Abstract
Background Intradermal capsaicin is a human pain model that produces reliable pain and sensitization. This model facilitates controlled testing of analgesic efficacy via a crossover design while minimizing confounding variables in clinical pain states and retaining sufficient power with small samples.
Methods To determine the lowest dose of capsaicin that produces consistent neurosensory measures, we administered 0.1, 1, 10, or 100 μg to healthy volunteers in a blinded manner (N = 19). Pain scores were recorded at 0, 5, 10, 15, and 60 minutes on a visual analog scale from 0 to 100. Areas and intensities of mechanical allodynia (foam brush stimulus) and pinprick hyperalgesia (von Frey test) were quantified at 15 and 60 minutes, as were flare areas.
Results Capsaicin produced dose-dependent increases in spontaneous pain (p = .013), the area and intensity of mechanical allodynia (p = .006 and p < .001, respectively), the area and intensity of pinprick hyperalgesia (p = .010 and p = .014, respectively), and the flare area (p = .010). The 10 μg dose produced greater spontaneous pain than the 1 μg dose (p = .017). The 100 μg dose produced greater spontaneous pain than the 10 μg, but the difference was not statistically significant.
Conclusion The 10 and 100 μg capsaicin doses produced robust pain measures across a range of modalities, and lower doses produced minimal effects. Whereas most studies use 100 μg, using a lower dose is reasonable and may facilitate detection of subtle analgesic effects—particularly with nonopioid analgesics—and drugs can be tested in lower doses, minimizing adverse side effects.
The use of the clinical pain state has evident limitations in evaluating analgesic interactions. For instance, pain following inadvertent or iatrogenic trauma is often multifactorial (eg, tissue and nerve injury and inflammation), treatment regimens involve multiple medications, and controlled crossover interventions typically cannot be conducted on the same subject. For this reason, human experimental models have been used that permit the ethical and well-controlled study of the response of the human volunteer to experimental pain stimuli. One such model is the intradermal injection of capsaicin that results in a transient report of an intense stinging sensation at the site of injection followed by a persisting tactile hypersensitivity in the area immediately surrounding the injection site (ie, secondary allodynia and hyperalgesia).
Intradermal capsaicin has been used extensively to study human pain, producing reliable pain and sensitization, allowing analysis of the constituents of various pain states. For instance, capsaicin-induced spontaneous pain corresponds to activity in C mechanoheat nociceptors via binding to vanilloid type 1 receptors.1-3In contrast, pinprick hyperalgesia (increased pain evoked by a previously painful stimulus) is mediated by Aδ and C afferent fibers, and mechanical allodynia (pain evoked by a previously nonpainful stimulus) is mediated primarily by Aβ fibers.4-6Because Aδ fibers normally convey only nonpainful sensations (ie, they are low-threshold fibers), it is thought that the discomfort caused by a stimulus that is normally nonpainful (ie, allodynia) is due to an alteration in the central processing of the pain input in the spinal cord.6This altered central processing is believed to underlie many clinical pain states (eg, postherpetic neuralgia, complex regional pain syndrome) and is therefore a target for novel analgesic agents.7-10Capsaicin-induced allodynia may thus be used as a potential “biomarker” of central sensitization.11
Numerous analgesic agents have been shown to significantly ameliorate capsaicin-mediated effects, including intrathecal adenosine,12intrathecal clonidine,13intravenous alfentanil,6,14,15intravenous ketamine,6,15intravenous infusion of a novel α-amino-3-hydroxy-5-methyl-4-isoxazoleproprionic acid/kainate antagonist,16oral gabapentin,17and subcutaneous lidocaine.18However, the painful stimulus from 100 μg of capsaicin is often resistant to many agents used clinically, including intramuscular amitriptyline,14intravenous clonidine,13intravenous lidocaine,19oral desipramine,20oral lamotrigine,21oral mexiletine,22and subcutaneous ketamine.18
One controversy of using intradermal capsaicin as a human pain model is that the stimulus from 100 μg may be too intense to detect certain drug effects, which may explain the minimal effect of some analgesic agents with this model. To determine the lowest dose of capsaicin to produce consistent neurosensory measures, we compared four doses of capsaicin to quantify spontaneous pain, the area and intensity of mechanical allodynia (foam brush stimulus), the area and intensity of pinprick hyperalgesia (von Frey test), and flare response in healthy human volunteers. It was hypothesized that each of these outcomes would be dependent on the dose of capsaicin administered. All outcomes were ascertained at fixed time intervals, and visual analog scale (VAS) pain scores were used to facilitate extrapolation of these findings to future analgesic efficacy studies using intradermal capsaicin as a pain model.
