![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
|
|
|
|
|
||
Vol. 281, Issue 3, 1381-1391, 1997
Department of Pharmacology, Faculty of Medicine, University of Chile, Santiago, Chile (D.B., C.P.), Laboratoire de Neurophysiologie, Hôpital Pitié-Salpétriêre, Paris, France (J.-C.W.), and INSERM U161, Paris, France (D.L.B.)
 |
Abstract |
|---|
 Top
Abstract
Introduction
Methods
Results
Discussion
References
|
|---|
A C-fiber reflex elicited by electrical stimulation within the
territory of the sural nerve was recorded from the ipsilateral biceps
femoris muscle in anesthetized rats. The temporal evolution of the
response was studied using a constant stimulus intensity (3 times
threshold), and recruitment curves were built by varying the stimulus
intensity from 0 to 7 times threshold. The intrathecal (i.t.) but not
i.c.v. administration of aspirin, indomethacin, ketoprofen and lysine
clonixinate resulted in dose-dependent depressions of the C-fiber
reflex. In contrast, saline was ineffective. Regardless of the route of
administration, the drugs never produced disturbances in heart rate
and/or acid-base equilibrium. When a constant level of stimulation was
used, 500 µg of aspirin i.t. induced a blockade of the reflex
immediately after the injection, followed by a partial recovery.
Indomethacin produced a stable depression, which reached 80 to 90%
with an i.t. dose of 500 µg. Ketoprofen and lysine clonixinate produced a more stable effect; the highest doses (500 µg) produced a
steady-state depression of approximately 50% for approximately 30 min.
When the recruitment curves were built with a range of nociceptive
stimulus intensities, all of the drugs except for indomethacin produced
a dose-dependent decrease in the slopes and the areas under the
recruitment curves without major modifications in the thresholds;
indomethacin also induced a significant dose-related increase in the
threshold. The orders of potency for both stimulation paradigms with
the i.t. route were the same, namely aspirin > indomethacin > lysine clonixinate 
ketoprofen. It is concluded that
nonsteroidal anti-inflammatory drugs elicit significant antinociceptive effects at a spinal level, which do not depend on the existence of a
hyperalgesic or inflammatory state. Such effects were not seen after
injections within the lateral ventricle.
| |
Introduction |
|---|
|
Top
Abstract
Introduction
Methods
Results
Discussion
References
|
|---|
In a previous paper, we reported
the actions of various i.v. administered NSAIDs on a C-fiber reflex
elicited by stimulation within the territory of the sural nerve and
recorded electromyographically from the biceps femoris muscle
(Bustamante et al., 1996a
). This test does not produce
inflammation at the site of nociceptive stimulation and was used to
investigate central analgesic actions because the reflex response is
elicited by direct stimulation of nerve terminals and fibers,
thus bypassing all transduction processes at a peripheral level; it
also allows the examination of responses to a wide range of stimulus
intensities (Strimbu-Gozariu et al., 1993; Falinower
et al., 1994; Guirimand et al., 1994). After i.v.
administration, NSAIDs exhibited an ability to affect essentially
reflexes elicited by suprathreshold stimuli. Although these effects
were weak, they appeared in the absence of any "central facilitation," at least when nontoxic doses were used, as in our previous study. Indeed, with our experimental approach there were no
inflammatory processes or ongoing sustained afferent barrages generated
by irritant sustained stimuli. However, we had no way of knowing
whether the drugs acted at a spinal level, a supraspinal level or both.
We have used i.t. and i.c.v. applications of NSAIDs to examine the
spinal and supraspinal components of the central effects that we
previously observed after i.v. administration. For the present study,
we chose a NSAID from each of four different families of drugs, namely
salicylates (aspirin), indol-acetic acid derivatives (indomethacin),
aryl-propionic acid derivatives (ketoprofen) and fenamates (lysine
clonixinate). After i.t. administration, all of these drugs exhibited
dose-related depressive effects on the responses to suprathreshold
stimuli. This was not the case after i.c.v. administration. These
results have been presented in abstract form (Bustamante et
al., 1996b).
| |
Methods |
|---|
|
Top
Abstract
Introduction
Methods
Results
Discussion
References
|
|---|
General procedure.
The general procedure was essentially
similar to that described previously (Strimbu-Gozariu et
al., 1993; Falinower et al., 1994; Guirimand et
al., 1994; Bustamante et al., 1996a).
Animal preparation. Experiments were carried out in male Sprague-Dawley rats weighing 250 to 300 g. During the surgical procedure, the animals were deeply anesthetized with 2.5% halothane in a nitrous oxide/oxygen mixture (2:1). The surgical procedure consisted of a performing a tracheotomy, cannulating a jugular vein and inserting either a polyethylene catheter i.t. to the lumbar enlargement of the spinal cord or a needle into the lateral ventricle for the administration of drugs.
