- Email: [email protected]
Topical application of clonidine relieves hyperalgesia in patients with sympathetically maintained pain
Pain, 47 (1991) 309-317
309
¢) 1991 Elsevier Science Publishers B.V. All rights reserved 0304-3959/91/$03.50
PAIN 01899
Topical application of clonidine relieves hyperalgesia in patients with sympathetically maintained pain K.D. Davis a,**, R.D. T r e e d e ~'*, S.N. Raja a Departments o f Neurosurgery,
b, R.A.
Meyer a,c and J.N. Campbell ~,c
b Anesthesiology and ~ The Applied Physics Laboratory, Johns Hopkins University, Baltimore, MD 21205 (U.S.A.)
(Received 4 December 1990, revision received 12 April 1991, accepted 15 April 1991)
Summary Patients with reflex sympathetic dystrophy or causalgia characteristically have ongoing pain and pain to light touch (hyperalgesia). Some of these patients obtain relief of their pain following interruption of sympathetic function to the affected area and, therefore, have sympathetically maintained pain (SMP). Evidence suggests that the pain and hyperalgesia in SMP are related to activation of peripheral adrenergic receptors. We wished to determine the contribution of alpha 1- and alpha2-adrenergic receptors in SMP and thus examined the effects of local application of adrenergic agents in patients with SMP. The alpha2-adrenergic agonist clonidine, available as a transdermal patch, was delivered topically to the patients' hyperalgesic skin. In four patients with SMP, clonidine eliminated or substantially reduced hyperalgesia to mechanical and cold stimuli. In three of these patients the effects were confined to the skin region beneath the patch, suggesting a peripheral and not central effect. The relief of hyperalgesia was not due to a local anesthetic effect since touch thresholds were unaffected. Topical clonidine did not relieve hyperalgesia of similar severity for two other patients whose hyperalgesia and pain were unaffected by sympathetic ganglion blocks (i.e., diagnosed as having sympathetically independent pain). In two SMP patients, intradermal injection of norepinephrine or phenylephrine (a specific alphat-adrenergic agonist) at a site treated with clonidine evoked intense pain and rekindled the pre-clonidine hyperalgesia at that site. It is likely that clonidine locally blocks the release of norepinephrine via activation of alpha 2 receptors on the sympathetic terminals. This study suggests, therefore, that SMP is mediated via alphal-adrenergic receptors located in the affected tissue. Key words: Reflex sympathetic dystrophy; Causalgia; Allodynia; alpha-Adrenergic receptors; Norepinephrine; Phenylephrine
Introduction Sympathetically maintained pain (SMP) refers to a chronic pain disorder in which the pain is dependent on sympathetic innervation of the painful part [22]. Thus, patients are diagnosed as having SMP if blockade of the sympathetic innervation to the affected area alleviates their pain [4,22]. Patients with SMP typically
* On leave from the Institute of Physiology, University Hospital Eppendorf, D-2000 Hamburg 20, F.R.G. * * Present address: Department of Physiology, University of Toronto, Toronto, Ontario, Canada. Correspondence to: Dr. James N. Campbell, Department of Neurosurgery, Meyer 7-113, Johns Hopkins University, Baltimore, MD 21205, U.S.A.
have both stimulus-independent (ongoing) pain and stimulus-dependent pain (hyperalgesia). Hyperalgesia is defined as a leftward shift of the stimulus-response function, such that a lowering of pain threshold and/or an increase in pain to suprathreshold stimuli are observed [5]. The decrease in pain threshold to mechanical stimuli may be such that lightly stroking the skin evokes pain, a phenomenon sometimes referred to as allodynia (here considered to be part of hyperalgesia). Several lines of evidence suggest that peripheral adrenergic receptors are involved in SMP. (1) Stimulation of the peripheral but not central cut end of the sympathetic chain reproduces pain in causalgia patients after sympathectomy [29]. (2) Local anesthetic blockade of the appropriate sympathetic ganglion or adrenergic blockade via intravenous administration of phentolamine rapidly abolishes sympathetically main-
310 tained pain and hyperalgesia [21,28]. (3) Depletion of peripheral catecholamines by regional intravenous guanethidine also relieves pain and hyperalgesia [13,17]. (4) Intradermal injection of norepinephrine rekindles the pain and hyperalgesia that had been relieved in patients by sympathectomy or sympathetic block [31] but does not cause pain or hyperalgesia in normal subjects. (5) The non-specific alpha-adrenergic antagonist phenoxybenzamine [12] and the specific alpha r adrenergic antagonist prazosin [1] can be effective in relieving pain in patients with SMP. (6) The betaadrenergic antagonist propranolol has little effect on SMP [21,25]. These data suggest that SMP may be a peripheral alpha cadrenergic receptor disease. Conventional treatments for SMP include repeated local anesthetic sympathetic blocks, intravenous regional guanethidine/reserpine blocks, surgical sympathectomy or oral sympatholytic therapy. However, each of these treatments carries with it a degree of risk, side effects and discomfort. Therefore, the topical application of an alpha-adrenergic blocking agent could be an alternate approach in the treatment of SMP. No alpha-adrenergic antagonist is commercially available for topical use at present. However, clonidine, an alpha2-adrenergic agonist commonly used in the treatment of hypertension, is available as a patch for transdermal administration. Clonidine activates pre-synaptic adrenergic autoreceptors resulting in a reduction of norepinephrine release [27]. This effectively decreases activation of the post-synaptic alpha~ receptors. In this study, the therapeutic benefit of peripheral alpha-adrenergic blockade with clonidine was examined in patients suspected to have SMP. To explore the involvement of peripheral alpha j- and alpha2-adrenergic receptors in SMP, these patients were treated with topical clonidine and, in some cases, were subsequently given intradermal injections of alpha-adrenergic agonists.
Methods
Patient selection and control subjects Six normotensive patients with chronic ongoing pain and cutaneous hyperalgesia to mechanical and cooling stimuli following soft tissue or nerve trauma entered the study. All patients had previously undergone sympathetic blocks (i.e., local anesthetic blockade of the appropriate sympathetic ganglia) to assess the involvement of the sympathetic nerves in their pain state. Additionally, in all but one patient (case 2), a systemic alpha-adrenergic block was performed via intravenous administration of phentolamine [see ref. 21 for details]. Four of the six patients (cases 1-4) experienced 70100% pain relief following the block and were considered to have SMP. The remaining two patients (cases 5
and 6) were considered to have sympathetically independent pain (SIP) since their pain was rcduccd by < 25% following local anesthetic, sympathetic ganglion block or the phentolamine block. Five normal subjects entered the study as controls. One subject was used to investigate the effect of clonidine on normal skin (i.e.. the top of the foot). The four remaining subjects were used to assess the effect of intradermal norepinephrine and phenylephrine in norreal skin. Topical clonidine Clonidine was administered to the hyperalgesic skin via a 7.0 or 10.5 cm 2 patch (Catapres-TTSR-2 and TTSR-3, Boehringer, Inge[heim). These patches deliver a systemic dose of 0.2 mg and (1.3 rag, respectively, of clonidine/day (i.e., 30 p,g/cm2/day) for a maximum of 7 days. A series of 2-7 patches were applied consecutively to different sites within the hyperalgesic zone. Each patch was left in place for 2-10 days within this affected zone. Prior to, during and immediately after drug application, the following parameters were monitored: heart rate, blood pressure, ongoing pain and pain to mechanical and cold stimuli (see below). Sensor)' testing To assess the local and systemic effects of clonidine, the following measures were used before application of the patch and immediately after removal of the patch. (1) Ongoing pain: stimulus-independent pain was assessed on a visual analog scale (VAS). The scale consisted of a 100 mm line where 'no pain' and 'most intense pain' were indicated on the left and right ends of the scale, respectively. (2) Hyperalgesia to mechanical and cooling stimuli: pain evoked by mechanical and cold stimuli was rated verbally on a scale from 0 (no pain) to 10 (most intense pain). The mechanical stimuli included innocuous brushing with a camel's hair brush and innocuous pressure from the 206 g weight of a 13 mm diameter thermally neutral probe. These mechanical stimuli are innocuous when applied to normal skin. Pain thresholds to mechanical stimuli were determined in the area within and surrounding the patch sites using calibrated von Frey filaments. The cooling stimulus consisted of a small drop of acetone lightly placed onto the patient's skin. Patients with SMP typically find that the 1-2 °C decrease in temperature induced by such a stimulus is painful [11]. (3) Tactile sensibility: to rule out local anesthetic effects, the detection (touch) thresholds were determined in the area within and surrounding selected patch sites using calibrated von Frey filaments. In addition, the patient's ability to distinguish the blunt from the sharp end of a pin was also assessed. The patients were instructed to close their eyes during the hyperalgesia and yon Frey threshold testing procedures.
311
Effects of intradermal norepinephrine and phenylephrine In two patients, intradermal injections of the nonspecific alpha-adrenergic agonist norepinephrine (5/xg in 10 ~1 normal saline) or the alphal-adrenergic agonist phenylephrine (10/zg in 10/xl normal saline) were made at a patch site immediately after removing the patch. Norepinephrine and phenylephrine injections were also made at control sites on the patients' normal limbs and into the normal skin of four healthy control subjects. Similar injection sites were chosen for the control subjects and patients. All injections were made in a single-blinded fashion. The patients and control subjects were instructed to rate the magnitude of any evoked pain and resultant hyperalgesia on a verbal scale from 0 to 10.
