fain, 49 (1992) 361-367 0 1992 Elsevier Science
361 Publishers
B.V. All rights reserved
0304-3959/92/$05.00
PAIN 02037
Cerebrospinal
fluid kinetics of epidural clonidine in man
C.J. Glynn a*c,M.A. Jamous ‘~~3’and P.J. Teddy ‘yb ’ Oxford
Regional Pain Relief Unit, ICRF Building, Churchill Hospital, Oxford OX3 7L.l (UK) and h Department of Neurological Surgery and ’ Nuffield Department of Anaesthetics, Radcliffe Infirmary, Oxford OX2 6HE (UK) (Received
8 July 1991, revision
received
12 November
1991, accepted
29 November
1991)
Ten patients with deafferentation pain after spinal cord injury were given 150 pg clonidine Summary epidurally. CSF and plasma samples were collected over the following 24 h, and drug concentrations were measured by radio-immunoassay. The results of only 6 patients are included in the pharmacokinetic analysis because the catheters were not in the epidural space in the remaining 4 patients. These analyses revealed a mean maximum CSF concentration of 228 ng/ml whereas the mean plasma concentration at all time points was less than 0.7 ng/ml. The elimination half-life of epidural clonidine was 66 k 2 min, while the absorption half-life was 31 t_ 7 min, T,,,,, was 60 + 7 min and C,,,,,was 228 + 56 ng/ml. The ratio of the area under the curve (AUC) for CSF and plasma was 52. One patient’s catheter was intrathecal and 3 were not in the epidural space. The measured plasma concentrations were similar after all injections. As 4 of 6 patients with epidural catheters obtained pain relief and all 3 patients with spasms obtained relief from epidural clonidine, these data suggest that clonidine has a place in the treatment of patients with spinal cord injury. Key words: Epidural
clonidine pharmacokinetics; analgesia; Spinal cord injury; (Man)
Alpha,
Introduction
The inhibition of nociception spinally by noradrenaline in animal models is well documented (Yaksh 1985). There is some evidence that the enkephalinergic systems in the medulla and the brain-stem, when stimulated, exert an inhibitory effect on the dorsal horn cells responsible for transmission of nociception (Reddy and Yaksh 1980). It is suggested that this inhibition is not transmitted via the spinal opioid system and that the most likely transmitters are serotonin and noradrenaline (Reddy and Yaksh 1980; Yaksh 1981, 1985). Clonidine, an alpha,-adrenergic agonist, given orally (Goldstein 1983) and intravenously has a demonstrable analgesic effect in man (Gordh and Tamsen 1983) and in animal models (Paalzow 1974), as well as a spas-
’ Present address: M.A. Jamous, National Spinal Injuries Stoke Mandeville Hospital, Aylesbury, Bucks, UK.
Centre,
Correspondence tot Chris Glynn, Oxford Regional Pain Relief Unit, ICRF Building, Churchill Hospital, Oxford OX3 7LJ, UK.
agonists; Clonidine; Epidural; Pharmacokinetics;
Spinal
molytic effect (Maynard 1986) in man. In addition it has been reported that epidural clonidine also provides analgesia in man in acute (Kalaia et al. 1986; Gordh 1988; Bonnet et al. 1989, 1990; Eisenach et al. 1989b; Lund et al. 1989; Mendez et al. 1990) and chronic pain (Tamsen and Gordh 1984; Glynn et al. 1986, 1988; Germain et al. 1987; Lund et al. 1987; Petros and Bowen-Wright 1987; Eisenach et al. 1989~). The fact that some patients with deafferentation pain following spinal cord injury failed to obtain any pain relief from epidural morphine, but were relieved of their pain by epidural clonidine (Glynn et al. 19861, suggests that the noradrenergic and opioid systems may be independent (Coombs et al. 1985, 1986). Clonidine is a lipid-soluble drug with a high volume of distribution and, because of its potency as a hypotensive drug, the doses used are relatively small and, therefore, the plasma concentrations are low (Dollery et al. 1976). For clonidine CSF concentrations, data is only available from 1 patient (Gordh 1988) and so it is important to confirm that clonidine is found in the CSF in sufficient quantities after epidural injection to have a spinal effect. It has been well documented that
362
the plasma concentration following epidural injection of all drugs studied is similar to that found after an intramuscular injection of the same dose of that drug (Glynn 1987). A recent study has confirmed that the plasma concentration of clonidine after epidural and intramuscular injections is similar and, in addition, both injections provided analgesia in patients with postoperative pain (Bonnet et al. 1990). Thus it is possible that the analgesia achieved following epidural clonidine is due to a central (Goldstein 1974; Paalzow 1974; Gordh and Tamsen 1983) or peripheral (Nakamura and Ferreira 1988) effect from the plasma concentration and not the presumed spinal effect (Yaksh and Reddy 1980; Yaksh 1981, 1985). Data has been published on the transfer of clonidine across the dura in sheep, following epidural and intravenous injections; this confirms a substantial transfer after epidural injection but very little transfer after intravenous injection (Castro and Eisenach 1989). Therefore it is important to know how much of the clonidine is transferred across the dura into the CSF after epidural injection in man.
Methods and patients Ten patients were studied (7 males and 3 females) who had sustained a spinal cord injury l-39 years ago (mean: 15 years). The males were 34-74 years old (mean: 54 years), and the females were 28-65 years old (mean: 43 years). Patients were given written information about the study and they consented to participate. The study had the approval of the local Ethics Committee. All patients were complaining of deafferentation pain which was presumed to be a result of their spinal cord injury, and 3 patients also had problems with muscular spasms. Only 1 female patient (no. IO) was taking oral clonidine for hypertension which was not discontinued prior to the study (Table Il. Despite extensive investigation no other cause, apart from spinal cord injury, could be found for their pain and all previous therapeutic interventions had been unsuccessful. The study was an open study; patients, had a 16-ga epidural catheter inserted under X-ray image-intensification at the Tl2-LI level and passed cephalad so that the tip was at the Tll vertebra. A
I
PATIENT Number
I *
2 3* 4
5 6
7 X’ Y* 10 * Patients
1 DATA Age
35 70 2x 70 41 74 34 65 52 36 not included
Sex
M M F M M M M F M F
Level of lesion C6-C7 L2-L3 Ll-L2 Conus C5-C6 Cervical Ch-C7 Cervical T7-T8 T8
in the pharmacokinetic
Post-injection Pain
Spasm
6/10 o/10 4/10 5/10 5/10 o/10 2/10 o/10 4/10 8/10
lO/lO
results.
IO/IO x/10 o/10 5/10 lO/lO
catheter
Dura mater
Epldural
CaudaEquina
catheter
SpinalCord
Fig. I. This figure shows an artist’s impression of the placement of the epidural and intrathecal catheters. The position of all catheters was confirmed by image intensification and a plain lateral X-ray of the thoraco-lumbar spine.
