- Email: [email protected]
Secondary hyperalgesia persists in capsaicin desensitized skin
Pain 84 (2000) 141±149 www.elsevier.nl/locate/pain
Secondary hyperalgesia persists in capsaicin desensitized skin Perry N. Fuchs a, 1, James N. Campbell a, b, Richard A. Meyer a, b,* a
Department of Neurosurgery, Johns Hopkins University, Room 5-109 Meyer Building, 600 North Wolfe Street, Baltimore, MD 21287, USA b Applied Physics Laboratory, Johns Hopkins University, Johns Hopkins Road, Laurel, MD 20723, USA Received 28 January 1999; received in revised form 10 June 1999; accepted 26 July 1999
Abstract Several lines of evidence suggest that secondary hyperalgesia to punctate mechanical stimuli arises from central sensitization to the input from primary afferent nociceptors. Conventional C-®ber nociceptors respond to heat stimuli and yet heat hyperalgesia is absent in the region of secondary hyperalgesia. This evidence suggests that the central sensitization to nociceptor input does not involve heat sensitive nociceptors. To test this hypothesis, we investigated whether desensitization of heat sensitive nociceptors by topical application of capsaicin led to an alteration in the secondary hyperalgesia. Two 2 £ 2 cm areas on the volar forearm, separated by 1 cm, were treated in 10 healthy volunteers. One of the areas was desensitized by treatment with 10% topical capsaicin (6 h/day for 2 days). The other site served as vehicle control. Hyperalgesia was produced 2 days later by an intradermal injection of capsaicin (50 mg, 10 ml) at a point midway between the two treatment areas. Secondary hyperalgesia to noxious mechanical stimuli was investigated by using a blade probe (32 and 64 g) attached to a computer-controlled mechanical stimulator. In the area of topical capsaicin treatment, there was a marked increase in heat pain threshold and decrease in heat pain ratings indicating a pronounced desensitization of heat sensitive nociceptors. However, touch threshold and pain to pinching stimuli were not signi®cantly altered. The intradermal capsaicin injection led to the development of a similar degree of secondary hyperalgesia at both the vehicle and capsaicin treatment areas. These results indicate that capsaicin insensitive nociceptive afferents play a dominant role not only in normal mechanical pain but also in secondary hyperalgesia to noxious mechanical stimuli. q 2000 International Association for the Study of Pain. Published by Elsevier Science B.V. Keywords: Hyperalgesia; Allodynia; Capsaicin; Nociceptors; Pricking pain; Sharpness
1. Introduction Cutaneous injury often leads to hyperalgesia, an altered state of sensibility that is characterized by a decrease in pain threshold and an increase in pain to suprathreshold stimuli. Primary hyperalgesia occurs at the site of injury and is characterized by enhanced pain to both heat and mechanical stimuli (Lewis, 1936; Hardy et al., 1950; Raja et al., 1984; Dahl et al., 1993). Secondary hyperalgesia develops in the uninjured area surrounding a cutaneous injury and is characterized by enhanced pain to mechanical, but not heat stimuli (Raja et al., 1984; Ali et al., 1996). Sensitization of peripheral nociceptors is thought to account for many aspects of primary hyperalgesia. In contrast, central sensitization is thought to account for secondary hyperalgesia (Hardy et al., 1950; Campbell et al., 1988; Meyer et al., * Corresponding author. Tel.: 11-410-955-2275; fax: 11-410-9551032. E-mail address: [email protected] (R.A. Meyer) 1 Present address: University of Texas Arlington, Department of Psychology, Box 19528, Arlington, TX 76019, USA.
1988; LaMotte et al., 1991; Simone et al., 1991; TorebjoÈrk et al., 1992). Two types of secondary hyperalgesia to mechanical stimuli have been identi®ed: the pain to light touching or stroking of the skin with a cotton swab has been termed stroking hyperalgesia whereas the enhanced pain to punctate stimuli such as von Frey probes has been termed punctate hyperalgesia (LaMotte et al., 1991). Whereas stroking hyperalgesia (allodynia) appears to be mediated by activity in low-threshold mechanoreceptors, several lines of evidence indicate that punctate hyperalgesia is mediated by different mechanisms. The duration and area of punctate hyperalgesia is greater than the duration and area of stroking-induced pain (LaMotte et al., 1991). The mechanism of punctate hyperalgesia appears to involve a central sensitization to the inputs of nociceptors (as opposed to low-threshold mechanoreceptors). Punctate hyperalgesia, but not stroking hyperalgesia, developed after intradermal capsaicin injection into the arm of a patient with a severe large ®ber neuropathy (Treede and Cole, 1993). Additionally, the pain produced by touching the skin with different wool fabrics was greatly increased in the region of secondary hyperalge-
0304-3959/00/$20.00 q 2000 International Association for the Study of Pain. Published by Elsevier Science B.V. PII: S 0304-395 9(99)00194-3
142
P.N. Fuchs et al. / Pain 84 (2000) 141±149
sia (Cervero et al., 1994). The pain was proportional to the prickliness of the fabrics. Since nociceptors, and not lowthreshold mechanoreceptors, exhibit a differential response to different wool fabrics (Garnsworthy et al., 1988), activity in nociceptors likely contributes to this form of secondary hyperalgesia to wool fabrics. Typical C-®ber nociceptors and many A-®ber nociceptors respond to heat and mechanical stimuli (e.g. Raja et al., 1999). Given that punctate hyperalgesia likely involves a central sensitization to nociceptor inputs, a major unresolved issue concerns the lack of coexistence of heat and mechanical hyperalgesia in the zone of secondary hyperalgesia. Two models have been developed to explain this disparity (Ali et al., 1996). One model of secondary hyperalgesia suggests a central sensitization to input from a subset of nociceptors that are mechano-sensitive but heat-insensitive. An injury, or capsaicin injection, may sensitize the central neurons that receive input from mechano-speci®c heatinsensitive nociceptors which innervate the skin surrounding the injury site. Since the nociceptors are mechano-speci®c, there is no corresponding zone of heat hyperalgesia. A second model of secondary hyperalgesia suggests that injury produces a disinhibition of input from polymodal nociceptors. Mechanical stimuli activate both low-threshold mechanoreceptors and polymodal nociceptors, whereas heat stimuli activate polymodal nociceptors. In normal skin, activity in the low-threshold mechanoreceptors may engage central inhibitory interneurons that act to decrease excitatory input from polymodal nociceptors. An injury may attenuate the inhibitory in¯uences of the low-threshold mechanoreceptors, thereby `releasing' the excitatory input of polymodal nociceptors. Consequently, innocuous mechanical stimuli are perceived as painful. Since heat stimuli do not activate these low-threshold mechanoreceptors, there is no disinhibition of the heat response, and therefore no heat hyperalgesia. Capsaicin serves as an excitotoxin on nociceptors. Capsaicin is believed to operate at a heat sensitive channel and is toxic to the terminals of heat sensitive nociceptors (Simone et al., 1998; Nolano et al., 1999). Evidence indicates that both topical and intradermal capsaicin leads to elimination of epidermal as well as dermal unmyelinated afferents. The present experiment was designed to test whether removal of input from capsaicin sensitive nociceptors leads to an alteration in punctate hyperalgesia. A preliminary report of these data has been presented (Meyer et al., 1998).
2. Methods Healthy volunteers were recruited and provided written consent for the experimental procedure approved by the Joint Committee on Clinical Investigation of the Johns
Hopkins Medical Institutions. Subjects received a stipend as reimbursement for participation. 2.1. General protocol design The protocol extended over a 5-day period. On day 1, subjects were trained in the psychophysical techniques to be used and pretreatment sensory measurements were made. Then, two sites on the volar forearm were treated for 6 h; one with vehicle and the second with topical capsaicin in order to produce desensitization of nociceptive afferents. On day 2, the two sites were again treated for 6 h with the same agent (vehicle or capsaicin) to complete the desensitization process. On day 3, the subjects were not tested so that any acute hyperalgesic effects of topical capsaicin could dissipate. On day 4, capsaicin was injected intradermally at a site midway between the two treatment sites to produce a large zone of secondary hyperalgesia that encompassed the two treatment sites. Measurements of secondary hyperalgesia were made at both treatment sites. On day 5, sensory measurements were made at the vehicle and topical capsaicin site to determine the extent of desensitization produced by the topical capsaicin. 2.2. Topical capsaicin to produce desensitization We wished to identify two locations on the volar forearm for the treatment sites that had similar initial heat pain properties. Two pairs of test sites were marked on the volar forearm of the non-dominant arm about halfway between the wrist and the elbow. The sites that composed a pair were separated by 3 cm. Heat pain thresholds and pain ratings to a suprathreshold heat stimulus (see below) were measured at each site and the pair of sites with the most similar pain thresholds and ratings was chosen for treatment. Once the test sites were selected, additional sensory measures were performed within the 2 £ 2 cm treatment area surrounding the test sites. As described in more detail below, measurements were made of touch detection threshold to von Frey probes, sharpness and pain incidence to a series of weighted needles, pain to a standardized pinch stimulus, cold threshold, heat pain threshold, and heat pain ratings to a suprathreshold stimulus. Following baseline sensory testing, capsaicin and vehicle were applied topically to the test sites. A 2 £ 2 cm cotton gauze was loaded with 0.7 g of a capsaicin cream (consisting of 10% capsaicin dissolved in ethoxydiglycol that was incorporated into a base composed of acetyl alcohol, stearic acid and fatty acid ester) and placed on one of the test sites. A second gauze loaded with a cream that contained just the base was applied to the other test site. There was a 1 cm gap between the two treatment areas. The proximal/distal location of the capsaicin treatment area was randomized. Dermicel tape was used to mask the boundaries of the treatment area and prevent the spread of capsaicin or vehicle. Once the capsaicin and vehicle were applied, the treatment areas were covered with an occlusive dressing for 6 h. At the end of the
P.N. Fuchs et al. / Pain 84 (2000) 141±149
6-h treatment period, the treatment areas were cleaned with alcohol. On the second day, the drugs were applied for an additional 6-h period to the same locations. 2.3. Measurements of desensitization Sensory testing was performed before treatment to con®rm that the test sites were comparable and then on day 5 to measure the extent of desensitization produced by the topical capsaicin. The capsaicin and the vehicle-treated areas were tested for the sensations of heat pain threshold, cold threshold, heat pain ratings to a suprathreshold heat stimulus, touch detection thresholds, sharpness detection and pinch pain ratings using the methods described below. Subjects were provided with two or three training stimuli on day 1 to familiarize them with the measurement techniques. 2.3.1. Thermal testing Thermal testing was performed using a computer controlled Peltier device with an 8 £ 8-mm contact surface (Thermal Devices Inc., Model LTS3). The force of the Peltier head contact with the skin surface was controlled by applying a ®xed weight to the Peltier head that was balanced just above the skin surface. Since the skin surface temperature lagged behind the control temperature of the Peltier system, all temperatures cited in the text were adjusted to the skin-surface calibration temperature (Tillman et al., 1995). 2.3.1.1. Heat pain threshold and cold threshold. The threshold for heat pain was determined using ramped heat stimuli. At the beginning of the heat pain threshold test, the temperature was stepped up from the 308C holding temperature to a baseline temperature of 358C for 3 s. Then the heat stimulus was increased at 0.858C/s. Subjects were instructed to push a button when the stimulus was perceived as painful. The computer automatically returned the temperature to the holding temperature of 308C when the button was pushed or when the cutoff temperature of 51.58C was reached. Heat-pain threshold was de®ned as the temperature of the stimulus at the time that the button was pushed. No correction was made for reaction time artifact. If a subject failed to reach a threshold before the cutoff temperature, then 51.58C was recorded as the heat pain threshold. The threshold for cold detection was determined in a similar manner, except that the baseline temperature started at 308C, and decreased at a rate of 0.858C/s to a 58C cutoff temperature. 2.3.1.2. Pain ratings to suprathreshold heat. Subjects used the technique of magnitude estimation to rate the intensity of pain to a suprathreshold heat stimulus. Subjects were free to choose an arbitrary number to rate the intensity of the ®rst stimulus and then they were instructed to use a ratio scale to rate the intensity of subsequent stimuli. For example, if the ®rst stimulus was
