doi: 10.1111/j.1472-8206.2004.00277.

ORIGINAL New metabolic and pharmacokinetic

ARTICLE
characteristics of thiocolchicoside and its
active metabolite in healthy humans
M. Trellua*, A. Filali-Ansarya, D. Françonb, R. Adamb, P. Lluelc,
C. Dubruca, J.P. Thénota
a
Sanofi-Synthélabo, Metabolim and Pharmacokinetic Department, 1 Av. Pierre Brossolette 91385 Chilly Mazarin Cedex,
France
b
Neurophysiology SNC Department, 31 Av. Paul Vaillant Couturier 92200 Bagneux Cedex, France
c
Mature Products Department, 82 Avenue Raspail 94255 Gentilly Cedex, France

Keywords:
bioequivalence,
metabolism,
muscle relaxant,
pharmacokinetics
Received 26 April 2004;
revised 29 June 2004;
1 accepted 5 July 2004
*Correspondence and reprints:
marc.trellu@sanofi-
synthelabo.com
ABSTRACT
Thiocolchicoside (TCC) has been prescribed for several years as a muscle relaxant
drug, but its pharmacokinetic (PK) profile and metabolism still remain largely
unknown. Therefore, we re-investigated its metabolism and PK, and we assessed the
muscle relaxant properties of its metabolites. After oral administration of 8 mg (a
therapeutic dose) of 14C-labelled TCC to healthy volunteers, we found no detectable
TCC in plasma, urine or faeces. On the other hand, the aglycone derivative obtained
after de-glycosylation of TCC (M2) was observed and, in addition, we identified, as the
major circulating metabolic entity, 3-O-glucuronidated aglycone (M1) obtained after
glucuro-conjugation of M2. One hour after oral administration, M1 plus M2
accounted for more than 75% of the circulating total radioactivity. The pharmaco-
logical activity of these metabolites was assessed using a rat model, the muscle
relaxant activity of M1 was similar to that of TCC whereas M2 was devoid of any
activity. Subsequently, to investigate the PK profile of TCC in human PK studies, we
developed and validated a specific bioanalytical method that combines liquid
chromatography and ultraviolet detection to assay both active entities. After oral
administration, TCC was not quantifiable with an lower limit of quantification set at
1 ng/mL, whereas its active metabolite M1 was detected. M1 appeared rapidly in
plasma (tmax ¼ 1 h) and was eliminated with an apparent terminal half-life of 7.3 h.
In contrast, after intramuscular administration both active entities (TCC and M1)
were present; TCC was rapidly absorbed (tmax ¼ 0.4 h) and eliminated with an
apparent terminal half-life of 1.5 h. M1 concentration peaked at 5 h and this
metabolite was eliminated with an apparent terminal half-life of 8.6 h. As TCC and
M1 present an equipotent pharmacological activity, the relative oral pharmacological
bioavailability of TCC vs. intramuscular administration was calculated and
represented 25%. Therefore, to correctly investigate the PK and bioequivalence of
TCC, the biological samples obtained must be assayed with a bioanalytical method
able to specifically analyse TCC and its active metabolite M1.

methoxy group with a thiomethyl group. This compound

INTRODUCTION
Thiocolchicoside (TCC) is a semi-synthetic derivative of
colchicoside (a naturally occurring glucoside present in
the Gloriosa Superba plant) derived by substitution of a
has been used for over 35 years as a myorelaxant in the
treatment of painful muscle contractions in acute and
chronic rheumatic conditions, in traumatology and
especially in patients with acute low back pain [1]. In