MATERIALS AND METHODS
Subjects
The study protocol was approved by the Institutional Review Board of the University of California, San Diego. Nineteen healthy subjects (13 women and 6 men) were recruited. Their average age was 28 years (range 19-53 years), and their average weight was 65 kg (range 44-86 kg). After receiving a thorough explanation of the study, subjects provided written informed consent and were enrolled. Exclusion criteria included pregnancy, current painful condition, current use of analgesics, or lack of ability to understand the experimental protocol. All subjects were blinded to the capsaicin dose received.
Intradermal Capsaicin Injection
Capsaicin (8-methyl-N-vanillyl-6-nonamide) was prepared in four concentrations, each dissolved in a normal saline solution with a 20% cyclodextran vehicle. Concentrations were 10, 1, 0.1, and 0.01 mg/mL. A volume of 10 μL was injected intradermally into the volar aspect of the nondominant forearm with a sterile tuberculin syringe, giving doses of 100 μg (n = 7), 10 μg (n = 4), 1.0 μg (n = 5), and 0.1 μg (n = 3).
Spontaneous and Elicited Pain Scores
To quantify the intensity of ongoing pain, spontaneous pain scores were recorded at 0, 5, 10, 15, and 60 minutes after capsaicin injection on a VAS: a 100 mm continuum with “no pain” at the 0 mark and “worst imaginable pain” at the 100 mark. Subjects moved a marker from 0 to the spot corresponding to their pain level (eg, a mark at 60 mm would be scored as 60). Subjects did not see the numeric correlates to their pain perceptions at any point during the study.
To quantify the intensity of elicited pain to a brush and to pinprick (ie, a von Frey test), elicited pain scores were recorded at 15 and 60 minutes. First, the hyperalgesic region was gently stroked with a foam brush five times (one stroke per second), with pain rated on the fifth stroke via the VAS. Then a 5.18 mN von Frey hair was bounced in the hyperalgesic region five times (one bounce per second), and pain was rated on the fifth bounce on the VAS.
Areas of Mechanical Allodynia, Pinprick Hyperalgesia, and Flare
The areas of mechanical allodynia (to a brush), pinprick hyperalgesia (to a von Frey hair), and flare were mapped at 15 and 60 minutes after capsaicin administration. For mechanical allodynia, the stimulus was initiated in an area of skin that did not produce pain (ie, sufficiently distant from the initial injection site) and the brush was swept tangentially toward the injection site at a constant intensity and a rate of 1 cm/s. With eyes closed, the subjects would indicate when the brush first began to cause a change in sensation (ie, pain, tenderness, or discomfort). To ensure accurate results, this method was demonstrated until each subject felt comfortable identifying the onset of allodynia (typically one trial).
The point of contact between the leading edge of the brush and the subject's skin was marked with a felt-tipped pen (a different color for each time interval). This process was repeated at a different angle until at least eight determinations of the borders of allodynia were outlined on the skin.
Analogous to the foam brush methodology, a standard 5.18 von Frey hair was applied at a rate of 1 cm/s beginning well outside the sensitized region and progressing toward the injection site. Subjects were asked to indicate when they sensed an increase in pain, and the site was marked with a felt-tipped pen of a different color. As with the brush stimulus, a trial was performed to ensure that the subjects could clearly identify the onset of hyperalgesia. This was repeated until at least eight points were outlined on the skin.
To demarcate the area of flare, a transparency was placed on the volar aspect of the subject's forearm and held with the experimenter's hand while the subject's forearm was completely extended at the elbow (this method provided optimal contact between the transparency and the subject's skin). After marking the injection site, the flare was visually inspected and traced directly onto the transparency. Immediately thereafter, the points delineating mechanical allodynia and pinprick hyperalgesia were transferred to the same transparency.
Subsequently, for mechanical allodynia and pinprick hyperalgesia areas, the eight points were connected with a marker, using a ruler to connect adjacent points with a straight line. Each area (foam brush, von Frey hair, and flare) was then determined (in cm2) by careful tracing with a mechanical planimeter.