16-ga epidural catheter was then inserted intrathecally at the L4-L5 or LS-Sl level and passed cephalad until the tips of the 2 catheters were juxtapositioned on either side of the dura using X-ray control (Fig. 11. A further plain lateral X-ray of the thoraco-lumbar spine was taken to confirm the position of the catheter. An intravenous cannula was inserted in a suitable peripheral vein for blood sampling. Patients were given 150 pg epidural clonidine diluted in 5 ml of physiological saline. Blood and CSF samples were withdrawn before, at zero time. and after the epidural injection at 5, 10, 15, 30, 45. 60. 90, 120, 150, 180,240, 300, 360, 540.720 and 1440 min. The CSF was collected as 2 ml samples in dry, sterile plastic tubes and the blood as 3 ml samples in lithium-heparin tubes. Blood samples were separated by centrifugation, and plasma and CSF samples were stored frozen until analysis. Clonidine concentrations were measured by differential radio-immunoassay (Boehringer Ingelheim: Arndts et al. 1981, 19831 (Table II). All patients were kept flat in bed for 3 h after the injection and their blood pressure was monitored. Pain relief and
TABLE
II
CSF AND EPIDURAL
PLASMA CONCENTRATIONS CLONIDINE
Data from 6 patients S.E.M. Time (min)
TABLE
lntrathecai
0 5 10 15 30 45 60 90 120 150 1x0 240 300 360 540 720 1440
AFTER
(nos. 2, 4, 5, 6, 7 and 101. Values
CSF (ng/ml)
Plasma
150
are mean+
(ng/mll
Mean
S.E.M.
Mean
S.E.M.
0.09 18.6 82.7 126 171 228 213 180 152 94.4 70.7 34.5 17.2 7.55 3.03
0.09 7.71 19.6 36.6 33.2 56.6 40.2 32.2 16.1 14.8 13 6.1 3.68 1.6 0.7 0.19 0.93
0.13 0.54 0.43 0.63 0.56 0.56 0.63 0.65 0.61 0.6 0.57 0.61 0.6 0.55 0.43 0.4 0.27
0.13 0.22 0.26 0.21 0.14 0.13 0.15 0.15 0.12 0.1 0.11 0.1 0.12 0.08 0.06 0.07 0.08
I .02 1.28
pg
363 Pt. No.
TABLE
III
CSF AND PLASMA CONCENTRATIONS CLONIDINE IN PATIENTS IN WHOM NOT IN THE EPIDURAL SPACE CSF and plasma Time
Patient
(min)
CSF
0
‘[,,[g 5 10 15
30
45 mln
60
2
3
4
hour
Fig. 2. This figure shows the CSF concentrations of clonidine achieved in all patients over the first 4 h, and it reveals 3 distinct groups. The first curve from 1 patient (no. 1) has a very high initial concentration with a relatively fast fall, indicating a CSF injection. The second curve involving 6 patients (nos. 2, 4-7 and 10) has a lower initial concentration followed by an increasing concentration for 45-60 min and then a slow decline. This is assumed to indicate an epidural injection. The third group involves 3 patients (nos. 3, 8 and 9), and the CSF concentration found is similar to that achieved from oral dosing and thus these injections were assumed not to be in the epidural space.
spasm was assessed globally at the end of the sampling using the visual analogue scale (VAS).
period
by
Statistical analysis Pharmacokinetic parameters were calculated by using the trapezoidal rule for area under the curve (AUC), while absorption half-life (T,,,,,,) and elimination half-life CT,,,,,) were calculated by non-linear regression. Time-to-maximum concentration CT,,,,,) and maximum concentration CC,,,,,) were determined according to standard equations. Statistical analysis, where appropiate, was with the Mann-Whitney U test and Fisher’s exact test; statistical significance was taken as P < 0.05.
5 10 15 30 45 60 90 120 150 180 240 300 360 540 720 1440
Plasma
0.10 0.40 0.56 0.53 0.70 0.93 0.76 0.69 0.81 0.50 0.65 0.52 0.52 0.51 0.38 0.32 0.19
Patients
nos. 3, 8 and 9
CSF
Plasma
(Mean)
(S.E.M.)
(Mean)
(S.E.M.)
0.1 0.1 0.1 0.1 0.2 0.3 0.3 0.3 0.4 0.4 0.5 0.4 0.4 0.4 0.4 0.3 0.2
0 0 0 0 0 0.1 0.1 0.1 0.2 0.2 0.2 0.2 0.1 0.1 0.1 0 0
0.1 0.7 0.7 0.7 0.6 0.7 0.6 0.6 0.6 0.7 0.7 0.6 0.6 0.6 0.5 0.4 0.3
0 0 0.2 0.1 0 0 0 0 0.1 0 0.1 0.1 0 0 0 0 0
plasma concentrations of 0.43 and 0.64 ng/ml, respectively, prior to receiving the epidural injection. The CSF clonidine concentrations of these 6 patients rose to a maximum mean concentration of 228 ng/ml after 45 min. They then fell significantly to 70.7 ng/ml at 3 h, to 17.2 ng/ml at 5 h and to 3 ng/ml at 9 h (P< 0.05, Mann-Whitney U test). The mean time-tomaximum CSF concentration was 60 + 7 min and the mean maximum CSF concentration was 228 k 56 ng/ml; the mean elimination half-life (T1,* .,) was 66 + 2 min whereas the calculated mean absorption
Clonidine
Results
The CSF clonidine concentration of patient no. 1 at 5 min (first sampling point) was very high (6690 ng/ml); therefore it must be assumed that the epidural catheter was intrathecal (Fig. 2, Table III). The CSF clonidine concentrations of 3 patients (nos. 3, 8 and 9) on the other hand were less than 1 ng/ml at all time points, and thus it was assumed that these 3 catheters were not in the epidural space (Fig. 2, Table III). The results of the remaining 6 patients (nos. 2, 4, 5, 6, 7 and 10) are presented in Fig. 3. Only patient no. 10 who was receiving oral clonidine for hypertension had CSF and
150 pg WERE
given ng/ml. no. 1
0.10 6 690 4290 2130 1080 732 157 90.7 15.0 4.77 3.25 1.48 85.0 26.5 5.24 0.85 0.60
AFTER CATHETERS
-
Plasma
.Ol ’
0
500
1000
1500
Time (min) Fig. 3. This figure shows the CSF and plasma concentrations of clonidine in the 6 patients in whom the catheters were in the epidural space.
364 TABLE
IV
PHARMACOKINETIC CLONIDINE CSF AUC T “lax T 1/2cl T I,‘? ,Ih\
AFTER
150
/Ig
EPIDURAL
29048+6716 (min/ny/ml) 6Ok6.71 (min) 66 of-1.63 (mid 31 rf-7.13 (min) 213 k55.6 (ng/ml)
C,,,;,, Plasma AUC CSF/plasma AUC
DATA
629 k X7 (min/ng/ml) ratio 5251
half-life CT,,, ab\) was 31 + 7 min. The CSF/plasma ratio of the area under the time-concentration curve was 52: 1 (Table IV>. The plasma clonidine concentrations rose to 0.63 ng/ml after 15 min. This concentration was maintained for 12 h and fell only by 50% after 24 h (Fig. 3). The plasma concentrations achieved in the remaining 4 patients were similar.