2004 Blackwell Publishing Fundamental & Clinical Pharmacology 18 (2004) 493–501 493
494
addition, this drug possesses analgesic and anti-inflam-
matory properties observed in animal models and its
good efficacy and its safety profile have been demon-
strated clinically in several studies [2,3]. Despite the fact
that it has been approved for marketing for many years,
its metabolism and pharmacokinetic profile (PK) have
not been completely investigated.
It was reported, that after oral administration, TCC is
rapidly absorbed from the gastrointestinal tract, with peak
plasma concentrations occurring within 1 h and with a
terminal half-life of 2–6 h [4–6]. However, as Sutherland
et al. [7] recently pointed out, the analytical methods used
in these studies, namely, radioimmunoassay (RIA) [4,5]
and gas chromatography coupled with mass spectrometry
(MS) after enzymatic hydrolysis of TCC to its aglycone,
3-demethylthiocolchicine [6], do not differentiate the
parent compound from potential pharmacologically act-
ive or inactive metabolites. Moreover, using a liquid
chromatography-MS-MS method, Sutherland et al. found
only traces of the parent drug in plasma of healthy
volunteers given 8 mg orally (the recommended thera-
peutic dose). In that study, concentrations were all below
0.30 ng/mL, the lower limit of quantification (LLOQ),
apart from the first sampling point, where it was approxi-
mately equal to the LLOQ. On the other hand,
these authors did find significant amounts of
3-demethylthiocolchicine, the aglycone metabolite of
TCC (M2), and they suggested that this substance was
the major TCC metabolite produced after oral administra-
tion of the drug. As a consequence, they argued that the
previously published PK data on TCC were incorrect.
However, Sutherland et al. did not investigate the phar-
macological activity of the aglycone, nor did they rule out
the possibility that there may be other TCC metabolites
that are pharmacologically active.
Given the limitations of the available data, we re-
investigated the metabolism, muscle relaxant properties
and PK of TCC. Our aims were (i) to investigate the
metabolism of TCC in healthy humans, (ii) to assess the
muscle relaxant properties of TCC and any other circula-
ting entities, and (iii) to investigate the PK of TCC using a
new bioanalytical method, one that is specific both for TCC
and any potentially active metabolites, that we developed.
MATERIALS AND METHODS
2 Tri-Carb Liquid scintillation Analyser model 1600TR
Metabolism study
Clinical protocol
Six healthy male volunteers (ages 18–35 years) were
included after giving their written informed participation

2004 Blackwell Publishing Fundamental & Clinical Pharmacology 18 (2004) 493–501
M. Trellu et al.
H3C
CH2OH
O O CH3
HO
O
HO O
OH
S
CH3
O
NH
O
*
CH3
14
Figure 1 Structure of C-thiocolchicoside (TCC). *14C-label
position.
consent. The study was conducted in agreement with the
declaration of Helsinki and the Good Clinical Practices.
The study started after the Ethical Review Board’s
approval. Each subject received an 8 mg oral dose of
14
C-labelled TCC (synthesized by Sanofi-Synthélabo,
Chilly Mazarin Cedex, France), containing up to 2 MBq
of 14C-labelled TCC (50 lCi) on the first day of the study
(Figure 1). Blood, urine, faecal and expired air samples
were collected at regular intervals over at least 120 h
after administration of the 14C-labelled drug; blood and
urine were also collected prior to administration of the
drug and at regular intervals between 0.25 and 120 h
postdose. Blood samples were centrifuged and the
resulting plasma stored at )20 C until analysed. Urine
fractions were kept in a refrigerator until the complete
fraction had been obtained. Samples of each fraction
were used for radioactivity assay and metabolic profiling.
Stools were collected from day 1 until at least 120 h after
drug intake and stored at )20 C until assayed. Before
being assayed, each stool was thawed and homogenized
with water. Radioactivity was measured in one aliquot
and the others were used for metabolic profiling. Expired
air samples were collected during the first 24 h imme-
diately after the corresponding blood sample had been
drawn and stored at 2–8 C until assayed.
Radioactivity measurements
Plasma, urine and faecal samples were analysed for
radioactivity by liquid scintillation counting with a
(Packard Instrument, Rungis, France), either directly or
after combustion of the sample into CO2. Expired CO2
traps were assayed once, urine and plasma samples in
triplicate, and faecal homogenates in quadruplicate.
 Metabolism and pharmacokinetic of thiocolchicoside
Identification of 14C-labelled TCC and its metabolites
in plasma, urine and faeces
14
C-labelled TCC and its metabolites in plasma, urine and
faeces were quantified by high-pressure liquid chroma-
3 tography (HPLC) after separation on a Kromasil C18,
5 lm, 250 · 4.6 mm column (Interchim, Montluçon,
France) with a Beckman 126 gradient system under
control of a microcomputer. The mobile phase A was
ammonium acetate 0.01 M, pH 5.0 (acetic acid) and
mobile phase B was CH3CN/MeOH (v/v), at a flow rate
of 1 mL/min. Compounds were detected on line with a
Diode Array Detector (model 168, Beckman Instruments
4 France, Gagny, France) and a Flow Scintillation Analyzer
(RadioMaticTM 500TR, Packard Instrument, Rungis,
France). UV detection was performed at a variable
wavelength using a Diode Array Detector at 383 nm,
and radioactivity was detected in 500 lL samples. TCC
metabolites were identified on the basis of MS analysis of
urine and faeces. HPLC/MS analysis was carried out on a
HP-MSD single-stage quadripole instrument (Hewlett-
Packard, Parc de la Vatine, France) equipped with an ion
spray interface operating in the positive mode. Structural
information and assignment of diagnostic fragments was
based on in-source collision-induced dissociation experi-
ments by varying the voltage between the end of the
transfer capillary and the first skimmer. HPLC separation
was conducted on a Kromasil C18, 150 · 2.1 mm
column at 200 lL/min with the mobile phase described
above. The structure of the metabolites in plasma was
ascertained by comparison of their retention times with
those of standards and by comparison with radiochroma-
togram profiles previously obtained with urine and faeces.
Pharmacological study
Animals and drugs
Male Sprague–Dawley rats (200–500 g) were purchased
from Charles River, France. TCC, M1 and M2 were
solubilized in distilled water with a drop of 0.05%
Tween-80 and administered intraperitoneally at doses of
3, 10 or 30 mg/kg. All experiments were performed in
accordance with the guidelines for animal experiments
and principles for the care and use of laboratory animals
established by Sanofi-Synthelabo Research Ethical com-
mittee following national and international directives.
Recording the polysynaptic reflex in normal rats
The flexor reflex was elicited by electric shocks to the
sural nerve innervation area in the right hindpaw of rats
via stainless steel needles inserted subcutaneously. The
flexor reflex was recorded as electromyogram (EMG)