143
given a value of 10, and the subsequent stimulus was twice as painful, the rating for the second stimulus should be 20. The suprathreshold stimulus consisted of a 388C, 3 s baseline followed by a step to 508C for 4 s. 2.3.2. Mechanical testing 2.3.2.1. Touch detection thresholds. Touch detection thresholds were determined using the up/down method of Dixon (1980) with six von Frey mono®laments that were calibrated to administer a force of 0.1, 0.5, 0.9, 3.2, 6.1 or 8.0 mN. Starting with 0.5 mN, the von Frey mono®lament was applied for approximately 1 s within the drug treatment area. If the subject failed to detect the stimulus, then the next higher force von Frey mono®lament was applied. When the subject detected the presence of the stimulus, the next lower von Frey was administered. The up/down test sequence continued for four additional trials after the initial detection. The 50% mechanical detection threshold was calculated using the procedure described in Dixon (1980). 2.3.2.2. Pain ratings to a pinch stimulus. Subjects used magnitude estimation techniques to report the pain intensity to a standardized pinch stimulus. The pinch was performed using an arterial clip. A small piece of skin was grasped with the arterial clip, and clip was left in place for 5 s. At the end of the 5 s, subjects rated the intensity of the pain produced by the arterial clip. The stimulus was applied four times within each treatment area. 2.3.2.3. Sharpness and pain to needle probes. Sharpness detection was determined using a weighted needle device (Chan et al., 1992). The tip of a 30-gauge needle (200 mm diameter) was ®led to produce a ¯at, cylindrical end. A cotton tip applicator was inserted into the Luer connection of the needle, and washers were placed on the shaft of the cotton tip applicator to achieve the desired force level for the stimulus. The entire assembly was then placed inside a 5 ml syringe so that the needle came out of the tip of the syringe and the assembly moved freely within the syringe. When the needle was applied to the skin surface, a reliable and consistent force was applied. Three forces were used: 100, 200 and 400 mN. Each stimulus was applied for about 1 s. Each force was applied 10 times within each of the treatment areas in a pseudo random order. We (and others, e.g. Greenspan and McGillis, 1991) have found in previous psychophysical studies that subjects distinguish sensations of sharpness and pain, and that sharpness is reported at stimulus intensities less than those that evoke pain. Accordingly, the subjects were instructed to indicate if the stimulus was sharp. If a stimulus was sharp, the subject then indicated if the stimulus was painful. 2.3.3. Axon re¯exive ¯are To assess whether the topical capsaicin treatment altered axon-re¯exive ¯are, two laser Doppler probes (Laser¯o,
144
P.N. Fuchs et al. / Pain 84 (2000) 141±149
Model PF 4001, 780 nm, 0.87 mW max; Probe: PF 308, 0.25-mm ®ber separation, TSI Inc., St. Paul, MN) were positioned to record skin blood ¯ow at the center of the two treatment areas. Doppler measurements were started 2-min before the capsaicin injection on day 4 (see below) and continued for 10 min after the injection. The size of the zone of visible ¯are was marked on the skin at 10 min after the capsaicin injection. 2.4. Intradermal capsaicin to produce secondary hyperalgesia On the 4th day, capsaicin (50 mg in a 10 ml volume; Ali et al., 1996) was injected intradermally midway between the two treatment areas to produce a large zone of secondary hyperalgesia that encompassed the two treatment areas. The capsaicin was prepared in a Tween 80 saline vehicle using the method described by LaMotte et al. (1991). Subjects rated their pain to the capsaicin injection for 10 min. At this time, the area of the bleb and the visible ¯are was marked. At 20 min after the injection, pain to computercontrolled force stimuli was measured (see below). Then, the zone of secondary hyperalgesia was mapped using an 200 mN von Frey mono®lament. The entire test sequence was repeated at 60 min after the injection. At the completion of the testing, the areas of drug treatment and markings for ¯are and borders of secondary punctate hyperalgesia were transferred to an acetate sheet. 2.5. Measurements of secondary hyperalgesia Two different techniques were used to determine whether the secondary hyperalgesia that developed at the desensitized site was different from that which developed at the vehicle treated site. Pain ratings to a computer-controlled constant-force mechanical stimulus were obtained for both treatment locations before and after the capsaicin injection. In addition, the zone of secondary hyperalgesia was mapped. 2.5.1. Pain ratings to a force controlled blade stimulus A computer-controlled mechanical stimulator was used to deliver force-controlled stimuli to predetermine locations on the volar forearm (Schneider et al., 1995). The stimulator consists of a servo-controlled linear motor capable of generating 1 kg of force over a 22-mm range. The motor is attached to a three-axis computer-controlled translation system that allows the probe to be positioned accurately at different locations on the volar forearm. The stimuli in this study were 34 or 64 g for 1 s duration and were applied to the skin using a 12 mm wide, 100 mm thick blade. These stimuli have been shown to activate cutaneous nociceptors in the primate (Slugg et al., 1995) and to produce pain in the zone of secondary hyperalgesia (Huang et al., 1997). Eight test sites were located at 1 cm intervals along a line that passed through the center of the treatment locations and the injection site (Fig. 1). This resulted in two mechanical test
Fig. 1. Schematic illustration of experimental paradigm. A 2 £ 2 cm area on the volar forearm was pretreated with capsaicin to produce desensitization. A second vehicle treated area served as the control. Two days later, intradermal injection of capsaicin between the two treatment sites produced a large zone of secondary hyperalgesia to mechanical stimuli that extended beyond the treatment zones. A computer-controlled mechanical stimulator system was used to apply controlled force stimuli via a blade probe to the sites indicated by the vertical lines. Subjects rated their pain to these stimuli using magnitude estimation techniques.
sites that were within each treatment location as well as two sites outside of each treatment location. No testing was done at the capsaicin injection site. The order of stimulus presentation (both for force and position) was randomized except for the constraints that (1) each location received two presentations of the 64 g stimulus and one presentation of the 32 g stimulus; and (2) two consecutive stimuli were not presented within 1 cm of each other. Within each test site location, the three stimulus presentations were separated by 1 mm to minimize any possible fatigue due to a previous mechanical stimulus presentation (Slugg et al., 1995). Each mechanical test sequence consisted of 24 stimulus presentations with a 20 s interstimulus interval. Subjects rated the pain to each stimulus using the technique of magnitude estimation. 2.5.2. Mapping of zone of secondary hyperalgesia To map the zone of secondary hyperalgesia, a 200 mN von Frey was applied well outside the expected area of hyperalgesia and then at 1 cm intervals along a radial path towards the injection site. The subjects were instructed to indicate when they felt that the stimulus became painful or if, originally painful, became notably more painful. The border of the secondary zone was con®rmed by retesting near the edge of the zone and then indicated on the skin with a marker pen. This procedure was repeated until a minimum of eight radial paths had been completed. 2.6. Data analysis 2.6.1. Normalization of data Subjects were free to use numbers of their own choice to rate the intensity of pain. Since a wide range of numbers were used, we needed to normalize the data in order to combine data across subjects. Data were normalized for a given subject by dividing that subject's ratings to a given stimulus by that subject's maximum rating to that stimulus.
P.N. Fuchs et al. / Pain 84 (2000) 141±149
145
Changes in laser Doppler measure of perfusion were normalized by dividing the average perfusion during the last 5-min of the 10 min post capsaicin injection period by the average perfusion during the 2 min before the capsaicin injection. 2.6.2. Analysis Data were analyzed using either Wilcoxon matched paired test, paired two-tailed t-tests or one-way analysis of variance (ANOVA). Signi®cance was P , 0:05. 3. Results Ten male subjects with age ranging from 23±42 years (mean 32 year) were enrolled in the study. One subject developed a very small zone of secondary hyperalgesia which extended less than 1.5 cm in the vehicle treated direction. This subject was excluded from subsequent analysis. All other subjects had a zone of punctate hyperalgesia that extended at least 3.5 cm in the vehicle treated direction. 3.1. Measurements of desensitization Sensory testing performed before treatment con®rmed that the test sites were comparable. Pre-treatment values for heat pain threshold, cold threshold, heat pain ratings, touch detection thresholds, sharpness detection thresholds, and pinch pain ratings at the capsaicin and the vehicle treatment locations were not signi®cantly different. 3.1.1. Thermal testing 3.1.1.1. Heat pain threshold. The median heat pain threshold at the capsaicin treated locations was signi®cantly higher than the heat pain threshold at the vehicle treated locations (P , 0:01, Wilcoxon matched pairs, Fig. 2a). All subjects had heat pain thresholds less than 498C at the vehicle treated area. In most subjects (8/9), the heat pain threshold at the capsaicin treated area was higher than the maximum temperature used (i.e. 51.58C). 3.1.1.2. Pain to suprathreshold heat stimuli. Topical capsaicin treatment also resulted in a signi®cant decrease in pain ratings to the suprathreshold stimulus (508C, 4 s) at the capsaicin treatment location compared to the vehicle treatment location (P , 0:05, Fig. 2B). At the capsaicin treatment location, eight of the nine subjects reported no pain to the suprathreshold heat stimulus. In contrast, only two subjects reported no pain to this stimulus at the vehicle treatment location. 3.1.1.3. Cold threshold. The mean cold detection threshold at the capsaicin treatment location (22:0 ^ 1:18C) was signi®cantly lower than at the vehicle treatment location (25:0 ^ 1:08C, P , 0:05). However, all
Fig. 2. Desensitization following treatment with topical capsaicin. (A) Heat pain thresholds were signi®cantly higher at the capsaicin treatment site than at the vehicle treatment site. Heat pain threshold was determined using a 0.858C/s ramp from 358C to 51.58C (median ^ 25%). (B) Heat pain ratings to the suprathreshold stimulus were signi®cantly lower at the capsaicin treatment site. Subjects rated the intensity of pain to a 508C, 4 s suprathreshold heat stimulus (mean ^ SEM). (C) Pain ratings to a pinch stimulus were not altered by the capsaicin treatment. Subjects rated the intensity of pain to an arterial clip applied to a fold of skin for 5 s (mean ^ SEM) (* P , 0:05; ** P , 0:01; Wilcoxon matched pairs).
subjects detected cold prior to the 58C cutoff temperature at both the capsaicin and vehicle treated test sites. 3.1.2. Mechanical testing 3.1.2.1. Touch detection thresholds and pain ratings to the pinch stimulus. The touch detection thresholds at the capsaicin (1:8 ^ 0:4 mN) and vehicle (2:1 ^ 0:3 mN) treated areas were not signi®cantly different (P . 0:40). Similarly, the pain ratings to the pinch stimulus were not signi®cantly different at the capsaicin and vehicle treated areas (P . 0:20, Fig. 2C). 3.1.2.2. Sharpness and pain to needle probes.
