European Journal of Pain 7 (2003) 449–456

www.EuropeanJournalPain.com

Effect of muscle relaxants on experimental jaw-muscle pain andjaw-stretch reflexes: a double-blind and placebo-controlled trial

Peter Svensson a,b,c,*, Kelun Wang b, Lars Arendt-Nielsen b

a Department of Clinical Oral Physiology, Dental School, Aarhus University, DK-8000 Aarhus C, Denmarkb Center for Sensory-Motor Interaction, Orofacial Pain Laboratory, Aalborg University, DK-9220 Aalborg E, Denmark

c Department of Oral and Maxillofacial Surgery, Aarhus University Hospital, DK-8000 Aarhus C, Denmark

Received 26 April 2002; accepted 3 February 2003

Abstract

A randomised, double-blind, placebo-controlled three-way cross-over study was performed to investigate the effect of two muscle

relaxants (tolperisone hydrochloride and pridinol mesilate) on experimental jaw-muscle pain and jaw-stretch reflexes. Fifteen

healthy men participated in three randomised sessions separated by at least 1 week. In each session 300mg tolperisone, 8mg pridinol

mesilate or placebo was administered orally as a single dose. One hour after drug administration 0.3ml hypertonic saline (5.8%) was

injected into the right masseter to produce muscle pain. Subjects continuously rated their perceived pain intensity on an electronic

10-cm visual analogue scale (VAS). The pressure pain threshold (PPT) was measured and short-latency reflex responses were evoked

in the pre-contracted (15% maximal voluntary contraction) masseter and temporalis muscles by a standardised stretch device (1mm

displacement, 10ms ramp time) before (baseline), 1 h after medication (post-drug), during ongoing experimental muscle pain (pain-

post-drug), and 15min after pain had vanished (post-pain). Analysis of variance demonstrated significantly lower VAS peak pain

scores (5:9� 0:4 cm) after administration of tolperisone hydrochloride compared with pridinol mesilate (6:8� 0:4 cm) and placebo

(6:6� 0:4 cm) ðP ¼ 0:020Þ. Administration of pridinol mesilate was associated with a significant decrease in PPTs compared with

tolperisone hydrochloride and placebo ðP ¼ 0:002Þ after medication, but not after experimental jaw-muscle pain. The normalised

peak-to-peak amplitude of the stretch reflexes were not significantly influenced by the test medication ðP ¼ 0:762Þ, but were in all

sessions significantly facilitated during ongoing experimental jaw-muscle pain ðP ¼ 0:034Þ. In conclusion, tolperisone hydrochloride

provides a small, albeit significant reduction in the perceived intensity of experimental jaw-muscle pain whereas the present dose had

no effect on the short-latency jaw-stretch reflex.

� 2003 European Federation of Chapters of the International Association for the Study of Pain. Published by Elsevier Science Ltd.

All rights reserved.

Keywords: Tolperisone; Pridinol mesilate; Randomised clinical trial; Temporomandibular disorders; Experimental muscle pain; Stretch reflex;

Pressure pain threshold

1. Introduction

Current pharmacological management of chronicmusculoskeletal pain disorders is to a large extent based

on clinical experience rather than on scientific evidence

and randomised clinical trials (RCTs) (Curatolo and

Bogduk, 2001). Moreover, the pathophysiological

mechanisms of chronic pain remain unknown and it can

therefore be difficult to select the most rational strategy

* Corresponding author. Fax: +45-8619-5665.

E-mail address: psvensson@odont.au.dk (P. Svensson).

1090-3801/$30 � 2003 European Federation of Chapters of the Internationa

All rights reserved.

doi:10.1016/S1090-3801(03)00013-2

to alleviate pain. A common and popular, yet unproved,

hypothesis of many musculoskeletal pain disorders in-

cluding the temporomandibular disorders (TMDs) is theso-called vicious cycle where activation of nociceptive

pathways is thought to lead to muscle hyperactivity and

spasms which again will cause more pain and subse-

quently setting up a positive feedback loop (Laskin,

1969; Travell et al., 1942; Travell, 1960). The vicious

cycle concept has been challenged since experimental

and clinical data generally do not support such a simple

reciprocal link between pain and muscle hyperactivity(for a review see Lund et al., 1991 and Svensson and

Graven-Nielsen, 2001). Nevertheless, therapies aimed to

l Association for the Study of Pain. Published by Elsevier Science Ltd.

Fig. 1. Schematic outline of the experimental protocol. PPT¼pressure

pain thresholds, VAS¼ visual analogue scale, MPQ¼McGill Pain

Questionnaire.

450 P. Svensson et al. / European Journal of Pain 7 (2003) 449–456

reduce muscle hyperactivity and relax tense musclescontinue to be advocated, e.g., the use of systemic

muscle relaxants.

A recent review of pharmacological management of

musculoskeletal pain concluded that systemic muscle

relaxants only showed limited effectiveness for chronic

neck pain and chronic low back pain for up to 4 weeks

(Moulin, 2001). Furthermore, another review on phar-

macological management stated that there is little sup-porting evidence in favour of muscle relaxants in chronic

TMD pain conditions (Dionne, 1997). For example, a

double-blind study, could not demonstrate a better effect

of carisoprodol compared with placebo in 60 patients

with chronic myofascial pain in the orofacial region

(Gallardo et al., 1975). However, it was pointed out that

there could be possible exceptions since some muscle

relaxants appeared to be more effective than placebo inmanagement of chronic musculoskeletal pain disorders

(for a review see Dionne, 1997).