Results The results of the clonidine applications in the control subject and six patients are presented below. The first patient was studied extensively and is thus presented in detail as an illustrative example.
•
zone
of hyperalgesia /
//
/
/
[[ %'
1
%ITF
--
I at patch 2
outside
5 ....
*verbal
ratings
patch
trolateral
/
/
~/
/ /
/
/
// /
/ /
/
//
i /
// // / / / / / / / /-//
/
/
// /
// //
~ / / /
/
~
//,/ //~
/j /1 d
"
I[ I-HRESHOLD (bars)l I PAIN RATING* lldetection pain llbrush pressure II II l[ -~ II 4 12 11 0 O il II II II
4
b
!! ii
2
10
on a O ( " n o
pain")
t o 10 ( " m o s t
II 6 . 5 II !! ii
10
0
intense
0
pain")
scale
Fig. 1. Example of the effect of a clonidine patch applied to the hyperalgesic skin of a patient with sympathetically maintained pain (case 1). Sensory testing was performed immediately following removal of the patch that had been in contact with a previously hyperalgesic site for 36 h. Site 1: at the patch site, pain to stroking the skin with a soft brush and to pressure (2N) from a thermally neutral probe was eliminated (right hand column of table). The touch detection and pain thresholds determined with van Frey probes (left hand column of table) were similar to those at a contralateral site (site 3) on normal skin. Site 2: immediately adjacent to the patch, the brush and pressure stimuli evoked pain and the pain threshold was lower than the contralateral normal skin site.
Case 1 The patient was a 30-year-old, normotensive, Caucasian female who sustained an injury to the sciatic nerve from hip replacement surgery 3 years prior to entering our study. Subsequently, the patient developed continuous pain and hyperalgesia in the anterolateral aspect of her left leg from the knee to the foot (Fig. 1). T h e r m o g r a p h y revealed that most of this area was 1 - 2 ° C colder than corresponding parts of the contralateral leg. Within the affected region, the patient experienced intense ongoing pain and was exquisitely sensitive to mechanical stimuli. Mild mechanical stimuli (e.g., light brushing or pressure) evoked severe pain. Cooling stimuli (e.g., acetone) were not detected when applied to some parts of the affected area and were painful when applied to other parts. The results of local anesthetic sympathetic ganglion blocks and i.v. phentolamine blocks confirmed that her ongoing pain and hyperalgesia were sympathetically maintained. This patient received a series of seven elonidine patches. A typical example of the effect of a clonidine patch is illustrated in Fig. 1. This patch was applied for 36 h to an area of skin which had previously been hyperalgesic to brushing, pressure and cooling stimuli. After removal of the patch, mechanical (brush, pressure) and cooling (a drop of acetone) stimuli applied to this region were detected but were not painful, although these stimuli still elicited pain when applied to the adjacent skin. The mechanical detection thresholds inside and outside the patch site were identical and were similar to the contralateral leg. Within the patch site, the pain threshold to van Frey filaments approxi-
mated the patient's ' n o r m a l ' pain threshold at a corresponding contralateral site and far exceeded the mechanical detection threshold. The ability to distinguish sharp from blunt stimuli was not affected at this site after the patch was removed. Outside the patch site, the pain threshold was low and approached the detection threshold. The patient's ratings of her stimulus-independent pain were not affected by the patch. Although the patient was pleased with the relief achieved from the patches, treatment was discontinued when the patches started to produce skin irritations. Six months following termination of the clonidine trial, she underwent a lumbar sympathectomy that completely eliminated her pain and hyperalgesia.
Extent of pain and hyperalgesia relief by topical clonidine A summary of the effect of four of the clonidine patches in which quantitative sensory testing was performed for case 1 is shown in Fig. 2. Each clonidine patch greatly reduced and, in many cases, completely eliminated the patient's hyperalgesia to mechanical stimuli. A confounding local anesthetic effect did not occur since detection thresholds were unchanged in the three patch sites tested (see Fig. 2A). The time course of the effect of each patch was not evaluated quantitatively. However, the patient did note a reduction of her mechanical hyperalgesia within 36-48 h after the patch was applied to the skin. After removing a patch, hyperalgesia was still absent for at least 12 h. Typically, the
312 ~20
A
O--•case
1
r - l - - ~ normot subject -o15 o
~I0
J~
g 5
i
ID C3
• []
0
BEFORE
2O
AFTER
B
?
/
•
O
~15 "O
[]
-6
•
P .12 F--
c: 5' O0¸
10 "6
BEFORE
AFTER
BEFORE
AFTER
C
8
EL
60
I
4.
03 2
10. .c o
n
D
8
13
o 6
> LIJ
I
~4
patient's hyperalgesia returned within less than a week of patch removal. Thc three remaining patients diagnosed as having SMP achieved relief of their hyperalgesia l'ollowing topical application of clonidine without any accompanying change in touch detection threshold (see Table 1: cases 2-4). The original zone of hypcralgcsia in two of these patients (cases 2 and 3) encompasscd more than half of the lower extremity. For these patients, topical clonidine only eliminated hyperalgesia at or near the site of drug application and did not alter stimulusevoked pain outside these sites. Furthermore, these patients did not report any reduction in their overall level of stimulus-independent pain. In case 4, the zone of pain and hyperalgesia radiated laterally from a small region above thc vermillion border in the maxillary nerve distribution. Following clonidinc application to the upper lip, complete relief of hypcralgcsia was obtained throughout the entirc affected zone. The patient's ongoing pain was reduced by 5()-75% and was confined to a very small region above the lip. One of the patients diagnosed as having SIP (see Table 1: case 5) had pain and hyperalgesia in the median nerve distribution of the left hand. Neither the patient's hyperalgesia nor ongoing pain werc improved by application of clonidine to the palmar aspect of the hand. Thc second patient diagnosed as having SIP (case 6) had pain and hyperalgesia in the foot. The clonidine patches were applied to the dorsum of the foot and resulted in only a slight reduction (by approximately 35%) in both ongoing pain and in mechanical hypcralgesia to brush stimuli. The hyperatgesia to cold stimuli was not affected by the clonidinc patches.
Side effects The side effects from these clonidine patches were minor. The patients occasionally complained of feeling sleepy, thirsty a n d / o r of having dry eyes. Following some of the patch applications, an erythematous rash was observed surrounding the patch site. This became a problem in two patients (cases 1 and 4), and further treatment had to be discontinued despite good hyperalgesia relief. No appreciable alteration in the blood pressure or heart rate was apparent in any of the patients.
~213_
Sensory effects of topical clonidine in normal skin
OBEFORE
AFTER
Fig. 2. Sensory testing before and after four different clonidine patches applied to different areas of hyperalgesic skin of a patient (case 1) with sympathetically maintained pain (solid circles) and after a single patch applied to a normal subject (open squares). Detection (A) and pain (B) thresholds at the patch site were determined with von Frey filaments. Pain evoked by light brushing (C) and firm pressure (D) was rated verbally on a 0 - 1 0 scale. Each line corresponds to results from a single patch.
To exclude further local anesthetic effects of clonidine, a clonidine patch was applied to the top of a normal subject's foot and sensory testing was performed prior to and 40 h after patch application. Neither the pain nor detection threshold to von Frey filaments were affected in this subject. Mechanical and cooling stimuli were detected but were not painful before or after clonidine application.
313
TABLE I S U M M A R Y OF I N D I V I D U A L CASE RESULTS SMP = sympathetically maintained pain; SIP = sympathetically independent pain; + = present; - = absent; nt = not tested. Case number
Age/sex Duration of symptoms (years) Extremity affected Mechanical hyperalgesia Pre-clonidine Post-clonidine Cold hyperalgesia Pre-clonidine Post-clonidine Touch detection threshold Pre-clonidine (bars) Post-clonidine (bars)
1
2
3
4
5
6
(SMP)
(SMP)
(SMP)
(SMP)
(SIP)
(SIP)
30/F 3 lower
45/F 1.5 lower
35/M 18 lower
52/F 2 face
52/M 7 upper
42/F 0.5 lower
+
+
+
+
+
_
_
--
+
+
+
+
+
+
+
-
nt
-
+
+
1
1.5
0.5
1
2
l
1.5
0.5
1
2
+
+
4 4
only mildly hyperalgesic (rated 1 out of 10) whereas brushing the adjacent untreated skin evoked a pain sensation rated as 4 out of 10. The norepinephrine injection evoked an intense, burning pain sensation that subsided after 4 min (see Fig. 3A). Twenty:five minutes after injection, the patient's mechanical hyperalgesia at the patch site was rekindled and was rated as 3.5 out of 10 (see Fig. 3B). Injection of norepinephrine
Effect of norepinephrine and phenylephrine To test the hypothesis that clonidine acts peripherally via reduction of norepinephrine release, an intradermal injection of norepinephrine (5 /zg in 10 /zl saline) was made at a previously hyperalgesic site that had been treated effectively with clonidine (case 1). Following removal of the clonidine patch and just prior to injection, lightly brushing the skin at this site was A
10. O~
8¸
c
6 I
N
-
C 0
=,--n
mm~_~
case 1: patch site case 1: contr, site normal subjects (n=4)
e - - e
o--o
4-
a_ 2 0 1
0 I
2 Time (min)
5 / , ~ j NE
E3 c o ra n~ O > LaJ I
..3
5 4
4
• --" •
case 1: patch site • case 1: adjacent site
3 2
r-
2 1
•
I-----
I
era
0 -10
-5
6 T NE 5 /a,g
~
10
15
20
25
Time ( m i n )
Fig. 3. Pain and hyperalgesia evoked by intrade[mal injection of norepinephrine (5 /zg in 10 p,l saline). A: the time course for evoked pain following the norepinephrine injection. Pain was rated verbally on a scale from 0 (no pain) to 10 (most intense pain), injections w e r e made into a clonidine patch site (filled squares) and a contralateral site (filled circles) for patient I, and into the lateral leg of four normal subjects (open circles, mean _+ S.E.M. are indicated). B: the norepinephrine injection rekindled hyperalgesia to mechanical stimuli at the patch site in patient 1. The pain evoked by light brushing at the patch site (filled squares) or adjacent to the patch site was rated verbally on a 0 - I 0 scale. Saline injections into the patch site did not evoke pain or hyperalgesia in the patient or subjects (data not shown).