Pain and spasm Two patients (nos. 2 and 6) in whom the catheter was in the epidural space obtained no analgesia, and 1 patient (no. 8) in whom the catheter was not in the epidural space also obtained no pain relief following injection of clonidine. Four of 6 patients in whom the catheters were in the epidural space obtained some analgesia from the epidural injection. One patient (no. 10) had good relief (8/10 on a VAS), 2 patients (nos. 4 and 5) had average relief (5/10), and 1 patient (no. 7) had minimal relief (2/10). Patient no. 1 in whom the catheter was intrathecal obtained good relief (6/10), whereas patients nos. 3 and 9 in whom the catheters were presumed not to be in the epidural space obtained average relief (4/10) from their injections. All 3 patients (nos. 5, 7 and 10) with spasms had excellent relief after the epidural injection of clonidine (8 or lo/lo). Patient no. 1, with the intrathecal catheter, also obtained excellent relief of his spasm (lo/lo); however, in 2 patients with spasm in whom the catheters were not epidural, patient no. 9 obtained average relief (5/10) and patient no. 8 obtained no relief.
Side effects There were no untoward side effects, although there was a fall in blood pressure of lo-20 mm Hg. in all
patients this did not cause any problems because all patients were supine for the first 3 h of the study. None of the patients complained of headaches during or after the study.
Discussion found These data show that the CSF concentrations 150 Fg were after the epidural injection of clonidine several hundred times greater than the plasma concentrations. Indeed the AUC of the CSF concentration curve is 52 times greater than the AUC for the plasma concentration curve. In this study, concentrations of clonidine found in the CSF and plasma in the 6 patients in whom the catheters were in the epidural space were similar to the concentrations found in 1 patient for whom the clonidine dose used was the same (only other published data: Gordh 1988). This study is a pharmacokinetic study and is not designed to address the question of the site of action of epidural clonidine. The other published pharmacokinetic data on the CSF concentration of clonidine after epidural injection is in the sheep (Castro and Eisenach 1989). The mean C”,;,, in this sheep study was 1024 ng/ml with an S.E. of 280 ng/ml. The dose of clonidine used was 2 times that used in this study (300 pg), but the C,,, achieved was 5 times that found in this study. Data from at least 2 of the animals suggest that there was a substantial leak of clonidine from the epidural space to the CSF around the hole in the dura made by the intrathecal catheter. The distance between the tip of the epidural catheter and the hole in the dura was 5 cm; the volume of epidural injectate was 10 ml, thus ensuring that a proportion of the injectate was over the hole in the dura. Therefore guaranteeing that the transfer across the dura would be greater because of the hole. The C,;,, achieved in these 2 animals was of the same order as that achieved from the intrathecal injection in the same study, namely 1970 ng/ml and 1327 ng/ml. Thus another explanation could have been that the 2 catheters were subarachnoid. In our study the distance between the tip of the epidural catheter and the hole in the dura made by the intrathecal catheter was at least 20 cm, and the volume of injectate was 5 ml, making any CSF contamination via the hole in the dura unlikely. The fact that the catheters did not appear to be in the epidural space in 4 of 10 patients, despite all X-ray evidence, underlines the importance of anaesthesia following epidural local anaesthetic as confirmation of epidural catheterisation. Unfortunately, it was not possible to assess anaesthesia in these patients because they were already numb from their spinal cord injury. Thus in this study it was only possible to assume a successful epidural placement on the X-ray evidence,
365
both image intensification and plain X-ray. There do not appear to be any anatomical reasons for these failures because only patient no. 3 had his lesion at the level of the insertion and the other 2 patients were cervical (no. 81 and thoracic (no. 9). The assumed misplacement of the epidural catheters in 4 of 10 patients is based on this pharmacokinetic data and data obtained from the injection of 5 mg of morphine in the same patients with the same catheters. The pharmacokinetic data obtained from the morphine study were similar and, as this data has already been published elsewhere (Glynn 19871, it was assumed that the catheters were not epidural in these 4 patients. The morphine study was either 24 h before or 24 h after the clonidine study. Thus we obtained similar results with 2 different drugs in the same patients using the same catheters (in preparation). Data from the patient in whom the injection was believed to be intrathecal and not epidural (Table III) are similar to data following intrathecal injections in sheep (Castro and Eisenach 1989). It is interesting to note that in both man (Fig. 2, Table III) and sheep (Castro and Eisenach 19891, the CSF concentration of clonidine was similar after 1 h irrespective of whether the injection was epidural or intrathecal. Data from this 1 patient seem to indicate that epidural injection may act as a depot (Fig. 2, Table III) because the CSF concentration is greater after 1 h following epidural injection than it is following intrathecal injection. There are no other published intrathecal clonidine pharmacokinetic data in man. Thus there may a clinical advantage in using the epidural route to provide this “depot” effect. The CSF concentrations achieved in the 3 patients in whom the catheters did not appear to be in the epidural space were similar to those found in the 1 patient who was taking oral clonidine (Fig. 2, Table III>. These data were also similar to the CSF concentration found in sheep after the intravenous injection of 300 pg of clonidine (Castro and Eisenach 1989). This is further evidence in favour of the catheters in these patients not being in the epidural space (Fig. 2, Table 1111.Plasma concentrations found after epidural and “parenteral” injections in this study are similar to those found by Bonnet et al. (1990) in their study of epidural and intramuscular clonidine in man. Plasma concentrations of less than 1 ng/ml, in 1 patient, were also reported by Gordh (1988) in the other published data of CSF and plasma concentrations of clonidine after epidural injection in man. Thus, it is impossible to differentiate among intrathecal, epidural, intramuscular, or any parenteral injection of clonidine using the plasma concentrations (Arndts et al. 1983). Data produced by Bonnet et al. (1990) confirm an analgesic effect of the intramuscular injection, in patients with postoperative pain which did not last as
long as the epidural injection. A possible criticism of this study is that epidural injection was not compared to parenteral injection in a double-blind manner. It would have been better to have given each patient, double-blind, parenteral and epidural injections of clonidine. Early published data suggested that epidural clonidine was not as effective as opioids in patients with postoperative pain (Kalaia et al. 1987; Gordh 19881, but recent evidence suggests that the dose required to provide analgesia for postoperative pain is greater than that used in the early studies (Bonnet et al. 1989, 1990; Eisenach et al. 1989b; Mendez et al. 1990). However, it appears to be more effective than epidural morphine in some patients with chronic pain, e.g., deafferentation pain (Glynn et al. 1986; Lund et al. 1987; Petros and Bowen-Wright 1987; Coventry and Todd 1989; Farcot et al. 1989). In this study, the fact that a degree of analgesia was achieved in 4 of 6 patients in whom the catheter was in the epidural space reinforces the previous suggestion (Glynn et al. 1986) that clonidine provides analgesia for some, but not all, spinal injury patients. Analgesia following epidural administration is unlikely to be due only to a spinal cord effect. The evidence in animal experiments is divided; Ossipov et al. (1989) state that clonidine has no central analgesic effect while Marwaha et al. (1983) and Murata et al. (1989) state that there is a central effect. Most of the experimental evidence is in favour of a spinal cord action for clonidine (Reddy and Yaksh 1980; Yaksh 1981, 1985) and this evidence, taken in conjunction with these pharmacokinetic data, is in favour of a substantial spinal cord action (Figs. 2 and 3). The clinical evidence is that there is an analgesic effect of clonidine given orally (Goldstein 1983) and intravenously (Gordh and Tamsen 1983) which infers a central site of action. In addition, recent evidence proposes a peripheral analgesic action for clonidine (Nakamura and Ferreira 19881. The fact that 2 patients (nos. 3 and 9) in whom the catheters were not in the epidural space obtained some analgesia from clonidine injection is evidence of a central or peripheral effect of the drug or alternatively a placebo response. In addition most patients became drowsy - a central effect. Pharmacokinetic data suggest that the site of action of epidural clonidine is inside the dura and thus both central and spinal components probably contribute to analgesia as is the situation with other drugs given epidurally (Coombs et al. 1985). The central component of the analgesia is a result of the combination of the plasma concentration of clonidine and the CSF concentration transferred centrally (Coombs et al. 198.51. The amount of clonidine transferred via the CSF is directly related to the lipid solubility of clonidine which has a partition coefficient of 75%, i.e., 3 parts in oil (spinal cord) to 1 part in water (CSF). This
should mean that most of the clonidine would be absorbed by the spinal cord but also that some would be transferred via the CSF to the brain. It has been suggested that the spasmolytic effect of clonidine is centrally mediated (Maynard 1986); these plasma pharmacokinetic data support this thesis. Five (1 intrathecal, 3 epidural, 1 parenteral) of 6 patients with muscle spasms obtained good to total relief of their spasms with injection of clonidine; 1 failure followed a parenteral injection (Table I). Experimental evidence in animals indicates that there are no adverse effects on spinal cord blood flow in pigs (Gordh et al. 1986a) or pregnant sheep (Eisenach et al. 1989a) and no long-term neuropathological effects in rat (Gordh et al. 1986b). This is reinforced by clinical experience in over 300 patients with chronic pain who have been given epidural clonidine either as a single injection or as repeated injections without any adverse short- or long-term side effects (unpublished observations). It would thus appear that the epidural injection of clonidine is unlikely to cause any long-term problems for the patient and does provide a degree of analgesia in certain types of pain.