Tolperisone hydrochloride (Mydocalm , Strathmann

AG, Germany) is a centrally acting muscle relaxant with

a chemical structure that partially resembles that of li-

docaine (Fels, 1996; Zsila et al., 2000). Thus, in addition

to the well-documented inhibitory effect of tolperisone

hydrochloride on spinal motoneurones and mono- andpolysynaptic reflex pathways in animal preparations

(Farkas et al., 1989; Kaneko et al., 1987; Ono et al., 1984;

Sakitama, 1993), clinical experience and some studies

have indicated beneficial effects on various painful

musculoskeletal pain disorders (Amir et al., 1978; Csa-

nyi, 1989; Inovay and Katona, 1991; Pratzel et al., 1996).

However, alleviation of pain could possibly also be re-

lated to the membrane stabilising potency of tolperisonehydrochloride (Ono et al., 1984). Preliminary electro-

physiological studies on cultured hippocampus neurons

have, indeed, indicated that tolperisone hydrochloride

can block voltage-gated neuronal sodium-channels in a

dose dependent manner and reduce the sodium influx

(Farkas et al., 2000). Another centrally acting muscle

relaxant – pridinol mesilate (Myoson, IPG Pharma,

Germany) – appears to have an inhibitory effect on reflexactivity of spinal motoneurones via an interaction with

the M1 and M4 subtype of the muscarinic receptor

(Keim, Mutschler and Lambrecht – unpublished obser-

vations). Furthermore, there is preliminary evidence

from a clinical trial that administration of pridinol

mesilate is associated with improvement in joint function

and pain scores in patients with chronic low back pain

(Pipino et al., 1991). Nevertheless, it is not clear howclosely the pain relieving effect is coupled to the effect of

systemic muscle relaxants on the excitability of reflex

pathways. We have recently shown a facilitation of the

short-latency stretch reflex in the masseter muscle when

exposed to experimental jaw-muscle pain (Svensson

et al., 2001; Wang et al., 2000, 2001). However, the

mechanisms for this facilitation are not clear but changes

in the fusimotor control linked to the nociceptive barragehave been suggested (Pedersen et al., 1997). We therefore

speculated that the jaw-stretch reflex and the facilitation

by pain would be decreased by administration of cen-

trally acting muscle relaxants.

Thus, the aim of the present study was to investigate

the effect of two centrally acting muscle relaxants on

both experimentally evoked jaw-muscle pain and mea-

sures of trigeminal reflex excitability in a double-blindand placebo-controlled trial.

2. Materials and methods

2.1. Subjects

The volunteers for this study were all healthy andunmedicated subjects without signs or symptoms of

TMD (Dworkin and LeResche, 1992). Fifteen men aged

between 22 and 29 years (mean age� SEM: 25:6� 0:6years) participated in this study. The study was con-

ducted in accordance with the Helsinki Declaration and

written informed consent was obtained from all subjects.

The local Ethics Committee and the Danish National

Board of Health approved the study.

2.2. Experimental protocol

All subjects participated in three sessions separated

by at least 1 week (Fig. 1). In each session, the pressure

pain thresholds (PPT) and jaw-stretch reflexes were re-

corded at baseline. Then the test medication (tolperisone

hydrochloride, pridinol mesilate or placebo) was ad-ministered orally as a single dose and after 60min the

PPT and jaw-stretch reflexes were recorded again (post-

drug). Subsequently experimental jaw-muscle pain was

evoked by injection of hypertonic saline and the subjects

rated the perceived pain intensity on visual analogue

scales (VAS). Jaw-stretch reflexes were recorded during

the ongoing pain (pain-post-drug). The subjects also

used a McGilll Pain Questionnaire (MPQ) to describethe quality of the saline-evoked pain. Finally, the PPT

and jaw-stretch reflexes were recorded when the subjects

had reported no pain for about 15min (post-pain).

P. Svensson et al. / European Journal of Pain 7 (2003) 449–456 451

2.3. Test medication

The study was performed as a randomised, double-

blind and placebo-controlled trial. The placebo tablets

were identical to the active medication in terms of their

size, colour, shape and smell. There were two different

kinds of placebo tablets, one for pridinol mesilate and

one for tolperisone hydrochloride. Thus, to guarantee

the blinding conditions each volunteers received a totalof 4 tablets at each session (double dummy design): two

active (4mg pridinol mesilate or 150mg tolperisone

hydrochloride) and two corresponding placebo tablets

or four placebo tablets.

The randomisation code table was prepared and

kept by the drug company (Strathmann AG, Germany)

with a copy in a sealed envelope at the study centre in

case of an emergency situation. After completion of theexperimental part of the study the randomisation code

for the sequence of the different test medications for

each subject was opened. The statistical analysis was

then performed in three different groups marked A, B

and C and first then the final randomisation code was

broken.