314
A
10 tn
8
c
:-7,
A----A e--o o--o
~&--&--&
c a s e 4: p a t c h site c a s e 4: contr, site normal subjects (n=4)
6
E~
o-o-o~m
n
•
2
ts I
2
[3
3
Time ( r a i n )
10 /zg phenyl
10 A--A
E
8.
n
,,/
6.
&
t
j
case
4
&
o4 .&-
20 -10
-'5
0 T
10 ug phenyk
1 &
1'o (5
2'o 2'5 3'0 3'5 ;o %
Time (min)
Fig. 4. Pain evoked by intradermal injection of phenylephrine (10/zg in 10/zl saline) A: as in Fig. 3A except phenylephrine injections were made into patient 4 at a patch site (filled triangles) and a control site at the volar forearm (filled circles), and into the volar forearm of four normal subjects (open circles, mean _+ S.E.M. are indicated). B: phenylephrine rekindled the patient's ongoing, stimulus-independent pain (filled triangles) and her pain to mechanical and cold stimuli (not shown). Pain ratings at time of injection are not shown. Ongoing pain was rated on a scale from 0-10.
into the normal limb of this patient and the normal limb of another SMP patient (case 4) also evoked intense pain but did not result in hyperalgesia (see Fig. 3). In contrast, intradermal injection of norepinephrine into the leg of four control subjects evoked a mild, short-lasting pain and no hyperalgesia (see Fig. 3A). Since topical application of an alpha2-adrenergic agonist relieved hyperalgesia and a non-specific agonist rekindled hyperalgesia, we hypothesize that alpha F adrenergic receptors play an important role in SMP. To test this hypothesis directly, the specific alpha F adrenergic agonist phenylephrine was injected intradermally (10 /xg in 10 /zl saline) at a clonidine patch site and a normal site (volar forearm) in a patient with SMP (case 4) following 3 days of clonidine treatment. This patient's pain and hyperalgesia, which were localized to a small region of the face, were relieved by clonidine. Injection of phenylephrine at the clonidine patch site evoked intense stinging pain that subsided approximately 3 min after injection (Fig. 4A). Approximately 25 min after injection, the patient's hyperalgesia to mechanical and cooling stimuli returned. At this time, her ongoing level of pain also increased substantially (Fig. 4B), Injection of phenylephrine at the control site in this patient evoked moderate pain of shorter duration (Fig. 4A). Only brief, mild pain sensations
and no hyperalgesia were evoked by the phenylephrine injections into the control subjects (see Fig. 4A).
Discussion
In this study topical application of clonidine significantly reduced mechanical and cold hyperalgesia at the site of drug administration in patients with SMP. The decrease in hyperalgesia was not due to a local anesthetic effect since touch detection thresholds were not altered by clonidine treatment in the patients and in the control subject. There was no change in hyperalgesia outside the area of drug application in patients with a large zone of hyperalgesia. In addition, the ongoing, stimulus-independent pain was not affected by the clonidine patches in these patients. Clonidine had little effect on pain or hyperalgesia in the patients whose pain was independent of the sympathetics (cases 5 and 6). Local administration of adrenergic agonists into the affected skin in patients with SMP evoked unusually sustained, intense pain compared to the pain evoked in normal subjects. These observations indicate an important role of peripheral alpha l-adrenergic receptors in SMP.
315 Clonidine is an alpha2-adrenergic agonist that may affect both central and peripheral adrenergic receptors. A central antinociceptive action of clonidine is thought to be mediated by alpha 2 (possibly post-synaptic) adrenergic receptors in the dorsal horn of the spinal cord [10,20]. Clinical studies have shown that intrathecal clonidine is effective in reducing pain associated with intractable cancer [6,9]. Max et al. [18] reported that the pain associated with postherpetic neuralgia could be significantly reduced with oral clonidine. Animal studies provide evidence for a descending noradrenergic pathway from the brain stem to the dorsal horn that can modulate nociceptive transmission [2,15,19,20,33,34]. In our study, the effects of transdermal clonidine appeared to be confined in three patients to the vicinity of the patch and, therefore, the site of action was not likely in the central nervous system. In the periphery, clonidine acts at presynaptic alpha2-adrenergic receptors. Activation of these autoreceptors inhibits the release of norepinephrine, thereby reducing noradrenergic neurotransmission [27,32]. The observed localized effects of clonidine make it likely that the antinociceptive action of the clonidine is via a reduction in peripheral release of norepinephrine. We also showed in two SMP patients that intraderreal injection of alpha-adrenergic agonists at a clonidine patch site evoked an intense burning pain and rekindled hyperalgesia. In contrast, similar injections in normal subjects produced brief, mild pain. A similar rekindling of hyperalgesia to mechanical stimuli was observed following injection of norepinephrine into the previously hyperalgesic skin of four SMP patients following sympathectomy or sympathetic block [31]. Although only one patient in our study was injected with the specific alpha 1 agonist phenylephrine, the results corroborate our interpretation that hyperalgesia relief following clonidine treatment is due to reduced norepinephrine release. Furthermore, these results are consistent with our hypothesis that, in SMP, the nociceptors develop an alphax-adrenergic sensitivity (see below). Use of specific adrenergic agonists applied intradermally may provide a valuable means to corroborate the diagnosis of SMP. Further study of the utility of this test will be worthwhile. The norepinephrine and phenylephrine injections into affected and unaffected skin in the patients evoked an unusually intense, prolonged pain response compared to the control subjects. Therefore, the alpha 1adrenergic receptor system in patients with SMP may be in an altered state. It is not clear, however, why the injections at sites remote to the hyperalgesic zone evoked such intense pain. Patients with SMP may be generally up-regulated with regard to production of alpha~ receptors, the function of which is associated with pain.
Clonidine appeared to have little effect on stimulus-independent (ongoing) pain in the patients with large areas of hyperalgesia. A differential effect of clonidine on hyperalgesia and ongoing pain could be due to several factors. First, ongoing pain and hyperalgesia may have different underlying mechanisms. This factor cannot be easily addressed since the relative contribution of various central versus peripheral mechanisms in the maintenance of hyperalgesia and ongoing pain in SMP is not well understood. Second, spatial summation may play a greater role in the maintenance of ongoing pain than of hyperalgesia. If the zone of hyperalgesia is large, each clonidine patch could affect hyperalgesia arising from the vicinity of the patch but would have little effect on the spatially summated level of ongoing pain. For the three patients with large areas of hyperalgesia, the clonidine patch covered less than 20% of the affected area and no change in the overall level of ongoing pain was observed. In the one patient with a relatively small zone of hyperalgesia, a significant reduction of ongoing pain was obtained. These findings are consistent with the hypothesis that spatial summation is important for the ongoing pain. Since clonidine had little effect on spontaneous pain and often eliminated hyperalgesia only at the relatively small patch sites, the potential therapeutic benefits of this treatment are limited. However, our results may contribute to the understanding of SMP mechanisms and encourage further research to develop topical agents such as creams or aerosols containing alphaadrenergic blockers. We have shown that clonidine may preferentially affect hyperalgesia in patients with SMP versus SIP. However, a possible pitfall of this interpretation is that in one of the patients with SIP the clonidine patches were applied to the palmar aspect (glabrous skin) of the hand, whereas the patches in the other patients were applied to hairy skin. Although quantitative data of the clonidine diffusion rates through glabrous and hairy skin are not available, it is known that the flux of many other compounds (e.g., lidocaine, acids, water) through glabrous skin is actually faster than through hairy skin [26]. Hence, we do not believe that the lack of effect of clonidine in this patient was due to diffusion barriers. Furthermore, in the other SIP patient (case 6), the clonidine was applied to hairy skin and had little effect on pain and hyperalgesia. Several mechanisms have been proposed to account for SMP. Levine et al. [16] suggested that under pathological conditions, norepinephrine released from sympathetic post-ganglionic neurons (SPGN) can activate alpha2-adrenergic receptors on the SPGN terminal. This would, in turn, stimulate release of prostaglandins from the SPGN terminals. It was proposed that the hyperalgesic effect of norepinephrine was mediated by prostaglandins. In contrast, our data indicate that acti-
316
vation of the presynaptic alpha 2 receptors with clonidine does not induce but actually reduces hyperalgesia. Roberts [22] proposed that SMP involves a sensitization of wide dyne.mic range neurons in the spinal cord that is maintained by sympathetic stimulation of lowthreshold mechanoreceptors (LTMs). This model does not account for the relief of pain following sympathetic blockade and the observation that activation of LTMs does not produce pain under conditions of a sympathetic block. Also, recent microneurographic evidence indicates that low-threshold mechanoreceptors are not activated by sympathetic stimulation in SMP [8]. Other peripheral sites such as the primary afferent terminals [14] may be involved in SMP. Primary afferents ending in neuromas develop alpha [3,7,30] but not bcta-adrenergic sensitivity [7,30]. The data presented here provides evidence for the involvement of peripheral alpha]-adrenergic receptors in sympathetically maintained pain. We postulate that, following certain injuries, cutaneous nociceptors develop an enhanced alpha t-adrenergic sensitivity and a response to sympathetic efferent activity. Activation of nociceptors by sympathetic stimulation has been observed following tissue injury [23,24]. This activity in nociceptors maintains the central pain signalling neurons in a dynamic sensitized state such that input from low-threshold mechanoreceptors is enhanced. This enhanced response accounts for the hyperalgesia to mechanical stimuli [5], Topical clonidine inhibits the release of norepinephrine from sympathetic terminals which in turn eliminates the ongoing activity of nociceptors and eliminates the central sensitization, thereby relieving hyperalgesia.