Coombs,
D.W..
Murphy,
Saunders,
C.A.,
dine for intractable Coventry.
D.M. C.T.,
Reid,
Clare.
R.A.
and pharmacokinetics Eisenach,
J.C., Castro,
clonidine
Epidural
70 (1989a) Eisenach,
Analg.,
analgesia
G.H.,
Dargie.
Dewan.
Lysak. S.Z.
Clin.
Pharmacol.
D.M.
clonidine
and its comparison
col. Meth., Arndts,
h (1981)
with reference
D., Doevendans,
man. Eur. J. Clin. Pharmacol., F.,
Touboul,
Boico.
O.,
C., Abhay.
sia with extradural Bonnet,
R. and Heinz,
24 (1983)
Rostaing.
clonidine.
F., Boico. O.,
Rostaing, analgesia
versus intramuscular
and Viscomi,
R.L..
Mercky,
Buzzanell,
C.M.,
Epidural
in
M..
Loriterne,
S., Loriferne,
63 (1989)
J.-L.. analge-
465-469.
J.F. and Saada.
in postoperative
patients:
Anesthesiology.
M..
epidural 72 (IYYO)
423-427. Budd, K.R.,
C. and Lysak. S.Z., Epidural
Clinical Anaesthesiology,
Vol. 1, Part 4. 1087.
pp. 915-933. Castro.
limites ct suivi a long terme.
H.. Fernette.
clonidine
L.
C.J.. Intrathecal
Flynn.
C.J..
Jamous,
Neron,
and
in the epidural
Flynn.
intrathecal,
thesiology, 71 (1989) Coombs.
D.W..
son, T.S. and Jensen. Anesthesiology,
and epidural
and dynamics 01
clonidine
in sheep. Anes-
418-425.
Saunders.
of intrathecal
J.C.. Pharmacokinetlcs
pain in patients
clonidine
L.E.,
Intrathecal
DADLE.and
62 (1985) 358-362.
D.. Savage. S.. Ragnar\morphine
tolerance:
intraventricular
use
morphine.
30
A., Antinociceptive
action of
and epidural M.A..
administration
Teddy,
196.
of opiates.
P.J., Moore.
R.A.
and L.loyd.
system in transmisalon
with spinal cord injury. l.ancct.
ii (IYXh)
ot
114%
I7.50. and
pain, Pain, 34 (IYXX) J.. Clonidinr
T.. Epidural
morphine
in patients
with
ill-
123- 128.
an analge.src. Hiol.
as
clomdine
after thoracotomy: Anesthesiol. Gordh.
P\ychint,.
clonidine
for treatment
IX (IYX.3)
01 postoperative
a double blind placebo controlled
Stand.,
pain
study. Acta
32 (198X) 702-7OY.
‘I‘. and Tamsen.
Gordh,
Sanders. R., A double blmd compdrl-
and
clonidinr
134(,.
Gordh.
A..
A
study of the
m man, Acta Anrathealol.
T.. Feuk. U. and Norlen,
Stand..
analgesic 27 (1983)
o(
clonidine
on in
Analg.,
K.. Effect of rpidural
etfect 72.
spinal cord blood flow and regional and central hemodynamics f15 (IYXha)
‘I‘.. Post. C. and Olsson, clomdine,
1312-1318. Y..
Evaluation
guanfacmr
of the toxicity ot
and a hubstance P antagonlht Analg.,
05 (lY86b)
1303-1311 Kalaia,
P.K..
Madam.
V.R..
tive analgrala,
lnd. J. Med.
Latha.
Comparison
evoked
D.B..
V. and
for postopera-
Effect of epldural
clonidine
to dermatomal
electrical
potentials
S., Greulich.
36 (19X7) 387.
A.. lljortscd. N.-C.
of the effects of extradural on
V.. Vardham.
clonidine
Reb.. X3 (19X6) 5.50-552.
and pain, Anacsthrtist,
C.. Qvitzau.
morphine
K.K..
Ii. and Hanben.
somatosrnsory
stimulation Lund.
R.. Batra.
C‘linlcal htudy on epldural
L.und, C.. Kehlrt,
postoperative
pain.
and Kehlct.
clonidine
II.,
with those of
atresa responses,
cardiopul-
monary function and motor and Aensory block. Br. J. Anaesth..
63
(19X9) 516519. J.. Kehne.
and Davis, Maynard,
M..
J.H., Commissarla.
R..
R.L..
clonidine
inhibits
Brain Res., 276 (1983)
37Y-3X2.
F.M.,
Spinal
Early
spahticity. Paraplegia. Mendez.
R.L.., Lachance.
Agressologie.
space. Pain, Suppl. 4 (1987)
Role of the spinal noradrenrrgic
coeruleus.
M.1. and Eisenach.
intravenous.
par voir epidude dcafferenta-
13’1-142.
Germain.
J.W..
C‘., Clonidine
ou sacree dans les douleurs
tion: possibilities. (1989)
Marwaha, Baillikre’s
clamdinc 71 (lY8Yb)
cancer pain: phase 1. Anesthe-
F. and Grasser,
rale thoraco-lumbar
on
M., Postoperative
Br. J. Anaesth.,
administration.
of clonidine
21-30.
S., Saada.
K. and Ghignone.
Clonidine-induced
B.. New aspect>
and pharmacodynamics
IY
siology, 7 I (1YXYc) 647-657. Farcot, J.M.,
Gode.
J.. Kirsten,
of the pharmacokinetics Bonnet.
methods, J. Pharma-
295-307.