2.4. Experimental jaw-muscle pain

Injection of 0.3ml hypertonic (5.8%) sterile saline

into the mid portion of the masseter muscle followed

the previously described procedures (Svensson et al.,

1998, 2001). The subjects continuously scored the per-

ceived pain intensity on an electronic 0-10 cm visual

analogue scale (VAS) with the lower extreme labelled

‘‘no pain’’ and the upper extreme labelled ‘‘most painimaginable’’. A computer sampled the VAS pain scores

every 5 s. After stop of the infusion the subjects were

furthermore asked to describe the quality of pain on a

Danish version of the McGill Pain Questionnaire

(MPQ). The pain rating indices (PRIs) of the sensory,

affective, evaluative and miscellaneous dimension of

pain were calculated in accordance with Melzack

(1975). The VAS peak pain score was for each subjectdefined as the highest value on the time – VAS curve

and the pain duration as the difference between the

VAS onset and VAS offset.

2.5. Pressure pain threshold

A pressure algometer (Somedic, Sweden) was used to

measure the PPTs. The PPT was defined as the amountof pressure (kPa), which the subjects first perceived to be

painful (Svensson et al., 1995). The probe with 1 cm

diameter was applied perpendicular to the central part

of the left and right masseter and temporalis muscles

(MAL, MAR, TAL and TAR). During the pressure

stimulation, the subjects kept their teeth in the intercu-

spal position with minimum voluntary contraction and

focused their attention on the experimental task. Thesubject pushed a button to stop the pressure stimulation

when the threshold was reached. The PPTs were deter-

mined in triplicate with a constant application rate of

30 kPa/s. There was a time lag of about 2min between

repeated measurements. The mean value was used for

further statistical analysis.

2.6. Jaw-stretch reflexes

Stretch reflexes were evoked in the jaw-closing mus-

cles with a standardised muscle stretcher (1mm dis-

placement, 10ms ramp time), which previously has been

described in details (Miles et al., 1993; Svensson et al.,

2000, 2001; Wang and Svensson, 2001; Wang et al.,

2000, 2001). In brief, bipolar disposable surface elec-

trodes (Neuroline 720-01-K, Medicotest, Ølstykke,Denmark) were placed bilaterally on the central part of

the masseter and anterior temporalis muscles (MAL,

MAR, TAL and TAR). The skin over the recording

positions was cleaned with alcohol and the impedance of

the electrodes was kept < 20kX. A ground electrode

soaked with saline was attached to the left wrist. The

EMG signals were amplified 2000–5000 times (Coun-

terpoint MK-2, Dantec, Denmark), filtered with band-pass 20Hz to 1 kHz, sampled at 4 kHz and stored for

later analysis. The subjects performed three maximal

clenches each lasting up to 3 s on the bar with their in-

cisor teeth to obtain the mean EMG value of the max-

imal voluntary contraction (MVC) in the four muscles.

The MVC calculated from the EMG activity of right

masseter (painful side) was used to construct a window

of 10% below and above the 15% MVC (i.e., 13.5–16.5%MVC). On-line calculation of the root-mean-square

(RMS) value in 200ms intervals of the EMG was per-

formed with the use of LabView. The subjects received

visual feedback from markers on a computer screen,

which changed from green to red upon crossing the

upper and lower limits of the window (Svensson et al.,

1998). If the ongoing EMG activity remained within the

pre-set window for more than 400ms, the program au-tomatically triggered the muscle stretcher. A total of

300ms EMG activity was recorded with 100ms pre-

stimulus and 200ms post-stimulus. Twenty trials with

an inter-stimulus interval about 10 s were recorded in

each condition (baseline, post-drug, pain-post-drug,

post-pain). In the pain-post-drug condition, the reflex

recordings were started 20–30 s after the injection of

hypertonic saline (Fig. 1).A special-purpose computer program processed the

reflex responses evoked in the EMG by the fast stret-

ches. First, the mean EMG in the pre-stimulus interval

()100 to 0ms) of the averaged and rectified signal was

calculated (lV). Then, the peak-to-peak amplitude

of the early reflex component, which appeared as a

biphasic potential in the average of the non-rectified

452 P. Svensson et al. / European Journal of Pain 7 (2003) 449–456

recordings, was measured (lV). Finally, the peak-to-peak amplitude was normalised with respect to the mean

pre-stimulus EMG activity, i.e., the ratio between the

amplitude and pre-stimulus EMG activity (Wang et al.,

2000).

2.7. Statistics

VAS peak pain scores, pain duration and MPQscores produced by injection of hypertonic saline were

compared with analyses of variance (ANOVA) with

repeated measures and one factor: test medication (three

levels: tolperisone hydrochloride, pridinol mesilate and

placebo). PPTs were analysed with a three-way ANOVA

with repeated measures and followed by pairwise mul-

tiple comparison procedures (Tukey tests). The factors

in the ANOVA were: test medication (three levels),muscle (four levels: MAL, MAR, TAL and TAR) and

Fig. 2. Perceived pain rated by the subjects ðN ¼ 15Þ on visual analogue

scales (VAS) following injection of hypertonic saline into the masseter

muscle in the three sessions with administration of tolperisone hydro-

chloride, pridinol mesilate or placebo. Mean values and SEM.