2004 Blackwell Publishing Fundamental & Clinical Pharmacology 18 (2004) 493–501
495
activity via needle electrodes inserted in the ipsilateral
posterior biceps femoris/semitendinosus muscles. The
stimulus parameters were 0.2–0.5 ms, 0.2 Hz and
activated mainly Ad fibres. The stimulus intensity was
defined by the recording conditions within each experi-
ment such that the stimulus was sufficient to produce a
stable response (8–12 mA). The EMG was amplified,
filtered (10–1000 Hz), rectified and integrated within a
time-window from 8–10 until 35–40 ms after the
stimulus. The magnitude of 20 consecutive reflex
responses was averaged before (control) and every
15 min after drug administration. The magnitude of
these reflexes after drug administration was expressed as
a percentage of the mean value of the predrug control
period. To analyse the drug effects, two-way ANOVA and
Dunnett’s test were performed to determine significant
variations from control values.
Development of an HPLC/UV method for
quantitative analysis of TCC and its metabolites
Extraction procedure
5 Extraction was performed with a Gilson AspecTM XLi auto-
mate (Gilson Medical Electronics, Villiers-le-Bel, France)
and Bond Elut C18 (50 mg) cartridges (Varian, Les Ulis,
France). Each plasma sample (1 mL) was spiked with
100 ng of internal standard (IS), and then diluted with
1 mL of purified water. The diluted sample was loaded
onto a cartridge previously conditioned with 1 mL of
methanol and 1.5 mL of purified water. The cartridge was
then washed with 1.5 mL of purified water, and the test
compounds eluted with 0.5 mL of methanol. The organic
phase was evaporated under nitrogen at 37 C. The dry
extract was dissolved in 0.040 mL of methanol to which
had been added 0.160 mL of 0.025 M KH2PO4 (pH 3.6).
A 0.15 mL aliquot was injected into the HPLC system.
Analytical conditions
The test compounds were speared by HPLC using a
6 YMC Pack ODS AQ column (5 lm, 150 · 4.6 mm) (AIT
France, Le Mesnil le Roi, France) and 0.025 M of KH2PO4
(pH 3.6)/acetonitrile (80/20, v/v) as a mobile phase,
with a flow rate at 1 mL/min. A Shimadzu UV spectro-
photometer at 320 nm was used for detection.
Calculation
Quantitative analysis of the samples was performed using
Multichrom 2.21 version software (VG Instruments, Fison
7 Instruments, Arcueil, France) running on a micro-Vax
3100.80. The integration method employed to determine
chromatographic peak areas is valley to valley.
496
Peak area ratio (product to quantify/IS) plotted vs. the
nominal concentration of the product to quantify was
used to generate the linear calibration curve (least
square regression). By interpolation from the calibration
equation, the considered ratio of product/IS corresponds
to a quantity x of the product, calculated in ng/g, taking
into consideration the sample amount. This value in
ng/g was similar to ng/mL considering that plasma
density is close to 1.
Pharmacokinetic study
Clinical protocol
Sixteen healthy nonsmoking male volunteers (age
26 ± 5 years) were included after giving their written
informed participation consent. The study was conduc-
ted in agreement with the declaration of Helsinki and the
Good Clinical Practices. The study started after the
Ethical Review Board’s approval. The study medication
was supplied by Sanofi-Synthelabo as capsules contain-
ing 4 mg TCC and as a solution for parenteral admin-
istration containing 4 mg TCC in 2 mL. Each subject
received a single 8 mg dose of the study drug, either
orally or intramuscularly (i.m.) on a randomized cross-
over design during two study periods, with a 2-day
washout period between each of the study periods. Blood
samples were collected immediately before drug admin-
istration and at regular intervals over the following 24 h
(n ¼ 16 for the oral dose; n ¼ 19 for the i.m. dose);
subjects were hospitalized for collection of these samples.
Blood was centrifuged and the resulting plasma separ-
ated and stored at )20 C until analysed.
As initially planned, plasma samples were first assayed
by RIA [5] (data not shown). Then, following the
criticisms of that method reported by Sutherland et al.
[7], we developed the specific HPLC method described
above and used it to re-assay plasma samples of six of the
subjects who were included in this study; these results
are presented in the Results and Discussion sections.
The following PK parameters were derived from data
obtained with the specific HPLC/UV method: time to
reach maximum plasma concentration (tmax), maximum
plasma concentration observed (Cmax), terminal elimin-
ation half-life (t1/2z), area under the plasma concentra-
tion–time curve from 0 to the time corresponding to the
last quantifiable concentration (AUClast), area under the
plasma concentration–time curve extrapolated to infinity
(AUC), apparent volume of distribution (Vz/F) and
apparent plasma clearance (Cl/F). PK parameters were
calculated using a noncompartmental approach with
WinNonlin V 3.1 Professional Network Edition (Phar-