In response
146
P.N. Fuchs et al. / Pain 84 (2000) 141±149
to the needle probes, the incidence of sharpness detection and the incidence of pain increased monotonically with the applied force at both the vehicle and capsaicin treated sites. However, the incidence of sharpness detection tended to be lower on the capsaicin treated skin than on vehicle treated skin. Although pain was still produced by the punctate probes, the incidence of pain reports was signi®cantly lower in capsaicin treated skin. For example, the 400 mN probe produced a signi®cantly higher incidence of pain in vehicle treated skin (53 ^ 10%) than in capsaicin treated skin (27 ^ 10%, P , 0:01), whereas the incidence of sharpness in the vehicle treated skin (91 ^ 5%) was not signi®cantly different from capsaicin treated skin (72 ^ 12%, P . 0:1). 3.1.3. Axon re¯exive ¯are The capsaicin injection between the two treatment areas resulted in the appearance of a visible ¯are that always extended into the vehicle treatment area but never extended into the capsaicin treatment area. In all subjects, a sharp edge to the ¯are was visualized that corresponded to the border of the capsaicin treatment area. The area of ¯are on the vehicle-treated side of the capsaicin injection (7:7 ^ 1:5 cm 2) was signi®cantly larger than the area on the capsaicintreated side (2:6 ^ 1:3 cm 2, P , 0:01). The laser Doppler measurements of ¯are revealed a 7:1 ^ 1:7-fold increase in perfusion at the vehicle site but only a 2:8 ^ 0:4-fold increase at the capsaicin treatment site (P , 0:05).
Fig. 3. Secondary hyperalgesia persists in the capsaicin desensitized area. Normalized pain ratings to the 64 g stimulus applied via the blade probe are plotted as function of time after the capsaicin injection. A signi®cant increase in pain ratings was observed at both the capsaicin desensitized area and the vehicle treatment area after the intradermal injection of capsaicin. Data for a given subject were averaged across the measurements within each treatment site (mean ^ SEM; ** P , 0:01, post vs. pre; D P , 0:05, capsaicin vs. vehicle).
on the vehicle treatment side of the capsaicin injection (21:7 ^ 2:5 cm 2 at 20 min, 19:7 ^ 3:7 cm 2 at 60 min) was not signi®cantly different from the area on the capsaicin treatment side (16:7 ^ 3:9 cm 2; 18:2 ^ 6:2 cm 2; P . 0:5).
3.2. Pain to capsaicin injection
4. Discussion
Intradermal injection of capsaicin produced an intense sensation of burning pain that started immediately with the injection. The pain reached a peak within the ®rst 5± 10 s and then gradually decreased during the next 10 min. The rate of decrease was greatest during the ®rst 5 min following the injection.
These experiments demonstrate that secondary hyperalgesia to noxious mechanical stimuli is little affected in capsaicin desensitized skin. The interpretation of this ®nding centers on an interpretation of what capsaicin does to cutaneous afferents. There are three sources of data that contribute to this understanding: Neurophysiological, histological, and psychophysical.
3.3. Measurements of secondary hyperalgesia 3.3.1. Pain to the computer-controlled mechanical stimulus The intradermal injection of capsaicin led to a pronounced mechanical hyperalgesia in both the vehicle and capsaicin treated areas (Fig. 3). The blade probe produced little pain before the capsaicin injection. Pain ratings were signi®cantly higher at the capsaicin and vehicle treatment locations after the capsaicin injection (P , 0:01). Thus, secondary hyperalgesia to punctate stimuli persisted in the capsaicin desensitized skin. Although the pain ratings in the capsaicin desensitized skin were signi®cantly lower than in the vehicle treated skin (P , 0:05) at the 20 min time point, this difference disappeared at the 60 min time point (Fig. 3). 3.3.2. Area of punctate hyperalgesia The area of punctate hyperalgesia was measured with a 200 mN von Frey probe. The area of punctate hyperalgesia
4.1. What nociceptive afferents survive capsaicin treatment From a neurophysiological perspective capsaicin appears to excite initially then deactivate nociceptors through a receptor mediated effect on a cation channel. Baumann et al. (1991) found that local injection of capsaicin induced a deactivation of C-®ber nociceptors to both heat and mechanical stimuli. This deactivation was restricted to that area in the receptive ®eld where the injection was administered. Other studies demonstrate that capsaicin opens a cation channel permeable to Na 1 and Ca 21 (Docherty et al., 1991; Bevan and Docherty, 1993). The mechanism of neurotoxicity of capsaicin is not clearly understood. However, evidence points to effects of excessive intracellular calcium and induction of proteases (Chard et al., 1995). Histological studies by Simone and collegues (1998) determined that even at doses as low as 2 mg capsaicin injected intradermally caused a localized loss of all intra-epidermal
P.N. Fuchs et al. / Pain 84 (2000) 141±149
®bers as determined with the pan-axonal marker PGP 9.5. Topical capsaicin was found in a separate study to have similar effects (Nolano et al., 1999). In the psychophysical study presented here capsaicin created a dissociated sensory loss, as manifest by heat analgesia, but only a small effect on pain to mechanical stimuli. At the capsaicin treatment location, eight of the nine subjects reported no pain to the supra-threshold heat stimulus (508C, 4 s). Others have reported similar ®ndings (e.g. SzolcsaÂnyi, 1977; Carpenter and Lynn, 1981; Jancso et al., 1985; Bjerring and Arendt-Nielsen, 1990; Simone and Ochoa, 1991; Davis et al., 1995; Beydoun et al., 1996; Simone et al., 1998; Fuchs et al., 1999). Yet, effects on pain to mechanical stimuli were more subtle. The pain to the punctate probe was signi®cantly less in the capsaicin treated skin, but the incidence of reports of pain only decreased by half to the highest force used (400 mN). The difference in reports of sharpness to a needle probe tended to be less in the capsaicin treated region, but this difference did not reach statistical signi®cance. There was no effect on pain to the pinch stimulus. In addition, there was no measurable effect on tactile sensibility. Others have also reported that pain to mechanical stimuli is only modestly (if at all) affected in capsaicin treated skin (e.g. Simone and Ochoa, 1991; Davis et al., 1995; Beydoun et al., 1996; Magerl et al., 1998; Simone et al., 1998). Our ®nding that the threshold for cold sensation decreased signi®cantly is consistent with the recent observation that the magnitude of cold sensation decreases after topical capsaicin treatment (Nolano et al., 1999). Topical application of capsaicin also leads to a decrease in itch sensitivity (ToÂth-KaÂsa et al., 1986; Handwerker et al., 1987; Simone and Ochoa, 1991) as well as a decreased response to bradykinin and capsaicin (Crimi et al., 1992) indicating that chemically sensitive afferents are also affected. In addition there is a marked decrease in axon re¯exive ¯are (e.g. Jancso et al., 1968; Carpenter and Lynn, 1981; SzolcsaÂnyi, 1988; Bjerring and Arendt-Nielsen, 1990). Sympathetic efferent ®bers appear not to be affected as evidenced by the lack of change in nicotine-induced axon re¯ex sweating (Izumi and Karita, 1988) in capsaicin treated skin. Given that the sensory changes are correlated with the striking loss of epidermal and even dermal ®bers (Nolano et al., 1999), the loss of heat sensibility is likely due to degeneration of C-®ber and A-®ber nociceptors responsive to heat stimuli. That pain to mechanical stimuli persists after capsaicin application suggests that a separate class of nociceptors responsive to mechanical stimuli exists that is resistant to the effects of capsaicin. Some have argued that the `capsaicin channel' is in effect a heat-transduction channel (e.g. Caterina et al., 1997). Thus nociceptors unresponsive to heat may, in effect, be protected from the toxic effects of capsaicin. The neurophysiological, histological, and psychophysical data taken together argue that heat sensitive nociceptors undergo local degeneration and that the persistent pain to mechanical stimuli arises from the input of
147
mechanically sensitive and heat insensitive nociceptive afferents. In a related study Magerl et al. (1998) undertook to determine the relative role of C- and A-®ber nociceptors as mediators of pain to mechanical stimuli in capsaicin desensitized skin. As in this study, capsaicin had only modest effects on pain to mechanical stimuli. However, a selective A-®ber block obtained by application of pressure over the super®cial radial nerve eliminated the remaining pain sensibility to mechanical stimuli. In another experiment, Ziegler et al. (1999) demonstrated that secondary hyperalgesia to punctate stimuli following capsaicin injection was not present when A-®ber conduction was blocked. These data support the conclusion that a class of A-®ber nociceptors insensitive to heat accounts for the pain to mechanical stimuli in capsaicin treated skin. 4.2. A-®ber nociceptors A-®ber nociceptors may be divided into three categories with regard to the heat response (Treede et al., 1998). Type I ®bers typically have an accelerating response to heat with a long receptor utilization period (time between activation and stimulus onset). The heat threshold of these ®bers for short (1±3 s) duration stimuli is high, often exceeding 538C. Because the heat tests in this study (and in the study by Nolano et al., 1999) would not be suf®cient to activate the majority of these ®bers, a role of type I A-®ber nociceptors in serving punctate hyperalgesia cannot be excluded. Type II A-®ber nociceptors have an adapting response to heat (similar to that of polymodal C-®ber nociceptors), with a short receptor utilization period. These ®bers serve ®rst pain sensation, a perception that is lost in capsaicin desensitized skin (Beydoun et al., 1996). Thus, these ®bers are unlikely to play a role in punctate hyperalgesia. A third variety of A®ber nociceptors respond to mechanical stimuli, but have no response to heat even at high intensities (Treede et al., 1998). These ®bers are leading candidates to serve punctate hyperalgesia. Little is known about the responsiveness of these ®bers to chemical stimuli. If indeed these ®bers are responsive to chemical stimuli, hyperalgesia to chemical stimuli in the secondary zone would be predicted. This has as yet not been tested. Though there was some change in painfulness of pinprick stimuli following the topical capsaicin administration, the decrease in the sensation of sharpness did not reach signi®cance. This would argue that sharpness may qualify as a separate sensation, and moreover that sharpness too, is served by A-®ber nociceptors (Garell et al., 1996). 4.3. Neural mechanisms of punctate hyperalgesia Two models to account for punctate hyperalgesia were presented in the Section 1. We assume, based on the evidence presented, that punctate hyperalgesia arises from a central sensitization to the input of nociceptive afferents. Moreover, the model must account for the absence of heat
148
P.N. Fuchs et al. / Pain 84 (2000) 141±149
hyperalgesia in the secondary zone. Now we have further data on which to base the model. The nociceptive afferents that provide the peripheral signals for punctate hyperalgesia survive capsaicin treatment. The evidence rules out the disinhibition model proposed in Section 1 and points to a class of A-®bers that is unresponsive to heat stimuli. Dougherty et al. (1998) studied the responses of primate spinothalamic projection cells that innervated areas of skin adjacent to a capsaicin injection. These cells acutely showed the expected sensitization to mechanical stimuli. The response to short duration heat stimuli was actually attenuated. This neurophysiological ®nding is in keeping with psychophysical ®ndings of Raja et al. (1984) who determined that shortly after a burn injury there was coexistent mechanical hyperalgesia and heat hypalgesia. Thus, the CNS determinants of the dissociated hyperalgesia to mechanical versus heat stimuli are already in place at the level of the dorsal horn spinothalamic projection cells. Mechanisms within the dorsal horn, therefore, likely account for punctate hyperalgesia. Four separate models of dorsal horn circuitry could explain these ®ndings.
tors with no heat sensitivity (or very high thresholds) encodes the punctate hyperalgesia that occurs in the zone of secondary hyperalgesia: (1) punctate hyperalgesia persists in capsaicin desensitized skin; (2) capsaicin appears to destroy afferent terminals in ®bers responsive to heat stimuli; (3) remaining pain-sensitivity to mechanical stimuli in capsaicin treated skin is eliminated by an A-®ber block; and (4) A-®ber nociceptors with selective sensitivity to mechanical stimuli exist. Future consideration of the differential dorsal horn connections of different types of nociceptors, anatomically, chemically, and physiologically, may illuminate further the mechanisms of punctate hyperalgesia.
