, 25, or 50 U (n = 32, n = 34, and n = 31,
respectively) of BTX-A (Botox) in up to 5 sites for a
maximum total of 50, 125, or 250 U. The control
group received saline placebo (amount not reported) in
up to 5 sites (n = 35).
Twenty-five subjects were randomized by Wheeler16
to receive BTX-A injections (231.2 � 50.1 U) in the
cervico-thoracic musculature. Total sites injected were
not reported. An additional 25 subjects received saline
placebo (206.8 � 39.1 U). The conversion of milliliters
to units of saline was not explained. Wheeler17 ran-
domized 11 subjects into 1 of 3 groups. The first group
underwent BTX-A injection with 50 U in 2 mL of
normal saline, the second group received 100 U in 2 mL
of normal saline, and the third group was injected with
2 mL of normal saline. These injections were conducted
in the subject’s most sensitive MTrP as identified via
examination.
Gobel randomized a total of 145 subjects into 2
groups. The experimental group received 40 units of
BTX-A (Dysport) at 10 sites (n = 75), while the control
group underwent saline placebo injection at 10 sites
(n = 70).
Qerama randomized a total of 30 subjects to 50 U of
BTX-A (Botox) (n = 15) or 0.25 mL of normal saline
(n = 15) in the infraspinatus muscle. Fourteen subjects
underwent BTX-A (Botox) injection with 50 U per site
for a total dose of up to 200 U and 100 U per side with
up to 2 muscles per side in the experimental group in a
study by Lew. The control group received saline placebo
(n = 15).
Four of the 7 studies disclosed funding support in
entirety or in part from Allergan, Inc.15,16,19,20; 1 study
was funded institutionally14; 2 studies did not disclose
any external funding sources.18
Statistical Analysis
The only study to calculate the population base to power
analysis was the study by Gobel et al., who also used
Wilcoxon–Mann–Whitney test to assess baseline equiv-
alence between treatment groups. To calculate the
differences in efficacy between the 2 treatment groups,
Gobel used Wilcoxon rank sum test or Fisher’s exact
test. When multiple tests were conducted, the P value
was corrected using the Bonferroni’s method. Ojala
et al. used an analysis of variance (ANOVA) and
samples t-test to compare the difference of change in
severity of neck and shoulder pain (SNP) and the
minimum pressure that produced pain (pain pressure
threshold, or PPT) between the different treatments. To
compare the verbal assessments, Ojala used chi-square
analysis and a 2-tailed nonparametric Mann–Whitney
U-test to evaluate the difference between the groups at
separate time points. The crossover effects were esti-
mated using repeated measures of ANOVA. Ferrante
et al. used chi-square analysis with Fisher’s exact test for
nonparametric data and Student’s t-test and ANOVA
for parametric data.
For the 36-item short-form Health Survey (SF-36),
Ferrante et al. used generalized estimation regression
models that adjusted for multiple values per patient, and
for each subscale of the SF-36, 3 separate intergroup
analyses were performed. Wheeler et al.16 analyzed
their data with a 2 9 5 unweighted means analysis,
which tested for the between-group effect of treatment.
Wheeler et al.17 used ANOVA for each outcome mea-
sure to identify effects of treatment group’s pre- and
post-treatment and the Kruskall–Wallis test because of
the small sample size at each of the testing periods to
assess the differences between groups. Qerama et al.
also compared medians using the nonparametric Mann–Whitney and Wilcoxon rank test. Kamali et al. used the
chi-square for measurements of original variables and
the Wilcoxon’s signed-rank test for comparisons
between groups. Lew et al. used the Wilcoxon rank
sum test (1-sided t-test) and the Wilcoxon signed-rank
6 � DESAI ET AL.
test (2-sided paired t-test) for the between- and within-
group comparisons, respectively.
Outcomes
The Ojala’s study reported no difference between small
doses of BTX-A and saline for MPS. 100% of enrolled
subjects completed the study and were available for
follow-up at 4 and 8 weeks. Outcome measures
included SNP and PPT via pain pressure dolorimetry.
No statistically significant differences in these measures
were observed following the interventions.
Ferrante analyzed visual analog scores (VAS), PPT,
SF-36 quality-of-life measure, and use of propoxyphene-
acetaminophen napsylate as a rescue medication at
weeks 1, 2, 4, 6, 8, and 12. Ferrante reported that
injection of BTX-A directly into MTrPs did not improve
cervico-thoracic MPS when compared with placebo.