Acknowledgements This study was supported by NIH NS-14447 and NS-26363 grants. K.D. Davis was supported by a Canadian MRC fellowship and R.D. Treede was supported by DFG Tr236/1-1. The authors thank Dr. M. Pappagallo for assistance with some of the intradermal injections.
References 1 Abram, S.E. and Lightfoot, R.W., Treatment of long-standing causalgia with prazosin, Reg. Anesth., 6 (1981) 79-81. 2 Besson, J.-M. and Chaouch, A., Peripheral and spinal mechanisms of nociception, Physiol. Rev., 67 (1987) 67-186. 3 Btumberg, H. and Jiinig, W., Discharge pattern of afferent fibers from a neuroma, Pain, 20 (1984) 335-353. 4 Bonica, J.J., Causalgia and other reflex sympathetic dystrophies. In: J.J. Bonica et al. (Eds.), Advances in Pain Research and Therapy, Vol. 3, Raven Press, New York, 1979, pp. 141-165. 5 Campbell, J.N., Raja, S.N. and Meyer, R.A., Painful sequelae of nerve injury. In: R. Dubner, G.F. Gebhart and M.R. Bond tEals.),
Proc. Vth World Congress o n Pain, Pain Research and Clinical Management, Vol. 3, Elsevier, Amsterdam, 1988, pp. 135 143. (~ Coombs, D.W., Saunders, R.L., Fratkin, J.D., Jensen, IA{. and Murphy, A.. Continuous intrathecal hydromorphine and cloni dine for intractable cancer pain, J. Neurosurg., 64 (1986) 890-894. 7 Dew)r, M. and J~inig, W., Activation of myelinated afferents ending in a neuroma by stimulation of the sympathetic supply in the rat, Neurosci. Left., 24 (19811 43-47, 8 Dotson, R., Ochoa, J.. Cline. M., Roberts, W.. Yarnitsky, D.. Simone, D. and Marchettini. P., Sympathetic effects on human low threshold mechanoreceptors, Soc. Neurosci. Abst., 1(, (199(I) 1280. 9 Eisenach, J.C., Rauck, R.L., Buzzanell, C. and Lysak, S.Z., Epidural clonidine analgesia for intractable cancer pain: phase 1, Anesthesiology, 71 (1989) 647-652. 10 Fleetwood-Walker, S.M., Hope, P.J,, lggo, A., Mitchell, R. and Molony, V., Characterization of receptors mediating the selective effect of noradrenaline on the cutaneous responses of identified dorsal horn neurones in the cat, J. Physiol. (Lond.), 343 (19831 67P. tl Frost, S.A., Raja, S.N., Campbell, J.N., Meyer, R.A. and Khan. A.A., Does hyperalgesia to cooling stimuli characterize patients with sympathetically maintained pain (reflex sympathetic dystrophy)? In: R. Dubner, G.F. Gebhart and M.R. Bond (Eds.), Proc. Vth World Congress on Pain, Pain Research and Clinical Management, Vol. 3, Elsevier, Amsterdam, 1988, pp. 151-156. 12 Ghostine, S.Y., Comair, Y.G., Turner, D.M., Kassell, N. anti Azar, C.G., Phenoxybenzamine in the treatment of causalgia, J. Neumsurg., 60 (1984) 1263-1268. 13 Hannington-Kiff, J.G., Intravenous regional sympathetic blocks with guanethidine, Lancet, i (1974) 1019-1020. 14 Jfinig, W., The sympathetic nervous system in pain: physiology and patbophysiology. In: M. Stanton-Hicks ted.), Pain and The Sympathetic Nervous System, Kluwer, Boston, MA, 1990, pp. 17--89. 15 Jones. S.L. and Gebhart. G.F., Characterization of coeruleospinal inhibition of the nociceptive tail-flick reflex in the rat: mediation by spinal aradrenoceptors, Brain Res., 364 (19861 315-33(I. I6 Levine, J.D., Taiwo, Y.O., Collins, S.D. and Tam, J.K., Noradrenaline hyperalgesia is mediated through interaction with sympathetic postganglionic neurone terminals rather than activation of primary afferent nociceptors, Nature, 323 (19861 158-16(I. 17 Lob, L. and Nathan, P.W., Painful peripheral states and sympathetic blocks, J. Neurol. Neurosurg. Psychiat., 41 (1978) 664-671. 18 Max, M.B., Schafer, S.C., Culnane, M., Dubner, R. and Gracely, R.H., Association of pain relief with drug side effects in postherpetic neuralgia: a single-dose study of clonidine, codeine, ibuprofen, and placebo, Clin. Pharmacol. Ther., 43 (1988) 363-371. 19 Proudfit, H.K. and Hammond, D.L., Alterations in nociceptive threshold and morphine-induced analgesia produced by intrathecally administered amine antagonists, Brain Res., 218 (19811 393-390. 211 Reddy, S.V.R. and Yaksh. T.I,., Spinal noradrenergic terminal system mediates antinociception, Brain Res., 189 (198111 391-401. 21 Raja, S.N., Treede, R.D., Davis, K.D. and Campbell, J.N., Systemic alpha-adrenergic blockade with phentolamine: a diagnostic test for sympathetically maintained pain, Anesthesiology, 74 ( 1991 ) 691-698. 22 Roberts, W., A hypothesis on the physiological basis for causalgia and related pains, Pain, 24 (1986)297-311, 23 Roberts, W. and Elardo, S.M., Sympathetic activation of A-delta nociceptors, Somatosens. Res., 3 (1985) 33-44. 24 Sanjue, H. and Jun, Z., Sympathetic facilitation of sustained discharges of polymodal nociceptors, Pain, 38 (19891 85-90, 25 Scadding, J.W., Wall, P.D., Wynn Parry, C,B. and Brooks, D.M., Clinical trial of propranolol in post-traumatic neuralgia, Pain, 14 (1982) 283-292.
317 26 Scheuplein, R.J. and Blank, I.H., Permeability of the skin, Physiol. Rev., 51 (1971) 702-747. 27 Starke, K., a-Adrenoreceptor subclassification, Rev. Physiol. Biochem. Pharmacol., 88 (1981) 199-236. 28 Treede, R.D., Raja, S.N., Davis, K.D., Meyer, R.A. and Campbell, J.N., Evidence that peripheral alpha-adrenergic receptors mediate sympathetically maintained pain. In: M. Bond et al. (Eds.), Proc. Vllth World Congress on Pain, Pain Research and Clinical Management, Vol. 5, Elsevier, Amsterdam, 1991, pp. 377-382. 29 Walker, A.E. and Nulsen, F., Electrical stimulation of the upper thoracic portion of the sympathetic chain in man, Arch. Neurol. Psychiat. (Chicago), 59 (1948) 559-560. 30 Wall, P.D. and Gutnick, M., Ongoing activity in peripheral nerves: the physiology and pharmacology of impulses originating from a neuroma, Exp. Neurol., 43 (1974) 580-593.
31 Wallin, G., Torebj6rk, E. and Hallin, R., Preliminary observations on the pathophysiology of hyperalgesia in the causalgic pain syndrome. In: Y. Zotterman (Ed.), Sensory Functions of the Skin in Primates, Pergamon, New York, 1976, pp. 489-502. 32 Weiner, N., Drugs that inhibit adrenergic nerves and block adrenergic receptors. In: A.G. Gilman, L.S. Goodman, T.W. Rail and F. Murad (Eds.), Goodman and Gilman's The Pharmacological Basis of Therapeutics, 7th edn., MacMillan, New York, 1985, pp. 181-214. 33 Yaksh, T.L., Pharmacology of spinal adrenergic systems which modulate spinal nociceptive processing, Pharmacol. Biochem. Behav., 22 (1985) 845-858. 34 Yaksh, T.L., Hammond, D.L. and Tyce, G.M., Functional aspects of bulbospinal monoaminergic projections in processing of somatosensory information, Fed. Proc., 40 (1981) 2786-2794.