1. Pre-synaptic regulation: The mechano-sensitive A-®ber nociceptors project to spinothalamic projection cells that receive convergent input from polymodal nociceptors. Collaterals from polymodal nociceptors that serve the injury zone have presynaptic connections with the central terminals of A-®ber nociceptors. This leads to enhanced neurotransmitter release through a mechanism such as primary afferent hyperpolarization. These collateral inputs do not affect the transmission of action potentials from polymodal nociceptors. Thus hyperalgesia is present only for mechanical stimuli. 2. Interneuron speci®city: The model here is similar except that the enhancement of input of the mechano-sensitive A-®ber nociceptors is postulated to occur as a result of sensitization of an interneuron that receives the A-®ber input. Polymodal nociceptors could sensitize these neurons through a post-synaptic mechanism. 3. Neurotransmitter speci®city: If polymodal and mechanospeci®c nociceptors utilize different transmitters at their junctions with spinothalamic projection neurons then selective sensitization could result from this difference. 4. Dendritic speci®city: Conceivably different parts of the dendritic tree of spinothalamic neurons could be specialized to receive the inputs of different afferents. If so, these post-synaptic regions could be selectively sensitized.
References
Intracellular recordings of dorsal horn cells could help resolve these models. 4.4. Conclusion The evidence presented here, when considered with other data, favors the hypothesis that a class of A-®ber nocicep-
Acknowledgements We wish to thank Mr. Timothy Hartke and Ms. Sylvia Horasek for their technical support, Dr. Gang Wu for reviewing the manuscript, and Dr. Sam Georgiou for preparing the capsaicin formulation. This research was supported by the NIH on grant NS-14447.
Ali Z, Meyer RA, Campbell JN. Secondary hyperalgesia to mechanical but not heat stimuli following a capsaicin injection in hairy skin. Pain 1996;68:401±411. Baumann TK, Simone DA, Shain CN, LaMotte RH. Neurogenic hyperalgesia: the search for the primary cutaneous afferent ®bers that contribute to capsaicin-induced pain and hyperalgesia. J Neurophysiol 1991;66:212±227. Bevan SJ, Docherty RJ. Cellular mechanisms of the action of capsaicin. In: Wood J, editor. Capsaicin in the Study of Pain, London: Academic Press, 1993. pp. 27±44. Beydoun A, Dyke DBS, Morrow TJ, Casey KL. Topical capsaicin selectively attenuates heat pain and A d ®ber-mediated laser-evoked potentials. Pain 1996;65:189±196. Bjerring P, Arendt-Nielsen L. Inhibition of histamine skin ¯are reaction following repeated topical applications of capsaicin. Allergy 1990;45:121±125. Campbell JN, Khan AA, Meyer RA, Raja SN. Response to heat of C-®ber nociceptors in monkey are altered by injury in the receptive ®eld but not by adjacent injury. Pain 1988;32:327±332. Carpenter SE, Lynn B. Vascular and sensory responses of human skin to mild injury after topical treatment with capsaicin. Br J Pharmacol 1981;73:755±758. Caterina MJ, Schumacher MA, Tominaga M, Rosen TA, Levine JD, Julius D. The capsaicin receptor: a heat-activated ion channel in the pain pathway. Nature 1997;389:816±824. Cervero F, Meyer RA, Campbell JN. A psychophysical study of secondary hyperalgesia: Evidence for increased pain to input from nociceptors. Pain 1994;58:21±28. Chan AW, MacFarlane IA, Bowsher D, Campbell JA. Weighted needle pinprick sensory thresholds: a simple test of sensory function in diabetic peripheral neuropathy. J Neurol Neurosurg Psychiat 1992;55:56±59. Chard PS, Bleakman D, Savidge JR, Miller RJ. Capsaicin-induced neurotoxicity in cultured dorsal root ganglion neurons: involvement of calcium-activated protease. Neuroscience 1995;65:1099±1108. Crimi N, Polosa R, Maccarrone C, Palermo B, Palermo F, Mistretta A. Effect of topical application with capsaicin on skin responses to bradykinin and histamine in man. Clin Exp Allergy 1992;22:933±939. Dahl JB, Brennum J, Arendt-Nielsen L, Jensen TS, Kehlet H. The effects of
P.N. Fuchs et al. / Pain 84 (2000) 141±149 pre- versus postinjury in®ltration with lidocaine on thermal and mechanical hyperalgesia after heat injury to the skin. Pain 1993;53:43±51. Davis KD, Meyer RA, Turnquist JL, Filloon TG, Pappagallo M, Campbell JN. Cutaneous injection of the capsaicin analogue. NE-21610, produces analgesia to heat but not to mechanical stimuli in man. Pain 1995;61:17±26. Dixon WJ. Ef®cient analysis of experimental observations. Annu Rev Phamacol Toxicol 1980;20:441±462. Docherty RJ, Robertson B, Bevan S. Capsaicin causes prolonged inhibition of voltage-activated calcium currents in adult rat dorsal root ganglion neurons in culture. Neuroscience 1991;40:513±521. Dougherty PM, Willis WD, Lenz FA. Transient inhibition of responses to thermal stimuli of spinal sensory tract neurons in monkeys during sensitization by intradermal capsaicin. Pain 1998;77:129±136. Fuchs PN, Pappagallo M, Meyer RA. Topical EMLA pre-treatment fails to decrease the pain induced by 1% topical capsaicin. Pain 1999;80:637± 642. Garell PC, McGillis SLB, Greenspan JD. Mechanical response properties of nociceptors innervating feline hairy skin. J Neurophysiol 1996;75:1177±1189. Garnsworthy RK, Gully RL, Kenins P, May®eld RJ, Westerman RA. Identi®cation of the physical stimulus and the neural basis of fabric-evoked prickle. J Neurophysiol 1988;59:1083±1097. Greenspan JD, McGillis SLB. Stimulus features relevant to the perception of sharpness and mechanically evoked cutaneous pain. Somatosens Motor Res 1991;8:137±147. Handwerker HO, Magerl W, Klemm F, Lang E, Westerman RA. Quantitative evaluation of itch sensation. In: Schmidt RF, Schaible HG, VahleHinz C, editors. Fine Afferent Nerve Fibers and Pain, New York: Weinheim, 1987. pp. 461±473. Hardy JD, Wolff HG, Goodell H. Experimental evidence on the nature of cutaneous hyperalgesia. J Clin Invest 1950;29:115±140. Huang JH, Ali Z, Campbell JN, Meyer RA. Pain ratings to punctate mechanical stimuli are similar through-out the zone of secondary hyperalgesia. Soc Neurosci Abstr 1997;23:1258. Izumi H, Karita K. Investigation of mechanisms of the ¯are and wheal reactions in human skin by band method. Brain Res 1988;449:328±331. Jancso N, JanscoÂ-GaÂbor A, SzolcsaÂnyi J. The role of sensory nerve endings in neurogenic in¯ammation induced in human skin and in the eye and paw of the rat. Br J Pharm Chemother 1968;32:32±41. Jancso G, ObaÂl FJ, ToÂth-KaÂsa I, Katona M, Husz S. The modulation of cutaneous in¯ammatory reactions by peptide-containing sensory nerves. Int J Tissue Reac 1985;7:449±457. LaMotte RH, Shain CN, Simone DA, Tsai E-FP. Neurogenic hyperalgesia: Psychophysical studies of underlying mechanisms. J Neurophys 1991;66:190±211. Lewis T. Experiments relating to cutaneous hyperalgesia and its spread through somatic nerves. Clin Sci 1936;2:373±423. Magerl W, Fuchs PN, Meyer RA, Treede R-D. Pain to punctate mechanical stimuli in humans is mainly mediated by capsaicin-insensitive A-®ber nociceptors. Soc Neurosci Abstr 1998;24:2085.
149
Meyer RA, Campbell JN, Raja SN. Antidromic nerve stimulation in monkey does not sensitize unmyelinated nociceptors to heat. Brain Res 1988a;441:168±172. Meyer RA, Magerl W, Campbell JN, Treede R-D, Fuchs PN. Normal punctate hyperalgesia in capsaicin desensitized skin. Soc Neurosci Abstr 1998b;24:2086. Nolano M, Simone DA, Wendelschafer-Crabb, Johnson T, Hazen E, Kennedy WR. Topical capsaicin in humans: parallel loss of epidermal nerve ®bers and pain sensation. Pain 1999;81:135±145. Raja SN, Campbell JN, Meyer RA. Evidence for different mechanisms of primary and secondary hyperalgesia following heat injury to the glabrous skin. Brain 1984;107:1179±1188. Raja SN, Meyer RA, Ringkamp M, Campbell JN. Peripheral neural mechanisms of nociception. In: Melzack R, Wall PD, editors. Textbook of Pain, 4th edition, Churchill Livingstone, 1999 in press. Schneider W, Slugg RM, Turnquist BP, Meyer RA, Campbell JN. An electromechanical stimulator system for neurophysiological and psychophysical studies of pain. J Neurosci Methods 1995;60:61±68. Simone DA, Ochoa J. Early and late effects of prolonged topical capsaicin on cutaneous sensibility and neurogenic vasodilatation in humans. Pain 1991;47:285±294. Simone DA, Sorkin LS, Oh U, Chung JM, Owens C, LaMotte RH, Willis WD. Neurogenic hyperalgesia: Central neural correlates in responses of spinothalamic tract neurons. J Neurophysiol 1991;66:228±246. Simone DA, Nolano M, Johnson T, Wendelschafer-Crabb G, Kennedy WR. Intradermal injection of capsaicin in humans produces degeneration and subsequent reinnervation of epidermal nerve ®bers: correlation with sensory function. J Neurosci 1998;18:8947±8959. Slugg RM, Meyer RA, Campbell JN. Mechanical receptive ®eld structure for cutaneous C-®ber and A-®ber nociceptors in the monkey determined with graded intensity stimuli. Soc Neurosci Abstr 1995;21:1161. SzolcsaÂnyi J. A pharmacological approach to elucidation of the role of different nerve ®bers and receptor endings in mediation of pain. J Physiol 1977;73:251±259. SzolcsaÂnyi J. Antidromic vasodilation and neurogenic in¯ammation. Agents Actions 1988;23:4±11. Tillman DB, Treede R-D, Meyer RA, Campbell JN. Response of C ®bre nociceptors in the anaesthetized monkey to heat stimuli: Correlation with pain threshold in humans. J Physiol 1995;485:767±774. TorebjoÈrk HE, Lundberg LER, LaMotte RH. Central changes in processing of mechanoreceptive input in capsaicin-induced secondary hyperalgesia in humans. J Physiol 1992;448:765±780. ToÂth-KaÂsa I, Jancso G, BognaÂr A, Husz S, ObaÂl FJ. Capsaicin prevents histamine-induced itching. Int J Clin Pharmacol Res 1986;6:163±169. Treede R-D, Cole JD. Dissociated secondary hyperalgesia in a subject with large ®bre sensory neuropathy. Pain 1993;53:169±174. Treede R-D, Campbell JN, Meyer RA. Myelinated mechanically-insensitive afferents from monkey hairy skin: Heat response properties. J Neurophys 1998;80:1082±1093. Ziegler EA, Magerl W, Meyer RA, Treede R-D. Secondary hyperalgesia to punctate mechanical stimuli: central sensitization to A-®bre nociceptor input. Brain 1999:in press.