The method of i.t. injection was the same as that described by Yaksh and Rudy (1976), which permits delivery of small amounts of drug into
the subarachnoid space. Basically, it consists of fixing the animals in
a stereotaxic frame, with the head ventroflexed by means of a metal
bar, and inserting a polyethylene (PE-10) catheter 8.5 cm down the
spinal subarachnoid space through a slit in the cisternal dura and
arachnoid membranes. The actual dorsal location of the i.t. catheter
was checked after a laminectomy at the end of each experiment. When it
was not located properly, i.e., dorsal to the lumbar
enlargement, the results were disregarded. After they were prepared,
the animals were removed from the head-holder and the experimental
protocol was followed.
When i.c.v. injections were to be made, the rats were placed in a
stereotaxic frame and a craniotomy was made over the lateral ventricle,
with care being taken not to disturb the vessels. A microinjection
needle was then placed stereotaxically at the following coordinates:
anteroposterior, +29.6 mm; lateral, +4.8 mm; vertical, +30 mm (with
zero corresponding to the interauricular axis). These animals were
maintained in the head-holder during the electrophysiological and
pharmacological procedures.
After surgery, the concentration of halothane was lowered to 0.9% in
100% oxygen. Throughout the experiments, the animals were artificially
ventilated and the heart rate was monitored. Respiratory rate (50 counts/min), O2, end-tidal CO2 (3.2-3.5%) and
halothane level (0.9%) were monitored continuously using a capnometer
(Capnomac II; Datex Instruments, Helsinki, Finland). The measurements
of CO2 and halothane were made by IR absorption and those
of O2 with a fast paramagnetic analyzer. These parameters were displayed digitally, and each was controlled by an alarm. Body
temperature was maintained at 37 ± 0.5°C by means of a
homeothermic blanket system.
Electrophysiological recordings.
This method has been
described previously (Strimbu-Gozariu et al., 1993;
Guirimand et al., 1994; Falinower et al., 1994). Electrophysiological recordings were made, in the ipsilateral biceps
femoris muscle, of reflex activity evoked by electrical stimulation of
the C-fibers within the receptive field of the sural nerve. A pair of
noninsulated platinum-iridium needle electrodes were inserted s.c. into
the medial part of the fourth and lateral part of the fifth toe. EMG
responses were recorded via another pair of noninsulated
platinum-iridium needles inserted through the skin into the biceps
femoris muscle.
Control period characteristics. All individual experiments began with a control period in which the characteristics of the responses were determined; 20 to 30 min after the end of the surgical preparation and the decrease in the level of anesthesia to 0.9% halothane, the application of 15-mA stimuli to the sural nerve resulted in stable supramaximal reflex responses with minimal spontaneous fluctuations. This was the prerequisite finding before the pharmacological procedures were started. A control recruitment curve was built by increasing the intensity of the stimulus. Reflex responses increased monotonically and reached a plateau at high intensities. The threshold of the C-fiber evoked response was determined as the intersection of the polymodal regression curve and the abscissa. Thereafter, a constant level of stimulation (3 times threshold) was applied. During the first 10 min, the stability of EMG responses was checked. The mean of the 50 successive reflex responses (corresponding to a 5-min period) that immediately preceded the first injection was considered as the mean control value. The constant level of stimulation was applied for 30 min after i.t. or i.c.v. injections. A new recruitment curve was then established.
Pharmacological procedures. To determine the effect and the relative potencies of the i.t. administered NSAIDs on a C-fiber reflex, several doses of each one were used, as follows: aspirin, 10, 50, 100 and 500 µg; indomethacin, 100, 200, 300 and 500 µg; ketoprofen, 100, 200, 300 and 500 µg; lysine clonixinate, 100, 200, 300, 500 and 1000 µg. All drugs were dissolved in saline (0.9% NaCl) to give a total volume of 10 µl for administration. The drugs were injected using a 50-µl Hamilton syringe, at a constant rate, over a 60-sec period. Saline (10 µl) was used as a control.
The effects of i.c.v. NSAIDs on the C-fiber reflex were determined using only one dose for each drug, namely 500 µg of aspirin, 250 µg of indomethacin, 250 µg of ketoprofen and 500 µg of lysine clonixinate. The drugs were dissolved in saline to give a total volume of 5 µl for administration. The drugs were injected using a 20-µl Hamilton syringe, at a constant rate, over a 60-sec period. Again, saline was used as a control. Only one dose of one drug was used in each rat. The rats were sacrificed with an overdose of pentobarbital at the end of the experiments.Processing of data. To analyze the results of the constant-stimulation paradigm, the EMG responses were expressed as percentages of the mean control value and the final individual results were expressed as means of 10 successive responses (corresponding to 1 min of the procedure). A graph of the temporal evolution over a 30-min period was built to evaluate the drug action at the constant level of stimulation; this allowed the magnitude of the effects to be calculated as the AUC by means of a trapezoidal procedure. This was then expressed as a percentage effect (percentage of AUC), where 100% represented complete blockade of the C-fiber EMG signal during the 30-min period.