2004 Blackwell Publishing Fundamental & Clinical Pharmacology 18 (2004) 493–501
M. Trellu et al.
sight, Cary, NC, USA). PK parameters were presented for
TCC and its active metabolite (M1) with indication of the
treatment, together with descriptive statistics: mean, SD
and CV.
RESULTS AND DISCUSSION
Metabolism of TCC after administration of a single
oral dose of 14C-labelled TCC
Excretion balance of total radioactivity
After oral administration of 8 mg 14C-labelled TCC, this
drug was well tolerated by the six healthy volunteers and
no adverse effect was observed. Radioactivity was rapidly
detected in plasma. The elimination balance of total
radioactivity was 98.7 ± 1.8% (19.9 ± 7.0% in the
urine and 78.8 ± 7.3% in the faeces). More than 90%
of the faecal excretion of radioactivity was completed
within 72 h in three subjects but was delayed in the
three others. Radioactivity in expired air was below the
limit of detection in nearly all samples.
Metabolic profiles
The radiochromatogram in Figure 2 shows typical
examples of the metabolic profile observed in urine,
faeces and plasma in comparison with a standard
solution corresponding to the 14C-TCC solution admin-
istered. However, no trace of unchanged TCC was
detected in urine, plasma or faeces. In plasma samples,
only M1 and M2 were detected, and the same was true
for urine. In addition, the faeces contained a third
metabolite, the di-demethylthiocolchicine (M3), which
was not detected in plasma or urine.
TCC
Administered solution
M1 M2
Plasma – 0.25 h
Urine – 0–3 h
M3
Faeces – 0–24 h
Figure 2 Radiochromatograms obtained from plasma, urine and
faeces samples after a single oral dose of 14C-thiocolchicoside (TCC)
in a healthy male subject. M1 ¼ 3-O-glucuronidated aglycone,
M2 ¼ aglycone-TCC and M3 ¼ di-demethylthiocolchicine.
Metabolism and pharmacokinetic of thiocolchicoside
Quantitative evaluation of TCC metabolites in
plasma, urine and faeces
The concentrations of TCC and of each of its two major
metabolites found in plasma, expressed as the percent-
age of circulating radioactivity, were quantified after a
single oral dose of the 14C-labelled drug over the period
from 0.25 to 3 h. In all subjects, TCC was below the
limit of quantification (<1%) throughout this period. Up
to 2 h, the concentration of M1 was at least threefold
higher than the concentration of M2 at all time points.
At 3 h, M1 was the only circulating entity observed in
the plasma. No traces of unchanged compound were
detected in urine up to 36 h, M1 accounted for
8.2 ± 2.7% and M2 for 3.0 ± 0.8% of the administered
dose. In faecal samples collected over 0–168 h, no
traces of unchanged compound were detected, M1
accounted for 1.8 ± 1.4% and the aglycone (M2)
for 35.7 ± 25.5% of the administered dose. The
di-demethylthiocolchicine (M3) found in the faeces
accounted for 16.0 ± 9.2% over the same collection
interval.
Metabolic pathways
The main metabolic pathways after oral administration
of TCC are shown in Figure 3. The labelled carbon atom
localized on the amide function of 14C-TCC was retained
in all metabolites identified. Thus, the metabolism of 14C-
TCC did not involve the labelling site on the amide
function, which is also confirmed by the absence of
radioactivity in expired air.
The two main entities present in the systemic circu-
lation are M1 and M2. De-glycosylation of the admin-
Figure 3 Proposed metabolic pathway of
thiocolchicoside (TCC) after a single oral
dose of 14C-thiocolchicoside. M1 ¼
3-O-glucuronidated aglycone,
M2 ¼ aglycone-TCC and M3 ¼
di-demethylthiocolchicine.