Table 1

PRI(S) PRI(A

Tolperisone hydrochloride 10:6� 1:6 1:0�Pridinol mesilate 12:0� 1:9 2:1�Placebo 11:6� 1:7 1:3�

ANOVA tests P ¼ 0:623 P ¼ 0

Pain rating indices (PRI) of sensory (S), affective (A), evaluative (E) an

tionnaire. ANOVA tests did not find significant differences between the test

time of measurement (three levels: baseline, post-drugand post-pain). The pre-stimulus EMG activity and

normalised jaw-stretch reflexes were analysed in a sim-

ilar way: test medication (three levels), muscle (four

levels) and time of measurement (four levels: baseline,

post-drug, pain-post-drug and post-pain). Furthermore,

as a measure of the intrasubject variability across the

three experimental sessions, the coefficient of variation

(CV ¼ standard deviation/mean in percentage) wascalculated for the different outcome parameters. The

significance level was set at P < 0:05. Mean val-

ues� SEM are presented in the text and figures.

3. Results

In total, the test medication was administered to 16subjects. One subject withdrew from the experiment

during first session (pridinol mesilate) due to moderate

dizziness. The symptom spontaneously disappeared after

6 h. Fifteen subjects finished all three sessions of the study.

Two out of the 15 subjects experienced slight dizziness

during the sessions with pridinol mesilate and placebo,

respectively. However, the relation to administration of

the test medication was unlikely according to the twosubjects and the experiments could be continued.

3.1. Experimental jaw-muscle pain

Injection of 0.3ml hypertonic saline produced in all

subjects a moderate to strong jaw-muscle pain, which

was centred on the masseter muscle and spreading to-

wards the temporomandibular joint, temple and teeth.

The pain reached its peak intensity after 20–30 s and

then gradually decreased lasting up to 600 s (Fig. 2). The

pain was consistently (> 30% of the subjects) described

on the MPQ as ‘‘hurting’’ (42%), ‘‘pressing’’ (40%),‘‘boring’’ (38%), ‘‘penetrating’’ (36%), ‘‘tingling’’ (31%)

and ‘‘intense’’ (31%). Less frequent words used to de-

scribe pain were ‘‘sickening’’ (29%), ‘‘shooting’’ (29%),

‘‘piercing’’ (27%), ‘‘taut’’ (27%), ‘‘tender’’ (24%), ‘‘nag-

ging’’ (24%) and ‘‘tiring’’ (22%). The pain rating indices

(PRIs) derived from the MPQ are shown in Table 1. The

PRIs were not significantly influenced by the test

medication (ANOVA: F < 2:551; P > 0:096). Theintraindividual coefficient of variation (CV) for the

) PRI(E) PRI(M)

0:4 1:7� 0:4 4:3� 0:7

0:7 1:7� 0:4 4:8� 0:7

0:5 2:5� 0:5 4:4� 0:9

:096 P ¼ 0:099 P ¼ 0:842

d miscellaneous (M) dimension of pain from the McGill Pain Ques-

medications.

Fig. 4. Ratios between peak-to-peak amplitudes of the jaw-stretch

reflexes and the pre-stimulus EMG activity at baseline, 60min after

administration of the three test medications (post-drug), during on-

going jaw-muscle pain (pain-post-drug), and after experimental pain

(post-pain). Mean values (SEM) from 15 subjects. The values from

four muscles (left and right masseter and temporalis) are pooled since

there were no significant interactions between muscles and test medi-

cation ðF ¼ 0:693; P ¼ 0:656Þ. *Indicates significantly higher values

for the pain-post-drug condition compared with baseline and the post-

drug condition (Tukey: P < 0:05).

P. Svensson et al. / European Journal of Pain 7 (2003) 449–456 453

PRIs were S ¼ 29� 4%, A ¼ 65� 18%, E ¼ 64� 19%and M ¼ 51� 11%.

There was a significant difference in VAS peak pain

scores following the test medications (ANOVA:

F ¼ 4:490; P ¼ 0:020). Administration of tolperisone

hydrochloride was associated with significantly lower

VAS peak pain scores (5:9� 0:4 cm) compared with

pridinol mesilate (6:8� 0:4 cm) and placebo (6:6� 0:4cm) (Tukey: P < 0:05). The duration of the saline-evoked jaw-muscle pain in the tolperisone hydrochloride

session (417� 27 s), pridinol mesilate session (370� 28 s)

and placebo session (429� 26 s) was not significantly

different (F ¼ 2:06; P ¼ 0:146). The intraindividual CV

for the VAS peak pain score was 12� 2% and for the

pain duration 20� 3%.

3.2. Pressure pain thresholds

Analyses of variance of the PPTs showed that there

was no main effect of any test medication (F ¼ 0:326;P ¼ 0:725) or time of measurement (F ¼ 0:534;P ¼ 0:592), but a significant difference between muscles

(F ¼ 6:894; P < 0:001). However, there was a significant

interaction between test medication and time of mea-

surement (F ¼ 4:968; P ¼ 0:002) and post hoc testsdemonstrated significantly lower PPTs following ad-

ministration of pridinol mesilate compared to tolperi-

sone hydrochloride and placebo only in the post-drug

condition (Tukey: P < 0:05) (Fig. 3). The intraindivid-

ual CVs for the four jaw-muscles were MAL ¼ 10� 2%,

MAR ¼ 12� 2%, TAL ¼ 13� 3% and TAR ¼10� 2%.

3.3. Jaw-stretch reflexes

The fast stretch of the contracting jaw-muscles

evoked an early (8–9ms) biphasic response in the EMG

in all subjects in accordance with previous reports

(Svensson et al., 2001; Wang et al., 2000, 2001).