The study reported no statically significant improvement
in VAS sum of changes from baseline, analysis of mean
weekly VAS scores, comparison of rescue dosing among
groups, or analysis of mean MTrP PPT. Interestingly, all
groups including placebo showed significant improve-
ments in VAS scores, while the use of rescue medication
and PPT demonstrated no observed differences among
treatment groups.
Neck Pain and Disability Visual Analogue Scale
(NPAD; 20 items including the VAS to measure neck
pain and associated problems), SF-36, and the Beck
Depression Inventory (BDI) were assessed by Wheeler16
at 4, 8, 12, and 16 weeks. Of the 50 subjects enrolled,
45 completed the study. The authors reported benefits in
both groups on Neck Pain and Disability Visual Ana-
logue Scale (NPAD). Improvements in subject and
physician global assessments of improvement (GAS),
PPT, BDI, and SF-36 were also noted but were not
specific to either treatment. Wheeler16 concluded that
single-dose treatment of BTX-A was not effective for
chronic neck pain.
All 33 subjects enrolled by Wheeler17 were available
for final evaluation at 1, 3, 6, 9, 12, and 16 weeks.
NPAD and PPT measurements were collected. All 3
groups showed decline in neck pain and disability by
VAS. All 3 groups showed an increase in PPT. The
placebo group did not reach pain-free status as much as
the BTX-A groups when given a second injection.
Wheeler17 concluded that although there was no statis-
tical significance in BTX-A over placebo, some subjects
were asymptomatic after second injection suggesting
possible benefit with sequential injections and accumu-
lating dosages.
Of the 145 subjects initially enrolled by Gobel, 24
were excluded from final analysis due to major protocol
deviations. Outcome measures included proportion of
subjects with mild or no pain at week 5, changes in pain
intensity, number of pain-free days per week, pain on
muscle palpation at each follow-up visit, and GAS. At
the end of the study, subject and physician preference for
repeated treatment was also determined. These mea-
surements were recorded at 4, 8, and 12 weeks. Gobel
noted that at week 5, significantly more patients with
Dysport injections had mild or no pain (51%). BTX-A
injection resulted in significantly greater change from
baseline in pain intensity from weeks 5 to 8. Fewer days
with pain per week were noted from week 5 to 12 in the
BTX-A group. No significant differences in duration of
daily pain or duration of sleep were noted. Mean pain
intensity score for all MTrPs was significantly lower in
the BTX-A group at week 4 until the conclusion of the
study. GAS rating favored BTX-A at weeks 4, 8, and 12.
Patients and physicians were significantly in favor of
repeated treatments with BTX-A. Gobel concluded that
subjects with upper back MPS treated with BTX-A had
significantly improved pain levels 4 to 6 weeks after
treatment.
All 30 subjects recruited by Qerama were available
for final evaluation at 3 days and 28 days after treat-
ment. Spontaneous and evoked pain by NRS, shoulder
movements, and interference pattern of electromyogra-
phy (EMG) during maximal voluntary contraction of
the infraspinatous muscle were measured. MEP activity
and interference pattern on EMG were reduced in the
BTX-A group. No change in PPT or pain was noted in
comparison with saline placebo. Qerama concluded that
the results did not support an anti-nociceptive or
analgesic role for BTX-A.
All 29 subjects enrolled in the Lew’s study were
available at 2, 4, 8, 12, 16, and 24 months. VAS, Neck
Disability Index (NDI), and SF-36 were collected. VAS
and NDI demonstrated a trend toward improvement but
were not significant. Statistically significant improve-
ments in bodily pain at 2 and 4 months and SF-36 at
1 month were noted, and statistically significant
improvements in VAS and NDI were seen in all groups
compared with baseline. These results led to the
conclusion by Lew that although this study had limited
power and population base, the trend toward improve-
ment justified further studies.
Botulinum Toxin in the Treatment for Myofascial Pain � 7
Complications
Ojala reported no significant difference in prevalence of
side effects between saline and BTX-A groups. Almost
half (48%, n = 15) after the first injections, compared
with 1 quarter (26%, n = 8) after the second injections
reported some side effects, most of which were minor
and short-lived. Pain at the injection site was commonly
reported. Other side effects included vertigo (1 after
placebo and 2 after BTX-A), sweating (1 after BTX-A),
fatigue of the hands (1 after placebo and 1 after BTX-A),
headache (2 after BTX-A), and swelling of the eyelids (1
after placebo).