Repeated Measures Comparison of 10 and 100 μg Doses
To account for potential interindividual variability of intradermal capsaicin, a subset of subjects (n = 3) were tested with both 10 and 100 μg doses of capsaicin. Two subjects who had received 10 μg and one subject who had received 100 μg returned 4 weeks after their initial study date to receive the alternate dose. Only data from the initial study visit were included in statistical analyses.
Vital Signs
Blood pressure, heart rate, and respiratory rate were recorded at baseline (prior to injection) and at 0, 5, 10, and 15 minutes after capsaicin administration, and all subjects were continuously monitored for safety.
Statistical Analysis
All spontaneous and elicited pain scores and areas of allodynia, hyperalgesia, and flare were analyzed using standard two-factor repeated measures analysis of variance (ANOVA) with time as a within-subject factor and dose as a between-subject factor. Post hoc comparisons were made as dictated by the outcome of the overall ANOVA using the Bonferroni correction for multiple comparisons. A two-sided test with a p value < .05 was considered statistically significant. Descriptive statistics are given as mean (standard error of measurement).
RESULTS
Spontaneous Pain Scores
Capsaicin produced spontaneous pain in a dose-dependent manner (Figure 1), evidenced by a significant main effect of dose (p = .013, F = 5.10, df = 3,15). For 0.1 μg, pain scores peaked immediately after capsaicin was administered, quickly dissipated, and were nonexistent by 10 minutes in each subject. Similarly, pain scores from 1.0 μg of capsaicin rose sharply before decreasing to 6.00 (2.1) by 15 minutes after an initial spike. However, the 10 μg dose produced a pain sensation of longer duration, with a mean score of 22.5 (6.5) at 15 minutes. Likewise, the highest pain scores were seen with 100 μg, averaging 37.7 (11) at 15 minutes. Post hoc analyses demonstrated that spontaneous pain scores in the 10 and 100 μg groups were each greater than values in the 1.0 μg group (p < .05). In addition, the 100 μg dose produced greater spontaneous pain than the 10 μg dose, but the difference was not statistically significant.
Area and Intensity of Mechanical Allodynia (Foam Brush Stimulus)
Dose-dependent effects of capsaicin were evident with the area of allodynia and elicited pain to the brush stimulus (Figure 2). Elicited pain intensity to a foam brush (on the VAS) exhibited the strongest dose-response relationship (p < .001, F = 10.0, df = 3,15). A stepwise increase in elicited allodynia pain scores occurred as the capsaicin dose was increased, an effect seen at both 15 and 60 minutes (see Figure 2A). In addition, data from the three subjects who received both 10 and 100 μg of capsaicin further support this trend, with 100 μg producing higher elicited pain scores than 10 μg at both 15 and 60 minutes in all three subjects (mean VAS scores of 33 vs 9.3 at 15 minutes and 16 vs 2.7 at 60 minutes).
The capsaicin-induced area of allodynia was found to have a significant main effect of dose (p = .006, F = 6.33, df = 3,15), with increases seen from 0.1 to 1.0 to 10 μg at both 15 and 60 minutes (see Figure 2B). Areas of allodynia between 10 and 100 μg were similar at 15 minutes, although at 60 minutes, they were larger in the 10 μg group (p < .05). Among subjects who received both 10 and 100 μg, the dose-dependent trend was maintained, with 100 μg producing larger areas of allodynia than 10 μg at each time point (mean of 9.1 vs 5.7 cm2 at 15 minutes and 6.0 vs 3.1 cm2 at 60 minutes).
Area and Intensity of Pinprick Hyperalgesia (von Frey Test)
The area and intensity of pinprick hyperalgesia were significantly associated with the dose of capsaicin (Figure 3). The intensity of pinprick hyperalgesia exhibited a dose-dependent effect overall (p = .014, F = 4.90, df = 3,15). This relationship was most pronounced at 15 minutes, and von Frey test-mediated pain scores in the 10 and 100 μg groups were comparable at 60 minutes. Comparison of 10 and 100 μg doses administered to the same subjects illustrates that the dose dependence persists among these higher concentrations, with 100 μg consistently producing higher elicited pain scores than 10 μg (mean VAS scores of 44 vs 17 at 15 minutes and 30 vs 8.3 at 60 minutes).