Secondary hyperalgesia persists in capsaicin desensitized skin Perry N. Fuchs a, 1, James N. Campbell a, b, Richard A. Meyer a, b,* a
Department of Neurosurgery, Johns Hopkins University, Room 5-109 Meyer Building, 600 North Wolfe Street, Baltimore, MD 21287, USA b Applied Physics Laboratory, Johns Hopkins University, Johns Hopkins Road, Laurel, MD 20723, USA Received 28 January 1999; received in revised form 10 June 1999; accepted 26 July 1999
Abstract Several lines of evidence suggest that secondary hyperalgesia to punctate mechanical stimuli arises from central sensitization to the input from primary afferent nociceptors. Conventional C-®ber nociceptors respond to heat stimuli and yet heat hyperalgesia is absent in the region of secondary hyperalgesia. This evidence suggests that the central sensitization to nociceptor input does not involve heat sensitive nociceptors. To test this hypothesis, we investigated whether desensitization of heat sensitive nociceptors by topical application of capsaicin led to an alteration in the secondary hyperalgesia. Two 2 £ 2 cm areas on the volar forearm, separated by 1 cm, were treated in 10 healthy volunteers. One of the areas was desensitized by treatment with 10% topical capsaicin (6 h/day for 2 days). The other site served as vehicle control. Hyperalgesia was produced 2 days later by an intradermal injection of capsaicin (50 mg, 10 ml) at a point midway between the two treatment areas. Secondary hyperalgesia to noxious mechanical stimuli was investigated by using a blade probe (32 and 64 g) attached to a computer-controlled mechanical stimulator. In the area of topical capsaicin treatment, there was a marked increase in heat pain threshold and decrease in heat pain ratings indicating a pronounced desensitization of heat sensitive nociceptors. However, touch threshold and pain to pinching stimuli were not signi®cantly altered. The intradermal capsaicin injection led to the development of a similar degree of secondary hyperalgesia at both the vehicle and capsaicin treatment areas. These results indicate that capsaicin insensitive nociceptive afferents play a dominant role not only in normal mechanical pain but also in secondary hyperalgesia to noxious mechanical stimuli. q 2000 International Association for the Study of Pain. Published by Elsevier Science B.V. Keywords: Hyperalgesia; Allodynia; Capsaicin; Nociceptors; Pricking pain; Sharpness
1. Introduction Cutaneous injury often leads to hyperalgesia, an altered state of sensibility that is characterized by a decrease in pain threshold and an increase in pain to suprathreshold stimuli. Primary hyperalgesia occurs at the site of injury and is characterized by enhanced pain to both heat and mechanical stimuli (Lewis, 1936; Hardy et al., 1950; Raja et al., 1984; Dahl et al., 1993). Secondary hyperalgesia develops in the uninjured area surrounding a cutaneous injury and is characterized by enhanced pain to mechanical, but not heat stimuli (Raja et al., 1984; Ali et al., 1996). Sensitization of peripheral nociceptors is thought to account for many aspects of primary hyperalgesia. In contrast, central sensitization is thought to account for secondary hyperalgesia (Hardy et al., 1950; Campbell et al., 1988; Meyer et al., * Corresponding author. Tel.: 11-410-955-2275; fax: 11-410-9551032. E-mail address: [email protected] (R.A. Meyer) 1 Present address: University of Texas Arlington, Department of Psychology, Box 19528, Arlington, TX 76019, USA.
1988; LaMotte et al., 1991; Simone et al., 1991; TorebjoÈrk et al., 1992). Two types of secondary hyperalgesia to mechanical stimuli have been identi®ed: the pain to light touching or stroking of the skin with a cotton swab has been termed stroking hyperalgesia whereas the enhanced pain to punctate stimuli such as von Frey probes has been termed punctate hyperalgesia (LaMotte et al., 1991). Whereas stroking hyperalgesia (allodynia) appears to be mediated by activity in low-threshold mechanoreceptors, several lines of evidence indicate that punctate hyperalgesia is mediated by different mechanisms. The duration and area of punctate hyperalgesia is greater than the duration and area of stroking-induced pain (LaMotte et al., 1991). The mechanism of punctate hyperalgesia appears to involve a central sensitization to the inputs of nociceptors (as opposed to low-threshold mechanoreceptors). Punctate hyperalgesia, but not stroking hyperalgesia, developed after intradermal capsaicin injection into the arm of a patient with a severe large ®ber neuropathy (Treede and Cole, 1993). Additionally, the pain produced by touching the skin with different wool fabrics was greatly increased in the region of secondary hyperalge-
0304-3959/00/$20.00 q 2000 International Association for the Study of Pain. Published by Elsevier Science B.V. PII: S 0304-395 9(99)00194-3
142
P.N. Fuchs et al. / Pain 84 (2000) 141±149
sia (Cervero et al., 1994). The pain was proportional to the prickliness of the fabrics. Since nociceptors, and not lowthreshold mechanoreceptors, exhibit a differential response to different wool fabrics (Garnsworthy et al., 1988), activity in nociceptors likely contributes to this form of secondary hyperalgesia to wool fabrics. Typical C-®ber nociceptors and many A-®ber nociceptors respond to heat and mechanical stimuli (e.g. Raja et al., 1999). Given that punctate hyperalgesia likely involves a central sensitization to nociceptor inputs, a major unresolved issue concerns the lack of coexistence of heat and mechanical hyperalgesia in the zone of secondary hyperalgesia. Two models have been developed to explain this disparity (Ali et al., 1996). One model of secondary hyperalgesia suggests a central sensitization to input from a subset of nociceptors that are mechano-sensitive but heat-insensitive. An injury, or capsaicin injection, may sensitize the central neurons that receive input from mechano-speci®c heatinsensitive nociceptors which innervate the skin surrounding the injury site. Since the nociceptors are mechano-speci®c, there is no corresponding zone of heat hyperalgesia. A second model of secondary hyperalgesia suggests that injury produces a disinhibition of input from polymodal nociceptors. Mechanical stimuli activate both low-threshold mechanoreceptors and polymodal nociceptors, whereas heat stimuli activate polymodal nociceptors. In normal skin, activity in the low-threshold mechanoreceptors may engage central inhibitory interneurons that act to decrease excitatory input from polymodal nociceptors. An injury may attenuate the inhibitory in¯uences of the low-threshold mechanoreceptors, thereby `releasing' the excitatory input of polymodal nociceptors. Consequently, innocuous mechanical stimuli are perceived as painful. Since heat stimuli do not activate these low-threshold mechanoreceptors, there is no disinhibition of the heat response, and therefore no heat hyperalgesia. Capsaicin serves as an excitotoxin on nociceptors. Capsaicin is believed to operate at a heat sensitive channel and is toxic to the terminals of heat sensitive nociceptors (Simone et al., 1998; Nolano et al., 1999). Evidence indicates that both topical and intradermal capsaicin leads to elimination of epidermal as well as dermal unmyelinated afferents. The present experiment was designed to test whether removal of input from capsaicin sensitive nociceptors leads to an alteration in punctate hyperalgesia. A preliminary report of these data has been presented (Meyer et al., 1998).
2. Methods Healthy volunteers were recruited and provided written consent for the experimental procedure approved by the Joint Committee on Clinical Investigation of the Johns
Hopkins Medical Institutions. Subjects received a stipend as reimbursement for participation. 2.1. General protocol design The protocol extended over a 5-day period. On day 1, subjects were trained in the psychophysical techniques to be used and pretreatment sensory measurements were made. Then, two sites on the volar forearm were treated for 6 h; one with vehicle and the second with topical capsaicin in order to produce desensitization of nociceptive afferents. On day 2, the two sites were again treated for 6 h with the same agent (vehicle or capsaicin) to complete the desensitization process. On day 3, the subjects were not tested so that any acute hyperalgesic effects of topical capsaicin could dissipate. On day 4, capsaicin was injected intradermally at a site midway between the two treatment sites to produce a large zone of secondary hyperalgesia that encompassed the two treatment sites. Measurements of secondary hyperalgesia were made at both treatment sites. On day 5, sensory measurements were made at the vehicle and topical capsaicin site to determine the extent of desensitization produced by the topical capsaicin. 2.2. Topical capsaicin to produce desensitization We wished to identify two locations on the volar forearm for the treatment sites that had similar initial heat pain properties. Two pairs of test sites were marked on the volar forearm of the non-dominant arm about halfway between the wrist and the elbow. The sites that composed a pair were separated by 3 cm. Heat pain thresholds and pain ratings to a suprathreshold heat stimulus (see below) were measured at each site and the pair of sites with the most similar pain thresholds and ratings was chosen for treatment. Once the test sites were selected, additional sensory measures were performed within the 2 £ 2 cm treatment area surrounding the test sites. As described in more detail below, measurements were made of touch detection threshold to von Frey probes, sharpness and pain incidence to a series of weighted needles, pain to a standardized pinch stimulus, cold threshold, heat pain threshold, and heat pain ratings to a suprathreshold stimulus. Following baseline sensory testing, capsaicin and vehicle were applied topically to the test sites. A 2 £ 2 cm cotton gauze was loaded with 0.7 g of a capsaicin cream (consisting of 10% capsaicin dissolved in ethoxydiglycol that was incorporated into a base composed of acetyl alcohol, stearic acid and fatty acid ester) and placed on one of the test sites. A second gauze loaded with a cream that contained just the base was applied to the other test site. There was a 1 cm gap between the two treatment areas. The proximal/distal location of the capsaicin treatment area was randomized. Dermicel tape was used to mask the boundaries of the treatment area and prevent the spread of capsaicin or vehicle. Once the capsaicin and vehicle were applied, the treatment areas were covered with an occlusive dressing for 6 h. At the end of the
P.N. Fuchs et al. / Pain 84 (2000) 141±149
6-h treatment period, the treatment areas were cleaned with alcohol. On the second day, the drugs were applied for an additional 6-h period to the same locations. 2.3. Measurements of desensitization Sensory testing was performed before treatment to con®rm that the test sites were comparable and then on day 5 to measure the extent of desensitization produced by the topical capsaicin. The capsaicin and the vehicle-treated areas were tested for the sensations of heat pain threshold, cold threshold, heat pain ratings to a suprathreshold heat stimulus, touch detection thresholds, sharpness detection and pinch pain ratings using the methods described below. Subjects were provided with two or three training stimuli on day 1 to familiarize them with the measurement techniques. 2.3.1. Thermal testing Thermal testing was performed using a computer controlled Peltier device with an 8 £ 8-mm contact surface (Thermal Devices Inc., Model LTS3). The force of the Peltier head contact with the skin surface was controlled by applying a ®xed weight to the Peltier head that was balanced just above the skin surface. Since the skin surface temperature lagged behind the control temperature of the Peltier system, all temperatures cited in the text were adjusted to the skin-surface calibration temperature (Tillman et al., 1995). 2.3.1.1. Heat pain threshold and cold threshold. The threshold for heat pain was determined using ramped heat stimuli. At the beginning of the heat pain threshold test, the temperature was stepped up from the 308C holding temperature to a baseline temperature of 358C for 3 s. Then the heat stimulus was increased at 0.858C/s. Subjects were instructed to push a button when the stimulus was perceived as painful. The computer automatically returned the temperature to the holding temperature of 308C when the button was pushed or when the cutoff temperature of 51.58C was reached. Heat-pain threshold was de®ned as the temperature of the stimulus at the time that the button was pushed. No correction was made for reaction time artifact. If a subject failed to reach a threshold before the cutoff temperature, then 51.58C was recorded as the heat pain threshold. The threshold for cold detection was determined in a similar manner, except that the baseline temperature started at 308C, and decreased at a rate of 0.858C/s to a 58C cutoff temperature. 2.3.1.2. Pain ratings to suprathreshold heat. Subjects used the technique of magnitude estimation to rate the intensity of pain to a suprathreshold heat stimulus. Subjects were free to choose an arbitrary number to rate the intensity of the ®rst stimulus and then they were instructed to use a ratio scale to rate the intensity of subsequent stimuli. For example, if the ®rst stimulus was