Three subjects in the Ferrante’s trial who received
BTX-A injections experienced flu-like symptoms, which
were transient and resolved during the course of the
study. No other adverse events (AEs) occurred.
Wheeler16 reported differences across groups with
regards to AE scores. Follow-up comparisons demon-
strated that more AEs were reported in the BTX-A
group compared with saline placebo during the 4th and
8th weeks. The most frequent AEs were weakness of the
injected muscles, pain or soreness of the injection sites,
and flu-like symptoms.
Wheeler17 reported mild AEs in the BTX-A group.
Two subjects reported transient ipsilateral arm heavi-
ness and numbness, which resolved with 1 week.
Additionally, 2 subjects noted transient discomfort
opposite the injection site. Two other subjects reported
a shift of their pain, and 1 to a mirror-image muscular
area. Moreover, 1 subject reported a shift of neck pain to
the midline.
Gobel reported at total of 65 AEs during the study (46
in 31 subjects in the Dysport group and 19 in 11 subjects
in the placebo group). Most AEs were mild or moderate
(48/64, data unavailable for 1 AE). The most common
AE was muscle soreness in both groups. No serious
adverse events (SAEs) were reported. Qerema reported
no severe or persistent AEs. Nine subjects in the BTX-A
group and 8 in the placebo group denied any AEs.
Headaches and neck pain were reported in 4 subjects in
the BTX-A group and 3 in the control group. Lew did
not report any SAEs although further description was
not provided.
DISCUSSION
The last decade has seen an exponential increase in the
use of BTX to treat MPS. The number of clinical
applications of BTX-A has also expanded, albeit with
varying results. Much of this variability in clinical
outcomes might be partly explained by the different
chemical makeup of various BTX preparations. The 3
aforementioned preparations have similar mechanisms
of action although their chemical formulation, clinical
potency, migration, diffusion, and safety profile are
different. This may result in problems of clinical and
economic bioequivalence.21 Dysport (Ispen Biopharm
Ltd.) is less expensive and therefore more economically
competitive. Compared with Botox (Allergan, Inc.),
Xeomin (Merz Pharma) and MyoBloc (Solstice Neuro-
sciences, Inc. [botulinum toxin type B]) have been shown
to have slightly lower efficacy.22
Due to the heterogeneous nature of each study, our
review of BTX in the treatment for MPS produced
mixed results. The vast heterogeneity of exclusion
criteria, concomitant interventions, diagnostic criteria,
patient population size, volume injected, injection tech-
niques, short-duration studies, and outcome measures
render any conclusions exceedingly difficult. The major-
ity of studies found negative results. The negative studies
were relatively similar with regard to their quality.
Wheeler16,17 scored the lowest in quality on both the
AHRQ and Cochrane assessment scales. These trials,
however, did adequately describe the randomization
and blinding process. The 2001 study did not control for
co-interventions, while the 1998 study lacked similar
groups in the study arms with regard to the most
important prognostic indicators. Wheeler17 included
subjects who were involved in automobile accidents
with a higher number of these subjects represented in the
BTX-A group. Furthermore, the 1998 study included
statistician supervision of the medical assistant in the
preparation of the study injections. Only 1 study, Gobel
et al., demonstrated a significant improvement in pain
levels 4 to 6 weeks after treatment. This particular study
also scored the highest in study quality on both the
AHRQ and Cochrane assessment scales.
Overall, the exclusion criteria were not uniform
between the studies. Ferrante et al. based their exclusion
criteria mainly on the number of totalMTrPs (restricting
to fewer than 5) and their location. This restricted the
study population to a milder form of MPS. Gobel et al.
required that patients have at least 10 MTrPs to be
included in the study, thereby creating a study popula-
tion with a more severe form of MPS compared with the
study population used by Ferrante.
In addition, the majority of the studies did not
standardize subjects’ concurrent analgesic medications
during the study period. Wheeler et al.17 excluded
8 � DESAI ET AL.
anesthetic or corticosteroid injections, while others
allowed patients to receive these injections during the
study period. Gobel et al. was the only trial that
controlled for co-interventions by excluding common
analgesic medications such as NSAIDS, opioids, and
parenteral or oral corticosteroids. Ferrante controlled
for co-interventions by ensuring that all enrolled
patients were maintained on a common analgesic
regimen. The analgesic regimen included ibuprofen
800 mg 4 times daily, amitriptyline, acetaminophen
prn (as needed), and PT. In addition, Ferrante’s was the
only study to utilize PT during the study period.
There were also considerable dissimilarities in the
diagnostic criteria of MPS and MTrP identification.