309
¢) 1991 Elsevier Science Publishers B.V. All rights reserved 0304-3959/91/$03.50
PAIN 01899
Topical application of clonidine relieves hyperalgesia in patients with sympathetically maintained pain K.D. Davis a,**, R.D. T r e e d e ~'*, S.N. Raja a Departments o f Neurosurgery,
b, R.A.
Meyer a,c and J.N. Campbell ~,c
b Anesthesiology and ~ The Applied Physics Laboratory, Johns Hopkins University, Baltimore, MD 21205 (U.S.A.)
(Received 4 December 1990, revision received 12 April 1991, accepted 15 April 1991)
Summary Patients with reflex sympathetic dystrophy or causalgia characteristically have ongoing pain and pain to light touch (hyperalgesia). Some of these patients obtain relief of their pain following interruption of sympathetic function to the affected area and, therefore, have sympathetically maintained pain (SMP). Evidence suggests that the pain and hyperalgesia in SMP are related to activation of peripheral adrenergic receptors. We wished to determine the contribution of alpha 1- and alpha2-adrenergic receptors in SMP and thus examined the effects of local application of adrenergic agents in patients with SMP. The alpha2-adrenergic agonist clonidine, available as a transdermal patch, was delivered topically to the patients' hyperalgesic skin. In four patients with SMP, clonidine eliminated or substantially reduced hyperalgesia to mechanical and cold stimuli. In three of these patients the effects were confined to the skin region beneath the patch, suggesting a peripheral and not central effect. The relief of hyperalgesia was not due to a local anesthetic effect since touch thresholds were unaffected. Topical clonidine did not relieve hyperalgesia of similar severity for two other patients whose hyperalgesia and pain were unaffected by sympathetic ganglion blocks (i.e., diagnosed as having sympathetically independent pain). In two SMP patients, intradermal injection of norepinephrine or phenylephrine (a specific alphat-adrenergic agonist) at a site treated with clonidine evoked intense pain and rekindled the pre-clonidine hyperalgesia at that site. It is likely that clonidine locally blocks the release of norepinephrine via activation of alpha 2 receptors on the sympathetic terminals. This study suggests, therefore, that SMP is mediated via alphal-adrenergic receptors located in the affected tissue. Key words: Reflex sympathetic dystrophy; Causalgia; Allodynia; alpha-Adrenergic receptors; Norepinephrine; Phenylephrine
Introduction Sympathetically maintained pain (SMP) refers to a chronic pain disorder in which the pain is dependent on sympathetic innervation of the painful part [22]. Thus, patients are diagnosed as having SMP if blockade of the sympathetic innervation to the affected area alleviates their pain [4,22]. Patients with SMP typically
* On leave from the Institute of Physiology, University Hospital Eppendorf, D-2000 Hamburg 20, F.R.G. * * Present address: Department of Physiology, University of Toronto, Toronto, Ontario, Canada. Correspondence to: Dr. James N. Campbell, Department of Neurosurgery, Meyer 7-113, Johns Hopkins University, Baltimore, MD 21205, U.S.A.
have both stimulus-independent (ongoing) pain and stimulus-dependent pain (hyperalgesia). Hyperalgesia is defined as a leftward shift of the stimulus-response function, such that a lowering of pain threshold and/or an increase in pain to suprathreshold stimuli are observed [5]. The decrease in pain threshold to mechanical stimuli may be such that lightly stroking the skin evokes pain, a phenomenon sometimes referred to as allodynia (here considered to be part of hyperalgesia). Several lines of evidence suggest that peripheral adrenergic receptors are involved in SMP. (1) Stimulation of the peripheral but not central cut end of the sympathetic chain reproduces pain in causalgia patients after sympathectomy [29]. (2) Local anesthetic blockade of the appropriate sympathetic ganglion or adrenergic blockade via intravenous administration of phentolamine rapidly abolishes sympathetically main-
310 tained pain and hyperalgesia [21,28]. (3) Depletion of peripheral catecholamines by regional intravenous guanethidine also relieves pain and hyperalgesia [13,17]. (4) Intradermal injection of norepinephrine rekindles the pain and hyperalgesia that had been relieved in patients by sympathectomy or sympathetic block [31] but does not cause pain or hyperalgesia in normal subjects. (5) The non-specific alpha-adrenergic antagonist phenoxybenzamine [12] and the specific alpha r adrenergic antagonist prazosin [1] can be effective in relieving pain in patients with SMP. (6) The betaadrenergic antagonist propranolol has little effect on SMP [21,25]. These data suggest that SMP may be a peripheral alpha cadrenergic receptor disease. Conventional treatments for SMP include repeated local anesthetic sympathetic blocks, intravenous regional guanethidine/reserpine blocks, surgical sympathectomy or oral sympatholytic therapy. However, each of these treatments carries with it a degree of risk, side effects and discomfort. Therefore, the topical application of an alpha-adrenergic blocking agent could be an alternate approach in the treatment of SMP. No alpha-adrenergic antagonist is commercially available for topical use at present. However, clonidine, an alpha2-adrenergic agonist commonly used in the treatment of hypertension, is available as a patch for transdermal administration. Clonidine activates pre-synaptic adrenergic autoreceptors resulting in a reduction of norepinephrine release [27]. This effectively decreases activation of the post-synaptic alpha~ receptors. In this study, the therapeutic benefit of peripheral alpha-adrenergic blockade with clonidine was examined in patients suspected to have SMP. To explore the involvement of peripheral alpha j- and alpha2-adrenergic receptors in SMP, these patients were treated with topical clonidine and, in some cases, were subsequently given intradermal injections of alpha-adrenergic agonists.
Methods
Patient selection and control subjects Six normotensive patients with chronic ongoing pain and cutaneous hyperalgesia to mechanical and cooling stimuli following soft tissue or nerve trauma entered the study. All patients had previously undergone sympathetic blocks (i.e., local anesthetic blockade of the appropriate sympathetic ganglia) to assess the involvement of the sympathetic nerves in their pain state. Additionally, in all but one patient (case 2), a systemic alpha-adrenergic block was performed via intravenous administration of phentolamine [see ref. 21 for details]. Four of the six patients (cases 1-4) experienced 70100% pain relief following the block and were considered to have SMP. The remaining two patients (cases 5
and 6) were considered to have sympathetically independent pain (SIP) since their pain was rcduccd by < 25% following local anesthetic, sympathetic ganglion block or the phentolamine block. Five normal subjects entered the study as controls. One subject was used to investigate the effect of clonidine on normal skin (i.e.. the top of the foot). The four remaining subjects were used to assess the effect of intradermal norepinephrine and phenylephrine in norreal skin. Topical clonidine Clonidine was administered to the hyperalgesic skin via a 7.0 or 10.5 cm 2 patch (Catapres-TTSR-2 and TTSR-3, Boehringer, Inge[heim). These patches deliver a systemic dose of 0.2 mg and (1.3 rag, respectively, of clonidine/day (i.e., 30 p,g/cm2/day) for a maximum of 7 days. A series of 2-7 patches were applied consecutively to different sites within the hyperalgesic zone. Each patch was left in place for 2-10 days within this affected zone. Prior to, during and immediately after drug application, the following parameters were monitored: heart rate, blood pressure, ongoing pain and pain to mechanical and cold stimuli (see below). Sensor)' testing To assess the local and systemic effects of clonidine, the following measures were used before application of the patch and immediately after removal of the patch. (1) Ongoing pain: stimulus-independent pain was assessed on a visual analog scale (VAS). The scale consisted of a 100 mm line where 'no pain' and 'most intense pain' were indicated on the left and right ends of the scale, respectively. (2) Hyperalgesia to mechanical and cooling stimuli: pain evoked by mechanical and cold stimuli was rated verbally on a scale from 0 (no pain) to 10 (most intense pain). The mechanical stimuli included innocuous brushing with a camel's hair brush and innocuous pressure from the 206 g weight of a 13 mm diameter thermally neutral probe. These mechanical stimuli are innocuous when applied to normal skin. Pain thresholds to mechanical stimuli were determined in the area within and surrounding the patch sites using calibrated von Frey filaments. The cooling stimulus consisted of a small drop of acetone lightly placed onto the patient's skin. Patients with SMP typically find that the 1-2 °C decrease in temperature induced by such a stimulus is painful [11]. (3) Tactile sensibility: to rule out local anesthetic effects, the detection (touch) thresholds were determined in the area within and surrounding selected patch sites using calibrated von Frey filaments. In addition, the patient's ability to distinguish the blunt from the sharp end of a pin was also assessed. The patients were instructed to close their eyes during the hyperalgesia and yon Frey threshold testing procedures.
311
Effects of intradermal norepinephrine and phenylephrine In two patients, intradermal injections of the nonspecific alpha-adrenergic agonist norepinephrine (5/xg in 10 ~1 normal saline) or the alphal-adrenergic agonist phenylephrine (10/zg in 10/xl normal saline) were made at a patch site immediately after removing the patch. Norepinephrine and phenylephrine injections were also made at control sites on the patients' normal limbs and into the normal skin of four healthy control subjects. Similar injection sites were chosen for the control subjects and patients. All injections were made in a single-blinded fashion. The patients and control subjects were instructed to rate the magnitude of any evoked pain and resultant hyperalgesia on a verbal scale from 0 to 10.