Fig. 3. Pressure pain thresholds at baseline, 1 h after administration of

the three test medications (post-drug) and after experimental pain

(post-pain). Mean values (SEM) from 15 subjects. The PPTs from four

muscles (left and right masseter and temporalis) are pooled since there

were no significant interactions between muscles and test medication

(F ¼ 1:121; P ¼ 0:357). � Indicates significantly lower PPTs compared

to tolperisone hydrochloride and placebo in the post-drug condition

(Tukey: P < 0:05).

Analyses of variance of the pre-stimulus EMG ac-

tivity showed that there was no main effect of any test

medication (F ¼ 1:932; P ¼ 0:146 or time of measure-

ment (F ¼ 0:184; P ¼ 0:907), but a significant difference

between muscles (F ¼ 127:6; P < 0:001). Analyses of the

peak-to-peak amplitude normalised with respect to pre-

stimulus EMG activity (i.e., ratio) indicated no signifi-cant main effect of test medication (F ¼ 0:275;P ¼ 0:762) or muscle (F ¼ 1:200; P ¼ 0:321) but a sig-

nificant effect of time of measurement (F ¼ 3:174;P ¼ 0:034). Post hoc tests demonstrated significantly

higher amplitudes during ongoing jaw-muscle pain

(pain-post-drug) compared with baseline and the post-

drug conditions (Tukey: P < 0:05) (Fig. 4). No signifi-

cant interactions between other factors in the ANOVAwere identified (F < 1:831; P > 0:069). The intraindi-

vidual CVs for the pre-stimulus EMG activity were

MAL ¼ 23� 4%, MAR ¼ 13� 3%, TAL ¼ 28� 4%

and TAR ¼ 22� 4% and for the normalised peak-to-

peak-amplitudes MAL ¼ 22� 3%, MAR ¼ 24� 6%,

TAL ¼ 34� 4% and TAR ¼ 45� 7%.

4. Discussion

The main finding in this experimental study was a

marginal hypoalgesic effect of tolperisone hydrochloride

on the perceived pain intensity of acute, experimental

jaw-muscle pain whereas there were no significant effects

of any test medication on measures of trigeminal reflex

excitability.

4.1. Muscle relaxants and muscle pain

Tolperisone hydrochloride has previously been as-

sessed for its efficacy on pain in the craniofacial region

454 P. Svensson et al. / European Journal of Pain 7 (2003) 449–456

(Amir et al., 1978). In the open study, which was per-formed on a small group of TMD patients (n ¼ 15),

150mg tolperisone hydrochloride for 4 weeks was found

to relieve pain in the majority of patients (88%) (Amir

et al., 1978). Similar results were also observed following

carisoprodol and occlusal splints and it was subse-

quently advocated that muscle relaxants could be effec-

tive for relieving TMD-related pain (Amir et al., 1978).

A double-blind and placebo-controlled study could,however, not demonstrate any significant differences

between carisoprodol and placebo in patients with

TMD-related pain (Gallardo et al., 1975). Dionne

(1997) noted that there appeared to be a discrepancy

between the common clinical use of systemic muscle

relaxants for relieve of musculoskeletal pain and the

results of randomised placebo-controlled clinical trials.

Furthermore, nonspecific central nervous system de-pression could, in part, account for a hypoalgesic effect

of muscle relaxants (Dionne, 1997). Tolperisone hy-

drochloride, however, appears to be a rather unique

muscle relaxants because few central nervous system

side effects are noted even at high doses (Dulin et al.,

1998). Daily administration of 150 or 450mg tolperisone

hydrochloride for 8 days did not impair a battery of

psychomotor tests including reaction time measure-ments in healthy subjects, which suggests that sedation

is not a major problem at clinical relevant doses. Fur-

thermore, few side effects like headache and fatigue (four

and two out of 24 subjects, respectively) were noted with

450mg tolperisone hydrochloride (Dulin et al., 1998).

This is in accordance with the present study with 300mg

tolperisone hydrochloride as a single dose without sig-

nificant side effects. It is therefore unlikely that non-specific central nervous system effects can account for

the observed hypoalgesic effect of tolperisone hydro-

chloride in the present study.

A recent RCT in patients with a clinical diagnosis of

painful reflex muscle spasms appear to support the no-

tion of a weak hypoalgesic effect of 300mg tolperisone

hydrochloride (Pratzel et al., 1996). In that study,

muscle spasms were defined as involuntary musclecontractions and local tenderness on manual palpation.

Although no operationalised criteria or direct electro-

physiological measurements of the muscle spasms were

presented it is interesting to note that treatment with

tolperisone hydrochloride compared with placebo after

10 and 21 days was associated with a significantly higher

increase in PPTs whereas no significant difference could

be found in the manually palpated muscle tone. In thepresent study no significant effects on PPTs in the jaw-

muscles could be observed following administration of

tolperisone hydrochloride. The reasons for these find-

ings are not known, but it could be, that multiple dosing

of tolperisone hydrochloride is needed in order to in-

crease PPTs as it was shown in the clinical trial of

Pratzel et al. (1996). Concerning pridinol mesilate a

significant reduction of the PPT compared with placeboand tolperisone was found. This effect of pridinol mes-

ilate was not expected and so far there is no explanation

for it. Nevertheless, the differential actions of tolperi-

sone hydrochloride and pridinol mesilate on VAS peak

pain scores and PPTs (Figs. 2 and 3) in the present study

support the notion of different receptor mechanisms for

the two types of muscle relaxants (see Section 1).