A majority of the studies used duration of pain for
enrollment rather than actual MPS diagnostic criteria.
Furthermore, the duration of pain varied greatly among
the studies; some studies required pain duration of
greater than 2 months, while others required pain of
greater than 6 months prior to enrollment. Ojala et al.’s
was the only trial that used a definition of MPS as a part
of their inclusion criteria. The definition that was
utilized by Ojala was the proposed diagnostic criteria
of MPS from the study by Tunks et al. which included
(1) a regional pain complaint, (2) pain or paresthesia in
the typical distribution of the MTrP, (3) a TB in the
muscle, (4) exquisite tenderness in the TB, and (5)
restricted range of movement in the affected muscle.
Criteria 1 to 4 had to be met by all patients. The subjects
also needed one of the following minor criteria: (1)
reproduction of pain complaint by pressure on the
MTrP, (2) LTR in the TB and (3) alleviation of theMTrP
by stretch.
There are several factors which lead to reasonable
deduction that some percentage of treated MTrPs in
these studies were in fact tender points, including the
lack of uniformity in applying MPS diagnostic criteria,
the lack of comments on how MPS was diagnosed, and
the known existence of inter-rater unreliability in
diagnosing MPS. The significant difference in treatment
of these discrete clinical entities (i.e., MTrPs) makes it
vital to pursue their accurate diagnosis. An MTrP is a
hyperirritable point in skeletal muscle that is associated
with a hypersensitive palpable nodule.23 Compression
of an MTrP causes referred pain to a distant region. In
contrast, a tender point only causes pain locally.24
Tender points are areas of tenderness occurring in
muscle, muscle–tendon junction, bursa, or fat pads.
When tender points occur in a widespread manner, they
are usually considered characteristic of fibromyalgia.25
Although there is no known cure for fibromyalgia,
symptoms and function may improve with patient
education, aerobic exercise, cognitive behavioral
therapy, and pharmacologic therapies, such as serotonin
norepinephrine reuptake inhibitors and alpha 2-delta
receptor ligands.26 Other treatments used for tender
points include acupuncture and counterstrain, which is
an osteopathic manipulation technique.
In addition to dissimilarities in diagnostic criteria, the
number of subjects included and volume injected also
varied greatly among the 7 studies. Only 2 studies had
greater than 100 patients in their trial.15,18 This begs the
question of whether the studies were adequately pow-
ered. In fact, the majority of studies enrolled < 50
subjects. Compounded by the profound variations in
other selection criteria, this issue significantly calls into
question the ability to generalize conclusions drawn
from these works. The volume injected also varied
greatly. BTX-A concentrations ranged from 15 U to
400 U. The smallest volumes were used by Ojala et al.
while the largest units were used by Gobel. Ojala’s was a
crossover study of 4 weeks only.
Injection techniques varied among the studies
reviewed. A myriad of injection techniques exists, and
there is considerable debate regarding optimal methods.
Less debatable is the fact that correct placement of the
injection is key to therapeutic efficacy. One method that
is fairly well accepted is needle insertion approximately 1
to 1.5 cm away from the MTrP. This technique facili-
tates the advancement of the needle into the MTrP at a
30-degree angle. The grasping fingers isolate the TB and
prevent it from rolling out of the trajectory of the needle.
A “fast-in, fast-out” technique should be used to elicit an
LTR, which can be described as a characteristic response
ofMTrPs.23 TBs are groups ofmuscle fibers that are hard
and painful on palpation.24 LTR is a brisk contraction of
the muscle fibers in and around the TB elicited by
snapping palpation or rapid insertion of a needle into the
MTrP.23 This LTR has been shown to predict the
effectiveness of the MTrP injection, which can help a
clinician identify a true MTrP.27 The efficacy of injec-
tions, however, is highly dependent of the clinician’s skill
to localize the active MTrP with a small needle.23
Outcome measures were also quite variable among
the studies reviewed. For example, outcome measures in
the Ojala’s study included SNP and PPT, whereas
Ferrante used VAS, PPT, and the SF-36. Wheeler et al.16
used the NPAD, SF-36, and BDI. This heterogeneity
among outcome measures makes it difficult to draw
conclusions based on the results.