Ther..
and Rose. J.C., Epidural
on rat spmal cord and nerve rootb. Anaesth. for
361 Publishers
B.V. All rights reserved
0304-3959/92/$05.00
PAIN 02037
Cerebrospinal
fluid kinetics of epidural clonidine in man
C.J. Glynn a*c,M.A. Jamous ‘~~3’and P.J. Teddy ‘yb ’ Oxford
Regional Pain Relief Unit, ICRF Building, Churchill Hospital, Oxford OX3 7L.l (UK) and h Department of Neurological Surgery and ’ Nuffield Department of Anaesthetics, Radcliffe Infirmary, Oxford OX2 6HE (UK) (Received
8 July 1991, revision
received
12 November
1991, accepted
29 November
1991)
Ten patients with deafferentation pain after spinal cord injury were given 150 pg clonidine Summary epidurally. CSF and plasma samples were collected over the following 24 h, and drug concentrations were measured by radio-immunoassay. The results of only 6 patients are included in the pharmacokinetic analysis because the catheters were not in the epidural space in the remaining 4 patients. These analyses revealed a mean maximum CSF concentration of 228 ng/ml whereas the mean plasma concentration at all time points was less than 0.7 ng/ml. The elimination half-life of epidural clonidine was 66 k 2 min, while the absorption half-life was 31 t_ 7 min, T,,,,, was 60 + 7 min and C,,,,,was 228 + 56 ng/ml. The ratio of the area under the curve (AUC) for CSF and plasma was 52. One patient’s catheter was intrathecal and 3 were not in the epidural space. The measured plasma concentrations were similar after all injections. As 4 of 6 patients with epidural catheters obtained pain relief and all 3 patients with spasms obtained relief from epidural clonidine, these data suggest that clonidine has a place in the treatment of patients with spinal cord injury. Key words: Epidural
clonidine pharmacokinetics; analgesia; Spinal cord injury; (Man)
Alpha,
Introduction
The inhibition of nociception spinally by noradrenaline in animal models is well documented (Yaksh 1985). There is some evidence that the enkephalinergic systems in the medulla and the brain-stem, when stimulated, exert an inhibitory effect on the dorsal horn cells responsible for transmission of nociception (Reddy and Yaksh 1980). It is suggested that this inhibition is not transmitted via the spinal opioid system and that the most likely transmitters are serotonin and noradrenaline (Reddy and Yaksh 1980; Yaksh 1981, 1985). Clonidine, an alpha,-adrenergic agonist, given orally (Goldstein 1983) and intravenously has a demonstrable analgesic effect in man (Gordh and Tamsen 1983) and in animal models (Paalzow 1974), as well as a spas-
’ Present address: M.A. Jamous, National Spinal Injuries Stoke Mandeville Hospital, Aylesbury, Bucks, UK.
Centre,
Correspondence tot Chris Glynn, Oxford Regional Pain Relief Unit, ICRF Building, Churchill Hospital, Oxford OX3 7LJ, UK.
agonists; Clonidine; Epidural; Pharmacokinetics;
Spinal
molytic effect (Maynard 1986) in man. In addition it has been reported that epidural clonidine also provides analgesia in man in acute (Kalaia et al. 1986; Gordh 1988; Bonnet et al. 1989, 1990; Eisenach et al. 1989b; Lund et al. 1989; Mendez et al. 1990) and chronic pain (Tamsen and Gordh 1984; Glynn et al. 1986, 1988; Germain et al. 1987; Lund et al. 1987; Petros and Bowen-Wright 1987; Eisenach et al. 1989~). The fact that some patients with deafferentation pain following spinal cord injury failed to obtain any pain relief from epidural morphine, but were relieved of their pain by epidural clonidine (Glynn et al. 19861, suggests that the noradrenergic and opioid systems may be independent (Coombs et al. 1985, 1986). Clonidine is a lipid-soluble drug with a high volume of distribution and, because of its potency as a hypotensive drug, the doses used are relatively small and, therefore, the plasma concentrations are low (Dollery et al. 1976). For clonidine CSF concentrations, data is only available from 1 patient (Gordh 1988) and so it is important to confirm that clonidine is found in the CSF in sufficient quantities after epidural injection to have a spinal effect. It has been well documented that
362
the plasma concentration following epidural injection of all drugs studied is similar to that found after an intramuscular injection of the same dose of that drug (Glynn 1987). A recent study has confirmed that the plasma concentration of clonidine after epidural and intramuscular injections is similar and, in addition, both injections provided analgesia in patients with postoperative pain (Bonnet et al. 1990). Thus it is possible that the analgesia achieved following epidural clonidine is due to a central (Goldstein 1974; Paalzow 1974; Gordh and Tamsen 1983) or peripheral (Nakamura and Ferreira 1988) effect from the plasma concentration and not the presumed spinal effect (Yaksh and Reddy 1980; Yaksh 1981, 1985). Data has been published on the transfer of clonidine across the dura in sheep, following epidural and intravenous injections; this confirms a substantial transfer after epidural injection but very little transfer after intravenous injection (Castro and Eisenach 1989). Therefore it is important to know how much of the clonidine is transferred across the dura into the CSF after epidural injection in man.
Methods and patients Ten patients were studied (7 males and 3 females) who had sustained a spinal cord injury l-39 years ago (mean: 15 years). The males were 34-74 years old (mean: 54 years), and the females were 28-65 years old (mean: 43 years). Patients were given written information about the study and they consented to participate. The study had the approval of the local Ethics Committee. All patients were complaining of deafferentation pain which was presumed to be a result of their spinal cord injury, and 3 patients also had problems with muscular spasms. Only 1 female patient (no. IO) was taking oral clonidine for hypertension which was not discontinued prior to the study (Table Il. Despite extensive investigation no other cause, apart from spinal cord injury, could be found for their pain and all previous therapeutic interventions had been unsuccessful. The study was an open study; patients, had a 16-ga epidural catheter inserted under X-ray image-intensification at the Tl2-LI level and passed cephalad so that the tip was at the Tll vertebra. A
I
PATIENT Number
I *
2 3* 4
5 6
7 X’ Y* 10 * Patients
1 DATA Age
35 70 2x 70 41 74 34 65 52 36 not included
Sex
M M F M M M M F M F
Level of lesion C6-C7 L2-L3 Ll-L2 Conus C5-C6 Cervical Ch-C7 Cervical T7-T8 T8
in the pharmacokinetic
Post-injection Pain
Spasm
6/10 o/10 4/10 5/10 5/10 o/10 2/10 o/10 4/10 8/10
lO/lO
results.
IO/IO x/10 o/10 5/10 lO/lO
catheter
Dura mater
Epldural
CaudaEquina
catheter
SpinalCord
Fig. I. This figure shows an artist’s impression of the placement of the epidural and intrathecal catheters. The position of all catheters was confirmed by image intensification and a plain lateral X-ray of the thoraco-lumbar spine.