To analyze the results with the recruitment curve paradigm, each individual EMG response was expressed as a percentage of the maximal C-fiber control response. The stimulus intensities were expressed as multiples of the threshold calculated during the control period, with the maximal response generally occurring at an intensity of 7 times threshold. More than 15 intensity points were used for building each recruitment curve. However, to simplify the processing of data, only nine intensity points, namely those obtained at 1, 1.5, 2, 2.5, 3, 4, 5, 6 and 7 times threshold intensity, were considered. In cases where one of these intensities had not been applied during the experiment, the response was estimated by linear regression between the two nearest intensity points. Such interpolations never involved intensities greater than 2 mA. The AURC values from the control (AURCcontrol) and after-drug (AURCtreated) curves were calculated by a trapezoidal approximation for each animal, and the percentage effect (percentage of AURC) was expressed as
![% AURC<IT>=</IT>[(AURC<SUB>control</SUB><IT>−</IT>AURC<SUB>treated</SUB>)/AUC<SUB>control</SUB>] × 100](/content/vol281/issue3/fulltext/1381/img001.gif)
|
).
One-way analysis of variance was used to analyze the efficacy of drug
treatment and to compare point by point the effects of drugs (or
saline) with controls over whole recruitment curves. Similar analyses
were made for the temporal evolution paradigm for the drug-treated
curves vs. saline. Post hoc comparisons of individual means were made using Fisher's a posteriori
least-significant difference test. Differences were considered
significant at P < .05. Results are expressed as means ± S.E.M. for the number of animals indicated in table 1.
|
Drugs. The following drugs were used: aspirin (lysine acetylsalicylate, Aspegic; Synthélabo, Meudon-la-Forêt, France), indomethacin (Indocid; Merck, Sharp & Dohme, Brussels, Belgium), ketoprofen (Profenid; Roda Mérieux, Santiago, Chile) and lysine clonixinate (pure substance; Roemmers, S.A.I.C.F., Buenos Aires, Argentina).
| |
Results |
|---|
|
Top
Abstract
Introduction
Methods
Results
Discussion
References
|
|---|
General Characteristics of the Reflex Responses and Effects of Saline Administration
Electrical stimulation (2 msec, 0.17 Hz, at a suprathreshold
intensity) within the territory of the sural nerve elicited a two-component EMG reflex response in the ipsilateral biceps femoris muscle. The first component, which had a short latency, a low threshold
and a short duration (<50 msec), was due to activation of myelinated
fibers and was not analyzed in this study. The second component, with a
longer latency (around 160 msec at 3 times threshold), a longer
duration and a higher threshold (6.1 ± 0.3 mA), resulted from
activation of cutaneous unmyelinated afferent fibers; it has been
termed the C-fiber reflex (Strimbu-Gozariu et al., 1993; Falinower et al., 1994; Guirimand et al., 1994)
and has been used to compare the pharmacological profiles of opioids
and NSAIDs administered by an i.v. route (Bustamante et al.,
1996a). In the present study, we focused on this nociceptive flexion
reflex, analyzing the EMG response in a 100- to 450-msec
poststimulus time window (see control example in fig.
1).
|
Saline administration had little effect on either the amplitude or the
duration of the reflex. In the constant-stimulus intensity paradigm,
after i.t. or i.c.v. saline administration, there were only minor
variations in the reflex amplitude during the 30-min recording period
(see the individual example in fig. 2A for i.t. treatment and the averaged results in figs. 3A and 6A
for i.t. and i.c.v. treatments, respectively); consequently, the
percentage variation in AUC at a constant stimulus intensity did not
reach statistical significance.
|
|
|
In the recruitment curve paradigm, the control and treatment curves
(1-7 times threshold) were almost superimposed for the i.t. route
(fig. 4A). However, after i.c.v. administration of 5 µl saline, the curve was slightly above the control. No statistical differences were found between the paired intensity points for the pre-
and post-saline curves in the 1 to 7 times threshold range (see fig.