The capsaicin-induced area of pinprick hyperalgesia was found to have a significant main effect of dose (p = .010, F = 5.42, df = 3,15). Stepwise increases were found between every level at 15 minutes and among the three lowest doses (0.1, 1.0, and 10 μg) at 60 minutes. Data from the cohort that received two doses again support the dose-dependent relationship among the higher concentrations of capsaicin because 100 μg produced larger areas of hyperalgesia than 10 μg at both time points (mean of 44 vs 23 cm2 at 15 minutes and 43 vs 23 cm2 at 60 minutes).
Area of Flare
The effect of capsaicin dose on the area of flare was significant (p = .010, F = 5.45, df = 3,15). In this case, expected increases were seen from 1.0 to 10 μg and from 10 to 100 μg at both 15 and 60 minutes, with a minimal flare response to 0.1 and 1 μg doses (Figure 4).
DISCUSSION
In the present study, intradermal capsaicin produced significant dose-dependent responses in all six primary outcomes: spontaneous pain, the intensity and area of mechanical allodynia (to a foam brush), the intensity and area of pinprick hyperalgesia (to a von Frey hair), and the flare area (neurogenic inflammation). This is the first time that these six markers of peripheral nociceptor activation and central sensitization have been shown to exhibit dose-dependent effects in human subjects when measured at fixed time points after capsaicin administration. These findings complement and extend the foundation laid by previous studies evaluating psychophysical and neurophysiologic manifestations of different doses of intradermal capsaicin.23-25
Spontaneous Pain and Neurogenic Inflammation
Spontaneous pain following intradermal capsaicin is mediated by VR-1 receptors on unmyelinated afferent C fibers.1-3,5,6Specifically, the intensity and duration of pain following capsaicin administration appear to be directly associated with the magnitude of C-fiber activation.1It has been demonstrated that capsaicin evokes increased magnitude and duration of pain as a function of dose, using normalized maximum magnitude estimates.24In the current study, VAS pain scores were also found to be dependent on capsaicin dose—specifically, 10 μg produced significantly greater pain scores than 1 μg, and a non-significant trend existed between the 10 and 100 μg doses (see Figure 1). As ongoing pain is a feature of many pain syndromes, the potential to reduce spontaneous pain intensity pharmacologically has direct clinical application. As capsaicin produces a clearly delineated dose-dependent relationship, with higher doses of capsaicin producing higher pain scores, this pain model allows production of a predetermined magnitude of pain intensity, allowing controlled assessment of analgesic efficacy.
Previously, neurogenic inflammation has been attributed to antidromic conduction at distal afferent axons (axon flare reflex)—a peripherally mediated event.26,27The area of flare is dependent on the capsaicin dose, with significant increases from 1 to 10 to 100 μg of capsaicin.24The current study confirms the dose-dependent relationship, demonstrating a substantial flare response following 10 and 100 μg doses (see Figure 4). Recently, it was suggested that neurogenic inflammation is composed of both peripheral and central sources—namely, axon reflex and dorsal root reflex—and this flare response, in turn, may contribute to secondary heat hyperalgesia.28,29It can be argued that a reduction in the direct vasodilatory effects comprising the flare response has the potential to ameliorate the development of secondary heat hyperalgesia as well. As capsaicin reliably produces neurogenic inflammation in a dose-dependent manner, this model may be used to analyze pharmacologic potential to specifically attenuate this component of pain.
Mechanical Allodynia and Pinprick Hyperalgesia
In the current study, the area and intensity of mechanical allodynia and of pinprick hyperalgesia were dependent on the capsaicin dose administered (see Figures 2 and 3). In addition, the repeated measures data demonstrate that 100 μg of capsaicin consistently produced greater values than 10 μg in each of these four variables among the three subjects who received both doses. These findings are consistent with a previous study demonstrating a direct relationship between the area of mechanical allodynia and the capsaicin dose.24To our knowledge, this is the first study illustrating a dose-dependent effect for the area and intensity of pinprick hyperalgesia, as well as the intensity of mechanical allodynia in humans.