16-ga epidural catheter was then inserted intrathecally at the L4-L5 or LS-Sl level and passed cephalad until the tips of the 2 catheters were juxtapositioned on either side of the dura using X-ray control (Fig. 11. A further plain lateral X-ray of the thoraco-lumbar spine was taken to confirm the position of the catheter. An intravenous cannula was inserted in a suitable peripheral vein for blood sampling. Patients were given 150 pg epidural clonidine diluted in 5 ml of physiological saline. Blood and CSF samples were withdrawn before, at zero time. and after the epidural injection at 5, 10, 15, 30, 45. 60. 90, 120, 150, 180,240, 300, 360, 540.720 and 1440 min. The CSF was collected as 2 ml samples in dry, sterile plastic tubes and the blood as 3 ml samples in lithium-heparin tubes. Blood samples were separated by centrifugation, and plasma and CSF samples were stored frozen until analysis. Clonidine concentrations were measured by differential radio-immunoassay (Boehringer Ingelheim: Arndts et al. 1981, 19831 (Table II). All patients were kept flat in bed for 3 h after the injection and their blood pressure was monitored. Pain relief and
TABLE
II
CSF AND EPIDURAL
PLASMA CONCENTRATIONS CLONIDINE
Data from 6 patients S.E.M. Time (min)
TABLE
lntrathecai
0 5 10 15 30 45 60 90 120 150 1x0 240 300 360 540 720 1440
AFTER
(nos. 2, 4, 5, 6, 7 and 101. Values
CSF (ng/ml)
Plasma
150
are mean+
(ng/mll
Mean
S.E.M.
Mean
S.E.M.
0.09 18.6 82.7 126 171 228 213 180 152 94.4 70.7 34.5 17.2 7.55 3.03
0.09 7.71 19.6 36.6 33.2 56.6 40.2 32.2 16.1 14.8 13 6.1 3.68 1.6 0.7 0.19 0.93
0.13 0.54 0.43 0.63 0.56 0.56 0.63 0.65 0.61 0.6 0.57 0.61 0.6 0.55 0.43 0.4 0.27
0.13 0.22 0.26 0.21 0.14 0.13 0.15 0.15 0.12 0.1 0.11 0.1 0.12 0.08 0.06 0.07 0.08
I .02 1.28
pg
363 Pt. No.
TABLE
III
CSF AND PLASMA CONCENTRATIONS CLONIDINE IN PATIENTS IN WHOM NOT IN THE EPIDURAL SPACE CSF and plasma Time
Patient
(min)
CSF
0
‘[,,[g 5 10 15
30
45 mln
60
2
3
4
hour
Fig. 2. This figure shows the CSF concentrations of clonidine achieved in all patients over the first 4 h, and it reveals 3 distinct groups. The first curve from 1 patient (no. 1) has a very high initial concentration with a relatively fast fall, indicating a CSF injection. The second curve involving 6 patients (nos. 2, 4-7 and 10) has a lower initial concentration followed by an increasing concentration for 45-60 min and then a slow decline. This is assumed to indicate an epidural injection. The third group involves 3 patients (nos. 3, 8 and 9), and the CSF concentration found is similar to that achieved from oral dosing and thus these injections were assumed not to be in the epidural space.
spasm was assessed globally at the end of the sampling using the visual analogue scale (VAS).
period
by
Statistical analysis Pharmacokinetic parameters were calculated by using the trapezoidal rule for area under the curve (AUC), while absorption half-life (T,,,,,,) and elimination half-life CT,,,,,) were calculated by non-linear regression. Time-to-maximum concentration CT,,,,,) and maximum concentration CC,,,,,) were determined according to standard equations. Statistical analysis, where appropiate, was with the Mann-Whitney U test and Fisher’s exact test; statistical significance was taken as P < 0.05.
5 10 15 30 45 60 90 120 150 180 240 300 360 540 720 1440
Plasma
0.10 0.40 0.56 0.53 0.70 0.93 0.76 0.69 0.81 0.50 0.65 0.52 0.52 0.51 0.38 0.32 0.19
Patients
nos. 3, 8 and 9
CSF
Plasma
(Mean)
(S.E.M.)
(Mean)
(S.E.M.)
0.1 0.1 0.1 0.1 0.2 0.3 0.3 0.3 0.4 0.4 0.5 0.4 0.4 0.4 0.4 0.3 0.2
0 0 0 0 0 0.1 0.1 0.1 0.2 0.2 0.2 0.2 0.1 0.1 0.1 0 0
0.1 0.7 0.7 0.7 0.6 0.7 0.6 0.6 0.6 0.7 0.7 0.6 0.6 0.6 0.5 0.4 0.3
0 0 0.2 0.1 0 0 0 0 0.1 0 0.1 0.1 0 0 0 0 0
plasma concentrations of 0.43 and 0.64 ng/ml, respectively, prior to receiving the epidural injection. The CSF clonidine concentrations of these 6 patients rose to a maximum mean concentration of 228 ng/ml after 45 min. They then fell significantly to 70.7 ng/ml at 3 h, to 17.2 ng/ml at 5 h and to 3 ng/ml at 9 h (P< 0.05, Mann-Whitney U test). The mean time-tomaximum CSF concentration was 60 + 7 min and the mean maximum CSF concentration was 228 k 56 ng/ml; the mean elimination half-life (T1,* .,) was 66 + 2 min whereas the calculated mean absorption
Clonidine
Results
The CSF clonidine concentration of patient no. 1 at 5 min (first sampling point) was very high (6690 ng/ml); therefore it must be assumed that the epidural catheter was intrathecal (Fig. 2, Table III). The CSF clonidine concentrations of 3 patients (nos. 3, 8 and 9) on the other hand were less than 1 ng/ml at all time points, and thus it was assumed that these 3 catheters were not in the epidural space (Fig. 2, Table III). The results of the remaining 6 patients (nos. 2, 4, 5, 6, 7 and 10) are presented in Fig. 3. Only patient no. 10 who was receiving oral clonidine for hypertension had CSF and
150 pg WERE
given ng/ml. no. 1
0.10 6 690 4290 2130 1080 732 157 90.7 15.0 4.77 3.25 1.48 85.0 26.5 5.24 0.85 0.60
AFTER CATHETERS
-
Plasma
.Ol ’
0
500
1000
1500
Time (min) Fig. 3. This figure shows the CSF and plasma concentrations of clonidine in the 6 patients in whom the catheters were in the epidural space.
364 TABLE
IV
PHARMACOKINETIC CLONIDINE CSF AUC T “lax T 1/2cl T I,‘? ,Ih\
AFTER
150
/Ig
EPIDURAL
29048+6716 (min/ny/ml) 6Ok6.71 (min) 66 of-1.63 (mid 31 rf-7.13 (min) 213 k55.6 (ng/ml)
C,,,;,, Plasma AUC CSF/plasma AUC
DATA
629 k X7 (min/ng/ml) ratio 5251
half-life CT,,, ab\) was 31 + 7 min. The CSF/plasma ratio of the area under the time-concentration curve was 52: 1 (Table IV>. The plasma clonidine concentrations rose to 0.63 ng/ml after 15 min. This concentration was maintained for 12 h and fell only by 50% after 24 h (Fig. 3). The plasma concentrations achieved in the remaining 4 patients were similar.