6A), although overall the increase in percentage of AURC was
significant (table 1). For both routes, the post/pre-saline threshold
ratio was nearly 1, as was the post/pre-saline slope ratio for the i.t.
route. In contrast, the post/pre-saline slope ratio for i.c.v.
treatment was greater than 1, reflecting a small facilitatory effect
after administration of saline into the ventricle (table 1).
|
General Effects of NSAIDs
No perturbations in respiratory gases or cardiac activity were observed during or after either i.t. or i.c.v. administration of any of the drugs tested. The i.t. administration of aspirin, indomethacin, ketoprofen and lysine clonixinate resulted in a dose-dependent depression of the EMG response. Figure 1 shows a typical individual example of the depressive effect of 500 µg aspirin on the C-fiber reflex elicited by an intensity 3 times threshold. Individual examples of the temporal evolution of the reflex responses to C-fiber activation (stimulus currents of 3 times threshold) after i.t. injections of saline and various NSAIDs can be seen in figure 2. In each case, the effect of NSAIDs (500 µg) was rapid in onset and, after a 5- to 10-min period of relative recovery, remained constant for the whole period of the recording. In contrast, the i.c.v. administration of aspirin, indomethacin, ketoprofen and lysine clonixinate resulted in minor nondepressive effects. The effects of the i.t. and i.c.v. administrations are detailed below for each drug for both the constant-stimulation and recruitment curve paradigms.
Effects of i.t. NSAIDs
Effects of i.t. aspirin. As illustrated in figure 2B with an individual example for 500 µg and in figure 3B with the averaged results for three selected doses, aspirin decreased, in a dose-related fashion, the C-fiber reflex elicited by an intensity of 3 times threshold. Whereas 10 µg had no effect (data not shown), 50 µg produced a small but significant depression of the C-fiber reflex. In contrast, after 100, 200 (data not shown) and 500 µg of aspirin, all time points were statistically different (P < .01) from the respective time points in the post-saline curve shown in figure 3A.
The percentage of AUC, which represents the overall effects seen over the 30-min recording period, clearly increased with dose, reaching significance (P < .001) after 100 µg. A dose of 500 µg produced a percentage of AUC of 88%, which represents almost complete abolition of the C-fiber reflex during the recording period. In the recruitment curve paradigm (fig. 4B), we observed a clear dose-related depressive effect in the 50- to 500-µg range. These depressive effects were dependent on the stimulus intensity and had no effect on the threshold (table 1). The post/pre-treatment slope ratios were nearly 1 after 50 and 100 µg but 0.15 after 500 µg (table 1), reflecting the drastic downward shift of the recruitment curve. In contrast to these depressive effects, a small facilitation was observed after 10 µg (data not shown in fig. 4B), as indicated by a negative percentage of AURC value and a post/pre-treatment slope ratio greater than 1 (table 1).Effects of i.t. indomethacin. As illustrated in figure 2C with an individual example for 500 µg and in figure 3C with the averaged results for three selected doses, indomethacin decreased, in a dose-related fashion, the C-fiber reflex elicited by an intensity 3 times threshold. None of the time points of the curve for 100 µg were statistically different from those for saline, except for those seen 1 to 2 min after the injections. In contrast, after 200, 300 (data not shown) and 500 µg, all time points were statistically different (P < .001) from the respective time points in the post-saline curve shown in figure 3A. The percentage of AUC clearly increased with dose, reaching significance (P < .001) after 200 µg and its maximum (84%) after 500 µg.
In the recruitment curve paradigm (fig. 4C), no effects were seen after 100 µg but there were clear dose-related depressive effects in the 200- to 500-µg range (see percentage of AURC in table 1). Interestingly, the thresholds were increased dose-dependently by i.t. indomethacin, with post/pre-treatment threshold ratios of >1 being obtained for all doses (significant for 300 and 500 µg) (table 1). There was a gradual decrease in the post/pre-treatment slope ratios as the dose increased (table 1).Effects of i.t. ketoprofen. As illustrated in figure 2D with an individual example for 500 µg and in figure 3D with the averaged results for three selected doses, ketoprofen decreased, in a dose-related fashion, the C-fiber reflex elicited by an intensity of 3 times threshold. None of the time points of the curves for 100 and 200 µg (data not shown) were statistically different from those for saline, except for those seen 1 to 4 min after the injections. In contrast, after 300 and 500 µg, all time points were statistically different (P < .05 and P < .01, respectively) from the respective time points in the post-saline curve shown in figure 3A. The percentage of AUC clearly increased with dose, reaching significance after 300 µg and its maximum (43%) after 500 µg.