2004 Blackwell Publishing Fundamental & Clinical Pharmacology 18 (2004) 493–501
497
istered drug results in the aglycone derivative M2, and
then glucuro-conjugation of the aglycone results in
3-O-glucuronidated aglycone M1. De-methylation of the
aglycone results in di-demethylthiocolchicine (M3).
About 38.7% of the 8 mg dose was recovered as
M2 (urine 0–36 h + faeces 0–168 h), 16.0% as
di-demethylthiocolchicine (faeces) and 10% as M1
(urine + faeces).
The present results demonstrate that TCC is not
detected in plasma samples of subjects who were
administered with a therapeutic dose of TCC orally,
confirming the report of Sutherland et al. [7]. At the
same time, we not only found the aglycone metabolite of
TCC (3-demethylthiocolchicine) but also a second, more
abundant metabolite, the glucuronidated conjugate of
the aglycone (M1).
Pharmacological activity of metabolites
The muscle relaxant activity of TCC and its two main
metabolites was investigated by recording the poly-
synaptic reflex in normal rats [8–10]. At 3 mg/kg, i.p. of
either TCC or M1, the effects on the polysynaptic reflex
were not statistically different from the effect of vehicle
(data not shown). At 10 mg/kg, i.p., TCC and M1
significantly reduced the polysynaptic reflex by about 60
and 80%, respectively (P < 0.01, Figure 4). The degree
of inhibition of this reflex induced by the two compounds
was not statistically different (P > 0.05). In contrast, the
effect of M2 on the polysynaptic reflex was not sta-
tistically different from that of the solvent and of the
control values (P > 0.05), up to the highest dose tested
(30 mg/kg, i.p., Figure 4).
498 M. Trellu et al.

Vehicle M1 10 mg/kg The results also demonstrated that the accuracy,

120 TCC 10 mg/kg M2 30 mg/kg
repeatability and intermediate precision, evaluated by
100 two analysts on 3 days, at LLOQ (1 ng/mL) and also on
80
Pre-drug period
quality control samples (2, 20 and 150 ng/mL) fulfilled
Variation (%)

60
the acceptance criteria.
**
** ** **
i.p.
40 Administration *
** * Pharmacokinetic profiles obtained with
** *
20 ** the HPLC/UV method
0 As planned in the protocol, subjects received a single
–30 –15 0 15 30 48 60 75 90
Time (min)
8 mg dose of TCC orally and a single 8 mg dose i.m. The
drug was well tolerated and no adverse events were
Figure 4 Time course of the effect of intraperitoneal administration
observed. The concentrations of TCC and M1 in the
of TCC (n), M1 (d) and M2 (m) on the polysynaptic reflex in normal
rats. Number of rats: three or four in each experiment. Abscissa: plasma samples from these subjects were assayed using
time in minutes for control ()30–0 min) and postdrug adminis- our specific HPLC/UV method.
tration periods (0–90 min). Ordinate: variation in percentage of the
area of the reflexes before and after administration of the drugs Oral administration
(100% is the mean of the areas of the 20 first reflexes; each point The PK profile (on both linear and semi-log scales) after
represents the mean area of 20 reflexes). Vertical arrow: i.p.
oral administration of the drug is shown in Figure 5.
injection. **P < 0.01, *P < 0.1, Dunnett’s test, control vs. postdrug
Again, TCC was below the limit of quantification
values. Data points after administration of TTC are not significantly
different from those after M1 administration, whereas data points
after administration of M2 are statistically different from post-TCC 70
TCC (all concentrations < LOQ)
and post-M1 values. 60 M1
Concentration (ng/mL)