Pain and spasm Two patients (nos. 2 and 6) in whom the catheter was in the epidural space obtained no analgesia, and 1 patient (no. 8) in whom the catheter was not in the epidural space also obtained no pain relief following injection of clonidine. Four of 6 patients in whom the catheters were in the epidural space obtained some analgesia from the epidural injection. One patient (no. 10) had good relief (8/10 on a VAS), 2 patients (nos. 4 and 5) had average relief (5/10), and 1 patient (no. 7) had minimal relief (2/10). Patient no. 1 in whom the catheter was intrathecal obtained good relief (6/10), whereas patients nos. 3 and 9 in whom the catheters were presumed not to be in the epidural space obtained average relief (4/10) from their injections. All 3 patients (nos. 5, 7 and 10) with spasms had excellent relief after the epidural injection of clonidine (8 or lo/lo). Patient no. 1, with the intrathecal catheter, also obtained excellent relief of his spasm (lo/lo); however, in 2 patients with spasm in whom the catheters were not epidural, patient no. 9 obtained average relief (5/10) and patient no. 8 obtained no relief.
Side effects There were no untoward side effects, although there was a fall in blood pressure of lo-20 mm Hg. in all
patients this did not cause any problems because all patients were supine for the first 3 h of the study. None of the patients complained of headaches during or after the study.
Discussion found These data show that the CSF concentrations 150 Fg were after the epidural injection of clonidine several hundred times greater than the plasma concentrations. Indeed the AUC of the CSF concentration curve is 52 times greater than the AUC for the plasma concentration curve. In this study, concentrations of clonidine found in the CSF and plasma in the 6 patients in whom the catheters were in the epidural space were similar to the concentrations found in 1 patient for whom the clonidine dose used was the same (only other published data: Gordh 1988). This study is a pharmacokinetic study and is not designed to address the question of the site of action of epidural clonidine. The other published pharmacokinetic data on the CSF concentration of clonidine after epidural injection is in the sheep (Castro and Eisenach 1989). The mean C”,;,, in this sheep study was 1024 ng/ml with an S.E. of 280 ng/ml. The dose of clonidine used was 2 times that used in this study (300 pg), but the C,,, achieved was 5 times that found in this study. Data from at least 2 of the animals suggest that there was a substantial leak of clonidine from the epidural space to the CSF around the hole in the dura made by the intrathecal catheter. The distance between the tip of the epidural catheter and the hole in the dura was 5 cm; the volume of epidural injectate was 10 ml, thus ensuring that a proportion of the injectate was over the hole in the dura. Therefore guaranteeing that the transfer across the dura would be greater because of the hole. The C,;,, achieved in these 2 animals was of the same order as that achieved from the intrathecal injection in the same study, namely 1970 ng/ml and 1327 ng/ml. Thus another explanation could have been that the 2 catheters were subarachnoid. In our study the distance between the tip of the epidural catheter and the hole in the dura made by the intrathecal catheter was at least 20 cm, and the volume of injectate was 5 ml, making any CSF contamination via the hole in the dura unlikely. The fact that the catheters did not appear to be in the epidural space in 4 of 10 patients, despite all X-ray evidence, underlines the importance of anaesthesia following epidural local anaesthetic as confirmation of epidural catheterisation. Unfortunately, it was not possible to assess anaesthesia in these patients because they were already numb from their spinal cord injury. Thus in this study it was only possible to assume a successful epidural placement on the X-ray evidence,
365
both image intensification and plain X-ray. There do not appear to be any anatomical reasons for these failures because only patient no. 3 had his lesion at the level of the insertion and the other 2 patients were cervical (no. 81 and thoracic (no. 9). The assumed misplacement of the epidural catheters in 4 of 10 patients is based on this pharmacokinetic data and data obtained from the injection of 5 mg of morphine in the same patients with the same catheters. The pharmacokinetic data obtained from the morphine study were similar and, as this data has already been published elsewhere (Glynn 19871, it was assumed that the catheters were not epidural in these 4 patients. The morphine study was either 24 h before or 24 h after the clonidine study. Thus we obtained similar results with 2 different drugs in the same patients using the same catheters (in preparation). Data from the patient in whom the injection was believed to be intrathecal and not epidural (Table III) are similar to data following intrathecal injections in sheep (Castro and Eisenach 1989). It is interesting to note that in both man (Fig. 2, Table III) and sheep (Castro and Eisenach 19891, the CSF concentration of clonidine was similar after 1 h irrespective of whether the injection was epidural or intrathecal. Data from this 1 patient seem to indicate that epidural injection may act as a depot (Fig. 2, Table III) because the CSF concentration is greater after 1 h following epidural injection than it is following intrathecal injection. There are no other published intrathecal clonidine pharmacokinetic data in man. Thus there may a clinical advantage in using the epidural route to provide this “depot” effect. The CSF concentrations achieved in the 3 patients in whom the catheters did not appear to be in the epidural space were similar to those found in the 1 patient who was taking oral clonidine (Fig. 2, Table III>. These data were also similar to the CSF concentration found in sheep after the intravenous injection of 300 pg of clonidine (Castro and Eisenach 1989). This is further evidence in favour of the catheters in these patients not being in the epidural space (Fig. 2, Table 1111.Plasma concentrations found after epidural and “parenteral” injections in this study are similar to those found by Bonnet et al. (1990) in their study of epidural and intramuscular clonidine in man. Plasma concentrations of less than 1 ng/ml, in 1 patient, were also reported by Gordh (1988) in the other published data of CSF and plasma concentrations of clonidine after epidural injection in man. Thus, it is impossible to differentiate among intrathecal, epidural, intramuscular, or any parenteral injection of clonidine using the plasma concentrations (Arndts et al. 1983). Data produced by Bonnet et al. (1990) confirm an analgesic effect of the intramuscular injection, in patients with postoperative pain which did not last as
long as the epidural injection. A possible criticism of this study is that epidural injection was not compared to parenteral injection in a double-blind manner. It would have been better to have given each patient, double-blind, parenteral and epidural injections of clonidine. Early published data suggested that epidural clonidine was not as effective as opioids in patients with postoperative pain (Kalaia et al. 1987; Gordh 19881, but recent evidence suggests that the dose required to provide analgesia for postoperative pain is greater than that used in the early studies (Bonnet et al. 1989, 1990; Eisenach et al. 1989b; Mendez et al. 1990). However, it appears to be more effective than epidural morphine in some patients with chronic pain, e.g., deafferentation pain (Glynn et al. 1986; Lund et al. 1987; Petros and Bowen-Wright 1987; Coventry and Todd 1989; Farcot et al. 1989). In this study, the fact that a degree of analgesia was achieved in 4 of 6 patients in whom the catheter was in the epidural space reinforces the previous suggestion (Glynn et al. 1986) that clonidine provides analgesia for some, but not all, spinal injury patients. Analgesia following epidural administration is unlikely to be due only to a spinal cord effect. The evidence in animal experiments is divided; Ossipov et al. (1989) state that clonidine has no central analgesic effect while Marwaha et al. (1983) and Murata et al. (1989) state that there is a central effect. Most of the experimental evidence is in favour of a spinal cord action for clonidine (Reddy and Yaksh 1980; Yaksh 1981, 1985) and this evidence, taken in conjunction with these pharmacokinetic data, is in favour of a substantial spinal cord action (Figs. 2 and 3). The clinical evidence is that there is an analgesic effect of clonidine given orally (Goldstein 1983) and intravenously (Gordh and Tamsen 1983) which infers a central site of action. In addition, recent evidence proposes a peripheral analgesic action for clonidine (Nakamura and Ferreira 19881. The fact that 2 patients (nos. 3 and 9) in whom the catheters were not in the epidural space obtained some analgesia from clonidine injection is evidence of a central or peripheral effect of the drug or alternatively a placebo response. In addition most patients became drowsy - a central effect. Pharmacokinetic data suggest that the site of action of epidural clonidine is inside the dura and thus both central and spinal components probably contribute to analgesia as is the situation with other drugs given epidurally (Coombs et al. 1985). The central component of the analgesia is a result of the combination of the plasma concentration of clonidine and the CSF concentration transferred centrally (Coombs et al. 198.51. The amount of clonidine transferred via the CSF is directly related to the lipid solubility of clonidine which has a partition coefficient of 75%, i.e., 3 parts in oil (spinal cord) to 1 part in water (CSF). This
should mean that most of the clonidine would be absorbed by the spinal cord but also that some would be transferred via the CSF to the brain. It has been suggested that the spasmolytic effect of clonidine is centrally mediated (Maynard 1986); these plasma pharmacokinetic data support this thesis. Five (1 intrathecal, 3 epidural, 1 parenteral) of 6 patients with muscle spasms obtained good to total relief of their spasms with injection of clonidine; 1 failure followed a parenteral injection (Table I). Experimental evidence in animals indicates that there are no adverse effects on spinal cord blood flow in pigs (Gordh et al. 1986a) or pregnant sheep (Eisenach et al. 1989a) and no long-term neuropathological effects in rat (Gordh et al. 1986b). This is reinforced by clinical experience in over 300 patients with chronic pain who have been given epidural clonidine either as a single injection or as repeated injections without any adverse short- or long-term side effects (unpublished observations). It would thus appear that the epidural injection of clonidine is unlikely to cause any long-term problems for the patient and does provide a degree of analgesia in certain types of pain.