In the recruitment curve paradigm (fig. 4D), no effects were seen after 100 µg but there were clear dose-related depressive effects in the 200- to 500-µg range (see percentage of AURC in table 1). Ketoprofen did not change the post/pre-treatment threshold ratios but did decrease, in a dose-dependent fashion, the post/pre-treatment slope ratios (table 1).Effects of i.t. lysine clonixinate. As illustrated in figure 2E with an individual example for 500 µg and in figure 3E with the averaged results for three selected doses, lysine clonixinate decreased, in a dose-related fashion, the C-fiber reflex elicited by an intensity of 3 times threshold. None of the time points of the curve for 100 µg were statistically different from those for saline, except for those seen 1 to 2 min after the injections. However, from 10 min after administration of 200 µg (data not shown), all time points were statistically different (P < .05) from the respective time points for saline. After 300, 500 and 1000 µg, all time points were statistically lower (P < .001) than the respective time points in the post-saline curve.
In the recruitment curve paradigm (fig. 4E), there was a clear dose-related depressive effect in the 100- to 1000-µg range (see percentage of AURC in table 1). Lysine clonixinate did not change the post/pre-treatment threshold ratios in the 100- to 300-µg range but increased them in the 500- to 1000-µg range. The drug did decrease, in a dose-dependent fashion, the post/pre-treatment slope ratios (table 1).Comparison of i.t. NSAIDs and Relative Potencies
The dose-response curves obtained in both the constant-stimulation
and recruitment curve paradigms are shown in figure 5. These were established on the basis of the variations of AUC and AURC,
respectively. The effective doses that produced 50% depression were
used for comparisons. Based on the ED50 values with 95%
confidence intervals, the following order of potencies was established:
1) in the constant-stimulation paradigm: aspirin [148 (126-164)
µg] > indomethacin [236 (182-266) µg] > ketoprofen [724
(669-754) µg] lysine clonixinate [639 (583-650) µg]; 2) in
the recruitment curve paradigm: aspirin [172 (149-187) µg] > indomethacin [250 (155-336) µg] > lysine clonixinate [586
(500-650) µg] ketoprofen [627 (525-699) µg].
|
In the constant-stimulation paradigm, the slope of the indomethacin curve was significantly higher (P < .05) than the slopes of the ketoprofen or clonixinate curves but not that of the aspirin curve; there were no significant differences between the values for the aspirin, ketoprofen and clonixinate curves. In the recruitment curve paradigm, there were no significant differences between the slopes of the dose-response curves, although a trend similar to that seen with the constant-stimulation paradigm was observed.
Effects of i.c.v. NSAIDs
A general overview of the results obtained with i.c.v. administration of the drugs shows that they did not produce depressive effects, despite the high doses used. As illustrated in figure 6 (left) for the constant-stimulation paradigm, none of the drugs were able to decrease the magnitude of the C-fiber reflex during the 30-min recording period. The apparent small facilitatory effects were not statistically different from the respective time points in the post-saline curve shown in figure 6A (left). Similar results were obtained with the recruitment curve paradigm (fig. 6, right). The post-treatment curves were slightly above the corresponding control curves, which possibly reflected a small facilitatory effect; however, only a few intensity points were statistically different from the controls. These data are summarized in table 1 in terms of percentage of AURC, post/pre-treatment threshold ratios and slope ratios for the recruitment curve paradigm; none of these parameters showed significant effects.
| |
Discussion |
|---|
|
Top
Abstract
Introduction
Methods
Results
Discussion
References
|
|---|
This study shows that i.t. but not i.c.v. administration of
various NSAIDs results in dose-dependent depression of a nociceptive flexion reflex elicited by activation of C-fiber afferents in the sural
nerve of anesthetized rats. This work is of particular interest for the
following reasons. 1) It involved the use of a simple noninvasive
method for recording and analyzing a C-fiber reflex that previously had
been shown to be very sensitive to i.t. µ-opioid agonists in
anesthetized animals that had undergone minimal surgical preparation
(Strimbu-Gozariu et al., 1993; Guirimand et al.,
1994). The model is also sensitive to the systemic administration of
high doses of NSAIDs (Bustamante et al., 1996a). Compared
with most pharmacological tests related to nociception and analgesia, which involve either threshold measurements or responses to a single
suprathreshold stimulus, we introduced a new parameter in these
studies, namely the use of stimulus-response recruitment curves. Thus,
information regarding the effects of drugs on the responses to
threshold and suprathreshold stimuli can be obtained. 2) The C-fiber
reflex was elicited by direct stimulation of nerve terminals and
fibers, thus bypassing all transduction processes at a peripheral
level. Moreover, because NSAIDs were injected directly into the spinal
cord, any effects obtained with such a method must be interpreted in
terms of central mechanisms. 3) No antinociceptive effects were
obtained when the NSAIDs were injected into the lateral ventricle. 4)
The depressive effects observed after the direct spinal administration
were achieved in the absence of any inflammatory processes or ongoing
sustained afferent barrages generated by an irritant sustained stimulus (in other words, in the absence of any "peripheral or central facilitation"); thus, these results firmly support a central
analgesic action of NSAIDs on the spinal transmission of nociceptive
signals. We discuss successively the antinociceptive effects of i.c.v. NSAIDs, the antinociceptive effects of i.t. NSAIDs, the effects of i.t.