In the present study there were no significant effectsof injection of hypertonic saline on the PPTs (Fig. 3).

This finding is in accordance with our previous studies

using hypertonic saline injections (Graven-Nielsen et al.,

1998a,b; Svensson et al., 1995) and suggests a lack of

sensitisation to mechanical stimuli when assessed with

PPTs on the local pain area. In contrast to hypertonic

saline, injection of the excitatory amino acid glutamate

into muscle tissue is associated with a significant de-crease in mechanical threshold both in animal prepara-

tions (Cairns et al., 2002) and in human studies

(Svensson et al., 2003). Thus, further studies may be

needed to examine the hypoalgesic effect versus the anti-

allodynic and anti-hyperalgesic properties of for exam-

ple muscle relaxants.

In summary, there is limited experimental and clinical

evidence in favour of a pronounced hypoalgesic effect ofsystemic muscle relaxants, but a small effect could be

detected for tolperisone hydrochloride in a highly stan-

dardised acute muscle pain model.

4.2. Muscle relaxants and jaw-reflexes

Both tolperisone hydrochloride and pridinol mesilate

have a strong and well-documented effect on spinal re-flex circuits and motoneurones when assessed in animal

preparations (Farkas et al., 1989; Ito et al., 1985; Ka-

neko et al., 1987; Ono et al., 1984; Sakitama, 1993). The

depression of spinal synaptic transmission by tolperi-

sone hydrochloride may be related to both pre-synaptic

and post-synaptic mechanisms in addition to a direct

membrane stabilising effect on primary afferent fibres

(Ono et al., 1984). However, less is known about thetrigeminal reflex excitability in human subjects. Takata

et al. (1996) demonstrated in an open study on five

healthy subjects that 100mg tolperisone hydrochloride

was associated with a marked depression of the tonic

vibration reflex measured in the masseter muscles. Vi-

bration (e.g., for 20 s) is a potent stimulus of primary

muscle spindles leading to a tonic excitation of moto-

neurones and activity in the EMG related to the dura-tion of the stimulus (Desmedt and Godaux, 1975). A

single rapid stretch (10ms) is also adequate to activate

primary muscle spindles and generate a short-latency

reflex (8–9ms) response in the pre-contracted masseter

muscle (Miles et al., 1993; Poliakov and Miles, 1994;

Wang et al., 2000; Wang and Svensson, 2001). We have

previously shown a facilitation of the short-latency

P. Svensson et al. / European Journal of Pain 7 (2003) 449–456 455

stretch reflex in the masseter muscle when exposed toexperimental jaw-muscle pain (Svensson et al., 2001;

Wang et al., 2000, 2001) in accordance with a significant

facilitation of the stretch reflex in the soleus muscle

during experimental pain (Matre et al., 1998). The

mechanisms for this facilitation are not entirely clear but

changes in the fusimotor control linked to the nocicep-

tive barrage have been suggested (Pedersen et al., 1997).

We had anticipated that the centrally acting muscle re-laxants would suppress the amplitude of the jaw-stretch

reflex response by similar mechanisms described above

for the spinal reflex circuits but this was not observed in

the present study. Furthermore, it was speculated that

the facilitation of the jaw-stretch reflex response by ex-

perimental muscle pain would be antagonised with the

concurrent medication of tolperisone hydrochloride or

pridinol mesilate. In accordance with our previousstudies on tonic jaw-muscle pain (Svensson et al., 2001;

Wang et al., 2000, 2001), the present more phasic type of

jaw-muscle pain was in all the sessions associated with a

significant increase in the normalised peak-to-peak am-

plitude of the jaw-stretch reflex, however, there were no

significant differences between the sessions with muscle

relaxants or placebo. One possible explanation for the

lack of effect of tolperisone hydrochloride could be thatthe jaw-stretch reflex is recorded during a slight (15%

MVC) voluntary, pre-contraction of the masseter mus-

cle, which could mask an effect on the trigeminal reflex

circuits in agreement with the suggestion of Romaniello

et al. (2000). However, this and our previous studies on

jaw-stretch reflexes (Wang et al., 2000, 2001; Svensson

et al., 2001) have all been able to demonstrate a signif-

icant pain-related increase in the reflex amplitude underthese experimental conditions. Moreover, in the present

study we found no significant differences in the pre-

stimulus EMG activity at any time point during the

experiments. Thus, we do not believe that methodolog-

ical concerns are the main reason for the lack of effect of

tolperisone hydrochloride on the jaw-stretch reflex.

Another possibility is that the doses and timing of the

medication in relation to the recording of the jaw-stretchreflexes were not optimal. Even if the average peak

plasma concentration of tolperisone has been identified

after 1.5 h (Miskolczi et al., 1987) and the elimination

half time is reported to be around 2.5 h (Dulin et al.,

1998) it is also known, that tolperisone hydrochloride

has a high interindividual variability of pharmacokinetic

parameters like peak plasma concentration and elimi-

nation half time.Finally, the variability of the outcome measures in

experimental pain studies needs to be considered (Yar-

nitsky et al., 1996). In the present study we employed

techniques, which previously have been sufficiently

sensitive to detect effects of experimental manipulations

in paired designs (Svensson et al., 1995, 2001). The in-

traindividual coefficient of variations (CV) for the PPTs

and saline-evoked pain were generally low (< 20%) andwithin the range of CVs for other psychophysical tech-

niques (Svensson et al., 1991; Yarnitsky et al., 1996).