Coombs,
D.W..
Murphy,
Saunders,
C.A.,
dine for intractable Coventry.
D.M. C.T.,
Reid,
Clare.
R.A.
and pharmacokinetics Eisenach,
J.C., Castro,
clonidine
Epidural
70 (1989a) Eisenach,
Analg.,
analgesia
G.H.,
Dargie.
Dewan.
Lysak. S.Z.
Clin.
Pharmacol.
D.M.
clonidine
and its comparison
col. Meth., Arndts,
h (1981)
with reference
D., Doevendans,
man. Eur. J. Clin. Pharmacol., F.,
Touboul,
Boico.
O.,
C., Abhay.
sia with extradural Bonnet,
R. and Heinz,
24 (1983)
Rostaing.
clonidine.
F., Boico. O.,
Rostaing, analgesia
versus intramuscular
and Viscomi,
R.L..
Mercky,
Buzzanell,
C.M.,
Epidural
in
M..
Loriterne,
S., Loriferne,
63 (1989)
J.-L.. analge-
465-469.
J.F. and Saada.
in postoperative
patients:
Anesthesiology.
M..
epidural 72 (IYYO)
423-427. Budd, K.R.,
C. and Lysak. S.Z., Epidural
Clinical Anaesthesiology,
Vol. 1, Part 4. 1087.
pp. 915-933. Castro.
limites ct suivi a long terme.
H.. Fernette.
clonidine
L.
C.J.. Intrathecal
Flynn.
C.J..
Jamous,
Neron,
and
in the epidural
Flynn.
intrathecal,
thesiology, 71 (1989) Coombs.
D.W..
son, T.S. and Jensen. Anesthesiology,
and epidural
and dynamics 01
clonidine
in sheep. Anes-
418-425.
Saunders.
of intrathecal
J.C.. Pharmacokinetlcs
pain in patients
clonidine
L.E.,
Intrathecal
DADLE.and
62 (1985) 358-362.
D.. Savage. S.. Ragnar\morphine
tolerance:
intraventricular
use
morphine.
30
A., Antinociceptive
action of
and epidural M.A..
administration
Teddy,
196.
of opiates.
P.J., Moore.
R.A.
and L.loyd.
system in transmisalon
with spinal cord injury. l.ancct.
ii (IYXh)
ot
114%
I7.50. and
pain, Pain, 34 (IYXX) J.. Clonidinr
T.. Epidural
morphine
in patients
with
ill-
123- 128.
an analge.src. Hiol.
as
clomdine
after thoracotomy: Anesthesiol. Gordh.
P\ychint,.
clonidine
for treatment
IX (IYX.3)
01 postoperative
a double blind placebo controlled
Stand.,
pain
study. Acta
32 (198X) 702-7OY.
‘I‘. and Tamsen.
Gordh,
Sanders. R., A double blmd compdrl-
and
clonidinr
134(,.
Gordh.
A..
A
study of the
m man, Acta Anrathealol.
T.. Feuk. U. and Norlen,
Stand..
analgesic 27 (1983)
o(
clonidine
on in
Analg.,
K.. Effect of rpidural
etfect 72.
spinal cord blood flow and regional and central hemodynamics f15 (IYXha)
‘I‘.. Post. C. and Olsson, clomdine,
1312-1318. Y..
Evaluation
guanfacmr
of the toxicity ot
and a hubstance P antagonlht Analg.,
05 (lY86b)
1303-1311 Kalaia,
P.K..
Madam.
V.R..
tive analgrala,
lnd. J. Med.
Latha.
Comparison
evoked
D.B..
V. and
for postopera-
Effect of epldural
clonidine
to dermatomal
electrical
potentials
S., Greulich.
36 (19X7) 387.
A.. lljortscd. N.-C.
of the effects of extradural on
V.. Vardham.
clonidine
Reb.. X3 (19X6) 5.50-552.
and pain, Anacsthrtist,
C.. Qvitzau.
morphine
K.K..
Ii. and Hanben.
somatosrnsory
stimulation Lund.
R.. Batra.
C‘linlcal htudy on epldural
L.und, C.. Kehlrt,
postoperative
pain.
and Kehlct.
clonidine
II.,
with those of
atresa responses,
cardiopul-
monary function and motor and Aensory block. Br. J. Anaesth..
63
(19X9) 516519. J.. Kehne.
and Davis, Maynard,
M..
J.H., Commissarla.
R..
R.L..
clonidine
inhibits
Brain Res., 276 (1983)
37Y-3X2.
F.M.,
Spinal
Early
spahticity. Paraplegia. Mendez.
R.L.., Lachance.
Agressologie.
space. Pain, Suppl. 4 (1987)
Role of the spinal noradrenrrgic
coeruleus.
M.1. and Eisenach.
intravenous.
par voir epidude dcafferenta-
13’1-142.
Germain.
J.W..
C‘., Clonidine
ou sacree dans les douleurs
tion: possibilities. (1989)
Marwaha, Baillikre’s
clamdinc 71 (lY8Yb)
cancer pain: phase 1. Anesthe-
F. and Grasser,
rale thoraco-lumbar
on
M., Postoperative
Br. J. Anaesth.,
administration.
of clonidine
21-30.
S., Saada.
K. and Ghignone.
Clonidine-induced
B.. New aspect>
and pharmacodynamics
IY
siology, 7 I (1YXYc) 647-657. Farcot, J.M.,
Gode.
J.. Kirsten,
of the pharmacokinetics Bonnet.
methods, J. Pharma-
295-307.
Ther..
and Rose. J.C., Epidural
on rat spmal cord and nerve rootb. Anaesth. for