NSAIDs on threshold and suprathreshold responses and some possible
mechanisms of action.
Antinociceptive effects of i.c.v. NSAIDs.
Direct i.c.v.
administration of NSAIDs has been reported to be active in animal
models of experimental pain where inflammatory processes are involved
and hyperalgesia is obvious. For example, Ferreira et al.
(1978) reported that i.c.v. aspirin (200-400 µg), indomethacin
(50-100 µg) or paracetamol (200-400 µg) decreased the
hyperalgesia elicited by an injection of carragenin into the plantar
surface of the rat paw. These results were confirmed with other tests
in rats; i.c.v. paracetamol, indomethacin, diclofenac, ibuprofen and
metamizol suppressed hyperalgesia after postischemic reperfusion
(Gelgor et al., 1992), i.c.v. indomethacin diminished electrically induced vocalization responses in arthritic rats (Okuyama
and Aihara, 1984), i.c.v. ketoprofen diminished thalamic neuronal
firing elicited by ankle mobilization in arthritic rats (Braga, 1990)
and increased the tail-flick latency (Rampin et al., 1988),
i.c.v. diclofenac decreased the number of abdominal constrictions
(Björkman et al., 1990, 1992) and salicylate inhibited the lame-walking reaction in adjuvant-induced, hind-paw-edematous rats
(Higushi et al., 1986).
), 100 µg
of ketoprofen or ibuprofen with the lame-walking reaction in
adjuvantinduced, hind-paw-edematous rats (Higushi et
al., 1986) and 250 to 500 µg of aspirin with the mouse-licking
behavior in the formalin test (Meglio et al., 1991). It is
possible that the supraspinal effects of NSAIDs are triggered only in
the presence of a hyperalgesic or inflammatory state, whereas we are
dealing presently with a model of acute nociception.
Antinociceptive effects of i.t. NSAIDs.
Antinociceptive
effects produced by the i.t. administration of various NSAIDs have been
reported in studies using several models of experimental pain. Malmberg
and Yaksh (1992a), using the formalin test, reported antinociceptive
effects of i.t. NSAIDs and concluded that their actions were related to
their ability to block prostanoid formation (Malmberg and Yaksh,
1992b). Ten micrograms of diclofenac have been reported to decrease the
number of abdominal constrictions in the abdominal constriction test in
mice (Björkman et al., 1990); antinociceptive effects
have also been reported with the tail-flick and paw-withdrawal tests in
rats (Taiwo and Levine, 1988; Wang et al., 1994) and the
colorectal distension model in rabbits (Jensen et al.,
1992). However, using the minimum alveolar anesthetic concentration as
an index of pain, Antognini (1993) reported a lack of effect, on the
response to a suprathreshold mechanical stimulus, of i.t. doses as
high as 500 µg/kg aspirin or indomethacin in rabbits. A general
analysis of the results reviewed above shows that the range of active
doses was variable, depending on the NSAID tested, the test used and the species. However, all active doses were in the 10- to 500-µg range, which is comparable to the effective doses in the present study.
). The preparation used by those authors was similar, in
terms of species and anesthesia, to that used presently and the dose
was close to the ED50 (236 µg) that we found. We have
mentioned already that the model used here seems very sensitive to
opioids, because low doses of i.t. morphine strongly depressed the
C-fiber reflex, with an ED50 of 0.2 µg (Strimbu-Gozariu
et al., 1993). Dickenson and Sullivan (1986) reported a
dose-dependent depression of the C-fiber evoked activity of dorsal horn
neurons after i.t. morphine, with an ED50 of 5.5 µg. The
apparently greater sensitivity of the present preparation to analgesic
drugs, compared with recordings from dorsal horn neurons, could be due
to the convergence of inputs onto motoneurons, as suggested by the
higher mean threshold for the C-fiber reflex, compared with the mean
threshold for C-fiber responses of dorsal horn neurons.
Jurna et al. (1992), using stimuli similar to those in the
present study, reported that i.t. aspirin (37.5-50 µg) or
indomethacin (50-100 µg) depressed the responses, to stimulation of
the sural nerve, of single cells in the ventrobasal complex of the
thalamus. As in the present study, those results were obtained with a
method that bypassed all transduction processes at the peripheral level and did not involve any inflammatory processes. Note that the relative
potencies of aspirin and indomethacin were similar to our findings,
although the efficacious doses were slightly lower.