Areas of mechanical allodynia (foam brush stimulus) and pinprick hyperalgesia (von Frey test) are believed to be mediated by central sensitization.1,5,6,30-32Mechanical allodynia results from neural plasticity in the dorsal spinal cord, such that a light touch activates Aβ fibers, which subsequently activate the ascending pain pathway.5,6Pinprick hyperalgesia appears to be transmitted by the smaller-diameter Aδ and C afferent fibers, with neuroadaptation again occurring at the level of the central nervous system.4,6Much of what is known about human pain pathways and mechanisms of pain states is the result of studies using intradermal capsaicin. Acting as a bridge between experimental and clinical pain states, the capsaicin model has the potential to dissect the pain experience into its pathophysiologic components.
Central sensitization is believed to be directly involved in the etiology of clinical pain states, including postherpetic neuralgia and complex regional pain syndrome.7-10As there is a direct link between clinical pain and central sensitization and central sensitization can be quantified with measures of allodynia and hyperalgesia, these are often used as primary outcomes in analgesic efficacy studies. Intensity measures of elicited pain to a foam brush (mechanical allodynia) and to the von Frey test (pinprick hyperalgesia) provide additional means of discerning pharmacologic efficacy in a component-specific manner. It has been proposed that emerging therapeutics should be evaluated and analyzed not only on the global pain experience but, more specifically, on those targets and elements of the pain experience at which they are aimed.33As pain pathways and constituents become more completely characterized, this will facilitate more targeted pharmacotherapies using a mechanism-based approach.
Sources of Variability
The vast majority of data follow a classic dose-response pattern, with higher doses of capsaicin producing effects of greater magnitude. However, a discrepant finding exists among the highest two doses for area of mechanical allodynia (see Figure 2B). Whereas 100 μg would be expected to produce larger areas than 10 μg, the areas were comparable at 15 minutes and were smaller in the 100 μg group at 60 minutes. This is most likely the result of between-subject variability, which accounts for most of the variability in the intradermal capsaicin pain model.11A previous study found that allodynia was shorter-lived and showed greater variability than pinprick hyperalgesia.6This is consistent with the current study, with the area of allodynia exhibiting greater variability than pinprick hyperalgesia at the later time point. When looking only at subjects who received both 10 and 100 μg doses, 100 μg produced larger areas in all three subjects at both time points, confirming the dose-dependent relationship between the capsaicin dose and the area of allodynia. Another study found that some subjects developed minimal areas of mechanical allodynia (to a brush) following intradermal capsaicin.34It is hypothesized that the between-subject variability is due to variations in the initial neural processing that leads to hyperalgesic phenomena, which may account for this variability in the area of allodynia.6Of note, a well-designed parallel-group study showed that gabapentin significantly reduced the capsaicin-mediated area of mechanical allodynia, illustrating the utility of the capsaicin model despite between-subject variability.17Minimizing between-subject variability will increase the sensitivity of the capsaicin model and maximize statistical power with small sample sizes.
There are several ways to optimize the capsaicin pain model. Prescreening subjects with a test dose of capsaicin to ensure adequate allodynia and hyperalgesia may be instituted as an inclusion criterion prior to entry into the study.34It has been suggested that a minimum mechanical allodynia area of 5 cm2 be present for 15 minutes as a criterion for an adequate responder.34Another way to reduce between-subject variability is to have subjects serve as their own controls via a crossover study design, whereby subjects are randomly assigned to the treatment or placebo arm and then crossover to the alternate group, with capsaicin administered twice. Overall, intradermal capsaicin is well tolerated and of limited variability such that a reduced number of human subjects would be required than that necessary to conduct an analgesic efficacy study of clinical pain.11
SUMMARY
In conclusion, intradermal capsaicin produces dose-dependent pain, area and intensity of mechanical allodynia and pinprick hyperalgesia, and flare in healthy human subjects. The 10 and 100 μg doses produced robust pain measures across a range of modalities, whereas lower doses produced minimal effects. Whereas most studies use 100 μg of capsaicin, using a lower dose is reasonable and may facilitate elucidation of various subtleties in future analgesic efficacy studies, particularly with regard to nonopioid analgesic agents. This pain model allows crossover designs, minimizes confounding variables in clinical pain states, and retains sufficient power with small sample sizes. The ability to analyze specific pharmacologic effects on different components of the pain experience may facilitate more targeted approaches to pain management.
ACKNOWLEDGMENTS
We thank Tony Yaksh, PhD, for provision of laboratory space for capsaicin preparation, Ben Atwater, MD, PhD, for graphical assistance, and Linda Sutherland and Peggy Mollen, RN, for technical assistance.