143
given a value of 10, and the subsequent stimulus was twice as painful, the rating for the second stimulus should be 20. The suprathreshold stimulus consisted of a 388C, 3 s baseline followed by a step to 508C for 4 s. 2.3.2. Mechanical testing 2.3.2.1. Touch detection thresholds. Touch detection thresholds were determined using the up/down method of Dixon (1980) with six von Frey mono®laments that were calibrated to administer a force of 0.1, 0.5, 0.9, 3.2, 6.1 or 8.0 mN. Starting with 0.5 mN, the von Frey mono®lament was applied for approximately 1 s within the drug treatment area. If the subject failed to detect the stimulus, then the next higher force von Frey mono®lament was applied. When the subject detected the presence of the stimulus, the next lower von Frey was administered. The up/down test sequence continued for four additional trials after the initial detection. The 50% mechanical detection threshold was calculated using the procedure described in Dixon (1980). 2.3.2.2. Pain ratings to a pinch stimulus. Subjects used magnitude estimation techniques to report the pain intensity to a standardized pinch stimulus. The pinch was performed using an arterial clip. A small piece of skin was grasped with the arterial clip, and clip was left in place for 5 s. At the end of the 5 s, subjects rated the intensity of the pain produced by the arterial clip. The stimulus was applied four times within each treatment area. 2.3.2.3. Sharpness and pain to needle probes. Sharpness detection was determined using a weighted needle device (Chan et al., 1992). The tip of a 30-gauge needle (200 mm diameter) was ®led to produce a ¯at, cylindrical end. A cotton tip applicator was inserted into the Luer connection of the needle, and washers were placed on the shaft of the cotton tip applicator to achieve the desired force level for the stimulus. The entire assembly was then placed inside a 5 ml syringe so that the needle came out of the tip of the syringe and the assembly moved freely within the syringe. When the needle was applied to the skin surface, a reliable and consistent force was applied. Three forces were used: 100, 200 and 400 mN. Each stimulus was applied for about 1 s. Each force was applied 10 times within each of the treatment areas in a pseudo random order. We (and others, e.g. Greenspan and McGillis, 1991) have found in previous psychophysical studies that subjects distinguish sensations of sharpness and pain, and that sharpness is reported at stimulus intensities less than those that evoke pain. Accordingly, the subjects were instructed to indicate if the stimulus was sharp. If a stimulus was sharp, the subject then indicated if the stimulus was painful. 2.3.3. Axon re¯exive ¯are To assess whether the topical capsaicin treatment altered axon-re¯exive ¯are, two laser Doppler probes (Laser¯o,
144
P.N. Fuchs et al. / Pain 84 (2000) 141±149
Model PF 4001, 780 nm, 0.87 mW max; Probe: PF 308, 0.25-mm ®ber separation, TSI Inc., St. Paul, MN) were positioned to record skin blood ¯ow at the center of the two treatment areas. Doppler measurements were started 2-min before the capsaicin injection on day 4 (see below) and continued for 10 min after the injection. The size of the zone of visible ¯are was marked on the skin at 10 min after the capsaicin injection. 2.4. Intradermal capsaicin to produce secondary hyperalgesia On the 4th day, capsaicin (50 mg in a 10 ml volume; Ali et al., 1996) was injected intradermally midway between the two treatment areas to produce a large zone of secondary hyperalgesia that encompassed the two treatment areas. The capsaicin was prepared in a Tween 80 saline vehicle using the method described by LaMotte et al. (1991). Subjects rated their pain to the capsaicin injection for 10 min. At this time, the area of the bleb and the visible ¯are was marked. At 20 min after the injection, pain to computercontrolled force stimuli was measured (see below). Then, the zone of secondary hyperalgesia was mapped using an 200 mN von Frey mono®lament. The entire test sequence was repeated at 60 min after the injection. At the completion of the testing, the areas of drug treatment and markings for ¯are and borders of secondary punctate hyperalgesia were transferred to an acetate sheet. 2.5. Measurements of secondary hyperalgesia Two different techniques were used to determine whether the secondary hyperalgesia that developed at the desensitized site was different from that which developed at the vehicle treated site. Pain ratings to a computer-controlled constant-force mechanical stimulus were obtained for both treatment locations before and after the capsaicin injection. In addition, the zone of secondary hyperalgesia was mapped. 2.5.1. Pain ratings to a force controlled blade stimulus A computer-controlled mechanical stimulator was used to deliver force-controlled stimuli to predetermine locations on the volar forearm (Schneider et al., 1995). The stimulator consists of a servo-controlled linear motor capable of generating 1 kg of force over a 22-mm range. The motor is attached to a three-axis computer-controlled translation system that allows the probe to be positioned accurately at different locations on the volar forearm. The stimuli in this study were 34 or 64 g for 1 s duration and were applied to the skin using a 12 mm wide, 100 mm thick blade. These stimuli have been shown to activate cutaneous nociceptors in the primate (Slugg et al., 1995) and to produce pain in the zone of secondary hyperalgesia (Huang et al., 1997). Eight test sites were located at 1 cm intervals along a line that passed through the center of the treatment locations and the injection site (Fig. 1). This resulted in two mechanical test
Fig. 1. Schematic illustration of experimental paradigm. A 2 £ 2 cm area on the volar forearm was pretreated with capsaicin to produce desensitization. A second vehicle treated area served as the control. Two days later, intradermal injection of capsaicin between the two treatment sites produced a large zone of secondary hyperalgesia to mechanical stimuli that extended beyond the treatment zones. A computer-controlled mechanical stimulator system was used to apply controlled force stimuli via a blade probe to the sites indicated by the vertical lines. Subjects rated their pain to these stimuli using magnitude estimation techniques.
sites that were within each treatment location as well as two sites outside of each treatment location. No testing was done at the capsaicin injection site. The order of stimulus presentation (both for force and position) was randomized except for the constraints that (1) each location received two presentations of the 64 g stimulus and one presentation of the 32 g stimulus; and (2) two consecutive stimuli were not presented within 1 cm of each other. Within each test site location, the three stimulus presentations were separated by 1 mm to minimize any possible fatigue due to a previous mechanical stimulus presentation (Slugg et al., 1995). Each mechanical test sequence consisted of 24 stimulus presentations with a 20 s interstimulus interval. Subjects rated the pain to each stimulus using the technique of magnitude estimation. 2.5.2. Mapping of zone of secondary hyperalgesia To map the zone of secondary hyperalgesia, a 200 mN von Frey was applied well outside the expected area of hyperalgesia and then at 1 cm intervals along a radial path towards the injection site. The subjects were instructed to indicate when they felt that the stimulus became painful or if, originally painful, became notably more painful. The border of the secondary zone was con®rmed by retesting near the edge of the zone and then indicated on the skin with a marker pen. This procedure was repeated until a minimum of eight radial paths had been completed. 2.6. Data analysis 2.6.1. Normalization of data Subjects were free to use numbers of their own choice to rate the intensity of pain. Since a wide range of numbers were used, we needed to normalize the data in order to combine data across subjects. Data were normalized for a given subject by dividing that subject's ratings to a given stimulus by that subject's maximum rating to that stimulus.
P.N. Fuchs et al. / Pain 84 (2000) 141±149
145
Changes in laser Doppler measure of perfusion were normalized by dividing the average perfusion during the last 5-min of the 10 min post capsaicin injection period by the average perfusion during the 2 min before the capsaicin injection. 2.6.2. Analysis Data were analyzed using either Wilcoxon matched paired test, paired two-tailed t-tests or one-way analysis of variance (ANOVA). Signi®cance was P , 0:05. 3. Results Ten male subjects with age ranging from 23±42 years (mean 32 year) were enrolled in the study. One subject developed a very small zone of secondary hyperalgesia which extended less than 1.5 cm in the vehicle treated direction. This subject was excluded from subsequent analysis. All other subjects had a zone of punctate hyperalgesia that extended at least 3.5 cm in the vehicle treated direction. 3.1. Measurements of desensitization Sensory testing performed before treatment con®rmed that the test sites were comparable. Pre-treatment values for heat pain threshold, cold threshold, heat pain ratings, touch detection thresholds, sharpness detection thresholds, and pinch pain ratings at the capsaicin and the vehicle treatment locations were not signi®cantly different. 3.1.1. Thermal testing 3.1.1.1. Heat pain threshold. The median heat pain threshold at the capsaicin treated locations was signi®cantly higher than the heat pain threshold at the vehicle treated locations (P , 0:01, Wilcoxon matched pairs, Fig. 2a). All subjects had heat pain thresholds less than 498C at the vehicle treated area. In most subjects (8/9), the heat pain threshold at the capsaicin treated area was higher than the maximum temperature used (i.e. 51.58C). 3.1.1.2. Pain to suprathreshold heat stimuli. Topical capsaicin treatment also resulted in a signi®cant decrease in pain ratings to the suprathreshold stimulus (508C, 4 s) at the capsaicin treatment location compared to the vehicle treatment location (P , 0:05, Fig. 2B). At the capsaicin treatment location, eight of the nine subjects reported no pain to the suprathreshold heat stimulus. In contrast, only two subjects reported no pain to this stimulus at the vehicle treatment location. 3.1.1.3. Cold threshold. The mean cold detection threshold at the capsaicin treatment location (22:0 ^ 1:18C) was signi®cantly lower than at the vehicle treatment location (25:0 ^ 1:08C, P , 0:05). However, all
Fig. 2. Desensitization following treatment with topical capsaicin. (A) Heat pain thresholds were signi®cantly higher at the capsaicin treatment site than at the vehicle treatment site. Heat pain threshold was determined using a 0.858C/s ramp from 358C to 51.58C (median ^ 25%). (B) Heat pain ratings to the suprathreshold stimulus were signi®cantly lower at the capsaicin treatment site. Subjects rated the intensity of pain to a 508C, 4 s suprathreshold heat stimulus (mean ^ SEM). (C) Pain ratings to a pinch stimulus were not altered by the capsaicin treatment. Subjects rated the intensity of pain to an arterial clip applied to a fold of skin for 5 s (mean ^ SEM) (* P , 0:05; ** P , 0:01; Wilcoxon matched pairs).
subjects detected cold prior to the 58C cutoff temperature at both the capsaicin and vehicle treated test sites. 3.1.2. Mechanical testing 3.1.2.1. Touch detection thresholds and pain ratings to the pinch stimulus. The touch detection thresholds at the capsaicin (1:8 ^ 0:4 mN) and vehicle (2:1 ^ 0:3 mN) treated areas were not signi®cantly different (P . 0:40). Similarly, the pain ratings to the pinch stimulus were not signi®cantly different at the capsaicin and vehicle treated areas (P . 0:20, Fig. 2C). 3.1.2.2. Sharpness and pain to needle probes.