However, the CVs for MPQ measures were higher

(29–65%) and could, in part, explain the lack of any

significant differences between the three sessions. The

CVs for the jaw-reflex measures were moderately low for

the masseter muscles (13–24%) but higher for the tem-

poralis muscles (22–45%) suggesting difficulties to findsignificant differences in the temporalis muscles. Never-

theless, a significant effect of jaw-muscle pain was found

on the normalised peak-to-peak amplitude, which indi-

cates that the present technique is adequate to detect

consistent effects in a paired design.

In conclusion, a single dose of 300mg tolperisone

hydrochloride provides a small, albeit significant re-

duction in the perceived intensity of experimental jaw-muscle pain but did not show a significant effect on the

short-latency jaw-stretch reflex in the present set-up.

Acknowledgements

The Danish National Research Foundation sup-ported the study and Strathmann AG, Germany pro-

vided the test medication.

References

Amir A, Sakuda M, Miyazaki T. Comparative effects of muscle

relaxants and occlusal bite plate in treatment of temporomandib-

ular joint dysfunction syndrome. J Osaka Univ Dent Sch

1978;18:1–10.

Cairns BE, Gambarota G, Svensson P, Arendt-Nielsen L, Berde CB.

Glutamate-induced sensitization of rat masseter muscle fibers.

Neuroscience 2002;109:389–99.

Csanyi L. High-dose Mydocalm therapy in certain myogenous

headaches. Ther Hung 1989;37:115–8.

Curatolo M, Bogduk N. Pharmacologic pain treatment of musculo-

skeletal disorders: current perspectives and future. Clin J Pain

2001;17:25–32.

Desmedt JE, Godaux E. Vibration-induced discharge patterns of single

motor units in the masseter muscle in man. J Physiol (Lond)

1975;253:429–42.

Dionne RA. Pharmacologic treatments for temporomandibular disor-

ders. Oral Surg Oral Med Oral Pathol Oral Radiol Endod

1997;83:134–42.

Dulin J, Kov�aacs L, Ramm S, Horvath F, Ebeling L, Kohnen R.

Evaluation of sedative effects of single and repeated doses of 50mg

and 150mg tolperisone hydrochloride. Results of a prospective

randomised double-blind placebo-controlled trial. Pharmacopsy-

chiatry 1998;31:137–42.

Dworkin SF, LeResche L. Research diagnostic criteria for temporo-

mandibular disorders. Review, criteria, examinations and specifi-

cations, critique. J Craniomandib Disord Facial Oral Pain

1992;6:301–55.

Farkas S, Tarnawa I, Berzsenyi P. Effects of some centrally acting

muscle relaxants on spinal root potentials: a comparative study.

Neuropharmacology 1989;28:161–73.

456 P. Svensson et al. / European Journal of Pain 7 (2003) 449–456

Farkas S, Kocsis P, Bielik N, Nagy J, Tarnawa I. Tolperisone, a

sodium channel inhibitor centrally acting muscle relaxant. In: 20th

European Winter Conference on Brain Research, Villars sur Ollon,

Switzerland, 2000, p. 126.

Fels G. Tolperisone: evaluation of the lidocaine-like activity by

molecular modelling. Arch Pharm Pharm Med Chem

1996;329:171–8.

Gallardo F, Molgo J, Miyazaki C, Rossi E. Carisoprodol in the

treatment of myofascial pain-dysfunction syndrome. J Oral Surg

1975;33:655–8.

Graven-Nielsen T, Babenko V, Svensson P, Arendt-Nielsen L.

Experimentally-induced muscle pain induces hypoalgesia in het-

erotopic deep tissues, but not in homotopic deep tissues. Brain Res

1998;787:203–10.

Graven-Nielsen T, Fenger-Grøn LS, Svensson P, Steengaard-Pedersen

K, Arendt-Nielsen L, Jensen TS. Quantification of deep and

superficial sensibility in saline-induced muscle pain – A psycho-

physical study. Somatosens Motor Res 1998;15:46–53.

Inovay J, Katona J. Several years of observation with Mydeton in the

treatment of lockjaw. Ther Hung 1991;39:185–7.

Ito T, Hori K, Karasawa T, Kadokawa T. Pharmacological studies of

1-(2,3-dimethyl-4-methoxyphenyl)-2-methyl-3-(1-pyrrolidinyl)-1-prop-

anone hydrochloridee (AD-2239), a centrally acting muscle relax-

ant. Arch Int Pharmacodyn 1985;275:105–22.

Kaneko T, Ono H, Fukuda H. Simultaneous evaluation of drug effects

on both the spinal cord and the descending pathways in rats. Arch

Int Pharmacodyn 1987;287:203–10.

Laskin DM. Etiology of the pain-dysfunction syndrome. J Am Dent

Assoc 1969;79:141–53.

Lund JP, Donga R, Widmer CG, Stohler CS. The pain-adaptation

model: a discussion of the relationship between chronic musculo-

skeletal pain and motor activity. Cand J Physiol Pharmacol

1991;69:683–94.

Matre DA, Sinkjaer T, Svensson P, Arendt-Nielsen L. Experimental

muscle pain increases the human stretch reflex. Pain 1998;75:331–9.