Results The results of the clonidine applications in the control subject and six patients are presented below. The first patient was studied extensively and is thus presented in detail as an illustrative example.
•
zone
of hyperalgesia /
//
/
/
[[ %'
1
%ITF
--
I at patch 2
outside
5 ....
*verbal
ratings
patch
trolateral
/
/
~/
/ /
/
/
// /
/ /
/
//
i /
// // / / / / / / / /-//
/
/
// /
// //
~ / / /
/
~
//,/ //~
/j /1 d
"
I[ I-HRESHOLD (bars)l I PAIN RATING* lldetection pain llbrush pressure II II l[ -~ II 4 12 11 0 O il II II II
4
b
!! ii
2
10
on a O ( " n o
pain")
t o 10 ( " m o s t
II 6 . 5 II !! ii
10
0
intense
0
pain")
scale
Fig. 1. Example of the effect of a clonidine patch applied to the hyperalgesic skin of a patient with sympathetically maintained pain (case 1). Sensory testing was performed immediately following removal of the patch that had been in contact with a previously hyperalgesic site for 36 h. Site 1: at the patch site, pain to stroking the skin with a soft brush and to pressure (2N) from a thermally neutral probe was eliminated (right hand column of table). The touch detection and pain thresholds determined with van Frey probes (left hand column of table) were similar to those at a contralateral site (site 3) on normal skin. Site 2: immediately adjacent to the patch, the brush and pressure stimuli evoked pain and the pain threshold was lower than the contralateral normal skin site.
Case 1 The patient was a 30-year-old, normotensive, Caucasian female who sustained an injury to the sciatic nerve from hip replacement surgery 3 years prior to entering our study. Subsequently, the patient developed continuous pain and hyperalgesia in the anterolateral aspect of her left leg from the knee to the foot (Fig. 1). T h e r m o g r a p h y revealed that most of this area was 1 - 2 ° C colder than corresponding parts of the contralateral leg. Within the affected region, the patient experienced intense ongoing pain and was exquisitely sensitive to mechanical stimuli. Mild mechanical stimuli (e.g., light brushing or pressure) evoked severe pain. Cooling stimuli (e.g., acetone) were not detected when applied to some parts of the affected area and were painful when applied to other parts. The results of local anesthetic sympathetic ganglion blocks and i.v. phentolamine blocks confirmed that her ongoing pain and hyperalgesia were sympathetically maintained. This patient received a series of seven elonidine patches. A typical example of the effect of a clonidine patch is illustrated in Fig. 1. This patch was applied for 36 h to an area of skin which had previously been hyperalgesic to brushing, pressure and cooling stimuli. After removal of the patch, mechanical (brush, pressure) and cooling (a drop of acetone) stimuli applied to this region were detected but were not painful, although these stimuli still elicited pain when applied to the adjacent skin. The mechanical detection thresholds inside and outside the patch site were identical and were similar to the contralateral leg. Within the patch site, the pain threshold to van Frey filaments approxi-
mated the patient's ' n o r m a l ' pain threshold at a corresponding contralateral site and far exceeded the mechanical detection threshold. The ability to distinguish sharp from blunt stimuli was not affected at this site after the patch was removed. Outside the patch site, the pain threshold was low and approached the detection threshold. The patient's ratings of her stimulus-independent pain were not affected by the patch. Although the patient was pleased with the relief achieved from the patches, treatment was discontinued when the patches started to produce skin irritations. Six months following termination of the clonidine trial, she underwent a lumbar sympathectomy that completely eliminated her pain and hyperalgesia.
Extent of pain and hyperalgesia relief by topical clonidine A summary of the effect of four of the clonidine patches in which quantitative sensory testing was performed for case 1 is shown in Fig. 2. Each clonidine patch greatly reduced and, in many cases, completely eliminated the patient's hyperalgesia to mechanical stimuli. A confounding local anesthetic effect did not occur since detection thresholds were unchanged in the three patch sites tested (see Fig. 2A). The time course of the effect of each patch was not evaluated quantitatively. However, the patient did note a reduction of her mechanical hyperalgesia within 36-48 h after the patch was applied to the skin. After removing a patch, hyperalgesia was still absent for at least 12 h. Typically, the
312 ~20
A
O--•case
1
r - l - - ~ normot subject -o15 o
~I0
J~
g 5
i
ID C3
• []
0
BEFORE
2O
AFTER
B
?
/
•
O
~15 "O
[]
-6
•
P .12 F--
c: 5' O0¸
10 "6
BEFORE
AFTER
BEFORE
AFTER
C
8
EL
60
I
4.
03 2
10. .c o
n
D
8
13
o 6
> LIJ
I
~4
patient's hyperalgesia returned within less than a week of patch removal. Thc three remaining patients diagnosed as having SMP achieved relief of their hyperalgesia l'ollowing topical application of clonidine without any accompanying change in touch detection threshold (see Table 1: cases 2-4). The original zone of hypcralgcsia in two of these patients (cases 2 and 3) encompasscd more than half of the lower extremity. For these patients, topical clonidine only eliminated hyperalgesia at or near the site of drug application and did not alter stimulusevoked pain outside these sites. Furthermore, these patients did not report any reduction in their overall level of stimulus-independent pain. In case 4, the zone of pain and hyperalgesia radiated laterally from a small region above thc vermillion border in the maxillary nerve distribution. Following clonidinc application to the upper lip, complete relief of hypcralgcsia was obtained throughout the entirc affected zone. The patient's ongoing pain was reduced by 5()-75% and was confined to a very small region above the lip. One of the patients diagnosed as having SIP (see Table 1: case 5) had pain and hyperalgesia in the median nerve distribution of the left hand. Neither the patient's hyperalgesia nor ongoing pain werc improved by application of clonidine to the palmar aspect of the hand. Thc second patient diagnosed as having SIP (case 6) had pain and hyperalgesia in the foot. The clonidine patches were applied to the dorsum of the foot and resulted in only a slight reduction (by approximately 35%) in both ongoing pain and in mechanical hypcralgesia to brush stimuli. The hyperatgesia to cold stimuli was not affected by the clonidinc patches.
Side effects The side effects from these clonidine patches were minor. The patients occasionally complained of feeling sleepy, thirsty a n d / o r of having dry eyes. Following some of the patch applications, an erythematous rash was observed surrounding the patch site. This became a problem in two patients (cases 1 and 4), and further treatment had to be discontinued despite good hyperalgesia relief. No appreciable alteration in the blood pressure or heart rate was apparent in any of the patients.
~213_
Sensory effects of topical clonidine in normal skin
OBEFORE
AFTER
Fig. 2. Sensory testing before and after four different clonidine patches applied to different areas of hyperalgesic skin of a patient (case 1) with sympathetically maintained pain (solid circles) and after a single patch applied to a normal subject (open squares). Detection (A) and pain (B) thresholds at the patch site were determined with von Frey filaments. Pain evoked by light brushing (C) and firm pressure (D) was rated verbally on a 0 - 1 0 scale. Each line corresponds to results from a single patch.
To exclude further local anesthetic effects of clonidine, a clonidine patch was applied to the top of a normal subject's foot and sensory testing was performed prior to and 40 h after patch application. Neither the pain nor detection threshold to von Frey filaments were affected in this subject. Mechanical and cooling stimuli were detected but were not painful before or after clonidine application.
313
TABLE I S U M M A R Y OF I N D I V I D U A L CASE RESULTS SMP = sympathetically maintained pain; SIP = sympathetically independent pain; + = present; - = absent; nt = not tested. Case number
Age/sex Duration of symptoms (years) Extremity affected Mechanical hyperalgesia Pre-clonidine Post-clonidine Cold hyperalgesia Pre-clonidine Post-clonidine Touch detection threshold Pre-clonidine (bars) Post-clonidine (bars)
1
2
3
4
5
6
(SMP)
(SMP)
(SMP)
(SMP)
(SIP)
(SIP)
30/F 3 lower
45/F 1.5 lower
35/M 18 lower
52/F 2 face
52/M 7 upper
42/F 0.5 lower
+
+
+
+
+
_
_
--
+
+
+
+
+
+
+
-
nt
-
+
+
1
1.5
0.5
1
2
l
1.5
0.5
1
2
+
+
4 4
only mildly hyperalgesic (rated 1 out of 10) whereas brushing the adjacent untreated skin evoked a pain sensation rated as 4 out of 10. The norepinephrine injection evoked an intense, burning pain sensation that subsided after 4 min (see Fig. 3A). Twenty:five minutes after injection, the patient's mechanical hyperalgesia at the patch site was rekindled and was rated as 3.5 out of 10 (see Fig. 3B). Injection of norepinephrine
Effect of norepinephrine and phenylephrine To test the hypothesis that clonidine acts peripherally via reduction of norepinephrine release, an intradermal injection of norepinephrine (5 /zg in 10 /zl saline) was made at a previously hyperalgesic site that had been treated effectively with clonidine (case 1). Following removal of the clonidine patch and just prior to injection, lightly brushing the skin at this site was A
10. O~
8¸
c
6 I
N
-
C 0
=,--n
mm~_~
case 1: patch site case 1: contr, site normal subjects (n=4)
e - - e
o--o
4-
a_ 2 0 1
0 I
2 Time (min)
5 / , ~ j NE
E3 c o ra n~ O > LaJ I
..3
5 4
4
• --" •
case 1: patch site • case 1: adjacent site
3 2
r-
2 1
•
I-----
I
era
0 -10
-5
6 T NE 5 /a,g
~
10
15
20
25
Time ( m i n )
Fig. 3. Pain and hyperalgesia evoked by intrade[mal injection of norepinephrine (5 /zg in 10 p,l saline). A: the time course for evoked pain following the norepinephrine injection. Pain was rated verbally on a scale from 0 (no pain) to 10 (most intense pain), injections w e r e made into a clonidine patch site (filled squares) and a contralateral site (filled circles) for patient I, and into the lateral leg of four normal subjects (open circles, mean _+ S.E.M. are indicated). B: the norepinephrine injection rekindled hyperalgesia to mechanical stimuli at the patch site in patient 1. The pain evoked by light brushing at the patch site (filled squares) or adjacent to the patch site was rated verbally on a 0 - I 0 scale. Saline injections into the patch site did not evoke pain or hyperalgesia in the patient or subjects (data not shown).