In response
146
P.N. Fuchs et al. / Pain 84 (2000) 141±149
to the needle probes, the incidence of sharpness detection and the incidence of pain increased monotonically with the applied force at both the vehicle and capsaicin treated sites. However, the incidence of sharpness detection tended to be lower on the capsaicin treated skin than on vehicle treated skin. Although pain was still produced by the punctate probes, the incidence of pain reports was signi®cantly lower in capsaicin treated skin. For example, the 400 mN probe produced a signi®cantly higher incidence of pain in vehicle treated skin (53 ^ 10%) than in capsaicin treated skin (27 ^ 10%, P , 0:01), whereas the incidence of sharpness in the vehicle treated skin (91 ^ 5%) was not signi®cantly different from capsaicin treated skin (72 ^ 12%, P . 0:1). 3.1.3. Axon re¯exive ¯are The capsaicin injection between the two treatment areas resulted in the appearance of a visible ¯are that always extended into the vehicle treatment area but never extended into the capsaicin treatment area. In all subjects, a sharp edge to the ¯are was visualized that corresponded to the border of the capsaicin treatment area. The area of ¯are on the vehicle-treated side of the capsaicin injection (7:7 ^ 1:5 cm 2) was signi®cantly larger than the area on the capsaicintreated side (2:6 ^ 1:3 cm 2, P , 0:01). The laser Doppler measurements of ¯are revealed a 7:1 ^ 1:7-fold increase in perfusion at the vehicle site but only a 2:8 ^ 0:4-fold increase at the capsaicin treatment site (P , 0:05).
Fig. 3. Secondary hyperalgesia persists in the capsaicin desensitized area. Normalized pain ratings to the 64 g stimulus applied via the blade probe are plotted as function of time after the capsaicin injection. A signi®cant increase in pain ratings was observed at both the capsaicin desensitized area and the vehicle treatment area after the intradermal injection of capsaicin. Data for a given subject were averaged across the measurements within each treatment site (mean ^ SEM; ** P , 0:01, post vs. pre; D P , 0:05, capsaicin vs. vehicle).
on the vehicle treatment side of the capsaicin injection (21:7 ^ 2:5 cm 2 at 20 min, 19:7 ^ 3:7 cm 2 at 60 min) was not signi®cantly different from the area on the capsaicin treatment side (16:7 ^ 3:9 cm 2; 18:2 ^ 6:2 cm 2; P . 0:5).
3.2. Pain to capsaicin injection
4. Discussion
Intradermal injection of capsaicin produced an intense sensation of burning pain that started immediately with the injection. The pain reached a peak within the ®rst 5± 10 s and then gradually decreased during the next 10 min. The rate of decrease was greatest during the ®rst 5 min following the injection.
These experiments demonstrate that secondary hyperalgesia to noxious mechanical stimuli is little affected in capsaicin desensitized skin. The interpretation of this ®nding centers on an interpretation of what capsaicin does to cutaneous afferents. There are three sources of data that contribute to this understanding: Neurophysiological, histological, and psychophysical.
3.3. Measurements of secondary hyperalgesia 3.3.1. Pain to the computer-controlled mechanical stimulus The intradermal injection of capsaicin led to a pronounced mechanical hyperalgesia in both the vehicle and capsaicin treated areas (Fig. 3). The blade probe produced little pain before the capsaicin injection. Pain ratings were signi®cantly higher at the capsaicin and vehicle treatment locations after the capsaicin injection (P , 0:01). Thus, secondary hyperalgesia to punctate stimuli persisted in the capsaicin desensitized skin. Although the pain ratings in the capsaicin desensitized skin were signi®cantly lower than in the vehicle treated skin (P , 0:05) at the 20 min time point, this difference disappeared at the 60 min time point (Fig. 3). 3.3.2. Area of punctate hyperalgesia The area of punctate hyperalgesia was measured with a 200 mN von Frey probe. The area of punctate hyperalgesia
4.1. What nociceptive afferents survive capsaicin treatment From a neurophysiological perspective capsaicin appears to excite initially then deactivate nociceptors through a receptor mediated effect on a cation channel. Baumann et al. (1991) found that local injection of capsaicin induced a deactivation of C-®ber nociceptors to both heat and mechanical stimuli. This deactivation was restricted to that area in the receptive ®eld where the injection was administered. Other studies demonstrate that capsaicin opens a cation channel permeable to Na 1 and Ca 21 (Docherty et al., 1991; Bevan and Docherty, 1993). The mechanism of neurotoxicity of capsaicin is not clearly understood. However, evidence points to effects of excessive intracellular calcium and induction of proteases (Chard et al., 1995). Histological studies by Simone and collegues (1998) determined that even at doses as low as 2 mg capsaicin injected intradermally caused a localized loss of all intra-epidermal
P.N. Fuchs et al. / Pain 84 (2000) 141±149
®bers as determined with the pan-axonal marker PGP 9.5. Topical capsaicin was found in a separate study to have similar effects (Nolano et al., 1999). In the psychophysical study presented here capsaicin created a dissociated sensory loss, as manifest by heat analgesia, but only a small effect on pain to mechanical stimuli. At the capsaicin treatment location, eight of the nine subjects reported no pain to the supra-threshold heat stimulus (508C, 4 s). Others have reported similar ®ndings (e.g. SzolcsaÂnyi, 1977; Carpenter and Lynn, 1981; Jancso et al., 1985; Bjerring and Arendt-Nielsen, 1990; Simone and Ochoa, 1991; Davis et al., 1995; Beydoun et al., 1996; Simone et al., 1998; Fuchs et al., 1999). Yet, effects on pain to mechanical stimuli were more subtle. The pain to the punctate probe was signi®cantly less in the capsaicin treated skin, but the incidence of reports of pain only decreased by half to the highest force used (400 mN). The difference in reports of sharpness to a needle probe tended to be less in the capsaicin treated region, but this difference did not reach statistical signi®cance. There was no effect on pain to the pinch stimulus. In addition, there was no measurable effect on tactile sensibility. Others have also reported that pain to mechanical stimuli is only modestly (if at all) affected in capsaicin treated skin (e.g. Simone and Ochoa, 1991; Davis et al., 1995; Beydoun et al., 1996; Magerl et al., 1998; Simone et al., 1998). Our ®nding that the threshold for cold sensation decreased signi®cantly is consistent with the recent observation that the magnitude of cold sensation decreases after topical capsaicin treatment (Nolano et al., 1999). Topical application of capsaicin also leads to a decrease in itch sensitivity (ToÂth-KaÂsa et al., 1986; Handwerker et al., 1987; Simone and Ochoa, 1991) as well as a decreased response to bradykinin and capsaicin (Crimi et al., 1992) indicating that chemically sensitive afferents are also affected. In addition there is a marked decrease in axon re¯exive ¯are (e.g. Jancso et al., 1968; Carpenter and Lynn, 1981; SzolcsaÂnyi, 1988; Bjerring and Arendt-Nielsen, 1990). Sympathetic efferent ®bers appear not to be affected as evidenced by the lack of change in nicotine-induced axon re¯ex sweating (Izumi and Karita, 1988) in capsaicin treated skin. Given that the sensory changes are correlated with the striking loss of epidermal and even dermal ®bers (Nolano et al., 1999), the loss of heat sensibility is likely due to degeneration of C-®ber and A-®ber nociceptors responsive to heat stimuli. That pain to mechanical stimuli persists after capsaicin application suggests that a separate class of nociceptors responsive to mechanical stimuli exists that is resistant to the effects of capsaicin. Some have argued that the `capsaicin channel' is in effect a heat-transduction channel (e.g. Caterina et al., 1997). Thus nociceptors unresponsive to heat may, in effect, be protected from the toxic effects of capsaicin. The neurophysiological, histological, and psychophysical data taken together argue that heat sensitive nociceptors undergo local degeneration and that the persistent pain to mechanical stimuli arises from the input of
147
mechanically sensitive and heat insensitive nociceptive afferents. In a related study Magerl et al. (1998) undertook to determine the relative role of C- and A-®ber nociceptors as mediators of pain to mechanical stimuli in capsaicin desensitized skin. As in this study, capsaicin had only modest effects on pain to mechanical stimuli. However, a selective A-®ber block obtained by application of pressure over the super®cial radial nerve eliminated the remaining pain sensibility to mechanical stimuli. In another experiment, Ziegler et al. (1999) demonstrated that secondary hyperalgesia to punctate stimuli following capsaicin injection was not present when A-®ber conduction was blocked. These data support the conclusion that a class of A-®ber nociceptors insensitive to heat accounts for the pain to mechanical stimuli in capsaicin treated skin. 4.2. A-®ber nociceptors A-®ber nociceptors may be divided into three categories with regard to the heat response (Treede et al., 1998). Type I ®bers typically have an accelerating response to heat with a long receptor utilization period (time between activation and stimulus onset). The heat threshold of these ®bers for short (1±3 s) duration stimuli is high, often exceeding 538C. Because the heat tests in this study (and in the study by Nolano et al., 1999) would not be suf®cient to activate the majority of these ®bers, a role of type I A-®ber nociceptors in serving punctate hyperalgesia cannot be excluded. Type II A-®ber nociceptors have an adapting response to heat (similar to that of polymodal C-®ber nociceptors), with a short receptor utilization period. These ®bers serve ®rst pain sensation, a perception that is lost in capsaicin desensitized skin (Beydoun et al., 1996). Thus, these ®bers are unlikely to play a role in punctate hyperalgesia. A third variety of A®ber nociceptors respond to mechanical stimuli, but have no response to heat even at high intensities (Treede et al., 1998). These ®bers are leading candidates to serve punctate hyperalgesia. Little is known about the responsiveness of these ®bers to chemical stimuli. If indeed these ®bers are responsive to chemical stimuli, hyperalgesia to chemical stimuli in the secondary zone would be predicted. This has as yet not been tested. Though there was some change in painfulness of pinprick stimuli following the topical capsaicin administration, the decrease in the sensation of sharpness did not reach signi®cance. This would argue that sharpness may qualify as a separate sensation, and moreover that sharpness too, is served by A-®ber nociceptors (Garell et al., 1996). 4.3. Neural mechanisms of punctate hyperalgesia Two models to account for punctate hyperalgesia were presented in the Section 1. We assume, based on the evidence presented, that punctate hyperalgesia arises from a central sensitization to the input of nociceptive afferents. Moreover, the model must account for the absence of heat
148
P.N. Fuchs et al. / Pain 84 (2000) 141±149
hyperalgesia in the secondary zone. Now we have further data on which to base the model. The nociceptive afferents that provide the peripheral signals for punctate hyperalgesia survive capsaicin treatment. The evidence rules out the disinhibition model proposed in Section 1 and points to a class of A-®bers that is unresponsive to heat stimuli. Dougherty et al. (1998) studied the responses of primate spinothalamic projection cells that innervated areas of skin adjacent to a capsaicin injection. These cells acutely showed the expected sensitization to mechanical stimuli. The response to short duration heat stimuli was actually attenuated. This neurophysiological ®nding is in keeping with psychophysical ®ndings of Raja et al. (1984) who determined that shortly after a burn injury there was coexistent mechanical hyperalgesia and heat hypalgesia. Thus, the CNS determinants of the dissociated hyperalgesia to mechanical versus heat stimuli are already in place at the level of the dorsal horn spinothalamic projection cells. Mechanisms within the dorsal horn, therefore, likely account for punctate hyperalgesia. Four separate models of dorsal horn circuitry could explain these ®ndings.
tors with no heat sensitivity (or very high thresholds) encodes the punctate hyperalgesia that occurs in the zone of secondary hyperalgesia: (1) punctate hyperalgesia persists in capsaicin desensitized skin; (2) capsaicin appears to destroy afferent terminals in ®bers responsive to heat stimuli; (3) remaining pain-sensitivity to mechanical stimuli in capsaicin treated skin is eliminated by an A-®ber block; and (4) A-®ber nociceptors with selective sensitivity to mechanical stimuli exist. Future consideration of the differential dorsal horn connections of different types of nociceptors, anatomically, chemically, and physiologically, may illuminate further the mechanisms of punctate hyperalgesia.