Botulinum Toxin in the Treatment for Myofascial Pain � 9
Most significantly, none of the 7 studies used a true
placebo-control group. A true placebo group involves a
sham injection. In the studies reviewed, normal saline
was injected into the members of the control group. It is
important to understand the difference between inject-
ing normal saline, sham injections, and TDN. It has been
suggested that injection of normal saline can have
therapeutic value. The saline injections may have
resulted in some clinical improvement in the control
group. Although a sham injection is an ideal placebo, its
use may not be practical in this setting. There was no
sedation of subjects in these studies, so subjects in the
control group might have realized that they received a
sham injection instead of BTX-A. This awareness by the
subjects would make the use of a sham injection
technically challenging.
Kamanli et al.28 investigated the differences between
efficacies of local injection of lidocaine, TDN, and low-
dose injections of BTX-A. In this study, injection of local
anesthetics, treatment with TDN, and the injection of
BTX-A had therapeutic efficacies of different degrees in
the MTrPs of all patients. When looking at the thera-
peutic effect on MTrPs, all of the variables measured by
VAS and disability scores showed no significant differ-
ence between lidocaine and BTX-A injection. However,
when compared with TDN, lidocaine injection was
more effective than BTX-A. PPT and pain scores
significantly improved in all 3 groups. Kamanli con-
cluded that injection of lidocaine is more practical and
rapid than TDN because it causes fewer disturbances
during the injection. A lidocaine injection is more cost
effective than BTX-A injection andmay be the treatment
of choice in MPS. However, BTX-A could be of use in
patients with MPS resistant to conventional treat-
ments.28
In addition, Kamanli found that to increase efficacy,
injections should be complemented with a strict PT
program as opposed to repeating injections.28 There is a
potential for lasting therapeutic efficacy when BTX is
combined with PT. In the 7 studies cited in this article,
Ferrante’s was the only study to utilize PT. Wheeler
et al.16 concluded that a single BTX-A injection session
without PT is not an effective treatment for chronic neck
pain. It is possible that repeated injections of BTX-A
have an additive effect and would be a more plausible
treatment for MPS. Future research should be aimed at
studying the effectiveness of repeat dosing of BTX-A
combined with PT.
Hong investigated the effects of injection with a local
anesthetic agent or TDN into an MTrP of the upper
trapezius muscle.27 Patients treated with TDN had a
greater intensity and longer duration of postinjection
soreness compared with those treated with a 0.5%
lidocaine injection. Hong concluded that it is essential to
elicit LTRs during injection to obtain an immediate and
desirable effect.
Overall, none of the studies used a true placebo-
control group. Therefore, the negative outcomes of the
studies may not demonstrate true failure of the BTX-A
interventions. Given this analysis, we believe that these
studies do not answer the question of whether or not
BTX-A is an effective treatment modality for MPS. The
heterogeneity of the various study parameters prevents
pooling the data from all 7 studies to perform a
meaningful meta-analysis. More research needs to be
performed to investigate BTX-A as a treatment for MPS.
CONCLUSIONS
Although of short duration, our analysis identified
Gobel et al.’s study as the highest quality study among
these prospectively randomized investigations. This was
due to appropriate identification of diagnostic criteria,
excellent study design, and objective endpoints. The
other studies had significant failings due to deficiencies
in 1 or more major criteria. It is readily apparent that
higher quality, rigorously standardized studies are
needed to more appropriately investigate this promising
treatment modality.
REFERENCES
1. Simons DG, Travell JG, Simons LS. Travell & Simons’s
Myofascial Pain and Dysfunction: The Trigger Point Manual,
vol 1, 2 ed. Baltimore: Williams & Wilkins, 1999.
2. Borg-Stein J. Treatment of fibromyalgia, myofascial
pain, and related disorders. Phys Med Rehabil Clin N Am.
2006;17:491–510, viii. Review.
3. Tough EA, White AR, Cummings TM, Richards SH,
Campbell JL. Variability of Criteria Used to Diagnose Myo-
fascial Trigger Point Pain Syndrome – Evidence from Review
of the Literature. Clin J Pain 2007;23:278–286.4. Gerwin RD. A study of 96 subjects examined for both
fibromyalgia and myofacial pain. J Musculoskeletal Pain.
1995;13(Suppl 1):121–125.5. Strine TW, Hootman JM. US national prevalence and
correlates of low back and neck pain among adults. Arthritis
Rheum 2007;57:656–665.6. Shah JP, Danoff JV, Desai MJ, et al. Biochemicals
associated with pain and inflammation are elevated in sites
near to and remote from active myofascial trigger points. Arch
Phys Med Rehabil 2008;89:16–23.
10 � DESAI ET AL.