50

Validation of the new HPLC/UV method for 40

investigation of TCC and its active metabolite
30
The determination of the concentration of TCC and that of
its active metabolite (M1) in human plasma was per- 20
formed using the method based on HPLC with spectro- 10
photometric UV detection after solid phase extraction.
0
Taking the limit of quantification to be the lowest
0 4 8 12
validation concentration with precision and accuracy Time (h)
within 20%, the current assay was validated from 1 to
200 ng/mL with a matrix size of 1 mL. The method is 100

specific and relative to endogenous compounds. More-

Concentration (ng/mL)

over, TCC and its metabolite were stable for up to 5 weeks

in stock solutions, and they were stable in plasma
following three freeze/thaw cycles, for up to 40 months
10
at )20 C. TCC at 2 ng and 150 ng/mL and M1 at
150 ng/mL were stable in human plasma for at least 24 h
at 37 C. Nevertheless, at a concentration of 2 ng/mL, a
loss of 27% of M1 was observed when the plasma was kept
at 37 C for 24 h. Thus, in the final analytical method, the 1
samples were extracted just after thawing at ambient 0 4 8 12
temperature. At the two concentrations studied, TCC and Time (h)
M1 were stable for 24 h at 20 ± 5 C. The test compound
9 Figure 5 Mean (+SEM) plasma concentrations of thiocolchicoside
was also stable in processed samples for 72 h on the auto- (TCC) and its active metabolite M1 obtained after administration of
sampler at 20 ± 5 C. For TCC and M1 (at 60 ng/mL), the a single 8 mg oral dose of TCC to six healthy subjects. Plasma
process efficiency of the method was 90 and 82%, samples were assayed using the HPLC/UV method described in the
respectively. text. Linear and semi-log scales.

2004 Blackwell Publishing Fundamental & Clinical Pharmacology 18 (2004) 493–501

Metabolism and pharmacokinetic of thiocolchicoside 499

200 TCC with a median tmax of 1 h, and it was eliminated with an

180 M1 apparent terminal half-life of 7.3 ± 5.1 h.
Concentration (ng/mL)

160
140 Intramuscular administration
120 The mean TCC and M1 plasma concentrations vs. time
100 obtained after a single 8 mg i.m. dose of TCC in the same
80 six subjects are shown in Figure 6. The mean PK
60
parameters of TCC and M1 observed in the plasma of
40
these subjects are presented in Table I. TCC was absorbed
20
rapidly (tmax ¼ 0.4 h) and eliminated with an apparent
0
0 4 8 12 terminal half-life of 1.5 ± 0.2 h. The concentration peak
Time (h) of M1 occurred at 5 h and it was eliminated with an
1000 apparent terminal half-life of 8.6 h. TCC concentrations
were five-fold higher than those of M1. TCC was
moderately distributed with an apparent volume of
Concentration (ng/mL)

distribution (Vz/F) of 42.7 ± 6.8 L, and its apparent

100
10
1 When administered by this route, concentrations of TCC
0 4
Time (h)
Figure 6 Mean (+SEM) plasma concentrations of thiocolchicoside
(TCC) and its active metabolite M1 obtained after administration of
a single 8 mg i.m. dose of TCC to six healthy subjects. Plasma
samples were assayed using the HPLC/UV method described in the
text. Linear and semi-log scales.
(<1 ng/mL) and only the pharmacologically active
metabolite M1 was detected. The mean PK parameters
of M1 after single-dose oral administration are presented
in Table I. M1 appeared rapidly in the subjects’ plasma,
systemic clearance was 19.2 ± 1.1 L/h.
Thus, the present data show that, after oral adminis-
tration, TCC is not detected in plasma whereas M1
appears rapidly and is eliminated with an apparent
terminal half-life of slightly <8 h. In contrast, after i.m.
administration, both TCC and M1 are detected in plasma.
8 12
are fivefold higher than those of M1. TCC is much more
rapidly excreted than M1.
We suggest that intestinal metabolism after oral
administration of TCC may explain the difference in the
PK profile observed after intramuscular administration.
The brush-border membrane of intestinal enterocytes is
known to contain a number of hydrolases, including
glycohydrolases. These intestinal enzymes may be
responsible for the hydrolysis of the sugar moiety of
TCC, this de-glycosylation leading to the formation of the
TCC aglycone derivative (M2). M2 formed via intestinal
metabolism would be the entity absorbed and then