314
A
10 tn
8
c
:-7,
A----A e--o o--o
~&--&--&
c a s e 4: p a t c h site c a s e 4: contr, site normal subjects (n=4)
6
E~
o-o-o~m
n
•
2
ts I
2
[3
3
Time ( r a i n )
10 /zg phenyl
10 A--A
E
8.
n
,,/
6.
&
t
j
case
4
&
o4 .&-
20 -10
-'5
0 T
10 ug phenyk
1 &
1'o (5
2'o 2'5 3'0 3'5 ;o %
Time (min)
Fig. 4. Pain evoked by intradermal injection of phenylephrine (10/zg in 10/zl saline) A: as in Fig. 3A except phenylephrine injections were made into patient 4 at a patch site (filled triangles) and a control site at the volar forearm (filled circles), and into the volar forearm of four normal subjects (open circles, mean _+ S.E.M. are indicated). B: phenylephrine rekindled the patient's ongoing, stimulus-independent pain (filled triangles) and her pain to mechanical and cold stimuli (not shown). Pain ratings at time of injection are not shown. Ongoing pain was rated on a scale from 0-10.
into the normal limb of this patient and the normal limb of another SMP patient (case 4) also evoked intense pain but did not result in hyperalgesia (see Fig. 3). In contrast, intradermal injection of norepinephrine into the leg of four control subjects evoked a mild, short-lasting pain and no hyperalgesia (see Fig. 3A). Since topical application of an alpha2-adrenergic agonist relieved hyperalgesia and a non-specific agonist rekindled hyperalgesia, we hypothesize that alpha F adrenergic receptors play an important role in SMP. To test this hypothesis directly, the specific alpha F adrenergic agonist phenylephrine was injected intradermally (10 /xg in 10 /zl saline) at a clonidine patch site and a normal site (volar forearm) in a patient with SMP (case 4) following 3 days of clonidine treatment. This patient's pain and hyperalgesia, which were localized to a small region of the face, were relieved by clonidine. Injection of phenylephrine at the clonidine patch site evoked intense stinging pain that subsided approximately 3 min after injection (Fig. 4A). Approximately 25 min after injection, the patient's hyperalgesia to mechanical and cooling stimuli returned. At this time, her ongoing level of pain also increased substantially (Fig. 4B), Injection of phenylephrine at the control site in this patient evoked moderate pain of shorter duration (Fig. 4A). Only brief, mild pain sensations
and no hyperalgesia were evoked by the phenylephrine injections into the control subjects (see Fig. 4A).
Discussion
In this study topical application of clonidine significantly reduced mechanical and cold hyperalgesia at the site of drug administration in patients with SMP. The decrease in hyperalgesia was not due to a local anesthetic effect since touch detection thresholds were not altered by clonidine treatment in the patients and in the control subject. There was no change in hyperalgesia outside the area of drug application in patients with a large zone of hyperalgesia. In addition, the ongoing, stimulus-independent pain was not affected by the clonidine patches in these patients. Clonidine had little effect on pain or hyperalgesia in the patients whose pain was independent of the sympathetics (cases 5 and 6). Local administration of adrenergic agonists into the affected skin in patients with SMP evoked unusually sustained, intense pain compared to the pain evoked in normal subjects. These observations indicate an important role of peripheral alpha l-adrenergic receptors in SMP.
315 Clonidine is an alpha2-adrenergic agonist that may affect both central and peripheral adrenergic receptors. A central antinociceptive action of clonidine is thought to be mediated by alpha 2 (possibly post-synaptic) adrenergic receptors in the dorsal horn of the spinal cord [10,20]. Clinical studies have shown that intrathecal clonidine is effective in reducing pain associated with intractable cancer [6,9]. Max et al. [18] reported that the pain associated with postherpetic neuralgia could be significantly reduced with oral clonidine. Animal studies provide evidence for a descending noradrenergic pathway from the brain stem to the dorsal horn that can modulate nociceptive transmission [2,15,19,20,33,34]. In our study, the effects of transdermal clonidine appeared to be confined in three patients to the vicinity of the patch and, therefore, the site of action was not likely in the central nervous system. In the periphery, clonidine acts at presynaptic alpha2-adrenergic receptors. Activation of these autoreceptors inhibits the release of norepinephrine, thereby reducing noradrenergic neurotransmission [27,32]. The observed localized effects of clonidine make it likely that the antinociceptive action of the clonidine is via a reduction in peripheral release of norepinephrine. We also showed in two SMP patients that intraderreal injection of alpha-adrenergic agonists at a clonidine patch site evoked an intense burning pain and rekindled hyperalgesia. In contrast, similar injections in normal subjects produced brief, mild pain. A similar rekindling of hyperalgesia to mechanical stimuli was observed following injection of norepinephrine into the previously hyperalgesic skin of four SMP patients following sympathectomy or sympathetic block [31]. Although only one patient in our study was injected with the specific alpha 1 agonist phenylephrine, the results corroborate our interpretation that hyperalgesia relief following clonidine treatment is due to reduced norepinephrine release. Furthermore, these results are consistent with our hypothesis that, in SMP, the nociceptors develop an alphax-adrenergic sensitivity (see below). Use of specific adrenergic agonists applied intradermally may provide a valuable means to corroborate the diagnosis of SMP. Further study of the utility of this test will be worthwhile. The norepinephrine and phenylephrine injections into affected and unaffected skin in the patients evoked an unusually intense, prolonged pain response compared to the control subjects. Therefore, the alpha 1adrenergic receptor system in patients with SMP may be in an altered state. It is not clear, however, why the injections at sites remote to the hyperalgesic zone evoked such intense pain. Patients with SMP may be generally up-regulated with regard to production of alpha~ receptors, the function of which is associated with pain.
Clonidine appeared to have little effect on stimulus-independent (ongoing) pain in the patients with large areas of hyperalgesia. A differential effect of clonidine on hyperalgesia and ongoing pain could be due to several factors. First, ongoing pain and hyperalgesia may have different underlying mechanisms. This factor cannot be easily addressed since the relative contribution of various central versus peripheral mechanisms in the maintenance of hyperalgesia and ongoing pain in SMP is not well understood. Second, spatial summation may play a greater role in the maintenance of ongoing pain than of hyperalgesia. If the zone of hyperalgesia is large, each clonidine patch could affect hyperalgesia arising from the vicinity of the patch but would have little effect on the spatially summated level of ongoing pain. For the three patients with large areas of hyperalgesia, the clonidine patch covered less than 20% of the affected area and no change in the overall level of ongoing pain was observed. In the one patient with a relatively small zone of hyperalgesia, a significant reduction of ongoing pain was obtained. These findings are consistent with the hypothesis that spatial summation is important for the ongoing pain. Since clonidine had little effect on spontaneous pain and often eliminated hyperalgesia only at the relatively small patch sites, the potential therapeutic benefits of this treatment are limited. However, our results may contribute to the understanding of SMP mechanisms and encourage further research to develop topical agents such as creams or aerosols containing alphaadrenergic blockers. We have shown that clonidine may preferentially affect hyperalgesia in patients with SMP versus SIP. However, a possible pitfall of this interpretation is that in one of the patients with SIP the clonidine patches were applied to the palmar aspect (glabrous skin) of the hand, whereas the patches in the other patients were applied to hairy skin. Although quantitative data of the clonidine diffusion rates through glabrous and hairy skin are not available, it is known that the flux of many other compounds (e.g., lidocaine, acids, water) through glabrous skin is actually faster than through hairy skin [26]. Hence, we do not believe that the lack of effect of clonidine in this patient was due to diffusion barriers. Furthermore, in the other SIP patient (case 6), the clonidine was applied to hairy skin and had little effect on pain and hyperalgesia. Several mechanisms have been proposed to account for SMP. Levine et al. [16] suggested that under pathological conditions, norepinephrine released from sympathetic post-ganglionic neurons (SPGN) can activate alpha2-adrenergic receptors on the SPGN terminal. This would, in turn, stimulate release of prostaglandins from the SPGN terminals. It was proposed that the hyperalgesic effect of norepinephrine was mediated by prostaglandins. In contrast, our data indicate that acti-
316
vation of the presynaptic alpha 2 receptors with clonidine does not induce but actually reduces hyperalgesia. Roberts [22] proposed that SMP involves a sensitization of wide dyne.mic range neurons in the spinal cord that is maintained by sympathetic stimulation of lowthreshold mechanoreceptors (LTMs). This model does not account for the relief of pain following sympathetic blockade and the observation that activation of LTMs does not produce pain under conditions of a sympathetic block. Also, recent microneurographic evidence indicates that low-threshold mechanoreceptors are not activated by sympathetic stimulation in SMP [8]. Other peripheral sites such as the primary afferent terminals [14] may be involved in SMP. Primary afferents ending in neuromas develop alpha [3,7,30] but not bcta-adrenergic sensitivity [7,30]. The data presented here provides evidence for the involvement of peripheral alpha]-adrenergic receptors in sympathetically maintained pain. We postulate that, following certain injuries, cutaneous nociceptors develop an enhanced alpha t-adrenergic sensitivity and a response to sympathetic efferent activity. Activation of nociceptors by sympathetic stimulation has been observed following tissue injury [23,24]. This activity in nociceptors maintains the central pain signalling neurons in a dynamic sensitized state such that input from low-threshold mechanoreceptors is enhanced. This enhanced response accounts for the hyperalgesia to mechanical stimuli [5], Topical clonidine inhibits the release of norepinephrine from sympathetic terminals which in turn eliminates the ongoing activity of nociceptors and eliminates the central sensitization, thereby relieving hyperalgesia.