Melzack R. The McGill Pain Questionnaire: major properties and

scoring methods. Pain 1975;1:277–300.

Miles TS, Poliakov AV, Flavel SC. An apparatus for controlled stretch

of human jaw-closing muscles. J Neurosci Methods 1993;46:

197–202.

Miskolczi P, Vereczkey L, Frenkl R. Gas–liquid chromatographic

method for the determination of tolperisone in human plasma:

pharmacokinetic and comparative bioavailability studies. J Pharm

Biomed Anal 1987;5:695–700.

Moulin DE. Systematic drug treatment for chronic musculoskeletal

pain. Clin J Pain 2001;17:86–93.

Ono H, Fukuda H, Kudo Y. Mechanisms of depressant action of

muscle relaxants on spinal reflexes: participation of membrane

stabilizing action. J Pharm Dyn 1984;7:171–6.

Pedersen J, Sj€oolander P, Wenngren BI, Johansson H. Increased

intramuscular concentration of bradykinin increases the static

fusimotor drive to muscle spindles in neck muscles of the cat. Pain

1997;70:83–91.

Pipino F, Menarini C, Lombardi G, Guerzoni P, Ferrini A, Pizzoli A,

Grangie A, Beltrame A, Sorbilli G, Gottardo R, Cilento F. A direct

myotonolytic (Pridinol Mesilate) for the management of chronic

low back pain: a multicentre, comparative clinical evaluation. Eur J

Clin Res 1991;1:55–70.

Poliakov AV, Miles TS. Stretch reflexes in human masseter. J Physiol

(Lond) 1994;476:323–31.

Pratzel HG, Alken RG, Ramm S. Efficacy and tolerance of repeated

oral doses of tolperisone hydrochloridee in the treatment of painful

reflex muscle spasm: results of a prospective placebo-controlled

double-blind trial. Pain 1996;67:417–25.

Romaniello A, Cruccu G, McMillan AS, Arendt-Nielsen L, Svensson

P. Effect of experimental pain from trigeminal muscle and skin on

motor cortex excitability in humans. Brain Res 2000;882:120–7.

Sakitama K. The effects of centrally acting muscle relaxants on the

intrathecal noradrenaline-induced facilitation of the flexor reflex

mediated by group II afferent fibers in rats. Jpn J Pharmacol

1993;63:369–76.

Svensson P, Arendt-Nielsen L, Nielsen H, Larsen JK. Effect of chronic

and experimental jaw muscle pain on pain–pressure thresholds and

stimulus–response curves. J Orofac Pain 1995;9:347–56.

Svensson P, Bjerring P, Arendt-Nielsen L, Kaaber S. Variability of

argon laser-induced sensory and pain thresholds on human oral

mucosa and skin. Anesth Prog 1991;38:79–83.

Svensson P, Cairns BE, Wang K, Hu JW, Graven-Nielsen T,

Arendt-Nielsen L, Sessle BJ. Glutamate-induced sensitization of

pressure pain thresholds in human masseter muscle. Pain

2003;101:221–7.

Svensson P, De Laat A, Graven-Nielsen T, Arendt-Nielsen L.

Experimental jaw-muscle pain does not change heteronymous H-

reflexes in the human temporalis muscle. Exp Brain Res

1998;121:311–8.

Svensson P, Graven-Nielsen T. Craniofacial muscle pain: review of

mechanisms and clinical manifestations. J Orofac Pain 2001;15:

117–45.

Svensson P, Macaluso GM, De Laat A, Wang K. Effects of local and

remote muscle pain on human jaw reflexes evoked by fast stretches

at different clenching levels. Exp Brain Res 2001;139:495–502.

Svensson P, Miles TS, Graven-Nielsen T, Arendt-Nielsen L. Modu-

lation of stretch-evoked reflexes in single motor units in human

masseter muscle by experimental pain. Exp Brain Res 2000;132:

65–71.

Takata Y, Nakajima T, Yamada Y. Quantitative evaluation of the

tonic vibration reflex (TVR) in the masseter muscle. J Oral

Maxillofac Surg 1996;54:1307–13.

Travell J. Temporomandibular joint dysfunction. J Prosthet Dent

1960;10:745–63.

Travell J, Rinzler S, Herman M. Pain and disability of the shoulder

and arm. Treatment by intramuscular infiltration with procaine

hydrochloridee. J Am Med Assoc 1942;120:417–22.

Wang K, Svensson P. Influence of methodological parameters on

human jaw-stretch reflexes. Eur J Oral Sci 2001;109:86–94.

Wang K, Arendt-Nielsen L, Svensson P. Excitatory actions of

experimental muscle pain on early and late components of jaw-

stretch reflexes. Arch Oral Biol 2001;46:433–42.

Wang K, Svensson P, Arendt-Nielsen. Effect of tonic muscle pain

on short-latency jaw-stretch reflexes in humans. Pain 2000;88:

189–97.

Yarnitsky D, Sprecher E, Zaslansky R, Hemli JA. Multiple session

experimental pain measurement. Pain 1996;67:327–33.

Zsila F, Holl�oosi M, Gergely A. Stereochemical study of tolperisone, a

muscle relaxant agent, by circular dichroism and ultraviolet

spectroscopy. Chirality 2000;12:720–6.