Table I Pharmacokinetic parameters of TCC and of its active metabolite M1 after a single 8 mg oral and i.m. dose given to healthy subjects.

b
Administration route Compound t1/2z (h) tmax (h) Cmax (ng/mL) AUClast (ng.h/mL) AUC (ng.h/mL) Cl/F (L/h) Vz/F (L)

Orala M1
Mean (SD) 7.3 (5.1) 1.0 61.0 (20.3) 114 (38) 126 (46) – –
CV% 69.9 33.3 33.7 36.0
Intramuscular TCC
Mean (SD) 1.5 (0.2) 0.4 174.8 (28.3) 414 (26) 417 (25) 19.2 (1.1) 42.7 (6.8)
CV% 11.8 16.2 6.2 6.0 5.7 16.0
M1
Mean (SD) 8.6 (12.2) 5.0 11.7 (4.8) 68 (25) 83 (35) – –
CV% 142.0 41.4 36.9 42.5

a
No TCC PK parameters calculated: all TCC concentrations were below the limit of quantification (1 ng/mL).
b
Median value.

2004 Blackwell Publishing Fundamental & Clinical Pharmacology 18 (2004) 493–501

500
metabolized to its glucuronidated-conjugate (M1) during
first pass through the liver (Figure 3). The formation of
M1 cannot be due to direct oxidation of the alcohol
moiety of TCC, as TCC is not metabolized in vitro by
human and rat hepatocytes (data not shown). Thus,
after oral administration, M1 is the only active entity
observed that could reach the systemic circulation. In
contrast, after intramuscular administration, TCC does
not undergo this intestinal and/or hepatic first pass effect
and thus becomes the main circulating entity.
Because of the significant intestinal metabolism of
TCC, the relative bioavailability of TCC observed after
oral administration is extremely low when compared
with that observed after i.m. administration. However, as
TCC and M1 have equipotent pharmacological activity,
the ratio of the sum of the AUCs of the active entities
after oral administration vs. intramuscular administra-
tion was calculated. As defined by this ratio, the
pharmacological bioavailability of TCC after oral admin-
istration relative to that observed after i.m. administra-
tion is around 25%.
CONCLUSION
In conclusion, our results confirm the report of Suther-
land et al. [7] showing that the previously published
data on the metabolism and PK of TCC are incorrect,
these results having been obtained using either a
nonspecific RIA [4,5] or a GC-MS method that only
measures the aglycone moiety [6]. Moreover, even the
specific LC/MS-MS method, which quantifies the TCC
and the aglycone, did not assay all the pharmacolog-
ically active metabolic entities [7,11]. Nevertheless, we
confirmed Sutherland et al. finding that the drug is not
detectable in the plasma of healthy human subjects after
oral administration. We then went further and, for the
first time, identified the two main circulating TCC
metabolites, M1 and M2 that are produced after oral
administration. One hour after oral administration, they
accounted together for more than 75% of the circulating
total radioactivity. We then demonstrated that of these
two metabolites, only M1 is at least as active as a muscle
relaxant compared with the parent compound, whereas
the aglycone (M2) is not active at all.
Using the specific HPLC/UV bioanalytical method
that we developed and validated, allowed us to quantify
both of its pharmacologically active entities (TCC and
M1), we showed that after oral administration of TCC
to humans, M1 but not TCC was observed in the
systemic circulation and, therefore, it is responsible for