Acknowledgements This study was supported by NIH NS-14447 and NS-26363 grants. K.D. Davis was supported by a Canadian MRC fellowship and R.D. Treede was supported by DFG Tr236/1-1. The authors thank Dr. M. Pappagallo for assistance with some of the intradermal injections.
References 1 Abram, S.E. and Lightfoot, R.W., Treatment of long-standing causalgia with prazosin, Reg. Anesth., 6 (1981) 79-81. 2 Besson, J.-M. and Chaouch, A., Peripheral and spinal mechanisms of nociception, Physiol. Rev., 67 (1987) 67-186. 3 Btumberg, H. and Jiinig, W., Discharge pattern of afferent fibers from a neuroma, Pain, 20 (1984) 335-353. 4 Bonica, J.J., Causalgia and other reflex sympathetic dystrophies. In: J.J. Bonica et al. (Eds.), Advances in Pain Research and Therapy, Vol. 3, Raven Press, New York, 1979, pp. 141-165. 5 Campbell, J.N., Raja, S.N. and Meyer, R.A., Painful sequelae of nerve injury. In: R. Dubner, G.F. Gebhart and M.R. Bond tEals.),
Proc. Vth World Congress o n Pain, Pain Research and Clinical Management, Vol. 3, Elsevier, Amsterdam, 1988, pp. 135 143. (~ Coombs, D.W., Saunders, R.L., Fratkin, J.D., Jensen, IA{. and Murphy, A.. Continuous intrathecal hydromorphine and cloni dine for intractable cancer pain, J. Neurosurg., 64 (1986) 890-894. 7 Dew)r, M. and J~inig, W., Activation of myelinated afferents ending in a neuroma by stimulation of the sympathetic supply in the rat, Neurosci. Left., 24 (19811 43-47, 8 Dotson, R., Ochoa, J.. Cline. M., Roberts, W.. Yarnitsky, D.. Simone, D. and Marchettini. P., Sympathetic effects on human low threshold mechanoreceptors, Soc. Neurosci. Abst., 1(, (199(I) 1280. 9 Eisenach, J.C., Rauck, R.L., Buzzanell, C. and Lysak, S.Z., Epidural clonidine analgesia for intractable cancer pain: phase 1, Anesthesiology, 71 (1989) 647-652. 10 Fleetwood-Walker, S.M., Hope, P.J,, lggo, A., Mitchell, R. and Molony, V., Characterization of receptors mediating the selective effect of noradrenaline on the cutaneous responses of identified dorsal horn neurones in the cat, J. Physiol. (Lond.), 343 (19831 67P. tl Frost, S.A., Raja, S.N., Campbell, J.N., Meyer, R.A. and Khan. A.A., Does hyperalgesia to cooling stimuli characterize patients with sympathetically maintained pain (reflex sympathetic dystrophy)? In: R. Dubner, G.F. Gebhart and M.R. Bond (Eds.), Proc. Vth World Congress on Pain, Pain Research and Clinical Management, Vol. 3, Elsevier, Amsterdam, 1988, pp. 151-156. 12 Ghostine, S.Y., Comair, Y.G., Turner, D.M., Kassell, N. anti Azar, C.G., Phenoxybenzamine in the treatment of causalgia, J. Neumsurg., 60 (1984) 1263-1268. 13 Hannington-Kiff, J.G., Intravenous regional sympathetic blocks with guanethidine, Lancet, i (1974) 1019-1020. 14 Jfinig, W., The sympathetic nervous system in pain: physiology and patbophysiology. In: M. Stanton-Hicks ted.), Pain and The Sympathetic Nervous System, Kluwer, Boston, MA, 1990, pp. 17--89. 15 Jones. S.L. and Gebhart. G.F., Characterization of coeruleospinal inhibition of the nociceptive tail-flick reflex in the rat: mediation by spinal aradrenoceptors, Brain Res., 364 (19861 315-33(I. I6 Levine, J.D., Taiwo, Y.O., Collins, S.D. and Tam, J.K., Noradrenaline hyperalgesia is mediated through interaction with sympathetic postganglionic neurone terminals rather than activation of primary afferent nociceptors, Nature, 323 (19861 158-16(I. 17 Lob, L. and Nathan, P.W., Painful peripheral states and sympathetic blocks, J. Neurol. Neurosurg. Psychiat., 41 (1978) 664-671. 18 Max, M.B., Schafer, S.C., Culnane, M., Dubner, R. and Gracely, R.H., Association of pain relief with drug side effects in postherpetic neuralgia: a single-dose study of clonidine, codeine, ibuprofen, and placebo, Clin. Pharmacol. Ther., 43 (1988) 363-371. 19 Proudfit, H.K. and Hammond, D.L., Alterations in nociceptive threshold and morphine-induced analgesia produced by intrathecally administered amine antagonists, Brain Res., 218 (19811 393-390. 211 Reddy, S.V.R. and Yaksh. T.I,., Spinal noradrenergic terminal system mediates antinociception, Brain Res., 189 (198111 391-401. 21 Raja, S.N., Treede, R.D., Davis, K.D. and Campbell, J.N., Systemic alpha-adrenergic blockade with phentolamine: a diagnostic test for sympathetically maintained pain, Anesthesiology, 74 ( 1991 ) 691-698. 22 Roberts, W., A hypothesis on the physiological basis for causalgia and related pains, Pain, 24 (1986)297-311, 23 Roberts, W. and Elardo, S.M., Sympathetic activation of A-delta nociceptors, Somatosens. Res., 3 (1985) 33-44. 24 Sanjue, H. and Jun, Z., Sympathetic facilitation of sustained discharges of polymodal nociceptors, Pain, 38 (19891 85-90, 25 Scadding, J.W., Wall, P.D., Wynn Parry, C,B. and Brooks, D.M., Clinical trial of propranolol in post-traumatic neuralgia, Pain, 14 (1982) 283-292.
317 26 Scheuplein, R.J. and Blank, I.H., Permeability of the skin, Physiol. Rev., 51 (1971) 702-747. 27 Starke, K., a-Adrenoreceptor subclassification, Rev. Physiol. Biochem. Pharmacol., 88 (1981) 199-236. 28 Treede, R.D., Raja, S.N., Davis, K.D., Meyer, R.A. and Campbell, J.N., Evidence that peripheral alpha-adrenergic receptors mediate sympathetically maintained pain. In: M. Bond et al. (Eds.), Proc. Vllth World Congress on Pain, Pain Research and Clinical Management, Vol. 5, Elsevier, Amsterdam, 1991, pp. 377-382. 29 Walker, A.E. and Nulsen, F., Electrical stimulation of the upper thoracic portion of the sympathetic chain in man, Arch. Neurol. Psychiat. (Chicago), 59 (1948) 559-560. 30 Wall, P.D. and Gutnick, M., Ongoing activity in peripheral nerves: the physiology and pharmacology of impulses originating from a neuroma, Exp. Neurol., 43 (1974) 580-593.
31 Wallin, G., Torebj6rk, E. and Hallin, R., Preliminary observations on the pathophysiology of hyperalgesia in the causalgic pain syndrome. In: Y. Zotterman (Ed.), Sensory Functions of the Skin in Primates, Pergamon, New York, 1976, pp. 489-502. 32 Weiner, N., Drugs that inhibit adrenergic nerves and block adrenergic receptors. In: A.G. Gilman, L.S. Goodman, T.W. Rail and F. Murad (Eds.), Goodman and Gilman's The Pharmacological Basis of Therapeutics, 7th edn., MacMillan, New York, 1985, pp. 181-214. 33 Yaksh, T.L., Pharmacology of spinal adrenergic systems which modulate spinal nociceptive processing, Pharmacol. Biochem. Behav., 22 (1985) 845-858. 34 Yaksh, T.L., Hammond, D.L. and Tyce, G.M., Functional aspects of bulbospinal monoaminergic projections in processing of somatosensory information, Fed. Proc., 40 (1981) 2786-2794.