Finally, the potential of aspirin to exhibit a direct spinal effect for
alleviating pain in humans has been confirmed in clinical trials with
patients, including those suffering from cancer pain. Devoghel (1983)
successfully used 500 µg/kg, whereas Pellerin et al.
(1987) administered 120 to 750 mg of i.t. aspirin. Interestingly, because we used rats with body weights of 250 to 300 g, the
ED50 for aspirin was in the 420 to 656 µg/kg range, which
is close to the efficacious dose used by Devoghel (1983).
Effects of i.t. NSAIDs on threshold and suprathreshold
responses.
The study using recruitment curves revealed that the
spinal administration of NSAIDs did not produce significant
modifications of the C-fiber reflex elicited by the lower (<2.5 times
threshold) stimulus intensities. In contrast, all NSAIDs clearly
depressed the response to the higher stimulus intensities, in a
dose-dependent fashion. Using the same approach, similar results were
found after i.v. administration of the same drugs (Bustamante et
al., 1996a). Interestingly, µ-opioid agonists not only produced
a shift in the encoding functions of the spinal cord but above all
reduced the gain of these functions, which produced a blockade of the C-fiber reflex regardless of the stimulus intensity; for instance, 1 µg i.t. or 10 mg/kg i.v. morphine completely suppressed all EMG
responses elicited by the activation of C-fibers (Strimbu-Gozariu et al., 1993; Guirimand et al., 1994, 1995a).
Interestingly, high doses of the opioid receptor partial agonist
buprenorphine depressed the C-fiber reflex, but only at weak stimulus
intensities (Guirimand et al., 1995a). Thus, from the use of
the recruitment curve paradigm, it appears that analgesic drugs exhibit
at least three pharmacological profiles, acting on responses to low,
high or all stimulus intensities.
). In
other words, the order of potency was identical for indomethacin, ketoprofen and lysine clonixinate but aspirin moved from last to first
place when the route of administration was changed from i.v. to i.t.
This observation is consistent with the fact that the rates of
diffusion of acetylsalicylic acid and its metabolite salicylic acid
through the blood-brain barrier are slower than those of ketoprofen and
indomethacin (Netter et al., 1985; Ochs et al.,
1985; Bannwarth et al., 1989, 1990). The constant-stimulus intensity paradigm allowed us to investigate the temporal evolution of
the depressive action of each of the NSAIDs. In the present study, the
effects of the i.t. drugs had an extremely rapid onset, suggesting that
the active pharmacological sites of action were reached easily with
this route of administration. This was not the case after i.v.
administration, where the effects of aspirin were characterized by a
slow progressive increase, whereas the actions of indomethacin,
ketoprofen and lysine clonixinate peaked quickly, in a manner very
similar to that of the effects described in the present study.
Possible mechanism(s) of action.
As previously stated, the
C-fiber reflex was elicited by stimulation of nerve terminals and
fibers, thus bypassing all transduction processes at a peripheral
level. In addition, the drugs were applied directly over the spinal
cord. As a result, the NSAIDs tested in the present study are likely to
have had a central action at the spinal level, as proposed previously
(for reviews, see Brune, 1994; McCormack, 1994a,b; Björkman,
1995). However, the hypothesis that NSAIDs act at higher levels within
the central nervous system and activate descending inhibitory
influences on the spinal transmission of nociceptive information has
been proposed for metamizol and aminophenazone (Carlsson and Jurna,
1987; Carlsson et al., 1986, 1988). It has also been
suggested that monoaminergic mechanisms, e.g., through
bulbospinal noradrenergic or serotonergic inhibitory systems, might
account for the efficacy of NSAIDs (Shyu et al., 1984;
Groppetti et al., 1988; Taiwo and Levine, 1988;
Tjölsen et al., 1991). Interestingly, the effects of
i.t. paracetamol in the paw-pressure test were abolished by the i.t.
administration of tropisetron, a 5-hydroxytryptamine3
receptor antagonist (Pelissier et al., 1995). A supraspinal
site of action on the spinal transmission of nociceptive information is
strongly suggested in humans, at least for ketoprofen, because the
drug-induced increase in the threshold of the RIII reflex
seen in normal volunteers was not seen in paraplegic patients with
chronic spinal cord sections (Willer et al., 1989). However,
the hypothesis that NSAIDs act at higher levels within the central
nervous system and activate descending inhibitory influences on the
spinal transmission of nociceptive information is not supported by the
present findings, where NSAIDs administered via the i.c.v.
route were without effect.