7. Ferguson LW, Gerwin R. Clinical Mastery in the
Treatment of Myofascial Pain, 1st ed. Philadelphia: Lippincott
Williams & Wilkins; 2004.
8. Simons DG. Review of enigmatic MTrPs as a common
cause of enigmatic musculoskeletal pain and dysfunction.
J Electromyogr Kinesiol 2004;14:95–107.9. Pickett A. Re-engineering clostridial neurotoxins for
the treatment of chronic pain: current status and future
prospects. BioDrugs. 2010;24:173–182.10. Daemen MA, Kurvers HA, Kitslaar PJ, et al. Neuro-
genic inflammation in an animal model of neuropathic pain.
Neurol Res 1998;20:41–45.11. Cheshire WP, Abashian SW, Mann JD. Botulinum
toxin in the treatment of myofascial pain syndrome. Pain
1994;59:65–69.12. Moore RA, Gavaghan D, Tramer MR, Collins SL,
McQuayHJ. Size is everything – larger amounts of information
we need to overcome random effects in estimating direction
and magnitude of treatment effects. Pain 1998;78:209–216.13. van Tulder M, Assendelft W, Koes B, Bouter LM.
Method guidelines for systematic reviews in the Cochrane
Collaboration Back Review Group for Spinal Disorders. Spine.
1997;22:2323–2330.14. Ojala T, Arokoski JP, Partanen J. The effect of small
doses of botulinum toxin a on neck-shoulder myofascial pain
syndrome: a double-blind, randomized, and controlled cross-
over trial. Clin J Pain 2006;22:90–96.15. Ferrante FM, Bearn L, Rothrock R, King L. Evidence
against trigger point injection technique for the treatment of
cervicothoracic myofascial pain with botulinum toxin type A.
Anesthesiology 2005;103:377–383.16. WheelerAH,GoolkasianP,GretzSS.BotulinumtoxinA
for the treatment of chronic neck pain. Pain 2001;94:255–260.17. Wheeler AH, Goolkasian P, Gretz SS. A randomized,
double-blind, prospective pilot study of botulinum toxin
injection for refractory, unilateral, cervicothoracic, paraspinal,
myofascial pain syndrome. Spine. 1998;23:1662–1666; dis-
cussion 1667.
18. Gobel H, Heinze A, Reichel G, Hefter H, Benecke R.
Dysport myofascial pain study group. Efficacy and safety of a
single botulinum type-A toxin complex treatment (Dysport)
for the relief of upper back myofascial pain syndrome: results
from a randomized double-blind placebo-controlled multicen-
tre study. Pain 2006;125:82–88.19. Qerama E, Fuglsang-Frederiksen A, Kasch H. A
double-blind, controlled study of botulinum toxin A in chronic
myofascial pain. Neurology. 2006;67:241–245.20. Lew HL, Lee EH, Castaneda A, Klima R, Date E.
Therapeutic use of botulinum toxin type A in treating neck and
upper-back pain of myofascial origin: a pilot study. Arch Phys
Med Rehabil 2008;89:75–80.21. Sławek J, Car H, Bonikowski M, et al. Are botulinum
toxin type A preparations really the same medication? A
comparison of three botulinum toxin A for variations in
labelled neurological indications Neurol Neurochir Pol
2010;44:43–64. Review.
22. Beylot C. Different botulinum toxins and their speci-
fications. Ann Dermatol Venereol 2009;136(Suppl 4):S77–S85. Review.
23. Lavelle ED, Lavelle W, Smith HS. Myofascial trigger
points. Anesthesiol Clin. 2007;25:841–851, viiiii. Review.
24. Yap EC. Myofascial pain–an overview. Ann Acad Med
Singapore 2007;36:43–48.25. Borg-Stein J, Stein J. Trigger points and tender points:
one and the same? Does injection treatment help? Rheum Dis
Clin North Am 1996;22:305–322.26. Smith HS, Barkin RL. Fibromyalgia syndrome: a
discussion of the syndrome and pharmacotherapy. Am J Ther
2010;17:418–439.27. Hong CZ. Lidocaine injection versus dry needling to
myofascial trigger point. The importance of the local twitch
response. Am J Phys Med Rehabil 1994;73:256–263.28. Kamanli A, Kaya A, Ardicoglu O, Ozgocmen S, Zengin
FO, Bayik Y. Comparison of lidocaine injection, botulinum
toxin injection, and dry needling to trigger points in myofascial
pain syndrome. Rheumatol Int 2005;25:604–611.
Botulinum Toxin in the Treatment for Myofascial Pain � 11