2004 Blackwell Publishing Fundamental & Clinical Pharmacology 18 (2004) 493–501
M. Trellu et al.
the muscle relaxant activity of the parent compound.
The most likely explanation is that TCC is hydrolysed in
the intestine and the corresponding aglycone (M2) is
then metabolized as a result of the hepatic first pass
effect into its glucuronidated conjugate M1. In contrast,
after intramuscular administration, TCC is not subjected
to this intestinal and/or hepatic first pass effect and
both the parent compound and M1 are detected in the
plasma. This new method allowed us to accurately
determine the PK profiles of M1 and TCC after both
routes of administration of the drug, thereby elimin-
ating the discrepancies that had been reported previ-
ously, in particular, the differences in terminal half-life
values between the oral route (5 h) and the i.m. route
(2.7 h) [5,12] and in time-related changes in the PK of
TCC (a decrease in total clearance) after repeated i.m.
administration [12]. These discrepancies could be due
to the fact that the active entities differ according to the
route of administration and also to the fact that RIA
does not discriminate the unchanged parent compound
from its metabolites.
We therefore recommend that in order to perform PK
studies of TCC, including determination of bioequiva-
lence, the biological samples must be assayed with a
bioanalytical method that can specifically assay both
TCC and its pharmacologically active glucuronidated
conjugate metabolite.
REFERENCES
1 Tuzun F., Unalan H., Oner N. et al. Multicenter, randomized,
double-blinded, placebo-controlled trial of thiocolchicoside
in acute low back pain. Joint Bone Spine (2003) 70
356–361.
2 Pietrogrande V., Cherubino P., Peretti G. et al. Studio poli-
centrico randomizzato, doppio cieco: Muscoril capsule e fiale vs.
Pbo in patologie ortopedico-traumatologiche. Ort. Traum. Oggl.
8 (1992) XII 132–137.
3 Marcel C., Rezvani Y., Revel M. Evaluation du thiocolchicoside
en monothérapie dans le lumbago douloureux. La Presse Méd.
(1990) 19 1133–1136.
4 Sandouk P., Bouvier d’Yvoire M., Chretien P., Tillement J.P.,
Scherrmann J.M. Bioavailablity of oral and intramuscular
thiocolchicoside in healthy volunteers. Biopharm. Drug Dispos.
(1994) 15 87–92.
5 Sandouk P., Chappey O., Bouvier d’Yvoire M., Schermann J.M.
Pharmacokinetics of thiocolchicoside in humans using a specific
radioimmunoassay. Ther. Drug Monit. (1995) 17 544–548.
6 Perucca E., Poitou P., Pifferi G. Comparative pharmacokinetics
and bioavailablity of two oral formulations of thiocolchicoside, a
GABA-mimetic muscle relaxant drug, in normal volunteers.
Eur. J. Drug. Metab. Pharmacokinet. (1995) 20 301–305.
Metabolism and pharmacokinetic of thiocolchicoside
7 Sutherland F.W.C., Smit M.J., Herbst L. et al. Highly specific and
sensitive liquid chromatography-tandem-mass spectrometry
method for the determination of 3-desmethylthiocolchicine in
human plasma as analyte for the assessment of bioequivalence
after oral administration of thiocolchicoside. J. Chromatogr. A
(2002) 949 71–77.
8 Chen D.F., Bianchetti M., Wiesendanger M. The adrenergic
agonist tizanidine has differential effects on flexor reflexes of
intact and spinalized rat. Neuroscience (1987) 23 641–647.
9 Turski L., Stephens D.N. Effect of the beta-carboline Abecarnil
on spinal reflexes in mice and on muscle tone in genetically
spastic rats: a comparison with diazepam. J. Pharmacol. Exp.
Ther. (1993) 267 1215–1220.

2004 Blackwell Publishing Fundamental & Clinical Pharmacology 18 (2004) 493–501
501
10 Advokat C., Duke M. and Zeringue R. Dissociation of ())
Baclofen-induced effects on the tail withdrawal and hindlimb
flexor reflexes of chronic spinal rats. Pharmacol. Biochem.
Behav. (1999) 63 527–534.
11 Ferrari M.P., Gatti G., Fattore C., Fedele G., Novellini R.
Comparative bioavailability and tolerability study of two
intramuscular formulations of thiocolchicoside in healthy
volunteers. Eur. J. Drug Metab. Pharmacokinet. (2001) 26
257–262.
12 Weinling E., Sandouk P., Debray M., Scherrmann J.M. Single
and repeated dose pharmacokinetics of intramuscular thiocol-
chicoside in healthy volunteers. Int. J. Clin. Pharmacol. Ther.
(1999) 37 503–509.