1. Pre-synaptic regulation: The mechano-sensitive A-®ber nociceptors project to spinothalamic projection cells that receive convergent input from polymodal nociceptors. Collaterals from polymodal nociceptors that serve the injury zone have presynaptic connections with the central terminals of A-®ber nociceptors. This leads to enhanced neurotransmitter release through a mechanism such as primary afferent hyperpolarization. These collateral inputs do not affect the transmission of action potentials from polymodal nociceptors. Thus hyperalgesia is present only for mechanical stimuli. 2. Interneuron speci®city: The model here is similar except that the enhancement of input of the mechano-sensitive A-®ber nociceptors is postulated to occur as a result of sensitization of an interneuron that receives the A-®ber input. Polymodal nociceptors could sensitize these neurons through a post-synaptic mechanism. 3. Neurotransmitter speci®city: If polymodal and mechanospeci®c nociceptors utilize different transmitters at their junctions with spinothalamic projection neurons then selective sensitization could result from this difference. 4. Dendritic speci®city: Conceivably different parts of the dendritic tree of spinothalamic neurons could be specialized to receive the inputs of different afferents. If so, these post-synaptic regions could be selectively sensitized.
References
Intracellular recordings of dorsal horn cells could help resolve these models. 4.4. Conclusion The evidence presented here, when considered with other data, favors the hypothesis that a class of A-®ber nocicep-
Acknowledgements We wish to thank Mr. Timothy Hartke and Ms. Sylvia Horasek for their technical support, Dr. Gang Wu for reviewing the manuscript, and Dr. Sam Georgiou for preparing the capsaicin formulation. This research was supported by the NIH on grant NS-14447.
Ali Z, Meyer RA, Campbell JN. Secondary hyperalgesia to mechanical but not heat stimuli following a capsaicin injection in hairy skin. Pain 1996;68:401±411. Baumann TK, Simone DA, Shain CN, LaMotte RH. Neurogenic hyperalgesia: the search for the primary cutaneous afferent ®bers that contribute to capsaicin-induced pain and hyperalgesia. J Neurophysiol 1991;66:212±227. Bevan SJ, Docherty RJ. Cellular mechanisms of the action of capsaicin. In: Wood J, editor. Capsaicin in the Study of Pain, London: Academic Press, 1993. pp. 27±44. Beydoun A, Dyke DBS, Morrow TJ, Casey KL. Topical capsaicin selectively attenuates heat pain and A d ®ber-mediated laser-evoked potentials. Pain 1996;65:189±196. Bjerring P, Arendt-Nielsen L. Inhibition of histamine skin ¯are reaction following repeated topical applications of capsaicin. Allergy 1990;45:121±125. Campbell JN, Khan AA, Meyer RA, Raja SN. Response to heat of C-®ber nociceptors in monkey are altered by injury in the receptive ®eld but not by adjacent injury. Pain 1988;32:327±332. Carpenter SE, Lynn B. Vascular and sensory responses of human skin to mild injury after topical treatment with capsaicin. Br J Pharmacol 1981;73:755±758. Caterina MJ, Schumacher MA, Tominaga M, Rosen TA, Levine JD, Julius D. The capsaicin receptor: a heat-activated ion channel in the pain pathway. Nature 1997;389:816±824. Cervero F, Meyer RA, Campbell JN. A psychophysical study of secondary hyperalgesia: Evidence for increased pain to input from nociceptors. Pain 1994;58:21±28. Chan AW, MacFarlane IA, Bowsher D, Campbell JA. Weighted needle pinprick sensory thresholds: a simple test of sensory function in diabetic peripheral neuropathy. J Neurol Neurosurg Psychiat 1992;55:56±59. Chard PS, Bleakman D, Savidge JR, Miller RJ. Capsaicin-induced neurotoxicity in cultured dorsal root ganglion neurons: involvement of calcium-activated protease. Neuroscience 1995;65:1099±1108. Crimi N, Polosa R, Maccarrone C, Palermo B, Palermo F, Mistretta A. Effect of topical application with capsaicin on skin responses to bradykinin and histamine in man. Clin Exp Allergy 1992;22:933±939. Dahl JB, Brennum J, Arendt-Nielsen L, Jensen TS, Kehlet H. The effects of
P.N. Fuchs et al. / Pain 84 (2000) 141±149 pre- versus postinjury in®ltration with lidocaine on thermal and mechanical hyperalgesia after heat injury to the skin. Pain 1993;53:43±51. Davis KD, Meyer RA, Turnquist JL, Filloon TG, Pappagallo M, Campbell JN. Cutaneous injection of the capsaicin analogue. NE-21610, produces analgesia to heat but not to mechanical stimuli in man. Pain 1995;61:17±26. Dixon WJ. Ef®cient analysis of experimental observations. Annu Rev Phamacol Toxicol 1980;20:441±462. Docherty RJ, Robertson B, Bevan S. Capsaicin causes prolonged inhibition of voltage-activated calcium currents in adult rat dorsal root ganglion neurons in culture. Neuroscience 1991;40:513±521. Dougherty PM, Willis WD, Lenz FA. Transient inhibition of responses to thermal stimuli of spinal sensory tract neurons in monkeys during sensitization by intradermal capsaicin. Pain 1998;77:129±136. Fuchs PN, Pappagallo M, Meyer RA. Topical EMLA pre-treatment fails to decrease the pain induced by 1% topical capsaicin. Pain 1999;80:637± 642. Garell PC, McGillis SLB, Greenspan JD. Mechanical response properties of nociceptors innervating feline hairy skin. J Neurophysiol 1996;75:1177±1189. Garnsworthy RK, Gully RL, Kenins P, May®eld RJ, Westerman RA. Identi®cation of the physical stimulus and the neural basis of fabric-evoked prickle. J Neurophysiol 1988;59:1083±1097. Greenspan JD, McGillis SLB. Stimulus features relevant to the perception of sharpness and mechanically evoked cutaneous pain. Somatosens Motor Res 1991;8:137±147. Handwerker HO, Magerl W, Klemm F, Lang E, Westerman RA. Quantitative evaluation of itch sensation. In: Schmidt RF, Schaible HG, VahleHinz C, editors. Fine Afferent Nerve Fibers and Pain, New York: Weinheim, 1987. pp. 461±473. Hardy JD, Wolff HG, Goodell H. Experimental evidence on the nature of cutaneous hyperalgesia. J Clin Invest 1950;29:115±140. Huang JH, Ali Z, Campbell JN, Meyer RA. Pain ratings to punctate mechanical stimuli are similar through-out the zone of secondary hyperalgesia. Soc Neurosci Abstr 1997;23:1258. Izumi H, Karita K. Investigation of mechanisms of the ¯are and wheal reactions in human skin by band method. Brain Res 1988;449:328±331. Jancso N, JanscoÂ-GaÂbor A, SzolcsaÂnyi J. The role of sensory nerve endings in neurogenic in¯ammation induced in human skin and in the eye and paw of the rat. Br J Pharm Chemother 1968;32:32±41. Jancso G, ObaÂl FJ, ToÂth-KaÂsa I, Katona M, Husz S. The modulation of cutaneous in¯ammatory reactions by peptide-containing sensory nerves. Int J Tissue Reac 1985;7:449±457. LaMotte RH, Shain CN, Simone DA, Tsai E-FP. Neurogenic hyperalgesia: Psychophysical studies of underlying mechanisms. J Neurophys 1991;66:190±211. Lewis T. Experiments relating to cutaneous hyperalgesia and its spread through somatic nerves. Clin Sci 1936;2:373±423. Magerl W, Fuchs PN, Meyer RA, Treede R-D. Pain to punctate mechanical stimuli in humans is mainly mediated by capsaicin-insensitive A-®ber nociceptors. Soc Neurosci Abstr 1998;24:2085.
149
Meyer RA, Campbell JN, Raja SN. Antidromic nerve stimulation in monkey does not sensitize unmyelinated nociceptors to heat. Brain Res 1988a;441:168±172. Meyer RA, Magerl W, Campbell JN, Treede R-D, Fuchs PN. Normal punctate hyperalgesia in capsaicin desensitized skin. Soc Neurosci Abstr 1998b;24:2086. Nolano M, Simone DA, Wendelschafer-Crabb, Johnson T, Hazen E, Kennedy WR. Topical capsaicin in humans: parallel loss of epidermal nerve ®bers and pain sensation. Pain 1999;81:135±145. Raja SN, Campbell JN, Meyer RA. Evidence for different mechanisms of primary and secondary hyperalgesia following heat injury to the glabrous skin. Brain 1984;107:1179±1188. Raja SN, Meyer RA, Ringkamp M, Campbell JN. Peripheral neural mechanisms of nociception. In: Melzack R, Wall PD, editors. Textbook of Pain, 4th edition, Churchill Livingstone, 1999 in press. Schneider W, Slugg RM, Turnquist BP, Meyer RA, Campbell JN. An electromechanical stimulator system for neurophysiological and psychophysical studies of pain. J Neurosci Methods 1995;60:61±68. Simone DA, Ochoa J. Early and late effects of prolonged topical capsaicin on cutaneous sensibility and neurogenic vasodilatation in humans. Pain 1991;47:285±294. Simone DA, Sorkin LS, Oh U, Chung JM, Owens C, LaMotte RH, Willis WD. Neurogenic hyperalgesia: Central neural correlates in responses of spinothalamic tract neurons. J Neurophysiol 1991;66:228±246. Simone DA, Nolano M, Johnson T, Wendelschafer-Crabb G, Kennedy WR. Intradermal injection of capsaicin in humans produces degeneration and subsequent reinnervation of epidermal nerve ®bers: correlation with sensory function. J Neurosci 1998;18:8947±8959. Slugg RM, Meyer RA, Campbell JN. Mechanical receptive ®eld structure for cutaneous C-®ber and A-®ber nociceptors in the monkey determined with graded intensity stimuli. Soc Neurosci Abstr 1995;21:1161. SzolcsaÂnyi J. A pharmacological approach to elucidation of the role of different nerve ®bers and receptor endings in mediation of pain. J Physiol 1977;73:251±259. SzolcsaÂnyi J. Antidromic vasodilation and neurogenic in¯ammation. Agents Actions 1988;23:4±11. Tillman DB, Treede R-D, Meyer RA, Campbell JN. Response of C ®bre nociceptors in the anaesthetized monkey to heat stimuli: Correlation with pain threshold in humans. J Physiol 1995;485:767±774. TorebjoÈrk HE, Lundberg LER, LaMotte RH. Central changes in processing of mechanoreceptive input in capsaicin-induced secondary hyperalgesia in humans. J Physiol 1992;448:765±780. ToÂth-KaÂsa I, Jancso G, BognaÂr A, Husz S, ObaÂl FJ. Capsaicin prevents histamine-induced itching. Int J Clin Pharmacol Res 1986;6:163±169. Treede R-D, Cole JD. Dissociated secondary hyperalgesia in a subject with large ®bre sensory neuropathy. Pain 1993;53:169±174. Treede R-D, Campbell JN, Meyer RA. Myelinated mechanically-insensitive afferents from monkey hairy skin: Heat response properties. J Neurophys 1998;80:1082±1093. Ziegler EA, Magerl W, Meyer RA, Treede R-D. Secondary hyperalgesia to punctate mechanical stimuli: central sensitization to A-®bre nociceptor input. Brain 1999:in press.