), possibly against prostaglandin biosynthesis, as
suggested by the following lines of evidence. Intrathecal
administration of prostaglandins has been reported to increase
nociceptive responses (Yaksh, 1982; Taiwo and Levine, 1988; Uda
et al., 1990; Minami et al., 1994a,b; Nishihara
et al., 1995; Saito et al., 1995). Activation of
high-threshold nociceptive afferent inputs has been found to release
prostaglandins within the spinal cord (Ramwell et al., 1966;
Coderre et al., 1990), which, in turn, can enhance
Ca++ conductance and the subsequent release of substance P
from dorsal root ganglion cells (Nicol et al., 1992).
Furthermore, microdialysis studies in conscious rats have revealed that
formalin injections into the hind paw elicit the spinal release of
prostaglandin E2, in a dose-dependent fashion, which is
reduced by pretreatment with morphine (Malmberg and Yaksh, 1995a) or
cyclooxygenase inhibitors (Malmberg and Yaksh, 1995b). Finally, the
i.t. administration of prostaglandin E receptor antagonists produces
antinociceptive effects in the formalin test in rats (Malmberg et
al., 1994).
One cannot exclude the possibility that other and/or complementary
mechanisms might explain the results reported herein. In fact, an
almost complete dissociation between analgesic efficacy and the potency
to inhibit cyclooxygenase has been demonstrated for some NSAIDs (Brune
et al., 1991, 1992; McCormack and Brune, 1991). It is
possible that cyclooxygenase inhibitors have additional mechanisms of
action (Vane, 1994); such possibilities were reviewed extensively
(McCormack, 1994b; Bannwarth et al., 1995; Björkman, 1995).
Because i.t. NSAIDs block the behavioral signs of hyperalgesia elicited
by an i.t. excitatory amino acid such as NMDA and 
-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid or substance P
(Hunskaar et al., 1985; Malmberg and Yaksh, 1992b;
Björkman et al., 1994), they have the potential to
impair both the release and the effects of neurotransmitters or
neuromodulators involved in the spinal transmission of nociceptive
information. Direct or indirect effects of NSAIDs on NMDA receptors are
suggested in the present study by the depressive action of the drugs on the responses to suprathreshold stimuli, with minimal effects on
responses to lower levels of stimulation. The NMDA receptor is blocked
by Mg++ ions in a voltage-dependent fashion (Asher and
Nowak, 1987), which means that only very strong excitatory signals
cause a depolarization of neurons sufficient to activate NMDA
receptors. When the NMDA receptor is activated by glutamate, the rise
of Ca++ ions in the cell activates phospholipase
A2, resulting in the formation of arachidonic acid from
membrane phospholipids (Dumuis et al., 1988). Because
arachidonic acid potentiates NMDA currents, increases the release of
glutamate from presynaptic terminals and impairs its glial reuptake
(Chan et al., 1983; Rhoads et al., 1983; Miller
et al., 1992), it follows that arachidonic acid would amplify the activation of NMDA receptors. Whether a NSAID-sensitive mechanism is involved in such an action remains to be investigated; for
this purpose, a study of the effects of glucocorticoids on the C-fiber
reflex could be interesting.
In summary, we have reported the antinociceptive potency of i.t.
NSAIDs, belonging to different families of drugs, in a model of acute
pain. These antinociceptive effects occurred without any facilitatory
or hyperalgesic components. These findings suggest a crucial role for
cyclooxygenase products in the spinal transmission of acute nociceptive
signals when the intensity (afferent barrage) reaches a sufficient
level. Because other analgesic drugs, such as buprenorphine, depress
C-fiber responses only to low intensities of stimulation (Guirimand
et al., 1995a,b), it is tempting to suggest a rational
therapeutic association for these two types of compounds.
| |
Acknowledgments |
|---|
The authors thank Drs. J. M. Benoist and S. W. Cadden for advice on the preparation of the manuscript.
| |
Footnotes |
|---|
Accepted for publication February 3, 1997.
Received for publication October 3, 1996.
1 This work was supported by INSERM. D.B. was supported by a fellowship from Roemmers Labs, Argentina.
Send reprint requests to: Dr. D. Le Bars, INSERM U-161, 2, rue d'Alésia, 75014 Paris, France.
| |
Abbreviations |
|---|
AUC, area under the curve; AURC, area under the recruitment curve; EMG, electromyographic; i.t., intrathecal; NMDA, N-methyl-D-aspartate; NSAID, nonsteroidal anti-inflammatory drug.
| |
References |
|---|
|
Top
Abstract
Introduction
Methods
Results
Discussion
References
|
|---|
-aminobutyric acid and glutamate uptake and (Na++K+)-ATPase activity in brain slices and synaptosomes by arachidonic acid.
J. Neurochem.
40: 309-316, 1983[Medline].
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
 |
 |
|
 |
 |
 |
1.2 ± 1.7% (data not shown). The regression lines were as follows:
indomethacin, y = 113 log x